Biologic of Atmospheric Pollutants
ICUL ATE
LYCYCLIC
JRGANIC
MATTER
NATIO
ADEMY OF SCIENCES
-------
Biologic Effects of Atmospheric Pollutants
PARTICULATE
POLYCYCLIC
ORGANIC
MATTER
Committee on
Biologic Effects of
Atmospheric Pollutants
DIVISON OF MEDICAL SCIENCES
NATIONAL RESEARCH COUNCIL
NATIONAL ACADEMY
OF SCIENCES
WASHINGTON, D.C. 1972
-------
NOTICE: The study repotted herein was undertaken under the aegis of the National Re-
search Council with the express approval of the Governing Board of the NRC. Such approval
indicated that the Board considered that the problem is of national significance, that eluci-
dation of the problem required scientific or technical competence, and that the resources of
NRC were particularly suitable to the conduct of the project. The institutional responsibili-
ties of NRC were then discharged in the following manner:
The members of the study committee were selected for their individual scholarly compe-
tence and judgment with due consideration for the balance and breadth of disciplines. Re-
sponsibility for all aspects of this report rests with the study committee, to whom sincere
appreciation is expressed.
Although the reports of our study committees are not submitted for approval to the
Academy membership nor to the Council, each report is reviewed by a second group of
scientists according to procedures established and monitored by the Academy's Report
Review Committee. Such reviews are intended to determine, inter alia, whether the major
questions and relevant points of view have been addressed and whether the reported findings,
conclusions, and recommendations arose from the available data and information. Distribution
of the report is approved, by the President, only after satisfactory completion of this review
process.
The work on which this publication is based was performed pursuant to Contract No.
CPA 70-42 with the Environmental Protection Agency.
Available from
Printing and Publishing Office
National Academy of Sciences
2101 Constitution Avenue, N.W.
Washington, D.C. 20418
ISBN 0-309-02027-1
Library of Congress Catalog Card Number 72-76309
Printed in the United States of America
-------
PANEL ON POLYCYCLIC ORGANIC MATTER
JAMES N. PITTS, JR., Statewide Air Pollution Control Center,
University of California, Riverside, Chairman
ROY E. ALBERT, Institute of Environmental Medicine, New York
University Medical Center, New York
RAYMOND J. CAMPION, Products Research Division, Esso Re-
search & Engineering Company, Linden, New Jersey
BERTRAM W. CARNOW, Department of Preventive Medicine,
University of Illinois College of Medicine, Chicago
ERNEST H. Y. CHU, Biology Division, Oak Ridge National Labora-
tory, Oak Ridge, Tennessee
JOHN C. CRAIG, Department of Pharmaceutical Chemistry, Univer-
sity of California, San Francisco
T. TIMOTHY CROCKER, Department of Community and Envi-
ronmental Medicine, University of California Medical School, San
Francisco
CHARLES HEIDELBERGER, McArdle Laboratory, University of
Wisconsin, Madison
GEORGE M. HIDY, Science Center, North American Rockwell
Corporation, Thousand Oaks, California
AVERILL A. LIE BOW, Department of Pathology, University of
California Medical School, La Jolla
EDWARD D. PALMES, Institute of Environmental Medicine, New
York University Medical Center, New York
RAYMOND R. SUSKIND, Department of Environmental Health,
University of Cincinnati College of Medicine, Cincinnati, Ohio
BENJAMIN L. VANDUUREN, Institute of Environmental Medicine,
New York University Medical Center, New York
ELIZABETH E. FORCE, Division of Medical Sciences, National
Research Council, Washington, D.C., Staff Officer
111
-------
CONSULTANTS
EDWARD J. BAUM, Oregon Graduate Center, Portland
SAMUEL S. EPSTEIN, Case Western Reserve University School of
Medicine, Cleveland, Ohio
CHRISTOPHER S. FOOTE, Department of Chemistry, University
of California, Los Angeles
SHELDON K. FRIEDLANDER, Department of Environmental
Engineering, California Institute of Technology, Pasadena
HARRY v. GELBOIN, Chemical Branch, National Cancer Institute,
National Institutes of Health, Bethesda, Maryland
JAMES E. GILL, Biomedical Division, Lawrence Radiation Labora-
tory, University of California, Livermore
DIETRICH HOFFMANN, Division of Environmental Toxicology,
American Health Foundation, New York
WILLIAM D. MacLEOD, Jr., Department of Pharmaceutical Chemis-
try, University of California, San Francisco
PAUL MEIER, Department of Statistics, University of Chicago,
Chicago, Illinois
HERBERT L. RATCLIFFE, Penrose Research Laboratory, Zoologi-
cal Society of Philadelphia, Philadelphia
O. CLIFTON TAYLOR, Statewide Air Pollution Control Center,
University of California, Riverside
CONTRIBUTORS
ROSWELL K. BOUTWELL, Department of Oncology and Biochem-
istry, University of Wisconsin, Madison
DAVID L. COFFIN, Experimental Pathology Section, Biology Re-
search Branch, Division of Health Effects Research, Air Pollution
Control Office, Durham, North Carolina
IAN T. HIGGINS, School of Public Health, University of Michigan,
Ann Arbor
MORRIS KATZ, Department of Chemistry, York University, Toronto,
Ontario, Canada
IV
-------
JOHN B. LITTLE, Department of Physiology, Harvard School of
Public Health, Boston, Massachusetts
NATHAN MANTEL, Biometry Branch, National Cancer Institute,
National Institutes of Health, Bethesda, Maryland
PAUL NETTESHEIM, Carcinogenesis Program, Biology Division,
Oak Ridge National Laboratory, Oak Ridge, Tennessee
LEO ORRIS, Institute of Environmental Medicine, New York
University Medical Center, New York
RICHMOND T. PREHN, Institute for Cancer Research and Depart-
ment of Pathology, School of Medicine, University of Pennsylva-
nia, Philadelphia
COMMITTEE ON BIOLOGIC EFFECTS OF
ATMOSPHERIC POLLUTANTS
ARTHUR B. DuBOIS, Department of Physiology, School of Medi-
cine, University of Pennsylvania, Philadelphia, Chairman
VINTON W. BACON, College of Applied Science and Engineering,
University of Wisconsin, Milwaukee
ANNA M. BAETJER, Department of Environmental Medicine,
School of Hygiene and Public Health, The Johns Hopkins Uni-
versity, Baltimore, Maryland
W. CLARK COOPER, School of Public Health, University of Cali-
fornia, Berkeley
MORTON CORN, Graduate School of Public Health, University of
Pittsburgh, Pittsburgh, Pennsylvania
BERTRAM D. DIN MAN, School of Public Health, University of
Michigan, Ann Arbor
LEON GOLBERG, Institute of Experimental Pathology and Toxi-
cology, Albany Medical College, Albany, New York
PAUL B. HAMMOND, Department of Physiology and Pharmacology,
College of Veterinary Medicine, University of Minnesota, St. Paul
SAMUEL P. HICKS, Department of Pathology, University of Michi-
gan, Ann Arbor
VICTOR G. LATIES, Department of Radiation Biology and Biophys-
ics, University of Rochester Medical Center, Rochester, New York
-------
ABRAHAM M. LILIENFELD, Department of Chronic Diseases,
School of Hygiene and Public Health, The Johns Hopkins Univer-
sity, Baltimore, Maryland
PAUL MEIER, Department of Statistics, University of Chicago,
Chicago, Illinois
JAMES N. PITTS, JR., Statewide Air Pollution Control Center,
University of California, Riverside
GORDON J. STOPPS, The Environmental Health Branch, Health
Studies Service, Ontario Department of Health, Toronto, Ontario,
Canada
O. CLIFTON TAYLOR, Statewide Air Pollution Control Center,
University of California, Riverside
JAROSLAV J. voST AL, Department of Pharmacology and Toxi-
cology, University of Rochester Medical Center, Rochester, New
York
T. D. BOAZ, JR., Division of Medical Sciences, National Research
Council, Washington, D.C., Executive Director
VI
-------
Acknowledgments
The preparation of this document, which deals with a very complex
subject, in so short a time was possible only because of the excep-
tional and dedicated efforts of those involved—particularly, the mem-
bers and consultants of the Panel on Polycyclic Organic Matter, all
of whom contributed their time generously. The report represents a
team effort and has been critically evaluated in toto by the entire
Panel.
Responsibility for generating drafts was divided among three task
forces—on environmental appraisal; studies of POM in animals,
mammalian cells, and vegetation; and human effects—to facilitate the
exchange of information and ideas among smaller groups of researchers
working in common areas of experience. The contributions of the
chairmen of these task forces-Drs. John C. Craig, Charles Heidelberger,
and Raymond R. Suskind, respectively—to the development of this
document were noteworthy.
Credit must be given to the persons who drafted the chapters.
Those who contributed to the chapters on structure and nomenclature
of POM, sources, atmospheric physics, chemical reactivity, collection,
separation, detection, identification, and quantitation include:
Drs. Edward J. Baum, Raymond J. Campion, John C. Craig, Christopher
vii
-------
viii Acknowledgments
S. Foote, James E. Gill, George M. Hidy, Dietrich Hoffmann, Morris
Katz, and William D. MacLeod, Jr.; the chapters on the effects of
POM in animals, mammalian cells, and vegetation: Drs. Roswell K.
Boutwell, Ernest H. Y. Chu, David L. Coffin, T. Timothy Crocker,
Samuel S. Epstein, Harry V. Gelboin, Charles Heidelberger, Dietrich
Hoffmann, Averill A. Liebow, John B. Little, Nathan Mantel, Paul
Nettesheim, Leo Orris, Richmond T. Prehn, Herbert L. Ratcliffe,
O. Clifton Taylor, and Benjamin L. Van Duuren; and the chapters on
the evaluation of effects of particulate POM in humans: Drs. Roy
E. Albert, Bertram W. Carnow, Ian T. Higgins, Averill A. Liebow,
Paul Meier, and Raymond R. Suskind.
A significant factor in expediting the development of this docu-
ment was the support given to the Panel by Dr. Robert J. M. Horton,
Senior Research Adviser, Dr. Eugene Sawicki, and Dr. Francis G.
Hueter of the Environmental Protection Agency.
The document was reviewed by the parent Committee on the Bio-
logic Effects of Atmospheric Pollutants, by several anonymous edi-
tors selected by the National Academy of Sciences, by the Academy's
Report Review Committee, by the Academy's Advisory Center on
Toxicology, and by the National Research Council's Divisions of
Biology and Agriculture, Chemistry and Chemical Technology, and
Physical Sciences. The report was edited by Mr. Norman Grossblatt,
Editor for the Division of Medical Sciences. Their substantial contri-
butions to this document are gratefully acknowledged.
Finally, I should like to acknowledge the efforts of Miss Elizabeth E.
Force, of the Division of Medical Sciences, who served as manager for
the Panel activities. Her rare combination of managerial ability, tech-
nical competence, and personal enthusiasm did much to promote the
effectiveness of the entire operation.
CHARLES L. DUNHAM
Chairman, Division of Medical Sciences
-------
Preface
The 1967 amendments to the Clean Air Act of 1963 required that the
Secretary of Health, Education, and Welfare
from time to time, but as soon as practicable, develop and issue to the States
such criteria of air quality as in his judgment may be requisite for the protection
of the public health and welfare. . . . Such criteria shall.. . reflect the latest
scientific knowledge useful in indicating the kind and extent of all identifiable
effects on health and welfare which may be expected from the presence of an
air pollution agent. . . .
A critical step in implementing these requirements of Congress has
been the issuance of Air Quality Criteria Documents by the National
Air Pollution Control Administration (N APCA), more recently des-
ignated the Air Pollution Control Office Technical Center of the
Environmental Protection Agency (EPA). Air Quality Criteria Docu-
ments already published are on particulate matter, oxides of sulfur,
hydrocarbons, carbon monoxide, photochemical oxidants, and oxides
of nitrogen. Until recently, these documents generally were prepared
by the combined efforts of inhouse NAPCA staff and consultants—
both individuals and teams from private companies.
In the spring of 1970, the Division of Medical Sciences, National
Academy of Sciences-National Research Council, entered into a
ix
-------
x Preface
contract with the EPA to produce background documents for pollu-
tants, including particulate poly cyclic organic matter (POM). To
facilitate the development and production of these four documents
and to ensure a critical "overview" in evaluating the content of the
documents, the Academy set up the Committee on Biologic Effects
of Atmospheric Pollutants (HEAP), under the chairmanship of Dr.
Arthur B. DuBois. Ad hoc panels of experts were formed to evaluate
each of the selected pollutants.
The comprehensive, evaluative, and multidisciplinary aspects of the
problem faced by the Panel on Polycyclic Organic Matter are clearly
illustrated in the following "Statement of Work," taken from the con-
tract with the Academy:
Prepare an open-ended series of comprehensive state-of-the-art technical reports,
which will reflect the latest scientific knowledge useful in indicating the kind and
extent of all identifiable effects on human health and welfare which may be ex-
pected from the presence of a variety of pollutants in the ambient air. In de-
veloping these reports, consideration will be given to the sources, chemical
and physical characteristics of the pollutants, the techniques available for their
measurement in the ambient air, their prevalence in contaminated air and pos-
sible modifying conditions such as: reaction time, effects of other pollutants
simultaneously present, and meteorological conditions. Documentation of the
effects of these pollutants on human health and well-being, on animals, on vege-
tation, on materials and on man's environment in general is deemed to be of pri-
mary importance. These reports will contain detailed comment on dose/response
relationship and margins of safety to be used in establishing air quality standards.
They will indicate groups of persons in the general population known to.be or
likely to be particularly sensitive to exposure.
It is important to recognize that at no time have the scientists,
physicians, economists, etc., who have dealt with previously prepared
criteria documents felt that there was ample unequivocal evidence to
support all their conclusions and recommendations. This was true even
for perhaps the best understood of the common gaseous pollutants-
carbon monoxide and oxides of sulfur. Despite the serious and wide-
spread deficiencies in our knowledge of POM and its biologic effects
on man, the Panel members and consultants attempted to make the
most effective use of reliable knowledge available today, not only from
the published literature, but also from the current unpublished ex-
perimental data from a number of laboratories. Furthermore, there
was unanimous agreement that, although there are wide areas of total
ignorance, misunderstanding, or disagreement in the literature on the
carcinogenic activity of poly cyclic organic compounds, recognizing and
evaluating such gaps in vital knowledge serve the positive function of
-------
Preface xi
providing a basis for the preparation of specific recommendations for
future research.
This document is not simply a review of the pertinent literature.
Such reviews and surveys have already been produced (e.g., Prelimi-
nary Air Pollution Survey of Organic Carcinogens. A Literature Re-
view, published in 1969 by the Department of Health, Education, and
Welfare768). Although they are helpful in illuminating the scope of the
literature, they are not intended to synthesize or evaluate information.
This report attempts to interpret, evaluate, and reconcile the immense
amount of information available, especially that concerning the car-
cinogenic effects of POM.
-------
Contents
1 Introduction 1
2 Structure and Nomenclature of Polycyclic Aromatic
Hydrocarbons and Aza-Arenes
3 Sources of Polycyclic Organic Matter 13
4 Atmospheric Physics of Particulate Polycyclic Organic
Matter 36
5 Chemical Reactivity of Polycyclic Aromatic Hydro-
carbons and Aza-Arenes 63
6 Historical and Theoretical Aspects of Chemical
Carcinogenesis 82
7 Experimental Design in Carcinogenesis Tests 87
8 In vivo Tests for Carcinogenesis and Cocarcinogenesis 95
9 Modification of Host Factors in in vivo Carcinogenesis
Tests 118
10 Distribution, Excretion, and Metabolism of Polycyclic
Hydrocarbons 132
11 In vitro Approaches to Carcinogenesis 142
12 Indirect Tests for Determining the Potential Carcino-
genicity of Polycyclic Aromatic Hydrocarbons 143
13 Teratogenesis and Mutagenesis 15 \
xii
-------
14
15
16
17
18
19
Appendix A:
Appendix B:
Appendix C:
Appendix D:
References
Index
Vegetation and Polycyclic Organic Matter
Introduction to Appraisal of Human Effects
Characteristics of Human Disease Related to Poly-
cyclic Organic Matter
Clinical and Epidemiologic Studies
General Summary and Conclusions
Recommendations for Future Research
Collection of Airborne Particles for Analysis
of Polycyclic Organic Matter
Separation Methods for Polycyclic Organic
Matter
Detection, Identification, and Quantitation
Regression Analysis
160
166
172
191
237
247
253
261
277
304
307
355
Xlll
-------
Introduction
Evidence of the induction of lung cancer via inhalation in man is ex-
tensive. The best-documented and most decisive evidence is related to
occupational exposures, but there is also evidence that the generally
low concentrations of some pollutants found in community air may
be associated with an increased risk of lung cancer.
There are many examples—some strikingly unequivocal, others only
suggestive—of human lung cancer caused by environmental factors. The
major hazard is currently attributed to cigarette smoking, but other
environmental sources are suspect. Indeed, the lung may lead the list
of human organs for the variety and number of instances in which en-
vironmental agents are involved in cancer induction, with the skin
second on the list. That the lung should share this role with skin reflects
its direct contact with environmental agents.
Many etiologic agents in cigarette smoke have been proposed, e.g.,
polycyclic hydrocarbons, arsenic, nitrosamines, and polonium. Poly-
cyclic aromatic hydrocarbons like benzofa] pyrene are thus candidates
for promoting lung cancer. This is not to say, however, that other
initiating carcinogens do not contribute.
Laboratory studies can have a variety of purposes aimed at increased
understanding of the human problem; among these are determining
1
-------
2 PARTICULATE POLYCYCLIC ORGANIC MATTER
whether an environmental agent is related to lung cancer in man, pin-
pointing the active agent(s) among a number of suspected environ-
mental agents, determining dose-response patterns or relations, and
developing a better understanding of the biologic course of the disease.
Present knowledge indicates that fractions of particulate POM con-
tain only two classes of compounds that are known animal carcino-
gens-the polycyclic aromatic hydrocarbons and their neutral nitrogen
analogues, the aza-arenes (e.g., indoles and carbazoles). Numerous
types of POM exist in urban air, such as pyrene, anthanthrene,
benz[a] anthracene, benzofluoranthenes, dibenzanthracenes, chrysene,
phenylenepyrene, benzoperylene, coronene, fluoranthene, and alkyl
derivatives of these compounds, as well as benzopyrene. For this
reason and because there are experimental data on the carcinogenic
effects of benzo[a] pyrene and other polycyclic aromatic hydrocarbons,
much attention has been directed in this report to the evaluation of these
compounds in carcinogenesis.
Laboratory data indicate that cancer can be produced by the super-
imposition of ozonized gasoline on influenza in mice; and the combi-
nation of polycyclic aromatic hydrocarbons with sulfur dioxide pro-
duces cancer in rats. Thus, irritants may be very important co-reacting
factors. Host-related and other co-acting factors are discussed and
evaluated in this report. The problems inherent in the extrapolation
of experimental data to humans are also considered.
The major thrust of this report has been the working hypothesis
that there is an urban-rural difference in lung cancer rates that may
stem from community air pollution and that may be attributable to
polycyclic hydrocarbons known to be carcinogenic, such as benzofa]-
pyrene. Epidemiologic evidence linking community air pollution to
lung cancer has been ambiguous to date, but a modest contribution
of air pollution to lung cancer is possible. Much effort has been ex-
pended by some of the Panel members in the statistical re-evaluation
of existing epidemiologic data in an attempt to delineate the associa-
tion of cigarette smoking and urban factors with lung cancer.
For complete evaluation of the hazards imposed on man by poly-
cyclic aromatic hydrocarbons in the air, it is necessary to identify the
sources of pollution and the types of substances they emit, to evaluate
the physical and chemical reactivity of these substances in the atmo-
sphere, and to describe existing methods for collecting, separating,
detecting, and quantitating polycyclic aromatic hydrocarbons.
These areas of research are examined in detail in this report.
The report has been structured along the same functional lines as
-------
Introduction 3
the Panel's activities. The environmental appraisal of POM (sources,
characteristics, etc.) appears in Chapters 2-5. Studies of the effects
of POM in animals, mammalian cells, and vegetation are discussed in
the next nine chapters. The human effects of POM are described,
on the basis of extensive epidemiologic studies, in Chapters 15, 16,
and 17. The methodology of collecting and analyzing atmospheric
PO M is set forth in Appendixes A, B, and C.
-------
Structure and Nomenclature
of Polycyclic Aromatic
Hydrocarbons and Aza-Arenes
The nomenclature used throughout this presentation is that adopted
by the International Union of Pure and Applied Chemistry (IUPAC)
and by Chemical Abstracts Service. The most important rules, all of
which are described in detail in The Ring Index,5*6 are the following:
1. The structural diagram is written to present the greatest possi-
ble number of rings in a horizontal row.
2. Horizontal and vertical axes are then drawn through the center of
the horizontal row, and the molecule is oriented in such a way as to
place the maximal number of rings in the upper right quadrant and the
minimal number of rings in the lower left quadrant.
3. The carbon atoms are numbered in a clockwise direction start-
ing with the carbon atom that is not part of another ring and is in the
most counterclockwise position of the uppermost ring farthest to
the right; carbon atoms common to two or more rings are not num-
bered.
4. The faces of the rings are now lettered in alphabetical order be-
ginning with "a" for the side between carbon atoms 1 and 2 and con-
tinuing clockwise around the molecule; ring faces common to two
rings are not lettered.
-------
Structure and Nomenclature 5
5. In naming a compound formed by the addition of a component,
the numbers and letters are placed in square brackets and placed im-
mediately after the name of the added component, showing where a
substituent group is attached or where a ring is fused to the face of
the molecule. If a ring is fused to more than one face of the molecule
simultaneously, this is indicated by using the appropriate letters to
denote the faces so involved.
6. The structural formulas used show aromatic rings as plain hexa-
gons and a methylene group as CH2.
Most of the polycyclic aromatic hydrocarbons and polycyclic
azaheterocyclic compounds (aza-arenes) listed in Tables 2-1 and 2-2 are
organic materials identified in the urban atmosphere, generally as
suspended particles.388
The property of these compounds that is most relevant to the
human health status of an exposed population and mainly discussed
in this report is their potential carcinogenic activity. For quick ref-
erence, the carcinogenicity of a compound in this chapter is indicated
by a simple code:
— not carcinogenic
± uncertain or weakly carcinogenic
+ carcinogenic
++, +++, ++++ strongly carcinogenic
The indications of carcinogenicity refer to the Public Health Ser-
vice (PHS) survey of compounds tested for carcinogenicity.346'699'700
Because many of these compounds are referred to in the PHS survey
by the old Richter nomenclature, which differs from the modern
nomenclature, the older names are given here in parentheses. Also,
starred (*) compounds indicate disagreement with standard num-
bering.
-------
6 PARTICULATE POLYCYCLIC ORGANIC MATTER
TABLE 2-1 Polycyclic Aromatic Hydrocarbons
Compound
Structure
Carcinogen icity
8
Anthracene
m.p., 216 C; b.p., 340 C
Benz [a] anthracene
10
m.p., 158 C; sublimes 9
(1,2-benzanthiacene)
7,12-Dimethylbenz [a] anthracene
(9,10-dimethyl-l,2-benzanthracene)
Dibenz [ a J ] anthracene
(1,2-7,8-dibenzanthracene)
Dibenz [a,h] anthracene
(1,2-5,6-dibenzanthracene)
Dibenz [a,c] anthracene
(1,2-3,4-dibenzanthracene)
Phenanthrene
m.p., 101 C;b.p., 340 C 8
7
1
10
-------
Structure and Nomenclature
TABLE 2-1 Polycyclic Aromatic Hydrocarbons-Continued
Compound
Structure
Caicinogenicity
Benzo[c] phenanthrene
(3,4-benzphenanthrene)
Fluorene
m.p., 116 C;b.p.,293 C
Benzo [ a ] fluorene
(1,2-benzfluorene)
Benzo [b] fluorene
(2,3-benzfluorene)
H2
Dibenzo[a,h] fluorene]
(1,2-6,7-dibenzfluorene)
Dibenzo[a,g] fluorene
(1,2-5,6-dibenzfluorene)
Benzo [c] fluorene
(3,4-benzfluorene)
Dibenzoja.c] fluorene
(1,2-3,4-dibenzfluorene)
-------
8 PARTICULATE POLYCYCLIC ORGANIC MATTER
TABLE 2-1 Polycyclic Aromatic Hydrocarbons-Continued
Compound
Structure
Carcinogenicity
Fluoranthene
Ci6H10
m.p., 110 C;b.p., 393 C
Benzo [ b ] fluoranthene
(2,3-benzofluoranthene)
Benzo [ j ] fluoranthene
(7,8-benzofluoranthene)
Benzo [k] fluoranthene
(8,9-benzfluoranthene)
Benzo [ghi] fluoranthene
10
Aceanthrylene
CieHji
m.p., 113 C
Benz[j] aceanthrylene
= cholanthrene
-------
Structure and Nomenclature
TABLE 2-1 Polycyclic Aromatic Hydrocarbons—Continued
Compound Structure
Carcinogen icity
3-Methylcholanthrene
Naphthacene
= benz[b] anthracene
m.p., 341 C; sublimes
Naphtho [2,1,8-qra] naphthacene
= naphtho[2,3-a]pyrene
(2',3'-naphtho-l,2^>yrene)
CH
10 1
Pyrene
m.p., 150 C;b.p.,>360 C
Benzo[a]pyrene
(1,2-benzpyrene)
(3,4-benzy pyrene *)
Benzo[e]pyrene
(4,5-benzpyrene)
(1,2-benzpyrene*)
Dibenzo(a,l] pyrene
(2,3-4,5-dibenzpyrene)
(1,2-3,4-dibenzpyrene*)
-------
10 PARTICULATE POLYCYCLIC ORGANIC MATTER
TABLE 2-1 Polycyclic Aromatic Hydrocarbons-Continued
Compound
Structure
Carcinogen icity
Dibenzo[a,h] pyrene
(1,2-6,7-dibenzpyrene)
(3,4-8,9-dibenzpyrene*)
Dibenzo[a,i] pyrene
(2,3-6,7-dibenzpyrene)
(4,5-8,9-dibenzypyrene*)
Dibenzo[cdjk] pyrene
= anthanthrene
Indeno[ 1,2,3-cd] pyrene
(O-phenylenepyrene)
Chrysene
= 1,2-benzophenanthrene
CigHij
m.p., 254 C;b.p.,448 C
10
Dibenzo(b,def ] chrysene
= dibenzoja.h] pyrene
(3,4-8,9-dibenzpyrene*)
Dibenzo[def,p] chrysene
= dibenzo[a,l] pyrene
(1,2-3,4-dibenzpy rene *)
-------
Structure and Nomenclature 11
TABLE 2-1 Polycyclic Aromatic Hydrocarbons—Continued
Compound
Structure
Carcinogen icity
Dibenzo [ def ,mno ] chry sene
= anthanthrene
= dibenzo[cdjk]pyrene
Perylene
m.p., 273 C;b.p., ca.
500 C
Benzojghi] perylene
12
12
11
Coronene
C^Hn
m.p., 438 C;b.p.,525 C
-------
12 PARTICIPATE POLYCYCLIC ORGANIC MATTER
TABLE 2-2 Aza-Arenes
Compound
Structure
Carcinogenicity
Dibenz[aJ]acridine
(1,2-7,8-dibenzacridine)
Dibenz|a,h] acridine
(1,2-5,6-dibenzacridine)
Dibenz[c,h] acridine
(3,4-5,6-dibenzacridine)
Carbazole
C,2H9N 6
m.p., 246 C;b.p.,355 C
Benzo [ a ] carbazole
(1,2-benzcarbazole)
Dibenzo [a,g] carbazole
(1,2-5,6-dibenzcarbazole)
Dibenzo [c,g] carbazole
(3,4-5,6-dibenzcarbazole)
Dibenzo[a,i] carbazole
(1,2-7,8-dibenzcarbazole)
^^ H
6^
-------
Sources of Polycyclic
Organic Matter
Polycyclic organic matter (POM) can be formed in any combustion
process involving fossil fuels or, more generally, compounds contain-
ing carbon and hydrogen. The amount of POM formed will vary widely;
efficient, controlled combustion favors very low POM emissions,
whereas inefficient burning favors high emissions. This chapter cata-
logs the more obvious sources of POM emissions to the atmosphere
and points out the uncertainties in our information on various source
contributions.
MECHANISM OF POM FORMATION
Although the mechanism of POM formation in combustion pro-
cesses is complex and variable, a relatively clear picture of the overall
reaction has emerged, owing primarily to Badger.21 Chemical reac-
tions in flames proceed by free-radical paths; in POM formation, a
synthetic route is postulated, as shown in Figure 3-1. Radical species
containing one, two, or many carbon atoms can combine in rapid
fashion at the high temperatures (500-800 C) attained in the flame
front. This pyrosynthesis of pyrolysis products is obviously a func-
13
-------
14
PARTICULATE POLYCYCLIC ORGANIC MATTER
(VII)
FIGURE 3-1. Mechanism of benzo[a] pyrene formation. (After Badger. )
tion of many variables, not the least of which is the presence of a
chemically reducing atmosphere, common in the center of flames.
In these conditions, radical chain propagation is enhanced, allowing
the buildup of a complex POM molecule, such as benzo[a] pyrene.
It is important to note that, although methane itself can lead to
POM,102 the formation of these large molecules is favored by the pres-
ence of higher-molecular-weight radicals and molecules in the fuel.
Obviously, it is unnecessary to break the starting material down com-
pletely to a two-carbon radical in order to form benzo[a] pyrene.
Any component of the combustion reaction that can contribute
intermediate pyrolysis products of the structure required for benzo-
[a] pyrene synthesis would be expected to lead to increased yields of
benzo[a] pyrene. Thus, Badger and Spotswood22 have shown that in the
pyrolysis of alkylbenzenes, including «-butylbenzene, enhanced
benzo[a] pyrene formation is due primarily to increased concentra-
tions of intermediate structures of types III, IV, and V (Figure 3-1).
Badger also showed conclusively that specific aromatic and diole-
finic compounds serve as precursors for other polycyclic organic prod-
ucts. The mechanism in Figure 3-1 is a pathway to benzo[a] pyrene
formation, but similar routes could be devised, with somewhat dif-
ferent intermediates, to lead to most of the known POM produced
in combustion processes. Badger's work, with its reliance on calcu-
lated C—C and C—H bond energies to predict favored pathways and
the experimental confirmation of these steps with radioisotopic
labeling, provides a clear-cut mechanism for POM formation in the
combustion process.
As pointed out by Hoffmann and Wynder,388 the use of nitrogen
atmospheres in Badger's experiments has been criticized for its lack
-------
Sources 15
of relevance to the actual combustion of organic molecules. None-
theless, the conditions are similar to those of the oxygen-deficient
environment in flames, and the data are in good qualitative agreement
with observed POM combustion products.
NONTECHNOLOGIC SOURCES OF POM
Uncontrolled combustion, such as that in forest fires, would be ex-
pected to produce POM. Although the requirements for apprecia-
ble POM formation can be met in these fires, data on actual emission
rates are lacking. The only other nontechnologic source of airborne
POM is agricultural burning, but, because it is often planned'by man,
its contribution is covered below, under "Refuse Burning."
TECHNOLOGIC SOURCES OF POM
Large quantities of POM are generated in the vast number of tech-
nologic activities that prevail in our society. The contribution of any
particular source depends on many factors, including geography,
urbanization, and climate; thus, nationwide emission inventories can
be misleading. Olsen and Haynes572 have summarized the available data.
Man-made POM emission sources can be broadly separated into
transportation or mobile sources and stationary sources. In the trans-
portation category, a major emitter is the conventional gasoline-
powered automobile, although all combustion engines contribute to
the overall atmospheric POM burden. Because they are ubiquitous and
are known to be contributors to POM concentrations in most urban
areas, motor vehicles require close scrutiny. The category of stationary
sources of POM encompasses a wide variety of processes that can be
local contributors to POM concentrations. It has been customary to
subdivide this category into heat and power generation, refuse burn-
ing, and industrial activities. Indoor POM emissions must also be con-
sidered in this assessment.
Transportation Sources
GASOLINE-POWERED VEHICLES
A significant mobile source of atmospheric POM is the conventional
automobile, powered by a spark-ignited internal-combustion engine.
-------
16 PARTICULATE POLYCYCLIC ORGANIC MATTER
The technical literature on vehicular effects is sparse, mainly because
of the difficulties associated with the type of experimentation re-
quired. In the last few years, results of investigation in this area have
been published;528 heightened interest in air pollution has resulted
in the initiation of comprehensive programs, which are in various
stages of completion. The available published literature and some
of the preliminary results from current programs are reflected in this
summary. Unpublished or incomplete data are included with the
recognition that additional tests may vitiate some of the preliminary
judgments.
An assessment of the current vehicular benzo[a] pyrene contribu-
tion, compiled from nationwide fuel-consumption data, is summarized
in Table 3-1.
As will be apparent in the following discussion, most efforts have
been directed at estimating the automobile contribution, with less
emphasis on trucks and buses. One study343 is available on benzo[a]-
pyrene emission from gasoline-powered trucks; it shows a wide varia-
tion in emission factors, from 70 to 1,500 Mg/gal-
The contributions of gasoline-powered vehicles can be separated
into vehicular effects and fuel-composition effects. The first category
includes the effects of air : fuel ratio or mixture stoichiometry, emis-
sion control devices, operating modes, deterioration, and combustion-
chamber deposits. The second includes effects of such variables as
aromaticity, fuel POM level, additives, and lubricants.
Effects of Vehicular Characteristics Efficient combustion is
enhanced by the presence of excess air, i.e., air: fuel ratios greater
than stoichiometric. Air: fuel ratios less than stoichiometric lead to
TABLE 3-1 Estimated Benzo[a] pyrene Emission in the United States
Vehicle Type
Gasoline-powered
Automobiles
Trucks
Diesel-fuel-powered
Trucks and buses
Total
Fuel Consumed,
gal/year
56.4 X 10'
24.2 X 10'
5.8 X 10'
Benzo [a] pyrene
Emission Factor,
Mg/gal
170"
-500"
62"
Benzo[a] pyrene
Emission,
tons/year
10
-12
0.4
-22
" Data from Hangebrauck et al."3
b Data from Begeman and Colucci.42
-------
Sources
17
the products of incomplete combustion, such as carbon monoxide and
unburned and oxygenated hydrocarbons. Before the current concern
for reducing vehicular emissions, most vehicles operated with fuel-
rich carburetion to promote smooth performance and readily ac-
cessible power. Modifications in post-1967 vehicles have resulted in
"leaner" fuel-air mixtures and, the data suggest, in significantly lower
POM emissions. Table 3-2 shows a compilation of available data;
vehicular variables will obviously influence the values shown but
should not affect the trends indicated. Automobile exhaust POM is
generally referred to in terms of the benzo[a] pyrene emission, pri-
marily because of its cited carcinogenicity and the fact that more data
are available on this material than on any other. Data are becoming
available on prototype emission control devices, such as thermal
reactors and catalytic converters, and a preliminary figure is included
for comparison.
It is apparent from these data that the introduction of presently
used emission control devices resulted in about an 85% reduction in
benzo[a] pyrene emissions from the pre-1965 levels. The data in
Table 3-2 have been selected from representative recent research on
vehicles operating over cyclic test conditions that approximate driv-
ing patterns found in actual customer use. Prototype emission con-
trol devices, such as thermal reactors and catalyst systems, result in
a continuing downward movement of vehicular POM emissions.
Other variables discussed in this section are less important than the
vehicle effects.
The effect of oxidizing and reducing atmospheres on incomplete
combustion and POM formation is important in estimating vehicular
emissions. Recent data42 indicate that benzo[a] pyrene production at
an air: fuel ratio of 10 : 1 is 30 times higher than at a ratio of 14 : 1.
TABLE 3-2 Automotive Benzo[a]pyrene Emission Factors
Benzo [a] pyrene
Emission Factots,
Source fig/gal of Fuel Consumed
Uncontrolled car (1956-1964) 170"
1966 Uncontrolled car 45-70*
1968 Emission-controlled vehicle 2(W30C
Advanced systems <10d
a Data from Hangebrauck et al.343 and Begeman and Colucci.42
6 Data from Gross.3'6
c Data from Begeman and Colucci42 and Gross.3''
a Estimated from Hoffman et al3S 2 and Faust and Sterba.264
-------
18 PARTICULATE POLYCYCLIC ORGANIC MATTER
Hoffman et al.382 suggest that benzo[a] pyrene emission is 10 times
higher at "rich" carburetion (2.85% CO) than at "lean" carburetion
(0.9-1.4% CO). This effect is the central reason for the lower POM
emissions from current emission-controlled vehicles. The effect of
engine operating temperature is closely related to this aspect; cold
engines operate in a "choked" or "rich" condition, indicating that
POM emissions would be maximized in cold starts and minimized
in hot engine operation.
A clear trend toward higher POM emissions with increasing engine
life has been documented by several workers. Hangebrauck et al.343
observed a sharp increase in benzo[a] pyrene emission rates of auto-
mobiles as they approached the 50,000-mile age, the rates being about
5 times higher than those of lower-mileage (e.g., 5,000-mile) vehicles.
Begeman and Colucci,42 who studied oil consumption effects, report
a tenfold increase in benzo[a] pyrene emission when oil consumption
is increased from 1,600 miles/qt to 200 miles/qt. These workers also
found that benzo[a] pyrene from the combustion chamber was pref-
erentially concentrated in the crankcase; eight times more benzo[a]-
pyrene entered the crankcase than left the exhaust system (at normal
oil consumption rates). These data help to explain the higher POM
emission rates of older vehicles. As cylinder wear increases, the lubri-
cant concentration in the upper cylinder increases and the heavy
lubricant molecules provide convenient intermediates for POM for-
mation.
As vehicles accumulate mileage in normal consumer use, deposits
form in the combustion chamber. The nature and composition of
these deposits have been shown to influence total hydrocarbon ex-
haust emissions; as mileage is accumulated, total emissions increase
until a stabilized condition is reached at several thousand miles.
Gross316 has shown that the condition of the deposit exerts a signifi-
cant effect on POM emission levels; the POM emission levels are about
twice as high in a vehicle with stabilized deposits from operation with
leaded fuel as in the same vehicle with stabilized deposits from opera-
tion with unleaded fuel. Another study382 sees no effect of combus-
tion-chamber deposits on POM emissions.
Effects of Fuel Composition The presence in fuel of precursors
of radical intermediates would be expected to facilitate POM forma-
tion; i.e., the pyrosynthetic path would be shortened. Conjugated
dienes and aromatics in the fuel should provide the maximal enhance-
ment of fuel-related POM formation. The literature does point to fuel
-------
Sources 19
composition as having an important effect, but simple judgments as
to the advantages and disadvantages of compositional modifications
are confounded by the number of variables in the vehicle-fuel-exhaust-
system relation.
Early research on the effects of fuel components on POM emissions
pointed clearly to increased aromatic content of fuel as a cause of
higher POM exhaust emissions. For example, Boubel and Ripperton79
showed that a benzene-fueled engine produced 10-30 times more
POM than an engine using «-hexane, cyclohexane, or hexene-1. Hoff-
mann and Wynder390 reported that higher emissions of benzo[a]-
pyrene and benz[a] anthracene resulted from blends of 50% o-xylene
and 50% benzene than from gasoline, pure paraffins, and pure olefins.
Hoffman et al.382 diluted unleaded, high-aromatic gasoline with pure
isooctane and achieved dramatic reduction in benzo[a] pyrene emis-
sions. Begeman40 reported higher benzo[a) pyrene emissions with test
fuels containing high POM and aromatic concentrations than with
commercial gasolines. Most of these studies have been carried out with
synthetic blends, as opposed to gasolines of conventional compositions.
To assess accurately the role of fuel in the question of atmospheric
POM, it is imperative to use realistic compositions in consumer driv-
ing conditions.
Gross316 reports that, when full-boiling-range fuels were used in
well-maintained vehicles operated under federal requirements for
testing 1968-1971 vehicles, POM emissions increased by 36-74%
in an uncontrolled vehicle and 8-34% in an emission-controlled vehi-
cle as fuel aromaticity was increased from 12 to 46%.
A more realistic picture of the effect of gasoline composition can
be obtained using Gross's data:316 When engine tests using a leaded
fuel of low to intermediate aromaticity, with stabilized combustion-
chamber deposits, are compared with tests using an unleaded fuel of
high aromaticity, also with stabilized deposits, no dramatic effect of
fuel aromaticity is apparent. Thus, potential increases in POM
emissions due to higher fuel aromaticity are offset by changes in the
nature of the combustion-chamber deposits when unleaded fuel is
used. These data must be regarded as tentative, because this program
is in progress, and other research in progress does not support these
conclusions.382 However, the benefits to be gained with future
control devices operating on unleaded fuel seem to outweigh greatly
the effects of fuel composition, such as aromaticity. Research in this
area is aimed at unraveling those effects, which now appear to be more
complicated than previously assumed.
-------
20 PARTICULATE POLYCYCLIC ORGANIC MATTER
The POM content of gasoline has been shown to affect POM emis-
sions from vehicles. Begeman and Colucci41 estimate that as much as
36% of the benzo[a] pyrene in the exhaust gas can be attributed to
the fuel benzol a] pyrene content; Gross316 estimates that 15-30%
increases in POM emissions can be obtained when fuel POM is varied
between the concentration extremes found in the field. The actual
effect was smaller in the controlled vehicle than in the uncontrolled
vehicle. These results are consistent with the Badger mechanism
outlined previously: High-molecular-weight fuel components and
lubricant losses to the combustion chamber will result in higher
POM emissions.
DIESEL-FUEL-POWERED VEHICLES
Any critical discussion of the relative contribution of diesel engines
to the atmospheric POM concentration must be qualified by the sup-
position that the vehicle is operated under rated load conditions,
i.e., is not overloaded. In normal use, the most objectionable features
of diesel operation (soot formation, odor, etc.) are apparent when
the engine is overfueled. Begeman and Colucci42 have determined
that a diesel engine, operated on a bus-driving cycle, emitted 62 jug
of benzo[a] pyrene per gallon of fuel. Reckner et al.626 showed that
PO M emissions from a diesel test engine increased with load up to
half-load, leveled off, and then dropped sharply at full load. Idle
operation resulted in high POM emissions, presumably because of lower
combustion-chamber temperatures.
Oil consumption in diesel engines can be somewhat higher than
in spark-ignited engines. However, Begeman and Colucci42 point out
that, because diesel combustion chambers do not operate under
vacuum, lubricating oil should not be drawn into the ignition area;
thus, POM from this source is probably not significant.
The only published data626 on the effects of fuel characteristics
on diesel POM emissions indicate that fuel aromaticity is not related
to exhaust POM levels. The test fuels ranged between 5 and 23%
aromatic content, with the fuel POM consisting primarily of pyrene,
anthracene, and fluoranthene. There was no detectable benzo[a]-
pyrene.
Although the variables discussed above were evaluated in engine-
laboratory conditions, it is apparent that actual on-the-road operation
of diesel-powered vehicles can result in higher POM emissions, owing
to overloading, poor maintenance, and so on. The objectionable fea-
-------
Sources 21
tures of diesel truck and bus operation, such as smoking and odor,
might be associated with higher POM emissions, and additional re-
search in this area should be fruitful.
MISCELLANEOUS TRANSPORTATION SOURCES
Quantitative data are generally lacking for diverse mobile sources of
POM, such as aircraft engines and various nondiesel two-cycle engines,
e.g., lawnmowers, outboard motors, and motorcycles. Aircraft and
turbine engine operation has apparently never been surveyed for
POM emissions.
A study of POM in the exhaust gas from two-cycle engines has been
reported.408 Two-cycle engines, which do not have crankcases, operate
on a mixture of premixed oil and fuel, the oil being the sole source
of lubrication in the system. The data suggest that these engines yield
large amounts of benzo[a]pyrene, with an emission factor of 11,000
Mg/gal for an oil: fuel ratio of 1 : 33, and that benzo[a] pyrene yields
are a direct function of oil concentration in the fuel. These findings
are consistent with the effect of oil consumption on exhaust-gas
benzo[a] pyrene found in four-cycle engines.42 In the most extreme
conditions, the oil: fuel ratio reported by Begeman and Colucci42
was 1 : 29. It seems obvious that the presence of higher-molecular-
weight components than normally found in the gasoline boiling
range has a positive effect on POM formation.
The major noncombustion transportation source of POM is prob-
ably the degradation of automobile tires in use. Carbon blacks, used
in tire manufacturing, contain POM and other high-molecular-weight
organic compounds (S. S. Epstein, personal communication). Mar-
chesani et al.519 estimate that 4.3 tons of rubber particles from tires
are emitted per day per million people in the United States. The benzo-
[a] pyrene contribution from the degradation can be roughly esti-
mated from the analytic data of Falk et al.,260 an emission rate of
0.3 Ib/day per million people is projected. Although tire degradation
does not appear to be a significant source of benzo[a] pyrene, the
ultimate burning of used tires and vehicles (which are categorized as
refuse burning) may be of far greater importance.
EMISSION CONTROL PROCEDURES
The emission control devices on cars since the 1968 models have re-
duced benzo[a] pyrene emission factors by about 85%, compared with
-------
22 PARTICULATE POLYCYCLIC ORGANIC MATTER
uncontrolled vehicles. This reduction is due to the more efficient
combustion associated with "leaner" air-fuel mixtures. It is antici-
pated that these methods will continue in use until the mid-1970's,
when more stringent controls on total emissions will be required. At
that time, such devices as catalysts or thermal reactors will probably
be required, and the mixture stoichiometry will depend on the partic-
ular systems chosen. These devices should result in additional reduc-
tions in POM emission, as shown in Table 3-2. Reductions in emission
of POM to the atmosphere from automobiles can be projected through
the 1980-1990 period, as the older cars are removed from service and
a greater proportion of vehicles are equipped with advanced emission
control systems. Although current work centers on the effects of
control devices, operating conditions, and fuel composition, surveil-
lance of vehicles in normal customer usage, including those in poor
operating condition, might support the extrapolations made in this
section. The emissions from heavier vehicles (such as gasoline- and
diesel-powered trucks and buses) seem significant, but little work has
been done on defining the emission factors of these vehicles. Knowledge
of these factors is clearly needed. There should also be continued ef-
forts to clarify the effects of fuel composition changes, such as trends
toward lead removal from gasoline. Although these factors seem to
be less important than the vehicle-emission-control system, close
scrutiny of them is nonetheless desirable.
Stationary Sources
Polycyclic organic matter is emitted from a vast number of diverse
stationary sources. Although the complexity and variety of POM
preclude a rigorous assessment of their contribution, it can be seen
from compilations of analytic data taken by the U.S. Public Health
Service in most of the urban areas of the country that some urban
areas close to significant POM sources are subjected to high atmo-
spheric POM concentrations. A recent comprehensive review of POM
sources by Hangebrauck et al.343 summarizes the current knowledge
of the relative contributions of the various stationary sources. Although
emissions from stationary sources consist of a variety of chemical
entities, the practice of using benzo[a] pyrene as an indicator of other
POM is suggested, owing to the dual factors of the demonstrated car-
cinogenicity of benzo[a] pyrene and the relatively large amount of
published data on it.
-------
Sources
23
HEAT AND POWER GENERATION
Coal, oil, gas, and wood are burned in a variety of installations. Hange-
brauck et a/.343 concluded that the most important source of benzo-
[a] pyrene of these four was the inefficient combustion of coal in
hand-fired residential furnaces. Data on all four as producers of
benzo[a] pyrene are shown in Table 3-3. That efficiency of combus-
tion, and not the fuel used, is the controlling factor is emphasized
by the low benzo[a] pyrene emission factor found in power plants
burning crushed or pulverized coal. Oil- and gas-burning units used
for institutional and home heating, as well as steam for process heat-
ing, were also shown to be sources of low POM emission. Although
these data are consistent with our knowledge of POM formation
processes—i.e., reducing conditions and insufficient oxygen—caution
should be used in extrapolating data from some 75 individual sources
to the nation as a whole. As will be pointed out, high ambient air
TABLE 3-3 Estimated Benzo[a]pyrene Emission from Heat and Power
Generation Sources" in the United States
Type of Unit
Gross Heat,
BTU/hr
Benzo [a] pyrene
Emission Factor,
jig/106 BTU
Benzo [a] -
pyrene
Emission,
tons/year
Coal
Hand-stoked residential
furnaces 0.1 X 10'
Intermediate units (chain-
grate and spreader
stokers) 60-250 X 106
Coal-fired steam power
plants 1,000-2,000 X 106 20^100
1,700,000-3,300,000 420
15^0
10
1
Oil
Low-pressure air-atomized 0.7 X 106
Other 0.02-21 X 106
Gas
Premix burners
Wood
0.01-9 X 10s
900
100
20-200
50,000
2
40
"Data from Hamburg,336 Hangebrauck ef al, 343 Muhich etal.,S5° U.S. Department of
Agriculture,764 U.S. Department of Health, Education, and Welfare,7 6 7'7 7" U.S. Depart-
ment of the Interior,1" Wadleigh,7" and L. McNab (personal communication).
-------
24 PARTICULATE POLYCYCLIC ORGANIC MATTER
POM concentrations in a particular region can be associated with local
fuel practices. It is apparent that substitution of energy sources that
are inherently more efficient than coal combustion in residential
units may be a short-term solution in areas of high POM emission.
No firm data are available in the literature on the extent of POM
emission from wood-burning combustion units. The growing popu-
larity of home fireplaces, as well as rural heating demands, call for
evaluation of this factor. An emission factor of about 50,000 Mg of
benzo[a] pyrene per million BTU is used to estimate the wood-burning
contribution.767
The total benzo[a] pyrene emissions from heat and power genera-
tion sources shown in Table 3-3 must be regarded as speculative.
This applies most directly to the coal- and wood-burning residential
usage figures. The possible error in these approximations is such that
the contribution of heat and power generation sources to the atmo-
sphere cannot be quantified; an estimate of 500 tons of benzofa]-
pyrene emitted per year to the atmosphere appears justified as an
upper limit.
REFUSE BURNING
The intentional combustion of solid wastes as a method of disposal,
as well as accidental or naturally occurring uncontrollable combus-
tion processes, can contribute significantly to overall POM emissions.
Such sources of POM should come under increasing scrutiny in view
of the increasing solid-waste disposal problem in the United States
today. Unfortunately, the very diversity and nature of these sources
has led to great uncertainties as to their actual contributions to atmo-
spheric POM concentrations.
The review by Hangebrauck et a/.343 cites benzo[a] pyrene emission
factors for municipal and commercial incineration of such wastes as
those collected from households, business, and restaurants, as well
as for burning of municipal and agricultural refuse and junked auto-
mobile parts. Benzo[a] pyrene emissions from these sources vary
widely and reflect the importance of efficient combustion in reducing
POM emissions. Large (50-250 tons/day) municipal incinerators had
benzo[a] pyrene emission factors of 0.1-6 jug/lb of charged refuse,
and commercial (3-5 tons/day) incinerators had factors of 50-260
jLtg/lb. Data show a benzo [a] pyrene emission factor of about 150
jug/lb of charged refuse for open burning of municipal wastes, as well
as for grass clippings, leaves, etc. Significantly, the destruction of auto
-------
Sources 25
components in test "open-burning" facilities yielded a benzo[a] pyrene
emission factor of 1.3 X 104 //g/lb of refuse.343
In their summation, Hangebrauck et al.343 conclude that about 20
tons of benzo[a] pyrene are emitted from these sources per year. More
recently,336'550'764'774-777.798 significantly higher emissions from these
sources have been suggested, reflecting higher estimates of total nation-
wide refuse burning, rather than appreciably different emission factors.
These newer data are compiled in Table 3-4. The largest single iden-
tified contributor listed is coal refuse bank burning (L. McNab, per-
sonal communication), a commonplace occurrence in mining areas.
These banks of coal-mining refuse (coal, shale, calcite) can be spon-
taneously ignited and will burn for long periods in sufficient combus-
tion conditions.
In general, the tonnage figures ascribed to the various refuse burn-
ing classifications must be regarded as order-of-magnitude approxi-
mations. The highly speculative nature of the emission factors used in
the publications cited does not inspire a high level of confidence in
the derived estimates. An estimate of 600 tons of benzo[a] pyrene
TABLE 3-4 Estimated Benzo[a] pyrene Emission from Refuse-Burning in the
United States
Benzo [a] pyrene
Source of Benzo [a] pyrene Emission, tons/year
Enclosed incineration
Municipal
-------
26 PARTICULATE POLYCYCLIC ORGANIC MATTER
emitted per year appears to be the best available value on the basis
of current knowledge.
INDUSTRIAL ACTIVITIES
The major direct petroleum-industry source of POM is the catalytic
cracking process by which organic molecules in crude oil are broken
down into the lighter components used in the manufacture of motor
gasoline, heating oil, aviation fuel, etc. The cracking takes place in
the presence of a catalyst, which can become deactivated through the
deposition of carbon, or coke, on the active sites. It is in the regenera-
tion of the catalyst, through the combustion of the coke on the cata-
lyst surface, that benzo[a] pyrene and other POM are formed. These
emissions are finally passed either to the atmosphere or to a carbon
monoxide waste-heat boiler. The latter device, originally designed to
make use of the waste heat from carbon monoxide gas, functions as
a direct-flame afterburner and removes almost all the POM from the
effluent being emitted to the atmosphere.
As can be seen from the data in Table 3-5, the contribution of
catalytic cracking processes to the atmospheric POM concentrations
is a function of the proportion of units equipped with carbon mon-
oxide waste-heat boilers. The various catalytic cracking units listed in
the table represent both the moving-bed catalytic systems [Thermofor
(TCC) and Houdriflow (HCC)] and the fluidized-bed system [fluid
catalytic cracking (FCC)]. A recent survey627 suggests that a greater
proportion of the units are equipped with carbon monoxide boilers
than were so equipped in 1967.343 The results of the later survey show
an annual contribution of 6 tons of benzo[a] pyrene from refinery
catalytic cracking operations. The total represents a reduction to about
one third of previously published estimates.
Another petroleum-industry process of interest is the air-blowing
of asphalt. This procedure is designed to yield materials of higher
softening point for roofing applications. The effluent from air-blow-
ing may contain many hydrocarbons, including POM. In the one
test of an actual process,343 however, very little benzo[a] pyrene was
found in the benzene-soluble fraction of the particulate matter. The
total contribution of asphalt air-blowing is estimated at less than 0.03
ton of benzo[a] pyrene per year.
Catalytic cracking of petroleum and air-blowing of asphalt are the
most obvious sources of POM emission in the petroleum industry,
but there may be miscellaneous other processes that have not been
-------
Sources 27
TABLE 3-5 Estimated Benzo[a]pyrene Emission from Catalytic Cracking
Sources'2 in the United States
Type of Cracking Unit6
FCC
no boiler
CO boiler
Subtotal
HCC
no boiler
CO boiler
Subtotal
TCC (air-lift)
no boiler
CO boiler
Subtotal
TCC (bucket-lift)
no boiler
CO boiler
Subtotal
Total
Petroleum Consumption,
million barrels/year
424
1,230
1,654
14
55
69
27
118
145
17
75
92
1,960
Benzo[a]pyrene
Emission, tons/year
0.08
0.02
0.10
3.4
0.0
3.4
2.4
0.0
2.4
0.0
0.0
0.0
5.9
" Data from Hangebrauck et al.3"3 and Oil and Gas Journal.6"
6 FCC, fluid catalytic cracking; HCC, Houdriflow moving-bed system; TCC, Thermofar mov-
ing-bed system.
evaluated. For example, the common refinery practice of flaring
waste gas might be a source of POM; modern combustion controls
on flares would be expected to remove these sources from consider-
ation.
The industrial emissions cited above were measured directly; that
is, the effluent stream itself was analyzed for POM. Many other
industrial sources are not amenable to such direct sampling, and
indirect means-often imprecise-have been used to estimate their
emissions.
Some of the processes considered include coke production in the
iron and steel industry; carbon black, coal-tar pitch, and asphalt-hot-
road mix processes; and general chemical processes. The analytic
procedure used in evaluating all such sources has been to sample the
atmosphere in the immediate vicinity of an expected emission source
-------
28 PARTICULATE POLYCYCLIC ORGANIC MATTER
or complex of sources. This necessarily leads to less accurate estimates
than direct sampling. Except for coke production, none of the indus-
trial processes considered above contributes significant amounts of
benzo[a] pyrene to the total atmospheric concentration. This conclu-
sion is obscured somewhat by the presence of other local emission
sources already discussed, such as residential coal-fired furnaces.
There is evidence711 that high benzo[a] pyrene emissions are asso-
ciated with the gaseous discharge of coke ovens. In the United States,
recent activities of the National Air Sampling Network and the Penn-
sylvania State Department of Public Health support the belief that
iron and steel works do contribute to higher atmospheric benzo[a] -
pyrene concentrations in the areas surrounding them. Corresponding
studies in areas outside the United States459'740 lead to similar con-
clusions. An emission factor, admittedly crude, for benzo[a] pyrene
emission from coke effluents has been calculated711 at 1.8 g/ton of
coke. Application of this factor to estimated nationwide coke dis-
charges results in a predicted emission of 192 tons/year.
It is obvious that many more industrial processes may contribute
to atmospheric benzo[a] pyrene. They will constitute localized sources
and can be expected to lead to increased atmospheric concentrations
at local sampling sites.
Industrial emission of benzo[a] pyrene is summarized in Table 3-6.
INDOOR POM EMISSION
Although the outdoor environment has received a fair amount of
study in terms of POM, little is known of the sources and magnitude
of the indoor burden. The possible sources in residential structures are
improperly vented furnaces and incinerators, tobacco smoke, and leak-
age from the outdoors. In industrial plants, many of the processes re-
ferred to previously can, if not controlled properly, emit POM to the
indoor environment.
TABLE 3-6 Summary of Estimated Industrial Benzo[a] pyrene Emission in
the United States
Benzo [a] pyrene Emission,
Source of Benzo [a] pyiene tons/year
Petroleum 6
Asphalt air-blowing <1
Coke production 200
-------
Sources 29
In the only published work, Stocks and co-workers726'733 studied
polycyclic hydrocarbons and smoke in garages and offices and reported
the concentrations of benzo[a] pyrene, benzo[ghi] perylene, pyrene,
and fluoranthene. Their data suggest that office sites have 25-70%
lower POM concentrations than found in the immediate outdoor
environment; the POM concentrations in bus and car garages were at
least as high as and usually somewhat higher than those of the ambient
air.
One major source of nonindustrial indoor POM pollution is tobacco-
smoking.831 During smoking, the mainstream smoke is inhaled and,
in the interval between puffs, the sidestream smoke escapes into the
environment. The use of one unfiltered cigarette (85 mm) releases
30-50 mg of "tar," which contains 0.10-0.15 Mg of benzo[a] pyrene,
0.20-0.30 Mg of pyrene, and 0.25 Mg of chrysene. In a medium-size
room (40 m3), three smokers can pollute the air with 2-4 Mg of benzo-
[a] pyrene, 5-8 Mg of pyrene, and 6 Mg of chrysene per 1,000 m3 of
air. Depending on ventilation and smoking activities, the indoor pollu-
tion by cigarette sidestream smoke can be significantly higher; for
example, Galuskinova288 reported 28-144 Mg of benzo[a] pyrene
per 1,000 m3 for a beer hall in Prague.
The most important source of personal air pollution is the main-
stream smoke of tobacco products. For example, a smoker of 30 un-
filtered cigarettes (popular 85-mm U.S. brand) inhales about 1.0
Mg of benzo[a] pyrene daily. Stated differently, 1,000 m3 of cigarette
mainstream smoke contain about 100,000 Mg of benzo[a] pyrene,
compared with 0.01-74 Mg/1,000 m3 found in polluted air. Although
the concentration of benzo[a] pyrene in a given volume of air and
in the same volume of cigarette smoke are not directly comparable,
they do provide some insight into the relative importance of tobacco-
smoking in pollution with POM.
EMISSION CONTROL PROCEDURES
Efforts aimed at improving the combustion efficiency of most of the
processes covered in this section would be obvious first steps toward
control of POM emission. However, the main contributors to the
POM emission of heat and power generation are small coal furnaces
and wood burning. Neither is economically amenable to better con-
trols, so alternate fuel sources would be the preferred solution.
With regard to refuse burning, more efficient incineration equip-
ment in commercial, industrial, and apartment-building sources would
-------
30 PARTICULATE POLYCYCLIC ORGANIC MATTER
be appropriate. Their relative contribution diminishes in comparison
with open burning, especially coal refuse burning and, to a lesser ex-
tent, forest and agricultural burning. Coal refuse burning could sim-
ply be eliminated by proper attention to refuse accumulation prac-
tices, and intentional forest and agricultural burning could be dis-
continued.
Polycyclic organic matter emissions from catalytic cracking in the
petroleum industry are well on their way to effective control through
the increasing use of carbon monoxide waste boilers. The contribution
from coking emissions in the iron and steel industry must be more ac-
curately assessed.
The contribution of stationary sources to the total POM emission
inventory, although poorly quantified, appears to be large. Latest
estimates of gross tonnage are to be viewed as a first approximation
and may be valid to within a factor of 10. Even with this qualification,
the stationary-source contribution probably accounts for 90% of the
total nationwide POM emission. It is important to note that emission
by stationary sources is usually highly localized, in contrast with that
by mobile sources, and results in high atmospheric POM concentra-
tions in the vicinity of major emitters.
Owing primarily to these localized emissions, comprehensive epi-
demiologic studies should be initiated in geographic areas that are
subject to high atmospheric concentrations of POM. A particularly
fruitful study might be done in the Appalachian Mountain-Mississippi
River area.
Stationary-source emission factors must be validated and the analyses
extended to include as many additional kinds of POM (i.e., other than
benzo[a] pyrene) as feasible. In particular, the importance of hand-
fired furnaces burning coal or wood must be critically evaluated. Coal
refuse burning is in the same category; the large, highly speculative
value chosen for this contribution requires verification. Alternate
disposal methods for coal refuse should be developed in the interim.
TABLE 3-7 Summary of Benzo[a] pyrene Emission by Stationary Sources in
the United States
Benzo[a]pyiene Emission,
Source of Benzo [a ] pyrene tons/year
Heat and power generation ~500
Refuse burning ~600
Coke production 200
-------
Sources 31
Emission of POM from coke production also requires scrutiny; alter-
nate manufacturing practices in the iron and steel industry should be
developed in case such emission must be controlled.
The best available current data suggest the stationary-source benzo-
[a] pyrene emission shown in Table 3-7.
GENERAL NATURE OF POM EMISSIONS
Individual POM Emissions
The polycyclic organic molecule mentioned most prominently here has
been benzo[a] pyrene. This material has been identified as a prominent
constituent of most of the processes discussed and has also been shown
to be a potent carcinogen. Although these facts confirm the importance
of benzo[a] pyrene, many other materials emitted in the same pro-
cesses have some carcinogenic activity.
It has been felt that benzo[a] pyrene could be used as an indicator
molecule, implying the presence of a number of other components
of similar structure. Several workers153'667'670'673 have reported
numerous types of POM in urban air, including pyrene, anthanthrene,
benz[a] anthracene, benzofluoranthenes, dibenzanthracenes, chrysene,
phenylenepyrene, benzoperylene, coronene, fluoranthene, and alkyl
derivatives of these compounds, as well as benzopyrenes. (See Table
2-1 for some of these materials and their structures and properties.)
There have been attempts to develop relations between these indi-
vidual compounds (such as the ratio of benzofa] pyrene to pyrene
and coronene to pyrene) as a function of their source. For example,
ratios shown in Table 3-8 have been determined for vehicular emis-
sions, industrial emissions, refuse burning, and heat generation.
It is obvious that these ratios can vary widely as a function of
emission source. Before benzo[a] pyrene can be used as an accurate
barometer of the entire class of POM, more information on these
ratios, as well as on the carcinogenic significance of the other POM
molecules, will be required.
Area-Concentration Relations
It is evident that three major stationary sources-coal-fired and
wood-fired residential furnaces, coal refuse fires, and coke production-
account for more than 90% of the annual nationwide benzo[a] pyrene
emission. Of the remaining sources, the transportation contribution
-------
32 PARTICULATE POLYCYCLIC ORGANIC MATTER
TABLE 3-8 Ratios of Individual POM Molecules by Emission Source
Pyrene: Benzo[ghi]perylene: Benz [a] anthracene:
Emission Source Benzo[a]pyrene Benzo[a] pyrene Benzo[a]pyrene
Automobiles"
Trucks
Gasoline-powered*
Diesel-fuel-powered 6
Catalytic cracking c
Incinerators0
Heat generation a
7:1-24:1
50:1-90:1
<1: 1-50:1
<1:1-23:1
6:1-16:1
1:1-1,000:1
2:1-5:1
_
-
0.3:1-3:1
0.2:1-1:1
-
1:1-2:1
_
-
-
_
-
"Data from Gross,316 Hangebrauck e/ai,343 Hoffmaneta/.,3'4 Kotinef al,"
and Sawickief ai674
6Data from Hangebrauck et al.342
c Data from Hangebrauck et al.34 3
d Data from Falk et al.2 60
is probably the most significant, in that it pervades all segments of
the nation. It is instructive to consider the predominant areas of the
country with regard to these stationary sources, as in Table 3-9.
When the areas of major emissions are grouped, it is obvious that
POM emissions are very high through the southeastern region along
the Appalachian Mountains, as well as in the area to the immediate
west as far as the Mississippi River and north to the Great Lakes. Al-
though more quantitative extrapolations are not warranted, we can^
consider the aerometric data now available through the National
Environmental Research Center of the Environmental Protection
Agency as indicative of the major urban areas in which POM concen-
trations may constitute health problems. A survey of the data for the
winter of 1969, in which benzo[a] pyrene concentrations are reported,*
is enlightening. Of the 40 U.S. cities in which the winter benzo[a] py-
rene concentration exceeded 5 jug/1,000 m3, 34 are in the region just
defined, as are 44 of the 53 cities with concentrations in excess of
4 jug/1,000 m3. The densely populated Northeast and the Far West
are conspicuous by their relatively low atmospheric benzo[a] pyrene
concentrations; the Los Angeles Basin, with its high vehicle and human
population densities, has concentrations of 1.5-3 jug/1,000 m3.
In the only other set of determinations of relative source contri-
butions to the atmospheric POM burdens, Colucci and Begeman153
have calculated that, in Detroit, motor vehicles contribute 5% of the
*Data from National Aerometric Data Bank, P.O. Box 12055, Research Triangle
Park, North Carolina 27709.
-------
Sources
33
TABLE 3-9 Contributions to National Totals of Benzo[a] pyrene by Source and
State"
Benzo [a] pyrene
Emission Source
Coal-fired
furnaces
Coal refuse
burning
Coke production
Stale
Illinois
Ohio
Wisconsin
Michigan
Indiana
W. Virginia
Pennsylvania
Kentucky )
Colorado V
Virginia \
Pennsylvania
Ohio
Indiana
Alabama )
Maryland >
W. Virginia )
Fraction of U.S. Total, %
State Group of States
22
12
10
8
6
45
25
20
29
16
14
20
58
90
79
a Based on data from Hamburg,336 L. McNab (personal communication), Muhich etaL,ssa
VS. Department of Agriculture,'"'' U.S. Department of Health, Education, and
Welfare,"7'"4 and U.S. Department of the Interior.7'"
benzo[a] pyrene in the downtown area, 18% of that in the freeway area,
and 42% of that in the atmosphere in the suburbs. Similar studies were
made in New York City151 and Los Angeles,152 but the same types
of calculations were not possible. In both cases, the data would permit
only correlation techniques, which indicate positive statistical relations
of benzo[a] pyrene with both automotive and stationary sources.
These data indicate that, in the absence of other major sources, as
in some suburban locations, the vehicular contribution may be as
high as 50%. Aerometric data indicate that, when this relative vehicu-
lar contribution is high, the local atmospheric POM concentrations
are low.
The implications of these trends are evident: Epidemiologic data
should be obtained in areas of high and low POM concentration to
establish the effect of atmospheric contamination by POM. Until
such studies are made, the nature and degree of sour.ce controls re-
quired will be unknown.
-------
34 PARTICULATE POLYCYCLIC ORGANIC MATTER
SUMMARY AND RECOMMENDATIONS
Polycyclic organic matter can be formed in any combustion process
involving hydrocarbons. Naturally occurring POM emission to the
atmosphere does not appear to be significant. The major technologic
emissions include those from transportation sources and such station-
ary sources as heat and power generation, refuse burning, and indus-
trial processes.
The internal-combustion engine is a ubiquitous source of POM.
Current efforts to reduce total vehicular emissions have reduced POM
emissions, and projections of future control levels point toward a con-
tinuing and marked decline. However, such projections presuppose
properly maintained and operated vehicles; close scrutiny should be
directed at the effects of deterioration of automobile emission control
devices and the use of diesel-fueled vehicles in overloaded conditions.
Research efforts to determine the effects of fuel composition and of
advanced emission control devices should be continued. Polycyclic
organic matter emissions from aircraft should be assessed, as well as
those from local mobile sources, such as two-cycle engines.
POM emissions from major stationary sources are poorly quantified.
Available data suggest that coal-fired residential furnaces, coal refuse
bank burning, and coke production from the iron and steel industry
are responsible for the bulk of the nationwide POM emission. How-
ever, serious reservations may be expressed as to the validity and mag-
nitude of these data. It is noted that atmospheric concentrations of
PO M are high in areas in which the cited sources are concentrated. In
addition, effective control procedures for these processes are lacking.
Substitution of alternate fuels or more efficient combustion processes
and discontinuance of coal refuse storage practices seem to be the
only appropriate methods for the restriction of coal-related emission;
the emission associated with coke production requires additional re-
search on control procedures and source analysis.
Current data suggest the following relative contributions of major
source categories to the total POM emission inventory (expressed in
terms of annual estimated benzofajpyrene emissions): heat and
power generation, 500 tons/year; refuse burning, 600 tons/year; coke
production, 200 tons/year; and motor vehicles, 20 tons/year.
These data represent nationwide; estimates based on extrapolations
from individual source emissions. In specific areas, the relative con-
tribution of any given source may differ significantly from that im-
-------
Sources 35
plied by the nationwide figures. For example, the vehicular source may
be the major contributor in suburban areas where other major sources
are absent. Epidemiologic studies using source inventory data and am-
bient atmospheric concentrations are required to assess the importance
of control measures in both high and low atmospheric POM areas.
-------
Atmospheric Physics of
Particulate Polycyclic
Organic Matter
The presence of aerosols in the atmosphere, even in locations very re-
mote from population centers, is well known. Although many sources
of natural aerosol particles have been identified,420 their magnitude re-
mains uncertain. Some typical natural sources of particles are listed in
Table 4-1, which classifies material as "primary" in origin (emitted as
particles directly from a source) or as "secondary" in origin (formed
by condensation or chemical reaction in the atmosphere itself)- From
a preliminary assessment indicated in Table 4-1, it appears that, glob-
ally, three major sources are dust rise caused by wind, sea spray,
and vegetation. The hydrocarbon aerosols from vegetation are believed
to be primarily terpenes.874
Superimposed on the natural background material is a significant
amount of aerosol produced by man. Some major identified sources of
anthropogenic aerosols are listed in Table 4-2. (The table excludes
cigarette smoke, which, although probably small in tonnage, repre-
sents a significant hazard.) This list suggests the importance of com-
bustion as a contributor to the particle population, especially in urban
areas. Although polycylic organic matter represents only a very small
fraction of the total amount of particulate matter in the atmosphere
or associated with combustion sources, it constitutes a very important
36
-------
Atmospheric Physics 37
TABLE 4-1 Some Sources of Natural Aerosols in the Atmosphere0
Source
Estimated Aerosol Production Rate,
tons/day
Primary
Dust rise by wind
Sea spray
Forest fires (intermittent)
Volcanic dust (intermittent)
Extraterrestrial (meteoritic dust)
Secondary
Vegetation: hydrocarbons
Sulfur cycle: SO4=
Nitrogen cycle: NO3
Volcanoes: volatile SO2 and H2 S (intermittent)
Maximal total
20,000-1,000,000
3,000,000
400,000
10,000
50-550
500,000-3,000,000
100,000-1,000,000
1,000,000
700,000
1,000
-10,000,000
" Modified from Hidy and Brock.3"
minor fraction, because of its potential hazard to human and animal
life. In urban areas, it is expected that localized anthropogenic sources
will far exceed natural sources, in contrast with the proportions of the
TABLE 4-2 Some Important Sources of Anthropogenic1 Aerosols in the
Atmosphere"
Source
Estimated Aerosol Production Rate,
tons/day
Primary
Combustion and industry
(potentially containing traces of POM)6 100,000-300,000
Dust rise by cultivation (intermittent)6 100-1,000
Secondary
Hydrocarbon vapors (incomplete combustion, etc.;
may involve traces of POM) 7,000
Sulfates (SO,, H, S-»SO=) 300,000
Nitrates (NOJC-»NO3') 60,000
Ammonia 3,000
Maximal total -700,000
" Modified from Hidy and Brock.3"
6 Assumes 90% emission control of emissions from coal-burning installations.
c United States only.
-------
38 PARTICULATE POLYCYCLIC ORGANIC MATTER
atmosphere as a whale, as indicated in Tables 4-1 and 4-2. Unfortu-
nately, there is little quantitative information on the relative contri-
butions of sources in polluted areas.
Because of the high melting and boiling points of materials classi-
fied as POM (see Chapter 2), the bulk of POM is believed to be linked
with aerosols. It is not known in what form such material generally is
present in aerosols. Available evidence, however, indicates that benzo-
[ajpyrene is identified primarily with soot particles.154'751 POM may
exist as particles of relatively pure material, or it may be adsorbed in
small amounts on other particles. Adsorption is especially important
because of the possibility that POM is more easily assimilated biologi-
cally if it is associated with soot, dust, or other particles (see, for ex-
ample, Kotin and Falk449).
Because much anthropogenic POM is identified with combustion
(see Chapter 3), it is likely that such material is emitted as vapor from
the zone of burning. Either in a stack or in an exhaust pipe, it will
cool and condense on existing particles or form very small particles
itself. If it enters the atmosphere as vapor, it will undoubtedly be
adsorbed on existing particles while undergoing condensation. In fact,
small particles in the atmosphere may act as nucleation centers for
forming POM aerosols when such vapors are supersaturated.
As POM is mixed with aerosols in the atmosphere, it is spread
among particles of widely varied sizes by collision processes. POM-
containing particles are dispersed in air by turbulence and may be trans-
ported great distances from their origin by winds. They are eventually
removed from the atmosphere by sedimentation or deposition, such
as on rocks, buildings, and vegetation. Removal is enhanced by wash-
out from under rain clouds and by rainout from within clouds.
All factors involved in the aging of atmospheric aerosols may be
important in the consideration of POM as a health hazard. Collision
processes change the original size of POM-containing particles, whereas
transport tends to disperse such material over broad areas. Dispersion
tends to dilute POM, making exposure to it less likely, while at once
increasing the likelihood of exposure to it in areas remote from the
original source. Removal processes are significant, in that they rid the
air of PO M, taking it out of circulation and making it less likely to be
inhaled; however, such removal may contaminate edible vegetation,
and POM may thus be ingested by man and animals.
The U.S. Public Health Service has recently published a general review
of the properties of atmospheric aerosols, their measurement, and their
health hazard;775 the bulk of that material will not be repeated here.
-------
Atmospheric Physics 39
Keeping in view the possible significance to health of POM-containing
aerosols, however, the physical properties of POM and related materials
are reviewed here, and then the fate of POM in the atmosphere
in relation to physical changes, transport, and removal of aerosols.
PHYSICAL PROPERTIES
Concentration and Particle Size Distribution
The concentration of aerosols with respect to a range of particle size
is crucial in the evaluation of the penetration of material into the
respiratory system.
Evaluation of particle size distribution is somewhat arbitrary, in
that particle dimensions are difficult to define uniquely. Aerosols in the
atmosphere contain particles of many different shapes, ranging from
spheres to long fibers. Particle size may be defined by two general
methods. The first refers to the dispersed material in terms of an
equivalent geometric, projected area. The second defines size in terms
of some property of the particles, such as settling rate, optical scat-
tering cross section, or ratio of electric charge to mass. Reported
data on size distributions differ somewhat even for the same sample,
depending on the detector used. In this respect, comparisons of data
taken at different times and places may be ambiguous. Much of the
size distribution information is reported in terms of a radius or
diameter of an "equivalent sphere," without careful definition of this
dimension in relation to the sampling or detection method. Despite its
ambiguities, "equivalent" size is adopted here for simplicity in dis-
cussing recent results of size distribution observations.
Because of the great variety of urban sources for aerosols, it would
be difficult, if not impossible, to assess and classify individual emissions.
Some very limited measurements have been made from major sources
like automobile exhausts. However, these have been rather crude and
have remained largely unreported by investigators.
In the case of automobile exhaust, Mueller et a/.,548 using different
kinds of samplers, indicated that about 68% by weight of aerosol
emitted was smaller than 0.3 /xm. Recent preliminary data from the
University of Minnesota aerosol laboratories suggest that automobile
exhaust contains more than 10s particles/cm3, with a mean radius as
small as 100 A. The automobile, then, contributes significant amounts
of aerosols that will remain in the air for extended periods, although
the total mass emitted may be small.
-------
40 PARTICULATE POLYCYCLIC ORGANIC MATTER
Most of the data in the literature775 on aerosol distributions from
stationary sources are confined to particles larger than the submicrom-
eter range. However, it is known that combustion processes produce
copious quantities of nuclei that measure a few tens of Angstroms.
Almost nothing is known about either the size distribution or the
chemical composition of particles from stationary sources that are
smaller than 0.1 fj.m.
Only recently have measurements become available that afford
a reliable idea of the range of particle size in urban atmospheres. It
has been found, for example, that aerosols may be found in city air
ranging in size from a few tens of Angstroms to hundreds of microm-
eters. It is not yet known whether there is significant fractionation
with respect to size of hydrocarbon material.
To measure aerosol particles in the atmosphere over the wide range
of particle size known to exist, more than one instrument is required.
The large-particle fraction—greater than 0.5 Mm in diameter—may be
observed with a variety of techniques, involving either indirect sam-
pling (optical counters) or direct collection of particles (impactors,
filters, or centrifuges). Such equipment has been described in detail
elsewhere.449 For particles less than 0.5 jum in diameter, there are only
a few methods for observation. Among them are the Aitken nuclei
counters,409 small adiabatic expansion chambers allowing nuclei
50 A or more in diameter to grow sufficiently in supersaturated
water vapor to be counted optically, and the electrostatic analyzer,409
a device relying on the specificity of the relation between an equilib-
rium electric charge and particle size (mass) for very small particles.
The latter device appears to operate most satisfactorily at an equiva-
lent radius of 0.02-0.1 jum.
Using a hybrid analyzer system, Husar and colleagues409 have de-
veloped a method for measuring the aerosol size distribution from
about 100 A to over 10 pm in diameter. Some results of this system are
shown in Figure 4-1 for the urban atmosphere in Minneapolis and
in Figure 4-2 for Pasadena. Some typical nonurban data from other
methods374 have been added in Figure 4-2 for comparison. Pasadena
and Minneapolis have their own characteristic patterns, the main dif-
ference between them being in the extent and intensity of photo-
chemical smog production. The Los Angeles area (which includes
Pasadena) has a much more photochemically reactive atmosphere
than Minneapolis. In both cases, the curves illustrate that urban at-
mospheres contain enormous numbers of very small particles, com-
pared with nonurban atmospheres. During the day, the Los Angeles
-------
Atmospheric Physics
41
E
n
I
u
108
107
106
105
104
103
5l Q
*lt> 1Q2
10°
10
1-1
10
,-2
10
,-3
o70ft
A 170 ft
V70ft 1200CST
• 170ft
,-4
10
-2
10
1-1
10°
Dp, Particle diameter
FIGURE 4-1 Particle size distribution of aerosols in Min-
neapolis, Minnesota, taken February 22, 1967. Data are
shown for two different heights above the ground at 11 a.m.
and 12 noon. (Derived from Peterson et al.591)
-------
42
PARTICULATE POLYCYCLIC ORGANIC MATTER
Aerosol Samples
Light to Moderate Smog
Pasadena, California
Sept. 3, 1969
Time (PDT|
00400
A 0820
V1240
O1900
10-
10-' 10°
D Particle Diameter (jam)
10'
FIGURE 4-2 Particle size distribution in Pasadena, Cali-
fornia, on the basis of the Minnesota Aerosol Analyzing
System. Data of K. T. Whitby. (Reprinted with permission
from Hidy and Friedlander.3"74 ) Results of Blifford63 for
nonurban air as measured with an impactor are shown for
comparison.
-------
Atmospheric Physics 43
atmosphere appears to be enriched in small particles, compared with
the night. However, the small-particle concentration in Los Angeles
is generally below that in Minneapolis. The diurnal difference in Los
Angeles is speculatively associated with increased man-made produc-
tion during the day. However, the difference between Los Angeles
and Minneapolis appears to be linked with the high photochemical
reactivity of the Los Angeles atmosphere. Evidently, the condensation
of vapors on particles larger than 0.1 f/m in radius during the day-
light evolution of photochemical smog overshadows the production of
Aitken nuclei in more reactive air (see, for example, Whitby et al.81S).
There is considerable natural variation from place to place and from
tune to time, at least over portions of the observed aerosol size distri-
bution. However, it has been observed that the upper end of the spec-
trum often, but not always, has a remarkable regularity, particularly
in data taken for many cases averaged together. The large-particle por-
tion of the spectrum (greater than 0.1 /xm) often tends, on the aver-
age, to follow a power law form:
dN
n(Dp) = —-= const 0 Df~\ (1)
rfi/p
in which Dp = particle diameter; n(Dp )dDp = the number of particles
in the diameter range, Dp to (Dp + dDp); N = total number of parti-
cles; and 0 = volume fraction. On the basis of many observations141'581
of urban aerosols, the constant in this equation is approximately 0.40,
but it may vary from 0.24 to 0.56. The power of the diameter ranges
between —3 and —5.421 There is no theoretical explanation, as yet,
for the apparent regularity in the upper portion of the "average"
aerosol particle size spectrum. However, some speculation has been
reported.371'421
Because n(Dp) has some regularity, particularly above 0.05 pm in
diameter, it is possible to estimate crudely the dosage to various parts
of the respiratory system expected in urban areas with limited knowl-
edge of the size distribution. Serious questions remain, however, about
the relation between an equivalent spherical diameter, as measured by
the optical and electric devices of the Whitby analyzer, and a parti-
cle size and shape more relevant to deposition on the surface of the
lung.
Particle Shape and Density
The characterization of a particle with respect to shape permits trans-
formation of projected area to, for instance, total area. With informa-
-------
44 PARTICULATE POLYCYCLIC ORGANIC MATTER
tion for particle density, one can calculate particle weight. Because of
the applicability of different particle dimensions to atmospheric con-
ditions or to the effects produced in biologic receptors by particulate
pollution, there is great interest in characterizing particle shapes to
transform from one particle size dimension to another. The signifi-
cance of the different classifications of particle size, shape, and weight
in the evocation of biologic responses is discussed by Davies.183
Hodkinson381 reviewed the effects of particle shape on measurement
of size and concentration. Cartwright119 has illustrated the utility of
a specific particle shape—the ellipsoid of revolution—in characteriz-
ing quartz dust. To measure the particle radius directly associated
with deposition by aerodynamic forces, such devices as the aerosol
centrifuge have been built. Perhaps the best known is that designed by
Goetz et al.300 Other promising devices have also been developed.380'725
The most accurately reported characterization of particle shapes is
probably that of Kotrappa,457 for which a spinning spiral duct725
was used to classify particles of coal, uranium dioxide, and thorium
dioxide aerodynamically. Kotrappa's work illustrates the use of
shape factors to describe irregularly shaped particles. The dynamic
shape factor, K, of an irregularly shaped particle moving at its terminal
settling velocity is the ratio of the drag force action on it to the drag
experienced by a spherical particle of the same mass and density mov-
ing at the same velocity. The Stokes diameter, Z>st, of a particle is de-
fined as the diameter of a spherical particle of the same density and
terminal settling velocity as the irregular particle. The aerodynamic
diameter, Dae, of an irregularly shaped particle is defined as the diam-
eter of a sphere of unit density with the same terminal settling veloc-
ity as the particle. The volume shape factor, av, is defined by:
volume of particle = avDp 3. . . , (2)
in which Dp is the diameter of a circle of the same area as the pro-
jected area of the particle.* The surface shape factor, as, is given by:
surface of particle = asjDp2 • f%\
In Kotrappa's studies^ the volume shape factor, av, remained rela-
tively constant at 0.38 for the coal sample whose projected area diam-
eter was 0.5-4.5 pm. The dynamic shape factor, K, was relatively
constant at 2.0 for particles with diameter less than 2.5 //m but in-
*This is identical with .Dp used in Figures 4-1 and 4-2 for spheres only.
-------
Atmospheric Physics 45
creased slowly for larger particles. The ratio of the projected area diam-
eter to Stokes diameter (Df/Dsl) was constant at 1.45 for £)p less
than 2.5 jum but decreased slowly to 1.35 forZ)p of 4.0 /xm. For
coal,
-------
46 PARTICULATE POLYCYCLIC ORGANIC MATTER
ume—is also important; but it has not been measured directly, be-
cause satisfactory techniques are largely unavailable. Once the size
distribution is known, however, the superficial surface area can be
estimated.
Because of the shape of the aerosol size distribution curve, it has
been found that the total number concentration is strongly associated
with particles less than 0.1 jurn in radius, whereas the volume (or mass
fraction) is associated mainly with particles larger than 0.1 Mm. The
large-particle fraction provides the largest dosage in mass, so it has
generally been accepted as the convenient observable quantity for
monitoring. It is usually measured by means of filters or impactors.775
Some typical values for particle number (N) and mass concentra-
tion (m) are listed in Table 4-3. The organic fraction of the mass con-
centration as measured by the benzene-soluble component is also listed,
with the benzofa] pyrene fraction for comparison.* Of the organic
fraction, a variety of organic compounds have been identified, includ-
ing some materials classified as POM.164 However, the identified
fraction represents only 10% of the organic components of urban
aerosol. Although the total number concentration is often very large
in cities, the mass concentration varies less and rarely exceeds about
200 Mg/m3 in the United States. The benzene-soluble fraction of this
is only a few percent of the total mass, and the concentration of
benzo[a] pyrene is far lower. Even in remote areas, there is a con-
tribution of organic material, as expected from Table 4-1. However,
it is uncertain whether the benzo[a] pyrene in these areas comes from
natural or anthropogenic sources.
Data from the National Air Surveillance Network (NASN) from
1967 to 1969 indicate that the concentrations of benzo[a] pyrene
are highest during the first and last quarters of the year (Table 4-4),
which is consistent with increased fuel consumption in stationary
sources during these periods. In contrast, maxima in the nonurban
stations occur more or less randomly during the year.
The frequency distribution of benzo[a] pyrene concentrations found
for different NASN sites in 1969 is shown in Figure 4-3. Comparison
of the annual average distribution with the distribution for the first
quarter shows the relation of increased benzo[a] pyrene concentration
with the increase in fuel burning that is associated with the winter
months. The curves in Figure 4-3 indicate that the average urban con-
*The benzene-soluble extract is not necessarily equivalent to the total amount
of organic material in the sample, but it is taken to be representative of such a
fraction.
-------
Atmospheric Physics
47
TABLE 4-3 Typical Values of Aerosol Concentration for Different Geographic
Areas (Annual Averages)"
Location
Nomirban
Continental
General
California
Oregon
Colorado
Indiana
Maine
New York
So. Carolina
Maritime
General
Pacific offshore
Oahu, Hawaii
Urban
Continental
General
Los Angeles
Portland
Denver
Minneapolis
Chattanooga
New York
Greenville, S.C.
Maritime
Honolulu, Hawaii
San Juan, Puerto
Rico
N,
particles/cm3*
103-104
103-10"
-
102-104
-
—
—
_
10" -10"
102-104
102-10"
103-106
103-10'
—
io3-ie5
103-105
-
_
-
102-104
-
Mass
Concen-
tration (m),
Mg/m3C
20-80
39
47
14
39
18
29
40
_
19-146
10-49 d
MOO
93
72
110
70
105
105
76
40
77
Benzene-
Soluble
Fraction of m,
Mg/m3
1.1-2.2
2.8
0.9
1.1
2.1
1.2
1.8
2.7
_
1. 5-6.1 d
0.7-6.3*6
7
12.5/
6.6
9.0
6.1
6.9
8.9
7.4
2.3
6.9
Benzo[a]pyrene
Fraction of m,
ng/m3
_
0.48
0.09
0.11
0.25
0.12
0.25
0.43
_
_
-
1.87
2.60
2.52
1.18
4.18
3.63
7.49
0.59
1.42
"Data based on 1969 National Air Surveillance Network observations, except for maritime
data, which are based on Junge,4>0 Holzworth,39' Barger and Garrett,3 * and G. M. Hidy
(unpublished data).
* Aitken nuclei.
c Geometric means.
d Short-term data.
e Chloroform-extractable.
^Recent measurements suggest that this is only about half the noncarbonate carbon frac-
tion in the Los Angeles aerosol.5"'
centration of benzo[a] pyrene in the first quarter is about 2.5 ng/m3
and is an order of magnitude larger than the nonurban concentration.
The average benzofa] pyrene concentration in U.S. cities has de-
creased since 1950 by a factor of nearly 3. Over the same period, the
-------
48
PARTICULATE POLYCYCLIC ORGANIC MATTER
TABLE 4-4 Seasonal Observations of Benzo[a] pyrene and Benzanthrone in Some
U.S. Cities"
Mass Concentration, ng/m
3
Benzo [a] pyrene
Site
Los Angeles
Medford, Oregon
Albuquerque, N.M.
Ashland, Ky.
Chicago
Nashville, Tenn.
Philadelphia
Pittsburgh, Pa.
Greenville, S.C.
Missouri (nonurban)
Pennsylvania
(nonurban)
Quarter
1
2.98
2.60
1.02
21.17
7.20
5.73
6.33
21.32
19.60
0.24
2.52
2
0.79
2.18
0.23
6.38
3.21
1.76
1.69
18.27
2.84
0.16
0.83
3
0.64
1.45
0.29
6.21
1.60
0.77
1.41
6.04
0.66
0.17
1.04
4
3.05
9.97
2.95
9.80
3.52
2.93
6.68
9.37
4.91
0.08
0.54
Benzanthrone
Quarter
1
4.48
8.14
1.47
12.17
4.86
4.76
11.02
9.28
15.52
0.47
0.83
2
1.87
1.69
0.57
3.69
3.38
2.04
1.64
4.75
2.70
0.23
0.57
3
1.74
2.92
0.67
5.47
2.31
1.62
1.60
3.10
1.56
0
1.01
4
6.10
9.69
3.34
6.64
3.78
5.68
3.65
3.91
8.46
0.47
1.00
" Preliminary data from the National Ait Surveillance Network, 1969 (J. B. Clements, En-
vironmental Protection Administration, personal communication).
average concentration in nonurban sites has also decreased somewhat.
The overall reduction in average benzo[a] pyrene concentration is
probably associated with nationwide changes in fuel usage and im-
proved furnace design in stationary sources, with more burning of
natural gas and fuel oil instead of coal.
The Environmental Protection Agency has found that benzanthrone
(7H-benz[d,e] anthracen-7-one) may be an additional indicator of
POM from combustion processes. During 1969 and 1970, samples
from some NASN stations have been analyzed for benzanthrone. For
comparison, some results for seasonal variation over 1969 are tabu-
lated in Table 4-4. In many cases, the benzanthrone concentrations are
larger than the benzo[a] pyrene concentrations. The highest concentra-
tions of benzanthrone occur during the winter and fall quarters. To
date, there are no data that describe benzanthrone as a carcinogenic
agent.
A recent study of Colucci and Begeman152 is an example of a more
detailed short-term urban survey of POM than is available from NASN.
From 1964 to 1965, these investigators found that the concentrations
of benzofa] pyrene and benz[a] anthracene were 4l/2 times greater in
central Los Angeles than at two suburban sites. However, the subur-
-------
Atmospheric Physics
49
10 i-
1.0
0.1
I
I
I
I I I I
I
I
I
I
J
99
15 20 50 70 90 98
Percentage of sites with less than a given concentration of benzo[a] pyrene
FIGURE 4-3 Frequency distribution of benzo[a] pyrene concentration in air
for all U.S. urban and nonurban sites in first quarter with annual average urban
distribution (1969 NASN data).
ban site downwind of the downtown area (on the average) appeared
to have systematically higher benzo[a] pyrene concentrations than the
upwind site. Daily concentrations reported in the Los Angeles area
ranged from 0.1 ng/m3 to over 10 ng/m3, depending on the season.
Benz[a] anthracene concentrations were F/2 times larger than the
benzo[a] pyrene concentrations. Annual average benzo[a] pyrene
concentrations were similar to the NASN data for downtown Los
Angeles. The POM concentrations increased substantially in winter.
Benzofa] pyrene concentrations were higher at night, in contrast with
-------
50 PARTICULATE POLYCYCLIC ORGANIC MATTER
those of other pollutants. All pollutants were higher on weekdays
than on weekends. Benzo[a] pyrene concentration was found to be
correlated with carbon monoxide and lead concentrations, with co-
efficients ranging from 0.6 to 0.9. Benzo[a] pyrene concentration was
also related significantly to those of hydrocarbon vapors, oxides of
nitrogen, and vanadium (a nonautomotive pollutant). Despite the
strong relation to lead, the statistics in the study failed to reveal a
clear identification of benzofa] pyrene emissions with automotive
or stationary combustion sources.
Because of the regularities in the size distribution function, it
appears possible to use correlations of the type suggested by Pasceri
and Friedlander581 and Clark and Whitby141 to estimate properties
of the distribution crudely. The aerosol distribution may be shown in
dimensionless variables. In Figure 4-4, all the observed data for the
distribution in terms of concentration n(Dp )dDp in the range Dp to
(Dp + dDp) are scaled to N and 0, as defined by the spectrum itself.
Thus, in principle, the entire distribution function may be derived
from observations of the total number concentration of particles
and the total volume fraction (or mass concentration). Because this
idea has been tested with only very limited data, it cannot yet be
accepted as a "universal" law. It is of interest in this connection that
the direct correlation between light scattering from aerosols and the
mass concentration129 depends on the existence of the Junge subrange,
or a distribution approximating this power law over the range from
0.1 jLim to about 10 jum in radius. Tests of this correlation in a num-
ber of cities suggest that the integrating nephelometer, for example,
may be a promising instrument for observing mass concentration
because of the relation between light scattering and the size spec-
trum.
An important feature of the Junge subrange, given by Eq. (1), is
that the mass concentration should be constant where this form
holds. A test of the sensitivity of size distribution observed as mea-
sured by such devices as those of Husar et a/.409 is to measure the
mass identified with each size range directly. Corn165 has reported
such data; he found that, at least in Pittsburgh, the Junge subrange
with a power law exponent of —4 was not verified. Instead, the dis-
tribution appeared to fall off with an exponent less than —4 from
0.5 to 3 /xm in radius.
There is very little information about the actual surface area of
aerosols in urban air. One of the few studies that has been made has
been carried out by Corn et al,166 who found that the specific sur-
-------
Atmospheric Physics
101
10°
icr1
51
1-2
10
10-3
10~4
10~7
10-8
10~9
Q
• Clark and Whitby
(Minneapolis)
Friedlander and Pasceri
(Baltimore)
Q Junge (Frankfurt)
A Junge (Zugspitze)
O Cartwright (Sheffield)
Q Cartwright (Buxton)
10-
10-
10°
102
FIGURE 4-4 Normalized correlations of aerosol size
spectra observed in polluted or partially polluted air at the
ground. (Modified from Clark and Whitby.141)
-------
52 PARTICULATE POLYCYCLIC ORGANIC MATTER
face area of aerosols in Pittsburgh varied seasonally from 1.90 m2/g
in the spring to 3.05 m2 /g in the winter.
POM and Particle Size Distribution
There is practically no information on the particle size or mass dis-
tribution of POM in an aerosol. In one study, DeMaio and Corn193
found that more than 75% of the weight of selected poly cyclic hydro-
carbons was associated with aerosol particles less than 2.5 jum in
radius. However, Thomas et a/.751 determined that the amount of
benzo[a] pyrene per unit weight of soot was constant in the sources
they tested.
A potentially important factor in limiting the size of the smallest
POM-containing particles is the so-called Kelvin effect—the equilibrium
vapor pressure varies with the radius of curvature of the particle and
with the surface energy of the volatile material. Unfortunately, there
is no information available on the surface energy at the POM-air or
POM-solid interfaces. Therefore, it is not possible to calculate the
Kelvin effect.
A problem in determining the size fractionation in POM-containing
aerosols may arise from current sampling methods. It has been found,
for example, that polycyclic hydrocarbons have high volatilities in the
presence of air flow.639 It is therefore possible that some of the POM
may be lost by vaporization from smaller particles, owing to the
Kelvin effect during the sampling process.
Adsorption and Elution of POM
Mechanisms for the adsorption of POM on particle surfaces have
not been studied extensively. However, one investigator814 has specu-
lated that the benzo[a] pyrene and other arenes appear to be adsorbed
primarily on the surface of soot by hydrogen bonding.
Perhaps as important as the ability of POM to be adsorbed on aero-
sols is its ability to be desorbed or eluted from particles in biologic
fluids. Studies of Falk et a/.263 have indicated that POM cannot be
readily eluted by plasma from soots less than 0.04 /urn in diameter.
Larger soot will release adsorbed polycyclic aromatic hydrocarbons
in the presence of plasma and cytoplasmic proteins. As particle size
increases, elution becomes more rapid and more extensive.
More recent studies of Kutscher et a/.469 have dealt with the abil-
ity of bovine serum to separate benzopyrene from three kinds of soots:
-------
Atmospheric Physics 53
fine-grained Corax L (mean size, 0.028 jim), inactive MT soot (0.4 /wn),
and Hame soot 101 (0.115 jum). The results indicate that a maximum
of 10% benzopyrene was eluted from Corax L, whereas soots MT
and 101 yielded 20% and 13%, respectively. The first traces from
Corax L were detected after 7 hr; it took only 15 min for soots MT
and 101 to appear in the serum. Incubations of single components of
albumin, alpha, beta, or gamma globulin, and Corax L that had ad-
sorbed benzopyrene showed that only albumin would elute the ad-
sorbed material.
An investigation by Pylev613 has demonstrated the significance
of particle surface area in lung retentivity. This investigator adsorbed
different amounts of benzo[a] pyrene on soots ranging in size from
128 to 3,047 A, with specific surface areas of 10-250 m2/g. He ex-
posed rats to the soots and evaluated retentivity after dissection. The
amount of benzo[a] pyrene left in the rats' lungs differed markedly,
depending on the particle surface area. The higher the specific sur-
face area of particles, the more difficult the POM was to elute. Py-
lev concluded that it is necessary to know, in addition to particle size,
more about the composition of the material on which POM may be
adsorbed.
DYNAMICS OF AEROSOLS NEAR THE GROUND
Once in the atmosphere, suspended particles will evolve by the mecha-
nisms listed in Table 4-5. Each of the processes listed can influence
TABLE 4-5 Processes Affecting Evolution of Aerosols in a Unit Volume of
Lower Atmosphere"
1. Growth or change in particles by homogeneous or heterogeneous chemical reactions of
gases on the surface of particles
2. Change in particles by attachment and adsorption of trace gases and vapors to aerosol
particles
3. Net change by collision between particles undergoing Brownian motion or differential
gravitational settling
4. Net change by collision between particles in the presence of turbulence in the suspending
gas
5. Gain or loss in concentration by diffusion or convection from neighboring air volumes
6. Loss by gravitational settling
7. Removal at the earth's surface on obstacles by impaction, interception, Brownian motion,
and turbulent diffusion
8. Loss or modification by rainout in clouds
9. Loss by washout under clouds
" After Hidy.37'
-------
54 PARTICULATE POLYCYCLIC ORGANIC MATTER
the observed size distribution of aerosols, as well as their chemical
composition as a function of size.
CoDision Processes
The coagulation of aerosol particles tends to create large particles
while continually depleting the smaller particles. Coagulation requires
that two or more particles collide and stick together. Relatively little
is known about adhesion during collision, so coagulation of aerosols
is generally assumed to occur after collision, with a probability of near
unity. There is no sound basis for such an assumption, but the hypoth-
esis is indirectly supported by the concept of Brownian motion in
many (not all) experiments.
Collisions between particles may result from their Brownian mo-
tion if they are small enough. Otherwise, differences in their velocities
in air are required. Other than collision by thermal agitation, some
mechanisms include turbulent air motion, differential settling, pho-
retic forces, and electric forces. Collisions by turbulence involve the
fact that particles, because of their large inertia, cannot follow exactly
the local eddying motion and may experience a local shear as a result
of interaction between eddies of different size. During the fallout of
large particles they may sweep out smaller particles in their path; this
constitutes the common "scavenging mechanism" in the atmosphere.
Relative motion between particles resulting from thermal or concen-
tration gradients during evaporation or condensation of water vapor
is possible, but is believed to be a second-order effect. Electric forces
between particles also may play an important role in coagulation,
particularly for particles less than 10 Aim in radius. Although con-
siderable effort has been devoted to development of an understanding
of electrification, no extensive calculations have quantitatively eval-
uated the significance of electric forces, compared with other collision
processes.
Because the atmosphere is frequently in a turbulent state, particles
tend to be transported rapidly in and out of fixed regions of air space.
Thus, particles are transferred in and out of an aerosol cloud by the
winds and by diffusion processes associated with turbulence and
Brownian motion. In the atmosphere, diffusion by turbulence far ex-
ceeds that by thermal agitation. In addition to these diffusion pro-
cesses, particles will disappear from a volume of air by sedimentation.
Particles are much more massive than the suspending gas, so there is
-------
Atmospheric Physics 55
always a tendency for aerosol clouds to lag behind air motion and to
settle out to the earth's surface in the absence of upward air motion.
Removal from the Atmosphere
Aerosol may be removed from the atmosphere by several mechanisms.
Deposition of large particles by gravitational settling is important. Over
the radius range of 0.5-20 jura, inertial forces acting on aerosols cause
departures in particle trajectories from the air flow around obstacles,
thereby allowing deposition by impaction. Obstacles may include rocks,
buildings, fences, and vegetation. Because of the finite size of aerosols,
they may be intercepted by obstacles even if their trajectories do not
depart significantly from the motion of the suspending air. If parti-
cles are less than 0.2 jum in radius, their thermal agitation may be suf-
ficient to allow them to diffuse around an obstacle if they pass very
near it, so turbulent diffusion around obstacles is important in par-
ticle removal.
It is possible to estimate roughly the deposition rate of particles
on an obstacle if the air-flow field near it is known. Methods have
been discussed in recent monographs.2821373 Little information is
available, however, from which one could deduce the removal rate
by buildings or other solid structures in actual atmospheric condi-
tions. Some limited data have been reported for deposition rates on
vegetation. Chamberlain,128 for example, has measured the deposition
of radioactive particles on flat grass surfaces, and Neuberger et al.561
and Rosinski and Nagamoto 642 have investigated deposition on
trees. The study of Neuberger et al. suggests that conifers are better
natural filters than deciduous trees.
Role of Rain Clouds
The development of rain clouds, superimposed on the removal mech-
anisms just discussed in an otherwise dry atmosphere, sometimes in-
fluences aerosols containing POM significantly. In contrast with other
removal mechanisms, rain clouds are intermittent and their geographic
frequency varies widely. Nevertheless, their effects have to be accounted
for in any evaluation of the fate of POM in the atmosphere.
Aerosols provide centers for nucleation of water droplets in the
atmosphere after the air becomes supersaturated with water vapor.
Aerosols inside clouds are captured in droplets by a variety of mecha-
-------
56 PARTICULATE POLYCYCLIC ORGANIC MATTER
nisms, ranging from scavenging to electric and phoretic interactions.
As a result of these processes, called "rainout," the size distribution
and the chemical composition as a function of size may be modified.
Because of the great difficulties in interpreting and observing natural
aerosols inside and outside clouds, practically nothing is known about
the nature of such in-cloud modifications.
When precipitation begins to fall from clouds, the hydrometeors
will sweep out smaller particles during their fall toward the ground, as
in the case of the dry scavenging mechanism. This washout is believed
to be significant in removing many pollutants, including POM, from
the atmosphere.
There remains some controversy about the importance of washout,
compared with other processes, based on current theoretical and ex-
perimental work. Theory indicates that the collection efficiency of
spheres falling through other spheres is a direct function of particle
size. In particular, for Stokes's flow, only very few particles smaller than
about 20 jum in radius are collected on falling spheres of much greater
size. Some laboratory data support this conclusion, but observation in
the atmosphere suggests a stronger action of precipitation in removal
of aerosols. Of course, the theory for interacting spheres can be only
qualitatively extrapolated to nonspherical material. For example, the
collection of smoke particles on snowflakes is undoubtedly different
from rain falling through liquid spheres. There is little quantitative
information on the effectiveness of washout with interacting non-
spherical particles.
Relative Significance of Aging and Removal Processes
Until recently, efforts have been devoted mainly to identifying and
characterizing mechanisms involved in the evolution of aerosols in
the atmosphere. There remains considerable uncertainty regarding
some of these mechanisms, particularly those related to rain clouds
and those requiring chemical transformations. Nevertheless, it is pos-
sible to evaluate the relative significance of many mechanisms.
A recent study,371 using semiquantitative theoretical arguments,
has indicated the importance of some factors in aerosol behavior near
the ground. The data in Table 4-6, from a typical urban atmosphere,
suggest that coagulation by Brownian motion, turbulence, and diffu-
sion are most important factors in shaping the size spectrum for par-
ticles less than 0.1 jxm in radius. For the large-particle fraction, the
collision mechanisms are weak, and turbulent diffusion with sedimen-
-------
Atmospheric Physics 57
TABLE 4-6 Processes Contributing to the Aging of Urban Aerosols at the
Ground Level (particles lost/cm3 per sec)a
Older of Magnitude of Contribution
Process
Convective diffusion6
Thermal coagulation6
Scavenging by particles (1 0 /urn Np = 10"' /cm3 ) c
Turbulent coagulation (e = 1,000 cm2/sec3)<*
Sedimentation removal6
Washout (100 urn Np = 10-3/cm3)
0.05 Mm
in radius
10- '-i
1
io-3
io-3
io-6
io-3
0.5pm
in radius
io-3
io-3
io-3
io-6
io-6
5 Mm
in radius
io-6
10~6
io-4
10-'
-
"Derived from Hidy.371
60.05 nmNp = 10s/cm3; 0.5 nmNp = 102/cm3; 5 Mm^p = 10~l cm3.
cNp = concentration of scavenging particle.
d e = dissipation rate of turbulent kinetic energy.
tation should dominate aerosol behavior. In the middle range of size,
collisions by Brownian motion and other mechanisms are weak, but
turbulent diffusion remains important. Although weak, turbulent
coagulation may be important in allowing particles to be transmitted
up the size spectrum from the range dominated by Brownian motion
to a range effectively terminated by removal through sedimentation.
The effect of washout by precipitation is evaluated tentatively as
less important than dry mechanisms near the ground.371 However,
above an altitude of about 300 m through typical cloud-top levels,
washout is likely to become more and more important in removing
atmospheric aerosols. These conclusions remain speculative, but fur-
ther studies should improve the accuracy of evaluation.
ATMOSPHERIC DISPERSION OF POM
Microscale Diffusion
It is important in evaluating the hazards of POM to have some knowl-
edge of the extent of dispersion of this material far from its sources.
Classic calculations802 using the Gaussian plume diffusion model with-
out chemical reaction of the contaminant have been applied extensively
in evaluating dispersion a few miles downstream from single sources.
These computations can be made for steady wind patterns over smooth
terrain for various classes of atmospheric density stratification, pro-
-------
58 PARTICULATE POLYCYCLIC ORGANIC MATTER
vided empirical diffusion coefficients can be estimated. This model
always predicts a considerable reduction in ground-level concentration
of a pollutant coming from a ground-level or elevated source.
One of the great difficulties with the single-plume models has been
their inability to account for topography quantitatively. Some recent
work of Hinds,377 however, has suggested that the diffusion coeffi-
cient in such calculations can be modified empirically (on the basis
of observations) to evaluate better the influences of rough topography.
Recently, attempts have been made to extend micrometeorology-
scale diffusion calculations to include chemically reactive pollutants.
An illustrative model was posed by Friedlander and Seinfeld,281 who
investigated the behavior of bimolecular reactions of pollutants under-
going diffusion modeled by the Lagrangian similarity hypothesis. In
this calculation, secondary pollutant concentrations at ground level
vary according to both the reaction rate and the diffusion rate. In
particular, it is sometimes possible to quench some reactions by
dilution.
An alternative approach to the Friedlander and Seinfeld analysis
has been proposed for photochemical smog by Eschenroeder and Mar-
tinez,251 who follow the chemical reactions along streamlines of the
wind field.
Neither of these models is realistic in the sense of having great
practical utility for estimating reaction rates and dispersion of
materials like POM, but they do represent a first step in recognizing
the importance of linking chemistry and air motion in evaluating air
pollution.
Larger-Scale Diffusion
On meteorology scales encompassing major urban areas, the Gaussian
plume models cannot supply an adequate quantitative assessment of
atmospheric dispersion. Account must be taken of the local wind
field and the surface distribution of sources. An example of a first
attempt to simulate the dispersion over a large urban area, the Los
Angeles Basin, has been reported recently by Lamb and Neiburger.470
Their experimental results suggest that inert contaminant levels can
be verified qualitatively. However, the data suffer from simplifications
of the local meteorology that will undoubtedly have to be considered
in more realistic studies. In particular, some effects, like surface
heating in cities, tend to change the mixing and wind field in air over
urban areas. The disturbance in the local meteorology caused by a
city has been studied in a preliminary way,78'270'598 but further
-------
Atmospheric Physics 59
work must be done to appreciate this factor fully in the dispersion of
pollutants.
On scales of tens of kilometers, attempts have been made to evaluate
transport and pollutant accumulations over air sheds. One example
of Reiquam630 uses a calculation that breaks up the air-shed volume
into well-mixed boxes with local sources. The analysis, again, is highly
simplified but leads to qualitative verification with sample air sheds.
There is concern currently for the mechanisms of dispersion over
synoptic meteorology scales (1,000 km horizontally). Some efforts
to evaluate such problems are beginning to appear in the literature.
Kao and Henderson,425 for example, have applied turbulence theory
to investigate classes of synoptic-scale diffusion in the atmosphere.
Holzworth's study392 offers some estimates of large-scale weather in-
fluences on community air pollution in the United States.
Qualitatively, one would expect an increasing penetration of pol-
lution into remote areas as time goes on. Indeed, this is suggested in
some respects for aerosol, in the trends of increasing mass concen-
trations measured by the National Air Surveillance Network over the
last 10 years.503 In contrast with the total particle loading increase in
remote areas, NASN data for the last 3 years suggest a decrease in benzo-
[a] pyrene in many of the remote sampling areas. This may be due to
many factors, with meteorology being only one. Perhaps equally
important are improvements in combustion processes and the use of
different fuels.
In any case, it is clear that significantly increased concentrations
of POM, as measured by benzo[a] pyrene, remain largely localized in
urban areas. Even in the cities, however, POM will be considerably
diluted in the ambient atmosphere from its source concentrations in
most circumstances.
LIFETIME OF POM IN THE ATMOSPHERE
On the basis of very limited evidence, aerosols are expected to remain
in the lower atmosphere for 5-30 days (see, for example, Junge420).
More recent evidence from Esmen and Corn252 suggests that, for ur-
ban aerosols in Pittsburgh air, the residence time without precipita-
tion is some 4-40 days for particles less than 1 p.m in diameter and
0.4-4 days for particles 1-10 /mi in diameter. With removal associated
with rainfall, these times are believed to be somewhat shorter.
Chemical reactivity of benzo[a] pyrene in the atmosphere, associated
-------
60 PARTICULATE POLYCYCLIC ORGANIC MATTER
with degradation taking place on soot in sunlight, yields a chemical
half-life of less than a day, as indicated by data of Thomas et al.7S1
However, half-lives of several days have been reported, apparently
without solar radiation, in earlier work of Falk et al.2S8 These re-
sults for the chemically determined lifetime of benzo[a] pyrene in the
atmosphere are ambiguous. However, one may guess that exposure to
sunlight of POM on the external surfaces of soot or dust could lead
to rapid degradation, whereas POM adsorbed in pores of particles may
persist for considerably longer periods.
Exposures to relatively large doses of POM occur in urban areas,
where combustion sources are most highly concentrated. Air masses
probably remain over a city for less than a day in most circumstances.
A residence time of 1 day for air in the Los Angeles Basin can be rouglv-
ly estimated from emissions of carbon monoxide and carbon monoxide
concentrations given by Lemke et al.,489 assuming an inversion height
of 500 m and assuming that carbon monoxide is conserved in such an
air mass. Thus, it appears that the time for chemical transition of POM
will be roughly the same as or longer than the meteorologic residence
time in urban centers. One can then expect some contamination of
suburban and sparsely populated areas as a result of dispersion and
transport of POM-contajning particles by the winds before chemical
transition can take place fully. This conclusion appears to be consistent
with at least some of the N ASN data in Table 4-3.
IMPORTANT AREAS OF UNCERTAINTY
From the standpoint of evaluating POM as a health hazard, the prin-
cipal missing information lies in the chemical nature of POM-contain-
ing aerosols. There is a need to know how this material is distributed,
with respect to particle size. In addition, more information is required
on the physical and chemical evolution of POM as it interacts with
other matter in the atmosphere. Of particular interest is the nature
of POM adsorbed on different classes of atmospheric aerosols. As
can be seen from this brief survey, very little is known about the
actual form of POM in the atmosphere. The connection with any
measure of adverse health effects is extremely difficult to evaluate
in the absence of specific information on POM after it has aged in
the air. If there is indeed a link between aerosols of soot or similar
materials and hematite in carcinogenesis, it is essential to identify
such synergistic materials for control purposes.
-------
Atmospheric Physics 61
SUMMARY AND RECOMMENDATIONS
Form of POM in Air
Polycyclic organic matter detected in the atmosphere is associated
exclusively with particulate matter, especially soot. It is uncertain
whether POM condenses out as discrete particles after cooling or con-
denses on surfaces of existing particles after formation during com-
bustion. In any case, knowledge of the behavior of aerosols is essen-
tial to understanding the fate of POM in the atmosphere.
Sources and Properties of Atmospheric Aerosols
As a first approximation, natural sources of aerosols contribute ap-
proximately 10 times as much as anthropogenic sources to the global
burden of suspended particulate matter. However, localized emission
inventories of these pollutants, particularly in urban areas, indicate
that anthropogenic sources predominate. It is necessary to refine the
estimates for specific categories of sources, particularly for the global
inventory of aerosols.
Particle size is the physical property with the greatest influence
on the behavior of POM-containing aerosols. Generally, the particle
size spectrum of atmospheric aerosols extends from less than 0.01
' Mm to greater than 10 Mm. There is considerable variation in the
size-concentration distribution with location in space and time, but
there is some regularity, on the average, in the particle equivalent
diameter range 0.1 /zm
-------
62 PARTICULATE POLYCYCLIC ORGANIC MATTER
100 to 200 Mg/m3» as measured by high-volume sampling. The ben-
zene-soluble portion of this material is approximately 10% by weight,
but can vary from 8 to 14% in urban areas. The POM component is
much smaller than the benzene-soluble fraction. The density and
specific surface of the urban aerosol, on the basis of very limited
data from one city, are 1.8-2.1 g/cm3 and 1.90-3.05 m2/g. Specific
surface appears to vary with season; density does not.
Additional data relative to the physical properties of the atmo-
spheric aerosol are needed. Simple and inexpensive instrumentation is
required to obtain, in particular, size-weight concentration data dur-
ing short intervals (minutes).
Lifetime of POM in Air
Because POM is carried by suspended particulate matter, its longevity
in air depends on the lifetime of the carrier aerosol in air and on
chemical alteration of POM itself. Initial estimates of atmospheric
residence times of particles less than 5 pm in diameter exceed 100 hr
in dry atmospheric conditions. Chemical reactivity in the presence of
sunlight may lead to transition of POM adsorbed on soot to other
material in several hours. Without sunlight, its lifetime may be much
longer. Meteorologic factors suggest that air will remain over a city
for less than a day. Therefore, it appears that the time for clinical
transition will be about the same as or greater than meteorologically
controlled residence times in a particular urban atmosphere.
-------
Chemical Reactivity of
Polycyclic Aromatic
Hydrocarbons and Aza-Arenes
Chemical reactions of polycyclic organic matter in the atmosphere
are important because such reactions appear to represent a major
mode of removal of polynuclear compounds and because the products
of the reactions may in some instances be health hazards themselves.
The low vapor pressure of most polycyclic compounds has restricted
laboratory studies on them to solutions, with few exceptions. In the
atmosphere, however, most polynuclear compounds are adsorbed on
particulate matter. Reactions of these compounds in the adsorbed
state appear to occur particularly readily, and the general nature of the
reactions is predictable on the basis of the few studies that have been
conducted. There is less information on reactions in the vapor phase
or in aerosol solution, but there is no reason to expect them to differ
from those in solution.
The following review is not comprehensive; rather, it is limited to
types of reactions likely to be of atmospheric importance. Each type
is illustrated by only one or two examples, which were chosen be-
cause they illustrate important features of the reactions or because
they are significant from a health standpoint. Tipson has reviewed the
oxidations of polycyclic aromatic hydrocarbons,753 and the general
reactions of polycyclic hydrocarbons have been surveyed by Gar.139
63
-------
64
PARTICULATE POLYCYCLIC ORGANIC MATTER
GENERAL REACTIVITY CONSIDERATIONS
A great deal of theoretical and experimental work on the reactivity
of polycyclic aromatic hydrocarbons has been carried out because
of suggested correlations between reactivity and carcinogenicity.94'
i8i,i96,6ii Qnjy a few reievant conclusions will be cited here. Reac-
tions may be classified as substitution (in which a hydrogen atom is
replaced by another atom) or addition (in which unsaturation is
destroyed). Multiple substitution may occur, and addition is often
followed by elimination (giving net substitution). Many of the primary
products of these reactions undergo further reactions, resulting in more
complex changes, such as quinone formation or bond cleavage. The
various types of reaction are summarized below.
X,Y
x + HY
1,4-Addition
X.Y
1,2-Addition
X,Y
X,Y
-HY-
The type of reaction that a given compound will undergo depends
both on the reagent and on the compound.94'139'181'196'611'753 Reac-
tions of linear hydrocarbons tend to take place at the anthracene
9,10-like position (1,4-addition or substitution) or the 1,2 position
(1,2-addition). Reactions of angular hydrocarbons tend to take place
at the anthracene 9,10-like position (1,4-addition or substitution) or
-------
Chemical Reactivity
65
the phenanthrene 9,10-like double bonds (1,2-addition). Reactions of
more condensed ring systems tend to take place at positions adjacent
to ring fusions. These tendencies are illustrated below.
Anthracene
1,4-addition
^9 1 xl,2-addition
~ ~r,
4
1,4-addition
substitution
Benz [a]anthracene
1,4-addition ,
+1-
1,2-addition
Pyrene
' ., substitution
1,2-addition
Benzo[a]pyrene
(secondary substitution)
3 -»-(secondary
substitution)
1,2-addition
1,4-addition
substitution
1° 5
main substitution
Dibenz [ a,h] anthracene
1,4-addition
*,.
substitution
1,2-addition
1,2-addition
1,4-addition
substitution
An alkyl or alkoxyl substituent strongly increases the reactivity in
most of the reactions described for these compounds, although it will
often prevent substitution at the position at which it is attached.
Electron-withdrawing substituents generally make reactions more dif-
ficult.
PHOTOOXIDATION
Gollnick and Schenck have reviewed photooxidation of polycyclic
aromatic hydrocarbons and dienes.304 Tricyclic or larger hydrocarbons
have strong absorption in ultraviolet radiation at wavelengths longer
than 300 nm (present in solar radiation at ground level), and most
-------
66
PARTICULATE POLYCYCLIC ORGANIC MATTER
are very readily photooxidized. Photooxidation is probably one of
the most important processes in the removal of polycyclic hydrocar-
bons from the atmosphere, and it is the only one on which substantial
data have been obtained under simulated atmospheric conditions,
using adsorbed hydrocarbons.
Photooxidation in Solution
The most common photooxidation reaction of polycyclic aromatic
and aza-aromatic hydrocarbons in solution is formation of endoper-
oxides. For example, 9,10-dimethylanthracene (1) yields the 9,10-
endoperoxide (2),716 7,12-dimethylbenz[a] anthracene (3) yields the
7,12-endoperoxide (4),158 and dibenz[b,h] acridine (5) yields the
1,4-endoperoxide (6).253
hu, O2
CH
hu,O2
CH
Photolysis or pyrolysis of peroxides of this sort produces a variety
of results, including dealkylation and ring cleavage.716 This process
proceeds by a free-radical mechanism (cleavage of the O-O bond) and
-------
Chemical Reactivity
67
initiates autoxidation.139 This is shown by the possible products of
pyrolysis of 9,10-endoperoxide (2) below.
CH
Alternatively, some endoperoxides on pyrolysis regenerate oxygen
in the singlet state, essentially in a reversal of the formation reaction.805
Some compounds undergo this reversal reaction rapidly at room tem-
perature.304 When the bridgehead carbon atoms bear hydrogen atoms,
the initial endoperoxides are readily converted to quinones. In many
cases, only the quinones can be isolated. For instance, compound (6)
rearranges readily to the quinone (7), and only the quinone (9) was
isolated from photooxidation of benz[b] acridine (8).253
-------
68
PARTICULATE POLYCYCLIC ORGANIC MATTER
It should be pointed out that diones can also be produced when,
for steric reasons, no endoperoxide can be formed. For example,
benzo[a]pyrene (10) yields a mixture of the 6,12-dione (11), the
1,6-dione (12), and the 3,6-dione (13).522 The mechanism of this oxi-
dation may well differ from that of the usual one and may involve
electron transfer. Endoperoxides are rarely formed from hydrocarbons
like benzo[a] pyrene; normally, two anthracene 9,10-like positions, not
at ring junctures, are required.
(10)
(12)
Photooxidations involve energy transfer from the triplet state of
the aromatic compound (A), yielding singlet oxygen (1O2), which
reacts with compound A, yielding peroxide (AO2 )276 [see Eqs. (1)
and (2)]. Singlet oxygen generated by various chemical sources also
reacts with anthracenes to yield endoperoxides.276 Singlet oxygen,
directly detected in gas-phase systems, has been formed by energy
transfer from benzene,271'721 naphthalenes,431'803 and benzalde-
hyde.159'462 Although no photooxidizable aromatic has been reacted
in the vapor phase, singlet oxygen reacts with cyclohexadiene552
and with dimethylfuran299 to yield the normal endoperoxide. adducts
in the vapor; therefore, aromatic compounds would be expected to
undergo reactions in the gas phase identical with those undergone in
solution. Further research is needed on this point.
(1)
(2)
AO,
-------
Chemical Reactivity
69
If the 9,10 positions of an anthracene (14) are no more than mono-
substituted (14), a photodimer (A2) may be formed by reaction of
the excited singlet of A with ground state A (15). The extent of
dimerization depends on the solvent; dimerization normally does not
take place in conditions in which anthracene does not fluoresce.304
Monosubstituted anthracene (14)
Photooxidation of Adsorbed Aromatic Hydrocarbons
Only a few reports of photooxidation of adsorbed aromatic hydro-
carbons have appeared, but they are sufficient to demonstrate that
the reactivity of adsorbed hydrocarbons is considerably greater than
that of hydrocarbons in solution. Anthracene (16) adsorbed on silica
gel or alumina is very rapidly oxidized to anthraquinone (17); the
endoperoxide does not seem to be intermediate in this oxidation.446
Further oxidation gives l,4-dihydroxy-9,10-anthraquinone (18).
Even naphthalene (normally inert) can be oxidized under these condi-
tions. The degree of oxidation of anthracene depends somewhat on
the adsorbent.797
OH
On thin-layer chromatography (TLC) plates (either alumina or
silica geLas adsorbents), in room or ultraviolet light, anthracene,
naphthacene, benz[a]anthracene, dibenz[a,c] anthracene, dibenz[a,h]-
anthracene,"pyfene, benzo[a]pyrene, benzo[e]pyrene, benzofghi]-
perylene, and coronen^are oxidized; phenanthrene, chrysene, and
triphenylene are inert. Pyrene (19) is oxidized to the 1,6-dione (20)
and the 1,8-dione (21) under these conditions.413 These studies pro-
vide potential models for oxidation of hydrocarbons adsorbed on
-------
70 PARTICULATE POLYCYCLIC ORGANIC MATTER
participate matter in the atmosphere and indicate the difficulties
faced by the analytical chemist.
(19) (20) (21)
The results gain significance from the reports of formation of radi-
cal cations from hydrocarbons adsorbed on various substances; many
of these oxidation reactions may occur via the radical cations.
There have been several reports of destruction of polycyclic aro-
matic hydrocarbons adsorbed on soot or smoke.260!262'745'751 Most
of the destruction requires light, although it is accelerated by syn-
thetic photochemical smog.260'262 It is difficult to calculate specific
rates from the data presented, but exposure for roughly 40 min to
light of one fourth the intensity of noon sunlight caused 35-65%
loss of benzo[a] pyrene in airborne smoke samples in an irradiated
flow reactor.745>7S1 Exposure for 6 hr to sunlight caused 15-50%
loss of benzo[a] pyrene in smoke on filters.745 In a different study,
10% of benzo[a] pyrene adsorbed on soot or a filter was destroyed
in 48 hr of exposure to light of unstated intensity, but 50% was
destroyed in 1 hr of exposure to light and synthetic smog (contain-
ing the unnaturally high oxidant concentration of 30 ppm).262 These
studies imply that the chemical half-life of benzofa] pyrene and of
other polycyclic aromatics260'262 may be limited to only hours or
days in the atmosphere. Further experimentation in simulated atmo-
spheric conditions with careful attention to wavelength distribution
of the light used and chemical characterization of the products is
highly desirable.
ACTION OF OZONE
Polycyclic Aromatic Hydrocarbons
Ozone reacts readily with polycyclic aromatic hydrocarbons.24'541
Several modes of reaction have been identified. One important reac-
tion is cleavage of phenanthrene-like double bonds, which in oxida-
tive conditions eventually results in the formation of diacids; another
-------
Chemical Reactivity
71
involves oxidation at anthracene 9,10-like positions to yield quinones;
a third is a more complex nuclear oxidation; and a fourth involves
side-chain oxidation. Of the large body of literature, only a few
particularly pertinent examples will be given here. Laboratory studies
are normally carried out at low temperature in an inert solvent and are
followed by either reductive or oxidative workup, but atmospheric
reactions probably take a very similar course and would be expected
to yield products similar to those of oxidative workup. Most of the
primary products are also subject to further oxidation. Material
balances are often rather poor.
Ozonolysis of benz[a] anthracene (22) provides a typical exam-
ple; the products are compounds (23)-(25) and oxygen.25
(22)
Product (23) is produced by cleavage of the reactive phenanthrene-
like double bond. The fate of the initial unstable molozonide (26)
depends somewhat on the conditions; under atmospheric conditions,
it would almost certainly be converted to compound (23). Aldehydes
and various polymeric peroxides may be intermediate.
Benz(a J anthracene
(22)
-------
72
PARTICULATE POLYCYCLIC ORGANIC MATTER
Quinone (24) and phthalic acid (25) are believed to be, in part,
products of decomposition of an initial trioxide (27) formed by ad-
dition of oxygen at the meso positions of benz[a] anthracene (22). A
trioxide of some stability has indeed been isolated from the ozonolysis
of 9,10-dimethylanthracene.250
Benz(a]anthracene
(22)
Quinone and
Phthalic acid
However, a second route to quinone is also postulated, involving
electrophilic attack of ozone on aromatic hydrocarbons, giving hy-
droxylated compounds and oxygen; the hydroxylated compounds
undergo rapid further oxidation, leading ultimately to quinone (24).
It is likely that the oxygen formed is in the singlet excited state.
Benzfa] anthracene
(22)
orO3
-»• Quinone
(24)
This mode of attack may be the most important one in the oxida-
tion of benzo[a]pyrene (10) to yield a mixture of the 3,6-dione (13)
and the 1,6-dione (12) in a 3 : 1 ratio, accompanied by a trace of the
4,5-dione (28);S4° presumably, the 4,5-dione comes from the unsta-
ble molozonide, analogous to the formation of compound (26). A
series of one-electron oxidations should not be ruled out as the mech-
anism of the ketone-forming processes.
-------
Chemical Reactivity
73
(12)
(trace)
The ozonolysis (O3) of 3-methylcholanthrene (29) illustrates the
side-chain oxidation that may occur with alkylated aromatics; in ad-
dition to minor products of more extensive oxidation [O], the major
products are compounds (30) and (31).544 Similar dealkylation occurs
with 7,12-dimethylbenz[a] anthracene.543
o co2H
3-Methylcholanthrene
(29)
COjH
(30)
(31)
Another type of oxidation involves attack at saturated C—H
bonds; for instance, fluorene is oxidized to fluorenone; anthrone,
to anthraquinone.125
Fluorene
[O]
[O]
Anthrone
Anthraquinone
-------
74
PARTICULATE POLYCYCLIC ORGANIC MATTER
Autoxidation of alkanes is initiated by ozone at ambient tempera-
tures.679"681 This reaction probably involves abstraction of hydrogen,
yielding an initial hydrotrioxide, which can break down to yield
radicals that can serve as initiators;816 this may be the mechanism of
side-chain oxidation, as given below.
RH
03
• Chains + Oj
Some of the oxygen produced from the reaction of isopropyl alco-
hol or diisopropyl ether with ozone is in the singlet state.553 This
oxidative reaction also occurs in the gas phase (3. N. Pitts, Jr., personal
communication). At least one other reaction of ozone, that with
tertiary phosphites, produces oxygen in its singlet state.804
Polycyclic Aza-Arenes
Aza-arenes and aromatic hydrocarbons react similarly, but aza-arenes
are generally less reactive. Acridine yields compounds (32) and (33);
phenanthridine yields compounds (34) and (35).542 Trialkylamines,
which have strongly basic nitrogen, are oxidized to N-oxides and
undergo various oxidations of alkyl groups attached to the nitrogen.26
CO,H
CO,H
CO2H
Phenanthridine
ACTION OF MISCELLANEOUS OXIDANTS
One-Electron Oxidation
One-electron oxidation of many polycyclic aromatic hydrocarbons
occurs readily;11'619 the primary products in most cases are radical
cations.5'6'811 These radical cations are unstable and react rapidly
with water or other nucleophiles (including unoxidized hydrocar-
-------
Chemical Reactivity
75
bon) or with oxygen. Acridities react similarly.6'811 The overall
reaction leads to formation of quinones (with water and oxygen),
nucleophile adducts, or dimeric hydrocarbons. s>6'811 Benzo[a]-
pyrene (illustrated below) is oxidized anodically to the radical cation,
which reacts with water to yield easily oxidized intermediates that
are further oxidized to the 1,6-dione and two other quinones; some
dimer is also formed, mostly at the electrode surface. The similarity
of the overall process to photooxidation and to part of the ozonolysis
process should be noted.
Benzo[a]pyrene
Radical cation
O
1,6-Dione and
other quinones
Dimer
Side-chain oxidation of alkyl-substituted aromatics can occur.5-6'811
As illustrated below, durene is anodically oxidized (—e~) in acetoni-
trile (CH3CN) to side-chain substitution product (36), with acetoni-
trile acting as a nucleophile.226 If water is present, the corresponding
benzyl alcohol (37) is formed.
CH3 O
• ii
CH2-NH-C-CH3
CH3
Durene
-------
76
PARTICULATE POLYCYCLIC ORGANIC MATTER
Although these reactions involve complex sequences of oxidations,
nucleophilic attack, and proton loss, the primary product in most
cases is the radical cation.5 Similar reactions occur with a wide variety
of one-electron oxidants. For example, 7,12-dimethylbenz[a] anthra-
cene [compound (3)] with any one of a variety of one-electron metal
salt oxidants yields mixtures in various proportions of the 7,12-endo-
peroxide [compound (4)] and compounds (38)-(43).279
CH3
CH
OH
Radical cations are formed from many polycyclic aromatic hydro-
carbons simply by treatment with strong Lewis acids, particularly in
the presence of oxygen.2 Irradiation greatly increases the degree of
electron transfer, so light can directly cause the formation of radi-
cal cations.' Adsorption on an alumina or silica surface can produce
radical cations directly, particularly in the presence of oxygen or
iodine.335-641'817'818 The adsorbed radical cations react to yield di-
mers or react with nucleophiles, as described above; for example,
benzo[a] pyrene (10) yields quinones and dimer. In the presence of
nuclear bases or pyridine, substitution products are formed, such as
compound (44).817'818 This reaction with nuclear bases may be
-------
Chemical Reactivity 77
important in the binding of carcinogenic hydrocarbons to biologic
materials.817'818
+
Pyridine
Quinones and
dimer
Many reactions of polycyclic aromatic hydrocarbons probably
proceed through radical cations. As a note of possible biologic impor-
tance, reaction of radical cations with electron donors regenerates
the hydrocarbon, partly in an excited state;176-361 this process repre-
sents a mechanism for producing excited hydrocarbon and (by reaction
of the excited hydrocarbon with oxygen) singlet oxygen in the absence
of light [see Eq. (3)]. This conceivably accounts for the formation of
endoperoxide (4) in the one-electron oxidation of 7,12-dimethylbenz-
[a] anthracene (3). Furthermore, singlet oxygen is known to oxidize
nucleic acid derivatives; such reactions with bound carcinogenic hydro-
carbons would produce singlet oxygen near genetic material and thus
might cause genetic damage.275 This possibility requires further research.
>'02
(3)
Peroxides, Radicals, and Other Oxidants
Polycyclic aromatic hydrocarbons react readily with peroxides; the
products are those of substitution or of further oxidation to yield
products similar to those described in previous sections. Benzoyl-
peroxide [(C6HSCO2)2 ] reacts with benzo[a]pyrene, which is ex-
tremely reactive, to yield the 6-benzoyloxy derivative (45 ).637 In the
presence of oxygen, the reaction might well lead to further oxidation.
-------
78
PARTICULATE POLYCYCLIC ORGANIC MATTER
(10)
Benzo[a)pyrene
Perbenzoic acid (C6HSCO3H) oxidation of dibenz[a,h] anthracene
(46) produces the 7-ketone (47) and the 5,6-epoxide (48), which
reacts further. Oxidation with the more reactive peracetic acid
(CH3CO3H) yields the 7,14-quinone (49), a diacid (50), and the
5,6-quinone(51).78S
(48)
(50)
(51)
The products of attack at the 5,6 double bond are a consequence
of the tendency of peracids to induce the formation of epoxides.
The attacks at the 7 and 14 positions of dibenz[a,h] anthracene with
peracids and at the 6 position of benzofa] pyrene with benzoyl-
peroxide may represent either electrophilic or free-radical substitution
reactions. Radicals react rapidly with polycyclic aromatic hydro-
carbons.222'445 Anthracene reacts with benzoylperoxy radicals,
formed in benzaldehyde autoxidation, to terminate the chain and
produce anthraquinone.222
-------
Chemical Reactivity
79
(final product)
Anthracene
(16)
Anthraquinone
(17)
Oxygen atoms react rapidly with aromatic compounds and behave
like electrophilic reagents.319 Although poly cyclic aromatics have not
been studied in this regard, alkylbenzenes yield phenols. Toluene is
typical, yielding a mixture of cresols and phenol (dealkylation).
Ethylbenzene yields a small amount of acetophenone, in addition to
similar products. Polycyclic aromatic hydrocarbons will certainly
react similarly.
Considerable evidence indicates that destruction of aromatics in
soot is accelerated by photochemical smog, although which con-
stituents are responsible is not known.260'262 Soil bacteria oxidize
benzo[a] pyrene; in some conditions, 80% can be destroyed in 8
days.597
ACTION OF NITROGEN OXIDES
Polycyclic aromatic hydrocarbons, especially the larger ones, are
extremely sensitive to electrophilic substitution and to oxidation.
Nitrogen oxides or dilute HNO3 can either add to, substitute in, or
oxidize polycyclic aromatic hydrocarbons. Anthracene (16) is oxi-
dized by dilute aqueous HNO3 or nitrogen oxides to anthraquinone;793
NO2 adds to and substitutes in anthracene, yielding compound (53)
[by loss of HNO2 from compound (52)] ,33 Benzo[a] pyrene is nitrated
in minutes at room temperature with nitric acid diluted by acetic acid
and benzene to give the mononitro compound.267
o
dil. aq.
HNO3
or NO,
Anthracene
(16)
Anthracene
O
Anthraquinone
fl-HNO,
-------
80 PARTICULATE POLYCYCLIC ORGANIC MATTER
ACTION OF SULFUR OXIDES
Reactions of polycyclic hydrocarbons with atmospheric SO2, SO3, or
H2 SO4 should be very facile, particularly in aerosols or when adsorbed
Sulfonation of pyrene proceeds readily at room temperature with con-
centrated H2 SO4 to produce a mixture of disulfonic acids;795 the con-
centration of H2 SO4 in aerosol droplets may be high enough to sulfo-
nate the more reactive aromatics. SO2 reacts with aromatic compounds
in smoke, with or without light.745 The products of reaction with SO2,
SO3, and H2 SO4 are sulfinic and sulfonic acids; these compounds are
water-soluble and hence will no longer appear in the benzene-soluble
fraction.
PHOTODYNAMIC ACTIVITY AND SINGLET OXYGEN
Photodynamic toxicity (ability to kill or damage organisms in the pres-
ence of light and oxygen) is associated with many carcinogenic and
some noncarcinogenic hydrocarbons.233 The photodynamic action of
many compounds has been reviewed.717 Some part of the photody-
namic toxicity is probably caused by singlet oxygen, generated by
energy transfer from the sensitizer.275'277 Singlet oxygen is also
formed by several reactions of ozone and possibly by reactions of
radical cations. In addition, some photoperoxides dissociate ther-
mally to yield singlet oxygen,805 and peroxyacetylnitrate hydrolysis
also produces singlet oxygen.720 All these reactions provide the pos-
sibility of causing nonphotochemical biologic damage of the sort
associated with photodynamic action.
It has been concluded that singlet oxygen is probably not an im-
portant atmospheric oxidant of olefins,367 but that conclusion does
not take into account any of the methods of formation of singlet oxy-
gen mentioned here. Furthermore, the likelihood that some of these
mechanisms may produce singlet oxygen near or within the human
organism suggests that it cannot be ignored as a matter of environ-
mental concern.
SUMMARY AND CONCLUSIONS
Polycyclic aromatic compounds are highly reactive. There is evidence
that they are degraded in the atmosphere by photooxidation, by re-
-------
Chemical Reactivity 81
action with atmospheric oxidants, and by reaction with sulfur oxides.
Comparative data on reactions in solution, vapor, and adsorbed phases
are very limited, and most of the work has been done in solution. In
the few cases for which there are comparative data, the reactions are
similar in the different phases.
Reactions may be particularly facile when the compounds are ad-
sorbed on particulate material, such as soot. Chemical half-lives may
be only hours or days under intense sunlight in polluted atmospheres.
Further research is needed on the products of chemical reaction of
POM in typical atmospheric conditions and on the possible biologic
activity of these products. Most of the likely atmospheric reactions of
hydrocarbons produce oxygenated compounds. Such oxygenated
compounds as 7H-benz[d,e] anthracen-7-one (54) are found in urban
air,662 and oxygenated fractions of air extracts seem to be carcino-
genic.233 More definite information on chemical half-lives in various
conditions is essential.
Several mechanisms involving polycyclic aromatic hydrocarbons
and other pollutants may cause reactive oxidizing species to be de-
livered to genetic and other biologic material. These mechanisms de-
serve further study.
-------
Historical and Theoretical
Aspects of Chemical
Carcinogenesis
Chemical carcinogenesis was first discovered in 1776, when the Brit-
ish physician Percival Pott called attention to the high incidence of
cancer of the scrotum in the chimney sweeps of London and correct-
ly attributed the disease to their continual contact with soot. In the
almost 200 years since then, pure chemicals have been shown to in-
duce cancer in a wide variety of animals and sites. As our civilization
has become more industrialized, it has become increasingly con-
taminated with a number of cancer-producing chemicals in particulate,
as well as nonparticulate, form. Poly cyclic aromatic hydrocarbons
are found in particulate air pollutants and in condensates of the
cigarette smoke used as a self-pollutant. Among these compounds,
benzo[a] pyrene is a powerful skin carcinogen and has been detected
and analyzed in samples of polluted air. This compound is by no means
the only polycyclic aromatic carcinogen found in polluted air. If
these and other types of carcinogenic compounds were removed from
the environment, many human cancers would be prevented.
Not all the chemical carcinogens present in polluted air have been
identified. And a dominant role of the polycyclic aromatic hydro-
carbons in the causation of human cancer has not yet been established.
A great deal has been learned about the mechanisms of chemical
82
-------
Chemical Carcinogenesis 83
carcinogenesis, but the final answers still elude us. Although cancer
is a disease of organized tissues, it appears to be more productive at
this point to concentrate on the possible cellular and molecular mech-
anisms of chemical carcinogenesis, with particular reference to poly-
cyclic aromatic hydrocarbons.
Chemical carcinogens can be considered to act by two cellular
mechanisms: transforming or converting normal cells into cancer
cells349'350 and selecting for pre-existing cancer cells, as proposed by
Prehn.602 Although it may not seem too difficult to determine which
of these mechanisms is correct, the matter could not be settled by
studies in whole animals and had to await the development of reliable
and quantitative systems for producing chemical carcinogenesis in cells
in culture. Mondal and Heidelberger537 have succeeded in transform-
ing individual single normal cells to malignant cells with 3-methyl-
cholanthrene with very high efficiency, thus ruling out the selection
hypothesis in this system.
The possibility that the carcinogen activates or "switches on" a
latent oncogenic virus has recently been proposed in its most com-
prehensive form by Huebner and Todaro.399 They postulate that all
chemical carcinogens act through intermediary or oncogenic viruses
or their informational precursors. There is evidence that in some cases
chemical carcinogens do lead to the appearance of detectable onco-
genic viruses. These cases appear to involve primarily leukemogenesis
and some sarcoma formation. In many other situations, there is no
evidence of the participation of an oncogenic virus in chemical carci-
nogenesis. However, it may be impossible to disprove such participa-
tion. It should be pointed out that only the mouse mammary tumor
virus is now known to induce carcinomas. If viruses are "switched on"
by chemical carcinogens and give rise to the many carcinomas that
chemicals are known to induce, then they must be viruses that are as
yet unknown. The possibility that oncogenic viruses are involved in
chemical carcinogenesis must always be kept in mind; but, even if the
participation of a virus turns out to be ubiquitous in chemical car-
cinogenesis, it is still the chemical that triggers the process of cancer
cell formation. Therefore, for human health protection, environmental
chemicals should be reduced or eliminated to prevent potential carci-
nogenesis.
If it is assumed that chemical carcinogens themselves produce can-
cer without the intervention of an oncogenic virus, then two major
molecular mechanisms are possible—mutational and nonmutational.
The somatic mutation theory of cancer was first proposed by
-------
84 PARTICIPATE POLYCYCLIC ORGANIC MATTER
Boveri81 in the 1920's. In modern terms, the mutational mechanism
of chemical carcinogenesis requires that the chemical interact with the
genetic material of the cell (D N A) in such a way as to alter its pri-
mary sequence to lead to a nonlethal perpetuated change that would
be inherited by all progeny cells. This has been discussed by many,
including Potter.601 It has been shown with polycyclic aromatic com-
pounds that there is a weak physical binding to DNA, which can be
converted into covalent binding by ultraviolet irradiation and various
chemical treatments, as studied by Ts'o and his colleagues.758 Prob-
ably more significant with respect to the process of carcinogenesis
is the fact that every chemical carcinogen of every type that has been
properly studied has been found to bind covalently with the DNA,
RNA, and protein of target tissues; this subject has been thorough-
ly reviewed by Miller.532 Because the hydrocarbons themselves
cannot undergo such covalent binding, it is clear that they must
be metabolically converted by target cells into chemically reactive
molecules, which may be considered as the "ultimate" carcinogens.
In the case of polycyclic hydrocarbons, their binding to the DNA
of mouse skin after topical application has been measured by
Brookes and Lawley93 and Goshman and Heidelberger.307 They
found that there was a rough quantitative correlation between the
carcinogenic activities of various hydrocarbons and the amount
bound to mouse skin DNA. The binding of carcinogenic hydrocar-
bons to DNA does not, however, prove that they act through a muta-
genic mechanism. The mutagenic activities of many carcinogenic and
noncarcinogenic compounds have been determined in many systems,
including recent studies in bacteriophage T4 by Corbett et al.163 In
none of these studies has the correlation between mutagenesis and
carcinogenesis been good, but that may be a result of deficiencies in
the various test systems used, rather than of an intrinsic lack of cor-
relation between the two biologic processes.
A nonmutational molecular mechanism for chemical carcinogenesis
would require a perpetuated epigenetic change, which would result
in altered gene expression. This would most likely be a derepression
of genetic information already present in the genome of the cell. A
theoretical molecular model for such a derepression was proposed
in 1963 by Pitot and Heidelberger,596 on the basis of the observation
that carcinogenic hydrocarbons are covalently bound to a specific
protein fraction in mouse skin.744 The derepression of an oncogene
has also been proposed by Huebner and Todaro399 in their compre-
hensive theory of carcinogenesis. Braun87 also supports the view that
-------
Chemical Carcinogensis 85
carcinogenesis results from a nonmutational mechanism. Immunologic
evidence for derepressions accompanying carcinogenesis has been ob-
tained by Abelev3 in mouse hepatomas and by Gold et al301 in hu-
man gastrointestinal carcinomas. These workers demonstrated that
the tumors contained fetal antigens that were repressed in the adult
animal but reappeared in the tumors. If carcinogenesis is the conse-
quence of a nonmutational mechanism, then there is a possibility of
reversion of cancer cells to normal. Indeed, there is some evidence
for this in a few cases, including neuroblastomas and the teratocar-
cinomas studied by Pierce.594
Another intriguing phenomenon accompanying carcinogenesis,
for which there is at present no satisfactory theoretical explanation,
is the acquisition of new transplantation antigens on the surfaces of
chemically induced tumors. This has been extensively studied with
hydrocarbon-induced sarcomas in mice by Prehn,604 Klein,438 and
Old etal.;5"71 with hydrocarbon-induced mouse skin carcinomas by
Pasternak et a/.;584 with azo dye-induced rat hepatomas by Baldwin
and Barker;28 and with in vitro hydrocarbon carcinogenesis by
Mondal et al.538 These tumor-specific antigens of chemically induced
tumors are individual and non-cross-reactive. Although the total num-
ber of such antigens is not known, the number exceeds 25 in the case
of hydrocarbon-induced sarcomas. The significance of these antigens
to the process of carcinogenesis remains unexplained.
CONCLUSIONS
1. Chemical carcinogens appear to transform normal cells into
cancer cells directly.
2. Chemical carcinogens may or may not "switch on" a latent
oncogenic virus that is responsible for the cancer induction.
3. If chemical carcinogens transform normal to cancer cells with-
out the intervention of an oncogenic virus, they can do so either by
a mutational or by a nonmutational mechanism. There is some evi-
dence to support each possibility.
4. Hydrocarbon-induced tumors have individual transplantation
antigens.
5. The response of a host to carcinogenic stimuli is determined by
its immunologic and hormonal status, its exposure to some drugs,
and its nutritional state. A variety of unknown factors in its envi-
ronment may also alter its response to carcinogenic stimuli.
-------
86 PARTICULATE POLYCYCLIC ORGANIC MATTER
RECOMMENDATIONS
Further and more definitive work must be done on the interrelations
between chemical carcinogens and oncogenic viruses. The question of
whether carcinogenesis involves a mutational or a nonmutational
mechanism can be settled only by genetic experiments. Hence, much
development must be carried out aimed at understanding the genetics
of mammalian cells. Further biologic, biochemical, and molecular bio-
logic research must be carried out in whole animals and cell cultures
undergoing chemical carcinogenesis and in various cell-free systems.
An understanding of the fundamental mechanisms of chemical car-
cinogenesis should provide new means for the eradication of cancer
through prophylaxis, induction of loss of malignancy, and perhaps
even reversion of malignant cells to normal cells.
-------
Experimental Design in
Carcinogenesis Tests
STATISTICAL DESIGN AND ANALYSIS
Laboratory testing of carcinogens has been rather straightforward
because of its emphasis on identifying strong carcinogens; experi-
ments could be short and could use small numbers of test animals.
However, the role of weak environmental carcinogens in imposing
subtle but potentially important threats to human populations has
required changes in experimental procedures.
Mantel and Bryan517 describe a method of extrapolating to a con-
servative safe dosage of an agent, whether or not carcinogenic activ-
ity was found. The dosages and number of test animals used may
influence the result of the extrapolation. In the extrapolation proce-
dures, a worst true risk at the test level used consistent with the ob-
served risk is determined; e.g., no tumors observed among 100 ani-
mals would be consistent at the 99% level of probability with a true
risk of 4.5%. This worst true risk is extrapolated backwards by some
arbitrarily shallow rule (e.g., 1 probit per tenfold change in dose) to
an arbitrary level of "virtual" safety (e.g., a risk of one in 100 mil-
lion). The "safe" dosage, where the observed risk is none in 100,
would be equal to 1/8,300 times the test dosage. An alternative, but
87
-------
88 PARTICULATE POLYCYCLIC ORGANIC MATTER
parallel, approach to setting "safe" dosage, which can be yet more
conservative, is described in Gross et al.317
A more general purpose of an experimental design is to identify
the carcinogenic properties of the agent tested. For strong carcino-
gens, it should reveal that tumors are elicited early and at a relatively
low dosage; for weak carcinogens, that few tumors are ever obtained,
even at a high dosage. Data resulting from such a general-purpose de-
sign could be variously used, e.g., for identifying weak or strong car-
cinogens and environmental or nonenvironmental carcinogens and for
correlating carcinogenic activity with other physicochemical proper-
ties of agents. Such a general-purpose design seems appropriate if
one wishes to identify carcinogenic activity in atmospheric pollu-
tants, using either crude particulate matter or specific extracts.
A general-purpose design used by Hadidian et a/.327 and outlined
in Mantel516 permits testing over a wide dosage range to identify both
weak and strong carcinogens, with test animals (rats) followed for 1 l/z
years, so that both early- and late-appearing tumors would be found.
In this design, testing was begun by using an agent at five different,
but increasing, log dilutions. On the basis of the observed 30-day toxi-
city, six additional half-log dosages were inserted, depending on the
toxic level, above, between, or below the original test dilutions. For
both initial and additional test groups, treatment was maintained for a
full year. The study was intended to use three male rats and three
females per dosage. The small groups were expected to be satisfactory,
because the information in the experiment would be based on the total
number of rats in all the groups, not on those at a single dosage. Be-
cause of a particular interest in weak carcinogens, the test design
was modified so that 15 rats of each sex were used at the next-highest
dilution of the retained nontoxic dosages. From a review of the result-
ing data, however, it was clear that the unmodified and modified ex-
periments led to essentially the same conclusions.
The complex results, which involved a wide variety of tumors, days
of tumor appearance, days of death or sacrifice, and so on, were anar
lyzed by a simplifying approach. A reference tumor rate was estab-
lished for controls that survived at least 1 & years, so that the expected
number of tumors, overall or for a particular site, among 30 male
or 30 female rats could be determined. If the 30 rats of a sex treated
with a particular agent (for all dosages combined) showed clearly above-
expected numbers of tumors, any adjustment for early mortality
among the rats could only increase, rather than diminish, the dif-
ference. The fact that tumors are competitive with each other, so
-------
Experimental Design in Carcinogenesis Tests 89
that death from a tumor at one site can preclude seeing a tumor at
another site, was handled by considering the total of tumors or
tumorous rats. Data on actual tumor sites were, however, available
for interpretation. It became necessary to eliminate such spontaneous
tumors from the totals in order to bring out neoplastic effects for
the agents tested; thus, a comparison of 21 tumors observed versus
18.5 expected could become 16 observed versus 1.6 expected when
interstitial cell tumors were excluded. This modification served as
a crude form of age adjustment.
The foregoing illustrates the importance of and need for flexibility
in the analysis of data from a long-term carcinogenesis experiment. The
resulting data will ordinarily not be neatly packaged, and there will
probably be no completely valid way of dealing with all the compli-
cations. In the present instance, the method of analysis cut across the
complications so as to fasten on a simple indicator of carcinogenic
effect. Flexibility was illustrated in the elimination of spontaneous
interstitial cell tumors, whose high incidence in a long-term experiment
had not been anticipated. If an agent showed active carcinogenicity,
eliciting early tumors, the rats would be prevented from surviving to
the more advanced ages at which spontaneous tumors arose; overall,
then, the total number of tumors might be increased only moderately.
The remedial device of eliminating the spontaneous interstitial cell
tumors brought out the carcinogenic effect clearly.
Another large-scale, long-term test of carcinogens, in mice, was
reported by Innes et a/.412 The interest focused on pesticides and
some other industrial chemicals. Agents were first tested for toxicity
to determine a maximal tolerated dosage for each. With the empha-
sis on bringing out the existence of carcinogenic effects, only this
maximal dose was used in testing, which was done with 72 mice di-
vided among two strains and two sexes. Summary totals of mice with
tumors were used in making comparisons with controls, but detailed
information of kinds of tumors observed was also kept. The statisti-
cal analysis reported included both a significance test—i.e., whether
the observed increase in tumor incidence was real-and a relative-
risk measure of that increase. The relative-risk measure varied with the
dosage administered for testing and, because of dosage differences,
is not strictly comparable between agents.
In the two studies just described, the agents tested were adminis-
tered orally to weanling rats or week-old mice. For other studies, a
wider variety of routes could be of interest. The use of more sensitive
neonatal test animals or of less sensitive young adult animals may be
-------
90 PARTICULATE POLYCYCLIC ORGANIC MATTER
of value. It may be of interest to test some agents in conjunction with
known carcinogens. With limited resources, just which agents to test
constitutes a problem of priorities. In a general chemical-testing pro-
gram, does one give highest priorities to chemicals already in use—
with priority related to the extent of such use—or does one give high-
est priorities to new chemicals, which can bring about an abrupt
change in the environment? The answer may be to have a program
large enough to test all agents, in time, with more detailed testing of
particular agents. In the case of atmospheric pollutants, the priority
problem applies to the program as a whole, rather than to individual
pollutants.
A characteristic of both studies described is that they emphasized
testing at relatively high dosages, although the first involved some low
dosages as well. For various reasons, low-dosage studies will yield
unsatisfactory and perhaps ambiguous results. For example, suppose
one tests at dosages comparable with actual human exposure; one
would likely get negative results in a reasonably large experiment, even
if the risk for humans were intolerably high—say, 0.1 or 0.01%. (Be-
cause of the large number of agents in the environment, the risk for
each, when not otherwise justified, must be kept extremely small.)
A low-dosage experiment cannot resolve the question of the existence
of thresholds. At a very low dosage, the absence of induced tumors would
not demonstrate a threshold, because it could be a chance occurrence
where the tumor risk is low; and the occurrence of tumors would not
belie the existence of a threshold, because they could be spontaneous,
rather than induced, tumors. If testing is done at several dosages and
the induced tumors are rather few, compared with spontaneous tumors,
proponents of a threshold theory could find support in the apparent
constancy of the tumor rate. An alternative and equally mistaken in-
terpretation could be that the dose-response curve is so shallow that
extreme extrapolation to much lower dosages would be necessary
to attain a safe level. Experimentation to prove or disprove the exis-
tence of thresholds seems pointless.
In summary, a carcinogenicity-testing program is shaped by a
variety of factors, including the purpose of the program. The purpose
may be to determine safe dosages, to identify agents with interesting
properties for future investigation, to identify specifically potent car-
cinogens, or even to identify specifically weak carcinogens. Screening
experiments with many dosage levels can be informative, even if the
numbers of subjects at individual dosages are small. Data from a car-
cinogenicity experiment can be complex, but there may be ways of
-------
Experimental Design in Carcinogenesis Tests
91
simplifying the data for analysis. Flexibility of analysis may be re-
quired to overcome unanticipated problems, such as a high rate of
spontaneous tumors among controls. The existence or nonexistence
of thresholds in carcinogenesis cannot be established solely by testing
programs, and experimentation solely to such ends would be wasteful.
IS THERE A THRESHOLD DOSE IN EXPERIMENTAL
CHEMICAL CARCINOGENESIS?
Although dose-response studies have been carried out with many
chemicals and the problem of quantitative carcinogenesis has been
extensively studied for a few selected carcinogens,308'440 it has
not been possible to reach agreement as to whether there is a threshold
dosage above which carcinogenesis is produced. The dilemma is re-
lated to the shape of the dose-response curve. Curve A in Figure 7-1
represents an idealized linear relation, and curve B an idealized thresh-
old relation. Because it is impossible, at very low doses, to obtain
reliable data without enormous numbers of animals, a curve like C
(which is a hybrid between A and B) might be obtained with insuf-
ficient numbers of animals. Therefore, the concept of a threshold dose
is probably meaningless, and it would be prudent, because of these un-
certainties of measurement, to extrapolate dose-response curves to
zero in a linear fashion.
To gain insight into this situation, it should be remembered that
there are probably at least two events involved in chemical carcino-
Incidence
dose •• dose •• dose •
FIGURE 7-1 Possible dose-response curves for agent dose and carcinogenesis.
A, idealized linear relation; B, idealized threshold relation; C, hybrid curve be-
tween A and B.
-------
92 PARTICULATE POLYCYCLIC ORGANIC MATTER
genesis. The first is the primary insult induced intracellularly by the
carcinogenic chemical. This molecular process is (according to the
available evidence) rapid and irreversible. This phase is followed by
the biologic process or processes involved in the conversion of the
premalignant cell(s) to malignant cell(s) or clone(s) of cells, which in
turn results in a tumor. With some carcinogens, the amount of mate-
rial required to bring about the primary molecular event is so small
that it is experimentally difficult to determine an apparent no-effect
dosage.
Some investigators maintain that for the so-called second phase of
carcinogenesis—i.e., the actual induction of a tumor—there is a
threshold dosage. Druckrey216 studied the response to dimethylamino-
azobenzene given orally to rats and found that the lowest dosages,
0.1 and 0.3 mg/day, in 150 rats did not induce any tumors; at 1 mg/
day and at higher dosages, tumors were induced. Similar studies were
also carried out with aromatic hydrocarbons.97 However, several fac-
tors have to be considered in experiments aimed at determining no-
response dosage. Because it is the usual pattern that the lower the dos-
age the longer the induction time, it follows that the age of the ani-
mal at the time of carcinogen treatment is important. A low dosage
or a weak carcinogen may require a period of treatment beyond the
life-span of the animal; hence, the duration of treatment, especially
with low dosages of weak carcinogens, is important.
The group size is extremely important; as far as human exposure
is concerned, an agent cannot be regarded as harmless on the grounds
that it did not induce tumors in a group of 50 humans (this figure is
used because in many carcinogenicity assays only 50 animals are used
per group). If one considers a weak carcinogen as one that causes
tumors in 0.1% or 0.01% of the test animals, 1,000 and 10,000 animals,
respectively, will have to be used to detect these agents, and that as-
sumes that a finding of only one animal with a tumor will be regarded
as significant. It is impractical at present to test substances on groups of
this size, so compounds are usually tested at much higher dosages than
normally encountered in the environment and much smaller test
groups are used; this is necessary in an attempt to reduce the gross
insensitivity of animal test systems as a function of restricted sample
size. Positive results thus obtained are justifiably extrapolated to hu-
man populations exposed to marginal doses.
The problem of threshold values is further complicated by the con-
siderable evidence that the effect of a single exposure is irreversible
and that the effect of several exposures is cumulative. In addition, in
-------
Experimental Design in Carcinogenesis Tests 93
environmental exposure the human population comes into contact
with carcinogens, cocarcinogens, and less well-defined cofactors. Cig-
arette-smoke condensate constitutes a good example of an environ-
mental cocarcinogen or tumor-promoting agent.792
Neither epidemiologic nor experimental data are adequate to fix
a safe dosage of any chemical carcinogen below which there will defi-
nitely be no tumorigenic response in humans.95'370'518 For these
reasons, synthetic chemicals, such as food additives and pesticides
that are known to be carcinogenic, must not be deliberately added
to the environment. With regard to air pollutants, which contain a
variety of defined and undefined carcinogens, the lowest possible ex-
posure must always be insisted on.
ATTEMPTS TO EXTRAPOLATE ANIMAL DATA TO MAN
Attempts to extrapolate from experimental data to human exposure
are fraught with problems. One is that animal test groups are neces-
sarily limited in numbers to, say, 50 per dosage group. As noted be-
fore, restricted test groups necessarily are grossly insensitive. For ex-
ample, if an environmental agent produces cancer in one of 10,000
men, and if it is assumed that sensitivity to the carcinogen in question
is similar in man and rodents, then test groups of 10,000 rats or mice
would be required to obtain one cancer. For statistical significance,
50,000 rodents would be required. Furthermore, in any particular
instance, humans may be more or less sensitive to the carcinogen in
question than are rodents.
Apart from the gross insensitivity of animal test systems as a
function of restricted sample size, a wide range of possible interac-
tions and synergisms would obtain in natural human exposure, in con-
trast with artificial laboratory systems. Thus, it is impossible to de-
termine safe levels of human exposure to any known or unknown
carcinogens on the basis of supposed no-effect levels in practical num-
bers of animals. Human experience has provided valuable post hoc
information from epidemiologic studies.
For these reasons, animal test systems using the latest sensitive
procedures are the only methods available for defining and antici-
pating human hazards from POM. A number of factors that influence
such tests are described in Chapters 8 and 9.
Classic test methods have used adult rodents, but the use of new-
born animals appears to offer greatly increased sensitivity, which will
-------
94 PARTICULATE POLYCYCLIC ORGANIC MATTER
result in the conservation of valuable and scarce test materials. Routes
of administration, such as inhalation and intratracheal instillation of
materials adsorbed to particles, may serve as simulant models for hu-
man exposure to air pollutants. Preliminary information suggests that
carcinogenesis experiments in primates are practical and might yield
results that can be extrapolated to man. Coordinated studies in ex-
perimental animals of toxicity, teratogenicity, mutagenicity, and car-
cinogenicity offer an optimal evaluation of the hazards of air pollutants.
The use of in vitro cell or organ cultures has the potential of pro-
viding additional, and possibly more sensitive, methods to detect
toxic, carcinogenic, and mutagenic activities of air pollutants. Deter-
mination of the levels of production of hydrocarbon-metabolizing
enzyme in various selected human cells and tissues may also assist in
assessing the sensitivity or resistance of humans to air pollutants.
In all test systems, it is essential not only to test for the activity
of benzo[a]pyrene and other polycyclic organic air pollutants, but
also to test crude materials and their fractions and to isolate and
characterize their carcinogenic components.
-------
8
In vivo Tests for
Carcinogenesis and
Cocarcinogenesis
PRACTICAL ASPECTS OF TESTING FOR HYDROCARBON
CARCINOGENESIS IN MICE AND RATS
A major purpose of experimental air pollution carcinogenesis studies
has been to determine whether urban air pollutants might prove to
be carcinogenic to different animal species and tissues and, if so,
which of the compounds is primarily responsible for the activity. Once
agents that are carcinogenic to experimental animals have been identi-
fied, attempts can be made to reduce and, if possible, to eliminate
their emission into the environment.
Established Polycyclic Carcinogens
Bioassays on mouse skin,no»147>388>404>450'453'582>632 subcutaneous mouse
tissue,404'487'488 and mouse cervix72 and in newborn mice241 have
shown that the particulate matter of city air can be carcinogenic to
experimental animals. Fractionation studies on urban pollutants sug-
gest that polycyclic aromatic hydrocarbons may play an important
role in the overall carcinogenicity and tumor-initiating activity of
urban pollutants in experimental animals.388'404'633'829 Carcinogens
95
-------
96 PARTICULATE POLYCYCLIC ORGANIC MATTER
are also present in nonpolycyclic aromatic compound fractions, such
as oxyneutrals or aliphatics. At present, experimental evidence that
neutral and basic N-heterocyclic hydrocarbons contribute to the overall
carcinogenic activity of urban air is lacking. Nevertheless, the well-
established carcinogenicity of some N-heterocyclic hydrocarbons346-
699,700 and tneir identification in urban air require that they be con-
sidered as carcinogenic agents in POM.
Application to the Skin of Mice
There are several reasons for choosing mouse skin as a test organ, in-
cluding the relatively low cost of the bioassay and the relatively low
demands for pure chemical compounds, especially if one uses inbred
mouse strains with high susceptibility to these carcinogens.830
Several review articles80-88'830 and books 17.227,403,812,828 have dis.
cussed details of the experimental conditions for mouse skin bio-
assay. In brief, the more important factors are mouse strain and
sex, solvent, dosage and mode of application, and method of obser-
vation.
•MOUSE STRAINS FOR CARCINOGJEJSICITY TESTING
Genetically homogeneous strains of mice have been developed by
selective inbreeding. The a'dvantage of inbred strains is that, unlike
random-bred animals of the same species, they show a constant bio-
logic response. Some of the commonly used strains of mice are A,
BALE, C 3H, C57BL, DBA, and Swiss. Some of these strains are sus-
ceptible to spontaneous tumors in particular organs. Strain A mice,
for example, develop spontaneous lung tumors and have often been
used for lung tumor induction by chemicals. A skin-tumor-susceptible
strain, STS, is also available, but it has a high incidence of sponta-
neous: mammary tumors, which makes it less,desirable for long-term
testing. When using inbred strains for carcinogenicity assay, it is im-
portant that no-treatment control groups be used to assess the sponta-
neous tumor incidence. For the widely used inbred strains, the sponta-
neous tumor incidences are well established.190
Some workers still prefer to use random-bred mice when evaluat-
ing new compounds for carcinogenicity, because the induction of
tumors in a given inbred strain of mice may be open to question. The
use of random-bred mice is probably also more relevant to human ex-
posure to environmental agents. With random-bred strains, it is desir-
-------
In vivo Tests for Carcinogenesis and Cocarcinogenesis 97
able to use larger groups of animals when testing new compounds for
carcinogenicity.
A few years ago, Boutwell developed the STS strain of mice, which
appears to be especially suitable for the assay of weak carcinogens,
tumor-initiators, and tumor-promoters. However, data have not been
available to demonstrate whether reproducible results can be obtained
with this strain. To obtain statistically significant data in air pollution
Carcinogenesis, a given sample must be tested on the skin of at least
20 mice. Because male mice are known to fight, most investigators
use only female mice for the skin tests.
SOLVENT
In general, the highest activity on mouse skin is obtained when acetone
is used as a solvent. However, mixtures of benzene in acetone must
occasionally be used to dissolve the organic pollutants. Dioxane,
H-hexane, and other solvents are not recommended for the mouse
skin test.
CONCENTRATION
A standard rule for the concentration of organic pollutants that must
be applied to mouse skin for a carcinogenic response cannot be speci-
fied. However, it appears appropriate to apply a dose that contains at
least 20% of the amount of benzo[a] pyrene required to induce a low
tumor yield in a given strain of mice. In the case of female Ha/ICR/Mil
(Swiss random-bred albino) mice, this requires a 0.001-0.002%
acetone solution of benzo[a] pyrene. In the past, the test solutions were
often painted onto the shaven intrascapular area of the back (about
2X2 cm) with a No. 5 camel's-hair brush. However, application of
a 50-/xl solution should produce more reproducible data than those
obtained with skin painting. This can be done manually with a micro-
pipette or, preferably, with an automatic applicator.828
Environmental respiratory carcinogens are often derived from com-
bustion. They contain traces of carcinogens, tumor-initiators, and
tumor-promoters. The most reproducible data are obtained if one
starts the application during the second telogen (resting) phase of the
hair cycle, when the mice are about 6-8 weeks old.47'48 In the past,
test solutions with organic pollutants have been applied three times
a week to the skin of mice. Under these conditions, one has to test
the material for some 8-12 months to obtain a significant number of
-------
98 PARTICULATE POLYCYCLIC ORGANIC MATTER
papilloma- and epithelioma-bearing mice. However, solutions of up to
20% organic pollutants were reported not to induce acute toxic ef-
fects in mice, and significant tumor yields might be obtained earlier
if the solution is applied five times a week.
MODE OF OBSERVATION
The weight of the mice should be recorded at the outset of the ex-
periment and again every 2 weeks (or at least every 4 weeks). Benign
skin tumors should be recorded when they have attained a diameter
of 1 mm. They enlarge by nodular growth (papilloma) or by lateral in-
vasion (carcinoma). Some may not enlarge; some will regress. Tumors
that remain 1 mm or grow larger for 21 consecutive days are counted
and become the raw tumor-yield data for tumor-bearing mice. Micro-
scopic lateral invasion of a tumor into adjacent skin is considered as
transformation into a carcinoma. Continued growth of such lesions,
however, is required before they can be recorded as macroscopically
observed carcinomas.
It is recommended that, on termination of the experiments, in
chemical carcinogenesis, the animals be examined for the histologic
nature of the skin lesion, for internal pathologic conditions, and es-
pecially for adenomas of the lung. The tumors on the test mice are
then compared with the incidence of spontaneous tumors in the con-
trol group.
The factors outlined above are specifically important in the bioas-
say of organic pollutants. Most other details are given in standard re-
views and books of chemical carcinogenesis17'227'403'812'828 and in
handbooks of breeding and management of laboratory animals.472
Subcutaneous Administration
Subcutaneous injection into mice and especially rats is the most widely
used form of parenteral administration in chemical carcinogenesis
testing. It can be chosen for a great variety of substances, including
the ones that are highly reactive, such as some alkylating agents, and
the ones that may decompose if administered orally or applied to the
skin. For testing a pure chemical, only a single injection into the sub-
cutaneous tissue of the nape of the neck is necessary. Therefore, only
small amounts of material are required—for example, 100 mg per mouse
for organic pollutants. Because a single monthly injection suffices
-------
In vivo Tests for Carcinogenesis and Cocarcinogenesis 99
for the screening of air pollutants, this method saves time and person-
nel, compared with the mouse skin test.
Homburger and Hsueh393'394 have demonstrated that the subcu-
taneous tissue of Syrian hamsters is susceptible to carcinogenic
aromatic hydrocarbons. One inbred line gave the most rapid tumor
production yet described in any rodent. With 0.5 mg 7,12-dimethyl-
benz[a] anthracene, the first tumor appeared after 5 weeks, and the
mean latency was 9 weeks. If the high susceptibility of connective
tissue to carcinogens is confirmed with other hydrocarbons, this
Syrian hamster inbred line may become a valuable test animal in air
pollution carcinogenesis testing.
Oral Administration
For testing by oral administration, the test substance can be incor-
porated into the diet, dissolved in the drinking water, or force-fed by
a stomach tube. In general, the oral administration of polycyclic
aromatic hydrocarbons, especially those present in the respiratory
environment, is a relatively poor way of determining their carcino-
genicity. However, Shay et al.696 and Muggins et a/.406 reported that
some strains of rats, especially the Sprague-Dawley, are highly suscep-
tible to some types of carcinogens (polycyclic aromatic hydrocarbons
and aromatic amines) when the substances are force-fed in an oily
vehicle. Even a single large dose induces mammary carcinoma rapidly
(2-6 months). If a week carcinogen is used, the test may require
repeated feeding. The agent to be tested should have low water and
high lipid solubility. This feeding method with Sprague-Dawley rats
merits special consideration for a rapid screening test of polycyclic
aromatic hydrocarbons.
Inhalation
The induction of a significant number of lung adenomas in mice
(C57BL) has already been demonstrated with artificial smog (ozonized
gasoline) by Kotin et a/.451'456 and Nettesheim etal.560 The latter
group also reported that male mice have a significantly higher lung
adenoma incidence than female mice when exposed to ozonized
gasoline or an insoluble chromium oxide dust. Gardner289 reported
a slight increase of lung adenomas in mice exposed to Los Angeles
ambient atmosphere.
-------
100 PARTICULATE POLYCYCLIC ORGANIC MATTER
Bladder Implantation
During the last decade, several studies have been reported on the
implantation of paraffin or cholesterol pellets containing potential
carcinogens into the bladders of mice.10-509 Recently, the technique
has been further developed and standardized by Ml,419 Boyland and
Watson,86 and Bonser et a/.74 With this technique, the sensitivity of
the bladder epithelium to tumorigenic stimuli can be used for routine
testing of some potential carcinogens. However, bladder implantation
is clearly not an appropriate primary route for testing air pollutants.
In view of recent epidemiologic evidence that the incidence of bladder
cancer is higher in areas of heavy air pollution, bladders should be
examined by appropriate techniques96 after administration of air
pollutants by conventional routes.
Factors Influencing Hydrocarbon Distribution in the Host
PARTICLE SIZE
In polluted air, particles occur in sizes ranging from 0.001 to 10,000
jum in diameter. In general, controlled combustion, the major source
of carcinogenic POM in polluted air, produces particles 0.1-10 jum
in diameter.388 In several studies, "lung-damaging" components are
considered to be particles 0.25-10 jum in diameter.388 Particles less
than 0.25 (an in diameter are retained to a low degree in the lung,
and particles with diameters greater than 10 Mm lodge in the upper
respiratory tract and thus do not reach the bronchi. Although air
particles can be separated according to size, separated particles have
thus far not been properly analyzed for polycyclic aromatic hydro-
carbons. Because practically every combustion leads to the forma-
tion of traces of carcinogenic polycyclic hydrocarbons and the parti-
cles generated by combustion are primarily 0.1-10 jum in diameter,
it can be assumed that POM will be partially deposited in the bron-
chial tree, especially at the bifurcation.
RETENTION AND ELUTION OF PARTICLES
During normal breathing, the lung retains particles 0.25-5 jum in
diameter, with maximal retention of 80% of 1-p.m particles and less
than 5% retention of particles smaller than 0.1 pm and larger than
5 jum. Falk et a/.258 demonstrated experimentally that benzofa] pyrene
-------
In vivo Tests for Carcinogenesis and Cocarcinogenesis 101
and other polycyclic aromatic hydrocarbons are readily eluted from
soot samples recovered from the human lung. Carcinogenic hydro-
carbons in polluted air appear to be adsorbed primarily on particles
that, according to their size range, are compatible with deposition of
and retention of a portion of the carcinogenic particles in our urban
respiratory environment.
Asbestos fiber has received considerable attention in recent years
as an air pollutant.59 Oil containing carcinogenic hydrocarbon is
absorbed by these fibers, especially the 7-chrysotile form. These hy-
drocarbons are first retained with the particles, readily eluted, and
then retained in the lung tissue. This process of retention of carcino-
gens in lung tissue is accelerated if the respiratory air contains irritants
that impinge on the bronchial epithelium.
CHANGES IN CILIARY MOVEMENT AND MUCOUS VISCOSITY
Several theories have been proposed for the importance of cilia in
toxicity in respiratory carcinogenesis. In air pollution carcinogenesis,
the concept of Hilding375 appears most relevant. This author emphasizes
that, in the pathogenesis of bronchogenic cancer, in both man and
animal, without ciliary stasis and concomitant mucous stagnation,
subsequent metaplasia from ciliated to squamous epithelium and then
to epithelial cancer is not likely to occur. These changes in mucous
and ciliary movement, changes in mucous viscosity, and changes in
the underlying basal layer of the epithelium are caused by irritants
in the respiratory air. The irritant effect seems to be essentially non-
specific, in that chemical, physical, and viral agents are capable of
inducing the changes. The large spectrum of irritants in our respira-
tory environment originates primarily from polluted air and cigarette
smoke. These two inhalants also contain carcinogenic hydrocarbons
and therefore increase the likelihood of combining biologic action
through either simultaneous or sequential inhalation.
The most widely found and more important volatile irritants in
our environment are formaldehyde, acrolein, formic acid, acetic acid,
peroxy acids, volatile phenols, ozone, nitrogen oxides, sulfur dioxide,
and hydrogen cyanide.
Conclusions and Recommendations
Three major groups of agents carcinogenic to experimental animals
have been identified or suggested in urban pollutants: polycyclic aro-
-------
102 PARTICULATE POLYCYCLIC ORGANIC MATTER
matic hydrocarbons, to a minor extent N-heterocyclic hydrocarbons,
and oxygenated neutral compounds of unknown structure. Until now,
the most valuable data from carcinogenic bioassays of organic pol-
lutants were obtained with tests on mouse skin and connective tissue.
To gain statistically significant bioassay data, at least 20 g of organic
particulate matter is needed. That amount of material can be filtered
from air in a reasonable time only with special and expensive equip-
ment. Other methods of assay requiring less material should be
explored.
The urban environment is known to contain traces of carcinogenic
agents, which should impinge directly on the bronchial epithelium
of the experimental animal.
The application of artificial atmospheres in model systems, such as
those using benzo[a] pyrene and sulfur dioxide,474 requires detailed
exploration. However, direct inhalation studies in animals with ambient
atmospheres may not provide useful data because of the high dosage
and enormous animal populations that would be required.
It is suggested that the carcinogenicity of POM be assayed by
direct intratracheal instillation in hamsters and rats in a model system
recently developed by Saffiotti.648 Direct intratracheal instillation
in rats may also be used to test a wide range of fractions and sub-
fractions of POM with Freund's adjuvant.
COCARCINOGENESIS, ANTICARCINOGENESIS, AND RELATED
ASPECTS
Research on organic air pollutants and their possible contribution
to lung cancer in man is probably far behind current thoughts in
tobacco use and its relation to lung cancer in man. Some years ago,
there was a single aim and purpose in tobacco research: to quantitate
benzo[a] pyrene in cigarette tars so that methods could be devised
for removing it from the smoke. During the last decade, it has become
evident that, if benzo[a] pyrene has any role at all in tobacco carcino-
genesis, it is not necessarily the most significant and that other car-
cinogens, and particularly cocarcinogens, must play an important
role. The same is probably true of organic air pollutants.
Definitions
"Cocarcinogenesis" is the process whereby cancer is induced in ani-
mal or man by the combined action of two or more agents, either
-------
In vivo Tests for Carcinogenesis and Cocarcinogenesis 103
by a single exposure or, as is more common in both the laboratory
and the environment, by repeated exposures. "Tumor-initiating
agents" are agents (which may or may not be carcinogenic by them-
selves) that, when given as a single exposure—and in the case of carcin-
ogens at a subcarcinogenic dose—and followed by a low exposure
to a tumor-promoting agent, result in tumors either at the site of ap-
plication of the initiating agent or at a distant site. "Tumor-promoting
agents" by themselves, when given in repeated doses, are usually non-
carcinogenic or at most very weakly carcinogenic. The sequential
process of induction of tumors by a single dose of an initiating agent
followed by repeated low-level exposure to tumor-promoting agents
is referred to as "two-stage carcinogenesis." The latter term is some-
what misleading, in that mechanistically multiple steps are most prob-
ably involved in this process of tumor induction, and the term refers
more specifically to the two stages of treatment with chemicals.
In cocarcinogenesis experiments, two or more agents are applied
simultaneously or alternately in single doses and repeatedly. "Anti-
carcinogenic agents" are sometimes carcinogenic but more commonly
noncarcinogenic; when they are given in single or multiple doses be-
fore, during, or after treatment with a carcinogen, they partially or
completely inhibit the induction of tumors at the site of application
or at distant sites.
Two-Stage Carcinogenesis
The process of two-stage carcinogenesis as defined above is, as an
experimental model, limited to one test system—i.e., mouse skin784 -
although some studies suggest that it may also occur in other sys-
tems.635 Nevertheless, it merits serious consideration in relation to the
induction of human lung cancer as caused by air pollutants and en-
vironmental exposures to other chemical agents; the example of
cigarette smoke condensate as a tumor-promoting agent in two-stage
carcinogenesis will be given below. A typical two-stage carcinogenesis
experiment on mouse skin is summarized in Table 8-1 with the usual
experimental observations. The agent used in these experiments as a
tumor-promoter is phorbol myristate acetate, a complex tetracyclic
terpenic lipophilic-hydrophilic ester derived from the plant product
croton oil.784'788 The notable features here are the rapid rate of papil-
loma induction, the high multiplicity of papillomas, and the fact that
each agent alone induces very few or no papillomas. Other, less potent
tumor-promoting agents are known, e.g., phenol, anthralin, Tweens,
-------
104 PARTICIPATE POLYCYCLIC ORGANIC MATTER
TABLE 8-1 Two-Stage Tumor Induction on Mouse Skin"
Primary treatment
Single subcaicinogenic dose of initiating agent (7,12-dimethylbenz[a] anthracene,
benzo[a] pyrene, or utethan)
Secondary treatment
Between 3 and 380 days after primary treatment, repeated application of phorbol myris-
tate acetate, 0.5-25 Mg, three times a week
Observations
1. First papillomas 30-70 days from beginning of secondary treatment
2. 50-100% of animals bear papillomas by 250 days
3. Number of papillomas per tumor bearer: 1-12 (some even higher)
4. 40-60% of animals bear squamous carcinomas after a year or more on test
5. Control groups with secondary treatment alone usually develop no.tumors; some
develop late tumors
6. Groups with primary treatment alone usually develop no tumors
7. Few, if any, tumor regressions
"Derived from Van Duuren.784
Spans, and dodecane. The most noteworthy of these is anthralin;
compounds of similar structure may occur as air pollutants.
One of the most important characteristics of two-stage carcino-
genesis is its potentially insidious character. An organ or tissue may
receive a single subcarcinogenic exposure to a carcinogen or one
exposure to a noncarcinogenic initiating agent and later undergo re-
peated low-level exposures to environmental (in this case airborne)
tumor-promoting agents, and the sequence may result in tumors. This
persistence of the initiating effect has been demonstrated several
times in laboratory experiments on mouse skin, as shown in Table 8-2.
It is clear from these experiments that an interval of more than a year
between initiation and promotion resulted in rapid induction of pap-
illomas. It must be remembered that this interval represents for the
mouse approximately half its life-span.
There is at present little or no information, to link the process of
two-stage carcinogenesis (as demonstrated on mouse skin) to cancer
of the lung in man as induced by organic air pollutants. Recent labora-
tory experiments by inhalation in rats have shown that combined
treatment with benzo[a] pyrene and sulfur dioxide induces squamous
carcinoma of the lung; neither agent alone has resulted in these tu-
mors.468 Convincing evidence of two-stage carcinogenesis—even using
the mouse skin model—is not available for air pollutant chemicals,
although it has been clearly demonstrated69'792 for cigarette smoke
-------
TABLE 8-2 Persistence of Initiating Effect"
Primary Treatment
and Route
7, 12-dime thy lbenz[ a] anthracene, skin
7,1 2-dimethylbenz[a] anthracene, 150
Mg, skin
Urethan, 20 mg, subcutaneous
Urethan, 20 mg, intraperitoneal
Interval between
Primary and
Secondary
Treatment, days
301
380
246
246
Secondary
Treatment (skin)
Croton oil, two times a week
Croton resin, 25 Mg, three times a week
Croton resin, 25 Mg, three times a week
Croton resin, 25 Mg, three times a week
Interval from
Beginning of
Secondary
Treatment to
First Papil-
loma, days
_
37
32
42
No. Mice with
Papillomas/
Total No. Mice
9/22
6/11*
7/20
8/20
Duration of
Test, days
420
500
456
456
"Derived from Van Duuren.784
6Two animals with squamous carcinoma.
-------
106 PARTICULATE POLYCYCLIC ORGANIC MATTER
condensate. From these findings, it is clear that cigarette smoke con-
densate is at best a weak tumorigen, but it is a moderately active
tumor-promoting agent. In a recent experiment, 30 of 60 mice exhib-
ited papillomas; eight of the 30 also bore squamous carcinomas of
the skin if the tar was applied after a single subcarcinogenic (50-/zg)
dose of 7,12-dimethylbenz[a] anthracene. When the tar was applied
alone at 50 mg per application, five times a week, only five of 60
mice bore papillomas; no animals in this group bore carcinomas after
390 days on test (B. L. Van Duuren, unpublished data).
Some agents are not by themselves carcinogenic, but result in sub-
stantial tumor yields when followed by repeated exposure to tumor-
promoting agents. Urethan is the classic compound in this category;
it is not carcinogenic for mouse skin (although it does result in lung
adenomas in mice), but it is a potent initiating agent for mouse skin,
whether applied topically or given systemically. Urethan and related
compounds are not known to be air pollutants, but several noncar-
cinogenic initiating agents are possible air pollutants. A number of
known aromatic hydrocarbon air pollutants—previously considered
noncarcinogenic or borderline carcinogens—were recently shown to be
active tumor-initiating agents. Compounds in this class include dibenz-
[a,c] anthracene, chrysene, and benz[a]anthracene.789
Cocarcinogenesis and Anticarcinogenic Agents
The problem of Cocarcinogenesis is difficult to deal with from the
point of view of environmental carcinogenesis, even in, a discussion
of laboratory models. Whenever two or more agents are applied simul-
taneously or sequentially, they may interact in a variety of ways so
as to alter the effect of each other. This modification of each other's
effects results at times in an increase in the biologic effect, such as
tumor induction; at times, it results in a decreased effect, owing to
competitive reactions at sites concerned with biologic activity, or
owing to induction or inhibition of enzymes, which results in the
detoxification of an otherwise carcinogenic agent. The agent |3-naphtho-
flavone has recently been shown to inhibit lung adenoma induction
by benzo[a] pyrene enzymes.807 Several other examples are known.787
Polycyclic aromatic hydrocarbons of weak or mild carcinogenicity
are present in polluted air and cigarette smoke. These have been
shown by Falk et al.261 to inhibit subcutaneous sarcoma production
by benzo[a] pyrene. The significance of this finding is that a given
mixture of POM in polluted air may have a smaller net carcinogenic
-------
In vivo Tests for Carcinogenesis and Cocarcinogenesis 107
effect than would be expected from the action of the potent carcin-
ogens present in the mixture.
Airborne Alkylating Agents
Several reports have discussed the presence of "neutral-nonaromatic"
fractions of air pollutant concentrates,448 but none of the reported
studies has proceeded with the fractionation, isolation, and chemical
identification of these materials. Because olefin hydrocarbons con-
stitute a considerable portion of organic air pollutants, it is reasonable
to expect that their oxidation products, whether spontaneous or photo-
chemically induced, may be present. The main source of these ole-
fins is automobile exhaust. Several schematic pathways have been
proposed, suggesting possible products of olefin oxidation and peroxi-
dation. Laboratory studies on olefin oxidation and ozonization have
resulted in the isolation of epoxides, peroxides, aldehydes, and ke-
tones.485 Some of the products and possibly reactive intermediate
species are shown in Figure 8-1 . It has become apparent, also, that
oxides of nitrogen and sulfur dioxide interact with olefins so that
a great variety of these products can be expected. Some of these com-
pounds are most probably carcinogenic, cocarcinogenic, or both. Ex-
tensive studies have been carried out on the carcinogenic activity of
many epoxides (mono-, di-, and poly-), |3-lactones, hydroperoxides,
and peroxides in a variety of test systems, including skin application
in mice, subcutaneous injection in mice and rats, intraperitoneal in-
jection in mice, intratracheal instillation in rats, and intraga^tric intu-
bation in rats. These studies have made it possible to derive some
structure-carcinogenicity correlations and predictions of the possible
(OZONIZATION)
R
3 COO n
ROCH or HO-C
I
r n
«' \\ »
S 1 VROC
r "
R1R2+CO2 ROM + CO
FIGURE 8-1 Olefin oxidation in air. (After Leighton.485)
-------
108 PARTICULATE POLYCYCLIC ORGANIC MATTER
carcinogenicity of compounds in this series not yet tested by long-
term bioassay.780 The chemical reactivity of these compounds has
been correlated with carcinogenic activity; stereochemical factors,
such as molecular flexibility, and the importance of interatomic dis-
tance as related to carcinogenic activity and possible intracellular tar-
get sites were also considered.
The epoxides, hydroperoxides, and (3-lactones are the most active
carcinogens within this series, but very few peroxides have been tested
for carcinogenic activity, so it is not possible to draw conclusions
about them. It is difficult to visualize the formation of |3-lactones
from olefins; however, some /3-lactones can be formed by spontaneous
dimerization of such pyrolytic products as ketene and its analogues,
as shown in Figure 8-2. The parent compound diketene (/3-methylene-
j3-propiolactone) is inactive as a carcinogen, probably owing to its high
chemical reactivity; but its less reactive saturated analogue, |S-methyl-
0-propiolactone, is carcinogenic.780 Thus, the possibility that |3-lac-
tones are carcinogenic air pollutants cannot be ruled out.
It should be pointed out that, in the extensive network of stations
for the collection of air pollutants, apparently only particulate mat-
ter is collected. It is expected that most of the potentially deleterious
agents just discussed will pass through the particulate-matter filters
in vapor form.
Aromatic Hydrocarbon Oxidation Products
The metabolism of carcinogenic aromatic hydrocarbons has been
extensively studied in a variety of animal species,142 but relatively
little has been done in explorations on the oxidation products, mostly
CH., CH, CH.
\ / II
|| PYROLYSIS ||
O O
ACETONE KETENE
DIKETENE
FIGURE 8-2 Formation of diketene from acetone.
-------
In vivo Tests for Carcinogenesis and Cocarcinogenesis 109
photochemical, of aromatic hydrocarbons in air. This is an important
area, because it has been demonstrated that some phenolic compounds-
e.g., phenol and 1,9-dihydroxy-8-anthrone—are tumor-promoting
agents. Further knowledge of these oxidation products is pertinent also
with respect to the detoxification of carcinogenic aromatic hydrocar-
bons in the air, depending on climatic and meteorologic conditions. In
addition, some of the conceivable oxidation products, such as hydro-
peroxides and peroxides, formed photochemically, may be carcino-
genic even if the parent hydrocarbons are noncarcinogenic. Benz[a]-
anthracene, for example, is a weak carcinogen. It is conceivable that
its hydroperoxide or peroxide, shown in Figure 8-3, is carcinogenic.
However, neither of these compounds is a known air pollutant. A
large variety of phenols and quinones of aromatic hydrocarbons are
expected as photochemical oxidation products. Some of these com-
pounds have been tested for carcinogenic activity.351 The quinones
are usually not carcinogenic, but some of the phenols are weakly
carcinogenic. It has been suggested that aromatic hydrocarbon epox-
ides are proximal carcinogens in aromatic hydrocarbon carcinogene-
sis.780 Some have been synthesized, e.g., 5,6-dihydro-5,6-epoxydibenz-
[a,h] anthracene.563 However, these compounds are highly reactive
and hence unstable, and they have not been isolated as metabolic
products of aromatic hydrocarbons in vivo or from air pollutants. It
is likely that they occur in both.
Conclusions and Recommendations
The polycyclic aromatic hydrocarbons and heterocyclics constitute a
group of known carcinogens that are present in the particulate phase of
polluted air. However, the extent of their contribution to the inci-
Q^O
Peroxide
QO
O
LOOH
Hydroperoxide
FIGURE 8-3 Peroxy (peroxide and hydroperoxide)
compounds of benz[a] anthracene.
-------
110 PARTICULATE POLYCYCLIC ORGANIC MATTER
dence of human lung cancer is unknown. A variety of other agents
probably contribute to the human health hazard. They include tu-
mor-promoting agents, cocarcinogens, noncarcinogenic tumor-initiat-
ing agents, and carcinogens other than aromatic hydrocarbons. The
unknown carcinogens include direct-acting alkylating agents (such as
epoxides and lactones), peroxides and hydroperoxides of olefins, and
aromatic hydrocarbons.
The role played by tumor-inhibiting agents or anticarcinogenic
agents in the health effects of air pollutants is at present poorly un-
derstood. Research is needed on the chemistry and biologic activity of
air pollutant cocarcinogens and tumor-promoting agents, such as poly-
phenols and paraffin hydrocarbons, and on the oxidation products
of airborne olefins and aromatic hydrocarbons, including the nature of
the epoxides, hydroperoxides, peroxides, and lactones formed and
their biologic properties.
EXPOSURE OF THE LUNG TO POLYCYCLIC
HYDROCARBONS
There are relatively few reports of induction of tumors by administra-
tion of carcinogenic polycyclic compounds via the airway. Because
of the attendant difficulty, there have been few studies in which ex-
posure of the respiratory tract has been carried out in a manner rele-
vant to the problem of air pollution.148'449'734
Exposure by Inhalation of Crude Material
Exposure of mice by inhalation of road sweepings, chimney soot,
air dusts, etc., produced pulmonary adenomas. One study was carried
out with sweepings from an asphalt road. Not only did the pulmonary
adenoma incidence rate increase but skin cancer was noted in exposed
animals.108 Other investigators produced increased numbers of pulmo-
nary adenomas with road dust524 and chimney soot.686'687
Mice have been exposed to aerosols of more purified materials con-
sisting of the neutral fraction of coal tars and the same fraction with
the addition of the acidic and phenolic fractions. No tumors occurred
in the C3H control mice; the incidence of both squamous metaplasia
and pulmonary adenomas was significantly greater in the animals that
were given both fractions than in those receiving only the neutral
fraction.760
-------
In vivo Tests for Carcinogenesis and Cocarcinogenesis 111
Inhalation of purified asphalt did not produce lung tumors in guinea
pigs or mice,702 whereas skin painting of similar material led to skin
cancer.703
Exposure to Ozonized Gasoline Vapor
A number of experiments have been performed in which mice were
exposed to "simulated auto smog." This atmosphere was produced
by the passage of nitrogen through leaded gasoline, reaction of the
vapor with ozone, and introduction of the effluent into exposure
chambers.451 Exposure to this atmosphere increased the incidence
of tumors in A strain mice and pulmonary tumors in C57B mice.454
Intratracheal Instillation of Carcinogens
An appreciable incidence of lung tumors was produced by the in-
tratracheal administration of 7,12-dimethylbenz[a] anthracene sus-
pended in a balanced saline solution containing 4% casein and India
ink powder.614 Studies were carried out with other agents, including
benzo[a] pyrene adsorbed on various types of purified carbon parti-
^613,615-618,691,692,694 Jhe method was successful for the prodUC-
tion of malignancies of the lung with 7,12-dimethylbenz[a] anthra-
cene; but with benzo[a] pyrene, 30% of these rats had adenocarci-
nomas and 67% had squamous cell carcinomas.
The adsorption of benzol a] pyrene on hematite (F2O3) particles
has produced tumors successfully when equal parts of both were trit-
urated in a mortar, suspended in saline, and injected via the trachea
into Syrian hamsters.649"656 The carcinogen/hematite/Syrian hamster
model has been successful in inducing a large number of cancers of
the tracheobronchial tree and lung parenchyma, which mimic those
occurring naturally in human beings.654 Furthermore, dosage effects
are apparent from both the quantity administered and the number of
weekly intratracheal instillations. The hamster is remarkably well
suited for this model, in that it js uniquely free of inflammation and
spontaneous tumors of the lung. The hamster model has been extended
to a primate, indicating that the intratracheal instillation method is
effective in producing squamous carcinoma of the lung in another order
of mammal. A cancer incidence as high as 76% was reported with benzo-
[a] pyrene and hematite. Methods using the addition of Tween 60
with benzo[a] pyrene have yielded as many as 50% tracheobronchial
tumors. This method has also been confirmed for a combination of
-------
112 PARTICIPATE POLYCYCLIC ORGANIC MATTER
asbestos and benzo[a] pyrene.366'534 The addition of Tween to the
carcinogen may be objected to on the grounds that Tween is a promot-
ing agent and may be carcinogenic itself.191 Other methods of induc-
tion of lung cancer by intratracheal implantation of pellets impreg-
nated with 3-methylcholanthrene or benzo[a] pyrene have yielded
45 and 30% cancers, respectively.474 Experiments have been carried
out in which pure hydrocarbon carcinogens in colloidal suspension
in gelatin, unaccompanied by other agents, were instilled into the
trachea in hamsters. These studies yielded only metaplasia with benzo-
[a ] pyrene but a very high incidence of poorly differentiated adeno-
carcinomas and other tumors with 3-methylcholanthrene.474
Another promising model was developed by Yasuhira.833 With or
without pretreatment with complete Freund's adjuvant, a carcinogen
(3-methylcholanthrene) was instilled once into the tracheas of Sprague-
Dawley and Wistar rats. More than 66% of the rats developed squamous
cell carcinomas and fibrosarcomas of the lung 50-400 days after treat-
ment.
The successful induction of tumors by the intratracheal injection
of benzo[a] pyrene appears to be related to the addition of some other
physical factor, which in all probability prolongs the residence time
of the carcinogen at the target site. Thus, no tumors resulted from
injection of benzo[a] pyrene alone,474 whereas its adsorption on car-
bon particles,613-618'691'692 hematite,649'656 or asbestos534 produced
a striking carcinogenic effect. It has been postulated that the carcino-
genic effect was made possible by the adherence of the hydrocarbon
to the surface of the particle, from which it was slowly released into
the tissue, thereby achieving a more prolonged and constant action. It
has been shown that the retention of the benzo [a] pyrene in the lung
is proportional to the amount of hematite used, and the rate of elimi-
nation suggests a prolongation of retention owing to the presence of
the inert particle.649
The elimination of benzo [a] pyrene from the lungs appears related
to the size of the carbon particle on which it is adsorbed, with elimi-
nation progressing more slowly from smaller particles.694 In experi-
ments using tritiated benzo[a] pyrene in hamsters, a significant slow-
ing of the clearance of radioactivity from the lung was found after a
14-day period when the benzo[a] pyrene was incorporated on carbon
or asbestos, compared with the clearance after administration of benzo-
[a] pyrene alone. The adsorption on the inert particles also resulted in
a more pronounced and prolonged increase in the number of pulmonary
alveolar macrophages.618 Similar prolongation of retention of the car-
cinogen might also be expected with the intratracheal implantation
-------
In vivo Tests for Carcinogenesis and Cocarcinogenesis 113
of impregnated pellets, although the role of trauma cannot be ruled
out in this method of application of the carcinogen.474
Exposure to a Combination of Carcinogens and Gaseous Pollutants
The study perhaps most relevant to air pollution inhalation used a
combination of sulfur dioxide and benzo[a] pyrene aerosol in rats and
hamsters.474 In these experiments, animals were exposed to 10 ppm
of sulfur dioxide for 6 hr/day, 5 days a week, plus a combination of
10 mg/m3 of benzo[a] pyrene and 3.5 ppm of sulfur dioxide for 1
hr/day, 5 days a week. Appropriate controls were used. No signifi-
cant alterations beyond cellular metaplasia were found in hamsters.
In rats, however, malignancies of the lung were induced. Two of 21
rats receiving only the benzo[a] pyrene-sulfur dioxide mixture for
1 hr/day developed squamous cell tumors of the lung, whereas five
of the 21 rats receiving the same mixture plus the 6-hr/day exposure to
10 ppm of sulfur dioxide developed lung tumors. These data indicate
that an atmosphere containing a benzo[a] pyrene-sulfur dioxide mix-
ture is carcinogenic to the lung of the rat and suggest that additional
exposure to sulfur dioxide causes an increment in such cancers.
Conclusions
Purified polycyclic compounds, such as benzo [a] pyrene, have pro-
duced tumors of the tracheobronchiolar tree or lung parenchyma only
when adsorbed on particles and delivered below the larynx. In inhala-
tion experiments, the addition of an irritant, such as sulfur dioxide,
to an aerosol of benzo[a] pyrene has induced lung carcinomas in rats.
Because many potentially interacting influences-including solid par-
ticles, irritant chemicals, and gases—are ubiquitous in polluted air,
these may be cofactors as important for the induction of pulmonary
cancer as the polycyclic hydrocarbons themselves.
TESTING FOR CARCINOGENICITY IN PRIMATES
Primates as natural hosts for tumors and as experimental animals in
Carcinogenesis assays have the putative advantage of a phylogenetic
relation to man.302 That advantage is tempered by the initial cost
and the expense of maintenance of primates and by their long life-
span, which may require many years of observation for study of a
disease with long latent periods. Data on primates as laboratory ani-
-------
114 PARTICULATE POLYCYCLIC ORGANIC MATTER
mals, breeding in captivity, with natural and acquired infections, and
as subjects for physiologic studies are available.302
Members of the suborder Prosimii appear to be closer to a rodent-
like precursor, whereas the suborder Anthropoidea—which includes
simians (both Old and New World monkeys), the apes, and man—en-
compasses the more highly evolved primates. This notion would raise
the possibility that the suborder Prosimii is not as good a potential
surrogate for man as the suborder Anthropoidea.7'168
Luther504 and Kent435 summarized the world literature on carcino-
genesis in primates to 1960. Kent435 reviewed both spontaneous and
induced neoplasms of simians and concluded that spontaneous tumors
increase with age and closely resemble their human counterparts.
Vadova and Gel'shtein779 observed that the greater proportion of
epithelial than of sarcomatous tumors among simians parallels the
distribution in man. Petrov592 reviewed 23 monkeys that survived 2
years after the injection of radium and 22 monkeys that survived 2
years after injection of carcinogenic hydrocarbons into marrow cavi-
ties of long bones. The first group yielded eight tumors, the second
group, one. All tumors were sarcomas of cartilage, bone, or reticu-
loendothelial cells. Kent435 regarded the production of bone neo-
plasms with ionizing radiation as evidence that monkeys are sus-
ceptible to experimental tumor production.
Skin and Subcutaneous Tissues
Pfeiffer and Allen593 administered methylcholanthrene, dibenzanthra-
cene, and benzo[a] pyrene in oil or saline to adult simians by injection
into subcutaneous tissues and mammary glands, by intravenous injec-
tion, by mouth, by implanting pellets in abdominal and pelvic viscera,
and by painting on skin and cervix uteri. At all sites of injection or
implantation, fibrotic reactions, granulomas, and accompanying
chronic or acute inflammation were produced without proven neo-
plasm. When administration was by skin painting or cervical canal
instillation, papillomatous masses and hyperkeratinization were ob-
served without neoplasms. After multiple exposures to hydrocarbons
by various modes of inoculation and with observations continued for
up to 10 years in some animals, the authors concluded that failure to
produce cancer suggested greater activity of protective mechanisms
than in rodents, and possibly detoxification.
Review of this experiment reveals that very few animals were
exposed repeatedly to the same agent at the same site for periods of
over 2 years. That is significant, in view of the 3-4 years required by
-------
In vivo Tests for Carcinogenesis and Cocarcinogenesis 115
Sugiura et a/.735 to induce squamous carcinomas in three of six rhesus
monkeys with high-boiling, catalytically cracked oil applied by re-
peated skin painting. Moreover, the monkeys used by Pfeiffer and
Allen593 were estimated to be 3-10 years old at the outset of treat-
ment. The work of Pfeiffer and Allen should not be regarded as defini-
tively negative, although the possibility that primates resist the effects
of hydrocarbons better than rodents is worthy of confirmatory test-
ing. Kelly et a/.433 induced hepatic cell carcinomas within 27 months
in macaques and cebus monkeys by giving them JV-nitrosodiethylamine
orally and in cercopithecus monkeys by giving it intraperitoneally.
The macaques were exposed from birth onward; the other species
began exposure before they were a year old.
Levy493 produced a fibrosarcoma in a marmoset 10 months after
the subcutaneous injection of 2 mg of methylcholanthrene. Adamson
et al.7 produced fibrosarcomas in three of six tree shrews and one of
12 galagos inoculated subcutaneously as young adults or as newborns
with benzo[a]pyrene or methylcholanthrene. Noyes reported induc-
tion of fibrosarcoma and rhabdomyosarcoma in a marmoset with ben-
zo[a] pyrene or dimethylbenzanthracene inoculated at different sites
in the same animal565 and fibrosarcomas in three tree shrews inocu-
lated with benzo[a] pyrene.566 The relative frequency of reports of
sarcomagenesis with polycyclic organic hydrocarbons in prosimians,
compared with simians, may reflect a true difference in susceptibility
to tumor induction or a lack of comparability in experimental plans
used in the two suborders.
The foregoing review deals with polycyclic aromatic hydrocarbons
and crude materials known to contain them. Chemical carcinogens of
other classes have been even less widely studied.
Pulmonary Tissues
Pulmonary carcinogenesis in simian primates was reported by Vor-
wald,796 who administered beryllium dust repeatedly to young adult
Macaca mulatto monkeys over a period of more than 5 years. The
earliest lung cancer was observed at 4.5 years after one intramural.
injection followed by numerous intrabronchial instillations of beryl-
lium oxide as a powder suspended in a saline solution. Bronchogenic
neoplasms were also developed after repeated inhalation of beryllium
sulfate aerosol.
Squamous carcinoma of the lung and/or bronchi reportedly oc-
curred in 50% (three of six) of the galagos after weekly tracheal in-
stillation of benzo[a] pyrene and ferric oxide dust. These agents were
-------
116 PARTICULATE POLYCYCLIC ORGANIC MATTER
administered for 67-69 weeks beginning at the time of weaning (12
weeks old).168 Hamsters developed squamous or anaplastic carcino-
mas of lung after exposure to the same carcinogen-particulate prepara-
tion by the same investigators168 following the schedule reported by
Saffiottie/a/.650
Nonneoplastic alveolar lesions in galago and hamster lungs were
histologically similar. This raised the question of whether the galago
(and possibly other prosimians) may resemble rodents in terms of
pathogenetic response.7'168
Conclusions and Recommendations
Experimental carcinogenesis caused by polycyclic aromatic hydrocar-
bons or crude products known to contain them has been achieved in
simian and prosimian primates. Nitrosamines have been shown to be
carcinogenic in simians. Subhuman primates are therefore susceptible
to experimental chemical carcinogenesis.
Not all attempts to produce tumors in simians with polycyclic
aromatic hydrocarbons have succeeded, raising questions of resis-
tance due to metabolic or immune characteristics of the suborder
Anthropoidea and of appropriate choice of age and method of appli-
cation. Primates of the suborder Prosimii appear more susceptible to
carcinogenesis by this class of compounds and have shorter latent
periods.
Pulmonary carcinoma in simians has been produced by particles of
beryllium salt but not by polycyclic aromatic hydrocarbons. Pulmonary
cancer has been produced in rodents and prosimian primates by intra-
tracheal instillation of particulate preparations of polycyclic hydro-
carbons.
Domestically bred simian primates should be tested with methods
identical with those which succeeded in prosimian primates and rodents
to provide data on the relative susceptibility to systemic, skin, and
bronchial carcinogenesis by polycyclic hydrocarbons in the suborder
Anthropoidea, which includes man.
EXPOSURE OF OUTDOOR ZOO ANIMALS TO ATMOSPHERIC
POLLUTANTS
The only available records on which to base an evaluation of the pos-
sible effects of atmospheric pollutants on outdoor zoo animals over
-------
In vivo Tests for Carcinogenesis and Cocarcinogenesis 117
long periods are the necropsy reports of the Philadelphia Zoological
Gardens from November 1901 to December 1969. These were made
available to the Panel by Dr. Herbert L. Ratcliffe, Director Emeritus
of the Penrose Research Laboratory of that institution. He and Dr.
Herbert Fox were the pathologists in charge during the 68-year period.
The data from necropsies on a total of 21,000 animals suggested
a recent increase in the incidence of lung tumors in avian species. There
were 350 malignant neoplasms among 19,000 animals in the necropsy
series, of which 27 were tumors of the lung. Cancer of the lung in
birds, especially ducks and geese, had increased significantly (p<0.05)
in the second half of the period of observation, compared with the
first, although this was not the case in mammals. Birds represented
approximately 60% of the animals that came to necropsy.
These results represent an analysis of raw data and must be inter-
preted with caution, especially because the total numbers of pulmo-
nary neoplasms were small.
The data from the Philadelphia Zoological Gardens provide sugges-
tive evidence of the existence of environmental pulmonary cancer in
birds. A controlled experiment is therefore indicated. It could consist
of the exposure of a flock of a particular species (for example, spar-
rows) to an urban atmosphere that is highly polluted, and in a spe-
cific geographic area where the prevalence of lung cancer in man is
unduly high. The control group would consist of a sheltered flock of
the same origin and age distribution that would breathe purified air,
but would otherwise be subjected to identical dietary and other en-
vironmental conditions. A comparative study of city sparrows with
country sparrows of the same species—i.e., from contrasting states of
atmospheric pollution—would also be satisfactory, although less
well controlled.
-------
Modification of Host Factors
in in vivo Carcinogenesis Tests
HOST IMMUNE STATUS IN CHEMICAL CARCINOGENESIS
The topic of host immune status in chemical carcinogenesis can be
divided into two distinct aspects: the role of immunity or hypersensi-
tivity to a carcinogenic chemical itself and the role of immunity to
cancer cells as a surveillance mechanism.
Hypersensitivity to chemical carcinogens themselves has not been
extensively investigated, and no definite conclusions can be reached
on the basis of current evidence. There is one unconfirmed report to
the effect that small, repeated, subcarcinogenic inoculations of a
polycyclic hydrocarbon may produce resistance to the carcinogenic
effects of larger later doses. Preliminary unpublished data by H. Peck
and co-workers suggest that the carcinogenic action of a hydrocarbon
may be decreased in rats immunized with a hydrocarbon-protein
conjugate. Guinea pigs are relatively resistant to chemical carcinogene-
sis, and this may be related to the ease with which skin hypersensi-
tivity can be induced in them. Unfortunately, this entire area is char^
acterized by contradictory reports and a lack of solid information.
Further investigation is definitely needed.
Immunity to cancer cells themselves and its effect on carcinogene-
118
-------
Modification of Host Factors 119
sis have been extensively investigated. Although much information has
been accumulated, the importance of immunologic surveillance in
chemical oncogenesis is still debated. There is much evidence that such
a mechanism does exist and can modify the course of chemical on-
cogenesis, but there is also evidence that immunosurveillance in
many systems is weak and largely ineffectual.
The evidence of an immunologic surveillance mechanism in chemi-
cal carcinogenesis is of three main types. No one of these types of
evidence is by itself conclusive, but in the aggregate they constitute
a strong case.
1. It is now generally believed that most, and perhaps all, neo-
plasms have tumor-specific antigens that can arouse an immune re-
sponse. The universality of such antigens has become almost a reli-
gious dogma in some circles. It is obvious that a nearly universal im-
munogenicity on the part of tumor cells is prerequisite for effective
immunosurveillance.437
2. The second type of evidence of immunosurveillance is histo-
logic. It is well established that several types of early neoplasms are
characterized by extensive lymphocytic infiltration. Such infiltration,
by analogy with the well-known homograft reaction, can be inter-
preted as evidence of an immune resistance to the developing neo-
plasm; it is particularly remarkable in early squamous cell neoplasms
of the skin and in early malignant melanomas. The kind of breast
carcinoma in the human that is heavily infiltrated with lymphocytes
(medullary carcinoma) has an unusually favorable prognosis.
3. The third type of evidence of immunosurveillance is the
apparent correlation between clinical or experimental conditions of
altered immunocompetence and the incidence of neoplasia. This
correlative evidence can be further subdivided into two broad cate-
gories: experiments of nature and experiments of man. A number of
experiments of nature are known in which deficits in immunologic
reactivity occur. These include various congenital thymic abnormali-
ties and Down's syndrome, as well as that most widespread of all
clinical syndromes, aging. In all these conditions, there appears to
be a decreased ability to mount an effective delayed-hypersensitivity
type of immune response to an antigenic challenge.305'437 An in-
creased likelihood of neoplasia is most strikingly manifested in aging.
One unconfirmed study suggests that people endowed with a hyper-
active immune mechanism may have a low incidence of cancer;273 it
was a retrospective study and showed that allergic disorders—such as
-------
120 PARTICULATE POLYCYCLIC ORGANIC MATTER
eczema, hay fever, and asthma—were less prevalent in the histories
of cancer patients than in a control population. The most widely used
man-made experimental modality to decrease immunocompetence
has been thymectomy in newborn mice. It has been shown in several
studies that neonatal thymectomy renders mice more sensitive to the
action of carcinogens administered later. Induction of sarcomas,
papillomas, and hepatomas has been influenced by this means.437
Early returns suggest that patients undergoing immunosuppressive
therapy for the purpose of kidney transplantation have a higher than
expected incidence of cancer. Immunocompetence, with respect to
lymphocyte-mediated immunity, can sometimes be artificially aug-
mented by the administration of BCG vaccine. This procedure may
lower tumor incidence in both man and mouse. It is probably more
than coincidental that, with perhaps one exception, all the known
chemical carcinogens are profoundly immunodepressive.437 Thus,
skin allografts can be induced to survive in mice treated with carcino-
genic dosages of a polycyclic hydrocarbon carcinogen. There are
claims that dosages too low to be immunosuppressive may still be
carcinogenic; this point needs further investigation.
Although the three types of evidence cited make, in the aggregate,
a strong case for the importance of immunologic surveillance in
chemical carcinogenesis, there are several arguments against this
thesis. These arguments are not decisive, but they do suggest that
immunosurveillance is a relatively unimportant mechanism of homeo-
stasis in many systems. The first of these lines of evidence concerns
the immunogenicity of tumor cells. Obviously, an efficient immuno-
surveillance mechanism would require that most tumor cells be highly
immunogenic. As pointed out, it is now commonly held that all
tumor cells have tumor-distinctive surface antigens and that these
are immunogenic in the host. The evidence is not entirely convincing.
For example, adenomas induced by urethan in mice appear to have
relatively little immunogenicity.603 Baldwin has not been able to
detect evidence of tumor antigens in some of the rat tumors he has
induced with 2-acetylaminofluorene.30 Many spontaneous tumors in
both rats and mice have very little, if any, immunogenicity.29'571
Other examples of at least very weak immunogenicity could be cited.
This relative lack of immunogenicity may or may not reflect a paucity
of tumor-distinctive surface antigens. Perhaps enhancing antibodies
or other masking mechanisms prevent or diminish immunogenicity.
Whatever the mechanism, many and possibly most so-called spontaneous
neoplasms seem to have little functional immunogenic capacity.
-------
Modification of Host Factors 121
Despite this apparent lack of widespread potent immunogenicity,
one could still argue that immunosurveillance is a very effective mech-
anism; indeed, the very lack of immunogenicity in many neoplasms
may suggest its effectiveness. Immunoselection may bring about a
situation in which the tumors that reach a clinically detectable size
represent a small surviving minority of the real total, most of which
were suppressed by an efficient immune reaction. If that is so, neo-
plasms that arise in a host owing to a deficient immunosurveillance
system should usually be highly antigenic and immunogenic. How-
ever, this is not the case, as is shown by studies of tumors arising in
tissue cultures or in diffusion chambers.
Although it has been shown that there is an immunologic surveil-
lance mechanism, some evidence suggests that it is relatively weak as a
defense against the development of neoplasia. Many tumors have
little or no immunogenicity, and this cannot be attributed entirely to
immunoselection. Even if potentially immunogenic, a neoplasm may
fail to immunize the host until late in its course of growth. Immuno-
depression does not regularly result in an increment in the develop-
ment of epithelial neoplasia, which would be highly suggestive of de-
pression of a major and critical defense mechanism against tumor
development. Other types of surveillance having little to do with
specific acquired immunity may also constitute a major defense
against incipient, chemically induced neoplasia.
AGE AND CARCINOGENESIS TESTING
Although many studies have been carried out in newborn and young
adult mice, not many studies have been carried out in aged animals.
There have been even fewer comparative studies on animals of different
ages. The study of chemical carcinogenesis in aged laboratory animals
is complicated by a number of factors: Aged animals are not always
readily available; the life-span is too short for chronic carcinogenicity
studies; and the incidence of spontaneous neoplasms and other disease
states increases with age and interferes with the interpretation of ex-
perimental results. Some studies suggest that aged animals are more
susceptible to skin carcinogenesis,790 whereas others suggest little
difference between young adult and aged animals in tumor induction
by aromatic hydrocarbons.529 In one study, 7,12-dimethylbenz[a]-
anthracene was given by intragastric intubation to rats less than 2
weeks old, 5-8 weeks old, and 26 weeks old. The overall tumor in-
cidences were very similar in the two older groups but considerably
higher in the group treated when they were less than 2 weeks old.
-------
122 PARTICULATE POLYCYCLIC ORGANIC MATTER
In some kinds of carcinogenesis studies, age is vitally important in
determining the results, in that it is related to such factors as hor-
monal status and immune response. It has been shown, with a variety
of carcinogens and routes of administration, that newborn mice
(usually 12-24 hr old) are more susceptible to carcinogenesis than
young adult mice.190 A noteworthy case is liver carcinogenesis using
aromatic hydrocarbons: Adult mice are normally resistant to liver
carcinogenesis, but hepatomas developed in 21 of 25 8-day-old male
suckling mice that were fed 3-methylcholanthrene.439
The importance of hormonal status in chemical carcinogenesis
is apparent from experiments on mammary tumor induction with
7,12-dimethylbenz[a] anthracene in rats 50-55 days old.407 With
rats 75-100 days old in the same experimental protocol, the time to
induction of mammary tumors is much longer and the tumor inci-
dence is lowered.
The effect of age on 3-methylcholanthrene carcinogenesis has been
examined in guinea pigs. The carcinogen was injected subcutaneously,
and the animals were examined for sarcoma induction. The animals
were in four age groups: young, mature, old, and senile. There was
no significant difference among the first three groups, but in the
senile animals there was a 50% reduction in sarcoma induction.
The frequency of metastases from induced tumors also declined with
age.65
In a number of additional studies, newborn mice have been found
to be more susceptible than adults to chemical carcinogenesis, par-
ticularly in organs distal to the site of administration. Klein441 pro-
duced a high incidence of hepatomas in week-old male mice with
2-acetylaminofluorene at a dosage lower than had been reported to
be hepatocarcinogenic. Kelly and O'Gara432 carried out an extensive
study of carcinogenesis after a single subcutaneous injection of 0.06
mg of dibenz[a,b] anthracene and 0.1 mg of 3-methylcholanthrene into
day-old mice; they found a high incidence of a great variety of lung
tumors, fibrosarcomas, leukemias, sebaceous gland adenomas, and
hepatomas that appeared in 8-32 weeks. Toth et a/.755 compared the
oral administration of dimethylnitrosamine in adult and its subcu-
taneous administration in newborn B ALB/c mice; in addition to a
number of other types of tumors, hepatomas were induced only in
the newborn mice.
Epstein et al.241 tested the carcinogenic activity in newborn mice
of air pollutants collected from six cities in the United States. The
mice received 15-25 mg of material subcutaneously. Some toxicity
-------
Modification of Host Factors 123
was observed, and a number of mice developed hepatomas and
pulmonary adenomas. Epstein and Mantel243 found that the herbicide
maleic anhydride produced hepatomas after subcutaneous injection
into newborn mice, and Epstein et a/.234 found similar results with
the fungicide griseofulvin.
The differences in the carcinogenicity of compounds between new-
born and adult mice appear in some cases to be paralleled by differences
in metabolism. Domsky et a/.210 showed that 7,12-dimethylbenz[a]-
anthracene injected subcutaneously in olive oil was retained at the
site of injection for a much longer time in newborn than in adult
mice. The same thing was found by Mirvish et a/.535 when urethan
was injected intraperitoneally.
Experimental chemical carcinogenesis in newborn animals has
been reviewed by Toth,754 who cautions that the issue of the
greater susceptibility of newborn animals is a complex question still
open to controversy.
It appears, then, that newborn mice in many cases are more sus-
ceptible to chemical carcinogenesis, particularly in organs distal to
the site of injection, than are adult mice. Increased sensitivity to
chemical agents, with the requirement of smaller doses, suggests
that greater use of newborn mice for carcinogenicity tests of air pol-
lution fractions would be advisable.
NUTRITION AND CARCINOGENESIS
During the 1930's and 1940's, there was considerable optimism over
the possibility that a solution to the cancer problem might'be achieved
through the science of nutrition. It seemed reasonable that, just as
the normal organism is responsive to its nutritional state, cancer
cells might be found to require specific nutrients. However, extensive
studies in vivo and investigations in cell and organ culture showed
that malignant tissue has no unique nutritional requirement that can
be exploited.741
Several generalizations about nutrition and carcinogenesis seem
valid. As was the case with established cancer, no unique and specific
nutritional factor could be demonstrated whose absence prevented
the formation of tumors. Nor could specific nutritional factors be
found whose presence protected the host against tumor formation.741
Relatively minor inhibition of tumor formation has been achieved
by manipulating the level of known nutritional elements, but none of
-------
124 PARTICIPATE POLYCYCLIC ORGANIC MATTER
these procedures has developed to the point of practical application.741
Current research suggests that ample vitamin A intake imparts some
protection to epithelial tissues against chemical carcinogens, as demon-
strated by the protection afforded by vitamin A in hamsters against
lung cancer produced by instillation of hydrocarbons and other par-
ticles.655
In the case of liver cancer induced by azo dye in rats, a specific
nutritional effect was discovered. Diets low in riboflavin resulted in
the full carcinogenic potential of these dyes. In contrast, diets adequate
in riboflavin facilitated efficient degradation of the azo linkage by
liver enzymes, thereby producing metabolites of the dye incapable of
inducing liver cancer.533
Although the protective effect of adequate riboflavin nutrition on
azo dye carcinogenesis is specific, it serves as a model for possible
nutritional effects on other carcinogens. The existence of dietary
factors specific for the metabolic activation or inactivation of most
classes of carcinogens—especially carcinogenic hydrocarbons—has
been thoroughly investigated, and none has been found. However, in
view of the ubiquity of carcinogens in our environment and the fact
that an adequate nutritional state generally facilitates the detoxification
of foreign molecules, adequate nutrition for all should be an impor-
tant goal in cancer prevention.
A role in cancer formation has been determined for caloric in-
take.741 If food is always present in a cage, it is characteristic for
animals to become obese. Food intake may be reduced to 65 or 75%
of the ad libitum level with generally beneficial results, particularly
as measured by longevity. In addition to fewer deaths from infection
and degenerative diseases in general, the incidence of spontaneous
cancer is lowered. In some experiments, deaths from spontaneous
mammary tumors occurred in two thirds of the fully fed female mice,
whereas none of the calorically restricted mice developed mammary
cancer. Inhibition of formation of every type of spontaneous and
chemically induced tumor that has been tested has been achieved by
caloric restriction. The inhibition appears to be inversely related to the
strength of the carcinogenic stimulus and directly related to the degree
of caloric restriction.
The effect of caloric restriction may also be observed in man.
Evaluation of insurance statistical studies indicates that persons who
are overweight when past middle age are more likely to die of cancer
than are persons of average weight or less.742 Insurance statistics then
support the conclusion that a considerable portion of potential can-
cers in man might be prevented or substantially delayed by avoiding
-------
Modification of Host Factors 125
overeating. It should be emphasized that this conclusion is justified
with respect only to the formation of cancer; there is no evidence
that caloric restriction is a practical way to affect the growth of an
established tumor.
Aryl hydrocarbon hydroxylase activity is not detectable in the
small intestine and lungs of rats fed purified diets but is measurable
in these tissues in animals fed crude diets. Therefore, dietary induc-
tion of enzyme is necessary. There is evidence that some vegetables
may be responsible (L. Wattenberg, personal communication).
RESPIRATORY INFECTION AND PULMONARY
CARCINOGENESIS
Experimental data on the possible role of infection in respiratory
carcinogenesis are extremely meager. There have been only a few at-
tempts to study the problem, the first by Campbell,109 who infected
mice with influenza virus and later exposed them to dust containing
tar. Kotin447 and Nettesheim et al.560 tested the effect of influenza
viral infection oh the incidence of lung tumors in mice chronically
exposed to ozonized gasoline. Leuchtenberger and Leuchtenberger491
and Harris and Negroni344 infected mice with influenza virus and ex-
posed them to cigarette smoke. The effect of influenza viral infec-
tion on the incidence of spontaneous lung tumors in mice was studied
by Steiner and Loosli,723 and Imagawa et a/.410 explored the effects
of influenza virus on urethan-induced lung tumor formation in mice.
The results obtained in these studies do not permit any definite con-
clusions. Except in the experiments of Nettesheim et al., no measures
were taken to exclude spontaneous respiratory infections, and it is
well known that conventional mice are heavily infected with a num-
ber of respiratory agents. The data reported by the various investiga-
tors are contradictory. Campbell found a reduction in the incidence
of tar-induced lung tumors after viral infection, but the number of mice
included in the study was small; Nettesheim et al. found that influenza
virus decreased the incidence of gasoline-fume-induced lung tumors;
and Steiner and Loosli found that the virus decreased the incidence
of spontaneous lung tumors. In contrast, Kotin and Imagawa et al.
found that viral infection increased the incidence of lung tumors in-
duced by gasoline fumes and urethan, respectively. The findings of
the Leuchtenbergers, using tobacco smoke, and Harris and Negroni,
using tobacco smoke and benzo[a] pyrene, are ambiguous. The mouse
lung-tumor system is less than ideal, in that, with few exceptions
-------
126 PARTICIPATE POLYCYCLIC ORGANIC MATTER
(e.g., Kotin et a/.), squamous hyperplasia and alveologenic adenomas
were produced in the mice, and no laboratory has yet developed a
method that will consistently produce tumors topographically and
morphologically similar to those most often seen in man.
However, the inconclusiveness of the available data can at this time
be regarded only as a consequence of insufficient experimental ef-
forts. In fact, the seemingly contradictory results of Kotin and Nettes-
heim et al. suggest that respiratory agents might play an important
role in the pathogenesis of lung cancer. The designs of the experiments
of Nettesheim et al. and Kotin (with ozonized gasoline fumes) were
almost identical, with two exceptions: Kotin used repeated viral in-
fection, whereas the animals of Nettesheim et al. were infected only
once. In the study of Kotin, control animals, as well as animals ex-
posed only to ozonized gasoline, showed a high incidence of pneumo-
nitis, and it is not unreasonable to assume that the mice used for the
adaptation of the three types of viruses used in the study were also
contaminated with various unidentified respiratory agents. Viral and
bacterial agents isolated from commercially raised mice and rats
have caused not only chronic bronchial pneumonia, but also exten-
sive squamous metaplasia of bronchiolar and alveolar epithelium. The
mice used in the studies of Nettesheim et al., however, were derived
from a germfree colony and were kept free of pathogens throughout the
experiment. No squamous metaplasia or squamous cell tumors developed,
and the incidence of pulmonary adenomas and adenocarcinomas was
reduced by influenza viral infection. It is therefore conceivable that
the squamous cell tumors observed by Kotin et al. after viral infec-
tion and smog exposure developed from chronic pulmonary lesions
caused by adventitious microbial agents.
A cocarcinogenic effect of respiratory infection appears to be an
attractive hypothesis. The respiratory system has a number of very
efficient protective devices: a protective layer of mucus, protective
layers of nonproliferating superficial cells, mucociliary clearance, and
alveolar clearance by phagocytes. In addition, the tracheobronchial
tree and the alveoli are lined with an epithelium with a low prolifera-
tive rate, and lymphatic tissue is closely associated with various parts
of the respiratory tract. Acute and chronic respiratory infections dis-
turb the mucus production, the integrity of superficial cell layers,
the ciliary action, the deep-lung clearance by pulmonary macrophages,
the normal regeneration and differentiation of epithelial cells, and the
local immunologic surveillance mechanisms. Respiratory infection
could result in easier penetration of the carcinogen to susceptible
-------
Modification of Host Factors 127
basal cell layers (disruption of mucous blanket and the superficial
cell layers) and in protracted residence and accumulation of carcino-
genic particles in various parts of the respiratory tract (perturbation of
the mucociliary clearance mechanism and the deep-lung clearance).
Cell necrosis and vigorous subsequent regeneration, often with dis-
turbed cellular differentiation (metaplasia), after respiratory infection
could render the target tissues more susceptible to malignant trans-
formation. Increase in cell proliferation has been shown to be an ef-
fective "promoter" in a number of tumor systems.
Respiratory infections may suppress local and systemic immuno-
competence, and thus the immunologic surveillance that would nor-
mally suppress the growth of malignant clones could be rendered
nonfunctional (antigen competition). This hypothesis is supported by
some recent experiments in which the incidence of either spontaneous
or chemically induced tumors was compared in germfree and con-
ventional animals. These experiments suggest that both the type and
the number of tumors developing in experimental animals are af-
fected by the bacteriologic status of the animals. The tumor systems
used in these kinds of studies are the mouse lung tumor103 and liver
tumor,634 leukemia,799 and myeloma.527 Although such data are
open to different interpretations, one reasonable explanation is that
it is the difference in immunologic status of the germfree animal that
is responsible for the lower tumor response.
In summary, because of the lack of sufficiently controlled experi-
mentation, no conclusive experimental evidence is available to support
or refute the hypothesis of a cocarcinogenic effect of respiratory in-
fection. However, because respiratory infections are detrimental to
a number of local and systemic defense systems and have a profound
effect on cell proliferation and differentiation, the hypothesis of co-
carcinogenicity of respiratory infections is very attractive and needs
extensive study.
INTERACTION BETWEEN PHYSICAL AND CHEMICAL
CARCINOGENESIS
Of the several physical factors that have been associated with the
carcinogenic process in vivo, ionizing radiation is the only one on
which there is a significant body of experimental data. Radiation
alone is a potent carcinogen, capable of inducing tumors in nearly all
tissues of most species.121'283 Because radiation produces chemical
-------
128 PARTICULATE POLYCYCLIC ORGANIC MATTER
alterations in cellular DN A that may lead to chromosome damage
and mutational changes, it has been thought to initiate the carcino-
genic process. The interactions between ionizing radiation and chemical
carcinogens have been studied primarily in three tissues—skin, mam-
mary gland, and lung.
A cocarcinogenic effect in skin has been observed when irradiation
has been preceded or followed by topical applications of various chemi-
cal agents, including 3-methylcholanthrene, benzo[a]pyrene, dibenzan-
thracene, croton oil, and cigarette-smoke condensate. In the case of
the first three chemicals, there is no convincing evidence that the effect
is more than additive. In the induction of skin tumors by locally act-
ing carcinogens, however, systemic factors may play an important
role. In one investigation,68 for example, the carcinogenic effect of
topically applied cigarette-smoke condensate was considerably in-
creased by local irradiation of the skin at points some distance from the
site of exposure to the chemical.
Shellabarger697 has investigated the effects of systemically ad-
ministered 3-methylcholanthrene and external x irradiation, singly
or in combination, on mammary carcinogenesis in rats. The agents were
given 10 days apart and in either order. Both 3-methylcholanthrene
and x irradiation given singly led to a significant incidence of tumors,
but the effect of combined administration was clearly additive. A
similar effect of combined treatment had been observed in earlier
studies on the leukemogenic action of these agents in mice.283
The interactions of radiation and chemical agents in respiratory
carcinogenesis are particularly important in air pollution control, con-
sidering the possibility of simultaneous human exposure by inhala-
tion to radioactive particles and chemical agents. A number of problems
have been encountered in attempts to produce pulmonary cancer in
experimental animals by ionizing radiation alone, and the induction of
tumors by inhalation exposure has proved to be especially difficult;
all these studies have recently been reviewed by Bair,27 Many of the
successful experiments have involved the implantation of radioactive
wires or pellets into the lung parenchyma to deliver an intense local
dose of radiation (about 104-106 rads) or the use of microcurie
amounts of radioactivity sufficient to deliver high doses of radiation
to large volumes of lung tissue. Although these studies have generally
shown a clear relation between radiation dose and tumor incidence,
the results have been complicated by the surgical trauma to the lung
or, more important, when large lung volumes were exposed, by the
direct radiation injury of functional lung tissue (including inflam-
-------
Modification of Host Factors 129
matory changes, necrosis, and fibrosis) caused by the high radiation
doses.
In three recent studies, significant numbers of lung cancers were
produced with relatively low doses, in the range to which man might
be exposed. Gross et al.318 produced lung cancers in 43% of rats and
2% of hamsters exposed to 3,000 or 4,000 rads of external chest x
irradiation. Little et a/.,500 using 15 weekly intratracheal injections of
the alpha-emitting radionuclide polonium-210 adsorbed onto hema-
tite particles, produced lung cancers in 48% of hamsters that received
lifetime doses of 225 rads to the whole lungs. Yuile and co-workers834
found primary lung tumors in 3-13% of rats that received whole-lung
doses of 71-538 rads after a single inhalation exposure to polonium-
210 in a sodium chloride aerosol. A continuing study of the importance
of various physical and biologic factors in lung carcinogenesis from
inhaled radioactivity is being carried out by Bair and his colleagues in
several species of animals.27'658 On the basis of their results and the
other available experimental data, these workers concluded that, for
deposited particles, alpha radiation is more carcinogenic than beta or
gamma radiation, that nonuniform irradiation of the lung by radio-
active particles is more carcinogenic than external irradiation, and
that the local radiation doses required for a substantial tumor incidence
must be very high near the radioactive particle or source. Although
these conclusions appear generally valid for the high-incidence portion
of the dose-response curve, the relative importance of "hot-spot"
or point-source irradiation of relatively small tissue volumes—com-
pared with a lower dose more uniformly distributed to all critical
cells in the lung—is not entirely clear at present (the results of Gross
et al.318 are a case in point). The available data suggest, however, that,
in the case of human lung exposure, the most hazardous source of ra-
diation exposure consists of highly active alpha-emitting particles
small enough to be inhaled and deposited in the lower respiratory
tract.
As far as the interactions of radiation and chemical agents in
respiratory carcinogenesis are concerned, the experimental data from
animals are limited to two investigations. Gross and co-workers318
exposed rats and hamsters to 3,000 or 4,000 rads of external radia-
tion to the chest and, beginning 8 weeks later, a series of intratracheal
injections of 7,12-dimethylbenz[a] anthracene. They found that the
lung cancer prevalence in irradiated animals that received injections of
7,12-dimethylbenz[a] anthracene was no higher than in irradiated
animals that did not receive injections. Although the injected dose when
-------
130 PARTICULATE POLYCYCLIC ORGANIC MATTER
administered by itself gave rise to no tumors in the rats, it exhibited
a 6% tumor prevalence in the hamsters. The authors suggest that the
failure of 7,12-dimethylbenz[a] anthracene to increase the yield of
radiation-induced tumors substantially may be because each agent
exerts its primary carcinogenic action on a different tissue within the
lungs. A similar conclusion may be. drawn from the studies by Little
et al.,500 who administered either polonium-210 or benzo[a] pyrene
adsorbed onto hematite carrier particles to hamsters by intratracheal
injection. Either agent given in sufficient dosages led to a high inci-
dence of bronchogenic cancer, but the benzo[a] pyrene-induced tur
mors varied in histologic type and arose primarily from the trachea
and large bronchi, whereas the radiation-induced tumors were mostly
peripheral and were all mixed adenocarcinomas and squamous cell
carcinomas. In other experiments (B. N. Grossman, J. B. Little, and
W. F. O'Toole, unpublished observations), low doses of the two car-
cinogens adsorbed onto the same carrier particles have been adminis-
tered simultaneously (Table 9-1). Although the data are limited, the
effects of the two agents administered together appear to be additive,
rather than synergistic.
Thus, although these two investigations do not show a synergis-
tic effect between radiation and chemical agents in lung cancer, the
lack of an effect may be due to technical factors, such as dosage of the
specific agents, or to the differences between the target tissues for
the carcinogens within the lung. Anatomically, the origin of the car-
cinogenic process initiated by external radiation 8 weeks before may
not be exposed to intratracheally injected 7,12-dimethylbenz[a]-
anthracene. However, owing to their differing chemical natures and
solubilities, the deposition and clearance patterns of benzo[a] pyrene
and polonium-210 after combined intratracheal administration in
hamsters may be very different, and the two agents may thus act
at differing sites in the lung. Another unknown variable is the impor-
TABLE 9-1 Comparison of Two Carcinogens Administered Singly and
Simultaneously
Carcinogen
Polonium-210
Benzo [a] pyrene
Polonium-210
+ benzo [a] pyrene
No. Hamsters
Developing Tumors
30
26
24
Time of Appearance
of First Tumor
40th week
64th week
38th week
Tumor Incidence
(After First Tumor), %
48
9
73
-------
Modification of Host Factors 131
tance of the chronic inflammatory process often associated with high
doses of radiation to the lung. Further experimental work is needed
for the relation between physical and chemical carcinogenesis in the
lung to be established.
A final piece of evidence bearing on this subject comes from the
epidemiologic studies of lung cancer among the Colorado Plateau ura-
nium miners exposed by inhalation to alpha radiation in the air of
the mines.27'265 The data indicate that the very high incidence of
lung cancer in these miners, which is related to the degree of exposure
to radioactivity in the mines, is associated primarily with the group
who are also cigarette smokers. Only two cancers have occurred thus
far among nonsmoking miners in whom 11 would be expected if
cigarette smoking played no role in the excess cancer deaths; the
excess respiratory cancer deaths per 10,000 person-years of observa-
tion were 10 times greater for cigarette smokers than for nonsmok-
ers. Furthermore, among the group of 761 underground uranium
miners who are American Indians (most of whom do not smoke),
the incidence of lung cancer was not significantly increased. These
observations suggest a synergistic effect in man between exposure to
alpha radiation from inhaled radioactive particles and components of
cigarette smoke.
In summary, it has been found that ionizing radiation and polycy-
clic aromatic hydrocarbons combined produce an additive carcino-
genic effect at various sites. It has been difficult to induce pulmonary
tumors with irradiation because of difficulties in delivering the radia-
tion to the lungs. Progress has been made by the use of polonium-210
adsorbed on hematite particles, and a combination of this and benzo-
[a] pyrene (also adsorbed on hematite) delivered by intratracheal in-
stillation produced an additive carcinogenic effect.
-------
10
Distribution, Excretion, and
Metabolism of Polycyclic
Hydrocarbons
DISTRIBUTION AND EXCRETION OF POLYCYCLIC
HYDROCARBONS
Before the use of isotopically labeled carcinogenic hydrocarbons,
fluorescence was applied to the study of the tissue distribution and
excretion of these compounds. One of the earliest such studies was
that of Peacock,588 who injected colloidal suspensions of anthracene,
dibenz[a,h] anthracene, and benzo[a] pyrene intravenously into fowls
and rabbits and found that fluorescent material was rapidly cleared
from the blood and excreted into the bile. Some unidentified fluores-
cent material was also seen in the livers. Doniach et a/.212 found that,
after inoculation of benzo[a] pyrene into mice and rabbits, the fluores-
cence of the kidneys, lungs, and liver changed from the characteris-
tic violet of the hydrocarbon to blue. Blue fluorescence also persisted
for several weeks in the skin after a topical application of the ben/o-
[a] pyrene. Chemical characterization of the blue-fluorescing hydro-
carbons was unsuccessful.
The distribution of radioactivity in mice after administration of
the first sample of dibenz[a,h] anthracene that was ever labeled with
carbon-14 was reported in 1948 by Heidelberger and Jones.354 They
132
-------
Distribution, Excretion, and Metabolism 133
measured the radioactivity in a number of tissues after intravenous
injection of a colloidal suspension of the hydrocarbon and found that
the material was first concentrated in the liver and then excreted via
the bile and intestinal tract into the feces. Similar studies were done
with other routes of administration, and some radioactivity was de-
tected in subcutaneous sarcomas induced in mice by the labeled car-
cinogen. Heidelberger et al.35S carried out various fractionations of
the tissues and excreta from the mice they studied.354 Relatively
small amounts of unchanged hydrocarbon were found in tissues, along
with larger quantities of acidic and phenolic compounds, which un-
doubtedly represented detoxication products.355 Heidelberger
and Weiss356 studied the rates of disappearance of radioactivity from
the sites of subcutaneous injection in tricaprylin of three labeled hy-
drocarbons; the disappearance rates increased in this order: dibenz-
[a,h] anthracene (half-life, 12 weeks), methylcholanthrene (half-life,
3V4 weeks), and benzo[a]pyrene (half-life, 1% weeks). The relative
carcinogenic activities of these three compounds at low doses injected
subcutaneously were directly proportional to the durations of their
retention at the site.356
Kotin etal.*52 studied the elimination of [14C]benzo[a]pyrene
from rats and mice and confirmed the biliary excretion seen in the
earlier cases. Only radioactivity was measured and no attempt was
made to fractionate the tissues. The solvent used for subcutaneous
injection was found to affect the retention of the radioactivity at
the site. Similar excretory patterns were demonstrated after intra-
tracheal instillation in rats.452
The biliary excretion in rats of hydroxylated derivatives of benzo-
[a] pyrene was decreased after administration of piperonylbutoxide,
a methylenedioxyphenyl derivative that is a potent inhibitor of
microsomal enzyme mixed-function oxidases.257
The distribution of radioactivity in rats after administration by
stomach tube of dibenz[a,h] anthracene, methylcholanthrene, and
7,12-dimethylbenz[a] anthracene was comprehensively studied by
Daniel et al.179 They also found biliary excretion into the feces and
a rather prolonged retention of the radioactivity in body fat, ovaries,
and adrenals. Shabad has reviewed in English some of the spectro-
fluorescence research in the Soviet Union on the distribution of car-
cinogenic hydrocarbons; elimination from the skin and elimination
after intratracheal instillation have been measured.693 Dontenwill
et al.213 have carried out a thorough study of the elimination of ben-
zo[a] pyrene after intratracheal instillation into Syrian hamsters. The
-------
134 PARTICULATE POLYCYCLIC ORGANIC MATTER
rate of resorption from the lungs was measured when the compound
was dissolved in solvents or adsorbed on hematite.
It is evident that no definitive study on the metabolism, tissue dis-
tribution, and excretion of carcinogenic hydrocarbons has yet been
carried out. Measurement only of radioactivity of a compound yields
no information on the chemical nature of the radioactive material.
Fractionation of the radioactive material has been carried out, but
no characterization of the compounds has been achieved. Studies
aimed at the complete characterization of metabolites and excretion
products, including tissue distribution and binding to macromolecules,
have not yet been attempted but should be carried out.
METABOLISM OF POLYCYCLIC AROMATIC
HYDROCARBONS
The metabolism of carcinogenic and noncarcinogenic hydrocarbons has
been investigated for a number of years in intact animals, in various
types of cells in culture, and in cell-free systems. The original studies
were done with fluorescence spectra as the primary analytic tool, but
recently the use of radioactive labeled hydrocarbons has provided
much more precise quantitative data, particularly when combined
with thin-layer chromatography. It has been known for a long time
that simple aromatic compounds are metabolically hydroxylated to
phenols, which are then conjugated with glucuronic acid or sulfate
to yield water-soluble, easily excreted products.
Although mouse skin is the tissue most commonly used in testing
for carcinogenic activity, little work has been done on the metabo-
lism of hydrocarbons in mouse skin. Heidelberger et al.352 studied
the metabolic conversions of dibenz[a,h] anthracene injected into
mice and identified several quinonoid metabolites. Boyland and
Sims83 investigated the metabolism of benz[a] anthracene given in-
traperitoneally to rats, mice, and rabbits; they identified several
phenols and dihydrodiols and their conjugates. Later, they studied
the metabolism of a number of poly cyclic hydrocarbons in rat liver
slices and homogenates. Although these carcinogens do not normally
induce tumors in liver, that organ is convenient for metabolic studies.
The liver microsomal system will be considered in detail below.
Several more recent studies have been carried out with cells in
culture. Diamond et al.199 and Duncan et a/.220 have measured the
conversion of polycyclic aromatic compounds to unidentified water-
-------
Distribution, Excretion, and Metabolism
135
soluble metabolites in various types of cells. Sims707 has character-
ized a number of metabolites produced from several hydrocarbons
in mouse embryo cells and has found them to be similar to those pre-
viously identified in liver homogenates. In no case, however, have
all the metabolites of any hydrocarbon yet been identified.
An example of what is known about the metabolism of an impor-
tant carcinogenic hydrocarbon is shown in Figure 10-1, derived
largely from the work of Sims.706 Benzo[a] pyrene gives rise to the
following metabolites: the monophenols (3- and 6-hydroxy-), the
H OH
9',10-dihydroxy-
9',10-dihydrobenzo[a] pyrene
1,2-dihydroxy- QH
1,2-dihydrobenzo [a] pyrene 6-hydroxybenzo [a] pyrene
3-hydroxybenzo[a] pyrene
1.6-dihydroxybenzo[a] pyrene
0
benzo[a] pyrene-1,6-dione
FIGURE 10-1 Metabolism of benzol a] pyrene.
benzo[a] pyrene-3,6-dione
-------
136 PARTICULATE POLYCYCLIC ORGANIC MATTER
diphenols (3,6- and 1,6-dihydroxy-), the corresponding quinones
(3,6-dione and 1,6-dione), and two dihydrodiols (1,2- and 9', 10-).
None of these metabolites is carcinogenic. A number of unidenti-
fied conjugated products were also present. It is of interest that the
so-called K regions (4,5 and 11,12 bonds), which the Pullmans have
postulated to be the critically reactive sites610 and are most reactive
toward some chemical reagents, appear to be metabolically inert;
Sims707 found no evidence of the presence of either the 4,5- or the
11,12-dihydrodiol. The details of the metabolism of other polycyclic
hydrocarbons are beyond the scope of this document.
In the case of almost every chemical carcinogen that has been prop-
erly examined, there is evidence of covalent binding of some deriva-
tive of the carcinogen to DNA, RN A, and protein. This work has
been thoroughly reviewed by Miller.532 In the case of polycyclic
carcinogens, the covalent binding to mouse skin DNA and RNA
has been studied by Brookes and Lawley,93 Goshman and Heidel-
berger,307 and Brookes and Heidelberger.92 In contrast with the
aromatic amines, the chemical nature of whose binding to nucleic
acids and proteins Miller has established,532 the chemistry of bind-
ing of hydrocarbons to nucleic acid is not yet known, but it is under
active investigation. The binding of aromatic hydrocarbons to pro-
teins has been studied most extensively by Abell and Heidelberger4
and by Tasseron et al.744 These authors have identified and partially
purified a soluble protein fraction from mouse skin to which the
polycyclic hydrocarbons are bound in direct proportion to their car-
cinogenic activities. The function of this protein fraction is being
investigated. It is not now known which, if any, of these macromole-
cules is the primary cellular target of carcinogenic action.
A number of carcinogens are chemically reactive, including alkyl-
ating agents, aliphatic epoxides, lactones, and methylnitrosoureas.
These compounds react directly with nucleophilic sites in D N A,
RNA, proteins, and probably other molecules; the chemistry of the
interactions has been reviewed by Miller.532 By contrast, many of
the more, important carcinogens, including polycyclic aromatic hydro-
carbons, are not chemically reactive. Therefore, it seems highly prob-
able that they must be converted metabolically into chemically reac-
tive compounds that are capable of reacting directly with the
macromolecules of the cell. This process, referred to as "metabolic
activation," is likely to be obligatory if polycyclic hydrocarbons are
to induce cancer.
Three types of metabolic intermediates have been proposed as the
active metabolic form of carcinogenic hydrocarbons. Boyland and
-------
Distribution, Excretion, and Metabolism 137
Sims in 196483 proposed that epoxides are formed from hydrocarbons
and are the reactive form. It is known that dihydrodiols, such as have
been identified as benzo[a] pyrene metabolites (Figure 10-1), can
be formed from epoxides. Thus, it seems likely that epoxides can be
visualized as the metabolic precursors of dihydrodiols and probably
also of phenols. Very recently, Grover and Sims321 have shown that
the K-region epoxides of phenanthrene and dibenz[a,h] anthracene
can react covalently in the test tube with DNA and histones. This,
plus other evidence that is accumulating—largely from work in cell-
free systems—suggests strongly that epoxides are the activated form
of carcinogenic hydrocarbons.
Dipple et al.203 have proposed, alternatively, that the active form
is a cation resulting from attack on an electronically delocalized part
of the molecule, whereas Wilk and Girke818 suggest that the active
metabolite may be a radical cation. Clearly, more work is needed to
establish firmly the nature of the ultimate carcinogenic form of poly-
cyclic hydrocarbons.
This question is of more than purely academic interest. It is now
almost generally accepted that polycyclic aromatic hydrocarbons must
be metabolically activated if they are to induce cancer. Hence, sur-
veys of human tissues for their ability to carry out this reaction may
be of practical value in assessing the contribution of this class of
compound, found so extensively in air pollution, to human cancer.
The enzymic nature of this conversion has been studied most fruit-
fully in the rat liver microsomal system. It has been possible to alter
the in vivo metabolism and carcinogenicity of polycyclic aromatic
hydrocarbons by treating the animals with various agents. The alter-
ation is most likely due to effects on the microsomal enzyme sys-
tem to be discussed below.
An interesting effect of metabolism on carcinogenicity and tox-
icity was discovered by Boyland et al.85 It had been known that
7,12-dimethylbenz[a] anthracene, in addition to being highly car-
cinogenic, also produced severe adrenal necrosis. It had also been
known that it was metabolized to two hydroxymethyl derivatives:
7-hydroxymethyl-12-methylbenz[ a] anthracene and 12-hydroxy-
methyl-7-methylbenz[a] anthracene. Although the former derivative
was much less carcinogenic than 7,12-dimethylbenz [a] anthracene,
it produced adrenal necrosis at a much lower dose; the latter deriva-
tive was inactive in both respects.
It has been demonstrated that K-region epoxides of several poly-
cyclic hydrocarbons are much more active than the parent hydro-
carbon and the corresponding dihydrodiols and phenols in produc-
-------
138 PARTICULATE POLYCYCLIC ORGANIC MATTER
ing malignant transformation in vitro in hamster embryonic and
mouse prostate cells.322 This constitutes strong evidence that epoxides
are the metabolically activated ultimate carcinogenic form of the
polycyclic hydrocarbons. It has also been suggested, on the basis of
nuclear magnetic resonance considerations, that activation of benzo-
[a] pyrene could theoretically occur at carbon-6.124
ARYL HYDROCARBON HYDROXYLASE: AN ENZYME
SYSTEM
The primary enzyme system responsible for the metabolism of poly-
cyclic hydrocarbons is a multicomponent complex generally localized
in the microsomal fraction of the cell.157'292'297 This enzyme sys-
tem has been called "benzpyrene hydroxylase" but is more suitably
named "aryl hydrocarbon hydroxylase," in that it metabolizes a
variety of polycyclic hydrocarbons. It also metabolizes steroids and
chemicals of exogenous origin, such as drugs, pesticides, and preser-
vatives.
The enzyme system contains reduced nicotinamide adenine dinu-
cleotide phosphate-cytochrome C reductase, cytochrome B5, cyto-
chrome P4SO, cytochrome P4SO reductase, and other unknown com-
ponents. It converts polycyclic hydrocarbons to epoxides,416 phenols,
dihydrbdiols, and quinones.84'706 Some phenols and dihydrodiols
are conjugated to yield glucuronides or sulfates by conjugating en-
zymes. Little is known about the mechanism of enzyme action. The
enzyme system is found in about 90% of the tissues of the monkey,
mouse, hamster, and rat.
An important feature of the system is its inducibility. Enzyme
level depends on exposure to polycyclic hydrocarbons, a variety of
such drugs as phenobarbital, various steroids, flavones, and nutritional
and hormonal conditions.1S7'292'297'557 Polycyclic hydrocarbons in-
duce the hydroxylase in the liver, lung, gastrointestinal tract, kidney,
and skin. The level and inducibility are genetically determined and
vary in different strains of mice, the hepatic enzyme being inducible
in the Swiss, C57, c3H, and A strains, but not in AKR/N or DBA
strains (according to Nebert and Gelboin557 and S. H. Yuspa et al,
unpublished data). The enzyme system is also induced transplacen-
tally in hamsters and rats: Exposure of the mother to polycyclic hy-
drocarbons increases enzyme concentration in fetal tissues and pla-
centa. There is a high correlation between the placental content and
-------
Distribution, Excretion, and Metabolism 139
the smoking habits of pregnant women.S59>813 The enzyme is also
present and inducible in cells grown in culture derived from whole
embryos of hamster, rat, and mouse; in cells from individual ham-
ster fetal tissues, such as liver, lung, small intestine, and kidney; and
in several cell lines, such as.mouse 3T3, HeLa cells, and mouse epi-
dermal cells. The enzyme is absent or very sparse in most established
cell lines. The mechanism of induction and various characteristics of
the enzyme system—such as cofactor requirements, kinetic behavior,
and half-life-have been studied.555'5S6'558
The microsomal enzyme catalyzes reactions in vitro that cause a
covalent binding between benzopyrene and DN A or protein.291-320
In cell culture, the enzyme converts polycyclic hydrocarbons to toxic
products.293 The level of enzyme activity correlates positively with
the susceptibility of the cells to the cytotoxicity of benzo[a] pyrene,
suggesting that the enzyme converts benzo[a] pyrene to cytotoxic
metabolites. One of the metabolites of benzo[a] pyrene, 3-hydroxy-
benzo[a] pyrene, is cytotoxic to cells that are either susceptible or
resistant to benzo[a] pyrene.293 A compound, 7,8-benzoflavone, that
inhibits the metabolism of benzo[a] pyrene and 7,12-dimethylbenz-
[a] anthracene in hamster embryo cell cultures, liver microsomes,
and skin homogenates also protects the cells against the toxic effects
of these carcinogens.198 This inhibitor markedly reduces mouse skin
tumorigenesis caused by repeated treatment with 7,12-dimethyl-
benzfa] anthracene or by a single treatment followed by weekly ad-
ministration of croton oil.294'807 These findings indicate that this
enzyme system is responsible for the activation of 7,12-dimethyl-
benz[a] anthracene to its carcinogenic form.293 These results with
the inhibitor of 7,12-dimethylbenz[a] anthracene tumorigenesis
may be related to an activation step involving hydroxymethyl forma-
tion, rather than ring hydroxylation. Thus, it is possible that inhi-
bition of some types of hydroxylation of the polycyclic hydrocar-
bons results in an increased tumorigenicity, whereas an increase in
other types of hydroxylation, perhaps ring hydroxylation at some
positions, may result in reduced tumorigenicity.294
Tumorigenesis is markedly influenced by preinduction of the enr
zyme. For example, Huggins et a/.405 found that pretreatment of
rats with small amounts of polycyclic hydrocarbons decreases 7,12-
dimethylbenz[a] anthracene-induced tumor formation in the mammary
gland. Wattenberg and Leong808 reported similarly that the inducer
5,6-benzoflavone inhibits 7,12-dimethylbenz [a] anthracene-induced
tumorigenesis in the lung and mammary gland of rodents. The pro-
-------
140 PARTICULATE POLYCYCLIC ORGANIC MATTER
tective effects of pretreatment with inducers may be due to increased
enzyme contents in the target tissue, which may more rapidly elimi-
nate the carcinogen or convert it to a less carcinogenic form. Induc-
ers also cause an increase in enzyme content in the liver, the major
site of metabolism, and this may lower the concentration of the
carcinogen in the target tissue.
The enzyme plays a central role in polycyclic hydrocarbon car-
cinogenesis and, because it is found in most tissues and its activity
is affected by environmental conditions, the nature of its role in
hydrocarbon metabolism and activation needs to be clarified. The
enzyme system converts hydrocarbons to products that are largely
noncarcinogenic, but in the process of conversion it may be.respon-
sible for carcinogen activation. The most difficult and important
question to answer is that of the relation between the possible enzy-
mic activation of the carcinogen and its enzymic detoxification to
inactive products. The role of related enzymes, such as the epoxide
hydrase and the conjugating enzymes, also needs clarification. Al-
though there is evidence that polycyclic hydrocarbons require acti-
vation if they are to be active, it is not certain whether the micro-
somal enzyme system is responsible for the activation or whether all
or only some of the carcinogenic hydrocarbons require activation. In
addition to inducers of the enzyme, there are compounds, like 7,8-
benzoflavone, that inhibit the enzyme.
Thus, it seems that the tools are at hand to examine more closely
the role of the enzyme in polycyclic hydrocarbon tumorigenesis.
When this role is clearly illustrated, a most important second step
will be to clarify the nature of the environmental factors—such as
pesticides, drugs, nutritional level, and hormonal states—that en-
hance the detoxification activity of the enzyme system, reduce the
activation of the carcinogen, or both. Recently, an epoxide has been
isolated as an intermediate in the microsomal hydroxylation of
dibenz[a,h] anthracene.20'690
In summary, the metabolism of polycyclic aromatic hydrocarbons
has been studied in various tissues of test animals, in cells in culture,
and in cell-free systems. The following types of metabolites have
been characterized in vivo: phenols, dihydrodiols, quinones, and
various water-soluble conjugated products. This metabolism is car-
ried out primarily by the drug-metabolizing enzyme system of the
microsomes. This complex system is found in many tissues of many
species and is inducible by polycyclic hydrocarbons and other types
of compounds, such as pesticides and drugs. The enzyme system can
-------
Distribution, Excretion, and Metabolism 141
also be inhibited by several compounds. It can increase or de-
crease the toxicity of hydrocarbons and is probably responsible for
their metabolic activation to chemically reactive carcinogens.293
Assay of this enzyme system in various human cells and tissues could
provide valuable information on their susceptibility or resistance to
hydrocarbon carcinogenesis. There is a need to determine the com-
plete metabolic profile of several carcinogens in various tissues.
There is also a great need for a reliable source of purified metabolites
and potentially reactive intermediates, as well as specific inducers
and inhibitors of the enzyme system.
-------
11
In vitro Approaches to
Carcinogenesis
CHEMICAL CARCINOGENESIS
It has been known for many years that normal cells in culture can
be transformed into cancer cells by a number of DNA and RNA
oncogenic viruses. The first report of chemical transformation
in vitro was by Berwald and Sachs in 1963.52 They described the
production with polycyclic hydrocarbons of characteristic mor-
phologic alterations in the growth patterns of hamster embryonic
fibroblasts in culture. This alteration involved a disorientation of
the arrangement of the cells and the formation of criss-crossed
piled-up colonies. The hydrocarbons were more toxic to the normal
fibroblasts than to transformed cells. Berwald and Sachs later dem-
onstrated that mass cultures of the morphologically transformed cells
gave rise to tumors in hamster cheek pouches, whereas the control
hamster cells did not.53 Quantitative cloning techniques were devel-
oped, and it was found that there was a proportionality between the
number of transformed clones produced by a given hydrocarbon and
its carcinogenic activity. Huberman and Sachs,398 using benzo[a]-
pyrene, developed quantitative dose-response curves, which were
interpreted as indicating that transformation was a one-hit process
and was separate from the toxicity exerted by the compounds.
142
-------
In vitro Approaches to Carcinogenesis 143
DiPaolo et al.201 confirmed the results of the Sachs group and
also demonstrated that individual clones of transformed cells could
give rise to tumors (all fibrosarcomas) in hamsters.202 Kuroki and
Sato467 have obtained malignant transformation in cultures of em-
bryonic hamster fibroblasts with 4-nitroquinoline-Ar-oxide and its
derivatives.
It is well known659 that embryonic mouse fibroblasts readily un-
dergo "spontaneous" malignant transformation in culture, as Ber-
wald and Sachs also found. However, spontaneous transformation
in the embryonic hamster fibroblasts was not observed. In spite of
this, it seemed desirable for various reasons to try to develop another
system for chemical carcinogenesis in vitro. Lasnitzki475 found that
polycyclic hydrocarbons produced profound histologic alterations in
organ cultures of mouse ventral prostate. Roller and Heidelberger638
obtained even more striking morphologic changes suggestive of ma-
lignancy; but, on inoculation of these altered pieces of prostate into
isologous mice in a variety of conditions, no tumors were produced.
Thus, carcinogenesis in vitro had not been accomplished.
Although the organ cultures of mouse ventral prostate treated
with carcinogenic hydrocarbons were not malignant, Heidelberger
and lype353 succeeded in obtaining from them lines of cells that
grew in a disoriented fashion and produced tumors on inoculation
into isologous mice. This indication of successful carcinogenesis
in vitro led to an effort to cultivate the mouse prostate cells directly
in cell culture. Chen and Heidelberger131 succeeded in obtaining a
number of cell lines from dispersed untreated organ cultures of adult
C3H mouse ventral prostates. Unlike most other mouse cell lines,
these only rarely underwent spontaneous transformation. These
cells grew exponentially until they reached a monolayer, at which
point no further growth occurred. When 107 of these control cells
were injected subcutaneously into irradiated C3H mice, no tumors
were obtained. When these cells were treated in culture with 3-methyl-
cholanthrene, their growth rate increased, they did not reach a satura-
tion density, they piled up in a disordered array, and they produced a
100% incidence of fibrosarcomas on subcutaneous injection of 1,000
cells into nonirradiated C3H mice.132
These results led to the development of a quantitative system with
the prostate cells in which transformation and toxicity were both
measured in sparsely plated cells; the cells gave rise to individual
colonies that coalesced into a monolayer and then formed individual
piled-up colonies that were very easy to score.133 It was shown that
each piled-up colony was capable of yielding tumors in mice, thus
-------
144 PARTICULATE POLYCYCLIC ORGANIC MATTER
fully justifying the scoring of these morphologically transformed
colonies as truly malignant. With this quantitative system, it was
shown that there was an excellent correlation between the number
of piled-up, malignant, transformed colonies and the carcinogenic
activities of nine polycyclic hydrocarbons. Furthermore, there was
no dose-response relation between the processes of transformation
and toxicity. It appears that systems like these might be used to
screen for potentially carcinogenic activities in fractions of all sorts
obtained from air pollution.133
This system of carcinogenesis in vitro with hydrocarbons has al-
ready furnished considerable information on fundamental cellular
mechanisms of hydrocarbon carcinogenesis. As mentioned before, it
has been possible to produce 100% transformation of clones derived
from single prostate cells, thereby proving (at least in this system)
that the carcinogen directly transforms nonmalignant into malignant
cells and eliminating the possibility that the hydrocarbon selects for
pre-existing malignant cells.537 Such experiments would not be pos-
sible in whole animals.
It has been demonstrated by Mondal et al.*38 that individual
transformed clones obtained from the same dishes have surface trans-
plantation antigens that are non-cross-reactive, as. is the case with
hydrocarbon-induced sarcomas in vivo. This provides added confidence
that this system is a valid model for polycyclic hydrocarbon carcino-
genesis.
In vitro systems have been used productively by Inbar and Sachs411
to detect differences in the surface properties of normal and trans-
formed cells by the use of Concanavalin A—a protein that binds to
glycoproteins on cell surfaces. When it is used quantitatively during
the process of carcinogenesis in vitro in embryonic hamster embryo
cells, there is an alteration in the topology of the membrane such
that masked binding sites are released.
It has been found in many of the studies alluded to here that poly-
cyclic hydrocarbons exert more toxicity to normal than to trans-
formed cells.S2'53>133'353'398 Diamond has also shown that this is
true in rodent cells, whether transformed by chemicals or by viruses.I97
Although it does not seem to be true in the case of primate cells in
cell culture,197 monkey and human respiratory epithelium and mu-
cosal connective tissues in organ culture are as sensitive as rodent
tissues to toxicity and induction of metaplasia by polycyclic aromatic
hydrocarbons.167
The embryonic fibroblast and the prostate cell systems for in
vitro chemical carcinogenesis have interesting differences and similar-
-------
In vitro Approaches to Carcinogenesis 145
ities and complement one another. Inasmuch as both systems are
transformed by carcinogenic hydrocarbons, they must contain the
enzymes to activate the hydrocarbons. Therefore, intensive studies
are now under way in several laboratories to study the metabolism
of hydrocarbons and their binding to macromolecules in direct re-
lation to the process of malignant transformation. It is clear that
these systems will be used very fruitfully in the future for the inves-
tigation of many theoretical and practical problems in chemical car-
cinogenesis that cannot be studied in whole animals.
ORGAN CULTURE
Easty225 describes the organ culture approach to studying carcino-
genesis, its applications thus far in viral and chemical carcinogenesis,
and its uses in studying cell differentiation and organogenesis. Other
reviews of the organ culture system in a variety of applications are
available in Willmer's three-volume text.820 Moscona, Trowell, and
Willmer,547 Grobstein,315 Wolff,826 Lasnitzki,477 Fell and Rinal-
dini,266 Bang,31 and Rapp and Melnick623 discuss special applica-
tions of organ and cell cultures.
Organ culture is a technique for maintaining organs or organ pieces
in vitro so as to retain normal histologic associations among cell types
and to preserve cell differentiation and growth at rates resembling
those in vivo. Adult organ pieces can be maintained for 2-4 weeks.
The cells of an organ in culture are regarded as responsive to test mate-
rials in the same way as cells of that organ would be in vivo, but ade-
quate proof that this is so is not commonly obtained. As a useful ref-
erence point, histologic, histochemical, and cytologic alterations in
organ culture explants can be compared with the alterations produced
in vivo after comparable times of exposure to test materials.
Dose-response analysis of the relative biologic activity of test mate-
rials can be undertaken in organ culture, with better control of con-
centration and duration of exposure than in vivo. Human tissues can
be exposed directly to compounds that could not be safely given to
living human subjects, thus permitting assessment of human tissue
reactions.
Metabolism of native and environmental chemicals by a target
organ can be observed. Metabolic products can be measured in culture
medium or in homogenates of explanted tissue. Penetration of sub-
strate and binding of substrate or its metabolites can be recorded by
cell type by using autoradiography with radioactive substrates.
-------
146 PARTICIPATE POLYCYCLIC ORGANIC MATTER
Most applications of this system to study the biologic effects of
atmospheric pollutants have been made with crude materials or pure
compounds already studied in animals. Results in organ culture
have demonstrated the parallelism between organ responses in vivo
and in vitro.For example, respiratory170'480 and prostatic482'638
tissues undergo epithelial metaplasia, pleomorphism, or devitaliza-
tion in approximate proportion to the toxic or carcinogenic activity
of polycyclic organic air pollutant materials, as established in ani-
mals 186,241,404,633,662 Environmental materials that have been tested in
organ culture include cigarette smoke and cigarette-smoke conden-
sates,476'478'479 benzene-soluble organic materials extracted from par-
ticles trapped by air filters, fractions of these extracts, and a number
of carcinogenic, weakly carcinogenic, or noncarcinogenic polycyclic
aromatic hydrocarbons.169'204
Carcinogenic polycyclic aromatic hydrocarbons applied to such
target tissues as the mouse prostate482'638 or rodent airways170'479'
481,574 produce similar epithelial metaplasias in vivo and in organ
culture. In further examples, neoplastic lesions that followed expo-
sure to benzo[a] pyrene were inhibited by vitamin A in living ani-
mals138' 65S and in organ culture,171 whereas similar inhibition of
prostatic lesions produced by methylcholanthrene occurred in vitro
with vitamin A.482 This series of studies established the comparabil-
ity of in vivo and in vitro responses and demonstrated the value of
integration of in vitro and in vivo methods in defining factors in lung
and prostate carcinogenesis.
The value of the organ culture system in studies of polycyclic
aromatic air pollutants lies in the production of histologically typical
early lesions by direct action of the test agent. Such lesions indicate
that the agents act proximately; hence, enzymes or other cellular
components necessary for metabolic activation of polycyclic aro-
matic materials are present in the target tissue in vitro and therefore
in vivo. The toxic hazard of such pollutants can be assessed, and
potential carcinogenic hazard can be regarded as worthy of study.
The production of lesions permits dose-response estimates of early
tissue responses in man and lower animals by criteria of cell physiol-
ogy and histopathology, in which abnormalities more subtle than cell
death can be identified. Moreover, the organ culture system may
offer the only direct means of comparing human and animal target
tissue responses to toxic or carcinogenic POM.
Evidence of neoplastic transformation of epithelia has not been
tested often in organ culture with chemical carcinogens. Failure in
-------
In vitro Approaches to Carcinogenesis 147
tests so far attempted may have resulted frorri the use of unfavorable
methods. In any event, failure has led to reduced confidence in this
system for in vitro study of epithelial carcinogenesis.148'256
The value of in vitro production of neoplastic transformation or
other cellular changes in detecting environmental health hazards or
in setting air quality criteria cannot be stated definitively, but no
single method of testing is perfect. Animal tests are time-consuming
and require large amounts of material. In vitro methods take less
time and material and can give evidence of toxicity, carcinogenicity,
and mutagenesis, but they may be too sensitive. For the present,
in vivo and in vitro methods need to be compared as to their relative
utility.
Cultured cells have been used as "first-line" preliminary screening
devices for detecting biologic activity and toxicity of air pollu-
tants.643'644 The highly developed cell culture methods for identifying
neoplastic transformation of fibroblasts52'53'201'202'353'398'467'537
by carcinogenic chemicals already place this in vitro system in a posi-
tion to detect carcinogenic potential of air pollutants. Animal tissues
in vitro have been compared with tissues in intact animals in compa-
rable conditions of acute exposure169 and provide evidence that
in vitro tests give close short-term parallels with in vivo reactions.
RECOMMENDATIONS
Animals and cultures should be compared by use of one exposure
system for evaluation of biologic response to atmospheric pollutants.
If in vitro methods are tested in parallel with animal models, deci-
sions as to usefulness of in vitro methods can be made, or improved
in vitro methods can be developed for use in evaluation of environ-
mental hazards.
Animal studies of air pollutants have provided important data,
but carryover to man is lacking. In vitro comparisons of human and
animal tissues should be conducted to attempt to bridge this gap.
Organ cultures may be more useful for this type of comparison than
cultured cells, because the organized differentiated tissue resembles
the tissue of the intact animal.
-------
12
Indirect Tests for Determining
the Potential Carcinogenicity
of Polycyclic Aromatic
Hydrocarbons
SEBACEOUS GLAND SUPPRESSION
In 1954, W. E. Smith et al.709 reported the efficacy of the sebaceous
gland suppression test in determining the carcinogenicity of some
petroleum fractions. The disappearance of the sebaceous glands could
be correlated directly with the carcinogenicity of these fractions.
Pullinger609 was the first to note that sebaceous glands disappear
from mouse skin within a few days after application of some car-
cinogens. Simpson and co-workers704'705 found that, if a carcinogen,
such as 3-methylcholanthrene, were dissolved in lanolin, it would be
noncarcinogenic and would not suppress sebaceous glands. On the
basis of these observations, the test was used extensively as a guide
to the fractionation of cigarette-smoke condensate by a number of
investigators458' 539>737 and most recently by Chouroulinkov et al.134
as a screening method for various types of cigarette-smoke condensate.
A close correlation between the results of long-term carcinogenicity
testing and sebaceous gland suppression was reported.134
Bock and Mund71 used whole mounts of mouse skin and found
that the sebaceous gland suppression effect was pronounced for poly-
cyclic hydrocarbon carcinogens and was parallel to the carcinogenic
148
-------
Indirect Tests for Determining Carcinogenicity 149
activity of this group of compounds. They applied test solution by
pipette twice daily for 3 days on the dorsal area of Swiss mice 55-65
days old. The mice were sacrificed 4 days after the last application.
Potent carcinogens like 3-methylcholanthrene and 7,12-dimethyl-
benz[a] anthracene showed the highest suppression index. The non-
carcinogens, phenanthrene and 1,9-benzanthrone, were inactive. How-
ever, chrysene and benz[a] anthracene also had pronounced suppression
effects, even though their carcinogenicity with regard to mouse skin
is weak. Furthermore, other carcinogens, such as jS-naphthylamine,
had no effect. Bock and Mund later extended their studies to 103
compounds70 and found high levels of sebaceous gland suppressor
activity associated with the benzfa] anthracene structure. It was em-
phasized, however, that sebaceous gland suppression was not always
associated with carcinogenic activity. The potent carcinogen, 7,9-
dimethylbenz[c] acridine, was demonstrated to be a weak sebaceous
gland suppressor, whereas colchicine, which is not a skin carcinogen,
was a moderately active suppressor. The subject of sebaceous gland
suppression by aromatic hydrocarbons has been extensively reviewed
by Bock.67
In summary, the sebaceous gland suppression test is not a reliable
indicator of carcinogenicity but may have limited use in predicting
the carcinogenicity of some groups of compounds, such as substituted
benz[a] anthracenes.
PHOTODYNAMIC ASSAY
Photodynamic activity is demonstrated by the immobilization and
death of Paramecium caudatum, a ciliate, when exposed to other-
wise harmless long-wave ultraviolet radiation after incubation with
photosensitizing poly cyclic compounds in a pure state or in crude
organic mixtures.238'239'246""248 The time required for immobilization
of 90% of P. caudatum is considered as the end point and reflects
concentrations of the photosensitizing agents.
The photodynamic activities of 240 polycyclic aromatic com-
pounds, determined with P. caudatum, have been compared with
their in vivo induction of zoxazolamine hydroxylase activity in
rats.240 A highly significant association was demonstrated between
photodynamic and enzyme-inducing activities. A significant statisti-
cal association between photodynamic activity and carcinogenicity
of polycyclic compounds of wide structural range has been demonr
-------
150 PARTICIPATE POLYCYCLIC ORGANIC MATTER
strated, but the photodynamic assay cannot identify a particular
246
polycyclic compound as being carcinogenic or noncarcmogenic.
The photodynamic assay permits differentiation of pollutants
from different sources and of various fractions of pollutant extracts
derived from any one source.247'249 Relative photodynamic potency,
expressed as apparent micrograms of benzo[a] pyrene per 1000 m3
of air, bears no relation to atmospheric concentrations of particles,
organic compounds, or derived fractions. For the aromatic fraction,
which contains nearly all of whatever benzo[a] pyrene is present in
the parent organic extract, photodynamic potencies are strongly and
positively correlated with benzo[a] pyrene concentrations. The assay
has been applied to organic atmospheric pollutants and six fractions
thereof from more than 100 different sources in the United States,
exemplifying a wide spectrum of urban and rural pollutant charac-
teristics. Pollutants were assayed over a range of 1-100 pg/ml, using
benzo[a] pyrene concentrations of 0.001-100 jug/ml as a standard.
Very high photodynamic activity and steep dose-response slopes
have been demonstrated in basic fractions of organic particulate pol-
lutants from more than 50 U.S. cities. This is of particular interest
in light of the presumptive isolation of dialkylated azaheterocyclic
carcinogens from basic fractions (E. Sawicki, personal communica-
tion) and the very high photodynamic activities of such carcino-
gens.246
More recently, a composite neutral fraction of organic extracts of
particulate atmospheric pollutants has been separated chromato-
graphically into 217 subfractions; nine polycyclic aromatic hydrocar-
bons and five polycyclic carbonyl compounds were identified in these
subfractions. The distribution of photosensitizing polycyclic com-
pounds in these subfractions has been determined with a photody-
namic bioassay using/*, caudatum.2*4
The economy, rapidity, and simplicity of the photodynamic bio-
assay, which can be conducted on less than 1-mg amounts of organic
extracts, are attractive. The data suggest that the bioassay provides
a biologic index of potential carcinogenic hazard attributable to poly-
cyclic compounds. Evaluation of this concept, however, demands
correlated photodynamic carcinogenic and chemical studies on numer-
ous samples and fractions of organic atmospheric pollutants collected
from sources exemplifying a wide epidemiologic spectrum of res-
piratory tract cancer incidence. Such studies are in progress.
-------
13
Teratogenesis and Mutagenesis
TERATOGENESIS
Teratology is the study of congenital malformations. These are
generally defined as structural abnormalities that can be recognized at
or shortly after birth and can cause disability or death.241 Generically,
teratology also includes microscopic, biochemical, and functional
abnormalities of prenatal origin. The incidence of human congenital
malformations is unknown in the absence of a comprehensive national
registry; it has been variously estimated at about 3-4% of total live
births. Three major categories of human teratogens have so far been
identified: viral infections, x irradiation, and such chemicals as mercu-
rials and thalidomide. Although the teratogenicity of various chemi-
cals has been experimentally recognized for several decades, only af-
ter the thalidomide disaster of 1962 were legislative requirements for
teratogenicity testing established.
Teratogenic effects of chemicals and other agents should, of
course, be identified in experimental animals, rather than in human
beings after accidental or unrecognized exposure. Test agents should
be administered to pregnant animals during active embryonic organo-
genesis. Shortly before anticipated birth, embryos should be removed
151
-------
152 PARTICIPATE POLYCYCLIC ORGANIC MATTER
by cesarean section and examined. Characteristics to be considered
in test and concurrent control animals include the incidences of ab-
normal litters, of abnormal fetuses per litter, of specific congenital
abnormalities, and of fetal mortality; maternal weight gains in preg-
nancy; and maternal and fetal organ : body weight ratios. Addition-
ally, some pregnant animals should be allowed to give birth in order
to identify abnormalities that may be manifest only in the perinatal
period.241'454'456
Agents to be tested for teratogenic effects and their known metab-
olites should be administered singly and repeatedly to two or more
mammalian species of more than one order, in various nutritional
conditions, during active organogenesis, by a variety of routes, and
at dosages reflecting possible human exposure. Of interest in this
connection is the total lack of data in the available literature on tera-
togenicity testing by the respiratory route; respiratory exposure is
particularly important for pesticide aerosols and vapors, besides
being the obvious route for testing air pollutants.
To date, there are no available data on teratogenicity testing of
air pollutants by any route. Therefore, community atmospheric
pollutants and defined components thereof should be tested in at
least two mammalian species by inhalation and by parenteral ad-
ministration. Test materials should be administered acutely, sub-
acutely, and chronically.
MUTAGENESIS
Mutagenicity Testing: In vivo Methods
Recent recognition of genetic hazards due to chemicals has been
paralleled by the development of a variety of methods for testing
mutagenicity. Submammalian tests-in bacteria, bacteriophage,
Neurospora, plants, and Drosophila-help to elucidate basic mech-
anisms. However, in view of the wide range of metabolic and bio-
chemical differences between these systems and man, submam-
malian tests should be used to provide data ancillary to more relevant
test systems. Of these, three in vivo mammalian tests are practical
and sensitive: dominant lethal assay, host-mediated assay, and in vivo
cytogenetics. Results from such tests may be extrapolated to man
with a relatively high degree of confidence.172'766
-------
Teratogenesis and Mutagenesis 153
DOMINANT LETHAL ASSAY
Dominant lethal mutants are convenient indicators of major genetic
damage that have been used in mammals for measuring effects of x
rays34 and, more recently, chemical mutagens.35'123'229'235'236'245'295'
636 Data on induction of dominant lethal mutants in mammals may
be appropriately extrapolated to man, especially inasmuch as most
recognizable human mutations are due to dominant autosomal
traits.763 The genetic basis for dominant lethality is the induction
of chromosomal damage and rearrangements, such as translocations
and aneuploidies, resulting in nonviable zygotes. Evidence of zygote
lethality induced in mammals by x rays and by chemical mutagens
has been obtained embryologically368'712'713 and cytogenetically242'
295,647 ^ j Bateman, personal communication), respectively. Ad-
ditional evidence of the genetic basis of dominant lethality is derived
from the associated induction of sterility and heritable semisterility
in F! progeny of males exposed to x irradiation444'712 and to chemi-
cal mutagens;122'237'2SS translocations have been cytologically dem-
onstrated in such semisterile lines in mice123'237'443'708 and in ham-
sters (K. S. LaVappa and G. Yerganian, personal communication).
The induction of dominant lethal mutations in animals can be
assayed with a high degree of sensitivity and practicality after acute,
subacute, or chronic administration of test materials, either orally
or by any parenteral route, including respiratory.232 After adminis-
tration of a drug, male rodents are mated sequentially with groups
of untreated females over the duration of the spermatogenic cycle.
For mice, the entire duration of spermatogenesis is approximately 42
days, comprising the following stages: spermatogonial mitoses, 6
days; spermatocytes, 14 days; spermatids, 9 days; testicular sperm,
5.5 days; and epididymal sperm, 7.5 days. Thus, matings within 3
weeks after single drug administration represent samplings of sperm
exposed during postmeiotic stages, and matings 4-8 weeks later,
samplings of sperm exposed during premeiotic and stem cell stages.
The classic form of the dominant lethal assay involves autopsy of
females approximately 13 days after timed matings, as determined by
vaginal plugs in mice and vaginal cytology in rats and enumeration of
corpora lutea and total implants (as evidenced by living fetuses and
early and late fetal deaths).
A modified test in mice allows the determination of effects of
drugs on pregnancy rates. Corpora lutea counts—which are not only
-------
154 PARTICULATE POLYCYCLIC ORGANIC MATTER
notoriously difficult to carry out but inaccurate in mice—can be
omitted. Numbers of total implants in test animals can be related to
those in controls, yielding a simple measure of preimplantation losses.
With such modified procedures and computerized data handling, large
numbers of test agents can be simply and rapidly tested for mutagenic
activity.8
Dominant lethal mutations are directly measured by enumeration
of early fetal deaths and indirectly by preimplantation losses. Re-
sults are best expressed as early fetal deaths per pregnant female,
rather than the more conventional mutagenic index (i.e., early fetal
deaths X 100 divided by total implants). The latter index can be
markedly altered by variation in the number of total implants. Pre-
implantation losses offer a presumptive index of mutagenic effects,
but there is no precise parallelism between preimplantation losses
and early fetal deaths; these should be regarded as concomitant and
not alternate measures.
HOST-MEDIATED ASSAY
In spite of the universality of the genetic code, some compounds are
actively mutagenic in animals but not mutagenic in microorganisms;
conversely, some compounds are mutagenic in microorganisms but
are detoxified in mammalian systems. The host-mediated assay was
developed to determine the influence of in vivo mammalian factors
in activating or detoxifying chemical mutagens. In this assay, the
chemical under test is administered to the mammal, which then re-
ceives an injection (by another route) of indicator microorganisms,
thus simplifying measurement of mutation frequencies. The micro-
organisms are later recovered and scored for induction of mutants.
Comparison of the mutagenic action of the test agent in the micro-
organism directly and in the host-mediated assay indicates the in-
fluence of host biochemical metabolism in activating or detoxify-
ing the potential mutagen. 284~287 The formation of mutagenic meta-
bolic products from dimethylnitrosamine and cycasin with this pro-
cedure has been reported. Although this is an indirect test, it is the
only practical method for detecting point mutations in vivo.
In vivo CYTOGENETICS: CHROMOSOMAL ABERRATIONS
Experimental animals and man, exposed acutely, subacutely, or
chronically to pollutants by any route, can be investigated cytogenet-
-------
Teratogenesis and Mutagenesis 155
ically for structural and numerical chromosomal aberrations. Chinese
hamsters are favored in cytogenetic studies, although rats are more
commonly used. Cytogenetic effects on metaphase or anaphase prep-
arations of somatic (bone marrow, spleen, and embryo homogenates)
or germinal cells can be studied singly or serially after recovery periods.
High quality of standardized preparations and the use of coded slides
are critical. Distinctions between chromosome and chromatid breaks
and between gaps and open breaks are probably less important than
hitherto assumed, inasmuch as these effects are generally parallel.
Chromosomal aberrations are regarded as indicators of induced genetic
instability and correlate well with mutational frequencies.
Data on mutagenicity testing of air pollutants in in vivo mammalian
systems—the dominant lethal assay, the host-mediated assay, and
in vivo cytogenetics-are scanty. There are no published data on mu-
tagenicity testing by inhalation. High concentrations of benzo[a] py-
rene administered parenterally to male mice induced dominant lethal
mutations in F1 embryos; however, an organic extract of particulate
atmospheric pollutants and three derived fractions were not found
to be mutagenic.245 Trimethylphosphate, used as a fuel additive in
gasoline at a concentration of approximately 250 mg/gal, was muta-
genic in mice after oral or parenteral administration;3 cumulative
effects were also demonstrated. Evaluation of potential human
hazards requires data, as yet unavailable, on the concentration of
unreacted trimethylphosphate and of any biologically active pyroly-
sis products in automobile exhaust.
Consideration of potential biologic hazards due to environmental
contaminants like air pollutants extends to chronic toxic effects,
including mutagenesis, carcinogenesis, and teratogenesis. Such effects
may be induced directly by components of air pollutants them-
selves, or indirectly after interactions between air pollutants and other
environmental pollutants, irrespective of route of exposure.
Test systems must be designed to reflect the role of microsomal
enzyme function in activation and detoxification and the role of
possible interactions between test agents (administered by any route)
and between dietary factors and other chemicals, such as unintentional
and intentional food additives, drugs, and air pollutants. In testing
air pollutants by inhalation, it is necessary to investigate the effects
of defined components by themselves and in combination with other
defined and undefined components, including their reaction products.
Pollutants must be tested at higher concentrations than those of
general human exposure;766'770 irrespective of route of administration,
-------
156 PARTICULATE POLYCYCLIC ORGANIC MATTER
maximally tolerated dosages are recommended for this purpose as the
highest dosage in dose-response studies. Testing at high doses is es-
sential to the attempt to reduce the gross insensitivity imposed on
animal tests by the routinely small sample groups-e.g., 50 or so rats
or mice per dosage per chemical, compared with the millions of hu-
mans at presumptive risk.
Mutagenicity Testing: In vitro Methods
A method of detecting point mutations in mammalian somatic cells
is to use in vitro tissue-culture systems. The potential of using mam-
malian somatic cells in vitro for genetic studies has long been recog-
nized, but substantial progress was not made until improved and
simplified techniques for mammalian cells were developed by Puck
and associates.607'608 These methods made possible quantitative anal-
ysis of genetic variations in cell populations via the plating technique
for mammalian cells.
It was demonstrated137'423 that gene mutations are induced by
treatment of Chinese hamster cells in cultures with alkylating agents.
In addition, physical agents, such as x rays and ultraviolet radiation,
and other chemical agents, such as carcinogens, have been shown to
induce forward and back mutations at several genetic loci in these
cells.90'91 >135.136.424 Thus, the in vitro cell culture offers a new sys-
tem for testing the mutagenicity of chemicals in the human environ-
ment. The question whether somatic mutation may cause cancer can
now be re-examined more critically, because both carcinogenesis
and mutagenesis have been shown to occur experimentally in the
same target cell system in vitro. Furthermore, human somatic cells
from normal and neoplastic tissues can also be tested directly.
Chu and co-workers136 have tested, in Chinese hamster cell cul-
tures, the mutagenicity of a few selected groups of chemical carcino-
gens and their related compounds and derivatives. Table 13-1 lists
the compounds tested so far and their relative carcinogenicity (based
on animal studies) and mutagenicity. The genetic marker assayed in
the hamster cells was the change from 8-azaguanine sensitivity to
resistance. The results obtained thus far indicate that there is a direct
relation between the degree of carcinogenicity and mutagenicity and
that metabolically activated derivatives of the test compounds often
play important roles in mutagenic action. It has recently been demon-
strated that epoxides of polycyclic hydrocarbons are much more mu-
tagenic to mammalian cells than are the corresponding hydrocarbons,
dihydrodiols, and phenols (C. Heidelberger, personal communication).
-------
Teratogenesis and Mutagenesis 157
TABLE 13-1 Relative Carcinogenicity and Mutagenicity of Selected
Compounds0
Test Compound Carcinogenicityb Mutagenicity
Benzo[e]pyrene - -
Benzo[a]pyrene + -
3-Hydroxybenzo[a] pyiene - ±
Dibenz[a,c] anthracene - -
Dibenz[a,h] anthracene ± ±
7,12-Dimethylbenz[a] anthracene +-H- -H-+
2-Acetylaminofluorene + —
Af-Hydroxy-2-acetylaminofluorene + -
JV-Acetoxy-2-acetylaminofluorene +++ -M-+
"Derived from Chu et al.136
*Key:
- = not carcinogenic (or mutagenic)
± = uncertain or weakly carcinogenic (or mutagenic)
+ = carcinogenic (or mutagenic)
+++ = strongly carcinogenic (or mutagenic)
Parallel results have been obtained in the induction of mutations
with the same series of compounds at the adenine-3 region of Neuror
spora (H. V. Mailing, personal communication). Similarly, N-acetoxy-
2-acetylaminofluorene has been shown to be mutagenic in T4 bac-
teriophage,163 transforming DNA in Bacillus subtilis507 andEscheri-
chia collSS1
Clearly, these results are promising, but more data using more rep-
resentative compounds and additional genetic loci will be needed be-
fore a more definitive conclusion may be drawn. The use of mammal-
ian cells in vitro for a combined and coordinated test for chemical
mutagenesis and carcinogenesis may be expected to yield significant
information on cellular mechanisms of cancer formation. However,
data on mutagenesis derived from somatic cells in vitro are limited by
the present inability to identify the factors involved by conventional
genetic techniques.
ASSOCIATIONS BETWEEN MUTAGENICITY AND
CARCINOGENICITY
It is now generally accepted that mutagenesis involves a change
in the structure of DNA. Whether such a change is essential for
chemical carcinogenesis remains unknown. Most chemical substances
that react with nucleic acids also react with proteins. That has often
made it difficult to identify the significant cellular receptors respon-
-------
158 PARTICULATE POLYCYCLIC ORGANIC MATTER
sible for the biologic effects of carcinogens. Because of the uncer-
tainty, two general molecular mechanisms of chemical carcinogenesis
have been proposed: somatic mutation resulting from the binding
of a chemical to DN A and alteration of its structure; and modifica-
tion of gene expression, which could occur in several ways, including
derepression.396 There is no compelling evidence for or against the
nonmutational theory of chemical carcinogenesis, but recent evidence
from mammalian cell systems that lends some support to the somatic-
mutation theory is considered in the paragraphs that follow.
The somatic-mutation theory of carcinogenesis, which is generally
quoted as originating with Boveri,81 has received intermittent support
from various authors up to the present. However, a review of the
evidence pertinent to this concept led Burdette101 to conclude that
a general correlation between mutagenicity and carcinogenicity could
not be established. The principal objection to the theory was that a
number of chemical carcinogens had not been found to be muta-
gens and well-established mutagens had not been shown to be car-
cinogens. It was, nevertheless, pointed out that these arguments were
not conclusive, inasmuch as some chemicals may be demonstrably
mutagenic only after metabolism and may differ in their ability to
yield particular types of mutations. Furthermore, the prolonged
testing necessary to eliminate the possibility that known mutagens
are carcinogenic has not been carried out in many cases.
If any association between mutagenicity and carcinogenicity is
sought, it seems desirable that the experimental tests be carried out
in the same species, preferably mammals. Various test systems now
available for mutagenicity in mammals have been discussed. The ensu-
ing discussions deal primarily with the mutagenicity tests with chemi-
cal carcinogens in mammalian systems.
Mutations may be classified into chromosomal alterations, point
mutations of nuclear genes, and mutations of extranuclear genes. It
has been shown that most, if not all, chemical carcinogens have the
capacity to induce mitotic and chromosomal abnormalities.56 Varia-
tions in chromosome number and structure have been reported in
several neoplasms induced by carcinogenic hydrocarbons in several
species of rodents. Nevertheless, there is no established correlation
between karyotypic abnormalities and the initiation of neoplasia.
More recently, it has been shown that some carcinogenic hydrocar-
bons, but not the structurally related noncarcinogenic compounds,
can induce chromosome aberrations in mammalian cells both in vivo
and in vitro.426'427 The latter type of study is promising, but more
-------
Teratogenesis and Mutagenesis 159
extensive tests using a larger selection of components would be
desirable. It is clear that the cytogenetic effects of environmental
chemicals could provide a convenient and valid indication of genetic
damage in cells and organisms. But it must be pointed out that many
chemicals can induce point mutations but not chromosome muta-
tions.
In 1948, Darlington180 proposed a plasmagene theory of the
origin of cancer. He believed that the cancer determinants that arise
in the cytoplasm are due to mutations in either hereditary plasma-
genes, infectious viruses, or proviruses. More recent evidence indi-
cates not only that some viruses are tumorigenic, but also that chemi-
cal agents might activate or modify the infectious viruses or proviruses
(oncogenes?) that are present in the cell.399 In addition, in view of the
demonstrations of extrachromosomal inheritance in protozoa and
fungi, epinuclear hereditary factors may be present in mammalian
cells. Efforts in this area, particularly with respect to the possible
alterations of these epinuclear factors in relation to cancer, may turn
out to be rewarding.
Chu135 has recently devised a modified version of host-mediated
assay by placing Chinese hamster cells in dialysis bags that are then
implanted surgically into the peritoneal cavities of rats. After various
periods, the hamster cells are removed for mutagenic studies in vitro.
Both point and chromosomal mutations can be assayed this way. Al-
though the procedures are still being improved, this test system may
be useful for the study of mutagenicity of carcinogens.
Point mutations in the germ cells of mammals have been experi-
mentally induced and quantitatively analyzed. A great majority of
dominant lethal mutations probably involve chromosomal aberrations.
The test procedure designed for the analysis of radiation-induced,
specific-locus mutations in mice has been readily adapted for muta-
genicity testing of chemicals (e.g., see Ehling228). Although this spe-
cific-locus method can be extended to include chemical carcinogens,
the costs for mammalian breeding experiments are so prohibitive as
to make routine testing impractical, at least at the initial stages of
screening for chemical hazards. However, somatic mutations in in-
tact mammals may be more relevant to carcinogenesis, and methods
for their detection are urgently needed.
-------
14
Vegetation and Polycyclic
Organic Matter
Polycyclic organic matter is widely distributed in water, soil, air, and
plants. Widespread occurrence of POM has been reported in many
plants and plant products, such as tobacco smoke,156'160"162'792
snuff,111'113 peat,298 wood soot,463 charred biscuits,463 stack gases
from pulp mills,66 nonurban soil,9'66 roasted coffee beans,130'465
plant tissues,77'312'324'325 pyrolyzed cellulose, lignin, pectin,296
tobacco leaves,114'792 wood-smoked foods,496'736 incinerator efflu-
ents,343 and marine fauna and flora.107'111'510 Combustion of almost
any organic material contributes polycyclic organic compounds and
their partial oxidation products in trace amounts to the environment
for probable contamination of every receptive surface; and POM is
produced by baking, barbecuing, broiling, or frying of many foods.325
Although several polycyclic compounds generated by burning of
vegetation are known to be carcinogenic when applied externally to
particularly susceptible animal tissue, few have been adequately tested
by ingestion by experimental animals.325 Tolerance or resistance to
low concentrations of the carcinogenic materials has undoubtedly
developed through the process of natural selection in animal species
that consume smoke-contaminated foods or are exposed repeatedly
to the by-products of combustion. Man has apparently acquired some
160
-------
Vegetation and POM 161
degree of tolerance to POM through continual exposure to the ex-
ternal environment and to foods. If the limits of their tolerance are
exceeded by additional exposures, the carcinogenic threshold of a
particular compound or combination of compounds may also be
exceeded, and that will result in the development of cancer. It is
important to examine the possible contribution of vegetation to
the total burden of carcinogenic or potentially carcinogenic com-
pounds in the environment.
POM apparently does not induce cancer-like tumors in plant tis-
sue, although a variety of tumors do occur naturally. The most com-
mon of these unusual growths is crown gall, which is initiated by a
microorganism.17>8S The ability of bacteria to produce crown gall is
related to their production of jS-indoleacetic acid. Cultivated mush-
rooms exposed to fumes emitted from coal tar, diesel oil, and a com-
ponent fraction of tar acids developed tumorous growths.254 Tumors
were also produced on mushrooms by incorporating soot and diesel
oil in the nutrient medium. Attempts to produce similar tumors by
direct application of carcinogenic polycyclic aromatic hydrocarbons
to the mycelium of the mushroom failed.
Both the carcinogenic benzo[a] pyrene and its inactive isomer
benzo[e] pyrene have been found to be fairly abundant, even in rural
soils remote from major highways and industries.76 These benzopy-
renes are among the pyrolytic products of wood, and they also occur
in the transformation of plant organic matter to peat and lignite.298
The POM content of soils is also increased by exposure to industrial
effluents, products of oxygen-deficient burning of vegetative matter,
deposits of petroleum products, and exhaust gases from the automo-
bile. Mallet and Heros511 detected benzofa] pyrene in tree leaves and
in decaying organic matter under the same trees. They suggested that
the benzofa] pyrene was absorbed through the tree roots and was
translocated through the transpiration stream to the leaves. These
observations and the report by Guddal323 that several hydrocarbons
may be absorbed from contaminated soils by plants were substantiated
by Dorr's experiments,215 which showed that benzo[a] pyrene was
absorbed from soil and water cultures by barley roots and was trans-
located to the shoots.
In polluted atmospheres,155'156-218'324'662'664'752 POM may contami-
nate plants used for food by settling on surfaces or by absorption into
tissues. Grimmer and Hildebrandt313 found that grain samples from
the Ruhr district contained 10 times more carcinogenic polycyclic
matter than samples from nonindustrial areas. Grimmer312 reported
-------
162 PARTICULATE POLYCYCLIC ORGANIC MATTER
increased concentrations of phenanthrene, anthracene, pyrene, an-
thanthrene, fluoranthene, benz[a] anthracene, chrysene, benzola]-
pyrene, benzofe] pyrene, perylene, benzo[ghi] perylene, dibenz[a,h]-
anthracene, and coronene in samples of lettuce, kale, spinach, leeks,
and tomato collected from the field. The concentrations varied widely
between fields and between locations in the same field. The maxi-
mal concentration of benzofa] pyrene in lettuce was 12.8 Mg/kg of
tissue, whereas the maximal concentration in kale was twice that.
Spinach, leeks, and tomato samples contained benzofa]pyrene at
7.4, 6.6, and 0.2 jug/kg, respectively. It was possible to remove ap-
proximately 10% of the benzopyrene from the vegetables by washing
them in cold water, but Grimmer312 reported that the remainder of
soot film could not be washed off with water. Plants with the small-
est amount of surface area exposed to the atmosphere, such as toma-
toes and leeks, had the least benzo[a] pyrene. Howard and Fazio397
indicated that little information was available on the extent of con-
tamination of our food supply via air pollutants. Gunther et al,32S
found anthracene (25 ppm) and five unidentified polycyclic com-
pounds in orange rinds obtained from an area adjacent to a heavily
traveled highway. Boiling73 reported higher concentrations of POM
in wheat, corn, oats, and barley grown in industrial surroundings
than in those grown in more remote areas; and Gra'f and Diehl309
identified eight polycyclic compounds in various plant leaves, with
concentrations of benzo[a] pyrene as high as 40 parts per billion (ppb)
in some leaves.
Borneff et al.77 conclusively demonstrated the biosynthesis of
POM in plants. Algal cultures were grown in nonlabeled and 14C-
labeled acetate as the sole carbon source. Data showed that POM
was synthesized and that plants have a normal concentration of about
10 /ig of benzo[a] pyrene per kilogram; the total amount may in
some cases be greater than 100 Mg/kg. According to Borneff et al.,77
the health risk in orally introduced polycyclic compounds has not
been completely clarified. Wynder et al.832 indicated that the con-
sistent uptake of polycyclic compounds by vegetation may exert an
influence on the incidence of human intestinal neoplastic disease.
Hakama and Saxen334 found a significant correlation between the
consumption of cereals and the occurrence of gastric cancer. Grim-
mer312 found that benzo[a] pyrene in 23 samples of cereal grain
varied from 0.2 to 0.4 Mg/kg, and flour and bread made from some
of these samples appeared to retain most of the benzofa] pyrene.
Grimmer also found that the water extract from 1 kg of tea contained
-------
Vegetation and POM 163
about 4 fig of benzo[a] pyrene; however, 11 coffee samples contained
hardly any hydrocarbons.
In addition to the polycyclic compounds synthesized in plant tissue
or deposited on plants by polluted air, the preparation of foods may
increase the total burden of carcinogenic compounds. Davies and
Wilmshurst185 reported the formation of 0.7 jug of benzopyrene per
kilogram of starch heated to 370-390 C and suggested that tempera-
tures reached during the toasting of bread (390-400 C) may be capa-
ble of producing polycyclic compounds. Chassevent and Heros130
found little benzopyrene in commercially roasted coffee, but consid-
erably more was found in "home"-roasted coffee, which included the
endosperm of the seed. Kuratsune and Hueper465 reported that poly-
cyclic aromatic hydrocarbons were found in roasted coffee but did
not specify whether the samples tested contained the endosperm.
Benzopyrene is apparently present in a wide variety of both cooked
and uncooked foods. According to Raven and Roe,625 0.3-2.1 Mg/kg
of Icelandic smoked meat and fish was found by Bailey and Dungal;23
Lijinsky and Shubik496 reported 8 jug/kg of smoked salmon; and
Gorelova and Dikun306 found up to 10.5 Mg/kg of home-smoked
sausages. Much of this benzopyrene was no doubt produced by py-
rolysis of fat in the meat, but condensed smoke from the oxygen-
deficient combustion of organic fuel used to produce smoke as a meat
preservative was also a contributor.
Wynder and Hoffmann831 found 5.3, 4.4, 2.4, and 1.4 jug of
benzo[a] pyrene per 100 cigarettes made from Virginia, Turkish, burley,
and Maryland tobaccos, respectively. The quantity of benzo[aj pyrene
isolated by various workers around the world varies widely. A review
of these results by Wynder and Hoffmann831 shows a range from
0.2 Mg/100 cigarettes reported in Denmark to a maximum of 12.25
jug/100 cigarettes for one sample in the United States.
As might be expected, smoke from tobacco products other than
cigarettes also contains polycyclic compounds. A significantly higher'
benzo[a] pyrene content for the smoke of pipe tobacco (especially
when compared with cigarette tobacco smoked in a pipe) suggested
that additives for pipe tobacco, particularly sugars, may become pre-
cursors of benzofa] pyrene on pyrolysis.390 Extensive studies of
tobacco pyrolysis have been made in an effort to determine whether
specific compounds in tobacco can be considered precursors of POM
in tobacco smoke.
Polycyclic compounds in processed tobacco itself may be derived
from polluted air or from tobacco processing (curing, aging, etc.).
-------
164 PARTICIPATE POLYCYCLIC ORGANIC MATTER
Studies by Lyons505 and Bentley and Burgan46 showed that traces of
polycyclic compounds found in tobacco do not contribute an ap-
preciable amount to the total aromatic hydrocarbons in tobacco
smoke. However, Bentley and Burgan45 and Wynder and Hoffman827
reported up to 12 ppb and 20 ppb of benzota] pyrene, respectively,
in tobacco. Van Duuren et al.792 found tumor-promoting agents in
tobacco leaf and in the smoke condensate. More recent work by
Van Duuren et al. 789>791 also pointed out that aromatic hydrocar-
bons may be involved in a two-stage process of carcinogenesis. Sev-
eral noncarcinogenic polycyclic compounds were found to function
as initiating agents with croton seed oil as a promoter of tumor for-
mation. In addition to the indicated tumor-promoting property of
croton seed oil, Hecker348 reported that the oil may contain a carci-
nogenic principle. The significance of polycyclic content of tobacco
on the emission of carcinogenic hydrocarbons from cigarette smoke
has been questioned by several researchers because of the small
amounts detected in the tobacco. Campbell and Lindsey113 used the
analysis of cherry laurel leaves for polycyclic compounds as a check
against results obtained with tobacco to show that there was essen-
tially no difference between them.
Food may become contaminated with polycyclic hydrocarbons
and other carcinogens when crops are sprayed for pest control. To
test this assumption, Gunther et al.32S selected oranges growing in
southern California, because fruits remain on the trees almost all
year and orchards are planted both in heavily polluted regions and in
areas relatively free of pollutants. The studies were designed to deter-
mine residue persistence of five selected polycyclic hydrocarbons
(3-methylcholanthrene, dibenz[a,h] anthracene, benzo[a] pyrene,
dibenzo[a,i] pyrene, and anthracene) in and on Valencia orange
rind. The first four of these occur in agricultural environments from
air pollution, from industrialization, and from petroleum oil pest-
control operations. With the exception of dibenzo[a,i] pyrene (per-
sistence half-life, 12 days), the compounds that had penetrated into
the rind to the extent of 1-12% were considered to be long-lived,
with persistence half-lives of 120-200 days in the field. Degradation
half-lives were uniformly 1-2 days. Apparently, 85% of the degrada-
tion losses, presumably by volatilization and oxidation, occurred in
a few days with little penetration into the fruit pulp, and there was
no evidence of translocation into twig tissue.
Anthracene and five unidentified fluorescing materials were found
in Valencia oranges grown near a major highway where there was rela-
-------
Vegetation and POM 165
tively high air pollution.325 Surface contamination of the fruits was
apparently degraded rapidly, but any portion of the polycyclic com-
pounds that became incorporated into cuticular oils and waxes per-
sisted for long periods.
In summary, no information was found to indicate that carcino-
genic polycyclic hydrocarbons affect vegetation. Polycyclic com-
pounds absorbed by roots from contaminated solutions, by foliage
from polluted atmospheres, and by aquatic plants from contaminated
bodies of water are added to the traces of these compounds produced
metabolically. One researcher has reported abnormal growths on
mushrooms grown on contaminated media. Burning of vegetation
and some plant products may produce significant quantities of several
carcinogenic hydrocarbons. The increased concentrations of these
materials in organic soils and in sediments in large bodies of water
suggest that many of the polycyclic compounds are produced in de-
cayed organic matter.
The following recommendations are appropriate:
1. Encourage research to determine the contribution of pestici-
dal sprays, herbicides, and polluted atmosphere to accumulation of
polycyclic hydrocarbons in and on vegetation.
2. Determine the influence of traces of carcinogenic materials in
vegetable foods on the incidence of cancer in man and animals.
3. Produce a reasonably accurate estimate of the amounts of car-
cinogens generated by wild fire and by the combustion of solid waste.
4. Investigate the effect of long-term exposure and massive dosages
of polycyclic compounds on plant growth, development, and repro-
duction.
-------
15
Introduction to Appraisal of
Human Effects
A historical review of man's reaction to airborne pollutants containing
what are now recognized as polycyclic aromatic hydrocarbons reveals
that occupational incidents provided the first evidence of cause-effect
relations. The degree of exposure to such materials is often much greater
in occupational settings than that encountered in community air pol-
lution. The first recorded description of an occupational disease related
to the burning of fossil fuel appeared in 1775, when Percivall Pott, a
surgeon at St. Bartholomews Hospital in London, published a paper600
about cancer of the scrotum in chimney sweeps and related the disease
to their constant exposure to soot. Since then, and particularly since
the large-scale use of fossil fuels has expanded, components and degrada-
tion products of such fuels after burning, refining, distilling, or cracking
have been demonstrated to have a close association with a high incidence
of skin cancer affecting the scrotum and other heavily exposed skin
areas. For example, a high incidence of cancer of the skin has been ob-
served among workmen in coal-tar industries and gas plants (particularly
in operations in oil and shale refineries) and among machine operators
using lubricating oils (from particular sources) in the textile industries
and in machine shops. Clinical incidents of this kind have been well de-
scribed and reviewed by Henry.359-360 In recent years, polycyclic or-
166
-------
Introduction to Appraisal of Human Effects 167
ganic materials proved to be carcinogenic in experimental animals have
been found in the derivatives of fossil fuel associated clinically with
skin cancer. It seems logical for skin problems to be described first, in-
asmuch as the skin is the most vulnerable of organ systems and its
diseases are easy to identify.
Descriptions of lung disease related to dust or airborne particles were
made in the sixteenth century. More than 300 years later, Harting and
Hesse345 demonstrated that lung cancer was prominent among the
pulmonary diseases from which miners were suffering. The ores
from the Joachimsthal and Schneeberg mine areas were eventually
shown to contain radioactive dusts; it was from the pitchblende from
this area that Marie Curie first extracted radium. Although there is
no evidence that POM played a role in the lung cancer of the miners,
this is an excellent historical example of the devastating biologic effect
of an airborne pollutant in man.
It was not until the twentieth century that cancer of the lung was
shown to be associated with coal processing operations, such as coking,
and the manufacture of illuminating gas. Substantial increases in lung-
cancer mortality rates over those of the general population have been
noted by Kawai et a/.430 in gas-generator workers and by Alwens et
a/.,13 Kuroda,466 the Kennaways,434 and Doll205 in coke-oven
workers and gas-retort workers. In most instances, polycyclic aro-
matic compounds, such as benzo[a] pyrene, have been recovered from
the airborne contaminants.
There has been strong suspicion for some time that the smoking of
tobacco is related to human disease, particularly lung cancer. From
1939 to 1964, at least 29 retrospective epidemiologic studies of lung
cancer were published.769 It was studies like those of Hammond and
Horn341 that provided the evidence that lung cancer could be cor-
related positively with cigarette smoking. Since 1964, several highly
significant reports have related the frequency of illness to smoking.54'
207,208,339,422
ENVIRONMENTAL POLLUTION: THE PUBLIC
HEALTH PROBLEM
Consideration of the importance of physical, chemical, and thus environ-
mental factors in disease first emerged from recognition that disease
patterns in Americans had changed and were different from those af-
fecting people in underdeveloped countries. In the latter, deaths appear
-------
168 PARTICULATE POLYCYCLIC ORGANIC MATTER
to be related largely to infection, with tuberculosis, pneumonia, and
diarrheal diseases still the major killers. In the United States, the major
causes of death include cardiovascular diseases, cancer, and stroke,''6
as well as chronic bronchitis and emphysema.
In 1969, lung cancer was the greatest single cause of cancer deaths
in men, killing almost 50,000. A rare cause of death in women 30
years ago, it was responsible for 10,000 deaths in women in 1969.
Such increases have occurred in almost every industrial nation in the
world, and Clemmesen144 predicts that lung cancer will achieve epi-
demic proportions within the next decade.
The relation between cigarette smoking and lung cancer771'772 has
stimulated concern over the role of air pollution in cancer, because
some urban air pollutants, including POM, are similar to those found
in cigarette smoke. Epidemiologic studies of occupations in which sim-
ilar substances were present in large quantities also revealed a sharp
increase in skin and lung cancer in persons with long exposures. The
rising incidence of and poor prognosis in lung cancer, even with early
diagnosis, has led to recognition of the necessity for defining all signifi-
cant etiologic factors. Primary prevention is the only effective means
of control.
DATA SOURCES AND PROBLEMS IN
INTERPRETATION OF HEALTH EFFECTS DATA
The difficulty in defining causal and dose-effect relations arises from
the peculiar natural history and characteristics of lung cancer. It has
a long latent period-possibly as long as 30 years—and its peak inci-
dence occurs after the age of 50. The long interval between initial
exposure to a cancer-inducing agent and the appearance of detectable
disease makes etiologic analysis difficult. This is particularly true in
view of the many changes that may occur in the person in terms of oc-
cupation, residence, nutrition, habits, and socioeconomic status. For
example, in the United States in the last 30 years, there has been a
remarkable migration from rural town and farm areas to large cities
and, more recently, from cities to suburbs. In addition, the nature and
amount of environmental contaminants have changed, with the develop-
ment of new industries and the closing down of old ones and changes
in modes of transportation and types of fuels. Many of these factors
apply as well to the pathogenesis of cancers in other body systems.
The use of laboratory animals is a valuable source of material. It
-------
Introduction to Appraisal of Human Effects 169
makes it possible to study short- and long-term biologic effects of daily
doses of single and multiple substances under different conditions in
a controlled environment not analogous to the human condition. Ani-
mal studies with airborne pollutants, particularly polycyclic aromatic
hydrocarbons, have indicated that pathologic changes, such as neo-
plasms and inflammatory responses, occur through the painting, in-
jecting, or implanting of the material on or in the skin and through
intratracheal instillation, implantation, or inhalation.
In vitro methods have also provided methodologies for studying
the induction of neoplastic or toxic changes in single cells, small cell
populations, or whole organ cultures.
Lung cancer is associated with many etiologic factors, some of which,
like cigarette smoking and occupational exposure to asbestos, are well
known. There are a number of methodologic approaches to the study
of the etiology and pathogenesis of lung cancer. Data are available
from epidemiologic studies of the disease in workers exposed to high
concentrations of known carcinogenic substances in different industries.
Some insight into cause-effect relations may be obtained by reviewing
available data on occupational exposures. But the populations under
observation are biased, in that they consist mainly of young, healthy
persons exposed for only a portion of each day or week. They do not
include those who, because of increased sensitivity to the pollutant or
development of related diseases, are forced out of the industry. In
addition, lung cancer affects mostly older people, many of whom might
have left the industry that was causally related to the disease and would
thus be lost to follow-up.
The epidemiologic method of studying the effect of air pollution on
the incidence of lung cancer involves the comparison of lung cancer
death rates in communities that have demonstrably different levels of
pollution. The largest environmental differences are found in contrasts
between different countries, but the interpretation of such contrasts
is made difficult by virtue of the wide differences between countries in
smoking habits and other characteristics. Comparisons between urban
and rural areas within a country are attractive, in that they tend to
maximize differences in levels of pollution. Contrasts between urban
areas offer the important advantage that direct measures of air pol-
lution are generally available only for urban centers.
Each such comparison requires a suitable method of adjustment for
the major known extraneous variables. Because the lung-cancer death
rate depends heavily on age and sex and because age distributions vary
markedly, only age-specific or age-standardized and sex-specific lung
-------
170 PARTICULATE POLVCYCLIC ORGANIC MATTER
cancer death rates should be compared. This is generally possible. Cig-
arette smoking has been shown to be correlated with lung cancer. How-
ever, adjustment for amount of cigarette smoking is difficult and in
any case uncertain. Detailed cigarette-consumption statistics of specific
areas are often difficult to obtain, and consumption specificity by age
and sex is virtually undocumented. Details of smoking practice-in
particular, butt length-may be important. Unfortunately, no usable
documentation of such differences is available.
Despite difficulties, some types of comparison strongly support the
proposition that urban pollution is related to an increased lung cancer
death rate. It is much more difficult to relate the increment in deaths
directly to specific contaminants, and efforts to do so have led to
variable results.
The problem of assessing the effect of POM on the incidence of
lung cancer is compounded further by other factors that make it more
difficult to carry out and evaluate epidemiologic studies. Areas to be
considered are listed below.
1. The adequacy of measurements of POM
In the past and even today, measurements were carried out in only
a few areas of large cities. In Chicago, for example, Carnow117 found
that wind patterns often create conditions in which levels of pollu-
tants may be remarkably different in various areas on different occasions,
so limited measurements might not truly represent the degree of expo-
sure of individuals, particularly if values are expressed as city wide or
annual averages.
2. Whether the increase in incidence rates is real or apparent
The extensive routine use of x rays by physicians and in mass sur-
veys for tuberculosis in the last 25 years and physicians' increased
awareness of this disease are undoubtedly factors in the reporting of
disease incidence. Greater use of methods for making tissue diagnosis-
such as bronchoscopy, bronchial and pleural biopsy, and exfoliative
cytology-and greater reliance on autopsy also increase the number of
detected cases of carcinoma, which in the past might have been con-
sidered tuberculosis, pneumonia, or other diseases. Greater access to
physicians in urban and rural areas increased the case-finding potential.
Aging of the population naturally contributes additional cases in per-
sons who, in the past, might have died earlier in life from other dis-
eases. Kotin,448 Lew,494 Clemmesen,144 Kreyberg,461 and others have
-------
Introduction to Appraisal of Human Effects 171
reviewed and carefully considered these issues. There seems little room
for doubt that a substantial portion of the observed increase is real.
3. Whether a rural-urban difference can be ascribed to persons who
die in large cities to which they have come from rural areas for diag-
nosis and treatment
Studies by Stocks,728 Haenszel,329 and others have shown that migra-
tion to cities for diagnosis and treatment is not a factor in urban-rural
differences.
4. Whether the apparently significant higher incidence in lower
socioeconomic groups is related to factors other than the urban factor,
i.e., inadequate medical care, poor nutrition, type of heating, or great-
er occupational exposure to carcinogens
This question has been studied by Manos and Fisher,515 Cohart,149 and
others; the urban-rural differences persist after adjustment for these
factors.
5. Whether differences can be ascribed to ethnic factors
Graham et al.310 documented high rates of lung, prostate, and gastric
carcinoma in Poles. Religious differences may also be significant, as
suggested by the work of MacMahon,506 who found upper respiratory
cancer rates in Catholics and Protestants to be 3 or 4 times those in Jews
but found little difference between foreign- and native-born groups.
This may be related to cultural, dietary, occupational, or genetic factors.
Racial differences were also studied but require further investigation.
Duchen219 found no increased incidence in Caucasians contrasted with
Bantu natives that could not be explained by differences in longevity,
whereas Hoffman and Gilliam383 found a higher lung cancer rate in
Caucasian than in Negro males.
-------
16
Characteristics of Human
Disease Related to Polycyclic
Organic Matter
LUNG CANCER
In no instance has exposure to a specific polycyclic aromatic hydro-
carbon been proved to have caused a tumor in man. That does not,
however, deny the risk of exposure. There is now good evidence of the
overwhelming importance of cigarette smoking in the etiology of lung
cancer in man, and polycyclic aromatic hydrocarbons in cigarette
smoke have been considered as an identifiable group of components in
this connection. Although the effects of known dosages of specified
substances acting alone should perhaps be assessed first, the possibility
of additive or potentiating effects of other factors must also be con-
sidered. For example, absence of evidence of carcinogenic effect of
atmospheric polycyclic aromatic hydrocarbons in nonsmokers does
not preclude an effect in the pathogenesis of lung cancer in smokers.
Carcinogenic polycyclic aromatic hydrocarbons from a variety of
sources are known to be present in urban atmospheres. To assess car-
cinogenic potential, the composition of the atmosphere must be con-
sidered. If the ambient air concentration of a supposedly carcinogenic
substance, "A," is found to correlate with the higher incidence of lung
172
-------
Characteristics of Human Disease Related to POM 173
tumors in a specific human population, but not in all, one of several
inferences can be drawn:
1. "A" is not a lung-specific carcinogenic agent and another sub-
stance, "X," or substances, may in fact be the responsible agent.
2. "A" is carcinogenic for the lung but only if:
a. its effect is added to that of another carcinogenic sub-
stance or substances;
b. it follows or accompanies other substances that possess an
initiating or cocarcinogenic effect;
c. a carrier substance, which may itself be inert, keeps "A" in con-
tact with the target cells;
d. other conditions are favorable for bringing "A" into effective
contact with the susceptible tissue; or
e. there is some combination of the above.
Carcinogenicity of air pollutants has been demonstrated not only
by bioassay of crude benzene extracts of deposited soot,241 >487 but
also by identification of specific substances known to be carcinogenic
in experimental animals, notably benzo[a] pyrene.
Irritant or toxic gases are known to exist in various concentrations
in the atmosphere—e.g., sulfur dioxide, oxides of nitrogen, and ozone.
Such gases are known to have effects other than simple irritation of the
conjunctiva—specifically, a potentiating action on the carcinogenic
properties of polycyclic aromatic hydrocarbons.12'19 This has been
demonstrated by the higher incidence in C AF/Jax mice of pulmonary
adenomas produced by simultaneous exposure to ozone and carcino-
gens.734 Probably more significant is the role of sulfur dioxide in
benzo[a] pyrene carcinogenicity in rats, inasmuch as some of the tumors
have been squamous.474
The interaction of a presumed carcinogenic inhalant and a target
tissue—in this case the lung—can be considered in the framework of
an "ideal model" for inhalation carcinogenesis in man. An ideal model is
one that provides in quantitative terms knowledge of the response of
a specific tissue to a defined agent and of the factors governing the
response. Such a model requires examination of data from analyses of
atmosphere and tissues, from physiologic and biochemical studies, from
animal experiments, and from epidemiologic studies. Inferences for
man from experimental studies involving animals, as considered here,
must be drawn with particular caution.
-------
174 PARTICIPATE POLYCYCUC ORGANIC MATTER
Components of a model for inhalation carcinogenesis in man are
given in Table 16-1.
Determinants of Concentration in Tissue at Risk
According to most investigators,449 polycyclic aromatic hydrocarbons
in the atmosphere are bound to particles that when condensed can be
characterized as "soot." The size distribution of the particles (0.125-
2.5 Mm in diameter) is well within the range likely to be aspirated into
the lower respiratory tract. 1S2'77S
Hatch347 has stated that the effective dose of an inhaled pollutant
is the dose that reaches the critical site in the body. Duration of con-
tact is also important. These are determined by the vector sums of
delivery of the pollutant and its clearance or degradation.
Particles larger than 5 Aim in diameter for the most part become en-
trapped within the upper respiratory tract or are removed by ciliary
action or coughing from the lower respiratory tract, usually carried
within mucin. Some material whose size makes it aspirable into the
lower respiratory tract is removed in the same way as the larger parti-
cles. Precise details depend on the properties of the material, the physio-
TABLE 16-1 Components of Model for Inhalation Carcinogenesis in Man
A. Composition of polluted atmosphere
1. Concentration of each pollutant
2. Physical state of each pollutant
3. Presence of potentiating agents
4. Presence of "inert" particles or absorbents
B. Factors determining concentration at target tissue
1. Physiologic characteristics
2. Protective mechanisms
ciliary function
mucous barrier
3. Clearance mechanisms
phagocytosis
leaching
ciliary action
removal by blood and lymph
metabolic transformation
C. Mechanism of carcinogenesis
D. Factors modifying reaction at target site
1. Synergistic or antagonistic substances
2. Host factors
previous disease
genetic factors
-------
Characteristics of Human Disease Related to POM 175
logic characteristics of the subject, and the subject's state of health.
These variables are difficult to establish accurately and represent one
of the deviations from the "ideal model."
Impairment of ciliary transport mechanisms by chemicals, including
pollutants in the atmosphere or in cigarette smoke, could be important
in increasing the concentration of damaging substances within the lower
respiratory tract. Such ciliostatic effects have been studied by Hilding376
and Dalhamn et al.m Table 16-2 is a partial list of ciliostatic sub-
stances whose effects have been investigated. Living pathogenic agents
can have a similar effect.449
Clearance of fine particles of polycyclic substances can be remarkably
rapid. In mice, methylcholanthrene is cleared within 6 hr of inhalation
and within 24 hr of intratracheal instillation.618 In some experimental
animals, presumably inert carrier substances have been shown to have
an important effect on the concentration-time determinants of the dam-
aging action of inhalants. The experiments of Boren75 have shown that
carbon functioning as an absorbent greatly increases the damaging ac-
tion of nitrogen dioxide on the lung. When tritiated benzo[a] pyrene
is incorporated in carbon or asbestos, clearance from the lungs of
hamsters is slowed.694 Increase in the carcinogenic effegt of benzo[a]-
TABLE 16-2 Compounds in or Related to Constituents of Polluted Urban Air
and Cigarette Smoke That Can Inhibit Ciliary Activity"
Compounds in Air Pollutants Compounds in Cigarette Smoke
Paraffin Nicotine
2-Methylpentane Pyridine
Olefins Ammonium hydroxide
2-Methylbutene-2 Methylamine
2-Methylpentene-2 Trimethylamine
Aromatic Acetonitrile
Benzene Thiocyanic acid
Aldehydes Methanethiol
Formaldehyde Phenol
Propionaldehyde
Acid
Formic acid
Peroxides
Acetyl peroxide
Peracetic acid
Epoxides
Propylene oxide
Cyclohexene oxide
a Derived from Kotin and Falk.449
-------
176 PARTICULATE POLYCYCLIC ORGANIC MATTER
pyrene by means of carbon and carrier particles613'616 and hematite650
has also been shown.
The rapid clearance of fine particles of polycyclic aromatic hydro-
carbons from the lower respiratory tract has already been mentioned.
In animal experiments, the major mechanism operative shortly after
inhalation or intratracheal instillation is probably ciliary action. Phago-
cytosis by macrophages, however, must play a role, and this process
becomes more important when the polycyclic aromatic hydrocarbons
are adsorbed on relatively inert carrier particles, such as those of car-
bon or hematite. The phagocytes, too, can be moved by ciliary action
or can travel through the tissues into lymphatics to regional and ulti-
mately more distant lymph nodes or into the bloodstream. The material
can also be released if it remains intact and can then become subject
to phagocytosis again.
A process of interest is the "leaching" of polycyclic aromatic hydro-
carbons. This was inferred from analysis of human tissues by Falk,
Kotin, and Markul.258 They found no benzo[a] pyrene in the residue
of soot within lymph nodes that, on the basis of the composition of
soot in the atmosphere, would be expected to contain a considerable
percentage of this carcinogen. The fate of the leached material is not
known. There did not appear to be a correlation between the amount
of soot and the presence of metaplastic or neoplastic change. Further
studies on this important subject are warranted.
Modifiers of the Reaction at the Target Site
How such gases as ozone and sulfur dioxide potentiate the effects of
carcinogenic polycyclic aromatic hydrocarbons in the lung is largely
unknown, and it is not clear how tissues altered by disease may enter
a "precancerous" state. Metaplastic changes and continuing regenera-
tive activity are common to injured and diseased tissue. One interpreta-
tion is that cells in mitosis are more susceptible to carcinogenic agents.
A well-known example is the frequent occurrence of carcinoma of the
skin in burn scars.
More relevant to the problem under consideration, especially be-
cause the observations were made on human lungs, is the demonstra-
tion of the relation between chronic interstitial pneumonia, honey-
combing, and cancer of the lung. Honeycombing is the revision of
pulmonary architecture that takes place in the healing phases of inter-
stitial pneumonia where loss of alveoli and interstitial fibrosis accom-
panied by hyperplasia of smooth muscle and by cellular infiltration
-------
Characteristics of Human Disease Related to POM 177
have taken place simultaneously. On the thick walls of the labyrinths
that remain, there is ingrowth of epithelium that can be astonishingly
hyperplastic and even metaplastic. In a series of 153 consecutive re-
sected lung tumors, some 22% were associated with honeycombing
and atypical proliferation.530 In many of these specimens, transitions
could be traced from the latter to obvious neoplastic change. Of the
associated tumors, 83% arose in the periphery of the lung, and there
was radiographic as well as anatomic evidence of such origin in some
cases. A peripheral lung carcinoma has its center of mass clearly beyond
a segmental bronchus. All the patients who had both honeycombing
and carcinoma were men, and all for whom the relevant information
was available were cigarette smokers. Of persons with honeycombing,
58% had prior histories indicative of pneumonia 5 or more years be-
fore developing pulmonary carcinoma, whereas only approximately
one third of patients with other lung cancers had such a history. Of
the other patients in this series with pulmonary reactions, 89% were
male and 16% were nonsmokers. In a control necropsy series, the inci-
dence of honeycombing was 4.7%. Among a total of 403 control patients,
86% of patients with this lesion were men. Four of 19 persons with
honeycombing (21%) also had cancer of the lung; all were men.
In the relation of lung cancer to honeycombing, cigarette smoking
and sex were associated factors. Focal interstitial pneumonia with
honeycombing is not invariably present in smokers, nor is it confined
to males. Indeed, a high incidence of peripheral tumors has been noted
in lungs with honeycombing.36 These studies illustrate that many fac-
tors can enter into pulmonary carcinogenesis in man and may be as
relevant to consideration of the effects of atmospheric pollutants as
they are to consideration of the effects of smoking.
Problems of Drawing Inferences for Man from Animal Data
Numerous pitfalls are inherent in any attempt to extrapolate data
bearing on dose-response relations from organ to organ or from species
to species. Thus, dosage data, insofar as they can be calculated from
skin painting or injection experiments241 and applications to cervical
epithelium,72 are inapplicable to the tumorigenic dose for the lung.
Some of the external factors that modulate interspecies variation are
age, sex, size, hormonal status, state of health, and type and extent of
supportive treatment. Rail620 has cited examples indicating that pre-
dictions as to effective dose of drugs are sometimes in proportion to
body surface area for a number of diverse species. There is no evidence
-------
178 PARTICIPATE POLYCYCLIC ORGANIC MATTER
that this applies to carcinogenesis. Even among individuals of a single
species, the presence of disease may alter (usually potentiate) the
development of tumors.
Considering the numerous variables that are known and anticipating
that some may be totally unknown, it is best to focus attention largely
on the target tissue, the lung, and to consider the air as the primary
route of entry of carcinogenic agents. This is not to deny the possibility
that an additional moiety of the same or a complementary factor might
reach the pulmonary tissue through some other route, especially the
bloodstream.
The initial reaction occurs between the agent and tissue. But complex
interactions involving chemical, physical, physiologic, and pathologic
characteristics must also be considered. Cocarcinogenesis657 compounds
the problem.
Some of the problems in the interpretation of experimental data
are illustrated by studies of the relation of cigarette smoke to lung
cancer, and these are relevant to investigations designed to determine
the effects of atmospheric polycyclic aromatic hydrocarbons. Auerbach,
Hammond, and associates18'340 have reported experiments in which
beagles were trained to "smoke" through a tracheostomy over periods
of approximately 2!/2 years. Animals exposed to unfiltered smoke in
amounts thought to be comparable with heavy human exposure devel-
oped not only pulmonary lesions interpreted as emphysema by objec-
tive criteria "blindly" applied, bwt also a high incidence of peripheral
bronchiole-alveolar tumors, some of which were interpreted as in-
vasive. The findings were considered dosage-related, inasmuch as animals
smoking filter-tip cigarettes, and therefore estimated to receive approx-
imately half the "tar" and considerably reduced nicotine, developed
less emphysema and fewer tumors, of which only two were localized
squamous lesions of the bronchi interpreted as locally invasive carcino-
mas. This is especially remarkable in that, in all of Auerbach's previous
work in human smokers, stress was laid on squamous metaplasia,
squamous carcinoma in situ, and invasive squamous carcinoma of the major
bronchi. In this respect, the canine lesions are more like those of the
patients with honeycombing in the surgical series studied by Meyer and
Liebow,530 especially because all the patients in that series were male
smokers. It is also noteworthy that at least some of the emphysema and
fibrosis in the dogs would undoubtedly fulfill the criteria of "honey-
combing" as defined in the work on the patients.
The recent work of Auerbach and associates can be criticized on the
following grounds:
-------
Characteristics of Human Disease Related to POM 179
1. The controls were not subject to the same conditions as the
smoking dogs—although they had tracheostomies, they were not made
to "smoke" unlighted cigarettes, and they were not made to stand.
2. The smoking dogs had intercurrent infections, and some were
given antibiotic treatment for these complications.
3. At least two smoking dogs had evidence of aspiration of food,
and another animal had lesions replete with "brown pigment and fat."
4. Two control animals had bronchiole-alveolar tumors at the age
of approximately 5 years, although they must be extremely rare in
dogs at this age.
It may therefore be questioned what part of the effect in these
experiments can be attributed to smoking and what part to other con-
ditions imposed. Possible factors include the lesser degree of cleanliness
of tubing in animals smoking cigarettes without filters and the hyper-
secretion in the smoking dogs. The sequence might be increased secre-
tion in the smokers, with aspiration leading to infection; pulmonary
damage; regenerative changes; and bronchiole-alveolar tumors.
In support of the importance of smoking is the fact that the effect
of smoking given numbers of filter-tip cigarettes was less than that of
smoking equal numbers of ordinary cigarettes. It is possible, therefore,
that the results of these experiments suggest a potentiating effect of
smoking on pulmonary damage related to other causes, as in Meyer and
Liebow's observations in man.
The applicability of these observations to problems of air pol-
lution, with its potential for pulmonary damage, and the potentiating
effects of cigarette smoking can be considered only suggestive at this
time.
It is frustrating, but also perhaps significant, that it has proved more
difficult to induce squamous and undifferentiated tumors in animals
than peripheral "pulmonary adenomas." The latter are least like
tumors in man associated with environmental agents, which are pre-
dominantly squamous and undifferentiated. More attention should
therefore be paid to the experimental squamous tumors and to the
factors in their induction.
Squamous cell carcinomas were produced in rats by the impaction
within bronchi of pellets consisting of carcinogen, either pure or diluted
within cholesterol in concentrations from 0.1% to close to 100%. Dose-
response relations are illustrated in Figure 16-1.474
It has not been possible in experiments reported to date to produce
epidermoid carcinomas in experimental animals given pure carcinogen
-------
PARTICULATE POLYCYCLIC ORGANIC MATTER
100
Log Concentration, %
FIGURE 16-1. Dose-response relations after exposure of the lungs of rats to a
graded series of concentrations of 3-methylcholanthrene and benzo[a] pyrene in
pellets impacted within bronchi. Concentration refers to percentage of carcinogen
in cholesterol carrier. Each pellet weighed 3-5 mg. The ordinate indicates the
percentage of animals developing bronchogenic carcinomas after correction for
early mortality. (Reprinted with permission from Laskin et a/.474)
intratracheally, but such tumors developed when benzo[a] pyrene was
given with the detergent, Tween 60.366
The first squamous lung tumors induced by the intratracheal instilla-
tion of a carcinogen were produced by Pylev614 and Shabad693 in rats
with 9,10-dimethyl-l,2-benzanthracene suspended in balanced saline
solution containing 4% casein and India ink powder. Similar results
were obtained with benzofa] pyrene introduced in a mixture contain-
ing purified carbon particles.613'616 The experiments of Saffiotti,
Cefis, and Kolb650 are especially noteworthy; they produced squamous
tumors in hamsters, which are relatively free of intercurrent pulmo-
nary disease and spontaneously occurring tumors. Benzo[a]pyrene was
administered mixed with equal parts of hematite. The particles were
ground to a size range averaging less than 3 jim in diameter, and 6 mg
of the dust were instilled once a week for 15 successive weeks. Some
-------
Characteristics of Human Disease Related to POM 181
76% of the animals developed tumors, mostly of the squamous type,
but including anaplastic carcinomas and adenocarcinomas.
Of great interest are lung tumors produced by inhalation of vapors
of ozonized gasoline and influenza virus.456 Although the exact com-
position of substances in the ozonized gasoline vapor is unknown,
these experiments suggest a potentiating effect of viral infection in
pulmonary carcinogenesis. However, similar experiments carried out
by Nettesheim et al. with influenza virus had an opposite effect.560
Of greatest relevance are the studies of Laskin, Kuschner, and Drew474
on the combined effect of sulfur dioxide and benzo[a] pyrene. Among
21 rats that received 534 exposures to an atmosphere containing 10
ppm of sulfur dioxide and 494 exposures to an atmosphere contain-
ing 3.5 ppm of sulfur dioxide plus a benzo[a] pyrene concentration of
10 mg/m3, five developed squamous cell carcinoma. Of 21 rats that
received only the 494 treatments of the combination of the lower
concentration of sulfur dioxide plus benzo[a] pyrene, two developed
lung carcinoma. Both the cytologic characteristics and the presence
of renal metastases confirmed the malignancy of the tumors. One
criticism of these experiments is the high incidence of bronchopul-
monary lesions to which infection probably contributed.
The difficulties of extrapolation from animals to man are at once
obvious when an attempt is made to compare the dosage schedule in
the several experiments just summarized with what might be received
in the distal pulmonary parenchyma of man breathing a polluted atmo-
sphere. Let it be assumed that the atmosphere contains benzo[a] pyrene
in a concentration of 30 /zg/1,000 m3, which is within the range at un-
favorable seasons in Birmingham and Detroit, and that this concentra-
tion remains constant through the year.667 If it is assumed that all the
inhaled carcinogen is retained, that the tidal volume is 500 ml, and
that the respiratory rate is 14/min, then it would take about 99 days
to retain 30 pig and 272 years to retain 30 mg. Especially after apply-
ing a correction for the difference in body surface area between ro-
dents and man, it is evident, for example, that the animals in the
experiments of Saffiotti etal.650 on hamsters and Laskin era/.474 on
rats received enormously greater dosages of benzo[a] pyrene adsorbed
on particles and in a much shorter time. Of course, the amount de-
posited in the various subjects would be determined by particle size;
by structural and physiologic factors, including tidal volume and
respiratory rate; by efficiency of the clearance mechanism; and pos-
sibly by the facilitating effects of viral infections or other lower respi-
ratory tract disease. Additional factors are the effects of particles in
-------
182 PARTICULATE POLYCYCLIC ORGANIC MATTER
delaying clearance, the presence of cocarcinogens or other carcinogens
in the atmosphere, and the damaging effects of gases, cigarette smoke,
and other substances. The overwhelming importance of cigarette smoke
is generally accepted.
Conclusions
The gaps in knowledge concerning inhalation carcinogenesis become
evident when it is considered that not even the concentration and
physical state of damaging elements at the human receptor site have
been adequately established.1 There is compelling evidence that carrier
substances are important, but their exact role can only be surmised.
The assumption that lung tumors are the consequence of a single
pollutant is almost certainly wrong, and there is much to support the
idea of synergism or cocarcinogenesis, especially with respect to
cigarette smoking. Pre-existing pulmonary disease can also be a pre-
disposing factor. Thus, the factors for lung cancer must be more com-
plex than suggested by the simplified schema of Table 16-1.
It is probably significant that the successful experimental produc-
tion of pulmonary squamous tumors by intratracheal insufflation or
by inhalation has required the simultaneous introduction of inert
particles or additional pulmonary injury produced by toxic gases
or viruses.
Examination of the dosage of carcinogens used in experiments that
have been successful in producing squamous cell carcinoma in the lungs
of animals makes it obvious that these doses have been much higher
than those to which man is likely to be subjected throughout his life-
time with any known specific agent. It is possible, however, that the
lungs of rats and hamsters are more resistant to carcinogenesis by poly-
cyclic aromatic hydrocarbons.
On the basis of available evidence, the best approach to assessing the
significance of atmospheric pollutants in the etiology of pulmonary
carcinoma in man is the epidemiologic one. Urban-rural differences
undoubtedly offer clues to the problems of pulmonary carcinogenesis
in man. Special attention should be paid to imperfections or contradic-
tions in the correlations of measurements of air pollution or of any
specific suspected agent, such as benzo[a]pyrene, with the incidence
of lung cancer.99'100'112'330'332'400'460'662'667'670-731."2 There is gen.
eral agreement on the increment in the incidence of lung cancer pro-
duced by smoking, although interpretations vary. This is of great
interest, because urban air and cigarette smoke have carcinogenic sub-
-------
Characteristics of Human Disease Related to POM 183
stances and some damaging gases in common.783 However, the com-
plexities of the problem of pulmonary carcinogenesis are compounded,
rather than simplified, by that fact.
Problems for Investigation
The epidemiologic approach remains open; although potentially it is
highly significant, it must be used with utmost caution to avoid post
hoc ergo propter hoc reasoning. The well-established urban-rural dif-
ferences in the prevalence of lung cancer must harbor important etio-
logic clues and are worthy of the most thorough investigation and the
most circumspect analysis. In addition, a possibly fruitful epidemio-
logic approach would be to compare disease in locations where there
are extremes of photochemical pollutants or extremes of polycyclic
aromatic hydrocarbons. In all such studies, the most careful attention
must be paid to the effects of cigarette smoking in quantitative terms.
Further studies of the composition both of the atmosphere and of
cigarette smoke, with a search for similarities and differences, should
be carried out. Exposure factors are incompletely known and have
great relevance.
Base-line data with respect to some pollutants could be obtained for
man by analysis of tissues of human beings of past centuries. Accurate
dating is possible, because clothing, coins, etc., often remain intact.
Human physiologic characteristics that have a bearing on pulmonary
carcinogenesis have been incompletely studied. An example is the fate
and clearance of particles from the respiratory tract. The effects of
some gases can also be investigated. Furthermore, advantage should be
taken of known industrial exposures and accidents that might have
relevance to the problem.
Further investigations of experimental models of pulmonary car-
cinogenesis that appear to be most relevant to the human disease are
indicated. These should be directed mainly toward inhalational exposure
to carcinogenic polycyclic aromatic hydrocarbons known to be present
in the atmosphere, with due attention to quantitative factors and to
adequacy of controls. The following are to be considered in the design
of such experiments:
1. The similarity of the experimental species to man,
2. Degrees of exposure in urban atmospheres (this is not to deny
the use of greater exposures as a first approach),
3. The high probability that lung cancer is a multifactorial disease
-------
184 PARTICULATE POLYCYCLIC ORGANIC MATTER
and that the most significant factor is cigarette smoke (experiments
designed to investigate interacting factors must be planned in a manner
permitting analysis), and
4. The use of germfree animals, once a suggestive model has been
established (this is desirable, because the role of infection is problemat-
ical and difficult to control in ordinary circumstances).
CHRONIC BRONCHITIS AND EMPHYSEMA
Chronic bronchitis and emphysema are two pulmonary diseases that
should be considered when evaluating the human health effect of urban
pollution.
The etiology of chronic bronchitis is not known,269'303'629'645 and
the disease has not been produced experimentally in animals by irritants,
at least in its severe stages, in which substantial airway obstruction
occurs. POM cannot be excluded with certainty as an etiologic factor
in chronic bronchitis, particularly because it might be associated with
disease when combined with other atmospheric pollutants. The re-
sponse of humans to complex mixtures of pollutants in which POM is
merely one component is not known.
The belief that exposure to POM may be of etiologic importance
in chronic pulmonary disease is based on two epidemiologic observa-
tions: Workers producing coal gas in England had an increased incidence
of both chronic bronchitis and lung cancer, compared with the general
population, but the increase in the incidence of chronic bronchitis
was considerably greater than that for lung cancer and these workers
had enormous exposures to POM, particularly benzo[a]pyrene; and
an urban-rural gradient for lung cancer and chronic bronchitis parallels
that for POM. Of course, it has been demonstrated that benzo[a] -
pyrene, a known carcinogen present in fossil fuels, is not particularly
irritating to the normal lungs of experimental animals, except at very
high dosages.269'303'629'645
Chronic bronchitis and emphysema are different forms of cumula-
tive injury that become overt disease when the damage reaches a level
sufficient to cause appreciable disability. The severity of the preclini-
cal stages of a disease like emphysema can be shown by autopsy
studies to correlate well with levels of urban pollution.415 However,
the role of POM cannot be ascertained from these data.
Stocks730 reported on data collected from 26 localities in northern
England and Wales. Standardized mortality for chronic bronchitis and
pneumonia in 1950-1953 was correlated with annual concentrations
-------
Characteristics of Human Disease Related to POM 185
of smoke, benzofa] pyrene, benzo[ghi] perylene, pyrene, and fluoran-
thene. A statistical process of successive elimination was applied to
discover which hydrocarbon was responsible for the demonstrated cor-
relation of mortality rates with smoke concentrations. For lung cancer
and chronic bronchitis, ben/o[a] pyrene was found to be the sub-
stance of prime importance. For pneumonia, benzo[a] pyrene was not
important. This seems to cast doubt on the significance of benzo[a]-
pyrene in relation to bronchitis mortality.
In a later report, Stocks729 correlated lung cancer and chronic
bronchitis mortality with consumption of cigarettes and solid and
liquid fuel in 20 European countries. The consumption of solid or
liquid fuel did not appear to be as important as cigarette smoking in
bronchitis mortality rates in these countries.
The evidence indicates that POM concentrations in polluted urban
air do not significantly influence the pathogenesis or outcome of non-
neoplastic lung diseases, such as emphysema and chronic bronchitis.
However, the total urban pollution content appears to be a significant
factor in disease development.
SKIN DISORDERS
Although it is apparent that, in occupational exposures in man, the
most common pathologic response to high dosages of POM is cutane-
ous neoplasia, various nonneoplastic skin responses have been observed.
They include nonallergic dermatitis, cell-mediated hypersensitivity
(allergic contact dermatitis), phototoxic and photoallergic reactions,
pilosebaceous responses (such as folliculitis and acne), and pigment
changes (such as hypermelanosis and hypomelanosis). It should be
clearly understood that these pathologic responses in man have not
been reported as related to community airborne PO M, but to the use
of PO M in work or at home.
Nonallergic Dermatitis
Nonallergic dermatitis in man caused by materials containing poly-
cyclic aromatic hydrocarbons is reportedly associated with the same
kinds of work exposures that may produce skin cancer. The materials
include derivatives of fossil fuels, such as coal tar, pitch, creosote,
and asphalt from coal; paraffin distillates, high-boiling petroleum
residues, asphalt, and lubricating, cutting, and coolant oils from petro-
leum; and shale oil.683 (pp- 300~335) Many polycyclic aromatic car-
-------
186 PARTICULATE POLYCYCLIC ORGANIC MATTER
cinogens, such as benzo[a]pyrene and 7,12-dimethylbenz[a]anthracene,
are primary irritants for animal skin in concentrations as low as 1% in
equal parts of acetone and olive oil or 0.5% in ethanol or in pharma-
ceutical-grade white mineral oil (R. R. Suskind, personal commu-
nication).
The inflammatory response of the skin to irritants and antigenic
agents in man is known as an eczematous reaction. It is characterized
clinically by erythema, swelling, and vesiculation. One cannot dis-
tinguish histologically between skin reactions provoked by a chemical
irritant and by a sensitizer. In man, the histologic features of the acute
process (acute contact dermatitis) are intercellular and intracellular
epidermal edema, which may lead to vesiculation or blister formation;
vascular dilatation in the upper dermis; and edema of the dermis. The
cellular infiltrate is usually composed of neutrophils and lymphocytes.
In chronic contact dermatitis, the epidermis is thickened and there is
elongation of the epidermal rete ridges and a marked thickening of
the protective horny layer (hyperkeratosis and acanthosis). Some micro-
scopic vesicles may be present in the subacute process but as a rule are
absent if the problem is longstanding. The upper dermis may contain
a moderate to large number of cells that are predominantly lympho-
cytes. Histocytes, fibroblasts, and eosinophils may also be found. The
infiltrate is usually perivascular. Neutrophils are rare. The number of
capillaries seen in sections may be increased and the walls of arterioles
thickened.
The skin response in mice and guinea pigs with single or repeated
exposure to benzo[a] pyrene is characterized by an inflammatory
response in which erythema and edema are primary events. This is fol-
lowed by epidermal and some dermal necrosis, hair loss, and depigmen-
tation.
Cell-Mediated Hypersensitivity
When the skin of guinea pigs and mice is exposed to a carcinogen, such
as benzofa] pyrene, a significant degree of immunoblast (pyronino-
philic cell) response in regional nodes can be induced.274 The im-
munoblast response is a characteristic primary event when a mam-
malian host is exposed to antigenic molecules. When guinea pigs
that have been repeatedly exposed to benzo[a] pyrene are challenged
with appropriate low concentrations (0.001-0.005% in ethanol),
a delayed hypersensitivity reaction of the skin-allergic contact derma-
titis-is produced (R. R. Suskind, personal communication). The
-------
Characteristics of Human Disease Related to POM 187
clinical response in the skin in 24-48 hr is characterized by erythema
and edema that persists longer than a primary irritant response. Al-
lergic contact dermatitis in man has been reported from therapeutic
coal-tar preparations, but it is rare. The list of polycyclic organic
materials known to sensitize after skin contact is sizable, e.g., anthra-
quinone and its derivatives, bisphenols (such as bithional and hexa-
chlorophene), tetrabromfluorescein, tetraiodofluorescein, mercaptoben-
zothiazole, (3-naphthol, a-naphthylamine, phenothiazines, phthalic
anhydride, rhodamine, rotenone, and halogenated salicylanilides
(such as tetrachlorsalicylanilide). However, these are rarely if ever
found as community air pollutants.
Cutaneous Photosensitization
Exposure to POM in the presence of solar radiation or ultraviolet radia-
tion from other sources may produce an inflammatory skin response
in man. There are essentially two types: Phototoxic reactions are dose-
dependent; no cell-mediated hypersensitivity state prevails; and clin-
ically they present as exaggerated sunburn. Photoallergic reactions are
not dose-dependent; a cell-mediated hypersensitivity mechanism is
involved; and clinically they take the form of eczematous allergic con-
tact dermatitis. The antigenic agent may be the original airborne ma-
terial, an ultraviolet-mediated degradation product, or a physiologic
metabolite. Examples of naturally occurring phototoxic agents are
furocoumarins like 8-methoxypsoralen, found in celery rot, and 5-
methoxypsoralen, found in oil of bergamot,585 a common ingredient
of perfumes and scents. The furocoumarins are the phototoxic agents
in such plants as cow parsnip, St. John's wort, mustard, and figs. Con-
tact with these plants may produce photodermatitis. Aerosol solvents,
such as methylated naphthalenes, used in insecticide application, are
notorious for their phototoxic potential.221
In the production and preparation of some drugs in industry, hos-
pitals, clinics, etc., contact with the dust or solutions of the drugs may
produce photodermatitis. Polycyclic organic compounds that are
photosensitizers include phenothiazine derivatives, such as chlorproma-
zine and promethazine; hydrochlorthiazide; fluorescent dyes, such as
eosin and trypaflavine; and antibiotics, such as demethylchlortetra-
cycline.
In industrial uses, components of pitch, coal tar, and creosote, to
which roadbuilders, roofers, gas workers, coke-oven workers, etc., are
exposed, may produce a phototoxic reaction that is characterized by
-------
188 PARTICIPATE POLYCYCLIC ORGANIC MATTER
a short induction period (a few hours) and the appearance of an exag-
gerated sunburn. The redness and scaling subside after removal from
exposure. After decline of the inflammation, hypermelanosis is often
observed. The photoallergic dermatitis cannot be distinguished histo-
logically from allergic contact dermatitis. Photosensitive eruptions are
usually limited to the sun- or ultraviolet-exposed areas of the skin-
commonly the face and hands.
Pilosebaceous Reactions
Some types of PO M and mixtures in which they are found are
known to induce changes predominantly in hair follicles and
sebaceous glands. When the inflammatory reaction is limited to or
around the follicle, the skin problem is called "folliculitis"; when
the entire pilosebaceous apparatus is involved, it is known as "acne."
Acne is characterized primarily by lesions involving the pilose-
baceous apparatus of the skin, such as comedones, milia, cysts, nodules,
follicular inflammation, erythematous papules, excessive oiliness, pus-
tules, and abscesses. Although the problem is usually associated with
puberty (acne vulgaris), such lesions may be provoked in adults by a va-
riety of chemical agents. Among the known causative substances are
components of crude petroleum, cutting oils, coal tar and some of its
products, and chlorinated aromatic compounds, such as chlorinated
naphthalenes, chlorinated diphenyls, and chlorinated diphenyloxides
like "dioxin." The latter is a common name for a group of potent acne-
genie compounds that are chlorinated dibenzodioxins, for example,
2,3,7,8-tetrachlorodibenzo-p-dioxin.
Occupational acne can also occur in persons involved in the man-
ufacture of inorganic and organic chlorinated compounds in which
acnegenic substances are produced, either inadvertently in small
amounts or as intermediate compounds.738 These may gain access to
the skin through the air or by direct handling.
PETROLEUM AND ITS DERIVATIVES
Workers in oil fields and in refineries who may have prolonged
skin contact with either the crude oil or the heavier oil fractions may
develop acneiform lesions, folliculitis, or both in the exposed areas.738
Weeks or months of contact with the material in relatively unhygienic
conditions are usually necessary. These conditions are now uncommon
among oil-field and refinery workers in this country.
-------
Characteristics of Human Disease Related to POM 189
By far the most common sources of acneiform skin eruptions
and folliculitis are the cutting oils used in machine-tool operations,
such as cutting, grinding, milling, boring, and honing. The acnegenic
oil is carried to the skin in mist, in aerosol form, or by direct contact.
COAL-TAR PRODUCTS
Persons who maintain unhygienic contact with coal-tar oils and pitch-
such as coal-tar plant workers and handlers of roofing, roadbuilding,
and construction materials like pitch and creosote-may develop acne,
folliculitis, or both.738 Because some of these materials also photo-
sensitize the skin, exposure to sunlight may provoke photochemical
dermatoses, including exaggerated sunburn and melanosis.
Other types of dermatoses that may be provoked by petroleum
and coal-tar fractions are contact dermatitis (nonallergic dermatitis or
allergic sensitization), papillomas, keratoses, and cancer.
CHLORINATED HYDROCARBON COMPOUNDS
The chlorinated aromatic compounds are among the most potent
acnegenic materials. The chlorinated naphthalenes, diphenyls, and
diphenyloxides have unusual dielectric and flameproofing proper-
ties and are used as electric wire and cable insulations, as well as
condenser dielectrics. Mixtures of the chlorinated naphthalenes and
diphenyls are called "halowaxes." These hydrocarbons, with three or
more chlorine atoms substituted for hydrogen atoms, are acnegenic.
Dioxin was found to be responsible for outbreaks of severe acne in
U.S. and West German plants manufacturing trichlorophenoxyacetic
acid, a widely used weed killer.738
PATHOGENESIS AND PATHOLOGY
As in acne vulgaris, the primary cellular response in chemically induced
acne is proliferation of the follicular epithelium, which lines the se-
baceous duct and follicle pore. Hyperkeratinization of the duct and
pore cells results in plugging of the orifice, which prevents normal
extrusion of sebum. A modulation of the sebaceous cells occurs. The
lipid-bearing cells of the sebaceous gland appear to be replaced by
keratinizing cells, and the process later produces a cyst or sac filled
with keratin and retained sebaceous lipid. These are the events that
occur in the evolution of comedones and keratin cysts.738
-------
190 PARTICULATE POLYCYCLIC ORGANIC MATTER
Any of the three categories of hazardous materials already discussed
may produce plugging, cyst formation, folliculitis, furuncles, and ab-
scesses; and it is not possible to distinguish between the reaction pat-
terns of the different causative agents. It is generally agreed, however,
that petroleum products that produce comedones and cysts will also
provoke more severe and widespread follicular inflammatory reaction
and furunculosis than will coal-tar products. In the latter case, increased
melanin formation is seen much more often than in response to petro-
leum products, and the retention cysts are usually smaller than in
other types of acne. Melanosis is also seen in acne provoked by chlo-
rinated hydrocarbons.
Pigment Disturbances
Two types of pigment reactions to POM are possible: hyperpigmenta-
tion and hypopigmentation.
Most of the agents that induce hyperpigmentation and contain POM
are photosensitizing. They include coal-tar products, low-boiling petro-
leum fractions (such as methylated naphthalenes used in insecticide
sprays221), essential oils from plant sources containing furocoumarins,
and such dyes as tetrabromfluorescein. Again, ultraviolet radiation is a
critical factor, and the chemical agent enhances pigment darkening
and pigment synthesis. Most of the phototoxic agents will induce
hypermelanosis of the skin.
Decrease in the color of the skin may result from damage to the
melanocytes or interference with melanin biosynthesis or maintenance.
Hence, severe irritation or chemical burns may result in temporary or
permanent loss of pigment as a result of cell death. There are no re-
corded incidents of a polycyclic aromatic hydrocarbon inducing depig-
mentation or hypopigmentation on a biochemical basis, as in the case
of the reaction of skin to the monobenzyl ether of hydroquinone or
the reaction of hair to mephenesin carbamate, a muscle relaxant.
Persons who have chronic photodermatitis in which the inflammatory
component is severe and prolonged may develop irreversible pigment
loss (leukoderma). This is the consequence of decreasing the popula-
tion of melanocytes by cell death.
-------
17
Clinical and Epidemiologic
Studies
OCCUPATIONAL SKIN EFFECTS
Most of the data regarding cutaneous effects in man of exposure
to POM are found in reports of occupational incidents. There is no
documentation that particulate materials containing POM in com-
munity air have caused any adverse skin effects.
In the occupational problems that have been described, the hazard-
ous material reaches the skin either by direct contact or as an aerosol,
dust, or mist. Studies of industrial exposures in which the skin is af-
fected do not attempt to differentiate exposures by air from exposures
by other means. This aspect should be considered carefully in attempt-
ing to extrapolate the information derived for occupational problems
to the possible hazards of community air pollution.
Numerous materials are recognized as carcinogenic for man, and
polycyclic aromatic hydrocarbons have been identified in some of
these. It is likely, however, that many of the actual carcinogens con-
tained in combustion and distillation products of carbonaceous sub-
stances are either identical or similar. Most of these are polycyclic
aromatic hydrocarbons of the benzo[a] pyrene and benzanthracene
types. The few studies in which clinical evaluations have been cor-
191
-------
192 PARTICIPATE POLYCYCLIC ORGANIC MATTER
related with chemical analyses support the belief that the carcinogens
in question are polycyclic aromatic hydrocarbons.
Hendricks et al.358 showed that cancer of the scrotum occurred in
wax pressmen but affected only workers with extensive, continuous,
or prolonged exposure to slack or crude wax containing high concen-
trations of polycyclic aromatic hydrocarbons. Workers exposed to
finished waxes that were low in aromatics did not develop cancer.
W. E. Smith et al.710 confirmed these epidemiologic suspicions when
they demonstrated that only the aromatic portions of the crude wax
were carcinogenic in mice. A convincing discussion of evidence in-
criminating benzo[a] pyrene as a carcinogen for man has been presented
by Falk et al.259 However, many studies, especially in earlier years,
related only such complex chemical mixtures as pitch and tar to the
cancer problem, and it is only on circumstantial evidence that the
polycyclic aromatics can be suspected of having been the actual car-
cinogens in those instances.
A variety of factors influence the induction of cutaneous cancer in
man, including degree or level of exposure, concentration of carcino-
gen, and duration of exposure; factors affecting absorption of car-
cinogen or cocarcinogen, such as vehicle, presence of surfactant, and
concomitant or prior physical or chemical injury to epidermal barrier;
factors affecting the carcinogenic activity of material on target tissue,
such as chemical or physical cocarcinogens, long-chain hydrocarbons,
ultraviolet radiation, and ionizing radiation; and genetic pigmentation,
the primary factor in determining skin cancer as related to exposure
to ultraviolet radiation.
Pitch, Tar, and Asphalt
Cyclic hydrocarbons are found in so-called high-temperature tar dis-
tillation, which is carried out above 370 C. The use of horizontal re-
torts is said to increase the amount of polycyclic aromatic hydro-
carbons by a factor of 4. Skin carcinomas are often observed in occupa-
tions in which "high-temperature tar" is the exposure material, and they
are not considered to be associated with low-temperature tar fractions,
which do not contain polycyclic hydrocarbons.402
Cancer associated with these products has been recognized since
1876, when Volkmann described skin cancer in a worker employed in
a Saxony tar distillery. It has since been the subject of many reports,
including those by Bridge and Henry,89 Haagensen,326 Carozzi,118
Staemmler,719 and Uytdenhoef.778
-------
Clinical and Epidemiologic Studies 193
Henry360 analyzed the official data issued by H. M. Chief Inspector
of British Factories from 1920 through 1945. Pitch, tar, and tar prod-
ucts were responsible for 2,229 (59.4%) of the 3,753 cases reported.
Shale oil, mineral oil, and bitumen were associated with 1,515 cases
(40.3%). The occupations in which skin cancers due to pitch, tar, or
tar products were reported were diverse. Most of the cases, however,
were among persons occupied in tar distilling (538), briquette man-
ufacture (364), coal-gas manufacture (305), and pitch loading (36).
Neve562 observed more than 2,000 kangri cancers in Kashmir. The
kangri is an earthenware bowl 6 in. or more in diameter held against
the skin as a personal heating appliance. The bowl is filled with hot
wooden embers and cooled by sprinkling with water. The tempera-
ture may reach 150 F. It has been maintained that the cancers are due
to repeated burns,640 but it is likely that the carcinogenic agent is a
tarry distillation product of woodcoal whose action may be increased
by small burns402 or continuous exposure to soot. Similar factors
may be associated with skin carcinomas related to the use of the kairo
in Japan.401
Fractionation and Distillation Products of Oils
Occupational cancers due to contact with fractionation and distillation
products of oil have been recognized for many years. Some oils and
their products appear to be more carcinogenic than others. In general,
the data from epidemiologic surveys and animal experiments are cor-
related with the presence or use of poly cyclic aromatics.58 Other
factors play a significant role in carcinogenesis; for example, the con-
centrations of accelerators and cocarcinogens influence the rate of
tumor development in shale-oil refinery exposures.
The hazards to health previously found in such persons as petro-
leum refinery and shale-oil workers, cotton mule spinners, and ma-
chinists exposed to cutting oils can be eliminated by good hygienic
practices.
Skin cancer in the shale-oil industry has been studied in Scotland,
where Bell43 first recorded the occurrence of two cases of scrotal
cancer among paraffin pressmen. In the previous year, Volkmann had
described three skin cancer cases among workers distilling paraffin
wax from lignite (brown coal). Scott684 reported that 89 cases of
epithelioma had been observed in a Scottish shale refinery from 1900
to 1928 among workmen employed as wax pressmen or in the distilla-
tion of shale oil. He calculated that, inasmuch as the oil company
-------
194 PARTICULATE POLYCYCLIC ORGANIC MATTER
employed 5,000 men, the cancer incidence was 2% over the 28 years
surveyed. However, because most of those workers had no prolonged
or significant exposures, the actual risk among the exposed was con-
siderably higher.
Henry,360 in his analysis of cases of occupational skin cancer of-
ficially recorded in Great Britain from 1920 to 1945, cited 52 cases of
skin cancer in 42 workmen engaged in oil refining. In all but four of
these cases, the contact had been with shale oils. Most of the cancers
were on the exposed hands and forearms, but 30% occurred on the
scrotum.
Twort and Twort759 treated mice with oils from various sources.
They found Scottish shale oil to be the most potent carcinogen, the
unfinished lubricating fraction being more carcinogenic than the
finished oils or the crudes. They felt that the carcinogenicity could be
removed by treatment with sulfuric acid purification. Schwartz et
a/683 (pp. 726-737) refer to ^g observation that shale oil is the most
carcinogenic and cite evidence that the incidence with various oils is
inconsistent with the assumption that benzenoid hydrocarbons are
carcinogenic and are almost all destroyed by treatment with sulfuric
acid, by oxidation and reduction processes, or by refining. Bingham
and Horton57 found that, when mice were painted with various crude
or partially refined oil fractions, carcinogenesis was correlated best
with the chromatographic mass spectrometric analyses of four-
and five-ring aromatic compounds.
MULE SPINNING
Some of the most dramatic associations between lubricating oils and
skin epitheliomas were noted in the 1920's among mule spinners* in
the cotton industry in Great Britain. The relation was first estab-
lished by Southam and Wilson,715 who analyzed 141 cases of
scrotal carcinoma observed at the Manchester Royal Infirmary over
a 20-year period. Sixty-nine of the cases were in mule spinners in
local cotton mills. Leitch486 showed that 20% of the fatal cases of
scrotal carcinoma in England occurred in mule spinners, an industry
employing approximately 23,000 men. He found a yearly average of
11.8 cases of fatal cancer, or 50 fatalities per 100,000 spinners
*A mule spinner in the textile industry is an operator of a mule, a spinning machine that
makes thread or yarn from fibers. Lubricant oils used in these machines were derived
from oil shale and petroleum. The clothes of the spinner, which were changed infrequently
became saturated with oil from the oil mist, from the oily surface of the machine, and
from spillage during maintenance.
-------
Clinical and Epidemiologic Studies 195
employed per year. Southam714 calculated the incidence of scrotal
cancer at 250 cases per 100,000 spinners per year, and estimated that
at that time 50 new cases of scrotal carcinoma were seen each year.
Henry360 found that, of 3,753 cases of cutaneous carcinoma noti-
fied as occupational in Great Britain from 1920 to 1945, 1,389 oc-
curred in the cotton industry. Of these, 1,296 (93.3%) were in persons
who worked in the mule-spinning room. Another 48 persons who had
been classified as having other occupations had been mule spinners at
some time (however short) in their lives. Of the skin cancers in the
cotton industry, 28.8% were on exposed surfaces, 10.7% on covered
sites other than the scrotum, and 60.4% on the scrotum. In all cases,
the carcinomas in the cotton industry were believed due to contact
with mineral oils, presumably containing POM.
Skin cancers in mule spinners have rarely been reported outside the
British Isles. Until Heller357 studied the incidence of scrotal cancer
among mule spinners in the United States in the 1920's, the disease
had been reported only in immigrants. By studying hospital records
and death certificates and questioning plant physicians, he was able
to find records of only two cases of scrotal cancer in mule spinners
who had not been employed in this occupation outside the country.
He concluded that the incidence was insignificant in mule spinners
in the United States and attributed this to the "refined character
of the oil" used. Since 1953 in Great Britain, the use of noncar-
cinogenic oil, with a reduced polycyclic aromatic hydrocarbon con-
tent, has been obligatory under the Mule Spinning (Health) Special
Regulation.268 Only seven cases of skin cancer among mule spinners
were reported since that time.15 In 1966, one death from scrotal can-
cer was reported; the victim was a man who had been occupationally
exposed before 1953.16
PETROLEUM REFINING
Heller357 described 20 cases of cancer caused by industrial mineral
oils in the United States. Eight were in employees of a refining
company using crude oils from Illinois and Indiana. Eleven were
gleaned from the records of the Memorial Hospital, New York,
and the New York Skin & Cancer Hospital. Eight tumors were on
the scrotum, and all except one of the others were on exposed parts.
He was unable to document cancers in workers handling refined lu-
bricating oils.
Hendricks et a/.358 and Lione and Denholm497 studied the inci-
-------
196 PARTICULATE POLYCYCLIC ORGANIC MATTER
dence of cancer among wax pressmen in one refinery from 1937 to
1957. Eleven cases of scrotal cancer were recorded, the incidence
among those with 10 or more years of service being many times that
among the general male population, although the incidences of other
forms of cancer were not increased. These cases were directly asso-
ciated with extensive, continuous, and prolonged exposure to slack
or crude wax. Workers exposed to finished wax had no scrotal car-
cinoma. Only the pressmen were in contact with the aromatic oils.
These data were consistent with the work of Smith et a/.,710 who
demonstrated that only the aromatic portions of the crude wax were
carcinogenic in animals.
LUBRICATING AND CUTTING OILS
Cruickshank and Squire174 studied British workers exposed to mineral
oils in the engineering industries. Of 138 workers using cutting oils,
60% of those exposed for more than 15 years had multiple hyperkera-
toses on their hands and one had a scrotal cancer. They later investi-
gated the records of scrotal carcinoma in the United Birmingham
Hospitals between 1939 and 1948. Thirty-four cases had occurred—12
in those exposed to oil in the engineering industry, 13 in workers ex-
posed to tar pitch, etc., and nine that could not be allocated to a
definite etiologic association.
Cancer of the hands and forearms in those attending the United
Birmingham Hospitals between 1941 and 1950 was studied by Cruick-
shank and Gourevitch.m Of 44 patients, 18 gave a history of occupa-
tional exposure to various oils and six of exposure to pitch, and three
were in other occupations.
Mastromatteo521 described six cases of squamous cell carcinoma in
a single Canadian plant employing just over 1,000 workers. Five of the
six cases occurred in machine-tool operators and were attributed to the
carcinogenicity of cutting oils.
CREOSOTE
Creosote, a fractionation product of tar, has been associated with a
number of reported cases of occupational cancer. The actual car-
cinogen has not been identified in the case reports, but this material
contains aromatic hydrocarbons.
Reports of groups of cases include those of O'Donovan569 (epi-
theliomas in four timber picklers) and of Bridge and Henry89 (cancer
in four timber picklers and six brick tile or pipe pressers). Henry360
-------
Clinical and Epidemiologic Studies 197
reported cases between 1920 and 1945 in 34 workmen, of whom 14
were employed in creosoting timber, eight in creosote storage, 11 at
brick or pottery presses, and one in the manufacture of a creosote dis-
infectant.
ANTHRACENE
Anthracene is a tricyclic aromatic obtained from the crude oil of coal
tar. A number of reports have described cases of cancer due to "anthra-
cene." Leymann in 1917495 noted that an annual report issued in 1902
by chemical plants in Oppeln, Silesia, recorded the occurrence of a
variety of skin lesions in 22 of 30 workers in an anthracene plant. Three
had undergone operations for scrotal cancers. Bridge and Henry89 men-
tion four cases of epithelioma in anthracene workers. Henry360 notes
the occurrence of five cases of epithelioma in workers at synthetic dye-
works engaged either in the purification or (in one case) the loading
of boxes of anthracene. Repeated exposure of susceptible mice and
rats to anthracene fails to produce skin cancer. Because it is generally
held that man is less susceptible than rodents to polycyclic aromatic
hydrocarbons, it is unlikely that anthracene itself is the carcinogen in
anthracene plants; a product, more closely related to benzo[a] pyrene,
is probably responsible.
SOOT
Percivall Pott described cancer of the scrotum in chimney sweeps in
1775. By 1788, laws regulating activities in the chimney sweep trade
had been passed in England. Henry359 cites an investigation of 1,631
cases of scrotal cancer, of which 121 were in chimney sweeps, 125 in
metal workers, and 575 in textile workers. It is difficult to derive actual
figures of incidence from these early reports, but Henry calculated the
incidence of scrotal carcinoma from death certificates as 754.7 per
million. Surprisingly, the Annual Report of H. M. Chief Inspector of
Factories in 196415 noted that scrotal cancer in chimney sweeps was
still occurring; the deaths of five former chimney sweeps due to scrotal
cancer were reported between 1962 and 1964.
Factors That Influence POM Carcinogenesis in Man
Animal experiments have demonstrated that many factors can influence
the susceptibility of the skin to specific identifiable carcinogens, such
as benzo[a] pyrene and 7,12-dimethylbenz[a] anthracene. The prime
-------
198 PARTICULATE POLYCYCLIC ORGANIC MATTER
factor of chemically induced cancer in man is the degree of exposure
(how much and for how long), which depends, e.g., on the concentra-
tion of chemical carcinogen and the aggregate duration of exposure.
Other factors can be regarded as cocarcinogenic.
Salaman and Roe657 list the factors that may be considered as cocar-
cinogenic in various circumstances. When a carcinogen is applied to the
skin, the vehicle may determine the rate of absorption and the amount
absorbed. Factors that block detoxification or excretion of the car-
cinogen may also increase carcinogenesis. Altered physiologic states,
such as hyperemia of the skin or subcutaneous tissue, may increase
carcinogenesis, and agents that induce these changes may be classified
as cocarcinogens. In experimental animals, carcinogenesis may be
slowed or even inhibited by such conditions as dietary restrictions,
stress, or altered hormonal balance.
CHEMICAL COCARCINOGENS
Suskind and Horton739 discuss the influence of chemical accelerators
like tt-dodecane, a straight-chain hydrocarbon found in petroleum prod-
ucts. By applying a single dose of a strong carcinogen to mice and then
applying dodecane, Horton et al.396 were able both to decrease the
latent time before onset of tumor formation and to increase tenfold
the percentage of mice ultimately developing tumors.
The carcinogenic properties of combustion and distillation products
may depend also on other aromatic substances, such as epoxides. An
excellent review of animal experiments demonstrating the carcinoge-
nicity of epoxides has been written by Van Duuren.780
The knowledge derived from experiments about the carcinogenic
potency of petroleum and distillates on the skin has been summarized
by Bingham and Horton.57 A carcinogen, most probably a four- or five-
ring polycyclic aromatic hydrocarbon, must be present. In occupational
exposures, fractions not containing these compounds have not been
shown to be carcinogenic. The potency of a fraction may be influenced
to a great extent, however, by long-chain aliphatic and aromatic hydro-
carbons (e.g., w-dodecane, dodecylbenzene, and diamylnaphthalene).
In addition, some sulfur-containing compounds not precisely identified
appear to influence and hasten the development of tumors. Some epi-
demiologic information is in close accord with these experimental ob-
servations. Thus, industrial exposures to coal tar and pitch high in poly-
cyclic carcinogens result in tumors usually within 20-24 years of
commencing that work. Spindle oils in which carcinogen concentration
-------
Clinical and Epidemiologic Studies
199
and accelerator concentration are low result in tumors that have a
long latent period (50-54 years).32 In the case of the paraffin wax pro-
cessing using material that is low in carcinogen but high in accelerator,
the latent period is as short (20 years) as in the case of exposures to
coal tar, which is high in carcinogen and negligible in accelerator
(Table 17-1).
ULTRAVIOLET RADIATION
It is known that ultraviolet radiation of wavelengths shorter than 320
nm is a significant carcinogen in itself and is responsible for most cases
of skin cancer. Blum64 summarizes the evidence on sunlight as an
etiologic agent in human skin cancer. The limiting factor is genetic,
i.e., the relative amount of melanin pigment in the skin.
The possible interactions between POM and ultraviolet radiation
are particularly important. All the present evidence of such interactions
is derived from animals. Epstein231 has shown that ultraviolet radiation
of 280-320 nm will accelerate 7,12-dimethylbenz[a] anthracene car-
cinogenesis in hairless mice. Using carcinogenic amounts of ultraviolet
radiation (12.06 X 107 ergs/cm2) in mice, a single application of
7,12-dimethylbenz[a] anthracene before initiation of ultraviolet ex-
posures accelerated the rate of tumor appearance and their growth, as
well as the incidence of tumors per mouse.
Santamaria and Giordano660 shed further light on the relation be-
TABLE 17-1 Carcinogenesis in Man and Rodent: Relation of Latent Period to
Carcinogen Content of Fossil-Fuel Fractions and Long-Chain Acceleration"
Material
Paraffin distillate
(slack wax)
Spindle oil
Coal tar
Pitch
White mineral
ott
Relative
Carcinogen
Content
Low
Low
High
Very high
None
Probable
Percentage
of Long-
Chain Ac-
celerators
High
Low to
moderate
Negligible
Negligible
Low
Average
Latent
Period
in Mice,
weeks6
20-35
60
12-27\
Noncarcinogenic
Average
Latent
Period
in Man,
years
20 or more
40-50C
23 c
Noncarcinogenic
" Derived from Suskind and Horton.739
6 Three applications per week.
0 Based on data from Henry.35*360
-------
200 PARTICULATE POLYCYCLIC ORGANIC MATTER
tween ultraviolet radiation and polycyclic hydrocarbon carcinogenesis.
They quote the conflicting results of previous studies on the action of
ultraviolet radiation on polycyclic hydrocarbon carcinogenesis in re-
lation to acceleration. They include the work of Findlay,272 Maisin
and De Jonghe,508 Vies et a/.,794 and Clark.140 In relation to inhibition,
they discuss the work of Doniach and Mottram,211 Morton et a/.,545'546
Kohn-Speyer442 (who reported no effect of ultraviolet radiation on
cancer induction), Seelig and Cooper,688 and Rusch et a/.646 In the
papers cited, little attention is given to the total quantity of ultra-
violet energy and wavelength, which are critical factors in
carcinogenesis.
Santamaria and Giordano660 found that the carcinogenic activity
of polycyclic hydrocarbons was to a large degree associated with their
photodynamic action, which can be demonstrated in in vitro experi-
ments at molecular, subcellular, and cellular levels. They also studied
changes in electrophoretic patterns of human serum proteins produced
by various polycyclic hydrocarbons in the presence and absence of
ultraviolet radiation. Again, there was a strong association between
photodynamic activity and carcinogenic activity.
In vivo studies were then performed by painting benzo[a]pyrene on
the skin of mice. Groups of mice were irradiated with various exposures
of long-wavelength (>320 nm) ultraviolet radiation. Blum showed that,
although these wavelengths are not carcinogenic in themselves, they
will actively excite benzo[a]pyrene molecules. Small doses of ultraviolet
radiation were found to increase tumor incidence, the most effective
dosage being 8 X 1010 ergs/cm2 twice a week. Greater doses were
associated with inhibition of carcinogenesis and with more severe
tissue damage. It was interpreted that this tissue damage, rather than
alteration of the carcinogen, led to the inhibition of carcinogenesis.
Conclusions
The role of POM in the production of human skin cancer from occupa-
tional exposures seems well established. It appears that levels of ex-
posure to carcinogens or cocarcinogens in industry are very different
from those due to community air pollution. In most instances, the
specific kind of POM has not been identified, nor has the importance
of airborne transmission of the carcinogenic agent been considered
in contrast with other routes of exposure.
The data do not allow the construction of any accurate dose-response
relations, although carcinogenesis may occur at a high level of exposure
-------
Clinical and Epidemiologic Studies 201
to carcinogen or cocarcinogen. There is no information on the effect of
clearly lower concentrations of chemicals, such as would be found in
polluted community air.
The effect of other factors, such as cocarcinogens, is obviously im-
portant. This is illustrated by the effect of high concentrations of
straight-chain hydrocarbons with low concentrations of polycyclic
hydrocarbons in the production of shale-oil cancers in man.
Because the vast majority of human skin cancers are due to the effects
of ultraviolet radiation, any factor that influences or is influenced by
the biologic activity of ultraviolet radiation in carcinogenesis could be
epidemiologically important. It is in this context that experiments
showing synergism between ultraviolet radiation and POM have great
significance.
Obviously, additional research is needed before the effects of
ambient-air polycyclic compounds on the incidence of skin cancer can
be assessed. Such research must be related to the actual conditions of
human exposure and must take into account compounds other than
POM that may be present in ambient air and the concomitant effects
of ultraviolet radiation.
OCCUPATIONAL PULMONARY DISEASE
Studies of industrial populations have always been important in identify-
ing etiologic factors in disease. Such studies are particularly valuable in
establishing dose-response relations when the epidemiologic investiga-
tions of disease patterns are coupled with quantitative estimates of job-
related exposures. There is ample industrial evidence that some poly-
cyclic organic compounds are carcinogenic. Polycyclic organic matter
is produced mainly by the combustion of fossil fuels, and the major
industry-related cancer experience has been in industries that are
heavily involved with the combustion or distillation products of coal.
The first cancers related to POM were of the skin—e.g., scrotal cancer
in chimney sweeps379'600 and facial epithelioma in workers exposed
to coke, coal, tar, and pitch.104"106 However, a number of early in-
vestigators speculated that lung cancer was related to exposure to coal
tar. In 1936, the Japanese described an unusual lung cancer experience
in men engaged in coal carbonization for the production of gas: 12 of
the 15 cases of cancer occurred in the lung.429'430 At that time, lung
cancer was relatively rare in Japan, accounting for only 3% of all
malignant neoplasms. Kennaway and Kennaway434 reported an approxi-
-------
202 PARTICULATE POLYCYCLIC ORGANIC MATTER
mately threefold excess lung cancer mortality for gas-production men,
chimney sweeps, and several categories of gas workers from 1921 to
1938. The excess of lung cancer in gas workers was later confirmed
by Dolletal.206
Lloyd502 has found that the lung cancer death rate was 2.5 times
higher than expected in coke-oven workers. Most of the lung cancers
occurred in men who worked on top of the coke ovens: Those employed
5 or more years at full-time topside jobs had a tenfold excess risk of
lung cancer.
Although it seems clear that there can be an excess risk of develop-
ing lung cancer in the coal-tar occupations, quantitative exposure data
are very meager. There is only one study that provides some data on
the incidence of lung cancer with respect to the magnitude of exposure
to POM.206'483 A selected population of 11,449 employees of the British
Gas Works Industry was followed for an 8-year period. The study pop-
ulation included only employees who had at least 5 years of employ-
ment and were 40-65 years old at the beginning of the observation
period. All but 0.4% of the men were successfully followed. The work-
ers were classified into three broad categories of exposure—heavy,
intermediate, and no exposure. Relative to the group without signifi-
cant exposure, the heavily exposed workers had a 69% higher lung
cancer incidence and, unexpectedly, a 126% higher death rate from
bronchitis. There were no differences in smoking habits between the
three groups of workers and the general population.
The concentrations of benzo[a] pyrene and other poly cyclic aromatic
hydrocarbons in gas-works retort houses of several types were mea-
sured.483 The tarry fumes that escaped from retorts contained extremely
high concentrations of polycyclic hydrocarbons, but, in general, men
were exposed to these fumes only very briefly. The mean concentration
of benzo[ a] pyrene determined from long-period samples at sites
representative of normal working conditions in three works was 3 jug/m3,
over 100 times the normal level in London. Above the retorts in an
old horizontal retort house, the concentration was approximately
216 Mg/m3, about 10,000 times that in the city, and the "top-man"
working .there could be exposed to this in the normal course of his
duty.
Figure 17-1 shows the above data as a crude dose-response curve
relating the level of exposure to benzo[a] pyrene to the lung cancer
mortality ratios of the urban nonsmokers, the average British gas-
worker, 206>483 and the topside coke-oven worker.502 Urban concen-
trations of benzo[a] pyrene vary markedly among cities, but it has
-------
Clinical and Epidemiologic Studies
203
14 -
13 -
12
11
o
I 10
I 9
3
o 8
* 7
<3 6
01
i 5
1
British gasworkers
10°
10"
101 102 103
Benzo[a] pyrene Concentration, /ng/1,000 m3
FIGURE 17-1 Dose-response curves: (A) for cigarette smokers (packs per day);337
and (B) relating level of exposure to benzo[a] pyrene to lung cancer mortality ratios
of urban nonsmokers, average gasworker, and topside coke-oven workers206'483'502
(prepared by crude extrapolation of data).
been estimated roughly that the city dweller is exposed to a concen-
tration of about 8.5 Mg/1 >000 m3. This estimate is plotted in Figure
17-1 with the corresponding mortality ratio of 2, which represents
the risk relative to the rural non-cigarette-smoker. The British gas-
workers experienced 69% more lung cancer than the general popu-
lation and were exposed to an average concentration of 3,000 jug/
1,000 m3 of benzo[a] pyrene.206'483 For equivalence to urban atmo-
spheric exposure, which takes place on a 24-hr/day basis, the occupa-
tional benzo[ a] pyrene concentration is reduced by a factor of 3 to
1,000 /ng/1,000 m3 and plotted in Figure 17-1 with the corresponding
mortality ratio of 3.4 (i.e., 1.69 X 2.0). The increased mortality ratio,
of course, represents the comparative response of the gasworker and
general populations in which 68% and 61%, respectively, smoked ciga-
rettes; in Figure 17-1, this excess mortality is being applied to urban
nonsmokers. This is justified by the expectation that an added car-
-------
204 PARTICULATE POLYCYCLIC ORGANIC MATTER
cinogen exposure would constitute an even greater hazard to cigarette
smokers than to nonsmokers, so that the degree of increased risk is
probably overestimated for nonsmokers.
According to Lloyd,502 the topside coke-oven workers have about
a tenfold increase in lung cancer risk. There are no benzo[a]pyrene
exposure data for this population, but the exposure is probably not
greater than that found by Lawther et a/.483 above the gasworks re-
torts at 216,000 jug/1,000 m3. This value, corrected for a 24-hr/day
exposure, is reduced by a factor of 3 to 72,000 MS/1,000 m3. This
estimate is given in Figure 17-1 for coke-oven topside workers.
It is invariable in dose-response relations for both human and animal
cancers that any increase in the dose of a carcinogen over that suf-
ficient to cause a detectable response produces, at the very least, a
proportional increase in cancer incidence. On this basis, the benzo[a]-
pyrene lung cancer mortality ratio curve in Figure 17-1 lacks plausibil-
ity, because a dose increment of two orders of magnitude—from about
10 /Ltg/1,000 m3 to 1,000 /ug/1,000 m3-hardly increases the lung can-
cer mortality ratio of the average British gasworker relative to the
urban dweller.
For comparison, a dose-effect curve for cigarette smokers is also
shown in Figure 17-1. This curve is a plot of Hammond's data for
lung cancer mortality ratios in American males 55-60 years old con-
suming 0-10, 11-20, and 21-40 cigarettes per day.337 The maximal
number of cigarettes smoked per day in each category (i.e., l/2, 1, and
2 packs per day) is expressed in terms of urban-air benzo[a] pyrene con-
centrations that, according to the assumptions and calculations of
Sawicki et a/.,667 produce equivalent lung exposures. The principal
point to be noted is that, when the cigarette consumption increases by
a factor of 4, from lk to 2 packs of cigarettes per day, there is a rise in
mortality ratio from 3.6 to 13.6.
Doubts can be raised about whether the twofold increase in lung
cancer associated with urbanization and atmospheric pollution is dire
predominantly to benzo[a] pyrene. It may be important to distinguish
between the type of pollution found in the retort houses and that in
urban air. In a retort, coal is distilled in the absence of air, and the
products include tar, carbon monoxide, other combustible gases, and
some hydrogen sulfide, but very little sulfur dioxide or black smoke.
When coal is burned in a domestic fire, distillation products, including
tar, are emitted each time the fire is refueled, but combustion is ac-
companied by emission of smoke and sulfur dioxide. In more efficient
heating appliances, little tar or smoke is produced, and the main prod-
ucts are carbon dioxide and sulfur dioxide. Urban exposures un-
-------
Clinical and Epidemiologic Studies 205
doubtedly involve a greater duration than industrial benzo[a]pyrene
exposures. The urban exposures begin at birth, and there is evidence
that newborns are more sensitive to polycyclic aromatic carcinogens
than adults.
Other considerations warrant caution in accepting benzo[a] pyrene
as a major pulmonary carcinogen at current urban atmospheric con-
centrations. Lung cancer incidence has steadily increased since the
1940's, and yet, qualitatively, the carcinogen content of urban atmo-
spheric pollution caused by combustion products of coal has been on
the decline. Recent tissue-culture evidence indicates that the benzo[a]-
pyrene content of particles recovered from city air accounts for less
than 1% of its carcinogenic activity.278
NONOCCUPATIONAL NEOPLASTIC PULMONARY EFFECTS
Evidence of the incidence of environmentally related lung cancer in
humans is derived largely from epidemiologic studies. Four groups
of studies can be distinguished and will be discussed here: The
first compares urban metropolitan populations with rural populations
and examines the overall differences in lung cancer death rates, in
most cases without examining specific etiologic factors; the second
compares lung cancer death rates in migrants with those in their
countries of origin and those in the countries to which they migrate
and examines the changes in rates in the migrating population group,
which change with changes in environment; the third compares demo-
graphic units—including countries, states in the United States, and cities
or metropolitan areas-and studies the relation between the lung cancer
death rates and various indices of pollution, using multiple-regression
techniques in an attempt to separate the effects of environmental and
other factors; and the fourth consists of sampling studies in which
characteristics of lung cancer decedents are determined by family
interviews and compared with the corresponding characteristics of the
remainder of the population. Such studies offer the prospect of a
relatively sharp discrimination between factors that are strongly re-
lated to lung cancer and factors that are only incidentally associated.
The general association between urbanization and increased lung
cancer is not in question. Its relation to environmental rather than, say,
genetic factors is almost equally certain. What characteristics of the
urban environment are primarily responsible is a subject of controversy.
The problem of identifying the causal factors is intensified by the dif-
ficulties of obtaining either accurate or extensive measures of exposure
-------
206 PARTICULATE POLYCYCLIC ORGANIC MATTER
for most factors and the close association of urban factors with one
another-e.g., an area high in benzo[a] pyrene will generally tend to
be high in sulfur dioxide and hydrocarbons.
To enable us to summarize the results of different studies on a
common scale, one of the common pollutants, benzo[a] pyrene, was
chosen as an index of urban air pollution. Benzo [a] pyrene is taken
as the primary index of air pollution, with the recognition that it is
only one of the polycyclic organic materials in the air. The choice
should not be taken to represent a conclusion that benzo [a] pyrene is
the causal agent in urban lung cancer. Its selection as an index is
plausible for several reasons: It appears in solid form in air, is usually
adsorbed on particles, and therefore can be filtered and collected. It
is relatively easy to measure and is well correlated with other POM.
It has been found to be carcinogenic in animals, and it is suspected of
being carcinogenic in man. The intent of the regression analyses in the
present study is to establish a relation between urban air pollution
levels (as indexed by benzo [a] pyrene) and lung cancer death rate.
The standard measure of benzo [a] pyrene concentration in air is
the number of micrograms per 1,000 m3 of air. A benzo [a] pyrene
unit is defined as 1 fig/1,000 m3 of air (or 1 ng/m3 of air).
Figure 17-2 shows the relation of smoke to cancer-producing sub-
stances (hydrocarbons). In an article published in 1963,612 Pybus esti-
mated that 2% of coal burned in London was discharged as smoke and
300 r-
0 20 60 100 140 180 220 260 300 340 380 420 460 500 540 580
Smoke, mg/1,000 m3 of air
FIGURE 17-2 Relation of concentration of smoke to concentration of cancer-
producing substances (hydrocarbons). (Based on data of Pybus.612)
-------
Clinical and Epidemiologic Studies
207
that 0.03% of the smoke was benzo[a] pyrene. Measurements by
Sawicki et al.667 and others show considerable seasonable variation,
with winter concentrations some 10-20 times the summer concentra-
tions. In over 100 U.S. sampling sites, Sawicki found an approximate
log normal distribution of benzo[a] pyrene concentrations for urban
sites, with a median winter-spring value of 6.6 ng/1,000 m3. A similar
distribution was found for nonurban sites, with a median value of
0.4 ng/1,000 m3, as shown in Figure 17-3. In comparing monthly
levels in areas with low and high benzo[a] pyrene, Sawicki found
marked seasonal variations, with peak levels in the winter months.
The relation of benzo[a] pyrene to season appears to be uniform in
both high and low areas. The urban and rural distributions given by
1001-
c 10
•*=
I
o
u
o>
I
I
«>
00
0.1
I I I I
23 20 40 60 80 90 95 100
Percent of Sites with Given Concentration
of Benzo[a] pyrene, pg/1,000 m3 of air
FIGURE 17-3 Frequency distribution of benzo[a] pyrene
concentrations in the air in urban and nonurban sites, Jan-
uary-March 1959, based on composite samples. (Derived
from Sawicki et a/.667)
-------
208
PARTICIPATE POLYCYCLIC ORGANIC MATTER
Sawicki used January-March data. However, rough calculation from
Figure 17-4 indicates that the January-March average is approximately
equal to the annual average in both cases.
Ideally, we should have available the average concentrations of all
suspect air pollutants for every environment in the study and for every
year during the last several decades. Unfortunately, direct measure-
ments of such pollutants have been rare in the past and are still sparse.
Thus, in comparing countries, Stocks729 uses such indices as amount
of coal burned annually per capita and smoke or sulfur dioxide concen-
tration. Where none of these is available, other indices, such as indus-
trialization and gross national product, have been used. In this discus-
sion, an attempt was made, wherever possible, to relate the measures
given to benzo[a] pyrene-the primary index of air pollution.
80 i-
N D J
Month
FIGURE 17-4 Changes in concentration of benzo[a] pyrene
in the air monthly, July 1958-June 1959, Los Angeles and
Birmingham. (Derived from Sawickiet a/.667)
-------
Clinical and Epidemiologic Studies 209
Ben/o[a] pyrene concentrations reported in the late 1950's, of course,
may not be a good indicator of earlier or later pollution levels. The
high benzo[a] pyrene concentrations in some areas have been reduced
in recent years, owing to replacement of coal burning by other energy
sources. Recent unpublished data from the National Air Surveillance
Network show a distribution of benzo[a] pyrene concentrations in
1967 that is substantially reduced from that in 1959. The median
January-March urban concentration in 1967 was approximately
2.5 units, in contrast with the 6.6 units in 1959; the comparable
rural values are 0.4 and 0.2. The benzo[a] pyrene concentrations in
different regions at a given time do give a reasonable index of long-
term differences in pollution levels.
Types of Rates and Vital Statistics
Lung cancer is predominantly a disease of older men, and the magni-
tude of a lung cancer index depends heavily on the population base
assumed. The studies discussed below vary considerably in the choice
of index. The simplest index is, of course, an age-race-sex-specific
lung cancer death rate, such as the rate for white males 45-54 years
old. (Most studies and tabulations use ICD categories 160-164—
cancer of the respiratory system—and, except where specifically noted,
this is the group of deaths covered by the term "lung cancer.") Where
a general index is called for, all races may be combined and the rate
for, say, males over 35 may be used. The choice of 35 seems reasonable,
in that few lung cancer deaths occur at lower ages. Because age distri-
butions vary, it is preferable, if age-specific rates are available, to cal-
culate the lung cancer death rate as age-adjusted to the age distribution
of some specified standard population.
The direct method of standardization (which applies the age-
specific rates to the distribution of ages in the standard population) is
readily understood and generally appropriate when available. However,
when the study population differs markedly from the standard popula-
tion, the direct standardized rates may be subject to excessive uncertainty
(e.g., a predominantly nonwhite study group standardized to the total
U.S. population), and an indirect index may be preferred. The standard-
ized mortality ratio (SMR) is simply the ratio of deaths observed in
the particular group to the deaths to be expected if standard rates are
applied. Often, the standard used is the mortality for the total group
under study, with SM R 's calculated for each subgroup.
For ready reference, Tables 17-2 and 17-3 show the age-race-sex
-------
210 PARTICIPATE POLYCYCLIC ORGANIC MATTER
TABLE 17-2 U.S. Population, 1960, Based on U.S. Census
No. Persons, 1,000's
Sex and Race
White
Male
Female
Nonwhite
Male
Female
Total
Under 35
Years
Old
44,931
44,378
6,513
6,803
102,625
35-44
Years
Old
10,564
11,000
1,192
1,326
24,082
45-54
Years
Old
9,114
9,364
979
1,028
20,485
55-64
Years
Old
6,850
7,327
686
709
15,572
65-74
Years
Old
4,702
5,428
414
453
10,997
75+
Years
Old
2,206
2,968
181
208
5,563
distribution of the U.S. population (1960 census) and the age-race-sex-
specific death rates for lung cancer (1959-1961 vital statistics). These
data were used to calculate the indices of lung cancer mortality pre-
sented in Tables 17-4 through 17-6.
Characterization of Regions
The ascertainment of an urban or pollution factor requires the
characterization of region of residence according to the expected de-
TABLE 17-3 Lung Cancer Death Rates (ICD 160-164), 1959-1961°
No. Deaths per Million Persons
Sex and Race
White
Male
Female
Nonwhite
Male
Female
Total
35-44
Years
Old
105
34
194
40
73
45-54
Years
Old
526
98
678
128
317
55-64
Years
Old
1,505
173
1,498
209
819
65-74
Years
Old
2,256
271
1,762
253
1,175
75+
Years
Old
1,836
399
1,344
284
995
a Sources: U.S. Census of Population (U.S. Department of Commerce, Census Bureau) and
Vital Statistics of United States (U.S. Department of Health, Education, and Welfare,
National Office of Vital Statistics).
-------
Clinical and Epidemiologic Studies 211
TABLE 17-4 U.S. Total Population, Lung Cancer Death Rate
(ICD 160-164)* U.S. Race- and Sex-Specific Death Rates for
Population of All Ages (per Million), 1959-1961b
Male Female
White Nonwhite White Nonwhite
369 292 66 49
" Rates given are averages of three rates for total United States: 1959,
I960, and 1961; the 1959 rate did not include Hawaii deaths.
° Sources: Vital Statistics of United States (U.S. Department of Health,
Education, and Welfare, National Office of Vital Statistics).
gree of its pollution problem. The most convenient characterization,
although in many ways it is less than satisfactory, is provided by the
designation of Standard Metropolitan Statistical Area (SMS A). An
SMS A is a county or group of contiguous counties (except in New
England) that contains at least one central city of 50,000 inhabitants
or more or "twin cities" with a combined population of at least 50,000.
In addition, other contiguous counties are included in an SMS A if, ac-
cording to specific criteria, they are essentially metropolitan in char-
acter and are socially and economically integrated with the central
city. In New England, towns and cities, rather than counties, are used
in defining SM S A's. Central cities are those named in the titles of the
areas. The entire territory of the United States has been classified as
either metropolitan ("inside SMS A's") or nonmetropolitan ("outside
SMS A's")- For a more detailed explanation and a listing of the com-
ponent areas of each SMS A, see U.S. Department of Commerce.765
Data on metropolitan-nonmetropolitan residence obtained from the
Current Population Survey776 are related to SMS A's as defined at the
time of the 1960 U.S. Census, except as otherwise noted. SMS A's are
in many cases heterogeneous with respect to population density and
pollution. For example, the Chicago SMS A includes Kane and Porter
TABLE 17-5 U.S. Total Population, Lung Cancer Dgath Rate:
Death Rates for Population Age 35 and Over (per Million)"
Male Female
White Nonwhite White Nonwhite Total
923 839 145 135 515
a Sources same as for Table 17-3.
-------
212 PARTICULATE POLYCYCLIC ORGANIC MATTER
TABLE 17-6 Lung Cancer Experience (ICD 160-164): Age-Specific Standardized
Mortality Ratios for Total U.S. Population, 1959-1961"
Standardized Mortality Ratio6
Males
Females
35-44
Years Old
1.56
0.48
45-54
Years Old
1.71
0.32
55-64
Years Old
1.84
0.22
65-74
Years Old
1.89
0.23
75+
Years Old
1.81
0.39
a Sources same as for Table 17-3.
h „ , ,. , , expected deaths, 1959-1961 „
" Standardized mortality ratio = . Sex-age-specific expected
observed deaths, 1959-1961
deaths for 1959-1961 = age-specific death rate, standard population (U.S. total, 1960) X
sex-age-specific population in 1960 X 3 (years).
Counties, both of which are predominantly rural. However, many
important characteristics, such as cigarette sales, are available for
SM SA's and not for more sharply defined urban areas.
A more relevant characterization is provided by the distinction be-
tween urban and rural areas. According to the 1960 Census definition,
the urban population comprises all persons living in (a) places of at
least 2,500 inhabitants incorporated as cities, boroughs, villages, or
towns (except towns in New England, New York, and Wisconsin); (b)
the densely settled urban fringe, whether incorporated or unincorpo-
rated, of urbanized areas; (c) towns in New England and townships in
New Jersey and Pennsylvania that contain no incorporated municipali-
ties as subdivisions and have either at least 25,000 inhabitants or
2,500-25,000 inhabitants and a density of at least 1,500 persons per
square mile; (d) counties in states other than the New England States,
New Jersey, and Pennsylvania that have no incorporated municipali-
ties within their boundaries and have a density of at least 1,500 per-
sons per square mile; and (e) other unincorporated places of at least
2,500 inhabitants.
Substantially the same definition was used for the 1950 Census, the
major difference being the designation in 1960 of urban towns in New
England and urban townships in New Jersey and Pennsylvania. In
censuses before 1950, the urban population was defined to comprise
all persons living in incorporated places of at least 2,500 inhabitants
and areas (usually minor civil divisions) classified as urban by some-
what different rules related to population size and density. In all
definitions, the population not classified as urban constitutes the
rural population.
-------
Clinical and Epidemiologic Studies 213
Changes in the size of the urban population from one census to
another are affected by two components: growth in areas classified
as urban at the beginning of the decade and reclassification of rural
territory as urban. Between censuses, it is possible to obtain measures
of only the first component from the Current Population Survey.776
Regular publication of data on urban-rural residence from the Current
Population Survey has been discontinued.
Urban-Rural Studies
A considerable number of studies have been carried out over the last
20 years to investigate lung cancer rates as they relate to urban living.
In many of them, no measure of pollution was considered; and in most,
no attempt was made to adjust for the cigarette factor. They provide,
therefore, only a crude comparison between urban and rural lung
cancer death rates.
Mancuso et a/.513 studied lung cancer death rates of urban and
rural populations in Ohio from 1947 to 1951. He found the highest
rates in the most urbanized areas and the lowest in rural counties.
(Table 17-7).
Manos and Fisher515 studied the age-adjusted mortality rates from
102 causes of death by degree of urbanization and by sex in the white
population in the United States from 1951 to 1959 (Table 17-8).
Among males, they found approximately twice the death rate in
highly urbanized as in nonmetropolitan counties. The trend in fe-
males was much less marked.
Hoffman and Gilliam383 examined lung cancer mortality distribu-
tion in the United States for 1948 and 1949 and reported urban-rural
differences in white and nonwhite males and females, as shown in
TABLE 17-7 Comparison of Lung Cancer Death Rates in Ohio by Urbaniza-
tion (Adjusted Rates, White Men, 25-64 Years Old, 1947-1951)"
Standardized
Population Area Mortality Ratio
Eight metropolitan counties
(cities of 100,000+) 1.23
Seven other urbanized counties
(50,000 or more persons in urban area) 0.82
Rural counties
(remainder) 0.69
" Derived from Mancuso et al.il3
-------
214
PARTICULATE POLYCYCLIC ORGANIC MATTER
TABLE 17-8 Age-Adjusted Lung Cancer Mortality Rates, 1951-1959 (of the
Trachea, Bronchus, and Lung-Primary)a
Mortality Rate (per 100,000 population)
Population Area
" Derived from Manos and Fisher.5's
White Males
White Females
Metropolitan counties
with central city
Metropolitan counties
without central city
All other counties
11.5
9.0
5.8
1.6
1.3
1.2
Table 17-9. In this study, urban mortality rates are approximately
twice the rural rates in both white and nonwhite populations.
Curwen et al.177 analyzed mortality due to cancer of the lung and
larynx in the country boroughs of London and the urban districts and
rural areas of different parts of England and Wales for the period
1946-1949.
As indicated in Table 17-10, standardized mortality ratios for cancer
of the lung in both sexes and of the larynx in males increased with
increasing population density and index of urbanization. In the same
study, a comparison of districts designated as rural but varying in pop-
ulation density revealed a similar relation, as shown in Figure 17-5.
Levin et al.,492 comparing lung cancer incidence in urban and rural
areas of New York State exclusive of New York City, in 1949-1951,
found the age-adjusted incidences per 100,000 of population to be as
shown in Table 17-11. Comparison of the lung cancer incidences of
metropolitan urban and metropolitan rural males reveals only a modest
difference, but the difference in incidence between metropolitan urban
and nonmetropolitan rural areas is striking.
TABLE 17-9 Lung Cancer Mortality Rates among U.S. Urban and Rural Popu-
lations, 1948-1949°
Mortality Rate, Total Age-Adjusted (per 100,000 population)
Population Area
Urban
Rural
White
Male
22.3
12.3
Nonwhite
Female
4.7
3.7
Male
16.9
7.3
Female
4.2
2.1
a Derived from Hoffman and Gflliam.38 3
-------
Clinical and Epidemiologic Studies
215
TABLE 17-10 Standardized Mortality Ratios for Cancer of the Lung in England
and Wales, 1946-1949"
Standardized Mortality Ratio6
Population Area
England and Wales
Greater London
Northern England
Midlands
Wales
Remainder
Males
1.0
1.37
1.0
0.93
0.79
0.83
Females
1.0
1.32
0.98
0.93
0.69
0.89
" Derived from Curwen era/.1"
6 Standardized mortality rates per 100,000 for males, 40.9; for females, 8.0.
Winkelstein et al825a studied the relation of air pollution and eco-
nomic status to various mortality indices, including lung cancer, in
Buffalo and its environs. Although the overall lung cancer death rates
for areas of differing pollution level were higher with higher pollution,
100 r-
80
CO
tr
£ 60
fi
o
1 40
20
I
I
I
0.10 0.20 0.30
Persons per Acre
0.40
FIGURE 17-5 Lung cancer in rural populations of England
and Wales. (Derived from Curwen et
-------
216 PARTICIPATE POLYCYCLIC ORGANIC MATTER
TABLE 17-11 Age-Adjusted Lung Cancer Rate in Urban and Rural New York
State, 1949-1951 (Exclusive of New York City)"
Age-Adjusted Lung Cancer Rate (per 100,000 persons)
Metropolitan Nonmetropolitan
Area
Urban
Rural
Males
29.2
23.9
Females
3.2
3.5
Males
20.8
15.2
Females
3.2
2.4
" Derived from Levin et al.*9*
the authors point out that pollution and low economic status are
highly correlated and, in studying the rates specific for economic
status, they failed to find any consistent relation between air pollution
and lung cancer deaths.
Hagstrom et al.333 reporting in the Nashville air pollution study,
found no association between pollution level and cancer of the lung.
They point out that a real relation might be obscured in this study
through the interaction with other factors, especially smoking, which
was not ascertained. In this and the Buffalo study, the comparisons
are of subareas within a fairly compact region. In consequence, the
variation in exposure levels for men in the different subareas is rather
less than is the case in comparisons between regions (especially in view
of within-region mobility of workers). A real effect may therefore be
correspondingly hard to distinguish.
Haenszel et al.331 show a distribution in Iowa similar to that in New
York when the age-adjusted cancer incidence per 100,000 of population
is examined for urban and rural areas by primary site and sex, and then
for urban and rural areas in metropolitan and nonmetropolitan coun-
ties with both sexes included (Tables 17-12 and 17-13).
Griswold et al.,314 carrying out a similar type of analysis of data for
lung and bronchus cancer in males from 1947 to 1951 in Connecticut,
also showed an urban-rural difference of similar proportions, i.e., an
urban incidence of 28.0 versus a rural incidence of 17.8.
Comparison of the New York, Iowa, and Connecticut studies reveals
somewhat higher rates than in the United States as a whole, as shown
in the study by Hoffman and Gilliam.383 Particularly, the New York
metropolitan-rural rates are much higher than the rates in the other
two studies or the national rates and are comparable with the U.S. rates
for urban white males. These data suggest that considerable care must
be exercised in defining urban and rural populations. It would appear
-------
Clinical and Epidemiologic Studies 217
TABLE 17-12 Age-Adjusted Lung Cancer Rate in Urban and
Rural Areas of Iowa2
Age-Adjusted Lung Cancer Rate (per 100,000
persons)
Primary Site
Respiratory system
Larynx
Lung and bronchus
Other respiratory
Male
Urban
32.8
2.9
29.0
0.9
Rural
12.1
1.4
10.2
0.5
Female
Urban
9.0
0.9
7.8
0.3
Rural
6.3
0.0
5.3
1.0
" Derived from Haenszel et
-------
218 PARTICULATE POLYCYCLIC ORGANIC MATTER
TABLE 17-14 Death Rates for Malignant Neoplasm of the Trachea, Bronchus,
and Lung among White Male Cigarette Smokers by Type of Area"
Population Area
Cities >1,000,000
Urban areas with
250,000-1,000,000
Urban areas <250,000
Rural areas
Death rate
(per 100,000 persons)
29.4
22.9
17.5
14.6
Percentage of
Cigarette Smokers
48.5
49.5
47.8
42.5
"Derived from Prindle.605
lung cancer death rates and measures of pollution. Taking the median
January-March urban benzo[a]pyrene concentration of 6.6 jug/1,000
m3 and rural concentration of 0.4 /zg/1,000 m3 of Sawicki et a/.,667
one may roughly associate a 100% increase in lung cancer death rate
with a 6.2-unit increase in benzo[a] pyrene or an increase of approxi-
mately 15% in deaths per unit increase in benzo[a] pyrene. However,
the 100% increase is associated with a contrast between the most
heavily urban and the most rural environments, so that the pollution
effect estimated from these studies should be somewhat less than 15%.
Migrant Studies
In the second group of studies, lung cancer death rates in migrants
from one country to another were compared with those in their fel-
lows in the home countries and with those in the people in the countries
to which they had migrated. If such migrants can be considered as
equivalent to random or representative samples of the populations of
the home countries, then differences in death rates from those in the
home countries can be ascribed to changes in environmental conditions.
Pollutants, including benzo[a] pyrene, vary considerably worldwide.
Concentrations in Great Britain are much higher than those in the
United States, and concentrations in Norway, Italy, New Zealand,
South Africa, and Australia are considerably lower than those in the
United States.731
Mancuso and Coulter512 studied lung cancer death rates in male
migrants 25-64 years old in Ohio; the smoking factor was not con-
sidered. Their results were as shown in Table 17-15. The findings sug-
gest that the lung cancer death rates of migrants are between that of
the mother country and that of the country to which they migrated.
British immigrants had a lower death rate than that observed for the
-------
Clinical and Epidemic logic Studies 219
TABLE 17-15 Lung Cancer Death Rates for Native and Immigrant Male Popu-
lations in Ohio, 25-64 Years Old"
Lung Cancer Death Rate (per 100,000 persons)
Population Group Migrant Group Nonmigiating Cohort Group
Native-born Americans
Immigrants from
United Kingdom
Immigrants from
Italy
29
32
19
_
55
16
" Derived from Mancuso and Coulter.5'2
English population, but greater than the rate for native Americans.
Italian immigrants had a slightly greater lung cancer death rate than
those remaining in Italy but considerably less than native white
Americans and immigrants from the United Kingdom.
In a study by Haenszel329 of migrants from specific countries, lung
cancer incidences agreed in essence with those of Mancuso and Coulter
in Ohio. The study also confirmed the observation that migrants coming
from countries with lower lung cancer death rates showed a displace-
ment in rates to a position intermediate between that observed for the
home country and that for the country to which they migrated.
Eastcott's study223 of migrants from the United Kingdom to New
Zealand is important, in that it attempts to separate the effects of the
urban and cigarette factors on cancer death rates. Without document-
ing individual smoking habits, it was determined that New Zealanders
were heavier smokers than persons in the United Kingdom.
Migrants from the United Kingdom had a 35% higher risk of lung
cancer than native New Zealanders if they came from the United
Kingdom before the age of 30 and a 75% higher risk if they migrated
after the age of 30. This was true, regardless of the fact that the
migrants generally increased the number of cigarettes smoked after
arriving in New Zealand.
Dean's studies187'189 compared lung cancer rates in British subjects
who migrated to South Africa and Australia with those in native-born
South Africans and Australians. South Africans are among the heaviest
cigarette smokers in the world, and British migrants to South Africa
tended to increase their consumption of cigarettes markedly. In spite
of this, they had a significantly lower lung cancer death rate than
persons remaining in England but a higher rate than native South
Africans, as shown in Table 17-16. The lung cancer mortality rates for
-------
220 PARTICIPATE POLYCYCLIC ORGANIC MATTER
TABLE 17-16 Lung Cancer Death Rates for White Male Natives of England,
Wales, and South Africa and United Kingdom Migrants to South Africa,
1947-1965"
Annual Lung Cancer Death Rate (per 100,000 persons)
Population Group 55-64 Years Old 65+ Years Old
Native white South Africans
United Kingdom migrants to
South Africa
Native white United Kingdom
50
112
135
112
172
219
" Derived from Dean.'• *
men 40 years old or older, as indicated in the Australian study, are
shown in Table 17-17. Variations in smoking practices (e.g., butt
length, depth of inhaling, and number of puffs) are important in as-
sessing the relation of cigarette smoking to lung cancer incidence.
Dean, comparing United Kingdom and South African smokers,
found that the average cigarette butt length was 25.3 mm for native
white South Africans and 25.2 mm for immigrants from the United
Kingdom. He also found that a greater percentage of South Africans
inhaled deeply and that South Africans took more puffs per
cigarette.
In Britain, Doll et al.209 observed an average butt length of about
20 mm, whereas Hammond (quoted in Doll et al.), in the United
States, noted a butt length of about 30 mm. Sampling methods for
the two studies were very different. Collection was by mail in Great
Britain, and butts discarded in public places (including restaurants)
were collected in the United States; the results are therefore difficult
to interpret.
TABLE 17-17 Age-Adjusted (40+Years Old) Lung Cancer
Death Rates, 1950-1958"
Lung Cancer Death Rate
Population Group (per 100,000 persons)
Native Australians 53
United Kingdom migrants to Australia 94
United Kingdom cohort group 154
" Derived from Dean.'8'
-------
Clinical and Epidemiologic Studies 221
In another study, Reid et a/.628 considered lung cancer death rates
in migrants from the United Kingdom and Norway and compared them
with those in persons remaining in the home country and native-born
Americans. Their results are summarized in Table 17-18. This large
and carefully detailed study confirms the findings from other, pre-
viously cited studies. Lung cancer death rates of migrants are inter-
mediate between those of native U.S. residents and those of persons
in the home countries.
The results of these migrant studies are very similar to the results
of the urban-rural studies reported, in that they show lung cancer death
rates paralleling the general pollution levels in the areas in question.
The data also suggest that exposure to pollution early in life may pro-
duce persistent effects. These studies lack data regarding the effect of
the smoking factor. Consideration of heavy smoking habits in New
Zealand, Australia, and South Africa, compared with the United
Kingdom, however, provides some basis for concluding that the
changes in lung cancer death rates are related to more than cigarette
smoking, i.e., to an urban factor.
Regression Studies
Regression studies attempt to separate the effects of factors that dif-
ferentiate urban and rural environments with the aim of identifying
urban factors that might be held responsible for the difference in lung
cancer death rates. The method usually adopted for attempting to
separate the effects of different factors statistically is multiple re-
gression. (See Appendix D for discussion of the regression analysis.)
TABLE 17-18 Age-Adjusted Death Rates from Lung
Cancer in Great Britain, Norway, and the United States"
Lung Cancer Death Rate
(per 100,000 persons)
Population Group
Great Britain residents
Great Britain-born U.S. residents
Norway residents
Norway-born U.S. residents
Native U.S. residents
Males
151.2
93.7
30.5
47.5
72.2
Females
19.3
11.5
5.6
10.7
9.8
a Derived from Reid etal""
-------
222 PARTICULATE POLYCYCLIC ORGANIC MATTER
This method is subject to severe limitations, of which the following
are most important:
1. Even if the relation has the assumed form, fluctuations in the
data due to random effects of other variables make it difficult to
determine the coefficients of the relation with reasonable precision.
Thus, the failure to find a significant result, according to the usual
statistical measures, is not necessarily evidence that there is no effect.
A real effect may readily be hidden in the sea of fluctuations induced
by extraneous variables.
2. Much more important from the point of view of confusion in
interpretation is the disguising of effects actually due to one variable
under the label of another. Particularly when variables are highly cor-
related (as is the case for most measures of urban crowding and pol-
lution) and when there are substantial measurement errors (as is the
case for measures of pollutants diffused through the air), the coef-
ficient of a regression relation may be very difficult to interpret, and
the addition or withdrawal of one variable in the equation may have
profound effects on the coefficients for the others. In consequence,
inappropriate analyses and interpretations are frequent, and no
analysis can lay claim to positive assurance.
In spite of all these reservations, the regression method provides the
best available technique for separating the effects of different variables,
and the evidence was studied and interpreted by this method, subject
to substantial qualification.
TABLE 17-19 Multiple Regression Analysis of Lung Cancer Death Rates for
Males in 19 Countries and Cigarette and Solid-Fuel Consumption'2
Regression Coefficients
Age Group,
years
Age-adjusted
25-34
35-44
45-54
55-64
65-74
Average
Death Rate
(per million
persons)
749.3
10.0
73.2
427.6
1,377.2
1,939.3
Constant
(C0)
330.0
2.8
9.7
164.0
704.6
810.0
Cigarettes,
1,000's per person
per year (avg. = 1.76)
110.0
2.0
23.0
78.0
138.0
321.0
Solid Fuel,
metric tons
per person
per year
(avg. = 1.55)
144.0
2.0
15.0
80.0
276.0
361.0
a Derived from B. W. Carnow and P. Meier (unpublished data).
-------
Clinical and Epidemiologic Studies 223
In 1966, Stocks731 reported analyses of a number of variables related
to lung cancer mortality in 1 9 countries. In view of the limited
analyses provided by Stocks, additional regression calculations were
performed for the same 1 9 countries. Cigarette consumption, known
for the mid-1 960's,731 was used, as were age-specific death rates from
Segi era/.689 Stocks's 1955-1958 solid fuel consumption was used as
a measure of pollution.
Earlier analyses of a number of variables reported by Stocks et a/.733
indicated that many were highly correlated with each other and some,
such as liquid fuel, seemed not to be related to lung cancer or other
indices of pollution. In accordance with these preliminary analyses,
the two variables -amount smoked per capita (or amount produced)
and solid-fuel consumption per capita— were chosen as the basic regres-
sion variables. Table 17-19 shows the results of the regressions of the
form
in which Y = age-sex-specific lung cancer death rates from 1958 to 1959,
Xt = cigarette consumption per capita, in thousands of cigarettes per
year per person over 15 years old, and
X-i = solid-fuel consumption per capita, in metric tons per person per
year.
On the assumptions that the lung cancer death rate is related both
to cigarette consumption and to solid-fuel consumption and that the
effects are at least approximately additive, Cl measures the increment
in lung cancer death rate per unit increase in cigarette consumption,
and C2 measures the increment per unit increase in solid-fuel con-
sumption. C0 is the regression coefficient constant.
For cigarette smoking, the coefficient is approximately 1 5% of the
average lung cancer death rate. For example, the coefficient for cig-
arettes in the male age-adjusted group is 110, which is 14.7% of the
average rate of 749.3. Taken at face value, this suggests an increment
in male lung cancer deaths of 1 5% per 1 ,000 cigarettes per year. To
convert the rate per 1 ,000 cigarettes per year to a rate per cigarette per
day, one multiplies it by 365/1 ,000, which gives a 5.4% increase in
lung cancer death rate per cigarette per day. This corresponds to approx-
imately a doubling in the lung cancer death rate corresponding to an
increase in smoking of a pack (20 cigarettes) per day. This estimate is
compatible with the variation in lung cancer death rates by smoking
category reported by Hammond.338
-------
224 PARTICULATE POLYCYCLIC ORGANIC MATTER
For solid-fuel consumption, the regression coefficient is approxi-
mately 20% of the average lung cancer death rate (i.e., 144/749.3).
This suggests an increment in male lung cancer deaths of 20% per
metric ton of coal burned per year per capita. Although Pybus612 esti-
mated total benzo[a] pyrene released per ton of coal burned, there is
no way to convert this to benzo[a] pyrene units.
The conclusion suggested by these results is that the products of
solid-fuel combustion or of some variable highly correlated with solid
fuel may be an important etiologic factor in lung cancer.
A study was undertaken by the Panel for the 48 contiguous states
of the United States with the following independent variables:
Lung cancer death rate/million persons,
Xl = cigarette sales per person over 15 years old (1963), and
Xi = benzo[a]pyrene, in benzofa]pyrene units (R. I. Larsen and J. B.
Clements, unpublished data), weighting urban and rural values measured in
each state according to the urban fraction of each state.
The results are shown in Table 17-20.
As noted in the analyses of regression coefficients for solid-fuel
consumption in the 19-country study,731 the regression coefficients
for benzo[a] pyrene for white males are around 5% of the average rates
(e.g., for age-adjusted white males, 47.5 is 5.5% of 867.5). Similar re-
sults are found for each age-specific group. This suggests that an in-
crease in urban pollution corresponding to an average benzo[a] pyrene
of 1 jug/1,000 m3 may result in an increase of 5% in the lung cancer
death rate. The coefficients for nonwhite males are larger—about 15%.
It should be kept in mind, in considering this group, that the nonwhite
population in many states is distributed between urban and rural
environments in a way quite different from the white majority. If, in
fact, nonwhites are found in heavily polluted areas of industrial states
but tend to be in rural areas of southern states, the proper benzofa]-
pyrene indices for nonwhite populations might be much more spread
than those in Table 17-20, and an appropriate weighting might lead to
a coefficient similar to that for whites.
The results for females, who generally exhibit much lower lung
cancer death rates, are more variable and mostly nonsignificant, ex-
cept for those 45-54 years old.
A final important qualification in comparing these results with
those of the urban-rural studies is based on the different benzo [a] -
pyrene levels used. The regression studies use 1969 levels that are
-------
Clinical and Epidemiologic Studies
225
TABLE 17-20 Multiple Regression Analysis on Lung Cancer Death Rates (per
1,000,000 population) in the United States in the White and Nonwhite
Population2
Average Benzo(a) pyrene,
Death Rate Tobacco Sales, $ Mg/1,000 m3 of air
Age Group, (per million Constant (avg. = 28.8) (avg. = 1.38)
years persons) (C,) (Ct) (C2)6
Male, white,
867.4
460.7
11.8
47.5
3544
45-54
55-64
65-74
Male, nonwhite,
age-adjusted
3544
45-54
55-64
65-74
Female, white,
age-adjusted
3544
45-54
55-64
65-74
Female, nonwhite
age-adjusted
3544
45-54
55-64
65-74
101.6
497.8
1405.8
2064.7
844.6
132.3
606.0
1367.6
1722.4
129.5
32.9
86.5
163.1
255.2
156.3
34.3
85.8
184.9
391.7
86.5
125.2
894.4
835.4
184.2
22.8
245.9
141.6
162.6
92.1
12.7
36.4
113.5
184.7
151.5
66.0
32.4
-375.0
1286.4
0.3
3.0
14.4C
35.3C
16.5C
1.6
7.8
34.0
38.8C
1.0C
0.8C
l.lc
2.0C
1.7
0.1
-1.0
-0.4
17.3C
-19.2
4.6
18.7
70.5
152.6C
133.1C
46.4C
97.7
179.0
318.8
7.1C
-1.1
14.2C
-6.2
16.3
1.2
-1.1
46.6C
44.5
-246.1
" Derived from B. W. Carnow and P. Meier (unpublished data).
b The estimate of benzo[a] pyrene concentration for each state was computed in the follow-
ing manner: Benzo [a] pyrene concentrations were measured quarterly in each year from
1967 to 1969 in various urban and nonurban places in the United States. For each place,
the four measurements in 1969 were averaged to obtain an estimate of the mean concen-
tration for that year. For the few places with insufficient data in 1969,1968 concentra-
tions were used. All cities in a given state were categorized into (1) urban places whose
population in 1960 was at least 1 million, (2) urban places whose population in 1960 was
less than 1 million, and (3) nonurban places. The estimates of mean 1969 concentration
were then averaged over all places with benzo[a] pyrene measurements in each of the
three categories. Each of the resulting three averages was considered to be representative
of all cities in the corresponding category, regardless of whether they had benzo[a] pyrene
measurements. (In this case, all cities in the first category-i.e., with population of 1 mil-
lion or greater-had measurements.) A weighted average of the three averages was obtained,
using the corresponding populations as weights. This weighted average was considered to
be the representative benzo[a] pyrene concentration for the entire state and was used as
the second independent variable in the regression analyses of this table. In some states, no
nonurban measurements were given; in those cases, the total population for the second
category was taken to be the population of the entire state less the total population of all
urban places with populations of at least 1 million each.
c Coefficient greater than 2 S.E.
-------
226 PARTICIPATE POLYCYCLIC ORGANIC MATTER
judged to be, perhaps, no more than half the 1950 levels (which may
actually be more relevant to current lung cancer deaths) quoted from
Sawicki in the section on urban-rural studies. If we adjust our results
correspondingly, we can conclude that, in place of 5% and 15%, coef-
ficients of 2% or 3% for benzo[a] pyrene unit for whites and 1% or
8% for nonwhites may be more appropriate.
Sampling Studies
In sampling studies, values of the variables of concern, such as smoking
habits arid residence, are determined for individual subjects, and con-
fidence in the interpretation of the results is correspondingly greater
than in regression studies. The problem of disguising is still present,
but to a far lesser degree.
Buell and Dunn" carried out a follow-up study of male American
Legion veterans in California. They compared lung cancer death rates
in subgroups classified by residence (urban, rural, inner-city) and
smoking habits. A substantial difference was found between lung
cancer death rates in the major cities and those in the smaller ones.
Differences in smoking habits between residents of different cities were
small, and, as shown in Table 17-21, differences in death rates between
cities are only slightly modified by adjustment for smoking.
From unpublished data of R. I. Larsen and J. B. Clements, benzo[a]-
pyrene concentrations for San Francisco and Los Angeles can be esti-
mated at 1.1 and 1.8 units, respectively. Concentrations for other Cali-
fornia cities are highly variable.
Dean,188 in a study in Northern Ireland, determined smoking
habits of lung cancer decedents and matched controls by interviewing
their families. Lung cancer death rates for rural areas were consistently
TABLE 17-21 California Veterans Survey (American Legion): Lung Cancer
Death Rates in Males, Age 25 and Older, Classified by Residence11
Lung Cancel Death Rate (per 100,000 persons)
Study Group
Adjusted for age only *
Adjusted for age and cigarette smoking6
Los Angeles
95.9
95.4
San Piancisco
Bay Area
104.5
102.0
All Other
California
Cities
75.3
75.5
a Derived from Buell and Dunn."
6 Adjusted to total American Legion population.
-------
Clinical and Epidemiologic Studies 227
lower than those for urban areas in every smoking category, as shown
in Table 17-22.
The levels of air pollution in Belfast have been measured, and the
annual average benzo[a]pyrene concentration in England and Wales
estimated from Stocks's data728 is about 30 /ug/1,000 m3. This high
value is associated with a considerable difference in lung cancer death
rates between inner Belfast and the country districts designated by
Dean as truly rural. The differences in death rates as a percent of the
Northern Ireland rates are 144%, 130%, 127%, and 87% for the four
smoking categories, corresponding to 5%, 4%, 4%, and 3% increases,
respectively, in lung cancer death rates per benzo[a]pyrene unit.
Hammond and Horn,341 reporting on a 44-month follow-up of a
sample population of U.S. white males, examined death rates from
lung cancer. These were well-established cases exclusive of adenocar-
cinoma. The death rates by city size, adjusted for age and smoking
history, are shown in Table 17-23.
From the Sawicki et a/.667 data using January-March data, which
approximate the yearly average, benzo[a]pyrene concentrations for
large cities and for rural areas can be estimated at 6.6 and 0.4 units,
respectively. Thus, the relation between lung cancer death rate and
estimated benzo[a]pyrene concentration is an increased rate of 52
- 39 = 13 per 100,000 of population, which is about 33% of the U.S.
white male rate for 1969, corresponding to a difference in pollution
level of 6.6 — 0.4 = 6.2 benzo[a] pyrene units, or a change of about
5% in the lung cancer death rate per unit.
Stocks and Campbell732 found that increased lung cancer in rural
areas was proportional to the number of cigarettes smoked. Studies in
urban Liverpool revealed increases over the rural rates in every category
of smokers; the differences were more pronounced in mild than in
heavy smokers. The findings are shown in Table 17-24. The largest
urban-rural differences in lung cancer death rates were in light smokers;
there was only a modest effect in heavy smokers. The rural standardized
death rates (SDR's) for the three categories of cigarette smokers were
70.8%, 36.3%, and 7.9% lower than the urban SDR's for light, moder-
ate, and heavy smokers, respectively; the difference in benzo[a] pyrene
concentration was 7.7 — 0.1 = 7 jug/1,000 m3.732 This leads to estimates
of increased lung cancer death rate per benzo[a] pyrene unit for light
(10%), moderate (5.5%), and heavy (1.1%) smokers. These estimates
cover a wide range and are considerably higher in the lighter smoking
categories than is the case in most other studies. This discrepancy
might reflect the unusually high benzo[a] pyrene concentrations in
-------
00
TABLE 17-22 Age-Standardized Lung Cancer Death Rates for Persons Age 35 and Older"
Lung Cancer Death Rate (pet 100,000 persons)
No. Cigarettes
Smoked
per Day
0
1-10
11-22
23t
Inner
Belfast
36
138
288
509
Outer Urban
Belfast Districts
40
140
207
430
" Derived from Dean.188
h . „ (inner Belfast) -
6 Fnnnk inn V -
21
121
171
515
(rural)
Small
Town
71
260
772
Environs
of Belfast
16
60
192
716
cEc
Rural
10
25
74
173
Northern
Ireland
18
87
169
383
Percent
Difference6
144
130
127
88
Percent
Difference per
Benzo[a]pyrene Unitc
4.8
4.3
4.2
2.9
percent difference
30 (Mg benzo[a]pyrene per 1,000 m3)
total No. Ireland
-------
Clinical and Epidemiologic Studies 229
TABLE 17-23 Lung Cancer Death Rates for White Males in the United States for
Smokers and Nonsmokers by City Size"
Population
Area
City of 50,000+
City of 10,000-50,000
Suburb or town
Rural
Lung Cancer Death Rate (per 100,000 persons)
Adjusted for
Age and Smoking
History No. Deaths Observed
52 83
44 59
43 67
39 52
" Derived from Hammond and Horn.3"'
the areas studied. It should, of course, be kept in mind that the number
of deaths on which these rates are based is small. The apparent effect
in nonsmokers is even greater (13%), but the numbers are clearly too
small to draw any useful conclusions. Only one sampling site was used
in each of the areas, and the urban-rural difference in benzo[a] pyrene
for the communities studied by Stocks and Campbell is therefore un-
certain.
Haenszel and associates330'332 conducted a study in which a 10%
sample of lung cancer deaths in white males and females was studied
and interviews with family members were conducted. A sample of
U.S. residents was also interviewed. Information collected included
age, sex, smoking habits, location, and duration of residence. As
shown in Table 17-25, the lung cancer death rate in males, adjusted
for age and smoking history, is higher in urban areas than in rural
areas, and the difference increases as the duration of residence in-
creases. This finding gives added support to the view that the urban-
rural difference in lung cancer mortality is related to direct environ-
mental effects. Haenszel notes that, among males, the urban-rural
difference is largest among heavy smokers. This finding was not dupli-
cated in the data for females, nor was it the case in some other
studies.379'732 It may be noted that the lung cancer death rate for
lifetime rural residents is half that for lifetime urban residents-in close
agreement with the simple urban-rural studies discussed earlier.
An unusually detailed, well-documented study was reported in
1968 by Hitosugi.379 The study was carried out in an area near Osaka
including some regions of high industrial pollution and some with rela-
tively low levels of pollution, as determined by measurements at
sampling sites. The method used was to interview families of the 259
-------
TABLE 17-24 Lung Cancer Death Rates from mid-1952 to mid-1954 According to Age, Smoking Category, and Population Area"
Lung Cancer Death Rate (per 100,000 persons)
to
o
Age 45-54
Smoking Category Rural Mixed Urban
Nonsmokers 0 0 31
Pipe smokers 0 0 104
Cigarette-light 69 57 112
Cigarette-moderate 90 83 138
Cigarette-heavy 117 214 205
Number of deaths 16 26 124
a Derived from Stocks and Campbell.732
6 Standardized death rate.
(urban SDR) - (rural SDR)
c Enuals 1 00 X — .
Age 55-64
Rural
0
30
70
205
626
26
Mixed
0
59
224
285
362
56
Urban
147
143
376
386
543
230
Age 65-74
Rural
70
145
154
362
506
27
d i7nnai.
Mixed
0
26
259
435
412
36
SDR6 Age 14-74
Urban Rural
336 14
232 41
592 87
473 183
588 363
183 68
Mixed
0
25
153
132
303
118
Urban
131
143
297
287
394
539
Difference0
89.3
71.3
70.7
36.2
7.9
Difference per
Unitd
12.8
10.2
10.1
5.2
1.1
percent difference
7.0 (ng benzofa] pyrene per 1,
000m3)
urban SDR
-------
Clinical and Epidemiologic Studies 231
TABLE 17-25 Standard Lung Cancer Mortality Ratios in White Males in Urban
and Rural Areas, Adjusted for Age and Smoking History"
Standard Mortality Ratio by Duration of Residence,b years
Current
Residence
Urban
Rural
All
Durations
113
79
-------
ts>
W
TABLE 17-26 Lung Cancer Death Rates for Males and Females, Age 35- 74, by Amount of Smoking and
Level of Air Pollution"
Lung Cancer Death Rate (pei 100,000
Smoking
Category,
cigarettes daily
Nonsmoker
Ex-smoker
1-14
15-24
25+
Males
Low
Pollution
11.5
(5)
26.2
(11)
10.6
(9)
14.7
(18)
36.3
(19)
Inter-
mediate
Pollution
3.8
(1)
42.6
(7)
14.2
(10)
19.1
(17)
15.8
(4)
High
Pollution
4.9
(1)
61.7
(7)
23.5
(14)
27.0
(17)
46.4
(9)
persons)2*
Total
7.9
(7)
36.0
(25)
15.3
(33)
19.1
(52)
44.0
(32)
Females
Low
Pollution
4.6
(15)
12.4
(2)
19.7
(13)
12.4
(1)
-
(0)
Inter-
mediate
Pollution
6.9
(12)
52.6
(2)
16.5
(6)
23.1
(2)
-
(0)
High
Pollution
3.8
(6)
124.0
(3)
15.3
(5)
24.0
(1)
-
(0)
Total
4.9
(33)
13.3
(6)
17.6
(24)
19.7
(4)
_
(0)
" Derived from Hitosugi.3"
b Numbers in parentheses are numbers of deaths.
-------
Clinical and Epidemiologic Studies 233
TABLE 17-27 Levels of Specific Air Pollutants Corresponding to Low, Inter-
mediate, and High Pollution Levels in Osaka, Japan0
Pollutant
Dust fall, tons/km2 per month
Sulfur dioxide, nig/day per 100 cm2
ofPbO2
Suspended participate matter, mg/m3
Benzo[a] pyrene, ^g/l,000m max.
Low
8.0
0,74
0.19
26.0
Intermediate
9.15
2.64
0.22
31.0
High
12.8
3.03
0.39
79.0
1 Derived from Hitosugi.3
Discussion
Although not all studies are completely consistent, it is clear that
increased lung cancer mortality is generally associated with urban
living. This relation does not appear to be accountable solely in terms
of differences in smoking habits.
The attribution of the urban-rural difference in lung cancer to one
or a few particular urban factors is far less certain, and, because the
data are subject to considerable variability and extraneous influences,
it is not possible to come to unequivocal conclusions. However, the
high correlation of urban factors—such as solid-fuel consumption,
smoke, and benzo[a]pyrene—suggests the reasonableness of using one
of these factors as an overall index of urban pollution. Benzo[a]pyrene
recommends itself as an index susceptible to direct measurement and
known, in other contexts, as a potent carcinogen. Accordingly, the
concentration of benzofa] pyrene was used as an index of urban pol-
lution. With fan- consistency, most of the studies in the four categories
discussed lead to an estimated increase of about 5% in the lung cancer
death rates among males, corresponding to an increase in urban pollu-
tion represented by 1 /xg of benzo[a] pyrene per 1,000 m3.
It is essential to consider the evidence that does not conform to this
pattern:
1. In most studies, relatively little effect of the urban environment
on the lung cancer death rate for women is found. It is possible that
a similar effect is present, but not clearly detectable, because of the
generally much lower rates in women. An alternative explanation
might be that the effects noted in men are due to occupational ex-
posure or to travel through local regions of very high pollution.
2. In some regions—such as Finland (particularly Helsinki) and the
-------
234 PARTICULATE POLYCYCLIC ORGANIC MATTER
United States (New Orleans)—high lung cancer death rates are found,
although the more familiar pollutants, including benzol a] pyrene, are
at a relatively low concentration. Special explanations have been sug-
gested (e.g., wood smoke in Finnish saunas and coffee roasting in
New Orleans), but it must be recognized that special explanations could
probably be found for anomalies in almost any region.
3. The possibility that air pollution is in fact correlated with some
entirely different variable (such as difference in smoking practices)
that is the real cause cannot be excluded. None of these studies con-
stitutes an experiment, and the more persuasive sampling studies are
of limited extent. The measurements of air pollution, especially, are
entirely inadequate to give reasonably good measures of individual
exposure.
The Nashville and Buffalo studies, each carried out within a single
region, found a strong correlation between cancer of the lung, bronchus,
and trachea and economic status; but no consistent relation with air
pollution was found.
Despite these reservations, a number of different types of studies
do lead to fairly consistent estimates of a substantial effect of air pol-
lution on lung cancer death rate, although more measurement of
pollutants and far more extensive sampling studies should be under-
taken. It appears both reasonable and prudent to take as a working
hypothesis the existence of a causal relation between air pollution and
lung cancer death rate at the rate of a 5% increase for each increment
of pollution as indexed by 1 benzo[a] pyrene unit.
This hypothesis leads to the estimate that a substantial reduction
in the pollution of highly urban environments would lead to a cor-
responding reduction in lung cancer death rate (e.g., a reduction of
air pollution corresponding to a reduction of benzofa] pyrene concentra-
tion from about 6 jug/1,000 m3 to around 2 jug/1,000 m3 might reduce
the lung cancer death rate by about 20%). Similar benefits might be
expected in all smoking categories.
Conclusions
Epidemiologic studies of lung cancer appear to show the following:
1. Lung cancer has emerged as the single greatest cause of cancer
death in males and a significant cause of death in females in the United
States, and its incidence has increased in the last 30 years.
-------
Clinical and Epidemiologic Studies 235
2. A major etiologic factor appears to be cigarette smoking; how-
ever, smoking does not completely account for the increased inci-
dence of the disease.
3. Urban dwellers have approximately twice as high an incidence
of lung cancer as those living in rural areas. Within urban communities,
the incidence is greater where more general industrial pollution is
present.
4. Polycyclic organic matter, found in cigarette smoke in high con-
centrations, causes cancer of the lung and other organs in experimental
animals, is present in the air in large quantities in industries whose
workers have high lung cancer rates, and is present in the air of urban
communities.
5. Generally, immigrants have an incidence of lung cancer between
that noted for their countries of origin and that of the countries to
which they migrate. The higher their ages when they migrate, the
closer their rates are to those of their cohorts who remain in the
countries of origin. In some of the studies, in which the home country
had a much higher cancer rate, the rates in persons who left it de-
creased significantly, even though their cigarette smoking increased.
These studies suggest a significant environmental effect operating
early in life for lung cancer development.
6. A variety of types of epidemiologic studies lead to an estimate
of the effect of pollution on lung cancer death rate of a 5% increase
per unit increase in urban pollution as indexed by benzo[a] pyrene
(one benzo[a] pyrene unit = 1 /ug of benzo[a] pyrene per 1,000 m3 of
air).
NONOCCUPATIONAL NEOPLASTIC DISEASE
OF OTHER ORGANS
The relation of POM to organs other than the lung is not well docu-
mented. In the Nashville Air Pollution Study,333 carcinoma incidence
was investigated in relation to sulfur trioxide, soiling dustfall, and
24-hr sulfur dioxide. In this study, socioeconomic factors in relation
to carcinoma were also evaluated, but occupation and cigarette smok-
ing were not considered. When the degree of exposure to air pollutants
was kept constant, an inverse relation between socioeconomic class
and mortality for cancer of the stomach, esophagus, and prostate was
observed. For cancer of the bladder, there was a direct relation to
socioeconomic class. The most consistent pattern was noted for sus-
pended particulate matter, as measured by the soiling index. For the
-------
236 PARTICULATE POLYCYCLIC ORGANIC MATTER
middle socioeconomic class, the death rate from all cancers combined
was somewhat higher for the area of higher pollution, as measured by
soiling. Direct relations were found between levels of pollution, as mea-
sured by soiling, and cancer of the esophagus, prostate, and bladder.
Bladder death rates were highest for areas of high pollution for all
four pollutants. For stomach cancer, significant differences were found
to be in accord with degree of dustfall.
Levin et a/.,492 in a study in New York, showed an urban excess for
cancer, not only of the respiratory system, but also of the esophagus,
intestines, and rectum. Schiffman and Landau677 demonstrated a signifi-
cant correlation between some indices of air pollution and death rates
from cancer of the stomach and esophagus, in addition to the lung,
bronchus, and trachea. Dorn214 revealed a lower mortality risk in non-
metropolitan, compared with metropolitan, counties for cancer of the
bladder.
Winkelstein and Kantor,824'825 in studies carried out in Buffalo,
found a significant relation between prostatic and stomach cancer and
particulate matter. There was no correlation, however, between lung
cancer death rates and particulate matter. Cancer of the prostate
appeared to be independent of economic status, at least in men under
70. In the middle economic grouping among white males 50-69 years
old, the mortality rate for prostatic cancer was 2.7 times as high in the
most polluted zone as in the least polluted. In regard to gastric can-
cer in white men and women 50-69 years old, mortality rates were
almost twice as high in areas of high suspended particulate air pol-
lution as in areas of low pollution. This appeared to be independent
of the effect of economic status and was not apparently attributable
to the ethnic distribution of the population in the study area.
Cohart150 found a significant association between socioeconomic
status of New Haven residents and the incidence of stomach cancer.
The peak ratio of observed to expected stomach cancer cases occurred
among males in the middle class, with a slight falling off among the
poor. Among females, however, there was a statistically significant
excess of observed to expected stomach cancer cases in the low socio-
economic class. Others have also shown a relation between the incidence
of stomach cancer and social class.145
In general, the results for organ systems other than the lungs appear
to be inconclusive. In most of the studies, smoking and dietary habits
and pollution levels were not considered, making it impossible to
evaluate the role of particulate POM in the induction of nonoccupa-
tional neoplastic diseases.
-------
18
General Summary and
Conclusions
SOURCES OF POLYCYCLIC ORGANIC MATTER
Polycyclic organic matter can be formed in any combustion process
involving compounds of carbon and hydrogen. Naturally occurring
POM emissions to the atmosphere do not appear to be significant.
Major technologic sources of POM emissions include transportation,
heat and power generation, refuse burning, and industrial processes.
The internal-combustion engine is a ubiquitous source of POM.
Current efforts and projections of future control levels point toward
a continuing decline in vehicular POM emissions.
The emissions from major stationary sources are poorly quantified.
Available data suggest that coal-fired furnaces, coal-refuse bank burn-
ing, and coke production from the iron and steel industry account
for the bulk of the nationwide PO M emission inventory. Atmospheric
POM concentrations are high in areas in which these processes are con-
centrated. Effective control procedures for these processes are lacking.
Current data suggest the following contributions of major source
categories to the total national POM emission inventory (expressed in
terms of annual estimated benzo[a]pyrene emissions): heat and power
237
-------
238 PARTICIPATE POLYCVCLIC ORGANIC MATTER
generation, 500 tons/year; refuse burning, 600 tons/year; coke pro-
duction, 200 tons/year; and motor vehicles, 20 tons/year.
These data represent nationwide estimates based on extrapolations
from individual source emissions. In specific areas, the relative con-
tribution of any given source may differ significantly from that
implied by the nationwide figures. For example, the vehicular source
may be the major contributor in suburban areas where other major
sources are absent.
ATMOSPHERIC PHYSICS OF PARTICULATE
POLYCYCLIC ORGANIC MATTER
Polycyclic organic matter detected in the atmosphere has been identi-
fied with particulate matter, especially soot. It is uncertain whether
POM condenses out as discrete particles after cooling or condenses on
surfaces of existing particles after formation during combustion.
Particle size, surface area, and density are physical properties that
have the greatest influence on the behavior of POM-containing aerosols.
Generally, the particle-size spectrum of the atmospheric aerosol ex-
tends from less than 0.01 pm to greater than 10 /urn. Very few measure-
ments of mean surface area and density are available, but they are
believed to range from 2 to 3 m2 /g and from 1 to 2 g/cm3, respectively,
for urban areas.
POM appears to be associated largely with particles less than 5 pm
in diameter. Although large local variations are tabulated by the
National Air Surveillance Network, suspended particulate-matter
concentrations in U.S. urban areas are 100-200 MS/m3, as measured
by high-volume sampling. The benzene-soluble portion of this material
is approximately 10% by weight, but can vary from 8 to 14% in urban
areas. The POM component, as measured by benzo[a]pyrene, is less
than the benzene-soluble fraction by a factor of 10 or more.
Because POM is carried by suspended particulate matter, its lon-
gevity in air depends on the lifetime of the carrier aerosol in air and on
chemical alteration of POM itself. Initial estimates of atmospheric
residence times of particles less than 5 jum in diameter exceed 100 la-
under dry atmospheric conditions. Chemical reactivity in the presence
of sunlight may lead to transition of POM to other material in several
hours. Without sunlight, its lifetime may be much longer.
-------
General Summary and Conclusions 239
CHEMICAL REACTIVITY OF POLYCYCLIC
AROMATIC HYDROCARBONS AND AZA-ARENES
Polycyclic aromatic compounds are highly reactive. Evidence suggests
that they are degraded in the atmosphere by photooxidation, reaction
with atmospheric oxidants, and sulfur oxides. Comparative data on
reactions in solution, vapor, and adsorbed phases are very limited, and
the great bulk of the available information pertains to solution re-
actions. In the few cases in which evidence is available, the reactions
in other phases are similar. Reactions may be particularly facile when
the compounds are adsorbed on such particulate material as soot.
Chemical half-lives may be only hours or days under intense sun-
light in polluted atmospheres.
Most of the likely atmospheric reactions produce oxygenated com-
pounds from the hydrocarbons. Several mechanisms involving aro-
matic hydrocarbons and other pollutants may cause reactive oxidizing
species to be delivered to genetic and other biologic material.
THEORETICAL ASPECTS OF CHEMICAL CARCINOGENESIS
Chemical carcinogens appear to transform normal cells directly into
cancer cells. Chemical carcinogens may or may not "switch on" a
latent oncogenic virus that is responsible for cancer induction. If the
chemicals transform normal cells into cancer cells without the inter-
vention of an oncogenic virus, they can do so either by a mutational
or by a nonmutational mechanism. There is some evidence in favor of
each possibility.
Tumors induced by carcinogenic hydrocarbons have individual anti-
gens, and the response of a host to such stimuli is determined by its im-
munologic and hormonal status, its exposure to particular drugs, and
its nutritional state. A variety of unknown host factors may influence
the response to carcinogenic stimuli.
EXPERIMENTAL DESIGN IN CARCINOGENESIS TESTS
The proper design of a carcinogenicity testing program can depend on
a wide variety of factors. Its purpose may be to determine safe con-
centrations of carcinogens, to identify agents with interesting prop-
erties for future investigation, to identify specifically potent carcinogens,
-------
240 PARTICULATE POLYCYCLIC ORGANIC MATTER
or even to identify specifically weak carcinogens. Screening experi-
ments with many dosages can be informative, even if the group sizes at
individual dosages are limited. Data from a carcinogenicity experiment
can be complex, but simplifying methods for analyzing the data may
exist. Flexibility of analysis may be required to overcome unantici-
pated problems, for instance, a high rate of spontaneous tumors among
controls. The existence of thresholds in carcinogenesis cannot be
established solely by testing programs, and experimentation solely to
such ends would be wasteful.
Neither epidemiologic nor experimental data are adequate to de-
termine a safe dosage of any chemical carcinogen below which there
will definitely be no tumorigenic response in humans. For these rea-
sons, synthetic chemicals, such as food additives and pesticides, that
are known to be carcinogenic must not be deliberately added to the
environment. One must always insist on the lowest possible exposure
to air pollutants, which contain a variety of defined and undefined
carcinogens.
It is impossible to determine safe levels of human exposure to any
known or unknown carcinogen on the basis of supposed no-effect
levels in practical numbers of animals. Therefore, high dosages must be
used to obtain statistically significant data. Intratracheal instillation
of particles may serve as a model for human exposure to air pollutants.
Human experience has provided valuable post hoc information from
epidemiologic studies.
IN VIVO TESTS FOR CARCINOGENESIS
AND COCARCINOGENESIS
The most reliable test systems for measuring carcinogenesis in mice
and rats include application to the skin (mice), subcutaneous injection
(also in hamsters), administration in feed, inlraperitoneal injection,
inhalation tests, and bladder implantation. Factors that influence hydro-
carbon distribution in the host are particle size, retention and elution
of particles, changes in ciliary movement, and mucous viscosity.
The polycyclic aromatic hydrocarbons and heterocyclic compounds
constitute a group of known carcinogens that are present in the partic-
ulate phase of polluted air. However, the extent of the contribution
of these agents to the incidence of human lung cancer is unknown. A
variety of other compounds probably contribute to the human health
hazard. These include tumor-promoting agents, cocarcinogens, non-
-------
General Summary and Conclusions 241
carcinogenic initiating agents, and carcinogens other than particulate
aromatic hydrocarbons. The last compounds, usually nonparticulate,
include direct-acting alkylating agents, such as epoxides and lactones,
and peroxides and hydroperoxides of olefins and of aromatic hydro-
carbons. The role played by tumor-inhibitory or anticarcinogenic
agents in the health effects of air pollutants is, at present, poorly
understood.
Purified polycyclic compounds, such as benzo[a]pyrene, have pro-
duced tumors of the tracheobronchiolar tree or lung parenchyma
only when adsorbed on particles and delivered below the larynx. In
inhalation experiments, the addition of an irritant, such as sulfur
dioxide, to an aerosol of benzo[a]pyrene has induced lung carcinomas
in rats. Inasmuch as many potentially interacting influences are ubiqui-
tous in polluted air-including solid particles, irritant chemicals, and
gases-these may be cofactors as important for the induction of pul-
monary cancer as the polycyclic hydrocarbons themselves.
MODIFICATION OF HOST FACTORS IN
IN VIVO CARCINOGENESIS TESTS
An immunologic surveillance mechanism of a host against his own
tumor exists, although some evidence suggests that it may be relatively
weak. Many tumors have little or no immunogenicity, and this cannot
be attributed entirely to immunoselection. Even when potentially
immunogenic, a neoplasm may fail to immunize the host until late in
its course. Furthermore, immunodepression does not regularly result
in an increment in neoplasia. Other types of surveillance having little
to do with specific acquired immunity may also constitute a major
defense against incipient, chemically induced neoplasms.
It appears that newborn mice are often more susceptible than adult
mice to chemical carcinogenesis (particularly in organs distal to the
site of injection). This increased sensitivity, together with the require-
ment of smaller doses, suggests that the greater use of newborn mice
for carcinogenicity tests of air pollution fractions would be advisable.
Subhuman primates are susceptible to experimental chemical carcino-
genesis. Not all attempts to produce tumors with polycyclic hydro-
carbons have succeeded, and the lack of success raises questions about
resistance and the appropriate choice as to age and method of applica-
tion. Primates of the suborder Prosimii appear to be more susceptible
than other primates to carcinogenesis by these compounds, and the
-------
242 PARTICULATE POLYCYCLIC ORGANIC MATTER
tumors have shorter latent periods than in other primates. Pulmonary
carcinoma in simians has been produced by particles of beryllium
salts, but not by poly cyclic hydrocarbons. However, the latter com-
pounds have produced lung cancer in prosimian primates by intra-
tracheal instillation of particulate preparations.
Control of caloric intake to maintain normal body weight has
been shown by human insurance statistics and experiments in mice to
lower the hazard of developing cancer. Good reasons exist for an
adequate intake of protein and vitamins to provide an optimal ability
for tissues to detoxify the wide variety of environmental carcinogens
to which man is exposed.
Because of a lack of sufficiently well-controlled experiments, no con-
clusive evidence is available to support, or refute, the hypothesis of a
cocarcinogenic effect of respiratory infection. However, because res-
piratory infections are detrimental to a number of local and systemic
defense systems, and because they have profound effects on cell pro-
liferation and differentiation, the hypothesis of cocarcinogenicity of
respiratory infections is attractive.
It has been found that ionizing radiation combined with polycyclic
aromatic hydrocarbons produces an additive carcinogenic effect at
various sites. The induction of pulmonary tumors by irradiation has
been difficult to achieve because of difficulties in delivering the radia-
tion to the lungs. Progress has been made through the use of polonium-
210 adsorbed on hematite particles, and a combination of this and
benzo[a] pyrene (also adsorbed on hematite) delivered by intratracheal
instillation produced an additive carcinogenic effect.
DISTRIBUTION, EXCRETION, AND METABOLISM
OF POLYCYCLIC HYDROCARBONS
No definitive study on the metabolism, tissue distribution, and excre-
tion of carcinogenic hydrocarbons has yet been carried out. Radioactive
labeling of carcinogenic hydrocarbons and fractionation of the radio-
activity have been carried out. However, characterization of the com-
pounds has not been achieved. Complete characterization of the metab-
olites and excretion products, tissue distribution, and binding to macro-
molecules has not yet been attempted.
A number of metabolites of compounds like benzo[a] pyrene have
been identified. These include various dihydrodiols, phenols, quinones,
and glutathione conjugates. It is likely that an epoxide is the metabolic
precursor of these compounds. In the case of 7,12-dimethylbenz[a]-
anthracene, there is metabolic hydroxylation of the methyl groups.
-------
General Summary and Conclusions 243
The metabolic products just referred to are produced primarily
through aryl hydrocarbon hydroxylase, the drug-metabolizing enzyme
system of the microsomes. This complex system is found in many
tissues of many species and is inducible by polycyclic hydrocarbons
and a variety of other compounds, such as pesticides and drugs. The
enzyme system can also be inhibited by several compounds. The system
can increase or decrease the toxicity of hydrocarbons, and it is prob-
ably responsible for their metabolic activation to a chemically reactive
ultimate carcinogen.
IN VITRO APPROACHES TO CARCINOGENESIS
Two reliable and quantitative cell culture systems are now available
for the study of hydrocarbon carcinogenesis. One uses hamster embryo
cells in primary or secondary culture; the other uses cell lines derived
from adult mouse prostates. These systems show an excellent cor-
relation between the carcinogenic activity and the frequency of ma-
lignant transformed colonies produced by a series of hydrocarbons,
making them potentially useful for screening carcinogenic activities
of related compounds obtained from polluted air. Considerable funda-
mental information pertaining to the cellular and molecular mechanisms
of chemical carcinogenesis has been obtained with these systems.
Organ cultures maintain, in vitro, differentiated tissue organization
that resembles that in tissue in the intact animal. When organ cultures
of human and mouse embryo tracheas are exposed to polycyclic hydro-
carbons and fractions from polluted air, marked histologic alterations
can be observed in the epithelial cells and their organization. More-
over, by this technique, selected human tissues in organ culture can be
used to assess some of the biologic effects of air pollutants.
INDIRECT TESTS FOR DETERMINING THE POTENTIAL
CARCINOGENICITY OF POLYCYCLIC AROMATIC
HYDROCARBONS
The suppression of sebaceous glands in mouse skin after application
of polycyclic hydrocarbons is not a reliable indicator of carcinogenicity.
But it may have limited use in predicting the carcinogenicity of some
groups of compounds, such as substituted benz[a] anthracenes.
The economy, rapidity, and simplicity of the photodynamic killing
of paramecia, which can be conducted on less than 1-mg amounts of
-------
244 PARTICULATE POLYCYCLIC ORGANIC MATTER
organic extracts, are attractive features. The data suggest that this bio-
assay provides a biologic index of potential carcinogenic hazard at-
tributable to polycyclic compounds. However, evaluation of this con-
cept demands correlated photodynamic, carcinogenic, and chemical
studies on numerous samples and fractions of organic atmospheric
pollutants collected from sources exemplifying a wide epidemiologic
spectrum of incidence of respiratory tract cancer.
TERATOGENESIS AND MUTAGENESIS
Polycyclic hydrocarbons have not been shown to be teratogenic,
although a number of other chemical carcinogens exhibit this bio-
logic activity. The teratogenicity of community atmospheric pol-
lutants and defined components thereof have not been tested as yet
in mammalian species by inhalation or by parenteral administration.
A number of systems for determining the mutagenicity of atmo-
spheric pollutants in animal species have been described. They include
the dominant lethal assay, the host-mediated assay, and in vivo cyto-
genetics.
Although there is an association between mutagenic and carcinogenic
activities in a number of compounds, there is as yet no proof that the
two processes are closely related or that the mechanism of chemical
carcinogenesis involves a somatic mutation.
Recent technical developments have made it possible to test the
mutagenicity of chemical carcinogens in Chinese hamster cells in cul-
ture, by scoring for the production of drug-resistant or auxotrophic
mutants. This leads to the possibility of studying mutagenesis and
carcinogenesis simultaneously in the same cells.
VEGETATION AND POLYCYCLIC ORGANIC MATTER
No information was found to indicate that carcinogenic polycyclic
hydrocarbons affected vegetation. However, absorption of polycyclic
compounds by roots from contaminated solutions, by foliage from
polluted atmospheres, and by aquatic plants from contaminated
bodies of water increased the traces of these compounds already pro-
duced metabolically.
Burning of vegetation and some plant products may produce signifi-
cant quantities of several carcinogenic hydrocarbons. Increased con-
-------
General Summary and Conclusions 245
centrations of these materials in organic soils and in sediments in large
bodies of water suggest that many of the polycyclic compounds may
be produced in decayed organic matter.
EFFECTS OF POLYCYCLIC ORGANIC MATTER ON MAN
There is clear evidence that airborne POM found in occupational
settings-especially in relation to the products of burning, refining,
and distilling of fossil fuels-is responsible for specific adverse biologic
effects in man. The effects include cancer of the skin and lungs, non-
allergic contact dermatitis, photosensitization reactions, hyperpigmenta-
tion of the skin, folliculitis, and acne. In concentrations found in
urban or nonurban air, POM does not appear to cause any of those
skin effects; similarly, there is no clear evidence that such materials as
benzofa] pyrene themselves in polluted air directly influence the path-
ogenesis of such nonneoplastic lung diseases as bronchitis and
emphysema.
There is convincing statistical evidence of a dominant relation
between cigarette smoking and lung cancer in man; one important
factor in that relation is the polycyclic aromatic hydrocarbons, such
as benzo[a] pyrene. Even in this lung cancer system, factors other than
polycyclic aromatic hydrocarbons, such as phenols, may act as cocar-
cinogens or as accelerators.
Both animal experiments and epidemiologic data indicate that pul-
monary cancer of environmental origin involves a complex series of
factors and events in which polycyclic aromatic hydrocarbons con-
stitute only one of the carcinogenic agents, that chemical cocarcinogens
are also involved, and that the effects of particles, injurious gases, and
coexistent viral and other pulmonary diseases must be considered.
The POM found in high concentration in cigarette smoke causes
cancer of the lung and other organs in experimental animals; it is also
present in the industrial environment, in which lung cancer rates
are high, and is found generally in the air of urban communities. Exam-
ination of epidemiologic studies shows that, although a major factor
in the causation of lung cancer in man is cigarette smoking, it does not
account completely for the increased incidence of this disease. It
appears that the incidence of lung cancer among urban dwellers is
twice that of those living in rural areas; and within urban com-
munities, the incidence is even greater where fossil-fuel products
from industrial usage are highly concentrated in the air.
-------
246 PARTICULATE POLYCYCLIC ORGANIC MATTER
It appears, then, that there is an "urban factor" in the pathogenesis
of lung cancer in man. The poly cyclic organic molecule mentioned
most prominently in this report has been benzo[a] pyrene. It was felt
that benzo[a] pyrene could be used as an indicator molecule of urban
pollution, implying the presence of a number of other poly cyclic or-
ganic materials of similar structure that may also have some carcino-
genic activity. The standard measure of benzo[aj pyrene concentration
in the air is the number of micrograms per 1,000 m3 of air. On the
basis of epidemiologic data set against information regarding the benzo-
[a] pyrene content of the urban atmosphere, one can develop a working
hypothesis that there is a causal relation between air pollution and the
lung cancer death rate in which there is a 5% increase in death rate for
each increment of urban air pollution. In this study, an increment of
pollution corresponded to 1 ng of benzo[a] pyrene per 1,000 m3 of air.
On the basis of this assumed relation, a reduction in urban air pollu-
tion equivalent to 4 benzo[a] pyrene units (i.e., from 6 /zg/1,000 m3
to 2 jug/1,000 m3) might be expected to reduce the lung cancer death
rate by 20%. These data, however, are not to be interpreted as indi-
cating that benzo[a] pyrene is the causative agent for lung tumors.
There is much to support the idea of synergism or cocarcinogenesis,
especially with respect to cigarette smoking. In addition, the carcnio-
genic significance of other polycyclic organic molecules in urban air
pollution should be determined.
Prospective epidemiologic work correlated with analytic environ-
mental surveillance has not been done that would provide insight into
the true role of polycyclic aromatic hydrocarbons in atmospheric pol-
lution as related to human disease in general and to lung cancer in
particular.
-------
19
Recommendations for Future
Research
SOURCES OF POLYCYCLIC ORGANIC MATTER
1. Close scrutiny should be directed to deterioration effects of
automobile control devices and the use of diesel-fueled vehicles
under overloaded conditions.
2. Research into the effects of fuel compositions and of advanced
emission control devices should be continued.
3. POM emissions from aircraft should be assessed.
4. Substitution of alternate fuels or more efficient combustion pro-
cesses and discontinuance of coal-refuse storage practices seem to be
appropriate methods for the restriction of coal-related POM emissions.
5. Emission associated with coke production requires additional
research on control procedures and source analysis.
ATMOSPHERIC PHYSICS OF PARTICULATE
POLYCYCLIC ORGANIC MATTER
1. Knowledge of the behavior of aerosols is essential to understand-
ing the fate of POM in the atmosphere. Additional data relative to the
247
-------
248 PARTICIPATE POJLYCYCLIC ORGANIC MATTER
physical properties of atmospheric aerosols are needed. Simple and
inexpensive instrumentation is required to obtain size-weight concen-
tration data, particularly during short periods (minutes).
2. Further information should be obtained on the chemical and
physical forms of PO M in air, as well as details of its association with
suspended particulate matter, especially with respect to particle size
and chemical composition.
CHEMICAL REACTIVITY OF POLYCYCLIC AROMATIC
HYDROCARBONS AND AZA-ARENES
1. More definite information on chemical half-lives under various
conditions is essential.
2. Further research is needed on the products of chemical reaction
under atmospheric conditions and on the possible biologic activity
of these products.
3. Mechanisms that deserve further study are those involving aro-
matic hydrocarbons, which may cause reactive oxidizing species to be
delivered to genetic and other biologic material.
STUDIES OF POLYCYCLIC ORGANIC MATTER
IN ANIMALS, MAMMALIAN CELLS, AND VEGETATION
1. Improved methods for studying the genetics of mammalian
cells should be developed.
2. Further biologic, biochemical, and molecular biologic re-
search should be done in the fundamental mechanisms of chemical
carcinogenesis, which could lead to the eradication of cancer by
prophylaxis and possibly the reversion of cancers to normality.
3. Further exploration of the use of artificial atmospheres, such
as benzo[a]pyrene and sulfur dioxide, for direct inhalation carcino-
genesis tests is necessary, as well as the further use of POM adsorbed
on particles to test a wide variety of air pollution fractions and sub-
fractions.
4. The chemistry and biologic activities of such airborne co-
carcinogens and tumor-promoting compounds as polyphenols and
paraffinic hydrocarbons should be studied further, as should the
activities of the oxidation products of airborne olefins and aromatic
hydrocarbons-in particular, the chemical nature and carcinogenic
-------
Recommendations for Future Research 249
and other biologic properties of the epoxides, hydroperoxides,
peroxides, and lactones.
5. An attempt should be made to demonstrate lung cancer in
outdoor animals (birds) exposed to a highly polluted urban environ-
ment in which lung cancer in man is unduly high, comparing with
a sheltered flock of the same origin and age distribution breathing
purified air but otherwise under identical dietary, sanitary, and other
conditions.
6. Greater use of newborn mice should be made in testing the
carcinogenicity of fractions of polluted air.
7. Tests of domestically bred simian primates by the same
methods that succeeded in prosimian primates and rodents should
be conducted to provide data on their relative susceptibility to sys-
temic, skin, and bronchial carcinogenesis.
8. A definitive study of the distribution, excretion, and metabo-
lism of a polycyclic carcinogenic hydrocarbon should be conducted,
with identification and characterization of all metabolites.
9. The microsomal aryl hydrocarbon hydroxylase system and the
effects of various inhibitors should be assayed in a variety of human
tissues.
10. A reliable source of standard preparations of polycyclic
hydrocarbons and their metabolites and of specific inducers and
inhibitors of the microsomal enzyme system should be provided.
11. Parallel tests for carcinogenicity of chemicals and fractions
of polluted air should be conducted in animals and cell cultures to
permit decisions on the usefulness of the in vitro systems.
12. Organ cultures should be used to study the histologic ef-
fects of POM on organized differentiated tissues, with particular
attention to epithelial cells.
13. Increased emphasis on the testing of teratogenic activity of
suspected carcinogenic fractions is needed.
14. Increased emphasis on testing for mutagenic activity of sus-
pected carcinogenic fractions in animal systems is needed in order
to gain further information on the relation between the processes
of carcinogenesis and mutagenesis.
15. A screening committee should be appointed to establish the
criteria for an environmental agent to be regarded as a carcinogen
or tumor-initiator. A list of the environmental carcinogens and
tumor-initiators should be compiled and made available.
16. The effects of traces of carcinogenic materials in vegetable
foods on the incidence of cancer in man and animals should be
-------
250 PARTICIPATE POLYCYCLIC ORGANIC MATTER
determined. The effects of long-term exposure to, and massive
dosages of, polycyclic compounds on plant growth, development,
and reproduction should also be investigated.
EPIDEMIOLOGIC STUDIES OF NONOCCUPATIONAL
NEOPLASTIC PULMONARY EFFECTS
1. More precise quantitation of exposure and definition of
ambient air concentrations of PO M and other possible carcinogens,
as well as all other major pollutants, is needed. This includes particu-
late matter, which may act as adsorbents and carriers of POM and
irritant gases that, by interfering with and slowing pulmonary clear-
ance, may increase the duration of contact between carcinogenic
materials and bronchial mucous membranes.
2. Sampling and quantitation data are needed for every major
city, so that a reasonable estimate of ambient air concentrations
may be obtained.
3. Much greater documentation of cigarette smoking is needed.
The exclusion from the 1970 Census of this major etiologic factor
in disease is unfortunate. Valid estimates of cigarette consumption
in major community areas, both urban and rural, in relation to
lung cancer and other major disease entities are not easily available.
4. Additional sampling studies of cigarette smoking, occupation,
and residence in well-defined populations are required.
5. Modern statistical methods should be used to examine the
role of airborne carcinogens other than POM in disease and to investi-
gate the relations of these and other environmental factors to non-
pulmonary neoplasms.
6. The environment-cancer association is weakened by the finding
of low benzo[a]pyrene concentration and high carcinoma incidence
in a number of communities. Studies seeking other etiologic agents
in these localities should be carried out.
7. More extensive investigation into the effects of airborne car-
cinogens should be carried out where they appear in industries in high
concentrations. This would be particularly valuable if workers with
heavy exposure to the same materials in different communities
throughout the country were compared.
8. Migrant studies reveal strong epidemiologic evidence of a relation
between airborne carcinogens and lung cancer. More detailed explora-
-------
Recommendations for Future Research 251
tion of the selection of the population subgroups that migrate is
necessary to rule out selection as a possible factor in these differences.
9. The results of studies carried out for this report suggest the
feasibility and desirability of further epidemiologic studies on airborne
carcinogenesis.
-------
Appendix A
Collection of Airborne Particles
for Analysis of Polycyclic
Organic Matter
Advances in methods of collection and separation of poly cyclic
aromatic hydrocarbons involving chromatography, absorption, and
fluorescence spectroscopy have led to the development of a number
of methods for the estimation of at least 25 poly cyclic aromatic hydro-
carbons. Of these, about eight are now estimated more or less routinely
from ambient air samples in several laboratories in a number of
countries. Data are available on the monthly and seasonal distribution
of these compounds in an increasing number of cities, some of which
show relatively high atmospheric concentrations. In the usual mode
of separation, more or less quantitative information can be obtained
on the concentrations of pyrene, fluoranthene, chrysene, benzo[a]-
pyrene, benzo[e] pyrene, benzo[k] fluoranthene, benz[a] anthracene,
perylene, benzo[ghi]perylene, coronene, and anthanthrene. Sawicki
and co-workers667 have examined the air of more than 130 urban and
nonurban areas in all sections of the United States and found that the
powerful carcinogen benzo[a] pyrene is universally present, in varying
concentration.
The atmospheric concentrations of polycyclic aromatic hydro-
carbons are so small, even in heavily populated cities, that the results
are reported in micrograms per 1,000 m3 of air sample. In general,
253
-------
254 Appendix A: Collection for Analysis of Airborne POM
benzo[a]pyrene increases markedly from a relatively low concentration
in the summer months to a maximum during the coldest months of
the heating season, the time of increased consumption of fuels for
heat and power. Concentrations may vary from less than 1 /zg/1,000 m3
in relatively clean air to over 100 /zg/1,000 m3 in smoky, polluted air
of a large city. Such trace quantities imply the sampling of extremely
large quantities of air to collect enough smoke or particulate matter by
filtration, or other means, for a sufficient sample for subsequent
analysis. Usually a minimum of about 10,000 m3 of air sample is re-
quired in order to collect by column chromatography a sufficient
quantity of POM for determination of individual compounds by
absorption spectroscopy.
SELECTION OF SAMPLING SITES
To assess the exposure of man to carcinogens in air, long-term average
concentrations are considered to be of prime importance, whereas
short-term fluctuations may have little meaning. Because airborne poly-
cyclic aromatic hydrocarbons and other carcinogenic substances are
derived from a variety of fluctuating sources, it is difficult to deter-
mine overall exposures of population groups except on the basis of
mean levels of individual carcinogens and mixtures of known com-
position and potency over long periods.
In general, pollution levels in cities and suburban environments
fluctuate widely, in accordance with prevailing meteorologic conditions.
Other influences on pollutant concentrations include topography,
nature and distribution of sources, distance from sources upwind and
downwind, and seasonal and annual variations.
Selection of sites to ensure representative sampling must be under-
taken with great care in order to incorporate the various influencing
factors, at least on a sound statistical basis. An illustration of this
approach is provided by the studies of smoke and polycyclic aromatic
hydrocarbon content of the air in two pairs of European cities (Belfast
and Dublin, and Oslo and Helsinki), as described by Waller and
Commins.800 They were undertaken as a pilot project under the super-
vision of the Cancer Study Group of the World Health Organization.
Each of the four cities was divided into five areas having roughly equal
populations, and the sampling sites were placed as close as possible to
the centers of population density of the areas. Great care was taken in
selecting sampling sites to avoid proximity either to individual
sources of pollution or to open spaces. At each location, the samplers
-------
Appendix A: Collection of Particles for Analysis 255
for the collection of suspended particulate matter (smoke and poly-
cyclic aromatic hydrocarbons) were operated for a period of 12
months.
The general considerations involved in the selection and location
of sampling sites have been discussed by Katz.428 For the study of
health effects, the sampling sites must ensure that the sample col-
lections are truly representative of air that is actually breathed by
the exposed population groups. Concentrations of pollutants vary
with height of the sampling points above ground level, so results
from sites on the roofs of buildings may differ materially from re-
sults from sites at breathing level. A representative number of sampling
stations for a given area may be established after a preliminary study.
This type of study should include information on the nature and magni-
tude of emissions from principal sources of pollution, a review of
available climatologic and meteorologic data, and the gathering of
some preliminary data on air pollutant concentrations in areas of
severe and slight pollution.
SELECTION OF EQUIPMENT AND FILTERS
The general practice in sampling and collection of polycyclic aromatic
hydrocarbons and related heterocyclic compounds of high molecular
weight involves filtration of air to collect the suspended particulate
matter. In the United States and Canada, the equipment used for
this purpose is a high-volume sampler, which consists of a suction fan
operated by a motor and equipped with a filter holder and calibrated
air flow gauge or manometer. The filter is a rectangular fiber glass
sheet of high collection efficiency and low resistance, 20 X 25 cm.
Current samplers are exposed inside a louvered box that holds the
filter surface, facing upward and horizontal, under a roof that pro-
tects the filter from rain and snow. The air sampling rate varies
from about 1.2 to 1.7 m3/min. Flash-fired fiber glass filters are
used; they have an efficiency of about 99.97%, despite their low
resistance to air flow. Particulate samples are normally collected
over periods of 24-48 hr.
In West Germany, the Institut fur Lufthygiene uses, in addition to
a high-volume sampler, the BAT I,* the BAT II,* and the Draeger
*Air sampling instruments, similar to the high-volume sampler, that collect air-
borne dust or particulate matter; coarse particles larger than 10 jum are removed
before passage of air sample through the system. Thus, only respirable particles
are collected by the filters.
-------
256 Appendix A: Collection for Analysis of Airborne POM
instrument. The air flow rate of the BAT I is 10 m3/hr; of the BAT
II, 100 m3/hr; and of the Draeger instrument, 3 m3/hr. The filters
used in the studies described by Waller and Commins800 were
fiber glass, Whatman GF/A type, 12.5 cm in diameter, which col-
lected particulate matter containing polycyclic aromatic hydro-
carbons by means of low-volume samplers with an air flow rate of
5 m3 /day. Each sampler was operated continuously for a month to
provide a sufficient sample of particulate material for gravimetric
estimation. The determinations of polycyclic aromatic hydrocarbons
were carried out on bulked samples for 6-month periods, mainly during
the summer and winter seasons.
It has been shown that some polycyclic aromatic hydrocarbons are
sufficiently volatile to evaporate during collection if long-period
sampling procedures at low flow rates are used. Although this is true
especially for the lower-molecular-weight hydrocarbons, it is impossible
to collect anthracene, phenanthrene, pyrene, and fluoranthene ef-
ficiently. It is evident that accuracy requires collecting samples at
comparatively high flow rates during short periods (about 24-48 hr)
and analyzing them in the laboratory as soon as possible. In special
cases of high air pollution levels, high-volume filters may become
blocked in less than a day. It has been suggested that samples be col-
lected at intermediate flow rates over a period of a week and that
weekly samples be pooled to yield material adequate for the analysis
of polycyclic aromatic hydrocarbons representing the four seasons
of the year.
FILTRATION AND STORAGE
The analytic procedure recommended by the Intersociety Committee
on Methods for Ambient Air Sampling and Analysis of the American
Public Health Association involves the collection of particulate matter
with high-volume air samplers on flash-fired fiber glass filters.746'747
Before use, the filter should be washed thoroughly with pentane to
make the blank as low as possible. The amount of particulate matter
collected depends on many variables, such as the particulate loading
in the atmosphere, sampler location, and volume and rate of air
sampled. On the average, in an urban area, the sampler will collect
approximately 250-350 mg of particulate matter while sampling
2,000-2,400 m3 of air during a 24-hr period. Of this quantity of
particulate matter, approximately 6-10% will be soluble in benzene.
-------
Appendix A: Collection of Particles for Analysis 257
Hence, this benzene extract or organic fraction will contain about
25 mg for an average high-volume sample. A quantity of benzene-
soluble material amounting to 50-150 mg is needed for analysis of
polycyclic aromatic hydrocarbons. To obtain this amount of material,
it is necessary to pool the organic fractions of several individual
fiber glass samples from a single site.
The procedure for weighing the filters before and after sample col-
lection should be standardized to control these conditions, preferably
at 25 C with a constant relative humidity below 50% in a conditioning
chamber. The upper limit of particle size collected by the high-
volume sampler is probably less than 50 jum in diameter in the standard
shelter of the U.S. National Ah- Sampling Network (NASN). The size
distribution of the material on the filter has not been measured directly.
However, it can be deduced that about 90% of the weight of suspended
particulate matter reported by the NASN can be associated with par-
ticles whose terminal settling velocities are less than that of a sphere
approximately 8 jum in radius with a density of 1 g/cm3.
With the BAT II instrument, the upper limit of size of the filtered
particles is stated to be about 5-7 pm in diameter; and with the
Draeger instrument, 10 jum. It is believed that sampling instruments
that eliminate the coarser particles (i.e., larger than 10 [Jan), which
cannot be inhaled, and collect the particles in the respirable range are
best suited for evaluation of the hazard to human health of poly-
cyclic aromatic hydrocarbons and related carcinogens. It is believed
that no significant amounts of polycyclic aromatic hydrocarbons,
especially those of high molecular weight, exist in the vapor phase in
the ambient atmosphere. Consequently, they can all be collected
by filtration of the particulate phase in appropriate sampling con-
ditions.
The volatility of some polycyclic aromatic hydrocarbons is of con-
cern in connection with filtration of particulate matter, especially at
low flow rates over periods of several weeks or longer. Filtration
media should consist of low-resistance fiber glass filters without or-
ganic binders. Membrane filters are generally unsuitable for air sampling
for analysis of polycyclic aromatic hydrocarbons, because they have a
high air flow resistance and undesirable solubility in organic sol-
vents. Accurate calibration of flow meters or flow gauges is essential
for the correct measurement of air volume during sampling.
Samples should be stored in the dark in cleaned glass vessels fitted
with ground-glass stoppers, preferably in a refrigerator. The use of
paper or polyethylene bags for storage is not recommended, because
-------
258 Appendix A: Collection for Analysis of Airborne POM
paper can absorb hydrocarbons from collected samples and poly-
ethylene contains oils and antioxidants that may contaminate the
samples. The organic fractions of samples remain stable for a period
of 5 years if stored in a dark refrigerator. Exposure to light, especially
ultraviolet, destroys some polycyclic aromatic hydrocarbons.262
STANDARDIZATION
Although much information is available on the atmospheric concentra-
tions of polycyclic aromatic hydrocarbons of cities and industrial areas,
the data derived by different methods of sampling and analysis can-
not be directly compared or properly assessed. Samples collected by
low-volume air flow filtration over periods of many weeks cannot be
compared directly with those obtained by high-volume filtration over
1 or 2 days, owing to losses by volatilization of some polycyclic aro-
matic hydrocarbons in low-volume filtration sampling. For assessment
of the potential hazards of airborne carcinogens to public health,
only the respirable portion of the suspended particulate matter is
useful. Samples in the required particle size range should be collected
by similar methods of sampling and analysis.
A sampling system for the assessment of urban aerosol particle
size-weight distribution has been described by O'Donnell et al.568 The
system includes a six-stage Andersen cascade impactor,14 a backup
filter, a vacuum pump, a critical-sized orifice, and a special cover that
simulates entry conditions of the NASN high-volume sampler. After
calibration of the Andersen sampler, particle size distributions can be
determined directly from the particle weights on the six collection
plates and the backup filter. To determine the total atmospheric
particle concentration, the wall losses on the top impactor sieve of
the Andersen sampler must be measured. Wall losses do not affect
size distribution results with atmospheric suspended particles smaller
than 9 nm in diameter, which is the size associated with deposition in
human lungs. Particles larger than 9 ion can be determined by mea-
suring these sieve losses. Alternatively, the fraction of particles larger
than 9 pm may be estimated by taking the difference between the
total atmospheric particle concentration, as measured by a high-
volume sampler, and the total concentration of the particles smaller
than 9 jum, measured by using five stages of the Andersen impactor
and the backup filter.
Results for total concentration obtained using the Andersen sampler,
-------
Appendix A: Collection of Particles for Analysis 259
corrected for sieve losses, were found to compare favorably with the
total concentration measurements obtained simultaneously with the
high-volume sampler. Only a slight error is caused by the differences
in collection efficiencies between the high-volume filter and the backup
membrane filter used with the Andersen sampler.
Table A-l lists the effective cutoff diameters reported for particles
collected on stages of the Andersen sampler during 24-hr sampling
periods.568 The average weight of particles collected in 24 hr was
137 MS/m3 of air sampled.
SUMMARY
High-volume filtration samplers are used routinely to collect atmo-
spheric particulate matter on fiber glass mats for periods of 24 hr or
more by the U.S. National Air Sampling Network. This type of
sampling is adequate for determining the concentration of POM and
of individual polycyclic compounds. However, such high-volume air
samples do not provide information relative to aerosol particle size-
weight distribution.
Emissions at various sources are sampled by filtration after passing
through a cooling train. No information is available on sample losses
in these sampling trains or on the errors associated with sample col-
lection.
TABLE A-l Effective Cutoff Diameters of Particles Collected on Stages of
Andersen Sampler (24-hr period)
Stage
Sieve 1
Stage 1
Sieve 2
Stage 2
Stage 3
Stage 4
Stages
Stage 6
Backup filter
Effective Cutoff
Diameter, /am
>9.2
9.2
_
5.74
3.3
1.76
0.98
0.50
<0.50
Stage Weight,
mg/24 hr
0.83
1.0
0.13
0.51
0.40
0.32
0.25
0.33
1.49
Fraction of
Sample Weight, %
15.8
19.0
2.5
9.7
7.6
6.1
4.7
6.3
28.3
Total - 5.26 100.0
-------
260 Appendix A: Collection for Analysis of Airborne POM
RECOMMENDATIONS
International standardization of sampling and analytic procedures
in this field is urgently required, so that data collected in different
countries can be evaluated. Only then can valid information be ob-
tained in epiderniologic and biologic studies for the establishment of
adequate criteria and standards of air quality on a worldwide basis.
Information on aerosol particle size-weight distribution is essential
for the prediction of pulmonary responses in man after inhalation of
aerosols.
-------
Appendix B
Separation Methods for
Polycyclic Organic Matter
In vivo bioassays have demonstrated that the greatest carcinogenic
activity of organic pollutants is associated with the neutral fractions
(nonacidic and nonbasic) that are enriched in polycyclic organic
matter.388>404 Present knowledge indicates that those fractions con-
tain only two classes of compounds that include known animal carcin-
ogens—the polycyclic aromatic hydrocarbons and their neutral
nitrogen analogs, the aza-arenes (e.g., indoles and carbazoles). The
basic aza-arenes constitute a third class of polycyclic organic pollutants
that includes some known animal carcinogens.346'699'700 Because the
basic compounds have in general represented only a minor portion of
organic pollutants, they have not been tested for carcinogenicity until
now. Hence, it cannot be stated with certainty that the basic aza-
arenes contribute to the overall carcinogenicity of the organic matter
from polluted air. Nevertheless, until proved otherwise in animal ex-
periments, the basic aza-arenes, polycyclic aromatic hydrocarbons,
and neutral aza-arenes should be analyzed in evaluating the carcino-
genic potential of organic pollutants.
261
-------
262 Appendix B: Separation Methods
STANDARD LABORATORY CONDITIONS
Precautions
All known carcinogenic polycyclic aromatic hydrocarbons absorb
light at wavelengths of 350-450 nm and, in the presence of oxygen,
may be oxidized.464'587 It is recommended that the collecting de-
vice be protected from direct exposure to sunlight and from the
radiation of fluorescent lamps. It is also advisable that analyses be
performed in laboratories that are illuminated with yellow light (no
radiation below 450 nm).
Hydrocarbon POM with four or more rings is quantitatively col-
lected by filtering samples in normal atmospheric conditions with
large-volume samplers.154 Commins demonstrated that, at normal en-
vironmental temperatures (up to 30 C) and with proper storage, de-
composition of even the most unstable aromatic hydrocarbon can be
avoided.154 During analysis, organic solvents have to be evaporated
to concentrate the POM. The loss of POM during this procedure can
be reduced or-even avoided by working at pressures above 12 mmHg
and at water-bath temperatures below 45 C. If one must concentrate
the PO M from large volumes, evaporation should take place in distil-
lation columns operated at least at a 2:1 reflux ratio. Often, especially
with gas or liquid chromatography, the PO M solution has to be con-
centrated to a rather small volume—perhaps only a few milliliters—in
which case the PO M solution should be concentrated by freeze-drying.
Some basic aza-arenes, to be isolated, must be concentrated by
extracting the POM solution with acids and then basifying with alkali,
preferably at low temperature (e.g., by external cooling with ice
water).
Internal Standards
Despite the greatest of precautions, there are some unavoidable
losses during the analysis of POM . These losses may vary from 10 to
40%, depending on the sample and the experience of the chemist. One
way to overcome these uncertainties is to standardize the analytic
method. It has been demonstrated, however, that samples of polycyclic
organic air pollutants can vary significantly in composition and con-
centration, even though the particles are collected at the same site but
in different pollution conditions. Therefore, the best method to
secure quantitative data for individual polycyclic organic compounds
-------
Appendix B: Separation Methods 263
is to impose internal standards at the beginning of the analysis. This
can be done by using the isotope-dilution method-a technique
that yields quantitative data for individual hydrocarbons accurate to
within 8%. Either the original sample is supplemented with one or
several hydrocarbons not present in the original sample but concen-
trated with the test carcinogenic polycyclic organic compounds, or
traces of a known carcinogenic hydrocarbon labeled with carbon-14
are added (tritium-labeled polycyclic aromatic hydrocarbons should
not be used, because an exchange between tritium and hydrogen can
occur during analysis).387
Reproducibility
In analyzing carcinogenic polycyclic hydrocarbons in the respiratory
environment, some factors that can affect the reproducibility of
analysis must be recognized, particularly the contamination of the
sample with POM from solvents, adsorbents, and the environment
and the poor quality of commercially available reference compounds.
These sources of error in PO M analysis cannot be overcome by the use
of an internal standard and must therefore be avoided by purifying
the solvents and agents to be used and by restricting the analytic work
to laboratories that are ventilated with filtered air. To avoid possible
cross-contamination, it is essential that reference compounds be
purified in a separate laboratory under a properly working hood.
DISTRIBUTION
Distribution between Solvents
The organic matter, or "tar," in gasoline and diesel engine exhaust
fumes is rich in aliphatic hydrocarbons. For example, an air pol-
lution sample from a Detroit collection site with high traffic density
consisted of more than 48% of nonpolar hydrocarbons (neutral
fraction N-l ).388 The corresponding figure for the N-l fraction from
exhaust "tar" from a standard test-stand run of a conventional gasoline
engine was more than 15%.384 Because high concentrations of paraffins
and olefins are known to reduce the efficiency of the column chroma-
tographic separation of POM, it is advisable to separate the bulk of
the aliphatic hydrocarbons from the POM. This can be done by
distribution between cyclohexane and nitromethane,389 cyclohexane
-------
264 Appendix B: Separation Methods
and dimethylsulfoxide (D. Hoffmann and G. Rathkamp, unpublished
data), «-hexane and dimethylsulfoxide, acetonitrile and H-hexane,328
or nitromethane and carbon disulfide.44 A good method involves
distribution between cyclohexane and dimethylsulfoxide and back-
extraction from the dimethylsulfoxide layer with cyclohexane after
addition of water (D. Hoffmann and G. Rathkamp, unpublished data).
With back-extraction, one has to evaporate only a relatively low-
boiling solvent.
Countercurrent Distribution
Several reports are concerned with the enrichment of POM from
environmental agents by countercurrent distribution. Demisch and
Wright195 suggest «-hexane and aqueous monoethanolammonium
deoxycholate as a solvent pair, and Mold et al.536 suggest cyclohexane
and methanol-water (9:1) containing 0.83% of tetramethyluric
acid. Selective separation systems like these may lead to the isolation
of hitherto unknown carcinogenic hydrocarbons from the respiratory
environment, especially if a 1,000-cell Craig countercurrent distribu-
tion apparatus is used. However, it is doubtful whether this method
has practical value for routine analysis of polycyclic organic pollutants.
Extraction of Basic Aza-Arenes
If all the basic aza-arenes must be concentrated in one fraction, the
concentrated solutions of the organic pollutants must be extracted
with strong inorganic acids—perhaps as concentrated as 10-15% sul-
furic acid. Lower acid concentrations do not lead to a quantitative
separation of these compounds from the neutral and acidic organic
pollutants.
COLUMN CHROMATOGRAPHY
Standard Conditions
The most widely used separation methods for the chemical analysis
of polycyclic aromatic hydrocarbons and aza-arenes are column and
ion-exchange chromatography. A large variety of adsorbents-such as
alumina, silica gel, Florisil, and cellulose acetate—and various ion-
exchange resins are selected for these separations. Adsorbents with
uniform particle size, either 80-100 or 100-120 mesh, and column
-------
Appendix B: Separation Methods 265
diameter: length ratios of at least 1 :25 are suggested. Laboratory
temperatures should be between 18 and 23 C and kept constant within
a few degrees. (The flow rate of columns is known to change in
laboratories whose temperatures change overnight or during week-
ends.)
The solvents to be used should be spectral-grade quality, or at
least purified through a distillation column with a reflux ratio of
2:1. It is advisable to consult Organic Solvents: Physical Properties
and Methods of Purification by Riddick and Bunger631 for the
purification of such solvents. Column separations generally begin with
a low-boiling alkane as the elution solvent. It is advisable to dry even
hydrocarbon solvents to avoid column deactivation (rc-hexane can
contain up to 0.01 vol.% water).
Several techniques are used for filling the column, mainly making
a slurry of the adsorbent in a hydrocarbon solvent or filling the column
partially with the solvent and pouring the adsorbent into it. In any
case, air bubbles have to be avoided during packing and operation.
The flow rate of the column is best controlled with a Teflon stop-
cock. A fraction collector is essential for establishing reproducible
separations. The separation can be based on either volume, timing,
or change of fluorescence on exposure to ultraviolet light of 365-nm
wavelength.
Maximal separation can be achieved by slowly and evenly increasing
the polarity of the solvent (gradient elution, e.g., increasing the pro-
portion of benzene in «-hexane with time). Further details on achieving
reproducible column chromatographic enrichments or separations of
POM are available in the literature.
Alumina
Alumina is one of the more strongly binding adsorbents. It is available
in various grades, sizes, pH's, and activities. One of the better standard-
ized preparations is Woelm alumina. In general, one should use alumina
with relative activities between 26 and 32 for the separation of poly-
cyclic aromatic hydrocarbons and neutral aza-arenes.369 The ratio of
POM concentrate to alumina will vary between 1 :100 and 1 :1,000,
depending on the degree of enrichment desired. It should be possible
to elute the polycyclic aromatic hydrocarbons in a reasonable time
with H-hexane with an increasing concentration of benzene up to
25%. Some investigators have successfully separated polycyclic aro-
matic hydrocarbons on very long alumina columns.143
These conditions may well be best for a laboratory that needs to
-------
266 Appendix B: Separation Methods
analyze only a few samples and is not specifically equipped for POM
analysis. However, the investigator should be aware that the analysis
may require a long time (5-10 days) and should therefore be performed
isothermally. Although it is known that some decomposition may oc-
cur on neutral alumina, this is significant only if one is also interested
in esters whose polarity is similar to that of poly cyclic aromatic hydro-
carbons. So far, alkaline alumina has not been applied to the analysis
of aza-arenes.
Silica Gel
Except for the silica gel developed specifically for thin-layer chroma-
tography, it is difficult to find commerically available silica gel that
contains no fluorescent materials. Therefore, silica gel has to be puri-
fied, either by washing it on the column before the actual separation
(which leads to partial deactivation) or by washing with a polar solvent
and later reactivating it. The sample: adsorbent ratio varies between
1 :50 and 1 :500. Another disadvantage of silica gel is its relatively
slow flow.
Florisil
Polycyclic aromatic hydrocarbons separate rather well on Florisil,
despite the short elution time. Florisil is weakly cationotropic (weakly
basic), which increases the possibility that traces of polycyclic aromatic
hydrocarbons may be partially photooxidized. If one is interested in
other neutral compounds with comparable polarity—such as indoles,
some esters, and DDT and DDD—one should avoid using Florisil to
prevent oxidation, hydrolysis, or dehydrohalogenation, respectively.
Chromatography on Cellulose and Modified Cellulose
Although the literature describes several powders of cellulose
derivatives—such as xanthates, succinates, and acetates—as promising
adsorbents, these chromatography materials so far have been tried
only once for the separation of polycyclic aromatic hydrocarbons.801
GEL FILTRATION
Gel filtration is of great value when used to supplement column
chromatography. During the last few years, gel filtration has led to the
-------
Appendix B: Separation Methods 267
isolation and identification of several natural products, biologically
active compounds, metabolites, and environmental agents.599 Wilk
et a/.819 first explored gel filtration for the separation of poly cyclic
aromatic hydrocarbons. Others have since found the system to be
most valuable for the separation of neutral aza-arenes385-386 and
aromatic hydrocarbons.624 The best results so far have been achieved
by separating concentrates of aromatic hydrocarbons isothermally
(at 32 C) on Sephadex LH-20 in long columns (ratio, about 1:1,000)
with propanol-2 as solvent. One disadvantage of the gel filtration
method is that it takes 1-3 days. Nevertheless, it is expected that gel
filtration will become more widely used for the analysis of poly-
cyclic organic pollutants.
THIN-LAYER CHROMATOGRAPHY
Since the development of thin-layer chromatography as a new analytic
tool, E. Sawicki and associates661'664'668 have skillfully applied this
technique to air pollution research. Today, various polycyclic air pol-
lutants are analyzed in accordance with the techniques developed by
them. Included are the analyses of polycyclic aromatic hydrocarbons,
aza-arenes, polycyclic carbonyl compounds, and phenols. Thin-layer
chromatography is quick, reproducible, and inexpensive; in the hands
of an experienced analyst, it leads to good separations of organic pol-
lutants, as reviewed recently.661
One point for discussion is the great emphasis on thin-layer chroma-
tography as a final step in isolating POM from urban air pollutants and
then assaying the extracts by spectroscopic methods. Mass spectrometry
and liquid chromatography, when applied as additional analytic tools
for the identification of aromatic hydrocarbons and aza-arenes, have
taught us that individual spots or bands from thin-layer chromatograms
are sometimes, in fact, mixtures of two or more compounds. Further-
more, during thin-layer chromatography and paper chromatography,
trace amounts of some polycyclic aromatic hydrocarbons and aza-
arenes may be decomposed. Therefore, internal standards should be
used to correct for the losses.
PAPER CHROMATOGRAPHY
Compared with thin-layer chromatography, paper chromatography,
although requiring significantly more time, affords a better separation.
-------
268 Appendix B: Separation Methods
Of course, evidence of purity or homogeneity based solely on chroma-
tographic methods should not be accepted as unequivocal. But either
method can serve as a useful guideline.
The method of Tarbell et a/.,743 using Whatman No. 1 paper and
dimethylformamide saturated with H-hexane, and the method of Spots-
wood,718 using acetylated paper, result in good separation of polycyclic
aromatic hydrocarbons. For the isolation of basic and neutral aza-arenes,
Van Duuren et a/.786 have developed a new paper chromatography sys-
tem. Some years ago, Sawicki664 reviewed in detail the use of paper
chromatography in air pollution research. A laboratory not equipped
with a gas chromatograph with electron capture detection or a liquid
chromatograph is well advised to use chromatography on acetylated
paper for the final separation of the carcinogenic hydrocarbons benzo-
[a]pyrene, benzo[b]fluoranthene, benzo[j]fluoranthene, indeno[l,2,3-
cdjpyrene, and chrysene.
PAPER AND THIN-LAYER ELECTROPHORESIS
E. Sawicki et a/.669 have developed methods for the separation of basic
aza-arenes by paper and thin-layer electrophoresis. For some of the aza-
arenes studied, encouraging degrees of separation were achieved, e.g.,
for the three benzacridines. These techniques are recommended as
additional methods for the qualitative analysis of basic aza-arenes in
polluted air.
Chemists, however, should be aware that paper, thin-layer, and col-
umn chromatography generally do not separate the individual alkylated
polycyclic aromatic hydrocarbons from each other. This separation can
be achieved only on gas chromatography columns with more than
25,000 theoretical plates and on liquid chromatography columns.
HIGH-SPEED LIQUID CHROMATOGRAPHY
Although liquid chromatography possesses many advantages for the
separation and isolation of organic compounds, the method has re-
mained unattractive owing to its long elution time and poor column effi-
ciency. Recent work, however, has shown that the speed and efficiency
of liquid chromatography can be greatly increased, to approach even that
of gas chromatography. Compared with gas chromatography, high-speed
liquid chromatography offers the advantage of operating at relatively
-------
Appendix B: Separation Methods 269
low temperatures (<100 C), thereby preventing thermal rearrangements
and decompositions. Depending on the chemical structure of the com-
pounds being analyzed, a differential refractometer detector or an
ultraviolet photometric detector (254 nm) is used. The detection limits
for most compounds are 1CT9 g for the refractometer detector and 10"1 °
g for the ultraviolet photometric detector for polycyclic aromatic hydro-
carbons that have at least four condensed aromatic rings. At present,
the use of high-speed liquid chromatography seems to be limited by the
high cost of an efficient instrument with a high-pressure pulse-free liquid
pump ($10,000+), the high cost of specific columns (which most labora-
tories are unequipped to prepare), and the rather cumbersome collection
of reagents, solvents, and standards required for the determination of
the identity and purity of unknowns. It appears for the time being that
gas chromatography and liquid chromatography will be used to comple-
ment each other.
Separation of polycyclic aromatic hydrocarbons by liquid chroma-
tography at 40 C and 80 atm on a 1-m column that is filled with a
specific-surface-porosity support and coated with a hydrocarbon poly-
mer has been reported.436 A mixture of water and methanol (6:4)
serves as the mobile phase. This system effectively separates the car-
cinogens chrysene, benzo[e] pyrene, and benzo[a] pyrene with reten-
tion times of 15, 22, and 27 min, respectively. Using 3-m columns with
the same stationary phase described above for carcinogens, but under
higher pressure and with a mixture of propanol-2 and water (1:4) as
the solvent, one can separate all 10 major four- and five-ring aromatic
hydrocarbons that are present in polluted air of New York City. Re-
cently, Ledford e?a/.484 separated arenes on a 1-m column filled with
Durapak OPN, with 0.25%"methylisobutylketone in heptane as the
moving phase. On Corning CPG glass beads treated with octadecyltri-
chlorosilane as the stationary phase and acetonitrile as the moving
phase, the same investigators completely separated benzo[e] pyrene
(6 min), perylene (7 min), benzo[b] fluoranthene (8 min), and benzo-
[a] pyrene (9 min) and isolated a pure specimen of benzo[a] pyrene
from a concentrate of tobacco smoke. These examples indicate the
great potential of high-speed liquid chromatography for the analysis
of carcinogenic polycyclic aromatic hydrocarbons; it most likely can
also be applied to the analysis of carcinogenic aza-arenes. High-speed
liquid chromatography is likely to become a major tool in the analysis
of organic air pollutants.
-------
270 Appendix B: Separation Methods
GAS CHROMATOGRAPHY
The gas chromatograph is a versatile instrument for analyzing mixtures
of organic air pollutants. Available at moderate cost ($2,000-8,000),
gas chromatography (G C) combines simplicity of operation with high
sensitivity. Useful application of GC to the analysis of POM dates from
1965, with noteworthy contributions from Wilmshurst,821 Cantuti
etal.,115 DeMaio and Corn,192 Carugno and Rossi,120 Chakraborty and
Long,127 andSearlefa/.685
Operating Principle
Gas chromatography vaporizes mixtures of organic compounds into
an inert, mobile vapor phase (carrier gas) and then transports it through
a long, narrow tube (column) containing an inert liquid (or soid) station-
ary phase in intimate contact with the mobile vapor phase. Organic
molecules travel along the column according to their relative
volatility and their physicochemical affinity for the liquid stationary
phase at the temperature of operation. The temperature and liquid
phase are selected for their ability to separate the sample mixtures in-
to their pure constituent compounds; after a characteristic interval
(retention time), the individual separated components emerge from the
column and are monitored by a sensing device (detector) that permits
the recording of their passage as a time-based analog signal rising
above the detector background signal (base line).
Application to Polycyclic Organic Matter
Successful gas chromatographic analysis of POM is a highly desired goal
because of its potential convenience, speed, and reproducibility.192>821
By operating on a physical principle different from that of column or
thin-layer chromatography, GC may separate mixtures of compounds
that cannot be resolved by the other procedures. It should be recognized,
however, that the complexity of the polycyclic organic fraction, con-
sisting of many closely related compounds, nearly precludes reliance on
a single separation technique. Thus, the use of GC in POM analysis must
be considered as complementing the other techniques for qualitative
and quantitative analysis.
Before the application of some refinements in 1965, analysis of POM
by GC was not very satisfactory.821 Indeed, the stringent requirements
of resolution, stationary-phase thermal stability, and detector sensitiv-
-------
Appendix B: Separation Methods 271
ity demanded by the presence of trace levels of closely related, rather
nonvolatile, poly cyclic organic compounds in polluted air continue to
challenge the analytic capabilities of GC.685
Detectors
Generally, the quantities of poly cyclic organic compounds that can be
sampled during a relatively short period are small. Therefore, only the
most sensitive GC detectors can be considered for analysis of POM.
For general hydrocarbon analysis, this has meant restriction to the use
of the flame ionization detector (FID). The FID functions by burning
the emerged (eluted), separated organic component in a hydrogen-
oxygen flame. Intermediate combustion states of the organic molecule
are ionized and appear as a current between two polarized electrodes.
The current is then amplified into an electric signal suitable for display
on an analog strip-chart recorder. The advantage of the FID is its wide
dynamic range of linear sample response. The most recent versions of
FID have sample sensing capability down to 1CT1 ° g, almost as sensi-
tive for POM as electron capture detection. Except for carbon atoms
oxidized to the carbonyl state or beyond, the signal of the FID detec-
tor is directly proportional to the number of carbon atoms being
burned, providing a quantitatively comparable detector response for
most polycyclic organic compounds.
The second kind of detector of interest uses electron capture (EC),
in which an eluted polycyclic organic component is irradiated by a
beta emitter. Because of the high temperatures involved, EC detectors
other than the conventional tritium version are required. When organic
compounds capable of capturing some of the electrons emitted pass
through the EC detector, their presence is detected by a diminution in
intensity of the normal stream of electrons collected from the beta
emitter. Whereas polycyclic organic compounds do indeed capture elec-
trons, the EC cross section may be expected to vary with molecular
structure; thus, the EC signal must be calibrated for each different
molecule for quantitative analysis. Those who wish to avoid this pro-
cedure frequently use the EC detector in parallel (split stream) with
the somewhat less sensitive but more general FID. With FID and EC
detection, the variable response of EC according to molecular struc-
ture has been used to advantage by Gantuti et al.115 to obtain tenta-
tive molecular identifications. A note of caution must be included for
the use of EC on particulate samples collected from the atmosphere:
EC detectors respond strongly 4:o the polyhalogenated hydrocarbons
-------
272 Appendix B: Separation Methods
used as pesticides. Hence, studies of comparative thin-layer chroma-
tographic and gas chromatographic behavior between pesticides and
POM may be necessary to avoid undesired interference from ambient
pesticides in particulate matter.
When the organic compounds separated by GC must be collected
for spectral identification, effluent stream splitters are generally satis-
factory.685 Occasionally, the less sensitive thermal conductivity detec-
tor is used127 to avoid the possibility of complications caused by the
stream-splitting inherent in the use of FID or EC. If this proves neces-
sary, special minimal dead-volume (micro) thermal conductivity detec-
tor cells (e.g., those manufactured by Perkin-Elmer Corp., Norwalk,
Conn.; or Carle Instruments, Inc., Fullerton, Calif.) are probably
necessary.
Submicrogram samples of eluted G C fractions may be detected and
collected with such detectors, whereas the conventional four-wire, semi-
diffusion type of thermal conductivity cells cut off above the micro-
gram level.
From time to time, it is to be expected that novel GC detectors will
be introduced; however, unless they can match or exceed the perfor-
mance of the detectors described here, their utility for GC analysis of
POM will be inconsequential.
Columns
A discussion of the types of GC columns useful in the analysis of POM
must include the nature of the liquid (or solid) partitioning phase and
the supporting surface. The relative nonvolatility of POM is a major
factor in its physicochemical association with airborne particulate sub-
stances. Therefore, it is not surprising that high temperatures are used
in GC analysis of POM. The GC partitioning phases used for such pur-
poses should have the highest thermal stability. Unfortunately, until
recently, the available liquid phases that were thermally the most stable
were barely adequate for the temperatures (>300 C) required. The sili-
cones (e.g., SE-30, SE-52, and QF-1) commonly used for lipid analysis
were often disposed in a dual-column configuration to compensate for
the excessive liquid-phase bleed that occurs above 250 C. Obviously,
columns operated in such a manner are going to be altered, if not de-
pleted, after several days of use. Two conventional methods for
reducing the overall bleed-lowering the column temperature and short-
ening the column—both work against effective POM analysis and rep-
resent, in part, the objections to work published before 1965 discussed
-------
Appendix B: Separation Methods 273
by Wilmshurst.821 What is required is a liquid partitioning phase that
possesses the requisite thermal stability to permit higher operating
temperatures and longer, more efficient (higher-resolving) columns.
Special precautions must also be taken in the selection of the
appropriate materials for the liquid-phase support and for the column
itself. Chromosorb W (Johns-Manville),676'821 or its equivalent,127'192
and Chromosorb G685 appear to be satisfactory. There are preferences
for minor variations of this flux-calcined diatomite, with one group
favoring the acid-washed version,82 another group a silanized version.676
Inasmuch as no definitive study has been reported on the relative merits
of the many diatomite versions available for analysis of POM at ex-
tremely low concentrations, it seems wise simply to use that which
has received the most extensive deactivation—e.g., acid washing and
silanization.685 Wilmshurst821 found the less reactive smooth spherical
glass beads relatively poor in resolution.
If open tubular (capillary) columns are used, the column itself
serves as the supporting agent for the liquid partitioning phase. The
openness and linearity of carrier gas flow make capillary GC suitable
for POM analysis, because low sample-carrying capacity of capillaries
is not a problem. Wilmshurst821 found Monel metal and copper unsatis-
factory as column material, whereas type 304 stainless steel appeared
more acceptable. Carugno and Rossi120 used glass capillaries rather
successfully. Because capillary columns used for POM analysis may
have a short lifetime, and the preparation of a workable column is
time-consuming, it would be unrealistic to claim that the preparation
and use of these highly resolving columns is ideal for POM analysis.
The subject must be studied further.
Future Prospects
There are several possibilities for improving the status of POM sepa-
ration and analysis by GC. Two liquid partitioning phases were intro-
duced recently: poly-m-phenoxylene (20 rings) and a polycarboarane-
siloxane (Dexsil-300 GC), which are thermally stable up to 250 C
and 350 C, respectively. Undoubtedly, analysis of POM will benefit
from substitution of these phases in GC procedures. The use of
inorganic phases (see Isbell and Sawyer414) for POM analysis is also an
interesting possibility. Silanization of glass (or even stainless-steel)
capillary walls can be used to minimize surface adsorptive or catalytic
effects. Finally, if diatomite supports are disadvantageous for GC
applications undertaken with packed columns, the recently introduced
textured (silanized) glass beads may suffice.
-------
274 Appendix B: Separation Methods
Comments
1. Despite earlier limitations, gas chromatography shows increasing
promise in qualitative and quantitative analysis of PO M.
2. Mixtures of polynuclear organics may be separated and detected
at nanogram levels if proper attention is given to deactivating the
surfaces of columns and diatomite support material thoroughly.
3. Recently introduced partitioning phases now enable the relatively
nonvolatile polycyclic organic compounds to be chromatographed at
300 C without damage to the partitioning phase.
4. Gas chromatography is best used in conjunction with other
confirmatory spectral or chromatographic techniques.
COMBINATION OF SEPARATION METHODS
Compared with such environmental agents as tobacco smoke, petro-
leum, and gasoline, POM can be isolated from organic air pollutants
and identified without great difficulty. In general, only two (at the
most, three) steps are needed for the isolation of these agents from
pollutants. In the past, combinations of column chromatography and
paper or thin-layer chromatography were more widely used, with or
without preliminary distributions of the organic matter between sol-
vents. During the last 5 years, various combinations have been developed,
including column or thin-layer chromatography with gas chroma-
tography and column chromatography followed by combined gas
chromatography and mass spectrometry.388>S73>661'664>668>676
The choice of method for the analysis of POM depends on such
factors as the degree of accuracy required by the scientist or the local
air pollution control authority, the availability of professional
and technical staff, the time allotted for analysis, and the availability
of funds for equipment and research.
SUMMARY
For the separation of polycyclic aromatic hydrocarbons and aza-arenes,
several precautions are suggested. These include the avoidance of
photooxidation and losses during the evaporation of solvents. Quan-
titative data should be secured with the aid of internal standards under
reproducible conditions.
-------
Appendix B: Separation Methods 275
Countercurrent distributions are cumbersome and of only limited
value. The distribution of organic matter between solvents is most
helpful for the separation of poly cyclic aromatic hydrocarbons from
paraffins. Higher concentrations of the latter negatively affect the
chromatographic separation of POM.
Chromatography generally uses columns filled with alumina, silica
gel, Florisil, or, occasionally, cellulose or modified cellulose. Separa-
tions should be completed under standard conditions, such as column
ratios, ratio of material to adsorbent, solvents, and flow rates. Column
chromatography results in high enrichment of polycyclic aromatic
hydrocarbons, but rarely leads to complete separation. In recent
years, gel filtration has been successfully explored.
Thin-layer chromatography is widely used; it works quickly, is
inexpensive, and can be used in every laboratory. Its limitations lie
primarily in the incomplete separation of mixtures of polycyclic
compounds. Paper chromatography requires, in general, 12-18 hr,
but gives better separations than thin-layer chromatography.
High-speed liquid chromatography offers several advantages, including
speed, high resolution, very sensitive ultraviolet detectors (1(T8 g for
benzo[a]pyrene), and the use of lower temperature than gas chroma-
tography. The technique needs further development. Nevertheless, it
is predicted that in a few years it will be widely used, especially when
a high-pressure pulse-free liquid pump becomes less expensive.
The most significant progress in the separation of POM has been
achieved by gas chromatography. With various column lengths and
diameters, liquid and solid phases and supports, and flame ionization
and electron capture detectors, gas chromatography is a most versatile
and sensitive technique. Polycyclic aromatic hydrocarbons may be
chromatographed at nanogram levels if the column and support sur-
faces are properly deactivated. New partitioning phases now permit
analysis at 300 C without interference from liquid-phase bleed.
PO M is actually analyzed with a combination of separation and
detection methods. The most widely used combination consists of
column chromatography followed by thin-layer or paper chromatography.
The most promising probably consists of column chromatography
followed by gas chromatography or, even better, by gas chromatography
plus high-speed liquid chromatography.
-------
276 Appendix B: Separation Methods
RECOMMENDATIONS
1. A national reference bank should be established to obtain and
purify all the polycyclic aromatic hydrocarbons and aza-arenes that
have been identified with certainty in the respiratory environment.
Melting points, ultraviolet and infrared absorption spectra, emission
and excitation spectra, relative gas-liquid chromatographic retention
times, and mass spectra should be recorded for all the purified agents.
If an agent has doubtful carcinogenic or tumor-initiating activity, the
same laboratory should be able to supply enough of the purified hydro-
carbon for biologic tests. The data so compiled should be made avail-
able to the scientific community. In some instances, small reference
samples should be made available to air pollution control laboratories.
2. A screening committee should be appointed to establish the
criteria for an environmental agent to be regarded as a carcinogen
or tumor-initiator. A list of the environmental carcinogens and tumor-
initiators should be compiled and made available.
3. A method should be developed for determining indirectly the
major sources of carcinogens in polluted urban air and their con-
tributions to the overall concentration.
4. A method should be developed for examining the concentration
and composition profile of mixtures of polycyclic aromatic hydro-
carbons in polluted air.
-------
Appendix C
Detection, Identification,
and Quantitation
ULTRAVIOLET ABSORPTION SPECTROSCOPY
Ultraviolet radiation is substantially absorbed by all polycyclic com-
pounds. The degree of absorption of light energy (or wavelength) is
a characteristic of a compound that may be used for identification
and determination of its concentration in a medium. The measurement
of such absorption has been widely used in air pollution analysis. The
advantages of this technique include the commercial availability of
high-quality spectrophotometers, the relative insensitivity of degree of
absorption to trace impurities, the fact that the absorption spectrum
of a mixture is usually the sum of the spectra of the components, the
strength of ultraviolet absorption bands (which obviates the use of
colorimetric reagents), and the high degree of sensitivity (which makes
it possible to determine and quantitate microgram amounts of com-
pounds). The disadvantages include the requirement of two or more
separatory steps for reliable identification and particularly quantita-
tion of air pollution samples and the low detection sensitivity (only
one-tenth to one-thousandth that of fluorimetric analysis).
Two principal laws of light absorption are relevant here. The first,
Bouguer's or Lambert's law, states that the proportion of light ab-
277
-------
278 Appendix C: Detection, Identification, and Quantitation
sorbed by a gas or liquid medium is independent of the intensity of
the incident light and that each successive layer of the medium absorbs
an equal fraction of the light that reaches it. The second, Beer's law,
recognizes that light can be absorbed by a molecule in vapor or liquid
phase only if the light collides with the molecule. The probability of
such a collision is proportional to the number of absorbing molecules
in the path of the light, and Beer's law states that the amount of light
absorbed in each successive layer is proportional to the concentration of
absorbing molecules and to the thickness of the layer. Combining these
two laws gives
I = I0 lQ-ebc,
where / - intensity of light transmitted through the solution,
international candles/cm3
70 = original intensity of incident light, international candles/cm3,
e = molar extinction coefficient,
b = length of light path, cm, and
c = concentration of absorbing molecules, moles/liter.
The ratio / : 70 is the transmittance of the medium ; and the absorbance,
the quantity usually measured by spectrophotometers, is
Note that the absorbance 04) is equal to the product of the concentra-
tion of the absorbing molecules, the path length of light in the medium,
and the molar extinction coefficient, and thus is proportional to each.
If more than one kind of molecule is present in the medium, the
absorption is nearly always additive, i.e.,
A - b(eicl
Molar extinction coefficients of polycyclic organic molecules are tabu-
lated in a number of sources.51'139'175'280'378'388'473'671'761'762
Spectral Characteristics of Polycyclic Organic Compounds
It is striking that all carcinogenic polycyclic aromatic hydrocarbons
and aza-arenes identified in urban air388 have highly structured ultra-
violet absorption spectra.139'761 Such spectra imply that the compounds
-------
Appendix C: Detection, Identification, and Quantitation 279
are planar in the ground state.51 If that planarity is required for their
carcinogenic activity (a possibility that requires more study), then
airborne carcinogens yet to be discovered will also have structured
absorption spectra. Highly structured spectra will be helpful in de-
tecting and identifying these compounds.
Clar's empirical classification of the bands in absorption spectra is
simple and adequate for analytic spectroscopy.139 The absorption
spectrum of benzo[a] pyrene in Figure C-l shows the three groups of
bands as labeled by Clar: The a. bands are weak (e £ 102 - 103) and
usually at the low-energy (long-wavelength) side of the spectrum; they
are sometimes partially covered by the neighboring p bands, which
are more intense (e ^ 104) and often have regular vibrational struc-
ture; the p bands occur at still higher transition energies (shorter
wavelengths), are more intense (e % 10s), and have less vibrational
structure than the other bands.
Although the total spectrum for each compound is unique, the indi-
vidual features are not. For example, both chrysene and 4,5-methylene
chrysene have a. bands at 361 nm,280 and both benzo[a] pyrene and
benzo[ghi]perylene have p bands at 382 nm.671 Therefore, identifica-
tion and quantitation of a polycyclic compound on the basis of only
one peak are unreliable.
Nonpolar inert solvents, such as pentane and cyclohexane, bring
out the spectral structures of polycyclic compounds better than polar
solvents, such as 95% ethanol, and are therefore the solvents of choice
for spectroscopic identification and quantitation. Aza-arene spectra
show a loss of structure and a shift to lower transition energies in
acidic solvents, so analysis in both neutral and acidic solvents is
useful.
Solvents appear to influence the positions of absorption bands
slightly. Sawicki et a/.671 give Xmax (maximal wavelength) for the
most intense
-------
280
Appendix C: Detection, Identification and Quantitation
1.0
0.9
0.8
5 0.7
z
UJ
_i
fE 0.6
0 0.5
UJ
O
0.4
CO
EC
O
/•> i
0.3
0.2
O.I
0.0
250 300 350
WAVELENGTH, nm
400
FIGURE C-l The absorption spectrum of benzo[a] pyrene at a concentration of
1(TS M in cyclohexane (7.5 jug of benzo[a] pyrene). In Clar's notation, the peak
at 403 nm is an a band, the peaks at 385, 364, 347, and 330 nm are p bands, and
the peaks at shorter wavelengths are (3 bands.139
-------
Appendix C: Detection, Identification, and Quantitation 281
length. The extinction coefficient and wavelength values of major
peaks are given in the figure legends and constitute a useful supple-
ment to the spectra, which have small and varied scales. Clar does not
include spectra of substituted aromatics or aza-arenes.
2. The Ultraviolet Atlas of Organic Compounds'161 is the most
elegant compilation of spectra. Spectra are plotted as log extinction
coefficient versus wave number (linear scale) and wavelength on trans-
parent paper with a uniform scale throughout, which greatly facilitates
comparison of spectra. Approximately half the known urban air car-
cinogens are included, and the format permits users to add spectra.
Compounds are indexed by structural group, by formula, and alpha-
betically. Included with each spectrum are the structure of the com-
pound, and values of the extinction coefficient, wavelength, and wave
number of each peak.
3. Ultraviolet Spectra of Aromatic Compounds by Friedel and
Orchin280 contains spectra of nearly all the known urban air car-
cinogenic polycyclic aromatic hydrocarbons, several aza-arenes, and
many substituted compounds. The spectra are plotted as log extinction
coefficient versus wavelength, and the scales are uniform and con-
venient. The introduction contains clear, concise discussions of nomen-
clature, different methods of plotting spectra, effects of substituents
and solvents on spectra, and multicomponent analysis. Unfortunately,
this excellent book is out of print.
4. Organic Electronic Spectral Data,™2 presently a six-volume set,
provides references to spectra published from 1946 to 1961 (additional
volumes covering more recent publications are in preparation). These
volumes list compounds by formula and name as designated by Chemf
ical Abstracts Index System, giving solvent Xmax (log e) and biblio-
graphic reference. This series, which already lists over 100,000 spectra,
will be increasingly valuable as a guide to spectra in the literature.
Three other compilations, although less useful, should be mentioned.
Absorption Spectra in the Ultraviolet and Visible Region, edited by
Lang,473 is a continuing series of spectra that are not arranged in any
systematic way and are of unknown reliability. The series does not
include even such well-known pollutants as benzo[a]pyrene, benzo[e]-
pyrene, and chrysene. Hirayama's Handbook of Ultraviolet and Visible
Absorption Spectra of Organic Compounds3'78 contains two tables, one
.giving absorption maxima from chemical structure, and the other,
absorbing chromophore from absorption maxima. The nomenclature
and structure classification in this book are unfamiliar, and the spectra
-------
282 Appendix C: Detection, Identification, and Quantitation
are pre-1960 and of uncertain quality. Because the data given are un-
reliable, one. would need to consult the original literature. Hirayama's
book is both less comprehensive and less current than Organic Electronic
Spectral Data. Finally, there are references in the literature175 to the
Sadtler Research Laboratories* collection of ultraviolet spectra, re-
puted to include 28,000 spectra, but this collection has proved
inaccessible.
Spectrophotometric Techniques
Many good dual-beam, ratio-recording spectrophotometers are avail-
able commercially. With only occasional calibration and mainte-
nance, these instruments can be used continually to yield accurate and
reliable data.
Wave number or wavelength calibration is generally straightforward;
holmium oxide filters and mercury vapor discharge lamps provide suit-
able spectra with little effort.621 A generally accepted method for
calibration of photometric accuracy is not yet available. Several chem-
ical standards have been recommended;621 potassium chromate in
potassium hydroxide has proved useful for testing photometric accuracy
and estimating stray light.531 The ease and persistence of spectrophoto-
metric intensity calibrations are clear advantages of spectrophotometry
relative to fluorimetry or spectrophotofluorimetry.
Because many polycyclic aromatic hydrocarbons and aza-arenes
have absorbance peaks as narrow as 4 nm, their spectra must be mea-
sured with instruments able to resolve 0.5 nm or better. Otherwise,
the spectra will be distorted in that true peak height will not be ob-
tained. Correct spectra, free from artifacts due to the spectrophotom-
eter, are essential for comparison with published spectra or tables.
The use of single-beam nonrecording instruments to obtain correct
spectra is tedious and can be recommended only if time and man-
power are abundant.
Many polycyclic compounds are fluorescent, so care must be
taken during absorption measurements to minimize the ratio of
fluorescence to transmitted light reaching the detector. Sample
absorbance should be kept below 1.0, and the sample should be placed
well away from the light detector. These precautions are often ignored
in reports of polycyclic compound absorption spectra.
Extracts of collected particulate matter prepared by any solvent
Spring Garden St., Philadelphia, Pennsylvania.
-------
Appendix C: Detection, Identification, and Quantitation 283
system contain a plethora of compounds that must be separated be-
fore quantitative spectrophotometry is possible. Column chroma-
tography yields fractions in solvents that may or may not be suitable
for spectral analysis. (Column chromatography is discussed in
Appendix B.) After chromatography, it is usually good practice to
evaporate the separating solvent at low temperature and redissolve the
residue in a solvent, like pentane or cyclohexane, that is inactive and
transparent in ultraviolet radiation up to 40,000 cm"1 (250 nm).3 The
solvent should be chosen to reveal spectral structure and to permit
comparison with standard spectra.
When separations are performed by thin-layer chromatography, ex-
traction from the layer is common practice; the choice of extracting
solvents is governed as above. A blank portion of the thin layer, at the
same distance relative to the solvent front (Rf) as the selected test
spot, should be extracted to provide the reference for dual-beam
spectrophotometry.
About 10 times greater sensitivity than usually reported for spectro-
photometry may be obtained by taking advantage of microcells and
limiting final extraction volume to about 50 jul. Adapters for such
microcells are commercially available.
In situ spectra from compounds on paper or thin-layer chromatOr
grams are sometimes used to detect and identify POM. Diffuse-
reflectance accessories to spectrophotometers may be used to obtain
reflectance spectra of spots on various adsorbents.666 Also, repeated
scanning of chromatograms with scanners set at incremental wave-
lengths permits plotting of reflectance spectra. Such spectra have
several major drawbacks: The relation between reflectance and absorb-
ance is unclear and is influenced by the choice of adsorbent; the spectra
are uncorrected; there is very little information on reflectance spectra
in the literature; and reproducibility is poor. For these reasons, reflec-
tance spectra are much less useful than absorption spectra of extracts.
Interpretation of Spectra
Spectra are often distorted by the presence of impurities or interfering
compounds; present practice in the interpretation of "unknown"
spectra appears to rely on the interpreter's experience more heavily
than is desirable. Incomplete separations are the major problem in
interpreting ultraviolet absorption spectra of air pollution extracts.
Interpretation of spectra of pure compounds is not a problem.
A satisfactory separation should yield an unknown fraction whose
-------
284 Appendix C: Detection, Identification, and Quantitation
absorption spectrum can be matched closely to the spectrum of a pure
reference compound. A close match implies that between the reference
and test compounds the absorbance ratio of any two peaks in one spec-
trum does not differ by more than 10% from the absorbance ratio of
the two corresponding peaks in the other spectrum. Such close agree-
ment is not usually found, even by procedures recommended very
recently.666 For example, thin-layer chromatography of an air sample
extract of benzo[a]pyrene shows a fraction whose ^385/^297 ratio
is 0.28; the corresponding ratio of pure benzo[a] pyrene is 0.43.
From the data of Sawicki et a/.,671 this ratio should be 0.50; the
Ultraviolet Atlas'161 data give the ratio as 0.47. These latter two values
are in satisfactory agreement with the pure benzo[a] pyrene ratio,
but differ significantly (^35% difference) from the value for the
benzo[a] pyrene fraction of the particulate air sample extract. Thus,
with the presently available separation methods, the accuracy of
spectrophotometric analysis of air sample extracts is questionable.
The calculated weights of the benzofa] pyrene fraction, based on
absorbance values at 382, 362, and 295 nm, are 1.4, 1.7, and 2.2 jug-
The reproducibility claimed for determinations at 382 nm is ± 7%.666
Partial separation into fractions containing several compounds
each may be adequate for the detection and identification of a partic-
ular compound if the absorption spectrum has one or more peaks
characteristic of that compound. A rough estimate of the amount of
the compound may be made by the base-line method.665 Such estimates,
however, should be labeled as only qualitative (± 50%), particularly if
only one absorption peak is used. There are two compelling reasons for
not trusting base-line estimates derived from only one peak: Dozens of
organic compounds have absorption peaks at any given wavelength,378
and drawing a base line under a peak presupposes that the background
is smooth, whereas most polycyclic compounds have highly struc-
tured spectra. For reliable detection and identification, at least two
absorption peaks must be used, preferably from different spectral
bands.
Once a compound has been identified in an extract and its molar
extinction coefficient gleaned from the literature (or from a spectrum
of a pure sample), the calculation of the compound concentration is
straightforward, provided that all the components of the extract have
been identified280 or, if not all components are known, that the frac-
tion of the absorbance at any given wavelength due to the specific com-
pound of interest is known.
Ideally, quantitation is based on single-component extracts. But
in practice, it may be necessary or desirable (to save time) to work
-------
Appendix C: Detection, Identification, and Quantitation 285
with multicomponent systems, in which case the experience and judg-
ment of the analyst carry a heavy burden. If a separation yields a close
match between spectra of standard and separated compounds, spec-
trophotometric quantitation should be accurate to within 10%. It
would be well if compilations of Air Quality Data™ indicated the
method of analysis with the expected accuracy of separation and
analysis. For this compilation, the method included separation by
thin-layer chromatography followed by fluorescence analysis. The
overall accuracy, limited by the characteristics of collection, was deter-
mined by controls to be ± 20% (E. Sawicki, personal communication).
Publication of benzol a] pyrene concentrations calculated to three or
four significant figures constitutes an unwarranted faith in the
analytic accuracy.
Knowing the concentration of a given compound in the final extract
should enable the analyst to determine the amount initially present in
the air sample. The entire extraction and separation must be quantitative
if the spectrophotometric data are to yield accurate concentrations.
LUMINESCENCE SPECTROPHOTOMETRY
Luminescence analysis has been reviewed in several publications re-
cently,38'60'82'363'57'7'835 and the reviews should be consulted for detailed
presentation of the matters raised here.
The books by Parker577 and Hercules363 give special attention to
techniques of luminescence analysis; Parker presents a great deal of
experimental detail and a list of earlier textbooks and monographs.
Becker38 discusses the theory of luminescence processes and, like
Parker, offers a convenient source of data. Sawicki published a review
of fluorescence analysis as it is related to the identification of air pol-
lutants.663 Pringsheim's classic text606 on luminescence analysis re-
mains a useful reference source of qualitative observations of visible
fluorescence. Luminescence analysis is a valuable technique when com-
bined with thin-layer chromatography.
The early literature is replete with inaccurate and false data re-
sulting from experimental error, so great care must be exercised in
evaluating published information on luminescence spectra, lifetimes, and
quantum yields. Furthermore, reports of analytic sensitivity and presenta-
tions of spectra are not readily compared. Complete knowledge of the
methods and criteria used in purifying standard compounds is re-
quired to evaluate and compare reported data. Valuable reference
sources of luminescence data have been published by Schmillen and
-------
286 Appendix C: Detection, Identification, and Quantitation
Legler678 and Zander.835 Convenient, uncritical guides to the literature
have been compiled by Passwater583 and Lipsett.498
"Fluorescence" is defined as the luminescence emitted in a radiative
transition from the lowest excited singlet state to the ground state.
Luminescence arising from a radiative transition from the lowest
excited triplet state to the ground state is termed "phosphorescence."
Measurements of luminescence involve the recording of two types of
spectra—emission spectra and excitation spectra. An emission spectrum
is obtained by irradiating, or exciting, a compound at the wavelength
of maximal absorption while scanning for emission at wavelengths
longer than that used for excitation. Excitation spectra are obtained
by analyzing the emission at the wavelength of maximal emission while
irradiating the compound with wavelengths shorter than that of maxi-
mal emission. Excitation spectra are very similar to absorption spectra;
in fact, excitation spectra can be converted to absorption spectra, and,
in the case of a pure compound, the two should be identical. Fluores-
cence and phosphorescence can be experimentally differentiated: Fluores-
cence is a relatively fast process and has a half-life of less than about 10"7
sec; phosphorescence is a relatively slow process and has a half-life of
over 1CT4 sec. Depending on solvent and temperature, the fluorescence
band more or less overlaps the electronic absorption band in poly-
cyclic aromatic hydrocarbons and aza-arenes, in many cases appearing
as the mirror image of the absorption band. Phosphorescence bands
appear at lower energy than the fluorescence bands. The energy dif-
ference between the para-band (* La in the Platt nomenclature and U
in the Moffitt nomenclature) absorption maximum and the onset of
phosphorescence (3La band) is about 10-12 X 103 cm"1 for the poly-
cyclic aromatic hydrocarbons and most aza-arenes of interest. These
differences permit selective observation of one or the other process by
proper choice of wavelength or by the use of shutter systems. It
is often possible to observe both the singlet-singlet and the very weak
singlet-triplet absorption processes by phosphorescence excitation
techniques. Experimentally, mirror-image symmetry is not always
observed between the singlet-triplet excitation band and the phosphores-
cence emission band.520
Instrumentation
Conventional instrumentation in emission spectroscopy consists of an
excitation train (comprising a high-intensity light source, a mono-
chromator or filter, and a quartz sample cuvette or sample tube that
may be placed in a quartz Dewar vessel for low-temperature work) and
-------
Appendix C: Detection, Identification, and Quantitation 287
an analyzing train (typically, a monochromator and a photomulti-
plier). In monitoring phosphorescence or delayed fluorescence selec-
tively, a shutter system is used for alternately exciting and viewing
the sample. The sample may be viewed at the face on which it is
being irradiated, at right angles to the direction of irradiation, or from
behind the cell in a line with the direction of irradiation. The type
of geometric arrangement used depends largely on the kind of specimen
to be examined—e.g., dilute solutions and gases, concentrated solutions,
opaque solids, and frozen solutions. Inner-filter effects, which are
instrumental artifacts due to excessive absorption of the exciting
light or absorption of the luminescence emitted, vary with the type
of geometric arrangement chosen and the specimen being examined.
A thorough discussion of the advantages and disadvantages of each
type of geometry in relation to each kind of specimen can be found
in Parker.577 (pp- 220-233) It should be noted that self-absorption
influences the apparent position of emission bands, as well as their
intensity,554 and the geometry chosen is an important factor in this
regard.
Many methods are available for alternately chopping the excitation
and luminescence beams. Mechanical chopping, using rotating slotted
disks or slotted cans, is most common. Electrooptical devices, such as
Kerr cells, have not been extensively used, owing to the unacceptable
attenuation of luminescence intensity when the cell is supposedly
"open," or transparent. Electronic gating of both lamp and photo-
multiplier has been accomplished with some success. Although gating
of the excitation source leads to difficulties, gating of the photo-
multiplier alone is acceptable in combination with a mechanical excita-
tion shutter, and this arrangement allows flexibility of time resolution
in viewing luminescence spectra. Winefordner has published a discussion
of time-resolved phosphorimetry and its major advantage, increased
selectivity of analysis.823 Time resolution is obtainable with a slotted-
disk spectrophotometer, but most commercial devices use a slotted
rotating can, and analytic applications of time-resolved instrumenta-
tion are rare.
Digital data acquisition has been applied to luminescence
analysis,217'499 although no attempt has been made to automate a
spectrophotometer. Aside from the obvious value of being able to
automate the lengthy process of correcting and evaluating spectra,
digital techniques permit the mathematical extraction of data from
complex spectra of mixtures. Methods and applications of the mathe-
matical techniques have been published.499'809
An increase in excitation intensity can lead to an increase in sensi-
-------
288 Appendix C: Detection, Identification, and Quantitation
tivity of luminescence analysis; monochromaticity of the excitation
source also determines, in part, the selectivity and sensitivity of
analysis. These factors, taken together with the sensitivity of fluores-
cence analysis to scattered light and the limitations imposed by the
bandwidth of the excitation source, would lead one to believe that
lasers would be especially useful in luminescence studies. Some work
has been done using lasers.S7S'698 A convenient tunable source, the
dye laser, has only recently become commercially available, and laser
sources may be used more widely in the future.
Most commercially available instruments are suitable for qualitative
luminescence analysis, although not all are capable of the high resolu-
tion required in some low-temperature techniques. These instruments
are not all obtainable with attachments for low-temperature phos-
phorescence and delayed-fluorescence studies; however, little thought
appears to have gone into this aspect of luminescence analysis, and
such attachments are generally inadequate. Another deficiency in
commercial instrumentation is the stray light characteristics of the
monochromators used in the less expensive devices. Because the
scattered and otherwise unwanted light output might depend heavily
on the monochromator wavelength settings, spurious luminescence
bands are often observed. These problems can be overcome through the
use of filters, as long as the resulting attenuation in luminescence
intensity is acceptable. An instrument by Farrand Optical can record
both corrected excitation and emission luminescence spectra at room
temperature, as well as at liquid nitrogen temperature. Instruments
manufactured by American Instrument and G. K. Turner are available
with attachments for automatic sensitivity correction of luminescence
spectra. These instruments are especially suitable for quantitative
luminescence analysis.
Sample Preparation
Luminescence processes are highly sensitive to impurity effects. Care
is required in separating and purifying samples to obtain the necessary
sample purity and to prevent the introduction of interfering sub-
stances. The solvent should be chosen with regard to ease of purifi-
cation and inertness, in addition to physical and spectroscopic properties.
Samples should be prepared in a clean laboratory; interfering sub-
stances like cigarette smoke are readily picked up in concentrations
sufficient to impair sensitivity or produce spurious spectra. Most
samples containing polycyclic aromatic hydrocarbons require efficient
-------
Appendix C: Detection, Identification, and Quantitation 289
degassing to remove oxygen before luminescence analysis, inasmuch as
oxygen severely quenches emission. Oxygen quenching affects room-
temperature analysis to a greater extent than low-temperature lumi-
nescence spectroscopy, because it is a diffusion-controlled process.
One approach toward minimizing impurity effects involves the use
of plastic hosts. Temperature quenching of fluorescence and phosphores-
cence appears to be determined mainly by the viscosity-dependent
migration of contaminants in samples,757 and high-viscosity plastics
promise to make ambient-temperature luminescence spectroscopy
more broadly useful. For instance, the intensity of phosphorescence
emission of dibenz[a,h] anthracene is found to be higher in a plastic
matrix at room temperature than in low-temperature glass.290 In
general, cross-linked matrices, like polycarbonate, polystyrene, and
polymethylmethacrylate, are most useful. There are some drawbacks
in the use of plastic samples: Emission spectra of compounds in
plastic matrices taken at room temperature are usually broad and dif-
fuse; and changes in relative intensities of bands and in spectral band
positions are noted. The possibility of increased luminescence intensi-
ties and relative insensitivity to temperature effects, however, make the
use of cross-linked plastic matrices attractive. An added advantage in
the use of plastic samples is that phosphorescence samples and sample
holders may be made so as to obtain the same reproducibility in
sample positioning as that obtained in ambient-temperature fluores-
cence or absorption measurements. This seemingly trivial problem has
often been a source of error.
Any separation procedure may be used before analysis of the com-
ponents of a mixture by emission spectroscopy. Of particular importance
is the analysis of mixtures after separation by thin-layer chromatography
and paper chromatography. Sawicki and Sawicki have published a review
of this method.661 Fluorescence measurements have been made in this
way on polycyclic aromatic hydrocarbons, aza-arenes, phenols, and
ketones. In general, a compound or suitable ionic species or reac-
tion product of the compound may be identified in situ by fluores-
cence color and position of the fluorescent spot. Quantitative analysis
has usually been carried out after further elution of the components
from the chromatogram.
Tentative methods based on column or thin-layer chromatography
combined with fluorescence analysis have been proposed for quantita-
tive analysis of polycyclic aromatic hydrocarbons—in particular benzo[a]-
pyrene and benzo[b] fluoranthene-in atmospheric aerosols.748"750 The
thin-layer chromatographic method has been used to elute the organic
-------
290 Appendix C: Detection, Identification, and Quantitation
fraction from particulate matter.749 Benzo[a]pyrene is identified on
the thin-layer chromatography plate by fluorescence methods and is
further eluted and analyzed as the cationic salt in sulfuric acid with a
fluorimeter. Lower limits of determination are reported as 3 ng of
benzofa] pyrene with a spectrophotofluorimeter and 10 ng with a
filter fluorimeter. Accuracy to within 5-10% is reported.
Thin-layer chromatographic fluorescence analysis has unique promise
in speed and convenience if complete qualitative and quantitative in situ
analyses can be performed. Several schemes have been published.661'663
In a recent analysis,756 anthracene, phenanthrene, pyrene, fluoranthene,
chrysene, benz[a] anthracene, benzo[a] pyrene, benzo[ghi]perylene,
dibenz[a,h] anthracene, and coronene were separated, identified, and
quantitatively analyzed by thin-layer chromatography combined with
in situ fluorescence excitation and emission measurements. Limits of
detection for these compounds were reported to be within 0.1-1 jug-
Several complications remain in the direct analysis of luminescence from
thin-layer chromatography plates. Light scattering and luminescence from
the plate itself are particularly troublesome in such techniques.
Luminescence from a crystalline or adsorbed compound often differs
drastically from that of the same compound in liquid solutions or in
low-temperature glass. Finally, luminescence spectra of separated com-
ponents on thin-layer chromatography plates are often found to dif-
fer significantly from luminescence spectra of standard compounds
obtained under the same conditions. At present, unambiguous identi-
fications from in situ luminescence spectra appear to require consider-
able experience.
Luminescence Analysis
Enough information is contained in low-temperature luminescence
spectra so that comparisons with spectra obtained from chemical
standards may aid in identifying unknown compounds or chromato-
graphic bands or spots. Information can also be obtained concerning
functional groupings attached to the molecule by noting whether the
compound is predominantly fluorescent or phosphorescent.
Vibrational analysis of luminescence spectra of polycyclic aromatic
hydrocarbons has been reviewed.567 Although there is no specific
method of achieving optimal results, several factors that affect the
resolution of the band structure can be identified. The observation of
characteristic vibrational progression lines is favored by low temperature,
high viscosity, low solute concentration, high component purity, and
the use of nonpolar solvents, such as «-alkanes.
-------
Appendix C: Detection, Identification, and Quantitation 291
Variation in pH certainly changes the nature of the solute, especially
if ionic species can be produced, and spectral changes are expected,
which may aid in analysis. A review of pH effects in fluorescence and
phosphorescence analysis has been published.682 Some polycyclic
aromatic hydrocarbons and aza-arenes are readily analyzed, at ap-
propriate pH, as the anionic or cationic salts. It is of special interest
that many compounds become more acidic or basic upon photoexcita-
tion. At appropriate pH, optical absorption or photoexcitation spectra
may be characteristic of neutral molecules, whereas the photolumines-
cence is characteristic of the cationic or anionic species. Spots on
thin-layer chromatography plates can be rapidly identified by exploiting
this property. For instance, 7H-benz[d,e] anthracen-7-one undergoes
fluorescence characteristic of its cation on a thin-layer chromatography
plate treated with trifluoroacetic acid. Selective agents, such as trifluoro-
acetic acid and tetraethylammonium hydroxide, facilitate similar
identification of acridines arid carbazoles.
In sufficiently dilute solutions and in the absence of interfering
substances (compounds that absorb light in the same region as the com-
pound of interest and compounds that quench or sensitize luminescence
of the compound of interest), emission intensity is a linear function of
concentration. Excitation intensity and detector response are com-
plicated functions of instrument characteristics; for analytic work,
it is common to use calibration curves determined in conditions exactly
the same as those to be encountered in the analysis. For polycyclic
aromatic hydrocarbons, linear calibration curves are generally obtained
for concentrations below 1 (Ts mole/liter when using the hydrocarbon
para-band C1 La, Platt; U, Moffitt) or (3 band C1 Bb, Platt; X, Moffitt)
for excitation and for concentrations below 1CT3 mole/liter when
using the a band (' Lb, Platt; V, Moffitt). Phosphorescence curves are
found to be linear over a greater range of concentration than are fluor-
escence curves.
The spectroscopic procedures themselves are highly sensitive and
accurate. The limiting factors in the analysis of polycyclic matter
appear to be the reliability and efficiency of the extraction and separa-
tion procedures used before luminescence analysis. The overall accuracy
of analytic procedures involving spectroscopic measurements is com-
plicated by the difficulty in distinguishing between many of the
polycyclic compounds. For instance, benzo[a] pyrene, benzofghi]-
perylene, and benzo[k] fluoranthene exhibit very similar electronic
spectra. If the sample preparation procedures cannot be relied on to
separate the components of a mixture thoroughly, then the use of no
one spectrophotometric technique can be considered completely re-
-------
292 Appendix C: Detection, Identification, and Quantitation
liable. Common practice involves the measurement of one band in
one type of spectrum (i.e., absorption, fluorescence, or phosphores-
cence), and this cannot be considered acceptable.
Luminescence Sensitization and Quenching
Absorption of energy by one compound, called the "donor," and
energy transfer from the excited donor to another compound, called
the "acceptor," may lead to luminescence of the acceptor. This type
of luminescence is known as "sensitized luminescence." It requires
an overlap of the emission spectrum of the donor with the absorption
spectrum of the acceptor and an excited-state energy of the acceptor
that is lower than that of the donor. The donor is known as the
"sensitizer."
Sensitization methods have unique potential for monitoring
small quantities of compounds. Parker demonstrated the determination
of impurities in polycyclic aromatic hydrocarbon standards by
sensitized delayed fluorescence.577 If several impurities are present,
it is possible to excite the one with the lowest triplet energy first,
and it should be possible to excite a limited number of impurities pro-
gressively and resolve the spectra with mathematical techniques. As
an example, anthracene and benz[a] anthracene may be determined
in pyrene by energy transfer to both impurities, using pyrene as the
sensitizer. The presence of anthracene can be independently determined,
using acridine orange as a sensitizer. High concentrations of components
are required and solutions must be deoxygenated, which makes this
method tedious. Zander835 has proposed doing this in crystallized aro-
matic hydrocarbons at low temperature—a technique that is much more
sensitive and does not require deoxygenation. No applications to air
pollution analysis have been reported.
Organic electroluminescence364'836 can be used both quantitatively
and qualitatively for the analysis of such polycyclic compounds as
anthracene, 9-phenylanthracene, 9,10-diphenylanthracene, pyrene, and
coronene. Electroluminescence techniques have many disadvantages,
compared with photoluminescence and other analytic methods. Lu-
minescence may arise by annihilation of radical ions with other
oxidizing or reducing agents. High concentrations (about 10~3 M) are
usually required, although spectra have been observed at very low con-
centration (about 1(T7 M). A potential for handling complex mixtures
without separation of components seems apparent. Electroluminescence
has been observed for polycyclic aromatic hydrocarbons, substituted
derivatives, and heterocyclics.
-------
Appendix C: Detection, Identification, and Quantitation 293
Organic electrochemistry presents unique opportunities for the
analysis of complex mixtures49'514'589 but does not appear to have
been exploited in air pollution analysis. Polarographic oxidation and
reduction potentials are available for many polycyclic hydrocar-
bons,49'514'595 and, although several compounds may have the same
reduction potential, organic polarography would be a uniquely rapid
method of quantitative analysis if coupled with other techniques for
unambiguous identification of components.
Quenching techniques have also been demonstrated to be of value
in pollution analysis.633'675 By using compounds like acidic or alkaline
nitromethane, acetophenone, and carbon disulfide, the luminescence of
polycyclic compounds can be selectively quenched, leading to the
analysis of azaheterocyclic compounds in the presence of polycyclic
hydrocarbons, aromatic amines, and imino heterocyclic compounds.
Evaluation of Sample Purity
Excitation spectrophotometry is an extremely sensitive form of
absorption spectroscopy. A change in the spectral distribution of
luminescence when the wavelength of excitation is changed or the
inability to match the absorption and excitation spectrum is an
indication of impurity. Because the problem of purity in both the
luminescence standards and samples can be a source of error in
luminescence studies, more widespread use of excitation spectro-
photometry for analysis of sample purity would be desirable.
Luminescence decay, simple to measure in phosphorimetry, is also
an excellent index of purity. A logarithmic plot of detector response
is linear if the substance is pure (in the absence of heavy-atom sol-
vents or solvent inhomogeneity). A good discussion of the technique
of making lifetime measurements has been published recently by Demas
and Crosby.194
Correction of Spectra
No comparisons may be made between spectra obtained and analyses
performed in different laboratories, unless data related to the calibra-
tion curve are known and reported. Many procedures have been advanced
for constructing calibration curves.363'577 The most convenient calibra-
tion procedures rely on the use of luminescence standards. There is
little agreement on the choice of an appropriate compound. Quinine
bisulfate in 1 N sulfuric acid is most widely used.
Luminescence spectra must be corrected for lamp intensity, which
-------
294 Appendix C: Detection, Identification, and Quantitation
decreases with wavelength; for optical elements, such as mirrors, mono-
chromator gratings, and lenses; and for response of photomultiplier
tubes in the detection system. If analysis is carried out with fixed
molecular orientation, which is common when the solute is dissolved
in high-viscosity solvent and adsorbed on a surface or joined to a
polymer by chemical reaction, corrections must be made for the
polarization of luminescence, because the horizontal and vertical com-
ponents of light are not transmitted equally through monochromators.
Numerical analysis indicates that large errors can result if this cor-
rection is not applied.576
Several methods of plotting corrected spectra are used.230'577 The
most common is to plot relative quanta per unit frequency interval
against either wavelength or wave number.
Complications and Limitations
Fluorescence analysis is particularly sensitive to light scatter and
Raman emission. Low-temperature luminescence techniques are
influenced by bubbling and "snow" in the coolant, cracking of the
sample glass, and repeatability of sample placement. All emission
techniques are influenced by luminescence of solvents and sample
cells, instrument noise, accuracy and repeatability of wavelength
settings, scatter of light in monochromators, and random fluctuation
in source intensity.
The occurrence of photochemical dimerizaton and oxidation
reactions places limitations on sensitivity and applicability of lumines-
cence analysis. Photochemical reactivity of the luminescent compound
limits the intensity of excitation that can be used. Photochemical prod-
ucts may also contribute positive or negative error in quantitative
analysis, because some products are themselves luminescent. Photo-
chemical reactivity, then, may be a special problem in luminescence
sensitization, in that a relatively large amount of energy may be trans-
ferred to the compound under study with this technique. Photochemi-
cal reactions involving other components of the sample may occur.
Parker and Hatchard have reported such a reaction of benzo[a] pyrene
and pyrene with a polymer host matrix, polyvinylpyridine butylbro-
mide.578
Proper choice of solvent, pH, viscosity, and addition of some metal
ions in particular cases increase both the sensitivity and the selectivity
of analysis. Luminescence intensity and band position are functions
of solvent and temperature. Fluorescence can be quenched through
-------
Appendix C: Detection, Identification, and Quantitation 295
complex formation and van der Waals interactions with polar solvents.224
The fluorescence bands of 9,10-diazaphenanthrene, phenanthrene,
naphthalene, and pyrene are found to blue-shift (and sharpen) below
120 K owing to solvent orientation effects.20'418'S7° Failure to com-
pensate for temperature fluctuations or to permit temperature equilibrium
is a serious source of error. A study of the effects of temperature fluctua-
tions in low-temperature glasses has been published by Leubner.490 Van
Duuren has published reviews of temperature and solvent effects on
luminescence spectra of aromatic compounds.781'782 A review of
environmental effects in fluorescence analysis was recently published.810
Excimer (excited dimer) formation is a special problem in solid
luminescence. In liquid solutions, it is sensitive to both temperature
and solvent at low hydrocarbon concentrations,395 and significant
errors can occur in sensitive analytic work.
Quenching by oxygen is a serious problem and may be very selective.
Berlman50 and Parmenter and Rau580 have published useful data on
quenching of luminescence by oxygen. Pyrene fluorescence is
especially susceptible to oxygen quenching. Although luminescence
is not influenced by some metal ions,311 the presence of heavy metal
anions, such as Br~ and I~, and especially of paramagnetic ions, such
as Co2"1" and Mn2+, does alter luminescence characteristics in specific
cases.39 Metal ions favor intersystem crossing from the excited singlet
state to the excited triplet state and therefore often enhance phos-
phorescence intensity at the expense of fluorescence intensity. Thus,
addition of some metal ions or compounds, such as ethyliodide and
ethylbromide, to solutions containing the compound under investiga-
tion often enhances the sensitivity of analysis.
Luminescence properties of crystalline compounds on dry thin-
layer chromatography plates depend as much on the crystal structure as
on the compound. Efficient energy-transfer processes in crystals ensure
that fluorescence bands will be diffuse, and excimer luminescence,
seen as a structureless and broad band, is often noted. There is reason
to believe that crystal phosphorescence should be too weak to be ob-
served, if it occurs at all. Crystal luminescence is very sensitive to
lattice distortions and dislocations. Band position arid intensity,
fine structure, and photochemical reactivity are drastically altered
when a crystal is "stressed."37'590 Intensity of fluorescence depends
on the crystal size and on the presence of oxygen.724 Some compli-
cations can be handled by annealing or by carefully choosing solvents
to control crystal size, but these factors present difficult problems for
in situ analyses of crystalline or solid-state systems. The luminescence
-------
296 Appendix C: Detection, Identification, and Quantitation
properties of crystals are beyond the scope of this discussion and
have been reviewed elsewhere.61'822 Birks and Cameron have pub-
lished a useful paper, which discusses crystal fluorescence of many
carcinogenic hydrocarbons.62
Comparison of Absorption and Emission Methods
In terms of utility in analytic applications, luminescence and absorp-
tion spectrophotometry are compared in three major respects: ease,
selectivity, and sensitivity. Although progress has brought a high degree
of sophistication to luminescence instrumentation and procedure, the
analyst is required to control directly many instrumental variables
that are often controlled automatically in absorption spectrophotom-
eters. In general, there are no standard methods of obtaining and
reporting luminescence spectra, although many proposals regarding
method, instrumentation, and presentation have been put forward.
Therefore, more care and effort are required of the analyst in inter-
preting, reporting, and especially comparing luminescence data. How-
ever, the analyst has control over more experimental conditions in
measuring luminescence. The wavelength of excitation and the wave-
length of viewing luminescence can be independently selected. These,
as well as conditions of solvent, temperature, and concentration of
added sensitizers or quenchers, can be chosen so as to favor the
observation of light emission of a particular type or from a particular
compound. This has been used to advantage in the analysis of com-
plex mixtures of polycyclic aromatic compounds without prior
separation of the components. For instance, benzo[a] pyrene may
be detected and quantitatively measured in the presence of as many
as 40 polycyclic compounds under appropriate conditions.671'672
Indeed, measurements of polycyclic compounds in the air have been
made in this manner.579
The sensitivity of luminescence analysis may be reported in a variety
of ways, leading to some confusion in attempting to make comparisons
without considering details of instrumentation and procedure. Theo-
retical and experimental comparisons have been made of the limits of
detection of various polycyclic aromatic hydrocarbons by several
methods.126'523'835
It is clear that emission techniques are more sensitive than absorption
spectrophotometry and that sensitivity depends in part on the instru-
mentation used.126'835 Concentrations in the nanogram-per-milliliter
range appear detectable by luminescence techniques. Absorption
techniques appear to be less sensitive by a factor of 102-103. This
-------
Appendix C: Detection, Identification, and Quantitation 297
may be compared with the subnanogram sensitivities claimed for gas
chromatographic and mass spectrometric techniques. Whereas the
data indicate that phosphorescence analysis and fluorescence analysis
are competitive procedures, in general, many factors combine to make
them complementary.
It is well to note that a combination of photoluminescence with
other techniques may lead to even more selective and sensitive analyses.
It appears that several techniques offer promise of quick routine
analyses of complex systems, but they have yet to be exploited.
Summary and Conclusions
Emission and absorption spectral analysis of polycyclic organic
compounds offers nanogram sensitivity and accuracy within 10%.
Clean separations are critical. Failure to achieve clean separation re-
sults in substantial analytic uncertainty. Some techniques of emission
analysis have potential for circumventing extensive separation and
should be pursued. There is not yet a good combination of separation
and spectral analysis that permits analytic accuracy to within 25%.
Compilations of air quality data should include methods of separation
and analysis and estimates of the accuracy of both.
INFRARED AND RAMAN SPECTROSCOPY
Infrared and Raman spectra reveal features characteristic of individual
bands or functional groups. Strong infrared bands are related to vibra-
tions that cause changes in dipole moment; strong Raman bands are
related to vibrations that cause changes in molecular polarizability.
Therefore, the two spectra are complementary and are often used to-
gether. Although single-ring compounds, which exhibit considerable
symmetry, have been extensively investigated, polycyclic compounds
have received only scant attention from vibrational spectroscopists.622
The uniqueness of individual bands is unsettled. It is not known
whether there are spectroscopically active vibrations that are character-
istic of such ring structures as chrysene or benzol a] pyrene as a whole.
Vibrational bands that are related to individual functional groups are
not unique, but the total spectrum of a molecule is unique. Thus, the
chief use of vibrational spectroscopy will be in identification of un-
knowns by revealing the presence of carbonyl or other functional
groups and by matching known and reference spectra.
Band strengths in vibrational spectroscopy are one hundredth to one
-------
298 Appendix C: Detection, Identification, and Quantitation
tenth as great as those in electronic (ultraviolet or visible) spectroscopy.
Infrared band strengths are often not proportional to concentration.
Infrared quantitative analysis requires large samples and careful calibra-
tion. Raman band strengths are proportional to concentration, but are
even lower in general than infrared band strengths and require hours of
exposure time to obtain spectra.
No special techniques have been developed for infrared or Raman
analysis of polycyclic compounds. In the few cases where vibrational
spectra are mentioned, common techniques appear to have been used.
Internal reflection spectroscopy is a technique whereby the sample,
layered on an interface between two transparent media, is tested by
analyzing the infrared light as it is totally internally reflected at the
interface. This technique yields spectra that are essentially the same
as conventional infrared spectra. The advantages lie in the small (micro-
gram) amount of material required and the ability to use liquid or solid
samples free of any suspending "matrix." This technique has been ap-
plied to pesticide365 and asbestos564 analysis and appears to be useful
for trace analysis of any nonvolatile compound.
The disadvantages of vibrational spectroscopy are the relatively weak
bands, the fact that infrared band strengths are not proportional to con-
centration, the requirement for a vibrationally transparent medium, and
the lack of unique polycyclic spectral features. The disadvantages far
outweigh the advantages of these techniques. Internal reflection spectros-
copy may become useful and deserves more attention. At present,
either ultraviolet absorption or fluorescence is preferable to vibrational
spectroscopy for analysis of POM.
MASS SPECTROMETRY
Organic mass spectrometry is practical over a range of sensitivities some-
what greater than that available with the more convenient and less ex-
pensive ultraviolet spectrophotometers and spectrofluorimeters. Reso-
lution with the less expensive equipment is to the nearest integral mass
number, so the molecular weights of the parent molecule and its charac-
teristic fragmentation products are expressed in nominal atomic mass
units (amu). Because sensitivity and resolution are inversely related,
this type of resolution affords the highest sensitivity, extending to
10"1 * g. As resolution is increased to the range where millimass (1CT3
amu) differences in the various elemental masses may be discerned, the
ultimate sample sensitivity may fall to 10~7 g. Briefly, the difference
-------
Appendix C: Detection, Identification, and Quantitation 299
between low mass resolution and high resolution is the difference be-
tween determining the nominal molecular weights of the parent mole-
cule and daughter fragments and determining the atomic composition
of these entities. In many cases, it is desirable that the latter be deter-
mined directly and unambiguously for the identification to be unequiv-
ocal.
Operating Principle
Organic mass spectrometry is based on high-vacuum ionization of or-
ganic molecules, generally by bombardment with 70-eV electrons. The
ions thus generated under vacuum may be deflected and focused vari-
ously by electric or magnetic fields such that inertial (mass) differences
allow the ions to be dispersed and displayed according to their own
masses. Under electron bombardment at 70 eV, most organic molecules
receive more than enough energy to dislodge an orbital electron, there-
by creating positively charged molecular ions. The excess kinetic energy
thus imparted to the molecular ion may then be released by various
mechanisms that involve rupture of the parent ion into smaller neutral
and ionized fragments. Those fragments and the parent molecular ion
constitute the principal features of the mass spectrum, which is simply
a statistical record of the abundance of ions at particular masses through-
out the mass range scanned. Because of the tendency of an ionized mole-
cule or fragment to decompose according to its intrinsic structural
features, the statistically consistent fragmentation pattern at 70 eV re-
veals structural detail that in many instances can be used to recon-
struct the original molecule.
If fragmentation predominates over the stability of the molecular
ion to the extent that the molecular ion cannot be discerned (e.g.,
in the case of substituted carbazoles),98 alternative approaches to ionif-
zation may be desired. Presently, they include low-ionizing-voltage
(7-15 eV) electron bombardment, field ionization, and chemical ioni-
zation procedures, all of which enhance the stability of the molecular
ion by minimizing its surplus kinetic energy.
Application to Analysis of Polycyclic Organic Matter
The use of mass spectrometry itself in the analysis of POM is analogous
to the use of other spectroscopic methods, such as ultraviolet fluores-
cence, in the examination of chromatographic fractions.685 In the
analysis of polycyclic organic chromatographic fractions, mass spectrom-
-------
300 Appendix C: Detection, Identification, and Quantitation
etry possesses the advantage over ultraviolet of being able to detect all
organic species present, not just those with the appropriate configuration
of conjugated unsaturation. Hence, mass spectrometry may be used
to determine the purity of separated polycyclic organic compounds. 146>685
However, except for some aza-arenes,98 the molecular stability
of polycyclic organic compounds on ionization is so great that the
excess kinetic energy imparted by the 70-eV electrons can be accom-
modated within the ionized molecule. This results in a minimum of
molecular fragmentation in the mass spectrometer and, except for mo-
lecular ions with multiple charges, a minimum of distinguishing qualitative
features among spectra of isomeric polycyclic organic compounds. Thus,
mass spectrometry is useful chiefly in determining the molecular weights
and purity of polycyclic organic chromatographic fractions. In this re-
gard, the low-ionizing-voltage technique has been used to simplify the
spectra taken on mixtures by producing mainly the unfragmented,
singly charged molecular jon.417'695'701 If high-resolution mass spec-
trometry is available, exact unambiguous atomic compositions of the
molecular ions present in mixtures may be determined.417'695
In the handling of samples for mass spectrometric analysis, poly-
cyclic organic compounds separated by various forms of chromatog-
raphy present little difficulty. Because they are generally so involatile,
they may be analyzed by the most sensitive of sampling techniques—
the direct-insertion probe. In the case of paper or thin-layer chromatog-
raphy, if absolutely necessary, the polycyclic organic sample may be
analyzed without prior elution from the chromatographic matrix—i.e.,
the matrix plus sample may be inserted directly into the evacuated mass-
spectrometry ionization chamber, where the sample may be slowly
vaporized by moderate heating. If polycyclic organic compounds are
collected from the gas chromatograph in melting-point capillaries, the
tubes themselves may be introduced with the probe.525 Because the
sample is at least partially consumed, mass spectrometry should be
performed after other nondestructive spectral analyses have been satis-
fied.
Coupled Gas Chromatography and Mass Spectrometry
As indicated above, mass spectrometry and other spectroscopic pro-
cedures generally are applied to polycyclic organic compounds in
combination with chromatographic separations. Naturally, the more
manipulations required in the various stages, the more time-consuming
the overall procedure, with attendant chances for losses and contamina-
-------
Appendix C: Detection, Identification, and Quantitation 301
tion. Efficient direct coupling of chromatographic separation with
spectral identification would circumvent such mechanical problems
and accelerate the analysis. With the recently introduced improvements
in gas chromatographic analysis of polycyclic organic compounds and
with an inert, efficient sample-transferring interface, the use of coupled
gas chromatography and mass spectrometry (cc-MS)501'525'526'806
for the analysis of POM is now logical and timely.
Coupled GC-MS, combining the high separational capability of gas
chromatography with the molecular definitiveness of mass spectrom-
etry, affords unparalleled speed, sensitivity, and completeness of
identification of the constituents in trace amounts of complex mixtures.
In fact, the ability of coupled GC-MS to produce analytic data "on the
fly" far exceeds the ability of technicians to process the data. Generally,
moderate to heavy GC-MS sample processing requires the assistance of
automated data acquisition and reduction to utilize the capacity of
GC-MS systems fully. Sampling sensitivity depends on whether low or
high resolution is used. As to the choice between low- and high-resolu-
tion GC-MS, it is almost always preferable to run the low-resolution
GC-MS profile before high-resolution analysis, inasmuch as it is not
possible to obtain statistically reliable ion abundances to the same
degree of sensitivity with high-resolution mass spectrometry as with
low-resolution mass spectrometry. High-resolution mass spectrometry
represents a greater investment in apparatus; nevertheless, whenever
the unambiguous elemental composition, of any compound and its
fragmentation products is required417'695 and there is sufficient
sample (more than 10~7 g), it is the method of choice.
Summary
Mass spectrometry affords a sensitive means of determining the
probable identity and relative purity of polycyclic organic chromato-
graphic fractions. Generally, this has been accomplished by the low-
voltage ionization method, which produces spectra containing mainly
singly charged, intact molecular ions. The masses determined for these
ions indicate the nominal molecular weights of the species present when
low-resolution spectra are taken. If high-resolution spectrometry is used,
the empirical formulas of the species are also determined. Recent suc-
cess in coupling gas chromatography with mass spectrometry suggests
benefits in applying this sensitive and selective dual technique to the
analysis of polycyclic organic material.
-------
302 Appendix C: Detection, Identification, and Quantitation
NEW AND PROMISING SPECTROSCOPIC TECHNIQUES
Fourier transform spectroscopy combines an old optical instrument—
the older mathematical apparatus of Fourier analysis—with modern
computer technology. The interferometer, in contrast with the spectrom-
eter, is a nondispersive device in which light is first split into two
beams and then recombined after the beams have traveled paths of
somewhat different lengths. By measuring the recombined-beam energy
as a function of the split-beam path-length difference, one obtains an
interferogram, which may be transformed into the more familiar
spectrum plot of energy versus frequency (or wavelength) or may be
used as is. Fourier transform spectroscopy may be applied throughout
the electromagnetic spectrum, but the infrared region has received much
of the attention. The most striking advantage of the technique using
new commercial instruments is its high speed. It can be coupled to gas
chromatograph output to obtain spectra of fractions as they pass by in
about 11 min, in contrast with the hours required with dispersing
instruments.
Correlation spectroscopy is a dispersive technique whereby several
spectral features are scanned for at once; spectrometers are modified
by the addition of multiple slits, screens, and rotating refractive plates.
The attractive feature of this technique is the possibility of selective
detection of individual compounds in mixtures,184 which could reduce
the time and effort now required for separation.
Electron spectroscopy measures energy of electrons ejected from
the sample under bombardment with x rays362 or with photons of
energy greater than about 10 eV.ss The energy of the ejected electrons
is governed by the identity of the atoms and their environment (e.g.,
oxidation state) in the sample molecule. For this reason, this technique
is best for structural determinations. The sample must be stable in a
vacuum of 10~s torr; microgram quantities are adequate for analysis.
Hercules362 states that the technique may be used to obtain element
ratios accurate to within about 5%.
CONCLUSIONS
It is not possible to point to either absorption or emission spectroscopy
as clearly superior for use as a standard procedure for spectral analysis
of air samples. There are too many interrelating uncertainties of cost,
-------
Appendix C: Detection, Identification, and Quantitation 303
complexities of instrument operation and interpretation, interference
of trace compounds, and uncertainties of separation methods. New
spectroscopic methods—such as chemiluminescence, electron spectros-
copy, Fourier transform spectroscopy, and correlation spectroscopy-
promise procedures that will circumvent separation difficulties.
-------
Appendix D
Regression Analysis
The method of multiple regression, widely used to estimate the in-
fluence of one or more variables on some quantity of interest, is
subject to serious limitations of interpretation when the observed
variables cannot be controlled by the observer and can be associated
in unknown ways with other hidden but influential variables. The
method does, nonetheless, enable one to apply a standard numerical
expression to trends in the data that would otherwise have to be
described more informally and with greater difficulty.
The procedure is most easily understood in the case of a single
independent variable, such as amount of smoking (cigarettes per
annum per person over age 15), whose relation to the quantity of
interest, such as age-standardized lung cancer death rate, is sought.
Figure D-l shows the graph of the 48 lung cancer death rates versus
the 48 tobacco sale values for the 48 contiguous states discussed in Table
17-20. Although there is considerable scatter, there is also an appre-
ciable tendency for states with high tobacco sales to have high lung
cancer death rates. A variety of methods might be used to express the
average relation. If, however, it is decided to express the relation as a
linear function, one plausible choice of best-fit line is the one that
minimizes the overall discrepancy in the sense of least squares, i.e., if
304
-------
Appendix D: Regression Analysis
305
a
OC
<3
01
c
3
Point representing
ith score
fitted line
Cigarette Consumption
FIGURE D-l Average relation between lung cancer death rate and cigarette
consumption for 48 contiguous states.
£>j = [lung cancer death rate (LCDR) for state i] -(value predicted
from fitted line). There is a unique line that minimizes the quantity
S = S Dj2 This line is called the fitted regression of Y (lung cancer
death rate) on X (cigarette consumption), and its slope is called the
regression coefficient (of Y on X).
One reasonable approach to the question of whether the slope found
in this way is "significant" is to ask how likely it is—given the values of
YI ,. .. , Y48, and those of Xl... , Z48, which are at hand-that a
completely random pairing of Y's with X's would give as steep a slope
as that provided by the actual pairing. To a good approximation, that
question can be answered by comparing the regression coefficient
with its standard error. If the slope is greater than twice the calculated
standard error, it is likely that something beyond random pairing of
7's with X's is going on (in the sense that fewer than 5 times in 100
would such a random pairing give as steep a slope).
Two generalizations of the above procedure are, in fact, used in the
text. First, it is hoped to fit a linear relation between Y and two or more
^'s, e.g., Xi = cigarette consumption, and X2 = average benzo[a] pyrene
concentration. Now there is a three-dimensional plot, with coordinate
axes representing Xlt X^, and Y. The trend of the points is now fitted
-------
306 Appendix D: Regression Analysis
by passing a plane (rather than a line) through the scatter, choosing the
unique plane for which S = S D\2 is minimal, where Dj = (LCDR) —
(value predicted from fitted plane).
Second, because for some states some of the population groups (e.g.,
female, nonwhite) are very small, the relation between Y and the X's
might be expected to be less pronounced than in states where the popu-
lation is larger. Appropriate weights, wi; can be assigned in such a case,
and one then seeks to minimize Sw = S H>J D{*.
The general principles of the analysis remain unchanged.
It is appropriate to emphasize that the regression method constitutes
merely one of a number of possible ways to fit a linear function to the
data. It has an advantage over some other methods, in that it is easy to
calculate a standard error to use in judging whether the regression
coefficient is larger than a random pairing of Vs with X's might plausibly
provide. The method in no way precludes effects of hidden variables,
nor does it, in itself, imply any causal connection, even when the exis-
tence of a relation is beyond doubt.
-------
References
1. Aalbersberg, W. Ij., G. J. Hoijtink, E. L. Mackor, and W. P. Weijland. Com-
plexes of aromatic hydrocarbons with strong Lewis acids. J. Chem. Soc.
1959:3055-3060.
2. Aalbersberg, W. Ij., G. J. Hoijtink, E. L. Mackor, and W. P. Weijland. The for-
mation of hydrocarbon positive ions in strong proton donors. J. Chem. Soc.
1959:3049-3054.
3. Abelev, G. I. Antigenic structure of chemically-induced hepatomas. Progr. Exp.
Tumor Res. 7:104-157, 1965.
4. Abell, C. W., and C. Heidelberger. Interaction of carcinogenic hydrocarbons
with tissues. VIII. Binding of tritium-labeled hydrocarbons to the soluble
proteins of mouse skin. Cancer Res. 22:931-946, 1962.
5. Adams, R. N. Anodic oxidation pathways of aromatic hydrocarbons and
amines. Accounts Chem. Res. 2:175-180, 1969.
6. Adams, R. N. Electrochemistry at Solid Electrodes. (Monographs in Electro-
analytical Chemistry and Electrochemistry.) New York: Marcel Dekker, Inc.,
1969.402pp.
7. Adamson, R. H., R. W. Cooper, and R. W. O'Gara. Carcinogen induced tumors
in primitive primates. J. Nat. Cancer Inst. 45:555-559, 1970.
8. Agricola, G. De Re Metallica. Translated by H. C. Hoover and L. H. Hoover.
New York: Dover Publications, Inc., 1950. 672 pp.
9. Alekseeva, T. A. The use of luminescence spectral methods for study of the
fractional composition of aromatic dispersed bitumens. Tr. Vses. Nauch.
Issled. Geologorazved. Neft. Inst. 33:271-277, 1962. (in Russian)
307
-------
308 References
10. Allen, M. J., E. Boyland, and G. Watson. Experimental bladder cancer, pp.
34-35. In British Empire Cancer Campaign. Thirty-fourth Annual Report,
1956. London: British Empire Cancer Campaign, 1956.
11. Allison, A. C., M. E. Peover, and T. A. Gough. Polarographic measurements of
electron donation and acceptance by carcinogenic compounds. Nature 197:
764-765, 1963.
12. Altshuller, A. P., and J. J. Bufalini. Photochemical aspects of air pollution: A
review. Environ. Sci. Technol. 5:39-64, 1971.
13. Alwens, W., E. E. Bauke, and W. Jonas. Auffallende Haufung von Bronchialkrebs
bei Arbeitern der chemischen Industrie. Arch. f. Gewerbepath. u. Gewerbehyg.
7:69-84, 1936.
14. Andersen, A. A. A sampler for respiratory health hazard assessment. Amer. Ind.
Hyg. Assoc. J. 27:160-165, 1966.
15. Annual Report of H.M. Chief Inspector of Factories on Industrial Health, 1964.
London: H.M. Stationery Office, 1969. 31 pp.
16. Annual Report of H.M. Chief Inspector of Factories on Industrial Health, 1966.
London: H.M. Stationery Office, 1967. 45 pp.
17. Arcos, J. C., M. F. Argus, and G. Wolf. Chemical Induction of Cancer. Struc-
tural Bases and Biological Mechanisms. 4 vols. (2nd ed.) New York: Academic
Press Inc., 1968.
18. Auerbach, O., E. C. Hammond, D. Kirman, and L. Garfinkel. Effects of
cigarette smoking on dogs. II. Pulmonary neoplasms. Arch. Environ. Health
21:754-768, 1970.
19. Ayres, S. M., and M. E. Buehler. The effects of urban air pollution on health.
Clin. Pharmacol. Ther. 11:337-371, 1970.
20. Baba, H., and C. Mugiya. Temperature dependence of fluorescence spectra and
fluorescence polarization of some organic compounds in polar solvents. The
Franck-Condon excited states. Bull. Chem. Soc. Jap. 43:13-19, 1970.
21. Badger, G. M. Mode of formation of carcinogens in human environment. Nat.
Cancer Inst. Monogr. 9:1-16, 1962.
22. Badger, G. M., and T. M. Spotswood. The formation of aromatic hydrocarbons
at high temperatures. Part IX. The pyrolysis of toluene, ethylbenzene, propyl-
benzene and butylbenzene. J. Chem. Soc. 1960:4420-4427.
23. Bailey, E. J., and N. Dungal. Polycyclic hydrocarbons in Icelandic smoked food.
Brit. J. Cancer 12:348-350, 1958:
24. Bailey, P. S. The reactions of ozone with organic compounds. Chem. Rev.
58:925-1010, 1958.
25. Bailey, P. S., J. E. Batterbee, and A. G. Lane. Ozonation of benz[a] anthracene.
J. Amer. Chem. Soc. 90:1027-1033, 1968.
26. Bailey, P. S., J. E. Keller, D. A. Mitchard, and H. M. White. Ozonation of amines.
Adv. Chem. Ser. 77:58-64, 1968.
27. Bair, W. J. Inhalation of radionuclides and carcinogenesis, pp. 77-101. In
M. G. Hanna, Jr., P. Nettesheim, and J. R. Gilbert, Eds. Inhalation Carcino-
genesis. AEC Symposium Series, No. 18. Washington, D.C.: U.S. Atomic
Energy Commission, 1970.
28. Baldwin, R. W., and C. R. Barker. Tumour-specific antigenicity of aminoazo-
dye-induced rat hepatomas. Int. J. Cancer 2:355-364, 1967.
29. Baldwin, R. W., C. R. Barker, and M. J. Embleton. Immunology of carcinogen-
-------
References 309
induced and spontaneous rat tumours, pp. 503-505. In J. Dausset, J. Hamburger,
and G. Mathe, Eds. Advance in Transplantation. Proceedings of the First Inter-
national Congress of the Transplantation Society, Paris, 27-30 June 1967.
Copenhagen: Munksgaard, 1968.
30. Baldwin, R. W., and M. J. Embleton. Immunology of 2-acetylaminofluorene-
induced rat mammary adenocarcinomas. Int. J. Cancer 4:47-53, 1969.
31. Bang, F. B. Effects of invading organisms on cells and tissues in culture, pp.
151-261. In E. N. Willmer, Ed. Cells and Tissues in Culture. Methods, Biology,
and Physiology. Vol. 3. London: Academic Press Inc., 1966.
32. Barger, W. R., and W. D. Garrett. Surface active organic material in the marine
atmosphere. J. Geophys. Res. 75:4561-4566, 1970.
33. Barnett, E. deB., J. W. Cook, and H. H. Grainger. Studies in the anthracene
series. Part III. J. Chem. Soc. 121:2059-2069, 1922.
34. Bateman, A. J. Mutagenic sensitivity of maturing germ cells in the male mouse.
Heredity 12:213-232, 1958.
35. Bateman, A. J. Testing chemicals for mutagenicity in a mammal. Nature 210:
205-206, 1966.
36. Batsakis, J. G., and H. A. Johnson. Generalized scleroderma involving lungs
and liver with pulmonary adenocarcinoma. Arch. Path. 69:633-638, 1960.
37. Baum, E. J. Photochemistry of organic crystals—anthracene. In J. N. Pitts, Jr.,
Ed. Excited State Chemistry. New York: Gordon & Breach, Science Pub-
lishers, Inc. (in press)
38. Becker, R. S. Theory and Interpretation of Fluorescence and Phosphorescence.
New York: John Wiley & Sons, Inc., 1969. 283 pp.
39. Beer, R., K. M. C. Davis, and R. Hodgson. Formation of excited charge transfer
complexes in the quenching of anthracene fluorescence by anions. J. Chem.
Soc. 1970D: 840-841.
40. Begeman, C. R. Carcinogenic aromatic hydrocarbons in automobile effluents,
pp. 163-174. In Vehicle Emissions (Selected SAE Papers). (Paper 440C
presented January 1962 at the SAE Automotive Engineering Congress.) New
York: Society of Automotive Engineers, Inc., 1964.
41. Begeman, C. R., and J. M. Colucci. Benzo(a)pyrene in gasoline partially per-
sists in automobile exhaust. Science 161:271, 1968.
42. Begeman, C. R., and J. M. Colucci. Polynuclear Aromatic Hydrocarbon
Emissions from Automotive Engines. SAE Paper 700469. Detroit: Society of
Automotive Engineers, 1970. 13 pp.
43. Bell, J. Paraffin epithelioma of the scrotum. Edinburgh Med. J. 22:135-137,
1876.
44. Bell, J. H., S. Ireland, and A. W. Spears. Identification of aromatic ketones
in cigarette smoke condensate. Anal. Chem. 41:310-313, 1969.
45. Bentley, H. R., and J. G. Burgan. Polynuclear hydrocarbons in tobacco and
tobacco smoke. Part I. 3:4-Benzopyrene. Analyst 83:442-447, 1958.
46. Bentley, H. R., and J. G. Burgan. Polynuclear hydrocarbons in tobacco and
tobacco smoke. Part II. The origin of 3:4-benzopyrene found in tobacco and
tobacco smoke. Analyst 85:723-727, 1960.
47. Berenblum, I., N. Haran-Ghera, and N. Trainin. An experimental analysis
of ihe~"hair cycle effect" in mouse skin carcinogenesis. Brit. J. Cancer
12:402-413, 1958.
-------
310 References
48. Berenblum, I., and P. Shubik. The persistence of latent tumour cells induced
in the mouse's skin by a single application of 9:10-dimethyl-l :2-benzanthra-
cene. Brit. J. Cancer 3:384-386, 1949.
49. Bergman, I. The polarography of polycyclic aromatic hydrocarbons and the
relationship between their half-wave potentials and absorption spectra.
Trans. Faraday Soc. 50:829-838, 1954.
50. Berlman, I. B. Handbook of Fluorescence Spectra of Aromatic Molecules.
New York: Academic Press Inc., 1965. 258 pp.
51. Berlman, I. B. On an empirical correlation between nuclear conformation
and certain fluorescence and absorption characteristics of aromatic com-
pounds. J. Phys. Chem. 74:3085-3093, 1970.
52. Berwald, Y., and L. Sachs. In vitro cell transformation with chemical carcino-
gens. Nature 200:1182-1184, 1963.
53. Berwald, Y., and L. Sachs. In vitro transformation of normal cells into tu-
mor cells by carcinogenic hydrocarbons. J. Nat. Cancer Inst. 35:641-661,
1965.
54. Best, E. W. R. A Canadian Study of Smoking and Health. Ottawa: Depart-
ment of National Health and Welfare, 1966. 137 pp.
55. Betteridge, D., and A. D. Baker. Analytical potential of photoelectron
spectroscopy. Anal. Chem. 42:43A-44A, 46A, 48A, 50A, 52A, 54A, 56A,
Jan. 1970.
56. Biesele, J. J. Mitotic Poisons and the Cancer Problem. Amsterdam: Elsevier
Publishing Co., 1958. 214 pp.
57. Bingham, E., and A. W. Horton. Environmental carcinogenesis: Experimental
observations related to occupational cancer, pp. 183-193. In W. Montagna
and R. L. Dobson, Eds. Advances in Biology of Skin. Vol. VII. Carcino-
genesis. Oxford: Pergamon Press, 1966.
58. Bingham, E., A. W. Horton, and R. Tye. The carcinogenic potency of cer-
tain oils. Arch. Environ. Health 10:449-451, 1965.
59. Biological effects of asbestos. Ann. N.Y. Acad. Sci. 132(Art. l):l-766, 1965.
60. Birks, J. B. Photophysics of Aromatic Molecules. New York: John Wiley
& Sons, Inc., 1970. 704 pp.
61. Birks, J. B. Scintillations in organic solids, pp. 433-508. In D. Fox, M. M.
Labes, and A. Weissberger, Eds. Physics and Chemistry of the Organic Solid
State. Vol. 2. New York: John Wiley & Sons, Inc., 1965.
62. Birks, J. B., and A. J. W. Cameron. Crystal fluorescence of carcinogens and
related organic compounds. Proc. Roy. Soc. London 249A:297-317, 1959.
63. Blifford, I. H., Jr. Tropospheric aerosols. J. Geophys. Res. 75:3099-3103,
1970.
64. Blum, H. F. Sunlight as a causal factor in cancer of the skin of man. J. Nat.
Cancer Inst. 9:247-258, 1948.
65. Blumenthal, H. T., and J. B. Rogers. Spontaneous and induced tumors in
the guinea pig, with special reference to the factor of age. Progr. Exp. Tumor
Res. 9:261-285, 1967.
66. Blumer, M. Benzpyrenes in soil. Science 134:474-475, 1961.
67. Bock, F. G. Early effects of hydrocarbons on mammalian skin. Progr. Exp.
Tumor Res. 4:126-168, 1964.
68. Bock, F. G., and G. E. Moore. Carcinogenic activity of cigarette-smoke con-
-------
References 311
densate. I. Effect of trauma and remote X irradiation. J. Nat. Cancer Inst.
22:401-411, 1959.
69. Bock, F. G., G. E. Moore, and S. K. Crouch. Tumor-promoting activity of
extracts of unburned tobacco. Science 145:831-833, 1964.
70. Bock, F. G., and R. Mund. A survey of compounds for activity in the sup-
pression of mouse sebaceous glands. Cancer Res. 18:887-892, 1958.
71. Bock, F. G., and R. Mund. Evaluation of substances causing loss of sebaceous
glands from mouse skin. J. Invest. Derm. 26:479-487, 1956.
72. Bogacz, J., and I. Koprowska. A cy to-pathologic study of potentially carcin-
ogenic properties of air pollutants. Acta Cytol. (Baltimore) 5:311-319, 1961.
73. Boiling, H. Sostanze cancerogene nei cereali sottoposti ad essiccazione con
gas di combustione. Tec. Molitoria 15:137-142, 1964.
74. Bonser, G. M., E. Boyland, E. R. Busby, D. B. Clayson, P. L. Grover, and
J. W. Jull. A further study of bladder implantation in the mouse as a means of
detecting carcinogenic activity: Use of crushed paraffin wax or stearic acid
as the vehicle. Brit. J. Cancer 17:127-136, 1963.
75. Boren, H. G. Carbon as a carrier mechanism for irritant gases. Arch. Environ.
Health 8:119-124, 1964.
76. Borneff, J., and R. Fischer. Cancerogene Substanzen in Wasser und Boden.
VIII. Untersuchungen an Filter-Aktivkohle nach Verwendung im Wasserwerk.
Arch. Hyg. Bakt. 146:1-16, 1962. (summary in English)
77. Borneff, J., F. Selenka, H. Kunte, and A. Maximos. Experimental studies on
the formation of polycyclic aromatic hydrocarbons in plants. Environ. Res.
2:22-29, 1968.
78. Bornstein, R. D. Observations of the urban heat island effect in New York
City. J. Appl. Meteorol. 7:575-582, 1968.
79. Boubel, R. W., and L. A. Ripperton. Benzo(a)pyrene production during con-
trolled combustion. J. Air Pollut. Control Assoc. 13:553-557, 1963.
80. Boutwell, R. K. Some biological aspects of skin carcinogenesis. Progr. Exp.
Tumor Res. 4:207-250, 1964.
81. Boveri, T. The Origin of Malignant Tumors. Translated by M. Boveri. Balti-
more: The Williams & Wilkins Co., 1929. 119 pp.
82. Bowen, E. J. Luminescence in Chemistry. London: D. Van Nostrand Co.,
Ltd., 1968.254pp.
83. Boyland, E., and P. Sims. Metabolism of polycyclic compounds. 24. The
metabolism of benz[a] anthracene. Biochem. J. 91:493-506, 1964.
84. Boyland, E., and P. Sims. The metabolism of benz[a] anthracene and dibenz-
[a,h] anthracene and their 5,6-epoxy-5,6-dihydro derivatives by rat-liver
homogenates. Biochem. J. 97:7-16, 1965.
85. Boyland, E., P. Sims, and C. Huggins. Induction of adrenal damage and can-
cer with metabolites of 7,12-dimethylbenz(a)anthracene. Nature 207:816-
817, 1965.
86. Boyland, E., and G. Watson. 3-Hydroxyanthranilic acid, a carcinogen pro-
duced by endogenous metabolism. Nature 177:837-838, 1956.
87. Braun, A. C. The Cancer Problem. A Critical Analysis and Modern Synthe-
sis. New York: Columbia University Press, 1969. 209 pp.
88. Braun, A. C., and H. N. Wood. The plant tumor problems. Adv. Cancer Res.
6:81-109,1961.
-------
312 References
89. Bridge, J. C., and S. A. Henry. Industrial cancers, pp. 258-268. In Report
of the International Conference on Cancer, London, 17th-20th July, 1928.
New York: William Wood and Company, 1928.
90. Bridges, B. A., and J. Huckle. Mutagenesis of cultured mammalian cells by
x-radiation and ultraviolet light. Mutat. Res. 10:141-151, 1970.
91. Bridges, B. A., J. Huckle, and M. J. Ashwood-Smith. X-ray mutagenesis of
cultured Chinese hamster cells. Nature 226:184-185, 1970.
92. Brookes, P., and C. Heidelberger. Isolation and degradation of DNA from
cells treated with tritium-labeled 7,12-dimethylbenz(a)anthracene: Studies
on the nature of the binding of this carcinogen to DNA. Cancer Res. 19:
157-165, 1969.
93. Brookes, P., and P. D. Lawley. Evidence for the binding of polynuclear aro-
matic hydrocarbons to the nucleic acids of mouse skin: Relation between
carcinogenic power of hydrocarbons and their binding to deoxyribonucleic
acid. Nature 202:781-784, 1964.
94. Brown, R. D. A theoretical treatment of the Diels-Alder reaction. Part I.
Polycyclic aromatic hydrocarbons. J. Chem. Soc. 1950:691-697.
95. Brues, A. M. Critique of the linear theory of carcinogenesis. Science 128:
693-699, 1958.
96. Bryan, G. T., and P. D. Springberg. Role of the vehicle in the genesis of
bladder carcinoma in mice by the pellet implantation technic. Cancer Res.
26:105-109, 1966.
97. Bryan, W. R., and M. B. Shimkin. Quantitative analysis of dose-response data
obtained with carcinogenic hydrocarbons. J. Nat. Cancer Inst. 1:807-833,
1941.
98. Budzikiewicz, H., C. Djerassi, and D. H. Williams. Structure Elucidation of
Natural Products by Mass Spectrometry. Vol. I. Alkaloids. San Francisco:
Holden-Day, Inc., 1964. 233 pp.
99. Buell, P., and J. E. Dunn, Jr. Relative impact of smoking and air pollution
on lung cancer. Arch. Environ. Health 15:291-297, 1967.
100. Buell, P., J. E. Dunn, Jr., and L. Breslow. Cancer of the lung and Los An-
geles-type air pollution. Prospective study. Cancer 20:2139-2147, 1967.
101. Burdette, W. J. The significance of mutation in relation to the origin of
tumors. A review. Cancer Res. 15:201-226, 1955.
102. Burrows, I. E., and A. J. Lindsey. Formation of polycyclic aromatic hydro-
carbons by pyrolysis of simple aliphatic hydrocarbons. Chem. Ind. 1961:1395.
103. Burstein, N. A., K. R. Mclntire, and A. C. Allison. Pulmonary tumors in
germfree mice: Induction with urethan. J. Nat. Cancer Inst. 44:211-214,
1970.
104. Butlin, H. T. Three lectures on cancer of the scrotum in chimney-sweeps
and others. Lecture 1. Secondary cancer without primary cancer. Brit. Med.
J. 1:1341-1346, 1892.
105. Butlin, H. T. Three lectures on cancer of the scrotum in chimney-sweeps
and others. Lecture 2. Why foreign sweeps do not suffer from scrotal cancer.
Brit. Med. J. 2:1-6, 1892.
106. Butlin, H. T. Three lectures on cancer of the scrotum in chimney-sweeps
and others. Lecture 3. Tar and paraffin cancer. Brit. Med. J. 2:66-71, 1892.
107. Cahnmann, H. J., and M. Kuratsune. Determination of polycyclic aromatic
-------
References 313
hydrocarbons in oysters collected in polluted water. Anal. Chem. 29:1312-
1317, 1957.
108. Campbell, J. A. Cancer of skin and increase in incidence of primary tumours
of lung in mice exposed to dust obtained from tarred roads. Brit. J. Exp.
Path. 15:287-294, 1934.
109. Campbell, J. A. Influenza virus and incidence of primary lung tumours in
mice. Lancet 2:487, 1940.
110. Campbell, J. A. Lung tumours in mice. Incidence as affected by inhalation
of certain carcinogenic agents and some dusts. Brit. Med. J. 1:217-221,
1942.
111. Campbell, J. M., and R. L. Cooper. The presence of 3,4-benzpyrene in snuff
associated with a high incidence of cancer. Chem. Ind. 1955:64-65.
112. Campbell, J. M., and L. Kreyberg. The degree of air pollution in Norwegian
towns. Brit. J. Cancer 10:481-484, 1956.
113. Campbell, J. M., and A. J. Lindsey. Polycyclic aromatic hydrocarbons in
snuff. Chem. Ind. 1957:951.
114. Campbell, J. M., and A. J. Lindsey. Polycyclic hydrocarbons extracted from
tobacco: The effect upon total quantities found in smoking. Brit. J. Cancer
10:649-652, 1956.
115. Cantuti, V., G. P. Cartoni, A. Liberti, and A. G. Torri. Improved evaluation
of polynuclear hydrocarbons in atmospheric dust by gas chromatography.
J. Chromatogr. 17:60-65, 1965.
116. Cardiovascular Diseases in the U.S., Facts and Figures. New York: American
Heart Association, 1965.
117. Carnow, B. W. Pulmonary disease and air pollution. A Chicago problem.
Chicago Med. 71:581-586, 1968.
118. Carozzi, L. Le cancer professionnel. Arch. Elect. Med. 42:85-93, 118-142,
155-160, 1934.
119. Cartwright, J. Particle shape factors. Ann. Occup. Hyg. 5:163-171, 1962.
120. Carugno, N., and S. Rossi. Evaluation of polynuclear hydrocarbons in ciga-
ret smoke by glass capillary columns. J. Gas Chromatogr. 5:103-106, 1967.
121. Casarett, G. W. Experimental radiation carcinogenesis. Progr. Exp. Tumor
Res. 7:49-82, 1965.
122. Cattanach, B. M. A genetical approach to the effects of radiomimetic
chemicals on fertility in mice, pp. 415-426. In W. D. Carlson and F. X.
Fassner, Eds. Effects of Ionizing Radiation on the Reproductive System.
Proceedings of an International Symposium held at Colorado State Univer-
sity, Fort Collins, Colorado. New York: The Macmillan Co., 1964.
123. Cattanach, B. M., C. E. Pollard, and J. H. Isaacson. Ethyl methanesulfonate-
induced chromosome breakage in the mouse. Mutat. Res. 6:297-307,
1968.
124. Cavalieri, E., and M. Calvin. Molecular characteristics of some carcinogenic
hydrocarbons (benzpyrene/dimethylbenzanthracene/3-methylcholanthrene/
nucleic acids). Proc. Nat. Acad. Sci. U.S.A. 68:1251-1253, 1971.
125. Cavill, G. W. K., A. Robertson, and W. B. Whalley. The chemistry of fungi.
Part VIII. The oxidation of methylene groups in compounds analogous to
O-dimethylcitromycin. J. Chem. Soc. 1949:1567-1570.
126. Cetorelli, J. J., W. J. McCarthy, and J. P. Winefordner. The selection of
-------
314 References
optimum conditions for spectrochemical methods. IV. Sensitivity of absorp-
tion, fluorescence, and phosphorescence spectrometry in the condensed
phase. J. Chem. Educ. 45:98-102, 1968.
127. Chakraborty, B. B., and R. Long. Gas chromatographic analysis of poly-
cyclic aromatic hydrocarbons in soot samples. Environ. Sci. Technol. 1:828-
834, 1967.
128. Chamberlain, A. C. Interception and retention of radioactive aerosols by
vegetation. Atmos. Environ. 4:57-78, 1970.
129. Charlson, R. J., N. C. Ahlquist, and H. Horvath. On the generality of cor-
relation of atmospheric aerosol mass concentration and light scatter. Atmos.
Environ. 2:455-464, 1968.
130. Chassevant, F., and M. Heros. Recherche du benzo3,4,pyrene dans le cafe
vert et torrefie et dans les sous-produits de torre faction. Cafe, Cacao, The
7:349-358, 1963.
131. Chen, T. T., and C. Heidelberger. Cultivation in vitro of cells derived from
adult C3H mouse ventral prostate. J. Nat. Cancer Inst. 42:903-914, 1969.
132. Chen, T. T., and C. Heidelberger. In vitro malignant transformation of cells
derived from mouse prostate in the presence of 3-methylcholanthrene. J.
Nat. Cancer Inst. 42:915-925, 1969.
133. Chen, T. T., and C. Heidelberger. Quantitative studies on the malignant
transformation of mouse prostate cells by carcinogenic hydrocarbons in
vitro. Int. J. Cancer 4:166-178, 1969.
134. Chouroulinkov, I., P. Lazar, C. Izard, C. Libermann, and M. Guerin. "Seba-
ceous glands" and "hyperplasia" tests as screening methods for tobacco tar
carcinogenesis. J. Nat. Cancer Inst. 42:981-985, 1969.
135. Chu, E. H. Y. Mammalian cell genetics. III. Characterization of X-ray-in-
duced forward mutations in Chinese hamster cell cultures. Mutat. Res. 11:
23-34, 1971.
136. Chu, E. H. Y., E. G. Bailiff, and H. V. Mailing. Mutagenicity of chemical
carcinogenesis in mammalian cells, pp. 62-63. In Tenth International Can-
cer Congress, Houston, 1970. Abstracts. Houston, Texas: Medical Arts
Publishing Co., 1970.
137. Chu, E. H. Y., and H. V. Mailing. Mammalian cell genetics. II. Chemical
induction of specific locus mutations in Chinese hamster cells in vitro. Proc.
Nat. Acad. Sci. U.S.A. 61:1306-1312, 1968.
138. Chu, E. W., and R. A. Malmgren. An inhibitory effect of vitamin A on the
induction of tumors of forestomach and cervix in the Syrian hamster by
carcinogenic polycyclic hydrocarbons. Cancer Res. 25:884-895, 1965.
139. Clar, E. J. Polycych'c Hydrocarbons. 2 vols. New York: Academic Press
Inc., 1964.974pp.
140. Clark, J. H. The effect of long ultraviolet radiation on the development of
tumors induced by 20-methylcholanthrene. Cancer Res. 24:207-211, 1964.
141. Clark, W. E., and K. T. Whitby. Concentration and size distribution mea-
surements of atmospheric aerosols and a test of the theory of self-preserv-
ing size distributions. J. Atmos. Sci. 24:677-687, 1967.
142. Clayson, D. B. Chemical Carcinogenesis, pp. 164-171. Boston: Little,
Brown and Co., 1962.
-------
References 315
143. Cleary, G. J. Discrete separation of polycyclic hydrocarbons in air-borne
participates using very long alumina columns. J. Chromatogr. 9:204-215,
1962.
144. Clemmesen, J. Bronchial carcinoma—a pandemic. Dan. Med. Bull. 1:37-
46, 1954.
145. Clemmesen, J., and A. Nielsen. The social distribution of cancer in Copen-
hagen, 1943-1947. Brit. J. Cancer 5:159-171, 1951.
146. Clemo, G. R. Some constituents of city smoke. Tetrahedron 23:2389-2393,
1967.
147. Clemo, G. R., and E. W. Miller. The carcinogenic action of city smoke. Chem.
Ind. 1955:38.
148. Coffin, D. L. Health Effects of Airborne Polycyclic Hydrocarbons. Presented
at the Symposium on Human Health and Vehicle Emissions. Detroit: Society
of Automotive Engineers, 1971. 23 pp.
149. Cohart, E. M. Socioeconomic distribution of cancer of the lung in New
Haven. Cancer 8:1126-1129, 1955.
150. Cohart, E. M. Socioeconomic distribution of stomach cancer in New Haven.
Cancer 7:455-461, 1954.
151. Colucci, J. M., and C. R. Begeman. Carcinogenic air pollutants in relation to
automotive traffic in New York. Environ. Sci. Tech. 5:145-150, 1971.
152. Colucci, J. M., and C. R. Begeman. Polynuclear Aromatic Hydrocarbons
and Other Pollutants in Los Angeles. Presented at the 2nd International
Clear Air Congress, Washington, D.C., Dec. 6-11, 1970.
153. Colucci, J. M., and C. R. Begeman. The automotive contribution to air-
borne polynuclear aromatic hydrocarbons in Detroit. J. Air Pollut. Control
Assoc. 15:113-122, 1965.
154. Commins, B. T. Interim report on the study of techniques for determination
of polycyclic aromatic hydrocarbons in air. Nat. Cancer Inst. Monogr. 9:
225-233, 1962.
155. Commins, B. T. Polycyclic hydrocarbons in rural and urban air. Int. J. Air
Pollut. 1:14-17, 1958.
156. Commins, B. T., R. L. Cooper, and A. J. Lindsey. Polycyclic hydro-
carbons in cigarette smoke. Brit. J. Cancer 8:296-302, 1954.
157. Conney, A. H. Pharmacological implications of microsomal enzyme induc-
tion. Pharmacol. Rev. 19:317-366, 1967.
158. Cook, J. W., and R. H. Martin. Polycyclic aromatic hydrocarbons. Part
XXIV. J. Chem. Soc. 1940:1125-1127.
159. Coomber, J. W., and J. N. Pitts, Jr. Singlet oxygen in the environmental
sciences. VIII. Production of O2 ('Ag) by energy transfer from excited benz-
aldehyde under simulated atmospheric conditions. Environ. Sci. Technol.
4:506-510, 1970.
160. Cooper, R. L., and A. J. Lindsey. Atmospheric pollution by polycyclic
hydrocarbons. Chem. Ind. 1953:1177-1178.
161. Cooper, R. L., and A. J. Lindsey. 3:4-Benzpyrene and other polycyclic
hydrocarbons in cigarette smoke. Brit. J. Cancer 9:304-309, 1955.
162. Cooper, R. L., A. J. Lindsey, and R. E. Waller. The presence of 3,4-benz-
pyrene in cigarette smoke. Chem. Ind. 1954:1418.
-------
316 References
163. Corbett, T. H., C. Heidelberger, and W. F. Dove. Determination of the mu-
tagenic activity to bacteriophage T4 of carcinogenic and noncarcinogenic
compounds. Molec. Pharmacol. 6:667-679, 1970.
164. Corn, M. Nonviable particles in the air, pp. 47-94. In A. C. Stern, Ed. Air
Pollution. Vol. 1. Air Pollution and Its Effects. (2nd ed.) New York: Aca-
demic Press Inc., 1968.
165. Corn, M. Urban aerosols: Problems associated with evaluation of inhala-
tion risk. In Assessment of Airborne Particles. Proceedings of the 3rd Uni-
versity of Rochester Conference on Environmental Toxicology, June 1970.
Springfield, 111.: Charles C Thomas, 1971.
166. Corn, M., T. L. Montgomery, and N. A. Esmen. Suspended particulate mat-
ter: Seasonal variation in specific surface areas and densities. Environ. Sci.
Tech. 5:155-158, 1971.
167. Crocker, T. T. Effect of benzo[a]pyrene on hamster, rat, dog and monkey
respiratory epithelia in organ culture, pp. 433-443. In M. G. Hanna, Jr.,
P. Nettesheim, and J. R. Gilbert, Eds. Inhalation Carcinogenesis. Proceedings
of a Biology Division, Oak Ridge National Laboratory, conference held in
Gatlinburg, Tenn., Oct. 8-11, 1969. AEC Symposium Series 18. Oak Ridge,
Tenn.: U.S. Atomic Energy Commission, Division of Technical Information,
1970.
168. Crocker, T. T., J. E. Chase, S. A. Wells, and L. L. Nunes. Preliminary report
on experimental squamous carcinoma of the lung in hamsters and in a pri-
mate (Galago crassicaudatus), pp. 317-328. In P. Nettesheim, M. G. Hanna,
Jr., and J. W. Deatherage, Jr., Eds. Morphology of Experimental Respiratory
Carcinogenesis. Proceedings of a Biology Division, Oak Ridge National Lab-
oratory, conference held in Gatlinburg, Tenn., May 13-16, 1970. AEC
Symposium Series 21. Oak Ridge, Tenn.: U.S. Atomic Energy Commission,
Division of Technical Information, 1970.
169. Crocker, T. T., and B. I. Nielsen. Effect of carcinogenic hydrocarbons on
suckling rat trachea in living animals and in organ cultures, pp. 765-787.
In L. Seven, Ed. Lung Tumours in Animals. Proceedings of the Third
Quadrennial International Conference on Cancer. June 24-29, 1965. Perugia,
Italy: University of Perugia, 1966.
170. Crocker, T. T., B. I. Nielsen, and I. Lasnitzki. Carcinogenic hydrocarbons.
Effects on suckling rat trachea in organ culture. Arch. Environ. Health
10:240-250, 1965.
171. Crocker, T. T., and L. L. Sanders. Influence of vitamin A and 3,7-dimethyl-
2,6-octadienal (citral) on the effect of benzo(a)pyrene on hamster trachea
in organ culture. Cancer Res. 30:1312-1318, 1970.
172. Crow, J. F. Chemical risk to future generations. Sci. Citizen 10:113-117,
1968.
173. Cruickshank, C. N. D., and A. Gourevitch. Skin cancer of the hand and
forearm. Brit. J. Ind. Med. 9:74-79, 1952.
174. Cruickshank, C. N. D., and J. R. Squire. Skin cancer in the engineering in-
dustry from the use of mineral oil. Brit. J. Ind. Med. 7:1-11, 1950.
175. Crummett, W., and R. Hummel. Ultraviolet spectrometry. Anal. Chem.
42:239R-248R, 1970.
176. Cruser, S. A., and A. J. Bard. Electrogenerated chemiluminescence. III.
-------
References 317
Intensity-time and concentration-intensity relation and the lifetime of radi-
cal cations of aromatic hydrocarbons in JV.Af-dimethylformamide solution.
J. Amer. Chem. Soc. 91:267-275, 1969.
177. Curwen, M. P., E. L. Kennaway, and N. M. Kennaway. The incidence of
cancer of the lung and larynx in urban and rural districts. Brit. J. Cancer
8:181-198, 1954.
178. Dalhamn, T., M.-L. Edfors, and R. Rylander. Retention of cigarette smoke
components in human lungs. Arch. Environ. Health 17:746-748, 1968.
179. Daniel, P. M., O. E. Pratt, and M. M. L. Prichard. Metabolism of labelled
carcinogenic hydrocarbons in rats. Nature 215:1142-1146, 1967.
180. Darlington, C. D. The plasmagene theory of the origin of cancer. Brit. J.
Cancer 2:118-126, 1948.
181. Daudel, P., and R. Daudel. Chemical Carcinogenesis and Molecular Biology.
New York: John Wiley & Sons, Inc., 1966. 158 pp.
182. Dautrebande, L. Physiological and pharmacological characteristics of liquid
aerosols. Physiol. Rev. 32:214-275, 1952.
183. Davies, C. N. Size, area, volume and weight of dust particles. Ann. Occup.
Hyg. 3:219-225, 1961.
184. Davies, J. H. Correlation spectroscopy. Anal. Chem. 42:101A-105A, 108A,
110A, 112A, May 1970.
185. Davies, W., and J. R. Wilmshurst. Carcinogens formed in the heating of food
stuffs. Formation of 3,4-benzopyrene from starch at 370-390°C. Brit. J.
Cancer 14:295-299, 1960.
186. Day, T. D. Carcinogenic action of cigarette smoke condensate on mouse
skin. An attempt at a quantitative study. Brit. J. Cancer 21:56-81, 1967.
187. Dean, G. Lung cancer among white South Africans. Brit. Med. J. 2:852-
857, 1959.
188. Dean, G. Lung cancer and bronchitis in Northern Ireland, 1960-2. Brit.
Med.J. 1:1506-1514,1966.
189. Dean, G. Lung cancer in South Africans and British immigrants. Proc. Roy.
Soc. Med. 57:984-987, 1964.
190. Delia Porta, G. Some aspects of medical drug testing for carcinogenic
activity, pp. 33-47. In R. Truhaut, Ed. Potential Carcinogenic Hazards from
Drugs: Evaluation of Risks. (UICC Monograph Series, Vol. 7.) Berlin:
Springer-Verlag, 1967.
191. Delia Porta, G., P. Shubik, K. Dammert, and B. Terracini. Role of poly-
oxyethylene sorbitan monostearate in skin carcinogenesis in mice. J. Nat.
Cancer Inst. 25:607-625, 1960.
192. DeMaio, L., and M. Corn. Gas chromatographic analysis of polynuclear
aromatic hydrocarbons with packed columns. Application to air pollution
studies. Anal. Chem. 38:131-133, 1966.
193. DeMaio, L., and M. Corn. Polynuclear aromatic hydrocarbons associated
with particulates in Pittsburgh air. J. Air Pollut. Control Assoc. 16:67-
71, 1966.
194. Demas, J. N., and G. A. Crosby. Photoluminescence decay curves: An
analysis of the effects of flash duration and linear instrumental distortions.
Anal. Chem. 42:1010-1017, 1970.
195. Demisch, R. R., and G. F. Wright. The distribution of polynuclear aromatic
-------
318 References
hydrocarbons between aqueous and non-aqueous phases. Can. J. Biochem.
Physiol. 41:1655-1662, 1963.
196. Dewar, M. J. S. A molecular orbital theory of organic chemistry. VI. Aro-
matic substitution and addition. J. Amer. Chem. Soc. 74:3357-3363, 1952.
197. Diamond, L. Effect of carcinogenic hydrocarbons on rodent and primate
cells in vitro. J. Cell. Physiol. 66:183-197, 1965.
198. Diamond, L., and H. V. Gelboin. Alpha-naphthoflavone: An inhibitor of
hydrocarbon cytotoxicity and microsomal hydroxylase. Science 166:1023-
1025, 1969.
199. Diamond, L., C. Sardet, and G. H. Rothblat. The metabolism of 7,12-
dimethylbenz(a)anthracene in cell cultures. Int. J. Cancer 3:838-849, 1968.
200. Dickey, F. H., G. H. Cleland, and C. Lotz. The r61e of organic peroxides in
the induction of mutations. Proc. Nat. Acad. Sci. U.S.A. 35:581-586, 1949.
201. DiPaolo, J. A., P. Donovan, and R. Nelson. Quantitative studies of in vitro
transformation by chemical carcinogens. J. Nat. Cancer Inst. 42:867-874,
1969.
202. DiPaolo, J. A., R. L. Nelson, and P. J. Donovan. Sarcoma-producing cell
lines derived from clones transformed in vitro by benzo[a]pyrene. Science
165:917-918, 1969.
203. Dipple, A., P. D. Lawley, and P. Brookes. Theory of tumour initiation by
chemical carcinogens: Dependence of activity on structure of ultimate car-
cinogen. Bur. J. Cancer 4:493-506, 1968.
204. Dirksen, E. R., and T. T. Crocker. Ultrastructural alterations produced by
polycyclic aromatic hydrocarbons on rat tracheal epithelium in organ cul-
ture. Cancer Res. 28:906-923, 1968.
205. Doll, R. The causes of death among gas-workers with special reference to
cancer of the lung. Brit. J. Ind. Med. 9:180-187, 1952.
206. Doll, R., R. E. W. Fisher, E. J. Gammon, W. Gunn, G. O. Hughes, F. H.
Tyrer, and W. Wilson. Mortality of gasworkers with special reference to cancer
of the lung and bladder, chronic bronchitis, and pneumoconiosis. Brit. J.
Ind. Med. 22:1-12, 1965.
207. Doll, R., and A. B. Hill. Mortality in relation to smoking: Ten years' obser-
vations of British doctors. Brit. Med. J. 1:1399-1410, 1460-1467, 1964.
208. Doll, R., and A. B. Hill. Mortality of British doctors in relation to smoking:
Observations on coronary thrombosis. Nat. Cancer Inst. Monogr. 19:205-
268, 1966.
209. Doll, R., A. B. Hill, P. G. Gray, and E. A. Parr. Lung cancer mortality and
the length of cigarette ends. An international comparison. Brit. Med. J.
5118:322-325, 1959.
210. Domsky, I., W. Lijinsky, K. Spencer, and P. Shubik. Rate of metabolism of
9,10-dimethyl-l,2-benzanthracene in newborn and adult mice. Proc. Soc.
Exp. Biol. Med. 113:110-112, 1963.
211. Doniach, I., and J. C. Mottram. On the effect of light upon the incidence of
tumours in painted mice. Amer. J. Cancer 39:234-240, 1940.
212. Doniach, I., J. C. Mottram, and F Weigert. The fluorescence of 3:4 benz-
pyrene in vivo. Part II. The inter-relationship of the derivatives formed in
various sites. Brit. J. Exp. Path. 24:9-14, 1943.
213. Dontenwill, W., H. Elmenhorst, G. Reckzeh, H.-P. Harke, and L. Stadler.
-------
References 319
Experimentelle Untersuchungen Uber Aufnahme, Abtransport und Abbau
cancerogener Kohlenwasserstoffe im Bereich des Respirationstraktes. Z.
Krebsforsch. 71:225-243, 1968.
214. Dorn, H. F. Morbidity and mortality from bladder tumors in North and
South America. Acta Un. Int. Cancr. 18:553-559, 1962.
215. Dc5rr, R. Die Aufnahme von Alkaloiden und Benzpyren durch intakte
Pflanzenwurzeln. Naturwissenschaften 52:166, 1965.
216. Druckrey, H. Pharmacological approach-to carcinogenesis, pp. 110-130. In
G. E. W. Wolstenholme and M. O'Connor, Eds. Ciba Foundation Symposium
on Carcinogenesis. Mechanisms of Action. Boston: Little, Brown and Co.,
1958.
217. Drushel, H. V., A. L. Sommers, and R. C. Cox. Correction of luminescence
spectra and calculation of quantum efficiencies using computer techniques.
Anal. Chem. 35:2166-2172, 1963.
218. Dubois, L., A. Corkery, and J. L. Monkman. The chromatography of poly-
cyclic hydrocarbons. Int. J. AirPollut. 2:236-252, 1960.
219. Duchen, L. W. Bronchogenic carcinoma. Incidence and pathology as seen
on the Witwatersrand. S. Afr. J. Med. Sci. 19:65-74, 1954.
220. Duncan, M., P. Brookes, and A. Dipple. Metabolism and binding to cellular
macromolecules of a series of hydrocarbons by mouse embryo cells in
culture. Int. J. Cancer 4:813-819, 1969.
221. Dunn, J. E., and F. S. Brackett. Photosensitizing properties of some petro-
leum solvents. Indust. Med. 17:303-308, 1948.
222. Dunn, J. R., W. A. Waters, and I. M. Roitt. The retardation of benzaldehyde
autoxidation. Part VI. The actions of a number of polycyclic aromatic hy-
drocarbons of carcinogenic interest. J. Chem. Soc. 1954:580-586.
223. Eastcott, D. F. The epidemiology of lung cancer in New Zealand. Lancet 1:
37-39, 1956.
224. Eastman, J. W. The dependence of fluorescence on solvent and temperature.
Spectrochim. Acta 26A: 1545-1557, 1970.
225. Easty, G. C. Organ culture methods. Methods Cancer Res. 5:1-43, 1970.
226. Eberson, L., and K. Nyberg. Studies on electrolyte substitution reactions.
II. Anodic acetamidation, a novel anodic substitution. Tetrahedron Lett.
22:2389-2393, 1966.
227. Eckardt, R. E. Industrial Carcinogens. (Modern Monographs in Industrial
Medicine, 4.) New York: Grune & Stratton, Inc., 1959. 164 pp.
228. Ehling, U. H. The multiple loci method in chemical mutagenesis, pp. 156-
161. In F. Vogel and G. Rohrborn, Eds. Chemical Mutagenesis in Mammals
and Man. Berlin: Springer-Verlag, 1970.
229. Ehling, U. H., R. B. Cumming, and H. V. Mailing. Induction of dominant
lethal mutations by alkylating agents in male mice. Mutat. Res. 5:417-428,
1968.
230. Ejder, E. Methods of representing emission, excitation, and photoconduc-
tivity spectra. J. Opt. Soc. Amer. 59:223-224, 1969.
231. Epstein, J. H. Ultraviolet light carcinogenesis, pp. 215-236. In W. Montagna
and R. L. Dobson, Eds. Advances in Biology of Skin. Vol. VII. Carcino-
genesis. Proceedings of a Symposium on the Biology of Skin held at the
University of Oregon Medical School, 1965. Oxford: Pergamon Press, 1966.
-------
320 References
232. Epstein, S. S. A "catch-all" toxicological screen. Experientia 25:617, 1969.
233. Epstein, S. S. Photoactivation of polynuclear hydrocarbons. Arch. Environ.
Health 10:233-239, 1965.
234. Epstein, S. S., J. Andrea, S. Joshi, and N. Mantel. Hepatocarcinogenicity
of griseofulvin following parenteral administration to infant mice. Cancer
Res. 27:1900-1906, 1967.
235. Epstein, S. S., E. Arnold, E. Steinberg, D. Mackintosh, H. Shafner, and
Y. Bishop. Mutagenic and antifertility effects of TEPA and METEPA in
mice. Toxicol. Appl. Pharmacol. 17:23-40, 1970.
236. Epstein, S. S., W. Bass, E. Arnold, and Y. Bishop. Mutagenicity of trimethyl-
phosphatein mice. Science 168:584-586, 1970.
237. Epstein, S. S., W. Bass, E. Arnold, Y. Bishop, S. Joshi, and I. D. Adler. Sterility
and semisterility in male progeny of male mice treated with the chemical mutagen
TEPA. Toxicol. Appl. Pharmacol. 19:134-146, 1971.
238. Epstein, S. S., and M. Burroughs. Some factors influencing the photodynamic
response of Paramecium caudatum to 3,4-benzpyrene. Nature 193:337-338,
1962.
239. Epstein, S. S., M. Burroughs, and M. Small. The photodynamic effect of the
carcinogen, 3,4-benzpyrene, on Paramecium caudatum. Cancer Res. 23:35-
44, 1963.
240. Epstein, S. S., N. P. Buu-Hoi, and D. P. Hien. On the association between
photodynamic and enzyme-inducing activities on polycyclic compounds.
Cancer Res. 31:1087-1094, 1971.
241. Epstein, S. S., S. Joshi, J. Andrea, N. Mantel, E. Sawicki, T. Stanley, and
E. C. Tabor. Carcinogenicity of organic particulate pollutants in urban air
after administration of trace quantities to neonatal mice. Nature 212:1305-
1307, 1966.
242. Epstein, S. S., S. R. Joshi, E. Arnold, E. C. Page, and Y. Bishop. Abnormal
zygote development in mice after paternal exposure to a chemical mutagen.
Nature 225:1260-1261, 1970.
243. Epstein, S. S., and N. Mantel. Hepatocarcinogenicity of the herbicide
maleic hydrazide following parenteral administration to infant Swiss mice.
Int. J. Cancer 3:325-335, 1968.
244. Epstein, S. S., N. Mantel, and T. W. Stanley. Photodynamic assay of neutral
subfractions of organic extracts of particulate atmospheric pollutants.
Environ. Sci. Technol. 2:132-138, 1968.
245. Epstein, S. S., and H. Shafner. Chemical mutagens in the human environ-
ment. Nature 219:385-387, 1968.
246. Epstein, S. S., M. Small, H. L. Falk, and N. Mantel. On the association
between photodynamic and carcinogenic activities in polycyclic compounds.
Cancer Res. 24:855-862, 1964.
247. Epstein, S. S., M. Small, J. Koplan, N. Mantel, H. L. Falk, and E. Sawicki.
Photodynamic bioassay of polycyclic air pollutants. Arch. Environ. Health
7:531-537, 1963.
248. Epstein, S. S., M. Small, J. Koplan, N. Mantel, and S. H. Hunter. Photo-
dynamic bioassay of benzo[a] pyrene with Paramecium caudatum. J. Nat.
Cancer Inst. 31:163-168, 1963.
249. Epstein, S. S., M. Small, E. Sawicki, and H. L. Falk. Photodynamic bioas-
-------
References 321
say of polycyclic atmospheric pollutants. J. Air Pollut. Control Assoc.
15:174-176,1965.
250. Erickson, R. E., P. S. Bailey, and J. C. Davis, Jr. Structure of the mono-
Szonide of 9,10-dimethylanthracene. A transannular ozone adduct. Tetra-
hedron 18:388-395, 1962.
251. Eschenroeder, A., and J. R. Martinez. Mathematical Modeling of Photo-
chemical Smog. AIAA Paper 70-116. Presented to the American Institute
of Aeronautics and Astronautics 8th Aerospace Sciences Meeting, New
York, 1970. llpp.
252. Esmen, N. A., and M. Corn. Residence time of particles in urban air. Atmos.
Environ. 5:571-578, 1971.
253. Etienne, A., and A. Staehelin. Sur quelques a-azanthrancenes et quelques
benzacridines. Bull. Soc. Chim. France 21:748-755, 1954.
254. Evans, H. J. The Effects of External Agents on Differentiation in the Cul-
tivated Mushroom. Ph.D. dissertation, University College of Wales, 1955.
255. Falconer, D. S., B. M. Slizynski, and C. Auerbach. Genetical effects of
nitrogen mustard in the house mouse. J. Genet. 51:81-88, 1952.
256. Falk, H. L. Carcinogenesis, mutagenesis, and teratogenesis, pp. 155-164.
In Man's Health and the Environment-Some Research Needs. Report of
the Task Force on Research Planning in Environmental Health Science. U.S.
Department of Health, Education, and Welfare. Washington, D.C.: U.S.
Government Printing Office, 1970.
257. Falk, H. L., and P. Kotin. Pesticide synergists and their metabolites: Poten-
tial hazards. Ann. N.Y. Acad. Sci. 160:299-313, 1969.
258. Falk, H. L., P. Kotin, and I. Markul. The disappearance of carcinogens from
soot in human lungs. Cancer 11:482-489, 1958.
259. Falk, H. L., P. Kotin, and A. Mehler. Polycyclic hydrocarbons as carcino-
gens for man. Arch. Environ. Health 8:721-730, 1964.
260. Falk, H. L., P. Kotin, and A. Miller. Aromatic polycyclic hydrocarbons in
polluted air as indicators of carcinogenic hazards. Int. J. Air Pollut. 2:201-
209, 1960.
261. Falk, H. L., P. Kotin, and S. Thompson. Inhibition of Carcinogenesis. The
effect of polycyclic hydrocarbons and related compounds. Arch. Environ.
Health 9:169-179, 1964.
262. Falk, H. L., I. Markul, and P. Kotin. Aromatic hydrocarbons. IV. Their
fate following emission into atmosphere and experimental exposure to
washed air and synthetic smog. A.M.A. Arch. Ind. Health 13:13-17, 1956.
263. Falk, H. L., A. Miller, and P. Kotin. Elution of 3,4-benzpyrene and related
hydrocarbons from soots by plasma proteins. Science 127:474-475, 1958.
264. Faust, W. J., and M. J. Sterba. Minimizing Exhaust Emissions—A Realistic
Approach. ASTM Symposium, Toronto, Canada, June 1970. Philadelphia:
American Society for Testing Materials, 1971.
265. Federal Radiation Council. Radiation Exposure of Uranium Miners. Wash-
ington, D.C.: Federal Radiation Council, 1968. 31 pp.
266. Fell, H. B., and L. M. Rinaldini. The effects of vitamins A and C on cells
and tissues in culture, pp. 659-699. In E. N, Willmer, Ed. Cells and Tissues
in Culture. Methods, Biology, and Physiology. Vol. 1. London: Academic
Press Inc., 1965.
-------
322 References
267. Fieser, L. F., and E. B. Hershberg. The orientation of 3,4-benzpyrene in
substitution reactions. J. Amer. Chem. Soc. 61:1565-1574, 1939.
268. Fife, J. G. Carcinoma of the skin in machine tool setters. Brit. J. Ind. Med.
19:123-125, 1962.
269. Filley, G. F. Pulmonary Insufficiency and Respiratory Failure. Philadelphia:
Lea&Febiger, 1967. 162pp.
270. Findlay, B. F., and M. S. Hirt. An urban-induced meso-circulation. Atmos.
Environ. 3:537-542, 1969.
271. Findlay, F. D., C. J. Fortin, and D. R. Snelling. Deactivation of O2 ('Ag).
Chem. Phys. Lett. 3:204-206, 1969.
272. Findlay, G. M. Ultra-violet light and skin cancer. Lancet 2:1070-1073, 1928.
273. Fisherman, E. W. Does the allergic diathesis influence malignancy? J. Allergy
31:74-78, 1960.
274. Fjelde, A., and J. L. Turk. Induction of an immunological response in local
lymph nodes by chemical carcinogens. Nature 205:813-815, 1965.
275. Foote, C. S. Mechanisms of photosensitized oxidation. Science 162:963-
970, 1968.
276. Foote, C. S. Photosensitized oxygenations and the role of singlet oxygen.
Accounts Chem. Res. 1:104-110, 1968.
277. Foote, C. S., Y. C. Chang, and R. W. Denny. Chemistry of singlet oxygen.
X. Carotenoid quenching parallels biological protection. J. Amer. Chem.
Soc. 92:5216-5219, 1970.
278. Freeman, A. E., P. J. Price, R. J. Bryan, R. J. Gordon, R. V. Gilden, G. J.
Kelloff, and R. J. Huebner. Transformation of rat and hamster embryo
cells by extracts of city smog. Proc. Nat. Acad. Sci. U.S.A. 68:445-449, 1971.
279. Fried, J., and D. E. Schumm. One electron transfer oxidation of 7,12-
dimethylbenzfa] anthracene, a model for the metabolic activation of carci-
nogenic hydrocarbons. J. Amer. Chem. Soc. 89:5508-5509, 1967.
280. Friedel, R. A., and M. Orchin. Ultraviolet Spectra of Aromatic Compounds.
New York: John Wiley & Sons, Inc., 1961. 52pp.
281. Friedlander, S. K., and J. H. Seinfeld. A dynamic model of photochemical
smog. Environ. Sci. Technol. 3:1175-1181, 1969.
282. Fuks, N. A. The Mechanics of Aerosols. New York: The Macmillan Co.,
1964.408pp.
283. Furth, J., and E. Lorenz. Carcinogenesis by ionizing radiations, pp. 1145-
1201. In A. Hollaender, Ed. Radiation Biology. Vol. 1. Part 2. New York:
McGraw-Hill Book Co., Inc., 1954.
284. Gabridge, M. G., A. Denunzio, and M. S. Legator. Cycasin: Detection of
associated mutagenic activity in vivo. Science 163:689-691, 1969.
285. Gabridge, M. G., A. Denunzio, and M. S. Legator. Microbial mutagenicity
of streptozotocin in animal-mediated assays. Nature 221:68-70, 1969.
286. Gabridge, M. G., and M. S. Legator. A host-mediated microbial assay for
the detection of mutagenic compounds. Proc. Soc. Exp. Biol. 130:831-
834, 1969.
287. Gabridge, M. G., E. J. Oswald, and M. S. Legator. The role of selection in
the host-mediated assay for mutagenicity. Mutat. Res. 7:117-119, 1969.
288. Galuskinova, V. 3,4-Benzpyrene determination in the smoky atmosphere of
social meeting rooms and restaurants. A contribution to the problem of the
noxiousness of so-called passive smoking. Neoplasma 11:465-468, 1964
-------
References 323
289. Gardner, M. B. Biological effects of urban air pollution. III. Lung tumors
in mice. Arch. Environ. Health 12:305-313, 1966.
290. Geacintov, N., G. Oster, and T. Cassen. Polymeric matrices for organic phos-
phors. J. Opt. Soc. Amer. 58:1217-1229, 1968. Erratum: ibid. 59:367, 1969.
291. Gelboin, H. V. A microsome-dependent binding of benzo[a]pyrene to DNA.
Cancer Res. 29:1272-1276, 1969.
292. Gelboin, H. V. Carcinogens, enzyme induction, and gene action. Adv. Can-
cer Res. 10:1-81, 1967.
293. Gelboin, H. V., and F. J. Wiebel. Studies on the mechanism of aryl hydro-
carbon hydroxylase induction and its role in cytotoxicity and tumorenicity.
Ann. N.Y. Acad. Sci. 171:529-549. 1971.
294. Gelboin, H. V., F. Wiebel, and L. Diamond. Dimethylbenzanthracene tumor-
igenesis and aryl hydrocarbon hydroxylase in mouse skin: Inhibition by
7,8-benzoflavone. Science 170:169-171, 1970.
295. Generoso, W. M. Chemical induction of dominant lethals in female mice.
Genetics 61:461-470, 1969.
296. Gilbert, J. A. S., and A. J. Lindsey. Polycyclic hydrocarbons in tobacco
smoke: Pipe smoking experiments. Brit. J. Cancer 10:646-648, 1956.
297. Gillette, J. R. Factors that affect the stimulation of the microsomal drug
enzymes induced by foreign compounds. Adv. Enzyme Regul. 1:215-223,
1963.
298. Gilliland, M. R., A. J. Howard, and D. Hamer. Polycyclic hydrocarbons
in crude peat wax. Chem. Ind. 1960:1357-1358.
299. Gleason, W. S., A. D. Broadbent, E. Whittle, and J. N. Pitts, Jr. Singlet
oxygen in the environmental sciences. IV. Kinetics of the reaction of oxy-
gen ('Ag) with tetramethylethylene and 2,5-dimethylfuran in the gas phase.
J. Amer. Chem. Soc. 92:2068-2075, 1970.
300. Goetz, A., H. J. R. Stevenson, and O. Preining. The design and performance
of the aerosol spectrometer. J. Air Pollut. Control Assoc. 10:378-383,
414,416, 1970.
301. Gold, P., M. Gold, and S. O. Freedman. Cellular location of carcinoembryonic
antigens of the human digestive system. Cancer Res. 28:1331-1334, 1968.
302. Goldsmith, E. I., and J. Moor-Jankowski. Experimental medicine and surgery
in primates. Opening remarks. Ann. N.Y. Acad. Sci. 162:5-6, 1969.
303. Goldsmith, J. R. Effects of air pollution on human health, pp. 547-615.
In A. C. Stern, Ed. Air Pollution. Vol. 1. Air Pollution and Its Effects.
(2nd ed.) New York: Academic Press Inc., 1968.
304. Gollnick, K., and G. O. Schenck. Oxygen as a dienophile, pp. 255-344.
In J. Hamer, Ed. 1,4-Cycloaddition Reactions. The Diels-Alder Reaction
in Heterocyclic Syntheses. New York: Academic Press Inc., 1967.
305. Good, R. A., and J. Finstad. Essential relationship between the lymphoid
system, immunity, and malignancy. Nat. Cancer Inst. Monogr. 31:41-58, 1969.
306. Gorelova, N. D., and P. P. Dikun. 3,4-Benzopyrene content of sausage and
smoked fish prepared using fuel gas and coke. Gig. Sanit. 30(7): 120-122,
1965. (in Russian)
307. Goshman, L. M., and C. Heidelberger. Binding of tritium-labeled polycyclic
hydrocarbons to DNA of mouse skin. Cancer Res. 27:1678-1688, 1967.
308. Gottschalk, R. G. Quantitative studies on tumor production in mice by
benzpyrene. Proc. Soc. Exp. Biol. Med. 50:369-373, 1942.
-------
324 References
309. Gra'f, W., and H. Diehl. Uber den naturbedingten Normalpegel kanzerogener
polycyclischer Aromate und seine Ursache. Arch. Hyg. Bakt. 150:49-59,
1966. (summary in English)
310. Graham, S., M. L. Levin, A. M. Lilienfeld, and P. Sheebe. Ethnic derivation
as related to cancer at various sites. Cancer 16:13-27, 1963.
311. Green, B. Influence of pH and metal ions on the fluorescence of polycyclic
hydrocarbons in aqueous DNA solution. Eur. J. Biochem. 14:567-574,
1970.
312. Grimmer, G. Cancerogene Kohlenwasserstoffe in der Umgebung des Men-
schen. Erdoel Kohle 19:578-583, 1966.
313. Grimmer, G., and A. Hildebrandt. Content of polycyclic hydrocarbons in
crude vegetable oils. Chem. Ind. 1967:2000-2002.
314. Griswold, M. H., C. S. Wilder, S. J. Cutler, and E. S. Pollack. Cancer in Con-
necticut, 1935-1951. Hartford: Connecticut State Department of Health,
1955. 141pp.
315. Grobstein, C. Differentiation: Environmental factors, chemical and cellular,
pp. 463-488. In E. N. Willmer, Ed. Cells and Tissues in Culture. Methods,
Biology, and Physiology. Vol. I. London: Academic Press Inc., 1965.
316. Gross, G. P. First Annual Report on Gasoline Composition and Vehicle
Exhaust Gas Polynuclear Aromatic Content. Durham, N.C.: U.S. Depart-
ment of Health, Education, and Welfare, 1970.
317. Gross, M. A., O. G. Fitzhugh, and N. Mantel. Evaluation of safety for food
additives: An illustration involving the influence of methyl salicylate on rat
reproduction. Biometrics 26:181-194, 1970.
318. Gross, P., E. A. Pfitzer, J. Watson, R. T. P. deTreville, M. Kaschak, E. B.
Tolker, and M. A. Babyak. Experimental carcinogenesis. Bronchial intra-
mural adenocarcinomas in rats from x-ray irradiation of the chest. Cancer
23:1046-1060, 1969.
319. Grovenstein, E., Jr., and A. J. Mosher. Reaction of atomic oxygen with
aromatic hydrocarbons. J. Amer. Chem. Soc. 92:3810-3812, 1970.
320. Grover, P. L., and P. Sims. Enzyme-catalysed reactions of polycyclic hydro-
carbons with deoxyribonucleic acid and protein in vitro. Biochem. J.
110:159-160, 1968.
321. Grover, P. L., and P. Suns. Interactions of K-region epoxides of phenanthrene
and dibenz[a,h] anthracene with nucleic acids and histone. Biochem. Phar-
macol. 19:2251-2259, 1970.
322. Grover, P. L., P. Sims, E. Huberman, H. Marquardt, T. Kuroki, and C. Hei-
delberger. In vitro transformation of rodent cells by K-region derivatives of
polycyclic hydrocarbons. Proc. Nat. Acad. Sci. U.S.A. 68:1098-1101, 1971.
323. Guddal, E. Isolation of polynuclear aromatic hydrocarbons from the roots of
Chrysanthemum vulgare Bernh. Acta Chem. Scand. 13:834-835, 1959.
324. Gunther, F. A., and F. Buzzetti. Occurrence, isolation, and identification
of polynuclear hydrocarbons as residues. Residue Rev. 9:90-113, 1965.
325. Gunther, F. A., F. Buzzetti, and W. E. Westlake. Residue behavior of poly-
nuclear hydrocarbons on and in oranges. Residue Rev. 17:81-104, 1967.
326. Haagensen, C. D. Occupational neoplastic disease. Amer. J. Cancer 15:641-
703, 1931.
327. Hadidian, Z., T. N. Fredrickson, E. K. Weisburger, J. H. Weisburger, R. M.
-------
References 325
Glass, and N. Mantel. Tests for chemical carcinogens. Report on the activity.
of derivatives of aromatic amines, nitrosamines, quinolines, nitroalkanes,
amides, epoxides, aziridines, and purine antimetabolites. J. Nat. Cancer
Inst. 41:985-1036, 1968.
328. Haenni, E. O., J. W. Howard, and F. L. Joe, Jr. Dimethyl sulfoxide: A superior
analytical extraction solvent for polynuclear hydrocarbons and for some
highly chlorinated hydrocarbons. J. Assoc. Offic. Agr. Chem. 45:67-70,
1962.
329. Haenszel, W. Cancer mortality among the foreign-born in the United States.
J. Nat. Cancer Inst. 26:37-132, 1961.
330. Haenszel, W., D. B. Loveland, and M. G. Sirken. Lung-cancer mortality
as related to residence arid smoking histories. I. White males. J. Nat. Cancer
Inst. 28:947-1001, 1962.
331. Haenszel, W., S. C. Marcus, and E. G. Zimmerer. Cancer Morbidity in Urban
and Rural Iowa. Public Health Monograph 37; Public Health Service Publi-
cation 462. Washington, D.C.: U.S. Government Printing Office, 1956. 85 pp.
332. Haenszel, W., and K. E. Traeuber. Lung-cancer mortality as related to resi-
dence and smoking histories. II. White females. J. Nat. Cancer Inst. 32:803-
838, 1964.
333. Hagstrom, R. M., H. A. Sprague, and E. Landau. The Nashville air pollution
study. VII. Mortality from cancer in relation to air pollution. Arch. Environ.
Health 15:237-248, 1967.
334. Hakama, M., and E. A. Saxen. Cereal consumption and gastric cancer. Int.
J. Cancer 2:265-268, 1967.
335. Hall, W. K. The formation of cation radicals on the surface of silica-alumina
catalysts. J. Catalysis 1:53-61, 1962.
336. Hamburg, F. C. Economically feasible alternatives to open burning in railroad
freight car dismantling. J. Air Pollut. Control Assoc. 19:477-483, 1969.
337. Hammond, E. C. Quantitative relationship between cigarette smoking and
death rates. Nat. Cancer Inst. Monogr. 28:3-8, 1968.
338. Hammond, E. C. Smoking in relation to mortality and morbidity. Findings
in first thirty-four months of follow-up in a prospective study started in
1959. J. Nat. Cancer Inst. 32:1161-1188, 1964.
339. Hammond, E. C. Smoking in relation to the death rates of one million men
and women. Nat. Cancer Inst. Monogr. 19:127-204, 1966.
340. Hammond, E. C., O. Auerbach, D. Kirman, and L. Garfinkel. Effects of
cigarette smoking on dogs. I. Design of experiment, mortality, and findings in
lung parenchyma. Arch. Environ. Health 21:740-753, 1970.
341. Hammond, E. C., and D. Horn. Smoking and death rates—report on 44
months of follow-up of 187,783 men. Part I. Total mortality. Part II.
Death rates by cause. J.A.M.A. 166:1159-1172, 1294-1308, 1958.
342. Hangebrauck, R. P., R. P. Lauch, and J. E. Meeker. Emissions of polynu-
clear hydrocarbons from automobiles and trucks. Amer. Ind. Hyg. Assoc.
J. 27:47-56, 1966.
343. Hangebrauck, R. P., D. J. von Lehmden, and J. E. Meeker. Sources of Poly-
nuclear Hydrocarbons in the Atmosphere. Public Health Service Publication
999-AP-33. Cincinnati: U.S. Department of Health, Education, and Welfare,
1967. 48 pp.
-------
326 References
344. Harris, R. J. C., and G. Negroni. Production of lung carcinomas in C57BL
mice exposed to a cigarette smoke and air mixture. Brit. Med. J. 4:637-641,
1967.
345. Harting, F. H., and W. Hesse. Der Lungenkrebs, die Bergkrankheit in den
Schneeberger Gruben. Viertejjahresschr. Gerichtl. Med. Oeff. Sanitaetsw.
31:102-132,313-337, 1879.
346. Hartwell, J. L., and P. Shubik. Survey of Compounds Which Have Been
Treated for Carcinogenic Activity. Public Health Service Publication 149.
(2nd ed.) Washington, D.C.: U.S. Government Printing Office, 1951. 583 pp.
347. Hatch, T. F. Significant dimensions of the dose-response relationship. Arch.
Environ. Health 16:571-578, 1968.
348. Hecker, E. Cocarcinogenic principles from the seed oil of Croton tiglium
and from other euphorbiaceae. Cancer Res. 28:2338-2348, 1968.
349. Heidelberger, C. Chemical carcinogenesis, chemotherapy: Cancer's continu-
ing core challenges. G. H. A. Clowes Memorial Lecture. Cancer Res. 30:1549-
1569, 1970.
350. Heidelberger, C. Studies on the cellular and molecular mechanisms of hydro-
carbon carcinogenesis. Eur. J. Cancer 6:161-172, 1970.
351. Heidelberger, C., M. E. Baumann, L. Griesbach, A. Ghobar, and T. M. Vaughan.
The carcinogenic activities of various derivatives of dibenzanthracene. Cancer
Res. 22:78-83, 1962.
352. Heidelberger, C., H. I. Hadler, and G. Wolf. The metabolic degradation in the
mouse of l,2,5,6-dibenzanthracene-9,10-C14. III. Some quinone metabo-
lites retaining the intact ring system. J. Amer. Chem. Soc. 75:1303-1308,
1953.
353. Heidelberger, C., and P. T. lype. Malignant transformation in vitro by carcino-
genic hydrocarbons. Science 155:214-217, 1967.
354. Heidelberger, C., and H. B. Jones. The distribution of radioactivity in the
mouse following administration of dibenzanthracene labeled in the 9 and
10 positions with carbon 14. Cancer 1:252-260, 1948.
355. Heidelberger, C., M. R. Kirk, and M. S. Perkins. The metabolic degradation
in the mouse of dibenzanthracene labeled in the 9 and 10 positions with
carbon 14. Cancer 1:261-275, 1948.
356. Heidelberger, C., and S. M. Weiss. The distribution of radioactivity in mice
following administration of 3,4-benzpyrene-5-C14 and 1,2,5,6-dibenzanthra-
cene-9,10-C14. Cancer Res. 11:885-891, 1951.
357. Heller, I. Occupational cancers. J. Ind. Hyg. 12:169-197, 1930.
358. Hendricks, N. V., C. M. Berry, J. G. Lione, and J. J. Thorpe. Cancer of the
scrotum in wax pressmen. I. Epidemiology. A.M.A. Arch. Ind. Health 19:
524-529, 1959.
359. Henry, S. A. Cancer of the Scrotum in Relation to Occupation. London:
Oxford University Press, 1946. 120 pp.
360. Henry, S. A. Occupational cutaneous cancer attributable to certain chemi-
cals in industry. Brit. Med. Bull. 4:389-401, 1947.
361. Hercules, D. M. Chemiluminescence from electron-transfer reactions. Ac-
counts Chem. Res. 2:301-307, 1969.
362. Hercules, D. M. Electron spectroscopy. Anal. Chem. 42:20A-28A, 30A,
32A, 34A-35A, 38A-40A, Jan. 1970.
-------
References 327
363. Hercules, D. M. Fluorescence and Phosphorescence Analysis. Principles and
Applications. New York: John Wiley & Sons, Inc., 1966. 258 pp.
364. Hercules, D. M. Organic electro-luminescence. In A. Weissberger and B. W.
Rossiter, Eds. Physical Methods of Organic Chemistry. Part II. Electro-
chemical Methods. (4th ed.) New York: Wiley-Interscience. (to be published)
365. Hermann, T. S. Identification of trace amounts of organophosphorous pesti-
cides by frustrated multiple internal reflectance spectroscopy. Appl. Spec-
trosc. 19:10-14, 1965.
366. Herrold, K. M., and L. J. Dunham. Induction of carcinoma and papilloma
of the tracheobronchial mucosa of the Syrian hamster by intratracheal in-
stillation of benzo[a]pyrene. J. Nat. Cancer Inst. 28:467-491, 1962.
367. Herron, J. T., and R. E. Huie. Reactions of O2 'Ag with olefins and their
significance in air pollution. Environ. Sci." Technol. 4:685-686, 1970.
368. Hertwig, P. Vererbbare Semisterilitat bei Mausen nach RSntgenbestrahlung,
verursacht durch reziproke Chromosomentranslokationen. Z. Indukt. Abstamm.
Verebungsl. 79:1-27, 1940.
369. Hesse, G., I. Daniel, and G. Wohlleben. Aluminiumoxyde fur die chromato-
graphische Analyse und Versuche zu ihrer Standardisierung. Angew. Chem.
64:103-107, 1952.
370. Heston, W. E., and M. A. Schneiderman. Analysis of dose-response in rela-
tion to mechanism of pulmonary tumor induction in mice. Science 117:109-
111,1953.
371. Hidy, G. M. The dynamics of aerosols in the lower troposphere. In Assess-
ment of Airborne Particles. Proceedings of the 3rd University of Rochester
Conference on Environmental Toxicology, June 1, 1970. Springfield, 111.:
Charles C Thomas, 1971.
372. Hidy, G. M., and J. R. Brock. An assessment of the global sources of tropo-
spheric aerosols, p. 113. (abstract) In The 2nd International Clean Air Con-
gress; Proceedings Digest. Washington, D.C., Dec. 6-11, 1970.
373. Hidy, G. M., and J. R. Brock. The Dynamics of Aerocolloidal Systems. New
York: Pergamon Press, 1970. 371 pp.
374. Hidy, G. M., and S. K. Friedlander. The nature of the Los Angeles aerosol,
p. 60. (abstract) In The 2nd International Clean Air Congress; Proceedings
Digest. Washington, D.C., Dec. 6-11, 1970.
375. Hilding, A. C. Ciliary streaming in the bronchial tree and the time element
in carcinogenesis. New Eng. J. Med. 256:634-640, 1957.
376. Hilding, A. C. On cigarette smoking, bronchial carcinoma and ciliary action.
I. Smoking habits and measurement of smoke intake. New Eng. J. Med.
254: 775-781, 1956.
377. Hinds, W. T. Diffusion over coastal mountains of Southern California. Atmos.
Environ. 4:107-124, 1970.
378. Hirayama, K. Handbook of Ultraviolet and Visible Absorption Spectra of
Organic Compounds. New York: Plenum Press Data Division, 1967. 642 pp.
379. Hitosugi, M. Epidemiological study of lung cancer with special reference to
the effect of air pollution and smoking habits. Inst. Public Health Bull. 17:
237-256, 1968.
380. Hochrainer, D., and P. M. Brown. Sizing of aerosol particles by centrifuga-
tion. Environ. Sci. Technol. 3:830-835, 1969.
-------
328 References
381. Hodkinson, J. R. The effect of particle shape on measures for the size and
concentration of suspended and settled particles, Amer. Ind. Hyg. Assoc.
1.26:64-71, 1965.
382. Hoffman, C. S., Jr., R. L. Willis, G. H. Patterson, and E. S. Jacobs. Polynu-
clear Aromatic Hydrocarbon Emission from Vehicles. Presented to the
160th National Meeting of the American Chemical Society, Los Angeles,
California, March 1971.
383. Hoffman, E. F., and A. G. Gilliam. Lung cancer mortality. Geographic dis-
tribution in the United States for 1948-1949. Public Health Rep. 69:1033-
1042, 1954.
384. Hoffmann, D., E. Theisz, and E. L. Wynder. Studies on the carcinogenicity
of gasoline exhaust. J. Air Pollut. Control Assoc. 15:162-165, 1965.
385. Hoffmann, D., and G. Rathkamp. Quantitative determination of 1-alkylin-
doles in cigaret smoke. Anal. Chem. 42:366-370, 1970.
386. Hoffmann, D., G. Rathkamp, and S. Nesnow. Quantitative determination of
9-methylcarbazoles in cigarette smoke. Anal. Chem. 41:1256-1259, 1969.
387. Hoffmann, D., and E. L. Wynder. A study of air pollution carcinogenesis. II.
The isolation and identification of polynuclear aromatic hydrocarbons from
gasoline engine exhaust condensate. Cancer 15:93-102, 1962.
388. Hoffmann, D., and E. L. Wynder. Chemical analysis and carcinogenic bioas-
says of organic particulate pollutants, pp. 187-247. In A. C. Stern, Ed. Air
Pollution. Vol. 2. Analysis, Monitoring, and Surveying. (2nd ed.) New York:
Academic Press Inc., 1968.
389. Hoffmann, D., and E. L. Wynder. Short-term determination of carcinogenic
aromatic hydrocarbons. Anal. Chem. 32:295-296, 1960.
390. Hoffmann, D., and E. L. Wynder. Studies on gasoline engine exhaust. J.
Air Pollut. Control Assoc. 13:322-327, 1963.
391. Holzworth, G. C. Atmospheric contaminants at remote California coastal
sites. J. Meteorol. 16:68-79, 1969.
392. Holzworth, G. C. Large scale weather influences in community air pollution
potential in the United States. J. Air Pollut. Control Assoc. 19:248-254,
1969.
393. Homburger, F. Chemical carcinogenesis in the Syrian golden hamster. A
review. Cancer 23:313-338, 1959.
394. Homburger, F., and S. S. Hsueh. Rapid induction of subcutaneous fibro-
sarcoma by 7,12-dimethylbenz(a)anthracene in an inbred line of Syrian ham-
sters. Cancer Res. 30:1449-1452, 1970.
395. Horrocks, D. L. Effect of solvent on excimer fluorescence yields. J. Chem.
Phys. 51:5443-5448, 1969.
396. Horton, A. W., D. T. Denman, and R. P. Trosset. Carcinogenesis of the skin.
II. The accelerating properties of aliphatic and related hydrocarbons. Can-
cer Res. 17:758-766, 1957.
397. Howard, J. W., and T. Fazio. A review of polycyclic aromatic hydrocarbons in
foods. Ind. Med. Surg. 39:435-440, 1970.
398. Huberman, E., and L. Sachs. Cell susceptibility to transformation and cyto-
toxicity by the carcinogenic hydrocarbon benzo[ a] pyrene. Proc. Nat. Acad
Sci. U.S.A. 56:1123-1129, 1966.
-------
References 329
399. Huebner, R. J., and G. J. Todaro. Oncogenes of RNA tumor viruses as
determinants of- cancer. Proc. Nat. Acad. Sci. U.S.A. 64:1087-1094, 1969.
400. Hueper, W. C. Carcinogens in the human environment. Arch. Path. 71:237-
267,355-380, 1961.
401. Hueper, W. C. Chemically induced skin cancers in man. Nat. Cancer Inst.
Monogr. 10:377-391, 1963.
402. Hueper, W. C. Occupational Tumors and Allied Diseases. Springfield, 111.:
Charles C Thomas, 1942. 896 pp.
403. Hueper, W. C., and W. D. Conway. Chemical Carcinogenesis and Cancers.
Springfield, 111.: Charles C Thomas, 1964. 744 pp.
404. Hueper, W. C., P. Kotin, E. C. Tabor, W. W. Payne, H. Falk, and E. Sawicki.
Carcinogenic bioassays on air pollutants. Arch. Path. 74:89-116, 1962.
405. Huggins, C., L. Grand, and R. Fukunishi. Aromatic influences on the yields
of mammary cancers following administration of 7,12-dimethylbenz(a)-
anthracene. Proc. Nat. Acad. Sci. U.S.A. 51:737-742, 1964.
406. Huggins, C., L. C. Grand, and F. P. Brillantes. Mammary cancer induced by
a single feeding of polynuclear hydrocarbons, and its suppression. Nature
189:204-207, 1961.
407. Huggins, C., and N. C. Yang. Induction and extinction of mammary can-
cer. Science 137:257-262, 1962.
408. HUnigen, E., N. Jaskulla, and K. Wettig. Die Herabsetzung Krebsfbrdernder
Schadstoffe in Ottomotoren-Abgasen durch Kraftstoffzusatz und Schmier-
stoffauswahl, pp. 191-193. Proceedings of the International Clean Air Con-
gress. Part 1. London, 1966.
409. Husar, R., N. Barsic, M. Tomaides, B. Y. H. Liu, and K. T. Whitby. Five
spectra and miscellaneous experiments, pp. 3-133. In K. T. Whitby, Ed.
Aerosol Measurements in Los Angeles. Part 3. Minnesota Aerosol Analyzing
System and Miscellaneous Experiments Particle Laboratory Publication 141.
Minneapolis: University of Minnesota, 1970.
410. Imagawa, D. T., M. Yoshimori, and J. M. Adams. The death rate in mice with
pulmonary tumors induced with urethan and influenza virus. Proc. Amer.
Assoc. Cancer Res. 2:217, 1957. (abstract)
411. Inbar, M., and L. Sachs. Structural difference in sites on the surface mem-
brane of normal and transformed cells. Nature 223:710-712, 1969.
412. Innes, J. R. M., B. M. Ulland, M. G. Valerio, L. Petrucelli, L. Fishbein,
E. R. Hart, A. J. Pallota, R. R. Bates, H. L. Falk, J. J. Gart, M. Klein, I.
Mitchell, and J. Peters. Bioassay of pesticides and industrial chemicals for
tumorigenicity in mice. A preliminary note. J. Nat. Cancer Inst. 42:1101-
1114, 1969.
413. Inscoe, M. N. Photochemical changes in thin-layer chromatograms of poly-
cyclic, aromatic hydrocarbons. Anal. Chem. 36:2505-2606, 1964.
414. Isbell, A. F., Jr., and D. T. Sawyer. Gas-solid chromatography with salt-
modified porous silica beads. Anal. Chem. 41:1381-1387, 1969.
415. Ishikawa, S., D. H. Bowden, V. Fisher, and J. P. Wyatt. The "emphysema
profile" in two midwestern cities in North America. Arch. Environ. Health
18:660-666, 1969.
416. Jerina, D. M., J. W. Daly, B. Witkop, P. Zaltzman-Nirenberg, and S. Uden-
-------
330 References
friend. 1,2-Naphthalene oxide as an intermediate in the microsomal hy-
droxylation of naphthalene. Biochemistry 9:147-155, 1970.
417. Johnson, B. H., and T. Aczel. Analysis of complex mixtures of aromatic
compounds by high-resolution mass spectrometry at low-ionizing voltages.
Anal. Chem. 39:682-685, 1967.
418. Jones, P. R., and S. Siegel. Temperature effects on the phosphorescence of
aromatic hydrocarbons in poly(methylmethacrylate). J. Chem. Phys. 50:
1134-1140, 1969.
419. Jull, J. W. The induction of tumours of the bladder epithelium in mice by
the direct application of a carcinogen. Brit. J. Cancer 5:328-330, 1951.
420. Junge, C. E. Atmospheric radioactivity, pp. 209-288. In Air Chemistry and
Radioactivity. International Geophysics Series. Vol. 4. New York: Academic
Press Inc., 1963.
421. Junge, C. E. Comments on "Concentration and size distribution measure-
ments of atmospheric aerosols and a test of the theory of self-preserving
size distributions." J. Atmos. Sci. 26:603-608, 1969.
422. Kahn, H. A. The Dorn study of smoking and mortality among U.S. Veterans:
Report on eight and one-half years of observation. Nat. Cancer Inst. Monogr.
19:1-125, 1966.
423. Kao, F-T., and T. T. Puck. Genetics of somatic mammalian cells. VII. Induc-
tion and isolation of nutritional mutants in Chinese hamster cells. Proc. Nat.
Acad. Sci. U.S.A. 60:1275-1281, 1968.
424. Kao, F-T., and T. T. Puck. Genetics of somatic mammalian cells. IX. Quanti-
tation of mutagenesis by physical and chemical agents. J. Cell Physiol.
74:245-258, 1969.
425. Kao, S.-K, and D. Henderson. Large-scale dispersion of clusters of particles
in various flow patterns. J. Geophys. Res. 75:3104-3113, 1970.
426. Kato, R. Chromosome breakage induced by a carcinogenic hydrocarbon in
Chinese hamster cells and human leukocytes in vitro. Hereditas 59:120-141,
1968.
427. Kato, R., M. Bruze, and Y. Tegner. Chromosome breakage induced in vivo
by a carcinogenic hydrocarbon in bone marrow cells of the Chinese hamster.
Hereditas 61:1-8, 1969.
428. Katz, M. Measurement of Air Pollutants. Guide to the Selection of Methods.
Geneva: World Health Organization, 1969. 123 pp.
429. Kawai, M., H. Amamoto, and K. Harada. Epidemiologic study of occupational
lung cancer. Arch. Environ. Health 14:859-864, 1967.
430. Kawai, M., T. Matsuyama, H. Amamoto, and M. Nakamura. A study of occu-
pational lung cancers of the generator gas plant workers in Yawata Iron and
Steel Works, Japan. J. Labour Hyg. Iron Steel Ind. 10:5-9, 1961. (in Japanese)
431. Kearns, D. R., A. U. Khan, C. K. Duncan, and A. H. Maki. Detection of the
naphthalene-photosensitized generation of singlet ('A g) oxygen by para-
magnetic resonance spectroscopy. J. Amer. Chem. Soc. 91:1039-1040, 1969.
432. Kelly, M. G., and R. W. O'Gara. Induction of tumors in newborn mice with
dibenz[a,h] anthracene and 3-methylcholanthrene. J. Nat. Cancer Inst 26'
651-679, 1961.
433. Kelly, M. G., R. W. O'Gara, R. H. Adamson, K. Gadekar, C. C. Botkin, W. H.
-------
References 331
Reese, Jr., and W. T. Kerber. Induction of hepatic cell carcinomas in monkeys
with ,/V-nitrosodiethylamine. J. Nat. Cancer Inst. 36:323-351, 1966.
434. Kennaway, E. L., and N. M. Kennaway. A further study of the incidence of
cancer of the lung and larynx. Brit. J. Cancer 1:260-298, 1947.
435. Kent, S. P. Spontaneous and induced malignant neoplasms in monkeys. Ann.
N.Y. Acad. Sci. 85:819-827, 1960.
436. Kirkland, J. J. High-speed liquid chromatography with controlled surface po-
rosity support. J. Chromatogr. Sci. 7:7-12, 1969.
437. Klein, G. Experimental studies hi tumor immunology. Fed. Proc. 28:1739-
1753, 1969.
438. Klein, G. Tumor-specific transplantation antigens. G. H. A. Clowes Memorial
Lecture. Cancer Res. 28:625-635, 1968.
439. Klein, M. Development of hepatomas in inbred albino mice following treat-
ment with 20-methylcholanthrene. Cancer Res. 19:1109-1113, 1959.
440. Klein, M. Induction of skin tumors in the mouse with minute doses of 9, 10-
dimethyl-l,2-benzanthracene alone or with croton oil. Cancer Res. 16:123-
127, 1956.
441. Klein, M. Influence of low dose of 2-acetylaminofluorene on liver tumori-
genesis in mice. Proc. Soc. Exp. Biol. Med. 101:637-638, 1959.
442. Kohn-Speyer, A. C. Effect of ultra-violet radiation on the incidence of tar
cancer in mice. Lancet 2:1305-1306, 1929.
443. Koller, P. C. Segmental interchange in mice. Genetics 29:247-263, 1944.
444. Koller, P. C., and C. A. Auerbach. Chromosome breakage and sterility in the
mouse. Nature 148:501-502, 1941. (letter to the editor)
445. Kooyman, E. C., and E. Farenhorst. The relative reactivities of polycyclic aro-
matics towards trichloromethyl radicals. Trans. Faraday Soc. 49:58-67,
1953.
446. Kortum, G., and W. Braun. Photochemische Reaktionen des Anthracens in
adsorbiertem Zustand. Justus Liebigs Ann. Chem. 632:104-115, 1960.
447. Kotin, P. The influence of pathogenic viruses on cancers induced by inhala-
tion. Can. Cancer Conf. 6:475-498, 1966.
448. Kotin, P. The role of atmospheric pollution in the pathogenesis of pulmonary
cancer. A review. Cancer Res. 16:375-393, 1956.
449. Kotin, P., and H. L. Falk. Atmospheric factors in pathogenesis of lung cancer.
Adv. Cancer Res. 7:475-514, 1963.
450. Kotin, P., and H. L. Falk. The role and action of environmental agents in the
pathogenesis of lung cancer. I. Air pollutants. Cancer 12:147-163, 1959.
451. Kotin, P., and H. L. Falk. The role and action of environmental agents in the
pathogenesis of lung cancer. II. Cigaret smoke. Cancer 13:250-262, 1960.
452. Kotin, P., H. L. Falk, and R. Busser. Distribution, retention, and elimination
of CI4-3,4-benzpyrene after administration to mice and rats. J. Nat. Cancer
Inst. 23:541-555, 1959.
453. Kotin, P., H. L. Falk, P. Mader, and M. Thomas. Aromatic hydrocarbons. I.
Presence in the Los Angeles atmosphere and the carcinogenicity of exhaust
extracts. A.M.A. Arch. Ind. Hyg. Occup. Med. 9:153-163, 1964.
454. Kotin, P., H. L. Falk, and C. J. McCammon. The experimental induction of
pulmonary tumors and changes in the respiratory epithelium in C57BL mice
-------
332 References
following their exposure to an atmosphere of ozonized gasoline. Cancer
11:473-481, 1958.
455. Kotin, P., H. L. Falk, and M. Thomas. Aromatic hydrocarbons. II. Presence
in the particulate phase of gasoline-engine exhausts and the carcinogenicity
of exhaust extracts. A.M.A. Arch. Ind. Hyg. Occup. Med. 9:164-177, 1954.
456. Kotin, P., and D. V. Wiseley. Production of lung cancer in mice by inhalation
exposure to influenza virus and aerosols of hydrocarbons. Progr. Exp. Tumor
Res. 3:186-215, 1963.
457. Kotrappa, P. Shape factors for aerosols of coal, UO2 and Th2 in the respirable
size range. In Assessment of Airborne Particles. Proceedings of the 3rd Univer-
sity of Rochester Conference on Environmental Toxicology, June 1970.
Springfield, 111.: Charles C Thomas, 1971.
458. Kracht, H. J., U. E. Klein, and M. Baghirzade. Erfahrungen mit dem
TalgdrUsenschwundtest. Verh. Deutsch. Ges. Path. 45:170-174, 1961.
459. Kreyberg, L. 3:4-Benzopyrene in industrial air pollution. Some reflexions.
Brit. J. Cancer 13:618-622, 1959.
460. Kreyberg, L. Occurrence and aetiology of lung cancer in Norway in the light
of pathological anatomy. Brit. J. Prev. Soc. Med. 10:145-158, 1956.
461. Kreyberg, L. The occurrence of lung cancer in Norway. Brit. J. Cancer
8:209-214, 1954.
462. Kummler, R. H., and M. H. Bortner. Production of Oj ('A g) by energy trans-
fer from excited benzaldehyde. Environ. Sci. Technol. 3:944-946, 1969.
463. Kuratsune, M. Benzo[a]pyrene content of certain pyrogenic materials. J.
Nat. Cancer Inst. 16:1485-1496, 1956.
464. Kuratsune, M., and T. Hirohata. Decomposition of polycyclic aromatic hydro-
carbons under laboratory illuminations. Nat. Cancer Inst. Monogr. 9:117-125,
1962.
465. Kuratsune, M., and W. C. Hueper. Polycyclic aromatic hydrocarbons in roasted
coffee. J. Nat. Cancer Inst. 24:463-469, 1960.
466. Kuroda, S. Occupational pulmonary cancer of generator gas workers. Ind.
Med. Surg. 6:304-306, 1937.
467. Kuroki, T., and H. Sato. Transformation and neoplastic development in vitro
of hamster embryonic cells by 4-nitroquinoline-l-oxide and its derivatives.
J.Nat. Cancer Inst. 41:53-71, 1968.
468. Kuschner, M. The causes of lung cancer. The J. Burns Amberson Lecture.
Amer. Rev. Resp. Dis. 98:573-590, 1968.
469. Kutscher, W., R. Tomingas, and H. P. Weisfeld. Untersuchungen tiber die
Schadlichkeit von Russen unter besonderer Berucksichtigung ihrer cancero-
genen Wirkung. 5. Mitteilung: Uber die Ablosbarkeit von 3,4-Benzpyren
durch Blutserum und einege Eiweissfaktoren des Serums. Arch. Hyg. Bakt.
151:646-655, 1967. (summary in English)
470. Lamb, R. G., and M. Neiburger. An interim version of a generalized urban
air pollution model. Atmos. Environ. 5:239-264, 1971.
471. Lane, W. R., and B. R. D. Stone. Structure and density of particulate
aggregates, pp. 417-426. In Proceedings of the International Conference on
Mechanisms of Corrosion by Fuel Impurities. Marchwood, near Southampton,
Hampshire, England. London: Butterworth & Co., 1963.
-------
References 333
472. Lane-Petter, W., Ed. Animals for Research. Principles of Breeding and
Management. New York: Academic Press Inc., 1963. 531 pp.
473. Lang, L., Ed. Absorption Spectra in the Ultraviolet and Visible Region. New
York: Academic Press Inc., 1961-.
474. Laskin, S., M. Kuschner, and R. T. Drew. Studies in pulmonary carcinogenesis,
pp. 321-350. In M. G. Hanna, Jr., P. Nettesheim, and J. R. Gilbert, Eds. In-
halation Carcinogenesis. AEC Symposium Series, No. 18. Washington, D.C.:
U.S. Atomic Energy Commission, 1970.
475. Lasnitzki, I. Growth pattern of the mouse prostate gland in organ culture and
its response to sex hormones, vitamin A, and 3-methylcholanthrene. Nat.
Cancer Inst. Monogr. 12:381-403, 1963.
476. Lasnitzki, I. Observations on effects of condensates from cigarette smoke
on human foetal lung in vitro. Brit. J. Cancer 12:547-552, 1958.
477. Lasnitzki, I. The action of hormones on cell and organ cultures, pp. 591-658.
In E. N. Willmer, Ed. Cells and Tissues in Culture. Methods, Biology, and
Physiology. Vol. 1. London: Academic Press Inc., 1965.
478. Lasnitzki, I. The effect of a hydrocarbon-enriched fraction of cigarette smoke
condensate on human fetal lung grown in vitro. Cancer Res. 28:510-516,
1968.
479. Lasnitzki, I. The effect of a hydrocarbon-enriched fraction from cigarette
smoke on mouse tracheas grown in vzfro.Brit. J. Cancer 22:105-109, 1968.
480. Lasnitzki, I. The effect of 3:4-benzopyrene on human foetal lung grown in
vitro. Brit. J. Cancer 10:510-516, 1956.
481. Lasnitzki, I. The effects of actinomycin D and methylcholanthrene on the
cytology and RNA and protein synthesis in prostatic epithelium grown in
vitro. Cancer Res. 29:318-326, 1969.
482. Lasnitzki, I. The influence of A hypervitaminosis on the effect of 20-methyl-
cholanthrene on mouse prostate glands grown in vitro. Brit. J. Cancer
9:434-441, 1955.
483. Lawther, P. J., B. T. Commins, and R. E. Waller. A study of the concentra-
tions of polycyclic aromatic hydrocarbons in gas works retort houses. Brit. J.
Ind. Med. 22:13-20, 1965.
484. Ledford, C. J., C. P. Morie, and C. A. Glover. Separation of polynuclear
aromatic hydrocarbons in cigaret smoke high-resolution liquid chromatog-
raphy. Tobacco Sci. 14:158-160, 1970.
485. Leighton, P. A. Photochemistry of Air Pollution. New York: Academic
Press Inc., 1961. 300pp.
486. Leitch, A. Mule-spinners' cancer and mineral oils. Brit. Med. J. 2:941-943,
1924.
487. Letter, J., and M. J. Shear. Production of tumors in mice with tars from city
air dusts. J. Nat. Cancer Inst. 3:167-174, 1942.
488. Leiter, J., M. B. Shimkin, and M. J. Shear. Production of subcutaneous
sarcomas in mice with tars extracted from atmospheric dusts. J. Nat. Cancer
Inst. 3:155-165, 1942.
489. Lemke, E. E., C. Thomas, and W. E. Zwaicker. Profile of Air Pollution in
Los Angeles County. Los Angeles County Air Pollution Control District,
1969.
-------
334 References
490. Leubner, I. H. Comment. (Observed phosphorescence lifetimes and glass
relaxation at 77°K.) J. Phys. Chem. 73:2088-2090, 1969.
491. Leuchtenberger, C., and R. Leuchtenberger. In L. Seven, Ed. Lung Tumours
in Animals. Proceedings of the Third Quadrennial International Conference
on Cancer. June 24-29, 1965. Perugia, Italy: University of Perugia, 1966.
492. Levin, M. L., W. Haenszel, B. E. Carroll, P. R. Gerhardt, V. H. Handy, and
S. C. Ingraham, II. Cancer incidence in urban and rural areas of New York
State. J. Nat. Cancer Inst. 24:1243-1257, 1960.
493. Levy, B. M. Induction of fibrosarcoma in the primate Tamarinus nigricollis.
Nature 200:182-183, 1963.
494. Lew, E. A. Cancer of the respiratory tract. Recent trends in mortality. J. Int.
Coll. Surg. 24:12-27, 1955.
495. Leymann. Steinkohlenteer- oder Steinkohlenteer-Pechkratze und -Krebs.
Zentralbl. Gewerbehyg. 5:2-7, 35-40, 51-55, 170-174, 1917.
496. Lijinsky, W., and P. Shubik. The detection of polycyclic aromatic hydro-
carbons in liquid smoke and some foods. Toxicol. Appl. Pharmacol.
7:337-343, 1965.
497. Lione, J. G., and J. S. Denholm. Cancer of the scrotum in wax pressmen. II.
Clinical observations. A.M.A. Arch. Ind. Health 19:530-539, 1959.
498. Lipsett, F. R. Energy transfer in polyacene solid solutions. VIII. A bibliog-
raphy for 1968. Molec. Cryst. Liquid Cryst. 6:175-204, 1969.
499. Lipsett, F. R., G. Bechthold, F. D. Blair, F. V. Cairns, and D. H. O'Hara.
Apparatus for measurement of luminescence spectra with a digital recording
system. Appl. Opt. 9:1312-1318, 1970.
500. Little, J. B., B. N. Grossman, and W. F. O'Toole. Respiratory carcinogenesis
in hamsters induced by polonium-210 alpha radiation and benzo{a] pyrene, pp.
383-392. In P. Nettesheim, M. G. Hanna, Jr., and J. W. Deatherage, Jr., Eds.
Morphology of Experimental Respiratory Carcinogenesis. AEC Symposium
Series, No. 21. Springfield, Va.: National Technical Information Service,
Department of Commerce, 1970.
501. Littlewood, A. B. The coupling of gas chromatography with methods of
identification. I. Mass spectrometry. Chromatographia 1968:37-42.
502. Lloyd, J. W. Long-term mortality study of steelworkers. V. Respiratory
cancer in coke plant workers. J. Occup. Med. 13:53-68, 1971.
503. Ludwig, J. H., G. B. Morgan, and T. B. McMullen. Trends in urban air quality.
Trans. Amer. Geophys. Union 51:468-475, 1970.
504. Luther, M. Cancer in Subhuman Primates. Public Health Service Publication
1138. Bibliography Series, No. 44. Washington, D.C.: U.S. Government
Printing Office, 1962. 61 pp.
505. Lyons, M. J. Assay of possible carcinogenic hydrocarbons from cigarette
smoke. Nature 177:630-631, 1956.
506. MacMahon, R. The ethnic distribution of cancer mortality in New York City,
1955. ActaUn. Int. Cancr. 16:1716-1724, 1960.
507. Maher, V. M., E. C. Miller, J. A. Miller, and W. Szybalski. Mutations and
decreases in density of transforming DNA produced by derivatives of the
carcinogens 2-acetyl-aminofluorene and.N-methyl-4-ammoazobenzene.
Molec. Pharmacol. 4:411-426, 1968.
-------
References 335
508. Maisin, J., and A. De Jonghe. Au sujet de 1'action de la lumiere et de 1'ozone
sur certains corps cancerigenes. C. R. Soc. Biol. 117:111-114, 1934.
509. Maisin, J., and E. Picard. Production experimentale d'un epithelioma
epidermoi'de de la vessie chez le rat blanc. C. R. Soc. Biol. 91:799-801, 1924.
510. Mallet, L. Recherche des hydrocarbures polybenzeniques du type benzo-3.4-
pyrene dans la faune des milieux marins (Manche, Atlantique et M6diterranee).
C. R. Acad. Sci. (Paris) 253:168-170, 1961.
511. Mallet, L., and M. H6ros. Pollution des terres vegetales par les hydrocarbures
polybenzeniques du type benzo-3.4-pyrene. C. R. Acad. Sci. (Paris) 254:958-
960, 1962.
512. Mancuso, T. F., and E. J. Coulter. Cancer mortality among native white, for-
eign-born white, and nonwhite male residents of Ohio: Cancer of the lung,
larynx, bladder, and central nervous system. J. Nat. Cancer Inst. 20:79-105,
1958.
513. Mancuso, T. F., E. M. MacFarlane, and J. D. Porterfield. Distribution of
cancer mortality in Ohio. Amer. J. Public Health 45:58-70, 1955.
514. Mann, C. K., and K. K. Barnes. Electrochemical Reactions in Nonaqueous
Systems. New York: Marcel Dekker, 1970. 560 pp.
515. Manos, N. E., and G. F- Fisher. An index of air pollution and its relation to
health. J. Air Pollut. Control Assoc. 9:5-11, 1959.
516. Mantel, N. Some statistical viewpoints in the study of carcinogenesis.
Progr. Exp. Tumor Res. 11:431-443, 1969.
517. Mantel, N., and W. R. Bryan. "Safety" testing of carcinogenic agents. J. Nat.
Cancer Inst. 27:455-470, 1961.
518. Mantel, N., W. E. Heston, and J. M. Gurian. Thresholds in linear dose-response
models for carcinogenesis. J. Nat. Cancer Inst. 27:203-215, 1961.
519. Marchesani, V. J., T. Towers, and H. C. Wohlers. Minor sources of air pollutant
emissions. J. Air Pollut. Control Assoc. 20:19-22, 1970.
520. Marchetti, A. P., and D. R. Kearns. Investigation of singlet-triplet transitions
by the phosphorescence excitation method. IV. The singlet-triplet absorption
spectra of aromatic hydrocarbons. J. Amer. Chem. Soc. 89:768-777, 1967.
521. Mastromatteo, E. Cutting oils and squamous-cell carcinoma. Part I. Incidence
in a plant with a report of six cases. Brit. J. Ind. Med. 12:240-243, 1955.
522. Masuda, Y., and M. Kuratsune. Photochemical oxidation of benzo[a]pyrene.
Air Water Pollut. 10:805-811, 1966.
523. McCarthy, W. J., and J. D. Winefordner. Phosphorimetry as a means of
chemical analysis, pp. 371-442. In G. G. Guilbault, Ed. Fluorescence: Theory,
Instrumentation, and Practice. New York: Marcel Dekker, Inc., 1967.
524. McDonald, S., Jr., and D. L. Woodhouse. On the nature of mouse lung adeno-
mata, with special reference to "the effects of atmospheric dust on the inci-
dence of these tumours. J. Path. Bact. 54:1-12, 1942.
525. McFadden, W. H. Introduction of gas-chromatographic samples to a mass
spectrometer. Separation Sci. 1:723-746, 1966.
526. McFadden, W. H. Mass-spectrometric analysis of gas-chromatographic ef-
fluents, pp. 265-332. In J. C. Giddings and R. A. Keller, Eds. Advances in
Chromatography. Vol. 4. New York: Marcel Dekker, Inc., 1967.
527. Mclntire, K. R., and G. L. Princler. Prolonged adjuvant stimulation in germ-
-------
336 References
free BALB/C mice: Development of plasma cell neoplasia. Immunology 17:
481-487, 1969.
528. McKee, H. C., and W. A. McMahon. Polynuclear Aromatic Content of Vehicle
Emissions. Project 21-2139, TR-1. San Antonio, Texas: Southwest Research
Institute, 1967. 37 pp.
529. Meranze, D. R., M. Gruenstein, and M. B. Shimkin. Effect of age and sex on
the development of neoplasms in Wistar rats receiving a single intragastric in-
stillation of 7,12-dimethylbenz(a)anthracene. Int. J. Cancer 4:480-486,
1969.
530. Meyer, E. C., and Liebow, A. A. Relationship of interstitial pneumonia,
honeycombing and atypical epithelial proliferation to cancer of the lung.
Cancer 18:322-351, 1965.
531. Mihalyi, E. Estimation of stray-light and fluorescence effects in differential
spectroscopy. Arch. Biochem. 11-0:325-330, 1965.
532. Miller, J. A. Carcinogenesis by chemicals: An overview. G. H. A. Clowes
Memorial Lecture. Cancer Res. 30:559-576, 1970.
533. Miller, J. A., and E. C. Miller. The carcinogenic aminoaza dyes. Adv. Cancer
Res. 1:340-396, 1953.
534. Miller, L., W. E. Smith, and S. W. Berliner. Tests for effect of asbestos on
benzol a] pyrene Carcinogenesis in the respiratory tract. Ann. N.Y. Acad.
Sci. 132:489-500, 1965.
535. Mirvish, S., G. Cividalli, and I. Berenblum. Slow elimination of urethan in
relation to its high carcinogenicity in newborn mice. Proc. Soc. Exp. Biol.
Med. 116:265-268, 1964.
536. Mold, J. D., T. B. Walker, and L. G. Veasey. Selective separation of poly-
cyclic aromatic compounds by countercurrent distribution with a solvent
system containing tetramethyluric acid. Anal. Chem. 35:2071-2074, 1963.
537. Mondal, S., and C. Heidelberger. In vitro malignant transformation by
methylcholanthrene of the progeny of single cells derived from C3H mouse
prostate. Proc. Nat. Acad. Sci. U.S.A. 65:219-225, 1970.
538. Mondal, S., P. T. lype, L. M. Griesbach, and C. Heidelberger. Antigenicity
of cells derived from mouse prostate cells after malignant transformation
in vitro by carcinogenic hydrocarbons. Cancer Res. 30:1593-1597, 1970.
539. Moore, G. E., and F. G. Bock. A summary of research techniques for in-
vestigating the cigarette smoking-lung cancer problem. Surgery 39:
120-130, 1956.
540. Moriconi, E. J., B. Rakoczy, and W. F. O'Connor. Ozonolysis of polycyclic
aromatics. VIII. Benzo[a] pyrene. J. Amer. Chem. Soc. 83:4618-4623,
1961.
541. Moriconi, E. J., and L. Salce. Ozonation of polycyclic aromatics. XV. Car-
cinogenicity and K- and/or L-region additivity towards ozone. Adv. Chem.
Ser. 77:65-73, 1968.
542. Moriconi, E. J., and F. A. Spano. Heteropolar ozonization of aza-aromatics
and their N-oxides. J. Amer. Chem. Soc. 86:38-46, 1964.
543. Moriconi, E. J., and L. B. Taranko. Ozonolysis of polycyclic aromatics. X.
7,12-Dimethylbenz[a] anthracene. J. Org. Chem. 28:1831-1834, 1963.
544. Moriconi, E. J., and L. B. Taranko. Ozonolysis of polycyclic aromatics. XI.
3-Methylcholanthrene. J. Org. Chem. 28:2526-2533, 1963.
-------
References 337
545. Morton, J. J., E. M. Luce-Clausen, and E. B. Mahoney. The effect of visible
light on the development of tumors induced by benzpyrene in the skin of
mice. Amer. J. Roentgen. 43:896-898, 1940.
546. Morton, J. J., E. M. Luce-Clausen, and E. B. Mahoney. Visible light and skin
tumors induced with benzpyrene in mice. Cancer Res. 2:256-260, 1942.
547. Moscona, A., O. A. Trowell, and E. N. Willmer. Methods, pp. 19-98. In
E. N. Willmer, Ed. Cells and Tissues in Culture. Methods, Biology, and Phys-
iology. Vol. 1. London: Academic Press Inc., 1965.
548. Mueller, P. K., H. L. Helwig, A. E. Alcocer, W. K. Gong, and E. E. Jones.
Concentration of fine particles and lead in car exhaust, pp. 60-77. In Sym-
posium on Air-Pollution Measurement Methods. ASTM Special Technical
Publication 352. Philadelphia: American Society for Testing Materials, 1964.
549. Mueller, P. K., R. W. Mosley, and L. B. Pierce. Carbonate and Non-carbonate
Carbon in Atmospheric Particles. Air Industrial Laboratory Report 72. Berke-
ley: California State Department of Health, 1970. 16 pp.
550. Munich, A. J., A. J. Klee, and P. W. Britton. 1968 National Survey of Com-
munity Solid Waste Practices. Public Health Service Publication 1866. Cin-
cinnati: U.S. Department of Health, Education, and Welfare, 1968.
551. Mukai, F., and W. Troll. The mutagenicity and initiating activity of some
aromatic amine metabolites. Ann. N.Y. Acad. Sci. 163:828-836, 1969.
552. Murray, R. W., and M. L. Kaplan. Gas-phase reactions of singlet oxygen from
a chemical source. J. Amer. Chem. Soc. 90:4161-4162, 1968.
553. Murray, R. W., W. C. Lumma, Jr., and J. W.-P. Lin. Singlet oxygen sources in
ozone chemistry. Decomposition of oxygen-rich intermediates. J. Amer.
Chem. Soc. 92:3205-3207, 1970.
554. Nakamizo, M., and Y. Kanda. Fluorescence spectra of organic compounds
in solution—I. On the positions of the O,O-bands of the fluorescence spec-
tra. Spectrochim. Acta 19:1235-1248, 1963.
555. Nebert, D. W., and H. V. Gelboin. Substrate-inducible microsomal aryl-
hydroxylase in mammalian cell culture. I. Assay and properties of induced
enzyme. J. Biol. Chem. 243:6242-6249, 1968.
556. Nebert, D. W., and H. V. Gelboin. Substrate-inducible microsomal aryl hy-
droxylase in mammalian cell culture. II. Cellular responses during enzyme
induction. J. Biol. Chem. 243:6250-6261, 1968.
557. Nebert, D. W. j and H. V. Gelboin. The in vivo and in vitro induction of
aryl hydrocarbon hydroxylase in mammalian cells of different species, tis-
sues, strains, and developmental and hormonal states. Arch. Biochem. Biophys.
134:76-89, 1969.
558. Nebert, D. W., and H. V. Gelboin. The role of ribonucleic acid and protein
synthesis in microsomal aryl hydrocarbon hydroxylase induction in cell
culture. The independence of transcription and translation. J. Biol. Chem.
245:160-168, 1970.
559. Nebert, D. W., J. Winker, and H. V. Gelboin. Aryl hydrocarbon hydroxyl-
ase activity in human placenta from cigarette smoking and nonsmoking
women. Cancer Res. 29:1763-1769, 1969.
560. Nettesheim, P., M. G. Hanna, Jr., D. H. Doherty, R. F. Newell, and A. Hell-
man. Effects of chronic exposure to artificial smog and chromium oxide
dust on the incidence of lung tumors in mice, pp. 305-320. In M. G. Hanna,
-------
338 References
Jr., P. Nettesheim, and J. R. Gilbert, Eds. Inhalation Carcinogenesis. AEC
Symposium Series, No. 18. Washington, D.C.: U. S. Atomic Energy Commis-
sion, 1970.
561. Neuberger, H., C. L. Hosier, and W. C. Kocmond. Vegetation as aerosol fil-
ter, pp. 693-702. In S. W. Tromp and W. H. Weihe, Eds. Biometeorology.
Vol. 2. Proceedings of the Third International Biometeorological Con-
gress held at Pau, S. France, 1-7 September 1963, organized by The Inter-
national Society of Biometeorology. Oxford: Pergamon Press Ltd., 1967.
562. Neve, E. F. Causation of cancer. Practitioner 122:355-359, 1929.
563. Newman, M. S., and S. Blum. A new cyclization reaction leading to epoxides
of aromatic hydrocarbons. J. Amer. Chem. Soc. 86:5598-5600, 1964.
564. Noonan, F. M., and B. Linsky. Internal reflection spectroscopy applied to
air pollution. Atmos. Environ. 4:125-128, 1970.
565. Noyes, W. F. Carcinogen induced neoplasia with metastasis in a South
American primate, Saguinus oedipus. Proc. Soc. Exp. Biol. Med. 131:223-
225, 1969.
566. Noyes, W. F. Carcinogen-induced sarcoma in the primitive primate, Tupaia
glis. Proc. Soc. Exp. Biol. Med. 127:594-596, 1968.
567. Nurmukhametov, R. N. Electronic absorption and luminescence spectra of
aromatic polycyclic hydrocarbons. Russ. Chem. Rev. 35:469-486, 1966.
568. O'Donnell, H., T. L. Montgomery, and M. Corn. Routine assessment of the
particle size-weight distribution of urban aerosols. Atmos. Environ. 4:1-7,
1970.
569. O'Donovan, W. J. Epitheliomatous ulceration among tar workers. Brit. J.
Derm. 32:215-228, 245-252, 1920.
570. Offen, H. W., and D. E. Hein. Environmental effects on phosphorescence.
VI. Matrix site effects for triphenylene. J. Chem. Phys. 50:5274-5278, 1969.
Old, L. J., E. A. Boyse, D. A. Clarke, and E. A. Carswell. Antigenic prop-
erties of chemically induced tumors. Part II. Antigens of tumor cells. Ann.
N.Y. Acad. Sci. 101:80-106, 1962.
572. Olsen, D., and J. L. Haynes. Preliminary Air Pollution Survey of Organic
Carcinogens. A Literature Review. National Air Pollution Control Adminis-
tration Publication APTD 69-43. Raleigh, N.C.: U.S. Department of Health,
Education, and Welfare, 1969. 117 pp.
573. Oro, J., and J. Han. Application of combined chromatography-mass spec-
trometry to the analysis of aromatic hydrocarbons formed by pyrolysis of
methane. J. Gas Chromatogr. 5:480-485, 1967.
574. Palekar, L., M. Kuschner, and S. Laskin. The effect of 3-methylcholan-
threne on rat trachea in organ culture. Cancer Res. 28:2098-2104, 1968.
575. Panteleev, V. V., M. L. Petukh, O. I. Putrenko, T. A. Yankovskaya, and
A. A. Yankovskii. The sensitivity of emission spectral analysis when using a
laser. Zh. Prikl. Spektrosk. 12:1106-1108, 1970.
576. Paoletti, J., and J.-B. Le Pecq. Corrections for instrumental errors in mea-
surement of fluorescence and polarization of fluorescence. Anal Biochem
31:33-41,1969.
577. Parker, C. A. Photoluminescence of Solutions. Amsterdam: Elsevier Publish-
ing Company, 1968. 544 pp.
-------
References 339
578. Parker, C. A., and C. G. Hatchard. Photoreaction of benz[a]pyrene in solu-
tions containing polymer. Photochem. Photobiol. 5:699-703, 1966.
579. Parker, C. A., and W. T. Rees. Determination of 3:4-Benzpyrene in the At-
mosphere of a Submarine. AML Report A/68 (M). Poole, England: Admiralty
Materials Laboratory, 1966. 9 pp.
580. Parmenter, C. S., and J. D. Rau. Fluorescence quenching in aromatic hy-
drocarbons by oxygen. J. Chem. Phys. 51:2242-2246, 1969.
581. Pasceri, R., and S. K. Friedlander. Measurements of the particle size distri-
bution of the atmospheric aerosol: II. Experimental results and discussion.
J. Atmos. Sci. 22:577-584, 1965.
582. Passey, R. D. Experimental soot cancer. Brit. Med. J. 2:1112-1113, 1922.
583. Passwater, R. A. Guide to Fluorescence Literature. 2 vols. New York:
Plenum Press Data Division, 1967. 736 pp.
584. Pasternak, G., A. Graffi, F. Hoffman, and K.-H. Horn. Resistance against
carcinomas of the skin induced by dimethylbenzanthracene (DMB A) in mice
of the strain XVH/Bln. Nature 203:307-308, 1964.
585. Pathak, M. A., F. Daniels, Jr., and T. B. Fitzpatrick. The presently known
distribution of furocoumarins (psoralens) in plants. J. Invest. Derm. 39:
225-239, 1962.
586. Patterson, A. M., L. T. Capell, and D. F. Walker. The Ring Index. A List of
Ring Systems Used in Organic Chemistry. (2nd ed.) Washington, D.C.:
American Chemical Society, 1960. 1425pp.
587. Payne, W. W. Methods of sampling, separation, and analysis of air pollu-
tants: Special requirements for carcinogenic bioassay. Nat. Cancer Inst.
Monogr. 9:75-80, 1962.
588. Peacock, P. R. Evidence regarding the mechanism of elimination of 1:2-
benzpyrene, l:2:5:6-dibenzanthracene, and anthracene from the blood-
steam of injected animals. Brit. J. Exp. Path. 17:164-172, 1936.
589. Peover, M. E. Electrochemistry of aromatic hydrocarbons and related sub-
stances, pp. 1-51. In A. J. Baird, Ed. Electroanalytical Chemistry. Vol. 2.
New York: Marcel Dekker, Inc., 1967.
590. Perkampus, H. H., and L. Pohl. Uber die Fluoreszenzspektren diinner
Filme aromatischer Kohlenwasserstoffe. Z. Phys. Chem. 40:162-188, 1964.
591. Peterson, C. M., H. J. Paulus, and C. H. Foley. The number size distribution
of atmospheric particles during temperature inversions. J. Air Pollut. Con-
trol Assoc. 19:795-801, 1969.
592. Petrov, N. N. Results of experiments in carcinogenesis in monkeys over a
20-year period (1939-1960). Piobl. Oncol. 6:1709-1715, 1960.
593. Pfeiffer, C. A., and E. Allen. Attempts to produce cancer in rhesus mon-
keys with carcinogenic hydrocarbons and estrogens. Cancer Res. 8:97-
127, 1948.
594. Pierce, G. B. Differentiation of normal and malignant cells. Fed. Proc. 29:
1248-1254, 1970.
595. Pietrzyk, D. J. Organic polarography. Anal. Chem. 42:139R-152R, 1970.
596. Pitot, H. C., and C. Heidelberger. Metabolic regulatory circuits and carcino-
genesis. Cancer Res. 23:1694-1700, 1963.
597. Poglazova, M. N., G. E. Fedoseeva, A. J. Khesina, M. N. Meissel, and L. M.
-------
340 References
Shabad. Destruction of benzo(a)pyrene by soil bacteria. Life Sci. 6:1053-
1062, 1967.
598. Pooler, F., Jr. Airflow over a city in terrain of moderate relief. J. Appl.
Meteor. 2:446-456, 1963.
599. Porath, J. Molecular sieving and adsorption. Nature 218:834-838, 1968.
600. Pott, P. Chirurgical Observations Relative to the Cataract, the Polypus of
the Nose, the Cancer of the Scrotum, the Different Kinds of Ruptures, and
the Mortification of the Toes and Feet. London: L. Hawes, W. Clarke, and
R. Collins, 1775.208pp.
601. Potter, V. R. Biochemical perspectives in cancer research. Cancer Res. 24:
1085-1098, 1964.
602. Prehn, R. T. A clonal selection theory of chemical carcinogenesis. J. Nat.
Cancer Inst. 32:1-17, 1964.
603. Prehn, R. T. Specific isoantigenicities among chemically induced tumors.
Ann. N.Y. Acad. Sci. 101:107-113, 1962.
604. Prehn, R. T. Tumor-specific antigens of putatively nonviral tumors. Cancer
Res. 28:1326-1330, 1968.
605. Prindle, R. A. Some considerations in the interpretation of air pollution
health effects data. J. Air Pollut. Control Assoc. 9:12-19, 1959.
606. Pringsheim, P. Fluorescence and Phosphorescence. New York: Interscience
Publishers, Inc., 1949. 794 pp.
607. Puck, T. T., and P. I. Marcus. A rapid method for viable cell titration and
clone production with HeLa cells in tissue culture: The use of x-irradiated
cells to supply conditioning factors. Proc. Nat. Acad. Sci. U.S.A. 41:432-
437, 1955.
608. Puck, T. T., P. I. Marcus, and S. J. Cieciura. Clonal growth of mammalian
cells in vitro. Growth characteristics of colonies from single HeLa cells
with and without a "feeder" layer. J. Exp. Med. 103:273-284, 1956.
609. Pullinger, B. D. The first effects on mouse skin of some polycyclic hydro-
carbons. J. Path. Bact. 50:463^*71, 1940.
610. Pullman, A., and B. Pullman. A quantum chemist's approach to the mech-
anism of chemical carcinogenesis, pp. 9-24. In Jerusalem Symposia on
Quantum Chemistry and Biochemistry. Vol. 1. 1969.
611. Pullman, A., and B. Pullman. Electronic structure and carcinogenic activity
of aromatic molecules. New developments. Adv. Cancer Res. 3:117-169, 1955.
612. Pybus, F. C. Cancer and atmospheric pollution. Newcastle Med. J. 28:31-
66, 1963.
613. Pylev, L. N. Effect of the dispersion of soot in deposition of 3,4-benzpyrene
in lung tissue of rats. Hyg. Sanit. 32:174-179, Apr.-June, 1967.
614. Pylev, L. N. Experimental induction of lung cancer in rats by intratracheal
administration of 9,10-dimethyl-l,2-benzanthracene. Bull. Exp. Biol. Med.
(U.S.S.R.) 52:1316-1319, 1961.
615. Pylev, L. N. Experimental induction of lung cancer in rats with 3,4-benz-
pyrene. Vestn. Akad. Med. Nauk S.S.S.R. 19(11):41-45, 1964. (In Russian)
616. Pylev, L. N. Induction of lung cancer in rats by intratracheal insufflation of
cancerogenic hydrocarbons. Acta Un. Int. Cancr. 19:688-691, 1962.
617. Pylev, L. N. Late appearance of tumors in rats after the administration of
-------
References 341
9,10-dimethyl-l,2-benzanthraceneinto the lungs. Vopr. Onkol. 10(8): 53-
60, 1964. (in Russian)
618. Pylev, L. N., F. J. C. Roe, and G. P. Warwick. Elimination of radioactivity
after intratracheal instillation of tritiated 3,4-benzopyrene in hamsters.
Brit. J. Cancer 23:103-115, 1969.
619. Pysh, E. S., and N. C. Yang. Polarographic oxidation potentials of aromatic
compounds. J. Amer. Chem. Soc. 85:2124-2130, 1963.
620. Rail, D. P. Difficulties in the extrapolation of the results of toxicity studies
in laboratory animals to man, pp. 62-73. In Physiological Characterization of
Health Hazards in Man's Environment. Washington, D.C.: National Academy of
Sciences, 1967.
621. Rand, R. N. Practical spectrophotometric standards. Clin. Chem. 15:839-
863, 1969.
622. Rao, C. N. R. Chemical Applications of Infrared Spectroscopy, pp. 163-
166. New York: Academic Press Inc., 1963.
623. Rapp, F., and J. L. Melnick. Cell, tissue and organ cultures in virus research,
pp. 263-316. In E. N. Willmer, Ed. Cells and Tissues in Culture. Methods,
Biology, and Physiology. Vol. 3. London: Academic Press Inc., 1966.
624. Rathkamp, G., and D. Hoffmann. Fluorenes and Fluoranthenes in Cigarette
Smoke. Presented to the 24th Tobacco Chemists' Research Conference,
Montreal, Canada, 1970.
625. Raven, R. W., and F. J. C. Roe, Eds. The Prevention of Cancer. London:
Butterworth & Co., 1967. 397 pp.
626. Reckner, L. R., W. E. Scott, and W. F. Biller. The composition and odor of
diesel exhaust. Proc. Amer. Petrol. Inst. 45 (Sect. 3): 133-147, 1965.
627. Refining survey. Oil Gas J. 67:115-137. 1969.
628. Reid, D. C., J. Cornfield, R. E. Markush, D. Seigel, E. Pedersen, and W.
Haenszel. Studies of disease among migrants and native populations in
Great Britain, Norway, and the United States. III. Prevalence of cardiores-
piratory symptoms among migrants and native-born in the United States.
Nat. Cancer Inst. Monogr. 19:321-346, 1966.
629. Reid, L., Ed. The Pathology of Emphysema. Chicago: Year Book Medical
Publishers, 1967. 372 pp.
630. Reiquam, H. An atmospheric transport and accumulation model for air-
sheds. Atmos. Environ. 4:233-247, 1970.
631. Riddick, J. A., and W. B. Bunger. Techniques of Chemistry. Vol. 2. Organic
Solvents: Physical Properties and Methods of Purification. (3rd ed.) New York:
John Wiley & Sons, Inc., 1970. 1072 pp.
632. Roe, F. J. C. Comparison of carcinogenicity of tobacco smoke condensate
and particulate air pollutants and a demonstration that their effects may
be additive, pp. 110-111. In K. H. Weber, Ed. Alkylierend Wirkende Verbin-
dungen. Hamburg: Forschungsinstirut im Berband der Cigaretten-industrie,
1968.
633. Roe, F. J. C. The induction of skin cancer in mice by combinations of ciga-
rette smoke condensate and particulate matter from London air, p. 759.
In Ninth International Cancer Congress. Abstracts of Papers. Tokyo, Japan,
Oct. 23-29, 1966.
-------
342 References
634. Roe, F. J. C., and G. A. Grant. Inhibition by germ-free status of development
of liver and lung tumours in mice exposed neonatally to 7,12-dimethyl-
benz(a)anthracene: Implications in relation to tests for carcinogenicity.
Int. J. Cancer 6:133-144, 1970.
635. Roe, F. J. C., and W. E. H. Peirce. Further studies on the tumour-promoting
action of citrus oils, pp. 212-213. In British Empire Cancer Campaign.
Thirty-eighth Annual Report, 1960. Part II. London: British Empire Cancer
Campaign, 1960.
636. Rbhrborn, G. Mutagenicity tests in mice. I. The dominant lethal method
and the control problem. Humangenetik 6:345-361, 1968.
637. Roitt, I. M., and W. A. Waters. Action of benzoyl peroxide on polycyclic aro-
matic hydrocarbons. J. Chem. Soc. 1952:2695-2705.
638. Roller, M. R., and C. Heidelberger. Attempts to produce carcinogenesis in
organ cultures of mouse prostate with polycyclic hydrocarbons. Int. J.
Cancer 2:509-520, 1967.
639. Rondia, D. Sur la volatilite des hydrocarbures polycycliques. Air Water
PoUut. 9:113-121, 1965.
640. Rook, A. J., D. S. Wilkinson, and F. J. Ebling, Eds. Textbook of Derma-
tology. 2 vols. Oxford: Blackwell Scientific Publications, 1968. 2980pp.
641. Rooney, J. J., and R. C. Pink. Formation and stability of hydrocarbon radi-
cal-ions on a silica-alumina surface. Trans. Faraday Soc. 58:1632-1641, 1962.
642. Rosinski, J., and C. T. Nagamoto. Particle deposition on and reentrain-
ment from coniferous trees. Part I. Experiments with trees. Kolloid-Z. 204:
78-84, 1966.
643. Rounds, D. E. Environmental influences on living cells. Arch. Environ.
Health 12:78-84, 1966.
644. Rounds, D. E., A. Awa, and C. M. Pomerat. Effect of automobile exhaust on
cell growth in vitro. Arch. Environ. Health 5:319-324, 1962.
645. Royal College of Physicians, London. Committee on Smoking and Atmospheric
Pollution. Air Pollution and Health. Summary and Report on Air Pollution
and Its Effect on Health. London: Pitman Medical and Scientific Publish-
ing Co., 1970. 80pp.
646. Rusch, H. P., B. E. Kline, and C. A. Baumann. The nonadditive effect of
ultraviolet light and other carcinogenic procedures. Cancer Res. 2:183-
188, 1942.
647. Russell, L. B., and W. L. Russell. Pathways of radiation effects in the mother
and the embryo, pp. 50-59. In Cold Spring Harbor Symposia on Quantita-
tive Biology. Vol. XIX. The Mammalian Fetus: Physiological Aspects of
Development. Long Island, N.Y.: The Biological Laboratory, 1954.
648. Saffiotti, U. Experimental respiratory tract earcinogenesis and its relation
to inhalation exposures,.pp. 27-54. In M. G. Hanna, Jr., P. Nettesheim,
and J. R. Gilbert, Eds. Inhalation Carcinogenesis. Proceedings of a Biology
Division, Oak Ridge/ National Laboratory, conference held in Gatlinburg,
Tenn., Oct. 8-11, 1969. AEC Symposium Series, No. 18. Oak Ridge,
Tenn.: U.S. Atomic Energy Commission, Division of Technical Information
1970.
649. Saffiotti, U., S. A. Borg, M. I. Grote, and D. B. Karp. Retention rates of
-------
References 343
participate carcinogens in the lungs. Studies in an experimental model for
lung cancer induction. Chicago Med. School Q. 24; 10-17, 1964.
650. Saffiotti, U., F. Cefis, and L. H. Kolb. A method for the experimental in-
duction of bronchogenic carcinoma. Cancer Res. 28:104-124, 1968.
651. Saffiotti, U., F. Cefis, and L. H. Kolb. Bronchiogehic carcinoma induction
by particulate carcinogens. Proc. Amer. Assoc. Cancer Res. 5:55, 1964.
(abstract)
652. Saffiotti, U., F. Cefis, L. H. Kolb, and M. I. Grote. Intratracheal injection
of particulate carcinogens into hamster lungs. Proc. Amer. Assoc. Cancer
Res. 4:59, 1963. (abstract)
653. Saffiotti, U., F. Cefis, L. H. Kolb, and P. Shubik. Experimental studies of
the conditions of exposure to carcinogens for lung cancer induction. J. Air
Pollut. Control Assoc. 15:23-25, 1965.
654. Saffiotti, U., F Cefis, and P. Shubik. Histopathology and histogenesis of
lung cancer induced in hamsters by carcinogens carried by dust particles,
pp. 537-546. In L. Severi, Ed. Lung Tumours in Animals. Proceedings of the
Third Quadrennial International Conference on Cancer. June 24-29, 1965.
Perugia, Italy: University of Perugia, 1966.
655. Saffiotti, U., R. Montesano, A. R. Sellakumar, and S. A. Borg. Experi-
mental cancer of the lung. Inhibition by vitamin A of the induction of tra-
cheobronchial squamous metaplasia and squamous cell tumors. Cancer 20:
857-864, 1967.
656. Saffiotti, U., R. Montesano, and N. Tompkins. Benzo(a)pyrene retention
in hamster lungs: Studies on particle size and on total dust load. Proc. Amer.
Assoc. Cancer Res. 8:57, 1967. (abstract)
657. Salaman, M. H., and F. J. C. Roe. Cocarcinogenesis. Brit. Med. Bull. 20:
139_144; 1964.
658. Sanders, C. L., Jr., R. C. Thompson, and W. J. Bair. Lung cancer: Dose
response studies with radionuclides, pp. 258-303. In M. G. Hanna, Jr., P.
Nettesheim, and J. R. Gilbert, Eds. Inhalation Carcinogenesis. Proceedings
of a Biology Division, Oak Ridge National Laboratory, conference held in
Gatlinburg, Tenn., Oct. 8-11, 1969. AEC Symposium Series, No. 18. Oak
Ridge, Tenn.: U.S. Atomic Energy Commission, Division of Technical
Information, 1970.
659. Sanford, K. K. "Spontaneous" neoplastic transformation of cells in vitro:
Some facts and theories. Nat. Cancer Inst. Monogr. 26:387-408, 1965.
660. Santamaria, L., and G. G. Giordano. Effects of long-wave ultraviolet hydro-
carbon carcinogenesis, pp. 569-580. In F. Urbach, Ed. The Biologic Ef-
fects of Ultraviolet Radiation. Oxford: Pergamon Press, 1969.
661. Sawicki, C. R., and E. Sawicki. Thin-layer chromatography in air pollution
research, pp. 233-293. In A. Niederwieser and G. Pataki, Eds. Progress in
Thin-layer Chromatography and Related Methods. Vol. 3. Ann Arbor, Mich.:
Ann Arbor Science Publishers, Inc., 1972.
662. Sawicki, E. Airborne carcinogens and allied compounds. Arch. Environ.
Health 14:46-53, 1967.
663. Sawicki, E. Fluorescence analysis in air pollution research. Talanta 16:1231-
1266, 1969.
-------
344 References
664. Sawicki, E. The separation and analysis of polynuclear aromatic hydro-
carbons present in the human environment. Chem.-Anal. 53:24-26, 28-30,
56-62,88-91, 1964.
665. Sawicki, E., R. C. Corey, A. E. Dooley, J. B. Gisclard, J. L. Monkman, R. E.
Neligan, and L. A. Ripperton. Tentative method of routine analysis for
polynuclear aromatic hydrocarbon content of atmospheric particulate mat-
ter. Health Lab. Sci. 7:45-55, 1970.
666. Sawicki, E., R. C. Corey, A. E. Dooley, J. B. Gisclard, J. L. Monkman, R. E.
Neligan, and L. A. Ripperton. Tentative method of spectrophotometric
analysis for benzol a] pyrene in atmospheric particulate matter. Health Lab.
Sci. 7(suppl.):68-71, 1970.
667. Sawicki, E., W. C. filbert, T. R. Hauser, F. T. Fox, and T. W. Stanley.
Benzo(a)pyrene content of the air of American communities. Amer. Ind.
Hyg. Assoc. J. 21:443-451, 1960.
668. Sawicki, E., W. Elbert, T. W. Stanley, T. R. Hauser, and F. T. Fox. Separation
and characterization of polynuclear aromatic hydrocarbons in urban air-
borne particulates. Anal. Chem. 32:810-815, 1960.
669. Sawicki, E., M. Guyer, and C. R. Engel. Paper and thin-layer electrophoretic
separations of polynuclear aza heterocyclic compounds. J. Chromatogr. 30:
522-527, 1967.
670. Sawicki, E., T. R. Hauser, W. C. Elbert, F. T. Fox, and J. E. Meeker. Poly-
nuclear aromatic hydrocarbon composition of the atmosphere in some
large American cities. Amer. Ind. Hyg. Assoc. J. 23:137-144, 1962.
671. Sawicki, E., T. R. Hauser, and T. W. Stanley. Ultraviolet, visible and fluo-
rescence spectral analysis of polynuclear hydrocarbons. Int. J. Air Pollut.
2:253-272, 1960.
672. Sawicki, E., and H. Johnson. Characterization of aromatic compounds by
low-temperature fluorescence and phosphorescence: Application to air
pollution studies. Microchem. J. 8:85-101, 1964.
673. Sawicki, E., S. P. McPherson, T. W. Stanley, J. Meeker, and W. C. Elbert.
Quantitative composition of the urban atmosphere in terms of polynuclear
aza heterocyclic compounds and aliphatic and polynuclear aromatic hydro-
carbons. Int. J. Air Water Pollut. 9:515-524, 1965.
674. Sawicki, E., J. E. Meeker, and M. J. Morgan. The quantitative composition
of air pollution source effluents in terms of aza heterocyclic compounds
and polynuclear aromatic hydrocarbons. Int. J. Air Water Pollut. 9:291-
298, 1965.
675. Sawicki, E., T. W. Stanley, and H. Johnson. Quenchofluorometric analysis
for polynuclear compounds. Mikrochim. Ichnoanal. Acta 1965:178-192.
676. Sawicki, E., T. W. Stanley, S. McPherson, and M. Morgan. Use of gas-liquid
and thin-layer chromatography in characterizing air pollutants by fluorome-
try.Talanta 13:619-629, 1966.
677. Schiffman, R., and E. Landau. Use of indexes of air pollution potential in
mortality studies. J. Air Pollut. Control Assoc. 2:384-386, 1961.
678. Schmillen, A., and R. Legler. Landolt-Boernstein Numerical Data and
Functional Relationships in Science and Technology, Group II: Atomic
and Molecular Physics. Vol. 3. Luminescence of Organic Substances. New
York: Springer-Verlag, 1967.
-------
References 345
679. Schubert, C. C., and R. N. Pease. Reaction of paraffin hydrocarbons with
ozonized oxygen: Possible role of ozone in normal combustion. J. Chem.
Phys. 24:919-920, 1956.
680. Schubert, C. C., and R. N. Pease. The oxidation of lower paraffin.hydro-
carbons. I. Room temperature reaction of methane, propane, n-butane
and isobutane with ozonized oxygen. J. Amer. Chem. Soc. 78:2044-2048,
1956.
681. Schubert, C. C., and R. N. Pease. The oxidation of lower paraffin hydro-
carbons. II. Observations on the role of ozone in the slow combustion of
isobutane. J. Amer. Chem. Soc. 78:5553-5556, 1956.
682. Schulman, S. G., and J. D. Winefordner. Influence of pH in fluorescence
and phosphorescence spectrometric analysis. Talanta 17:607-616, 1970.
683. Schwartz, L., L. Tulipan, and D. J. Birmingham. Occupational Diseases of
the Skin. (3rd ed.) Philadelphia: Lea& Febiger, 1957. 981 pp.
684. Scott, A. Cancers in mineral oil refineries, pp. 275-279. In Report of the
International Conference on Cancer, London, 17th-20th July, 1928. New
York: William Wood and Company, 1928.
685. Searl, T. D., F. J. Cassidy, W. H. King, and R. A. Brown. An analytical
method for polynuclear aromatic compounds in coke oven effluents by
combined use of gas chromatography and ultraviolet absorption spectrometry.
Anal. Chem. 42:954-958, 1970.
686. Seelig, M. G., and E. L. Benignus. Coal smoke soot and tumors of the lung
in mice. Amer. J. Cancer 28:96-111, 1936.
687. Seelig, M. G., and E. L. Benignus. The production of experimental cancer
of the lung in mice. Amer. J. Cancer 33:549-554, 1938.
688. Seelig, M. G., and Z. K. Cooper. Light and tar cancer. An experimental
study, with a critical review of the literature on light as a carcinogenic
factor. Surg. Gynec. Obstet. 56:752-761,1933.
689. Segi/M., M. Kurihara, and T. Matsuyama. Cancer Mortality for Selected
Sites in Twenty-four Countries. No. 5 (1964-1965). Dept. of Public Health,
Tohoku University School of Medicine, Sendai, Japan, 1969.
690. Selkirk, J. K., E. Huberman, and C. Heidelberger. An epoxide is an inter-
mediate hi the microsomal metabolism of the chemical carcinogen, dibenz-
(a,h)anthracene. Biochem. Biophys. Res. Commun. 43:1010-1016, 1971.
691. Shabad, L. M. Experimental cancer of the lung. J. Nat. Cancer Inst. 28:
1305-1332, 1962.
692. Shabad, L. M. Experimental cancer of the lungs. Fed. Proc. (Transl. Suppl.)
22:T331-336, 1963.
693. Shabad, L. M. Studies in the U.S.S.R. on the distribution, circulation,
and fate of carcinogenic hydrocarbons in the human environment and the
role of their deposition in tissues in carcinogenesis: A review. Cancer
Res. 27:1132-1137, 1967.
694. Shabad, L. M., L. N. Pylev, and T. S. Kolesnichenko. Importance of the
deposition of carcinogens for cancer induction in lung tissue. J. Nat. Cancer
Inst. 33:135-141, 1964.
695. Sharkey, A. G., Jr., J. L. Shultz, T. Kessler, and R. A. Friedel. High-
resolution mass spectrometry has advantages in determining organic con-
taminants in air and water. Res./Develop. 20:30-32, 1969.
-------
346 References
696. Shay, H., E. A. Aegerter, M. Gruenstein, and S. A. Komarov. Development
of adenocarcinoma of the breast in the Wistar rat following the gastric instil-
lation of methylcholanthrene. J. Nat. Cancer Inst. 10:255-266, 1949.
697. Shellabarger, C. J. Effect of 3-methylcholanthrene and X irradiation, given
singly or combined, on rat mammary carcinogenesis. J. Nat. Cancer Inst.
38:73-77, 1967.
698. Shigorin, V. D., and G. P. Shipulo. Laser excitation of organic molecules
phosphorescence. Zh. Prikl. Spektrosk. 12:331-333, 1970.
699. Shubik, P., and J. L. Hartwell. Survey of Compounds Which Have Been
Tested for Carcinogenic Activity. Supplement 1. Public Health Service
Publication 149-1. Washington, D.C.: U.S. Government Printing Office,
1957.388pp.
700. Shubik, P., and J. L. Hartwell. Survey of Compounds Which Have Been
Tested for Carcinogenic Activity. Supplement 2. Public Health Service
Publication 149-2. Washington, D.C.: U.S. Government Printing Office,
1969.655pp.
701. Shultz, J. L., R. A. Friedel, and A. G. Sharkey, Jr. Analyses of coal-tar pitch
by mass spectrometry. Fuel 44:55-61, 1965.
702. Simmers, M, H. Petroleum asphalt inhalation by mice. Effects of aerosols
and smoke on the tracheobronchial tree and lungs. Arch. Environ. Health
9:727-734, 1964.
703. Simmers, M. H., E. Podolak, and R. Kinosita. Carcinogenic effects of
petroleum asphalt. Proc. Soc. Exp. Biol. Med. 101:266-268, 1959.
704. Simpson, W. L., C. Carruthers, and W. Cramer. Loss of carcinogenic activity
when methylcholanthrene is dissolved in anhydrous lanolin. Cancer Res.
5:1-4, 1945.
705. Simpson, W. L., and W. Cramer. Sebaceous glands and experimental skin car-
cinogenesis in mice. Cancer Res. 3:515-518, 1943.
706. Sims, P. Qualitative and quantitative studies on the metabolism of a series of
aromatic hydrocarbons by rat-liver preparations. Biochem. Pharmacol. 19:
795-818, 1970.
707. Sims, P. The metabolism of some aromatic hydrocarbons by mouse embryo
cell cultures. Biochem. Pharmacol. 19:285-297, 1970.
708. Slizynski, B. M. Pachytene analysis of Snell's T(5:8)a translocation in the
mouse. J. Genet. 50:507-510, 1952.
709. Smith, W. E., N. S. Cooper, and E. L. Wynder. Biological tests of fractions
derived from cigarette smoke condensates. Proc. Amer. Assoc. Cancer Res.
1:45, 1954. (abstract)
710. Smith, W. E., D. A. Sunderland, and K. Sigiura. Experimental analysis of
the carcinogenic activity of certain petroleum products. A.M.A. Arch. Ind.
Hyg. Occup. Med. 4:299-314, 1951.
711. Smith, W. M. Evaluation of Coke Oven Emissions. Presented to the 78th
General Meeting of the American Iron and Steel Institute, New York City,
May 28-29, 1970.
712. Snell, G. D., E. Bodemann, and W. Hollander. A translocation in the house
mouse and its effect on development. J. Exp. Zool. 67:93-104, 1934.
713. Snell, G. D., and D. I. Picken. Abnormal development in the mouse caused
by chromosome unbalance. J. Genet. 31:213-235, 1935.
714. Southam, A. H. Mule-spinners cancer, pp. 280-283. In Report of the Inter-
-------
References 347
national Conference on Cancer, London, 17th-20th July, 1928. New York:
William Wood and Company, 1928.
715. Southam, A. H., and S. R. Wilson. Cancer of the scrotum: The etiology,
clinical features, and treatment of the disease. Brit. Med. J. 2:971-973,
1922.
716. Southern, P. F., and W. A. Waters. The thermal decomposition of some
me^o-substituted anthracene photo-oxides. J. Chem. Soc. 1960:4340-4346.
717. Spikes, J. D., and R. Straight. Sensitized photochemical processes in bio-
logical systems. Ann. Rev. Phys. Chem. 18:409-436, 1967.
718. Spotswood, T. M. Chromatography of polycyclic aromatic hydrocarbons
on acetylated paper. J. Chromatogr. 2:90-94, 1959.
719. Staemmler, M. I. Referat Uber Beruf und Krebs. Verb. Deutsch. Ges.
Path. 30:188-238, 1937.
720. Steer, R. P., K. R. Darnall, and J. N. Pitts, Jr. Base-induced decomposition
of peroxyacetyl nitrate. Tetrahedron Lett. 43:3765-3767, 1969.
721. Steer, R. P., J. L. Sprung, and J. N. Pitts, Jr. Singlet oxygen in the environ-
mental sciences. Evidence for the production of O2 ('A g) by energy transfer
in the gas phase. Environ. Sci. Technol. 3:946-947, 1969.
722. Stein, F., N. A. Esmen, and M. Corn. The shape of atmospheric particles
in Pittsburgh air. Atmos. Environ. 3:443-453, 1969.
723. Steiner, P. E., and C. G. Loosli. The effect of human influenza virus (type
A) on incidence of lung tumors in mice. Cancer Res. 10:385-392, 1950.
724. Stevens, B., and E. Hutton. Delayed fluorescence from micro crystalline
aromatic hydrocarbons. Proc. Phys. Soc. 81:893-897, 1963.
725. StSber, W., and H. Flachsbart. Size-separating precipitation of aerosols in
a spinning spiral duct. Environ. Sci. Tech. 3:1280-1296, 1969.
726. Stocks, P. Air pollution and cancer mortality in Liverpool Hospital region
and North Wales. Int. J. Air Pollut. 1:1-13, 1958.
727. Stocks, P. Cancer and bronchitis mortality in relation to atmospheric deposit
and smoke. Brit. Med. J. 1:74-79, 1959.
728. Stocks, P. Epidemiology of cancer of the lung in England and Wales. Brit.
J. Cancer 6:99-111, 1952.
729. Stocks, P. Lung cancer and bronchitis in relation to cigarette smoking and
fuel consumption in twenty countries. Brit. J. Prev. Soc. Med. 21:181-185,
1967.
730. Stocks, P. On the relations between atmospheric pollution in urban and
rural localities and mortality from cancer, bronchitis and pneumonia, with
particular reference to 3:4 benzopyrene, beryllium, molybdenum, vanadium
and arsenic. Brit. J. Cancer 14:397-418, 1960.
731. Stocks, P. Recent epidemiological studies of lung cancer mortality, cigarette
smoking and air pollution, with discussion of a new hypothesis of causation.
Brit. J. Cancer 20:595-623, 1966.
732. Stocks, P., and J. M. Campbell. Lung cancer death rates among non-smokers
and pipe and cigarette smokers. An evaluation in relation to air pollution
by benzpyrene and other substances. Brit. Med. J. 2:923-939, 1955.
733. Stocks, P., B. T. Commins, and K. V. Aubrey. A study of polycyclic hydro-
carbons and trace elements in smoke in Merseyside and other northern localities.
Int. J. Air Water Pollut. 4:141-153, 1961.
734. Stokinger, H. E., and D. L. Coffin. Biologic effects of air pollutants, pp.
-------
348 References
445-546. In A. C. Stern, Ed. Air Pollution. Vol. 1. Air Pollution and Its
Effects. (2nd ed.) New York: Academic Press Inc., 1968.
735. Sugiura, K., W. E. Smith, and D. A. Sunderland. Experimental production of
carcinoma in rhesus monkeys. Cancer Res. 16:951-955, 1956.
736. Sulman, E., and F. Sulman. The carcinogenicity of wood soot from the
chimney of a smoked sausage factory. Cancer Res. 6:366-367, 1946.
737. Suntzeff, V., A. B. Croninger, E. L. Wynder, E. V. Cowdry, and E. A.
Graham. Use of sebaceous-gland test of primary cigarette-tar fractions and
of certain noncarcinogeriic polycyclic hydrocarbons. Cancer 10:250-254,
1957.
738. Suskind, R. R. Acne: Occupational, pp. 563-573. In P. D. Cantor, Ed. Trau-
matic Medicine and Surgery for the Attorney. Vol. 6. Psychiatry. Skin and
its Appendages. Washington, D.C.: Butterworth Inc., 1962.
739. Suskind, R. R., and A. W. Horton. Etiologic factors and the pathogenesis
of premalignant and malignant lesions of the skin, pp. 171-192. In S.
Rothman, Ed. The Human Integument: Normal and Abnormal. Washington,
D.C.: American Association for the Advancement of Science, 1959.
740. Tanimura, H. Benzo(a)pyrene in an iron and steel works. Arch. Environ.
Health 17:172-177, 1968.
741. Tannenbaum, A. Nutrition and cancer, pp. 517-562. In F- Horn burger, Ed.
The Physiopathology of Cancer. (2nd ed.) New York: Paul B. Hoeber, Inc.,
1959.
742. Tannenbaum, A., and H. Silverstone. Nutrition in relation to cancer. Adv.
Cancer Res. 1:451-501, 1953.
743. Tarbell, D. S., E. G. Brooker, A. Vanterpool, W. Conway, C. J. Claus, and
T. J. Hall. A system for paper chromatography of 3,4-benzpyrene, some
derivatives and other polycyclic aromatic hydrocarbons. J. Amer. Chem.
Soc. 77:767-768, 1955.
744. Tasseron, J. G., H. Diringer, H. Frohwirth, S. S. Mirvish, and C. Heidelberger.
Partial purification of the soluble protein from mouse skin to which carcino-
genic hydrocarbons are specifically bound. Biochemistry 9:1636-1644, 1970.
745. Tebbens, B. D., J. F- Thomas, and M. Mukai. Fate of arenes incorporated
with airborne soot. Amer. Ind. Hyg. Assoc. J. 27:415-422, 1966.
746. Tentative method of analysis for polynuclear aromatic hydrocarbon content
of atmospheric particulate matter. Health Lab. Sci. 7 (Suppl.):31-44, 1970.
747. Tentative method of analysis for suspended particulate matter in the atmo.-
sphere: (high-volume method). Health Lab. Sci. 7:279-286, 1970.
748. Tentative method of chromatographic analysis for benzo[a]pyrene and
benzo[k] fluoranthene in atmospheric particulate matter. Health Lab.
Sci. 7 (Suppl.):60-67, 1970.
749. Tentative method of microanalysis for benzo[ a] pyrene in airborne particulates
and source effluents. Health Lab. Sci. 7 (Suppl.): 56-59, 1970.
750. Tentative method of routine analysis for polynuclear aromatic hydrocarbon
content of atmospheric particulate matter. Health Lab. Sci. 7 (Suppl ):
45-55, 1970.
751. Thomas, J. F., M. Mukai, and B. D. Tebbens. Fate of airborne benzo[a] pyrene.
Environ. Sci. Tech. 2:33-39, 1968.
752. Thomas, J. F., B. D. Tebbens, E. N. Sanborn, and J. M. Cripps. Fluorescent
-------
References 349
spectra of aromatic hydrocarbons found in polluted atmosphere. Int. J. Air
Pollut. 2:210-220, 1960.
753. Tipson, R. S. Review of Oxidation of Polycyclic, Aromatic Hydrocarbons.
National Bureau of Standards Report 8363. Washington, D.C.: U.S. Govem-
. ment Printing Office, 1964. 89 pp.
754. Toth, B. A critical review of experiments in chemical carcinogenesis using
newborn animals. Cancer Res. 28:727-738, 1968.
755. Toth, B., P. N. Magee, and P. Shubik. Carcinogenesis study with dimethyl-
nitrosamine adminstered orally to adult and subcutaneously to newborn
BALB/c mice. Cancer Res. 24:1712-1719, 1964.
756. Toth, L. Spektralfluorometrische in situ-analyse polycyclisher aromaten
nach trennung auf acetylierten •celluloseschichten. I. Mit qualitative und
quantitative auswertung. J. Chromatogr. 50:72-82, 1970. (summary in English)
757. Tsai, S. C., and G. W. Robinson. Phosphorescence and the true lifetime of
triplet states in fluid solutions. J. Chem. Phys. 49:3148-3191, 1968.
758. Ts'o, P. O. P., S. A. Lesco, and R. S. Umans. The physical binding and the
chemical linkage of benzpyrene to nucleotides, nucleic acids, and nucleo-
histones, pp. 106-135. In Jerusalem Symposia on Quantum Chemistry
and Biochemistry. Vol. 1, 1969.
759. Twort, C. C., and J. M. Twort. The carcinogenic potency of mineral oils.
J. Ind. Hyg. 13:204-226, 1931.
760. Tye, R., and K. L. Stemmer. Experimental carcinogenesis of the lung. II.
Influence of phenols in the production of carcinoma. J. Nat. Cancer Inst.
39:175-186, 1967.
761. Ultraviolet Atlas of Organic Compounds. 5 vols. New York: Plenum Press,
1967-1968.
762. Ungnade, H. E., Ed. Organic Electronic Spectral Data. New York: Inter-
science Publishers, Inc., I960-.
763. United Nations. Report of the United Nations Scientific Committee on the
Effects of Atomic Radiation, p. 99. General Assembly; Official Records:
Twenty-first Session. Supplement 14 (A/6314). New York: United Nations,
1966.
764. U.S. Department of Agriculture, Forest Service, Division of Cooperative
Forest Fire Control. 1968 Wildfire Statistics. Washington, D.C.: U.S.
Department of Agriculture, 1969. 48 pp.
765. U.S. Department of Commerce, Census Bureau. Estimates of Population
of Selected Standard Metropolitan Statistical Areas. July 1, 1962; July 1,
1963. Washington, D.C.: U.S. Department of Commerce, 1964. 8 pp.;6 pp.
766. U.S. Department of Health, Education, and Welfare. Mutagenicity of pes-
ticides, pp. 565-653. In Report of the Secretary's Commission on Pesticides
and Their Relationship to Environmental Health. Parts I and II. Washington,
P.C.: U.S. Government Printing Office, 1969.
767. U.S. Department of Health, Education, and Welfare. Nationwide Inventory
of Air Pollutant Emissions. 1968. National Air Pollution Control Adminis-
tration Publication AP-73. Washington, D.C.: U.S. Government Printing
Office, 1970. 36 pp.
768. U.S. Department of Health, Education, and Welfare. Preliminary Air Pol-
lution Survey of Organic Carcinogens. A Literature Review. National Air
-------
350 References
Pollution Control Administration Publication APTD 69-43. Raleigh, N.C.:
U.S. Department of Health, Education, and Welfare, 1969. 117 pp.
769. U.S. Department of Health, Education, and Welfare. Smoking and Health.
Report of the Advisory Committee to the Surgeon General of the Public
Health Service. Public Health Service Publication 1103. Washington, D.C.:
U.S. Government Printing Office, 1964. 387 pp.
770. U.S. Department of Health, Education, and Welfare. Teratogenicity of
pesticides, pp. 655-677. In Report of the Secretary's Commission on
Pesticides and Their Relationship to Environmental Health. Parts I and II.
Washington, D.C.: U.S. Government Printing Office, 1969.
771. U.S. Department of Health, Education, and Welfare. The Health Consequences
of Smoking: 1969 Supplement to the 1967 Public Health Service Review.
Washington, D.C.: U.S. Government Printing Office, 1969. 98 pp.
772. U.S. Department of Health, Education, and Welfare. The Health Consequences
of Smoking, a Public Health Service Review: 1967. Public Health Service
Publication 1696, Revised 1968. Washington, D.C.: U.S. Government Print-
ing Office, 1968.227pp.
773. U.S. Department of Health, Education, and Welfare, National Air Pollution
Control Administration. Air Quality Data from the National Air Sampling
Networks and Contributing State and Local Networks. 1966 Edition.
NAPCA Publication APTD 68-9. Durham, N.C.: U.S. Department of Health,
Education, and Welfare, 1968. 157 pp.
774. U.S. Department of Health, Education, and Welfare, Public Health Service,
Consumer Protection and Environmental Health Service, Environmental
Control Adminstration, Bureau of Solid Waste Management. Technical-
Economic Study of Solid Waste Disposal Needs and Practices. Public Health
Service Publication 1886. Washington, D.C.: U.S. Department of Health,
Education, and Welfare, 1969. 4 pp.
775. U.S. Department of Health, Education, and Welfare, Public Health Service,
National Air Pollution Control Administration. Air Quality Criteria for
Particulate Matter, NAPCA Publication AP-49. Washington, D.C.: U.S.
Government Printing Office, 1969. 211 pp.
776. U.S. Department of Labor, Labor Statistics Bureau. Current Population
Survey. Washington, D.C.: U.S. Government Printing Office, 1964. 18 pp.
777. U.S. Department of the Interior, Bureau of Mines. Automobile Disposal,
a National Problem. Washington, D.C.: U.S. Government Printing Office,
1967.569pp.
778. Uytdenhoef. Quelques considerations sur les dermatoses professionnelles.
Arch. Med. Soc. Hyg. 2:830-847, 1939.
779. Vadova, A. V., and V. I. Gel'shtein. Spontaneous tumours in catarrhine
monkeys according to the data obtained in the monkey colony of the Sukhumi
medico-biological station, pp. 137-158. In I. A. Utkin, Ed. Theoretical and
Practical Problems of Medicine and Biology in Experiments on Monkeys.
Translated by Ruth Schacter. New York: Pergamon Press Inc., 1960.
780. Van Duuren, B. L. Carcinogenic epoxides, lactones, and halo-ethers and
their mode of action. Ann. N.Y. Acad. Sci. 163:633-651, 1969.
781. Van Duuren, B. L. Effects of the environment on the fluorescence of aro-
matic compounds in solution. Chem. Rev. 63:325-354, 1963.
-------
References 351
782. Van Duuren, B. L. The fluorescence spectra of aromatic hydrocarbons
and heterocyclic aromatic compounds. Anal. Chem. 32:1436-1442, 1960.
783. Van Duuren, B. L. The polynuclear aromatic hydrocarbons in cigarette-
smoke condensate. II. J. Nat. Cancer Inst. 21:623-630, 1958.
784. Van Duuren, B. L. Tumor-promoting agents in two-stage carcinogenesis.
Prog. Exp. Tumor Res. 11:31-68, 1969.
785. Van Duuren, B. L., I. Bekersky, and M. Lefar. The peracid oxidation of
dibenz[a,h] anthracene. J. Org. Chem. 29:686-689, 1964.
786. Van Duuren, B. L., J. A. Bilbao, and C. A. Joseph. The carcinogenic nitrogen
heterocyclics in cigarette-smoke condensate. J. Nat. Cancer Inst. 25:53-61,
1960.
787. Van Duuren, B. L., and S. Melchionne. Inhibition of tumorigenesis. Prog.
Exp. Tumor Res. 12:55-94, 1969.
788. Van Duuren, B. L., and L. Orris. The tumor-enhancing principles of Croton
tiglium L. Cancer Res. 25:1871-1875, 1965.
789. Van Duuren, B. L., A. Sivak, B. M. Goldschmidt, C. Katz, and S. Melchionne.
Initiating activity of aromatic hydrocarbons in two-stage carcinogenesis. J.
Nat. Cancer Inst. 44:1167-1173, 1970.
790. Van Duuren, B. L., A. Sivak, and L. Langseth. The tumor-promoting activity
of tobacco leaf extract and whole cigarette tar. Brit. J. Cancer 21:460-463,
1967.
791. Van Duuren, B. L., A. Sivak, L. Langseth, B. M. Goldschmidt, and A. Segal.
Initiators and promoters in tobacco carcinogenesis. Nat. Cancer Inst.
Monogr. 28:173-180, 1968.
792. Van Duuren, B. L., A. Sivak, A. Segal, L. Orris, and L. Langseth. The tumor-
promoting agents of tobacco leaf and tobacco smoke condensate. J. Nat.
Cancer Inst. 37:519-526, 1966.
793. Varma, P. S., and J. L. Das Cupta. Comparative study of the preparation of
anthraquinone. Q. J. Indian Chem. Soc. 4:297-298, 1927.
794. Vies, F., A. de Coulon, and A. Ugo. Recherches sur les proprietes physico-
chimiques des tissus en relation avec l'6tat normal ou pathologique de 1'or-
ganisme. XXI. Influence de 1'obscuritfe et de la luminere sur la cancerisa-
tion parle goudron. Arch. Phys. Biol. 12:255-277, 1935.
795. Vollmann, H., H. Becker, M. Corell, and H. Streeck. Beitrage zur Kenntnis
des Pyrens und seiner Derivate. Justus Liebigs Ann. Chem. 531:1-159, 1937.
796. Vorwald, A. J. Medical aspects of beryllium disease, pp. 167-200. In H. E.
Stokinger, Ed. Beryllium. Its Industrial Hygiene Aspects. New York:
Academic Press Inc., 1966.
797. Voyatzakis, E., D. Jannakoudakis, T. Dorfmuller, C. Sipitanos, and G. Stalidis.
Action de la lumiere ultraviolette sur les hydrocarbures polybenzenique
adsorbes. Anthracene, naphthalene. C. R. Acad. Sci. (Paris) 251:2696-2707,
1960.
798. Wadleigh, C. H. Wastes in Relation to Agriculture and Forestry. Miscellaneous
Publication 1065. Washington, D.C.: U.S. Department of Agriculture, 1968.
112pp.
799. Walburg, H. E., Jr., G. E. Cosgrove, and A. C. Upton. Influence of microbial
environment on development of myeloid leukemia in x-irradiated RFM mice.
Int. J. Cancer 3:150-154, 1968.
-------
352 References
800, Waller, R. E., and B. T. Commins. Studies of the smoke and polycyclic
aromatic hydrocarbon content of the air in large urban areas. Environ. Res.
1:295-306, 1967.
801. Waltz, P., and M. Hausermann, Sur un traitment du tabac en vue de diminuer
la teneur en hydrocarbures polycycliques de la fumee de cigarettes. A. Praeven-
tivmed. 8:111-124, 1963.
802. Wanta, R. C. Meteorology and air pollution, pp. 187-226. In A. C. Stern,
Ed. Air Pollution. Vol. 1. Air Pollution and Its Effects. (2nd ed.) New
York: Academic Press Inc., 1968.
803. Wasserman, E., V. J. Kuck, W. M. Delavan, and W. A. Yager. Electron para-
magnetic resonance of 'A oxygen produced by gas-phase photosensitization
with naphthalene derivatives. J. Amer. Chem. Soc. 91:1040-1041, 1969.
804. Wasserman, E., R. W. Murray, M. L. Kaplan, and W. A. Yager. Electron
paramagnetic resonance of 'A oxygen from a phosphite-ozone complex.
J. Amer. Chem. Soc. 90:4160-4161, 1968.
805. Wasserman, H. H., and J. R. Scheffer. Singlet oxygen reactions from photo-
peroxides. J. Amer. Chem. Soc. 89:3073-3075, 1967.
806. Watson, J. T. Gas chromatography and mass spectroscopy, pp. 145-225.
In L. S. Ettre and W. H. McFadden, Eds. Ancillary Techniques of Gas
Chromatography. New York: John Wiley & Sons, Inc., 1969.
807. Wattenberg, L. W., and J. L. Leong. Inhibition of the carcinogenic action of
benzo(a)pyrene by flavones. Cancer Res. 30:1922-1925, 1970.
808. Wattenberg, L. W., and J. L. Leong. Inhibition of the carcinogenic action of
7,12-dimethylbenz(a)anthracene by beta-naphthoflavone. Proc. Soc. Exp.
Biol. Med. 128:940-943, 1968.
809. Weber, G. Enumeration of components in complex systems by fluorescence
spectrophotometry. Nature 190:27-29, 1961.
810. Wehry, E. L. Structural and environmental factors in fluorescence, pp.
37-132. In G. G. Guilbault, Ed. Fluorescence: Theory, Instrumentation, and
Practice. New York: Marcel Dekker, Inc., 1967.
811. Weinberg, N. L., and H. R. Weinberg. Electrochemical oxidation of organic
compounds. Chem. Rev. 68:449-523, 1968.
812. Weisburger, J. H., and E. K. Weisburger. Tests for chemical carcinogens.
Methods Cancer Res. 1:307-398, 1967.
813. Welch, R. M., Y. E. Harrison, A. H. Conney, P. J. Poppers, and M. Finster.
Cigarette smoking: Stimulatory effect on metabolism of 3,4-benzpyrene by
enzymes in human placenta. Science 160:541-542, 1968.
814. Went, F. W. Organic matter in the atmosphere, and its possible relation to
petroleum formation. Proc. Nat. Acad. Sci. U.S.A. 46:212-221, 1960.
815. Whitby, K. T., R. S. Husar, and B. Y. H. Liu. The Aerosol Spectra of Los
Angeles Smog. Presented to the American Chemical Society Kendall Award
Symposium, Los Angeles, March 29, 1971.
816. Whiting, M. C., A. J. N. Bolt, and J. H. Parish. The reaction between ozone
and saturated compounds. Adv. Chem. Ser. 77:4-14, 1968.
817. Wilk, M., W. Bez, and J. Rochlitz. Neue Reaktion der Carcinogen Kohlenwasser-
stoffe 3,4-Benzpyren, 9,10-Dimethyl-l,2-Benzanthracen und 20-Methylcholan-
thren. Tetrahedron 22: 2599-2608, 1966. (abstract in English)
-------
References 353
818. Wilk, M., and W. Girke. Radical cations of carcinogenic alternant hydrocarbons,
amines, and azo dyes, and their reactions with nucleobases, pp. 91-105. In
E. D. Bergman and B. Pullman, Eds. Physico-chemical Mechanisms of Car-
cinogenesis. Proceedings of an International Symposium Held in Jerusalem,
21-25 October, 1968. Jerusalem: Israel Academy of Sciences and Humanities,
1969.
819. Wilk, M., J. Rochlitz, and H. Bende. Saulenchromatographie von polycyclischen
aromatischen Kohlenwasserstoffen an lipophilem Sephadex LH-20. J. Chroma-
togr. 24:414-416, 1966.
820. Willmer, E. N., Ed. Cells and Tissues in Culture: Methods, Biology, and
Physiology. 3 vols. London: Academic Press Inc., 1965-1966. 2423 pp.
821. Wilmshurst, J. R. Gas chromatographic analysis of polynuclear arenes. J.
Chromatogr. 17:50-59, 1965.
822. Windsor, M. W. Luminescence and energy transfer, pp. 345-431. In D. Fox,
M. M. Labes, and A. Weissberger, Eds. Physics and Chemistry of the Organic
Solid State. Vol. 2. New York: Interscience Publishers, Inc., 1965.
823. Winefordner, J. D. Time-resolved phosphorimetry. Accounts Chem. Res.
2:361-367, 1969.
824. Winkelstein, W., Jr., and S. Kantor. Prostatic cancer: Relationship to suspended
particulate air pollution. Amer. J. Public Health 59:1134-1138, 1969.
825. Winkelstein, W., and S. Kantor. Stomach cancer. Positive association with
suspended particulate air pollution. Arch. Environ. Health 18:544-547, 1960.
825a. Winkelstein, W., S. Kantor, E. W. Davis, C. S. Maneri, and W. E. Mosher. The
relationship of air pollution and economic status to total mortality and selected
respiratory system mortality in men. I. Suspended particulates. Arch. Environ.
Health 14:162-171, 1962.
826. Wolff, E. Embryogenesis in vitro, pp. 531-589. In E. N. Willmer, Ed.
Cells and Tissues in Culture: Methods, Biology, and Physiology. Vol. 1.
London: Academic Press Inc., 1965.
827. Wynder, E. L., and D. Hoffmann. Present status of laboratory studies on
tobacco carcinogenesis. Acta Path. Microbiol. Scand. 52:119-132, 1961.
828. Wynder, E. L., and D. Hoffmann. Selected laboratory methods in tobacco
carcinogenesis. Methods Cancer Res. 4:3-52, 1968.
829. Wynder, E. L., and D. Hoffmann. Some laboratory and epidemiological
aspects of air pollution carcinogenesis. J. Air Pollut. Control Assoc. 15:
155-159, 1965.
830.Wynder, E. L., and D. Hoffmann. The epidermis and the respiratory tract
as bioassay systems in tobacco carcinogenesis. Brit. J. Cancer 24:
574-582, 1970.
831. Wynder, E. L., and D. Hoffmann, Eds. Tobacco and Tobacco Smoke. Studies
in Experimental Carcinogenesis. New York: Academic Press Inc.,
1968.730pp.
832. Wynder, E. L., J. Kmet, N. Dungal, and M. Segi. An epidemiological investi-
gation of gastric cancer. Cancer 16:1461-1496, 1963.
833. Yasuhira, K. Experimental induction of lung cancer in rat and mouse with
20-methylcholanthrene in Freund's adjuvant. Acta Path. Jap. 17:475-493,
1967.
-------
354 References
834. Yuile, C. L., H. L. Berke, and T. Hull. Lung cancer following polonium 210
inhalation in rats. Radiat. Res. 31:760-774, 1967.
835. Zander, M. Phosphorimetry. The Application of Phosphorescence to the
Analysis of Organic Compounds. Translated by T. H. Goodwin. New York:
Academic Press Inc., 1968. 206 pp.
836. Zweig, A. Electron transfer luminescence in solution. Adv. Photochem. 6:
425-451, 1968.
-------
Index
Aerosol concentration
of benzanthrone, 48
of benzofaj pyrene, 46-49
by geographic areas, 47
particle size distribution and (See
Particle size distribution)
seasonal observations of, 46-49
urban short-term survey of, 48-50
Aerosol dynamics
aging of urban aerosols, 56, 57
collision processes, 54
removal from atmosphere, 55-57
aging and, 56
mechanisms for, 55
rain cloud role in, 55, 57
Aerosol sources
anthropogenic aerosols, 36, 37
natural aerosols, 36, 37
Ail sampling. See Collection of air sample;
Separation methods; Spectral analysis
Animal data extrapolation
in carcinogenesis tests
in vitro cultures and, 94
in post hoc epidemiologic studies, 93
of primates, 94
test group size and, 93
problems in lung cancer studies (See
Lung cancer)
Aryl hydrocarbon hydroxylase, 138-141
inducibility of, 138
in vitro activity of, 139
role in carcinogenesis, 140
Atmospheric physics. See Physics, atmo-
spheric
Automotive emissions. See Transportation
sources
Aza-arenes
carcinogenicity of, 5,12
chemical reactivity of, 63-81
one-electron oxidation, 75
ozone reactivity, 74
photooxidation in solution, 66-69
nomenclature of, 4, 5
structure of, 4, 5,12
Bronchitis, chronic, and POM, 184
Cancer. See Lung cancer; Neoplastic disease
Carcinogenesis, chemical (See also In vitro
approaches; Lung, exposure in vivo)
chemical carcinogens, 82-86
benzo[a] pyrene, 82
355
-------
356
Index
experimental tests of (See Caicinogenesis
tests, experimental design)
historical aspects of, 82
mechanisms of, 82-86,159
cellular, 83
molecular (See Molecular mechanisms)
and mutagenesis, 157-159
mutagenicity testing of carcinogens,
158,159
oncogenic viruses and, 83,142
transplantation antigens and, 85,144
Caicinogenesis tests. See Animal data
extrapolation; Caicinogenesis tests,
experimental design; Host factors;//!
vitro approaches;In vivo tests; Poly-
cyclic aromatic hydrocarbons
Carcinogenesis tests, experimental design of
analysis of, 89, 91
dosage of agent, 87-90
administration route, 89
low dosage, 90
maximal tolerated dosage, 89
range of, 88
"safe" dosage, 87
threshold dose (See Threshold dose)
general-purpose design in, 88,90
reference tumor rate in, 88
Carcinogenicity. See also Carcinogenesis
tests
of aza-arenes, 12
of polycyclic aromatic hydrocarbons, 6-11
Chemical reactivity of POM, 63-81, 239
general reactivity, 64
with nitrogen oxides, 79
with oxidants, 74-79
one-electron oxidation, 74
peroxides, 77
radicals, 78
with ozone (See Ozone reactivity)
photodynamic activity, 80
and singlet oxygen, 80
photooxidation (See Photooxidation)
with sulfur oxides, 80
Chromatography methods. See Separation
methods
Chronic bronchitis, and POM, 184
Cigarette-consumption, regression studies
of, in lung cancer, 222-226
Clinical studies. See Neoplastic disease:
Pulmonary disease; Skin effects
Cocarcinogenesis
and anticarcinogenic agents, 106
definition of, 102
two-stage Carcinogenesis studies, 103-106
on mouse skin, 103,104
persistence of initiating effect, 104,105
Collection of air samples
equipment selection, 255
filter selection, 255
filtration of sample, 256
particle size and, 257
sample size and, 256
sampling site selection, 254
standardization of sampling, 258
Andersen sampler and, 258
lack of, 258
particle diameters and, 259
storage of sample, 257
Combustion emissions: See Emissions;
Sources of POM
Cutaneous effects of POM. See Skin
disorders; Skin effects
Detection of POM. See Spectral analysis
Disease, POM and
chronic bronchitis, 184
emphysema, 184
lung cancer (See Neoplastic disease)
skin disorders (See Skin disorders)
Dispersion of POM, atmospheric
larger-scale diffusion, 58
microscale diffusion, 57
Distribution of hydrocarbons
fluorescence studies of, 132
host factors in, 100
ciliary movement changes, 101
elution of particles, 100
mucous viscosity changes, 101
particle size, 100
retention of particles, 100
radioactivity distribution studies of,
132-134
Electrophoresis methods. See Separation
methods
Emissions of POM. See also Sources of POM
area-concentration relations of, 31-33
areas of major emissions, 32, 33
relative area contributions of
benzo[a] pyrene, 32, 33
individual emissions, 31, 32
general nature of, 31
ratios of as a function of source, 31, 32
types of, 31
Emphysema, POM and, 184
-------
Index
357
Epidemiologic studies. See Lung cancer
mortality rates
Excretion of polycyclic hydrocarbons
fluorescence studies of, 132
radioactivity studies of, 132-134
Formation of POM, mechanism of, 13-15
Health data
assessment, problems in, 168-171
ethnic factors, 171
incidence rates, interpretation of, 170
measurements, adequacy of, 170
lung cancer characteristics, 168
socioeconomic factors, 171
urban-rural differences, 171
sources of, 168
epidemiologic studies, 169
in vitro methods, 169
lab animals, 168
Historical review
of chemical carcinogenesis, 82
of human effects of POM, 166
Host factors, modification of in in vivo tests
age, 121-123
and hormonal status, 121
immune status of host, 118-121
immunological surveillance mechanism,
evidence of, 119-121
nutrition, 123-125
role of caloric intake, 124
physical interaction with carcinogens,
127-131
of ionizing radiation, 127-131
respiratory infection, 125-127
Human effects of POM. See Disease, POM and
Hydrocarbons. See Polycyclic aromatic
hydrocarbons
Identification of POM, in air samples. See
Spectral analysis
Industrial sources of POM emissions. See
Stationary sources of POM
In vitro approaches to carcinogenesis
in chemical carcinogenesis, 142-145
cell transformation in, 142-144
cellular mechanisms of carcino-
genesis in, 144
oncogenic viruses in, 142
quantitative system in, 143
surface transplantation antigens in, 144
in organ cultures, 145-147
compared with in vivo methods,
146,147
dose-response analysis, 145
metabolism of chemicals in, 145
technique in, 145
value of, 146
In vivo carcinogenesis tests. See also Lung,
exposure in vivo
for cocarcinogenesis (See Cocarcinogenesis)
host factors, modification of in (See Host
factors)
in mice and rats, 95-102
by bladder implantation, 100
distribution of hydrocarbons in host
(See Distribution of hydrocarbons)
by inhalation, 99
by oral administration, 99
polycyclic carcinogens in, 95
by skin application (See Mouse skin)
by subcutaneous administration, 98
in primates, 114
of pulmonary tissues, 115
of skin and subcutaneous tissues, 114
in zoo animals outdoors, 116
necropsy reports, 117
Lifetime of POM, atmospheric, 59
Luminescence spectrophotometry. See
Spectral analysis
Lung, exposure in vivo
to carcinogens with gaseous pollutants, 113
by inhalation of crude material, 110
by intratracheal instillation of carcinogens,
111-113
benzo(a]pyrene, 112
benzo[a]pyrene on hematite, 111
7,12-dimethylbenz[a]anthracene, 111
3-methylcholanthrene, 112
to ozonized gasoline vapor, 111
Lung cancer. See also Pulmonary disease
animal data extrapolation problems in,
177-182
of dose-response relations, 177-182
external factors of, 177
in smoking experiments, 178
in squamous cell carcinoma, 179-181
in viral infection, 181
determinants of tissue particle concentra-
tion in, 174-176
ciliary transport mechanisms, 175
clearance of inhaled particles, 175
dose, effective, 174
"leaching" process, 176
-------
358
Index
particle size distribution, 174
incidence of, 168 {Seealso Urban-rural
studies)
model of inhalation carcinogenesis, 173,
174
modifiers of lung reaction, 176
honeycombing, 176
mortality (See Lung cancer mortality rates)
problems for investigation, 183
factors in experiment design, 183
Lung cancer mortality rates
demographic-unit regression studies of,
221-226
by age-race-sex, 224-226
in cigarette consumption, 222-226
limitations of method in, 222
of males in 19 countries, 222-224
regression analysis in, 304-306
in solid-fuel consumption, 222-226
dose-response curve relating benzo[a]-
pyrene to, 202-205
indices of, 209-213
by age 35 and over, 211
by age-race-sex, 210
by race and sex, 211
migrant studies of (See Migrant studies)
rates, types of, 209-213
sampling studies of, 226-233
and benzo[ajpyrene concentration, 227
by city of residence, 226
by city size, 227, 229
by duration of residence, 229, 231
by level of pollution, 231-233
by smoking category, 227-232
urban-rural differences in, 226-229, 231
standardized mortality ratios, 212
urban-rural studies of (See Urban-rural
studies)
Man, POM effects on, 245. See also Disease,
POM and
Metabolism of polycyclic aromatic hydro-
carbons
aryl hydrocarbon hydroxylase in (See
Aryl hydrocarbon hydroxylase)
of benzo[a]pyrene, 135
metabolites of, 135
covalent binding findings and, 136
chemical reactivity of carcinogens and,
136
in culture cells, 134
metabolically activated carcinogenic forms,
137
metabolic intermediates, 136
in mouse skin, 134
Migrant studies, of lung cancer mortality
in Australia, 219, 220
of British migrants, 219, 220
in New Zealand, 219
of United Kingdom migrants, 219
in South Africa, 219, 220
of white British male migrants, 219, 220
in United States, 218, 219
of British migrants, 221
of immigrant males, 218, 219
of Norwegian migrants, 221
Molecular mechanisms of chemical carcino-
genesis, theoretical
Mouse skin, as test organ
concentration of pollutants used on, 97
mouse strains used for, 96
observation of, 98
solvent used for, 97
Mutagenesis
and carcinogenesis, 157-159
mechanisms of carcinogenesis (See
Molecular mechanisms)
mutagenicity testing (See Mutagenicity
testing)
mutations, classification of, 158
Mutagenicity testing of carcinogens
point mutations and, 156,159
in vitro methods, 156
relative mutagenicity of compounds,
157
in vivo methods, 152-156,158,159
dominant lethal assay, 153
host-mediated assay, 154,159
in vivo cytogenetics, 154
Mutational mechanisms, 83,159
plasmagene theory, 159
somatic-mutation theory, 83,159
Neoplastic disease. See also Lung cancer;
Pulmonary disease; Skin effects
Nonpulmonary disease, nonoccupational,
235
bladder cancer, 235
esophageal cancer, 235, 236
intestinal cancer, 236
prostate cancer, 235, 236
rectal cancer, 235, 236
socioeconomic class and, 235, 236
stomach cancer, 235, 236
-------
Index
359
Nomenclature
of aza-arenes, 4, 5,12
of polycyclic aromatic hydrocarbons, 4-11
rules of, 4,5
Nonmutational mechanisms, 83-85
Occupational effects, of POM exposure. See
Pulmonary disease; Skin effects
Olefin hydrocarbons
0-lactones, 108
diketene formation, 108
olefin oxidation, 107
Ozone reactivity
with polycyclic aromatic hydrocarbons,
70-74
with polycyclic aza-arenes, 74
Particle size distribution
in automobile exhaust, 39
evaluation of, 39
measurement techniques for, 40
particle size distribution moments, 45-52
correlations with aerosol spectra, 50,
51
typical values of, 46,47
POM and, 52
from stationary sources, 40
in urban atmospheres, 40-53
variations in, 43
Particles of POM. See Particle size distribu-
tion; Physical properties of POM
Photodynamic activity of POM, 80
Photodynamic assay
as biologic index of potential carcino-
genicity, 150
in differentiation of pollutants, 150
method in, 149
Photooxidation of POM
of adsorbed aromatic hydrocarbons, 69
in solution, 66-69
Physical properties of POM
adsorption, 52
in aerosols (See Aerosols)
elution, 52
particle density, 43-4-5
effective density, 45
particle shape, 43-45
dynamic shape factor, 44
surface shape factor, 44
volume shape factor, 44
particle size distribution (See Particle
size distribution)
particle weight, 44
Physics of POM, atmospheric. See also
Aerosol concentration; Aerosol dynamics;
Physical properties of POM
areas of uncertainty in, 60
dispersion of POM, 57-59
lifetime, atmospheric, 59
Plants. See Vegetation
Polycyclic aromatic hydrocarbons. See also
Chemical reactivity; Distribution;
Excretion; Metabolism; Vegetation
carcinogenicity of, 5-11
nomenclature of, 4-11
oxidation products of, 108
peroxy compounds of benz[a] -
anthracene, 109
structure of, 4-11
tests for potential carcinogenicity of (See
Photodynamic assay; Sebaceous gland
suppression test)
Pulmonary disease, occupational
dose-response relations, 201-205
dose-response curve, 202
Pulmonary effects of POM, nonoccupa-
tional. See Lung cancer mortality rates
Quantification of POM. See Spectral analysis
Regression analysis, 304-306
Regression studies, in lung cancer, 221-226.
See also Lung cancer mortality rates
Research recommendations
in animal studies, 248
on atmospheric physics of POM, 247
on chemical reactivity of POM, 247
in epidemiologic studies, 250
in mammalian cell studies, 248
on sources of POM, 247
in vegetation studies, 248
Sampling of air. See Collection of air
sample; Separation methods
Sebaceous gland suppression test
of cigarette-smoke condensate, 148
and long-term carcinogenesis testing, 148
reliability of, 149
Separation methods for POM
column chromatography, 264-266
alumina in, 265
cellulose in, 266
florisil in, 266
silica gel in, 266
standard conditions in, 264
combination of methods, 274
-------
360
Index
distribution of sample, 263
aza-arene extraction, basic, 264
countercurrent distribution, 264
between solvents, 263
gas chromatography, 270-274, 300
application to POM of, 270
columns in, 272
detectors in, 271
future prospects in, 273
and mass spectrometry coupled, 300
operating principle of, 270
gel filtration, 266
laboratory standards, 262
internal standards, 262
precautions, 262
reproducibility factors, 263
liquid chromatography, high-speed, 268
paper chromatography, 267
paper electrophoresis, 268
thin-layer chromatography, 267
thin-layer electrophoresis, 268
Skin disorders, 185-201
cell-mediated hypersensitivity, 186
cutaneous photosensitization, 187
nonallergic dermatitis, 185
clinical characteristics of, 186
pigment disturbances, 190
pilosebaceous reactions, 188-190
from chlorinated hydrocarbons, 189
from coal-tar products, 189
pathogenesis and pathology of, 189
from petroleum and derivatives, 188
Skin effects of POM. See also Skin disorders
from asphalt exposure, 192
influencing factors on, 197
chemical cocarcinogens, 198
ultraviolet radiation, 199
from oil fractionation and distillation
products, 193-197
anthracene, 197
creosote, 196
cutting oils, 196
lubricating oils, 196
in mule spinning, 194
in petroleum refining, 195
soot, 197
from pitch exposure, 192
from tar exposure, 192
Solid-fuel consumption, regression studies,
in lung cancer, 222-226
Sources of POM, by combustion
emissions of POM, nature of (See Emissions)
nontechnologic sources, 15
forest fires, 15
POM formation, mechanism of, 13-15
of benzo[a]pyrene, 14
technologic sources
stationary sources (See Stationary
sources)
transportation sources (See
Transportation sources)
Spectral analysis, air sample
infrared spectroscopy, 197
luminescence spectrophotometry, 285-297
and absorption spectrophotometry, 296
analysis of luminescence, 290-292
complications in, 294-296
correction of spectra, 293
instrumentation of, 286-288
limitations in, 294-296
quenching techniques of, 293
sample preparation in, 288-290
sample purity evaluation, 293
sensitization methods, 292
mass spectrometry, 298-301
application to POM analysis, 299
and gas chromatography coupled, 300
operating principle of, 299
new spectroscopic techniques, 302
correlation spectroscopy, 302
electron spectroscopy, 302
Fourier transform spectroscopy, 302
Raman spectroscopy, 297
ultraviolet absorption spectroscopy,
277-285
of benzo[a]pyrene, 279, 280
interpretation of spectra in, 283-285
spectral characterisitics of POM, 278
spectrophotometric techniques, 282,
296
Stationary sources of POM
emission control procedures for, 29-31
heat generation, 23
benzo[a]pyrene emission from, 23
indoor POM emission sources, 28
garages, 29
offices, 29
tobacco-smoking, 29
industrial POM emission sources, 26-28
asphalt air-blowing, 26
benzo[a]pyrene emission from, 27, 28
catalytic cracking of petroleum, 26
coke production, 28
power generation, 23
benzo[a]pyrene emission from, 23
refuse burning, 24
benzo[a]pyrene emission from, 25
-------
Index
361
Structure, chemical
of aza-arenes, 4,5,12
structure diagrams, 12
of polycyclic aromatic hydrocarbons, 4-11
structure diagrams, 6-11
Teratogenesis
definition of, 151
human teratogens, 151
teratogenicity testing, 151
characteristics considered in, 152
Tests. See Carcinogenesis tests; In vitro
approaches;//! vivo tests; Mutagenicity
testing; Teratogenesis
Threshold dose, in carcinogenesis testing
dose-response studies, 91
dose-response curve, 91
existence of, 90
and group size, 92
and low dosage studies, 92
Transportation sources of POM
diesel-fuel-powered vehicles, 20
emission control procedures for, 21
gasoline-powered vehicles, 15-22
automotive emission factors, 17
benzo[a] pyrene emission by, 16
effects of fuel composition, 18-20
effects of vehicular characteristics,
16-18
miscellaneous sources, 21
Urban-rural studies
of lung cancer incidence, 214, 216, 217
by residence, 214, 216, 217
by sex and cancer site, 216, 217
of lung cancer mortality, 213-218
by economic status, 215
by pollution level, 215
by race, 213, 214
by residence of white male smokers,
217,218
by sex, 213, 214
standardized mortality ratios, 213-215
regional characterization in, 211-213
frequency distribution of benzo[a] -
pyrene concentrations, 207-208, 227
Standard Metropolitan Statistical Area,
211
urban-rural definitions, 212
urban size changes, 213
Vegetation
biosynthesis of POM in, 162
and carcinogens, 161-165
in foods, 163
in pesticides, 164
in polluted atmosphere, 161,165
in soil, 161
in tobaccos, 163
in tobacco smoke, 163
effect of POM on, 161,165
POM in, 160
------- |