Proliferative  Hepatocellular
                 Lesions of the Rat:

             Review and Future  Use in
                   Risk Assessment
                          Prepared for the
                       Risk Assessment Forum
                 U.S. Environmental Protection Agency
                          Washington, DC

                          February 1986
Esther Rinde, Ph.D.
Richard Hill, M.D., Ph.D.

)ffice of Pesticides and
 Toxic Substances
   Authors
      Arthur Chiu, M.D., Ph.D.
      Bernard Haberman, D.V.M., M.S.
      Office of Research and
        Development
|arry Anderson, Ph.D. (ODW)*
 lary Argus, Ph.D. (OPTS)
  \i\n Baumel, Ph.D. (OPTS)*
Judith Bellin, Ph.D. (OSWER)
Iteven Bayard, Ph.D. (ORD)
Jhao Chen, Ph.D. (ORD)
Irthur Chiu, M.D., Ph.D. (ORD)
Technical Panel
      David Cleverly, M.S. (OAR)
      Bernard Haberman, D.V.M., M.S. (ORD)*
      Richard Hill, M.D., Ph.D. (OPTS)
      Louis Kasza, Ph.D., D.V.M. (OPTS)
      Paul Milvy, Ph.D. (OPPE)
      Esther Rinde, Ph.D. (OPTS)
      Jackson Schad, M.S. (OSWER)
                    *Technical Panel Co-Chairman
                    Risk Assessment Forum Staff
           Dorothy Patton, Ph.D., J.D., Executive Director (Acting)
                Alan Ehrlich, Ph.D., Executive Secretary


          U.S. ENVIRONMENTAL PROTECTION AGENCY
                   WASHINGTON, DC 20460

-------
     IN MEMORY OF
DR. BERNARD HABERMAN

-------
Disclaimer

-------

-------
                              Contents
                                                                  Page
 List of Figures
 List of Tables

  I.  Introduction ....................................                  1
     A.  Background ....................................   ........... ^
     B.  Purpose and Scope .......................... .............     1
 II.  Chemical Hepatocarcinogenesis, A Model .....................      3
 III.  Characteristics of Proliferative Hepatocellular Lesions ........        5
     A.  Morphological Classification ....................... ........... .5
     B.  Background Incidence of Lesions in Untreated Animals. '.'.'.'.'.'.'.'.'. 7
     C.  Comparative Morphology ..............................  ..... 7
 IV.  Relationship Between Certain Hepatocellular Proliferative
     Lesions and Hepatocellular Carcinoma ...........................    9
     A.  Regression of Early Lesions .................... ............... B
         1 .  Foci ............................................ ......... g
         2.  Neoplastic Nodules ........................... ...........  10
     B.   Progression .................................. .............   1 1
         1 .  Foci ........................................ ...........  12
         2.  Neoplastic Nodules ................................. .....  12
 V.  Current Governmental Use of Data on Rat Hepatocellular Lesions ...  1 4
VI.  Future Use of Rat Hepatocellular Lesions by EPA ................    15
     A.   Rationale for Positions  ....................... " ' ] ] ..........  15
     B.   EPA Positions ....................... ............... ........  16
References ......................................                    1 o

-------
List of Figures
                                     Page
                                     ... 4
Number
  1.   Hypothetical scheme of neoplastic development in rat liver..
  2.   Terminology of rat liver lesions	5

-------
                          List of Tables
Number
                                                               Page
  1.   Proliferative Liver Lesions: Background Incidence in
      2-Year-Old Rats	8

  2.   Evidence for Progressions of Hepatocellular Lesions	11

-------

-------
                           I.  Introduction
 A.  Background

   When liver carcinogens are administered to rats, there is an increase in
 the frequency of certain morphological lesions and  often a chronology in
 their formation. Generally foci of cellular alteration are first to appear, then
 nepplastic nodules, and finally carcinomas are noted in the livers of the rats
 This time sequence suggests an interpretation of the carcinogenic process
 in which foci are a precursor stage in the development of neoplastic nod-
 ules, which  in turn are  precursors of carcinomas. However,  under  certain
 circumstances, if carcinogen administration ceases, many foci and neoplas-
 tic nodules  disappear, and the formation of carcinomas occurs at a low
 frequency. Thus, an alternative interpretation is that foci and neoplastic
 nodules may be independent of carcinoma development.
   Because of these general findings, it is difficult to evaluate  the  signifi-
 cance of increases in lesions in long-term tests in rats.  Since there is  no
 certain scientific knowledge or consensus about the biological significance
 of liver nodules, the EPA needs a policy for the interpretation of rat carcino-
 gen bioassays in which  liver nodules are an end point.
   Differences as to the way rat hepatocellular lesions might be used in risk
 assessment  arose within the Agency. In July 1984, the Office of Toxic Sub-
 stances submitted to the EPA Risk Assessment Forum a proposal to inves-
 tigate the nature and use of neoplastic nodules of the rat liver. Issues were
 formulated that dealt with the histological classification  of hepatocellular
 proliferative  lesions and the way these lesions would be used  for  hazard
 assessment  and dose-response  assessment. It was  agreed that the EPA
 should  have a clear position on these issues.
 B.  Purpose and Scope

  The classification of certain experimentally induced proliferative hepato-
 cellular lesions in the rat liver, other than carcinoma, has been the subject
 of great controversy. This paper addresses three questions that  are crucial
 in the evaluation of such rat tumor data and in the assessment of carcino-
 genic risk for humans. One question is whether to consider these lesions
 hyperplastic" or "neoplastic" in nature,  the latter classification being in-
 dicative of tumor formation.1 A second  question is to interpret the finding
that in some experimental systems, certain nonmalignant proliferative le-
sions may regress following removal of an inciting, carcinogenic stimulus
with only a small proportion of the lesions persisting. Finally, it is important
to know whether any of these lesions may progress to carcinoma.
  This paper begins with a composite view of chemically induced hepato-
carcmogenesis.  Proliferative hepatocellular lesions are  described  histo-
iDorland's Illustrated Medical Dictionary (1974) defines the term neoplasm as "... a mass of new
 nh^? 7 '  Per»'tf an,  9r°WS indePendent|y °f "s surrounding structures and which has no
 physologic use." Neoplasm has also been defined as ". . . an abnormal tissue that grows by
 cellular proliferation more rapidly than normal and continues to grow after the stimuli that
 initiated the new growth cease" (Stedman's Medical Dictionary, 1982). Dorland's defines hyper-
 plasia as  ... abnormal multiplication or increase in the number of normal cells in normal
 arrangement m tissue." Stedman's definition of hyperplasia is ". . . increase in the nCmS™or
 cells in  a tissue or organ, excluding tumor formatioq, where-by the bulk of the organ is in-
 crsssGu.
                                  1

-------
pathologically, and comparisons among species are made. Discussions are
developed on the regression of proliferative lesions following removal of a
carcinogenic stimulus, as well as their potential  for progression to car-
cinoma. Current uses of rat liver proliferative lesions in risk assessments by
governmental agencies are reviewed, and recommendations are made for
the future use of these lesions by EPA in cancer risk assessment.

-------
        II. Chemical Hepatocarcinogenesis, A Model

   This section presents concepts generally accepted to represent the natu-
 ral history of chemical hepatocarcinogenesis, and provides a background
 for the larger discussion of neoplastic nodules and foci of cellular alteration
 in evaluating chemical carcinogenesis. Only key elements are discussed.2
   Hepatocellular carcinoma, like other tumors, is generally recognized as
 the last stage in a multistage process, in which several stages can be de-
 scribed  in terms of the atypical phenotypic expression(s) of cells. The con-
 cept of a stepwise development in the carcinogenic process is mainly based
 on morphological observations, correlated with cellular behavior. However,
 the precise developmental stages in carcinogenesis have not been unequiv-
 ocally demonstrated, but are mainly inferred by post-hoc reconstruction.
   The first step in hepatocarcinogenesis (usually termed initiation)  is
 thought to be an irreversible alteration of one or more hepatocytes, either
 as a "spontaneous" event or induced by contact with an active chemical
 substance. The initial lesion becomes fixed following one or more cycles of
 cell proliferation, such that there appears to be lasting memory of its occur-
 rence in all following cell generations. In some systems initiation appears to
 be manifested by a resistance of the initiated cell to the growth-inhibiting
 effect of carcinogens on normal cells; thus the cell can proliferate into a
 focus, the first observable morphologic hepatocellular lesion (Scherer and
 Emmelot, 1976; Farber, 1980; Pitot and Sirica, 1980; Williams,  1982). Fol-
 lowing initiation, the progression of the initiated cell to cancer can be greatly
 accelerated by the administration of promoting  agents (Berenblum, 1975;
 Pitot and Sirica, 1980).
   Foci are microscopic lesions that can be distinguished from surrounding
 normal  liver tissue by a number of phenotypic markers. For example, rou-
 tine staining with hematoxylin  and eosin differentiates foci into several
 types (e.g., clear cells, eosinophilic foci, and basophilic foci). Foci can also
 differ in their histochemical properties, such as alterations in enzyme activ-
 ities (e.g., deficiencies in glucose-6-phosphatase and beta-glucuronidase;
 increases in gamma-glutamyl transpeptidase), and reduced iron accumula-
 tion. While these are general characteristics, individual lesions may differ in
 their phenotypic marker expression. Foci do not show any disruption of
 normal  hepatic lobular architecture, although some types of foci show ele-
 vated rates of cell division as compared with the  surrounding "normal"
 parenchyma  (Squire  and Levitt, 1975; Farber, 1980; Williams,  1980;
 Bannasch et al., 1980;  Pitot and Sirica, 1980; Emmelot and Scherer, 1980;
 Stewart, 1980; Bannasch et al., 1982).
   Neoplastic nodules usually appear later in time than foci, but before the
 advent of carcinoma; there may be overlap, however. Nodules are com-
 posed of cells that are similar to those of foci (e.g., show similar phenotypic
2Various authors have used a plethora of terms over the years to describe the various hepatocel-
 lular proliferative lesions and associated phenomena. Sometimes they have meant to indicate
 specific types of lesions or cellular processes. However, in order to keep this review simple in
 concept, we use the term "focus" (foci) for any of the various literature references that describe
 small, focal proliferative areas or areas of cellular alteration. We use the term "neoplastic nod-
 ules" to refer to any "nodular" proliferative lesions, whether the authors called them hyperplas-
 tic, neoplastic, or any other comparable terms. Lastly, processes leading to morphological rever-
 sal of proliferative lesions back to normal-appearing morphology (termed variably such things
 as phenotypic maturation, remodeling, reversion, re-differentiation) are called regression.

-------
alterations following routine and special histochemical staining) and show
elevated rates of cell  division. Nodules, however, are larger in size and,
unlike foci, compress  the surrounding parenchyma, from which they are
sharply demarcated  (Squire and Levitt, 1975;  Stewart, 1980; Williams,
1980; Farber, 1980).
  Still later following initiation and the development of foci and nodules,
hepatocellular carcinomas are seen to develop. Carcinogens usually induce
multiple foci or nodules and generally fewer carcinomas.
  Although the morphology and timing of the lesions suggests a progres-
sion from focus to nodule to carcinoma, the case has certainly not been
proven. Farber  (1984)  states that nodules are one identified source, while
Williams (1981) suggests that foci do not necessarily give rise to neoplastic
nodules, but may be direct precursors of carcinoma. Yet, unidentified le-
sions that are antecedents of carcinoma  may also exist.
  A hypothetical projection consistent with these observations is shown in
Figure 1.
  Progression of lesions to carcinoma is only part of the scheme. As will be
discussed later in this paper, under certain experimental conditions it ap-
pears that some foci and neoplastic nodules may  regress following removal
of a chemical stimulus.
Figure 1.   Hypothetical scheme of neoplastic development in rat liver.


                                         Normal Hepatocyte
                                               I
                                            Initiated Cell
                                             . Focus
            Neoplastic Nodule
                                            Carcinoma

-------
             III.  Characteristics of Proliferative
                     Hepatocellular Lesions
A.  Morphological Classification
  In recognition of the diversity of terms used  to describe proliferative
lesions of the rat liver, a conference was held in 1974 to develop a consen-
sus among researchers in the field (Squire and Levitt, 1975). Many of the
previous terms that had existed in the literature  up to that time, such as
nodular hyperplasia, hyperplastic nodule, hepatoma, and hepatic cell ade-
noma, were replaced  by other terms. Three hepatocellular proliferative di-
agnoses were identified: foci (or  areas) of cellular alteration, neoplastic
nodule, and hepatocellular carcinoma (see Figure 2).
Figure 2,    Terminology of rat liver lesions.
              Pre-1975
                       Since 1975
 1985(NTP)
Preneoplastic
           Hyperplastic Foci \
           Basophilic Hyperplasia
           Enzyme-Deficient
           Islands
           Hyperplastic Nodule
           Nodular Hyperplasia
Neoplastic
Benign
Malignant
Hepatoma
Hepatic Cell Adenoma
 Hepatoma
 Hepatic Cell
   Carcinoma
 Trabecular
   Carcinoma
                        Neoplastic
                         Nodule
                                  \
                                      V
                                 Hepatocellular
                                   Carcinoma
Hepatocellular
  Carcinoma
SOURCES: Squire and Levitt, 1975; Stewart et al., 1980; Maronpot et al., 1985.

  Foci were defined as small lesions of less than one liver lobule in size,
whereas areas of alteration were the size of a lobule or greater. The primary
distinguishing features involve the staining and texture of the cytoplasm,
leading to the use of descriptive terms such as, clear cell, eosinophilic (or
ground glass), and basophilic to describe the differences. Foci did not show
any disruption of the normal hepatic architecture. Because of the structural
similarities with neoplastic nodules, most participants agreed that foci and
areas of alteration may be part of a spectrum of lesions capable of progress-
ing to neoplastic nodules (Squire and Levitt, 1975).

-------
   Neoplastic nodules were defined as generally spherical lesions, several
 hepatic lobules in size, which show architectural distortion and sharp de-
 marcation from the surrounding liver over at least a portion of its periphery.
 Neoplastic nodules  also display compression of surrounding liver par-
 enchyma. It was recognized that nodules occur at a low incidence in control
 animals, could be induced by administration of carcinogenic substances,
 and "at the least, indicate an increased probability for the development of
 hepatocellular carcinoma" (Squire and Levitt, 1975). A majority of persons
 thought the previously used term, hyperplastic nodule, was a misnomer,
 and preferred the designation neoplastic nodule. Some question still ex-
 isted about the nature of neoplastic nodules, however, since the conferees
 concluded at that time that benign hepatocellular tumors (that is, those
 without malignant behavior potential) were not recognized in  rats and,
 therefore, the term adenoma was not recommended.
   The  hepatocellular carcinoma designation was applied to lesions with
 characteristic  histological and cytological features.  They are larger and
 more irregular than neoplastic nodules and may involve major portions of
 the liver lobes. At the periphery they compress or extend into the surround-
 ing parenchyma (Squire and Levitt, 1975).
   A second group of experts was assembled by the  National Academy of
 Sciences (NAS) (Stewart et al., 1980) to consider the nomenclature of rat
 liver proliferative lesions. This group affirmed and extended the classifica-
 tion suggested by the previous group (Squire and Levitt, 1975). Hepatocel-
 lular lesions were divided into focus and primary neoplasms (i.e., neoplastic
 nodule and various subtypes of carcinoma).
   The  NAS group recognized that although foci occur in aging control rats,
 an increase in their incidence following chemical  administration raises the
 suspicion that the substance may be carcinogenic. The neoplastic nodule
 was interpreted as a manifestation of the carcinogenic process, with the
 focus representing an earlier stage.
   The  nature and significance of neoplastic  nodules was captured  in the
 report by the following:

    Before and to some  extent after this workshop,  some pathologists
   termed this lesion a "hyperplastic  nodule." Some  who use this term
   regarded, and continue to regard, the lesion so designated as a stage of the
   neoplastic process; others consider it to be unrelated to neoplasia. Still
   others have called the neoplastic nodule a hepatoma or adenoma, thus
   implying that the lesion is a benign neoplasm. The  consensus of those
   attending the workshop,  however,  was that neoplastic nodules,  as
   described, do progress. End-stage benign neoplasms of hepatic cell origin
   have not been identified  with  certainty in the rat liver.  From these
   considerations, the use of the term "hyperplastic nodule" as a synonym for
   neoplastic nodule is to be discouraged. (Stewart et al., 1980).

   The  Squire and Levitt  and NAS papers developed a new classification
 scheme for hepatocellular proliferative lesions, but the exact position and
 significance of the neoplastic nodule was not totally clarified. The nodule
 was seen not as a hyperplastic lesion but as somewhat neoplastic in nature,
 yet still not an end-stage benign tumor.
  Bannasch  (1976) proposed that neoplastic nodules seem to include a
mixture of forms, some precancerous, some cancerous, and  various inter-
mediate cells.
  Dissatisfaction with the term  neoplastic nodule is reflected  in the discus-
sion of a recent international conference organized by the Society of Toxico-
logic Pathologists (Symposium Panel Discussion, 1982). An informal con-

-------
 sensus of the conference participants (determined by a show of hands)
 recommended that the term neoplastic nodule be abandoned.
   Most recently, after discussion with an outside group of experts in the
 field, the National Toxicology Program (NTP) has instituted a change in the
 classification of rat liver proliferative lesions in the chronic toxicity program
 (Maronpot et al., 1986; Maronpot, personal communication, 1985). Basi-
 cally, the NTP has retained the terms focus and hepatocellular carcinoma as
 used before, but has replaced neoplastic nodule with two terms: hyper-
 plasia and hepatocellular adenoma (see Figure 2). Hyperplasia is character-
 ized by mild compression of the surrounding parenchyma that is not noted
 in the focus of cellular alteration. Hyperplasia is viewed as the result of cell
 injury  with accompanying regeneration.  In contrast, hepatocellular  ade-
 noma shows clear-cut demarcation from  surrounding normal tissue  with
 compression and loss of the normal lobular architecture.
   The new NTP classification scheme is similar to that for some other ep-
 ithelial organs: cellular reactions include non-neoplastic hyperplasia as well
 as benign and malignant neoplastic lesions. The choice of terms should
 help to alleviate some of the problems associated with the term neoplastic
 nodule. The effect that the change in nomenclature will have on the propor-
 tion of neoplastic nodules that will now be called hyperplasia must await its
 use in a number of chronic studies. Although the NTP is applying the new
 terminology to the Fischer 344 rat, the new nomenclature could be extended
 to other strains. Finally, it is important to see how the proposed classifica-
 tion will hold  up in experimental studies concerned with hepatocarcinogen-
 esis in which  a host of biological indicators are used to assess the nature of
 the various proliferative lesions.

 B. Background Incidence of Lesions in Untreated Animals
  A few authors have reported the incidence of hepatocellular proliferative
 lesions in control animals in various strains of rats. A representative set of
 values is tabulated in this report (Table 1). Over a lifetime, it appears that the
 bulk of untreated animals develop foci, whereas many fewer develop neo-
 plastic nodules or carcinomas. For instance, for Fischer 344 rats, it seems
 that over 75% of control animals develop foci, whereas usually less than 5%
 and less than 1% of animals develop neoplastic nodules and liver-cell  can-
 cers, respectively (Goodman et al., 1979; Haseman et al., 1984; Popp et al.,
 1985).

 C. Comparative Morphology
  Proliferative liver lesions have been identified in a number of different
 species. Foci showing alteration in routine or special (e.g., enzymatic) histo-
 logical  staining have been  identified in the mouse,  rat, golden hamster,
 monkey, and  human  (Ward, 1984; Stenback et al., 1986). The same is  true
for "benign"  nodules and carcinomas. Both Ward (1984) and Popper et al.
 (1977)  point out the  similarity in proliferative lesions (i.e., foci, nodules
 [adenomas], and carcinomas) among species, although Ward claims  that
 "dysplastic" foci are more common in the monkey and human than in the
 mouse,, and Popper et al. indicate that human  lesions show a much more
 pronounced fibrotic reaction than those in rodents. This latter difference is
probably a reflection of the fact that normal rat liver contains only one-third
the hydroxyprolihe (a measure of collagen, a necessary component in fibro-
sis) of normal human liver.

-------
Table 1.    Proliferative Liver Lesions: Background Incidence in 2-Year-Old
           Rats
Animals with (%)
Strain
Wistar

Sherman
COBS
CD (Charles
River)
Fischer 344a

Fischer
344b
Fischer 344
(early NCI
experi-
ence)
Fischer 344
(recent
NTP ex-
perience)
Sprague-
Dawley0
(Dow
Chemical)
Wistar (Col-
worth
strain)

Number Neoplastic
Sex examined Foci nodules
M

F

M
F
M

M
F
M
F


M
F


M
F


M
F


7 100

173 16

20 15
20 50
9 100

85 78
79 87
1794
1754


2306
2356


171
172


45
45


—

	

—
—
—

11
4.6
1.3
2.7


3.4
3.0


3.5
7.0


13
11


Carcinomas Reference
—

	

—
—
—

1.2
0
0.39
0.39


0.8
0.2


2.9
1.7


16
29


Ogawa et al..
1981
Kimbrough
et al., 7975
Ulland et al.,
1977
Ogawa et al.,
1981
Popp et al.,
1985
Goodman et
al., 1979


Haseman et
al., 1984


Adapted
from
Thorpe et
al., 1982
Adapted
from
Thorpe et
al., 1982
"72 weeks old.
b111 weeks old; total from three laboratories.
Tote/ from two studies.

-------
     IV.  Relationship Between Certain Hepatocellular
    Proliferative Lesions and Hepatocellular Carcinoma

   Administration of hepatocarcinogenic substances to rats is associated
 with two different phenomena. One is the induction of proliferative hepato-
 cellular lesions, foci, and neoplastic nodules, and usually later, hepatocellu-
 lar carcinomas. The other is reversion of early lesions back to "normal"
 morphology, which occurs under some experimental conditions, if the car-
 cinogenic stimulus  is removed early in the process. This section summa-
 rizes some of the evidence dealing  with the potential of various early prolif-
 erative lesions either to progress to carcinoma or regress to normal
 appearance.


 A. Regression of  Early Lesions

   Early proliferative lesions induced by hepatocarcinogens often change
 following discontinuation of administration  of the carcinogen. The mor-
 phology of foci and nodules becomes more like that of normal adult liver.
 Many  of the biochemical  changes noted in these lesions reverse (e.g.,
 glucose-6-phosphatase activity is  regained, gamma-glutamyl  transpepti-
 dase activity declines). Due to this regression, the area becomes grossly
 indistinguishable from the surrounding liver (Kitagawa and Pitot, 1975-
 Farber, 1980).
 1. Foci

  Goldfarb and Zak (1961) reported regression of foci 4 weeks after with-
 drawal of 3'-methyl butter yellow, fed for 26 days. (When the same carcino-
 gen was fed for 40 days, neoplastic nodules and hepatocellular carcinomas
 were found.)
  Kitagawa (1976) demonstrated  after feeding N-2-fluorenylacetamide
 (2-AAF) that a majority of "hyperplastic" areas (it is assumed that the author
 was referring in part to foci) underwent regression. On the other hand, in a
 small group of foci, enzyme deficiency  persisted during the observation
 period. The author speculated that these areas appeared to be more impor-
 tant than the ones that had regressed, in relation to the later development
 of carcinoma.
  Williams and Watanabe  (1978) reported that over 95% of altered foci,
 induced by 2-AAF, regressed within 6 months after withdrawal of 2-AAF (but
 that a small fraction progressed to neoplastic nodules and carcinoma). Also
 after promotion with phenobarbital (following 2-AAF withdrawal) "over
 90% of livers developed carcinoma by 6 months."
  Regression of foci has also been reported with N-nitrosomorpholine
 (NNM) (Moore et al., 1983). There have also been reports in which foci did
 not regress. For instance, daily oral doses of NNM for 7 weeks at  doses of
 80-200 mg/L in drinking water resulted in increases in the number of foci
 which persisted for periods up to 50 weeks after cessation of dosing (Moore
 et al., 1982).
  Irreversible foci were also induced with a single dose  administration of
diethylnitrosamine (DEN) followed by partial hepatectomy (Scherer and
Emmelot, 1975) or followed by feeding of phenobarbital (Pitot et al., 1978).

-------
2. Neoplastic Nodules
  Sprague-Dawley rats fed 0.064% methyl  butter yellow for 26 days or
40 days and then put on a standard lab diet showed different carcinogenic
responses (Goldfarb and Zak, 1961). Neoplastic nodules formed in animals
treated for 40 days, but most underwent regression upon cessation of treat-
ment; many animals had developed hepatocellular carcinomas by the end
of the study. However, few nodules or carcinomas were noted after 26 days
of treatment. This compound induced hepatotoxic effects in addition to
parenchymal cell changes (e.g., lesions of bile ducts and cirrhosis). Confirm-
ing studies showing regression of nodules initially produced by methyl
butter yellow were reported by Goldfarb (1973).
  Teebor and Becker (1971) reported heterogeneity among neoplastic nod-
ules. They found that when Sprague-Dawley rats were fed 0.06% 2-AAF for
3 monthly cycles (3 weeks on treatment;  1  week on basal diet) and then
returned to control diet, many nodules were produced  which then re-
gressed. Only a low incidence of hepatocellular carcinomas developed over
time. However, in another group of rats that had been fed for only 1 addi-
tional monthly cycle, nodules persisted, and there was a high incidence of
animals with hepatocellular carcinomas.
  Kitagawa (1976) found that many neoplastic nodules induced by feeding
0.03% 2-AAF to Sprague-Dawley or Donryu rats for 12 weeks returned to a
more normal phenotype upon withdrawal of the agent.
  In contrast to the above experiments, Hirota and Williams (1979) failed to
find regression of neoplastic nodules in F344 rats fed 0.021% 2-AAF for 3,4,
or 5 cycles (4 weeks on treatment; 1 week on basal diet) and then returned
to control diet, although some nodules did show changes in enzyme stain-
ing that are often seen in nodules that regress.
  Moore et al. (1982) also failed to find regression of neoplastic nodules and
foci in male Sprague-Dawley rats that had been given NNM in drinking
water (80-200 mg/L) for 1  to 20 weeks and then sacrificed 8 weeks after
removal of the chemical. Some animals given NNM for 7 weeks were not
sacrificed until 50 weeks post-treatment; again, there was no indication of
regression. Unlike these findings, when the carcinogen is given for a limited
period of time and then stopped, when the  same laboratory (Moore et al.,
1983) evaluated NNM in a short-term exposure system,3 regression of nod-
ules and foci was observed.
   In a series of experiments using a related model,4 Farber and co-workers
have clearly indicated that nodules synchronously appear in great numbers
following treatment, only to regress following termination of 2-AAF (Solt et
al., 1977; Ogawa et al., 1979b; Enomoto and Farber, 1982; Farber, 1982;
Tatematsu et al., 1983). These results have been confirmed by Van der
Heijden and Dormans (1981). During regression, staining characteristics in
most nodules return to normal and the characteristic morphology of the
lesions gradually disappears, with the cells in the area again appearing
"normal." Tritiated thymidine labeling has demonstrated that regresssion
to  normalcy in this system is not attended with profound cellular  loss;
instead, existing cells in nodules  return to normal phenotypes (Kitagawa,
1976; Enomoto and Farber, 1982).
 SPartial hepatectomy, 3 days NNM (400 mg/L in drinking water), 2 weeks of 2-AAF (0.2 mg/kg in
  feed) with a single dose of carbon tetrachloride (1 mL/kg) midway during 2-AAF exposure.

 4A single dose of N-nitrosodiethylamine followed 2 weeks later by feeding 2-AAF for 1 week and
  partial hepatectomy.

                                10

-------
  Thus, it appears that both foci and neoplastic nodules can regress under
at least two circumstances following removal of a carcinogenic stimulus. In
"stop" experiments, where a carcinogen is given for a short period and then
stopped, both foci and nodules have reverted to normal morphology. In
addition, foci and nodules have also regressed in short-term experiments,
when a carcinogen is given for a period of time following administration of
another carcinogen to initiate the process and  a stimulus is given to cause
liver regeneration. Regression has not, however, been noted by every inves-
tigator, nor have all experimental systems been investigated.
  Even in those cases in which regression has been demonstrated, some
proliferative lesions persist. Farber has stated that despite  a high rate of
regression of nodules in his experimental model, 70% of animals develop
hepatocellular carcinomas by 9 months following cessation of treatment
(Ogawa et al., 1979a; Farber, 1982).
  For those proliferative lesions that regress, the question remains whether
they retain "memory" of the initial lesion. Farber (1982) hypothesized that
these hepatocytes revert back to an initiated cell; however, evidence for this
position is still lacking.

B. Progression

  The general acceptance of the developmental sequence:  focus to neo-
plastic nodule (NN) to hepatocellular carcinoma (HCC) (i.e., progression) is
based mainly on indirect evidence. Table 2 presents a summary of some of
this evidence in simplified form.
  Although there has been much discussion of  candidates for a phenotypic
marker of cells committed to develop carcinoma, none as yet has proven to
be a reliable indicator of the origin of hepatocellular carcinoma cells. Thus,


Table 2.    Evidence for Progressions of Hepatocellular Lesions
                                     Foci
                                                   NN
                                                               HCC
Incidence increased by carcino-
  gen administration                   +

Sequence in the development of
  lesions                             ^t

Proliferative lesion                     +

Number of lesions per liver           + + +

Morphology: compression of
  adjacent tissue                      —

Markers,

  Enzyme-staining abnormalities        +

  Decreased iron accumulation          +

Site of hepatocellular carcinoma
  formation                            ?

Transplantability                      -/?
                                                    +


                                                  2nd

                                                    +

                                               + + +/++?


                                                    +
3rd
                                                  +/_

                                                  +/_
                                                               + -f
n/a
                                11

-------
most of the conclusions in the literature on this aspect of hepatocarcinogen-
esis are based on inference, rather than direct evidence. Some of these are
presented.

1.  Foci
  It has been claimed that some carcinomas arise within foci (Schulte-
Hermann et al., 1982). However, the authors cautioned that there was no
proof that all foci progressed to malignancy.
  Cells of certain carcinogen-initiated  foci cells are able to proliferate (in-
crease in number) in the absence of the carcinogen for considerable periods
of time (Scherer and Emmelot, 1975).
  Bannasch et al. (1982)  observed that "foci became larger and acquired
phenotypic markers closerto neoplasia, independent of further action of the
carcinogen."
  Although foci undergo regression, it has  been suggested that these re-
verted  cells  may retain  some  permanent  carcinogen-induced alteration
which could make them susceptible to  progression if exposed to additional
carcinogens (Williams and Watanabe, 1978). These authors  describe an
experiment where  PB was given as a promotor after withdrawal  of AAF,
which resulted in increased persistence of foci, and up to 90% of livers so
treated developed carcinomas by 6 months.
  Scherer and Emmelot (1975) suggested there is a precursor-product rela-
tionship between foci and hepatocellular carcinoma, which they claim to
have demonstrated by a quantitative technique. (See Emmelot and Scherer,
1980 for a detailed  discussion.)
  It has also been claimed that only a small fraction of foci may develop into
nodules or carcinomas:  about 1 in  1000 following  administration of AAF
(Watanabe and Williams, 1978) and about 1  in 11,000 after partial hepatec-
tomy and dosing with methyl(acetoxy-methyl)nitrosamine (Kaufman et al.,
1985).

2. Neoplastic Nodules
  Cancer has been found to arise in nodules  (Goldfarb, 1973; Williams, 1976
[observed cancer arising  both inside and outside nodules]; Solt et al., 1977;
Popper et al., 1977; Farber, 1982). Some nodules have been found to be a
site for cancer development with five different carcinogens in different mod-
els (Farber et al., 1984); however, cancer cannot be said to arise exclusively
inside nodules.
  A  minority of nodules do not undergo regression following carcinogen
removal, but persist (Ogawa et al., 1979a; Van der Heijden and Dormans,
1981; Enomoto and Farber, 1982; Moore et al., 1983; Tatematsu et al., 1983;
Farber, 1984). These may play a role in the  development of carcinoma.

Autonomy
  Neoplastic nodules sometimes develop long after removal  of a carcino-
gen  and are progressively growing entities:  a carcinogen can be withdrawn
for 8 to 12 weeks,  and liver cancer will still appear weeks or months later
(Epstein etal., 1967). Several researchers have published experimental re-
sults that clearly demonstrate this concept of autonomous growth for neo-
plastic nodules. For many weeks after chemical withdrawal, neoplastic nod-
ules continue to enlarge or invariably  increase in number (Goldfarb, 1973;
Hirota  and Williams, 1979). In fact, Hirota  and Williams (1979) presented
data from studies  in F344 rats fed N-2-fluorenylacetamide for a limited
number of cycles, which clearly showed that nodules were persistent after
carcinogen withdrawal and had progressive growth ability (in situ), a char-
acteristic of neoplasms. Similar conclusions regarding persistence and au-

                                 12

-------
 tonomy were also reported for various experimental protocols (Scherer and
 Emmelot, 1975; Emmelot and Scherer, 1980).

 Transplantation
   Since on transplantation, hepatocellular carcinomas display continued
 growth, invasiveness, and metastasis,  several investigators have studied
 the behavior of "putative preneoplastic lesions" following transplantation.
 The results are not consistent.
   Lee et al. (1983) described the transplantation of carcinogen-altered hep-
 atocytes and their subsequent development into metastasizing cancer. Dis-
 sociated  cells  from livers  bearing  chemically  induced nodules were im-
 planted into spleens of syngenic rats.  Cells from "early nodules,"5 grew
 slowly in the spleen, but their evolution into cancer was  not observed  in
 studies lasting up to 2 years. Cells from "persistent nodules,"6 on the other
 hand, generated  hepatocellular carcinoma in the spleen within 18 months
 (Finklestein et  al., 1983; Farber, 1984).
   In another transplantation model, hepatocytes from nodules showed pro-
 gressive invasion into chick embryo heart fragments in the same way as did
 those from hepatocellular carcinoma. Control hepatocytes showed no inva-
 sion (Wanson  et  al., 1981).
   In still another series of studies summarized by Williams (1980), earlier
 attempts at transplantation of neoplastic nodules were unsuccessful. Nod-
 ule transplants placed in the mammary fat pads of syngenic rats were found
 to persist for months, but did not display "continued growth, invasiveness,
 or metastasis," although cell division was noted.
   Ward et al. (1975) claimed that two well-delineated  nodules, induced by
 dietary aflatoxin B-i and transplanted into syngenic recipients, grew slowly.
 In a later study, Ward and Lynch (1984)  were unsuccessful in transplanting
 a hepatocellular "adenoma" and a hepatocellular carcinoma.
   In summary, there is some  information that indicates that foci and nod-
 ules may continue to grow and possibly progress toward carcinoma. The
 case is stronger for the progression of some nodules  to carcinoma than it
 is for foci, but in neither case is the evidence conclusive. It is also recognized
 that under certain experimental conditions many foci and  nodules regress
 after the administration of a hepatocarcinogen has been discontinued.
 Thus,  even if some foci or nodules do progress, all such lesions are not
 necessarily obligate precursors of carcinoma.
   One  can  conclude that foci and nodules are part  of the carcinogenic
 process. It is generally recognized that foci are hyperplastic lesions, but it  is
 still difficult to discern whether rat liver nodules are hyperplastic or neoplas-
 tic lesions. A position consistent with the data is that at least some nodules
 may be neoplastic (e.g., those that persist after carcinogen removal) and
 have the potential to progress to carcinoma.
5Those nodules induced in the short-term test system employed by Farber and co-workers; see
 footnote 4.
6The small percentage of nodules that do not regress following removal of a carcinogenic stim-
 ulus.

                                 13

-------
      V.  Current Governmental Use of Data on Rat
                    Hepatocellular Lesions
  Several governmental agencies have adopted procedures for the use of
proliferative lesions in experimental animals in the evaluation of chemicals
for carcinogenicity.
  The National Toxicology Program (NTP) commonly pools the incidence of
certain benign and their corresponding malignant neoplasms in evaluating
the carcinogenic hazard of chemicals in long-term animal tests (McConnell
et al., 1986); this practice has been endorsed by an advisory group to the
NTP (1984). For rat hepatocellular lesions, separate incidence figures are
generated  for animals bearing neoplastic nodules, hepatocellular car-
cinomas, and for the combination of the two. Carcinogenicity determina-
tions are based on treatment versus control comparisons and dose-trend
relationships for all these groupings. Neoplastic nodules have been inter-
preted by NTP as if they were benign neoplasms (Maronpot and Boorman,
1982); chemically related increases in any of the three groupings constitute
"clear" evidence of carcinogenicity. The NTP also collects information on
the frequency of different types of foci  of cellular alteration among animals
in their long-term studies.
  Recently, NTP has proposed to change the classification of rat hepatocel-
lular proliferative lesions in their chronic toxicity testing program (Maronpot
et al.,  1986). The major difference in their new classification scheme is
replacement of the term neoplastic nodule with two terms: hyperplasia and
hepatocellular adenoma. At the end of a 2-year trial period, NTP will evalu-
ate the adequacy of this new nomenclature.
  The Office of Science and Technology Policy (1985) report and the EPA
Proposed Guidelines for Carcinogen Risk Assessment (U.S. EPA, 1984) sup-
port the pooling of benign and malignant tumors of the same histological
type when  scientifically defensible. The EPA Guidelines go on to state that
the contribution of benign tumors to the total extrapolated risk should be
expressed when pooling has been used. In qualitative hazard assessment,
the Guidelines accord less weight to cases where the only neoplastic re-
sponse attributed to chemical exposure is the production of benign tumors.
Such data constitute limited evidence of animal carcinogenicity,  and such
chemicals are generally put into Group C (Possible Human Carcinogen). The
EPA Proposed Guidelines also state that decisions as to whether to proceed
with quantitative risk assessment on Group C substances would be made on
a case-by-case basis.
  The Food and Drug Administration, Occupational Safety and Health Ad-
ministration, and Consumer Product Safety Commission  have used rat neo-
plastic nodules in the assessment of carcinogenicity  of  chemical sub-
stances,  but  no specific  policies have been established.  Hepatocellular
carcinomas and nodules are used in a weight-of-evidence determination as
to whether a substance may pose a carcinogenic hazard. All agencies  use
foci in their hazard assessments to help in the interpretation of hepatocellu-
lar neoplastic responses.
                                 14

-------
     VI.  Future Use of Rat Proliferative Hepatocellular
                          Lesions by EPA
 A.  Rationale for Positions
   Although the disagreements seem to have narrowed in recent years,
 there are still differences of opinion within the scientific community on the
 biological significance of certain parenchyma! cell proliferative lesions in rat
 liver. Several factors  have influenced this situation, including questions
 about nomenclature and possible progression and regression of these le-
 sions under certain circumstances. As the National Academy of Sciences
 (NRC, 1983) pointed out, even in the absence of scientific consensus, Fed-
 eral agencies have the responsibility for developing inference guidelines
 (i.e., science policiesj  which lay out the positions to be taken on specific
 issues together with their rationale.
   From the accompanying review it seems apparent that scientists have
 described a number of proliferative lesions of rat hepatocytes which are in
 some way part of the carcinogenic process. Morphological entities similar
 to those found in the rat have been described in the livers of other rodents,
 monkeys, and humans. Foci appear to be some type of preneoplastic lesion,
 but controversy still surrounds the term neoplastic nodule. Some consider
 the  nodule to be a hyperplastic lesion while others interpret it to be a
 neoplastic entity (a benign or possibly a malignant  lesion), and the term
 may include entities from all such morphological  designations.
  Foci are commonly found in the livers of control rats at the end of long-
 term studies;  neoplastic nodules and hepatocellular carcinomas are less
 common. Administration of carcinogens, but not of noncarcinogens, in-
 creases the incidence  of these three lesions. These lesions are also in-
 creased in animals initially given a carcinogen and then a promoter. There
 seems to be agreement that an  increase in foci and nodules in treated
 animals is some indication that the liver is at increased risk of cancer forma-
 tion.
  The appearance of  morphological entities generally follows a  time-
 dependent pattern following administration  of a hepatocarcinogen: foci
 form first, followed by  nodules, and  still later, by carcinomas.
  Although there is some evidence supporting the notion of progression
 from a latent-initiated cell to focus to nodule and then to carcinoma, that
 relationship cannot be said to be established. Carcinoma sometimes seems
 to arise from a nodule; a focus or some yet unidentified precursor or a
 combination of the ab'ove may  be other sources.
  Regression experiments serve two roles: they help to elucidate steps in
the carcinogenic process, and they may give insights  into phenomena that
 may occur under conditions of limited or intermittent exposure to carcino-
genic substances. Human exposures to environmental toxicants may be
 limited or intermittent and, therefore, regression experiments may be rele-
vant to risk evaluation.
  It also appears that under certain experimental conditions, early prolifer-
ative lesions (i.e., some foci and nodules) that have been  induced by hepa-
tocarcinogens will largely regress  following removal of the carcinogenic
stimulus. Whether these lesions return to a latent-initiated state or back to
a totally normal cell, or whether the cells may be  lost or destroyed under
some circumstances, is not known. To the extent that proliferative lesions

                                15

-------
regress, however, there may be some modulation of the potential hazard for
cancer development.
  Even in those cases in which the bulk of the foci and nodules have been
shown to regress following carcinogen removal, some lesions persist. The
persistent lesions are often associated with carcinoma formation at a later
time. It also seems possible that some of the persistent lesions may actual ly
progress to carcinoma.
   Federal guidance for carcinogen risk assessments, as developed  in the
Office of Science and  Technology Policy (1985) document, recommends
that the incidence of animals bearing malignant and corresponding benign
lesions should be combined when scientifically defensible. It also points out
the importance of reviewing the incidence of preneoplastic lesions in  target
organs as being an adjunct to the assessment of a chemical's carcinogenic-
ity. EPA's Proposed Guidelines for Carcinogen  Risk Assessment (1984) af-
firm these positions, but state that an increase in benign tumors alone is
accorded less weight in assessing carcinogenic hazards than an increase in
malignant tumors or a combination of malignant and corresponding benign
tumors.  Increases in benign tumors only are given a weight-of-evidence
classification of Group C (Possible Human  Carcinogen). Decisions as to
whether to proceed with a  quantitative risk assessment are made on  a
case-by-case basis.
   Thus,  in regard to rat liver proliferative lesions, it appears that Federal
agencies are using foci and neoplastic nodules along  with hepatocellular
carcinomas in risk assessments on chemicals, and it seems appropriate to
continue such  practices. Accordingly, the Agency provides the following
guidance for interpreting liver toxicity data in rats.

B. EPA Positions
   In formulating a policy position on the interpretation of neoplastic nod-
ules, one needs to be cognizant of the existing scientific evidence as well as
the inconsistencies in the data, the absence of information in certain  areas,
and the lack of a consensus in the scientific community.
   Morphologically described proliferative lesions of rat parenchymal cells
will be used in  carcinogen risk assessments in conjunction with the EPA
 Proposed Guidelines for Carcinogen Risk Assessment (1984) as follows:
   1.  Foci of cellular alteration are to some extent indicative of an ongoing
carcinogenic process and should be used in risk assessments. Accordingly,
 EPA adopts the following position:
    An analysis of the incidence of foci of cellular alteration among treat-
   ment groups can contribute to the interpretation of a given rat toxicity
   study and, thus, becomes part of the hazard identification step of the
   carcinogenic risk assessment. Dose-response and time-to-occurrence
   parameters are useful in this qualitative analysis.
   2.  Neoplastic nodules are increased in animals  receiving carcinogens,
 and that occurrence is almost always associated with an increase in hepato-
 cellular carcinoma. However, it appears that there is not a one-to-one rela-
 tionship between nodules and  carcinomas.  Although some  nodules may
 have "malignant potential," others may only be "hyperplastic" lesions. Still
 other nodules'may regress following cessation of carcinogenic administra-
 tion.
   Therefore, the exact contribution of neoplastic  nodules to the overall
 incidence of hepatocellular tumors in the rat  is unclear at this time. EPA
 adopts the following position until better means of articulating uncertainties
 are developed.
     Determination of carcinogenic hazard will be based upon consider-
   ation of the incidence  of hepatocellular carcinoma alone, neoplastic

                                 16

-------
   nodule alone, and a combination of carcinoma and nodule. The range
   of responses could vary from cases where there are very significant
   increases in the incidence of carcinomas backed up by increases in
   nodules, through  situations where only the combined frequency of
   animals with carcinomas or nodules is significant, to cases where in-
   creases are limited to nodules alone.
     (i) At one end, responses showing an overwhelming increase in car-
   cinomas alone will be interpreted as providing sufficient evidence of
   animal carcinogen/city when other criteria for the sufficient category
   are met. In the absence of human evidence and supporting information,
   such evidence will generally be given a weight-of-evidence designation
   of Probable Human Carcinogen  (Group B2), and quantitative risk as-
   sessment is appropriate.
     (ii) At the other end, where increases in lesions and statistical signif-
   icance are restricted to neoplastic nodules alone,  such data  will be
   interpreted  as only limited evidence of animal carcinogenioity. When
   neither human evidence nor supporting information accompany the
   increase in nodules, the weight-of-evidence designation is Possible Hu-
   man Carcinogen (Group C), but quantitative assessment is not war-
   ranted.
     (Hi) Situations within the two extremes will be evaluated on a case-
   by-case basis.  In determining carcinogenic hazard, different  factors
   should be considered in deciding whether there is sufficient or limited
   evidence of animal carcinogenicity.
    Some of these factors include the proportion of proliferative lesions
   that are malignant; the magnitude of increase in incidence of nodules
   and carcinomas; considerations of dose response and time-to-tumor;
   and the nature and the outcome of short-term tests and other support-
   ing information. In deciding whether or not to proceed with quantita-
   tion, consideration will be given to the overall weight of evidence. In
  cases where quantitation is performed, methods of expressing the pos-
  sible uncertainty in  the estimated risks should  be explored (e.g.,  ex-
  pressing risk as a range between that determined from extrapolation of
  carcinoma incidence alone and that from a combination of carcinomas
  and nodules).
  3.  Hepatocellular adenoma  will constitute a diagnostic classification in
some rat studies reviewed by the Agency  in the  future. The designation
adenoma includes a yet unknown proportion of lesions previously called
neoplastic nodules. The Agency concludes that:
    The EPA current position on benign tumors will apply to the designa-
  tion, hepatocellular adenoma. The pooled incidence of hepatocellular
  adenomas and carcinomas  will be  used in making qualitative judg-
  ments as to  whether a rat study demonstrates evidence of a carcino-
  genic hazard. Pooled incidence will also be used for dose-response
  assessments. In cases where adenomas alone reach significance, deci-
  sions as to whether to proceed with a dose-response assessment will
  be made on a case-by-case basis.
                                17

-------
                           References

Section I
Dorland's Illustrated Medical Dictionary, 25th ed. (1974) Philadelphia, PA:
  W.B. Saunders Co.
Stedman's Medical Dictionary, Illustrated, 24th ed. (1982) Baltimore, MD:
  Williams and Wilkins.

Section II
Bannasch, P.; Mayer, D.; Hacker, H.J. (1980) Hepatocellular glycogenesis
  and hepatocarcinogenesis. Biochim. Biophys. Acta 605:217-245.
Bannasch, P.; Moore, M.A.; Klimek, F.; Zerban, H. (1982) Biological markers
  of preneoplastic foci and neopiastic  nodules  in rodent liver. Toxicol.
  Pathol. 10:19-37.
Berenblum, I. (1975)  Sequential aspects of chemical carcinogenesis: skin.
  In: Becker, F.F., ed. Cancer: a comprehensive treatise. V. 1. New York, NY:
  Plenum  Press, pp. 323-344.
Emmelot, P.; Scherer, E. (1980) The first relevant stage in rat liver carcino-
  genesis. A quantitative approach. Biochim. Biophys. Acta 605:247-304.
Farber, E. (1980) The sequential analysis of liver cancer induction. Biochim.
  Biophys. Acta 605:149-166.
Farber, E. (1984) Cellular biochemistry of the stepwise development of can-
  cer with chemicals: G. H. A. Clowes Memorial Lecture.  Cancer Res.
  44:5463-5474.
Pitot, H.C.; Sirica, A.E. (1980) The stages of initiation and promotion in
  hepatocarcinogenesis. Biochim. Biophys. Acta 605:191-215.
Scherer, E.; Emmelot, P. (1976) Kinetics of induction and growth of enzyme-
  deficient islands involved in hepatocarcinogenesis. Cancer Res. 36:2544-
  2554.
Squire, R.A.; Levitt, M.H. (1975)  Report of a workshop on classification of
  specific hepatocellular lesions in rats. Cancer Res. 35:3214-3223.
Stewart, H.L.  (1980)  Histological typing  of liver tumors of the rat.  J. Natl.
  Cancer Inst. 64:178-190.
Williams, G.M. (1980) The pathogenesis of rat liver cancer caused by chem-
  ical carcinogens. Biochim. Biophys. Acta 605:167-189.
Williams, G.M. (1981) Liver carcinogenesis: the role for some chemicals of
  an epigenetic mechanism of liver-tumor promotion involving modifica-
  tion of the cell membrane. Food Cosmet. Toxicol. 19:577-583.
Williams, G.M. (1982) Phenotypic properties of preneoplastic rat liver le-
  sions and applications to detection of carcinogens and tumor promotors.
  Toxicol. Pathol. 10:3-10.

 Section III
 Bannasch, P. (1976) Cytology and cytogenesis of neopiastic (hyperplastic)
   hepatic nodules. Cancer Res. 36:2555-2562.
 Goodman, D.G.; Ward, J.M.; Squire, R.A.; Chu, K.C.; Linhert, M.S. (1979)
   Neopiastic and  nonneoplastic lesions in aging F344 rats. Toxicol. Appl.
   Pharmacol. 48:237-248.
 Haseman, J.K.; Huff, J.; Boorman, G.A. (1984) Use of historical control data
   in carcinogenicity studies in rodents. Toxicol. Pathol. 12:126-135.
 Kimbrough, R.D.; Squire, R.A.; Linder, R.E.; Strandberg, J.D.; Montalli, R.J.;
   Burse, V.W. (1975) Induction of liver tumors in Sherman strain female rats

                                 18

-------
   by polychlorinated biphenyl-Aroclor 1260. J. Natl. Cancer Inst. 55:1453-
   1459.
 Maronpot, R.R.; (1985)  Letter to  B. Haberman, Carcinogen  Assessment
   Group, August 14, 1985.
 Maronpot, R.R.; Montgomery, C.A., Jr.; Boorman, G.A.; McConnell, E.E.
   (1986) National Toxicology Program nomenclature for rat hepatoprolifer-
   ative lesions. Submitted to Toxicol. Pathol.
 Ogawa, K.; Onoe, T.; Takeuchi,  M.  (1981) Spontaneous  occurrence  of
   gamma-glutamyl transpeptidase-positive hepatocytic foci in 105-week-
   old Wistar and 72-week-old Fischer 344 male rats. J. Natl.  Cancer Inst
   67:407-412.
 Popp, J.A.; Scortichini, B.H.; Carvey, LK. (1985) Quantitative evaluation  of
   hepatic  foci  of alteration occurring spontaneously in Fischer 344 rats
   Fundam. Appl. Toxicol, 5:314-319.
 Popper, H.; Selikoff, I.J.; Maltoni, C.; Squire, R.A.; Thomas, LB. (1977)
   Comparison  of neoplastic hepatic lesions in  man and experimental ani-
   mals. In: H.H. Hiatt; J.D. Watson; J.A. Winsten. Origins of human cancer.
   Book C. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory DD
   1359-1382.
 Squire, R.A.; Levitt, M.H. (1975) Report of a workshop on classification  of
   specific hepatocellular lesions in rats. Cancer Res. 35:3214-3223
 Stenback, F.; Mori, H.;  Furuya, K.; Williams, G.M. (1986) The pathogenesis
   of dimethylnitrosamine-induced  hepatocellular cancer in hamster liver
   and a lack of enhancement by phenobarbital.  J. Natl. Cancer Inst 76-327-
   333.
 Stewart, H.L.; Williams G.M.;  Keysser C.H.; Lombard,  L.S.; Montali, R  J
   (1980) Histologic typing of liver  tumors of the rat. J. Natl. Cancer Inst
   64:177-207.
 Symposium Panel Discussion. (1982) Rodent liver nodules: significance and
   human cancer risk. Toxicol. Pathol. 10:208-226.
 ThMnoe\f'' Bolt' H'M-; Elcombe' C-R- Grasso,  P.; Paglialunga S.;  Roe,  F.
   (1982) Hepatocarcinogenesis in laboratory rodents: relevance for man
   Monograph No. 4. Brussels, Belgium: European Chemical Industry Ecol-
   ogy and  Toxicology Centre. 77 pp.
 Ulland, B.M.; Page, N.P.; Squire, R.A.; Weisburger, E.K.; Cypher, R.L (1977)
   A carcinogenicity assay of Mirex in Charles River CD rats. J. Natl. Cancer
   Inst. 58:133-140.
 Ward, J.M. (1984) Morphology of potential preneoplastic hepatocyte lesions
   and liver tumors in mice and a comparison with other species. In- Popp
   J.J. ed.  Mouse liver  neoplasia. New York, NY: Hemisphere Publishing
   Corp., pp. 1-26.
Section IV

Bannasch, P.; Moore, M.A,; Klimek, F.; Zerban, H. (1982) Biological markers
  of preneoplastic foci and  neoplastic nodules in rodent liver Toxicol
  Pathol. 10:19-37.
Emmelot, P.; Scherer, E. (1980) The first relevant stage in rat liver carcino-
  genesis. A quantitative approach. Biochim. Biophys. Acta 605-247-304
Enomoto, K.; Farber, E. (1982) Kinetics of phenotypic maturation of remod-
  eling of hyperplastic  nodules during liver carcinogenesis. Cancer Res
  42:2330-2335.
Epstein, S.; Ito, N.; Merkow,  L; Farber, E. (1967) Cellular analysis of liver
  carcinogenesis: the induction of large hyperplastic nodules in the liver
  with 2-fluorenylacetamide or ethionine and some aspects of their mor-
  phology and glycogen metabolism. Cancer Res. 27:1702-1711.

                                19

-------
Farber, E. (1980) The sequential analysis of liver cancer induction. Biochim.
  Biophys. Acta 605:149-166.
Farber, E. (1982) The biology of carcinogen-induced hepatocyte nodules and
  related liver lesions in the rats. Toxicol. Pathol. 10(2): 197-205.
Farber, E. (1984) Cellular biochemistry of the stepwise development of can-
  cer with  chemicals:  G. H.  A.  Clowes Memorial Lecture.  Cancer Res.
  44:5463-5474.
Farber, E.; Eriksson, L.C.; Roomi, M.W.; Cameron, R.G.; Hayes, M.A.  (1984)
  Chemical carcinogenesis: hepatocyte nodules with a special phenotype
  as a common step at the crossroads. Toxicol. Pathol. 12:288-290.
Finklestein, S.D.; Lee, G.; Medline, A.; Tatematsu, M.; Makowka, M.; Farber,
  E. (1983) An experimental method for rapid growth of liver in spleen. The
  survival and proliferation of chemically induced preneoplastic hepato-
  cytes in spleen. Am.  J. Pathol. 110:119-126.
Goldfarb, S. (1973) A morphological and histochemical study of carcinogen-
  esis of the liver in rats fed 3'-methyl-4-dimethylaminoazobenzene. Cancer
  Res. 33:1119-1128.                                        .     .
Goldfarb, S.;  Zak, F.G. (1961) Role of injury and hyperplasia in the induction
  of hepatocellular carcinoma. J. Am. Med. Assoc. 178:137-139.
Hirota, N.; Williams, G.M. (1979) Persistence and growth of rat liver neoplas-
  tic nodules following cessation of carcinogen exposure. J. Natl. Cancer
  Inst. 63:1257-1265.
Kaufman, W.K.; Mackenzie, S.A.; Kaufman, D.G. (1985) Quantitative rela-
  tionship between hepatocytic neoplasms and islands of cellular alteration
  during hepatocarcinogenesis in the male F344 rat. Am. J. Pathol. 119:171-
  174.
Kitagawa, T. (1976)  Sequential  phenotypic changes in  hyperplastic areas
  during hepatocarcinogenesis in the rat. Cancer Res. 36:2534-2539.
Kitagawa, T.; Pitot,  H.C. (1975) The regulation of serine dehydratase and
  gIucose-6-phosphatase in hyperplastic nodules of rat liver during diethyl-
  nitrosamine and N-2-fluorenylacetamide feeding. Cancer Res. 35:1075-
  1084.
Lee, G.; Tatematsu, M.; Makowka, L (1983) The development of hepatocel-
  lular carcinoma from  carcinogen altered hepatocytes implanted in the
  spleen. Identification of a new cell population with relevance to cancer.
  AACR Proc. #422.                                                .
 Moore, M.A.; Hacker, H.J.; Bannasch, P. (1983) Phenotypic instability in
  focal and nodular  lesions induced in a short term system in the rat liver.
   Carcinogenesis 4:595-603.
 Moore, M.A.; Mayer,  D.; Bannasch, P. (1982) The dose dependence and
   sequential appearance of putative preneoplastic populations induced in
  the rat liver by stop experiments with nitrosomorpholine. Carcinogenesis
   3:1429-1436.                                              fl.
 Ogawa, K.; Medline, A.; Farber, E. (1979a) Sequential analysis of  hepatic
   carcinogenesis: the comparative architecture of preneoplastic malignant,
   prenatal, postnatal and regenerating liver. Br. J. Cancer 40:782-790.
 Ogawa, K.; Medline, A.; Farber, E. (1979b) Sequential  analysis of  hepatic
   carcinogenesis: a  comparative study of the ultrastructure of preneoplas-
   tic, malignant, prenatal, postnatal and regenerating  liver. Br. J.  Cancer
   A*I "OO *3R
 Pitot] H.C.; Bareness, L.; Goldsworthy, T.; Kitagawa, T. (1978) Biochemical
   characterization of stages of hepatocarcinogenesis after a single  dose of
   diethylnitrosamine.  Nature 271:456-458.
 Popper, H.I.; Selikoff, J.; Maltoni, C.; Squire, R. A.; Thomas, LB. (1977)
   Comparisons of neoplastic hepatic lesions in man and experimental ani-
   mals. In: H.H. Hiatt, J.D. Watson, J.A. Winsten. Origins of human  cancer.

                                 20

-------
   Book C. Cold Spring Harbor,  NY: Cold  Spring Harbor  Laboratory,
   pp. 1359-1382.
 Scherer, E.; Emmelot, P. (1975) Foci of altered liver cells induced by a single
   dose of diethylnitrosamine and partial hepatectomy: their contribution to
   hepatocarcinogenesis in the rat. Eur. J. Cancer 11:145-154.
 Schulte-Hermann, R.; Tirnmermann-Trosiener, I.;  Schuppler, J. (1982) Re-
   sponse of liver foci in rats to hepatic tumor promoters. Toxicol. Pathol
   10:63-71.
 Solt, D.B.; Medline, A.; Farber, E.  (1977) Rapid emergence of carcinogen-
   induced hyperplastic lesions in a new model for the sequential analysis of
   liver carcinogenesis. Am. J. PathoK 88:595-618.
 Tatematsu, M.; Nagamine, Y.; Farber, E. (1983) Redifferentiation as a basis
   for remodeling of carcinogen-induced hepatocyte nodules to normal ap-
   pearing  liver. Cancer  Res. 43:5049-5058.
 Teebor, G.W.; Becker, F.F. (1971) Regression and persistence of hyperplastic
   hepatic nodules induced by N-2-fluorenylacetamide and their relationship
   to hepatocarcinogenesis. Cancer Res. 31:1-3.
 Van der Heijden, C.A.;  Dormans, J.A.M.A. (1981)  Short-term induction  of
   neoplastic nodules  in  the rat liver. II. Study of their development and the
   effects of withdrawal of 2-acetylaminofluorene. Carcinogenesis 2:147-
   156.
 Wanson, J.C.; de Ridder, L; Mosselmans, R. (1981) Invasiveness of hyper-
   plastic nodule cells from diethylnitrosamine-treated rat liver. Cancer Res
   41:5162-5175.
 Ward, J.M.; Sontag, J.M.; Weisburger, E.K.; Brown, C.A. (1975) Effect of
   lifetime exposure to aflatoxin BI  in rats. J. Natl. Cancer Inst.  55:107-113.
 Ward, J.M.; Lynch,  P.H.  (1984) Transplantability of naturally occurring be-
   nign and malignant neoplasms and age-asociated non-neoplastic lesions
   of the aging F344  rat as biological evidence for the histological diagnosis
   of neoplasms. Cancer Res. 44:2608-2615.
 Watanabe, K.; Williams, G.M. (1978) Enhancement of rat hepatocellular-
   altered foci by the liver tumor promoter phenobarbital: evidence that foci
   are precursors of  neoplasms and that the promoter acts on carcinogen-
   induced lesions. J. Natl. Cancer Inst. 61:1311-1314.
 Williams, G.M. (1976)  Functional  markers and  growth behavior of pren^o-
   plastic hepatocytes. Cancer Res. 36:2540-2543.
 Williams, G.M.; Watanabe, K. (1978) Quantitative kinetics of development of
   N-2-fluorenylacetamide-induced altered (hyperplastic) hepatocellularfoci
   resistant to iron accumulation and of their reversion or persistence follow-
   ing  removal of carcinogen. J. Natl. Cancer Inst. 61:113-121.
 Williams, G.M. (1980) The pathogenesis of rat liver cancer caused by chem-
   ical carcinogens. Biochim. Biophys. Acta 605:167-189.

 Sections V and VI

 Maronpot, R.R.; Boorman, G.A. (1982) Interpretation of rodent hepatocellu-
  lar proliferative alterations and  hepatocellular tumors in chemical safety
  assessment. Toxicol. Pathol. 10:71-80.
 Maronpot,  R.R.; Montgomery, C.A., Jr.; Boorman, G.A.; McConnell, E.E.
  (1986) National Toxicology Program nomenclature for rat hepatoprolifer-
  ative lesions. Submitted to Toxicol.  Pathol.
 McConnell, E.E.; Solleveld, H.A.; Swenberg, J.A.;  Boorman, G.A. (1986)
  Guidelines for combining neoplasms for evaluation of rodent carcinogen-
  esis studies. J. Natl. Cancer Inst. 76:283-289.
National  Academy of  Sciences (National Research Council). (1983)  Risk
  assessment in the Federal government: managing the process. Washington,
  DC: National Academy Press.

                                 21

-------
National Toxicology Program (NTP). (1984) Report of the NTP Ad Hoc Panel
  on Chemical Carcinogenesis Testing and Evaluation. U.S. Department of
  Health and Human Services. Public Health Service.
Office of Science and Technology Policy. (1985) Chemical carcinogens; a
  review of the science and its principles. Federal Register 50:10371-10442.
U.S. Environmental Protection Agency. (1984) Proposed guidelines for car-
  cinogen risk assessment. Federal Register 49:46294-46301.
                                 22

-------

-------

-------