EPA-600/ 9-79-001
June 1979
WATERBORNE TRANSMISSION
OF GIARDIASIS
Proceedings of a Symposium
September 18-20, 1978
Sponsored by the
Health Effects Research Laboratory
and the
Municipal Environmental Research Laboratory
Edited By
W. Jakubowski and J. C. Hoff
Do not remove. This document
should be retained in the EPA
Region 5 Library Collection.
U.S. ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF RESEARCH AND DEVELOPMENT
ENVIRONMENTAL RESEARCH CENTER
CINCINNATI, OHIO 45268
This document is available to the public through the National Technical Information Service,
Springfield, Virginia 22151
-------
DISCLAIMER
This report has been reviewed by the Health Effects Research Laboratory,
and the Municipal Environmental Research Laboratory, U.S.
Environmental Protection Agency, and approved for publication. Approval
does not signify that the contents necessarily reflect the views and policies of
the U. S. Environmental Protection Agency, nor does mention of trade
names or commercial products constitute endorsement or recommendation
for use.
-------
Ill
FOREWORD
The U. S. Environmental Protection Agency was created because of
increasing public and government concern about the dangers of pollution to
the health and welfare of the American people. Noxious air, foul water, and
spoiled land are tragic testimony to the deterioration of our national
environment. The complexity of that environment and the interplay
between its components require a concentrated and integrated attack on the
problem.
Research and development is that necessary first step in problem solution
and it involves definingthe problem, measuring its impact, and searching for
solutions. The primary mission of the Health Effects Research Laboratory
in Cincinnati (HERL) is to provide a sound health effects data base in
support of the regulatory activities of the EPA. To this end, HERL conducts
a research program to identify, characterize, and quantitate harmful effects
x of pollutants that may result from exposure to chemical, physical, or
<> biological agents found in the environment. In addition to the valuable
health information generated by these activities, new research techniques
and methods are being developed that contribute to a better understanding
of human biochemical and physiological functions, and how these functions
are altered by low-level insults.
The Municipal Environmental Research Laboratory develops new and
improved technology and systems for the prevention, treatment, and
management of wastewater and solid and hazardous waste pollutant
discharges from municipal and community sources, for the preservation and
treatment of public drinking water supplies, and to minimize the adverse
economic, social, health, and aesthetic effects of pollution. This publication
is one of the products of that research; a most vital communications link
between the researcher and the user community.
This report represents an attempt to determine the current state-of-
knowledge on the transmission of a pathogenic protozoan agent through
drinking water and to define future research needs. With a better
understanding of the health effects, measures can be developed to reduce
exposure and to provide adequate treatment in order to eliminate this
organism as a significant cause of waterborne disease.
R. John Garner, Director Francis T. Mayo, Director
Health Effects Research Municipal Environmental Research
Laboratory Laboratory
-------
IV
PREFACE
The occurrence of recent community-wide outbreaks of giardiasis
associated with drinking water supplies has resulted in the recognition that
basic information on the etiologic agent and mode of transmission is lacking.
The U. S. Environmental Protection Agency has a mandate under the Safe
Drinking Water Act of 1974 to identify contaminants that may cause adverse
health effects and to either promulgate maximum contaminant levels for
drinking water or specify treatment requirements. In order to determine
what measures are required to protect the public health, reliable data are
needed on the occurrence, epidemiology, and adequacy of available
treatment technology. The purpose of this Symposium was to bring together
scientists, public health administrators, and water suppliers to discuss
various aspects of giardiasis and to provide in a single publication a
comprehensive summary of current knowledge.
The Proceedings are organized into seven main sections corresponding to
the format of the Symposium and addressing the organism, the disease,
epidemiology, detection methods, other recent research, treatment
technology, and research needs. One section consists of abstracts of papers
presented at an evening session (Session VI - Other Recent Research) and
represents relevant research brought to the attention of the Symposium
coordinators after the program had been finalized.
In many cases, the Proceedings papers are more comprehensive than the
Symposium presentations because of the desire to provide thorough
coverage of a given topic. The section on epidemiology contains a paper not
presented at the Symposium. As a result of a suggestion made at the
Symposium, the coordinators sought and obtained consent from Dr. Lyle
Veazie to incorporate in the Proceedings an unpublished manuscript on a
suspected outbreak of giardiasis affecting an estimated 50,000 persons in
Portland, Oregon in 1954-55. Information concerning this event had
previously been reported only in the form of a letter to the New England
Journal of Medicine (281:853, Oct. 1969).
Edited discussions are included with papers and an attempt has been made
to identify each questioner where possible. A list of registrants is presented,
thus enabling the reader to contact any speaker for further information.
Walter Jakubowski
John C. Hoff
Editors
-------
ABSTRACT
The Health Effects Research Laboratory and the Municipal
Environmental Research Laboratory of the U. S. Environmental Protection
Agency sponsored a National Symposium on Waterborne Transmission of
Giardiasis in Cincinnati on September 18-20, 1978.
The Symposium brought together scientists, engineers, and federal, state,
and local public health officials for the purpose of determining the state-of-
knowledge regarding the etiologic agent and the disease as they relate to
water supplies. Sessions on the organism, the disease, epidemiology,
detection methodology, water treatment technology, and research needs
were held.
The Proceedings consist of 19 invited manuscripts, 6 abstracts of
contributed papers and transcripts of discussions following each
presentation. A panel discussion on research needs is included.
-------
ACKNOWLEDGMENTS
The assistance of the many individuals who contributed to the success of
this Symposium and the timely completion of the Proceedings is gratefully
acknowledged. Special appreciation is due to the speakers, for the quality of
their presentations and promptness in submitting their papers, the session
chairmen and the many participants who contributed to the discussions.
We also wish to acknowledge the assistance of the Office of Support
Services and in particular the efforts of Ms. Kathy Burleigh, of that Office, in
providing the many administrative services in connection with arrangingthe
Symposium. We also are indebted to Dixie White, Saundra Underwood and
Shirley Jakubowski for assisting with registration, and to Steve Waltrip and
Judy Barnick who helped in many ways to make the Symposium a success.
-------
Vll
CONTENTS
Foreword iii
R. J. Garner, F. T. Mayo
Preface iv
W. Jakubowski, J. C. Hoff
Abstract v
Acknowledgments vi
Keynote Address xi
G. G. Rtibeck
SESSION I: THE ORGANISM i
G. J. Jackson, Session Chairman
Giardia lamblia: Classification, Structure, Identification 2
N. O. Levine
Ultrastructural Aspects of Giardia 9
H. G. Sheffield
Surface Morphology of Giardia Cysts Recovered
from a Variety of H osts 22
A. S. Tonibes
SESSION II: THE DISEASE 38
W. Jakubowski, Session Chairman
Managing the Patient with Giardiasis: Clinical,
Diagnostic and Therapeutic Aspects 39
M. S. Wolfe
The Possible Use of an Indirect Immunofluorescent
Test Using Axenically Grown Giardia lamblia Antigens
in Diagnosing Giardiasis 53
G. S. Visvesvara, G. R. Healv
The Experimental Transmission of Giardia lamblia
among Volunteer Subjects 64
R. C. Rendtorff
Giardiasis in the Mouse: Clues to Host Immune
Mechanisms 82
D. P. Stevens
SESSION III: EPIDEMIOLOGY 91
M. G. Schultz, Session Chairman
The Presence and Absence of Giardia lamblia in
Studies on Parasite Prevalence in the U.S.A 92
G. R. Healv
-------
VIII
Animal Reservoirs and Cross-Species Transmission
of Giardia 104
R. B. Davies, C. P. Hibler
Waterborne Outbreaks of Giardiasis 127
G. F. Craun
Waterborne Giardiasis 150
D. Juranek
Water Supply Problems Associated with a Waterborne
Outbreak of Giardiasis 164
E. C. Lippy
An Outbreak of Gastroenteritis Associated with
Giardia lamblia 174
L. Veazie, I. Brownlee, H. J. Sears
SESSION IV: DETECTION METHODOLOGY 192
J. C. Hoff, Session Chairman
Methods for Detection of Giardia Cysts in Water Supplies 193
W. Jakubowski, T. H. Ericksen
The Propagation of Giardia Trophozoites In Vitro 211
E. A. Meyer
Induction on Giardia Excystation and the Effect of
Temperature on Cyst Viability as Compared by Eosin-
Exclusion and In Vitro Excystation 217
A. K, Bingham, E. L. Jarroll, Jr., E. A. Meyer,
S. Radulescu
SESSION V: WATER TREATMENT
TECHNOLOGY 230
D. W. Liechty, Session Chairman
Disinfection Resistance of Giardia Cysts: Origins
of Current Concepts and Research in Progress 231
J. C. Hoff
Water Filtration Techniques for Removal of Giardia
Cysts and Cyst Models 240
G. S. Logsdon, J. M. Symons, R. L. Hoye, Jr.
SESSION VI: OTHER RECENT RESEARCH
(ABSTRACTS) 257
W. Jakubowski, Session Chairman
In Vitro Model for Analyzing the Attachment/ Release
of Trophozoites of Giardia'. Evidence for the
Involvement of Contractile Proteins and the Effect
of Atabrine 258
L. S. Erlandsen, D. E. Feelv, J. V. Schollmeyer, D. G. Chase
-------
IX
Report of Endosymboisis in Giardia muris and Comparison
of Organelle Distribution in Giardia muris and Giardia
lamblia 261
P. Nemanic, R. L. Owen, D. P. Stevens, J. C. Mueller
Intestinal Distribution, Attachment and Relationship
of Giardial Trophozoites to Peyer's Patches During
Acute Infection in the Immunologically Intact Host 264
R. L. Owen, P. Nemanic, D. P. Stevens
Giardiasis in Vail, Colorado 267
W. Jones, T. Edell
The Presence of Giardia lamblia Cysts in Sewage and
Sewage Sludges from the Chicago Area 268
J. C. Fox, P. R. Fitzgerald
Axenically Cultivated Giardia lamblia: Growth,
Attachment, and the Role of L-Cysteine 270
F. D. Gillin, L. S. Diamond
SESSION VII: RESEARCH NEEDS 273
G. G. Robeck, Session Chairman
Giardiasis and the Safe Drinking Water Act 274
C. Hendricks
Panel Discussion: Current Research and Future
Research Needs 287
G. Robeck, L. J. McCabe, M. G. Schultz, T. M. Vernon,
Jr., K. Kawata
REGISTRATION LIST 298
-------
-------
\l
KEYNOTE ADDRESS
GORDON G. ROBECK
COORDINATOR OF DRINKING WATER RESEARCH
U. S. ENVIRONMENTAL PROTECTION AGENCY
As Mr. F. Mayo indicated in his welcoming speech, I have a dual role in
the area of drinking water research: one is to direct the treatment activities
within the Municipal Environmental Research Laboratory, and the second
is to try to help some of our planners in Washington, D.C. at headquarters to
select priorities and thus allocate available resources.
The administrator of the U. S. EPA Office of Research and Development,
Dr. Stephen J. Gage, has been trying to integrate the various research needs
through a Research Committee. There are many subumts in his office that he
is experimenting with now to see how well this integrating process will work.
He has formed five Research Committees and named managers for each of
these to try to elicit the needs of the cities, regions and states, and to see what
the operating programs in Washington need concerning existing legislative
mandates. It is in the capacity of a Research Committee Manager that I
speak to you today.
As you know, EPA is a regulatory agency, but we are delighted to be able
to spend a certain amount of the agency's resources on research, designing
our studies to provide a basis for intelligent regulatory action.
As you have heard from our two welcoming speakers, we are here today
because we have had an upsurge in the occurrence of giardiasis, particularly
in certain areas of the United States. It is somewhat puzzling why it is
regionalized, and ironically, in those areas that we frequently call protected
watersheds. It is the usually relatively good mountain water that seems to be
contributing to the outbreaks. Although better reporting of disease may be a
factor, we became particularly apprehensive when we noticed that even
where we had reasonably good treatment systems in place we still had
outbreaks. As an engineer, that concerned me and made me realize that it
was time for a better understanding of several scientific and technical
questions, as well as for a better enforcement and training program.
However, we had no specific goal or standard for enforcement, so we
realized that we had to do more research in order to advise the policy makers
and regulators.
For instance, is a combination of the current turbidity limit of one to five
and the coliform limit of a monthly average of one adequate to protect the
consumer against giardiasis? We have had some outbreaks where both limits
have been met, so we are not totally convinced that we have an adequate
safety factor built into the microbiological and physical limits.
-------
XII
EPA's Office of Drinking Water has been contemplating the use of a
section in the Safe Drinking Water Act that allows setting a treatment
requirement when it is difficult to monitor for a specific contaminant. That
option, of course, may involve requiring coagulation and filtration, followed
by disinfection, particularly for surface waters. I hope this symposium will
help guide us to some sensible conclusion with regard to the needs for future
research to support necessary Agency action.
I have been asked to set the tone for this symposium, and to do so 1 must of
course briefly review what we know from the literature that is currently
available and what we know also from recent studies that many of you have
participated in, and lastly, where we are going in the Agency.
What we know about the organism is that it was reported in the scientific
literature as a known entity almost three hundred years ago. It had long been
considered non-pathogenic because many people shed the organism and did
not seem to be ill. Now, most public health people consider it as a disease to
be reckoned with, and, we certainly think we know in general terms what the
causative agent is and perhaps how to control the disease through treatment
as far as water being the medium for transmission.
As most of you know, it has two stages in its life cycle, one the
reproductive trophoroite, and the other the cyst stage. It is as a cyst, of
course, that it is usually shed by most people. The populations of cysts from
different hosts, and even from the same host, actually exhibit variations in
size and shape, and we do know that there is a difference in the internal
structure of some of these. However, wedo not know the number of species
of these organisms that exist, nor the host range of the species. Other
mysteries are the mechanisms that trigger excystment and encystment.
Through the use of the electron microscope, serologic studies and animal
infection studies, we may be able to shed some light on further characterizing
this organism and determining the relationship of possible host reservoirs to
waterborne transmission.
Now, what do we know about the disease itself? The infection is
manifested in a variety of ways ranging from asymptomatic to acute disease
with severe diarrhea, weight loss, and sometimes retarded growth in
children. The occurrence of the disease as well as the efficacy of treatment
may depend upon inherent characteristics of the host including
immunologic status and the underlying disposition of the individual, or, on
the other side of the equation, upon the virulence of the organism itself. The
infectious dose appears to be quite low and a large number of cysts can be
shed from infected individuals. The serologic response may depend upon the
severity of the infection and possible adaptive mechanisms in the organism
itself.
We need better methods for diagnosis in order to more accurately assess
the true incidence of infection and disease. Also, there are only two drugs of
choice for treatment in this country now. One has been shown to be
carcinogenic in mice and rats and the other may result in side effects more
severe than the infection itself. The possibility that a Giardia infection may
-------
Mil
predispose individuals toward ulcers and gallbladder disease or other
conditions should be explored.
The other major area that we want to discuss in this Symposium is the
epidemiology of the disease. Here we have studied and observed worldwide
that giardiasis is endemic in some places and epidemic in others. The
underlying attitude for many years was that it might be a tropical disease, but
this has been somewhat negated by evidence that there have been infections
and outbreaks in the Yukon, in Russia and in the northern mountainous
areas of the United States. These places obviously are not tropical.
We had gradually accumulated the impression that giardiasis was a
disease of hikers, campers, hunters, skiers and fishermen who drink
untreated water from mountain streams, but in truth we have had more than
20 outbreaks that have been documented in the United States alone since the
outbreak in Aspen, Colorado in 1965. Between 1966 and 1974, we have had
14 outbreaks affecting about 400 people, and they generally have occurred in
small municipal and semi-public systems in recreational areas. However, in
the last four years there have been community-wide outbreaks associated
with municipal water supplies resulting in over 6300 cases.
The epidemiologists also tell us that the organism infects a wide variety of
domestic and wild animals that may act as reservoirs of infection and
certainly sources of contamination of water supplies. Beavers, for instance,
were implicated in the Camas, Washington outbreak of 1976. Therefore, we
need to determine the sources of contamination of our water supplies,
understand the seasonal occurrence in domestic and wild animals and
human populations, and also determine the significance of water supplies for
transmitting and maintaining the infection throughout the population.
How well can we quantify and identify this organism? Actually, there is no
good cultural method for detecting and enumerating the organism, and just
to give some perspective as to where we are in our inability, just imagine
gathering data on coliform levels in water by doing microscopic counts of the
organisms. That is basically where we are with Giardia, During the Rome,
New York outbreak of 1974-1975, the Center for Disease Control developed
a sand filter that they hoped would allow them to concentrate the cysts in
large enough numbers to be able to detect them microscopically and to
determine their infectivity. This method was successful but the size of this
unit, unfortunately, made it very difficult to transport and to collect samples
in areas a considerable distance away from Atlanta. The Health Effects
Research Laboratory here at EPA developed a portable water sampling
device that has since been successfully used to detect the cysts in the Camas,
Washington; Berlin, New Hampshire; and, Vail, Colorado outbreaks.
However, this method remains to be evaluated under a variety of sampling
conditions. The microscopic detection of cysts does not supply information
about the species of origin or viability, but recent work at the University of
Oregon may lead the way to development of a cultural technique and a
method for determining viability.
-------
XIV
What do we know about controlling the occurrence of the disease,
presuming that it can be spread not only by humans contaminating
watersheds, but by infected animals, which may be very difficult to control?
Outbreaks have occurred, as I have mentioned, in several states, but there
was no treatment to speak of or else there had been a breakdown in the
existing treatment. The effect of various disinfectants, such as chlorine, or
some of the suggested new ones such as ozorie, ultraviolet radiation or
chlorine dioxide, are not well known. We had been trying to extrapolate
from Entamoeba histolytica data, and you will be discussing later this week
the appropriateness of that laboratory information to the Giardia situation.
We certainly need to determine the optimum parameters for removal of cysts
by flocculation, filtration and by disinfection, particularly in the small
systems. Throughout the last year or so we have found through
implementation of the Safe Drinking Water Act that most noncompliance
situations develop in the small systems. Therefore, we are turning more and
more of our resources and research toward addressing that problem.
It seems that many people have reasonsed in the past that we have the
technology for controlling not only organisms but certain factors such as
turbidity created by clay and other runoff products, as well as certain
chemicals, but the newness of the Safe Drinking Water Act, and its
application to such small systems, soon taught us that most of the small
utilities were simply not aware of the feasibility of using some of the available
technology. We must have good training programs, and less expensive
systems which are easier to operate than current systems. We also must make
sure that there are skilled people commensurate with the sophistication of
equipment.
We feel that the Administrator of EPA will find the results of this
conference a great aid to his decision making process. If we can develop the
appropriate information, we may be able to show that treatment of all
surface waters by coagulation and filtration is a necessity, and if operated
properly, we can minimize the amount of disinfectant used. In the past,
superchlorination and dechlorination have been used as a control measure
for Giardia lamblia, but this could result in creating trihalomethanes, and
this then means that one public health practice is resulting in another public
health problem.
1 want to once again thank you for attending and 1 look forward to hearing
your presentations and the ensuing discussions.
-------
SESSION I - THE ORGANISM
Chairman - George J. Jackson,
Food and Drug Administration, Washington, D.C.
Giardia lamblia: Classification,
Structure, Identification
N. D, Levine
Ultrastructural Aspects of Giardia
H. G. Sheffield
Surface Morphology of Giardia Cysts
Recovered from a Variety of Hosts
A. S. Tombes
-------
Giardia Lamblia: Classification, Structure,
Identification
Norman D. Levine
College of Veterinary Medicine and Agricultural Experiment Station, Univ.
of Illinois, Urbana, Illinois
ABSTRACT
Giardia lamblia Kofoidand Christiansen, 1915 is a flagellated protozoon belonging to the
class Zoomastigophorasida, order Diplomonadonda, family Hexamitidae. It is found in
the small intestine of mas and other mammals. Its trophozoites are 9 to 21 /jm long, 5 to 15
/jm wide and 2 to 4 /xm thick. They have a piriform body with a broaely rounded anterior
end, 2 nuclei, 2 slender median rods, 8 flagella in 4 pairs, a pair of darkly staining median
bodies, and a large, ventral sucking disk Its cysts are ovoid, 8 to 12 x 7 to 10 jum, with 4
nuclei and remnants of the organelles. Its trophozoites are easily identified in moist
scrapings of the intestinal mucosa from their shape and characteristic motion. They can
also be identified in fixed and stained smears from their structure. Its cysts can be
concentrated by zinc sulfate (but not sugar) flotation, and can be identified from their
structure
CLASSIFICATION
Giardia lamblia is a protozoon which has generated much interest in
recent years. It has been taken as the emblem of the International Congresses
of Protozoology, the first of which was held in Prague, Czechoslovakia in
1961 and the fifth in New York City in August, 1977. Giardia lamblia has
also emerged as a cause of human disease. The classification of the genus
would be a simple matter if it were based on modern knowledge alone, but
the nomenclature of its species is another matter.
To elucidate; Leeuwenhoek was the first to see Giardia—in 1681 in his
own stools. The genus itself was named by Joseph Kunstler in 1882; Raphael
Anatole Emile Blanchard called it Lamblia in 1888, but he was 6 years too
late, and this generic name is a synonym of Giardia. Despite this fact, some
Europeans still use the name Lamblia. Reichenow(l) compromised; he
called the frog species Giardia agilis and the human one Lamblia intestinalis.
Vilem Dusan Fedorovic Lambl found the human organism in 1859 and
called it Cercomonas intestinalis. In 1879 Giovanni Battista Grassi found
Giardia in the house mouse and called it Dimorphus muris; in 1881 he found
it in man and called it Megastoma entericum. Unfortunately, he also used
this same name for giardias in Mus, Apodemus and the cat, and thus
compromised it. The species name intestinalis was found to be a homonym
of the organism that Diesing had called Cercomonas intestinalis in 1850, and
the species name enterica was suppressed as a synonym of Giardia muris.
-------
N. D. Levine 3
Most Europeans still call the human species Giardia intestinalis, ignoring the
fact that this species name is a homonym. Corradetti(2) made a good case for
calling it Giardia enterica. Although not stated by Corradetti(2), the median
bodies of the human form are curved bars resembling the claw of a claw-
hammer while those of G. muris are small and rounded. Since this is the case,
G. enterica cannot be a synonym of G. muris, and the name is available for
the human species.
There had been so much confusion about the availability of the specific
names, intestinalis and enterica, that Charles Wardell Stiles established the
name Giardia lamblia in a letter to Kofoid and Christiansen(3); therefore,
this is the name that we use.
Using the light microscope, Filice(4) was unable to find any structural
difference between the giardias of the laboratory rat and a number of wild
rodents. On reviewing the literature, he discovered that almost no acceptable
cross-transmission studies existed between some species. He appears to have
concluded that there are only 2 species in mammals, each having a number of
races and that these species differ only in their median bodies. G. muris
occurs in the house mouse, rat and hamster, and has rounded median bodies.
G. duodenalis (Davaine, 1875) occurs in the rabbit, cottontail, Norway rat
(G. simonij, golden hamster, guinea pig, man, and presumably ox, dog, cat,
etc., and has median bodies shaped like the claws of a claw-hammer. Since
duodenalis was the first name (barring intestinalis and enterica) given to a
claw-hammer Giardia, this was the species name he used. Bemrick(S) agreed
with this point of view, and accepted only the names G. muris and G.
duodenalis for the mammalian species.
Many individuals,- -myself among them—have used different names for
the claw-hammer giardias from different hosts. We speak of G. lamblia or G.
intestinalis or G. enterica from man, G. canis from the dog, G. cati from the
cat, G. bovis from the ox, G. duodenalis from the rabbit, G. simonifrom the
rat and mouse, etc. This practice, which is a matter of convenience, ignores
the possibility that they are all the same species and can be transmitted from
one host to another. At present it would appear that cross-transmission
between various hosts is common. This, of course, is one of the reasons for
our present symposium. If this is the case, I suppose that the purists among
us will want to call the organism Giardia enterica (Grassi, 1881).
Now, forgetting this aspect of the problem, how about the classification of
the genus? Giardia is a flagellate, and is classified as follows:
Phylum-SARCOMASTlGOPHORA Honigberg & Balamuth, 1963
With a single type of nucleus, vesicular; sexuality, if present,
syngamy; with flagella, pseudopodia, or both types of locomotory
organelles; typically no spore formation.
Subphylum-MASTIGOPHORA Diesing, 1866
One or more flagella typically present in trophozoites; asexual
reproduction basically by longitudinal binary fission; sexual
reproduction unknown in many groups.
-------
4 WATERBORNK GIARDIASIS/THE ORGANISM
Class-ZOOMASTIGOPHORASIDA Calkins, 1909
Chromatophores absent; 1 to many flagella; amoeboid forms with or
without flagella in some groups; sexuality absent in most groups;
predominantly parasitic.
Order-DIPLOMONADORlDA Wenyon, 1926
With 2 karyomastigonts, each with 4 flagella; at least 1 of these
flagella is recurrent; with 2 nuclei; without mitochondria or Golgi
apparatus; with cysts; free-living or parasitic.
Family-HEXAMITIDAE Kent, 1880
With at least 1 recurrent flagellum per karyomastigont.
This is the family within which Giardia occurs, and the only family in the
order. It contains 5 genera. Brugerolle(6,7) made an electron microscope
study of the group and divided this family into 2 subfamilies, based on
whether there are functional cytostomes or not. In the subfamily
Hexamitinae Kent, 1880, each recurrent flagellum lies in a hollow tube or
depression which forms a functional cytostome; whereas in the subfamily
Octomitinae the recurrent flagella lie in the cytoplasm itself and there are no
cytostomes.
In the subfamily Hexamitinae there are 3 genera: Trepomonas Dujardin,
1838, with 2 flagella directed forward and 6 directed backward and lying in 2
lateral fossae or cytostomes; Hexamita Dujardin, 1838, with a broad body,
subspherical nuclei, and 6 flagella directed forward and arising external to
the nuclei, and with the 2 recurrent flagella in cytostomal tubes; and
Spironucleus Lavier, 1936, like Hexamita but with an elongate body and
elongate, S-shaped nuclei.
In the subfamily Octomitinae there are 2 genera: Octomitus von
Prowazek, 1904, with 6 anterior and 2 recurrent flagella, without a sucking
disk, and Giardia Kunstler, 1882, with the same number of flagella as
Octomitus, but with a sucking disk.
The evolutionary relationships of these genera are shown in Fig. 1, taken
from Brugerolle(6,7). Those relationships are different from illustrations
based on light microscope studies which are found in textbooks. Note that
Brugerolle considered that Trepomonas arose from Enteromonas by
longitudinal doubling. He was so sure of this that he put Enteromonas in the
Diplomonadorida even though it is not doubled. Levine(8), called it Incertae
Sedis because he didn't know where to put it. Brugerolle said that the 5
genera—Trepomonas, Hexamita, Spironucleus, Octomitus and Giardia—
form a linear series, each one arising from the one below it. The parasitic
genera arose from the free-living ones, which are shown to the left of the
dashed lines.
All species of Trepomonas are free-living, some species of Hexamita are
free-living and some are parasitic and all species of Spironucleus, Octomitus
and Giardia are parasitic. The diagram (Fig. 1) indicates that Enteromonas
is free-living but this is not true. It has been found in the intestine of a number
of mammals, including man, pigs, rats and rabbits. The forms that
-------
/V. D. Levine
DIPLOMONADIDA
llbres
Enteromonas
FIG. 1. Evolutionary relationships of genera in the family Hexamitidae(6).
Brugerolle studied came from the intestines of a salamander and the
domestic rabbit.
I think that I have confused you enough. Fig. 2 is a drawing of the
organism under discussion today—Giardia lamblia or whatever you want to
call it. The drawing shows both the trophozoite and the cyst.
STRUCTURE
The fine structure of Giardia as seen through the electron microscope will
be discussed by others (see Sheffield, these Proceedings). However, 1 think
that 1 might review for you what you would see with the light microscope.
The body is rather pear-shaped, with a broad anterior end which comes to
a blunt point posteriorly. It is bilaterally symmetrical. There is a large
sucker, or sucking disk, on the ventral side; the dorsal side is convex. There
are 2 anterior, vesicular nuclei, and 2 recurrent flagella running back
-------
WATERBORNE GIARDIASIS/THE ORGANISM
FIG. 2. Trophozoite(A)andcyst(B)ofG\*T&\*
lamblia. A after Filice (4); B from
Becker & Frye(lO).
through the middle of the body. These have been referred to as median rods
or axostyles; however, they really are not axostyles. There are 4 pairs of
flagella. (They have been cut off in the drawing—and are actually much
longer than shown here). There are 2 median bodies composed of bundles of
microtubules arranged either irregularly or sometimes united in ribbons.
They disappear in the course of division and perhaps have something to do
with formation of the new sucking disk.
The cysts have 4 nuclei when mature, plus a variable number of fibrillar
remnants of the trophozoite organelles.
IDENTIFICATION
Giardia can be identified either while alive—at least the trophozoites can
be—or after fixation and staining. To identify the live trophozoites, make a
scraping of the duodenum, mix with a little physiological NaCl solution, and
examine under the oil immersion objective of the microscope. They are 9 to
21 p.m long, 5 to 15 /urn wide and 2 to 4 fj,m thick. They can be readily
recognized from their appearance and from their sort of flopping motion.
The cysts are ovoid to ellipsoidal, 8 to 12 x 7 to 10 ^m.
Trophozoites and cysts can also be identified after fixation and staining.
For fixation Schaudinn's fixative (2 parts saturated mercuric chloride
solution and 1 part 95% alcohol, plus 5% acetic acid added immediately
before use) is recommended. For staining, Heidenhain's hematoxylin is best.
Directions for its preparation and use are given in Levine(8).
Cysts can be concentrated by flotation with a solution of greater density.
Such flotation solutions are used routinely for fecal examination, but can
also be adapted to use with water samples. They are worthless for
-------
N. D. Levine 1
trophozoites, however, and suitable only for cysts, which may or may not be
present in water.
Not all flotation solutions are suitable. Sugar solutions are unsatisfactory,
and I suspect that NaCl solutions are also. The cyst wall is actually quite
delicate, and 1 have watched Giardia cysts that have been concentrated by
sugar flotation shrivel and become unrecognizable in a matter of minutes. I
recommend zinc sulfate flotation. Hampel(9) developed a modification of
this technic, layering 1% potassium alum solution on the surface of the zinc
sulfate solution, recentrifuging, and staining with 1% eosin solution.
REFERENCES
1. Reichenow, E. 1949-1953. Lehrbuch der Protozoenkunde. Fischer, Jena, 6th ed., 1213 pp
2. Corradetti, A. 1935. Giardta entenca (Grassi, 1881), the correct name for the Gtardia of
man. J. Parasit., 21:310-311.
3. Kofoid, C. A. and E. B. Christiansen. 1915. On the life-history of Giardia Proc. Nat. Acad
Sci., 1-547-552.
4. Filice, F. P. 1952. Studies on the cytology and life history of a Giardia from the laboratory
rat. Univ. Calif. Publ^Zool., 57:53-143.
5. Bemrick, W. J. 1962'The host specificity of Giardia from laboratory strains of Mus
musculus and Rail us norvegicus J Parasit., 48.287-290.
6. Brugerolle, G. 1975. Etude ultrastructurale du genre Enleromonas da Fonseca
(Zoomastigorphorea) et revision de 1'ordre des Diplomonadida Wenyon. J. Protozool ,
22:468-475.
7. Brugerolle, G. 1976. Contribution a 1'etude cytologique des proto?oaires zooflagelles
parasites: Proteromonadida, Retortamonadida, Diplomonadida, Oxymonadida,
Tnchomonadida. Thesis, Universite de Clermont 11, Clermont-Ferrand, France.
8. Levine, N. D. 1973. Protozoan Parasites of Domestic Animals and of Man. Burgess,
Minneapolis, 2nd ed., ix + 406 pp.
9. Hampel, M. 1969 An improved method for flotation of cysts of Giardia lamblia. Parasit.
Hung, 2:74.
10. Becker, E. R. and W. W.Frye. 1927. Some protozoa found in the feces of cattle. Proc. Iowa
Academy of Science, 34:331-333
Discussion
W. JAK.UBOWSKI: You mentioned median body differences in the
muris and lamblia trophozoites. Do these differences carry over into the cyst
stage?
N. LEVINE: As far as 1 know, they do.
W. JAKUBOWSK1: Then you believe it would be possible to
differentiate cysts of muris and other species by the appearance of the
median bodies?
N. LEVINE: If you get them early enough, but later on the median bodies
will degenerate and you will not find them at all.
W. JAKUBOWSKI: You also mentioned the maturation period required
for cysts and the presence of four nuclei in mature cysts. Can you say
anything about how long this maturation period might take and what
changes go on?
N. LEVINE: I cannot tell you. It takes a while.
B. MYERS: In speciation, how do you distinguish rodent Giardia from
Giardia in beaver, dog and so on?
N. LEVINE: If they have claw hammer type median bodies, the only way
that I can tell them apart is by measurements, and that is not very good. I
have looked at Giardia in the dog, various ruminants and humans, and I
cannot tell them apart. They all look the same except for muris. But in mice
-------
8 WATKRBORNK GIARDIASIS/THE ORGANISM
and rats, also, you have not only muris but simoni with claw hammer median
bodies.
SPEAKER: Could you give a little more detail on those flotation
techniques? What was the problem with the sugar flotation?
N. LEVINE: The sugar must penetrate and shrink the cysts. They shrivel
up and you cannot recognize them at all after a few minutes. That is why 1
think you have to use zinc sulfate for flotation.
The technique itself is rather simple. Use sucrose in water, and a little
carbolic acid (phenol) as a preservative. Then use half of that and half of the
fecal suspension in water or physiological salt solution, and either centrifuge
it for five minutes or let it stand for about a half hour. 1 use an ordinary test
tube that has been ground off perfectly flat on top and place a round cover
slip on it, and then pick off the cover slip and place it on a slide for
examination.
D. STEVENS: This is a minor detail, but we have used a sucrose flotation
technique, and we watched cysts immediately after removing them from the
interface and placing them in physiological saline, and we have not noted the
same problem in morphology, but this is with a rodent strain. As long as you
pull them right off, there does not seem to be a problem.
N. LEVINE: Yes, it ought to work all right. I never tried that. I have just
used the standard flotation technique.
G. HEALY: 1 remember reading a review paper in French on the species
of Giardia, (Ansari, M.A.R., Pakistan J. Health, 4:131, 174, 1954-55), and
the thing that bothered me was that he had some pretty good evidence for
tremendous differences in cyst size of various species. I wonder what your
thought is about this in terms of having differentiation based on the claw
hammer or the smooth type of median bodies.
N. LEVINE: I am suspicious of anything that depends entirely on size
because it is possible that the size may be just a race character or be
influenced by the host, so I just do not know.
-------
The Ultrastructural Aspects of
Giardia
Harley G. Sheffield
Extramural Activities Program
National Institute of Allergy and Infectious Diseases
National Institutes of Health
Bethesda, Maryland
ABSTRACT
Transmission and scanning electron microscopy has provided new information on the
structure and possible function of the various organelles in the trophozoite and cyst of
Giardia spp. The adhesive disk consists of numerous parallel microtubules lying adjacent
to the ventral plasma membrane The microtubules are attached to each other by
filamentous bridges and each has a ribbon-like structure extending dorsally from it.
Contraction of the adhesive disk, attachment by the ventrolateral flange, or negative
pressure produced by flagellar motion under the disk has been proposed as a means of
attachment of the organism to the host intestine Median bodies have been identified as
groups of microtubules. Numerous vacuoles along the periphery of the organism have been
proposed as feeding organelles or secretory vesicles involved in encystment. These various
organelles are also seen in the cyst but are often disorganized In addition, the cyst has a
thin fibrous cyst wall to protect it trom the environment outside of the host. This paper
reviews the ultrastructure of the trophozoite and cyst
Giardia lamblia, which infects man, and various other species of the genus
Giardia have been of interest to light microscopists for many years. In fact,
G. lamblia was among the first of the parasitic protozoa to be seen with the
microscope. Judging from his detailed description, Leeuwenhoek probably
observed the trophozoite of G. lamblia while studying his own feces( 1). Over
the years, the morphological characteristics of Giardia have been studied by
many workers. The distinct appearance of the trophozoite has amused
biology students as well as being used as the emblem for the International
Congresses of Protozoology. When electron microscopes became available
to protozoologists, Rossi-Espagnet and Piccardo(2) were the first to
examine Giardia. However, their technique of utilizing whole cells did not
yield sufficient resolution to augment existing knowledge derived by light
microscopy. In this presentation, the morphology of the trophozoite and the
cyst of Giardia as determined by transmission and scanning electron
microscopy will be reviewed.
The trophozoite, when viewed from the dorsal surface, is pear-shaped
with the anterior end broadly rounded and the posterior end pointed. The
dorsal surface is convex whereas the ventral surface is somewhat concave.
Eight bilaterally paired flagella are present. Their basal bodies arise near the
-------
to
WATERBORNE GIARDIASIS/THE ORGANISM
midline at the level of the bilaterally situated nuclei. Two of the flagella
emerge anterolaterally, two posterolaterally, two ventrally and two
caudally. The distinctive sucking disk occupies most of the anterior ventral
surface (Fig. 1).
FIG. 1. Giardia sp. in dog small intestine. 10,SOOX. TropHozoite attached to microvillar
border of an epithelial cell. Abbreviations for all figures: AS, axostyle, AX,
axoneme; CW, cyst wall; L, lacuna; MB, median body; MV, micro villi;PM, plasma
membrane; S, endosymbiont; SD, striated disk; SL, striated layer; V, vacuole; VF,
ventrolateral flange.
The fine structure of the sucking disk was first demonstrated by
Cheissin(3) who studied Lamblia (Giardia) duodenalis obtained from
rabbits and grown in culture. He described the sucking disk as having two
lobes which were fused between the nuclei. The pellicle of the sucking disk
had a characteristic striation due to alternating light and dense lines. These
lines, or "ridges" as he termed them, were interpreted to be infoldings of an
inner membrane which was continuous and located just internal to the
ventral plasma membrane. The ridges were concentric within the centers of
the lobes but were longitudinally arranged at the margins. At the midline,
anterior to the nuclei, was an area devoid of the striations which was termed
the "naked area". Morecki and Parker(4) studied Giardia lamblia obtained
from human biopsies. They also found that the sucking disk occupied the
ventral portion of the entire cephalic pole, except for the naked area, and a
caudal extension of the latter termed the ventral groove. They also described
two additional lobes which were situated on either side of the ventral groove.
These lobes fused at their caudal ends. The structure of the sucking disk
consisted of ridges which were perpendicular to the ventral surface and
extended 100 to 200/xm into the cytoplasm. In contrast to the observations of
Cheissin(3), Morecki and Parker(4) described the ridges as electron-opaque
-------
H. G. Sheffield 11
lines 20-30nm in thickness which were interconnected at their bases by
circular profiles. They did not find evidence of striation within the ridges.
The ultrastructure of G. muris trophozoites was reported by Friend(5),
who described the sucking disk (which he termed "adhesive disk") to consist
of a single layer of microtubules each of which had a thin fibrous band
extending dorsally from its wall. The microtubules were situated about 200A
from the ventral cell membrane. They were about 250 to 300A in diameter
and were connected to adjacent microtubules by slender filamentous
bridges. When cut in cross-section, the fibrous bands, or ribbons, were
observed extending dorsally to the endoplasm and had a thickness of 200A.
Longitudinal sections indicated a fine cross-banding at the ribbons.
Lamblia (Giardia) muris was also studied by Soloviev and Chentsov(6).
Their work corroborated that of Friend(5) and added that the ribbons were
approximately 2000A in height and were spaced about 30()A apart. The
thickness of 80A was somewhat less than that observed by Friend.
Thus far workers agreed, in general, that the sucking disk was composed
of perpendicular ribbons attached to microtubules aligned along the ventral
plasma membrane. This type of structure does not suggest one of motility
but rather one of support. For this reason, Friend(5) chose to rename the
structure calling it a "striated disk". If the striated disk serves only a
cytoskeletal function, what is the mechanism by which the parasite
maintains its association with the microvillar surface of the intestinal cells?
One possible answer may lie in the function of the marginal lip of the pellicle.
This structure was first observed by Cheissin(3) and later described by
Friend(5) as the ventrolateral flange. It is an extension of the cytoplasmic
dorsal portion of the parasite to the periphery beyond the edges of the
striated disk. The ventrolateral flange consists of a finely striated layer which
separates the striated disk from the endoplasmic portion of the cell on its
medial side and extends outward forming a thin flange covered by the
plasma membrane. The striated portion is composed of alternating light and
dark bands. Nearer the periphery the bands are more dense. In random
cross-sections, the configuration of the ventrolateral flange varies
considerably, both in the same organism and between organisms, indicating
it to be a flexible structure (Fig. 2).
Involvement of the ventrolateral flange in attachment of the organism to
the host mucosa was suggested by Friend(5). The structure was seen in
intimate contact with the microvillar border of the epithelial cells and the
inner portion of the flange, when cut tangential to the fine striations,
displayed a structure and periodicity similar to that of paramyosin obtained
by various extraction methods. Friend did not identify the material in the
flange but suggested that if it were a contractile protein it would give the
necessary motility to the flange.
In his study of attachment of G. muris Holberton(7) questioned the
traditional acceptance of the ventral disk (striated disk) as the organelle of
attachment. However, his findings supported the old view and he concluded
that Friend's(5) hypothesis was incorrect. In a detailed description of the
-------
12
WATERBORNK GIARDIASIS/THE ORGANISM
FIG. 2. Giardia muris. 74,OOOX. Section through lateral portion of trophoxoite showing the
ribbons and microtubules of the striated disk, the ventrolateral flange and the
striated layer extending into the flange.
ventral disk and the association of it with the host mucosa, Holberton
pointed out that the incurved edge of the ventral disk penetrates the host
enteric surface coat and often becomes situated between the epithelial cell
microvilli. The ventrolateral flange was observed to lie against the
microvillar surface coat but rarely in contact with the plasma membrane of
the cell. The ventral disk was reported to be, with little doubt, the organelle
of attachment. Study of the microtubules and their side-arms indicated that
a sliding action could take place but the degree of motion would be limited
by the large number of cross-bridges between the ribbons of the disk.
Further examination of other parameters such as disk diameter, curvature
of the ventral surface and inflection of the disk rim did not support the
hypothesis that mechanical deformation of the disk was responsible for
adhesion. Thus, while disagreeing with Friend on the role of the ventro-
lateral flange in attachment, Holberton also concluded that the ventral disk
was a rigid supportive structure.
Although the electron microscope had provided much new information
on Giardia structure, it was the light microscope that facilitated further
investigation of adhesion of the parasite to the substrate. In the study
mentioned above, Holberton(7) observed, by phase microscopy, living
trophozoites obtained from mice. In motile cells all flagella seemed to
contribute to the motion. However, when the parasite became attached, the
anterior, posterolateral and caudal flagella were quiescent whereas the
ventral flagella beat with a vigorous, synchronous rhythym. Such beating
produced a strong fluid flow through the ventrocaudal groove. This fluid
flow was postulated to produce an area of lowered pressure in the region of
-------
H.G.Sheffield 13
the ventral disk which resulted in the adhesive force necessary to hold the
parasite against the mucosa. This postulate was further supported by the fact
that when ventral flagellar motion ceased, the parasites became detached
from the substrate. The mathematical basis for such an attachment
mechanism was presented in a later study by Holberton(S).
An additional hypothesis for parasite attachment was briefly presented by
Mueller, Jones and Brandborg(l 1). Using scanning electron microscopy in
their study of G. lamblia, views of both the ventral surface of the parasite and
surface views of the intestinal brush border where parasites had been
attached, indicated that the striated disk was not a bilaterally symetrical,
bifobed disk but an overlapping spiral of parallel lamellae. They suggested
that adhesion of the disk was effected by contraction of the microtubules of
the spiral causing the edges of the disk to be pulled inward and down between
the microvilli of the intestinal cells. Support for their interpretation is found
in the work of Brugerolle(12) who, in a detailed study of the striated disk,
reported that it was formed by a helicoidal helix of microtubules b'earingthe
ribbon-like structures.
Another structure, seen mainly in Giardia trophozoites, that has received
attention is the median body. This organelle, readily visible by light
microscopy, was first studied by Cheissin(3). He described it as a group of
straight or slightly curved double membranes which, when cut in cross-
section, displayed circular profiles indicating them to be tubules. He found
no connections between the median bodies and other organelles of the
parasite and concluded that the median bodies were probably lipoprotein
reserves playing a role in parasite energetics. Similar descriptions of the
median bodies of G. lamblia consisting of groups of randomly arranged
microtubules frave been reported by Takano and Yardley(9); Morecki and
Parker(4); and Brooks et al(10).
Friend(5) studied the median body of G. muris in detail. He found a
similar random arrangement of microtubules which did not have origin or
insertion into other structures. In addition, he noted the presence of ventral
extensions from the median body which were composed of 6 to 14
microtubules in regular rows. In comparing the organization, structure,
incidence and attachments of the median body with the axostyle of other
protozoa, Friend concluded that the two were not analogous. Considering
their location in the caudal portion of the organism and the fact that they
disappear upon encystment, he suggested that they served a supporting
function in the posterior portion of the parasite which did not contain the
striated disk.
Near the dorsal surface of all Giardia thus far examined, there are a series
of vacuoles which, in section, range in shape from circular to oval. They tend
to be aligned in a single row separating the granular cytoplasm from the
plasma membrane. Similar vacuoles are seen in the naked area of the ventral
surface and posterior to the lobes of the striated disk(5). Fhend(5) compared
the structure and location of the vacuoles with mucocysts of other protozoa
and suggested that they might be involved in the'encystment process. In
-------
14 WATERBORNE GIARDIASIS/THE ORGANISM
1968, Bockman and Winborn(13) reported the uptake of ferritin by G.
muris. One hour after injection of ferritin into the intestinal lumen of
infected hamsters, all vacuoles contained a high concentration of ferritin.
Ferritin was also found attached to the cell membrane. Because of the
observed close association of the vacuoles with the cell membrane, Bockman
and Winborn(13) suggested that a connection between the vacuole interior
and the intestinal lumen might exist. Recent studies (personal
communication, Robert L. Owen, VA Hospital, San Francisco, California)
indicate that the vacuoles are interconnected and form a system of
intercytoplasmic canals (Fig. 3). The peripheral location of the vacuoles,
both dorsal and ventral, provides them intimate contact with the contents of
the intestinal lumen and the mucosal surface of the epithelial cells. This and
the ferritin uptake study(14) suggest that the vacuoles could be involved in
food uptake. On the other hand, their similarity to secretory organelles of
other protozoa supports their role in encystment. Obviously, more work
should be done to elucidate the function of this important organelle.
Although Giqrdia has abundant ribosomes in its cytoplasm, it is lacking
other common cytoplasmic organelles. Thus far, no one has identified
typical mitochondria, Golgi vesicles nor endoplasmic reticulum, although
Owens (personal communication) has seen what appears to be rough
endoplasmic reticulum in G. muris trophozoites.
Transmission of Giardia occurs via the cyst. Trophozoites which are
sensitive to environmental changes form cysts in the intestine. These cysts
are enclosed by a resistant wall and are able to survive until ingested by the
new host. Multiplication of the organism also occurs during the cyst stage.
Although the cyst is important in the etiology of giardiasis, little work has
been done on its ultrastructure. Soloviev and Chentsov(14) reported the
general architecture of the cyst of Lamblia (Giardia) muris. They observed a
homogenous cyst wall of 0.3 to 0.4/um thickness which was separated from
the enclosed parasite. The cytoplasm contained ribosomes, vacuoles,
flagellar axonemes and rounded formations surrounded by a double
membrane. The rounded formations may represent bacteria as recently
observed by Owens (personal communication). Soloviev and Chentsov(14)
did not see endoplasmic reticulum, mitochondria, nor Golgi apparatus but
did report fragments of the striated disk dispersed in the cytoplasm.
Similar findings were reported from G. lamblia by Sheffield and
Bjorvatn(15). Cysts of G. lamblia isolated from a patient with a heavy
infection, were surrounded by a cyst wall 0.3jum in thickness and composed
of thin fibrous elements interspersed with fine particles (Fig. 4). The same
fine fibrous network was observed in the cyst wall of G. muris (Sheffield,
unpublished observations). The mechanism of protection of the parasite by
the cyst wall could not be ascertained from the electron micrographs but one
might speculate that the fibrous network might impede large molecules while
the membrane at the inner edge of the fibrous layer might function in
controlling the passage of small molecules. In G. lamblia a thin cytoplasmic
layer was associated with the cyst wall and separated from it by the plasma
-------
H. G. Sheffield
15
CW
FIG. 3.
Giardia muris cyst. 25,7SOX.
Vacuoles along periphery of the
organism have the appearance of
canaliculi in the cytoplasm.
FIG. 4. Giardia lamblia cyst. 52,800X.
Cyst wall is composed of fine
fibrous material with interspersed
small particles. Note plasma
membrane of parasite and
peripheral vacuoles containing
membrane-bound granules.
(Sheffield and Bjorvatn. 1977.
Courtesy of American Journal of
Tropical Medicine and Hygiene).
membrane. In opportune sections, the direct relationship of the main
cytoplasmic mass of the parasite and that underlying the cyst wall was clearly
evident (Fig. 5). As reported by Soloviev and Chentsov(14) and in our
unpublished observations of the murine giardia, the parasite's central mass
appeared to be separate from the thin layer adjacent to the cyst wall. Direct
connection has not been seen, although in some areas protrusions from the
thin layer extend inward and appear to contact the main body (Fig. 6).
The possibility, suggested above, that the vacuoles around the periphery
of the trophozoite are involved in cyst formation is not supported by
observations on the structure of the cyst. In both those cysts in which the
parasite fills the interior and those in which it has separated from the cyst
wall, there are numerous vacuoles close to the plasma membrane. These
vacuoles have the same appearance and contain the same flocculent material
seen in the vacuoles of the trophozoite. If they functioned as secretory
vesicles for cyst wall formation, one would not expect to find them in
abundance subsequent to wall formation. Likewise, if they functioned in
nutrition of the parasite, one would not expect to find them in the cyst which
is apparently a non-feeding stage.
The various organelles have a random arrangement in the cyst cytoplasm.
Most noteworthy, and unusual, are the fragments of the striated disk. These
structures present various appearances depending upon the angle of
sectioning (Fig. 7-9). When cut perpendicular to the microtubule, one sees
the circular profile of the microtubule with the ribbon extending from it.
-------
16
WATKRBORNK GIARDIASIS/THE ORGANISM
CW
FIG. 5. Ciardia lamblia cyst. 20,900X. Main portion of cystoplasm is separated from
cyst wall by large lacunae. A thin layer of cystoplasm (arrow) lines the inside of
the cyst wall and is connected to the main body.
FIG. 6. Giardia muris cyst. 25.000X.
Projections from the thin cyto-
plasmic layer extend inward
(arrow) and appear to contact
main cytoplasmic body. Note
median body microtubules and
fragments of the striated disk.
FIG. 7. Giardia lamblia cyst. 22.300X.
Section through the axonemes of
three pairs of flagella. Two groups
of randomly oriented micro-
tubules represent the median
bodies (arrows) and flat rows of
microtubules adjacent to central
pair of axenomes are portions of
the axostyle. Fragments of the
striated disk are also present.
-------
H. G. Sheffield
17
FIG. 8. Giardia lamhlia cyst. S2.800X.
Cross-sectioned fragments of the
striated disk. Note cross bridges
connecting the microtubules and
the fine filaments between the
ribbons. (Sheffield and Bjorvatn,
1977. Courtesy of American
Journal of Tropical Medicine and
Hygiene).
FIG. 9. Giardia lamhlia cyst. 49.300X.
Longitudinally sectioned striated
disk fragments. Mote periodicity
in ribbon structure. (Sheffield and
Bjorvatn, 1977. Courtesy of
American Journal of Tropical
Medicine and Hygiene).
Although the -median body has been considered a structure of the
trophozoite by earlier workers, Sheffield and Bjorvatn(15) found a group of
randomly arranged microtubules in the vicinity of the flagellar axonemes of
G. lamblia and related it to the median body. The compactness of the
microtubules was less than that seen by Friend(5) in G. muris, so the
apparent absence in stained light microscope preparations may be due to an
altered configuration during reorganization in the cyst. Short rows of
microtubules, reported by Friend(5) to be ventral extensions of the median
body, were also seen in G. lamblia cysts by Sheffield and Bjorvatn(15).
However, the rows were not closely associated with the median body
microtubules but commonly were seen coursing parallel and adjacent to two
flagellar axonemes and were described as components of the axostyle.
Comparable rows of microtubules were found in cysts of G. muris (Sheffield,
unpublished observations) (Fig. 10). Occasionally, double rows of
microtubules occurred on each side of the axonemes suggesting a possible
duplication of the structure prior to division of the parasite.
Two previously undescribed structures were observed in G. muris cysts.
The first consisted of a series of dense lines that were either straight or
slightly curved and grouped together in a disorderly array of angles (Fig. 11).
It is possible that they represent developing elements of the striated disk. The
other structure appeared as a somewhat rounded mass occuring often at
several places in the cytoplasm (Figs. 10, 12). Structurally, the mass had a
-------
18
WATERBORNK GIARDIAS1S/THE ORGANISM
FIG. 10. Giardia muris cyst. 35,600X. Striated layer of ventrolateral flange is present as are
portions of striated disk and axonemes offlagella. Note bacteria-like endosymbiont
in cystoplasm.
FIG. 11. Giardia muris cyst. 34JOOX. Randomly arranged dark lines (arrow) may represent
forming elements of the striated disk. Dense amorphous structures are often seen in
association with the dark lines.
-------
H.G.Sheffield 19
FIG. 12. Giardia muris 39,3SOX. Areas of light and dark lines with definite periodicity
resemble the striated layer of the ventrolateral flange and may indicate assembly of
this material in the cyst.
definite periodicity produced by evenly spaced light and dark elements.
When cut across these elements, a crystalline appearance was seen. The
structure corresponds closely to that of the striated area of the ventrolateral
flange demonstrated by Friend(5) in G. muris- Whether the structure is a
remnant from the trophozoite or a newly formed component is not known.
Undoubtedly, the electron microscope has provided us with much
information on the substructure of Giardia. At the same time, new questions
have arisen. While the cytological aspects of the parasite have been well
described, the functional ones need further study. It is hoped that the recent
increased interest in giardiasis, as evidenced by this symposium, coupled
with new techniques to provide experimental infections and large numbers
of parasites will facilitate elucidation of the functional role of the various
organelles presently known.
REFERENCES
1. Dobell, C. 1932. Antony van Leeuwenhoek and his "little animals", pp. 224-225 Harcourt
Brace and Company, New York.
2. Rossi-Espagnet, A. and M. G. P/ccardo 1957 Osserva/iom su alcune strutture
morfologiche e moriofun/ionah di Giardia inte^ltnalit studiata al microscopico
elletronico. Riv. Parasitol 18.211-217
3. Cheissin, E. M. 1964 Ultrastructure of l.ambliaduodenali\ I. Body surface, sucking disk
and median bodies. J Protozooi 11:91-98.
4 Morecki, R., and J. G. Parker. 1967 Ultrastructural studies of the human Giardia tamblia
and subjacent jejunal mucosa in a subject with steatorrhea, Gastroenterology 52'151-164
5 Friend, D. S. 1966 The fine structure of Giardia mum. J. Cell Biol 29-317-332
6 Soloviev, M. M., and Yu. S. Shentsov. 1966 Med Parasitol. 35-667-672.
7 Holberton, D. V. 1973 Fine structure of the ventral disk apparatus and the mechanism of
attachment in the flagellate Giardia muris. J. Cell Sci 13:11-41.
8. Holberton, D. V. 1974. Attachment ot Giardia - A hydrodynamic model based on flagellar
activity J. Exp Biol. 60207-221
-------
20 WATERBORNE GIARDIASIS/THE ORGANISM
9. Takano, J., and J. H. Yardley. 1965. Jejunal lesions in patients with giardiasis and
malabsorption. An electron microscope study Bull Johns Hopkins Hosp. I 16:413-429.
10 Brooks, S. E. H., J. Audretsch, C. G. Miller, and B. Sparke. 1969 Electron microscopy of
Giardia lamblia in human Jejunal biopsies. J Med. Microbiol. 3 196-199, Plates XV-X1X
11 Mueller, J. C., A. L. Jones, and L. I.. Brandborg. 1973. Scanning electron microscope
observation in human giardiasis. Scanning Electron Microscopy 1973 (Part III).
Proceedings of the Workshop on Scanning Electron Microscopy in Pathology. ITT
Research Institute, Chicago, Illinois, April, 1973
12. BrugeroHe, G. 1975 Contribution a 1'etude cytologique et phyletique des diplo/oaires
(Zoomaitigophorea, Diplozoa, Dangeard 1910 V. Nouvelle interpretation de
1'orgamsation cellulaire de Giardia. Protistologica 11'99-109
13. Bockman, D. E., and W. B. Winborn. 1968 Electron microscopic localization of
exogenous ferntin within vacuoles of Giardia mitris. J. Protozool 15.26-30.
14. Soloviev, M. M., and Vu. S. Shentsov. 1970. infrastructure of cysts of Lamblia muni,.
Parasitologia Acad. Nauk 4.510-514.
15. Sheffield, H. G., and B. Bjorvatn. 1977. Infrastructure of the cyst of Giardia lamblia Am.
J. Trop Med. Hyg. 26-23-30.
16. Roberts-Thomson, I. C., D. P. Stevens, A. A. F. Mahmoud, and K. S. Warren. 1976
Giardiasis in the mouse: An animal model. Gastroenterology 71.57-61
17 Bingham, A. K., and E. A. Meyer. 1978. The in vitro excystation of Giardia Fourth
International Congress of Parasitology. Short Communications, Section, B, pp. 95-96.
Warsaw.
Discussion
F. SCH AEFER: With regard to cyst vacuolar areas, were those perhaps
artifacts around the shell of the cyst wall due to flotation techniques that
might have been used in preparing the cysts for electron microscopy?
H. SHEFFIELD: That is a very appropriate question in view of our
earlier discussion of sucrose and flotation problems. I do not believe that
these spaces are really caused by different isotonic pressures of flotation
solutions. We see them in different cysts within the same preparation. In the
same EM section you may see cysts in which the cytoplasm is very closely
applied to the cyst wall, and in other cases where there are large open areas in
between.
All of these open spaces, or lacunae, are all lined with a membrane, and we
think if there had been a shrinking or a pulling of cytoplasm apart in some
way we would not find any membrane around it.
D. JURANEK.: Are there any ultrastructural differences between Giardia
lamblia and G. murisl
H. SHEFFIELD: 1 really cannot report any differences at this point. We
have not looked at a large number of G. muris cysts, and generally speaking,
they look quite similar, but I would hate to say that they are different or
similar.
S. ERLANDSEN: One comment about the activity of the flagella that I
think needs to be kept in mind. In Dr. H olberton's work he was using quartz
halogen lamps for illumination, and I think most of the trophozoites that he
was observing were rather sick and that is why only the ventral flagella were
actually beating. All of the flagella beat with a high frequency, with the
exception of the caudal pair which appear to be inactive. This flagella
activity may be very important in terms of attachment to the surface.
Have you seen any structure close to the nuclei that might be associated
with the production of material within the lumen of the vacuoles or vesicles
located near the surface?
-------
H. G. Sheffield 21
H. SHEFFIELD: We have not observed vesicles which you might relate
to Golgi vesicles in the ventral area of the parasite. The vesicles seen in my
experience are always around the periphery, and 1 have never seen similar
structures internally. Around the naked area on the ventral surface they look
very similar.
-------
22
Surface Morphology of Giardia Cysts Recovered
from a Variety of Hosts
A. S. Tombes, S. S. Landfried, L. D. Williams
Department of Biology, George Mason University, Fairfax, Virginia
ABSTRACT
The surfaces of Giardia cysts, collected from feces of laboratory mice, guinea pigs, dogs,
mule deer, and humans, have been studied with scanning electron and phase microscopy
The cysts appear to be elliptical and slightly asymmetric, measuring approximately 10 by
6/^m The surface was relatively smooth without any characteristic features and cystslrom
all hosts appeared to be indistinguishable.
Giardia lambda has been known by several names for more than 100
years. Organisms possessing the exceptional morphological features of this
species have been identified from the following hosts: human (1), cat
(1,2,12), dog (3,4), horse (5), cattle (1,6), non-human primates (1,7), sheep
and goat (1), wild and laboratory rats and mice (1,8,9), rabbits (1,10,12,13),
guinea pig (1,11), chinchilla (1,12,13), and hamster (1). Giardia from these
hosts have been variously considered to represent the same or a different
species and have been named accordingly (1,14).
Based on clinical impressions, a potential for pathogenicity in human
hosts had been long adduced for this organism. However, until this decade
Giardia did not appear to represent a significant pathogen in the United
States although the medical practice of eastern European countries had
reported individual infections and recognized epidemics in institutionalized
populations for a much longer period.
The reported outbreaks of G. lamblia infections in the United States, with
evidence for importation and transmission of the organism by municipal
water supplies, have provoked questions about the source of these
infections, the presence of a reservoir in nature (15), and the methods of
detecting small numbers of Giardia cysts in large quantities of water (14).
In recognizing the source of Giardia infections, considerable assistance
would be provided if one or more of the distinguishing characteristics of
Giardia cysts from a given host could be identified, especially if those hosts
reasonably could be inferred to contaminate water supplies. The present
work has examined the surface morphology of Giardia cysts, obtained from
several hosts, with the prospect of identifying any distinguishing surface
features.
MATERIALS AND METHODS
Fecal samples from humans were received from laboratories in Virginia
and South Carolina. All samples from animals were obtained through the
-------
A. S. Tombes 23
courtesy of Dr. Charles Hibler, Colorado State University, Fort Collins,
with the exception of the laboratory mice sample which was obtained
through the courtesy of Dr. Robert Russell, Naval Research Laboratories,
Bethesda, Md. Some samples were mailed after being suspended in 10%
formalin solution. Alternately, specimens were obtained as fresh material
and then fixed (in formalin or glutaraldehyde) or mixed to form a slurry in
tap water when processed soon after collection. An aliquot of 0.5 to 1 .Oml of
these supensions was initially passed through three Kimwipes to remove
large particulate matter. Then the solution was filtered through a 20/xm pore,
47mm diameter polyethylene membrane (Tetko, Inc.) in a standard steel
parabola filtration funnel. One drop of this filtered solution was usually
removed and examined by phase contrast microscopy. Finally, a syringe was
used to pass this filtered solution through 13mm diameter, polycarbonate
membranes (Bio-Rad Laboratories) in two Swinney filters mounted in
series, an Sjurh pore membrane preceding a 5jum pore membrane. The filtrate
was discarded. Additional fixation and dehydration solutions were passed
through the membranes as needed using a syringe.
If the cysts and fecal matter were not already fixed, 5ml of 2.5%
glutaraldehyde were slowly passed through the membranes. The material
collected on the membranes from all specimens was then osmicated by
adding 2ml of 1% osmium tetroxide with the syringe and slowly filteringthe
solution through the membrane over a period of 20 minutes. The membranes
were then rinsed with 5ml cacodylate buffer. For dehydration 5ml of 50%
ethanol were added, 2ml being filtered through the membranes to insure
contact with the sample. After a ten minute contact period, the remaining
ethanol was filtered through the membranes. The dehydration was
continued similarly with increasing concentrations of 75, 95, and 100%
ethanol. After the last ethanol solution had passed through the membranes,
the Swinney holders were disassembled and the membranes carefully
removed and air dried for two minutes. The dried membranes were attached
to 13mm scanning electron microscope (SEM) stubs with double-stick tape,
then coated with gold in preparation for viewing in the SEM.
Some specimens were also critical-point dried rather than air dried. After
the 100% ethanol had filtered through the membranes, the Swinney holders
were opened quickly, the membranes immediately placed on 13mm steel
membrane holders in 100% ethanol, and a 3/um pore polycarbonate
membrane placed directly on top of each 5^m or 8jum pore membrane. The
membrane holders were secured, quickly placed in the critical-point drier,
and dried using carbon dioxide as the transition fluid.
Examination of the gold-coated membrane was made in the ETEC
scanning electron microscope located at the Bureau of Biologies, Food and
Drug Administration, Washington, D.C.
RESULTS
Laboratory Mouse
Fecal pellets from a laboratory strain of mice with chronic giardiasis
provided large concentrations of cysts. Fig. 4 and 5 show typical examples of
-------
24
WATKRBORNK GIARDIAS1S/THE ORGANISM
cysts in phase microscopy. No other cyst-like forms were seen showing the
size and well-known internal features of Giardia. Figs. 1, 2, and 3 show the
corresponding structures as observed on a 5jum pore membrane in the SEM.
A relatively clean and high density collection of these forms was obtained
from the fecal specimens. Structure dimensions were 9.0 by 6.0;um giving a
length/width ratio (L/W) of 1.50. All considerations of shape and size,
including recovery from the membrane and re-identification of the internal
morphology by phase microscopy, indicated that the structures seen in SEM
were, in fact, those of Giardia cysts.
FIG. 1-5. Giardia cysts from the mouse. Fig. 1 shows ten Giardia cysts preparedfromfeces of
the laboratory mouse and photographed on a S jimpore membrane at 450x, and
in Fig. 2 at 7SOx. The three cysts at the lower left are photographed again in Fig. 3
at 3000x. Two phase micrographs of similar cysts from the same source are shown
in Fig. 4 and 5 at 2000x. Scale markers are 10 /JL m in all photographs unless
noted to the contrary.
-------
A. S. Tombes
25
Guinea Pig
After processing, the guinea pig feces yielded a noticeably lower
concentration of cysts than those from the laboratory mouse; however, the
cyst could be located and identified by phase microscopy (Fig. 9) and the
corresponding structure was seen in SEM (Fig. 6,7, and 8). The cysts were
the expected elliptical shape and approximately 10x6/um(L/W= 1.66). The
cysts appeared slightly asymmetric. No consistently occurring pores or
projections were present, but the cyst walls showed signs of cracking, the
possible consequence of chemical fixation and/or age of the cysts.
FIG. 6-9. Giardia cysts from the guinea pig. Fig. 6 shows three cysts on an S fJ.m pore
membrane photographed in the SEM at 7SOx, in Fig. 7 at 3000x and in Fig. 8 at
7500x. A similar cyst from the same source is photographed at 2000x in phase
microscopy in Fig. 9.
-------
26
WATERBORNE GIARDIASIS/THE ORGANISM
Dog
Cysts found in dog feces were noticeably variable in dimensions. Fig. 13
and 14 show phase micrographs of cysts with the dimension 13 by 11/im
(L/W = 1.18) and 12 by 10/im(L/W = 1.20). These two micrographs show
that a larger, less elliptical cyst is present together with a smaller, more
elliptical cyst. Fig. 10 and 11 show the more elliptical structure as seen in
SEM with a higher magnification given in Fig. 12. There are no
distinguishing surface features although the more spherical cyst appears to
have a smoother surface than the more elliptical type. The elliptical type is
slightly rougher and also more like the cysts from the guinea pig and mouse.
FIG. 10-14. Giardia cysts from the dog. Fig. 10 shows three cysts on aS jJ. m pore membrane
at 750*, in Fig. 11 at ISOOx, and Fig. 12 at lO.SOOx. Two similar cysts'/ram the
same source are shown in phase microscopy at magnifications oj 4000x and
2500X in Fig. 13 and 14, respectively.
-------
A.S. Tombes
27
Mule Deer
Giardia were extracted from feces of this host in relatively high
concentration and were predominately found on the 5/xm pore membrane
(Fig. 15,16,17). The cysts were identified in phase microscopy, although no
photographs are available. The 9 by 6/^m size and elliptical appearance are
similar to those from the mouse, guinea pig, and dog. The walls of the cysts
show some of the rough and flaking surface observed on previous cysts, and
again no unique morphological features were noted.
On the membrane at the base of approximately a fifth of the cysts,
material was noted which appeared to have melted, flowed evenly from the
FIG. 15-17. Giardia cysts from the mule deer. Fig. 15 shows over fifty cysts prepared from
feces of the mule deer andphotographed on a 5 /j.m pore membrane in the SEM
at 750*, in Fig. 16 at ISOOx, and in Fig. 17 at 4SOOx.
-------
28
WATERBORNE GIARDIASIS/THE ORGANISM
cysts, and solidified forming a disk around the point of attachment between
the organisms and the support membrane. Such rings are noticeable on
several of the cysts in Fig. 17. Even after this unexplained denuding no new
features are evident on the surface of those cysts. No attempt was made to
repeat this observation; thus its significance is unknown.
On Fig. 15 and 16 several spherical structures are apparent. These are
possibly cysts of other protozoa, which possess'smooth surfaces and are
between 5 and 6jum in diameter. If a slightly different procedure is followed
in collecting the filtrate prior to passage through the Swinney filters, a higher
FIG. 18-25. Unknown cysts from the mule deer. Figs. 18 and 19 show phase micrographs of
two unknown cysts from the feces of the mule deer which have some morpho-
logical similarities to Giardia. They are also shown in Figs. 20 and 21 on an
8/Am pore membrane and are photographed at 750x and at ISOOx respectively.
In the last 4 micrographs they are ona5/j.m pore membrane and photographed
at 750x in Figs. 22 and 24, and at ISOOx in Fig$. 23 and 25.
-------
A. S. Tombes 29
concentration of these structures is noticed with fewer Giardia. Two of these
cysts are shown on Fig., 18 and 19 which clearly demonstrate that they are
not Giardia. Fig. 20 illustrates the collection of cysts and debris on the 8jum
pore membrane, a portion of which is enlarged in Fig. 21 to show four of the
unknown cysts and a single Giardia cyst. The accumulation of cysts and solid
fragments on the 5yum pore membrane are shown on Fig. 22 and 24 with
areas magnified in Fig. 23 and 25.
Human
Four phase micrographs (Fig. 28,29,30, and 31) of Giardia lamblia cysts
are shown which illustrate a morphology similar to those already noted from
previous hosts. Fig. 26 and 27 are scanning electron micrographs of cysts
which are typical of those collected on both the 8 and 5;um pore membranes,
but the anticipated elliptical shape is not found. Rather, the cysts appear to
have become distorted, creating a shape which is more cuboidal than
elliptical. This is the only example where the results obtained by scanning
and phase microscopy are not complementary.
Non-Giardian Cysts
Three test protozoans were examined as internal controls to provide
confidence that the SEM techniques being employed would indeed
demonstrate pores, ridges, and surface indentations if such were present on
the cysts of Giardia. The three organisms chosen were Naegleria,
Acanthamoeba, and Sarcocystis which have particular relevance to this
study because their cysts may be isolated from water.
Naegleria fowleri is shown in Fig. 32,33 and 34 with micropores, 0.1 ^m in
diameter, and clearly evident surface indentations. Naegleria gruberi is
photographed in Fig. 35,36, and 37 with a large depression on a mature cyst
and micropores clearly evident on the walls of the empty cysts.
Acanthamoeba culbertsoni is photographed in Fig. 38 to 42 and shows
quite clearly the ratchet-like internal structure which is conspicuous
following the collapse of the cyst wall. When the cysts are critical-point-dried
and the walls do not collapse such internal structures are not evident.
Finally, Sarcocystis sp. is shown in an air dried preparation in Fig. 43 and
44, in a critical-point-dried preparation in Fig. 45, and in phase microscopy
in Fig. 46 and 47. The slightly raised ribbing along the surface of the cyst
wall, forming an oval hexahedron, is easily and consistently observed with
the SEM.
DISCUSSION
The examination of Giardia cysts from five hosts, (mouse, guinea pig, dog,
mule deer, and humans), demonstrated a degree of similarity in reference to
size, and appearance. The cysts appear as elliptical, slightly asymmetric
structures measuring approximately 10/um by 6^m. The one exception, from
humans, deviated from the elliptical shape of cysts when examined in the
SEM but not when examined in phase microscopy. This size alteration was
very probably a consequence of some procedural error which was
-------
30
WATKRBORNE GIARDIASIS/THE ORGANISM
FIG. 26-31. Giardia cysts from humans. Fig. 26 is an SEM micrograph of several Ciardia
cysts from human material on an 8 fimpore membrane at 750*. Fig. 27 is the
same material magnified to 3000x. Figs. 28, 29, 30, and 31 are phase
micrographs of cysts of the same sources at 2000x.
-------
A. S. Tombes
FIG. 32-37. Cysts of two Ngegleria species. CystsofN. fowleri (ATCC 30463) are shown in
Fig. 32 on an 8 fj. mpore membrane at 7SOx and in Figs. 33 and34at3000x The
initial photograph shows a typical field with two relatively intact and numerous
collapsed cysts. The large cyst In the lower left corner is shown enlarged in
Fig. 34 which along with Fig. 33 show micropores and surface irregularities
Cysts of N. gruberi (ATCC 30133) are shown in Fig. 35 on an 8 u.m pore
membrane at 750*, in Fig. 36 at 3000x, and in Fig. 37 at ISOOx. The cyst in the
upper right corner of Fig. 35 is shown in Fig. 36, and a group of collapsed cvsts
are shown in Fig. 37.
-------
32
WATKRBORNE GIARDIASIS/THE ORGANISM
FIG. 38-42. Cysts of Acanthamoeba culbertsoni (A TCC30172). Numerous cysts are seen in
SEM on an 8 (J- m pore membrane at 7SOx in Fig. 38. A similar group of cysts at
ISOOx is shown in Fig. 39, and a mature cyst (left) together with a rounded, less
mature cyst is shown at S900x in Fig. 46. Figs. 41 and42 show the appearance of
the two groups of cysts in situ on the surface of an agarplate as they appear in
phase microscopy at 300x and 600x, respectively.
-------
A.S. Tombes
33
FIG. 43-47. Cysts of Sarcocystis §p. Fig. 43 illustrates a large number of cysts at 750x. In
Fig. 44 most of the cysts have collapsed during air drying in the preparation
procedure. Figs. 44 and 45 are of the same magnification (3000x) but the cysts in
Fig. 45 were prepared by critical-point-drying which demonstrates improved
retention of the cyst shape. Figs. 46 and 47 are phase micrographs at 2000x. The
pattern of surface ribbing which stands out sharply in the scanning electron
micrographs is faintly delineated in the phase micrographs, especially in Fig. 46.
-------
34 WATERBORNK CIARDIASIS/THK ORGANISM
unknowingly repeated in each human sample and is not felt to represent a
biologically significant morphological characteristic. Research is underway
to clarify this matter.
The surface of all Giardia cysts is relatively smooth and even, without a
consistent pattern of pits, pores, or major depressions and without ridges,
projections, or a uniform pattern of granularity or roughness. This surface
smoothness must then be the major surface characteristic of the organisms.
Therefore, based on our observations of the surface of cysts, we cannot
support the hypothesis that Giardia from each host represents a separate
species. Consequently, it appears impossible at this time to determine the
source of a sample of cysts through an examination of the cysts with the
SEM. The possibility still exists, however, that with some modification of
the procedure for the detection of nematodes in drinking water (16) it will be
possible to detect Giardia cysts with the SEM in samples of water.
REFERENCES
1. Levine, N.D. 1961. Protozoan parasites of domestic animals and man. Burgess Publishing
Co., Minneapolis
2. Brightman, A. H., and G. F. Slonka. 1976. A review of five clinical cases of giardiasis in
cats. J. Am. Anim. Hosp. Assoc. 12:492-497.
3. Christie, D. W., R. S. Anderson, E. T. Bell, and G. L. Gallagher.
Ulceration of the ileum and giardiasis in a beagle. Vet. Rec. 88:214-215.
4. Bemrick, W. J. 1963. Observations on dogs infected with Giardia. J. Parasitol. 49:1031-
1032.
5. Bemrick, W. J. 1963 Giardia in north american horses. Vet. Med. 63:163-165.
6. Jensen, E. A., and D. M. Hammond. 1964. A morphological study of trichomonads and
related flagellates from the bovine digestive tract J. Protozool. 11:386-394
7. Myers, B. J. and R. E. Kuntz. 1968. Intestinal protozoa of the baboon, Papia doguera
Pucheran, 1856. J. Protozool. 15:363-365.
8. Rothenbacher, H., J. F. Kavanaugh, and F. A. Stormer. 1970. Giardiosis in a wild mouse
(Peromvscus leucopuii) colony. J. Amer. Vet. Med. Assoc. 157:685-688.
9. Boorman, G. A., P. H. C. Lina, C. Zurcher, and H. T. M. Nieuwerkerk. 1973 Hexamna
and Giardia as a cause of mortality in congenitally thymus-less (nude) mice. Clin. Exp.
Immunol. 15.623-627.
10 Danciger, M. and E. A. Meyer. 1971. The in vitro growth response of G/a^/fa trophozoites
from the rabbit J. Protozool 18:311-313.
11. Fortess, E. and E. A. Meyer. 1976. Isolation and axenic cultivation of Giardia trophozoites
from the guinea pig J. Parasitol. 62:689.
12 Meyer, E. A. 1970. Isolation and axenic cultivation of Giardia trophozoites from the
rabbit, chinchilla, and cat Exp. Parasit. 27'179-183.
13. Meyer, E. A. and B. L. Pope. 1965. Culture in vitro of Giardia trophozoites from the rabbit
and chinchilla. Nature (Lond). 207:1417-1418.
14. Jakubowski, W., T. H. Ericksen, and S. L. Chang. 1977. Detection, identification and
enumeration of Giardia cysts in water supplies. Proceedings, AWWA Water Quality
Technology Conference. Kansas City, Missouri, Dec. 4-7, 1977
15. Wright, K. W., R. W. Bacorn, V. Smith, and D. O. Lyman. 1975 Giardiasis in residents of
Rome, New York and in U. S. travelers to the Soviet Union. Morbidity and Mortality
Weekly Report. 24 366-367.
16 Tombes, A. S., A. R. Abernathy, D. M. Welch, and S. A. Lewis. 1978. Extraction of
nematodes from drinking water and the comparative detection efficiency of optical and
scanning electron microscopy. Scanning Electron Microscopy/1978. 11:297-302.
Discussion
R. OWEN: Did you do transmission electron microscopic correlation on
any of your filter material?
A. TOMBES: No, we have not.
-------
A. S. Tombes 35
R. OWEN: Dr. Erlandsen has described various ways of doing this. It
really is critical because unfortunately with the scanning microscope, when
you are looking at the surface, little round things look like little round things.
The stool has a lot of little round things in it, and its really hard to tell one
from another. It is essential before making any judgment about seeing
anything in scanning microscopy that you take the specimen and look inside.
You can take it off the stub even with gold on it, embed it in resin, and section
it as though it were cultured material. It is just like not telling books by their
covers.
The other problem that we have had is looking at material that has been
dehydrated, particularly on a filter. The evaporation rate of alcohol is so fast
that when you take your filter material and try to get it in your critical-point
dryer it tends to air dry before it gets there. Even though you go through the
critical-point drying procedure, essentially you have an air dried specimen.
Did you ever take the material out of the filter holder under alcohol and
keep it under alcohol continuously before putting it into the critical-point
dryer?
A. TOMBES: No, we did not.
R. OWEN: It is very hard to avoid air drying your material, and there may
be some surface characteristics in Giardia that may be lost. This is unlike
nematodes where you have enough size that they retain enough moisture to
keep from drying the wrong way.
A. TOMBES: I realize that situation. Commenting on the second question
first, if you are quick you can remove the membrane, keeping it under
alcohol while you move it to the dryer. With the Bomar critical-point dryer
you can move the membrane within 3 or 4 seconds so that it does not air dry.
Holding the membrane properly, you slip it onto a steel support that will
hold it in the alcohol while you secure the retainer during the drying process.
Consequently, I believe this eliminates the problem of unintentionally air
drying the cysts.
We saw with Sarcocystis that different cyst morphologies are observed
following air or critical-point drying. This is not the case with Giardia where
no consistent differences were observed between the 2 drying procedures.
R. OWEN: From your continuous correlation with phase microscopy, I
presume that what you are looking at are Giardia cysts, but I would feel
much better if at least once you had a transmission correlation so you could
be sure.
A. TOMBES: We have been devoting most of our time to scanning and
phase microscopy. If we go to TEM and section some of the material, it may
strengthen our belief that we have Giardia, but it would not help us in
deciding whether we have muris or another species and that was the objective
of our study.
S. ERLANDSEN: In most of the pictures there seemed to be a
tremendous amount of secondary emission coming off the cyst, and
essentially it was whited out on the screen. Have you used lower voltage
scanning microscopy to see surface detail?
A. TOMBES: Yes. In the dog cyst for example, the smooth one, you
would suspect that any morphology there would be lost because of the
secondary emission. We lowered the voltage on a number of occasions, but
then you lose detail. We have also altered the photographic settings on the
microscope to try to improve the detail.
-------
36 WATF.RBORNK CIARDIASIS/THK ORGANISM
J. HOFF: Many of us seem to have the need for obtaining good clean cyst
preparations. My question is one of general inquiry as to a possible way of
doing this for our own work in disinfection. We need highly purified
suspensions of cysts that retain viability and are free of both soluble and
particulate material that may exert a chlorine demand.
There is a recently developed centrifugal procedure called elutriation
which involves low speed centrifugation. The. fluid flows in a direction
counter-current to the centrifugal force, and some quite remarkable
separations of cell populations with different surface characteristics have
been achieved. I wonder if anyone has used this procedure?
A. TOMBES: 1 have not used the procedure. Has anyone here used the
procedure?
(No audible response.)
T. ERICK.SEN: Do you think that the presence of the 10 per cent formalin
in your fecal specimens might have had an effect on your cysts? There has
been experience with these under storage, particularly at room temperature,
where the cysts will disappear. Secondly, it is my understanding that you said
the cysts did not have pores. Is that correct?
A. TOMBES: That is correct, and is contrary to what we had talked about
earlier. The explanation is that we must have had an artifact in the
preparation of an earlier human specimen to give us the morphology that we
were seeing at the time.
1 am fairly certain that some of the fine cracking or flaking that is evident
on the Giardia cyst in the SEM is a consequence of the fixation procedure.
We are in the midst now of a study to determine the effects of shelf life on the
cysts in a variety of conditions to see if the 10% formalin does create
distortion with time, and what effects 2% glutaraldehyde might produce.
V. OLIV1ER1: Do you have any information on the per cent recovery for
the various steps in your procedure? Also, how many samples of human and
animal fecal material have you examined?
A. TOMBES: We have collected fecal material from eight mice, one
guinea pig, one dog, one mule deer, and three humans. When working with
sucrose gradients, we found a 15 or 20% recovery of cysts on the membrane.
We have not conducted a similar study using the procedure presented today.
I am sure we are losing many cysts in the Kimwipes and on the 20 /xm pore
screens; however, the objective of the study was to look at the cyst surface
and not try to determine the per cent recovery.
R. RENDTORFF: The micromanipulator made by Dr. Charles Reese of
NIH is a very simple technique to get cysts in very minute quantities of water.
In terms of number, in our studies, we went up to concentrations of over a
million. Those were not free of particulate matter because we did not use the
micromanipulator. We did, however, obtain accurate numbers of cysts up to
100 in small amounts of water very easily. 1 think sometimes modern
techniques are perhaps too refined.
A. TOMBES: There is no problem in getting a hundred cysts under a
phase microscope, and there is no problem in getting similar cysts at about
the same density on the membranes in the SEM. The problem concerns
picking up one cyst which is unquestionably Giardia lamblia or muris, and
then looking at that same cyst in the SEM.
R. RENDTORFF: 1 really think that could be done.
-------
A.S.Tombes 37
A. TOMBES: 1 do not doubt that it can be done, and if I could get a
micromanipulator to use, I would do it immediately.
R. W1DDUS: I can put you in touch with people that can actually do the
micromanipulation and have been working on it.
A. TOMBES: This is a very important point and goes back to what Dr.
Owen was saying at the beginning. That is the test, to know what we are
looking at intheSEM. I am the first to admit that the data are circumstantial
for we have been correlating phase microscopy of the cysts with what is seen
in the SEM.
-------
38
SESSION II - THE DISEASE
Chairman - Walter Jakubowski,
U.S. EPA, HERL, Cincinnati, Ohio
Managing the Patient with Giardiasis:
Clinical, Diagnostic and Therapeutic
Aspects
M. S. Wolfe
The Possible Use of an Indirect Immunofluorescent
Test Using Axenically Grown Giardia lamblia
Antigens in Diagnosing Giardiasis
G. S. Visvesvara and G. R. Healy
The Experimental Transmission of
Giardia lamblia Among Volunteer
Subjects
R. C. Rendtorff
Giardiasis in the Mouse: Clues to Host
Immune Mechanisms
D. P. Stevens
-------
39
Managing the Patient with Giardiasis:
Clinical, Diagnostic and Therapeutic Aspects
Martin S. Wolfe
Office of Medical Services, Department of State, Washington, D. C.
ABSTRACT
Giardia lamblia is the most common pathogenic intestinal parasite in the United States
and is a leading cause of morbidity in travelers returning from abroad. Clinical features
may differ in the recently acquired acute stage and in the established subacute and chronic
stages. Diagnosis can usually be made with three well performed stool examinations,
properly collected on alternate days, utilizing both direct smear and formol-ether
techniques. Examination of duodenal contents and intestinal biopsy only rarely confirm
the diagnosis when these stool examinations are negative. Not infrequently, giardiasis is
suggested by typical symptoms and exposure history, and parasitologic confirmation
cannot be made after reasonable diagnostic effort In this situation, an empiric trial of the
drug quinacrme often leads to gratifying clinical improvement. Qumacrine is the treatment
of choice, with metronida/ole and furazolidone being alternative available drugs in the
U nited States All three drugs, however, have potential problems associated with their use
Giardia lamblia is the most common pathogenic intestinal parasite in the
United States, according to the most recent Intestinal Parasite Surveillance
Report of the Center for Disease Control(l), and is also the most frequently
encountered intestinal pathogen in my practice in Washington, D.C. Over a
five year period from late 1969 to late 1974, 670 cases of giardiasis were
found at the Office of Medical Services of the Department of State, while
during the same period 508 cases of amebiasis were diagnosed. These 670
giardiasis cases, occurring primarily in returnees from 97 different countries,
and in a much smaller number of individuals who had not left the United
States, have been reviewed and are the primary source of data presented
here.
CLINICAL FEATURES
In experimental infections, the prepatent period, based on the first
detection of parasites in the stool, averaged 9.1 days(2). The most reliable
data on time interval between infection and onset of acute symptoms comes
from investigations of travelers returned from the Soviet Union. The mean
time period until the onset of illness from entrance into the Soviet Union has
been 12 to 15 days (ranging from 1 to 75 days)(3,4,5).
Symptomatology in giardiasis differs, depending on the duration of
infection when diagnosis is first made.
Acute Stage
Typically, there is the sudden onset of explosive, watery, foul diarrhea;
marked abdominal distention, foul gas, and belching; and nausea, anorexia,
vomiting, fatigue, and cramps which are usually upper or midepigastric.
-------
40 WATKRBORNE CilARDIASIS/THE DISEASE
Lowgrade fever and chills may precede or occur with the onset of diarrhea,
as may headache. Mucus occurs only rarely in the stool and when blood is
present in the stool it is invariably secondary to anal irritation from the
diarrhea. Again, some of the best available data on the symptomatology of
acute infections comes from investigations of individuals recently returned
from the Soviet Union. Table 1 shows the frequency of symptoms in a series
of 32 of these cases with Giardia lamblia found in the stool(3).
The acute stage usually lasts only three to four days and is often not
recognized at the time as being due to giardiasis. In some cases, the acute
stage may last for months, leading to malabsorption, debility, and
significant weight loss. This latter situation appears to be more common in
young children than adults and perhaps explains why giardiasis had
formerly been considered to be primarily a disease of childhood. The acute
symptoms of giardiasis may mimic those of acute amebic dysentery,
bacillary dysentery, bacterial food poisoning, and "travelers' diarrhea"
caused by enterotoxigenic E. coli. But the foul-smelling stool, flatus, and
belching, the marked abdominal distention, and the rarity of mucus and
blood in the stool, are more characteristic of giardiasis.
Subacute And Chronic Stage
Acute infections can develop into long-standing subacute or chronic
infections. In returnees from residence or lengthy travel abroad, the acute
Table I. Clinical features of 32 patients recently returned from the Soviet Union (1)
with proven giardiasis compared with 105 State Department cafes
with undetermined duration of proven injection
Symptom
Flatulence
Foul Stool
Cramps
Distention
Anorexia
Nausea
Weight Loss
Belching
Heartburn
Headache
Constipation
Vomiting
Fever
Chills
Diarrhea
Blood in Stool
Mucus in Stool
Fatigue
State Dept %
(n=105)
46 7
44 8
32 4
31 4
20 0
200
18 1
15 2
14 3
1 1 4
11 4
4 8
3 8
2 9
62 9*
0
3 8
28 6
USSR %
(n = 32)
56 5
52 2
59 4
56 2
59 4
304
—
34 4
17 4
—
71 9
0
4 4
87 5
* 52.4% had soft or mushy stools; 10.5% had watery stools
-------
M.S. Wolfe 41
stage is often not recognized or recalled, and they frequently present with
persistent or recurrent mild to moderate symptoms. In a series of 105 cases
seen in 1974 at the Department of State with uncertain duration of infection,
almost one-third to one-half presented with flatulence, mushy foul stools,
primarily upper intestinal cramps, and abdominal distention. Other
symptoms included anorexia, nausea, weight loss, belching, heartburn,
headache, and constipation. Vomiting, fever, and chills occurred rarely.
These symptoms are shown in Table 1 and are compared to the symptoms of
the 32 acute cases. Type of stool passed, frequency of bowel movements, and
virtual absence of blood, pus, and mucus in the stool for the 105 State
Department cases from 1974, is shown in Table 2. Of these 105 cases, 39%
were judged by our criteria to have mild symptoms (cramps, constipation,
flatulence, foul stool, mushy stool); 41% were judged to have moderate
degree of symptoms (above mild symptoms and/or belching, nausea,
anorexia, vomiting, distention, heartburn, weight loss, fatigue, chills, fever
and headache); and 6.7% were considered to have incapacitating severe
symptoms (above symptoms plus frequent watery diarrhea, marked weight
loss, and marked fatigue). No symptoms were present in 13.3% of this group.
Thus 80% of these cases were considered to have mild to moderate subacute
or chronic symptoms.
Table 2 Characteristic1, o/ the stools of 105
State Department caf.es with proven
No %
A
B
C
Type of Stool
Formed
Mushy
Watery
Unknown
Blood
Pus
Mucus
Frequency
Normal
Increased
Very Frequent
35
55
1 1
4
0
0
4
65
22
13
33 3
52 4
10 5
3 8
0
0
3 8
61 9
21 0
12 4
The typical non-acute stage patient presents with recurrent or less
commonly persistent brief episodes of loose foul stools which may be
yellowish, frothy, and float in the toilet water, accompanied by increased
distention and foul flatus. Between exacerbations, stools are usually mushy
or constipation may occur. Abdominal discomfort is often caused by the
marked distention and when cramps occur they are usually in the mid to
upper epigastrium. Sulfuric belching, known by Peace Corps Volunteers as
the "purple burps," and substernal burning are not uncommon. Anorexia,
-------
42 WATERBORNE GIARDIASIS/THE DISEASE
nausea, and a gurgling or uneasiness tn the epigastrium are frequent
complaints, but vomiting is unusual in this stage. There may be tenderness to
palpation in the right, and to a lesser extent, in the mid and left upper
quadrants. These symptoms often suggest ulcer, hiatal hernia, or gall
bladder disease. We have seen a few cases of urticaria associated with
giardiasis, and others have reported this sign(6), erythema multiforme(7),
and arthritic symptoms(S). Weight loss, lassitude or fatigue, headaches, and
myalgias may occur. In rare cases, these symptoms may persist for years, but
in the great majority of cases spontaneous disappearance of parasites and
symptoms occurs after variable periods of time. If eosinophilia occurs, it is
usually related to another concomitant cause.
Rare reports of cholecystitis(7,9) and pancreatitis(lO) associated with the
presence of G. lamblla parasites have been reported. Various types of
malabsorption have been confirmed (11.12), including steattorhea,
disaccharidase (particularly lactase and xylase) deficiency, Vitamin Bn
malabsorption, and protein-losing enteropathy.
A post-giardial lactose intolerance may develop, particularly in patients
from ethnic groups with a predisposition to lactase deficiency, following
apparent eradication of parasites with specific treatment. This must be
considered in individuals with negative post-treatment specimens who have
persistent mushy stools and excessive gas and distention, before
administering further anli-Giardia treatment.
The duration of the asymptomatic cyst-passing state has not been
determined.
DIAGNOSIS
Stool Examinations
In the majority of Giardia infections, a well-trained parasitology
technician should usually make a diagnosis by stool examination. In our
series of 670 parasitologically confirmed cases, utilizing direct smear and
formol-ether concentration tests, 76% were positive on the first specimen,
90% were found with two specimens, and 97% were found with three
specimens. An additional 1.7% were found on the fourth or subsequent
specimen. The zinc sulfate flotation method in our experience has not
significantly increased the positive yield over direct smear and formol-ether
concentration methods. Examination of stools on alternate days has given
us an increased positive yield over specimens examined on three consecutive
days and this is probably related to the intermittent passage of parasites due
to periods of their active multiplication. Some investigators claim that only
50% of proven cases can be confirmed by stool examination, but they usually
perform either a single stool examination by formol-ether technique or they
perform only direct saline smears without concentration(13).
Diagnosis appears to be easier in early acute than in established infections.
In the acute stage, stools are frequently watery or loose and may contain only
the more labile trophozoites, due to rapid bowel transit. It is thus important
to either immediately examine a wet smear or to preserve stool in formalin,
merthiolate-formalin, or polyvinyl alcohol (PVA) for later examination.
-------
M. S. Wolfe 43
After the acute stage has passed stools are more often semi-formed or
formed and contain the more hardy cyst form of the parasite. There is then
no urgency to immediately examine "warm stools" and direct smear and
formol-ether concentration examinations of unpreserved stools are often
adequate to make the diagnosis. Purging does not appear to aid the
diagnosis and routine culture methods are not presently available.
Not infrequently, we encounter returnees from abroad with a clinical
picture highly compatible with giardiasis in whom we are unable to prove the
presence of parasites by repeated stool examinations. It has been shown in
some returnees to Finland following visits to the Soviet Union that in the
first three weeks after infection, symptoms may be present before parasites
become detectable in the stool and repeated examinations may be necessary
to confirm the diagnosis(14). In given populations of Giardia-infecled
individuals in an endemic area with established infections, high, low, and
mixed patterns of cyst excretion have been observed(15), and this situation
may well occur in returnees from endemic areas. It has been suggested that
positive confirmation of infection in low and mixed excretors might require
examinations of two or three stools a week for four or five weeks, but this is
tedious, expensive, and often impractical.
Examination of Duodenal Contents and Biopsies
Where parasites cannot be found by stool examination in suspected cases,
duodenal fluid examination may allow for their recovery. Our experience
with duodenal tubal aspiration is limited, but this method has been reported
to be more reliable than stool examinations in making the diagnosis( 16). The
"Enterotest", a gelatin capsule containing a string, has been shown by some
workers to be a suitable substitute for duodenal intubation in obtaining
Giardia trophozoites(17). Other workers suggest that examination of
biopsies with Giemsa-stained sections and touch preparations from mucus
adherant to the biopsy section is perhaps the most sensitive technique to
diagnose cryptic infections not found by stool examinations or examination
of duodenal contents(16). In the face of negative stool examinations, we
have rarely been able to confirm an infection with these methods, perhaps
due to the patchy nature of loci of infection or the extreme paucity of
parasites in some symptomatic individuals. Two recent reports from India
have shown that stools can be positive when aspirates and biopsies are
negative(13,18). As duodenal tubal aspiration and intestinal biopsy are
bothersome and time-consuming procedures, when giardiasis is strongly
suggested by typical epidemiologic exposure and symptomatology, and
stool examinations are negative, I prefer to give an empiric course of
treatment with the most effective anl\-Giardia drug, quinacrine (Atabrine).
This frequently leads to gratifying improvement in these difficult to diagnose
cases. As quinacrine is not known to have other non-specific intestinal
effects, such as on enterobacteria, improvement strongly suggests that the
drug has acted on a cryptic Giardia infection.
-------
44 WATERBORNE GIARDIASIS/THE DISEASE
Barium Examination of the Small Bowel
Barium examination of the small bowel may show a characteristic pattern
of edema and segmentation(19), but these findings by themselves do not
confirm the diagnosis and it is vitually impossible to recognize intestinal
parasites for one week or longer after barium use. Many antibiotics,
antacids, kaolin products, paregoric, most enema preparations, and oily
laxatives can also cause a temporary disappearance from or masking of
parasites in the stool. Stool examination of duodenal contents should be
deferred for approximately five to ten days after use of these products.
Serological Testing
Culture techniques are described in other presentations in this
Symposium and the described successful efforts in culturing Giardia
trophozoites may soon lead to the availability of serologic tests for
giardiasis.
TREATMENT
In the United States, three drugs are available for treatment of giardiasis:
quinacrine (Atabrine), metronidazole (Flagyl), and furazolidone
(Furoxone). A number of other drugs, including tinidazole and
nimorazole(20,21), found useful abroad, are not licensed or available in this
country and will not be discussed here.
Quinacrine
Most workers recommend quinacrine as the treatment of choice for
giardiasis. We have obtained a cure rate of over 95% with quinacrine in over
100 patients given this drug and most workers report cure rates of at least
90% with it. One remarkable exception to these excellent results is a recent
report from London, where Giardia was eradicated in only 63% of patients
given this drug( 12). The dose recommended for those over eight years old is
100 mg three times a day for seven days. For younger children, the dose is 2
mg/ kg 3 times daily for seven days. In my experience, though quite effective
in younger children, tolerance to quinacrine is poorer in this group. Some
intestinal upset, mild headaches or dizziness, and yellow urine commonly
occur in all age groups receiving quinacrine. Rarely, vomiting, fever, and
severe skin rash occur. We have-noted a 1.5% incidence of toxic psychosis in
adult patients receiving this drug and patients must be forwarned of this
possibility. Yellowing of the skin and sclerae rarely occur in doses used for
giardiasis. Quinacrine is contraindicated in those with psoriasis and should
not be taken together with alcohol or with primaquine.
Metronidazole
Metronidazole (Flagyl) is frequently prescribed in this country for
giardiasis in an adult dose of 250 mg three times a day for seven days. The
recommended dosage for children is 5 mg/ kg 3 times daily for 7 to 10 days.
In a series of 21 adult cases with established infections, we cured 18, or85%.
In another group of 42 recently returned travelers from the Soviet Union
with giardiasis treated with metronidazole 250 mg 3 times daily for 7 days, 27
had one or more followup stool examinations four weeks after completion of
-------
M 5. Wotfe 45
treatment and 11 of these were still positive. A similar failure rate of 30% was
reported from Sweden in a group of recent returnees from the Soviet
Union(21). It would thus appear that metronidazole in the usual
recommended dose is less effective in acute than in established infections.
This is perhaps related in some way to the excellent absorption of
metronidazole. In England, metronidazole in a single 2.0 g dose on three
successive days produced a parasitological cure rate of 91%(12), but for
reasons to be described, and as the effectiveness and tolerance of these higher
doses of the drug are no better than with quinacrine, this regimen cannot be
recommended in this country. Although metronidazole in the accepted dose
in this country is somewhat less effective than quinacrine, tolerance is
generally better. Side effects may include nausea, headache, metallic taste,
dizziness, and dark discoloration of the urine. Metronidazole has a well
recognized disulfiram-like reaction when taken with alcohol. Metronidazole
has been found to be carcinogenic in mice, but not in hamsters, and is
mutagenic in bacteria(22), and some controversy has attended its use in non-
life-threatening situations. Unfortunately, treatment of giardiasis remains
an unapproved indication for metronidazole in this country, and its use in
this condition is somewhat debatable. Under these circumstances, it would
not appear wise to use this drug in the high daily doses being used in England
for giardiasis. Perhaps second thought should be given before it is used in
young children and in severely symptomatic pregnant women in whom
giardiasis should be treated, at least until it becomes approved for giardiasis.
Furazolidone
A third drug, which is not widely used here in giardiasis, is the antibiotic
furazolidone. It has been found to be effective and rather well-tolerated in a
number of studies abroad(20,23). Furazolidone is the only available anti-
Giardia drug in this country in suspension form as well as in tablets, making
it useful in children. In a group of 31 children below age 10 treated with
furazolidone, 24 (77%) were considered parasitologically cured by us.
Vomiting, diarrhea, nausea, or fever occurred in 10 of 24 of these cases
reviewed for side effects, and in three it was necessary to discontinue the
drug. There have been other reports of the rare occurrence of
hypersensitivity reactions, including hemolysis, hypotension, and urticaria.
We recently reported serum sickness in two adults who received tablets of
this drug manufactured in Latin America(24). Furazolidone is a monoamine
oxidase inhibitor, it can precipatate a disulfiram-like reaction, and it may
cause mild reversible intravascular hemolysis in glucose-6-phosphate
dehydrogenase deficient individuals. This last situation must be
differentiated from the frequently occurring brown color of the urine caused
by a metabolite. Also troublesome has been the induction of mammary
tumors in rats, which led the Food and Drug Administration to question its
use in humans. With satisfactory warnings included in a revised package
insert, furazolidone is approved in the United States for treatment of
giardiasis. Recommended dosage for adults is 100 mg 4 times daily and for
children 6 mg/kg/day divided into 4 doses for 7 days.
-------
46 WATERBORNE GIARDIASIS/THE DISEASE
Pregnancy
None of the three available anti-Giardia drugs has been proved safe during
pregnancy and if used at all during pregnancy, they should be administered
only to those women with severe symptoms definitely attributable to
giardiasis where benefit is judged to outweigh potential risk. When forced to
treat pregnant women, we have chosen quinacrine and have not recognized
any deleterious effects on the fetus.
Asymptomatic Carriers
Although many individuals with giardiasis are asymptomatic when
diagnosed and may never become symptomatic, there is the potential for
development of intermittent chronic symptoms in some; and particularly
with infected children or foodhandlers, there is the possibility of their
infecting others(25). I therefore recommend treatment of all infected persons
with no contraindication to use of the available drugs, such as pregnancy.
SOME PROBLEM AREAS
I am certainly -convinced that Giardia lamblia is a pathogen in many, but
not all of those infected, and that it is a leading cause of morbidity. In my
practice, I call it "my bread and butter parasite" as it is the most frequently
diagnosed cause of illness in my patients. Many others share my opinion.
Unfortunately, as with amebiasis, not everyone is yet convinced that it is a
definite pathogen and a satisfactory study showing increased morbidity in
those with giardiasis over a control population is not yet available.
1 have been unable in this presentation to discuss the pathogenesis of
Giardia in humans, which remains enigmatic. However, much exciting work
is being carried out in both humans and in animal models which should help
us better understand the pathogenesis. This includes the role of intestinal
bacteria and bile salt deconjugation(12); intraepithelial lymphocyte
counts(26), and immunoglobulin studies(27), and scanning electron
microscope evidence of involvement of the microvillous border and of
invasion of the bowel(28). Some of this work will be discussed elsewhere in
this Symposium.
An important unanswered problem for me is the patient with typical
symptoms and exposure history for giardiasis, in whom we are unable to
confirm the presence of parasites by available means, but who responds to
specific anti-Giardia treatment. It would appear that either our diagnostic
methods are not always sufficiently sensitive or that some individuals with
extremely light infections may respond very violently. Similar hyper-
reactivity is seen in the non-immune in other parasitic diseases such as
schistosomiasis and malaria.
One excellent drug (quinacrine) and two other quite useful drugs
(metronidazole and furazolidone) are available to us in this country, but all
have potential problems. Though more effective, quinacrine has the
potential to cause serious toxic effects in a small percentage of those taking
it, including toxic psychosis, vomiting, fever, and exfoliative dermatitis.
Metronidazole, though rather effective and well tolerated remains
-------
M.S. Wolfe 47
unlicensed for giardiasis, and as a suspected carcinogen and mutagen, there
are potential legal problems should long-term damaging effects eventually
be proven. Even now I hesitate to administer it to young children or
asymptomatic individuals or in pregnancy. Furazolidone is the least
effective available drug; it may cause more hypersensitivity and toxic side
effects; and it too has been suspected of having carcinogenic effects. It
appears unlikely that other presently available anli-Giardia drugs will be
approved for use here. A useful future contribution to treatment of giardiasis
would be a well controlled comparative study in this country of quinacrine,
metronidazole, and furazolidone.
REFERENCES
I. Center for Disease Control. 1977 Intestinal Parasite Surveillance.-Annual Summary 1976
2. Rendtorff, R. C. 1954. The experimental transmission of human intestinal proto/oan
parasites' Giardia lambha cysts given in capsules. Am. J. Hyg 59:209-220.
3. Walzer, P.O., Wolfe, M.S., and Schultz, M.G., 1971. Giardiasis in Russia. J Infect Dis
124.235-237, 1971.
4, Giardia lambha infections in travelers to the Soviet Union. 1974. Morbidity Mortality
Week Rep. 23.78-79.
5. Brodsky, R.E., Harrison, L.S., and Schultz, M.G. 1974. J. Infect Dis. 130:319-323,
6. Webster, B.H. 1958. Human infection with Giardia lamblia. An analysis of 32 cases
Am.J.Dig.Dis. 3-64-71.
7. Kononenko, Y.M. 1976. Erythema multiforme exudativum in a child with lambhal
cholycystitis. Pediatr. Akush Ginekol. 2.30-31, 1976.
8. Goobar, J.P. 1977. Joint symptoms in giardiasis Lancet 1 1010-1011.
9 Soto, J.M. and Dreiling, D.A. 1977. Giardia lambha. A case presentation of chronic
cholecystitis and duodenitis. Am. J. Gastroenterol. 67:265-269.
10. Kosyarska, J. 1977. Lambhasis as a cause of acute pancreatitis Wiad.l.ek. 30:875-877.
11. Hoskins, L.C., Winawer, S.J., Broitman, S.A., et al. 1967. Clinical giardiasis and intestinal
malabsorption Gastroenterology 53.265-279
12. Wright, S.G., Tomkins, A.M., and Ridley, D.S.1977. Giardiasis clinical and therapeutic
aspects. GUT 18:343-350
13. Nair, K.V., Sharma, M.P., Mithal, S., et al.1977. Comparative evaluation of diagnostic
methods in giardiasis, Indian J Med. Res. 66.417-419.
14. Jokipii, A.M.M. and Jokipii, L. 1977 Prepatency of giardiasis. Lancet L1095-1097
15 Danciger, M., and Lopez, M.: 1975. Numbers of Giardia in the feces of infected children
Am.J.Trop.Med Hyg. 24:237-242
16 Kamath, K.R., and Murugasu,R. 1974. A comparative study of four methods for detecting
Giardia lamhlia in children with diarrheal disease and malabsorption Gastroenterology
66.16-21.
17. Bezjak, B. 1972. Evaluation of a new technique for sampling duodenal contents in
parasitologic diagnosis. Am..! Dig Dis. 17 848-850
18. Madanogopalan, N., Rao, V.P., Somasundaram, A., et al, 1975 A correlative study of
duodenal aspirate and faeces examination in giardiasis, before and after treatment with
metronidazole Current Med. Res. and Opinion 3:99-103.
19. Marshak, R.H., Ruoff, M. and Lindner, A.E. 1968. Roentgen manifestations of giardiasis.
Am J Roentgenol Radium Ther Nucl Med 104.557-560.
20. Levi, G.C., deAvila, C.A., and Neto, V.A. 1977. Efficacy of various drugs for treatment of
giardiasis. A comparative study. Am.J.Trop.Med Hyg 26.564-565.
21. Andersson, T., Forssell, J., and Sterner, G. 1972. Outbreak of giardiasis: Effect of a new
antiflagellate drug, tmidazole Brit Med J. 2-449-451
22. Is Flagyl dangerous? 1975 Medical Letter Drugs Therapeut. 17:53-54.
23. Bassily, S., Farid, A., Mikhail, J.W., et al. 1970. The treatment of Giardia lambha infection
with mepacnne, metromda?ole, and furazolidone. J Trop.Med.Hyg. 73:15-18.
24. Wolfe, M.S. and Moede, A.L., 1978. Serum sickness with furazolidone.
Am.J.Trop.Med Hyg. 27.762-765
25. Black, R.E., Dykes, A.C., Sinclair, S.P. et al.1977. Giardiasis in day-care centers. Evidence
of person-to-person transmission. Pediatrics 60486-491.
-------
48 WATERBORNE GIARDIASIS/THE DISEASE
26. Wright, S.G. and Tompkins, A.M. 1977. Quantification of lymphocytic infiltrate injejunal
epithelium in giardiasis. Clm.Exp Immunol 29-408-412
27. Thompson, A., Rowland, R., Hecker, R., et al. 1977 Immunoglobulm-bearmg cells in
giardiasis. J. Clin. Pathol 30:292-294.
28 Erlandsen, S.L., and Chase, D.G. 1974 Morphological alterations in the microvillous
border of villous epithelial cells produced by intestinal microorganisms. Am J.Chn Nutr
27:1277-1286.
Discussion
M. SCHULTZ: I would like to emphasize the point that people with
clinical giardiasis are often not diagnosed by their physicians. We found that
this was very common early in the development of giardiasis in Americans
returning from the Soviet Union.
The story 1 would like to cite to illustrate this is the case of the wife of an
American .anatomist. She went with him to Leningrad, and when they
returned home she was ill. She had had an ileostomy in place before she took
the trip, and when she returned she had copious outpouring of fluid from her
ileostomy. She went to see her surgeon who decided to do a resection and
when it did not work, he did another resection.
Thereafter, the husband read a story in the New York times that giardiasis
was common in American travelers to Leningrad. He bought $1.98 worth of
Atabrine (quinacrine) at a drugstore, and without even testing his wife,
treated her and cured her promptly.
This is a case of prolonged suffering due to a physician not reaching a
correct diagnosis; 1 have seen many other instances. Those of us who are
disciples of this new cult of giardiasis need to preach the gospel about it. It is
now the most common intestinal protozoan in America, and an exceedingly
common illness in people traveling overseas, as Dr. Wolfe has just pointed
out.
M. WOLFE: I see patients all the time who have been previously seen by
internists and gastroenterologists. Some of these people have had extensive
and expensive workups including X-rays and biopsies, before stool
examinations were ever performed. Or because a single stool examination,
not performed well in a good laboratory, was negative, their physicians
immediately discounted the presence of parasites.
R. RENDTORFF: Dr. Wolfe, 1 commend you on a very fine
presentation, particularly on your description of the epidemiology and
clinical features of the disease. 1 am interested in whether the patients who
travel abroad have been on antibiotic therapy, which is a very common form
of therapy private physicians render to people who travel to far away places.
We have had the news of the value of doxycycline in ordinary traveler's
diarrhea disease. Were the patients you discussed on antibiotic therapy
which may have influenced this clinical symptomatology?
More importantly, have you used matched controls? I think it is critically
important that matched controls be used to determine clinical syndromes,
especially ones that have been around for a long time. Giardia has been here
longer than I have, and 1 am sure many of the older people here will
remember that the incidence of this disease used to be much higher than it is
today, and we did not have the kind of fear and clinical worry about the
disease that we do today.
-------
M. S, Wolfe 49
1 am not saying that I do not believe it is a pathogen under certain
circumstances; I think it is. When you mention such things as urticaria and
arthritis, it is reminiscent of what happened during the amebiasis scare in the
1940's when amebiasis was accused of everything from falling hair to
ingrown toenails. We are getting into a similar situation with Giardia. Do
you really think arthritis is a phenomenon caused by this parasite?
M. WOLFE: Concerning your point about antibiotics, doctors for my
State Department population do not recommend any sort of antibiotic or
drug prophylaxis for diarrhea. 1 have received many calls regarding
doxycycline prophylaxis and 1 steadfastly refuse to recommend or prescribe
it to anybody. All of our doctors and nurses overseas have followed this
policy.
1 think antibiotics do play a role in Giardia as does Flagyl
(metronidazole). After treatment with Flagyl and antibiotics, patients may
experience persistent diarrhea, and sometimes bloating and gas. When we
examine their stools with fungus cultures, we sometimes discover a Candida
overgrowth. This can be an after effect of antibiotics causing these
symptoms. 1 treat people with a possible fungus overgrowth with Mycostatin
(nystatin).
1 did mention that a good clinical study should be done with matched
controls. We have two sets of controls for the 670 cases we have been
discussing. These may not represent totally suitable controls since some
different questions were asked of the controls. They are matched primarily
for epidemiological purposes, not for clinical purposes. You need a
prospective study of those who have giardiasis, along with controls, asking
them the same questions. However, you will have to differentiate within the
controls between people who are symptomatic and totally asymptomatic.
We have a group of controls called the "worried well" who think they have
parasites. Many of them have parasitophobia and insist on workups every
year or every two years when they return from overseas. We are never able to
find anything on them, and although sometimes we treat them, they do not
get better. They are probably people with irritable bowels.
You have to be careful in selecting controls. In the CDC evaluation of
people coming back from the Soviet Union, less than half of the patients
were proven to have giardiasis. The other cases were called giardiasis on
clinical grounds. There are many people with illnesses that might mimic
giardiasis, who if used as controls, will have the same symptoms as your
proven Giardia cases.
It is a major problem finding suitable controls in a clinical situation where
people are coming in after travel because they have problems. Other
populations might serve better wherein you can use people who have not
been exposed to Giardia to the extent people living abroad have, and you can
be more certain that they do not have Giardia. If you then ask them the same
questions as you have asked those in proven Giardia cases, you might have a
better type of control.
Concerning your point about whether Giardia is a pathogen, I can say the
symptoms are there. We are seeing a lot of sick people with Giardia parasites.
We treat them, and the symptoms go away. I agree that some of the bizarre
symptoms of arthritis and erythema multiforme are very reminiscent of the
great debate that has gone on for years about amebiasis; whether
-------
50 WATERBORNE GIARDIASIS/THE DISEASE
constitutional symptoms such as fatigue and headache and dizziness and
other things considerably remote from the bowel are caused by these
parasites.
The evidence is that when you treat these people the urticaria goes away,
the fatigue goes away, and some of the headache and dizziness goes away.
We cannot prove it, but I am willing to accept that many of these symptoms
can be caused by both Giardia and amebae.
Whether the cause is a toxin or a metabolic breakdown product or some
tissue breakdown product, I do not know. When people come in with
myalgias and marked prostration, and we find that they have giardiasis or
amebiasis, and treat them with what we consider to be very good drugs,
many of these people improve. I think there is a very good evidence that these
parasites are toxic.
A. CONNELL: Is there any evidence that humans develop acquired
immunity in any form? For example, do people who have been treated, and
then travel abroad, get reinfected? If they do get reinfected, do they show
acute symptomatology or do they develop a chronic type of infection?
In the treatment of cryptic cases, do you see any parasites in the stool and
have you followed the stool in these patients with this type of therapy?
M. WOLFE: There seems to be some evidence from the data in Colorado
that there might be acquired immunity. In the 670 cases we have been
discussing, there are a fair number of repeaters. These are people who have
gone abroad for two-year assignments for the most part, or live in
Washington and travel abroad on temporary assignments. I have not yet
examined these cases carefully enough to see whether these repeated
infections give the same symptoms as the initial infection.
It is not often that we see people in the acute stage because they usually
have an undetermined duration of infection. They may have been abroad for
two years and might have become infected on the first day, the last day, or in
between. If they have gone through the typical acute stage that only lasts
three to four days, they pass this off as some kind of common diarrhea that
people experience.
It is very difficult to work with this population in relation to previous
problems. This is especially true if they have been infected for one or two
years and there is development of an immunity process. They may well have
developed some form of immunity different from that of the traveler who is
in and out of the Soviet Union on a one-or two-week trip and then develops
the symptoms shortly after he returns to the United States.
There are many intriguing questions which might be answered when we
have the proper tools and can measure certain types of immunity,
particularly at the tissue level in the small bowel.
A major problem is obtaining enough patients who are willing to subject
themselves to duodenal tubing and perhaps multiple intestinal biopsies
before and after treatment.
Your other question of so-called cryptic cases concerns people in whom
we find no parasitic etiology for their symptoms. However, we are not testing
for anaerobes and fungi such as is being done at the Hospital for Tropical
Disease in London. If we do not find parasites before treatment of these
people, then treat them and they get better, I do not really see an indication
to do any follow-up examinations. I do not think it is practical or very
worthwhile. Once we treat them and they get better, we release them. If they
-------
M. S. Wotfe 51
are still having problems after taking Atabrine (quinacrine) which we find to
have a 95% cure rate, the probability of it being giardiasis is not too high.
Then we can begin intestinal biopsies, malabsorption studies, and other
workup to determine the cause of the diarrhea.
T. N ASH: The first major problem 1 see is the treatment of asymptomatic
patients. Is there any general agreement on the role of the asymptomatic cyst
passer infecting other groups of people?
The second concerns those patients that present brief abdominal type
symptomatology. I wonder if the patient following that actually has
pancreatic enzyme elevations and symptomatic giardiasis. How many
people have studied or looked at that type of problem in a prospective
manner?
M. WOLFE: Certainly regarding children 1 see every reason to treat them
because we have many situations wherein if one child in a family is infected,
other siblings may become infected from him. It does not work the other way
around as much. An adult who is infected and is asymptomatic and washes
his hands, even if he is a food handler, is probably not going to infect other
people. Certainly if he is a food handler, though, public health logic calls for
him to be treated.
Whether or not symptoms will develop in these asymptomatic people over
a long period of time, I cannot say as I do not usually follow them but treat
them. It would be worthwhile to follow a group of these people and see
whether symptoms develop.
Pancreatic enzymes would be very worthwile to examine. Apparently, the
Russians and others in East Europe are doing a great deal of pancreatic
enzyme work. The Tropical Disease Bulletin contains many articles about
the influence of Giardia on liver and pancreatic function. These people are
performing some very discreet type of testing that no one has even begun to
do here.
D. JURANEK: Just to further emphasize your point about treating
asymptomatic patients, especially children, there is at least one study that
was done by the Center for Disease Control (Black, et al. Pediatrics, 60:486,
1977). We found that the introduction of a single infected child to a day care
center nursery-type situation, rapidly spreads the disease. In this case about
50% of the children developed giardiasis. We compared the prevalance of
this infection to socioeconomically matched children of the same age who
did not attend day care centers. The prevalence of giardiasis was about one
per cent.
M. WOLFE: Giardiasis is very prevalent in institutionalized children,
also.
D. JURANEK: We need to better define exactly which type of
asymptomatic cyst passer in epidemiologic terms (age, occupation, personal
hygiene and sanitation) poses a threat to others and make specific treatment
recommendations for that type or group of persons. There are too many
asymptomatic cyst passers who pose virtually no public health threat to
warrant a blanket statement that everyone be treated.
M. SCHULTZ: The work of Black, et al, was recently confirmed by
Keystone in Toronto, in a published paper in the Canadian Medical
Association Journal in August. He studied two nurseries in which Giardia
spread throughout very rapidly, and was significantly pathogenic. He
recommended that carriers be treated. He observed that Canadian born
-------
52 WATERBORNK GIARDIASIS/THE DISEASE
children have a much higher rate of giardiasis than children in the same
classroom who came from overseas. He speculated, although he did not
prove it in any way, that there might be some immunity in the children who
had previous infection, and that the children from abroad came into this
classroom and were susceptible to infection.
Lastly, on this point concerning immunity in people, Richard Wright, et al
(American Journal Epidem., 105:332, 1977) did an epidemiologic study in
Colorado which showed that native Coloradoans, who were long-term
residents, had lower carrier rates than short-term residents, implying that
there might be immunity in long-term residents.
M. WOLFE: As you may know, for years the Russians have been denying
that there is a problem with giardiasis in the Soviet Union. We have had very
little direct information and some indirect information from translations of
Russian literature. However, one notable piece of information in a study of a
nursery in Leningrad was that 90% of preschool children attending the
nursery were infected with giardiasis. Their literature certainly admits that a
problem exists.
W. JAK.UBOWSK.I: I would like to thank you for an interesting and
informative talk, and I think we could probably spend about two and a half
days just discussing the symptomatology and the diagnosis and treatment of
giardiasis.
1 had a close encounter with Giardia about 2 years ago, and I had a course
of treatment with metronidazole (Flagyl) and also with quinacrine
(Atabrine). Five months later the disease spontaneously cured itself. During
the course of my infection I had a concurrent yeast infection. In Berlin, New
Hampshire, the water concentrates that we examined had large numbers of
yeast cells. In your experience in examining stool specimens from infected
individuals have you come across any concurrent infections with other
organisms?
M. WOLFE: That is an interesting point and it opens up the whole
question of what is the possible role of concomitant organisms.
Enterobacteria seem to have some association as determined by studies done
in England and in India. Some people have postulated that there is some
organism, virus, bacteria, or maybe a fungus in the water in Leningrad that
makes the disease so much more common in travelers to Leningrad than it
seems to be in travelers to other parts of the world.
Speaking specifically of yeast, we see yeast frequently in the stool. It is
partially a function of how long an unpreserved stool has stood before being
examined, as a small amount of yeast might have multiplied before you
examine the specimen. It is possible there may be some inter-relationship
with intestinal yeast but 1 think special attention would have to be paid to
this in examining specimens, including fungus cultures of duodenal
aspirates, to determine the relation between Giardia and the fungus.
Some people post-treatment seem to get a disruption of their normal
bowel flora and an overgrowth of yeast, and this can give symptoms not
totally typical of Giardia but reminiscent of those that occur with both
Giardia and amebiasis. When you treat for this yeast infection with a non-
toxic 7 day course of Mycostatin (nystatin), some of these people do tend to
improve. What the interaction is between bacteria and yeast or other
organisms with Giardia is another interesting aspect that hopefully some of
you here will be able to investigate.
-------
53
The Possible Use of an Indirect Immunofluorescent
Test Using Axenically Grown Giardia Lamblia
Antigens in Diagnosing Giardiasis
G.S. Visvesvara and G.R. Healy
General Parasitology Branch, Bureau of Laboratories,
Center for Disease Control, Public Health Service, U.S. Department of
Health, Education, and Welfare, Atlanta, Georgia
ABSTRACT
The possibility of using an indirect Immunofluorescent (IIF) test for diagnosing
giardiasis is discussed. When we treated Giardia lamblia organisms in 104 M 2-
mercaptoethanol, before preparing antigens-slides, all nonspecific reactions were
eliminated Seventy-two of IDS sera (66.6Cf) from patients with symptomatic and
asymptomatic giardiasis had tilers of 32 or greater, whereas none of 53 normal sera had
liters as high as 32. The test-lo-test reproducibihty was very good, with no difference
greater lhan one fourfold dilution amongG. lamblia antibodies in positive sera were
completely absorbed by G. lamhlia antigens However, when positive sera were trealed
with Emamoeha hiMolvnca, Trichonionas vagmali\, or Ewheruhia coh ihe anli-G
lamblia liter did not decrease appreciably, which indicales thai ihe IIF reaclion is very
specific.
In addilion lo IIF, indirect hemagglutinalion (IHA), direct agglutination (DA), and
double diffusion in agarose gel (GO) were also evaluated The IIF test was found to be ihe
mosl useful lest and it could be used as an ad]unct to stool or duodenal aspirate
examination in diagnosing giardiasis
Giardia lamblia, a protozoan parasite of man, is found all over the world.
It is the most frequently diagnosed intestinal parasite in public health
laboratories in the U. S., according to an ongoing intestinal parasite survey
being conducted by the Center for Disease Cont rol (1). G iardiasis is endemic
in the U. S., and an estimated 3 to 7% of the adult population harbor the
parasite (2). G. lamblia normally resides in the duodenum and upper
jejunum and frequently causes acute or chronic diarrheal illness, and
sometimes leads to steatorrhea and malabsorption syndrome. In some cases
it may even cause lesions in the small intestine (3-5). In other instances the
infection may be lost spontaneously after 1 to 4 weeks of patency (6). The
reasons for this variation in host susceptibility are not yet fully understood.
The high incidence of giardiasis in hypogammaglobulinemic individuals (3);
the low infection rate among the permanent residents of Colorado (7) —
which is an endemic area, as evidenced by the high infection rates among
visiting skiers (8) — and the fact that experimentally infected mice develop
prolonged resistance to re-infection with G. muris (9) strongly suggest an
immunologically mediated phenomenon. Until recently, however, no
serological test was available to study anti-G. lamblia antibodies in patients'
-------
54 WATERBORNE GIARDIASIS/THE DISEASE
sera. Ridley and Ridley (10) recently described an indirect
immunofluorescence (11F) test using G. lamblia cysts as antigen. They found
a rough correlation between histological malformation and antibody titer.
However, they also reported that finding a reliable source of antigen is
always a problem. Another problem that we have had difficulty with is the
autofluorescence of the cysts. The successful development of an axenic
culture medium by Meyer (11) for the routine cultivation of G. lamblia
trophozoites solves these problems. Using the axenically cultivated G.
lamblia trophozoites as antigens, 4 tests were evaluated as potential
diagnostic procedures for detecting giardiasis: double diffusion (DD) in
agarose gel; direct agglutination; indirect hemagglutination; and I IF. Our
preliminary investigation revealed that the 11F test was the most promising.
In this report we describe the results of using the 1IF test for detecting anti-G.
lamblia antibodies in the sera of humans.
MATERIALS AND METHODS
Antigen Preparation
A strain of Giardia lamblia was obtained from Dr. E. A. Meyer. It had
been isolated from the duodenal aspirate of a human patient and was grown
axenically in Meyer's HSP-1 medium (11), which contains 15 to 20% human
serum. Unfortunately, the organisms grown in this medium were found to be
unsuitable because they reacted with even normal sera in the IIP test. We
then tried to grow the organisms in Meyer's medium containing bovine or
rabbit serum instead of human serum and later grew them in Diamond's TP-
S-l medium (12). Actively growing, 3 to 4-day-old Giardia lamblia were
dislodged from the walls of the 16- X 125-mm screw-capped tubes by
immersing the tubes in ice water for 5 to 6 min. The tubes were then
vigorously rolled between the palms and centrifuged at4°Cat 250 Xgfor 10
min in a refrigerated centrifuge (Damon, model CRU-5000 1EC Needham
Heights, Mass.).* The supernatant was aspirated, and the sediment from
several tubes was pooled and suspended in sterile normal saline in the ratio
of 50 ml saline to 0.5 ml of packed cells. The suspensions were then
centrifuged as above. The organisms were washed 6 times in this fashion, and
the sediment from the final wash was fixed in 1 % neutral formalin in normal
saline. The parasite suspension was adjusted to obtain 500 to 1000 organisms
per low-power (100X) microscopic field. A drop of this antigen suspension
was then placed in each of the 12 wells on a slide and allowed to dry at room
temperature. Prepared slides were wrapped in tissue paper, labelled, dated,
and stored at -70°C until used. Alternately, the pooled organisms from the
culture tubes were suspended in normal saline containing 10~4 M 2-
mercaptoethanol (2 ME—saline) and allowed to stand at 37°C for 2 hours.
They were then centrifuged as above and washed twice in saline containing 2
ME and a third time in normal saline. The sediment from the final wash was
suspended in 1% formalin, and slides were prepared and stored as above.
*Use oj trade names is for identification only and does not constitute endorsement by the Public-
Health Service or bv the U.S. Department of Health, Education, and Welfare
-------
G. S. Visvesvwa 55
Antigens for the DD test were prepared by washing the organisms 6 times
in normal saline, freezing the sediment after the last wash at -20°C,
sonicating it in a Bronwill sonicator (Biosonic 2A, Bronwill Scientific,
Rochester, N.Y.) for 2 min, and centrifuging it at 12,060 X g in a Sorval
refrigerated centrifuge (Ivan Sorval, Inc., Norwalk, Conn.) for 30 min. The
supernatant was collected, added in 0.5-ml aliquots to a number of vials, and
lyophilized.
Sera
Of the 161 sera tested, 16 were from the Clinic for Tropical and Parasitic
Diseases, Toronto, Canada; 19 were from persons with clinical symptoms of
the infection during a giardiasis epidemic in New Hampshire; 28 from
Cherokee Indian children (collected during 1965-66); 27 from patients seen
by private physicians, 6 from CDC personnel returning from trips abroad;
12 from the Prince Leopold Institute of Tropical Medicine, Antwerp,
Belgium; and 53 from normal persons. Of the 53 normal sera, 15 were from
CDC personnel and 38 from other normal healthy persons who had no
gastrointestinal problems but had donated blood for an arbovirus survey.
The sera used in the study were grouped as follows: group 1, 68 sera from
clinically symptomatic patients with acute, subacute, or chronic giardiasis
who were confirmed to have G. lamblia trophozoites and/or cysts in the
stools or trophozoites in the duodenal aspirates; group 2, 40 sera from
asymptomatic individuals who were excreting G. lamblia cysts; group 3, the
53 sera from normal individuals.
Indirect Immunofluorescence Test
Serial twofold dilutions of each serum (0.05 ml) beginning at 1:2 were
prepared in U-type microtitration plates (Linbro Scientific Co., Inc.,
Hamden, Conn.) using microtitration loops. Phosphate buffered saline
(PBS), pH 7.6, was the diluent. The antigen slides were removed from the
freezer, washed once in PBS, and allowed to dry. One drop of each dilution
of serum was transferred to individual wells in a 12-well slide, and the slides
were incubated at 37°C for 30 min in a moist chamber. The slides were then
washed three times in PBS at 10 min intervals and allowed to dry. A drop of
the conjugate, consisting of goat antihuman globulin labelled with
fluorescein isothiocyanate (FJTC), CDC F/P 48, protein 40 mg/1, was
added to each well of the slide. The conjugate was used at a dilution of 1:500
with a 1:500 Evans' blue counterstain in PBS, pH 7.6. The concentration of
the conjugate and the Evans' blue was arrived at by block titration against
known positive and negative sera.
The slides were again incubated at 37° C for 30 min in a moist chamber and
washed three times in PBS at 10 min intervals. They were then mounted with
buffered glycerin, pH 9.0, and covered with a cover slip. The slides were
examined with a Leitz Ortholux microscope equipped with a Ploem vertical
illuminator, 2 KP 490 exciter filters, and a K 530 barrier filter at 40X
magnification. The intensity of fluorescence was graded on a scale of 1 + to 4+
with a 1 + or greater reaction considered positive. In the negative reaction the
organisms did not show any fluorescence. Positive control serum was
-------
56 WATERBORNE GIARDIASIS/THE DISEASE
obtained from an individual who had clinical symptoms of giardiasis and
was passing Giardia trophozoites and cysts in the stool, and negative control
serum was obtained from an individual with no history of travel or
gastrointestinal disturbance. A PBS control was included in each test run.
The DD test was performed according to Kagan and Norman (13). Three
peripheral wells surrounding a central well were' cut in the agarose gel. The
anti-G. lamblia antiserum positive control made in rabbits was placed in the
central well and the test materials such as duodenal fluid, G. lamblia
antigens, and saline were placed in the peripheral wells.
RESULTS
Indirect Immunofluorescence (IIP)
Because antigens prepared from organisms grown in HSP-1 medium
fluoresced brightly even with normal sera, they were not suitable for the IIP
test. Antigens prepared from organisms grown in HSP-1 medium containing
either rabbit serum or bovine serum or those grown in TP-S-1 medium
fluoresced brightly at the 1:128 or 1:256 dilution of the positive control
serum. However, they also fluoresced faintly at the 1:64 dilution of the
normal serum. Even though the fluorescence patterns of the positive and
negative sera were quantitatively different, we found it difficult to make use
of this fact in the absence of a fluorometer. However, the 2 ME-treated
organisms were very satisfactory for our purpose, in that most of the normal
sera tested did not react at the 1:16 dilution, and most of the sera from
symptomatic patients fluoresced brightly at the 1:16 or higher dilution of the
serum. The optimal concentration of 2 M E (1 X 10~4M) and the optimal time
(2h) for keeping the organisms in 2 ME-saline in order to eliminate all
nonspecific reactions were determined by suspending and washing the
organisms at different concentrations of 2 ME (1 X 1(P, 1 X 10~4, 1 X 10~5, 1
X 10'* M) and for different time intervals (60, 90, 120, 150 min) respectively.
Fig. 1-5 show the types and degrees to which the positive sera fluoresced at
different dilutions relative to those of the negative control at a 1:16 dilution.
Neither the negative control nor the diluent control reacted at all at the 1:16
dilution, whereas the 1:16 dilution of the positive control serum fluoresced
brightly at 4+ uniformly over the entire surface including the flagella. At
higher dilutions the fluorescence was confined to the surface membrane and
the intensity was considerably less. The reciprocal of the lowest dilution of
the test serum that reacted at 1+ was defined as the titer of that serum. The
distribution of IIP tilers among the groups of sera tested is shown in Table 1.
Fifty-two of 53 (98%) of the sera from normal persons did not react with G.
lamblia antigens at a titer of 16, and none reacted at a titer of 32. Hence a titer
of at least 32 can be considered as diagnostic and indicative of infection.
Forty-four out of the 68 test sera (64.7%) from symptomatic patients were
positive for G. lamblia antibodies at a titer of 32 or greater. Of the sera from
patients with asymptomatic giardiasis, 28 of 40 (70%) had liters of 32 or
greater. Thus, if 32 is defined as the diagnostic titer, then ihe test has a
sensitivily of 66.6%.
-------
G. S. Visvesvara
57
NOTE TO FIGURES 1-5: Photomicrograph illustrating the types and degrees ofimmuno-
fluorescent staining of G. lamblia. Photomicrographic conditions as well as the
degree of enlargement and printing time were the same in all cases.
FIG. I. A 4+ reaction at the 1:16 dilution
of the positive serum. Note the
bright and uniform staining of all
the organisms.
FIG. 2. A 3+ reaction at the 1:64 dilution
of the positive serum. The staining
is confined to the surface mem-
brane in most cases.
FIG. 3. A 2+ reaction at the 1:256 dilution
of the positive serum. The staining
is confined to the surface mem-
brane, and the brightness is
considerably reduced.
-------
58
WATERBORNE GIARDIASIS/THE DISEASE
FIG. 4. A 1+ reaction at the 1:1024
dilution of the positive serum. The
staining of the surface membrane
is still discernible even though it is
dull.
FIG. 5. Negative reaction at the 1:16
dilution of the negative control.
In an effort to determine the test-to-test reproducibility and consistency of
the test results, we measured the IIP liters on 1 positive and 1 negative serum
on 12 different days. The negative serum was consistently nonreactive at the
1:16 dilution. The positive serum had a liter of 256 on 5 occasions, a tiler of
128 four limes, and a tiler of 512 Ihree times (Table 2). The lest-to-test
reproducibility was very good, with differences in results never ranging
higher than 1 fourfold dilution.
One serum wilh a high liter of 512 and 1 wilh a low liter of 64 were used lo
delermine the specificily of the tesl. A 0.5-ml aliquot of each serum was
mixed with 0.5 ml packed G. lambliatrophozoites, allowed loslandal 37° C
for Ih, and stored at 4°C overnight. Samples were Ihen centrifuged al 3000
RPM for 20 min, Ihe supernalanls were collected and designaled as Giardia-
absorbed sera, and Iheir I IF lilers determined. These 2 sera were nonreaclive
even al Ihe 1:2 dilution, indicating that all anti-Giardia antibodies had been
removed. In anolher experiment IIP tilers were determined for 3 sera (2 with
tilers of 256 and 1 with a liter of 512) after they were similarly absorbed
separalely wilh Entamoeba histolytica or Trichomonas vaginalis
Irophozoiles, or Ihe baclerium, Escherichia coli. These absorbed sera
-------
G. S. Visvesvara
59
6c
i
5
^
1^
^.
u.
• tnbution <>/
^
-i
-c
£
o
Z ro
CD
0
1—
K
CNJ
O
CN
LO
CO
LO
CN
CN
"xt
CD
C\l
CO
CD
CO
-vf
CN
£ ^
S 2
00 0 " "I
CD ^ ^ ^
CN CO O LO
•—CD o r-^
m CD o CN
<— CN
CD O O CD
<— CD O r~-
-- ^r o LO
CD ^ t- <^-
^t CNI CD CN
•st~ i— co ro
r~- LO oo o
ro LO
o <"
O r: ir,
_ -t-^ CD
~ ro co
? E m
E o S
2 a^
E |" § -5
i- ^ u o -5
/-\ -^r -y ,
-------
60 WATERBORNE GIARDIASIS/THE DISEASE
Table 2. Tesl-lu-tesl reprodiicibililv of IIP liters measured on
12 different davs on two sera with G Lamblia antigen.
liter
Sera tested
Positive
control
Negative
control
<16 16 32 64 128 256 512 1024
- - - 4 5 3 -
12
showed no appreciable decrease in IIP tilers indicating that they did not
contain cross-reacting antibodies.
Seronegative But Symptomatic Persons
Sera from 24 of 68 patients (35.3%) had liters of 16 or less, i.e., lower than
the defined diagnostic liter. These patients were clinically ill with diarrhea
and were stool positive for G. lamblia trophozoites and/ or cysts. The case of
a patient who had chronic giardiasis but was consistently seronegative is
described below.
Case of Chronic Giardiasis: A 29-year-old Caucasian female apparenlly
acquired giardiasis in Chile 3 years ago. Since lhal time, she has suffered
episodes of intermittent diarrhea and epigastric pain. She has been treated
with atabrine, metronidazole, and furazolidone. She recently completed a
course of metronidazole therapy (250 mg 3 times daily for 17 days). She was
stool positive for G. lamblia before she received this medication and was still
positive for G. lamblia the day after ihelrealmenl ended. One of her samples
of the duodenal aspirale was positive for G. lamblia trophozoite antigens
(Fig. 6) in the DD test. The duodenal sample reacted with the anti-G. lamblia
antiserum made in rabbits to form 2 precipitin lines, both of which became
confluent with 2 of the 7 precipitin lines produced by the homologous
antigen-antibody systems indicating antigenic isology. Paired samples of the
patient's serum obtained 3 months apart had normal levels of
immunoglobulins (Table 3). However, she does not have anli-Giardia
antibodies.
Table 3. Immunoglobuhn concentration of sera from a
patient with symptomatic giardiasis
Serum sample
1
2
Normal values
for 1 3 yr and
older
Immunoglobuhn (mg%)
igG
1310
1350
(800-1800)
IqA
120
120
(90-450)
IqM .
21 1
208
(60-280)
-------
G. S. Visvesvara 61
FIG. 6. Precipitin patterns obtained in
agarose gell showing interactions
between the antigens of G. lamblia
(GIA AG) and the giardiasis
patient's duodenal sample (DUO)
with the rabbit anti-G. lamblia
(ANTl GIA) serum. Normal
saline (SAL) served as the
negative control.
DISCUSSION
The data presented here indicate the usefulness of the I IF test as a
diagnostic tool for detecting antibodies to G. lamblia in serum. The IIP test
appears to be both specific and reproducible. Absorbing the positive serum
with homologous antigens (G. lamblia) completely removed the anti-G.
lamblia antibodies, whereas absorbing the serum with heterologous antigens
(e.g., E. histolytica, T. vaginalis, and Escherichia co/i) had no appreciable
effect on the anti-G. lamblia liter. The test-to-test reproducibility was found
to be good, with results never varying more than 1 fourfold dilution.
However, the major drawback of the test is its sensitivity; it cannot be used to
detect all cases of giardiasis. Only 72 of 108 (66.6%) patients with giardiasis
(including both symptomatic and asymptomatic) had diagnostic titers of at
least 32. Although no false positives were detected, many false negatives
were. A similar problem seems to exist in amoebiasis serology also. Healy et
al(14) state that amoebic antibody is not necessarily detected in every
infected individual. According to them, "Some people infected with the
organism show no antibody to antigens in the 1HA test." They think these
negative results probably stem from the fact that not all strains of E.
histolytica are tissue invaders. Neal et al(15) also showed that many
asymptomatic individuals who were passing cysts of noninvasive E.
histolytica in their stools did not have measurable IH A or CF antibodies to
these amoebae in their sera. A similar situation may apply to some of our
patients who are seronegative but positive for Giardia lamblia in stool
examinations.
Recently Ridley and Ridley(lO) described an 1IF test for detecting
circulating antibody against G. lamblia. They used G. lamblia precysts as
antigen and obtained positive results for 32 of 36 patients with giardiasis and
malabsorption, 0 of 2 patients with giardiasis and no malabsorption, and 10
of 34 patients with malabsorption but with no detectable G. lamblia
infection. However, they reported having difficulty in obtaining a
satisfactory and reliable source of antigen. This problem can be solved by
axenically cultivating G. lamblia trophozoites. The nonspecific dull
fluorescence that we encountered during our initial experiments with
axenically cultivated trophozoites was easily eliminated by pretreating the
organisms to be placed on slides with 2 ME-saline. The mode of action of 2
ME is not clearly understood, but it probably reacts with the bound serum
products and degrades them to unmask antigenic sites and make them
available to the anti-Giardia antibodies.
-------
62 WATERBORNE GIARDIASIS/THE DISEASE
Ament and Rubin(3) reported finding no parasites in four of seven
patients even after repeated stool examinations. Burke(5) estimated that
10% to 50% of infected individuals do not excrete the parasites in their feces.
In this situation, the IIP test would provide a valuable diagnostic tool.
ACKNOWLEDGEMENTS
We are indebted to: Dr. Philip F. Stuart of the Clinic for Tropical and
Parasitic Diseases, Toronto, Canada; Dr. Eddy Willaert of the Prince
Leopold Institute of Tropical Medicine and Hygiene, Antwerp, Belgium; to
many physicians throughout the U.S. for supplying us with sera and clinical
histories of patients; Mr. Charles Peters, Serum Bank, Center for Disease
Control, Atlanta, Georgia, for the normal serum samples; Miss Dorothy
Allain for performing the IHA test; Mrs. Floy Brand for excellent technical
help; Mr. Don Howard for taking the I1F photographs; Dr. Shirely E.
Maddison of_the Parasitic Immunochemistry Branch, CDC, for helpful
suggestions; and Dr. Alexander Sulzer of the Parasitic Serology Branch,
CDC, for the FITC-conjugated antihuman immunoglobulin and for
stimulating discussions.
REFERENCES
I. Center for Disease Control. 1977. Intestinal Parasite Surveillance Annual Summary 1976
2 Schultz, M. G. 1975. Giardiasis. J.A.M.A. 233'1383-1384
3 Ament, M.E., and Rubin, C. E. 1972. Relation of giardiasis to abnormal intestinal
structure and function in gastrointestinal immunodeficiency syndromes
Gastroenterology. 62.216-226.
4. Petersen, H. 1972. Giardiasis (Lambhasis). Scand J Gastroent. 7 (suppl. 14): 1-44
5 Burke, J. A. 1977. The clinical and laboratory diagnosis of giardiasis CRC Cnt Rev.Clm.
Lab. Sci. 8:373-391.
6. Danciger, M., and Lopez, M. 1975 Numbers of Giardia in the feces of infected children.
Am. J. Trop Med Hyg. 24:237-242.
7. Gleason, N. N., Horwitz, M.S., Newton, L.H., and Moore, G.T. 1970 A stoolsurvey for
enteric organisms in Aspen, Colorado. Am. J. Trop. Med Hyg. 19:480^84
8. Moore, G. T., Cross, W. M., McGuire, D., Mollohan, C. S., Gleason, N. N., Healy,G. R.,
and Newton, L. H. 1969 Epidemic giardiasis at a ski resort. New Engl. J. Med 281.402-
407
9. Roberts-Thomson, I. C., Stevens, D. P., Mahmoud, A. A. F., and Warren, K. S. 1976.
Acquired resistance to infection in an animal model of giardiasis. J. Immunol. 117'2036-
2037.
10. Ridley, M.J., and Ridley, D.S. 1976. Serum antibodies and jejuna! histology in giardiasis
associated with malabsorption. J. Clin. Path. 29:30-34.
11. Meyer, E. A. 1976. Giardia lambha isolation and axenic cultivation. Exp. Parasitol
39:101-105.
12. Diamond, L. S. 1968 Techniques of axenic cultivation of Entamoeba hMolvnca
Schaudinn, 1903 and Entamoeba /iislo/vlica-hke amoebae. J. Parasitol. 54:1047-1056.
13. Kagan, I. G., and Norman, L. 1970. Serodiagnosis of parasitic diseases, pp. 453^(86. In
Blair, J. E., Lennette, E H , and Truant, J P. (eds.). Manual of Clinical Microbiology.
American Society for Microbiology, Bethesda, Md.
14 Healy, G. R., Kagan, I. G. and Gleason, N. N. 1970. Use of the indirect hemagglutmation
test in some studies of seroepidemiology of amoebiasis in the Western Hemisphere. Hlth
Lab. Sci. 7.109-116.
15 Neal, R. A., Robinson, G.L., Lewis, W. P., and Kessel,J.F. 1968 Comparison of clinical
observations on patients infected with Entamoeba hi.\iolytica, with serological liters of
their sera and virulence of their amoebae to rats. Trans. Roy. Soc. Trop. Med. 62:69-75.
-------
G. S. Visvesvara 63
Discussion
Editors' note: This paper was presented by G. R. Healy.
T. NASH: You didn't mention the specificity of the conjugate you were
using. Is it anti-everything?
G. HEALY: Yes, it is anti whole gamma globulin.
T. NASH: What did you take as the end point of your liter in the IFAtest?
G. HEALY: We determined a 32 liter as positive. We very seldom had sera
that went out much beyond 1024.
T. NASH: It is conceivable lhat the organisms produced protease of one
type or another, that in order lo really solidify your absorption experiment,
you might want to abosrbsera that has other types of IFA antibodies. I have
done similar IFA work with only a handful of patients, not the hundred that
you are dealing with, and basically I cannot differentiate between the
negative control.
G. HEALY: Were you able to pick up more positives in your symptomatic
patienls?
T. NASH: I would nol be able lo slale any numbers. I have aboul 8 sera,
bul these were all worked up patients lhal we have had under invesligation
for malabsorption and at least as far as sera go, using specifically the IgA,
IgM and IgG, I was not able to tell very clearly. The IgG seemed to be Ihe
anlibody of importance; IgM and IgA were nol very helpful.
G. HEALY: I have no idea how soon after infeclion or symploms Ihese
sera were oblained. We do have several individuals at CDC who were
symptomatic giardiasis cases. We obtain serum from everybody in the lab
every 2 years in case we contracl some exotic disease. It is forlunate we had
sera from these people, in one case going back lo 5 years before infeclion.
Two of them were negative before and after they were infected, and have
been serologically negative since. It is very disappoinling because we had an
opportunity to do some prospective examinations of their samples.
G. HULDT: Your conjugate is polyvalent, but probably reacted mostly
against IgG. As this is a small intesline infection with probably very lillle
invasion into the tissues, wouldn't you expect antibodies to be mainly of the
secretory IgA class? Have you looked at all for class specificity of your serum
antibodies, and have you looked for secretory IgA in secretions, as for
instance human milk?
G. HEALY: No, we have not had the opportunity. I think your point is
well taken.
G. HULDT: We have tried. We have only been working in this area for a
very, very short time and our malerial is so small. In 15 palienls we found
remarkably high amounts of IgA circulating and also secretory IgA.
-------
64
The Experimental Transmission of
Giardia Lamblia Among Volunteer Subjects
Robert C. Rendtorff
Departments of Community Medicine and Microbiologv,
College of Medicine,
University of Tennessee Center for the Health Sciences,
Memphis, Tennessee
ABSTRACT
Data presented 25 years ago are re-examined in light of present day knowledge of
giardiasis. Prisoner volunteers exposed to Giardia lamblia by various routes were studied
intensively under strictly controlled environmental conditions. In all experiments adequate
controls remained negative. Early studies to determine the infective dose showed 5 men
failing to become infected after receiving a single cyst given orally in a gelatin capsule.
Varying doses from ten to one million cysts produced 100% infection except in a group
where only 25 cysts were given to each man There is evidence of low infectwity in this latter
source which produced 6 infections in 20 men exposed. Infections were produced with a
single cyst of Entamoeba co/i and there is no biological reason why single cysts of Giardia
woulti not also be infectious. An experiment designed to test for survival of cysts in tap
water held at 8°C showed survival for up to 16 days, the longest period tested. One may
conclude from this that water is a very efficient vehicle for the transmission of Giardia and
small numbers of cysts m water could readily produce infections in a susceptible host
Other important factors discussed include that a large percentage of infections clear
themselves spontaneously in a relatively short time; that the dose is not related to
persistence of infection; that there are possible strain differences in the mfectivity of cysts
and that stools may have long periods when they are negative in a person who is truly
positive. Aside from transient changes toward looser and more frequent stools generally
unrecognized by the volunteer, no clinical symptoms could be attributed to these
experimental Giardia infections.
Between 1950 and 1952 while I was with the Laboratory of Tropical
Diseases at NIH, I undertook a series of investigations designed to
contribute knowledge concerning poorly understood aspects of host-
parasite relationships of human intestinal protozoa. This work was done at
the then new Federal Correctional Institution at Seagoville, Texas. These
studies were designed to obtain basic information such as the number of
parasites necessary to establish the infection, the length of the prepatent
period, the persistence of the infections and other aspects of clinical,
epidemiological and biological interest concerning protozan intestinal
parasites. The original series of papers were published in 1954 (1-3).
It should be stated at the outset that our interest was mainly focused on the
problems of amoebiasis. But since we considered it improper at that time to
expose prisoners to Entamoeba histolytica, the experiments were confined
to non-pathogenic amoebae: Entamoeba coli, Endolimax nana and
lodamoeba butschlii. In these studies we also included Giardia lamblia,
-------
R. C. Rendtoiff 65
which at that time was not generally believed to be an invasive pathogenic
parasite of man. Giardia was thought in the 1950's to cause occasional
problems of diarrhea in children but its appearance was so common and, in
adults so lacking in clinical symptomatology, that most considered it a non-
pathogen. As a result, we felt safe in exposing prisoners to Giardia.
Fortunately our studies proved this to be correct for we had no evidence of
any serious pathogenic effects of the parasites in the volunteers. 1 will
elaborate on this more later. Now a word on how these studies were
conducted.
Experiments were conducted in the prison under conditions which
allowed for a large degree of control over environmental factors. The degree
of this control limited the number of volunteers studied to a maximum of 20
men per group because of physical factors. Information was obtained from
34 Giardia infections. This represents the only controlled series of human G.
lamblia infections produced experimentally.
Participants were prisoners who volunteered after being fully informed of
the conditions of the experiments and the risks involved. Very rigid controls
were in effect over food and environment during the critical periods of the
study. Each subject was quartered individually in a modern brick building
and was separated from other inmates. They were fed from a special kitchen
and dining room which were under very strict control to prevent transfer of
parasites through food and were remote from the main prison cafeteria.
Men selected for the experiments showed no parasites in a total of ten
stool specimens collected approximately every other day. They were then
studied further under the controlled environmental conditions before being
exposed to the parasites. In each feeding experiment some men were chosen
randomly to serve as controls and these received no Giardia.
Each study group was observed for a period of approximately 6 months,
the first 10 weeks of which were under the specially controlled environmental
conditions. The remainder of the time, approximately 3.5 months, the
volunteers lived under conditions existing in the prison at large. In short, the
experiments were conducted in such a manner that the transmission of
parasites outside of the experimental means was thought to be very unlikely.
This was substantiated by the fact that none of the numerous controls ever
acquired the infection.
A medical history was obtained on each volunteer and a complete physical
examination was performed at the outset of the experiment and after the
experiment was completed. Oral temperatures were recorded daily and
weight recorded twice weekly on all subjects during the period of intensive
observation. A physician who was available for sick call, administered
medicine only when definitely indicated in an attempt to avoid all drugs
which might possibly influence the experimental infections.
During the period of intensive observation each entire stool was collected
and the weight, color and consistency of each specimen were recorded. Most
specimens were examined within an hour after collection but in some cases
they were held at 8°C for periods up to 48 hours. All fresh stool specimens
were examined by the direct smear technic unstained in saline and stained
-------
66 WATERBORNE GIARDIASIS/THE DISEASE
with D'Antoni's iodine stain. They were also examined by means of a zinc
sulphate centrifugal flotation concentration technique. The number of
parasites present were counted and the numbers were arbitrarily grouped
into 3 categories of parasite density: light, moderate and heavy.
Sources of Giardia cysts were donors who were found free of bacterial
pathogens and who were followed long enough to be reasonably certain that
they harbored only Giardia. Cysts were processed in a manner resulting in
clean, undamaged parasites. When small numbers of cysts were given they
were counted by means of a micromanipulator in a manner in which we were
absolutely certain of the numbers up to and including 100 cysts. When cysts
were given in doses larger than 100, a suspension of cysts'was counted by
means of a hemocytometer and appropriate dilutions made to give the
estimated desired numbers. All experiments were conducted double blind so
that neither subjects nor technicians examining stools knew who belonged to
the experimental or control groups.
Various kinds of experiments were eventually undertaken but first the
subjects were given accurate known numbers of cysts in gelatin capsules in
order to establish the dose of Giardia infectious for man. Later experiments
were done on the survival of cysts in tap water and experiments were made
under simulated natural conditions to study fly transmission.
Table 1 gives the results of the first experiment in which 6 men received
Giardia with 2 each getting doses of 1, 10 and 100 cysts and 7 men remaining
as controls. All of the men receiving between 10 and 100 cysts became
infected and neither of the 2 receiving a single cyst nor the 7 controls became
infected. The prepatent period is that time in days from when the capsule was
swallowed until the parasites were detected by our stated routine methods of
stool examination. This ranged here from 9 to 15 days.
In the second experiment 3 men given single cysts failed to become
infected. Doses of 10,000, 100,000, 300,000, and 1,000,000 cysts were also
given to study the effect of massive doses. As can be seen in Table 2 all these
latter men became infected. The shortest prepatent period was 6 days, the
longest 11 days. The prepatent period did not seem well correlated to the
dose of cysts given. Four men serving as controls remained negative.
In the third experiment 25 cysts were given to each man as shown in Table
3. This and other experiments were designed for purposes not related or
pertinent to the study of infective dose but are included because of their
bearing on this. This particular experiment concerned the survival of cysts
stored overnight in tap water or in saline. Only one of the ten volunteers
given cysts became infected and none of the nine controls became infected.
The reason for the low rate of infection is not clear particularly in view of the
fact that in other studies two out of three men exposed to 100 Giardia cysts
stored for 16 days in aerated tap water became infected. From prior
experience one would expect a dose of 25 cysts to yield a very high infection
rate. In this experiment, perhaps the failure may have been due to variations
of infectivity of cysts since cysts used in this experiment were from a different
source. This of course is conjecture and we never actually tested this
hypothesis bearing on the infectivity of the parasite.
-------
R. C. Rendtorff
67
.o
E
a
6
S o
ra 5
o
|z
Q.
D
O
LD CD O
ooocnoroco
OLO'sTCNCDOOO
OOLOCOOCM OOOOOOO
^T CM IT)
O O O O
<- »- O O
O
CNOOCNCNCM^t
CN Csl CM
f\| ^
05
"o
C
o
0
-------
68
WATERBORNE GIARDIASIS/THE DISEASE
.0
I
a
•5
03
6
o
~§
§
§ s
c fc
ui
^
£ E
co —
Q- m
a
CO I—* CO ^J~ CN ^vj" CD "sT I • CO CO CO P^- CO CD CD CN CO
LOcocDcor-^cDcocjjcNcococDr^LO CDCDCDCD
ro
O o O o
ooooooooooo
ooooooooooo
o o o o q q q o q q q
o o o" o o o" o o o" o o"
<—<—<— cococooo
o u o
COCDCDOCDCDCDCMOTLn
LOOO
~~ I— CD
CM CO CN LT)
^ co CN r^ co
^ CM CO CO -=t
o
OB
aj
z
U
-------
R. C. Rendtorff
69
.£>
s
o
3
<*:
C
2E'
CD
nc
o CT
Z
Q.
n
o
6
CD
CDCNCslLOCO CO^CDLD
OOOOO OOCMOO OOOO OOO
0
o^ouo
CD »— O -— to
CMCNirO^J-CM
oo o en CM ^r
<— r\i ^ LO LO
CD r~~ ^— a) LO
CO^tOOCMCxI
LO oo o »- •*
LD LO CD CD CO
cDr-~.cnr~~ LDCDCM
C^CNCNCM CxIOslCNI
CD CO r~- O CD CN CO
^t LO LO LO LO CD CD
CD
c
CO
oo
T3
CD
LL-
00
O
o
-------
70 WATERBORNE GIARDIASIS/THE DISEASE
Experiment 4 was similar to the preceeding one in that men received doses
of 25 cysts each as shown in Table 4. One group of 5 men received cysts from
donor H; of this group 3 became positive. Another group of 5 men received
cysts from donor G; of these 2 became positive. Three controls in this
experiment remained negative.
Table 5 sums the results of all the experiments where men were infected
with known numbers of Giardia cysts. This shows that very small numbers of
cysts will produce infections. It seems very likely that in a susceptible host a
single cyst would cause infection. This fact is of extreme importance to this
conference where consideration is given to the possibility that drinking water
containing very small numbers of cysts might be infective to large numbers
of individuals. Certainly large doses are not needed to cause infection.
Before going on into some of the biological characteristics of the
experimental Giardia infections, other of our experiments should be noted.
One such experiment dealt with the housefly, Musca domestica, as a possible
vector of intestinal protozoan cysts. The method of exposing flies to cysts
assured that the laboratory reared flies did indeed ingest the cysts into their
alimentary tracts. In this particular experiment starved flies were exposed to
attractive suspensions of a mixture of E. coli and G. lamblia and
subsequently allowed to attack the food items. Within one hour after
exposure to the flies the foods were fed to the men. The foods were not fed
directly but were mixed with larger samples which were then fed to all of the
men at the same time at their evening meal. Foods involved were various
jellos, puddings, custards and milk and exposures were done during 9 of 11
consecutive days. Numbers of flies involved in the experiment varied from 30
to 38 at each exposure. Of the 16 volunteers who partook of the food
exposed to the flies, none became infected with Giardia lamblia and only a
single volunteer subsequently developed an E. coli infection. One of the
difficulties of this experiment was that the donor F of the G. lamblia cysts
was the same donor whose dosages of 25 cysts given by capsule produced
only a 10% infection rate. We of course were not aware of this at the time and
perhaps the experiment would have been successful had we used cysts from
another donor. Giardia cysts are certainly no larger than E. coli cysts and
one would have to assume that if the housefly can transmit E. coli it could
transmit Giardia. We concluded however, that the housefly was not a very
efficient transmitter of either E. coli or Giardia.
Of particular importance to this seminar are the water transmission
experiments. In 1 experiment a bottle type electrically cooled drinking water
dispenser was used along with disposable paper cups in the men's quarters.
This dispenser had been placed in the quarters prior to the experiments.
Unknown to the volunteers 2000 cysts each of Entamoeba coli and of
Giardia from donor F were placed within the water bottle. The cysts were
added to each of 3 bottles on alternate days in January of 1952. There was no
control over who drank the water, how much they consumed or exactly
when they drank it.
Of 20 men exposed to this water, 4, or 20%, subsequently became infected
with E. coli and none showed infection with Giardia. Again, there is doubt
-------
R. C. Rendtorff
71
o
I i
*
-C
(2
(D ^
!r CD
Q_ CL
O
O
CD
cr
o
>
CM
O
CN
CM CM CD ro
OOO
in
CM
LO
CM
O
(_)
O (J CJ
o u o
COLD'— ro^r^cDCO-^LO r^coi^-
CMCMOOrOCMCMCMCMOJCM CMCOOO
I
o
c
o
Q
O
O
c
o
Q
C
O
(J
-------
72 WATERBORNE GIARDIASIS/THE DISEASE
Table 5 Combined results of all experiment!, comparing the percentage of men
infected with the number of Giardia lamblia cv>ts given. Adapted from reference (1)
No cysts given
1
10
25
100
10.000
100,000
300,000
1,000,000
All doses
Control
No men
exposed
5
2
20
2
3
3
3
2
40
21
No men
infected
0
2
6
2
3
3
3
2
21
0
Per cent
of men
infected
0
100
30
100
100
100
100
100
52 5
0
about the infectivity of the Giardia cysts from donor F. One may presume
their infectivity was low. A fairly high rate of E. coli infections occurred and
there is no reason to doubt from a biological basis that the risk of exposure of
Giardia would not be equal to the risk of exposure to E. coli. Either host
factors or factors inherent in the parasitic cysts must explain the differences
of infectivity rates between these 2 species in this experiment.
In another water experiment using again both E. coli and G. lamblia, we
controlled the variables inherent in the preceeding experiment in a test to
show more directly the effect of storage in water upon cysts. One hundred
cysts of Giardia from donor H were added to 100 ml of aerated tap water
contained in a beaker which was covered with aluminum foil and placed in a
refrigerator at 8°C. Similar doses of cysts of E. coli were also prepared in a
like manner. Tests done for residual chlorine in this water proved negative.
Two control men received water held under similar situations containing no
cysts. Two volunteers were given the water with cysts on the day the beakers
were prepared. Three volunteers each received water after 1, 2, 4, 8, and 16
days of storage. The water was stirred in the beaker prior to the men drinking
it. On each day that the cysts were given to a particular subject, all of the
remaining subjects received plain water. The controls remained negative
throughout the experiment. Of the 17 men receiving the Giardia cysts all but
6 (64.7%) became infected, while 12 (70.6%) developed E. coli infections. The
prepatent period ranged from 9 to 22 days and averaged 13.1 days for
Giardia.
In this particular group of volunteers, we were able to re-expose those who
were refractory to infection to cysts of Giardia. These men had been followed
for a considerable period of time after the original exposures to convince us
that they were truly negative. They were then given 100 cysts of Giardia in
100 ml. of sterile tap water and 2 of 6 men thus re-exposed did show parasites
subsequently.
It would appear that where the dose is sufficient, that water is a very
efficient vehicle for the transmission of Giardia cysts and these will persist in
-------
R. C. Rendtorff 73
cool water for periods of at least 16 days. One could presume from the
persistent viability of cysts, that they must be infectious for considerably
longer than 16 days at least when held at 8°C.
There are many unanswered questions about the biology of Giardialhal
we did not explore in our experiments, which are interesting to speculate
upon. It has been mentioned that two of the six men re-exposed to Giardia
came down with infection. But even more interesting are the four who did
not come down with Giardia. Would this indicate that these men had a
resistance to the organism? It would have been interesting to subject them to
still higher doses of cysts. Subsequent epidemiological studies have implied
that resistance may be an important factor in human Giardia infections.
The biology of experimentally produced Giardia infections in man has
raised some other pertinent points. First, and most important, is the
spontaneous disappearance of the infections. This phenomenon is best
presented in graphic form as shown in Fig. 1 and 2, where the solid areas
above the line indicate light, moderate, or heavy presence of parasites in
stools examined on that day and the shaded areas below the line represent a
negative stool. One can argue that the level of parasites' density in the stools
is meaningless but the fact that stools became negative and stayed as such is
significant.
Among 25 men who were given known numbers of cysts in either the
capsule experiments or in the drinking water experiments, 21 (84%),
exhibited a spontaneous cure and only 4 showed persisting and apparently
chronic infections.
We purged 19 men, who were still available to us, 70 to 90 days after
exposure and examined the 2 to 5 stools each deposited on the day of the
purge. None showed parasites. One man was screened intermittently over a
year's time and never again became positive.
Among those infections showing spontaneous disappearances, the
duration of the parasites in the stool ranged from 5 to 41 days, with a mean of
18.7 days and median of 13 days. These results were obtained from the men
of the first two experiments, wherein men in Experiment 1 received cysts
from one donor; the men in Experiment 2 received cysts from a second
donor; and, the men in Experiment 3, on survival of cysts in water, received
cysts from a third donor (see Fig. 3 and 4). In this third experiment, the
parasites ranged from 25 to 41 days with a mean of 33.9 days and a median of
30 days. The volunteers of the second and third groups differed significantly
in the length of duration of their infection. Also, among the men of the last
group with the longer period of patency, the pattern of these infections
commonly showed a bimodal distribution. This is evident in at least 7 of the
11 men studied.
It should be pointed out that in the apparent persistent infections, there
were rather long periods when stools alternated between positive and
negative. Volunteers, identified as numbers 102 and 119, established this
pattern nicely, wherein periods of negative stools lasted up to 13 days after
which time they again became positive.
-------
74
WATERBORNE GIARDIASIS/THE DISEASE
E-f
I-
B
5 u.
0. >
5? £
r
i
Jt
•
II
o
1.
-------
R. C. Rendtorff
75
X
u
"5
•
11
-------
76
WATERBORNE GIARDIASIS/THE DISEASE
I
5
2
•a-fe
-5
1,
-------
K. C. Rendtorff
77
5
a*
*s
ll
o
£ i
-------
78 WATERBORNE GIARDIASIS/THE DISEASE
Finally, there appears to be little or no association between the dose of
cysts given and the pattern of the infection in terms of parasite shedding in
the stools. Higher doses of cysts did not lead to permanent infections.
Lessons to be learned so far from these experiments are: a large
percentage of infections disappear spontaneously; there may be biological
strain differences in the parasite as judged by the different behavior of
infections depending upon the donor source; dose is not related to
persistence of infection; and, finally, what has already been well recognized
in other intestinal protozoa infections, that Giardia also may have long
periods where stools are negative in a person who is truly positive.
Some of these characteristics of the Giardia infections have not been
properly considered in subsequent epidemiological and clinical studies done
by others. For example, 1 am not aware of any double blind clinical trial with
drugs recommended for giardiasis and in view of the great numbers of our
cases that clear themselves without any treatment such trials in my opinion
are definitely indicated.
Finally a word in regard to the clinical picture of the Giardia infection.
This is best presented by reiterating the original from our second paper(l):
"In a number of the Giardia infections a definite change in the stool pattern of the
individuals was noticed. This pattern is the correlation of frequency with weight and with
the consistency of the stool. It is rather constant for each individual but vanes widely
between individuals. Thus, in the Giardia infections of experiments 1 and 2 there were 4
cases of a marked change in the stool pattern, 5 cases of a moderate to questionable change,
and 6 cases where no change was discernible. These changes were always from the normal
pattern to one showing more frequent and looser stools and usually stools of greater
weight. These changes usually began abruptly with the first appearance of parasites and
lasted only 2 to 4 days In some cases they preceded or followed the first appearance of
parasites by a day or two. The change in stool pattern was always transient, lasting at the
most but a few days.
Changes in the stool pattern were sometimes very subtle and required actual plotting of
the data in graph form to be recognizable. In other instances, such as in volunteer No. 23,
these changes were very marked In this subject the change was noticeable on the day
parasites first appeared in the stool and it lasted for 4 days. During this 4-day period 9
stools were passed which averaged 189.7 grams in weight. Of these stools 77.7 per cent were
loose, 11.1 per cent were soft, and 11.1 per cent were semiformed. Of a total of 89 stools
collected before and after this period of change, the average weight was 127.2 grams and 1.2
per cent were loose, 11.6 per cent soft, 51.2 per cent semiformed, and 36 0 percent formed.
Changes in the stool pattern were rarely noticed by the individual and when they were
noticed there was no complaint other than the fact that the movements were looser than
usual.
In no instance did any clinical illness or subjective complaint occur among the subjects
which could be attributed with certainty or bear close relationship to Giardia infections.
Gastrointestinal complaints did occur among the infected individuals but these bore no
distinct relationship to the Giardia infections and also occurred among the non-infected
men. One possible exception to this statement was a single case of a man who complained
of nausea, vomiting, diarrhea, and abdominal discomfort associated with the first
appearance of parasites. However, this subject gave a history of similar episodes prior to
his experimental infection and was observed to have such episodes long after the Giardia
infection had disappeared spontaneously. Other features of his illness deemed it highly
improbable that the parasite had any causal relationships to his episodes of gastrointestinal
illness."
-------
R. C. Rendtorff 79
In closing I want to reiterate my thanks to my research staff and others of the
Public Health Service and the Federal Prison Service who made these
studies possible. I want also to especially thank the men who volunteered
and cooperated so admirably in this study.
REFERENCES
1. Rendtorff, R. C. 1954 The experimental transmission of human intestinal protozoan
parasites. 11. Giardw lamblia cysts given in capsules. Amer. J. Hyg. 59: 209-220.
2. Rendtorff, R. C. and C.J. Holt. 1954 The experimental transmission of human intestinal
protozoan parasites 111. Attempts to transmit Endamoeba colt and Giardia lamblia cysts
by flies. Amer J. Hyg. 60: 320-326.
3. Rendtorff, R. C. and C. J. Holt. 1954. The experimental transmission of human intestinal
proto/oan parasites. IV. Attempts to transmit Endamoeba coli and Giardia lamblia by
water. Amer. J Hyg. 60 327-328
Discussion
R. OWEN: Since the question exists as to whether it is the parasite or the
host that causes those cases which are symptomatic, I want to clarify the
source of the cysts that you used. Were any of the people from whom the
cysts were isolated people who had been sick?
R. RENDTORFF: They were normal individuals in the prison who
volunteered as donors and received benefits for the volunteering. They were
all tested, of course, for bacterial pathogens. I am sure they were not passing
any pathogens along with the Giardia. The difference in the Giardia sources
was discovered in the experiments and not recognized microscopically or
any other way.
R. OWEN: It is conceivable that this was a strain that was not virulent, if
such a strain exists, although it may be hard to determine.
R. RENDTORFF: We used many different donors in the experiments.
R. OWEN: There is an increasing awareness that Giardia can be
transmitted as a venereal disease, as can any of the other enteric pathogens.
Many studies indicate that there is a high degree of venereal transmission
within prison populations. Recrudescence of stool excretion in most prison
populations would be difficult to differentiate from venereal transmission or
retransmission of Giardia. The exception would be those prisoners who were
in solitary cells.
R. RENDTORFF: Are you referring to the variations in negative and
positive findings of parasites?
R. OWEN: Yes.
R. RENDTORFF: This is true of all the other parasites we used. It has
been experimentally known long before our work with monkeys using
Entamoeba coli and other intestinal protozoa. It has no relationship.
The experiments we have been discussing were performed in a model
prison institution, without bars, within the federal prison system. I do not
know if homosexuality was prevalent in that prison, as it is in the crowded
prisons today. Sexual mores were different 30 years ago. This activity was at
a minimum and unknown to me, and I really do not believe it ever occurred
in our experimental group. They were under very close observation and were
intent on achieving their meritorious good time for volunteering for the
experiments.
M. WOLFE: These were very unique studies and have become historical
in terms of volunteer infections of Giardia. You referred to use of clean,
-------
80 WATERBORNE GIARDIASIS/THE DISEASE
undamaged cysts and their survivability in stools kept for 24 to 48 hours
before cysts were separated. Were you able to test for viability of these cysts?
R. RENDTORFF: No. To my knowledge, there is no test for viability of
cysts that can be performed without killing the cyst.
M. WOLFE: I think there may be some vital stain tests which can be
performed to test viability.
R. RENDTORFF: 1 do not think you would want to use those cysts when
they are to be given to men. 1 really do not understand the essence of your
question.
M. WOLFE: The essence of my question is that maybe these cysts were
not all alive when you gave them to the men. M ay be that is why they had self-
limited infections with short-lived durations.
R. RENDTORFF: They were not all self-limited. Some lasted for long
periods of time.
M. WOLFE: Perhaps it took only one viable cyst out of twenty-five, or as
many as a hundred viable cysts in one patient to give him a persisting
infection. Others may not have taken in any viable cysts because of the
circumstances under which your experiments were performed.
R. RENDTORFF: I think there is a biological difference in the donors
which can determine low and high infectivity of cysts. This particular donor,
as a source, was apparently low in infectivity. However, in the others, I do
not see how this could possibly be the case. The cysts were of high viability
and infectivity because very low numbers of them produced infections that
lasted very long times, and the numbers of parasites found were enormous.
This was not just a low grade infection.
As few as 10 cysts produced infection; that is a very small number. In our
E. coli and other tests, 1 cyst produced good, solid infections. I cannot see the
question of viability in regard to this except in the case of that one donor
where there is some question about his cysts. I believe that strain differences
cause different infection rates.
W. JAKUBOWSK.I: As a result of your experiments with the prisoners,
along with other information you have obtained, is there any question in
your mind regarding the pathogenicity of the Giardia organism?
R. RENDTORFF: The results of our prison experiments convince me
that the parasite that these men had was non-pathogenic. That does not
mean 1 do not think the parasite could cause disease. I am right now working
on a very fascinating disease, Legionnaires' disease, which infects only
certain individuals from groups of people. Why only certain individuals are
infected with Legionnaires' disease or become infected by Giardia while
others do not, is unknown to me.
I think the infectivity rate is a host factor we have yet to define. We know
that the host factor is true for almost all diseases. There are some diseases
that attack children; others attack adults. The evidence of immunity is an
extremely interesting topic which indicates that this may be an extremely
important factor in the spread of giardiasis. It is a topic that we thought of
years ago; however, we did not refer to it in any of our publications.
We were convinced years ago that Giardia is spread very rapidly among
people. The evidence today, based on outbreaks of Giardia in orphanages
and places where young children are crowded together, indicates that
Giardia spreads like some viruses and other diseases found in young
children. It is not surprising that Giardia is transmitted from person to
-------
/?. C Rendtorff 81
person. It is important to understand this relation of the transmission of
Giardia in waterborne outbreaks among human populations. A great deal of
study needs to be performed in this area.
W. JAK.UBOWSKI: Independent of host differences, it is also being
considered that there might be virulence factors in different strains of the
organism. Would you care to comment on the possibility that we might be
dealing with more virulent strains of the organism resulting in an increase in
the number of outbreaks of the disease?
R. RENDTORFF: I do not know why a virulent strain would persist in
water rather than a non-virulent one. Virulence is a factor of the parasite, not
the host. The evidence for virulence, in our experiments, is that cysts from
one donor did not cause as great a number of infections as we expected.
Certain strains of E. histolyticavary in virulence remarkably as evidenced by
the severe outbreak at the Chicago World's Fair. Other strains apparently
are less infective, and there is no reason to believe that Giardia may not have
the same variance.
-------
82
Giardiasis in the Mouse:
Clues to Host Immune Mechanisms
David P. Stevens
Department of Medicine,
University Hospitals of Cleveland
and
Case Western Reserve University School of Medicine, Cleveland, Ohio.
ABSTRACT
The role of host immunity in human giardiasis is poorly understood. Prospective studies
in a murine model of" this infection have been facilitated by the development of
quantification techniques. Following primary infection, mice are resistant to reinfection.
Protective immunity is not achieved in congenitally athymic mice. Protection against
infection is transferred passively from immune mothers to suckling mice via breast milk
and is associated with specific unti-Giardia IgA. Immune lactatmg females lose their
intestinal resistance to Giardia but regain it after weaning. Studies in mice may lead to
appropriate investigations in human giardiasis designed to achieve immunoprophylaxis
for this infection.
OBSERVATIONS WHICH SUGGEST IMMUNITY IN HUMAN
GIARDIASIS
The influence of host immunity on the clinical course of giardiasis remains
undefined. Giardiasis may range from a short, clinically innocuous infection
to a prolonged, debilitating illness(l). It may resist repeated
chernotherapeutic efforts to eradicate it, or it may be short and
spontaneously self-limited (2). There remains no explanation for the striking
variability of clinical illness in this infection. Nevertheless, protective
immunity is suggested by the lower prevalence of cysts in stools of long-term
residents in areas of high endemicity such as Aspen, compared to brief
visitors to these locations(3). Epidemic illness in Aspen during an outbreak
was present in 11% of vacationing skiers(4) compared to 2% in permanent
residents(S) studied in a separate survey.
One of the observations in humans which suggests a role for the immune
response in host resistance to Giardia is the association of giardiasis with
immunodeficiency. As many as 80% of persons with variable immune
deficiency syndrome and diarrhea may have Giardia in the stool(6). In this
setting, giardiasis is usually accompanied by marked morphological changes
in jejunal histology.
(This work was supported by a grant-in-aid from the Rockefeller Foundation and Grant No.
Al-15351 from the U. S. Public Health Service.)
-------
D.P.Stevens 83
Preliminary studies into the demonstration of specific immunity to
giardiasis have been inconclusive. Serum immunoglobulin deficiency was
not observed in 70 otherwise healthy travelers with giardiasis(7). In another
report of preliminary studies, circulating antibodies of the IgG class were
detected by indirect immunofluorescent techniques. This work was
apparently handicapped by difficulties in obtaining adequate purified
specimens of cysts or trophozoites for antigen source(8). While one group
reported a reduction of duodenal IgA, with normal circulating IgA, in 10
subjects who had recovered from giardiasis(9), their collection and assay
methods were later criticized(lO) and the observations have not been
confirmed by subsequent investigators(l 1).
PROSPECTIVE STUDIES IN MURINE GIARDIASIS
Basic Model
Why the Mouse?
The mouse was chosen for prospective studies of giardiasis(12,13) for
practical reasons. Firstly, there are well-defined inbred murine strains on
which to perform immunological studies. Secondly, many basic
immunological investigations have been performed using the mouse.
Thirdly, considerable immunological technology on the mouse is available
on which to base investigations of its immune response to this protozoan.
Finally, the mouse is small, easily handled, and economical.
Quantification Techniques
Methods of quantification of infection were developed to measure
intensity and duration of infection. Quantification is not generally
considered a meaningful measurement in protozoan infection as it is in
helminth infections(14). However, the characteristics of giardiasis which we
have observed in mice makes quantification both appropriate and, indeed,
meaningful.
Isolation and quantification of cysts from the stool employs techniques
developed in this laboratory based on a method originally employed for
ultrastructural studies(15). Trophozoites are counted by flushing free
trophozoites from the entire small intestine by irrigation with a known
volume of physiologic saline followed by enumeration in a hemacytometer
counting chamber. Non-fasted mice are isolated in individual cages for two
hours. All stools obtained during this period are broken up in 3 ml of tap
water and isolated on 2.5 ml of 1 M sucrose (specific gravity 1.11) in a 75x12
mm plastic tube and centrifuged at 400 X g for 15 minutes at 20° C. Cysts are
removed from the water-sucrose interface by careful aspiration, washed by
resuspension in 4 ml of physiologic saline and centrifuged at 600 X g for 10
min. The cysts are resuspended in saline and counted in the hemacytometer
chamber. This technique yields a recovery rate of 65% of cysts in the original
specimen consistently over a range of 10,000 to 100,000 cysts(12).
Giardia muris Adapted from the Hamster
The Giardia strain employed in these studies was originally isolated from
naturally infected golden hamsters in the Case Western Reserve School of
-------
84 WATERBORNE GIARDIASIS/THE DISEASE
Medicine Animal Facility. The strain was recovered by direct peroral
inoculation of Giardia-infected stool into non-immune CF-1 mice.
Subsequent passage of this strain in stool extracts was carried out from
mouse to mouse. This strain has now been passaged over 50 times in
subsequent generations of CF-1 mice. Since its adaptation to mice in 1975,
the strain has been distributed to laboratories for studies elsewhere in the
United States and Australia. Subsequent attempts to adapt other hamster
Giardia strains to mice have been uniformly unsuccessful in our laboratory
(D.P. Stevens and J. Houser, unpublished) raising questions of why this
particular isolate passes readily. Ultrastructural studies, however, have
demonstrated unique morphologic aspects of this strain (P. Nemanic, R.
Owen, and D. P. Stevens, unpublished) and may shed light on criteria for
strain specificity in various Giardia species.
Susceptible Murine Hosts
Numerous mouse strains have been employed, but the bulk of the studies
to be described were performed in female CF-1 strain albino mice (Carworth
Farms Division, Charles River, Portage, Michigan). Susceptible inbred
strains have included Ajax, BALB/c, nude and others (13). Careful selection
of supplier and breeding stock is important since many colonies are naturally
infected with Giardia. CF-1 mice obtained from Charles River and inbred
strains obtained from Jackson Laboratories, Bar Harbor, Maine, have been
uniformly free of previous Giardia infection.
CF-1 mice are obtained at 12 to 14 g weight (5 to 6 weeks of age) and are
immediately screened for the presence of cysts in stools. Freedom from
infection at this age insures freedom from previous infection, since studies in
newborn animals indicate that susceptible suckling animals inoculated as
early as 2 days of age will remain infected when tested at 6 weeks of age (D.P.
Stevens, unpublished). Freedom from previous infection is also assured by
demonstration of susceptibility to infection of randomly selected animals
from each shipment.
Murine Model: General Characterization
When simultaneous measurements are made of trophozoites in small
bowel and cyst excretion in stool, the curves which describe these counts
over time are parallel (13). Therefore, the level of cyst excretion inthemouse
model may be interpreted to reflect intensity of trophozoite burden in the
small intestine. This observation, which reflects a difference from human
infection where numerous trophozoites may be found in the small bowel
when cysts are undetectable in stool lends itself to longitudinal studies of the
same mice. Nevertheless, while resolution of infection in this model occurs 6
to 8 weeks after inoculation when the sucrose flotation technique is used for
cyst quantification, absolute elimination of infection is probably not
achieved. Animals in our laboratory will excrete small numbers of cysts on
isolated occasions long after clearance of infection. Such animals, however,
are resistant to the development of measurable infection after reinoculation.
Moreover, previously infected animals in which cysts are undetectable in
stools can still be found to harbor small numbers of trophozoites in the
-------
D. P. Stevens 85
intestine if the entire small intestine is flushed with large volumes of saline
and concentration techniques are then employed (D.P. Stevens and D.M.
Frank, unpublished).
The studies described using this model were conducted using oral inocula
in the range of 500 to 1,000 cysts. The Infectious Dose50 (ID5o) in our
laboratory for this strain is, however, between 25 and 50 cysts (D.P. Stevens
and D.M. Frank, unpublished).
Comparisons of Murine and Human Giardiasis
An important question remains: does infection in the mouse result in
disease analogous to that observed in man? Jejunal morphology, while less
strikingly altered than that seen in severe giardiasis in man, is rendered
abnormal by Giardia infection in the mouse as determined by villusicrypt
ratios (12,13). The villusicrypt ratio is 3.1±0.1 in jejunum of uninfected
control mice in biopsies obtained 10 cm from the gastroduodenal junction.
The same measurement made at the peak of infection following inoculation
of 10,000 cysts shows a villus:crypt ratio of 2.1 which returns to normal after
resolution of the infection. Furthermore, weight gain is retarded in infected
animals compared to control animals. Mean weight gain for 16 g mice was
8.25±0.4 g in control animals while that for animals inoculated with 100,
1,000 and 10,000 cysts was 6.88±0.3, 6.13±0.7, and 5.0010.5 g respectively
over a 28 day period. These figures are significantly different from
uninfected animals for all groups (p<0.02). Although G. muris infection in
CF-1 mice results in moderate jejunal morphologic changes and retarded
weight gain, infection in mice appears clinically milder than the most severe
form observed in man and appears to be more uniform. Nevertheless, the
changes are sufficiently similar to suggest that the study of this model will be
relevant to an understanding of the interaction of host and parasite in
humans.
IMMUNE STUDIES IN THE MOUSE
Demonstration of Immunity
We established in CF-1 mice that resistance to reinfection developed after
resolution of the primary Giardia infection(16). Sixty-four previously
uninfected mice were divided into two groups, the first of which was
inoculated on day 0 with 1,000 G. muris cysts orally while the other group
served as uninoculated controls. At 6, 12 and 18 weeks animals were selected
by lot from each group and challenged with 1,000 G. muris cysts orally.
Animals which previously had been infected demonstrated resistance to
reinfection at each of the 3 time periods. The control animals went through
the usual expected pattern of peak excretion at 7 to 14 days followed by
gradual resolution at 6 to 8 weeks as repeatedly seen in primary infections in
other mice. The presence of resistance for up to 18 weeks in these animals
and up to 24 weeks in animals used in other experiments, as well as clearance
of infection following primary infection in all animals, indicates that
resistance to murine giardiasis is acquired after primary infection and
probably reflects immunity against this protozoan.
-------
86 WATERBORNE GIARDIASIS/THE DISEASE
Immunity is Thymus-Dependent
All inbred strains of mice do not resist reinfection uniformly. We have
observed this resistance in Ajax and BALB/c strains and the pattern of
infection is similar in these inbred strains to that seen in CF-1 animals. When
congenitally athymic (nude) mice were tested for resistance to reinfection,
however, typical resistance did not develop (17). Groups of 14 nude mice and
14 heterozygous thymus-intact animals were inoculated orally with either
1,000 G. muris cysts, or in the case of controls, a similarly processed but G.
mwra-free stool extract. Cyst excretion was measured twice weekly. Cyst
excretion in nude mice rose to peak levels at day 18 and persisted at high
levels until day 46 and then gradually returned to reduced but persistent
levels. In contrast, maximum cyst excretion occurred in thymus-intact
controls on days 7 to 14 and returned to undetectable levels by 42 days after
inoculation, a pattern identical to that seen in other immunologically-intact
strains.
Uninoculated animals in both groups remained free of Giardia infection
throughout the 130 days duration of the experiment. Wasting or fatal disease
was not observed in the athymic animals as has been described by others in
naturally infected(18) or prospectively infected nude mice(19). The
explanation for the differences observed in clinical illness in athymic mice in
this laboratory and those of others, may be associated with differences in
nude strains or may be a reflection of synergistic infection by other infectious
agents in the animals that became ill. It is recognized that nude mice are
extremely susceptible to a wide variety of infections(lS). We are confident
that additional pathogens were not introduced in our inocula since our
control animals were inoculated with identically processed but Giardia-free
stool extracts and they remained free of illness as reflected by their
maintenance of weight and absence of mortality.
At least one thymus-intact inbred strain has been studied in which Giardia
infection does not spontaneously resolve(19). This suggests that
explanations other than thymus-dependency may be involved in the
development of immunity against Giardia and may include the congenital
absence of an immune response to Giardia antigens. Moreover, preliminary
data derived from studies of FI hybrids of resistant and non-resistant strains
suggests that the ability to develop resistance to Giardia in the mouse may be
a dominant genetic trait(19).
The precise role which the thymus-dependent immune system plays in
resistance to Giardia remains speculative. Little morphologic evidence is
available that aatual cellular immune mechanisms are active in the
resolution of Giardia infection. Increased numbers of lymphocytes are
observed in the lamina propria of Giardia-'mtected mice(20). Demonstration
of juxtaposition of intraluminal lymphocytes and Giardia trophozoites in
the jejunum of infected mice by means of scanning electron microscopic
techniques (R. Owen, P. Nemanic, and D.P. Stevens, submitted for
publication), while provocative, does not clarify the mechanism of the
lymphocyte-trophozoite interaction.
-------
D. P. Stevens 87
A Role for Secretory IgA
Because the elaboration of secretory IgA is a thymus-dependent process, it
is reasonable to suspect that this local humoral immune mechanism may be
active in the immune response to Giardia. To approach this hypothesis,
studies were performed to test the ability of immune mothers to protect
G/arc/fa-inoculated offspring from infection(20). Accordingly, female CF-1
mice were divided into GYflraYa-inoculated and uninoculated groups. All
were observed weekly by quantification of cyst excretion to document the
establishment and resolution of infection in the experimental group and the
absence of infection in the control group. After resolution of primary
infection, resistance to reinfection was demonstrated in the immune group
by re-challenge with 1,000 cysts orally. The immune and non-immune
groups were then bred to Giardia-freeCF-l males. Within 1 week of delivery
of the resulting litters, the offspring of both groups were challenged orally
with 500 G. muris cysts each. Ten days later the offspring were sacrificed and
the intestine of each suckling mouse flushed with 5 ml of physiologic saline.
The trophozoites obtained from each animal were quantified in a
hemacytometer counting chamber. Infection occurred in all 27 mice suckled
to non-immune mothers while it was absent in 30 of 31 animals suckled to
immune mothers.
The possibility remained that protection might be transferred
transplacentally via IgG. Therefore similar experiments were performed
whereby litters from immune mothers were inoculated with 500 G. muris
cysts and then fostered on non-immune wet nurses. Similarly, offspring from
non-immune mothers were fostered on immune wet nurses. When
trophozoites were examined 10 days after inoculation in these offspring,
infection was undetectable in all animals suckled to immune mothers while it
was present in 12 of 14 animals suckled to non-immune animals. That
protection was absent when animals were no longer suckling, was
established by demonstration of infection after inoculation of weaned
animals regardless of the immune status of the animal on which they were
previously breast-fed. Preliminary studies using indirect
immunofluorescence studies which employed purified G. muris trophozoites
as antigen indicate that such protection is associated with the presence of
specific anti-Giardia IgA in breast milk of lactating immune CF-1 females (J.
Andrews and D.P. Stevens, unpublished). These data demonstrate that in
the mouse resistance to Giardia is passively transferred in maternal milk and,
moreover, suggest that the secretory IgA system may be involved in host
immunity which is developed against this organism.
Loss of Maternal Intestinal Resistance During Lactation
Previous studies have emphasized the role of the gut-associated lymphoid
tissue which provides the lymphoid population that homes to the mammary
glands and proliferates to lead to secretion of IgA in breast milk(22). We
studied the effect of lactation on maternal intestinal resistance to Giardia at
the time of lactation(21). Immune female CF-1 mice that had undergone
primary infection 5 months previously were bred to Giardia-tree CF-1
-------
88 WATERBORNE GIARDIASIS/THE DISEASE
males, and maternal cyst excretion was measured twice weekly. While stool
cyst excretion had remained absent for the 14 weeks following resolution of
primary infection, it spontaneously commenced at 5 to 8 days before the
onset of lactation. Maternal cyst excretion in these "immune" mothers rose
to levels comparable to those seen during primary infection and persisted as
long as lactation persisted. Immediately upon cessation of lactation by
removal from the mothers of the suckling litters, whether at 5, 10, 15 or 21
days after birth, maternal fecal cyst excretion returned to undetectable levels
in the mothers within 7 days. These observations suggest that maternal
intestinal resistance to Giardia in immune females is lost during lactation.
This may be secondary to migration of gut-associated lymphoid tissue to
mammary sites, although mechanisms other than trafficking of these
lymphocyte populations cannot be ruled out by these data.
FUTURE PROSPECTS
Current studies are underway in this and other laboratories using the
murine model of giardiasis to better understand Giardia antigens and the
detailed mechanisms of the immune response which they elicit. Because the
natural history of this protozoan infection and the host immune response are
so poorly defined in the human, the use of an animal model lends itself to a
better understanding of these aspects of this infection. Fortuitously, the
precise immune status of the host can be defined in prospective observations
in the mouse. It is therefore hoped that such studies will lead to experiments
and observations in human subjects which will be meaningful and lead
ultimately to the practical goal of immunoprophylaxis in giardiasis.
REFERENCES
1. Hoskins, L. C., S. J. Winawer, S. A. Broitman, L. S. Gottlieb, and N. Zamcheck. 1967.
Clinical giardiasis and intestinal malabsorption. Gastroenterology 53' 265-279.
2. Webster, B. H. 1958. Human infection in Giardia lambha. An analysis of thirty-two cases.
Am. J Digest. Dis. 3.64-71.
3. Shultz, M. G. 1975. Giardiasis. JAMA 233:1383-1384.
4. Moore, G. T., W. M. Cross, D. McGuire, C. S. Mollohan, N. N. Gleason, G. R. Healy and
L. H. Newton. 1969. Epidemic giardiasis at a ski resort. N. Engl. J. Med. 281:402-407.
5. Gleason, N. N., M. S. Horwitz, L. H. Newton and G. T. Moore. 1970. A stool survey for
enteric organisms in Aspen, Colorado. J. Trop. Med. Hyg. 19:480-484.
6. Ament, M. E., H. D. Ochs and S. D. Davis. 1973. Structure and function of the
gastrointestinal tract in primary immunodeficiency syndromes. A study of 39 patients.
Medicine 52:227-248.
7. Babb, R. R., O. C. Peck, and F. G. Vescia. 1971. Giardiasis: A cause of traveler's diarrhea.
JAMA 217:1359-1361.
8. Ridley, M. J. and D. S. Ridley. 1976. Serum antibodies and jejunal histology in giardiasis
associated with malabsorption J. Chn. Pathol. 29:30-34
9. Zinneman, H. H. and A. P. Kaplan. 1972. The association of giardiasis with reduced
intestinal secretory immunoglobulin A Am. J. Dig. Dis. 17:793-797.
10. McClelland, D. B. L., R. R. G. Warwick, and D. J. C. Shearman. 1973 IgA concentration.
Am. J. Dig. Dis. 18:347-348.
11. Jones, E. G. and W. R. Brown. 1974. Serum and intestinal fluid immunoglobulin in
patients with giardiasis Am. J. Dig. Dis. 19:791-796.
12. Roberts-Thomson, I. C., D. P. Stevens, A. A. F. Mahmoud and K. S. Warren. 1976.
Giardiasis in the mouse: an animal model. Gastroenterology 71:57-61.
13. Stevens, D. P. and I. C. Roberts-Thomson. 1978. Animal model of human disease:
giardiasis. Am. J. Pathol. 90:529-532.
-------
D.P.Stevens 89
14. Stevens, D. P. 1978. Quantitative techniques Clinics in Gastroenterology 7:231-238.
15. Sheffield, H. G. and B. Bjorvatn. 1974 infrastructure of the cyst of Giardta lamblia
Presented at the 23rd Annual Meeting of the American Society of Tropical Medicine and
Hygiene, Honolulu, Hawaii.
16. Roberts-Thomson, I. C., D. P. Stevens, A. A. F. Mahmoud and K. S. Warren. 1976.
Acquired resistance to infection in an animal model of giardiasis J. Immunol. 117:2036-
2037
17. Stevens, D.P., D.M. Frank and A.A.F. Mahmoud. 1978. Thymus dependency of host
resistance to Gmrdia muris infection: Studies in nude mice. J Immunol. 120:680-682.
18. Boorman, G.A., P.H.C. Lina, C. Zurcher and H.T.M. Nieuwerkerk. 1973. Hexamita and
Giardia as a cause of mortality in congemtally thymus-less (nude) mice Clm. Exp.
Immunol. 15:623-627.
19 Roberts-Thomson, I.C., and G.F. Mitchell. 1978. Giardiasis in mice. I. Prolonged
infections in certain mouse strains and hypothymic (nude) mice. Gastroenterology 75:42-
46.
20. MacDonald, T.T., and A. Ferguson. 1978. Small intestinal epithelial cell kinetics and
protozoal infection in mice. Gastroenterology 74.496-500.
21. Stevens, D.P. and D.M. Frank. 1978. Local immunity in murme giardiasis: Is milk
protective at the expense of maternal gut? Trans. Assoc. Am. Phys. 91:268-272.
22. Roux, M.E., M. McWilliams, J.M. Phillips-Quagliata, P. Weisz-Carrington and M.E.
Lamm. 1977. Origin of IgA secretory plasma cells in the mammary gland J. Exp. Med.
146.1311-1322.
Discussion
S. ERLANDSEN: Did you use the Giardia muris or Giardia lamblia
species in your experiments?
D. STEVENS: We are calling the strain which came from a hamster G.
muris. We subsequently looked at cysts from other hamsters, but we were
unsuccessful in infecting mice with these strains. I am uncertain if this is a
characteristic of this particular strain from this particular hamster.
Some of the morphologic observations that R. Owen has made on the
strain which he will discuss tomorrow night may be relevant, such as the
presence of the endosymbiont in some of our strains.
SPEAKER: The data on the offspring were extremely interesting. The
intestine is extremely different up through about three or four weeks of age,
and so you are seeing something which is perhaps a little unique in some
cases because that intestine is still capable of absorbing maternal
immunoglobulins.
Did you look at trophozoites in the intestine, since we are looking at cyst
production here, or cyst shedding? Were they gone at the same time? Would
this indicate perhaps a control over cyst production versus trophozoite
numbers or perhaps both?
D. STEVENS: When adults stop shedding cysts and the cysts are
undetectable in a two-hour stool collection, looking for trophozoites in the
intestine yields about a 65 per cent recovery rate. Consequently, the absence
of cyst excretion does not represent total immunity. Moreover, in maternal
infections, there were spontaneous infections which developed presumably
from trophozoites, present in the intestine. As in most protozoan infections,
the immunity, if it occurs—and we think it does—is certainly not absolute in
this animal.
T. NASH: Is this relevant to the immunodeficient patients who are
gammaglobulinemic? Is the abnormality the same or different?
D. STEVENS: The nudes obviously have a different congenitial
abnormality; however, some similarities may exist. But truly, it is not the
-------
90 WATERBORNK GIARDIASIS/THE DISEASE
same disease. The nudes have a different disease than the immunodeficient
patients that we see in that the former have a cellular immune deficiency and
the latter a humoral deficiency.
T. NASH: 1 am not sure that the G cell dependent has been properly
examined.
D. STEVENS: We did not set out to establish a model of
immunodeficiency. We were looking at the question of thymus dependency
of immunity in giardiasis.
G. HULDT: Did you put the thymus back into your nudes and measure
that effect on your production?
D. STEVENS: No. Dr. Roberts-Thomson has restored nudes
intravenously using thymocytes and has shown partial restoration of
resistance.
G. HULDT: Did you just look for Giardia in the mice or did you look for
antibodies?
D. STEVENS: The results of antibody testing are so preliminary that we
would be reluctant to discuss them in this forum; we may be better prepared
to discuss it later. Not bnly does the median body of the Giardia muris differ
from that of lamblia and other species, but Giardia muris is the one strain in
nature that Meyer or others have been unable to cultivate axenically, thus
making purified antigen very difficult to obtain. However, using some work
J. Andrews in our lab has done, we are now at the stage where we feel we may
be dealing with some purified antigen preparations, and in
immunofluorescent studies, we are pursuing experiments in the mouse
similar to those discussed with G. Healy this morning.
-------
91
SESSION III - EPIDEMIOLOGY
Chairman - Myron G. Schultz
Center For Disease Control, Atlanta, Georgia
The Presence and Absence of Giardia
lamblia in Studies on Parasite
Prevalence in the U.S.A.
G. R. Healy
Animal Reservoirs and Cross-Species
Transmission of Giardia
R. B. Davies, C. P. Hibler
Waterborne Outbreaks of Giardiasis
G. F. Craun
Waterborne Giardiasis
D.Juranek
Water Supply Problems Associated
with a Waterborne Outbreak
of Giardiasis
E. C. Lippy
An Outbreak of Gastroenteritis Associated
with Giardia lamblia
L. Veazie, I. Brownlee and H. J. Sears
-------
92
The Presence and Absence of Giardia lamblia in
Studies on
Parasite Prevalence in the U.S.A.
George R. Healy
General Parasitology Branch, Bureau of Laboratories, Center for Disease
Control, Public Health Service, U. S. Department of Health, Education,
and Welfare, Atlanta, Georgia
ABSTRACT
Giardia lamb/ia is currently agreed by most people to be a pathogenic protozoan parasite
of the upper digestive tract. In the past when stool surveys were done, the organism was
treated merely as another commensal gut parasite Different prevalence rates recorded in
various stool surveys probably resulted from true differences between various populations
but also from a number of other factors' 1) differences in stool examination techniques
utilized, 2) the use of preservatives in specimens, 3) the number of stools examined, and 4)
the inherent variability with which G lamblia is excreted. A diagnosis of G lamblia
infection is still largely dependent upon recognizable morphologic features of the
organism.
If we are to do a better job of determining the infection in an individual patient or the
number of people infected in a waterborne outbreak, stool techniques must be developed
which will detect the organism's presence whether or not its morphologic profile is intact.
Giardia lamblia has had a variable reputation as a parasite of the upper
digestive tract of man. An analysis of stool parasite surveys conducted over
the past 55 to 60 years in the United States indicates that of the
approximately 100 papers published, relatively few attempted to assess the
prevalence of G. lamblia itself (1,2). The organism has had to occupy second
place to Entamoeba histolytica, which was long believed by most workers
and still believed by a few until recently to be the only pathogenic protozoan
parasite of the gut.
Giardia was considered to be merely one of several commensal protozoa
of the gut (including Endolimax nana, lodamoeba butschlii, Chilomastix
mesnili, and Trichomonas hominis). However, the unique morphology and
size make Giardia relatively easy to spot under the microscope. The only
feature that Giardia shares with E. histolytica is epidemiologic, i.e., they are
both transmitted via a fecal-oral route, and their presence in stools indicates
fecal pollution. Today G.lamblia has reached the ultimate as an obligate
protozoan parasite of man. It is considered to be pathogenic and often but
not always, elicits a wide spectrum of upper digestive tract symptoms. Its
sole mode of transmission was long believed to be person-to-person via
direct fecal contamination or by contaminated vegetables or other fomites.
-------
G. R. Heafy 93
However, we now know that G. lamblia or other species of the genus, are
acquired from both modern municipal water supplies and from remote
uninhabited mountain streams. It has recently been grouped with
Treponema, Gonococcus, Shigella, and E. histolytica as a causative
organism of sexually transmitted diseases (3,4).
The fact that the presence of G. lamblia was noted in records of surveys
conducted many years ago was masked because the organism was not
believed to be of sufficient importance to include in index descriptors or key
words in abstracting journals.
Surveys to detect parasites of the gut were often stimulated by
parasitologists or public health workers with diverse phylogenetic interests;
that is, the helminthologists were concerned with the worms and the
protozoologists with the protozoa. For example, reports of the numerous
surveys for hookworm disease conducted under the auspices of the
Rockefeller Foundation many years ago contain very little information on
the prevalence of Giardia or other protozoa. Examinations of fresh stools
for hookworm eggs and the use of sodium hydroxide treatment in the Stoll
egg counting technique provide little information about the stool protozoa.
Therefore, the only survey reports likely to contain accurate prevalence data
for Giardia would be those conducted to detect E. histolytica.
Stool surveys have been conducted for various reasons: to assess the
prevalence of organisms in recently returned servicemen from wars, ranging
from World War 1 up to the Vietnam conflict (5-8), to recognize parasites in
children in orphanages and mental institutions (9-12), in population groups
following epidemics of diarrheal disease (13-15), in American Indians and
Eskimos throughout the United States (10,16-21), in college students, in
residents of communities with recognized poor personal or community
hygiene (22-28) and in some cases for reasons not clearly stated.
To this day, a confirmatory diagnosis of giardiasis depends on the results
of examining stool specimens which have been normally passed or obtained
by administering cathartics. The end result is still fecal protozoology! It is,
therefore, important to consider not only which populations were being
surveyed and for what purpose, but also, and perhaps more importantly,
what techniques were being used to assess the prevalence of the parasite.
Table 1 contains an overview of stool examination procedures, some
developed in the past, and some more recent, but all utilized today. The
category "L" or "other" includes a conglomeration of techniques such as
simple gravity sedimentation, brine flotation, or mixtures of stool with sand
to break up the fecal particles. The direct saline wet-mount technique is most
widely used. For over 70 years it has been the preferred technique and is still
used exclusively in many diagnostic laboratories for individual patient
evaluations and for detecting infections in large populations. A skilled
microscopist can use this technique to obtain much valuable information on
a fresh stool specimen. Cultivation is one of the least used techniques. In
fecal protozoology it is used routinely only to supplement other procedures
for diagnosing amebiasis. It has not yet been widely applied in diagnosing
-------
94 WATERBORNE GIARDIASIS/EPIDEMIOLOGY
Table I. Techniques - Slo'ol Examination
Protozoa
A Direct saline wet mount Saline, iodine
B Schaudmn's fixative Hematoxylm, trichrome
C Kohn's Chlorazol Black E
D Merthiolate-lodme-Formalm (MIF)
E Phenol-Alcohol-Formalm (PAF)
F Sodium acetate-Acetic acid-Formalin (SAF "Junod")
G Polyvmyl-alcohol fixative (PVA)
H Formalin (5 to 10%)
I Zinc sulfate flotation
J Formalin-Ether sedimentation
K Cultivation
L Other
giardiasis. The remaining techniques shown in Table 1 (B,C,D,E,F,G,H,J,)
were developed to provide the laboratory with the options of permanent
staining, concentration and preservation in order to overcome one of the
major problems in parasite fecal diagnosis. Autolysis, disintegration,
bacterial overgrowth, and all kinds of morphologic degeneration can occur
if an untreated stool specimen is allowed to stand at various temperatures for
a long time after it is obtained and before the diagnostician examines it.
The zinc sulfate and formalin-ether procedures evolved from efforts to
improve the diagnostic capabilities of the laboratory by maximizing the
recovery of protozoan cysts and helminth eggs from specimens.
Unfortunately in Giardia detection, the formalin-ether method is
questionable. There is good evidence that it concentrates Giardia cysts
poorly, causing much distortion, and that in most cases it is not as efficient as
a direct wet-mount or zinc sulfate procedure.
It is therefore important to realize that any published prevalence rates for
G. lamblia in population surveys must be interpreted in light of the manner
in which the specimen was collected, the techniques that were used, as well as
the competence of the examiner.
Many years ago Svenssen and Linder(29) concluded that a series of about
10 normally passed stools must be examined per patient if at least 90% of all
parasitic protozoan infections are to be detected. Later, Svenssen(30) and
Sawitz and Faust(31), among others, concluded that they could reduce the
number of stools examined to six and still detect from 70 to 90% of the
infections. However, in order to obtain these results, at least two techniques
would have to be used per stool from each patient. It is therefore quite
possible that with the one-technique-per-stool approach taken in many
surveys aimed primarily at detecting E. hystolytica, only 35 to 50% of the
Giardia infections present were found.
All things being equal, if a freshly passed specimen is examined correctly
using an appropriate procedure, and the microscopist is capable of
-------
G. R. Heafy 95
distinguishing the various morphologic features of the protozoa, about half
of the infections that exist will be detected. The advantages offered by such
preservatives as merthiolate-iodine-formalin (MIF) or 10% formalin, and
the advantages of a concentration method such as zinc sulfate, for example,
may improve these detection rates somewhat. However, all things are not
always equal and the advantages provided by a particular technique may be
negated by a poor examiner or an old specimen.
Table 2 shows G. lamblia infections detected among children in stool
surveys in five randomly selected situations. Prevalence rates of 4 to 22%
were recorded. Desowitz(32) recently used wet-mount examination in
detecting Giardia infections at a level considered to represent the normal rate
(4 to 5%). However, Weiner et al(33) in Philadelphia used the polyvinyl-
alchohol fixative, formalin-ether concentrate, and stained slides to
document Giardia infection prevalence rates of 8 to 20%. Many years ago
Owen et al(10) examined one stool from each of 83 Indian boys in a
procedure which involved emulsifying the specimen in water and then
centrifuging it and staining it with iodine. Boeck(l), on the other hand, used
the simple direct wet-mount examination with iodine and found that 22% of
50 Boston children were infected. Maxey(34) found that 16% of 89 children
were infected. The procedure he used was probably a direct wet-mount, but
he does not outline the exact technique in his paper.
Table 2. Giardia lamblia - Children - Stool -S'urm'i
Place
Hawaii
Philadelphia
Wyoming
Boston
Baltimore
Ethnic
group
Local
Foreign
Puerto Rican
White
Black
Indian
Local
Local
No
Specimens
275
1 15
167
169
49
83
50
89
%
Positive
4
5
20
8
8
22
22
16
Reference
(32)
(33)
(10)
( D
(34)
The survey by Boeck( 1) is noteworthy for at least two reasons: First, it was
designed primarily to detect Giardia infections; second, he found no other
protozoa in the stools from these 50 children. Boeck concluded that the
available evidence did not support the hypothesis that G. lamblia itself
caused diarrhea in either children or in adults, except perhaps in a few
obscure and sporadic cases.
The variability of G. lamblia infection rates as detected in several surveys
conducted primarily on adults is shown in Table 3. Various techniques were
used to document prevalence rates ranging from 2 to 15%. Boeck and
-------
96 WATERBORNE GIARDIASIS/EPIDEMIOLOGY
Stiles(35) evaluated more than 8,000 specimens by direct examination only.
Brooke and his coworkers(15) used poly vinyl-alcohol fixative, stained
slides, formalin preservation, and, in some cases, formalin-ether
concentration. Ballinger(36), using direct examination, zinc sulfate
flotation, and, with some of the specimens, water sedimentation, detected a
low 2% infection rate among hospital employees.
Table 3. Giardia lamblia - Stool Survev - Aduhs.
Place
Washington, D C
South Bend, Ind
New York, N Y
Aspen, Colo
Atlanta, Ga
Minneapolis, Minn
No
Specimens
8,029
600
427
419
4,000
169
%
Positive
7
4
2
5
5
15
Reference
(35)
(15)
(36)
( 2)
(38)
( 8)
The stool survey conducted in Aspen by Gleason et al(2) was another in
which a 5% Giardia infection rate was detected among a resident population
following an epidemic in that city(37). Although such an infection rate seems
high for this population group, it probably results from the fact that Gleason
and colleagues used chlorozol black, polyvinyl-alcohol fixative, trichrome
stain, plus formalin preservation and formalin-ether concentration. They
used virtually every currently known technique except for MIF and zinc
sulfate for detecting Giardia. After World War II, Jacobs and his co-
workers^ 8) found a 5% Giardia infection rate in a stool survey on 4,000
military personnel. They did direct examination, zinc sulfate flotations, and
in some cases, evaluated stained slides. In the temperate climate of
Minneapolis, Riley(8) used direct wet-mount examination, and in some
cases, stained slides in detecting a 15% Giardia prevalence rate. As can be
seen from the results of these modern surveys, prevalence rates for G.
lamblia vary with population group, time, and the investigator.
Detecting the presence of Giardia in stools is made more difficult by its
variable excretion pattern. When conditions are ideal for the "biologic
production" of this organism, it seems to outdo itself in fecundity. Many
years ago Tsuchiya(39) found Giardia in virtually all of the stool specimens
collected from two infected individuals over several days. The highest counts
he recorded were 7.05 x 108/day for one subject and 5.85 x 108/day for the
other. Porter(40) had previously documented cyst excretion counts in
patients as high as 1.44 x 10'°/day. Kofoid and Swezy(41) reported cyst
counts of as high as 3.9 x 109/day excreted by one individual. Such Giardia
concentrations should clearly provide enough organisms to be detected with
any accepted procedure applied to the 2 to 3 stools produced per day by the
average person. However, as stated earlier, excretion rates vary
substantially.
-------
C. K. Heafy 97
Rendtorff(42) showed one of the best-illustrated examples of the variable
excretion patterns of Giardia and other protozoa in his carefully conducted
human volunteer study reported in 1954. Dancigerand Lopez(43) examined
stool specimens from hospitalized children and provided one of the most
interesting examples of the variability of Giardia excretion.
From their studies, children could be divided into three groups on the
basis of their level of Giardia excretion: those who were high excreters; low
excreters; and mixed excreters. Those with high rates of excretion expelled
up to 7,000 G. lamblia/mg feces and regularly excreted as many as 1,000
cysts/ mg feces for several days. Those with low excretion rates varied greatly
in the excretion of cysts, having as few as 12 to 108 cysts/mg feces with
periods of several days during which no organisms were detected. On the
other hand, those with the mixed excretion rates had periods when no
organisms could be detected in stools and other periods when 400 to 900
cysts/mg feces were present.
The results of multiple stool examinations for Giardia performed by other
workers are shown in Table 4. Boeck and Stiles(35) found an 8% infection
rate during their initial examination and found a 15% infection rate after 6
examinations. Jeffery(44) in his excellent report on the 3-year study of
parasitism in a mental institution described a 5% infection rate detected by a
single examination, and a 35% rate detected after 9 examinations. Indeed, in
a 7-year-study on 61 patients, he documented a 33% prevalence rate after
examining 13 specimens(45). Borland(46) found a 3% prevalence rate on his
first examination of stools from an adult population but found a 10%
prevalence rate after the third examination.
Table 4 Repeat Stool Examinations - Giardia I.amblia
Number
Specimens
505
1 10
61
191
% Positive on
Initial Exam
8
5
10
3
% Positive on
Multiple Exams
(No Exams)
15(6)
35 ( 9)
33 (13)
10 ( 3)
Reference
(35)
(44)
(45)
(46)
Results of a comparable study done in our laboratory are shown in Table
5, which illustrates the bimonthly prevalence of intestinal protozoa in 78
children whose stool specimens were examined six times in a year. This study
was done following the initial survey(2i). We used direct wet-mount
examination and formalin-ether concentration on all specimens. PVA-
fixed, trichrome-stained slides were examined when the stools were soft,
loose, or watery. Individual bimonthly prevalence rates of 5 to 16% were
obtained. However, for the whole study year 19(24%) of these children were
infected at some time.
-------
98 WATERBORNE GIARDIASIS/EPIDEMIOLOGY
Table 5 Prevalence of intestinal parasites in 78 Cherokee
Indian second grade school children
Total
Prevalence
Parasites Prevalence at Each Month (%) for 6
May'66 July'66 Oct'66 Dec'66 Feb'67 Apr'67 Months (%)
E histolytica
E hartmanni
E coli
E nana
G lamblia
C mesnili
5
22
36
24
16
8
9
13
45
18
15
5
9
10
33
18
5
4
12
15
36
18
8
5
1 1
23
35
23
1 1
4
1
13
35
23
7
1
14
38
58
41
24
10
The results obtained on stool specimens positive for Giardia from each of
the 19 children can be seen in Table 6, which also contains the findings from a
May 1965 survey. In" some instances, we detected Giardia in every stool
specimen obtained from a particular child (No. 37 and 62). In some instances
we found organisms in only about half of the stools collected from a child
(No. 39), with some Giardia found in specimens obtained early in the study
and some found near the close of the study. In a few instances, only 1
specimen out of the 7 obtained from a particular child was positive (No. 61
and 95).
What do these findings mean? In addition to the difficulty of consistently
detecting G. lamblia organisms in stool surveys conducted in the United
States in the last 60 years, there is the problem of assessing the relevance of
these findings to the true prevalence of disease in the population. Survey
specimens merely show the surface of the real presence of the parasites in
individuals. Economic, time, and personnel constraints realistically prevent
our performing the examination of at least 6 specimens per individual
deemed adequate to detect infections. It is not even feasible to examine the
Table 6. Giardia lamblia detection in stool
specimens o/ 19 Indian school children
Examination
Date
May,
May,
July,
Oct
Dec,
Feb,
April,
1965
1966
1966
1966
1966
1967
1967
1415
X X
X X
X
X
X
X
2429
X
X
X
X
X X
X
37
X
X
X
X
X
X
X
39
X
X
X
X
4448
X X
X
X
X
X
Child
5053
X
XX
X
X
No
55565861 62658295
X X X X X
X XXX
X
X X
X X
X X
X
201
X
X
X = Organisms Detected
-------
G. R. Heafy 99
recommended 3 stools (the standard O and P x 3) in surveys. Indeed,
evidence in the clinical literature on individual patients with symptoms
suggestive of giardiasis, particularly that contained in the study by Kamath
and Murugasu(47) and the review by Burke(48), indicates that a number of
stool examinations may still be negative when it is possible to find Giardia
aspirated in duodenal contents as well as in biopsy material or mucosal
imprint smears.
In the future, I think we must develop more sophisticated techniques to
detect Giardia in the population. Our present technology, for example, of
using the filter developed by the Environmental Protection Agency (see
Jakubowski, these Proceedings), shows that we can probably more
accurately detect a few cysts in 20,000 gallons of water than we can in 2 g of
feces from an infected individual. I believe that the future of stool
examination in general depends upon the development of more sensitive
techniques which do not depend solely on the morphologic identity of an
organism. The recent report by Root and his coworkers in Mexico(49) on
using an enzyme-linked immunosorbent assay (EL1SA) procedure for
detecting E. hvstolytica antigen in stool specimens is exciting. If we can
adapt this technique and rely on its results in detecting G. lamblia in stool
specimens rather than on the morphology of the variably excreted organism,
we will have reached a milestone in enlarging our diagnostic capability.
REFERENCES
1. Boeck, W. C. 1927 Giardiasis in man Its prevalence and relation to diarrhea and to
gallbladder disease Arch. Intern. Med 39-134-158
2 Cleason, N. N., M. S. Horwitz, L. H. Newton, and G. T. Moore. 1970. A stool survey for
enteric organisms in Aspen, Colorado Am. J. Trop. Med Hyg. 19(3):480-484.
3. Mildvan, D., A. M. Gelb, and D. William. 1977. Venereal transmission of enteric
pathogens in male homosexuals Two case reports. J.A M.A 238'1388-I389
4. Meyers, J. D., H. A. Kuharic, and K. K. Holmes. 1978 Giardia lamblia infection in
homosexual men. Brit J. Ven Dis 53 54-55
5 Burrows, R. B. 1954. Intestinal parasitic infections in military food handlers U. S Armed
Forces Med J V.77-82
6. Burrows, R. B. 1961 Prevalence of amebiasis in the United States and Canada Am. J
Trop. Med Hyg. 10:172-184.
7. Kuntz, R. E. 1959. Intestinal protozoans and helminths in U S. military and allied
personnel, Naval Hospital, Bethesda, Maryland. Amer. J Trop Med Hyg 8:561-564
8 Riley, W. A. 1929. Proto/oal infestation of ex-service men in Minnesota. J A.M A.
92:1661-1662.
9 Boeck, W. C.I 921. A protozoan survey of an industrial school for boys and girls J. Parasit
7:191.
10. Owen, W. B., R. F. Honess, and J. R. Simon. 1934. Protozoal infestations of American
Indian children J.A.M A 102913-915
11 Brown, E. H. 1948. Giardia lamhlia The incidence and results of infestations of children in
residential nurseries Arch Dis Child. 23-119-128.
12 Cortner, J. A. 1961. Giardiasis in childhood Quart Rev. Pediat 16:218-220
13. Spencer, H. C., J. A. Hermos, G. R. Healy, D. M. Melvin, and E. Shmunes. 1976. Endemic
amebiasis in an Arkansas community. Am J. Epidemiol. 104(1) 93-99.
14. Sexton, D. J., D. J. Krogstad, H. C. Spencer, Jr., G. R. Healy, S. Sinclair, C. E. Sledge, and
M. G. Schultz. 1975 Amebiasis in a mental institution. Serologic and epidemiologic
studies Am J Epidemiol. 100(5)414-423.
15 Brooke, M.M., D. M. Melvin, R. Sappenfield, F. Payne, F.R. N. Carter, A. C.Offutt, and
W. W. Frye. 1955. Studies of a water-borne outbreak of amebiasis. South Bend, Indiana.
Ill Investigation of family contacts. Am. J. Hyg 62(3).214-226.
-------
100 WATERBORNE GIARDIASIS/EPIDEMIOLOGY
16. Fournelle, H. J., E. L.Wallace, and V.Rader, 1958. A. bacteriological and parasitological
survey of enteric infections in an Alaskan Eskimo area. Am. J Publ. Hlth. 48(11). 1489-
1497.
17. Fournelle, H. J., V. Rader, and C. Allen. 1966. A survey of enteric infections among
Alaskan Indians. Publ Hlth. Rpts. 81:798-803
18. Melvin, D. M., and M. M. Brooke. 1962. Parasitologic surveys of Indian reservations in
Montana, South Dakota, New Mexico, Arizona, and Wisconsin Am. J. Trop. Med. &
Hyg. ll(6):765-772
19. Hitchcock, D. J. 1951. Parasitological study on the Eskimos in the Kotzebue area of
Alaska. J. Parasit. 37(3):309-311.
20. Becker, D.A. 1968. Enteric parasites of Indians and Anglo-Americans, chiefly on the
Winnebago and Omaha reservations in Nebraska. Nebraska State Med. J. 53(6):293,
53(7):347, 53(8):380, 53(9):42I
21. Healy, G. R., N. N. Gleason, R. Bokat, H. Pond, and M. Roper. 1969. Prevalence of
ascanasis and amebiasis in Cherokee Indian school children Publ. Hlth 84(10).907-914.
22. Faust, E. C., and W. H. Headlee. 1935. Intestinal parasite infections of the ambulatory
white clinic population of New Orleans, Am. J. Trop. Med 16(l):25-38.
23. Jones, F. E., C. S. Smith, and D. E. Eyles. 1943. Epidemiological study of Endamoeba
histolylica and other intestinal parasites in the New Hope community of Tennessee. A
restudy after 21 years. Am J. Trop. Med. Hyg 3:266-275.
24. Eyles, D. E., F. E. Jones, and C. S. Smith. 1953. A study of Enlamoeba histolvlica and
other intestinal parasites in a rural West Tennessee community. Am. J. Trop. Med. Hyg.
2:173-190
25. Meleney, H. E. 1930. Community surveys for Endamoeba hutolvtica and other intestinal
protozoa in Tennessee: First report. J. Parasit. 16:146-153
26. Faust, E. C. 1930. A study of the intestinal protozoa of a representative sampling of the
population of Wise County, Southwestern Virginia. Am. J. Hyg. 11:371-384.
27. Jones, F. E., D.E. Eyles, and C. S. Smith. 1953 Intestinal parasitism in a plantation
population of the Yazoo-Mississippi delta. Mississippi Doctor:85-87.
28 Byrd, E. E. 1936. On the incidence of intestinal parasites in 537 individuals on the relief rolls
in the city of Athens, Georgia, and vicinity. Amer. J. Trop. Med. 16(l):39-45.
29 Svenssen, R., and F. J. Linder. 1934 The chances of detecting infection with intestinal
protozoa. Acta Medica Scandinavia 81:267-324.
30. Svenssen, R. 1935. Studies on human intestinal protozoa. Acta Med. Scand. Suppl. LXX
pp X+115.
31. Sawitz, W. G., and E. C. Faust. 1942. The probability of detecting intestinal protozoa by
successive stool examinations. Am. J. Trop. Med 22.130-136.
32. Desowitz, R. S., and N. H. Wiebenga. 1975. A survey of intestinal parasites in Oahu school
children Hawaii Med. J. 34:21-23.
33 Weiner, D., M. M. Brooke, and A. Witkow. 1959. Investigation of parasitic infections in
the central area of Philadelphia. Am. J. Trop. Med. Hyg. 8(6):625-629
34. Maxey, K. F, 1921. Giardia (lambha) mteslinalis, a common protozoan parasite of
children. Bull. Johns Hopkins Hosp. 32:166.
35. Boeck, W. C., and Stiles, C. W. 1923. Studies on various intestinal parasites (especially
amoebae) of man. Bull. 133. Hyg. Lab. United States Public Health Service.
36. Ballinger, J., M. Simon, and B. Weingarten. 1962. Intestinal parasitosis among hospital
employees J.A.M.A. 181-638-640.
37. Moore, G. T., W. M. Cross, C. D. McGuire, C. S. Mollohan, N. N. Gleason, G. R. Healy,
and L. H. Newton. 1969. Epidemic giardiasis at a ski resort. NewEng. J. Med. 281:402-407.
38 Jacobs, L., F. G. Wallace, and L. S. Diamond. 1946. Survey of intestinal parasites in
soldiers being separated from service. Bull U. S. Army Med. Dept 6:256-262.
39. Tsuchiya, H. 1931. A study of variabilities in dimensions and numbers of discharged cysts
of Giardia lambha (Stiles 1915) from day to day under normal conditions. Am. J Hyg.
13:544-567.
40. Porter, A. 1916. An enumerative study of the cysts of Giardia intesnnalis in human
dysenteric feces. Lancet 94: 1166-1169.
41. Kofoid, C. A., and O. Swezy. 1922 Mitosis and fission in the active and encysted phase of
Giardia entenca (Grassi) of man, with a discussion of the method of origin of bilateral
symmetry in the polymastigote flagellates Univ. Calif. Pub Zool. 20:199-234
42. Rendtorff, R. C. 1954. The experimental transmission of human intestinal protozoan
parasites Am. J. Hyg 59:209-220.
-------
G.K.Heafy 101
43. Danciger, M., and M. Lopez. 1975. Numbers of Giardia in the feces of infected children.
Am. J. Trop. Med. Hyg. 24:237-242.
44. Jeffery, G. M. 1960. A three-year epidemiologic study of intestinal parasites in a selected
group of mental patients. Am. J. Hyg 71(l):l-8.
45 Jeffery, G. M., and A. J. Harrison. 1963. Intestinal parasitism in a group of mental patients
during seven years of observation. J. Parasit 49(2):308-3I2
46. Borland, J. L. 1939. The minimum incidence of intestinal protozoa in a representative
sampling of the adult population in Florida Southern Med. J. 32(4):364-370.
47 Kamath, K. R., and R. Murugasu. 1974. A comparative study of four methods for
detecting Giardia lamblia in children with diarrheal disease and malabsorption. Gastroent
66.16-21.
48. Burke, J. A. 1977. The clinical and laboratory diagnosis of giardiasis. CRC Cnt. Rev. In
Clin. Lab Sci..373-391.
49 Root, D., F. X. Cole, and J. A. Williamson. 1978. The development and standardization of
an EL1SA method for the detection of Entamoeba hntoh'tica antigens in feces samples.
Arch. Invest Med. (Mex.) 9 Suppl. 1-203-210
Discussion
C. HENDRICKS: There are some very good data to show that giardiasis
can be a waterborne disease, but I think it is obvious that giardiasis can be
transmitted by person-to-person contact as well as foodborne. In your
opinion, to what extent do you think giardiasis may be transmitted by water,
and can you perhaps give some idea on what the degree of national concern
should be?
G. HEALY: 1 am a diagnostician and you may want to hear from some of
the epidemiologists who have a more intimate knowledge of the
epidemiology of Giardia. The evidence is certainly strong that it is
transmitted by water: it has been found in water and it has been associated
with diarrheal diseases in surveys that have been conducted that indicate that
it was transmitted through water. 1 think there should be national concern
for it.
M. SCHULTZ: The evidence for waterborne transmission will be
presented in papers by Dr. Juranek and Mr. Craun; if it is not, I will ask the
question again. But, I would like to challenge you on your comment that
there is evidence that giardiasis is foodborne. 1 do not know of any such
evidence.
G. JACKSON: As far as I know, there are no proven cases of foodborne
giardiasis. The cysts have been detected, presumably in a viable state, on
strawberries, but we have no direct evidence of a human case. 1 only saw a
newspaper story on this; however, in New York City, venereal transmission
has been implicated in giardiasis.
M. WOLFE: 1 do not know of any proven evidence that Giardia is
transmitted by food. It just stands to reason that people travelling in endemic
areas who are eating salads or breads, or are drinking beverages that have
been handled by infected food handlers, may become infected this way. I do
not think that giardiasis can be considered exclusively a waterborne disease.
The epidemiologrc evidence in this country and from the Soviet Union
certainly indicates that this is the major form of transmission. I really cannot
see it as the exclusive way that Giardia is transmitted. 1 believe it is
reasonable to say that a salad, mixed by fingers of an infected food handler,
after it has been washed in water, constitutes a foodborne rather than a
waterborne disease.
-------
102 WATERBORNE GIARDIASIS/EPIDEMIOLOGY
R. RENDTORFF: It is very important to consider this parasite in terms
of past and present. 1 think the survey that you gave, Dr. Healy, is very
important in this regard. I too feel that Giardia can be transmitted by a
number of routes, as are all parasitic protozoa. We tend to focus here on
Giardia and forget that there are other organisms of very similar size that are
also transmitted and have the same biological characteristics as Giardia. In
the future you may pay more attention to these, such as E. coli and
Entamoeba histolytica, in terms of how they might be transmitted through
water.
One thing that bothers me about the problem with giardiasis today is the
fraction of physicians who obtain a report from hospital laboratories of
Giardia lamblia in a patient when there is no symptom referable to
giardiasis. Should this patient be treated or shouldn't he be? It is an
expensive medical, economical problem. Giardiasis is fairly prevalent in our
population, as you can see by these statistics. If five per cent of us have it, it
takes a great deal of Atabrine (quinacrine) and a lot of treatment to get rid of
this organism from society, and we do not know whether we are really
getting rid of it this way.
I believe we have to interpret these statistics very carefully. Years ago,
when we did our experiments, the medical profession at large thought
Giardia was perfectly non-pathogenic, and it was rarely treated. In fact some
hospitals did not report what they considered to be non-pathogenic species.
If you look at hospital statistics, I think you will find that in years past,
Giardia was considered to be non-pathogenic and non-existent, while E.
histolytica was everywhere. One was overreported and the other was
underreported. Accordingly we must carefully interpret these data.
R. DAV1ES: There is a published report of foodborne giardiasis by
Gangarosa(Journ. Inf. Dis., 122:354, 1970.) His only data were derived from
one outbreak, and I believe it affected about 19 people. There was no
explanation in his text. That is the only published report I have ever seen in
reference to foodborne giardiasis. (Editor's Note: The article referred to by
Mr. Davies was adapted from the then National Communicable Disease
Center (NCDC) Foodborne Outbreak Annual Summary for 1969. At that
time, waterborne outbreaks were incorporated into the foodborne
summary. Although not mentioned in the Journ. Inf. Dis. paper, the NCDC
summary has water listed as the vehicle in this outbreak.)
D. JURANEK: We have had one foodborne outbreak reported in the
CDC Annual Summary of Foodborne Outbreaks, 1973; but, as 1 recall, the
epidemic was worked up inadequately. Although the epidemiologist
concluded that the epidemic may have been the result of foodborne
transmission, not much manpower or effort was devoted to the investigation
and the data are far from conclusive. Had the epidemiologist realized the
importance and controversial nature of his observations, I am sure that the
epidemic would have been worked up better.
S. ERL AN DSEN: Y ou are talking about a prevalence of about 24 per cent
on cases in your stool surveys. If we look at some of the animal literature two
studies come to mind: one performed in Czechoslovakia and one performed
in Japan, on essentially opposite sides of the world. They refer to prevalence
of Giardia in animals on the order of 70 to 85%. Would you think perhaps
that we can see a higher percentage than your 24% in terms of trophozoites
-------
G.R.Heafy 103
being normally found in the bowel but not showing any signs of
symptomatology, just a normal inhabitant?
G. HEALY: I think the prevalence rates would probably be much higher if
we had better techniques for detecting the organisms. We have to identify
them in stools, as trophozoites, or we have to look for the median bodies in
the quadrinucleate cyst form, and we have been doing the same thing now for
75 years. Our technology has reached the point where we probably should
utilize some of the more sophisticated techniques. The ELIS A technique for
antigen in stools is one that 1 think is especially good. Mr. David Root
(Millipore Corp., Bedford, Mass.) came with us when we did a survey of
Cherokee Indians this past year and showed us how this technique worked
for E. histolytica infection. With this technique, we detected E. histolytica
antigen in 2 Indian children. Our regular stool examination techniques had
shown that one was negative on initial examination.
The EL1SA test may be one way to establish the prevalence of the
organism by obtaining an index of its presence by antigens or proteins,
because they get broken up on the way down through the bowel. Our
detection system is qmte poor in detecting organisms in very small numbers,
whereas presumably a small amount of protein could be picked up by the
ELISA test.
T. VERNON: A comment on the generalization of prevalence surveys: it
concerns me that we would leave here with the figure of five percent. 1 think
it important to note that in the prevalence studies that have been done we
have not adequately characterized the survey population. Certainly the
Aspen residents with their 5% prevalence in the spring of 1966 were not
typical of the residents of the United States in general. The Cherokee Indians
in North Carolina are certainly not representative, either. Since that 5%
figure has been used rather frequently, I wanted to comment on that.
Incidentally, I did catch the tense of the verb you used when you said the
Aspen residents "are" drinking water contaminated with Giardia cysts.
Whatever is happening in other ski resorts, I do not know, but it certainly
should be stated, and the Aspen residents would agree, that Aspen has well-
filtered water.
G. HEALY: The situation was corrected. In reference to the prevalence
rate mentioned in those instances, I do not know what the prevalence of
Giardia is throughout the United States, and I could not give you a figure. It
would depend on which population you wanted to pick and how
representative that population is of the entire country. It does vary from
place to place.
-------
104
Animal Reservoirs and Cross-Species
Transmission of Giardia*
Robert B. Davies and Charles P. Hibler
Wild Animal Disease Center
Colorado State University
Fort Collins, Colorado
ABSTRACT
During 1975-77 a survey of wild and free-ranging domestic animals and man for Giardia
was completed in several areas of Colorado. Fecal samples obtained from animals and man
were examined by a zinc sulfate centrifugation technique. Of 744 samples from 33 species of
vertebrates, 65 (9%) were positive for Giardia. Six species of mammals were represented
with positive findings as follows: beaver (Castor canadensis)44 of 244 (18%); coyote (Cams
latrans) 2 of 34 (6%); cattle (Bos tarns) 6 of 58 (10%); domestic cat (Felis domesticus) 1 of 4
(25%); dog (Canisfamiltaris) 10 of 78 (13%) and man (Homo sapiens)2 of 32 (6%). Human
source Giardia cysts were given to laboratory animals, wild animals and domestic animals.
Hamsters (Mesocricetus auratus), domestic rabbit (Oryctolagus cunniculus), laboratory
mice (Mus musculm), deer mice (Peromyscus maniculatus), black bear (Ursus
amerwanus), wapiti (Cervus canadensis), mule deer (Odocoileus hemionus), white-tailed
deer (Odocoileus virgmianus), domestic sheep (Ovts aries), and cattle (Bos tarns) did not
become infected. Laboratory rats (Rattus norvegicus), gerbils (Gerbil/usgerbtllus), guinea
pig (Cavis porcellus), beaver, dog, raccoon (Procyon lotor), bighorn X mouflon sheep
(Ovis canadensis X O. musimon), and pronghom (Antilocapra americana) became
infected with Giardia. Cysts were first found in the feces from 6 to 34 days post exposure.
Animals were positive for Giardia for from 1 day to 3 months
The flagellated protozoan Giardia (Protozoa: Hexamitidae) was first
observed by Leeuwenhoek, who found the trophozoite in his own feces, and
was later described by Lambl in 1859(1,2). Europeans often call the parasite
Lamblia intestinalis, but Giardia lamblia is the preferred and generally
accepted name for the human form(3). A recent survey(4) has shown that G.
lamblia is the most common parasite of man in the United States, infecting
about 3.8% of the populace with prevalence ranging from 2 to 20 percent(5).
Giardia is about 3 times more common in children(6) and up to 100% of a
group of children may be infected(l). The pathogenicity of Giardia is a
controversial subject(6) in spite of numerous reports of Giardia caused
diarrhea(2,7,9) and one report of death(lO).
Giardia infections are acquired by ingesting viable cysts(l). Cysts may be
ingested on or in food(l,l 1), in water or during intimate contact(l). Several
outbreaks of giardiasis have been traced to municipal water supplies(5,9,12,
16). Two of these, in Aspen(12) and Vail(14), Colorado, were traced to
* Financial support for these studies was provided by the Center for Disease Control,
Atlanta, Georgia (Survey) and the Environmental Protection Agency, Cincinnati, Ohio
(Cross-transmission).
-------
K.B.Davies 105
contamination of the water supply by sewage. An outbreak of giardiasis in
Camas, Washington has been associated with Giardia infected beaver living
in streams supplying the Camas water system(17). Raw water samples from
the water system of Rome, New York, collected during an outbreak of
giardiasis, were fed to beagle puppies; infection with Giardia resulted in
experimental dogs while control dogs remained negative(15). These cysts
would probably infect humans as well. Previous work completed at
Colorado State University (CSU) has shown human-source Giardia will
infect dogs and cats and circumstantial evidence indicates the reverse also
occurs (Hibler, unpublished data).
Giardiasis exists in 2 forms in Colorado: endemic and epidemic. During
the last 5 years an average of 280 cases of human Giardia have been reported
each year; there have been 271 reported cases in 1978 as of 26 August and 75
of these occurred in the month of August( 18). Epidemics of giardiasis have
been reported from Aspen(12), Boulder(19), Vail(14), Estes Park(5) and
various resort lodges(20,21). In most cases the source of the cysts was not
determined but the possibility of animal reservoirs being the source has been
raised(20).
Very little cross transmission work has been done with Giardia. Previous
work has consisted of infecting laboratory rats and woodrats (Neotoma
fuscipes)with Giardia obtained from man(6). Grant and Woo(22) found that
G. muris from laboratory mice, G. simoni from laboratory rats and G.
peromysci from deer mice were very host specific while G. micron and G.
mesocricetus were not host specific. Human-source Giardia has been given
to puppies and infections with Giardia have been established(23).
Due to the lack of cross transmission work and the belief that Giardia is
very host specific, different specific names have been given to Giardia from
each host(6). Filice(24) compared some species of Giardia, found no
structural difference and concluded there were two species of Giardia which
infect mammals: G. muris in the mouse, rat and hamster and G. duodenalis
in all other mammals including man.
Because of the high number of cases and epidemics of human giardiasis in
Colorado and the possibility that wild or domestic animals infected with
Giardia contributed to these outbreaks, the present study was begun in 1975.
The study consisted of two parts: 1) a survey of wild and free-ranging
mammals, including man, along streams in areas which were both endemic
and had a history of epidemics of giardiasis, and 2) an attempt at cross-
transmission utilizing human-source Giardia to infect as many species of
laboratory animals, wild animals and domestic animals as possible.
METHODS AND MATERIALS
Survey of Wildlife
Fecal specimens were obtained from water, off the ground, from the
rectum of animals shot or killed by automobiles, from veterinary clinics, and
from backpackers and their pets. The greatest number of samples was
obtained by picking up fresh feces of various animals. Samples were not
picked up unless they were less than 12 hours old except for those of beaver
-------
106 WATERBORNE GIARDIASIS/EPIDEMIOLOGY
which were collected from the bottom of ponds. Beaver feces remained cool
and moist in these ponds. Dry or frozen feces were not collected. Samples
also were obtained from animals submitted for necropsy at the Wild Animal
Disease Center, CSU. Additional fecal samples were obtained from animals
live trapped and released or kill trapped. Fecal samples from dogs and cats
were obtained from veterinarians in Aspen and Granby, Colorado and from
the ground in the wild.
A survey of backpackers and their pets utilizing the Maroon Bells-
Snowmass Wilderness Area near Aspen, Colorado (Fig. 1) was completed in
1976. Each backpacker going through the U.S. Forest Service check station
on the Maroon Bells road was given a packet which included an explanation
of the study, a 60 ml polypropylene bottle containing 40 ml of 2% formalin
solution and a prepaid and addressed envelope. The backpacker was asked
to place approximately a teaspoon of feces into the bottle, mark the label as
to sex, age, and species of the source, and mail the bottle to the Wild Animal
Disease Center, CSU. Results of the fecal examinations were made available
to interested cooperators. Fecal samples collected in 1975 were placed in a
solution of 10% formalin. This method did not prove entirely satisfactory as
10% formalin disrupted many of the cysts. In 1976, samples were placed in
2% formalin or refrigerated in distilled water until examined.
Samples preserved in formalin were shaken to suspend all the particles
and a 15 ml round bottom centrifuge tube was filled. After centrifuging at
2300 rpm for 5 to 10 min, the supernatant was decanted and the sediment
examined. Samples in distilled water were treated the same as formalin-fixed
samples if excess water was present; if not, a marble sized piece of feces was
used for examination. All fecal samples were examined by a zinc sulfate
(ZnSO4) centrifugation technique.
ZnSO4 at a specific gravity of 1.18 is made by adding 331 g of ZnSO4 to
1000 ml of water; since ZnSO4 is hygroscopic, 331 g is seldom sufficient and a
hydrometer must be used to adjust the specific gravity. Lugol's iodine is
made by adding 10 g of potassium iodide and 5 g of iodine to 100 ml of
distilled water.
Five or 6 drops of Lugol's iodine is added to the sediment in the centrifuge
tube and thoroughly mixed. The tube is then filled half way with ZnSO4
solution and thoroughly mixed. ZnSO4 solution is added until the meniscus
bulges slightly and a coverslip is placed on top. Centrifuge at 2300 rpm for 3
min, remove the coverslip and place on a slide for microscopic examination.
THE SURVEY STUDY AREAS
Aspen and Hideaway Park in central Colorado (Fig. 1) were chosen as the
primary study areas because of a history of endemic and epidemic human
giardiasis in each area, and their year-round use by a large number of people.
Aspen is a small resort town in Pitkin County at an elevation of 2400 m
(7900 feet). Tourism during the summer and skiing during the winter months
are the main industries. The surrounding terrain is very steep and mostly
covered with trees or brush. Aspen utilizes water from Castle and Maroon
Creeks for the city water supply.
-------
R.B.Davies 107
Hideaway Park is a small resort town in Grand County at 2900 m (8800
feet) in elevation. It is the nearest town to the Winter Park Ski Area, which
has no overnight or restaurant facilities. Therefore, Hideaway Park is a
center of activity in the winter. During the summer it is a popular resort. The
water supply for Hideaway Park is the Little Vasquez River. The city of
Denver, Colorado, also takes water from the Vasquez River Basin. Several
secondary sampling areas near Hideaway Park also were examined. These
river drainages included Beaver Creek, Cow Creek (Cow Gulch), Williams
Fork River, Willow Creek, Corral Creek and the Fraser River.
Brush Creek, near Eagle in west-central Colorado (Fig. 1), was sampled
after a human who drank water from East Brush Creek developed giardiasis
9 days later. Both the east and west branches of Brush Creek flow through
private and public property. Fishing and camping are the main activities in
the area.
Vermejo Park is a large (193,500 ha or 480,000 acre) ranch in north-central
New Mexico owned by the Pennzoil Corporation. There is little human
activity on the.ranch. Fecal samples were gathered from several species of
animals which would have little chance of acquiring the parasite from man.
A small part of Rocky Mountain National Park (RMNP) was examined
because an outbreak of giardiasis in 8 people was reported(5). The people
were from California and stayed at the Cascade Cabins on Fall River within
the boundaries of RMNP. Fecal samples, from animals, found along Fall
River to its origin and from along Hidden Valley Creek, were examined.
Samples from animals from the entire state of Colorado were examined
whenever they became available. These are all listed under the heading
"Statewide".
EXPERIMENTAL INFECTIONS
Human feces containing Giardia cysts were obtained from the CSU
Student Health Center, Poudre Valley Memorial Hospital and the Northern
Colorado Diagnostic Laboratory. Beaver-source Giardia were obtained
from naturally infected beaver on Beaver Creek near Hot Sulphur Springs,
Colorado. Mule deer source Giardia were obtained from a naturally infected
fawn (part of a captive herd), near Kremmling, Colorado.
A small amount of feces from each sample was examined by the ZnSO4
centrifugation method to ensure the presence of Giardia cysts. If the cysts
appeared viable, the entire fecal sample was thoroughly mixed with distilled
water. This mixture was strained through a double thickness of cheese cloth,
the container rinsed with distilled water, and the fluid poured over the
filtered feces which was then discarded. The filtered liquid was poured into
50 ml round bottom test tubes and centrifuged at 2300 rpm for 5 minutes.
The supernatant was discarded and the pellet resuspended in distilled water
and centrifuged again. This process was repeated until the supernatant was
almost clear. The fluid was discarded and once again the pellet resuspended
in distilled water. This was allowed to stand for 1-5 minutes or centrifuged at
1000 rpm for 10 seconds to sediment larger particles. The supernatant was
then poured into a glass beaker and samples taken to determine the
-------
108 WATERBORNE GIARDIASIS/EPIDEMIOLOGY
COLORADO
FIG. 1. Locations of Giardia survey study areas in Colorado 1975-1977,
concentration of Giardia cysts. Fluid could be removed, after the cysts
settled to the bottom of the beaker, or added to obtain the desired number of
cysts per volume of fluid.
Animals were infected via stomach tube, orally or in water provided to
animals deprived of water for the previous 24 hours. Ten thousand cysts
were used as an infective dose whenever possible. Fecal samples were
examined daily from each animal, including controls, for a minimum of 14
days prior to infection and every day after infection for at least 21 days. All
fecal samples were examined by the ZnSCX centrifugation method. Animals
were monitored to determine the prepatent period (the time from infection
to the passing of cysts in the fe'ces) and the patent period (the length of time
cysts were found in the feces).
Laboratory mice and rats, hamsters, guinea pigs, gerbils and domestic
rabbits were obtained from the Experimental Animal Service at CSU or
local pet shops. Beaver, raccoon, deer mice and house mice were live trapped
in the wild. The large wild mammals, mule deer, white-tailed deer, wapiti,
pronghorn, bighorn X mouflon sheep and black bear were obtained from the
Colorado Division of Wildlife or were captive animals held in pens at the
CSU Foothills Campus. Large domestic animals (sheep and cattle) were
obtained and held in pens at the Foothills Campus. Specific Pathogen Free
(SPF) beagle puppies were obtained from the Collaborative Radiological
Health Laboratory (CRHL) on the Foothills Campus. All animals were kept
either at the Foothills Campus or in animal rooms on the main campus, fed
-------
R.B.Davies 109
dry commercial feed and/or alfalfa hay. Fort Collins city water was
provided as needed.
When feasible, animals were sacrificed and wet impression smears and
smears from intestinal scrapings were made from various levels of the
intestine. Wet smears were stained with Lugol's Iodine and examined
immediately. A fecal sample was taken from the lower colon at necropsy and
examined by the ZnSO4 centrifugation method.
RESULTS OF THE SURVEY
Overview
One of the primary objectives of the study was to determine if wild and / or
free ranging mammals found along streams utilized by municipalities and
outdoor recreation users for drinking water, were infected with Giardia. As
the primary complaints of giardiasis were from backpackers who thought
they acquired the parasite while backpacking, most samples collected during
1975 were near or above timberline. Samples were taken at lower elevations
during 1976.
A total of 744 fecal samples from 33 species of vertebrates was examined
by the ZnSO4 centrifugation method (Tables 1 and 2). Sixty-five of the 744
fecal samples (9%) were positive for Giardia with positive findings in
mammals as follows: beaver 18% (44 of 244); coyote 6% (2 of 34); cattle 10%
(6 of 58); domestic cat 25% (1 of 4); dog 13% (10 of 78) and man 6% (2 of 32).
During 1975, 164 samples were examined from three areas: Aspen, Middle
Park and Vermejo Park. Cattle (5 of 23), cats (1 of 2) and dogs (7 of 19) were
infected with Giardia (Table 1). Most of the 1975 samples were recovered at
high altitudes (over 2700 m or 9,000 feet). In 1976, sampling was
concent rated .at lower elevations and 52 of 580 fecal samples were positive
for Giardia (Table 2). These were beaver (44 of 244), coyote (2 of 34), dogs (3
of 59), cattle (1 of 35) and human (2 of 31).
Aspen
Aspen utilizes Castle and Maroon Creeks for its city water supply.
Samples from these streams and their major tributaries above the city water
siphons, and fecal samples from animals in the vicinity of the streams were
examined.
Maroon Creek is a typical mountain stream with fast flowing water. The
terrain on both sides is steep and covered with various types of vegetation.
Willow (Salix spp), aspen (Populus spp), pines (Pinus spp), spruce (Picea
spp), firs (Pseudotsuga) and juniper (Juniperus spp) are the main species
seen with numerous grasses and forbs present as an understory. Many slopes
are in reality cliffs and numerous talus slopes are present. Some open grass
covered avalanche paths are seen.
Aspen takes water from Maroon Creek, 270-360 m (300-400 yards) above
the T Lazy 7 Ranch. As far as is known, there are no human dwellings above
the T Lazy 7 Ranch along Maroon Creek. Numerous campgrounds are
maintained by the U.S. Forest Service along Maroon Creek up to and
including Maroon Lake. These campgrounds are used heavily all summer.
-------
110
WATERBORNE GIARDIASIS/EPIDEMIOLOGY
Table I. Number of fecal samples of New Mexico and Colorado mammals examined
for Giardia sp. in 1975-1976.*
AREA
Animal" Vermejo Park
Colorado
chipmunk
Cottontail rabbit
Golden-mantled
ground squirrel
White-tailed
jackrabbit
Marmot
Pika
Porcupine
Richardson's
ground squirrel
Red squirrel
Rock squirrel 2
Badger
Black bear
Pine marten
Raccoon
Red fox
Striped skunk
Wapiti 10
Mule deer
Bovine
Cat
Dog
Horse
Sheep
Human
12
Aspen
4
1
1
5
10
1
2
1
1
2
1
10
1
19(7)
10
20
1
90(7)
Middle Park
6
1
5
7
1
5
1
2
1
3
1
6
1
13(5)***
KD
8
62(6)
Total
10
1
1
1
10
17
2
7
1
2
2
1
1
3
1
1
18
2
23(5)
2(1)
19(7)
18
20
1
164(13)
* / July 1975 - 30 June 1976.
** See Appendix A for a list of Scientific Names.
*** Number fecal samples examined (number infected).
-------
R. B. Davits
111
Table 2 - Number of fecal samples of New Mexico and Colorado mammals examined
for Giardia sp. in 1976-1977.*
AREA
Animal" Aspen
Beaver 64(8)***
Cottontail
rabbit
Marmot 1
Muskrat 4
White-tailed
prairie dog
Richardson's
ground
squirrel 2
Black bear
Bobcat
Coyote
Mountain
lion
Raccoon
Striped skunk
Bighorn
sheep
Wapiti 3
Mule deer
White-tailed
deer
Bovine 2
Cat
Dog 20(3)
European
ferret
Horse
Human 30(2)
Brook trout
126(13)
Middle Brush
Park RMNP Creek Other
107(33) 41 32(3)
1
1 4 12
1
1
1
34(2)
6
13 2 9
2
20 32
6 6
35
1
22 11(1)
2
38 1
1
1
1
10
185(33) 86 43(4) 140(2)
Total
244(44)
1
1
21
1
2
1
1
34(2)
6
24
2
52
15
35
1
35(1)
2
59(3)
1
1
31(2)
10
580(52)
* July 1976 - 30 June 1977
** See Appendix A for a list of Scientific Names.
*** Number of fecal samples examined (number injected).
-------
112 WATERBORNE GIARDIASIS/EP1DEMIOLOGY
The area around Maroon Lake and Crater Lake are heavily used by day
hikers and backpackers.
Very little wildlife was observed along the streams serving as the water
supply for Aspen. Beaver were not seen above the confluence of East and
West Maroon Creeks. One old beaver colony exists at Maroon Lake but no
beaver were observed; however, several dams were found below thejunction
of the creeks. One of 29 (3.4%) beaver samples from Maroon Creek was
positive for Giardia. None of the 4 muskrat samples from the pond where the
positive beaver sample was found contained Giardia cysts. Most species of
wildlife were found in rocky areas which were not close to open water or
moist areas. Chipmunks, pika and marmot, were usually found high on the
slopes in rocks or on dry ground. Marmot were found close to Maroon
Creek only one time. All samples collected from these animals were negative.
Few mule deer or wapiti were seen along the river and all samples from these
animals were negative. On one occasion approximately 200 domestic sheep
were grazing at West Maroon Pass. Twenty samples from these animals were
negative for Giardia. None of a small herd of cattle, pastured near the
confluence of East and West Maroon Creek, was positive for Giardia. The
most common animal seen was the domestic dog. Dog feces was common
around campgrounds, Maroon Lake and most hiking trails which follow
rivers and creeks in the area.
Castle Creek is very similar to Maroon Creek except the valley bottom is
wider. Human dwellings are found along Castle Creek and most of its
tributaries. At the end of the road on Castle Creek is a mining operation and
trespass beyond this point is not allowed. Approximately 0.5 km (0.3 miles)
below the mine Pine Creek joins Castle Creek. A moderately used trail leads
along Pine Creek to Cathedral Lake. Animal fecal samples were collected
along Pine Creek and around Cathedral Lake. All pika, marmot, pine
marten and dog samples from the Pine Creek area were negative. Samples
from one beaver pond on Pine Creek were negative as were samples from a
beaver pond just below the confluence of Pine and Castle Creeks. All animal
samples collected above the confluence of Pine Creek and Castle Creek were
negative. A small campground with picnic tables is located near Cathedral
Lake and an outdoor toilet has been built about 60m (200 feet) uphill from
Pine Creek. Below the confluence of Pine Creek and Castle Creek is an
assortment of restaurants, dude ranches and private homes.
Owners of the Ashcroft Ski Touring Company would not allow collection
of samples from the beaver ponds or other animals on the property. This
organization had a dog kennel which was built along Devaney Creek. There
is also an outhouse built along the creek near the old Ashcroft townsite.
Below the Ashcroft property 22% (7 of 32) of beaver samples were positive
for Giardia whereas none of 3 samples collected upstream from the property
were positive. Express Creek flows into Castle Creekjust below the Ashcroft
property and originates near the summit of Taylor Pass. This is a popular
"four wheel drive" road. Several houses have been built along Express Creek
and several campgrounds, without sanitary facilities, are found along the
-------
R.B.Davies 113
Creek above the homes. All fecal samples from pika and marmot in this area
were negative for Giardia. One human stool found along Express Creek was
negative for Giardia.
Conundrum Creek joins Castle Creek 9.5 km (6 miles) below Express
Creek. Conundrum Creek, in part, originates from the Conundrum Hot
Springs which is a very popular site in the area. It has become so popular the
Forest Service has prohibited camping within 1.6 km (1 mile) of the springs.
The terrain along Conundrum Creek is very steep and clifflike along most of
the trail to the hot springs. There are also some very heavily timbered areas,
mostly spruce and fir, and some willow patches. Very little wildlife,
consisting of a small number of pika and marmots was observed along the
trail. Dog feces was found along the trail and human feces, too old for
sampling, was found near several old campsites often within 0.6 m (2 feet) of
the streams. The area surrounding the hot springs was littered with articles of
clothing, mostly underclothing, trash and toilet paper.
Beaver dams were not observed on Conundrum Creek or on Castle Creek
below Elk Mountain Lodge downstream to near Queens Gulch. Beaver
possibly inhabit this stretch of Castle Creek but much of it is on private land
and not readily available for examination. In the area below Queens Gulch, 3
of 6 fecal samples from beaver dams were positive for Giardia, as were 4 of 26
samples from the Elk Mountain Lodge area. No other animal species were
found positive for Giardia.
Dog feces was picked up whenever it seemed fresh enough to be a valid
sample (not dried out or exposed to freezing temperatures). Feces was found
along every trail and campground visited but few were fresh enough to be
used. During 1975, 10 samples were obtained from the dog pound in Aspen
operated by Lew Lou's Boarding Kennels, and 7 of these contained Giardia.
In 1977 additional samples were collected consisting of 11 samples from
kenneled dogs and 5 from pound dogs. All 11 samples from kenneled dogs
were negative and 60% (3 of 5) of the pound dog samples were positive for
Giardia. All dog feces picked up from trails and campgrounds was negative
for Giardia although other parasite ova were found.
During the summer of 1976, 130 Giardia-samp\ing packets were given to
backpackers going into the Maroon Bells-Snowmass Wilderness Area. Of
the packets, 25% (33 of 130) (3 samples from dogs and 30 from humans) were
returned. No dogs were infected, but 6.6% (2 of 30) human samples
contained Giardia (Table 3). The survey was repeated in 1977, but the road
to Maroon Lake was closed to all but camper traffic and all other visitors,
including backpackers, rode 143 the lake in buses and a very small number of
packets were given out. Of the 12 which were returned, all were negative for
Giardia.
Vermejo Park
Samples from two rock squirrels and 10 wapiti taken from Vermejo Park
in 1975 were negative for Giardia. In 1976, 34 coyote samples were collected
from this area and 6% (2 of 34) were infected with Giardia. This area is used
by very few people and there are no camping areas.
-------
114 WATERBORNE GIARDIASIS/EPIDEMIOLOGY
Table 3 - Results of 1976 human survey for Giardia in the Maroon Bells
area of Central Colorado.
Sample #
1
2
3
4
5
R
U
7
/
8
9
10
1 1
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
09
o z
00
o o
Age
29
24
28
28
33
R 1
•j i
0 1
Z. I
29
—
—
2
17
24
25
41
23
28
23
28
9
—
53
23
—
8
26
—
—
26
28
27
Sex
M
M
M
M
M
M
M
M
M
M
F
F
F
M
M
M
F
F
—
M
M
M
M
M
—
—
M
F
F
Results
Neg
Neg.
Neg
Neg
Neg
M & r\
IN cy
M & (~i
1 N tiy
Neg
Neg
Neg.
Neg
Neg
Neg
Positive
Neg
Neg
Neg
Neg
Neg
Neg
Neg
Neg.
Neg
Positive
Neg
Neg
Coccidia
Neg
Neg
Neg
Neg
Kl o <"i
Neg
M o n
Neg
Return City & State
Boulder, Colorado
Boulder, Colorado
Chicago, Illinois
Littleton, Colorado
San Francisco, California
Denver, Colorado
Fort Morgan, Colorado
El Paso, Texas (Dog)
Boulder, Colorado
Denton, Texas
San Francisco, California'
Littleton, Colorado
Denton, Texas
El Paso, Texas
Pasadena, California
Venice, California
Aspen, Colorado
Lakewood, Colorado
Lakewood, Colorado
Aspen, Colorado
Portland, Indiana
Ft Collins, Colorado (Dog)
Snowmass, Colorado
Glenwood Springs, Colorado
Glenwood Springs, Colorado (Dog)
Snowmass, Colorado
Aspen, Colorado
Aspen, Colorado
Rocky Mountain National Park
Following an outbreak of giardiasis in a group of tourists staying at the
Cascade Cabins on Fall River within Rocky Mountain National Park, the
entire length of Fall River from the location of the water siphon for the
Cascade Cabins to its headwaters was examined. Fall River winds its way
through a broad meadow for approximately 8 km (5 miles); above this it is a
very steep, rapid flowing creek in a narrow canyon originating in a cirque
basin. Human use of this entire stretch of river is very heavy. Wildlife in this
-------
R.B.Davies 115
area is varied and consists of beaver, muskrat, bighorn sheep, mule deer,
wapiti, raccoon and coyote. As we have been asked several times if fish could
carry Giardia we captured 10 brook trout from the pond where water for
Cascade Cabins was taken. All were negative. Horses are used infrequently
along the river and only one fresh sample was found; it was negative for
Giardia. One dog sample was found along the river and, even though it
appeared to be at least a day old, it was examined. Giardia cysts were not
found. One human female was observed defecating along Fall River, within
1.5m (5 feet) of the water in a very moist area, in spite of the presence of an
outdoor toilet within 185 m (200 yards) which was in plain view and clearly
marked with signs. This sample was negative for Giardia. Field crews from
the Center for Disease Control (CDC) and Environmental Protection
Agency (EPA) pumped water from Fall River and did detect Giardia cysts in
a sample from a beaver pond. Giardia was not recovered from beaver feces
from this location or upstream from it.
Near the visitor center at the top of Fall River Pass, there is an open
sewage plant for the visitor center. The possibility exists that material from
this sewage plant could leak into Fall River.
Brush Creek
Brush Creek is a tributary of the Eagle River and empties into the Eagle
River near the-town of Eagle. Near Eagle the terrain is fairly flat. As the river
is followed upstream towards the Sawatch Mountains the topography
becomes more severe and the river valley narrows. The slopes are steep but
not as severe as those near Aspen, nor are the ridges as towering. A Colorado
resident with a documented case of giardiasis stated the only untreated water
which he drank was from East Brush Creek. For this reason samples were
collected on East and West Brush Creek. Samples from beaver and cattle
were obtained on West Brush Creek. None of 14 beaver and 1 of 11 cattle
were infected with Giardia. These cattle, about 15-18 animals, were grazing
beside West Brush Creek and they were observed defecating into the creek.
East Brush Creek flows through a canyon which is a bit more rugged than
West Brush Creek. Cattle were not observed in the area but old feces were
found along the river. Of 18 beaver fecal samples picked up below Yeoman
Park Campground, 17% (3 of 18) were positive for Giardia.
East Brush Creek is a fast flowing creek until it reaches the upper portion
where it flows through a large meadow. Two campgrounds, Yeoman Park
and Fulford Cave, have been established here. Along the river near the
Yeoman Park Campground, human feces and toilet paper were found in
several locations. This stretch of river is densely vegetated with willow which
provides cover for people using the river for a toilet.
Middle Park
Middle Park is a large mountain basin in north-central Colorado. The
Continental Divide forms the park boundary on the north, east, and south
sides and the Gore Range on the west. The lowest point in Middle Park is
west of Kremmling where the Colorado River flows through Gore Canyon at
an elevation of 2200 m (7300 feet). Some of the surrounding mountain peaks
-------
116 WATERBORNE GIARDIASIS/EPIDEMIOLOGY
are in excess of 3900 m (13000 feet). The major vegetation type varies from
sagebrush to spruce-fir and alpine. The entire area is heavily used all year for
recreation by residents and many tourists use the area for vacationing.
Spring and summer activities are primarily hiking, backpacking, camping
and fishing. Hunting is a major activity during the fall as is wood gathering.
Hiking, camping and fishing are other major activities, as are snow-mobiling
and skiing in winter.
As stated before, backpackers were the first to complain of giardiasis in
this area. The first area to be examined was the Vasquez River above
Hideaway Park. Several outbreaks in visitors at guest lodges in the area had
been documented. Water for the lodges was taken from the Vasquez River.
This drainage is part of the Denver water collection system and outbreaks of
giardiasis have been documented in resorts using water out of the Denver
water siphon. These resorts use chlorinated but unfiltered water only.
The slopes along Vasquez Creek are fairly steep, and usually heavily
wooded with aspen, spruce and firs. Very little wildlife was observed along
the stream other than a few red squirrels. Higher on the slopes in rocky, dry
ground, pika, marmot, golden mantled ground squirrels and chipmunks
were found. All samples from these animals were negative for Giardia.
Beaver or beaver dams were not seen on the Vasquez watershed. One herd of
wapiti was found on the ridge north of the west branch of Vasquez Creek.
This herd numbered 31 cows and 17 calves. Six fecal samples were picked up
and all were negative for Giardia. Even though this is a fairly rugged area,
humans were seen every day. Feces of humans was not found near any open
water. None of the animal feces examined was positive for Giardia from
these areas.
The area around Hot Sulphur Springs was examined following reports of
giardiasis in area residents. Giardia cysts were found in 2 of 6 bovines near
Corral Creek and 3 of 5 on Cow Creek. As this is dry country, the cattle were
very close to the creeks and were observed defecating into both creeks.
Richardson's ground squirrels were collected within 30m (100 feet) of where
the positive cattle were grazing, and all 5 samples were negative for Giardia.
Since animals positive for Giardia were not found at higher elevations,
streams at lower elevations were examined. Beaver Creek flows into the
Colorado River at the west end of Byers Canyon about 4 km (2.5 miles)
southwest of Hot Sulphur Springs. Several campgrounds and summer
homes are found along Beaver Creek. The stream originates near Church
Park and is heavily utilized by humans. A few cattle and horses are seen
occasionally and deer and elk are seen year around at different elevations.
Beaver samples collected near the campground where Beaver Creek joins the
Colorado were positive for Giardia. Samples were then taken from beaver
dams all the way up Beaver Creek. Beaver samples from approximately one-
third of the dams were positive. All of the positive samples were obtained
from an area extending from the Beaver Creek Campground to the private
homes.
The following spring none of the beaver samples collected along Beaver
Creek was positive but by October, 42% of the samples were positive for
-------
R.B.Davies 117
Giardia. Of these, 11% (1 of 9) of the samples were positive from above the
campground but 52% (14 of 27) at the campground were positive. From
Williams Fork River, where a large percentage of miners was supposed to
be infected with Giardia, 4 beaver samples were negative. Beaver samples
also were collected from Willow Creek, which flows along Colorado
highway 125. Of the samples collected, 23% (9 of 39) contained Giardia cysts.
On the upper reaches of Willow Creek none of the samples were positive.
The first positive samples came from a beaver pond near the highway and
from this location downstream, positive samples were taken from every
beaver dam.
Many campgrounds are located along Willow Creek. Wells at Saw Mill
Gulch and an unnamed picnic area with hand pumps were declared "unsafe"
by the Grand County Environmental Health Department.
Just inside the mouth of Byers Canyon, approximately 400 m (450 yards)
west of Hot Sulphur Springs, two streams flow from the side of the road.
Samples taken from these springs by the Colorado Dept. of Health showed
unsatisfactory coliform results and were positive for Giardia cysts.
RESULTS OF EXPERIMENTAL INFECTIONS
Animals exposed to human-source Giardia produced varied results both
within groups of animals and among groups of animals (Table 4). Hamsters,
domestic rabbits, laboratory mice and deer mice remained uniformly
negative. Attempts to infect calf and adult wapiti, fawn and adult mule deer,
fawn and adult white-tailed deer, black bear, domestic sheep and cattle have
failed. Animals which did become infected and passed Giardia cysts were
laboratory rats, gerbils, guinea pigs, beaver, dogs, raccoons, bighorn X
mouflon sheep and pronghorn. Cysts recovered from these animals ranged
in size from 9.5 to 11.0 /um X 8.0 to 9.5 /Ltm. Cysts from pronghorn and
naturally infected mule deer were often distorted and were about half as wide
as non-deformed cysts.
The first group of rats exposed to human-source Giardia was kept in a
single cage and composite fecal samples were examined. Giardia cysts were
found on 22, 25 and 40 days (Table 4) post exposure (PE) and were of 2 sizes,
one 5/im long the other 10 /nm long, yet identical in all other aspects. The
second group of rats was kept in separate cages.One rat shed cysts for 1 day.
34 days PE. All others were negative on fecal examination and at 40 days
when the animals were sacrificed, trophozoites were not seen in any
intestinal smears. Feces from the infected rat was fed to a beagle pup which
became infected and started shedding Giardia cysts 8 days PE.
A total of 4 groups of gerbils (each group consisting of 4 exposed animals
and 1 in-group control animal) were exposed to human-source Giardia. In
the first group, 3 of the 4 exposed animals became infected between 13 and 18
days PE (Table 4). Thirty-three days PE the control animal had Giardia cysts
in the feces. The remaining exposed animal remained negative for 45 days.
One gerbil from 2 additional groups was positive for GiardiaS days PEfor 1
day. The last group of gerbils remained negative for Giardia for 42 days PE.
Trophozoites were not seen in necropsied animals. One group of guinea pigs
-------
118 WATERBORNE GIARDIASIS/EPIDEMIOI.OGY
Table 4 - Results of experimental infections mmg 'human (Homo sapiens) source Giardia.
Experimental Animal*
Results
Hamster (Mesocncetus auralus)
(4+1)
Domestic rabbit (Orycto/agus
cunicu/us) (4+1)
Laboratory mice (Mus muscu/us)
4 groups of (4+1)
Deer mice (Peromyscus manicu-
latus) 2 groups (4+1)
Rats (Rattus norvegicus)
(4+1)
(4+1)
Gerbils (Gerbillus gerbillus)
(4+1)
2 groups(4+1)
(4+1)
Guinea pig (Cavis porcellus)
(4+1)
Beaver (Castor canadensis)
(2 + 1)
(2+1)
Dog (Cams fami liar is)
(2 + 1)
(2+1)
Raccoon (Procyon lotor)
(3+1)
(3+1)
(3+1)
(D
Black bear (Ursus amer/canus)
(D
Negative - 34 days
Negative - 28 days
All negative - 40 days
Negative - 22 days
Composite fecals - positive - day
22, 25 and 40 (Cysts of 2 sizes)
One rat positive 34 days post-
exposure for 1 day
#1 positive day 22,26-29,32-33
#2 positive day 13,19-22,26-28
#3 Control-positive day 33-35
#4 Positive day 18-21,26
#5 Negative 45 days
One gerbil was positive for 1 day
8 days post-exposure
Negative - 42 days
One animal became positive on
day 21 and remained positive for
31 days All others were
negative
Negative 38 days
Both positive 25 days after
exposure, cysts seen for 22 days
Positive 6 days post-exposure for
14 days Animals treated
Positive 8 days post-exposure for
14 days Animals treated
Negative - 28 days
Negative - 43 days
Negative - 34 days
Positive 8 days post-exposure for
1 day - Negative for the next 20.
Negative - 28 days
-------
R. B. Davies
119
Table 4 - continuation . . .
Experimental Animal*
Results
Bighorn X mouflon sheep
(Ovis canadensis X O
musimon)
(4+1)
Wapiti (Cervus canadensis)
(2+1)
(2 + 1)
Mule deer (Odocoileus
hem/onus)
(2+2)
(4+2)
(3 + 6)
Prong horn (Antilocapra
americana)
(1 + D
Domestic sheep (Ovis ar/es)
(3+1)
Domestic cattle (Bos larus)
(4+1)
White-tailed deer (Odocoileus
virginianus)
(1+2)
(1+6)
#1 - Negative
#2 - Negative
#12 - Control - Negative
#29 - Positive 9 days for 4 days
#44 - Positive 9 days for 4 days
Negative - 30 days
Negative - 28 days
Negative - 28 days
Negative - 30 days
Negative - 22 days
Positive days 16-18 post-
exposure
Negative - 45 days
Negative - 45 days
Negative - 30 days
Negative - 22 days
*number of infected ammah followed by number of in-group controls
was exposed to human-source Giardia (Table 4). One animal had Giardia
cysts in the feces 21 days PE and remained positive for 31 days. As each
guinea pig was held in a separate cage no chance for transmission between
animals existed. The remaining 4 animals were negative 58 days PE.
Three beaver were maintained in captivity and all were negative for over
40 days before 2 were exposed to human-source Giardia. The control animal
and 1 previously exposed animal were each given 10,000 human source
Giardia cysts. Both of these animals had Giardia cysts in their feces 25 days
PE and remained positive for 22 days. Both animals then became negative
for 7 to 15 days after which they again were positive. This pattern was
followed for approximately 3 months. The control animal remained
negative.
Two groups of 2 dogs each were exposed (Table 4) to human-source
Giardia. In one group both infected dogs became positive at 6 days PE, and
-------
120 WATERBORNE GIARDIASIS/EPIDEMIOLOGY
the in-group control was positive 7 days later. The second group of infected
dogs started shedding Giardia cysts 8 days PE and the in-group control
became infected 7 days later. In both cases control animals kept in the same
room remained negative.
A total of 10 raccoons was infected with Giardia from a human source
(Table 4). One young raccoon became positive for Giardia 8 days PE for 1
day and only 3 cysts were found. One black bear cub (age about 6 months),
exposed to human-source Giardia, remained negative for 28 days.
Four bighorn X mouflon sheep were given human-source Giardia. Two of
the 4 plus the in-group control remained negative (Table 4). The other 2 shed
Giardia cysts from day 9 through 13 PE and both had loose stools during this
time. Feces from both animals were collected and given to beagle puppies.
The pup given Giardia cysts from sheep #29 became positive 10 days PE yet
Giardia cysts from sheep #44 failed to infect a pup. Control dogs remained
negative. One pronghorn fawn (2.5 months old) was positive 16, 17 and 18
days after exposure. He remained negative for an additional 10 days.
Control animals remained negative throughout this time.
Giardia cysts isolated from beaver feces collected from beaver ponds on
Beaver Creek near Hot Sulphur Springs, Colorado were given to laboratory
mice, rats, guinea pigs, and hamsters, all of which remained negative (Table
5). Cysts from the same source were given to beagle puppies and, 8 days later,
cysts were found in the feces of 4 of the infected pups but not the in-group
Table 5. Results of experimental infections using Beaver (Castor canadensis) source Giardia
Experimental Animal* Results
Mice (Mus muscu/us)
2 groups (4+1) All negative
Rats (Rattus norvegicus)
2 groups (4+1) All negative
Hamsters (Mesocricetus auralus)
(4+1) Negative - 30 days
Guinea pig (Cavia porcellus)
(4+1) Negative - 30 days
Dog (Cams familiaris)
(4+1) All 4 exposed animals positive 8
days after exposure - control
positive 9 days later - allpositive
21 days post-exposure
Human (Homo sapiens)
(3+1) #1 Positive 1 1 dayspost-exposure
#2 Positive 6 days post-exposure
#3 Negative-subject was on
Tetracyclme therapy
* Number of exposed animals followed by number of m-group control animals.
-------
R.B.Davies 121
control. The control animal was positive 9 days after cysts were found in the
4 experimental animals.
Cysts from beaver were also given to 3 human volunteers. One of the
volunteers had cysts in his feces 6 days PE, another at 11 days PE. The third
volunteer remained negative as did the control. It was later determined the
third volunteer was on tetracycline therapy.
Giardia cysts isolated from 1 of 4 naturally-infected, captive mule deer
fawns were given to 2 beagle puppies, which remained negative, and 1 human
volunteer who started shedding Giardia cysts in the feces 9 days after
exposure (Table 6).
Table 6 - Result;, of experimental infections using Mule deer
(Odocoileus hemionus) source Giardia
Experimental Animal' Results
Dog (Cams fami/iaris) Negative 28 days
(2 + 5)
Human (Homo sapiens) Positive 9 days
(1) post-exposure
* Number of exposed animals followed by number of in-group controls
Cyst output varied greatly in the animals. The number of cysts would be
very high one day and then quite low the following day. Dogs and beaver
produced the most consistent numbers of cysts.
DISCUSSION
Giardiasis is a growing problem in Colorado and will probably remain so
for many years. An ever increasing number of people and their pets using the
outdoors for recreation, lack of sanitary facilities and reluctance to use them,
inadequate sewage and water treatment systems are important factors in the
epidemiology of sylvatic giardiasis. There is no doubt that animal reservoirs
exist and contribute to the problem. However, except for dogs and cats,
infected animals were not found in the spring. Beaver and cattle were not
found positive for Giardia until mid to late summer and fall. This suggests
that these animals lose their Giardia infections over winter and become
reinfected the following summer.
Since prevalence of Giardia was high and beaver became infected with
human Giardia, beaver may be an excellent sentry animal for Giardia. One
such instance was found on the Eraser River where all beaver fecal samples
examined above the sewage treatment plant at Eraser, Colorado, were
negative for Giardia while beaver directly below the treatment plant were
positive. Thus, the sewage plant was probably a source of Giardia in the
Eraser River.
Beaver are an important reservoir of Giardia since they carry the infection
for at least 3 months and defecate into the water where Giardia cysts survive
very well. People often drink water from beaver dams and often collect water
-------
122 WATKRBORNE GIARDIASIS/EPIDEMIOLOGY
as it flows over the dam, not realizing beaver defecate on the other side of the
dam.
Cattle tend to congregate around water because much of their grazing
range is dry. They often defecate around or in water and also walk through
feces getting to the water. For this reason and the fact that Giardia was found
in them, cattle become important as reservoirs.
Dogs accompany man almost every where he goes including his
recreational trips. Dogs also have habits of coprophagy and a desire to roll in
feces. These activities increase the animal's chance of becoming infected with
Giardia and bringing the cysts back to their owners on the haircoat. Dogs
have been observed defecating near open streams; people, not wanting to see
or step in the feces, proceed to kick the feces into the stream.
Cats are probably not very important in sylvatic giardiasis away from the
home. Cats do not travel with backpackers and only rarely with campers.
Near home, however, a cat could become infected while out of the house and
then return home to infect its owners. Two such instances are known; a
family of 5 in Fort Collins had giardiasis and the family cat was shedding
large numbers of Giardia cysts in a normal appearing stool. A resident of
Hideaway Park, Colorado and her boyfriend both had severe giardiasis and,
when a stool from the female's cat was examined, large numbers of Giardia
cysts were found. The question remains as to whether the cats infected the
people or the cats obtained the infection from the people.
Humans are the most important component in the epidemiology of
giardiasis. The Snowmass-Maroon Bells Wilderness area has been aptly
described as "one of the world's largest open air toilets". People have been
observed defecating near streams with outdoor toilets in sight. Perhaps they
think this activity is a necessary part of the wilderness experience. Diapers
containing feces have been found in lakes and streams and there are reports
of people emptying recreational vehicle (RV) toilets into rivers. In each study
area a common complaint of county officials is the lack of sewage disposal
by owners of mountain homes. Instances of direct dumping of sewage into
streams have been found. The very limited survey of people going into the
Maroon Bells-Snowmass Wilderness Area showed people already infected
with Giardia are going into the back country. As the trails in the area follow
streams, it is easy to visualize how contamination of these streams, which
always lead toward a more populated area, can occur. People, especially day
hikers, generally are not in physical condition for strenuous hiking and these
people drink large amounts of water from the streams near the trails.
Wild ruminants, common inhabitants of mountainous watersheds, were
not infected with Giardia. Four cases of Giardia infection in captive mule
deer fawns were found in 1977; all deer had severe diarrhea for over 2
months. These animals were all orphans which were being hand-raised along
with other uninfected mule deer. One of the infected fawns died 3 weeks after
being treated for Giardia and being Giardia-irze for the 3 weeks. At necropsy
a marked absence of thymic tissue was noted. As a result of the thymic
atrophy this animal, and most likely the other 3 , was immunologically
incompetent and, therefore, susceptible to Giardia.
-------
R.B.Davies 123
Measurements of cysts recovered from experimentally infected animals
were within the limits for G. lamblia cysts(6) except for the small cysts which
were recovered from the first group of rats infected with human-source
Giardia. It is possible 2 species of Giardia were present, but since
trophozoites were not recovered, this question remains unanswered.
In several instances only 1 or 2 animals of a group would shed cysts after
infection. This is most likely due to the individual animals immune status at
the time of exposure. The immune response may also dictate the invasiveness
of the cysts shed by the host. A possible example is the bighorn sheep X
mouflon hybrid experiment in which 2 of the 4 infected animals shed cysts
and cysts from only 1 of the sheep were infective to beagle pups. The dog
which remained negative was later infected with human-source Giardia and
cysts were found in the feces 9 days after infection. Rats, hamsters, and
guinea pigs showed similar results. The host immune response may also be
responsible for some of the variability in cyst production which was
observed in infected animals.
It is possible that some of the animals were infected but, because they were
shedding only small numbers of cysts which were not recovered with the
ZnSO4 method, they were termed negative. However, Giardia cysts or
trophozoites were not recovered at necropsy from animals negative for
Giardia cysts using the ZnSO4 method.
It appears that human-source Giardia will not infect hamsters, laboratory
mice, deer mice, mule deer, wapiti or white-tailed deer. Inconsistent
infections resulted in rats, gerbils, beaver, guinea pigs, bighorn X mouflon
sheep and raccoons given human-source Giardia cysts. Additional infections
of domestic sheep and cattle, pronghorns and black bear must be completed
before definite conclusions are made concerning their ability to become
infected with human-source Giardia.
In most cases (pronghorn, raccoon, rats .and bighorn X mouflon sheep)
the patent period lasted only a few days. Gerbils had a longer patent period
but cysts were not detected every day (Table 4). The longest patent periods
were observed in beaver (at least 22 days) and beagle pups (at least 3
months). Shedding of cysts by beaver became somewhat cyclic after 22 days,
possibly reflecting reinfection or a true latent period. Dogs were monitored
for Giardia cyst production and while the number of cysts shed varied
greatly, the animals remained infected for over 3 months (unpublished data).
Infection of animals with Giardia cysts from beaver paralleled the results
of human-source Giardia in that mice and hamsters did not become infected.
Guinea pigs and rats did not become infected which was unlike the results
obtained using human-source Giardia. Beagle pups did become infected and
continued to shed cysts for 21 days when the dogs were terminated. In
humans 2 of 3 also became infected after ingesting beaver-source Giardia
cysts. The third person was on a regimen of 250 mg tetracycline 3 times per
day for a non-related infection. Tetracycline will alter the intestinal pH and
may have prevented establishment of Giardia.
-------
124 WATERBORNE GIARDIASIS/EPIDEMIOLOGY
The results from infections with rriule deer-source Giardia are confusing.
It is not known why these cysts resulted in infection of a human, but not in
beagle pups which were successfully infected with human-source Giardia 30
days later.
Human-source Giardia displayed a remarkable shelf life as long as the
cysts were cleaned and stored in distilled water in a refrigerator. Cysts were
stored for as long as 3 weeks and still resulted in patent infections in dogs. It
is not recommended that cysts be held this long as their internal structures do
break down.
Cysts which were distorted were recovered from mule deer with a natural
infection and a pronghorn with a human-source infection. The distortion
consisted of a constriction of the cyst wall and the trophozoites pulling away
from the cyst wall. When the feces from the animals was soft or diarrheicthe
cysts were not distorted but as the feces became formed and dryer distorted
cysts were found.
This study has presented prevalence data for a variety of domestic and
wild animal hosts which have not previously been studied for Giardia in
Colorado. Evidence is also provided that beaver and cattle are reservoirs for
sylvatic giardiasis in Colorado.
REFERENCES
1. Faust, E. C., P. F.Russell and R.C.Jung. 1974. Craig and Faust's Clinical Parasitology.
Lea and Febiger, Philadelphia, pp. 64-66
2. Stiles, M. A. and L. A. Kretchman. 1974. Giardiasis and the adult traveler. Postgrad. Med.
56(4): 134-138.
3. Petersen, H. 1972. Giardiasis (Lambliasis). Scan. J. Gastro. 7 (supplement 14).
4. Anonymous. 1978. Intestinal parasite surveillance. United States, 1976. MMWR 27(20):
167-168.
5. Anonymous. 1977. Giardiasis- California, Colorado. MMWR 26(7). 60.
6. Levine, N. D. 1973. Protozoan parasites of domestic animals and man. Burgess Publishing
Co., Minneapolis, pp. 118-122.
7. Payne, F. J., F. O. Atchley, M. A. Wasley and M. E. Wenning. 1960. Association of
Giardia lamblia with disease. J. Parasitol. 46: 742.
8. Babb, R. R., O. C. Peck and F. G. Veseia. 1971. Giardiasis: A cause of travelers diarrhea.
J.A.M.A. 217(10): 1359-61.
9. Schultz, M. G. 1975. Giardiasis. J.A.M.A. 233(13): 1383-1384.
10. McGrath, J., P. T. O'Farrell and S. J. Boland. 1940. Giardial steatorrhea: A fatal case with
organic lesion. Irish J. Med. Sci. (Dec. 1940): 802-816.
11. Gangarosa, E. J.and J. A. Donadio. 1970. Surveillance of food borne disease in the United
States. A comparison of data-1968-1969. J. Inf. Dis. 122(4): 354-358.
12. Gleason, N. N., M. S. Horwitz, L. H. Newton and G. T. Moore. 1970 A stool survey for
enteric organisms in Aspen, Colorado. Am. J. Trop. Med. and Hyg. 19(13): 480-484.
13. Thompson, R. G., D. S. Krandiker and J. Leek. 1974. Giardiasis - An unusual cause of
epidemic diarrhea. Lancet 1(858): 615-616.
14. Gietzan, T.J N. S. Hayner, P. Landis, T. M. Vernon, and D. O. Lyman. 1978. Giardiasis-
Vail, Colorado. MMWR 27(19): 155.
15. Hibler, C. P., K. MacLeod and D. O. Lyman. 1975. Giardiasis - In residents of Rome, New
York. MMWR 24(43): 366.
16. Wright, K. W., R. W. Bacorn, V. Smith and D. O. Lyman. 1975. Giardiasis -U.S. travelers
to the Soviet Union. MMWR 24(43):366,371.
17. Juranek, D. D. 1978. Personal communication.
18. Anonymous. 1978. Weekly Disease Summary. Colo. Dis. Bull. 6(34): 2.
19. Mayer, W. T. 1973. Epidemic giardiasis. Rocky Mtn Med. J. 70(10):48-49
20. Vernon, T. M. 1976. Personal communication.
21. Wright, R. A. and T. M. Vernon. 1976. Epidemic giardiasis at a resort lodge. Rocky Mtn.
Med. J. 73(4): 208-211.
-------
R. B. Davies
125
22. Grant, D. R. and P. T. K. Woo. 1978 Comparative studies of Giardia spp. in small
mammals in southern Ontario. II. Host specificity and infectivity of stored cysts. Can. J.
Zool. 56- 1360-1366.
23. Padchenko, I. K. and N. G. Stolyarchak. 1969. On the possible circulation of Lamblia in
nature. Prog. Protozool. 3:311-312.
24. Filice, F. P. 1952. Studies on the cytology and life history of a giardia from the laboratory
rat. Univ of Calif. Publ. Zool 57(2): 53-146.
25. Armstrong, D. M. 1972. Distribution of mammals in Colorado. Mus. Nat. Hist , Univ.
Kansas. Monogr. No 3. 415 pp.
APPENDIX A.
SCIENTIFIC NAMES(25) OF ANIMALS
LISTED IN TABLES 1 AND 2.
Badger
Beaver
Bighorn sheep
Black bear
Bobcat
Bovine
Brook trout
Colorado chipmunk
Cottontail rabbit
Coyote
Dog
Domestic sheep
European ferret
Golden-mantled ground squirrel
Horse
Human
Marmot
Mountain lion
Mule deer
Muskrat
Pika
Pine fnarten
Porcupine
Raccoon
Red fox
Red squirrel
Richardson's ground squirrel
Rock squirrel
Striped skunk
Wapiti
White-tailed deer
White-tailed jackrabbit
White-tailed prairie dog
Taxidea taxis
Castor canadensis
Ovts canadensis
Ursus amencana
Lynx rufus
Bos taurus
Salvehnus fontmalis
Eutamias quadnvitlatus
Sr/vilagus spp
Canii latrans
Can is familiar is
Ovis aries
Mustela putonus
Spermophilis lateralis
Equus caballus
Homo sapiens
Marmola flaviventris
Fe/is concolor
Odocodeus hemtonus
Ondatra zibelhica
Ochotona pnnceps
Manes amencana
Erelhizon dorsatum
Procvon lotor
Vulpes fulva
Tamiascmrus hudsomcus
Spermophilus nchardsoni
Spermophilus vanegatus
Mephitis mephitis
Cervus canadensis
Odocoileus virgmianus
Lepus townsendn
Cvnomes leucurus
Discussion
R. FREEMAN: The uniqueness of the mountain situation you referred to
intrigues me. For some 20 years I have worked in Canada. We certainly have
a tremendous amount of water routes which people follow and I am sure
they are drinking untreated water just as much out of the streams, as you
have experienced people drinking water in the mountains. What is so unique
about your situation that would not apply elsewhere in North America? I am
sure our physicians are equally qualified in diagnosing and if this condition
-------
126 WATKRBORNK GIARDIASIS/EPIDKMIOLOGY
were particular to certain areas only, physicians would be informed very
quickly.
Could it be that there is not enough sedimenting in areas with a great deal
of flowing water? We have flowing water in Canada, but most of our rivers
are short, ending in a deep lake where there is a lot of sedimenting. We
certainly have very thin soil, so it is not the same situation you have in
Colorado. However, due to the influx of Americans into Canada, 1 am sure
we have as much Giardia cyst potential. 1 do not understand the difference in
the number of cases found in your studies compared to those in Canada.
R. DAVIES: I have no idea what is so unique unless we have a great deal
more travelers who report their infections. Bill Samuel from the University
of Edmonton, says that Giardia cases around there are extremely high, and
most of his graduate students have developed Giardia infections. Possibly
giardiasis is being examined more closely in Colorado than anywhere else
and thus the numerous outbreaks of giardiasis appear to be unique.
L. McCABE: I would think the Canadian situation might not be unique. I
am told that the International Nickel Company lowered the pH to 3.5 in the
lakes receiving fallout from their stacks.
SPEAKER: Mr. Davies, would you please reiterate the material shown on
the last slides.
R. DAVIES: We took Giardia that we isolated from beaver found in the
Beaver Creek area. This was given to 4 beagle pups and they developed the
disease. We also exposed 3 humans. One became positive at 6 days post-
inoculation and 1 at 11 days post-inoculation. The third remained negative,
but he was on tetracycline therapy, which I have been informed can alter the
pH of the gut. This possibly prevented Giardia from becoming established.
D. JURANEK: How well screened were the individuals who were taking
Giardia cysts?
R. DAVIES: Since we have been working with Giardia, periodically we all
check ourselves to determine if we have become infected and are carriers.
Stool samples taken before our experiments began were negative, with the
exception of that from Dr. Hibler who had a previous Giardia infection. He
did become positive again.
J. HOFF: Were the cysts recovered from these various animals
morphologically similar to lamblia, or were there different morphological
types of cysts?
R. DAVIES: All of the cysts we have recovered from wild, free ranging
mammals have been morphologically similar, if not exact, to lamblia, and
the cysts all fall within the limits of those published for G. lamblia, including
everything obtained from the experimental infections.
W. JAKUBOWSKI: Have you determined whether or not there is a
difference in the percent of positive animals in areas receiving a lot of human
use as opposed to those areas receiving very little or no human use?
R. DAVIES: That is difficult to say. The areas that we examined were
chosen because of the heavy human use and the high risk in those areas. The
places where we did find animals which were not infected were along the
upper Colorado River. There is a lot of use by people who primarily are just
taking a little hike down to the river and back. Backpackers do not use this
area. I was surprised that we did not find anything positive there or up the
Williams Fork, an area used by many miners.
-------
127
Waterborne Outbreaks of Giardiasis
Gunther F. Craun
U. S. Environmental Protection Agency
Health Effects Research Laboratory
Field Studies Division,
Epidemiology Branch
Cincinnati, Ohio
ABSTRACT
Data are presented on waterborne outbreaks of giardiasis that have affected travelers to
foreign countries, particularly Americans to the Soviet Union, and that have been
documented in the United States, including 2 retrospective epidemiologic studies of
endemic giardiasis associated with consumption of untreated water.
During the past 5-year period, the most commonly identified pathogen in waterborne
outbreaks in the U mted States has been G. lamblia. Twenty-three waterborne outbreaks of
giardiasis affecting 7,009 persons have been reported in the United States since the
outbreak in Aspen, Colorado in 1965-1966. These have occurred primarily in mountainous
areas of this country particularly in New England, the Pacific Northwest, and the Rocky
Mountains and have generally involved small municipal water systems or semi-public
water systems in recreational areas. Most of the outbreaks occurred as the result of
consuming untreated surface water or surface water treated only with chlorine.
The data from waterborne outbreaks of giardiasis in the United States indicate that
simple disinfection as the only treatment for surface water sources is ineffective in
preventing waterborne transmission of this infection and that to protect against
transmission, all surface water should receive pretreatment, preferably with sedimentation,
and filtration in addition to chlormation.
The data also indicate that negative results of cohform tests do not provide assurance
that the water is free of Giardia cysts. In several outbreaks where disinfection was provided,
Giardia cysts were found in the absence of coliforms Since disinfection practices in surface
water systems where chlormation is the only treatment do not generally provide for high
concentrations of chlorine or long contact time, it is likely that Giardia cysts would survive
treatment whereas coliforms would not.
Giardia lamblia is a flagellated protozoan of the small intestine. In the past
its pathogenicity has been questioned; however, there is increasing clinical
and epidemiologic evidence that G. lamblia is a cause of acute illness and a
significant pathogen of man. Clinical manifestations of Giardia infection
can range from asymptomatic cyst passage to severe malabsorption.
Prominent symptoms include diarrhea, abdominal cramps, fatigue, weight
loss, flatulence, anorexia, and nausea. An incubation period of 1 to 8 weeks
is typical, and the mean duration of acute illness is often 2 to 3 months(l-4).
Giardiasis is endemic in the United States and in many other countries of
the world; however, only recently has G. lamblia been recognized as the
etiologic agent in several common source outbreaks. Although many
questions remain to be answered about this infection, there is much that has
been learned in recent years through epidemiologic studies of the endemic
and epidemic occurrence of giardiasis.
-------
128 WATERBORNE GIARDIASIS/EPIDEMIOLOGY
This report will review the data on waterborne outbreaks of giardiasis that
have affected travelers to foreign countries, particularly Americans to the
Soviet Union, and that have been documented in the United States,
including 2 retrospective epidemiologic studies of endemic giardiasis
associated with consumption of untreated water.
In discussing an outbreak of amebiasis that affected 150 American
occupants of aTokyo apartment building during December 1946to January
1947, Dayis and Ritchie(5) provided some early epidemiologic evidence of
the waterborne transmission of G. lamblia. The outbreak was attributed to
sewage contamination of the building's water supply, and Entamoeba
histolytica was recovered from 96 (64%) of the occupants and G. lamblia
from 116 (77%). There were 26 occupants who experienced diarrhea with
abdominal discomfort but whose stools were found to be negative for E.
histolytica. G. lamblia was found in the stools of 20 (80%) individuals in this
group suggesting that this organism may have been responsible for some of
the illness in this outbreak.
WATERBORNE GIARDIASIS IN TRAVELERS
The first reports of epidemic giardiasis among travelers to the Soviet
Union appeared in 1970 when 2 outbreaks affecting American travelers were
investigated by the Center for Disease Control (CDC)(6, 7). Twenty-three of
80 (29%) persons who had accompanied the U.S. Olympic Boxing Team on
a tour of the Soviet Union in February and March 1970 and 84 of 164(51%)
scientists visiting in May 1970 had become ill with giardiasis. G. lamblia was
found in fecal specimens from 9 patients and 2 asymptomatic persons in the
first group and in 23 patients and 3 asymptomatic persons in the second
group. The illness usually began toward the end of the trip or shortly after
return to the United States, and the principal symptoms included nausea,
abdominal cramps, diarrhea with stools that were frequently greasy and
malodorous, flatulence, anorexia, and fatigue. Retrospective epidemiologic
study of these outbreaks implicated Leningrad as the site of acquisition of
the infection and tap water as the probable mode of transmission.
Reports of epidemic giardiasis among American and other tourists to
Leningrad steadily increased after these 2 outbreaks, and the resulting
retrospective and prospective epidemiologic studies have confirmed the
relationship between epidemic giardiasis in tourists and the consumption of
tap water in Leningrad(8-19).
The CDC collected information by questionnaire on 1,419 members of 47
individual tour groups to the Soviet Union from 1969-1973(8, 15). Data were
obtained from tour groups who came to the attention of CDC because of
cases of giardiasis and from lists of tour members provided by travel agents
specializing in travel to the Soviet Union. Each traveler submitted
information concerning age, sex, illness, symptoms, place of lodging, cities
visited, and exposure to food and water. Questionnaires showed that 502
(36%) of these travelers were ill during the tour or shortly after their return to
the United States. Defining a case of giardiasis as a positive stool
examination and/or a diarrheal illness lasting for 1 week or longer, 324 of
-------
G.F.Craun 129
the 1,419 (23%) travelers were counted as cases. Rates of attack of giardiasis
according to cities visited showed Leningrad to be the only city implicated.
No specific hotel in Leningrad could be related to increased risk of infection.
Histories of tap water consumption implicated drinking water in Leningrad
as the likely source of infection. Of 1,082 persons who drank tap water, 274
(25%) developed giardiasis; of 297 who did not drink tap water, 44 (15%)
became infected. In 1 tour group of 400 where tap water consumption was
quantitated, risk of infection increased with the amount of tap water
consumed per day(8). Other food items such as ice cream, uncooked
vegetables, and fruit were not statistically related to an increased risk of
infection.
Epidemiologic study of another outbreak of giardiasis affecting 72 of 179
(40%) American travelers to the Soviet Union in March 1975 where tap
water was found to be significantly related to acquiring the infection also
found a positive correlation between the amount of water consumed and the
likelihood of becoming infected(18).
A prospective study of 96 staff members of the National Aeronautics and
Space Administration (NASA) who traveled to the Soviet Union in 1973 was
also conducted by CDC(8). Stool samples were examined for ova and
parasites before the tours left the United States and after their return. None
of the 96 NASA members had a Giardia-posiilve stool specimen before
travel to the Soviet Union. Twenty-one individuals traveled to Leningrad
and 17 (90%) reported illness; 75 individuals did not travel to Leningrad and
only 3 (4%) reported illness. Stool examinations after return to the United
States showed that 15 of the 17 (88%) individuals who traveled to Leningrad
and reported illness were positive for G. lamblia.
Another prospective study of travelers to the Soviet Union was conducted
by Jokipii and Jokipii(17) and included 60 students from Finland who
visited Leningrad in 3 separate tours in 1971 and 1973. Each group stayed at
a different hotel while in Leningrad. Stool specimens were examined 1 to 2
days before departure and upon return to Finland. All reported on use of tap
water and symptoms of illness. Giardia cysts were recovered from 2 students
prior to the trip. Both of the students had visited Leningrad within the past
year. Fifteen of 43 (35%) individuals who were negative prior to the trip had
stool specimens which were positive for Giardia cysts after their return from
Leningrad. The recovery of (7. lamblia from stools correlated with the
occurrence of symptoms and with the consumption of tap water in
Leningrad.
Giardiasis had not previously been reported in residents of Leningrad;
however, limited reports of giardiasis in Leningrad residents are now
available (see Wolfe, these Proceedings). It is difficult to assess if this is a
widespread problem among residents, and the question remains as to why
many tourists become infected through consumption of tap water while
residents apparently do not. If this is indeed the case, it has been suggested
that 1 or more of these possibilities might be responsible:
1. Water supplies for various hotels in Leningrad may be separated
from the supplies for residents.
-------
130 WATERBORNE GIARDIASIS/EPIDEMIOLOGY
2. Giardiasis may not be considered by local physicians in the
differential diagnosis of chronic diarrhea among residents.
3. There may be an acquired immunity among Leningrad residents as a
result of repeated exposure over the years.
Although epidemiologic evidence has clearly implicated Leningrad as the
site of infection and tap water as the probable vehicle of transmission of
giardiasis to tourists, it has not been possible to obtain data directly on water
supplies and treatment or water samples from Leningrad. There is 1 report in
the literature, however, that does offer limited information on the water
source and treatment for parts of Leningrad. Ryan and Grainge(20)
participated in an international conference of sanitary engineers held in the
Soviet Union and were able to visit 1 of the 5 water-treatment plants serving
the city of Leningrad. The Neva River is used as the source of water and the
raw water is reported to be of relatively good physical, chemical, and
bacterial quality. As reported by Ryan and Grainge in mg/1, the turbidity of
the raw water is generally 3-7 mg/1 except during spring floods when it is as
high as 70 mg/1. The bacteriological quality of the raw water ranges from 1 to
10 coliforms per liter. Apparently, because the raw water quality is so good,
many industries draw water directly from the river without treatment. It is
not known if the various hotels use river water directly, but it is suspected
that they do not because of the availability of several water-treatment plants
in the city. The water-treatment plant visited had 2 separate sections. One
was a conventional system and was not discussed. The second treatment
process was described as follows: prechlorination of 1.6 mg/1, screening,
rapid sand filtration, aeration for 6-8 min, ammonia feed of 1.2 to 1.6 mg/1,
alum feed, rapid mixing, settling, and fluoridation. Upward-flow filtration
rather than gravity filtration was utilized because the upward filtration
results in longer filter runs with colder water temperatures. For gravity
filters, it was found that a filtration rate of 7 to 8 m/h could be maintained
when the water was 20° C but only 4m/hatl°C.With upward filtration the 7
to 8 m/h could be maintained regardless of water temperature. The sand in
the filters is 2 m thick and of 0.5 to 2.0 mm sizes. It was reported that the
treated water contains a turbidity of 1.6 mg/1 and that coliforms in the
distribution system average 1/1. Although this is an incomplete description
of the water treatment, it does provide some basis for questioning the
adequacy of treatment and operation to protect against the waterborne
transmission of giardiasis. Effective disinfection of a water supply with
chlorine depends on the type of chlorine residual, the factors of contact time,
pH, water temperature, and the presence of suspended material(21,22). The
application of chlorine and ammonia, as practiced in this treatment plant,
produces chloramines or a combined available chlorine rather than a free
available chlorine for disinfection. Combined available chlorine forms are
less effective disinfectants. About 25 times as much combined available
residual chlorine is necessary to obtain equivalent bacterial kills as free
available residual chlorine under similar conditions of pH, water
temperature, and contact time; about 100 times longer contact time is
required to obtain equivalent bacterial kills for similar concentrations at the
-------
G.F.Craun 131
same pH and water temperature. It is significant to note that minimum
chlorine requirements for destruction of E. histolytica cysts are greater than
for bacterial inactivation and that diminished effect at a reduced water
temperature is more pronounced even under the best conditions of
operation. For example, the minimum recommended cysticidal free
available chlorine residual(22) after 30 min contact time at pH 7 for a water
temperature of 22 to 25°C is 3 mg/1 compared to 10 mg/1 under similar
conditions at 2 to 5°C. Although coliform levels are reportedly low in the
Leningrad water distribution system, it is certainly possible that sufficient
contact times and/ or concentrations of combined available chlorine are not
being provided to inactivate Giardia cysts, especially in the cold water
temperatures encountered during winter and early spring. The comparision
of turbidity values in raw and treated water indicate poor operation of the
filtration process possibly related to inadequate control of chemical
pretreatment of the raw water or the high filtration rates reported for the
upflow filters. This suggests that the filtration provided at certain times
would not be effective in removal of Giardia cysts.
While Leningrad is one location where there seems to be a continuing risk
of travelers acquiring giardiasis through drinking water, sporadic
waterborne outbreaks of giardiasis can occur elsewhere in the world. An
example occurred in October 1976 when reports of a high incidence of
diarrhea in a group of 1,400 Americans who had vacationed at the
Portuguese island of Madeira were received upon their return to the United
States(23). A mail questionnaire survey suggested waterborne giardiasis as
the etiology of the outbreak. Of 859 respondents, 27% developed an illness
resembling giardiasis, that is, diarrhea of 1 week duration or longer or
diarrhea of shorter duration but accompanied by abdominal distention. G.
lamblia was recovered from 27 of 58 (47%) ill patients who had a stool
examination for parasites; E. histolytica was isolated from 3 (5%) persons.
Drinking tap water was statistically associated with illness as was
consumption of ice cream and raw vegetables; however, the illness was felt to
have been most likely acquired through consumption of tap water because
contamination of ice cream and raw vegetables by tap water could not be
ruled out. It was reported that the drinking water was chlorinated, but
additional information on water source and other treatment was not
available. A follow-up survey of 90 Americans traveling there in the Spring
of 1977 showed only 4.5% developing an illness compatible with giardiasis
suggesting that the October outbreak was an isolated event rather than a
reflection of an ongoing problem.
ENDEMIC WATERBORNE GIARDIASIS IN THE UNITED STATES
Giardiasis occurs in both epidemic and endemic forms in the United
States, and studies in Colorado and Minnesota suggest consumption of
untreated drinking water can be an important cause of endemic infection.
Parasitologic surveys have shown G. lamblia to be the most frequently
identified parasite in Colorado. For example, a retrospective laboratory
survey conducted in 1973 in Colorado showed G. lamblia to be present in
-------
132 WATERBORNE CIARDIASIS/EPIDEMIOLOGY
40% of all positive examinations for ova and parasites(24). The majority of
infected residents had experienced an episode of chronic watery diarrhea
with a median duration of 3.8 weeks, bloating, flatulence, and weight loss
averaging 5.1 kg. A statewide telephone survey was also conducted of 256
infected residents identified from these laboratory records and 256 controls
matched by age, sex, race, and place of residence. An analysis of the data
obtained showed: (a) an increased incidence of giardiasis in persons between
the ages of 16 and 45 with males and females equally affected; and (b) a
higher proportion of cases than controls who visited Colorado mountains
(69% vs 47%), camped out overnight (38% vs 18%), and drank untreated
mountain water (50% vs 17%). Person-to-person transmission was found to
be relatively infrequent and the association between giardiasis and drinking
untreated mountain water was strengthened by the fact that other potential
sources of infection such as home water sources, swimming in unchlorinated
pools, domestic animal exposure, or out-of-state or foreign travel were not
associated with G. lamblia infections. Bacteriologic examination of water
samples from 16 mountain streams located in areas with no permanent
human habitation demonstrated fecal coliform contamination in each
stream; the highest concentration occurring between June and August (475
to 500 fecal coliforms/100 ml), a seasonal pattern which correlated with the
monthly distribution of G. lamblia infections. The investigators concluded
that giardiasis was endemic in Colorado and that drinking untreated
mountain water was an important cause of endemic infection.
A similar but smaller scale study conducted in Minnesota also suggested
that consumption of untreated water was an important factor in the endemic
occurrence of giardiasis in that state(25). Interviews of individuals identified
from state laboratory records as having stool examinations positive for G.
lamblia during January 1974 to February 1975, showed that 78 had no
history of recent foreign travel. Among this group of individuals without
foreign travel, it was found that 63% had consumed untreated water during
the period of study and 46% had not been out of Minnesota in the 2 months
before onset of symptoms. Although these data are interesting, this
particular study suffers from the lack of an appropriate control group and no
firm conclusions can be drawn.
WATERBORNE OUTBREAKS OF GIARDIASIS
IN THE UNITED STATES
A cooperative effort between the Health Effects Research Laboratory,
Environmental Protection Agency (EPA), in Cincinnati, Ohio and the CDC
in Atlanta, Georgia, to investigate, document, and report waterborne
disease outbreaks in the United States has been in existence since 1971. Local
and state health departments investigate waterborne disease outbreaks and
at times request assistance from CDC and the EPA. As part of the reporting
system, state epidemiologists and engineers in state water supply
surveillance agencies cooperate in providing data on waterborne outbreaks
to EPA and CDC annually. Data reported to EPA and CDC on waterborne
outbreaks of giardiasis are summarized periodically(26-28).
-------
G.F.Craun 133
The waterborne outbreaks of giardiasis reported in Table 1 are those in
which drinking water has been implicated epidemiologically as the vehicle of
transmission of the illness. Only outbreaks associated with water used for
drinking or domestic purposes are included. To be considered an outbreak,
at least 2 cases of giardiasis must be reported before a common source can be
noted and investigated. Single cases of giardiasis reported in backpackers,
which may be related to drinking untreated water, have not been included as
outbreaks.
In a few of the outbreaks, heavy bacterial contamination of the water or an
obvious human source of contamination was found. However, inmost of the
outbreaks, little or no bacterial contamination of the water was reported.
There have been only a few outbreaks where Giardia cysts have been isolated
from drinking water (Table 2). Brady and Wolfe(29) reported 5 cases of
giardiasis in 1973 that were epidemiologically associated with drinking water
from an underground cistern on a Tennessee farm. The cistern was
apparently contaminated by seepage from a pit privy, but no coliform data
were available. The association of illness with consumption of water from
the cistern was strengthened by the reported finding of trophozoites in water
samples taken from the cistern, but the methods used to isolate and identify
the trophozoites were not explicitly described. These findings have been
questioned by Rendtorff(30) and Wright(31). Rendtorff felt that the
investigators either identified Giardia cysts or, more likely, trophozoites of
another free-living flagellate mistaken for G. lamblia. Wright noted that
Giardia trophozoites, unlike the cysts, are fragile organisms whose survival
outside the gastrointestinal tract has not been shown. The first waterborne
outbreak of giardiasis where a G. lamblia cyst was found in the municipal
water supply .occurred in Rome, New York, during November 1974 to June
1975(32). A single Giardia cyst was found upon microscopic examination of
water sediments representing 1,059,8001 of water collected during early M ay
1975 from the raw water intake. This was also the first time that water from
an outbreak had been shown to infect laboratory animals, as giardiasis was
produced in specific pathogen-free dogs fed sediment samples collected from
the raw water intake. Giardia cysts were also isolated from drinking water in
the Camas, Washington, and Berlin, New Hampshire, outbreaks and these
are discussed in more detail later(33-35). An outbreak of giardiasis near
Estes Park, Colorado, in the Rocky Mountain National Park in June 1976
initially affected 9 of 17 persons attending a reunion(36, 37). The group had
stayed at summer cabins supplied by water from a small reservoir on the Fall
River; the water was chlorinated but not filtered. The outbreak prompted
additional studies of 48 other visitors who had stayed at the same cabins
during June 1 to July 28 and a control group of 42 who had stayed during the
same period at a' nearby lodge which received filtered, chlorinated water
from a municipal supply. Symptoms of giardiasis were reported in 37% of
those in the first group but in none of the controls. Giardia cysts were found
in a water filtrate sample taken from a beaver pond located upstream from
the water supply reservoir, but 11 water samples collected from upstream
water sources revealed no fecal coliform counts above 0.5/100 ml.
-------
134
WATERBORNE GIARDIASIS/EPIDEMIOLOGY
e
a
3
•«
5
i
c
~~
—
3
v
a
5c
!*)
1
3
o
^J
c
V.
1
I
~~-
_$J
5
£
-
CD
^>
O
CD
o5
o
CD
CO
>•
o
c
CD
O
CD
Q
,_ >-
2 "Q.
CD 0-
§c^
in
CD
c/)
CD
u
0)
CD
Q
C
o
rt
CJ
o
\
E
TO
C
SE
ft-)
E £
i_ CO C
o > i-
i! co CD
— i— CO
O CD >.
7\ -"^
u ro OT
c § £
CDn- °
? ° 3
I"
£*»!
— C "D
^0
CD C ^
> CO CJ
CO 03
°^^
C > t/D
o *-S
^3 cn
CT3 CO -C
c~ ? ~*~~'
- > g
^ CD >, CD
ca *^ — cj
1± CC Q. i_
^ § 0.3
°^ 3 0
°^ CO CO
Iss^
5 c. CD ca
5 3T O)
03 0 DM-
(/) co co O
^r
.>
CO
CM
^CD
^CD
0305
T~ •—
0 c
Si
0
TD
ca
O
0
C
CD
D.
CO
<
O c'
03 o 0
-
5 °
-^.
-
° *
O
3 —
0 5
^ >.
O^:
1 1
1 1
C L—
?
o
"O
5-D
g 03
0 m
_C ^
CO i=
« E
_QJ CD
Oc
Eo
ca u
CO CD
i_ _a
03 0
TO
> 03
> 0
> 1—
> 3
ca o
CC co
CD
ca
( —
0 c
8 2
E ro
i_ C
o z
t: O
"o Z
CJ CJ
o o 2 =2
1 "m -*-1 CX3 CD CO i '
°c° § § § c °
Q)_CD^--> > -" o 0)
T-^'rScD o CD cj ^
-_°-roo -t; o ^ E
co _^ co j- ca c^
~n O ~n c~ ^tl
= 45 ^ -
i- — ~n
i?-i CD -i i- ~
-C 0)_c_2 ^ 0) ^ Q- _£=
•t; 5 t; -c to ^(-t;
J> •* — ' J> C_) *— ^ i"5 ( ' • "^
> U > 0) c
,_ CD ,_ CD +± (
cp tr CD > ro c
i. 0) ;i! — CD
03 "O CD o *~
g ^5^03 C
CD -^ CD CD ^ <
O C3 "O >+-
ro 0 ro . 0 4
ti c T: _>~ CD
3 o r3 c co
oo ^ oo oz> c
Q_ Q- Q_ ^
CO C/5 C75 *
-t. ^^ f— , i— <-
J 03 "- CD >
£ ^ ro E « S
" C O) 05^
i 03 ^ i: >
TJ ^ § CD"CD >
-v g^raS^
= 2 ° S.E^ °
3. *^ CD CD C tl C
13 O to CD E: 3 O
o cZ3 oo 2w ^
^ °- -=5
^ C75 - ^
00 ^T O CM rn LO CM
CM CM CD r
- ^ CO
CM CM ^ cMro^ co co^-
r~~ r- £; r~~r-r~- "^ ^^
05 CD 2- 030)05 OT 0505
>• ^ Q. CJ (- >- CJ5 >Z
^^ CD CD^^ D 00.
2 S cy) f
0 ca
.u 5
"0 - O-j
n ° U c r
°° \n™
>- >- ow j±
O) CD 4-J ( ) Z)
0.-D '-^ 0 .-
c -= O -C c ca
t 3 CO ,-J, t QJ
co o CD _ ca c.
3^^ < Z<
CD h-
03 >
O "
- 0 U S c5
§° oc c
05 " CD 03
^ 0 ,0^ U
I^ll§ i
D CD o >- ca co
O CD CC X O < (75 Q lOLL LU
-------
G. F. Craun
135
3
C
p
£
^t
$
«i
1
c
^
2
T3
1
^
a
•S
o
c
1
5
^
1
s—
V
03
§
o
"co
D
a
m
o
02
>,
d
Q)
0
"CD
Q
QJ "Q
(0 C
^
QJ
CD
O
CD
CD
°
C
O
ro
o
o
\
O CD r-
$ ^^
~O O
— aZ CO
03^ CJ
isgl
>^l
> 0. 0 £
>- CD D- nS
D CO n ~
<" r- 2~0
-oiE c c
^§05 -
1 1 II
ro ^ CD co
5°^e
c c
0 0
U CJ
CD CD
"c 'c
CO CO
~0 T3
.c _c
I I
CD CD
CD CD
^ ^
CD CD
O O
ca CD
D "E D "c
co o co o
^- Q_
-^ CO
O CO
CN i
CO "^ ^"
i — r^ r^
ro CD ro
^ — CD
0 >- C
Q< ^
—
>
§" 8
CD <->
i_
^ O
CD O
1 S>"i
ro o 2
Q i o
D E O-Q
c c oo ^u_
JJ3 <-• tz <— jr
)— — ~o \ •'"^ — C
-0 ° ^ °^^ cS""
CD C ° CD CD O GLC
CD — c~ CO CO
Eo^coEtDE^. c
El§|||pl
c^|^c?°ic5zi
c c
o a) g
o > u
CD > CD
E u -
co CD !2
-o t -a
-C CO ^
5 "CD ^
m co r5
*i. CD <--
CD ^ ro
5 c §
CD D CD
CJ H- O
CD O CO
D C CO DC
CO O Z> CO O
2 — ^
co ^ o
r~~ co O
co
"*
-* ^Ln
en ^ r-- r~~-
D ro OT ro
CD ^ 4- -CD
c ^ Q- - ^ c
— i <- co z -3
_^
— T— CD ^~
_ _ ) >
- CD
C 52 Z
CD CD
-DO. CD
- E E
w CO 0
2 O CC
D
^ 1 °
\
5° E
3| 1
-"5 oo
n CJ CM
i_ i_
CD CD
CD CD
5 §
CD CD
0 CJ
CD CD
'D "D
CO CO
"S o
ro CD
CD CD
^ ^;
C C
"o "b
CO CO
13 ^
CL
CO
ro CN
m
CD r~~'
^ ro
^_,
d n
CO UL
CD"
f—
ro-°8
m
CO ^ °
*— o<5 C_)
CD
Q. CD "a
E - c
CD H— 2
O OO
E
t=
7, °
CD cj
||
co o
—- ^_
1^
!~§!
D?|S
T3
CO §
25^
- 2 S^
CD ^3 — —
•*— ' — O ^~
TO ..~ — M—
5^^ o
CD "- "° c
ro § C"S
IB 2§ c5 „
CD ^- l—
~O •">"'Lt= ^ ^
CD CO —
co CD ~5 ~D c
QJ CD c_) — ^
~~^> ^ c — c~
5> -i p CD H cj
^
o
0
CD
r^
ro
"~
CD
2
C
o
D)
1
ca
$
co"
CD
CD
0
u
CD •-
\
•r; ,—
CD °
|1
CD CD
1— CD
°^
II
c
o
CD
"c
CO
-5
£
i
05
CD
^
CD
CO
5 "^
CO 0
Q_
CO
r-~
CM
CD
r--
2
CD
C
—3
O
..
-Jf
CD
ci w
CD TO
OLLJ
-------
136
WATERBORNE GIARDIASIS/EPIDEMIOLOCY
^3
s of giardiasis in United Slates (continue
-*
<3
1>
-Q
3
O
C
^
-c
%l
?3
-Si
-Ci
£
O
CD
CO
O
C
Q)
CJ
Q
V Q_
CD Q-
^ GO
C/l
CD
CD
CJ
0)
CD
Q
c
0
CD
O
C
1
c
0
4—"
n
£
~o
CO
CD
Q.
E
CD
_
C
w
to
c
o
u
CD
CD
§
£
CD
CO
CO
tional treatment and surface
CD
^
CD
CO
CD
C
C
O
CD
CD
D
CO
CO
CD
Q.
jr
§
2
CD
g
and disinfection, deficiencies
>-
found in construction of gravr
CD
filter and operation of pressur
filter
Use of untreated well water
influenced by surface water
CL
co
i\
CD
CD
C
D
j:
ro
5
T3
C
0
CD
CL
E
CO
O
i
1
l__
Use of untreated surface wate
CL
co
LO
in
en
T —
!>*
n
"CD
O
I!
c
Surface water with dismfectio
*zr
^
CD
*^J-
CN
CT>
, —
"2**
^
CO
5 c?
_^ Q.
£ ~,
£"5
CD
O J>
^
C
O
,
CD
C
CD
C
0
2
-------
G.F.Craun 137
Table 2. Walerborne outbreaks of gardiavs in the Untied Slates where Giardia
have been isolated from water
Year Location Description
1 973 Tennessee Numerous trophozoites of G lamblia were
found in water samples collected from an
underground cistern
1 974 Rome, A single Giardia cyst was found after
New York filtering 1,059,800 liters of raw water
from the plant intake
1 976 Carnas, Giardia cysts were found in the raw water
Washington and distribution system
1 976 Camp, Giardia cysts were found m a water filtrate
Estes Park, sample taken from a beaver pond located
Colorado upstream from the water supply
reservoir
1 977 Berlin, Giardia cysts were recovered from both
New Hampshire raw water sources and several sites within
the distribution system
Twenty-three waterborne outbreaks of giardiasis affecting 7,009 persons
have been reported in the United States (Table 3). The first waterborne
outbreak documented in the United States occurred at Aspen,
Colorado(38), during December 1965 through January 1966. The outbreak
came to the attention of CDC when a physician at the Center developed
characteristic, symptoms of giardiasis after return from a ski holiday at
Aspen. His stools contained no bacterial pathogens, but cysts of G. lamblia
were abundant. A survey of 1,094 skiers who had vacationed in Aspen
during the 1965-1966 ski season showed that at least 123 (11%) had
developed similar symptoms. The association of G. lamblia with the illness,
the absence of other pathogens, and the response to treatment suggested that
G. lamblia was the agent responsible for illness.
Approximately half the city's water came from a distant mountain creek
serving the western side of the city and half from 3 wells serving the eastern
side. Chlorinators were provided at each water source, but coliform
contamination of the water system was noted intermittently during the 1965-
1966 ski season. Examination of the sparsely populated creek area revealed
no obvious possibility of sewage contamination; however, fluorescent and
detergent tracers placed in the sewage system were detected in 2 of the 3
wells. Engineering evaluation suggested sewage contamination of the wells
from leaking sewer mains near the wells. G. lamblia cysts were isolated from
sewage in the leaking sewer lines. A parasitologic survey of Aspen residents
showed a difference in prevalence of G. lamblia infection between the eastern
section (6.9%) and the western section (3.7%) but this was not statistically
significant. In retrospect, it is possible that the surface water source was also
-------
138 WATERBORNE GIARDIASIS/EPIDEMIOLOGY
Table 3. Walerborne outbreaks of giardiasis in United Slates
Year
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
Total
Outbreaks
1
—
—
—
1
1
—
4
4
4
1
3
4
23
Cases
123
—
—
—
19
34
—
124
73
4,930
9
639
1,058
7,009
contaminated with G. lamblia; however, this was not sufficiently evaluated
at the time.
There has been a steady increase in waterborne outbreaks of giardiasis in
the United States since 1971. Increased recognition of this disease because of
reports of giardiasis in travelers to the Soviet Union has no doubt been
responsible for increased surveillance, investigation, and reporting by public
health authorities in the United States.
Waterborne outbreaks of giardiasis have occurred primarily in
mountainous areas of this country particularly in New England, the Pacific
Northwest, and the Rocky Mountains (Table 4). These areas have
traditionally depended upon surface water sources relatively free of human
sewage contamination and of good bacterial quality, and treatment has been
minimal consisting primarily of disinfection only. Attention to maintaining
continuous disinfection has not been as great as in other areas of the country
where streams are known to be contaminated with human waste discharges.
Colorado has experienced more outbreaks than any other state and this
probably reflects increased surveillance and investigation. Attention focused
on Colorado because of the Aspen outbreak, endemic giardiasis, and reports
of giardiasis in backpackers has no doubt been responsible for increased
surveillance activities.
During the past 5-year period, the most commonly identified pathogen in
waterborne outbreaks in the United States was G. lamblia (Table 5). There
were 20 waterborne outbreaks and 6,833 cases of gardiasis during 1972-1977.
It is possible that some of the 100 outbreaks and 16,504 cases of acute
gastrointestinal illness of undetermined etiology were giardiasis; however,
-------
G.F.Craun 139
Table 4, Location of waterborne outbreaks of giardiasis in United Slates
State Outbreaks
Colorado 9
Utah 3
Montana 2
New Hampshire 2
Vermont 2
California 1
Idaho 1
New York 1
Washington 1
Tennessee 1
this is unlikely because the symptoms, incubation period, and self-limiting
nature of the illness in those outbreaks were not characteristic of giardiasis.
Comparing waterborne outbreaks of giardiasis to other waterborne
outbreaks shows the relative importance of this organism as a waterborne
pathogen (Table 6). In some years there have been relatively few cases of
waterborne giardiasis compared to other waterborne illness, but in other
years the number has been quite high. For instance, in 1974 and 1977, G.
lamblia was responsible for 59% and 27% respectively of the waterborne
illness. The Rome, New York outbreak accounted for most of the cases in
1974, and several large outbreaks occurred in 1977.
Waterborne outbreaks of giardiasis have generally affected small
municipal water systems or semi-public water systems in recreational areas
(Table 7). Municipal water systems are defined as public or investor-owned
water supplies that serve communities. Semi-public water systems, located
in areas not served by municipal systems, are those developed and
maintained at locations where the general public has access to drinking
water provided by industries, institutions, camps, parks, hotels, etc.
Individual water systems are those used by single residences in areas without
Table 5. Eliologv of walerborne outbreaks in United Slates, 1972-1977
Acute Gastrointestinal Illness
Chemical Poisoning
Giardiasis
Shigellosis
Hepatitis A
Salmonellosis
Typhoid
Enterotoxigemc E coli
Outbreaks
100
22
20
14
9
5
4
1
Cases
16,504
508
6,833
4,652
234
1,000
222
1,000
-------
140 WATERBORNE GIARDIASIS/EPIDEMIOEOGY
Table 6. Waterborne outbreaks of giardiasis compared to other
waterborne outbreak!, in United States, 1972-1977
Year
1972
1973
1974
1975
1976
1977
Total
Waterborne
Outbreaks of
Giardiasis
4
4
4
1
3
4
20
Other
Waterborne
Outbreaks
25
22
21
23
32
30
153
Waterborne
Giardiasis
124
73
4,930
9
639
1,058
6,833
Other
Waterborne
Illness
1,514
1,701
3,426
10,870
4,429
2,802
24,742
municipal systems or by persons traveling outside of populated areas (e.g.,
backpackers). The reporting of outbreaks involving individual water
systems is much less complete than for outbreaks in municipal or semi-
public water systems. The small number of documented outbreaks in
backpackers is due to either the lack of investigation or incomplete
investigation or reporting. When investigated, the outbreaks are often
reported in an anecdotal manner. In addition, a number of single cases of
giardiasis occur in which a common source cannot be investigated.
Most of the outbreaks occurred as the result of consuming untreated
surface water or surface water treated only with chlorine (Table 8). The
largest waterborne outbreak of giardiasis occurred in Rome, New York from
November 1974 to June 1975; 350 residents had laboratory-confirmed
giardiasis, and an epidemiologic study estimated that approximately 4,800
individuals were affected during the outbreak(32). Rome used a surface
water source with chlorination as the only treatment. The watershed, 185
square miles of heavily wooded rolling hills, was sparsely populated, but the
presence of human settlements in the watershed area suggested that the
water supply could have been contaminated by untreated human waste. As
previously noted, the infectivity of municipal water was confirmed by
producing giardiasis in specific pathogen-free dogs fed sediment samples of
raw water obtained from an inlet of a city reservoir, and a microscopic
examination of the water sediments uncovered a G. lamblia cyst in one
sample. Rome's raw water bacterial quality was generally good since the
Table 7. Waterborne outbreaks of giardiasis in various types
of water systems in United States
Type of System Outbreaks Cases
Municipal 10 6,695
Semi-public 10 266
Individual 3 48
Total 23 7,009
-------
G.F.Craun 141
Table 8 Water system deficiencies responsible /or waierborne outbreaks
of giardiasis in United Stales
Deficiency
Surface Water with
Chlormation Only*
Untreated Surface Water
Ineffective Filtration
Untreated Ground Water
Ground Water with
Chlormation Only
Contaminated Cistern
Unknown
Total
Outbreaks
10
6
2
2
1
1
1
23
Cases
5,307
186
1,350
19
123
5
19
7,009
* Includes one outbreak where filtration facilities were available but used intermittently
average raw water coliform count during the period June 1974 to June 1975
was 220/100 ml(39). Much higher than normal values (4,600 coliforms/100
ml), however, were experienced prior to the outbreak during August and
September 1974, suggesting that the system may have been contaminated
during this time. For water treatment, ammonia and chlorine were
introduced together to produce a chloramine for a total combined chlorine
residual of 0.8 mg/1. As also noted previously(21,22), chloramine is generally
a less effective disinfectant, requiring longer contact times and/or higher
concentrations than free chlorine. Shaw, et al (32), reported that coliform
counts in the Rome distribution system were negative at 13 of 15 sampling
points and 1 and 3 colonies/100 ml at the two other points. Results of routine
bacteriological sampling of the distribution system showed that from
November 1974 to June 1975, an average of 66 samples were collected each
month; all were negative for coliforms, except 4 samples on February 25
which contained 20, 20, 30, and 40 coliforms/100 ml (personal
communication, S. Syrotynski). Shaw, et al(32), also reported the total
bacterial count to be quite high in the distribution system, and it is possible
that this could have been responsible for the lack of high coliform counts,
since high total bacterial populations have been implicated as possibly
suppressing coliform growth in test media(40).
For those systems with disinfection, an attempt was made to determine if
the chlorination was interrupted or the facilities were operating but
providing an inadequate concentration of chlorine or contact time. Only 2
systems were identified where the chlorination facilities were defective and
chlorination was interrupted prior to the outbreak. The remaining systems
were apparently continuing to chlorinate with sufficient chlorine to
inactivate coliform organisms but insufficient chlorine for inactivation of
Giardia cysts.
For the untreated water system, coliform counts were available for only 2
outbreaks. In 1 that occurred in Colorado and affected 12 individuals, total
-------
142 WATERBORNE GIARDIASIS/EPIDEMIOLOGY
coliforms of 23/100 ml were detected'in the water system. The other
outbreak(41) involved a group of campers on a 2-week trip in the Uinta
Mountains of Utah during September 1974. Thirty-four of the 54 campers
had diarrhea during and after the trip; 22 (79%) of 28 symptomatic campers'
stools examined contained G. lamblia cysts. Epidemiological data and fecal
coliform counts implicated the remote mountain stream used as a water
source by the group as the vehicle of transmission. Water samples from the
stream had a mean fecal coliform count of 42/100 ml. Water was not
disinfected or boiled prior to consumption. The source of the infection was
not identified, but several active beaver ponds, grazing sheep, and a
shepherd were noted in the area. It was felt that perhaps wild or domestic
animals served as an alternate host and the source of Giardia in this
outbreak, but this could not be proven as cysts could not be identified in
water or in animal droppings.
The first documented waterborne outbreak of giardiasis involving a
filtered water supply occurred in Camas, Washington(33, 34), in the spring
of 1976 and affected 600 individuals of a population of 6,000. Camas used 2
water sources, mountain streams and deep wells. Giardia cysts were isolated
from the raw water entering the water treatment plant and from treated
water in 2 distribution system storage reservoirs. Well water was not found
to be contaminated. The watersheds for the surface water sources were well
isolated, had no human habitation, and had extremely limited human
activity. Several animals on the watershed were trapped and examined for
Giardia cysts. Three positive beavers were found within foraging distance of
the water intakes for Camas. Treatment for the surface water sources
consisted of a mixed-media pressure filter and disinfection; no
sedimentation was employed prior to filtration. Prior to the outbreak,
failure of the chlorination equipment occurred, and a number of deficiencies
were found in the condition and operation of the pressure filters, including
ineffective chemical pretreatment. It was reported that the treated water
produced by the treatment plant had met both turbidity and coliform
standards prior to and during the outbreak.
A second waterborne outbreak of giardiasis involving a filtered water
supply occurred in Berlin, New Hampshire(33,35) in the spring of 1977
affecting 750 of a population of 15,000. Berlin used 2 rivers for its water
supply, the Ammonoosuc and the Androscoggin. Giardia cysts were
recovered from the raw and treated water from both rivers and from 2 sites
within the distribution system, the regional hospital and city hall. The
Ammonoosuc River is located in the White Mountain National Forest;
however, access is not restricted, and an estimated 3,000 people used the area
for recreational activities during October, November, and December 1976.
Water from the Ammonoosuc was chlorinated and filtered under pressure
without sedimentation or flocculation. There was no pretreatment of the
water with conditioning chemicals. The physical plant was 30 years old and
numerous deficiencies, which could permit cysts to pass through the filters,
were found in the condition and operation of the pressure filters. For the
Androscoggin River, a new treatment plant with upflow clarification and
-------
G. F. Craun
143
rapid sand filtration was put into service just prior to the outbreak.
Difficulty in creating a proper weight floe and operating the upflowclarifier
were reported. An engineering evaluation revealed faulty construction of a
common wall separating filtered and unfiltered water, which allowed
unfiltered water to bypass the modern, conventional treatment plant. It was
felt that this was the primary reason for cyst passage into the distribution
system served by the modern treatment plant. Routine bacterial samples
collected in the distribution system prior to the outbreak showed that the
coliform standards had not been exceeded and that a free chlorine residual
was not maintained in the distribution system even though free chlorine
residuals of 0.3 and 0.7 mg/1 were found in finished water at the treatment
plants.
There were two small outbreaks involving untreated ground water
systems. In one, human sewage contamination of the well was documented.
The second outbreak involved a well that was adjacent to a stream. Because
there was a distinct joint in the concrete casing just below the river level, the
water in the well was felt to originate primarily from the stream. Algae were
identified in the well water and chemical analysis of the well water and river
water was quite similar. Fecal coliforms of 11 to 14/100 ml and total
coliforms of 14 to 110/100 ml were found in the well water. Turbidity of the
well water was 1.4-2.4 NTU. A beaver dam was located one-half mile
upstream from the well. The outbreak involving a ground water system with
chlorination as the only treatment was the Aspen outbreak. This outbreak
and the one involving the contaminated cistern in Tennessee have already
been discussed.
Waterborne outbreaks of giardiasis seem to involve 2 distinct groups of
people, either visitors or campers or the usual residents of the area (Table 9).
Table 9. Seasonal distribution of waterborne outbreaks of giardiasis in United Stales
Month
January
February
March
April
May
June
July
August
September
October
November
December
Outbreaks
0
1
3
1
1
4
3
2
3
0
2
3
Population
Visitors or
Campers
—
—
2
—
1
3
1
2
3
—
—
1
Affected
Usual
Residents
—
1
1
1
—
1
2
—
—
—
2
2
Total 23 13 10
-------
144 WATERBORNE GIARDIASIS/EPIDEMIOLOGY
There appears to be a seasonal trend of outbreaks in visitors during the
summer months. This implies there is either increased contamination of
these water supplies during this period, or if it is assumed that the supplies
are always contaminated, use by greater numbers of susceptible individuals
during this period. Outbreaks involving usual residents seem to occur during
any season, but most have occurred in the late fall through early spring.
There have been at least 2 major outbreaks of giardiasis in the United
States where failure to isolate G. lamblia from the suspected water source
has strongly influenced the investigators to reject drinking water as the
possible vehicle of infection. One outbreak occurred in Portland, Oregon,
from October 1954 to March 1955 and affected an estimated 50,000
persons(42,43). This outbreak was reported only in the correspondence
section of one journal and briefly mentioned in a review article on epidemic
giardiasis. A full report of the outbreak was reportedly rejected for
publication by a journal on the basis of insufficient proof of the etiology of
the illness and failure to isolate an organism from the suspected water
supply. In retrospect, it appears that this outbreak of gastroenteritis was
probably of mixed etiology with Giardia implicated in those cases with fairly
distinctive symptoms. No enteric pathogens or viruses were isolated from
clinical specimens nor were any protozoa found in unusual numbers except
Giardia. The source of infection and mode of transmission were never
satisfactorily determined although drinking water was apparently suspected.
The water supply is a surface water source with chlorination as the only
treatment.
The second outbreak occurred in Boulder, Colorado, from June to
August 1972 affecting a population of at least 297 individuals who were
positive for G. lamblia and had symptoms compatible with giardiasis(44).
The epidemiologic information, with one major exception, did implicate a
widespread common source, the exception being an age-specific attack rate
heavily weighted toward young adults. A review of the records revealed no
positive coliform samples, and based on a limited sampling of small volumes
of water, no Giardia cysts were found in the water supply. Water plant
operation records did indicate that because of high demand on 4 separate
occasions in May and June, surface water from one source bypassed the
filtration system and was treated with chlorine only for a total of 14 days. In
bypassing the filtration system, the water was processed through
microstrainers which cannot remove particles the size of Giardia cysts. Even
with this information about the water treatment process, the investigators
concluded that there was no evidence except a coincidence of dates to
suggest that the distribution of unfiltered water was implicated as the source
of infection. Unpublished data (personal communication, J. W. Hoffbuhr)
also noted that the water utility was experimenting with polyelectrolyte
addition prior to filtration and stopped using alum during a period before
the outbreak. Because the finished water quality deteriorated during this
period, alum feed was resumed. Further investigation of areas served by the
unfiltered water or specific dates of change in alum feed might have provided
answers to some questions raised during the investigation.
-------
G.F.Craim 145
SUMMARY
The data from waterborne outbreaks of giardiasis in the United States
indicate that simple disinfection as the only treatment for surface water
sources is ineffective in preventing waterborne transmission of this infection
and that to protect against transmission, all surface water should receive
pretreatment, preferably with sedimentation, and filtration in addition to
chlorination. In Rome, the absence of filtration and treatment of surface
water by simple chlorination only, left the residents unprotected against
waterborne giardiasis. In Camas and Berlin, outbreaks occurred even
though the water supplies were filtered. In these 2 outbreaks, deficiencies in
both the operation and installation of the treatment facilities allowed the
passage of cysts into the water system and subsequent illness. Lack of proper
chemical pretreatment in these situations emphasizes the importance of
chemical conditioning in filtration. The use of pressure filters in both of these
systems illustrates the operational problems and the unreliability of pressure
filter systems used for microbiological treatment of drinking water. While
pressure filters are routinely used for iron and manganese removal, they are
generally not recommended for microbiological treatment(45). Faulty
construction of a new conventional treatment plant in Berlin allowed
untreated .water to bypass the rapid sand filters and illustrates the
importance of good design and construction inspection. Waterborne
outbreak data, engineering experience, and filtration theory indicate that
well operated and properly functioning, conventional treatment plants
employing coagulation/flocculation, settling, and filtration should be
successful in preventing waterborne outbreaks of giardiasis. The giardiasis
outbreaks that have occurred in filtered water supplies do not contradict this
reasoning since numerous deficiencies have been found in the design,
installation, and operation of these filtration systems. Water filtration
theory indicates that organisms the size of Giardia cysts should be removed
by conventional sand filters if effective pretreatment of the water is
accomplished prior to filtration. Conventional treatment of surface water
generally includes coagulation/flocculation and settling prior to filtration,
or if the settling process is not used, the addition of appropriate chemicals for
conditioning of the water or the filter media. New State drinking water
regulations(46) adopted by Colorado in 1977 now require surface water
supplies to be filtered, as well as disinfected, to remove Giardia cysts.
Coliform organism identification is used as an indication of fecal
contamination of water supplies and is widely employed in routine
surveillance programs. Negative results have usually been interpreted as
providing assurance that the water is free of enteric pathogens. This
interpretation must be reevaluated in light of data available from
waterborne outbreaks of giardiasis. In the several outbreaks of giardiasis
where disinfection was provided, Giardia cysts were found in the water
supply in the absence of coliforms. Although adequate disinfection data are
not currently available, outbreaks have shown that chlorination at
conventional dosages and contact times normally employed in water
treatment are inadequate for destruction of Giardia cysts. If Giardia cysts
-------
146 WATERBORNE GIARDIASIS/EPIDEMIOLOCY
are as resistant to chlorination as cysts of E. histolytica, high concentrations
of chlorine and long contact times would be required for cyst inactivation.
Disinfection practices employed in systems using only disinfection do not
provide for high concentrations of chlorine or long contact times, and in
surface water systems where simple disinfection is the only treatment, it is
likely that Giardia cysts could survive the treatment whereas coliforms
would not. The coliform test in these situations could not provide assurance
that an outbreak of giardiasis would be prevented.
REFERENCES
1. Benenson, A. S. 1975. Control of communicable diseases in man. 12th ed. Amer. Public
Health Assoc., Washington, D. C
2. Wolfe, M. S. 1975. Giardiasis. J.A.M.A , 233.1362-1365
3 Schultz, M. G. 1975 Giardiasis J.A.M.A., 233:1383-1384.
4. Wolfe, M. S. 1978 Giardiasis. New Engl. Jour. Med., 298:319-321.
5. Davis, C. and L. S. Ritchie. 1948. Clinical manifestations and treatment of epidemic
amebiasis occurring in occupants of the Mantetsu apartment building, Tokyo, Japan.
Amer. Jour. T.rop. Med., 28:817-823.
6. Center for Disease Control. 1970. Giardiasis \n travelers. Morb. and Mort. Wkly. Rpt ,
19.455
7. Walzer, P. D., M. S. Wolfe, and M. G. Schultz. 1971 Giardiasis in travelers. Jour. Infect
Dis., 124-235-237.
8. Brodsky, R. E., H. C. Spencer, Jr., and M. G. Schultz. 1974. Giardiasis in American
travelers to the Soviet Union. Jour. Infect. Dis., 130.319-323.
9. Johnson, D.D. 1972. Enteritis secondary to Giardia tamblia in students traveling on tour in
Russia. Jour. Amer. Coll. Health Assoc., 20:207-208.
10 Anderson, T., J. Forssell, and G. Sterner. 1972 Outbreak of giardiasis. effect of a new
antiflagellate drug, timdazole. Br. Med. Jour., 2:449-451.
11. Jokipii, L. 1972. Giardiaasia Leningradista. Duodecim., 88:522-526.
12 Forssell, J., K. Lantorp, and G. Sterner. 1972. Giardiasis among Swedish tourists after a
visit to the Soviet Union. Lakartidningen, 69:1132-1137
13. Aust Kettis, A. and L.Magnius. 1973. Giardia lamblia infection in a group of students after
a visit to Leningrad in March 1970. Scand. Jour. Infect Dis., 5.289-292.
14. Fiumara, N. 1973. Giardiasis in travelers to the Soviet Union New Engl. Jour. Med.,
288:1410-1411.
15 Center for Disease Control 1974. Giardia lamblia infection in travelers to the Soviet
Union. Morb. and Mort. Wkly. Rpt., 23:78-79
16. Gendel, E. 1974 Giardiasis in Russia. New Engl. Jour. Med., 290:286.
17 Jokipii, L. and A. M. M. Jokipii. 1974. Giardiasis in travelers, a prospective study. Jour.
Infect. Dis., 130.295-299.
18. Center for Disease Control. 1975. Giardiasis—in residents of Rome, New York, and in
travelers to the Soviet Union. Morb. and Mort. Wkly. Rpt., 24:371.
19. Martin, J. F. and M. A. Martin. 1975. Giardiasis from Russia Brit. Med. Jour., 2.89.
20 Ryan, W. L. and J. W. Grainge. 1975. Sanitary engineering in Russia. Jour. Amer. Water
Works Assoc., 67-285-288.
21. U. S. Environmental Protection Agency. 1971. Manual for evaluating public drinking
water supplies. Gov't. Printing Office, Washington, D. C.
22 American Water Works Association. 1971. Water quality and treatment McGraw-Hill,
New York, N. Y.
23. Center for Disease Control. 1977. Outbreak of suspected giardiasis among travelers to
Madeira, 1976. Morb. and Mort. Wkly. Rpt , 26:426-427.
24. Wright, R. A., H. C. Spencer, R. E. Brodsky, and T. M. Vernon. 1977. Giardiasis in
Colorado, an epidemiologic study. Amer. Jour. Epidem., 105.330-336.
25. Weiss, H. B., D. A. Winegar, B. S. Levy, and J. W. Washburn. 1977. Giardiasis in
Minnesota, 1971-1975. Minn. Med., 60:815-820.
26. Craun, G. F. and L. J. McCabe. 1973. Review of the causes of waterborne disease
outbreaks. Jour. Amer. Water Works Assoc., 65:74-84.
-------
G.F.Craun 147
27. Craun, G. K., L.J. McCabe, and J. M. Hughes. 1976 Waterborne disease outbreaks in the
U S —1971-1974. Jour. Amer. Water Works Assoc., 68:420-425.
28. Craun, G. F. 1978. Outbreaks of waterborne disease in the United States, 1971-1977. 98th
Annual Conference Proceedings. Jour. Amer. Water Works Assoc., Denver, Colorado.
29. Brady, P. G. and J. C. Wolfe. 1974. Waterborne giardiasis. Ann. Intern. Med., 81:498-499
30. Rendtorff, R. C. 1975. Giardia in water. Ann. Intern. Med., 82:280.
31. Wright, R. A. 1975. Giardial infection from water. Ann. Intern Med., 82.589-590.
32. Shaw, P. K., et al. 1977. A communitywide outbreak of giardiasis with evidence of
transmission by a municipal water supply. Ann. Intern. Med., 87.426-432.
33 Center for Disease Control. 1977. Waterborne giardiasis outbreaks-Washington, New
Hampshire Morb and Mort. Wkly. Rpt., 26:169-175
34 Kirner, J. C., J. D. Littler, and L. A. Angelo. 1978 A waterborne outbreak of giardiasis in
- Camas, Washington Jour Amer. Water Works Assoc., 70:35-40
35 Lippy, E. C. 1978. Tracing a giardiasis outbreak at Berlin, New Hampshire. Jour. Amer
Water Works Assoc., 70-512-520
36. Center for Disease Control 1977. Giardiasis-Califorma, Colorado. Morb. and Mort.
Wkly. Rpt . 26 60
37. Center for Disease Control. 1977 Errata, Vol 26, No. 7. Morb and Mort. Wkly Rpt.,
2692
38. Moore, G. T., et al. 1969. Epidemic giardiasis at a ski resort New Engl Jour Med.,
281 402-407
39. Syrotynski, S. and T. A. Reamon. 1977. Giardta as related to water (unpublished
manuscript)
40. Geldreich, E. E., H. D. Nash, D. J. Reasoner, and R. H. Taylor. 1972. The necessity of
controlling bacterial populations in potable water, community water supply. Jour Amer.
Water Works Assoc., 64:596-602.
41. Barbour, A. G., C. R. Nichols, and T. Fukushima. 1976. An outbreak of giardiasis in a
group of campers Amer. Jour. Trop. Med and Hyg , 25.384-389.
42 Veazie, L. 1969. Epidemic giardiasis. New Engl. Jour. Med., 281.853.
43 Meyer, W. T. 1973. Epidemic giardiasis, a continued elusive entity. Rocky Mt. Med. Jour ,
70:48-49
44. Vernon, T. M. 1973. Giardiasis probe inconclusive Colo, Health, Jan.-Feb
45. Great Lakes-Upper Mississippi River Board of State Sanitary Engineers. 1976.
Recommended Standards for Water Works. Health Education Service, Albany, N. Y
46 Colorado Board of Health 1977 Primary Drinking Water Regulations, Section 17.1.2.
Discussion
E. GELDREICH: There are 2 points I wish you would comment on which
seemed to surface in your correlations with coliform occurrences or lack of
them and the presence of Giardia outbreaks.
It appears that there is a problem of sampling frequency for small supplies,
and when we investigate these small supplies in which outbreaks occurred,
we may find data for only 1 to 10 samples/ month. My concern is that we are
not monitoring supplies with sufficient samples to recognize that there is a
breakdown in treatment.
You have noted that in some outbreaks, coliforms at times range from 1 to
100/100 ml, and one wonders why there was no apparent justified response
to these signals by the water treatment operator. There was something
wrong, something breaking through the treatment barrier because we did
find coliforms in the water. Why didn't the water plant operator or the
engineers respond by doing something about it? Is it just constantly ignored
because there is only a low level of coliform contamination? If so, I think we
must act and sound the alarm. We cannot afford to ignore these low levels of
coliform occurrences.
G. CRAUN: I agree with your comment that not enough samples are
collected in the small water systems, and that more samples should probably
-------
148 WATERBORNE GIARDIASIS/EPIDEMIOI.OGY
be collected, that is, if you agree with the basic premise that by collecting
samples and doing coliform tests we are going to help prevent outbreaks.
Too many times the common reaction to finding a positive is that the
water sample was contaminated during collection, that either someone put
their finger in the bottle or they did not collect the sample properly. A single
positive sample does not seem to be taken as an indication that something
needs to be done.
E. GELDREICH: Finding coliforms is not going to solve any problems in
terms of outbreaks, but it should signal the water plant operator that
something is wrong with the treatment and there should be a proper
response. If you do find coliforms, 1 think there should be a sanitary
engineering survey done again on the treatment system.
G. CRAUN: I agree. The reason I mentioned it was because in those
systems that chlorinate only, the coliforms are disinfected but the Giardia
cysts may not be.
E. GELDREICH: We in EPA with the Office of Drinking Water are
looking at the frequency of sampling, and this certainly is an area of
weakness that must be explored.
T. VERNON: I thought that was a very informative presentation, and I
appreciate it. The Boulder story of 1972 is an interesting one and I think it
can be well characterized as an absence of concern about transmission by
water. Indeed there was a very hot late spring season in the Boulder area at
that time. There was a great deal of lawn watering going on, and the records
show that the slow sand filtration system was bypassed. There was a
microstrainer on the system. 1 understand the pore size of that microstrainer
would have been about 26 nm. There was no coagulation or sedimentation
of that water.
I think most of the controversy in the Colorado area about the Boulder
outbreak is whether indeed it was an epidemic at all. Many people believe in
retrospect that with all the publicity, people just began to pick up on the
endemic level of Giardia cysts. On the other hand, I believe most people in
Boulder would say that it was an outstanding event; such an occurrence had
not been experienced in preceding years, nor has it occurred in the years
since that time.
One other comment which certainly leads to our discussions tomorrow on
research needs is that we in Colorado have taken a step which is in a sense
beyond the level of technology in that we have required complete treatment
of surface waters in our regulations, interpreted to mean not only filtration
but also coagulation and sedimentation. We need to know a great deal more
to support our regulatory advances. This is unfortunately true in air
pollution and a number of other areas in which we are instituting regulations
without having the data base which we really need.
G. CRAUN: Thank you for your comments on Boulder. I was not aware
that the regulations also required coagulation to precede filtration. Very
good.
T. DEGAETANO: Is there any correlation between turbidity and
giardiasis?
G. CRAUN: I have looked for information on turbidity in systems
showing an outbreak but it generally is not available. Either the test was
never performed or I simply could not obtain the information.
-------
G.F.Craun 149
R. FREEMAN: All but 1 of the states that you put down on your list could
be considered mountain states, whether they are east or west. Are there any
outbreaks known that would be from the Plains States? Also, should we
really differentiate quite clearly if we are talking about giardiasis the disease
or the presence of the Giardia cysts in stools? 1 do not know if this is
important. It strikes me as being important that we should differentiate
between those 2 issues.
G. CRAUN: The outbreaks that are reported in the U.S. are ones in which
people were symptomatic. The studies of the endemic occurence depended
primarily on stool-positive individuals who may or may not have been
symptomatic.
The outbreaks occurred in generally mountainous areas in all the states
except Tennessee, where a contaminated cistern was involved, and there is
some question as to whether or not that was actually a waterborne outbreak.
I would offer an explanation of why they do not occur in the Plains, in the
Midwest, in the Mid-Atlantic and the Southeast. In these areas the water
sources are known to be contaminated, and appropriate treatment plants are
usually built. Based on studies by CDC there is person-to-person
transmission of giardiasis in daycare centers in the Southeast and places
other than the mountainous areas of the country, but 1 think waterborne
transmission is prevented because of recognition that the water supplies are
contaminated and filtration is practiced. In mountainous areas where the
outbreaks have occurred we have generally seen streams of fairly good
quality where only chlorination was provided but not much attention was
paid to maintaining the chlorinators because of a belief that the streams were
not contaminated.
E. MEYER: I was not in Portland, Oregon during 1954-55, but my
predecessor (L. Veazie) was, and 1 heard about the outbreak in some detail.
In those days every medical student was asked to examine his own stool
sample for parasites, and that year an extraordinarily high number of them
found Giardia. At the same time that this epidemic came across the city, it
was reported that an estimated 50,000 cases of this disease occurred, and
physicians were seeing Giardia that they had not seen since their school days
in stool samples.
My predecessor wrote the article and I am disheartened that she never
published it. 1 have the article, and if anybody wants it or if you want to have
it as an addendum to this symposium, you are welcome to it. It was at a time
when there were still many people who did not believe that Giardia could
cause disease, which could be one of the reasons that there was not much
statistical information on the incidence of these cysts in people before the
outbreak. In retrospect, it looks like the water supply could well have been
involved. It came during the winter at a time when the rains were particularly
heavy in Oregon and the water was sometimes turbid. Furthermore, 1 think
the State Department of Public Health at the time examined these patients
for viruses, for all kinds of bacteria, for other protozoa, and were not able to
come up with any other possible source of diarrheal disease, but they did
comment on the extraordinarily high incidence of giardiasis that they found.
(Editors Note: L. Veazie's manuscript on the Portland, Oregon outbreak
appears at the end of the Epidemiology Section on p. 174.)
-------
150
Waterborne Giardiasis
(Summary of recent epidemiologic investigations
and assessment of methodology)
D.Juranek
Parasitic Diseases Division, Bureau of Epidemiology,
Center for Disease Control, Public Health Service,
U.S. Dept. of Health, Education, and Welfare,
Atlanta, Georgia
ABSTRACT
The type and depth of epidemiologic investigations of waterborne outbreaks of
giardiasis have varied considerably. The purpose of this paper is to review the strengths and
weaknesses of epidemiologic investigations conducted in Rome, New York, 1975; Camas,
Washington, 1976; and Berlin, New Hampshire, 1977. It is concluded that 1) an alternative
case identification technique to the one presently used is badly needed, but until a new
technique becomes available the presently used criteria for defining a case should be
standardized so that data obtained in one outbreak are comparable to those obtained in
another, 2) greater care will be required in future investigations to insure that adequate
specimens are collected to rule out bacterial and viral causes of diarrhea, and 3) the SPF
dog model should continue to be used to demonstrate viability and mfectivity of Giardia
cysts recovered from water until an equally reliable laboratory method is identified and
made available The long-term solution to control of waterborne giardiasis lies in
improvements in and widespread use of water filtration However, there will continue to be
a need for an effective and safe method for disinfecting water on an emergency basis in
communities without water filtration and in those communities where established water
filters have failed
Although there have been numerous waterborne outbreaks of giardiasis in
the United States (see Craun, these Proceedings) the type and depth of
epidemiologic investigations performed in these outbreaks has varied
considerably, and therefore the data obtained have been of varying quality.
Among the more important factors contributing to the variable quality in
data are: 1) the period of time elapsing between detection of a possible
outbreak and initiation of an epidemiologic investigation; 2) number and
training of personnel participating in the epidemiologic investigation; 3)
type and availability of laboratory support services; and, 4) amount of
cooperation extended by the group or community experiencing the
outbreak. Early investigation of an outbreak is critical because much of the
information gathered during an epidemiologic investigation is based on a
respondent's memory of past events or activities. The shorter the time
between the occurrence of an event and administration of an epidemiologic
questionnaire, the more reliable the respondent's answers are likely to be.
The number and training of personnel participating in an investigation
undoubtedly have an impact. No longer can one person be expected to have
all the knowledge and expertise necessary to handle all aspects of a
-------
D.Jwanek 151
waterborne disease investigation. Specialists are needed, and failure to seek
their advice or assistance may seriously handicap an investigation. Similarly,
failure to obtain adequate laboratory and community support may also
undermine the investigation. In the recent investigations of waterborne
giardiasis in Rome, New York; Camas, Washington; and Berlin, New
Hampshire, increasing efforts were made to minimize the negative impact of
these variables by enlisting the assistance and cooperation of research
scientists, sanitary engineers, and health professionals at the university,
local, state, and federal levels in identifying, reporting, and investigating
outbreaks. While this multidisciplinary team approach has provided the
scientific community with a surprising amount of new information about
Giardia in a relatively short period of time, there is much more to be learned
about the epidemiology of giardiasis.
The purpose of this paper is to review the strengths and weaknesses of
epidemiologic investigations conducted in Rome, Camas, and Berlin. The
objectives are three-fold: 1) to identify epidemiologic approaches that have
been useful in the past so that useful data will not be omitted in future
investigations; 2) to point out our mistakes so that they will not be repeated
in the future; and, 3) to identify unresolved epidemiologic problems with the
hope that new and innovative field and/or laboratory approaches will be
found.
ROME, CAMAS AND BERLIN OUTBREAKS - BACKGROUND
These 3 outbreaks had several features in common: 1) they all occurred in
mountainous communities; 2) they all resulted from contaminated surface
water, not well water; 3) chlorine was used as a water disinfectant in all 3
cities; 4) all 3 outbreaks occurred in the spring of the year (suggesting that
spring thawing and subsequent overflow of stream beds may be related in
some way to the outbreaks) and, 5) Giardia cysts were recovered from
suspect water in all 3 outbreaks.
The Rome, New York, outbreak will probably be remembered for 2
reasons. First, it is the largest outbreak of giardiasis ever reported in the
United States. Approximately 5,000 persons (10% of the population) were
believed to have been ill. Second, it was the first investigation in which
Giardia cysts were actually recovered from water during an outbreak( 1). The
Camas outbreak differed from the one in Rome in that: 1) it was the first
outbreak involving a filtered water supply; 2) it was the first outbreak in
which Giardia cysts were readily demonstrated in samples of raw and treated
water; and 3) it was the first time that an animal reservoir (beavers) had been
implicated as a source of water contamination(2). The outbreak in Berlin
was unique among the 3 in that about half of the persons infected with
Giardia were asymptomatic. Berlin also differed from the other 2
communities in that it had 2 surface water supplies, each with its own water
treatment facility that included filtration. Giardia cysts were found in raw
and treated water from both water supplies, and there was some
epidemiologic evidence to suggest that there may have been 2 simultaneous
waterborne outbreaks in the city(3).
-------
152
WATERBORNE GIARDIASIS/EPIDEMIOLOGY
EPIDEMIOLOGIC DATA RELATED TO WATER
Epidemiologic data related to water are summarized in Table 1. The
amount of chlorine used in routine water disinfection is believed to have
virtually no effect on Giardia cysts. However, it is important that both total
and free residual chlorine levels in drinking water be measured in waterborne
outbreaks. The determination that adequate chlorination of water exists
provides additional epidemiologic evidence that a waterborne epidemic of
diarrheal disease is due to Giardia, since most bacterial and viral agents of
diarrheal disease would be killed by adequate levels of chlorine. The Rome,
New York, water supply was inadequately chlorinated. Free residual
chlorine could not be detected at the time of the investigation. The New York
State chlorination standard is 0.2 mg/1 of free chlorine or 1 mg/T of
chloramine throughout the distribution system. Inadequate chlorination
during this outbreak indicates that possibly some of the illness within the
community may have been due to bacteria or viruses. During the outbreak in
Camas, several temporary breakdowns of an automatic chlorinator also
resulted in undetectable levels of free residual chlorine in the water
distribution system on at-least 1 day(4). Adequate chlorine levels were found
in the Berlin water system throughout the epidemic.
Water filtration plays a key role in preventing waterborne outbreaks of
giardiasis. However, outbreaks have occurred in at least 2 communities that
used some type of water filtration procedure. Types of water filters and their
filtering efficiency are discussed elsewhere in these Proceedings. It must be
Table I Waterborne epidemics of giardiasis
ic data related to water)
1975
VARIABLE ROME. NY
1976
CAMAS, WASH
1977
BERLIN. N H
Type of water
Surface
Water chlorinated Yes - but
inadequate
Water filtered No
Type of filtration
Surface
Yes - but
temporary
interruptions
Yes
Pressure filter
Surface from 2
sources
Yes - satisfactory
at all times
Yes
1) Pressure filter
2) Gravity sand
filter
Giardia cysts re-
covered from
water
Recovery
method
Identification
method
Yes
CDC sand
filter
Inoculation of
SPF dogs
Yes
EPA filter
Microscopic
exam of filter
sediment
Yes
EPA fitter
Microscopic
exam of filter
sediment
-------
D.Juranek 153
emphasized that in the 2 outbreaks where water was filtered prior to
distribution, defects were found either in the filters themselves or in their
utilization. For water filters to be effective in removing Giardia cysts, they
must be properly constructed, properly maintained, and properly operated.
Because this is not always the case, it should never be assumed that the
presence of water filtration devices rules out the possibility of waterborne
giardiasis.
Epidemiologic evidence for waterborne transmission for Giardia has
existed for many years, but definitive proof, i.e., recovesy of the organims
from suspect water, managed to elude investigators until 1975. At the time of
the Rome, New York, outbreak, the Center for Disease Control (CDC) was
in the process of evaluating a water sampling sand filter. Although very little
was known about the capability of this filter, the CDC decided to use it in
Rome. The sand filter is capable of handling 95,000 liters (25,000 gal.) of
water per day. Filtration of the water in Rome yielded approximately 1 liter
of sediment in a 24 h period. The filter was operated for 10 consecutive days
providing 10 different sediment samples. Approximately 30 coverslip
preparations were made from each sample and examined in Rome by a CDC
microbiologist within 12 h after collection. No Giardia cysts, were found by
direct microscopy in Rome. However, 2 of 10 samples sent to Colorado State
University for inoculation into specific pathogen-free (SPF) beagles did
produce asymptomatic infections in the dogs. Aliquots of the 2 samples
found infective for dogs were then reexamined microscopically. After 40
coverslip preparations, the CDC lab was able to find 1 Giardia cyst in 1
sample; no organisms were seen in 20 coverslip preparations from the second
sample. In the other two outbreaks, Giardia cysts were easily recovered from
both raw and filtered water and identified by simple microscopic techniques.
A new piece of portable filtration equipment developed and operated by the
Environmental Protection Agency (EPA) investigative team is credited for
this great improvement in water sampling technology (see Jakubowski, these
Proceedings). The CDC sand filter was also used in Camas, Washington, but
no Giardia cysts were recovered using this equipment.
An attempt was made in Camas to demonstrate the infectivity of Giardia
cysts recovered from water, using SPF beagles. Unfortunately, specimens
were mishandled enroute to Colorado and when they were received by
Colorado State University obvious morphologic deterioration of the cysts
had occurred. The specimens were given to the beagles anyway but failed to
produce patent infections. SPF beagle studies were not performed during
the Berlin outbreak due to the unavailability of dogs at that time.
The contribution made by the SPF dog model in the Rome investigation
should not be overlooked in future outbreaks. Although use of SPF dogs is
relatively expensive, this biologic method for recovering Giardia yields
important information not available by other techniques. Direct
microscopic examination of filter sediment provides information about the
presence or absence of Giardia cysts in the water, but it provides no
information about the animal origin of the cyst, the viability of the cyst, or its
-------
154
WATKRBORNK CIARDIASIS/KPIDKMIOI OCY
potential infectivity for humans. Using in vitro excystment procedures (see
Bingham, these Proceedings), some assessment of cyst viability is currently
possible, but the relationship between excystment and infectivity of Giardia
cysts has not yet been established. Thus, the SPF dog model is the only
method that demonstrates both viability of Giardia cysts and their
infectivity. From the studies of Davies and Hibler (see Davies, these
Proceedings], there is also suggestive evidence that cysts found infective for
SPF dogs are potentially infective for humans.
SOURCE OF WATER CONTAMINATION
Once a Giardia cyst has been identified in water, the next logical question
is, "How did it get there?" Data on human and animal sources of water
contamination are summarized in Table 2. In Rome, Camas, and Berlin,
there was no evidence of sewage contamination of drinking water. The most
likely source of water contamination in Rome was never determined. Water
Department personnel were confident that there were no humans living in
the watershed area. However, aerial reconnaissance revealed several cabins
along the shore of the lake serving as the main storage reservoir. It was
recommended that the method of sewage disposal used by cabin occupants
be evaluated and a special search be made for evidence of sewage discharged
directly into the lake. To my knowledge this was never done. Similarly, a
stool survey to determine the prevalance of Giardia in humans and animals
in the watershed was proposed but never performed.
In Camas, drinking water was supplied by 2 small streams. The watershed
area was so steep and heavily forested that it could only be entered by
walking up the middle of the stream bed. The stream beds were explored to
their origins and the watershed area was also surveyed from the air. We
Table 2 Source of Giardia c\:\t contamination in walerhorne epidemics of giariiiaMS
SOURCE OF WATER
CONTAMINATION
1975
ROME.
N Y
1 976
CAMAS.
WASH
1977
BERLIN,
N H
Sewage cross-
connection
Human
Animal
No
Undetermined
but possible
Undetermined
but possible
No
Unlikely
Likely
No
Likely
Possible
Beaver
Otter
Mink
Porcupine
Coyote
Opposum
Nutria
3/7*
0/1
0/2
0/1
0/1
1/5
0/5
0/4
0/1
* No positive/No. examined
-------
D.Jwanek 155
found no evidence of human trespassers and no houses in the watershed
area. Since human contamination seemed unlikely, the possibility of an
animal reservoir was considered. Three of 7 beavers captured had Giardia in
their stool. This was the first time that Giardia had ever been recovered from
a beaver. Giardia cysts recovered from the beavers and those recovered from
the city water supply appeared morphologically identical to Giardia lamblia
obtained from infected humans. In addition, beaver Giardia were found to
be infective and cause disease in SPF beagle puppies at Colorado State
University.
The fact that infected beaver were found below, instead of above, the
water intakes weakens the epidemiologic hypothesis that they were the
source of infection (Fig. 1). However, beaver are known to migrate in the
spring, and we found several caches of freshly cut limbs above the water
intakes to attest to the recent presence of beavers in those locations. It is
hypothesized that beaver may have become infected through exposure to
water contaminated with human feces further down stream and that during
their spring migration the beavers contaminated the streams above the water
intakes, thus, precipitating the outbreak.
In Berlin there was ample opportunity for human fecal contamination of
raw water sources. Although 1 of 5 beaver captured in the water area was
infected with Giardia, there was no way to determine whether this animal
was an unlucky victim of water contaminated with Giardia of human origin
or whether beaver served as a major contributing source of Giardia in the
water.
CLINICAL-EPIDEMIOLOGIC PROBLEMS
Two particularly difficult clinical-epidemiologic problems have been
encountered in investigations of suspected waterborne giardiasis. The first is
development of an accurate case definition, i.e., the epidemiologist must
have some method for deciding who does or does not have giardiasis. The
second problem is to determine whether Giardia is the real cause of illness or
merely a commensal organism. In a clinical setting it is not difficult to
determine if a patient is infected with Giardia. A physician simply requests
multiple stool exams or, in some circumstances, obtains a duodenal aspirate
or biopsy. However, in an epidemiologic study, it is not possible to obtain 1
stool specimen, let alone 2 or 3, from each of 400-100(} survey respondents.
Since there is currently no suitable serologic test for giardiasis,
epidemiologists have had to rely heavily on a respondent's description of his
or her illness as a diagnostic method. Unfortunately, the symptom complex
in giardiasis is not pathognomonic for the disease. Several bacterial and viral
agents produce similar illnesses.
The chronic nature of diarrhea in giardiasis is the one clinical feature that
has been used to differentiate giardiasis from bacterial and viral enteritis.
The minimum duration of diarrhea accepted by investigators as being
indicative of giardiasis has varied from one outbreak to another (Table 3). In
Rome, a clinical case of giardiasis was defined as any person who had
-------
156 WATERBORNK GIARDIASIS/EPIDEMIOLOGY
CAMAS CITY WATER INTAKES
FIG. 1. City ofCamas Watershed, May 1976.
diarrhea for 5 days, or more. In retrospect, this case definition may be too
broad. Evidence is increasing that certain bacterial infections
(Campylobacter) and viruses (rotavirus) may cause diarrhea of long
duration. In an English study of Camp v/ofcoc/er-associated enteritis in over
1000 persons, 8% had persistent or recurring diarrhea lasting 2 or more
weeks(5). While the frequency of acute abdominal pain and blood in stools
can be useful in distinguishing Campylobacter enteritis from giardiasis,
these clinical features are not consistently present. In the above study less
than 15% of patients with Campylobacter infection reported these
symptoms or signs. Thus, many Campylobacter infections could be
-------
D.Juranek 157
mistaken for giardiasis. Similarly, while rotavirus is commonest in young
children and usually causes diarrhea for less than 10 days, it does occur in
persons of all ages and may produce diarrhea lasting 5 days to 3 weeks
(usually 8 days)(6). Since no specific search was made for Campylobacteror
rotavirus in Rome and chlorination of Rome water was inadequate, the
possibility exists that 1 of these organisms may have contributed to the size
of the outbreak. Tightening the clinical case definition for giardiasis to a
minimum of 10 days of diarrhea would not have completely excluded
bacterial and viral cases of enteritis from consideration; however it would
have reduced the probability of their being included in numbers large
enough to have had a significant influence on statistical analysis.
Table 3 Climcal-epidemiologic findings in recent walerborne outbreaks o
1975 1976 1977
ROME. NY CAMAS, WASH BERLIN, N H
(Pop 50.000) (Pop 6,000) (Pop 15,000)
Clinical case
definition Duration
of diarrhea in days ^5 ^10 ^7
Clinical attack rate
(No surveyed) 10 6% (1421) 3 8% (503) 5% (676)
Stool positivity rates
(No surveyed) 3 4% (645) 11% (18) 46% (74)
Stool positive but
asymptomatic
(No surveyed) NJIK 0% (9) 45% (51)
* Not done
It should be kept in mind that these observations and criticisms are being
made through the power of hindsight. At the time of the Rome outbreak our
concept of the public health significance and clinical spectrum of
Campylobacter and rotavirus was quite different from that of today. Based
on other epidemiologic data gathered during the outbreak, I feel confident
that Giardia was the principal etiological agent involved, but there is no
assurance that Giardia was the only cause of diarrheal illness lasting 5 or
more days.
In Camas, we attempted to increase the specificity of the case definition
for giardiasis by accepting as cases only those persons who had diarrhea for
10 or more days. In Berlin, data were analyzed using a clinical case definition
of diarrhea lasting 7 or more days; few people in Berlin had illnesses lasting
10 or more days. The clinical attack rates based on the above clinical case
definitions are shown in Table 3. To some extent the attack rates are
inversely proportional to the number of days of diarrhea used to define a
clinical case of giardiasis.
-------
158 VYATKRBORM; (JIARDIASIS/EPIDEMIOLOGY
One of the strong points of the Rome investigation was the large number
of stool specimens that were obtained from survey respondents (Table 3). Of
1421 persons surveyed, 645 submitted a stool specimen and 3.4% were
positive for Giardia. Unfortunately, the proportion of asymptomatic
persons surveyed who were positive for Giardia was not tabulated. This can
be a very useful figure, especially if calculated during the field investigation.
A high rate of stool positivity in asymptomatic persons is suggestive of an
endemic Giardia problem rather than an epidemic one. It alerts the
epidemiologist, early in the investigation, to the possibility that Giardia may
not be the cause of illness, thus making it possible to arrange special studies
to resolve the issue. We attempted to do this in the Camas and Berlin
investigations.
We requested stool specimens from 20 persons in Camas and received
them from 18. Nine of the 18 persons had a history of diarrhea of at least 2
days duration and 9 were well. Of the 18 persons, 2(11%) had Giardia in their
stools; both persons had a history of diarrhea fitting the clinical-case
definition, i.e., diarrhea lasting 10 or more days. Giardia was not found in
stools of asymptomatic persons.
The stool positivity rates in Berlin were quite different from those
obtained in the previous outbreaks. Eleven percent (74/676) of the survey
respondents submitted a stool specimen for examination, and Giardia was
found in 46% of these stools. Sixty-nine percent (51/74) of the persons
submitting a stool specimen were asymptomatic and yet 45% also had a
Giardia infection. By repeating the stool survey 9 months later, we were able
to demonstrate that the high rate of stool positivity in asymptomatic persons
was probably not a reflection of an endemic problem. In the second survey of
54 randomly selected asymptomatic individuals, only 2 (3.4%) were infected.
The data did sensitize us to the fact that a careful search for other possible
etiologic agents would be essential to the investigation. The methods used to
rule out other enteric pathogens are discussed in the next section.
The Berlin epidemic also provided us with a unique opportunity to
evaluate the role of person-to-person transmission of Giardia within
families. Eight stool-positive persons worked in Berlin but resided in nearby
towns. We examined the stools of all 23 of their household contacts and
found none positive for Giardia. This indicates that person-to-person
transmission of Giardia during a waterborne outbreak does not significantly
contribute to further dissemination of the parasite.
RULING OUT OTHER CAUSES OF DIARRHEA
Recent clinical and epidemiologic data support the concept that Giardia is
a pathogenic parasite of humans. However, it must also be acknowledged
that Giardia infection does not always result in clinical illness. For this
reason, finding Giardia in persons with diarrhea during an epidemic has not
been accepted as prima facie evidence that Giardia is the cause of illness. In
order to assess the relationship between Giardia infection and clinical illness
in epidemics, a number of epidemiologic approaches have been tried. One of
the most common approaches has been a retrospective analysis of
-------
D.Juranek 159
laboratory reports. An increase in the frequency of diagnosis of a particular
enteric pathogen in a community coincident with an increase in diarrheal
disease suggests that the organism may be related to the outbreak. This is a
very useful and practical technique, but it does have one drawback — most
local and many state laboratories do not include the more difficult to isolate
bacteria (Campvlobacter, enteropathogenic E. coli, Yersinia) and viral
agents (rotavirus, Norwalk, etc.) in their routine examinations of stool.
Therefore, it has been argued that an increased diagnosis of Giardia by a
laboratory during a waterborne outbreak of diarrheal illness may only
reflect fecal contamination of the water supply and that an undiagnosed
bacterial or viral agent is the actual cause of illness. The probability of this
occuring is highest where chlorination of water is not adequate to kill
bacterial and viral contaminants.
Another method that has been used to rule out other pathogens is to look
for them in ill patients with stool-confirmed giardiasis. If Giardia is not the
cause of illness, then one would hope to demonstrate the real cause. This is a
valid technique provided 1) the laboratory used is capable of culturingand
identifying the more difficult-to-isolate organisms, and 2) adequate
specimens from a control group (persons without diarrhea and without
Giardia in their stools) are obtained at the same time in case another
organism is detected in stool-confirmed cases of giardiasis. The second
provision has not been closely adhered to. We have been fortunate that no
other pathogens have ever been recovered in significant numbers from stool-
confirmed cases of giardiasis.
The third and probably best technique for determining the true cause of
diarrhea in an outbreak is to randomly select 10 to 20 acutely ill persons and
10 to 20 persons without illness for a complete parasitic, bacterial, and viral
stool examination. The specimens must be obtained at approximately the
same time (within 7 to 10 days). For this approach to succeed, it is imperative
that persons planning and/or executing specimen collection know the type
of specimen(s) to be collected (stool, rectal swab, serum) and become
thoroughly familiar with factors that make specimens unsuitable for
laboratory examination such as recent use of antidiarrheals or antibiotics,
recent barium studies, elapsed time between onset of illness and specimen
collection and between collection and preservation of specimens, type of
preservative (fixatives, transport culture media, freezing, etc.), and
appropriate packaging and transport time for laboratory samples.
In all 3 outbreaks discussed here, retrospective analysis of laboratory
reports indicated an increase in the number of stools positive for Giardia
concommitant with an increase in diarrhea in the community. The most
intensive search for other enteric pathogens in stool confirmed cases of
giardiasis was conducted during the Camas investigation. All 128 stool-
confirmed cases of giardiasis identified by local laboratories in that outbreak
were cultured for Salmonella and Shigella. Salmonella was isolated from
only 1 person. Seventy-five of the 128 confirmed cases of giardiasis had viral
cultures; all were negative. However, local laboratory examinations did not
-------
160 WATERBORNE GIARDIASIS/EPIDEMIOLOGY
include a search for toxigenic and invasive Escherichia coli, Yersinia,
Campylobacter, or rotavirus. These studies were performed at CDC on stool
specimens obtained from 9 ill persons and 9 well persons selected at random
from the community. Giardia was found in 2 of the 9 ill persons; no other
enteric pathogens were identified in either group.
In Berlin 20 fresh diarrheal stool specimens arriving at the local hospital
laboratory during the epidemic period were cultured for Salmonella and
Shigella and found to be negative; Giardia was identified in 11 of these stool
specimens. In addition, CDC examined a fresh stool specimen from each of
12 randomly selected Berlin residents with a recent history of diarrheal
illness; 5 of these specimens were collected within 1 week of illness. All
specimens were negative for Salmonella, Shigella, toxigenic and invasive E.
coli, pathogenic vibrios, Yersinia entemcolitica, and Campylobacter jejuni.
Virologic studies included examination of fresh frozen diarrheal specimens
obtained within 1 day of onset of illness in 6 Berlin patients ranging in age
from 11 to 70 years. Rotavirus-like particles were identified in an 11-year-old
boy who was also infected with G. lamblia.
FUTURE NEEDS AND CONSIDERATIONS
The clinical case definition of giardiasis used in epidemiologic
investigations and based on duration of diarrhea leaves much to be desired.
Alternative case identification techniques suitable for epidemiologic
investigations of giardiasis such as a serologic test or a method for detecting
Giardia antigen on rectal swabs are needed. During future epidemics of
giardiasis, evaluations should be made of newly developed diagnostic tests.
However, until new case identification techniques become available, an
attempt should be made to standardize the clinical case definition used for
epidemiologic purposes so that data obtained in one outbreak are
comparable with data obtained in another. Diarrhea lasting for 7 or more
days is suggested as the minimum duration of diarrhea that should be
considered as indicative of a clinical case of giardiasis. Of course, the
patient's illness must be otherwise compatible with giardiasis as well. When
there is evidence of inadequate chlorination of water or in the absence of
good laboratory support to rule out other enteric pathogens, the minimum
duration of diarrhea used in the case definition should be extended to 10 or
more days. It should be recognized that some Giardia infections will result in
diarrhea of a much shorter duration and will be excluded from study by these
case definitions.
The clinical differences in diarrheal disease caused by certain protozoan,
bacterial, and viral agents appears to be decreasing as the amount of new
clinical and epidemiologic information about enteric pathogens increases.
Therefore, greater care will be required in future investigations of
waterborne outbreaks of diarrhefal disease to insure that adequate specimens
are collected for parasitic, bacterial, and viral studies, regardless of the
suspected etiology of the outbreak.
We should continue to use the SPF dog model to demonstrate the viability
and infectivity of Giardia cysts recovered from water in future outbreaks, at
-------
D.Juranek 161
least until equally reliable laboratory methods, such as excystation of
Giardia cysts in vitro or vital staining, can be perfected and their correlation
with results obtained with SPF dogs is established. We should not lose sight
of the fact that Giardia cysts found in water may come from a variety of
different animals and that not all of these cysts are infectious for humans (see
Davies, these Proceedings). At the present time, the only method for
assessing whether Giardia cysts recovered from water are potentially
infectious for humans is to demonstrate that they are infectious for dogs.
More research is needed to elucidate those factors that contribute to the
development of clinical giardiasis. Parasite factors that deserve further study
include dose, strain differences relating to geographic regions, and origin
(human or animal). Host immunity undoubtedly influences the risk of
developing illness, but the exact mechanisms have not been determined.
Outbreaks of giardiasis provide excellent opportunities to study specific
immune factors that may play a role. Development of a simple technique for
evaluating host immunity for application in epidemiologic investigations
deserves further consideration.
The long term solution to control of waterborne giardiasis will
undoubtedly involve improvements in widespread use of water filtration.
However, effective management of emergency problems encountered in
waterborne outbreaks also will be needed, particularly in communities
where water filtration has never been implemented as well as in communities
where existing filtration equipment has failed and cannot be repaired
promptly. Under these circumstances, safe and effective methods for
destroying Giardia cysts in drinking water are certainly needed. Techniques
for making a field evaluation of the efficacy of existing water filtration
equipment would also be helpful. Emergency methods for handling animal
reservoirs implicated as a source of waterborne outbreaks deserve further
consideration. Killing these animals is neither a desirable nor an effective
emergency procedure. Perhaps an effective way to administer antigiardial
medication to target animals in food (e.g., apple pulp for beaver) could be
explored.
REFERENCES
1 Shaw, P.K., Brodsky R.E., I.yman, D.O., et al 1977 A community-wide outbreak of
giardiasis with documented transmission by municipal water Ann Intern Med 87.426-432
2. Dykes, A.C., Juranek D.D., Loren/, R.A., et al : Municipal waterborne giardiasis-beavers
implicated as reservoir. (In preparation)
3. Lopez C.E., Dykes A.C., Juranek D.D., et al : Commumtywide outbreak of waterborne
giardiasis associated with a high rate of asymptomatic infection. (In preparation)
4 Kirner J.C., Littler J.D., Angelo L.A.: 1978. A waterborne outbreak of giardiasis in
Camas, Washington J Amer Water Works Assoc 70'35-40.
5 Communicable Disease Surveillance Centre. 1978. Reports of campylobacter isolates in
1977. Communicable Disease Report 78(13), March 31.
6 Walker-Smith J. 1978 Rotavirus gastroenteritis. Arch Dis of Childhood 52 355-362.
Discussion
R. RENDTORFF: I would like to thank you for a fine presentation. My
only objection was that it did not come before the others so we could
-------
162 WATKRBORNK (ilARDIASIS/KPIDKMIOIXMIY
evaluate those better. The large number of cases reported in Rome, as you
pointed out, is extremely important to epidemiology.
One thing that intrigued me is that you attempted to determine if Giardia
were spread by the individuals who worked in Berlin, N.H. among those
within their households that lived out of the Berlin area. 1 am wondering if
the shedder was a food handler, and if your sampling from household
members was at the time when shedding occurred. In our volunteer
experiments some of these infections are extremely transitory and last but a
few days.
D. JURANEK: We did get into the investigation a little bit earlier than
some of the other outbreaks, so there were ongoing cases at the time the
investigation was conducted. As 1 recall, these 8 individuals that worked in
Berlin but lived outside the community were all men. 1 am making a
supposition that they may not have been food handlers in the household,
although they were excreting Giardia cysts shortly before we examined
household members. We only did 1 stool exam on household members,
however, and there is some question about the reliability of just doing a
single stool exam.
M. SCHULTZ: Also, in the studies that we did of the tourists that went to
Leningrad and came back with giardiasis, there was virtually no clinical
evidence of secondary spread from them to their family members.
T. VERNON: Recently there was a study conducted on next door
neighbor families who were matched to families in which there were cases,
and the prevalence rate in the families next door was the same as the
prevalence rate of the family members of the known cases.
D. PRICE: I am working with a migrant community, and I would like to
know how you would classify endemic and epidemic people in a migrant
situation in which you had fluctuating rises and falls in giardiasis.
D. JURANEK: That is going to be a difficult problem. I would suspect
you would have to follow these same people throughout their, migratory
pathway and see what happens, especially if you have treated them once and
then they go into another community and1 experience the same prevalence of
infection. You have to establish whether they are all getting giardiasis from
an environmental source or whether there is a certain amount of person-to-
person transmission within the families. In a migrant labor situation it would
not surprise me if there was some person-to-person transmission within the
household simply because they usually live under poor sanitation and
hygienic conditions.
D. PRICE: It is my understanding that the chlorine content has been
reduced in the last few years in water supplies. Many of the health
department examinations are not for free chlorine but are for total chlorine.
Does this give us an indication of what is actually going on in the water
supply, particularly in areas where you have a high organic content in the
water?
D. JURANEK: That is my impression, but 1 want to defer to the EPA
experts; they will have to address that.
G. HEALY: Your point about examining asymptomatic individuals in a
population with an outbreak is very important. Would you consider, as an
epidemiologist, that it is worthwhile to conduct a study for the detection of
Giardia cyst in water supplies, and do you know of any situations today
-------
D. Juranek \ 63
where a selected city is now being evaluated? Do you know if EPA has a filter
in use anywhere to determine the presence of Giardia cysts?
D. JURANEK: I believe there is one such study being performed under
EPA funding in the State of Washington.
-------
164
Water Supply Problems Associated
With a Waterborne Outbreak of Giardiasis
Edwin C. Lippy
Health Effects Research Laboratory
U.S. Environmental Protection Agency
Cincinnati, Ohio
ABSTRACT
A waterborne outbreak of giardiasis occurred in the spring of 1977, in Berlin, New
Hampshire. Investigation of the water system identified a source of contamination,
recovered Giardia cysts in samples of raw and treated water, and determined routes by
which cysts passed through the treatment plants. Corrective measures were implemented to
abate the outbreak. Other problems related to the operation of water systems and the
transmission of Giardia cysts are discussed.
A waterborne outbreak of giardiasis occurred in the spring of 1977 in
Berlin, New Hampshire. Results of an epidemiological study indicated that
nearly 7,000 of the city's 15,000 residents were infected with Giardia
lamblia(l). Large volume samples of raw and treated water collected from
the water system were positive for Giardia cysts and routes by which the cysts
passed through the two plants were found. A survey of one of the watersheds
used as a source of supply revealed a beaver lodge in the raw water storage
reservoir. One of 4 beavers inhabiting the lodge was infected with Giardia(2).
This paper will deal with some of the problems encountered in the
investigation of the Berlin Water System and other concerns for water
supplies which are challenged by Giardia cysts.
DESCRIPTION OF THE BERLIN WATER SYSTEM
Berlin is supplied water from two sources; the Upper Ammonoosuc River
and the Androscoggin River. The Upper Ammonoosuc River, located west
of Berlin, is impounded by Godfrey Dam which forms a water supply
reservoir of 99.4 x 103m3 (26 MG) capacity. The heavily wooded watershed is
located within the confines of the White Mountain National Forest and
under control of the National Forest Service. There are no point-source
discharges upstream of the reservoir; however, hunting, fishing, and other
forms of recreation are permitted. Water is transported from the reservoir
through a bottom intake to a treatment plant located on the western fringe of
the city as shown in Figure 1.
The Ammonoosuc Treatment Plant consists of eight pressure filters of
cylindrical shape that are 2.4m (8 ft) in diameter and 5.5m (18 ft) in length.
The filters contain graded anthrafilt media and are operated in parallel
fashion. Chemical conditioning of water prior to filtration is not practiced.
-------
£". C. Lippy
165
z
o
n
O
O
O
z
ANDROSCOGGIN
TREATMENT PLANT
AMMONOOSUC
TREATMENT PLANT
GODFREY DAM
AND
RESEVOIR
FIG. 1. Simple schematic diagram of Berlin, N.H, water system.
Filter effluent is chlorinated and the treated water is introduced into the
distribution system that mainly serves the western portion of the city.
The Androscoggin River, which is the second source of supply, flows
through the city and has a drainage area of approximately 3367 km2 (1,300
sq mi). Average flow at Berlin is 67m3/ sec (2,400 cfs). The drainage area is
mostly timberland and there are no significant point-source waste discharges
upstream of the treatment plant.
Water is pumped from the river to a recently completed treatment plant.
Conventional treatment is provided including chemical coagulation,
clarification, rapid sand filtration and chlorination. Treated water is
distributed mainly to the eastern portion of the city; however, there are
several river crossings and a zone of mixed water occurs in the system
adjacent to the Androscoggin River.
The city had not experienced any unusual water quality problems prior to
and during the outbreak as measured by conventional parameters for raw or
-------
166 WATERBORNK CIARDIASIS/EPIDKMIOLOdY
finished water. Coliform organisms were not detected in finished water
based on 5 weekly samples and historically were extremely low for both raw
water sources usually occurring in the range of less than 100 colonies/100 ml.
Turbidity values for raw water were normally less than 5 Nephelometric
Turbidity Units (NTU) with fluctuation occurring during periods of runoff.
The values were acceptable for finished water when compared with
applicable standards at that time. There was a slight color problem in
finished water from the Ammonoosuc source presumably resulting from
decaying vegetation.
WATER SUPPLY PROBLEMS
Giardia cysts were detected in large volume samples of raw and finished
water from the Ammonoosuc and Androscoggin treatment plants and in
samples of finished water from several locations in the distribution system.
The results presented a number of problems including: source of the
contamination, the reason for passage through the treatment plants,
abatement procedures that could be utilized to prevent further illness, and
other concerns that deserve attention.
Source of Contamination
It was reasonable to conduct a sanitary survey in the Ammonoosuc
watershed to determine the source(s) of contamination because of its size but
the prospect of inspecting the entire Androscoggin drainage area (1,300 sq
mi) was not seriously considered. Inspection of the perimeter of Godfrey
Reservoir showed a beaver lodge about 610m (2000 feet) upstream from the
Ammonoosuc plant intake. Four beavers were eventually trapped with
permission of the State Game Commission and autopsied at a local hospital.
Samples of material from the gastrointestinal tract of the animals were
initially examined and the findings were negative. The pathologist then
prepared slides from material extracted from the duodenal mueosa and
positively identified Giardia trophozoites in one of the beavers.
Trapping and sacrifice of wild animals during the outbreak was
unpopular. Resentment existed in the public and governmental agencies not
aware of the implication of beaver in a previous outbreak of giardiasis in
Camas, Washington(3). There are alternatives to sacrifice of animals such as
collection of fecal specimens from their habitat, or live trapping the animals
for study under captivity. If the animals are actively shedding cysts, the
results are soon available; however, they may be infected and not shedding
cysts as was the case with the Godfrey Reservoir beaver. There is also the
question of resources to support continued study of an animal in captivity
where its habits and shedding conditions may change, and the length of time
that one can wait for results. Collection of fecal samples from the vicinity of
the beaver lodge was not attempted. The water in this area was 3 to 4.6m (10
to 15 ft) deep with a temperature of about 2 to 3°C and there were no
volunteers available to obtain a fecal sample.
It would have been desirable to conduct profile sampling of the
Androscoggin River to locate sources of Giardia but there was not sufficient
laboratory support for sample analyses. A limited sanitary survey was made
-------
E. C. Lippy 167
and over 40 cease and desist orders were issued to homes and institutions
with individual sewage disposal violations.
Based on the limited information available it was concluded that the
source of cyst contamination in the Ammonoosuc watershed was beaver,
possibly infected by human carrier. The source of cyst contamination in the
Androscoggin River was unknown.
Cyst Passage Through Treatment
Giardia cysts were detected in raw and finished water at both treatment
plants during the outbreak using sampling methodology developed by
EPA(4). Treatment processes and facilities were evaluated to determine how
cysts passed through the plants.
The Ammonoosuc Plant relied solely on filtration for particle removal
while chemical coagulation and clarification were provided prior to
filtration at the Androscoggin Plant. The effectiveness of either form of
treatment for Giardia cyst removal had never been investigated. Cysts of E.
histolytica which are similar in size to Giardia cysts have been studied in pilot
plant experimentation^). The researchers found that E. histolvtica cysts
were removed by conventional treatment similar to that provided at the
Androscoggin Plant, and that they were not removed by simple filtration.
These results are somewhat reinforced by experimentation conducted by the
Center for Disease Control during studies of sampling methodology for
Giardia cysts in drinking water. The method consisted of passing a large
volume of water through a swimming pool filter in which the normally used
diatomaceous earth medium was replaced with 0.45 m (18 in) of 0.45 mm
sand. In a laboratory experiment, Giardia cysts at a concentration of 20/ 3.81
(1 gal) were dosed onto the filter and after passing 34,200 1 (9,000 gal), no
cysts were detected microscopically in the concentrated preparations of filter
backwash(6).
When filtration is used as the only treatment process, removal of applied
material depends on the, mechanism of straining. While forces of
sedimentation do occur in channels and voids in the filter bed, the
predominant mode of removal is physical straining or trapping of
paniculate matter. The trapping capability of a filter bed can be visualized
by considering the size of the voids. In a filter containing sand with a
diameter of 0.5 mm, the pore openings or voids will range in size from 0.1 to
0.2 mm. With coarse angular filter media such as carbon-based material, the
pore openings are larger, ranging from 0.3 to 0.6 mm. Particles that are easily
trapped are larger than the pore opening while smaller particles may pass
through. Giardia cysts measuring 10 to 14 /urn can be easily transported
through pore spaces that are 10 to 40 times their size. For filtration to be
effective in removing small particles, it should be preceded in the treatment
train by chemical conditioning or coagulation/flocculation. This enables
the formation of larger particles and permits filtration to reduce the applied
load of turbidity and microbial populations through mechanisms of strain-
ing, sedimentation and flocculation in the voids, and adsorption.
-------
168 WATERBORNE C.IARDIASIS/EPIDEMIOLOGY
The passage of cysts through the Ammonoosuc Plant was accommodated
by the condition of the filter beds. Pressure filters are enclosed vessels and
inspections of the filter beds are infrequent because access is difficult.
Inspection of the filters during the outbreak showed that three had severely
disrupted bed surfaces characterized by mounds, pyramids, and shrinkage
resulting in separation of the beds from vessel walls. The filters were plagued
with agglomeration of mud masses within the beds and extended
backwashing sequences were not successful in cleaning the media. One filter
had a large conical depression covering approximately one-third of the bed
area which resulted from a break in the main header of the air wash system.
During backwash, air that was normally distributed uniformly across the
bed cross-section for scouring action, was discharged full-force from the
header break. The disrupted bed conditions changed the profile of filter
surfaces and filtration occurred through only a few inches of media instead
of 61 cm (24 in) of a normal filter bed.
Passage of cysts through the Androscoggin Plant was not expected
because the type of treatment was thought to be effective for removal. The
passage of cysts was initially thought to be related to inefficient control of
coagulation and solids removal prior to filtration. This possibility was
investigated and resulted in a recommendation by the process designers that
other chemicals be used. However, further study of plant operation detected
escape of bubbles through the joints in the slab of the backwash channels
during the filter washing sequence (Figure 2). The washing sequence
included the introduction of air into the filter bottom to permit scouring and
cleansing of the filter media. The escape of air through the joint indicated
that water applied to the filters could be by-passing treatment and
contaminating filtered water. It also indicated that during the backwash
sequence, water laden with concentrated filterable material including
microorganisms would pass through the joint and contaminate filtered
water. Leakage through the joint was confirmed by static hydraulic test of
the backwash compartment and results showed that daily leakage for one
filter that was tested could be 45m3 (12,000 gal). The test result if
extrapolated to operation of all filters meant that nearly 187m3 (50,000
gal)/day of unfiltered water were pumped into the distribution system.
While repairs were underway to seal the joints, Giardia samplers were
installed on the effluent lines of a repaired filter and another that still had
faulty joints. The samplers were installed to test the capability for removal of
cysts. The sampler is a yarn-wound orlon filtering device that traps
particulate matter including Giardia cysts. The sampling results in Table 1
show the difference in operation of the two filters. The most notable
difference is the volume of water that was passed through the sampler from
the repaired filter as opposed to the one not repaired. At the termination of
the sampling period, the sampler on the unrepaired filter was nearly plugged
with particulate matter. Abatement ProCedures
The presence of cysts in finished water with no practical way of controlling
or identifying all sources of contamination in the watersheds and the
-------
E. C. Lippy
169
PATH OF FILTERED WATER
PATH OF AIR DURING
WASH SEQUENCE
FIG. 2. Isometric view of the Androscoggin plant filters (not to scale).
Table I Ammonoosuc plant filler le-,1. Berlin, New Hampshire, 1977
Filter No 1
(Not Repaired)
Filter No 2
(Repaired)
Volume Filtered (Gal)
285
928
Turbidity (NTU)
0 23
0 12
Colonies/100ml
3 (Start)
<1 (Stop)
<1 (Start)
<1 (Stop)
uncertainty associated with cyst removal by the Ammonoosuc Plant (and
the Androscoggin Plant until repairs were completed) reduced the
abatement measures to disinfection and boiling water.
There were a few problems associated with disinfection, the major one
being the lack of information on the destructive dose for Giardia cysts. In the
absence of specific data, concentration-contact time relationships for cysts
of E. histolytica developed by Chang in 1967 as shown in Figure 3 were used
(unpublished data, S.L. Chang). For disinfection to be effective at doses
normally applied in water treatment, the contact time requirement is beyond
the capability of most treatment plants. Therefore, the dose must be
increased to compensate for lack of contact time.
A chlorine contact basin was not incorporated into the design of the
Ammonoosuc Plant. The plant effluent was chlorinated and the
transmission main between the plant and the distribution system allowed
-------
170
WATKRBORNE GIARDIASIS/KPIDEMIOLOGY
10
LU
z
tr
O
_i
x
CJ
D
Q
CO
0.1
0 01
CYSTS OF
E. HISTOLYTICA
0 1
10
100
TIME (hr)
FIG. 3. Concentration-time relations for 99.999 percent destruction of E. coli and cysts
of E. histolytica by free chlorine at 2 to S°C (Chang, 1967).
about one hour of contact time. This required a chlorine residual of 6 mg/1
which was temporarily applied until emergency pre-chlorination facilities
were installed at Godfrey Dam. The installation at the dam permitted use of
8.8 km (5.5 mi) of transmission main for contact time which reduced chlorine
dosage requirements to more acceptable concentrations. The heavy dose of
chlorine initially applied at the Ammonoosuc Plant did provide an
unexpected benefit in water quality by eliminating a color problem.
Finished water storage facilities were incorporated into the design of the
Androscoggin Plant and were used as chlorine contact basins. Capacity of
the storage facilities was sufficient to provide about 7 hours of theoretical
contact time based on daily plant output. A chlorine residual of 2 mg/1 was
initially applied; however, a review of a day's operating results showed that
the contact time was less than 2 hours due to short-circuiting in the storage
facilities. Chlorine feed was then adjusted to meet the requirements shown in
Figure 3.
The issuance of a boil-water order is unpopular with water utility
management as well as the public. Management resents it because of adverse
public reaction and the public does not like to be inconvenienced. A boil-
-------
E.C.Lippy m
water order was issued in Berlin as a protective measure until repairs and
adjustments in treatment were made at the two plants and samples of water
were negative for cysts. Unfortunately, the repairs required a longer time
than expected and the boil order was in effect for six weeks. The time of boil
to destroy cysts was not known so a conservative estimate of 5 minutes was
initially recommended. However, there were objections because it was not
substantiated by specific data. The local hospital conducted some boiling
experiments and found that 15 to 20 minutes were required to rupture the
cysts. The recommendation to boil water was then changed from 5 to 20
minutes. The inconvenience of boiling water over a lengthy period of time
coupled with rising energy costs did create resentment and some individuals
threatened to stop paying their water bills.
OTHER PROBLEMS AND CONCERNS
A few other problems and concerns deserve discussion. The raw water
intake at Godfrey Reservoir was located at the base of the dam. The dam was
constructed in the 1930's when the design consideration was to locate the
intake at a position which would maximize use of the storage capacity of the
reservoir. Current practice requires installation of a number of outlet ports
to enable selective withdrawal from several elevations to obtain the best
quality of water in the reservoir as conditions change throughout the year.
The specific gravity of Giardia cysts is greater than water and settling occurs
under quiescent conditions of a reservoir. A bottom intake withdraws water
from the depths of a reservoir, the area of greatest cyst concentration. This
should be avoided if possible for water supply purposes by use of an intake
that permits withdrawal from near the reservoir surface.
A concern in water treatment pertains to filtration efficiency during the
period immediately following backwash. When a filter is washed and put
into operation it takes some time for the bed to stabilize. If effluent turbidity
is measured from the beginning of filter operation a spike is noted which
generally lasts for about 30 min. Turbidity levels of up to 50% of the turbidity
of applied water is measured in filtered water(7). The spike is termed initial
breakthrough and, because it normally has no great effect on plant effluent if
evaluated in terms of turbidity, it is accepted and no corrective action is
taken. It is counteracted in some plants by discharging filtered water to waste
during the spiking period. The effect of initial breakthrough is subdued in
most plants because multiple filters are discharging to a common effluent
manifold and the increase in turbidity from one filter is not sufficient to
affect the turbidity level as measured in the total plant output. The concern
with filter spike is that it may present the opportune time for passage of cysts
and contamination of the water system may occur on an intermittent basis.
A filter at the .Ammonoosuc Plant was sampled for turbidity after
backwashing, and the results shown in Table 2 demonstrate initial
breakthrough.
Another concern in water treatment is the disposal of filter backwash and
sedimentation basin sludge from plants challenged by Giardia. The removal
and concentration of cysts by water treatment processes may prevent an
-------
172 WATERBORNE GIARDIASIS/EPIDEMIOLOGY
Table 2 Ammonoowc plant filter test
Berlin, New Hampshire, 1977
Filtered Water
Sampling Sequence Turbidity (NTU)
Prior to backwash 0 75
Filter in operation
- Start
- 1 mm
- 2 mm
- 5 mm
- 1 0 mm
- 1 5 mm
1 55
4 25
2 1
1 5
1 35
1 4
outbreak from occurring in one community but the discharge of cyst-laden
backwash and sludge may create problems for downstream uses including
water supply, recreation and possibly shellfishing. The long-term survival of
cysts in water compounds the problem and cyst laden discharges could have
far-reaching effects. Also of concern is the discharge of inadequately treated
sewage in areas where outbreaks of giardiasis have been notable.
Information on wastewater discharges for three states in the northeast is
shown in Table 3 (personal communication, P. Karalekas) and indicates a
need for concern.
Finally, the heavy doses of chlorine required to destroy cysts create
problems with the formation of undesirable treatment by-products. The
reaction of certain organic materials with chlorine produces
trihalomethanes which are currently under investigation as carcinogens.
While chlorination may be providing a solution to an acute health problem
in the community, it may be contributing to the occurrence of a chronic
situation. While the public naturally reacts to odors and tastes resulting from
heavy doses of chlorination, the threat of cancer produces a different
reaction. The recommendation to boil water to counteract taste and odor
problems has the added benefit of eliminating trihalomethanes. A few
minutes of boiling drastically reduces trihalomethane concentrations as
shown in Table 4 (personal communication, T. Love). Therefore, boiling
Table 3. Wastewaler discharges
Maine, New Hampshire, Vermont, 1978
Sewage Discharges Backwash Water
State Raw or Chlorinated Only and/or Sludge No of Downstream
(No o1 communities) (No of plants) Water Supplies
Maine
New Hampshire
Vermont
68
40
27
12
2
5
0
2
27*
* All these communities use Lake Champlain as a source of water supplv.
-------
E.C.Lippy 173
Table 4 Tnhalomethane removal b\' boiling
Time of Boiling Trihalomethane Removal
(mm) (%)
<1 10
1 80
2 96
5 >99
water may be a necessary adjunct to chlorination for protection of public
health in some cases.
The problems and concerns presented above are not unique to the Berlin
Water System and it is hoped that their discussion will be of benefit to others
involved in waterborne outbreaks of giardiasis.
REFERENCES
I. Lopez, C. E.; Dykes, A. C.; Juranek, D. D.; Sinclair, S. P.; Conn, J.; Christie, R. W.;
Lippy, E. C.; Schultz, M. G., and Mires, M. H. Community wide outbreak of waterborne
giardiasis associated with high rate of asymptomatic infection. (Submitted for
publication)
2 Lippy, E. C. 1978. Tracing a giardiasis outbreak at Berlin, N H. J Am Water Works Assn.
70:512-520.
3 Kirner, J.; Littler, J. and Angelo, L. 1978. A waterborne outbreak of giardiasis in Camas,
Washington J Am. Water Works Assn 70:35-40.
4. Jakubowski, W.; Chang, S. L.; Ericksen, T. H.; Lippy, E. C.; and Akin, E. W. Large
volume sampling of water supplies for microorganisms. J. Am. Water Works Assn.,
70-702-706.
5. Specter, B. K.; Baylis, J. R.; and Cullans, O. 1934. Effectiveness of filtration in removing
from water, and of chlorine in killing, the causative organism of amoebic dysentery U.S.
Pub. Hlth. Rep 49:786-800.
6. Jakubowski, W.; Ericksen, T. H.; and Chang, S. L. 1977. Detection, identification and
enumeration of Giariiia cysts in water supplies, Proc AWWA Water Quality Technol.
Conf, Kansas City, Mo . December 1977.
7. American Water Works Association 1971. Water quality and treatment. 3rd ed. McGraw
Hill Co., N.Y. 654 p.
-------
174
An Outbreak of Gastroenteritis
Associated with Giardia Lamblia*
Lyle Veazie, Inez Brownlee, and H. J. Sears
Department of Bacteriology,
University of Oregon Medical School
Portland, Oregon
There is still controversy concerning the pathogenicity of the intestinal
flagellate, Giardia lamblia. For many years it was generally considered a
harmless commensal, because of its presence in many healthy individuals
and its lack of tissue invasiveness. Experimental infection of human
volunteers resulted in transitory changes in stool pattern in some of the
subjects, but not in clinical illness (1). There have been, however, numerous
reports of cases of intestinal disturbances associated with the presence of
Giardia, and without any other detectable cause (2-4). The relief of
symptoms in many of these patients after eliminating the giardias with
quinacrine has greatly strengthened the case for the pathogenicity of the
parasite (5-8).
The manifestations of Giardia infection are believed to be due to chronic
irritation of the duodenum resulting from the attachment of enormous
numbers of trophozoites to the surface of the mucosa (9). The most common
symptoms are diarrhea, flatulence and upper abdominal discomfort
(3,5,6,8). Steatorrhea is often noted, especially in children (9), who tend to
have both a higher carrier rate and a higher rate of symptomatic infection
than do adults. The disease is usually sporadic in incidence, though one
outbreak involving 44 persons, over half of them children, was reported in
England in 1942 (10).
The present report gives the results of an investigation made during an
outbreak of gastroenteritis which occurred in Portland, Oregon during the
fall and winter of 1954-55. The source, mode of spread, and nature of the
etiological agent were never satisfactorily determined. A striking increase in
the incidence of Giardia in the population at the time of the outbreak and its
correlation with cases of diarrheal disease are considered worthy of note,
especially since no published reports of similar epidemics could be found.
EPIDEMIOLOGY AND CLINICAL ASPECTS
The outbreak began in October, 1954 and continued until March, 1955,
reaching a peak in late December and early January. The Oregon State
*This paper was not presented but has been included in the Proceedings at the request of the
editors. The manuscript was prepared about 1959 but never published.
-------
L. Veazie '75
Board of Health estimated that at least 50,000 cases occurred during that
period. The outbreak appeared at the time to be confined to Portland and
the surrounding suburban communities, though later reports made it seem
probable that similar outbreaks occurred elsewhere in the state. Age, sex,
place of residence or work within the city seemed to play no role in the
distribution of cases, except that relatively more cases occurred among
adults than among children.
The outbreak was unique in the experience of the city, both in the numbers
of people involved and in the character and duration of symptoms. In 500
cases investigated by the epidemiologists, the average duration was 14.8
days, with an extreme range of 1 to 120 days. A third of the cases were of the
chronic, intermittent type. Abdominal discomfort, diarrhea, loss of appetite
and nausea were the most frequent symptoms. From histories obtained
during the course of our investigation, it appeared that a typical individual
experience was a series of episodes of diarrhea, occurring at intervals over a
period of several weeks. The stools during the acute attacks often were
watery, pale and fatty, containing no blood or pus. Though the acute
episodes of diarrhea often lasted only a few days at a time, the discomfort,
especially in the upper abdomen, the nausea and lack of appetite often
persisted between attacks. Most of those affected were able to continue their
usual activities; however, those afflicted more severely required
hospitalization for uncontrollable diarrhea and marked weight loss.
Neither preceding nor during the outbreak was there any reported
increase in the number of infections with salmonella, shigella or amebae. A
marked increase in the incidence of Giardia found in fecal specimens from
patients with diarrhea, especially in those specimens which were liquid, pale
and fatty in appearance, was noted by some of the clinical laboratories in
October. As the possibility of a relationship between the presence of the
flagellate and the current syndrome became known, many physicians began
to use quinacrine in treatment. Many of them reported excellent results, with
marked improvement in symptoms within 24 to 48 hours, even in cases
which previously had been treated unsuccessfully with a variety of other
drugs. Others reported equally good results with other methods of
treatment. Unfortunately, the number of reports containing parasitological
data was too small for any statistical analysis by the epidemiologists.
Physicians frequently prescribed quinacrine only on the basis of symptoms,
without ordering a stool examination. The lack of clear-cut evidence led the
State Board of Health to conclude that the outbreak was caused most likely
by an unidentified virus, but the unusual prevalence of Giardia lamhlia cysts
in stools of patients seemed worthy of record.
Because of the strong public interest in the outbreak, the differences of
opinion among members of the medical profession concerning the
significance of Giardia in relation to the disease, and the difficulty of
obtaining adequate data, an investigation was conducted among a group of
people from whom reasonably complete records could be obtained. The
study was initiated in January, 1955, at a time when the outbreak was
declining.
-------
176 WATKRBORNE GIARDIASIS/EPIDEMIOLOGY
PROCEDURES
The group chosen for the survey consisted of 81 persons, 68 men and 13
women, all of them students or employees of the Medical School. Twenty-
nine other individuals were also studied, but not included in the survey
group. Most of them were students or employees who were suffering from
gastroenteritis, and the others were members of the families of those found
to be harboring Giardia.
Each member of the survey group submitted up to 3 stool specimens for
examination, unless Giardia was found in one of the specimens. Those found
to be harboring the parasite were not required to submit further specimens.
The number of examinations performed on the patient-contact group varied
widely, but usually consisted of only 1 specimen. A few patients, in all of
whom Giardia had been found on the first examination, submitted several
specimens over a period of weeks. Each person examined filled out a
questionnaire, stating whether he had had a gastrointestinal disturbance of
any kind during the preceding 6 months, the nature of the symptoms,
duration, type of treatment, and the results of previous stool examinations.
Fresh, unpre'served specimens were used, and all were examined in iodine-
stained wet mounts after concentration by the formalin-ether technic. Those
specimens which were of softer than normal consistency were examined in
saline and iodine mounts before concentration. All soft specimens, as well as
those found to contain Giardia and all those from individuals with a history
of present or recent gastrointestinal illness, were also examined for bacterial
enteric pathogens and the serologicalO group of 10 Escherichia coli colonies
was determined. Iron hematoxylin stains were made from 20 of the
specimens which contained large numbers of Giardia.
RESULTS
The bacteriological studies revealed no enteric pathogens, and no pattern
of E. coli grouping appeared. The incidence of intestinal protozoa other
than Giardia did not differ from that found in similar groups in the past.
Of the survey group, 37% (30 of 81) were found to be harboring Giardia.
Forty-two percent of the survey group gave a history of a gastrointestinal
disturbance within the previous 6 months, and 21% were having symptoms
of diarrhea, flatulence, abdominal pain or discomfort at the time the
examinations were made. Since no significant difference was found in the
number of positive results in those without symptoms and those who had
recovered from previous symptoms before the examination was made, only
those who were symptomatic at the time of examination will be considered
separately. Table 1 shows the comparative incidence of Giardia in the two
groups. Included for comparison are the results obtained from 112 students
and employees examined at other times. Fifty of these 112 persons were
studied prior to the epidemic year, most of them having had 5 specimens
examined. The remainder were students in the class following the one
examined during the outbreak and were studied at the same time the
following year. Procedures for the examination of specimens were the same
for all groups.
-------
L. Veazie
Table 1. Comparative incidence of Giardia lamblia during
epidemic and non-epidemic periods.
Ill
Epidemic period
Giardia +
Giardia -
Total
Gastro-
intestmal
symptoms
present
15
2
17
Gastro-
intestinal
symptoms
absent
15
49
64
Total
30
51
81
Non-epidemic period
Gastro-
ntestmal
symptoms
present
0
2
2
Gastro-
intestinal
symptoms
absent
8
102
1 10
Total
8
104
1 12
Statistical analysis of the results showed that, in spite of the fact that only
half of those harboring Giardia were having symptoms during non-epidemic
periods, a highly significant association existed between the presence of
Giardia and symptoms during the epidemic period. The Chi square for the
association was 21.5, as compared with an expected 3.184 for a chance
probability of 0.05.
All but 5 of the 29 persons in the patient-contact group were having
symptoms at the time of examination. Giardia was found in 2 of the 5 normal
persons and in 16 of the 24 patients. Those with negative findings were
examined only once. Since one of the characteristics of the outbreak was the
chronicity of the symptoms and since several of those with symptoms were
examined during or immediately after a short, acute attack of diarrhea
which subsided spontaneously within a day or two, a comparison was made
of all those who had symptoms, both among the survey group and the
patients, based on the duration of illness. To simplify the analysis, an
arbitrary division was made based on duration of more or less than one
week. The results are shown in Table 2.
Of the 3 people in the "chronic" series who were negative, 1 was a baby
whose diarrhea followed a series of immunizing injections, and a second was
an adult who developed diarrhea after intensive antibiotic therapy for an
infection. None of the 3 was examined more than once. All of the people who
were having symptoms characteristic of the outbreak, and from whom at
least 3 specimens were obtained, were found to be excreting Giardia, usually
in enormous numbers. Twelve of them received quinacrine after the
parasites were found and all reported prompt and complete relief of their
symptoms, which in some of the cases had been present for 2 months. At
Table 2 Relation between duration of symptoms and presence of Giardia
in 41 i uses of gastroenteritis
Duration of symptoms
More than 1 week
Less than 1 week
No of
persons
31
10
No positive
for Giardia
28
3
% positive
90
30
-------
178 \\ATERBORNK GIARDIASIS/KPIDKMIOLOGY
least 7 of the 12 relapsed shortly after discontinuing therapy. In 2 of those,
the only ones that were studied adequately, the flagellates were not
demonstrable during the symptom-free period, but were found in large
numbers during relapses. Two others, who were not examined again,
reported relapses and responded promptly to quinacrine treatment.
The characteristics of the onset of symptoms were reported in 26 of the 28
chronic cases in which parasites were found. In only one of these was the
onset described as abrupt. The 25 others described onset as gradual or
insidious, with symptoms increasing in severity over a period of days or
weeks, or as maintaining from the beginning a fairly uniform, low-grade
course.
Only 14 people in the entire series were examined within 5 days of onset of
their first symptoms. Half of them were negative for Giardia. Two of these
patients were babies with mild to moderate diarrhea. Three of the adults had
attacks lasting only a few hours, with nausea and vomiting as prominent
symptoms. The others had attacks lasting 2 to 3 days. In all, symptoms
subsided without specific treatment and without known recurrences.
Three of the 7 patients with Giardia also had brief attacks, without specific
therapy or known recurrences, though 2 were still excreting the parasites
when re-examined a month later. The other 4 had a more prolonged course.
One recovered spontaneously after 9 days, though his stools still contained
Giardia on re-examination 6 weeks after recovery. Two had typical severe,
chronic cases, which responded promptly when quinacrine was finally
administered, but both later had at least 2 relapses. The fourth experienced
mild symptoms of abdominal discomfort and soft stools for at least a week
following a severe attack of diarrhea, but was not examined again.
In none of our cases was it possible to demonstrate to our complete
satisfaction that the patient was parasite-free until the onset of symptoms,
and developed chronic disease with the establishment of Giardia infection.
The patient who most nearly fulfilled the requirements had undergone
numerous parasitological examinations in the past, none of them positive,
and had large numbers of Giardia on the first examination following the
onset of symptoms. The disease in this case was typical in the nature of
symptoms, long duration and response to therapy. The last negative
examination, however, had been made several months before the attack.
Another reported having had slight diarrhea and minor malaise during the
second week in January, the symptoms subsiding without treatment. Three
stool specimens, examined between January 24 and February 2, were
negative for Giardia. Diarrhea and intestinal discomfort reappeared on
March 4, after a course of sulfonamides for a sinus infection. A stool
specimen examined on March 9, while the diarrhea was still severe,
contained enormous numbers of Giardia. The symptoms ceased within 24
hours after the administration of quinacrine. Both of these cases strongly
suggest the association between the parasites and the diarrheal attacks, but
neither provides proof of the association.
-------
L. Veazie 179
DISCUSSION
The results of this investigation, and a comparison of these results with
those of other periods, make it clear that there was an entirely abnormal
incidence of Giardia infection in the group studied at the time of the
outbreak. The flagellate was found in 44% of those studied during the
outbreak, in contrast to 7% of those examined during non-epidemic periods.
Even if only those who gave no history of present or past intestinal
disturbance are considered, the rate of infection was 24%, or over 3 times
that of the control figure.
It is apparent that the presence of the parasite did not necessarily result in
disease, since half of those infected were asymptomatic. This is in accordance
with the observations of other workers (3). The role of Giardia in the
transitory cases, either mild or acute, is difficult to assess, because of the
small numbers which were studied and the high carrier rate that existed at
the time of the study. Undoubtedly many, and very possibly all, of the acute,
brief episodes of vomiting and diarrhea were due to other causes, since they
were similar in symptomatology and course to cases which occur
sporadically every winter. It does not appear from our data that Giardia
played any role of consequence in these. Its association with the several mild
cases of transitory abdominal discomfort and a few unwontedly soft stools is
also undetermined, especially since its continued presence did not lead to
continued symptoms.
The high rate of repeated relapses among the quinacrine-treated cases is
contrary to the usual experience, and to the experience here with cases which
have occurred sporadically since the outbreak. It is quite possible that the
apparent relapses were actually reinfections, the people acquiring them
being highly susceptible and the possibilities for reinfection being extremely
great at the time.
We feel that our results do give support to the belief that Giardia was
responsible for the chronic syndrome which was the outstanding feature of
the outbreak. This belief is based on the constant presence of the parasites in
those suffering from this syndrome, the uniformity of their symptoms, the
favorable effect of quinacrine, and the similarity of the symptoms to those
described by others as characteristic of giardiasis. The infrequency with
which acute attacks were recorded as preceding or initiating the symptoms in
the chronic cases suggest that the enhanced pathogenicity of the parasite in
these cases was not due to the action of another acute infection in preparing
the way for its establishment and unrestrained multiplication. We have no
adequate explanation for the extreme variability in the results of Giardia
infection in different individuals, nor for the circumstances which led to its
exceptional prevalence during this period. We doubt, however, that the
experience here was unique, and suggest that giardiasis be more seriously
considered than it usually is both in sporadic cases in adults and in outbreaks
in which the symptomatology is similar to that described here.
-------
180 WATERBORNE GIARDIASIS/EPIDEMIOLOGY
SUMMARY
An outbreak of gastroenteritis, characterized by the exceptionally
prolonged course of many of the cases, is described. A survey made during
the outbreak revealed that only Giardia lamblia, of the known possible
etiological agents of such a syndrome, was detectable in the majority of those
affected. The results of the survey are presented and discussed, and the
reasons given for the conclusion that the prolonged attacks were probably
chiefly due to Giardia infection.
(Editors Note: Supplementary information developed by the Oregon
State Board of Health and supplied by the author follows the References).
REFERENCES
1. Rendtorff, R. C. 1954. The experimental transmission of human intestinal protozoan
parasites. 11 Giardia lamblia cysts given in capsules. Amer. J. Hyg 59.209-220.
2. Fraser, J. F., and Taylor, R. 1945. Diarrhoea due to Giardia lamblia Brit. Med. J. 2-184-
185.
3. Hartman, H. R., and Kyser, F. A. 1941 Giardiasis and its treatment J. Am Med. Assn.
116.2835-2839.
4. Maris. E. P., and Bushong, S. 1942. The diagnosis of Giardia inteMmahs infestations by
means of intestinal intubation Pennsylvania Med. J., 45.724-726.
5. Webster, B. H. 1958. Human infection with Giardia lamblia Am. J. Dig. Dis., 3-&4-71.
6. Cpurt, J. M., and Anderson, C. M. 1959 The pathogenesis of Giardia lamblia in children.
Med. J. Australia 2:436-438.
7. Katsampes, C. P., McCoord, A. B.,and Phillips, W. A. 1944. Vitamin A absorption test in
cases of giardiasis. Am. J. Dis Child. 67.189-193.
8. Kean, B. H. 1952. The flagellate diarrheas New York State J Med , 52:2655-2657
9. Veghelyi, P. V. 1940. Giardiasis Am. J Dis. Child. 59:793-804.
10 Ormiston, G., Taylor, J., and Wilson, G. S. 1942. Enteritis in a nursery home associated
with Giardia lamblia. Brit. Med J. 2:151-154.
-------
L. Veazie
181
DEPARTMENT OF HEALTH, EDUCATION, AND WELFARE
Public Health Service
RFf'ORl OF UlStASF GUI BREAK
PHS 767
STATE
OREGON
COUNTY J CITY OR TOWN
MULTNOMAH [METROPOLITAN AREA OF PORTLAND, OREGON
TYPE OF DISFA5E OUTBREAK
GASTROENTERITIS
NAME OF COMMUNITY. CAMP.
INSTITUTION
^OPULATION
UNDETERMINED
0*Tt OF ONSET
MPST TA5L ABOUT 11/20/54 LAST CASE
NUMBER OF CASFS XKKKKKX ESTIMATED
50,000
NUMBER OF DEATHS
UNCERTAIN
CASES INVESTIGATED EP 1 DEM 1 OLOG 1 CALL Y
CASES CONFIRMED BY LABORATORY EXAMINATION
LASLS WHICH CONTACTED A SUSPECTED VEHICLE
LE( "
NUMBER
537
-
-
L-l SHELLFISH
!_ j OTHER FOOD (stn
WATER L.J UNIDENTIFIED
' i YES } NO
- -i NATURAL
-J GROWING AREAS
| 1 FLOATS
"ND° E "
f J OTHERWI SE
s
WE AND POSITION
TAPES
OF INVESTIGATOR
AGENCY Portland Bureau ot
State Board of Health,
fTealth , Oregon
Bacteriology
-, M^rl-I^T ?^K^n1
This is a summary report of the outbreak previously described in a preliminary report
dated 1/31/55. While evidence is inconclusive, this is believed to be a contact spread
outbreak rather than a common source outbreak. There is now considerable evidence that
the syndrome has been noted in other parts of the state and probably out-of-state. While
an unidentified virus seems the most likely etiologic agent, the unusual prevalence of
Giard^a lamblia cysts in stools of patients seems worthy of record.
See attached data.
S.B. OSGOOD, M.D., DIRECTOR, EPIDEMIOLOGY SECTION, OREGON STATE BOARD OF HEALTH
2/17/55
-------
182 WATERBORNE GIARDIASIS/EPIDEMIOLOGY
RESULTS OF UNIVERSITY OF OREGON MEDICAL SCHOOL STUDY ON
GASTROENTERITIS OUTBREAK
* A A A X A * * * A * * A A * A A * A A A A A A A A
A. Virus Isolation
No virus isolated from stools of acute cases as of 2/15/55.
B- Enteric Pathogens (Single stool from 100 persons, usual cultural
technics).
No enteric pathogens isolated as of 2/2/55.
C. Intestinal Parasites
I. Specimens obtained from 23 patients with acute gastroenteritis,
68 students in one class of medical students and 9 members of
bacteriology department staff for a lotal of 100 cases.
II. Methods
All stools examined by a concentration method for parasites
and ova; when initial stool was negative, up to 3 specimens
were secured on the same case before any were called negative
for "i ;r :i :.
Histories were obtained from the "survey" group of 77 cases
to determine whether or not there was any history of gastro-
intestinal symptoms during the prior 3 months.
III. Results
A. No ''. hiat.i'7yti •; and very few intestinal par.ssites
other than ""^n^'l^c:. were found.
B. '"',',i:'.'j,i i were found in 15 of 23 (65%) of the "Patient"
group and in 25 of 77 (32.5%) of the "survey" group
for an overall percentage of 40% positive '"i-i,a^"li:. This
contrasts with less than 3% positive for f'jinc 4,~' \ in a
similar but smaller group of medical students examined
by the same technic about 1 year ago.
C. Of the 25 in the survey group positive for 'liif'J-it 16
(64%) had gastrointestinal symptoms (10 prolonged or
recurrent and 2 relieved by atabrine) whereas 9 (36%)
denied any symptoms and in 4 of these f;i.ardisi were
numerous.
D. tiiardia lomblia (chiefly c^ystsj were found in 31 (58%)
of the group that had symptoms and in 9 (20%) of the
group that had no symptoms.
/dl Epid Section-OSBH
2/17/55
-------
L. Veazie
183
W
O
fn U
O M .
H 1-.
• < Q
-------
184 WATERBORNE GIARDIASIS/EPIDEMIOLOGY
RESULTS OF HETEROPHILE ANTIBODY TESTS AND SEROLOGIC TESTS
FOR SYPHILIS ON KNOWN GASTROENTERITIS CASES
* * * * * * * * * *******
A. HETEROPHILE ANTIBODY TESTS
1 case, Giafdia present, antibody titer 1/80 two weeks after
onset and 1/40 three weeks after onset.
B. SEROLOGIC TESTS FOR SYPHILIS
17 cases, r,i..irdia known to be present in stools of 11, STS
negative on all 17 cases - blood taken 7 to 35 days after onset.
•(a few doctors had reported doubtful serologic tests for syphilis
and slight reactions to the heterophile antibody test on 1 or 2
cases of the current gastroenteritis syndrome)
/dl 2/17/55
-------
/,. Veazie 185
DIARRHEAL ILLNESS
Total Tabulation Including Cases From Private Doctors, OSBH Cases, Plus
State Office Building Cases
I. Multiple vs Single Case Households
A. Total Household - 253
B. Single Case Household - 127 (50%)
C. Multiple Case Household - 126 (50%)
II. Duration of Illness
A. "rivate M.D. Cases (142 cases)
Range - 1 - 120 days
Average - 18 days
B. OSBH (42 cases)
Range - 30 days
Average - 8 days
C. State Office Building (317 cases)
Range - 1 - 90 days
Average - 14.3 days
D. Total (Ii01 cases)
Range - 1 - 120 days
\verage - 14.8 days
F,. Chronic, intermittent type illness
1. Private M.D. - 52 of 142 cases (36.6%)
2. OSBH - 9 of 42 cases (21.4%)
3. State Office Building - 105 of 317 cases (33.1%)
III. Sex Distinction (537 cases)
A. Male - 217 (40.5%)
B. Female - 320 (59.5%)
IV. Age Distribution (461 cases) (Age distribution of Portland
population, 1950 census for
A. Cases comparison)
Less than 5 - 27 ( 5.86%) Under 5
5-9 - 28 ( 6.1%) 5-9
10-14 - 27 ( 5.86%) 10 - 14
15-19 - 23 ( 5.0%) 15 - 19
20 and over - 356 (77.0%) 20 and over - 73.4%
B. Exposures (740 total)
Less than 5-85 (11.5%)
5-9 - 72 ( 9.7%)
10-14 - 41 ( 5.5%)
15-19 - 44 ( 5.9%)
20 and over - 498 ( 67%)
-------
186 WATERBORNE GIARDIASIS/EPIDEMIOEOGY
- 2 -
C. Over-all attack rate in exposed household
1. Total exposed - 740
2. Total ill - 461
3. Rate - 62%
D. Age-specific attack rate
Less than 5 - 31.7%
5-9 - 39%
10-14 - 66%
15-19 - 52%
20 and over - 71.5%
V. Symptoms (358 cases)
Loss appetite - 246 ( 69%)
Nausea - 238 ( 66%)
Vomiting - 129 (36.0%)
Abdominal discomfort - 291 ( 81%)
Diarrhea - 275 ( 77%)
Muscle Pain - 101 (28.3%)
Fever - 92 (25.7%)
Headache - 152 (42.5%)
Cough - 62 (17.3%)
Sore Throat - 52 (14.5%)
Runny Nose - 51 (14.6%)
VI. Duration of Illness vs Therapy
I. Atabrine therapy (71 cases)
A. Average duration before treatment - 11.1 days
B. Average duration after treatment - 7.8 days
C. Average duration total illness - 18.9 days
II. Treated by means other than atabrine (103 cases)
A. Average duration before treatment - 5.8 days
B. Average duration after treatment - 9.7 days
C. Average duration total illness - 15.5 days
-------
L. Veazie 187
REPORT ON DIARRHEA! DISEASE IN STATE OFFICE EMPLOYEES
I. Attack rate among employees
A. 561 persons reported on
B. 173 persons ill
C. Attack rate - 30.8%
II. Duration of illness (known on 317 cases)
A. Average duration - 14.3 days
B. Range - 1 to 90 days
C. Chronic relapsing type of illness - 105 cases (33.3%)
III. Multiple vs Single Household cases (182 households)
A. Single Case Household - 95 (52%)
B. Multiple Case Household - 87 (48%)
2 cases - 53
3 cases - 18
4 cases - 14
5 cases - 1
IV. Sex Distribution (323 cases)
A. Male - 132 (41%)
B. Female - 191 (59%)
V. Age Distribution
A. Cases - 319
Less than 5 - 11 ( 3.45%)
5-9 - 13 ( 4.1%)
10-14 - 23 ( 7.2%)
15-19 - 20 ( 6.3%)
20 and over - 252 (79.0%)
B. Exposures - 485
Less than 5-47 (9.7%)
5-9 - 33 (6.8%)
10-14 - 27 (5.6%)
15-19 - 37 (7.6%)
20 and over - 341 ( 70%)
C. Over-all attack rate in exposed household
1. Total exposed - 485
2. Ill - 319
3. Rate - 66%
-------
188 WATERBORNE GIARDIASIS/EPIDEMIOLOCY
- 2 —
D. Age-specific attack rates
Less than 5 - 23.4%
5-9 - 39.4%
10-14 - 85%
15-19 - 54%
20 and over - 74%
VI. Treatment (known on 184 cases)
A. Non-specific (none, Pepto Bismol, etc.) - 108 (58.6%)
B. Bismuth and Paregoric 10
C. Kaopectate 22 37 (20%)
D. Antispasmodics 5
E. "Specific" therapy 40 (21.7%)
Aureomycin - 3
Creomycin - 2
Penicillin - 7
Terramycin - 2
Streptomycin - 1
Ilotycin - 1
Dramamine - 1
Sulfa - 1
Antihistamine - 3
Atabrine - 6
Unknown prescription - 13
VII. Symptoms (174 cases)
Loss of appetite - 119 ( 68%)
Nausea - 122 ( 70%)
Vomiting - 58 ( 33%)
Abdominal Discomfort - 154 (88.5%)
Diarrhea - 131 ( 75%)
Muscle Pain - 67 (38.5%)
Fever - 63 ( 36%)
Headache - 112 (64.4%)
Cough - 35 ( 20%)
Sore Throat - 45 ( 26%)
Runny Nose - 43 (24.7%)
-------
L. Veazie 189
February 16, 1955
ABSENTEEISM DATA IN RELATION TO PORTLAND GASTROENTERITIS OUTBREAK
1953 November
1954 November
1953 Dec ember
1954 December
1954 lanuarv
1955 lanuarv
1955 rtDTu.ir\
(1st 2 weeks)
1953 Ckt. 192 New 20'. Dec 227 Jan. 194
1954 Oct 200 Nov 221 Dec 266 Jan. 277
1955 Ian 318
(December 1954, 119 specified as "flu" and 3 «is "enteritis", January 1955, 139
"flu" and 7 as "enteritis)
(919 em-
ployed) 1953 Oct. 1944 Nov. 3841 Dec. 3748 Ian. 2488
(818 cm-
ployed) 1954 Oct 2585 u>v 2972 Dec 3504 Ian. 3064
1955 Ian 2976
V. National Biscuit Compaivv--Absenteeism, all causes (Year of 1954 - 2.6450
1954 January 2 617 December 1.967
1955 lanuary 1st week 2 467. 2nd week 2.457, 3rd week 2.65%, 4th week 2.76'!
/dl Epid Section-ObBH
2/17/55
-------
190
WATERBORNE GIARDIASIS/EPIDEMIOLOGY
"I
I*
: C
I*
I
u
Sfl
L
E
S3SVJ JO ON
-------
L. Veazie
191
Diarrheal Illness
(519 Cases by week of onset)
I OSBH CASES
D PRIVATE MD CASES
S STATE OFFICE
BLDG CASES
34 11 18 2510 9 16 23301 Ifi1 3 202712 1) 18
WEEK OF ONSET
-------
192
SESSION IV - DETECTION METHODOLOGY
Chairman - John C. Hoff,
U.S. EPA, MERL, Cincinnati, Ohio
Methods for Detection of Giardia
Cysts in Water Supplies
W. Jakubowski, T. H. Ericksen
The Propagation of Giardia
Trophozoites In Vitro
E. A. Meyer
Induction of Giardia Excystation and the
Effect of Temperature on Cyst Viability
as Compared by Eosin-Exclusion and In
Vitro Excystation
A. K. Bingham, E. L. Jarroll, Jr., E. A. Meyer, S. Radulescu
-------
193
Methods for Detection of Giardia Cysts
in Water Supplies
Walter Jakubowski and T. H. Ericksen
U. S. Environmental Protection Agency
Health Effects Research Labor at or v
Field Studies Division
Epidemiology Branch
Bacterial and Parasitic Diseases Group
Cincinnati, Ohio
ABSTRACT
The development and application of methods for detecting (jiardia cysts in water are
reviewed. The concentration of cysts in raw sewage was calculated to be between 9 6x 101
and 2.4x lO5 cysts, 1 when l-25%of the population is infected Evaluation and application
of a large volume sample collection and processing method are described. A step-by-step
analysis of the procedure indicates 42% of cysts lost in preparing a homogenate of the
primary Orion filter, 58 to 71% loss of cysts when using a 7 fj.m mesh nylon screening
process, and more than 90% loss if flocculation is used to clarify the homogenate
Suggested modifications for producing a fourfold increase in recovery efficiency are
described and evaluated. Sampling considerations, including sample volume, frequency,
flow rate, storage, shipment, processing, and interpretation of results are discussed. A
sample volume of 380 1 (100 gal) is suggested for occurrence surveys in water supplies when
concurrent epidemiologic and or watershed animal surveys are performed,
The detection, identification, and enumeration of microbiological
contaminants, and in the present instance specifically Giardia cysts in water
supplies, is of concern to workers in a number of different disciplines. The
microbiologist and the epidemiologist want to know the source of
contamination, the level of contamination, the relationship to indicator
organisms, and the reservoirs of infection. The engineer and the utility
operator want to know how to remove or inactivate Giardia by water
treatment. Those responsible for administering drinking water programs
want to know where and when the organism occurs in water, the suitability
and availability of monitoring methods, and whether or not maximum
contaminant levels or treatment requirements should be promulgated.
These data are required on this pathogen because recent outbreaks, in
particular those inCamas, Washington(l,2)and Berlin, New Hampshire(3),
have demonstrated that the coliform standard cannot be relied upon to
indicate the safety of a water supply with respect to Giardia. The possibility
that other organisms, such as yeasts, could act as alternate indicators of
Giardia contamination remains to be investigated. The application of
cultural techniques to provide answers to the questions being posed is not
now possible for Giardia in drinking water. In addition, certain basic
-------
194 WATERBORNE GIARDIASIS/DETECTION METHODOLOGY
information (e.g., how to determine species of origin, viability, or infectivity)
necessary for the interpretation of results obtained through microscopic
examination is lacking. The purpose of this paper is to describe the state-of-
the-art with regard to sampling methodology for the detection,
identification, and enumeration of Giardia cysts in water supplies.
DETECTION METHODS
At the present time, there is no cultural method for detecting Giardia cysts
in water samples. However, Meyer(4,5) and others have successfully
cultivated the trophozoite stage, and the recent development of an
excystation technique(6) may subsequently lead the way to development of
cultural methods. The available alternatives are animal feeding and direct
microscopic examination. Animal feeding experiments require special
handling and isolation facilities, and are time-consuming and expensive.
Their application to surveys and monitoring would, therefore, be limited by
economic and practical considerations. Animal feeding, however, is the only
method for demonstrating the infectivity of Giardia cysts.
Microscopic methods of detection are not quite as time-consuming as
animal feeding but they are tedious. In addition, identification of the
organism is dependent upon the training, experience, skill, and patience of
the examiner. The results of a proficiency-testing program reported(7) by the
Center for Disease Control (CDC) indicated that between 20% and 35% of
public and private laboratories could not correctly identify intestinal
protozoa. Furthermore, the identification of protozoan cysts is not a skill
that would be available in most water supply microbiological laboratories.
In performing microscopic analysis it is also important to have a clean
preparation as free as possible of inorganic and organic debris that could
mask the presence of the organisms or interfere with identification.
Nevertheless, recognizing these limitations, microscopic methods are the
only currently feasible approach for detection of Giardia in water samples.
THEORETICAL LEVELS OF CYSTS
IN SEWAGE AND WATER
In developing methods, it would be advantageous to have some
information on cyst levels that might be expected in sewage and water
samples. Giardia is an intestinal parasite not likely to multiply in a water
environment outside of a host. Tsuchiya(S) studied cysts produced by two
human carriers, A and B, for 36 and 28 days, respectively. An analysis of
these data indicated a mean of 1.1 x 106 cysts/g of stool for subject A with a
maximum observed level of 3.3 x 106cysts/g. Mean daily production of cysts
for subject A was 2.1 x 108 with a maximum of 7.1 x 108. Results for subject B
were similar (mean of 1.6 x 106/g with a maximum of 4.3 x 106, and daily
production of 2.0 x 108 with a maximum of 6.4 x 108). Tsuchiya did not
report the subjects' ages but it is assumed that they were adults.
Danciger and Lopez(9) studied cyst production in 15 children 3 to 7 years
of age. Three patterns of excretion were noted: high, with large numbers of
cysts in nearly all stools; low, with small numbers of cysts in 40% of the
-------
W. Jakubowski 195
stools; and mixed, where periods of high excretion alternated with shorter
periods of low excretion and an average of about 60% positive stools. The
number of cysts excreted ranged up to 2.2 x 106/g of formed stool with an
overall mean of 5.8 x 105/g. Total daily production of cysts on a per capita
basis was not reported.
The per capita daily usage of water from public supplies in the U. S. in
1975 was 638 1 (168 gal)(10). Water lost in the distribution systems and used
for public activities such as firefighting, street cleaning, etc. was estimated at
36% or 230 1 (60 gal). Industrial and commercial use was 31% or 198 1 (52
gal). This results in an estimated residential usage of 210 1 (55
gal)/person/day. Assuming that the average daily water usage for a family
of four is 840 1 (220 gal)/day, and that 2.0 x 108 cysts are shed daily by an
infected adult, the theoretical concentration of cysts in raw sewage would be
2.4 x 105 cysts/1 if one of the individuals (25%) was shedding cysts. If l%of
the population were infected, the concentration of cysts in raw sewage would
be 9.6 x 103 cysts/1. These calculations are based upon a uniform dispersion
of fecal material in sewage originating entirely from domestic sources.
Calculation of cyst levels in water supplies would then be dependent upon
the degree of raw domestic sewage contamination of a given water supply.
Various surveys have indicated the prevalence of Giardia-infected
individuals to be in the range of 1 to 24%(11).
The Camas, Washington outbreak(l) and reports of infections in campers
and hikers(12) suggest that water supplies may become contaminated with
Giardia by sources other than human fecal discharge. Beavers were
implicated in the Camas outbreak and the extent to which wild and domestic
animals may act as reservoirs of infection is currently being determined( 13).
Sparkling mountain streams in pristine locales may be subject to
contamination by wild animals. The role of man in disseminating infection
in wild animals needs to be determined by comparing infection rates in
animals in those areas open and closed to human use. In a survey conducted
in Washington State over the last 2 years (14), 7% of beaver and 33% of
muskrat examined were (j/a/W/'a-positive. It was noted that many of the
muskrats were trapped close to human dwellings and in areas which may
have received human sewage contamination.
PREVIOUS ATTEMPTS AT DETECTING CYSTS IN WATER
The basic assumption in application of microscopic methods to the
detection of Giardia cysts in water samples has been that some form of
sample concentration is necessary. In 1956, Chang and Kabler( 15) described
a membrane filter technique for concentrating Entamoeba histolytica cysts
in tap water. Five replicate experiments were conducted at four cyst
concentrations: 1, 3, 5, and 10 cysts/3.8 liters. Each sample was membrane
filtered and the- retained material was washed from the membrane with
about 2 ml of distilled water. This concentrate was examined
microscopically in a counting chamber. Mean cyst recoveries of 20%, 28%,
and 42% were reported at cyst concentrations of 3, 5, and 10/3.8 liters,
respectively. No cysts were recovered from the sample containing one
cyst/3.8 liters.
-------
196 WATERBORNE GIARDIASIS/DETECTION METHODOLOGY
During the investigation of the 1965-66 Aspen, Colorado giardiasis
outbreak, Moore et al(16) adapted the Chang and Kabler method'to the
examination of water and sewage samples for Giardia cysts. Two-liter water
samples from 10 sites throughout the city, and one-liter sewage samples from
three main trunk lines were collected. The samples were passed through
cheesecloth and filtered through 0.45 ^m porosity membrane filters. The
sediment on the filters was brushed into water, centrifuged, and preserved in
10% formalin for microscopic examination. No cysts were detected in the 10
water samples. Two sewage samples from lines draining the eastern section
of the city contained three Giardia cysts and one or 2 cysts of Entamoeba coli
and Endolimax nana/100 ml.
In 1972, R. A. Wright (personal communication) added unknown
quantities of Giardia cysts to one 1 of physiologic saline, filtered the samples
through 8 Mm porosity membrane filters by suction, and brushed and
washed the sediment into 50 ml of saline. In 5 trials, a few cysts were
invariably identified but quantitation was not attempted. Subsequently, in
the investigation of an outbreak, Wright used this technique on two 20 gal
stream water samples and one 5 gal sewage sample. The sewage sample was
prefiltered before going through the 8jum membrane. Giardia cysts were not
identified in these samples.
In 1973, Barbour et al(12), while investigating an outbreak in a group of
campers, filtered 22 1 of stream water through a membrane filter. Giardia
cysts were not found but coccidian oocysts were observed in small numbers.
With the exception of the Aspen, Colorado sewage samples, the membrane
filter method has not been successful in demonstrating cysts in water
samples. Whether this is a function of the inability to process a sufficient
volume of water, or of the failure to adequately "clean-up" the retained
material, or perhaps an indication that the organism was not present in these
water samples at the time of sampling, is not known.
By the time of the Rome, New York outbreak in 1974-75, and apparently
operating under the premise that larger volumes of water must be sampled to
detect cysts, the CDC had developed a large-volume sampling
technique(17). The method used a swimming pool filter in which sand was
the filtering medium. Ten samples of raw Rome, New York water,
representing a total volume of about 1.1 x lO6 liters (280,000 gal.) were
collected by pumping water through the filter at an average flow rate of 76
liters/ min (20 gpm) daily for 10 days. Filter backwash was collected each day
in two 210 liter (55 gal) drums and coagulated with alum. After a 24-hour
settling period, the supernatant fluid was removed and 1 to 2 liters of
sediment were collected. Aliquots of each sample were fed to beagle puppies,
examined microscopically at Rome, and sent to CDC for further processing.
The sediment samples examined at CDC were washed through two layers
of cheesecloth and an 80 jum wire screen by gravity, and then through a 50
fj.m porosity membrane filter by vacuum. Most of the supernatant was
removed following centrifugation at 1,000 rpm for 10 min, and selected
sediment samples were examined microscopically. Two of the ten samples
fed to dogs produced infection 20 and 34 days after feeding, and a cyst was
-------
W.Jakubowski 197
observed microscopically at CDC in one of these samples. This was the first
time that a cyst had been detected in a water supply, and the first time that
concentrates from a water supply had been found infective for laboratory
animals. The efficiency of this method has not been reported by CDC, but in
one experiment where cysts were added to water at a level of 20/3.8 liters (1
gal) and 34,200 liters (9,000 gal) were sampled, no cysts were detected by
microscopic examination (B. Wood, personal communication).
The CDC used this method in Camas without success but cysts were
detected by the U.S. Environmental Protection Agency (EPA) in an 855 liter
(225 gal) sample of distribution system water collected with the EPA method
(18,19) described in a later section, and also in samples of raw surface water
from the Camas supply. In simultaneous sampling by EPA and CDC in
Rocky Mountain National Park in 1976(19, 20), cysts were detected in
beaver pond water but only by EPA examination of the CDC concentrates.
This may have reflected the difference in location of intakes for the
respective samplers (the CDC intake was located on the pond bottom; the
EPA intake was suspended about 0.5 m above the pond bottom), or the fact
that the volume of water sampled by CDC was approximately 10 times
greater than that sampled by EPA.
Another method was developed by alert laboratory personnel at the
Androscoggin Valley Hospital in the Berlin, New Hampshire outbreak of
1977(3). The sampling device was a funnel consisting of a plastic bottle with
the bottom cut out. A filter pad made of layers of gauze was placed in the
neck of the bottle and laboratory tap water was filtered. Cysts were identified
in the material trapped by the gauze but the volume of water sampled was
not determined. A portion of the material retained by the filter was
examined by EPA and large numbers of yeast cells were also noted.
THE EPA METHOD
Development
In 1976, also operating under the assumption that some form of sample
concentration was necessary, EPA developed a method for detecting
Giardia cysts in water samples. In developing a method, one would logically
proceed by gathering information about the physical and chemical
characteristics of the organism and pertinent facts relating to its natural
history. An appropriate method would then be devised, evaluated in the
laboratory, and applied in the field. In this case, pre-existing field equipment
for large volume virus sampling was adapted for detecting G7ard/'acysts(18).
In February, 1976 the equipment was used for the first time in the field in
Montgomery Center, Vermont to collect samples for virus analysis during
an outbreak of gastroenteritis. One component of the virus apparatus is a 4
cm (10 in)-long, honeycomb yarn-wound 10 jum-porosity Orion filter used to
prevent clogging of the epoxy-fiberglass virus-adsorbing filter when
sampling turbid water. Since it was known that suspected waterborne
outbreaks of giardiasis had occurred previously in Vermont(21), it was
decided to process the Orion prefilter from a 1060 1 (279 gal) raw stream
water sample collected in Montgomery Center. The Orion filter was
-------
198 WATERBORNE GIARDIASIS/DETECTION METHODOLOGY
backwashed with about one 1 of distilled water. The backwash suspension
was passed through a series of nylon screens (30 to 50 pm mesh) to remove
larger particulates, and the material retained by a 7/xm mesh nylon screen
was examined microscopically. One Giardia cyst was detected in this
concentrate.
In May, 1976, the Camas, Washington giardiasis outbreak occurred.
Using the EPA method, cysts were detected in treated distribution system
water, in chlorinated influent at the filtration plant, and in raw water from 1
of the town's 2 surface supplies (1,18). In a preliminary survey of 6 sites in
Washington State in June, 1976, a Giardia cyst was detected in treated water
of the Hoquiam water supply, and a distorted cyst was detected in the raw
water of the Everett supply.
By this time, the sampling device, weighing less than 3 kg (about 6 Ib),
consisted of a garden hose inlet, the Orion filter holder, a plastic water meter,
and a 1.9 liter/min (0.5 gpm) limiting orifice at the end of the effluent hose
(Fig 1). The backwashing process was replaced with a procedure that
involved cutting the Orion fibers down to a depth of about 0.5 cm and
blending the fibers at low speed in a homogenizer for 10 seconds with 250 ml
of distilled water to separate the retained particulates from the filter
medium. The blending procedure was adopted because, in the preliminary
survey in Washington State, cysts were not detected in backwash material
but were observed in the filter homogenates. It was also subsequently
determined that the Orion filter medium was superior to cellulose acetate or
FIG. 1. USEPA Giardia sampling device. A—inlet hose; B—-filter housing; C—outlet hose;
D—water mete; E—limiting orifice flow controller.
-------
W. Jakubowski 199
polypropylene yarn-wound filters, and to epoxy-fiberglass filter tubes. The
cellulose acetate and polypropylene fibers had a tendency to break upon
homogenization thus contributing fibers to the concentrate that interfered
with cyst detection. The epoxy-fiberglass filters also exhibited this
disadvantage, and in addition, would clog during filtration at much lower
volumes than would the yarn-wound filters. Modifications subsequently
incorporated into the method included use of 7 p.m nominal porosity Orion
filters instead of 10 ;um, removal and blending of all fibers on the stainless
steel filter core instead of just the upper 0.5 cm layer of fibers, and use of acid,
alkaline, or acid-formalin flocculation for clarification of homogenates
containing high levels of interfering particulates.
For 1977 and most of 1978, the procedure followed in examining water
samples has been essentially that shown in Fig 2. Using this procedure, cysts
were detected in 1977 in spring water in Granby, Colorado, in well water and
surface water in Eagle County, Colorado, and in well water in Hot Springs,
Colorado. Cysts were detected in the raw surface water of 2 supplies and in
the distribution system water serving Berlin, New Hampshire, in additional
raw stream water samples in Hoquiam, Washington, and in lake water in
Conneticut. In 1978, a positive sample was obtained from stream water in
Vail, Colorado.
Blend Orion Filter with
1000 ml distilled water or saline
_L
I 1
Sediment <5% of total volume Sediment >5% of total volume
acidify with 2 ml cone. HC1
No Floe
Floe-* Add Formalin
| to 2-5%
gauze cotton
45 ym nylon screen
^-30 iim nylon screen
V
7 pm nylon screen »-wush with 1-2 ml
distilled water
examine
microscopically
FIG. 2. Processing scheme for GUrdia water filters.
Evaluation
The EPA method had been successfully used to detect cysts in water in
several outbreaks and surveys. However, a number of questions soon arose
concerning the methodology: What was the efficiency of the method? What
sample volume should be examined? How frequently should samples be
-------
200 VVATKRBORNK GIARDIASIS/DKTKC TION MKTHODOLOGY
collected? What flow rate should be used? H-ow should samples be stored and
shipped? How soon after collection must they be examined? The first
problem requiring resolution was: What method should be used to count the
cysts in experiments to determine efficiency of the concentrating technique?
S. L. Chang (personal communication) suggested a method in which 2
drops of cyst suspension were dispensed from a calibrated Pasteur pipet onto
a slide, a coverslip was applied, and cyst counts were done of 20 random
fields under low power (ca 300x). The cyst concentration was calculated by
multiplying the total observed count by the number of fields available in the
coverslip when using a Pasteur pipet calibrated to produce 40 drops/ ml. In
Dr. Chang's experience, this method was easier to use and produced more
consistent results than counts obtained with a cell counting chamber. A
preliminary evaluation of the coverslip and hemocytometer methods is
shown in Table 1. The hemocytometer counts were obtained with an
American Optical Improved Neubauer Hemocytometer. Two drops of cyst
suspension were placed in the middle of the chamber, a coverslip was set in
place, 20 squares were counted, and the total count was multiplied by an
appropriate factor (2 x 105) to give the number of cysts/ml.
Fifteen replicate samples of a cyst suspension were counted with both
methods. The mean counts obtained with the hemocytometer were more
Table I Comparison of Giardia < rsr counts b\ the covt'rslip and hemoi \lometer methods
Subsample
1
2
3
4
5
6
7
8
9
10
1 1
12
13
14
15
x-
S E Ma
Cyst Count
Coversl ip
3 12
3 34
2 79
2 33
3 22
1 85
2 42
2 68
3 01
2 03
2 22
1 96
1 71
1 44
2 15
2 42
0 15
1x10'')
Cell C ham ber
6 4
22 4
44 8
25 6
38 4
8 0
25 6
14 4
12 8
38 4
22 4
20 8
19 2
22 4
20 8
22 8
2 8
"Standard error of the mean
-------
W. Jakubowski 201
than 9 times higher than those observed with the coverslip method and no
satisfactory explanation could be found. However, the standard error of the
mean for the hemocytometer method was significantly greater than that of
the coverslip method even after adjusting for the higher counts observed
with the hemocytometer. The variability exhibited in counts made with the
hemocytometer would make interpretation of data difficult.
Modifications were made in both counting methods and another
evaluation was performed. For the coverslip method, 5 p.\ of cyst suspension
were placed on a slide, covered with a 22 x 22 mm coverslip, and sealed with a
1:1 parafin-vaseline mixture. The cyst density was calculated as follows:
cyst no. x 1,000^1 _ cyst no. x 200 _ cysts/
-------
202 WATERBORNE GIARDIASIS/DETECTION METHODOLOGY
The same cyst suspension was diluted 1:400 to obtain a sufficient volume
to analyze 30 replicate samples of 100 ml each on a HI AC (Pacific Scientific
Co., Montclair, California 91763) particle counter. The mean number of
particles counted on channels 4 thru 7 of the HI AC (representing particles 7-
20 fj.m in size) was 91,271 or 1.2 and 2.2 times higher than the counts
obtained with the hemocytometer and coverslip methods, respectively. Since
the HI AC cannot discriminate between cysts and debris particles of similar
cross-sectional surface area, it was not felt that this method would offer a
suitable alternative to microscopic counting methods.
Based upon the above analysis and the fact that the hemocytometer
requires only a small fraction of the time required with the coverslip method,
the hemocytometer was selected as the method of choice for enumeration of
cysts.
Late in 1977 we reported(19) on a series of 11 trials where about 1 * 106
cysts were added to 380 liter (100 gal) volumes of Cincinnati drinking water
which were processed for cyst recovery by the EPA method (Table 3). The
mean recovery of cysts in this series was 6.3% with a range of 3 to 15%. We
have since analyzed preliminary data for each step in the sample collection
and processing method in order to determine where improvements may be
made.
The first point at which cysts may be lost is during the collection of the
sample with the 7 pirn porosity Orion filter. Experiments have not been
performed using this filter to determine what fraction of Giardia cysts might
pass through it. However, data supplied by the manufacturer(22) indicate
that the honeycomb filters remove a minimum 95% of all particles at the
rated porosity when processing non-aqueous fluids with a viscosity 3.5 times
Table 3. Recovery of added Giardia
-------
W. Jakubowski 203
that of water at a flow rate of 11.4 liters/ min(3 gpm). Giardia cysts are
generally larger than 7 /zm, and since efficiency of filtration would increase in
aqueous fluids of lower viscosity and as particulates accumulated on the
filter, it was assumed that retention of cysts by this filter would exceed 95%.
The next step in the process is blending the filter fibers with a diluent to
recover retained particulates. As mentioned earlier, backwashing the filters
had been found less successful than the blending process in finding naturally-
occurring cysts in field samples. Distilled water, saline, and 1% beef extract
diluents have been examined for effects on recovery by filtering cysts
suspended in these media through 7 /urn mesh nylon screens. No apparent
advantage in terms of recovery was exhibited by any of these media
(unpublished data). In the absence of data on the effects of medium tonicity
on Giardia physiology, and since the concern was with detecting the
organisms as they occur in a hypotonic medium (drinking water), distilled
water was selected as the diluent.
Recovery of cysts from the Orion filter was determined as follows: 3.8 1 of
drinking water containing about 1 x 106 cysts (additive suspension) were
continuously added to 380 1 of drinking water by means of a proportioner
pump while filtration progressed through the 7 pm porosity Orion filter. A
mixing chamber was present in the system between the point of cyst addition
and the Orion filter holder. The number of cysts in the 3.8 1 of additive
suspension was determined by filtering another 3.8 1 portion of additive
suspension through a 7 p.m mesh nylon screen, collecting the cysts in 1 ml of
distilled water and counting with the coverslip method. This count
represented the challenge dose to the Orion filter. The number of cysts
recovered from the Orion filter was determined by passing the filter
homogenate through a 7 ^irn mesh nylon screen and performing counts in a
manner similar to that for the additive counts. The percent recovery was
determined by dividing the homogenate count by the additive suspension
count and multiplying by 100. The results are shown in Table 4.
A mean recovery of 58% of cysts was obtained from the Orion filter with a
range of recovery of 12 to 139%. Assuming that only an insignificant
percentage of cysts would pass through the filter (based on manufacturer's
data) the relatively low and highly variable recovery may be due to
difficulties in preparation of the homogenate. The homogenate is obtained
by blending 1 / 4 of the Orion filter at a time with 250 ml of distilled water at
low speed (17,000 rpm) for 10 seconds in a 473 ml (1 pint) blender jar. It is
extremely difficult to perform this procedure in exactly the same fashion
from sample to sample. The blender blades will often become entangled in
the Orion fibers thus reducing the blending speed and effectiveness of the
process. Cysts may also be lost while separating the particle suspension from
the Orion. In this procedure, the aqueous portion is separated from the fiber
mat by decantation after blending. The fiber mat is then placed into a plastic
bag with one corner removed and the remaining fluid is expressed from the
fibers. Cysts would undoubtedly adhere to or become entrapped in the filter
fibers at this point.
-------
204 WATERBORNE GIARDIASIS/DETECTION METHODOLOGY
Table 4 C\ s; recovery from Orion filler homogenate
Trial No
1
2
3
4
5
6
7
8
9
10
1 1
% Recovered
12
26
18
41
99
77
28
48
139
79
66
Mean 58
S E Ma 12
"Standard error of the mean
In a filter homogenate not requiring flocculation (one with <5%
sediment), the next step in the concentration process that has been evaluated
is the recovery of cysts on the 7 /urn nylon screen. Three trials were conducted
in which Orion filters were challenged with about 1 * io6 cysts using the
proportioner method described earlier and 380 liter volumes of drinking
water. Aliquots of the filter homogenate were concentrated with a 7 /um
mesh nylon screen and the cysts were collected in 1 ml of distilled water, or by
centrifuging the homogenate at 1,500 rpm for 10 min, suspending the
sediment in 1 ml of the supernatant fluid, and counting the cysts. The results
are presented in Table 5. Mean recovery with the nylon screen was 7.8%, and
with the centrifugation procedure, 32%. These results indicated that about
71% of the cysts in the homogenate were not being recovered with the nylon
screen procedure when compared to recovery by centrifugation.
Recovery of cysts with the nylon screen was investigated further in
another series of experiments in which polycarbonate (PC) membrane filters
and centrifugation were also compared. Replicate 50 ml samples containing
about 1 x IO6 cysts in water were vacuum filtered through 47 mm diameter 7
Mm mesh nylon screens or 8 /^m porosity PC membrane filters (Bio-Rad,
Richmond, California 94804) and retained cysts were collected in 1 ml of
Table 5. Cw recovery from Orion homogenate bv nylon screen vi ceninlugalion
7 Mr
Mean
% Recovery
n Nylon Screen
2 7
14 6
6 1
7 8
Centnlugation
40
24
32
32
% Loss
with Screen
93
40
81
71
-------
W. Jakubowski
205
water. Filtrates from the screens and PC filters were centrifuged at 1,500 rpm
for 15 min, and the sediment was resuspended in 0.5 ml of supernatant fluid.
Replicate 50 ml samples of cyst suspension were centrifuged without
filtration and concentrates were prepared in the same manner as for the
filtrates. All counts were performed with a hemocytometer. Percent recovery
was calculated by dividing the cyst counts obtained in the concentrates by
the cyst count of the original suspension and multiplying by 100. Results are
shown in Table 6.
The mean recoveries for 5 trials were 46%, 64%, and 132% for the screen,
PC filters and centrifugation, respectively. Centrifugation analysis of the
filtrates indicated that 16% of the cysts were passing through the screens and
0.9% through the PC filters. Adding the surface recoveries to those obtained
in the filtrates results in overall recoveries of 62% for the screens and about
65% for the PC filters. The loss of 35 to 38% of the cysts may have been due to
adsorption of cysts to screen and filter material with subsequent failure to
recover these with the small volume (1 ml) of wash water used. The > 100%
recoveries by centrifugation may have been due to counting error from the
heavy cyst suspension in the final concentrate (>1 * 106/0.5 ml).
Table 6 Comparison of nvlon screen, pn/\ carbonate filter,
and tenlrifiigalion /or recovery of Giardia ri'Ws
Trial No
1
2
3
4
5
~X
S E M a
1
2
3
4
5
X"
S E Ma
7 /jin
Nylon Screen
46
44
43
41
55
46
2 5
Screen Filtrate
12
15
13
19
20
16
1 6
% Recovery
8 /im
PC Filter
65
49
75
69
62
64
4 3
PC Filtrate
2
0 7
1 1
0 2
0 5
09
0 3
1 500 rpm, 1 5 mm
Centrifugation
1 18
149
122
133
140
132
5 7
"Standard error of the mean
-------
206 WATERBORNE CIARDIASIS/DETECTION METHODOLOGY
The experiments described in the above sections were performed with
cysts suspended in clear waters relatively free from other particulate matter.
Many waters contain significant quantities of suspended particulates that
are retained by the Orion filter and released into the homogenate thus
complicating detection of cysts. In the processing scheme shown in Fig. 2,
the presence of >5% sediment in the filter homogenate was arbitrarily
selected as the level at which clarification procedures (flocculation, gauze
and nylon screen straining) would be used. Evaluation of recovery efficiency
when these technics are employed has not been performed. Of the
clarification steps listed in Fig. 2, flocculation would be the one most likely
to result in significant cyst loss.
In trial no. 1 listed in Table 3, one aliquot of the homogenate was
processed only with the 7/nm mesh nylon screen. Another aliquot was
subjected to acid flocculation and strained through gauze and 45 and 30 nm
mesh nylon screens before collection on the 7 /um mesh screen. The recovery
of cysts was 0.6% when flocculation was used as compared to 8% with only
the 7 jum mesh screen, representing more than a one logarithm loss of cysts
by flocculation and preliminary straining.
Suggested Modifications To The EPA Method
The 7 /urn porosity Orion filter is probably more than 95% efficient in
retaining Giardia cysts in water samples. The blending technique for cyst
recovery from the filter is inadequate for high-efficiency reproducible cyst
recovery. The availability of a backwashable filter medium that could
process large volumes of water would greatly enhance a detection method
for Giardia. Preliminary contacts with filter manufacturers indicate that a
need for a backwashable filter does exist, at least in the pharmaceutical
industry, and developmental research is underway.
The use of the 7jum mesh nylon screen as the final step in collection of cysts
(Fig 2) is no longer recommended. Results presented in Tables 5 and 6
indicate 58 to 71% loss of cysts with these screens. A portion of the lost cysts
pass through the screens and others become enmeshed in, or adsorbed to, the
screen material. Using 8 /urn porosity PC filters results in an increase of about
50% in recovery over the nylon screen (Table 6), and centrifugation
produced a 4-fold increase in recovery as compared to the nylon screen
(Table 5). Recent experiments in our laboratory (unpublished data)
indicated up to 81% recovery of cysts on 5/um porosity PC filters. Therefore,
it is recommended that the final concentration process substitute either
centrifugation (1,500 rpm, 15 min) or membrane filtration (5 pm or less
porosity) for 7 /nm mesh nylon screen filtration.
The use of flocculation on a given sample must weigh decreased cyst
recovery against a possible increase in masking effects caused by the
presence of particulate matter. This decision should be based upon the
experience of the investigator and the nature and quantity of the particulate
matter. Homogenates from a single sample may be divided and processed by
different techniques to determine which results in optimum recovery. The
use of flotation techniques may be a satisfactory alternative to flocculation
-------
W. Jakubowski 207
as a means for sample purification but these techniques have not been
evaluated in our laboratory. Inorganic particulates, such as clays, should not
present a problem in separation because of their higher density. Flotation
may not prove suitable for cyst purification if organic particulates with
densities similar to Giardia cysts are present.
SAMPLING CONSIDERATIONS
Volume, Frequency, And Flow Rate
Questions of sample volume, frequency, and flow rate are not peculiar to
Giardia detection. These questions exist with regard to other
microbiological and inorganic contaminants of water supplies as well. Their
resolution is usually dependent upon the purpose of the survey, the efficiency
of the method, and economic and practical factors rather than consideration
of statistically sound sampling guidelines. In 1977, asample volume of 2,000
1 (500 gal) was suggested for surveys on the occurrence of Giardia cysts in
water supplies (19). This suggestion was based on the assumption that it
would be desirable to detect Giardia at a level of 1 cyst/20 1 and that the
efficiency of the method was 1%. It is our opinion now that the sample
volume should be reduced to 380 1 (100 gal) if the centrifugation or
membrane filtration process is substituted for the 7 /urn mesh nylon
screening, especially if turbid waters are being sampled (i.e., those resulting
in >5% sediment in the filter homogenate). There is no advantage in
collecting a 2,000 1 sample over a 380 1 sample if sensitivity in the larger
sample volume is decreased one logarithm by the need for flocculation or
other clarification techniques.
Flow rates of 11.4 liters/ min (3 gpm) at line pressures of 15 to 70 psig have
usually been used in Giardia sample collection. No evaluation of flow rate on
recovery of cysts has been performed. In outbreak investigations, attempts
are made to sample historic water within the distribution system. Conditions
at the outbreak site will usually dictate sample volumes collected and flow
rates used. There is currently no basis on which to establish frequency of
sampling recommendations.
STORAGE, SHIPMENT, AND PROCESSING
There are no reliable data on the effects of different storage temperatures
on Giardia cysts. A study performed in our laboratory on cysts stored at -70,
-20, 5, 20, and 37°C indicated no appreciable change in cyst counts over a 2
week period with the exception of those stored at-20°C (unpublished data).
Cyst counts decreased by about 40% after one week storage at -20° C;
possibly due to aggregation of the cysts. No attempt was made to determine
viability or infectivity of stored cysts. It would seem prudent to store and
ship filter samples at refrigerator temperatures (ca 5°C) and to process
samples as soon as possible after collection. Higher temperatures could
result in growth and activity of other organisms that might be detrimental to
cyst recovery.
Cyst concentrations in water samples may be estimated by multiplying the
number of cysts observed in a measured volume of concentrate by the
-------
208 WATKRBORNK GIARDIASIS/DKTKC TION METHODOLOGY
appropriate conversion factor for the counting chamber used and the
reciprocal of the fraction of concentrate examined. In our laboratory, a
sample is considered negative when no cysts are detected in four 5 jul aliquots
of the 1 ml sample concentrate.
SIGNIFICANCE OF RESULTS
Not much can be said concerning the safety of a supply if a cyst is detected
in a water sample. At the present time, there is no easy method for
determining the host species of origin of the cyst. Since Giardia attacks a
wide variety of animal species and little is known of the host range of Giardia
species, the interpretation of a positive finding is open to question. In other
words, a detected cyst may be host specific for a species other than man.
Furthermore, microscopic examination alone yields no information on the
viability or infectivity of detected cysts. Infectivity of cysts can be determined
only by animal feeding experiments. The recent development of an
excystation technique(6) may provide a means for determining viability.
However, the relationship of viability by excystation to infectivity remains
to be determined.
At the present time, it is suggested that the use of the concentration
method for detection of cysts in water samples be limited to outbreak
investigations or to examination of water supplies where concurrent surveys
of watershed animals and/or epidemiologic studies are being conducted.
The efficacy of treatment techniques for removal of cysts can be determined
only if raw and treated water samples are examined. A negative sample does
not necessarily mean that a water supply is safe since this finding may reflect
intermittent contamination of the supply, poor recovery efficiency with a
given water, or insufficient sample volume or sampling frequency.
SUMMARY
The development and application of methods for detection of Giardia
cysts in water samples have been reviewed. Microscopic detection methods
appear to be the only presently available practical approach for conducting
occurrence surveys in water supplies. The level of cysts in raw sewage was
estimated to be between 9.6 x 103 to 2.4 * 105 cysts /liter when 1 to 25% of the
population is infected.
Evaluation and application of a large volume sampling and processing
method developed by EPA has been described. A step-by-step analysis of the
procedure indicates a cyst loss of 42%to the Orion filter in the preparation of
the homogenate, 58 to 71% loss in the 7 /mm mesh nylon screening process,
and more than 90% loss in the flocculation step. Suggested modifications to
the method that may result in a four-fold increase in efficiency of recovery
(substitution of centrifugation or membrane filtration for the 7 pm mesh
nylon screen) were described and evaluated. Other modifications that
remain to be developed and investigated include the use of backwashable
filters and the substitution of flotation techniques for flocculation as a
clarification procedure.
A sample volume of 380 liters (100 gal) is suggested for occurrence surveys
in water supplies with concurrent epidemiologic and/or watershed animal
-------
W. Jakubowski 209
surveys. Interpretation of positive findings is dependent upon development
of methods for differentiating Giardia species derived from a variety of
animal hosts, ascertaining the host range of the organisms, and determining
the infectivity of detected cysts.
ACKNOWLEDGMENT
The advice and assistance of Dr. Shih-Lu Chang in developing the method
and examining numerous field samples are gratefully acknowledged. The
evaluation was made possible through the excellent technical assistance
provided by Dr. Frank W. Schaefer, Ms. Helen Wood, Ms Laura
Arozarena and Mr. Eugene Rice in preparation and purification of cyst
suspensions and in performing cyst counts.
REFERENCES
1. Kirner, J. C.,J. D. Littler, and L. A. Angelo. 1978 A waterborne outbreak of giardiasis in
Camas, Washington Jour AWWA. 70.35-40
2. Morbidity and Mortality Weekly Report. 1977 Waterborne giardiasis outbreaks -
Washington, New Hampshire Center for Disease Control, Atlanta, Georgia, 26.169.
3 Lippy, E.C. 1978 Tracing a giardiasis outbreak at Berlin, New Hampshire. Jour AWWA,
70512-520
4 Meyer, E. A. 1976 Giardia lamhlia Isolation and axenic cultivation. Exp. Parasitol ,
30-101-105
5. Meyer, E. A. 1979. The propagation of Giardia trophozoites in vitro. 1979 In W
Jakubowski and J C Hotf (ed.), Proc. National Symposium on WaterborneTransmission
of Giardiasis. U S. Environmental Protection Agency, Cincinnati, Ohio, Sept ember 18-20.
1978.
6. Bingham, A., E. L. Jarroll, Jr., E. A. Meyer, and S. Radulescu. 1979 Induction of Giardia
excystation and the effect of temperature on cyst viability as compared by eosm-exclusion
and in vitro excystation. In W. Jakubowski and J C Hoff (ed.), Proc. National
Symposium on Waterborne Transmission of Giardiasis, U.S Environmental Protection
Agency, Cincinnati, Ohio, September 18-20, 1978
7. Center for Disease Control. 1977 Intestinal parasite surveillance annual summary 1976,
Atlanta, Georgia
8. Tsuchiya, H. 1931 A study on variabilities in dimensions and numbers of discharged cysts
on Giardia lamhlia (Stiles, 1915) from day to day under normal conditions Am Jour.
Hyg., 13 544-567.
9. Danciger, M. and M. Lopez. 1975 Numbers of Giardia in the feces of infected children.
Am. Jour Trop Med. & Hyg 24:237-242
10. Murray, C. R. and E. B. Reeves. 1977 Estimated use of water in the United States in 1975
Geological Survey Circular 765. USGS, Department of Interior, Arlington, Virginia
11 Healy, G. R. 1979. The presence and absence of Giardia lamhlia in studies on parasite
prevalence in the U S A InW Jakubowski and J.C Hoff (ed.) Proc. National Symposium
on Waterborne Transmission of Giardiasis, U.S Environmental Protection Agency,
Cincinnati, Ohio, September 18-20, 1978
12 Barbour, A. G., C. R. Nichols, and T. Fukushima. 1976 An outbreak of giardiasis in a
group of campers Am. Jour. Trop Med. & Hyg., 25.384-389.
13 Davies, R. B. and C. P. Hibler. 1979 Animal reservoirs and cross-species transmission of
Giardia.lnW Jakubowski and J C Hoff (ed.) Proc National Symposium on Waterborne
Transmission of Giardiasis, U S Environmental Protection Agency, Cincinnati, Ohio,
September 18-20, 1978
14. Frost, F., B. Plan, D. W. Liechty, Y. Fichtenau, and J. Jernigan. 1978 Update on Giardia
in Washington State aquatic mammals. In Proc. 33rd Intern N W. Conf of Diseases in
Nature Communicable to Man, Vancouver, B. C., Canada, August, 21-23, 1978
15. Chang, S. L. and P. W. Kabler. 1956 Detection of cysts of Endamoeba huitolytica in tap
water by the use of membrane filter. Am Journ. Hyg , 64:170-180
16. Moore, G. T., W. M. Cross, D. McGuire, D. S. Mollohan, N. N. Gleason, G. R. Healy, and
L.H.Newton. 1969 Epidemic giardiasis at a ski resort New Eng Jour. Med., 281.402-407.
-------
210 WATERBORNE GIARDIASIS/DETECTION METHODOLOGY
17. Shaw, P. K. et al. 1977. A communitywide outbreak of giardiasis with evidence of
transmission by a municipal water supply. Ann. Intern. Med., 87.426-432.
18. Jakubowski, W., S. L. Chang, T. H. Ericksen, E. C. Lippy, and E. W. Akin. 1978. Large
volume sampling of water supplies for microorganisms. Jour. AWWA, 70'702-706.
19. Jakubowski, W., T. H. Erickscn, and S. L. Chang. 1978. Detection, identification and
enumeration of Giardia cysts in water supplies. In Proc. AWWA Water Quality
Technology Conference, Kansas City, Missouri, December 4-7, 1977.
20. Morbidity and Mortality Weekly Report. 1977. Giardiasis—Califorma, Colorado Center
for Disease Control, Atlanta, Georgia, 26:60,92:
21. Witherell, L. E. and L. M. Herbert. 1977. An epidemiological investigation of two
outbreaks of giardiasis in Vermont during 1974. Jour. New Eng. Water Works Assoc.,
91:102-118.
22. Bulletin 250M Rev. 2. Honeycomb filter tubes. Commercial Filters Div., The
Carborundum Co., Lebanon, Indiana.
Discussion
T. NASH: Taking the recoveries that you had in the techniques that were
used, can you calculate numbers of cysts you were finding per gallon in the
epidemic study?
W. JAKUBOWSKI: That is really impossible to say at this point. Initially
we were pleased just to be able to detect cysts. In many cases, this was after
examination of several cover slip preparations. Perhaps a volume in the
range of 20 microliters out of 1 milliliter of concentrate may have been
examined microscopically. Generally, when a cyst was detected in a sample
we did not look any further. We usually considered the sampling negative
after an examination of four 5-microliter portions yielded no cysts.
Initially, we were not trying to calculate levels of cysts. One of the highest
that we did detect was about 50 cysts in 225 gallons of distribution system
water. We did do some extrapolation to reach this figure — although we
cannot depend upon this extrapolation, since we did not have any
information on the efficiency of the method used in that particular water.
A. TOMBES: Your schematic showed that you washed off your 7 p.m
screen. Would you then withdraw 1 or 2 samples of 5 jul volumes to examine
for determining cyst concentration?
W. JAKUBOWSKI: Our guideline is that a sample is negative when four 5
H\ portions have been examined without detecting cysts. In positive samples,
further examination of concentrate aliquots were not performed when cysts
were detected. In other words, if the first 5 jul portion was negative and the
second 5 p\ portion contained 1 cyst, the level in the concentrate would be
estimated as 1 cyst/10 p\.
-------
211
The Propagation of Giardia Trophozoites In Vitro
Ernest A. Meyer
University of Oregon Health Sciences Center, Portland, Oregon
ABSTRACT
Giardia tropho/oites have proven difficult to propagate in vitro. Of the three
morphologic types (G duodenal^, G. muris, G. agilu,), only organisms of the G.
duodenalis type, of which Giardia from humans is an example, have been cultured
Although initially Giardia were isolated and cultured in the presence of yeast, then
separated from it, viable yeast is not required for Giardia propagation Cultures of these
organisms have been initiated m the absence of any other symbiont. All of the culture
media presently employed for Giardia propagation are complex and incorporate some type
of animal serum A brief account is presented of the methods used to isolate and propagate
tropho/oites of the G. duodeoah', type
For the purpose of studying Giardia isolated from water supplies it would
be ideal to be able to routinely induce excystation, then to establish and
study the resultant trophozoites in culture. Although Giardia excystation,
with the emergence of motile trophozoites, is now a process that can be
regularly induced in the laboratory, the establishment of these organisms in
culture remains difficult. Most of the Giardia cultures now available were
started from trophozoites obtained from human or animal small intestine.
Of all of the common intestinal protozoa, organisms in the genus Giardia
have proven among the most difficult to establish in culture. Even now, only
a relatively few of these organisms have been isolated, and difficulties with
their maintenance have resulted in their study in relatively few laboratories.
Of the three recognized morphologic types of Giardia, named G. agilis, G.
duodenalis, and G. muris by Filice(l), only isolates of the G. duodenalis type
have been successfully cultured in vitro. Apparently there are no reports of
the attempted culture of G. agilis. Several workers have reported
unsuccessful attempts to culture 6*. mitns (2-4). The failure of G. muris to
grow in media that support G. duodenalis suggests that, despite their
morphologic resemblance, G. duodenalis and G. muris have significant
physiologic differences.
Reports of in vitro maintenance and culture of Giardia trophozoites
began appearing more than a half century ago. Within a few years of each
other, three workers, Chatterjee(S) in 1927, Penso(6) in 1929, and
Poindexter(7) in 1931 independently reported keeping Giardia trophozoites
alive in vitro for up to 5 weeks.
In 1960, Iwata and Araki(8) reported the results of their efforts to culture
Giardia trophozoites from humans in vitro. They used a variety of complex
media, all of which contained pig liver infusion, rabbit serum and
antibiotics. The maximum period of survival they observed was 12 days.
-------
212 WATERBORNE GIARDIASIS/DETEC TION METHODOLOGY
Trophozoite multiplication was observed in all cases on the second culture
day and in some cases on the third or fourth days. In the most favorable
medium, trophozoites survived for 12 days. They believed that bacterial
overgrowth was responsible for termination of Giardia growth.
Karapetyan's work marks the breakthrough in Giardia cultivation. In
1960(9), he reported the method he had used to culture Giardia trophozoites
from humans in vitro for 7 months. He obtained trophozoites together with
Candida guillermondii, by duodenal intubation, concentrated them by
centrifugation, and inoculated the mixture into bottles previously seeded
with chick fibroblasts. His complex culture medium included serum
(human, horse or beef serum was successful), chick embryo extract, chick
amniotic fluid or a tryptic digest of meat, and Hanks or Earl's solution. The
culture medium was changed daily. During the first few weeks the fibroblasts
were gradually destroyed; thereafter the Giardia and Candida multiplied,
apparently symbiotically. In a modified system, Karapetyan cultured
Giardia from the rabbit and from man for up to 5 months in a modified
system in which Saccharomvces cerevisiae was used instead of Candida and
fibroblasts were not employed(lO). Karapetyan reported unsuccessful
attempts to axenically culture Giardia trophozoites. The failure of Giardia
to grow without yeast led Karapetyan to believe that there may be some
synergistic relationship between these organisms.
A number of workers later reported success with Karapetyan's method or
with modifications of it. Solov'ev(l l)in 1971 studied the effect of varying the
serum concentration in Karapetyan's medium on the organism's growth
response. He noted that the serum concentration could be reduced from 25
to 10 percent without significantly impairing the growth response. Several
French workers also reported successful Giardia culture using Karapetyan's
methods. Roux and Ecalle in 1968(12) and Ecalle(13) in a separate paper
that same year studied the effect of whole pancreatic juice and of three
pancreatic enzymes on the in vitro growth response of Giardia from the
rabbit. In 1971, Gayrel and Ecalle( 14) reported that changing the interval of
Giardia medium renewal changed the time of appearance of the organism's
exponential growth phase.
During the late 1950s, we in Oregon also were attempting to culture
Giardia trophozoites. We had not found a successful medium when
Karapetyan's papers became available. We immediately attempted to
duplicate Karapetyan's work but were unsuccessful. In retrospect we
identified our problem. We were using Karapetyan's medium, but we were
not using Karapetyan's organism. We were trying to culture G. muris
because it was readily available, not suspecting that this organism would
behave differently from G. duodenalis organisms from the human or rabbit.
We finally attempted to culture the rabbit organisms using Karapetyan's
medium, and the growth response was so striking that it was clear to us that
Karapetyan was onto something.Subsequently we found that by modifying
Karapetyan's method we could culture Giardia from the rabbit and
chinchilla(lS) monaxenically with 5". cerevisiae.
-------
E. A. Meyer
213
Axenic cultures of Giardia were only obtained in 1970(16). We
approached this problem by assuming that the presence of living yeast was
not essential, and that the contribution of the yeast might be either to
provide nutrients or, as a result of its growth, to lower the redox potential.
We induced the Giardia and yeast to grow in a yeast extract-containing
medium with a low redox potential, then attempted separation of the
Giardia.
We finally obtained axenic Giardia cultures by means of a U-tube(Fig. 1).
Giardia and yeast were inoculated in the right arm of the tube. The yeast,
being nonmotile, settled to the base of the right arm of the tube and remained
there. The (//ar^/m trophozoites, being motile, migrated across the tube. The
tube could be viewed from below with an inverted microscope and the
progress of the trophozoites across the tube could be determined. After
about a month Giardia could be seen at the base of the left arm of the tube
free of yeast. These Giardia, now free of yeast, were inoculated into tubes of
the same culture medium, where they failed to grow.
These organisms would grow across a dialysis membrane from viable 5.
cerevisiae (Fig. 2). Periodically we attempted to subculture these Giardia
axenically, and after a month of trying, axenic cultures of trophozoites from
the rabbit, chinchilla and cat were obtained( 16). Subsequently we succeeded,
using similar methods, in isolating and culturing Giardia from humans(17)
and guinea pigs(18).
Insert Fig. 2
The separation of Giardia from yeast is probably the most tedious part of
obtaining an axenic culture of Giardia. We know now that there are easier
ways of separating these organisms. One is the judicious use of an antifungal
Vaccine stoppers
-M3 medium
Supporting plate
-4
Viewing slot
FfG. 1. i!-tube used to obtain Giudi* tropkoioites fret of S. cerevMM.
-------
214 WATKRBORNK GIARDIASIS/DETECTION METHODOLOGY
Dialysis bag containing
S. cerevisiae
Giardia free of
S. cerevisiae
FIG. 2.
U-tube in which Giardia trophozoites are cultured across a dialysis membrane from
S. cerevisiae.
agent, slowly increasing the amount added to the culture until only a small
amount of Saccharomyces remains, at which time subcultures are often free
of yeast.
Another way of obtaining axenic Giardia cultures is to start without yeast.
Bingham (See Bingham, these Proceedings) has shown that it is possible to
initiate axenic cultures from treated cysts, first concentrated and purified
from human or monkey feces.
Other workers are culturing and studying these organisms. Visvesvara et
al(19) succeeded in adapting Giardia from humans to grow in media
containing either bovine or rabbit serum. This is an important contribution
because organisms grown in medium containing human serum are not
suitable for use in the indirect immunofluorescence test for the detection of
antibodies to Giardia in the serum of patients.
Finally, Lakhonina in Estonia(20) has reported axenically culturing 13
strains of Giardia from the rabbit; she and Teras(21) have reported that
adding 1 to 1.5% agar to Giardia media offers several advantages. It permits
slower growth of the organisms and medium changes at less frequent
intervals and it is a medium in which the ability of Giardia to produce acid or
gas from a variety of sugars can be tested.
Presently, then, methods are available to isolate and culture organisms of
the G. duodenalis type. The methods are difficult and we are working to
develop simplified procedures that will permit the widespread study of these
organisms. Giardia have not yet been cultured on a defined medium and this
is another objective to which we are directing our attention. Organisms of
the G. muris type remain a challenge to culture; no one up to now has
apparently even come close. And G. agilis, the representative of this genus
-------
E.A.Meyer 215
found in coldblooded vertebrates, remains largely unknown; a search of the
literature yields no reports of its attempted culture.
REFERENCES
1. Filice, F. P. 1952 Studies on the cytology and life history of a Giardia from the laboratory
rat. University of California Publications in Zoology 57 53-146.
2. Nozaki, Y. 1956 Survival of Giardia muri\ in artificial media (an attempt of the cultivation
of the organism). 1RYO 10-401-405
3. Hasslinger, M. A. 1966 Beitrag zum Lambhenproblem Zeitschnft fur Parasitenkunde
28.65-74.
4. Meyer, E. A. and Radulescu, S. 1975. Characteristics of Giardia trophozoites in culture.
Second European Multicolloquy of Parasitology, 'I rogir, p 30
5. Chatterjee, G.C. 1927 On cultivation of Gmrcha mieMinahs. Indian Medical Record 47 33.
6. Penso, G. 1929 The morphology, life history, culture and treatment of Giardia inie\imali!,.
Annali di medicana Navale (e colomale) 2:125-161
7. Poindexter, H. A. 1931 Studies on the cultivation of parasitic intestinal flora. Puerto Rico
Review of Public Health and Tropical Medicine 7.417-434.
8. Iwata, S. and Araki, T. I960 Studies on giardiasis. Bulletin of the Osaka Medical School
6.92-106.
9. Karapetyan, A. E. I960. Methods of Lamblia cultivation. Tsitologna. 2 379-384.
10 Karapetyan, A. 1962 In vitro cultivation of Giardia duodenal^. Journal of Parasitology
48-337-340
1 /. Solov'ev, M. M., Akimova, R. F. and Shmakova, V.I. 1971. Isolation and maintenance of
Lamblia duodenahs cultures on a modified Karapetyan medium. Meditsmskaia
Para/.itologna i Parazitarnye Bolezm 40:75-78.
12 Roux, G. and Ecalle, R. 1968. The effect of pancreatic juice on the proliferation in vitro of
Giardia duodenalis. Comptes Rendus Hebdomadairedes Seances de 1'Academie des
Sciences 266 2434-2436.
13 Ecalle, R. 1968 Influencedequelquesen/ymesdusucpancreatiquesurlaculturemvitroal
Giardia duodenalis. Comptes Rendus des Seances de la Societe de Biologie et de Ses
Fihales 162:2321-2323.
14. Gayrel, R. and Ecalle, R. 1971 Influence of periodic renewal conditions of media on
cultural parameters of Giardia duodtnaln. Comptes Rendus de la Societe de Biologic de
Toulouse 165728-731
15 Meyer, E. A. and Pope, B.L. 1965 Culture in vitro ot GV/Y/wtropho/oites from the rabbit
and chinchilla. Nature 207 1417-1418
16 Meyer, E. A. 1970 Isolation and axenic cultivation of Giardia tropho/oites from the
rabbit, chinchilla, and cat Experimental Parasitology 27:179-183
17 Meyer, E. A. 1976. Giardia lamblia isolation and axenic cultivation Experimental
Parasitology 39 101-105.
18. Fortess, E. and Meyer, E. A. 1976 Isolation and axenic cultivation of Giardia trophozoites
from the guinea pig. Journal of Parasitology 62.689.
19 Visvesvara, G. S., Healy, G. R. and Meyer, E. A. 1977. Comparative antigenic analysis of
Giardia lamblia tropho/.oites grown axemcally in medium supplemented with either
human, bovine, or rabbit serum 5th International Congress of Protozoology, New York
20. Lakhonina, G. 1978 On the feasibility of cultivating Giardia in solid media and
determining their saccharolytic properties Easti Nsv Teaduste Akadeemia Toimetised. 27
Koide Biologia I 47-54
21. Teras, J. and Lakhonina, G. 1975 Solid medium for prolonged axenic cultivation and
testing of biological properties of Giardia. Second European Multicolloquy of
Parasitology, Trogir pp. 30-31.
Discussion
M. WOLFE: In the relationship of yeast to trophozoite growth, if yeast
flora were present in the small bowel, it might help to support new growth of
Giardia and perhaps work synergistically towards some pathogenic effect.
Although I do not know what factors would induce yeast growth in human
beings, I believe this is a possibility that gastroenterologists should consider.
-------
216 WATERBORNE GIARDIASIS/DETECTION METHODOLOGY
Can you give us some idea of the effect of the presence, or absence, of yeast in
the fluid of the small bowel?
E. ME YER: It would not surprise me if yeast is present in the small bowel
and aids in the growth of Giardia. However, I believe that its presence is not
essential for Giardia growth in the small intestine. Bemrick (Bemrick, W. J.
and M. K. Grady, J. Parasitol. 51:685, 1965) wondered about the need for
yeast in the small intestine. As a result, he examined a large number of
sacrificed dogs to determine what per cent contained Giardia alone in the
small intestine, yeast alone, or Giardia plus yeast. His conclusion was that
there is no relationship to incidence of Giardia and yeast in the small
intestine although you often find the 2 together.
D. JURANEK: Is there a morphological difference between G. agilis and
G. muris that is obvious?
E. MEYER: G. agilis are much longer and much thinner. They have
teardrop shaped median bodies. G. muris and G. duodenalis in a stain, look
similar except for the median bodies, while the G. agilis is perhaps twice as
long and considerably narrower.
L. DIAMOND: In reference to the comment Dr. Wolfe made about the
relationship between yeast and Giardia, I would suspect there is no
relationship of the type under discussion.
Have you varied the amount of serum that you have used in attempts to
culture various types of Giardia! How many strains from humans have you
been able to isolate?
E. MEYER: I do not know if you can consider them strains or not. We
have cultured them from 2 or 3 different people. For the culture of
trophozoites, one of my students determined the requirement for serum. He
went from 0 to 95 per cent serum. Below 15 to 20 per cent, you encounter
more difficulties; however, there are reports that you can successfully use
10% serum. That has not been our experience.
L. DIAMOND: With histolytica there are some strains which require 15
to 30% serum for axenization. Once they have been axenized, then we can
drop down to the usual 10 per cent. We have cultured some at 5 per cent, but
I have never been able to axenize any at 5 per cent serum.
E. MEYER: If you are trying to axenize, it would be a good idea to vary
the serum concentration.
-------
217
Induction of Giardia Excystation
And
The Effect of Temperature on Cyst Viability
as Compared by Eosin-Exclusion and In Vitro
Excystation*
A. K. Bingham, E. L. Jarroll, Jr., E. A. Meyer
University of Oregon Health Sciences Center, Portland, Oregon
Simona Radulescu
Institute Cantacuzino, Bucharest, Romania
ABSTRACT
Factors involved in Giardia excystation have been investigated G«7/W;fl excystation has
been shown to be a pH-dependent process unaffected by the presence of gastric salts or
pepsin. A method for the m vitro excystation of Giardia has been developed and utilized to
induce excystation in cysts obtained from humans, monkeys, dogs, rats and mice.
The viability of Giardia cyst suspensions was compared by the excystation procedure
and eosm-exclusion Eosm-exclusion consistently indicated higher cyst viability than
could be demonstrated by in vitro excystation%
Using excystation as the criterion of viability, the effect of four storage temperatures
(-13, 8, 21 and 37°C) on Giardia survival in water was studied
Storage at 8°C permitted longest cyst survival upwards of 77 days. Cysts stored at
21° C retained their viability for 5 to 24 days, while those at 37° C never survived longer than
4 days. Freezing and thawing of Giardia cysts resulted in an almost complete loss of
viability although a low level of viability (less than 10J) persisted for at least 14 days.
EXCYSTATION OF GIARDIA
The information which is available describing excystation of intestinal
protozoans is largely based on in vivo and in vitro research on Entamoeba,
Eimeria, and Giardia (I). Of these genera, the latter has been studied least.
Many of the early descriptions of Giardia excystation were based on in
vivo observations. In 1925, Hegner (2) described Giardia excystation after
having observed the process in a patient's stool. Later, he reported observing
the process in the intestine of rats 20 min to 4J4 h following peroral
administration of Giardia lamblia cysts (3). From this experiment, Hegner
concluded that moisture, temperature, and certain unspecified digestive
juices were major factors in the excystation process. In 1937, Armaghan (4)
attempted to pinpoint the site of Giardia excystation by placing suspensions
of Giardia muris cysts directly into the stomach, duodenum, jejunum, ileum,
and caecum of several rats. After 8 days, she found that only those rats
* Financial support for these studies was provided by the Environmental Protection Agency,
Cincinnati, Ohio.
-------
218 WATERBORNK GIARDIASIS/DETECTION METHODOLOGY
which had cysts placed in the stomach or duodenum were excreting cysts,
while none of the other rats was found to be infected at autopsy 1 8 days later.
Deschiens (5) found dead trophozoites in gastric juice and aqueous
hydrochloric acid to which he had exposed Giardia cysts during an attempt
to determine if cysts survived gastric passage. This was apparently the first in
vitro excystation of Giardia. However, it was not until the work of Bingham
and Meyer (6) that a factor inducing Giardia excystation was demonstrated.
Cysts were exposed to solutions of varying pH, ranging in complexity from
complete synthetic gastric juice to aqueous hydrochloric acid. This
experiment, which is summarized in Table 1, showed that the hydrogen ion,
chloride ion, or a combination of the two was responsible for inducing
excystation.
Table 1. Percentage excvsiation of Giardia o'i/i exposed to complete
and component-varied synthetic gastric juice.
Solution
Complete (HCI + salts + pepsin)
HCI + salts
HCI only
Water (control)
HSP-3 (control)
pH
1 6
1 6
1 6
6 8
6 8
Mean* % excystation
± SE M
25 8 ± 7 3
23 0 ± 7 0
23 2 ± 7 0
0 1 ± 0 1
00± 00
* Means are derived from a total of 1191 to 1724 cysts which were counted in each solution.
A subsequent experiment using a variety of inorganic acids (Table 2)
confirmed that the hydrogen ion was required to initiate excystation
regardless of the counter ion used. These results are similar to findings
reported for other cyst-forming protozoans. For example, Kaushal and
Shukla (7) found that treatment of cysts with HCI and trypsin accelerated
excystation of Hartmannella culbertsoni.
More recently, unpublished data from our laboratory indicate that
additional factors are involved in excystation. Table 3 shows the excystation
response of cysts exposed from 5 minutes to 4 hours to aqueous solutions of
HCI at various pH values. The following were noted. First, lowering the pH
Table 2. Percentage excystation o^Giardia cvsls exposed to inorganic acids at pH 2.0 to 2.1.
Mean" % excystation
Acid ± S E M
Water (control) 0 2 ± 0 2
HCI 17 4± 2.2
HNO3 12 5 ± 3 0
HCI04 15 5 ± 2 0
H2SC>4 155126
H3PQ4 19.4 ± 2.8
* Means are derived from a total of 2096 to 2392 cysts which were counted in each solution.
-------
A.K.Bingham 219
decreases the exposure time required for excystation. Second, it appears that
for acid pH values there exists an optimum range of exposure times above
and below which significantly lower levels of excystation occur. Third, the
mean optimum exposure time within this range appears to increase as the pH
increases. Additionally, the mean optimum percentage of excystation is
greater at low pH (0.5 to 2.0) than at higher ones. The maximum level of
excystation at pH 0.5 and 2.0, while not significantly different from each
other, are significantly greater than those at pH 4.0 and 6.2 (p<0.01). The
higher mean optimum excystation percentage at pH 0.5 and 2.0 indicate that
the excystation process 1) is favored by low pH values and 2) has adapted to
proceed optimally at pH values which correspond to those occurring
normally in the human stomach.
The medium into which cysts are transferred after exposure to acid is also
critical to the excystation process. It can be seen from Table 4 that transfer
of acid-exposed cysts to either water or HC1 yields negligible excystation.
On the other hand, transferring such cysts to HSP-3 medium (8) routinely
used to culture Giardia trophozoites results in significant levels of
excystation. This suggests that while the hydrogen ion activates the
excystation process, subsequent transfer to a favorable medium is required
for the completion of the process. HSP-3, the only medium that has been
tested in this system, appears to be favorable but it is possible that other
physiological solutions may also prove satisfactory. Similar observations
have been made with other protozoans. For example, Yorke and Adams (9)
observed that suspending Entamoeba histolytica cysts in water at 37° C was
not deleterious to the cysts, but was lethal to excysting amoebae. These
amoebae were often destroyed by the water before they could escape their
envelopes.
Temperature has also been identified as important in the excystation
process. The temperature of both the acid solution and the subsequent
excystation medium (HSP-3) appears to be crucial. The results presented in
Table 5 demonstrate that high levels of excystation will occur when the
temperature of both is 37° C. Lower levels of excystation occurred when
certain temperature combinations were used, but no excystation was
observed when the temperature of the HSP-3 was held at 8°C regardless of
the temperature of the HC1.
The process of Giardia excystation appears to be similar to that of
excystation of metacercariae of the trematode, Paragonimus westermani,
which also excyst in host small intestine. Metacercariae of P. westermani
excysted when exposed to a medium consisting of an isotonic salt solution at
40°C and pH 9, following initial exposure to an acid medium of pH 3 (10).
EFFECT OF TEMPERATURE ON CYSTS
With in vitro excystation of Giardia as an established procedure, it is now
possible to re-assess earlier reports on other cyst characteristics such as
thermal sensitivity. Using eosin-exclusion as their criterion for viability,
Boeck (11) reported a thermal-death point of 64°C for G. lamblia cysts,
while Cerva(12) reported 50°C as the upper limit of temperature tolerance.
-------
220 WATERBORNE GIARDIASIS/DETECTION METHODOLOGY
1
Q
2:
3
•5
j^
£
^
0
CD
O
en
o
CM
LD
•-
O
LO
I
d
CO "3" CD LO
<3- CM CM
CM CM
O O LO O
*— CM ^f CM
i~\
! . i O(
, O CD Col
1 CM
i CO CM 03
1 LO
co o col col
ro 0 CM)
CM" OJcMJ0
* Cslj LO
i ^j '
' ^^
IJ i , i
CM ' ' '
.
S^^ 1
col co o ,
^H CM i
C\Jj
LO O O CN
O CN ^t CD
c
OJ
v
C-
4J
W
4J
T3
OJ
C
,
rt
Jo
u-
cd
~"
^
•J
*
?
3
C
6
o
o
X
CL
-D
•3
~3
~C
O
CJ
O
u
o
C
(D
U
£
•5
"1
c
o
c«
o
X
<~
0>
00
5
c
0
fci
u
a.
c
(L>
(«
1>
D.
1>
1
cd
*
O
o.
u
.c
o
•5
QJ
E
o
<—
c
1
.
'c
CJ
<«
C
op
"trt
o
c
OJ
u,
o
3!
a.
c
•^
00
f
cfl
c
_0
3
«
o
CO
O
Ml
C
O.
s
J
O
C
o
^0
c
Si
o
o
cu
*^
M
-a
u
c
i
fe
•o
O
c
cd
c
~n
CJ
X
*
*
d,
C
u
bfi
cd
cd
(Lt
O
CJ
(U
£
O
<^
c
u
1
-5
j-^
c
CJ
^5
c
oo
o
c
ej
>
c
o
cystati
t*-
o
OJ
00
=a
c
4J
CJ
9J
a.
E
3
E
D.
O
r;
CJ
"i
^_,
w
OJ
E
(U
3
O
D.
O
(D
oo
(U
>
+
I
C.
c
ob
cd
^_
tu
•£
o
CJ
S
e
0
c
'ts
TD
^
"^
cd
CJ
C
00
O
c
cd
CJ
C
o
cd
x
CJ
X
u
t—
O
(U
00
cd
C
a>
o
8.
£
a
E
Q.
O
D
00
flj
CU
<
i
-------
A.K.Bingham 221
Table 4 Percentage excvstalion f± S E M.) of Giardia cysts exposed to various
HCl solutions then transferred to either HCI solutions at the same pH as the original solution,
water at pH 6 8, or HSP-3 at pH 6 8
Incubation media
pH
0
2
4
6
5
0
0
2
0
0
0
0
Water
1 ± 0
1 ± 0
0± 0
1 ± 0
2
2
0
2
0
0
0
0
1
1
0
0
HCI
± 0
± 0
± 0
± 0
HSP-3
2
^
0
0
2
15
17
0
0±
4±
3±
3±
0 8
2 7
2 3
0 3
Additionally, Boeck (13), using the same criterion of viability, noted that at
temperatures from 12to22°C Giardia cysts remained viable for as long as 66
days. However, the eosin-exclusion method of determining cyst viability has
been criticized as unreliable for determining the viability of Entamoeba
histolytica cysts (14,15). Earlier workers preferred in vitro excystation as the
criterion of E. histolytica cyst viability.
In our laboratory, we have compared eosin-exclusion to in vitro
excystation as methods for evaluating Giardia cyst viability. Cysts were
purified by a modification of other procedures (16,17) as follows: Feces are
emulsified to a thin consistency in tap water, filtered through a layer of
cheese cloth to remove large particles, and 3 to 5 ml of filtrate overlaid on 3
ml of chilled 0.85M sucrose in a 15 ml conical centrifuge tube. The tube is
centrifuged at 600g for 5 min at room temperature and the water-sucrose
interface removed, diluted 1:10 with tap water, and recentrifuged for 5 min.
The pellet is resuspended in 3 ml of water and the sucrose gradient
centrifugation repeated. Following this the water-sucrose interface is again
removed, diluted 1:10 with water, filtered through a 20 m-aperture nylon
mesh (Tetko, Inc., Elmsford, N. Y.), and centrifuged for 5 min. The resultant
pellet is resuspended in tap water to the desired cyst concentration and
stored at 8°C. This procedure can be accomplished in 1 h yielding 30% cyst
recovery, with purity acceptable for light microscopic examination and
enumeration, and apparently undiminished cyst viability.
Three aliquots of Giardia cysts, purified from the same fecal specimen,
were stored in unchlorinated tap water (pH6.8) at 37, 21, and 8°C,
respectively. Cyst viability, evaluated by both methods, was assessed
Table 5 Excvslanon of Giardia cysts exposed to temperature-varied HCI fpH 2 0)
and transferred to temperature-varied HSP-3.
Acid temperature
(°C)
8
21
37
8
0
0
0
HSP-3 temperature' (°C)
21
0 5
3 0
15 6
37
1 5
5 4
100 0 '
* Values represent percentage of excystation with reference to the control tube (HCI 37C,
HSP-3 37C) which is arbitrarily 100%.
-------
222 WATERBORNE CIARDIASJS/DETECTION METHODOLOGY
periodically for as long as the cyst suspension lasted, or until excystation
and dye exclusion had decreased to zero for 5 days. The results of these
experiments are summarized in Figures 1, 2 and 3. In every case, the eosin-
exclusion test indicated higher levels of viable cysts than could be
demonstrated by the excystation procedure. In cysts incubated at 21 and
37° C (Figs. 1 and 2) eosin-staining (indicating dead cysts) eventually
reached 100% but not until several days (even weeks in some cases) after
excystation indicated the same thing. With cysts stored at 8°C for 77 days
(Fig. 3), eosin-staining indicated more than 80% viability while the
excystation procedure indicated viability of less than 1%. If one considers
the ability to excyst as a criterion of viability, then our results suggest that in
vitro excystation is a more sensitive indicator of cyst viability than is eosin-
exclusion. Furthermore, the results indicate that cysts which neither stained
with eosin nor excysted were either dead, or alive but incapable of excysting.
Giardia cysts stored at 8°C had higher levels of excystation for a longer
duration than those stored at 21 or 37° C. Cysts stored at 37° C never
survived longer than 4 days and usually the percentage of excystation was
greatly reduced even after 24 h (Figure 4). At 21°C, cyst survival ranged
from 5 to 24 days. In all cyst suspensions examined from fresh fecal
specimens and stored at 8 or 21°C, low initial levels of excystation were
observed. This period of low excystation, varying from 2 to 7 days and
followed by an increase in the level of excystation, suggests that a
100 r
•—• Eosm-exclusion
A A Excystation
3 10 12 14 16
Storage Time (days)
18 20 22 24
FIG. 1. Comparison of Giardia viability as determined by excystation and eosin-exclusion
of cysts stored at37°C.
-------
A. K. Bingham
223
100
90
80
.§ 70
I
I
• • Eosm-exclusion
A A Excystation
60
^ 40
'-§
•^ 30
O
£ 20
10
0
0246 8 10 12 14 16 18 20 22 24 26 28 30 32
Storage Time (days)
FIG. 2. Comparison of Giardia viability as determined by excystation and eosin-exclusion
of cysts stored at 21° C.
•—• Eosm-exclusion
A—A Excystation
26 10/4 18 22 26 30 34 38 42 46 50 77
Storage Time (days)
FIG. 3. Comparison of GUrdia viability as determined by excystation and eosin-exclusion
of cysts stored at S°C.
-------
224 WATERBORNE GIARDIASIS/DETECTION METHODOLOGY
100
90
80
70
60
50
40
30
20
10
—• 8°C
—A 37°C
FIG. 4.
10 12 14 16
Storage Time (hours)
The effect of storage at 8 and 37° C on Giardia cyst viability as determined by
excy station. Cysts were permitted to mature at *° Cfor I week prior to beginning of
the experiment.
maturation period is required before Giardia cysts reach their full
excystation potential. A maturation period of 800 h or more has been
reported for Acanthamoeba castellanii (18). Even though Dutta and Jehan
(19) found that they could induce A. castellanii excystation in less time than
the previous workers,- the cysts still required a maturation period of
approximately 3 days.
In an attempt to assess the effect of freezing and boiling on Giardia cysts, 3
aliquots of the cyst suspension were placed at temperatures of 100, 8 (as
control), and -13C, respectively, and viability determined periodically by the
in vitro excystation procedure. Cysts subjected to 100°C were killed
immediately (Table 6). Freezing cysts also greatly reduced their viability
(Figure 5), but it is interesting to note that very low levels of excystation were
observed even after low temperature storage for 14 days. This suggests that
Table 6 Percentage excyslation of Giardia cvsts exposed to boiling water.
Control
(8°C)
% excysting
29
1
Boiling
0
0
1
0
3
0
5
0
time (minutes)
10
0
15
0
20
0
25
0
30
0
-------
A. K. Bingham
225
50 r
30
20
10
•—• 8°C
A—A -|3°C
468
Storage Time (days)
10
12
14
FIG. 5.
The effect of storage at 8 and -13°C on Giardia cyst viability as determined by
excystation. Cysts were permitted to mature for 1 week prior to beginning of
experiment.
the freezing process itself and not the low temperature may be the principal
factor responsible for cyst death.
i
10
11
REFERENCES
I ackie, A.M. 1975 The activation of infective stages of endoparasites of vertebrates Biol.
Rev. 50.285-323.
2. Hegner, R.W. 1925 Excystation m Giardia lambha from man. Am J. Hyg 5:250-257.
3. Hegner, R.W. 1927. Excystation and infection in the rat with Giardia lambtia from man.
Am J. Hyg. 7:433-447
4 Armaghan, V. 1937. Biological studies on the Giardia of rats. Am J. Hyg. 26 236-258.
5. Deschiens, R. 1929 Chimisme gastnque et infectious parasitaires du tube digestf. Ann. de
1'lnstitute Pasteur. 43 1353-1369.
6. Binghant, A.K. and E.A. Meyer, 1979. Giardia excystation can be induced in vitro in acidic
solutions Nature. 277:301-302
7 Kaushal, D.C. and O.P. Shukla, 1977 Excystment of axemcally prepared cysts of
Harlmannella culberlsoni J. Gen. Microbiol 98:117-123.
8. Meyer, E.A. 1976. Giardia lambha' Isolation and axenic cultivation. Exp. Parasitol.
39:101-105.
9. Yorke, W. and A.R.D. Adams. 1926(a). Observations on Enlamoeba histolytica. 1.
Development of cysts, excystation, and development of excysted amoebae, in vitro. Ann.
Trop. Med. Parasitol. 20.279-303.
Yogore, M.G., B.J. Cabrera, T.P. Araullo, and E.F. Cabatega. 1959. Studies on
paragommiasis. VIII. On the excystation of Paragammus metacercanae. Philippine./. Sci.
88:61-80.
Boeck, W.C. 1921(a). The thermal-death point of the human intestinal protozoan cysts.
Am. J. Hyg. 1:365-387
12. Cerva, L. 1955. The effect of disinfectants on cysts of Giardia mtestinamis. Ceskoslovenska
Parasitol. 2:17-21.
13. Boeck, W.C% 1921(b). On the longevity of human intestinal protozoan cysts. Am. J. Hyg.
1:527-540.
-------
226 W ATERBORNE GIARDIASIS/DETECTION METHODOLOGY
14. Vorke, W. and A.R.D. Adams. 1926(b) Observations on Entamoeha histolvtica. II.
Longevity of the cysts in vitro, and their resistance to heat and to various drugs and
chemicals. Ann. Trop. Med. Parasitol. 20.317-325.
15. Chang, S.L. and G.M. Fair. 1941. Viability and destruction of the cysts of Endamoeba
histolvnca. J Am. Water Works Assoc. 33.1705-1715
16 Roberts-Thomson, I.C., D.P. Stevens, A.F. Mahmoud, K.S. Warren. 1976. Giardiasis in
the mouse' an animal model. Gastroenterology 71:57-61
17 Sheffield, H.G. and B.Bjorvatn. 1977. Ultrastructureof the cyst of Giardia lamblia. Am. J
Trop. Med. Hyg. 26(l):23-30.
18 Chambers, J.A. and J.E. Thompson. 1974. Age-dependent excystation of the proto7oan
Acanthamoeba castellmu. J. Gen Microbiol. 80:375-380.
19 Dutta, G.P. and M. Jehan. 1977. 'Age-dependent excystment' of cysts of soil amoeba.
Indian J. Exp Biol. 15.72-74
Discussion
D. STEVENS: What strain of Giardia did you use?
A. BINGHAM: It was a human strain.
D. STEVENS: Have you looked at differences between G. muris and G.
lamblia in terms of resistance of cysts to various physical factors?
A. BINGHAM: No. We only performed the excystation procedure with
G. muris, and we had good success. Using pH 2 for approximately 30
minutes, we obtained about 50 or 60% excystation with G. muris, which is
comparable to what we obtained with G. lamblia.
D. STEVENS: Have you compared excystation to viability as
demonstrated by infectivity of the cysts?
A. BINGHAM: We have not yet. It seems like a difficult problem, but we
may approach it in the future.
G. JACKSON: This is really a lovely series of experiments on excystation.
Did you perform any experiments on encystation?
A. BINGHAM: When I initially started experiments in the lab, I made a
few different attempts to induce encystation. I decided that excystation
might be a more fruitful area and, fortunately, it was for my thesis. Thus far
we have not had any success with our encystation attempts.
G. HEALY: I would like to add my congratulations for a beautiful piece
of work. 1 am intrigued by the openings in the end of the cysts where the
trophozoites come out. 1 do not remember seeing any electron micrographs
indicating a weakness in the cyst in this area. Do the trophozoites always
come out from the same area of the cyst?
A. BINGHAM: We have noticed some interesting things concerning that
area of the cyst. We have not been able to determine which end of the cyst, if
you can differentiate the ends, the trophozoite usually comes out of.
Fortunately we had the same donor for the entire period of our*
experiments. As this donor approached spontaneous remission we observed
increasing numbers of odd-shaped cysts from which trophozoites attempted
to excyst at the side of the cyst rather than through one end. Excystation was
never successful in these cysts.
We believe this to be a type of abortive excystation and may be related to
the immunological response of the host. The majority of viable-looking
cysts in good preparations will emerge from one end; we do not know which
end it is or if that makes any difference.
L. HIBBERT: Did you notice any thinning of the membranes on the end
of the cyst just after your hydrochloric acid treatment? 1 was wondering if
-------
A. K. Bingham 111
this might have been observed prior to your treatment with the hydrochloric
acid.
A. BINGHAM: No. With the light microscope the most detailed thing
that we have been able to see is the withdrawal of the trophozoite from the
cyst wall and then penetration of the wall, and we have not been able to
visualize what happens during that particular moment.
V. OL1V1ERI: Could you give a more detailed description of your
method of calculating per cent of excystation? Was it a method similar to
what R.P. Stringer (J. Parasitol, 58:306, 1972) used for Entamoeba
histolytica, counting empty versus full cysts?
A. BINGHAM: No, it was not. We found that counting empty Giardia
cysts is more difficult and less accurate than counting trophozoites. I assume
that with Entamoeba histolytica there is a great deal of activity and
movement immediately after excystation. However, with Giardia we found
that in most instances this is not the case. The trophozoites, once they have
excysted, will settle down immediately near the cyst and complete the
process of division, and usually they will stay there. With the cultures that
we have initiated, we have noticed that the trophozoites will settle down in
one spot and multiply there; they do not seem to move.
Another reason that we prefer trophozoites to empty Giardia cysts is that,
particularly when there is a moderate amount of debris, empty cysts are
sometimes difficult to see. On the other hand, the trophozoites have active
flagella and will cause a great amount of stirring. In fact, they clear out large
areas of debris around them with this flagellar motion, and are easily
observed.
R. OWEN: Would you care to speculate how giardiasis can then occur in
patients with achlorhydria?
A. BINGHAM: If we can bar the possibility that these people were
already infected or that they may have somehow consumed trophozoites
rather than cysts, then I would have to put forth the opinion that there is a
large variability in the sensitivity of cysts to various pH's. There seem to be
sub-populations of cysts within one fecal specimen with varying sensitivity
and responses to pH and exposure time to acid. There may be a small
percentage of cysts within any population of cysts which will excyst even
under near neutral conditions. We have seen low level, maybe 0.1 or 0.2 per
cent, excystation even in solutions which were close to neutrality, pH 6.2 to
6.8, and this may be one possible explanation for infection in these people.
F. GILLIN: I would like to know the approximate time course of the
excystation sequence.
A. BINGHAM: With very active cysts in a good, viable preparation, we
usually see excystation occurring within 10 to 30 minutes following transfer
into a HSP-3 medium. The majority will be out within 10 to 15 minutes.
They emerge very rapidly if they are going to emerge.
F. GILLIN: Have you tried to keep the emerged trophozoites growing in
culture?
A. BINGHAM: Yes, we have. We have established trophozoite cultures
from two humans and from monkeys axenically. We have also induced
excystation with Giardia cysts from dogs, cats, rats and mice thus far, and
we are very confident that we can probably do it with just about any cyst
from any species of animal.
-------
228 WATERBORNE GIARDIASIS/DETECTION METHODOLOGY
A. TOMBES: I have not seen any pores in the end of the cysts as you
indicated. However, on Isospora, there is a very definite pore probably a
half micron in diameter in one end of the cyst for the organism to come out.
Could you comment further on the odd shaped cysts that you saw at
different times?
A. BINGHAM: We saw cysts that ranged from the normal oval shape to
square, teardrop, or dumbbell shaped. Most of the time we saw these odd
shapes in preparations from this particular patient as he approached
spontaneous remission of the infection. We assume that this may have
something to do with the immune response.
From my recollection, the total viability with excystation in these
preparations is much lower than with the normal looking cyst. We have had
excystation as great as 60% with good batches of normal looking cysts,
whereas with these other batches excystation normally is 5 per cent or less.
J. HOFF: I would like to add our congratulations on the quality and
success of your research in this area. I am particularly happy since EPA has
been funding this research for the last couple of years. This research really
opens the way for addressing the applied type of question concerning water
treatment, that is, regarding these organisms and disinfection. We look
forward to good progress in this area in the future, particularly since you are
involved in this work now.
In the study during which you stored the cysts for a period of 24 hours and
over all showed no decline in viability there was considerable variation in the
per cent excystation from one sample to another. Is that due to errors in
counting? Does it indicate an extreme sensitivity to exposure to the acid? Do
you have any idea what caused the variations?
A. BINGHAM: We are convinced that at least a portion of it is due to
sampling variation since we are taking a relatively small sample. For most of
the results that we have shown, we had counted maybe 1,000 cysts at one
particular value. A total of 36,000 cysts were sometimes counted in one
experiment, but even that was only a small percentage of the total
population in a cyst preparation.
However, over weeks of storage we notice peaks of excystation followed
by troughs which will last for several days and then a peak again, with as
great as 30 or 40% difference between a trough and peak. I believe that this is
related to something inherent in the cysts, as well as to sampling variation.
In other words, maybe some cysts are ready to excyst at one particular time
whereas others are not. I do not know the reason for this. There might be
some kind of population effect there, possibly related to age. There may be
some maturation that occurs in the gut. One group of cysts may remain in
the gut for 2 days or longer, and are excreted mature and ready to excyst;
another sub-population may be excreted rapidly and are a day or two less
mature. So as a result, we see excystation peaks 1 or 2 days apart.
S. ERLANDSEN: You mentioned a latent time of approximately 2 to 7
days before excystation. What was the medium for storage of your cyst
material? Were these cysts that were kept at refrigerator temperatures, and
would then excyst after that latent period?
A. BINGHAM: The cysts were stored in tap water at 8°C and did excyst
after a latent period.
S. ERLANDSEN: We are concerned with fecal droppings in some of
these reservoir areas. In your attempts to examine feces, what type of cyst
-------
A. K. Bingham 229
viability might you expect with material that is left out to essentially
desiccate as we expect to take place in nature?
A. BINGHAM: When I began these experiments 1 initially did try leaving
fecal specimens in the refrigerator and using portions as I needed them
before we had developed a purification procedure. In my recollection the
cysts did not survive as long as purified cysts. It seems to me that the cysts
will last a little longer in water than they will in feces, and as far as
desiccation goes, we have not done anything in that area.
-------
230
SESSION V - WATER TREATMENT
TECHNOLOGY
Chairman - David W. Liechty, Department
of Social and Health Services, Olympia, Washington
Disinfection Resistance of Giardia Cysts:
Origins of Current Concepts and Research
in Progress
J. C. Hoff
Water Filtration Techniques for Removal
of Giardia Cysts and Cyst Models
G. S. Logsdon, J. M. Symons, R. L. Hoye, Jr.
-------
231
Disinfection Resistance of Giardia Cysts:
Origins of Current Concepts and
Research in Progress
John C. Hoff
Microbiological Treatment Branch
Drinking Water Research Division
Municipal Environmental Research Laboratory
U.S. Environmental Protection Agency
Cincinnati, Ohio
ABSTRACT
Our current concept of the resistance of Giardia lamblia cysts to disinfection, as applied
in treatment of drinking water, is that the cysts are highly resistant and that their control by
disinfection is not feasible. The origins of this concept appear to be based on early
disinfection studies and on their similarity to another pathogen, Entamoeba histolvlica
which has been shown to be highly resistant to disinfection. The results of more recent
studies indicate that the early studies in which E Hiatolvtica and G. larnblia cyst viability
was determined by dye exclusion provided results indicating erroneously high disinfection
resistance. Therefore, the early disinfection studies should be disregarded. The application
of a recently developed in vitro excystation method to Giardia disinfection studies and the
current Giardia disinfection research program are discussed.
The recent occurrences of waterborne outbreaks of giardiasis have
resulted in a recognition of our lack of knowledge of the effects of water
treatment processes on the etiologic agent, Giardia lamblia. The most
important single water treatment unit process for the control of pathogenic
microorganisms is disinfection. The initiation of chlorination in the
treatment of drinking water in the early part of this century caused a
dramatic reduction in the occurrence of cholera, typhoid fever and other
waterborne diseases.
Because it has been considered unimportant as a waterborne pathogen, G.
lamblia, in contrast to the enteric bacteria, viruses and other pathogens, has
received little study with regard to disinfection. Based on a few early studies
suggesting that the cysts were extremely resistant to disinfectants and
because of its similarity to another protozoon, Entamoeba histolytica, G.
lamblia is currently regarded as highly resistant to chlorine.
The purposes of this paper are to review the earlier literature, discuss the
present state of our knowledge regarding the disinfection resistance of G.
lamblia and E. histolytica cysts and outline an approach for generating
disinfection information on G. lamblia that will be useful in assessing the
adequacy of disinfection as a water treatment unit process for preventing
waterborne outbreaks of giardiasis.
-------
232 WATERBORNE GIARDIASIS/TREATMENT TECHNOLOGY
G. LAMBLIA DISINFECTION — SOURCES OF
CURRENT PERCEPTIONS
As indicated above, the prevailing concept conveyed in the current
literature relating to the G. lamblia cysts is that they are extremely resistant
to chlorine. A statement to the effect that "Giardia cysts resist 0.5%
chlorinated water for 2 to 3 days" appears in reference texts (1,2) and in at
least 2 recent review papers (3,4).
To those familiar with and involved in drinking water disinfection, a
chlorine concentration of 0.5% translates into an available chlorine level of
well over 1000 mg/1. Chlorine levels used in water treatment are usually on
the order of 2 to 3 orders of magnitude lower than this with contact times of
30 min to a few hours. Thus, the concept conveyed is that the cysts are
extremely resistant to disinfection and impossible to control by disinfectant
concentrations and contact times used in water treatment. Unfortunately,
the source of this statement was not cited in any of the reports.
A search of the early literature on protozoan cyst inactivation indicates
that the source of the statement may have been a paper published by Mills et
al (5). As part of a study of contamination of fruits and vegetables by
microorganisms these authors examined the resistance of a number of
protozoan cysts, including Giardia intestinalis (lamblia) to chlorinated
water, alcohol, and boiling. The information conveyed by their results is
that G. intestinalis and E. histolytica cysts WERE killed by exposure to
0.5% chlorinated water for 2 to 3 days. However, Liu (6) cited the earlier
work as follows: "According to Mills, Bartlett and Kessel( 1925) 0.5 percent
chlorinated water will kill most kinds of bacteria but not the protozoan cysts
in 2-3 days." It may have been this unfortunate misstatement repeated in
subsequent publications that resulted in the conveyance of the concept of
extreme resistance of G. lamblia cysts to chlorine.
Viability Based on Vital Staining
Although Penfold et al (7) apparently had succeeded in demonstrating E.
histolytica viability by microscopically observing excystment in a nutrient
medium, all of the early studies on the effects of environmental conditions
on survival of protozoan cysts were carried out using vital staining as an
indicator of cyst viability. In this test, eosin was usually used for staining.
Cysts which remained colorless when exposed to eosin were considered
viable while those penetrated and stained by the dye were considered dead.
According to Boeck (8), the basis for the use of this type of test as a measure
of protozoan cyst viability was the lack of other reliable measures of
viability such as in vitro excystment or animal infectivity. However, it is
evident that the validity of staining as a measure of viability was questioned
by many early workers. Both Boeck (8) and Root (9) quoted Wenyon and
O'Connor who had also used this test in studies published in 1917 as follows:
"It seems, therefore, clear that the eosm-staming cells are dead though it may be
argued that others which do not stain may be dead also or, at any late, nonmfectwe
Still, if we accept the eosin test as a criterion and regard all unstained cysts as living,
the error in judgement will be on the safe side."
-------
J. C. Hoff 233
In a more recent publication on the effects of disinfectants on E.
histolytica and G. lamblia cysts, Cerva (10) used staining with eosin for
measuring cyst viability after exposure to disinfectants. He reported that a 1
to 3% solution of chlorine or chloramine killed (as determined by vital
staining) only 10% of G. lamblia in 30 min to 24 hours. A 10% solution
killed 30 to 40% in 1 h in 6 of 10 experiments. Similar results, showing
extremely high resistance to chlorine, were obtained in comparative studies
on E. histolytica. However, E. histolytica disinfection studies conducted by
Chang and Fair (11) using culture technics to determine viability indicated
that these cysts were inactivated by much lower levels of chlorine, thus
demonstrating the lack of correlation between vital staining and culture for
determining cyst viability. Although data were not shown, Chang and Fair
reported that staining of cysts with eosin was grossly unreliable as a criterion
of cyst death as determined by cyst culture. In a later publication, Chang
(12) stated:
"I have observed from time to time that cysts were tound to be nonviable by the
culture method but 40 to 70 percent failed to take the stain (EOSIN). Also, an
examination of cysts stored in a cold room 42 to 50° K for four to five months after the
period of viability showed that 30 to 40 percent failed to be stained It is felt that while
the nuclei of cysts are no longer viable, the cystic wall may resist the infusion of
dyestuff into the cyst for a long time."
In their work on the effects of chlorine on E. histolytica cyst viability
Brady et al (13) also reported that in preliminary experiments:
"The degree of correlation between the indicated survival of cysts by the eosm-
staming technic and the actual survival by ability to establish cultures was not close
enough to permit the use of the former as a criterion "
The report of Bingham et al (14) has confirmed and quantified the
differences between measurement of G. lamblia cyst viability by eosin
staining and measurement by the in vitro excystation technic they have
developed (15). Their results indicate that eosin exclusion consistently
indicates higher cyst viability than can be shown by in vitro excystation.
If the above detective work is correct in tracing the source of current
information indicating that G. lamblia cysts can resist 0.5% chlorinated
water for 2 to 3 days, the concern generated by such statements should be
dispelled since the results on which it is based were both misstated and were
obtained using viability measurements that have been shown to provide
erroneous results.
E. histolytica Resistance to Disinfectants
As indicated previously, a second source of our current impression that G.
lamblia is highly resistant to disinfection stems from its similarity to E.
histolytica. Both are proto/oons, existing in two stages. One stage is a cyst
form in which the organism exists in a nonmotile state inside a more or less
impervious protective cyst wall which enables its survival outside the host in
unfavorable environmental conditions. In the other stage the organisms are
actually motile and extremely fragile, an amoeba in the case of E. histolytica
and a flagellate in the case of G. lamblia.
Development of culture media for the cultivation of E. hisiolvtica (16)
and of a cultural encystment method (17) coupled with increased interest in
-------
234 WATERBORNE GIARDIASIS/TREATMENT TECHNOLOGY
amebiasis, generated by several waterborne outbreaks in the U.S. and
military needs associated with stationing troops in tropical countries,
stimulated extensive disinfection studies on the cysts of this protozoon.
These studies employed a variety of disinfecting agents (chlorine,
chloramines, iodine, bromine, ozone) under a variety of conditions and
resulted in the formation of a large body of disinfection information on E.
histolytica.
In 1956, the National Research Council (18) prepared recommended
chlorine and chloramine bactericidal and cysticidal residuals for drinking
water treatment by the armed forces under field conditions. The
recommended levels were based on a 30 min contact period and provided for
different residual levels at various water pH's and temperatures. The
cysticidal recommendations were based on studies by Chang and Fair (11)
and Chang (12), as well as results of other unpublished studies most of which
had been conducted by Chang and others at the Department of Sanitary
Engineering and School of Public Health at Harvard University. Table 1
was prepared by Palin (19) based on the National Research Council Report
(18). The results indicate that levels of free chlorine needed to inactivate E.
histolytica cysts at temperatures of 22 to 25°C at low pH are reasonably
attainable in water treatment practice. However, at higher pH's at 22 to
25° C and at any pH at temperatures of 2 to 5°C, relatively high levels of
chlorine that would be considered impractical are necessary. Chloramines,
which are much weaker biocides than free chlorine, were considered
impractical for cyst control. These results are consistent with many other
bacterial and viral disinfection studies, reflecting the greater disinfecting
efficency of chlorine and other oxidants at higher temperatures and also the
influence of pH on disinfection rates through its influence on the chemical
disinfectant species present.
Our present concept of the disinfection resistance of E. histolytica cysts
stems largely from the research on which the National Research Council
report was based. Thus, White (20) stated:
"The best protection against this disease (amoebic dysentery) is removal of the
organisms from water supplies by filtration. They are large enough to be trapped by
sand filters, and require high doses of free residual chlorine (10 ppm) and long contact
time (one hour or more) to effect disinfection."
Table I. USPHS minimum cysticidal and bactericidal residuals'
(After 30 minutes contact)
Free Chlorine. HOCI + OCI"
pH
Value
6 0
7 0
8.0
9 0
Bactericidal
0-25°C
0 2
0.2
0.2
0.6
Cysticidal
22-25°C
2.0
2.5
5.0
20.0
Cysticidal
2-5°C
7.5
100
20.0
70.0
Combined
Chlorine. NH2CI
Bactericidal
0-25°C
2.0
2 5
3.0
3.5
1 Concentration required for inactivation of 99.999% of cysts. Adapted from (19).
-------
J. C. Hoff 235
Subsequently, Stringer (21) developed a method for determining E.
histolytica cyst viability based on microscopic observation of full and empty
cyst walls following incubation of cysts in excystation medium. He pointed
out that this viability assay offered a simple and reliable method for
determining the percentage of living cysts in a cyst population and provided
more quantitative results than the methods based on cyst growth in liquid
medium that had been used in previous E. histolytica disinfection studies.
Using this method, Kruse (22) and Stringer and Kruse (23,24) reexamined
the effects of chlorine, iodine, and bromine on amoebic cysts. They used a
standard exposure period of 10 min at 30°C over a range of pH 5 to 10, with
varying concentrations of each disinfectant under both clean and extraneous
disinfectant demand conditions. It is difficult to compare their results with
those in Table 1 based on the National Research Council's analysis because
of the different contact periods (10 min vs 30 min) and the endpoints chosen
(50 to 99.9% inactivation vs 99.999%). The 30° C temperature used by
Stringer and Kruse (23,24) was also considerably higher than in the previous
studies. However, Kruse (22) cited the National Research Council
recommendations for cyst chlorination at pH 7.0 as in agreement with the
results of his studies at this pH. Although the differences in experimental
conditions used make it difficult to correlate the results, they appear to be in
fairly good agreement and convey the concept that E. histolytica cysts are
among the most chlorine resistant waterborne pathogens known.
Depending on water plant operating conditions with regard to water
temperature, pH, and disinfectant contact time, it appears that E. histolytica
cyst inactivation would be practicable under some conditions but not under
others.
Thus, if E. histolytica cysts are a good model for indicating the
disinfection resistance of G. lamblia cysts, the concept that disinfection
cannot be relied on for preventing waterborne transmission of giardiasis is
well founded. Unfortunately, their validity as a model is not known and it
would appear that the only way the model can be validated is to conduct
disinfection research on G. lamblia.
CURRENT G. LAMBLIA DISINFECTION RESEARCH
The development of an in vitro excystation method for G. lamblia andits
successful application in the viability studies presented at this symposium
(14) indicate that disinfection studies on G. lamblia cysts are feasible.
In our laboratory, G. muris is being used as a G. lamblia model and
chlorine inactivation studies on this species have been initiated using the in
vitro excystation method. The cyst suspending medium used has been 0.05
M phosphate buffer or low turbidity, clean, natural water. The natural
water was filtered" through an 0.22 m porosity filter to remove particulates.
Both the buffer and natural water were rendered chlorine demand-free by
addition of excess chlorine, storage for 24 hours and destruction of the
remaining chlorine by exposure to ultraviolet light. This process satisfied
any chlorine demand caused by impurities in the water and thus aided in
maintaining a constant chlorine residual during the experiment.
-------
236 WATERBORNE GIARDIASIS/TREATMENT TECHNOLOGY
Prior to each experiment, chlorine was added to a previously refrigerated
suspending medium to the desired level (1 mg/1), the pH was adjusted, and -1
liter was placed in each of four 1 liter beakers. These were placed in a 5°C
circulating water bath. The contents of each beaker were continuously
stirred by overhead stirrers. The chlorine in the cyst control beaker was
neutralized by the addition of 1 ml of 10% sodium thiosulfate. At time zero
purified cysts calculated to provide a concentration of 100 to 200 cysts/ml
were added to each of the stirred beakers. At appropriate time intervals,
sodium thiosulfate was added to each beaker, the beaker removed from the
water bath and the contents filtered through 5 ' m porosity 47 mm
polycarbonate filters to separate the cysts from the suspending medium. The
cysts were washed from the filter surface into a 15 ml conical centrifuge tube
using 1 ml of water. The cysts were exposed to pH 2HClfor30min to trigger
excystation, reconcentrated by centrifugation and incubated in excystation
medium for 1 h at 37° C in sealed glass coverslip preparations and examined
for full and empty cysts and the presence of trophozoites as described by
Bingham and Meyer (14).
The results of these studies are shown in Table 2. Although the data are
preliminary, they do indicate that this approach can be used to provide
disinfection data. The differences in G. muris percent excystation after
exposure to 1 mg/1 free residual chlorine at pH 7.6 and 9.0 are consistent
with the well known decrease in disinfection efficiency of free chlorine as pH
increases because of the predominance of hypochlorite ion (OCW), a weak
disinfectant chemical species, at higher pH's. The results indicate that the
suspending medium used in cyst disinfection studies may be an important
factor. In bacterial and viral disinfection studies 0.05 M phosphate buffer is
conventionally used to insure pH stability. In one of the preliminary
experiments, plasmolysis, as indicated by shrinkage of the cytoplasm from
the cyst wall, occurred when the cysts were reconcentrated from 0.05 M
phosphate buffer and placed in the excystation medium. The low
excystation levels in nonchlorinated control cysts suspended in phosphate
buffer also indicated that this may be a problem. Higher cyst control
excystation was obtained when a natural water was used.
Table 2. Preliminary experiments on Giardia cvst mactivaliun bv chlorine
(5°C, 1 mg/1 free residual chlorine)
Percent Excystation
Cyst
Species
G muris
G lamblia
G mur/s
G muris
pH
6 0
7 0
7 6
9 0
Suspending
Medium
05M PCU
05M P04
water
water
Positive Chlorine
Control
0
2 5ab
22
21
10
0
Oa
ob
20
Exposure Time (Mm)
20
0
oa
0
6
30
0
oa
0
7
* Cytoplasm shrinkage observed
b Incomplete excystation? Blebs observed
-------
J. C. Hoff 237
The results of 1 preliminary experiment designed to determine the effect
of suspending media on excystation are shown in Table 3. The results bear
out the initial observations made in the disinfection experiments and
indicate that the suspending medium used in Giardia disinfection studies
can affect the results obtained.
Table 3. Preliminary observation* of effects of suspending medium
on G muns cvst viabihtv*
Suspending
Medium
0 5M P04
0 5M P04
0 5M P04
Water"
Water"
Water"
pH
6 0
7 0
8 0
6 0
7 0
8 0
Percent Cytoplasm
Excystation Shrinkage
7 +
4 +
6 +
9
15
20
Trophozoite
Activity
±
-
-
±
+ +
+ +
" Stored 1 hr , 5°C w stirring at indicated pH
b Clean natural water, filtered (0 22 Mm)
It is also possible that the cyst purification procedure involving
concentrating the cysts on 0.8 M sucrose by centrifugal force could
osmotically damage the cysts. The questions of optimum time in the
infectious cycle for collection of cysts from the infected animals, optimum
time for separation of the cysts from the fecal material, and whether or not
cysts undergo a maturation process need to be answered. In order to provide
reliable disinfection information, cysts in their optimum state of resistance
and infectivity should be used in disinfectant exposure studies. In this regard
it is interesting to note that Kruse1 (22) reported that E. histolvtica cysts
produced in vitro using an encystment medium were less resistant to
inactivation by chlorine than cysts produced in monkeys. Thus, while it
would appear that development of a G. lamblia encystment medium would
be advantageous in providing a reliable continuing supply of cysts for
experimental work, the results of disinfection studies using culturally
produced cysts might not accurately indicate the resistance of naturally
produced cysts.
In addition to the in vitro disinfection studies on G. muris, limited
disinfection studies using mouse infectivity assays for measuring viability
also are being conducted by Dr. Stevens group at Case Western Reserve
University. The results of these studies should make it possible to relate the
validity of in vitro results to actual host infectivity. The difficulties .in
obtaining a continuing and consistent supply of G. lamblia cysts make the
use of readily obtainable cysts such as those of G. muris especially attractive.
It is expected that, through the above efforts, disinfection information will
become available for use in devising water treatment processes that will
effectively prevent future waterborne giardiasis outbreaks.
-------
238 WATERBORNK GIARDIASIS/TREATMENT TECHNOLOGY
SUMMARY
Analysis of the results of early studies on Giardia cyst inactivation by
disinfectants indicates that the results of these studies should be disregarded
since the method used for viability determination (dye exclusion) has been
shown to give inaccurate results. The recently developed in vitro excystation
method appears promising for disinfection research on Giardia cysts and
research designed to provide disinfection information on this protozoon is
in progress.
ACKNOWLEDGMENT
The excellent technical assistance of Mr. Eugene Rice and Ms. Judith
Barnick in tracing and locating references in the early literature, preparing
and purifying cyst suspensions, and conducting the disinfection experiments
is gratefully acknowledged.
REFERENCES
1. Craig, C. F. and E. C. Faust. 1945. Clinical parasitology. Lea and Febiger, Philadelphia.
Pa
2. Stitt, E. R. 1945. Diagnosis, prevention, and treatment of tropical diseases. R. P. Strong
(ed.), Blakiston Co Philadelphia, Pa. 7th ed.
3. Wolfe, M. S. 1975 Giardiasis J.A M.A. 233:1362-1365
4. Burke, J. A. 1977. The clinical and laboratory diagnosis of giardiasis. In CRC Critical
Reviews in Clm Lab. Sci. 7-373-391
5 Mills, R. 0., C. L. Bartlett and J. F. Kessel. 1925 The penetration of fruits and vegetables
by bacteria and other particulate matter and the resistance of bacteria, protozoan cysts and
helminth ova to common disinfection methods. Am J. Hyg. 5:559-579,
6 Liu, K. 1928. The comparative lethal effects of certain chemicals on bacteria and cysts of
Entamoeba histolytica from human feces. China Med. J. 42-568-574.
7 Penfold, W. J., H. M. Woodcock and A. H. Drew. 1916 The excystation of K hislolytica
as an indication of the viability of cysts. Br. Med. J. 1.714-719.
8. Boeck, W. C. 1921. The thermal-death point of the human intestinal protozoan cysts. Am.
J. Hyg 1 365-387
9. Root, F. M. 1921. Experiments on the carriage of intestinal proto/oa of man by flies. Am.
J. Hyg. 1-131-153
10. Cerva, L. 1955 The effects of disinfectants on the cysts of Giardia mtestmahs Ceskosl
Parasit 2.17-21
11. Chang, S. L. and G. M. Fair. 1941. Viability and destruction of the cysts of Entamoeba
histolytica J Am. Water Works Assn 33 1706-1715.
12. Chang, S. L. 1944. Studies on Enlamoeba histolytica. Ill Destruction of cysts of
Entamoeba histolytica by a hypochlorite solution, chlorammes in tap water and gaseous
chlorine in tap water of varying degrees of pollution. War Med. 5:46-55.
13. Brady, F. J., M. F. Jones and W. L. Newton. 1943. Effects of chlonnation of water on
viability of cysts of Entamoeba histolytica War Med. 3-409-419.
14. Bingham, A. K., E. L. Jarroll, Jr. and E.A.Meyer. 1979. Induction of Giardia excystation
and the effect of temperature on cyst viability as compared by eosm-exclusionand in vitro
excystation. In W. Jakubowski and J. C. Hoff (ed.), Proc. Symposium on Waterborne
Transmission of Giardiasis, Cincinnati, Ohio, Sept. 18-20, 1978.
15 Bingham, A. K. and E. A. Meyer. 1979. Giardia excystation can be induced in vitro in
acidic conditions. Nature 277-301-302.
16. Boeck, W., C. and J. Drbohlav. 1925. The cultivation of Entamoeba histolvtica. Am. J.
Hyg. 5:371-401.
17. Stone, W. S. 1937. A method of producing encystment in cultures of Endamoeba
histolvtica. Am. J. Trop. Med. 17:539-551
18. Snow, W. B. 1956. Recommended chlorine residuals for military water supplies. J. Am.
Water Works Assn 48:1510-1514
19 Palin, A. T. 1975. Water disinfection-chemical aspects and analytical control. In J. D.
Johnson (ed.), Disinfection Water and Wastewater, Ann Arbor Sci. Ann Arbor, Mich.
20. White, G. C. 1972. Handbook of chlorination. Van Nostrand Reinhold Co., New York.
-------
J, C. Hoff 239
21 Stringer, R. P. 1972 New bioassay system for evaluating per cent survival of Entamoeba
hutolvtica cysts. J. Parasitol. 58 306-310.
22 Kruse1 C. W. 1971. Etiology of £ histolvtica transmission. Final Technical Report, U. S.
Army Medical Research and Development Command. Contract No DADA 17070-C-
0025. 52 p. Dec 1971.
23. Stringer, R. P. and Kruse1, C. W. 1970. Amoebic cysticidal properties of halogens in water
In Proc. Natl. Specialty Conf. on Disinfection, July 8-10, I970, Univ of Massachusetts,
Amherst Am. Soc Civil Eng.
24. Stringer, R. P. and Kruse1, C. W. 1975. Comparison of bromine, chlorine, and iodine as
disinfectants for amoebic cysts. In J D. Johnson (ed.). Disinfection Water and
Wastewater, Ann Arbor Sci., Ann Arbor, Mich
Discussion
L. LEONG: Are you aware of any research into the effects of ozone on
Giardia cysts?
J. HOFF: There has been some work on the effects of ozone on
Entamoeba histolytica. I am not aware of any that has been done on Giardia
lamblia.
M. WOLFE: It may be appropriate to recall that in the studies done
during World War II on field disinfection of small bodies of water, it was
conclusively shown that the amount of chlorine released by Halazone
tablets was inadequate to kill amebic cysts. Then the Globaline iodine tablet
was produced which did kill amebic cysts. Globaline is not readily available
on the commercial market, but there is something called Potable-Aqua
tablets (Badger Pharmacal, tetraglycine hydroperiodide) that contain
iodine. We are ignorant of what the effects of iodine tablets would be on
Giardia at this point until your studies are carried further. I think all we can
do is analogize to the situation that occurs with Entamoeba histolytica.
My current recommendation for travelers and to people hiking in the
Rocky Mountains, who may want to drink stream or other untreated water,
is to use iodine water purification tablets or tincture of iodine solution rather
than to rely on Halazone tablets. Some doctors and others are making
recommendations to travelers to use chlorine tablets, but I think it is
important to emphasize that iodine appears to be superior at this point.
Hopefully, the same findings with Giardia as have been shown for
Entamoeba histolytica will be confirmed. At this time I think it is wise to use
iodine for field disinfection.
J. HOFF: Assuming the similarity of G. lamblia to E. histolytica, I think
that is probably a good recommendation.
-------
240
Water Filtration Techniques for Removal
of Cysts and Cyst Models
Gary S. Logsdon, James M. Symons
and Robert L. Hoye
Physical and Chemical Contaminant
Removal Branch, Drinking Water
Research Division, Municipal Environmental
Research Laboratory, U. S. Environmental
Protection Agency, Cincinnati, Ohio
ABSTRACT
Because outbreaks of waterborne giariliasis have been documented only in recent years,
information on the removal of Giardia cysts by water filtration is limited to filtration
failures associated with outbreaks.
Filtration for removal of Entamoeba hislo/ytica cysts was studied in the 1930's and
1940's. Granular media filtration was effective for E. histolylica cyst removal if the water
was properly preconditioned Minimum preconditioning required was effective
coagulation. Sedimentation and diatomaceous earth filtration also were effective for £
histolytica cyst removal.
Recent experiments, although limited, showed that Giardia lamblia cysts were removed
by diatomaceous earth filtration. Preliminary laboratory results also suggest that
coagulation and granular media filtration will remove Giardia cysts.
Both previous work with granular media filtration and present work using 9 m
radioactive microspheres as cyst models show that good plant operating procedures are
necessary if water filtration plants are to be an effective barrier against the passage of cysts
into drinking water. Water must be properly coagulated before granular media filtration.
Diatomaceous earth filters should be precoated with 91 g (0.2 Ib) of diatomite/0.1 m2(ft2) of
filter surface, and body feed must be added to maintain the integrity of the filter cake
Conscientious, high quality filter plant operation is a key factor in providing public health
protection against waterborne giardiasis.
Outbreaks of waterborne giardiasis have been occurring more frequently
in recent years. This trend is a cause for concern among public health,
environmental, and water utility officials. This paper will discuss some
possible engineering solutions.
THE PROBLEM
Kirner et al in a discussion of a giardiasis outbreak at Camas,
Washington, stated "Chlorination in the manner being practiced by the city
was either ineffective in disinfecting the raw-water supply, where the
infective organism was concerned, or the failure of disinfection equipment
allowed enough organisms to survive to cause a problem" (1). They also
stated that turbidity and coliform count alone were not adequate
parameters for assessing the biological quality of filtered water.
For situations in which monitoring methods are not adequate to insure
production of potable water, the use of certain water treatment techniques
-------
G. Logsdon 241
may be appropriate. This concept is included in the Safe Drinking Water
Act. The efficacy of chlorination at Camas was uncertain, and other cysts,
such as E. histolytica are known to be much more resistant to chlorination
than are coliforms (2). Therefore, removal of Giardia cysts by filtration
should be a practical approach to providing safe drinking water where the
threat of Giardia contamination of water sources exists. This paper will
review literature on water clarification for cyst removal and present recently
obtained data on the subject.
PRINCIPLES OF FILTRATION
Water filtration has been practiced for many years. The concept of
conventional water treatment in which coagulation, flocculation and
sedimentation are used to condition water for filtration, was studied
carefully by G. W. Fuller at Louisville, Kentucky in the 1890's. The final
clarification step in conventional water treatment is filtration through
granular media.
Since World War II, design engineers have tended to use multi-media
(dual media or mixed media) filters rather than the sand filters studied by
Fuller. By carefully selecting media sizes and densities, engineers can design
filter beds with the largest type of media at the top of the filters, where the
pretreated water enters the porous bed. This facilitates storage of
particulates and results in longer filter runs, or ability to operate filters at
higher rates, or both.
Numerous water sources in the United States are low in suspended solids.
When the color of such waters is low, the coagulant chemical dose needed to
prepare the water for effective filtration is low. In such circumstances,
sedimentation may be very ineffective for removal of paniculate matter even
though the filtration process is very efficient. Beard and Tanaka (3)
demonstrated this by the use of particle counting. If the particulate content
of a coagulated water is low, filtration without sedimentation (direct
filtration) may be used to remove particulates from the water. Proper
coagulation is just as essential in direct filtration as it is in conventional
treatment.
Some low-particulate waters are treated by diatomaceous earth (DE)
filtration. The sizes of the DE particles used as filter medium vary, but the
range can be from several micrometers to forty or fifty micrometers.
Mechanical straining is the mechanism of removal for particles in the size
range of E. histolytica and Giardia cysts. Coagulation is generally not
employed in DE filtration.
LITERATURE REVIEW
A review of literature revealed no research on water filtration for Giardia
cyst removal. This is not surprising because the Giardia organism was not
implicated as a cause of waterborne disease until recently. The hazards of
waterborne amoebiasis have been known for decades, so both before and
during World War II water filtration studies for removal of E. histolvtica
cysts were carried out.
-------
242 WATERBORNE GIARDIASIS/TREATMENT TECHNOLOGY
Granular Media Filtration
Granular media filtration was in use in numerous water plants in the
1930's when an amoebic dysentery outbreak occurred in a Chicago hotel, so
investigations of the capabilities of this process for cyst removal were
performed.
Chicago Experimental Filtration Plant Research
After the Chicago outbreak occurred, experiments for removal of E.
histolytica cysts from water were carried out at the Experimental Filtration
Plant operated by the Chicago Department of Public Works. In preliminary
laboratory experiments, 4.1 cm (1.5 in) diameter filter columns were used.
Tests with these small columns showed complete removal of E. histolytica
cysts from water by coagulation and rapid sand filtration (4).
Further testing with a pilot plant filter having a surface area of 0.93m2 (10
sq ft) took place in 1935 and 1936 (5). In this research, cysts were added to
coagulated water that was then treated by direct filtration at 7.7 1pm/0.1m2
(2 gpm/ft2). During the filter runs, which ranged in length from 4 to 14
hours, samples were taken at regular intervals. The samples were
concentrated in a Foerst centrifuge, and the sediment was examined in a
Rafter counting cell. Five tests for recovery of cysts by the centrifuge
technique gave ratios of recovered/original ranging from 0.71 to 1.18.
Analysis of the filtered water showed the test filter was effective. A total of
433 1 of filtered water was centrifuged and the sediment was examined for
cysts. On the basis of the total volume of water filtered and the total number
of cysts applied, 178,000 cysts would have been dosed in 433 1. Only 4 cysts
were found in the filter effluent. This indicates a removal efficiency
exceeding 99.99%. Baylis et al (4) concluded that filtration of waterthrough
rapid sand filters, in the manner then extensively used in the United States (2
gpm/ft2), was an effective means of removing Entamoeba histolytica cysts
from water.
War Department Research, 1943-1944
The need to provide water free of E. histolytica cysts in World War 11
resulted in a cooperative filtration research program involving The Engineer
Board, U.S. Army, Fort Belvoir, Virginia, and the National Institute of
Health, U. S. Public Health Service. A report of this work was prepared by
the War Department (6), but the report was not widely circulated. This
paper will present a summary of that report, subsequently referred to as the
Army report.
An apparent contradiction develops when Baylis' conclusion that rapid
sand filtration was successful is compared to the Army report conclusion
that the complete removal of E. histolytica cysts was not accomplished when
the U. S. Army Portable Purification Unit sand filter was operated at flow
rates practical for field use. H. H. Black, Chief of the Water Supply
Equipment Branch that prepared the Army report, briefly discussed sand
filters in a paper on diatomite filtration, stating, "The need for better
filtration equipment arose from the failure of sand filters to remove cysts of
Entamoeba histolytica at the high rates of filtration employed in the field"
-------
G. Logsdon 243
(7). How Baylis and Black reached different conclusions about rapid sand
filtration can be understood if the objectives of treatment and process
operating conditions are considered for each case.
Operating conditions, and possibly water quality goals, were less
demanding at Chicago. The filter there was operated at 7.7 lpm/0.1m2
(2gpm/ft2) rather than 24.6 or 36.6 lpm/0.1m2 (6.4 or 9.5 gpm/ft2) as was
done in the Army experiments. Cysts in raw water at Chicago ranged from
362 to 2370 per gal. Cyst concentrations for Army tests involving
coagulation and filtration ranged from 2300 to 7000 per gal. In some of the
Army tests cysts were dosed on a batch basis with 1 to 3 million cysts being
added in 20 to 30 seconds. Whereas 4 cysts in 433 liters was acceptable filter
performance to Baylis, Black apparently sought total removal of the cysts
from the treated water. The Army report data on sand filtration are
summarized in Tables 1 and 2. In Baylis' filtration studies the 433 1 that
contained 4 cysts in the filter effluent would have had 178,000 in the influent
(99.9978 percent reduction). This gives 22 cysts passing the filter/million
applied, a much better result than 250/million applied, the best result
obtained with the Army Portable Water Purification Unit.
Although the Chicago experiment filter performed better than the Army
portable unit, for the circumstances relevant to Baylis' evaluation of the
sand filter, both his data and the best Army data show substantial
reductions of cysts. A key factor is the expected concentration of cysts in the
raw water. Cyst concentrations in a grossly contaminated small jungle
stream or pond, in an area where amoebiasis was endemic, would probably
be much greater than cyst concentrations in a lake or river in the USA.
Army equipment had to be able to purify any water that might be found in
.field conditions, whereas water treatment plants built in the USA are
supposed to be located so that grossly contaminated water does not reach
the plant intake. As a result of the differences in water quality standards
domestic water filtration plants would not have to have the performance
capabilities of the Army portable units. Whether or not sand filtration was
deemed effective must have been in some measure a reflection of expected
raw water quality.
Tables 1 and 2 show the influence of good operating practice on cyst
removal. Tests 1 and VI (A and B) involved filtration of uncoagulated water.
Test II was performed with poorly coagulated water. For these tests, the
numbers of cysts passing the filter/million applied were much higher than
for other tests with good coagulation and comparable experimental
conditions.
Diatomite Filtration
In the United States, the Army took the lead in developing diatomaceous
earth filtration for drinking water with a research and development program
during World War II. This work is described in the Army report (6). A
discussion of DE filtration principles and a summary of bacteriological data
developed in the project was also published by Black and Spaulding (7).
However, this paper, published in 1944, merely mentioned a need to remove
-------
244 WATERBORNE GIARDIASIS/TREATMENT TECHNOLOGY
-^
-a
^
1
'5
3
c
.0
a
1
3
a.
fe
5
S
a
§
Q,
i'
X
"^
s
g
S
C
£X
*u
1
7
"^
!-
V ^:
D
0 TD
V £
uT "Q
o
- ^
ui O
CD -
Q- —
O CL
pl
i w —
2! co S=
^LU
_o cj _
^ 15 o
03 ^- ^
-§ Q- S
2^^
°°lD
CD ^r (n
CJ u
2
CD
CC ^
Q ~i
CD CL
iE
c
'o 2
> ro
ro en
^ CD
O 0
O
c
CD O
oo ro
CO O
co co
0 0
0 0
^r LO
*— r^"
CD
O LO
t- CD
I—
O CO
O oo
O 00
LO LO
CO 00
CD CD
"O CD
o c
0 0
m
> -L
-------
G. Logsdon
245
-^
1
o
ion unu
'men/an
5 "a
(i %
C'~3
ft. §
i^. c
o §
£l
a s
0 SO
ft. C
14
~T v£.
-C 5,
•~ (3
s "^
c ^
Is
qj £•
•"--, °
o' c;
.3 fc
3 ^
<£
'N
-Si
-c
Ol §
ro _
c: o
o -Q
£,£
O) "^
in Q) a.
w Q. CL
^ | c
%vc
"c Q. • LU
o U -
— ro
CD ^ (H
O U
1_
^ o3 'c
1^1
z ^
o °
C rv
- - o
CD CO
O O
0 0
CO LO
co" r^"
CO CD
CO LO
CO CD
CO .-
O CO
O CO
O CO
o
LO
ro
CD
0
CT>
LO
r-
LO CD <— r^- co co <•
CO CO ^ CD LO ^ CD
CO CD CM CD CO CD CD
CD CD CO CD CD CD CD
O O O O C
O O O LO (
LO ^ O CN f
CD •<-" CD
T— r--
\
CO CO
3 ^r CD D
D O CD
D O CM
~-~ r-~ LO
sF LO
— co
CO
^
CO CO CO
CD LO O CD CD CM CD
QC CC LO CC
T3 T3 "D
CU CD
C C
.5 .5
E E
O *fr O
CD
C
_5
E
o
^ O LO ^ ._. CD ^
O X L°
00 §
00 CO'
LO
\ ^
CD
LO
co
CD
O LO
CM \ r~~
'
^i
CO
> :
>;>
-------
246 WATERBORNE GIARDIASIS/TREATMENT TECHNOLOGY
cysts of E. histolytica without giving actual test results. Experimental data
can be found in the Army report.
Experiments on diatomaceous earth filtration for cyst removal were
carried out with a variety of DE filter systems, ranging from devices to
provide batch filtration of as little as 3.8 1 (1 gal) of water to equipment
capable of filtering approximately 771pm (20 gpm). Data from various parts
of the Army report have been combined and summarized in Table 3. This
table includes data calculated by the present writers through the use of other
original data.
Newly calculated data include all of the data for the ratios of cysts
recovered from filtered effluent to cysts in the same volume of filter influent,
as well as some data on filter aid usage and flow rates. Some data were not
available in the Army report so no values are given in certain instances in
Table 3.
Data in Table 3 do support the conclusion the DE filtration could remove
essentially all cysts. The largest number of cysts recovered, 7, resulted from
using contaminated water (225,000 cysts in 35 gal) to precoat the filter.
Clean, filtered water should always be used for precoating DE filters.
CURRENT RESEARCH
Past research on cyst removal was related to treatment for E. histolytica
cysts. Research results demonstrated that the cysts could be removed and
these results suggested that filtration also should be effective for reducing
the concentration of Giardia cysts in water. To demonstrate this specifically,
a program of filtration research with Giardia cysts and cyst models is being
conducted.
Diatomaceous Earth Filtration Research
The Drinking Water Research Division's pilot plant facilities include a
DE filter, described fully in a Johns-Manville technical bulletin (8). This
O.lm2(l ft2) filter was used to perform treatment studies involving Giardia
lamblia cysts and radioactive bead cyst models. The test results, presented
more fully elsewhere (9,10), are summarized briefly in this text.
Experimental Methods
The DE test filter was set up for operation as a pressure filter with a slurry
feeder for body feed. The 0. Im2( 1 ft2) filter was operated at 3.81pm (1 gpm).
The filter system is shown in Fig. 1. The operation of the DE filter consisted
of three steps: precoating, filtration, and backwashing.
A diatomaceous earth slurry was prepared by adding the desired weight of
DE to tap water in the precoat tank and mixing the contents of the tank
during precoating. The weight of diatomite added to the precoat tank was
used to calculate precoat thickness expressed as pounds per square foot. The
slurry was recirculated until the water in the precoat tank appeared clear and
free of diatomite.
During the filtration process, a small amount of diatomaceous earth was
continuously added to the raw water in order to maintain good hydraulic
characteristics of the precoat filter cake. Without the body feed, the surface
-------
6". Logsdon
247
g
<±l
l^~
J.
a
3
o
S
diatom
-C
a
S
|
T
03
i>
5
g
^
CJC
.c
"3
i
1
•z
C'
c
C3
">
C
O
c
•s
o
be
.C
5
i
^
1
S
3
?
S
'"*'
\
"S "o
i— (D
CU —
g£
o CD
"o ro S
> <
CD *O
E £~
-2" §,
o — —
> Ll_
i
^~
0) CL
C O)
cn
CD
ro
a:
5
o c
- 1
O)
>--D
TJ a>
O Q)
CO LJ-
ro 7^
R x
CJ X
£5
«
CD
(-
0 O
0 O 0 0 0
o o o o o
o °. o 9 °.
LO CO •— O~ LO
00 OO <— CD <—
\ \ \ \ \
O OO i- O O
O CO •- O O
o o o o o o
000000
000000
o" CD" o" CD" o" CM"
LO T- ,- T-
co
LO
00
<- CD «- O --
'- <-
LO
co
,_ LO rv rv ,_
CM
O ^
00 ^ ^ CN ^
d>
CN
O
69" ""
T— ^
O <-
O ^ O o r-
CN O O LO CD
'. — . — . fY-i
1 T T T L'J
O O O O O
Q Q Q Q O
'— CN <3" LO CD
000
o o •-
LO CD CD
CD <-" »-" " " " "
r~- CM T-
r~- i— o
r- ^ 0 -- O •- 0
O O LO
O O oo
LO CD CO
•^f co" >-"""""
r^- CD CD
<—
CD LO CO ^ ^ ^ ^
en oo ^t
CD •tf CD
CM CD CD r- co O r^
*- CM 00 CM T— CM CN
CD CD CN CD | — . CO ^"
CM 00 •* CO CM OO "3-
>— r^ -vj- -st- CN oo oo
CO -vT 00 ^ CO LO 0)
CM CM CM CM •— CM CM
O O O O O O O
(^ CM CM
r — T— * —
r-- LO LO LO LO LO LO
o o o o o o o
Q Q Q Q Q Q Q
r^- co co ci 4- CM oo
". . o
^^ 4J
P 0
OJ rf
CM 0
d 2
-&Q
11
c -^
0 I-
B
— ti
<_o >
CO O
£ U
CU c<
E ^
.•a ^
Z|
h- 5
2 2
o ?
Q|
— ^
*2 ^
•2^ ^
c ^
| 1
aj ex.
"H. £
E i:
J ^
« ^D
3 «;
3 ^
S" 3
O ^
•o ^
o __.
K §
>^ "^
1 "S
*^ S
^ -v
(^ ^J
il ;
•£ ^_.S
cu -a "s «
o S u ° J>
Ecu v> m D-
> 3 « O.
« t» 3 g —
^ CJ 1> C f^
- ^ " >! (U
j= '(^ £ "2 3) ;
~ r~ Xi ™
W5 "O ^.' -3 ^
3 «j a 3 c 3
SSSOIE
^ _s/j C 00 O -^
C cd u cfl -C Q
^3 ^ C ^ — ^
2 1 g-.s 1 S
15 ^-S^ «
CO O — "O cO 3
S?^ 3 2 u £
1 E-sIa^-
S 2 T3 J *5 "=
> ^ ^ n ^
^ C pj ^ ^ "O
^ .5 ^ ci K/l (n
cu ^ ^5 O **
^ *^ "^ ^ CQ (/I
O *^ ^ i^ CU (U
••£ S .e 1
•5 c ~ * £ ^
U 3 S o '5 C!
§ ? u ^ 1 °
S ^ S J; ° c*
-o ?,5 8 <5 2
o ° ° o J: u
*" "o i! 2 - -
a u 3 o, i; c
* g s^l-S
-§ S ^ - en "S
OJ J3 c •*- • '
® -C ^
-------
248 WATERBORNE CIARDIASIS/TREATMENT TECHNOLOGY
UJ
a:
CL
SI
z>
CL
^\
^*
CL
\—
<
2
3NJ
(
\
\
=t
_> ^
u z
n: <
x \—
<
(_
c
c
c
v£
Qi LLJ
1— _l
UJ
1 31 D
^.
^
(
V
J O
[ Z
Pi ^
) —1 z
) O <
3 I H-
3
O
Z.
i-i DC
-i
00 U_
r
(?
3
|c
1 —
^ K
^-^IVODSad
LU
g
3
1—
<
3
^
I Z
Q^ L/> T^
' t-^ >— ' <[ >-*
PI pq PI
a: « o
S T ° - ^
) ^
3 o:
^ UJ
"• o i—
< <
* — * J
1C
1 1
~f — -
(NJ p, <
1— OL LU h-
L^. • LI 1 LU
_^ UJ h- u_ >
• • i— > a;
s: ^q « u_ p, zj
O _i
pq in
v> ^ °f^
^ °" ' J
^—^
. _ (^\ \
-, ^J 1 Jk
oil
v ^
UJ UJ Z
> a: <
>-i UJ I—
1— I
<_J D_ O
« Qi P
Pi 0 _J
< -i O
«
2,
s?
fc
t
|
'5,
•c
^
S
S
1
a
i
1
•s-
|
&
•5
i
S
•5
o
u.
-------
G. Logsdon 249
of the filter cake would soon be coated with clay, bacteria, and other
particulates found in natural waters and removed by the filter. The pores
would be plugged, and a high pressure drop would occur in the filter.
When body feed is added, the filter cake becomes thicker during the run,
but because the major constituent of the filter cake is DE rather than
removed particulates, the cake is porous, with good flow characteristics. In
order to properly adjust body feed dose, some DE filter plant operators add
body feed in proportion to raw water particulates. Turbidity is used as an
indicator of the paniculate content of water, so a body feed-to-turbidity
(mg/1 to NTU) ratio is sometimes used as a guideline for the needed dose of
body feed. An economical body feed-to-turbidity ratio would vary for
different raw waters. Body feed-to-turbidity ratios in this work ranged from
2:1 to 35:1. Raw water turbidity was generally 0.6 to 2 NTU.
A DE filter run generally is terminated because of head loss
considerations. Because the filter cake increases in thickness during the run,
filtered water turbidity should not rapidly increase (turbidity breakthrough)
as with granular media filters. Turbidity breakthrough can occur, however,
if the flow of water through the filter is interrupted. Interrupting the flow
can cause the diatomaceous earth cake to fall off the filter septum. This
should not be allowed to occur.
When a DE filter run is completed, the cake is removed from the septum,
spent diatomite is discharged, the septum is cleaned, and the filter is
prepared for another precoating. Because the spent diatomite in this
research contained radioactive particles or Giardia cysts, all spent diatomite
was saved in a 212 1 (55 gal) drum for disposal. The various grades of DE
used were commercial products obtained from Johns-Manville (Denver,
Colo).
An important feature of the filter system was a membrane sampling filter
located downstream from the DE filter. The purpose of this membrane filter
was to sample the entire effluent from the test filter and thus attain more
reliable data on removal of cyst models. Low turbidity water from an
inactive gravel pit was used for these tests.
Detecting and counting Giardia cysts in water is difficult, so a model for
the cysts was sought. Radioactive microspheres (New England Nuclear,
Boston, Mass.), 9 /nm in diameter, were used because they were very similar
in size to Giardia cysts and were easy to trace. The microspheres were resin
beads coated with cerium-141 so that the isotope would not leach. The
radioactivity of a single microsphere could be differentiated from
background in the low level counter used to measure gamma radioactivity
on the 5 jitm pore diameter membrane sampling filters. Most of the research
work utilized the radioactive microspheres.
Two confirmatory tests were performed with G. lamblia cysts. A cyst
suspension was injected into the raw water ahead of the DE filter. The DE
filter effluent was passed through a 7 /xm nylon screen (Tabler, Ernst and
Traber, Inc., Nitex). After the filter run was completed, the screen was
washed repeatedly with 1 to 2 ml of water. The wash water volume was
-------
250 WATERBORNE GIARDIASIS/TREATMENT TECHNOLOGY
recorded and the water was examined for cysts under the microscope. Cysts
were enumerated by the coverslip method and the hemocytometer method.
The efficiency of the effluent sampling method was determined by dosing a
quantity of cysts into the effluent line immediately prior to the sampling
filter. Water was pumped through the system and the cysts recovered from
the sample filter as described. Two trials indicated recovery efficiencies of 8.0
and 7.6%. A factor of 0.08 was used for calculations made to compare
numbers of cysts dosed in raw water and cysts recovered from filter effluent.
The removal of cysts by the diatomite filter assembly without filter aid was
determined by injecting cysts into the system, pumping water through the
system, and enumerating the cysts on the effluent sample filter. A series of 3
tests were conducted. The cyst recoveries were found to be 14.1, 9.0, and
6.1%, averaging 9.7%. Essentially no cyst removal took place, as cyst
recovery through the diatomite filter was similar to recovery when cysts were
added after the filter.
Results and Discussion
Test results showed that in most cases a substantial reduction in the
concentration of microspheres could be attained. Filter performance was
not related to turbidity reduction, a finding very similar to the statement on
p. 16 of the Army report. A probable explanation for this is that the cysts
were removed even though clays and other particulates that cause turbidity
were small enough to pass through the filter. The smallest cyst dimension
would be about 7 ^m, whereas clays and bacteria would be as small as 1 /urn,
and thus more likely to pass through the pores of the filter cake. Therefore,
performance is related to good operating procedures and not necessarily to
turbidity removal.
Data obtained from the filter runs in which microspheres were used are
plotted to show the differences in filter performance that are related to filter
operating procedures. Specifically, reduction of microsphere concentration
met or exceeded 99.9% in 21 tests when the diatomite precoat applied was at
least 91g/0.1m2 (0.2 lb/ft2) of filter surface, and when body feed was utilized.
In 2 tests with these operating conditions, reduction of the microsphere
concentration was less than 99.9%.
Filter operation with a precoat thickness of less than 91g/0.1m2 (0.2
lb/ft2) (2 runs), or with no body feed (3 runs), or with less than 91g/0.1m2
(0.2 lb/ ft2) of precoat as well as no body feed (3 runs) resulted in removal of
less than 99.9% of the microspheres. The test results show the importance of
properly establishing the filter cake by using an adequate amount of precoat
diatomite, and of maintaining the integrity of the cake during a filter run by
applying body feed.
Data indicated that filter efficiency was related to precoat thickness.
Therefore a series of tests was set up to evaluate this. The filter was operated
with precoat only. The tests were done with a single batch of water so that
turbidity differences would not influence the results, shown in Fig. 3.
Efficiency of microsphere removal increased with increasing precoat
-------
G. Logsdon
251
90
99
99. 9
I
CL
„ 99.999
99.9999
•».*
0 20 40 60 80 100
TURBIDITY REDUCTION PERCENTAGE
• 0.20 LB/FT2 OR MORE PRECOAT PLUS BODY FEED
• 0.20 LB/FT2 PRECOAT BUT NO BODY FEED
* LESS THAN 0.20 LB/FT2 PRECOAT PLUS BODY FEED
o LESS THAN 0.20 LB/FT2 PRECOAT AND NO BODY FEED
FIG. 2. Relationship of water quality and filter operation.
thickness up to 91g/0.1m2 (0.2 lb/ft2). Additional precoat did not improve
filter performance.
Results from the membrane sampling filters showed that the microsphere
removal capability of the DE filter usually improved during the run.
Ordinarily during a filter run the sampling filter was changed once.
Counting data showed that the relative number of radioactive beads caught
during the first portion of the run nearly always exceeded the proportion of
-------
252 WATERBORNE GIARDIASIS/TREATMENT TECHNOLOGY
LLJ
O
o
CL
LJJ
Q-
<
>
O
CL
111
O
cc.
o
90
99
99.9
99.99
0 0.1 0.2 0.3
HYFLO PRECOAT, POUNDS PER SQ FT
FIG. 3. Effect of precoat amount on microsphere removal.
the DE effluent that had passed through the filter (Table 4). Conversely, the
fraction of beads caught on the filter for the latter part of the run was nearly
always less than the proportion of effluent passing the sample filter. These
data suggested that addition of body feed and removal of particulates
produce a filter cake that is progressively more effective for paniculate
removal as the filter run continues.
Filtration research involving actual Glardia cysts was carried out to verify
the results with the microsphere studies. The Giardia cysts were dosed into
the diatomite filter influent and filtered through a very coarse filter aid
without the addition of body feed because these were the most inefficient
conditions for particle removal as indicated by filtration runs involving the
microspheres. The cysts were dosed immediately upon run initiation and the
-------
G. Logsdon 253
Table 4. Effect of run length on microsphere removal
% of Total
Volume of
Filtrate Samples % of Total Ratio of
and % Microspheres Recovered
Run Effluent Precoat & of Microspheres Recovered in To
No Sample No Amount Expected Effluent Sample Expected
28
29
34
58
59
37
57
56
55
43
1
2
1
2
1
2
1
2
3
1
2
1
2
1
2
1
2
1
2
3
1
2
JM 545/JM512
0 1 Ib/ft2 ea
JM 545/JM 512
0 1 Ib/ft2 ea
JM 545/HYFLO
0 1 Ib/ft2 ea
HYFLO
0 2 Ib/ft2
—
HYFLO
005 Ib/ft2
JM 503
0 10 Ib/ft2
JM 503
0 10 Ib/ft2
JM 503
0 10 Ib/ft2
JM 535
0 10 Ib/ft2
—
JM 560
0.20 Ib/ft2
62
38
73
27
65
35
21
31
48
23
77
77
23
53
47
58
42
23
29
48
60
40
63
37
88
22
0 1
99 9
59
30
1 1
42
58
28
72
80
20
79
21
97 7
0 6
1 7
72
28
1 02
0 97
1 20
0 81
001
2 86
2 81
0 97
0 23
1 83
0 75
036
3 13
1 51
043
1 36
0 50
4.26
02
04
1 2
0 7
run continued for a sufficient time to overcome the detention time of the
filter unit and piping, approximately 30 min.
One run utilized Qlg/O.lm2 (0.2 Ib/ft2) precoat of JM 560 with 2.2 x 10*
cysts dosed. Ten percent of the volume of the sample screen washing water
was examined. No cysts were detected. With corrections of recovery
efficiency (0.08) and volume sampled (0.10), if all cysts dosed had passed
through the DE filter, 18,000 should have been counted. A reduction of cysts
to less than one per 18,000 initially present is indicated. In a second run with
45g/0.1m2(0.1 Ib/ft2) of J M 545 as a precoat, l:3x 107 cysts were dosed but
again none were observed in the 15% of the filter washings examined.
Applying the factors (0.08 x 0.15), 1.6 x 105 cysts should have been observed
if all passed the DE filter and 8% were recovered from the sample filter. The
filter reduced the cyst concentration to less than 1 per 160,000 initially
present.
If a diatomaceous earth filter was as efficient for Giardia cyst removal as it
was for microsphere removal, a very small percentage of cysts would pass
-------
254 WATERBORNK GIARDIASIS/TREATMENT TECHNOLOGY
through the filter. Failure to observe cysts in the effluent suggests that cysts
are more easily removed than microspheres or that the cysts in the effluent
were not present in a large enough number to be detected by the sampling
methods employed in this study, therefore the use of the microsphere model
is conservative.
Conclusions from DE Tests
DE filtration for Giardia cyst removal is practical and effective provided
these filters are properly maintained and operated, both essential factors.
Thickness of DE precoat had a greater effect on microsphere removal
than did DE grade (particle size).
Precoat thickness for DE filters operated for Giardia cyst removal should
be at least 91g/0.1m2 (0.2 lb/ft2) to provide an effective barrier to cyst
passage.
DE filter efficiency for microsphere removal generally increased as the
filter runs progressed.
DE filters used in the production of potable water should not be operated
without body feed addition.
Measurement of effluent turbidities is an inadequate measure of the DE
filter's performance in regard to Giardia cyst removal so reliance must
instead be placed on use of proper treatment techniques with an adequate
amount of precoat and body feed diatomaceous earth.
Granular Media Filtration Studies
A program of laboratory and pilot plant research on Giardia cyst removal
by granular media filtration is now underway in the Drinking Water
Research Division, Preliminary jar test experiments in the laboratory are
designed to yield information on alum and polymer doses that will effect
reduction of Giardia cysts in granular media pilot plant filter tests.
The source of G, lamblia cysts for the first series of tests was an infected
human donor. After a number of apparently successful jar tests had shown
that coagulation, flocculation and sedimentation had reduced the cyst
concentration by 99%, the cysts were shown to be nonviable. Work with
these cysts was discontinued because of concern that test results with
nonviable cysts might not be representative of results that would be
obtained with viable cysts.
The Health Effects Research Laboratory has obtained mice infected with
viable Giardia muris cysts. The work is now being performed with G. muris
cysts. Preliminary results look promising. Analysis of a sample of G. muris
cysts in distilled water at pH 5.5 showed that the zeta potential (related to
electrophoretic mobility) was -26 millivolts with a standard deviation of 2.8
millivolts. Initial tests of flocculation and sedimentation also suggest that G.
muris cysts can be readily removed from water. If so, granular media
filtration also is likely .to be successful. This will be studied in the pilot plant.
FUTURE RESEARCH
The Drinking Water Research Division has funded a study of water
I treatment for Giardia cyst removal with research to be done by the
-------
G. Logsdon 255
Department of Environmental Health at the University of Washington. The
objective of this study is to develop information on the removal of Giardia
cysts from water by filtration and associated treatment processes such as
coagulation, flocculation, and sedimentation. Laboratory studies will be
conducted during the first year of the grant to obtain information on
treatment techniques that can remove the cysts from water. This
information will provide direction for field studies to be conducted in the
second year of research. Field studies will be done using pilot plants, full
scale facilities, or both. Granular media filtration and diatomaceous earth
filtration will be studied. The Department of Social and Health Services,
State of Washington, will cooperate in the field studies. This project, already
under way, is expected to continue into 1980. EPA studies at the
Environmental Research Center will complement the university research.
SUMMARY
Granular media filtration has been shown to be effective for removal of E.
histolytica cysts from water. In the 1930's Baylis reported removals
exceeding 99.99% when coagulated water was filtered through 61 cm (24 in)
of 0.5 mm effective size sand at 7.7 lpm/0.1m2 (2 gpm/ft2). U. S. Army
studies during WW II showed that removals exceeded 99.9% when
coagulation, flocculation and sedimentation preceded filtration at 24.4
lpm/0.1m2 (6.35 gpm/ft2) through 46 cm (18 in) of 0.36 mm effective size
sand. A preliminary measurement of the zeta potential of Giardia muris
cysts showed that these cysts are electronegative, as are most clays and
bacteria, so granular media filtration should remove Giardia cysts. Proper
operating procedures must be used at granular media filtration plants.
Army data showed that many more cysts passed through the sand filter
when good coagulation was not employed.
Diatomaceous earth filtration was shown by the US Army to effectively
remove E. histolytica cysts. Preliminary experiments with G. lamblia cysts
recently indicated DE filtration was just as effective for these cysts as for E.
histolytica. Research with radioactive microspheres showed a relationship
between filter operating procedures and effluent quality. To assure the most
effective filter performance for removal of cyst-sized particles, plant
operators should apply 0.2 pounds of precoat diatomite per square foot of
filter surface. Body feed should always be used.
Both granular media and diatomaceous earth filtration plants should be
able to remove cysts. Experimental data show that use of proper operating
techniques is essential if these plants are to perform effectively.
REFERENCES
1. Kirner, J.C., Littler, J. D.and L. A. Angelo. 1978. A waterborne outbreak of giardiasis at
Camas, Washington. Journal American Water Works Association. 70:35-40.
2. Laubusch, E. J. 1971. Chlorination and other disinfection practices, p. 200-203. In Water
Quality and Treatment, 3rd ed., American Water Works Association.
3. Beard, J. D II., and T. S. Tanaka. 1977. A comparison of particle counting and
nephelometry Journal American Water Works Association. 69:533-540.
-------
256 WATERBORNE GIARDIASIS/TREATMENT TECHNOLOGY
4. Spector, B. K., Baylis, J. R. and O. Gullans. 1934. Effectiveness of filtration m removing
from water, and chlorine in killing, the causative organism of amoebic dysentery. Public
Health Reports. 49:786-800.
5. Baylis, J. R., Cullans, O. and B. K. Spector. 1936. The efficiency of rapid sand filters in
removing the cysts of amoebic dysentery organisms from water. Public Health Reports.
51.1567-1575.
6. War Department Report 834. Efficiency of standard Army purification equipment and
diatomite filters in removing cysts of Endamoeba histolytica from water. 3 July 1944.
Submitted to The Engineer Board, Fort Belvoir, Virginia, and/ or The Chief of Engineers,
U. S. Army, Washington, D.C., by Water Supply Equipment Branch, Technical Division
111, The Engineer Board, Fort Belvoir, Virginia, and the National Institute of Health, U.S.
Public Health Service.
7. Black, H. H. and C. H. Spaulding. 1944. Diatomite water filtration developed for field
troops. Journal American Water Works Association 36:1208-1221.
8. Johns-Manville Products Corporation. 1968. Johns-Manville one square foot test filter
unit operating instruction, FF-113A.
9. Hoye, R. L., Jr., 1978. Removal of Giardia cysts and cyst models by diatomaceous earth
filtration. Unpublished M.S Thesis, University of Cincinnati.
10. Hoye, R., Logsdon, G. S. and J. M. Symons. 1978. Removal of Giardia cysts and cyst
models by diatomaceous earth filtration. Unpublished manuscript.
Discussion
E. L1PPY: Would you comment on the use of a DE filter as a sampling
unit for Giardia detection?
G. LOGSDON: I think it would be possible to do that if you could
separate the diatomaceous earth from the Giardia after the water had passed
through the DE filter. The Army work with amebic cysts and the DE filter
showed that the Army obtained practically complete removal of the cysts. A
number of those filter runs had no cysts detected in the effluent. I do not
know, however, how you would separate diatomite from cysts in order to use
this as a detection technique.
D. JURANEK: At the time that CDC was developing the sand filter,
diatomaceous earth was used as a first coat medium because that is what
came with the swimming pool filter that was used. You pointed out exactly
the problem; there is no good way to separate Giardia cysts from the
diatomaceous earth.
B. LIECHTY: I think it is really encouraging to see a research person
associated with EPA recognize the importance of quality plant operations. I
think there is a lesson to be learned from the waste water treatment program
where literally millions of dollars were spent to design and construct
treatment facilities that were very effective, on paper. Experience is now
showing that, in some cases, effluent standards are not being achieved, often
as the result of improper operation of the plants.
-------
257
SESSION VI - OTHER RECENT RESEARCH
(ABSTRACTS)
Chairman - Walter Jakubowski,
U.S. EPA, HERL, Cincinnati, Ohio
In Vitro Model for Analyzing the Attachment/
Release of Trophozoites of Giardia:
Evidence for the Involvement of Contractile Proteins
and the Effect of Atabrine
L. S. Erlandsen, D. E. Feely, J. V. Schollmeyer,
D. G. Chase
Report of Endosymbiosis in Giardia muris and
Comparison of Organelle Distribution in
Giardia muris and Giardia lamblia
P. Nemanic, R. L. Owen,
D. P. Stevens, J. C. Mueller
Intestinal Distribution, Attachment and Relationship
of Ciardial Trophozoites to Peyer's Patches During
Acute Infection in the Immunologically Intact Host
R. L. Owen, P. Nemanic, D. P. Stevens
Giardiasis in Vail, Colorado
W. Jones and T. Edell
The Presence of Giardia lamblia Cysts in Sewage
Sludges from the Chicago Area
J. C. Fox, P. R. Fitzgerald
Axenically Cultivated Giardia lamblia: Growth,
Attachment, and the Role of L-Cysteine
F. D. Gillin, L. S. Diamond
-------
258
In Vitro Model for Analyzing the Attachment/
Release of Trophozoites of Giardia: Evidence for the
Involvement of Contractile Proteins
and the Effect of Atabrine
Stanley L. Erlandsen, Dennis E. Feely,
Judy V. Schollmeyer, and David G. Chase
Departments of Anatomy and Lab Medicine and
Pathology, University of Minnesota, Minneapolis,
Minnesota, and Cell Biology Laboratory, V. A. Hospital
Sepulveda, California
The direct interaction and/or attachment of the trophozoites of Giardia
with the microvillous border of the intestinal epithelium appears to be a
common factor in postulated mechanisms for the pathogenesis of this
intestinal protozoan. Using a method for the isolation of trophozoites of
Giardia muris developed in our laboratory (Feeley and Erlandsen, 1978,
Anal. Rec. 190:393) the mechanism of attachment of trophozoites of Giardia
to surfaces in vitro has been analyzed by means of: 1) morphological
examination including scanning, transmission, and high voltage electron
microscopy, 2) functional studies using light microscopy cinematography
with phase and Nomarski interference optics, 3) immuno-biochemical
studies on the presence and distribution of contractile proteins in the
trophozoite, and 4) the effect of various drugs including Atabrine on the
attachment and/or release of trophozoites from surfaces. High voltage and
transmission electron microscopy have revealed the presence of fibrous
masses in the edge of the adhesive disc (lateral ridge), in rod-like masses
paralleling intracellular axonemes of the antero- and postero-lateral flagella,
and fibrous masses associated with the axostyle paralleling the caudal
flagella. Using SDS-gel electrophoresis, co-migrating bands similar to avian
contractile proteins have been identified for myosin, a -actinin, tubulin, and
acti'n. Immuno-fluorescent studies using antisera to avian a -actinin and
actin have demonstrated a -actinin localization in the lateral ridge of the
adhesive disc and in the rod-like fibrous masses associated with the postero-
lateral flagella. Actin localization was also seen in these same fibrous masses
as well as in the median body, an organelle composed of microtubules.
Cinematography of attached trophozoites revealed that the tail of the
trophozoite is capable of vigorous lateral movement and that dorsal flexion
of the tail occurs just prior to detachment from a surface. Collectively, these
observations suggest that the rim of the adhesive disc (lateral ridge) may
possess contractile properties which govern shape changes in the adhesive
-------
S. L. Erlandsen 259
disc. Attachment of the disc to the surface followed by a reduction in the size
of the disc due to contraction of the rim would create a negative pressure to
maintain attachment. Relaxation of the contractile elements in the rim
presumably during or prior to dorsal flexion of the tail of the trophozoite
would vent the negative pressure and permit detachment from the surface.
Studies on the effect of Atabrine on trophozoite attachment revealed that 1
/xg/ml of Atabrine (quinicrine-HCl) produced 22% detachment of
trophozoites in 10 min whereas 10 and 100 jug/ml produced 45 and 92%
detachment respectively. After 30 minutes of exposure, 44% detachment was
observed at a dosage of 1 /j.g/ ml while 85 and 99% detachment were observed
at 10 and 100 /^g/ml of Atabrine. The exact antiparasitic mechanism of
Atabrine is not clear although it is known to inhibit nucleic acid synthesis,
however, these data clearly demonstrate that it may act directly on the
attachment of trophozoites to surfaces through an as yet undefined
mechanism. (Supported in part by USPHS Grant No. AM 18344)
Discussion
B. WHEAT: Those were beautiful micrographs. What type of fixation,
dehydration, et cetera, did you use for scanning preparation?
SI ERLANDSEN: We use a variety of fixatives, mainly glutaraldehyde
type fixatives. Sometimes we have osmium associated with them in what we
call a cocktail fixation. It is not a good name for it, but they are actually in
the same vial together. These specimens are all critical-point dried. We have
not looked at Giardia with a freeze-drying technique.
D. STEVENS: My compliments also on your usual elegant presentation.
Do you have any thoughts on what the discriminatory character of
trophozoites is for surfaces, or do you think it is just a rather mindless
operation on the part of the trophozoite? 1 realize this goes beyond your
data, but 1 wondered if you had reached any conclusions. In the sense that
we see attachment in the intestine, in particular areas and not in others, is
there something in terms of proteins or pH?
S. ERLANDSEN: There is no answer at this particular point. We are
working with in vitro systems that will allow us to hopefully pinpoint some
of the answers that you would like to have on that subject. Perhaps the more
interesting question would be in relationship to something that E. Meyer
raised a little earlier, that Giardia may have an affinity for heat. The fact that
this organism likes warm surfaces such as epithelium might help to attract it,
and to counteract this one person even suggested that perhaps we should
drink a lot of ice water. A lot of people like to chew cracked ice, but in
Mexico that is not a very good thing to do because of other things in the
water.
D. STEVENS: How did you test the viability of the trophozoites?
S. ERLANDSEN: We tested the viability in four ways. First of all we
looked at dye exclusion using either trypan blue or erythrosin B. We have
done some labelling with isotopes and have looked at spontaneous release
over time. We have correlated each of these tests also with trophozoite
motility. Through all of these experiments we are able to keep viability at
90% up through about 180 min so we feel that for the very short acute
experiments that we are dealing with, viability is not a factor. All of our
work is usually done within a half hour to an hour at the most.
-------
260 WATERBORNK GIARDIASIS/OTHER RESEARCH
T. NASH: I noticed in the micrographs that there was nothing attached to
the organisms. Is this a common occurrence? Do you ever see anything
attached?
S. ERLANDSEN: In vivo we have seen things that look like bacteria
associated with them, but we have not seen those other than during early
stages of the isolation procedure. It appears to be an artifactual association
with the trophozoite flagella. In the second cycle during isolation, no
microorganisms are seen attached to Giardia , and if you go through a third
and a fourth cycle there are few if any microorganisms present.
T. NASH: I was particularly interested in the bacteria. I am interested in
any mucus, or white cells.
S. ERLANDSEN: No. I am looking forward to hearing Dr. Owen's paper
on that.
-------
261
Report of Endosymbiosis in Giardia muris and
Comparison of Organelle Distribution in
Giardia muris and Giardia lamblia
Paulina Nemanic, Robert L. Owen, John C. Mueller,
and David P. Stevens
Cell Biology Section, Veterans Administration Hospital,
San Francisco, California and Department of Medicine,
Case Western Reserve Univ. School of Medicine and
University Hospitals, Cleveland, Ohio
Ultrastructural observations of Giardia muris in the mouse model system
revealed endosymbiotic microbes not previously reported in Giardia.
Endosymbionts were 240 to 260 nm wide and 600 to 1400 nm long with
internal structure consistent with bacteria. Endosymbionts were not seen
entering Giardia but appeared to be passed on during giardial replication.
They were observed to divide within the cytoplasm of Giardia without
evidence of digestion by the host in either trophozoites or cysts and were
found in groups of mice serially infected with cysts passaged repeatedly
through CF1 mice. Phagocytosis of particles as large as endosymbionts was
not seen. Instead, a number of previously undescribed organelle features
point to a feeding mechanism for Giardia by means of small (40 nm)
micropinocytotic vesicles located immediately beneath the dorsal and
exposed ventral surfaces of the trophozoite. Vesicles fuse with larger
peripheral tubules found in the same area and previously presumed to be the
primary feeding vacuoles. Well-developed RER, previously unreported in
Giardia, was observed and confluence of RER cisternae with these tubules is
suggested as the source of digestive enzymes, with discharge of unusable
products via larger exocytotic vesicles along the dorsal surface of the
organism. Endosymbionts were concentrated centrally in the nuclear area
and were uncommon in the peripheral feeding regions. In comparison with
G. lamblia the same organelles as in G. muris were found but not the
endosymbionts. Replication and transmission of foreign DNA by
endosymbionts within Giardia strains must be investigated to determine
whether they alter trophozoite pathogenicity, metabolism, range of
infectivity, antigenic surface characteristics or host specificity as
endosymbionts have been shown to do in other protozoa.
(Editor's note: For the complete paper on this topic see Nemanic, P. C., R.
L. Owen, D. P. Stevens, and J. C. Mueller. 1979. Ultrastructural
observations on giardiasis in a murine model. II. Endosymbiosis and
-------
262 WATER CIARDIASIS/OTHER RESEARCH
organelle distribution in Giardia muris and comparison with Giardia
lamblia. Jour. Inf. Dis. August 1979, In Press)
Discussion
B. WHEAT: There is a technique that has been used with coccidia in
testing whether sporozoites actually penetrate cells by using ruthenium red,
and it stains very darkly and does not enter cells that are attached. You
might try putting that in your suspension and seeing if that helps.
P. NEMAN 1C: I will try it.
S. ERLANDSEN: Do you see any difference in the amount of RER
associated where these symbionts are present in the organism?
P. NEMANIC: Since all had endosymbionts, I cannot say.
S. ERLANDSEN: When you showed us the long tubules, there were
median bodies that had short dorsal ribbons extending up from them, and it
looked like the median bodies were dissolving; there was much more space
between the microtubules than one would expect to see in a microtubule
organelle like that. I wonder if that is not perhaps the beginning stage of
encystment and what you are seeing is very similar to what H. Sheffield was
describing with the long tubules. Perhaps this is the release of the cyst wall
itself through exocytosis of that material.
P. NEMANIC: That is possible. All of this was done in vivo, in the
animal; and we have some micrographs which show trophozoites next to
cysts, and whenever we find that, the cytoplasm nearest the dorsal edge is
very, very light and the trophozoite itself looks very condensed.
S. ERLANDSEN: Endosymbionts in other intestinal protozoa are
common. As far back as the turn of the century Dr. Dobell made a very
beautiful drawing of endosymbionts present in some of those protozoa,
including Hexamita intestinalis. The RER tends to be arrayed in fairly
substantial amounts around these endosymbionts, and it is very interesting
to note that these areas were referred to as higher metabolic centers within
the protozoans.
D. PRICE: Recently, 1 have noted a number of what appear to be ingested
bacteria in Giardia trophozoites that I see in light microscopy, and also I see
what appears to be some digestion of these organisms since they vary in size
from the time they enter until the time they seem to be extruded, and I
wonder what evidence you have to show that this is not a vacuole in which
the organism has been ingested, or whether it is an actual compartment in
which the organism is living.
P. NEMANIC: We never find any evidence of digested endosymbionts.
We have seen some evidence of what looks like division taking place within
the endosymbionts.
D. PRICE: All the other intestinal protozoa do ingest bacteria. Is this the
one exception?
P. NEMANIC: We have never seen any evidence that Giardia are
ingesting them, and one would not expect to find too many bacteria floating
around the upper duodenum.
S. ERLANDSEN: I think in some intestinal protozoa they are not always
ingested. If you consider the order Hypermastigida, there are
endosymbionts that metabolize cellulose, and you have a number of cases
like that where the organisms exist either in or on the host itself. Perhaps you
-------
P. Nemank 263
would comment on the trophozoite you saw that had 35 endosymbionts
within a single section. Did it look viable?
P. NEMANIC: It looked wonderfully viable. It looked just like all the
rest, only freckled.
A. TOM BES: Following along with that question about the digestion, did
you see any lytic activity?
P. NEMANIC: I have not seen anything that looks like a lysosome. I
expect that the peripheral tubules may be a heterogeneous group, that is,
they all look alike, but we are doing histochemical testing, and we may find
that perhaps some of these are digestive organelles. I have looked at literally
thousands of Giardia and have never once seen any evidence of ingestion of
bacteria. I have only seen a few invaginations filled with material that is
always different from the debris floating near the trophozoites.
-------
264
Intestinal Distribution, Attachment and Relationship
of Giardial
Trophozoites to Peyer's Patches During Acute
Infection in the
Immunologically Intact Host
Robert L. Owen, Paulina Nemanic, and David P. Stevens
Cell Biology Section, Veterans Administration Hospital, San Francisco,
California and Department of Medicine, Case Western Reserve University,
School of Medicine and University Hospitals, Cleveland, Ohio
Increased severity and incidence of symptoms of giardiasis in
immunodeficient humans and laboratory animals points to an abnormal
immune response to Giardia. We studied prospectively infected healthy mice
in a well-described model system, using ultrastructural techniques to
establish normal distribution of trophozoites, their relationships to
intestinal mucosa, particularly Peyer's patches and cecal lymphoid patches,
and structural indications of the normal reaction of intestine and intestinal
immune organs to Giardia muris.
We found that trophozoites which colonize only the proximal 25% of the
intestine adhere to microvilli of columnar cells near the bases of villi, wedge
into furrows in the epithelial surface, or lodge in mucus within the unstirred
layer. Density of trophozoites colonizing the jejunal epithelium correlated
with cyst excretion numbers prior to as well as at peak cyst shedding. Giardia
do not adhere to Peyer's patch M cells, which transport small particulate
material from the lumen to the lymphoid system, but appear to enter
intestinal lymphoid structures by incursions through defects in the lymphoid
follicle epithelial barrier. Lymphocytes leave the epithelium and attach to
Giardia and Hexamita in the lumen. We conclude that Giardia colonize only
the proximal 25% of the small intestine and maintain this location by
adhesion and localization in the unstirred layer. They produce no apparent
ultrastructural damage in normal mice but elicit a previously undescribed
intraluminal cellular immune response prior to clearance by the host.
(Editor's note: For the complete paper on this topic see Owen, R. L., P. C.
Nemanic, and D. P. Stevens. 1979. Ultrastructural observations on
giardiasis in a murine model. I. Intestinal distribution, attachment and
relationship to the immune system of Giardia muris. Gastroenterology
76:757-769.)
Discussion
W. JAKUBOWSKI: I would like to compliment you and Dr. Nemanic
and Dr. Erlandsen on some very fine electronmicroscopic work, but I must
-------
R. L. Owen 265
admit that I feel somewhat like an adolescent getting his first look at
Penthouse Magazine — I am fascinated by what 1 see, but I am not sure I
understand it. Based on your observations, would you or any of the others
care to postulate a mechanism for the production of symptoms in giardiasis?
R. OWEN: Well, we do a lot of fantasizing about this, as with Penthouse
Magazine, I suppose. You begin to feel a very intimate sort of relation with
Giardia, like thinking about taking them for walks and things of that nature.
When there is co-infection with a virus or something else that causes loss
of integrity of the barrier along the wall of the intestine, more of the Giardia
may wander into the tissue. In a sensitive host capable of a cellular immune
response, one can anticipate that there would be an inflammatory response
within the tissue which might disrupt or damage the villi of the intestine.
This is a question that we have dealt with tangentially in the last couple of
days: why do some people get sick and not others? Why can someone carry
Giardia for a long time and not have symptoms and then begin to have
symptoms? Dr. Wolfe suggested that perhaps there may be a co-infection
with some other agent so that what we see when there is symptomatic
giardiasis may be a summation effect. The Giardia are minding their own
business, eating a little bit of your lunch, and along comes some other factor -
infection, stress, irradiation or something else - and they wander into the
epithelium. Then there is a cellular immune response within the gut wall that
may disrupt the epithelium and create a change in the turnover rate so that
less mature epithelial cells are migrating up out of the crypts and covering the
surface of the villi, cells that are less developed and less able to absorb and
transport nutrients from the lumen of the host.
S. ERLANDSEN: Your results with the mouse seem to be quite different
from what has been reported to occur in the rat in terms of distribution. In
the studies that have been done (Hegner, R. and L. Eskridge, J. Protozool,
24:511, 1938) where they were sacrificing 20 to 50 animals per group, it was
found that a bimodal distribution of Giardia existed in the rat intestine,
trophozoites were not present in the first 30 to 40 cm of the loop of the
duodenum and jejunum, indicating that there was something in the pancreas
affecting them, and I think you are stating the opposite of that. It was also
shown that there is one high incidence of colonization in the jejunum and
another one down in the ileum as well. Do Dr. Stevens or you ever see
variations of this distribution? Do you ever find them lower, or have you
looked at a large enough series of animals to be sure that they do not get
down into other areas as well?
R. OWEN: With electron microscopy 10 animals can seem like a very
large series by the time you have gone through all the sectioning and
preparation. There are some studies on distribution that indicate that when
you change the diet one can change the distribution. In puppies, for instance,
you can move them up and down in the intestine by changing the diet, and by
changing the diet I presume what one changes is motility. When there is an
overgrowth or a massive infection, you find them everywhere: in the
gallbladder, the colon, the stomach. It may have to do with the numbers of
organisms that are present and the resistance of the host. We did not fast
these animals because we did not want to stress them. If the animals were
fasted, we might see something different, and obviously in other animals
other things have been seen.
-------
266 WATERBORNE CIARDIASIS/OTHER RESEARCH
One thing that we speculated about is that there might be some sort of
membrane configuration over the microvilli that could induce attachment,
that somehow Giardia might read the surface to get a clue that they were in
the upper jejunum. Dr. Marilyn Etzler (Etzler, M. E. and Branstrator,
M. L., Jour. Cell Biology, 62:329, 1974) at Davis has used lectins with
fluorescent labels and found that there are proximal intestinal membrane
configurations at the bottom of the villus and not at the top, and they coinci-
dentally happen to be just in the areas where the Giardia seem to hang on.
Whether the Giardia have evolutionary clung to this sort of membrane as
an indication of where they should attach we do not know, but this is what
we have seen.
S. ERLANDSEN: In your micrographs you had a radial spoke
arrangement on the adhesive disk. It looked like areas of elevations and
depressions, and that puzzles me greatly. Do you have any structural
correlation for this, or do you think there is some other explanation,
perhaps?
R. OWEN: I do not know.
D. STEVENS: In response to Dr. Erlandsen's earlier question about
looking at more animals for evidence of infection in the upper part of the
small bowel, when we first started working with a model 4 years ago, Dr.
Roberts-Thomson working in my lab, made the observations of those
villus:crypt ratios which we talked about yesterday, but I was not terribly
enthusiastic about their control. It disheartened me a bit that he only saw the
changes in the first 10 to 15 cm of the small intestine. The reduction in the
villus:crypt ratio did not occur beyond that point, so that all observations
thereafter were made arbitrarily at 10 cm. In retrospect, these observations
are probably valid and may be related to interaction with the adhesive
trophozoites.
S. ERLANDSEN: If one measures villus to crypt height, one has to be
very careful of liver and pancreatic function because it has been shown by
transposition of the intestine that the height of the villus is dependent upon
pancreatic and bile flow.
G. JACKSON: Just as an aside, with the intestinal nematodes one often
sees whole coats of host cells, both polymorphonuclear and
monomorphonuclear white cells, when the nematode projects into the lumen
and there is a host reaction.
-------
267
Giardiasis in Vail, Colorado
Wesley E. Jones and Timm A. Edell
Parasitic Diseases Division, Bureau of Epidemiology,
Center for Disease Control, Atlanta, Georgia and Field Services Division,
Bureau of Epidemiology, Center for Disease Control,
Denver, Colorado 80266
Within a 2-day period in April 1978, 37 cases of laboratory-confirmed
giardiasis from Colorado and other states were reported to the Parasitic
Diseases Division, Center for Disease Control. All were in residents of or
visitors to Vail, Colorado, during March or April 1978.
An epidemiologic investigation was initiated consisting of a questionnaire
and a stool survey of residents of the city and an adjacent control town.
Information was obtained on 741 residents of Vail; 429 (58%) gave a history
of diarrheal illness that occurred February, March, or April 1978. The
epidemic curve demonstrated a rise in the number of acute diarrheal illnesses
beginning on March 14-16 and reaching a peak on April 1-2. No statistically
significant differences in illness rates were noted by age, sex, or length of
residence. However, there were significant differences between those who
drank no water and those who drank 1 or more glasses per day (P<0.05).
None of the 7 separate Vail Valley water districts could be statistically
implicated as the source. Seventy-two stool examinations revealed 12 cases
(17%) of Giardia lamblia infection. None of 37 specimens from the control
town were positive (P<0.05). This outbreak emphasizes the continuing need
for investigative work into the epidemiology and pathophysiology of G.
lamblia infection.
-------
268
The Presence of Giardia lamblia Cysts in Sewage
and Sewage Sludges from the Chicago Area
J. Carl Fox and Paul R. Fitzgerald
Department of Parasitology, Microbiology and
Public Health, College of Veterinary Medicine,
Oklahoma State University, Stillwater,
and
Department of Veterinary Pathology and Hygiene,
College of Veterinary Medicine, University of Illinois, Urbana, Illinois.
Raw sewage and anaerobically digested sludge collected from 5 sewage
treatment plants.operated by the Metropolitan Sanitary District of Greater
Chicago were examined biweekly (February 1976 to May 1977) for the
presence of parasite ova or cysts. Eleven genera of parasites were found:
Ascaris sp., Toxocara sp., Toxascaris leonina, Trichuris sp., Enterobius
vermicularis, Taenia sp., Hymenolepis sp., Eimeria sp., Isospora sp.,
Entamoeba coli and Giardia sp. Giardia cysts were never observed by
sucrose flotation, however, they could be detected and counted by
examination of stained wet mounts. Buffalo black, methylene blue and
phyloxine B were effective stains for use with wet mounts. The Giardia cyst
counts ranged from 333 to 2000/gallon in some raw sewage samples. In most
cases there were too few to count. Cysts were not detected in anaerobically
digested sludge, even though the organic solids were much more
concentrated. This indicates that these cysts may not survive the anaerobic
digestion process.
Giardia cysts were most prevalent in sewage composed primarily of
domestic waste. Cysts were seldom found in sewage from industrial areas.
Cysts were most often encountered during the warm months (April -
September) but were occasionally present in raw sewage at other times. The
actual incidence of Giardia in sewage may be higher, as Giardia cysts are
difficult to detect when present in small numbers. Improved methods are
needed to detect these parasites in sewage.
Discussion
R. OWEN: Were these separate or combined storm and sanitary sewers
that you examined? Could the presence of dogs have contributed to your
findings?
J. FOX: It was all one system and dogs could have contributed to our
findings. If you walk down the streets of Chicago you have to tiptoe at times
since there is a large amount of dog feces on the streets which can be washed
into the sewers by rain.
-------
J.C.Fox 269
R. OWEN: Is there any suggestion from anything you have that some ot
those cysts you found in the two-week digested sludge were viable?
J. FOX: I do not know whether they were viable. I was looking at
formalin treated specimens. They looked just like the cysts that I found in
the raw sewage, but that was only at one plant, and that plant often had
problems with their digesters. They have shut them down sometimes and
used them just as holding tanks. I am sure that at times they gave samples
that were not representative of what they should have been giving me, which
was the digester drawoff that should have been two weeks old. I may have
received some new sewage. Also, digesters are charged every day. These
digesters hold three and a half million gallons. They put in one-fourteenth of
that volume each day and they draw out that much. 1 may have just received
a fresh specimen.
R. CYPESS: To indirectly answer Dr. Owen's question, serum samples
taken from swine that were fed sewage were analyzed in my laboratory for
the presence of Toxocara antibody. These animals developed an antibody
response suggesting that the embryonated eggs were definitely viable. There
is a difference in resistance of Toxocara and Giardia, of course, but there is
an indication that Toxocara can survive and remain viable in sewage.
J. FOX: The second slide 1 showed you, the Toxocara slide, was an egg
recovered from a holding basin in Fulton County, and that egg had already
been in that holding basin for two years in an unembryonated state. It was
removed and embryonated in the laboratory. It was there that long because
the sanitary district stopped shipping sewage down the river for awhile, so
there was that time period when they were not hauling anything there. That
sample was collected during that time and indicates that these eggs can
remain viable for a long time in an unembryonated state.
D. PRICE: If you fix Giardia in neutral formalin, they will last
indefinitely. I am sure CDC is still sending out material collected from
honey pots in Japan after World War II, so I think that they will stay in good
shape for a fairly long time in formalin.
J. FOX: I do not think Giardia is a problem in sewage sludges, but I do
think there is a possible problem with ascarids — probably Ascaris
lumbricoides. Cook County has one slaughter house which slaughters
swine, but that one slaughter house does not dump sewage into all the plants
that I was investigating, and I found the eggs at all the plants. I think there
are human factors contributing to the Ascaris eggs that we are seeing. These
eggs survived that long period in the holding basins, and they would
embryonate after we brought them back into the laboratory. We seldom
found an embryonated egg right out of the basin.
W. JAKUBOWSK.I: Would you care to speculate on the possibility of
examining sewage as a means of conducting surveys on prevalence of
Giardia in populations?
J. FOX: I think there is a very good possibility of using sewage plants. I
believe it is a valid approach to determine whether the cysts are present in the
community. At least then you know that there is a host somewhere
harboring Giardia. You can then pick out those communities and key in on
them and look for the source of Giardia cysts, whether they be from the dog
population or human populations, or whether they are prevalent in certain
parts of town or in certain water supplies.
-------
270
Axenically Cultivated Giardia lamblia:
Growth, Attachment and the Role of L-cysteine
Frances D. Gillin and Louis S. Diamond
Laboratory of Parasitic Diseases
National Institute of Allergy and Infectious Diseases
National Institutes of Health
Bethesda, Maryland
Giardia lamblia is grown axenically in vessels nearly filled with medium
and with little air phase. The media used are undefined, but all contain added
L-cysteine, probably as a reducing agent. We have studied the requirement
for L-cysteine and its effects upon the kinetics of trophozoite multiplication,
sensitivity to air, and attachment to the culture vessel in Diamond's TP-S-1
and new TY1-S-33 media using a strain of G. lamblia isolated by E. A. Meyer
and obtained from G. Visvesvara and G. R. Healey.
If L-cysteine was omitted from either medium, trophozoites did not grow,
and eventually lysed. If L-cysteine were required only as a reducing agent, it
should be replaceable by other reducing agents. In TYI-S-33 medium, the
requirement for L-cysteine was rather specific: only glutathione or N-acetyl-
L-cysteine supported slight growth. In TP-S-1 medium, lower
concentrations of L-cysteine satisfied the requirement and other sulfhydryl
compounds were partially effective: possibly because of higher endogenous
L-cysteine concentrations in this medium. Ascorbic acid was- totally
ineffective.
Rates of killing of trophozoites in the absence of L-cysteine, were
determined using a new quantitative method. Trophozoites were exposed to
various experimental conditions: (1) high or low oxygen tension (pO-i)
achieved by varying the relative liquid to air volume; and (2) presence or
absence of L-cysteine or D-ascorbate. After varying times, survival was
measured by clonal growth in a semi-solid agarose medium in culture tubes
with lowpO2 plus L-cysteine. Survival curves obtained under lowpOz (^ 30
mm Hg) were complex; 50% of the cells were killed in ~ 3-1/2 hours with
neither L-cysteine nor D-ascorbate and in ~ 11 hours with only D-ascorbate.
Under increased pQi (~ 140 mm Hg), trophozoites were killed exponentially
~ 50%/hour without L-cysteine or D-ascorbate. With both L-cysteine and
D-ascorbate, killing at the same rate was preceded by a lag of ~ 11 hours,
while with either one the lag was ~ 5 hours. Thus, while D-ascorbate did not
support growth of trophozoites, it did provide some protection from killing
by oxygen.
G. lamblia trophozoites in axenic culture exhibit a striking tendency to
attach to the walls of the culture vessel in addition to their ability to swim
-------
F.D.GiKn 271
freely. Attachment of trophozoites to intestinal epithelium in vivo may play
a role in the disease process. The proportions of attached and free
trophozoites were determined during growth of cultures in TP-S-1 initiated
(1) with only free cells or (2) with cells which had been attached, but were
detached by chilling. In both cases, a majority of the cells was attached until
the end of logarithmic growth phase.
Attachment of trophozoites to glass in TP-S-1 with L-cysteine was rapid
for the first 2 hours then continued more slowly. With only D-ascorbate,
attachment was somewhat depressed even before the cells began to show
decreased viability. Cells did not attach without L-cysteine or D-ascorbate.
Thus, L-cysteine appeared to fullfill at least 2 requirements in the axenic
cultivation of G. lamblia: first as a reducing agent; and second as a growth
factor- possibly as an essential amino acid.
Discussion
L. DIAMOND: These 2 media were designed originally for E. histolytica,
and there is a pitfall: in relation to E. histolytica, there is not much difference
whether you sterilize by filtration or autoclaving, but apparently
sterilization by autoclaving TP-S-1 or TYI-S-33 will result in no growth of
Giardia.
G. HEALY: I would like to commend you for a nice job. It was an elegant
series of experiments, and I can appreciate the work that has gone into it,
realizing the short period of time you had the cultures.
I can see problems in the future; the realization that people will soon be
discovering new trophozoites from cysts by the elegant work performed,
and that more cultures will soon be established. Having suffered the
admonitions of Dr. Diamond in naming E. histolytica strains, and seeing the
incredible mixup in naming Acanthamoeba and Naegleria isolates, I would
make a plea that Dr. E. Meyer, Dr. L. Diamond and Dr. Visvesvara get
together and work out some system for naming these isolates. I do not even
know what designation you use for this one except that it is something that
Dr. Meyer cultured after it was removed from somebody's duodenum.
F. GILLIN: Dr. P. Daggett of the American Type Culture Collection said
that when the culture from Dr. Meyer's lab was delivered he filled in a blank
for the designation of the strain and Portland 1 was written down.
G. HEALY: That proves the point that a system is needed for naming
these isolates.
E. MEYER: What per cent agar did you use in your semisolid medium?
F. GILLIN: First of all, it is agarose and it is a very low per cent, about 0.2,
so it really is semisolid. G. Lakhonina and J. Teras (Estonian Academy of
Science; Abs. 5th Intern. Congress of Protozoology, p. 188, 1977) used a
really solid 1.5 per cent agar medium.
E. MEYER: Yes, and the results were quite different from ours. They
observed that the organisms were confined to the interface between the
medium and the glass.
F. GILLIN: I was not aware of that, but that does not surprise me at all. I
think that inside such a solid agar matrix they really could not grow very
well. I might add that we see localized colonies within the agar matrix and
eventually there are halos between the agarose and the glass so that cells are
-------
272 WATER GIARDIASIS/OTHER RESEARCH
escaping. This does lead to potential problems in obtaining clones, which
can probably be overcome by dilution.
S. ERLANDSEN: Do you have any idea how many organisms you have
to have in order for the colonies to become visible?
F. GILLIN: You mean how many cells constitute a colony?
S. ERLANDSEN: Yes, and the fact that there may be many colonies that
you are not detecting because you cannot see them if you are just using a
macroscopic examination.
F. GILLIN: We have what we call colony forming or cloning efficiency in
which you take the number of cells that you put in, which is a statistical
figure arrived at by dilution from a counted sample, and divide by the
number of colonies you count. You get a perfectly linear dose-response
curve. The colony forming efficiency is in the range of 30 to 70%, usually
about 50%.
S. ERLANDSEN: There's about 30 to 50% that you sometimes do not
detect, and the question is whether they reproduce or whether they remain
viable and are not multiplying to give a visible colony.
F. GILLIN: Most bacteria or trichomonads have a 100% colony forming
efficiency. That is not the case here. I worked out a similar method for
growing E. histoiytica trophozoites in agar. The histolytica trophs prefer
agar to agarose, and they prefer a slightly higher concentration, about 0.5%.
You can look at the agar culture under an inverted microscope and estimate
the number of cells in a visible colony of E. histolytica, which is of course a
much larger cell. You may see single cells but if you let the culture grow
longer, the colony counts do not increase with time. I really do not have a
very good idea of how many cells there are per Giardia colony.
-------
273
SESSION VII - RESEARCH NEEDS
Chairman - Gordon G. Robeck,
U.S. EPA, MERL, Cincinnati, Ohio
Giardiasis and the Safe Drinking
Water Act
C. Hendricks
Panel Discussion: Current Research and
Future Research Needs
Moderator- G. G. Robeck, U.S. EPA
PANEL MEMBERS
L. J. McCabe, U.S. EPA
M. G. Schultz, DHEW, CDC
T. M. Vernon, Jr., Colorado Department of Health
K. Kawata, Johns Hopkins University
-------
274
Giardiasis and the Safe Drinking Water Act
Charles W. Hendricks
Office of Drinking Water,
U.S. Environmental Protection Agency,
Washington, D.C.
ABSTRACT
Historically, regulatory activity in the United States with respect to drinking water has
focused on the containment of infectious disease. The Safe Drinking Water Act of 1974
and its 1977 amendments are no exception. However, they do provide the U.S.
Environmental Protection Agency with additional tools to control waterborne disease
including options for establishing maximum contaminant levels or mandatory treatment
requirements. Based upon analyses of these regulatory means of controlling
contaminants, it is possible to regulate infectious contaminants such as Giardia on a case
by case basis, where risk can be established.
This technical symposium to review the waterborne transmission of
giardiasis has afforded the Agency, the Department of Health, Education
and Welfare, the waterworks industry and various individuals from
universities, state organizations, as well as the public, a unique opportunity
to investigate one of the microbiological agents involved in waterborne
gastroenteritis. The National Academy of Science volume, "Drinking Water
and Health," drew attention to what appeared to be a steady increase in
waterborne disease since the early 1950's. Whether this "increase" is due to
actual disease, improved reporting practices or both, is debatable; but, the
opportunity we have had here can be of major significance in determining a
means of control of infectious disease.
This report will center on the Safe Drinking Water Act, its provisions and
how it is currently being implemented to control pathogenic microbiological
agents. I will also discuss a series of regulatory options, any of which could
be taken under the Safe Drinking Water Act. Finally, I will suggest research
that is needed to answer fundamental questions concerning the treatment of
drinking water for the control of Giardia lamblia and other pathogenic
microorganisms.
LEGISLATIVE CONTROL OF DRINKING WATER
The concept of legislative control of infectious disease in the United States
began with the signing of the National Quarantine Act of 1878. This Act was
designed to prevent the introduction of infectious or contagious diseases
into the United States. Modifications of the National Quarantine Act in
1890 and 1893 provided the Treasury Department with authority to prevent
the introduction of contagious diseases across state lines and granted the
Department additional quarantine powers. The first Interstate Quarantine
-------
C. Hendricks 275
Regulations based on this legislation were published in 1894. Under these
regulations, activity primarily consisted of maintaining sanitary drinking
water aboard vessels. However, in 1912, due to severe outbreaks of intestinal
disease among passengers aboard steamships on the Great Lakes, the
Treasury Department issued Circular Number 49 which required U.S.
Public Health Service surveillance of sanitary conditions on interstate
carriers. Paragraph 15 was added to the Interstate Quarantine Regulations
at this time which required State certification that drinking water aboard
interstate vehicles must be incapable of causing disease. This paragraph, in
concept, was the yardstick by which the 1914 Treasury Standards for
drinking water were developed. The 1914 Treasury Standards prescribed
mandatory limits for bacteria and were applicable to about 9,000 interstate
carrier supplies. In addition to a limit of less than 2.2 coliform bacteria/100
cc, general populations of bacteria were not to exceed 100/cc. Later
revisions by the Treasury Department and by the U.S. Public Health Service
in 1925, 1943-1946 and 1962 added various chemical constituents to the
Standards as guidelines. In 1925 the standard plate count requirement was
dropped when a turbidity standard was established as an esthetic parameter.
Early in 1968 the first series of bills was introduced in Congress for the
protection of all public water supplies. Eventually the Safe Drinking Water
Act of 1974 emerged after 14 attempts. Although Public Law 93-523 amends
the Public Health Service Act (P.L.- 410), it applies to all public water
systems serving 25 persons or more on a regular basis, or an estimated
155,000,000 Americans. A more complete review of the history of drinking
water standards development may be found in Taylor, 1977 (1).
The Safe Drinking Water Act
With the passage of the Safe Drinking Water Act, Congress mandated the
U.S. Environmental Protection Agency (EPA) to provide safe drinking
water and to protect the public from waterborne illness and disease. With
respect to drinking water standards, Congress intended that the Act be
implemented in three separate phases. First, the EPA was to promulgate
Interim Primary Drinking Water Regulations (2) based on a revision of the
1962 Public Health Service Standards. These regulations were issued
December 24, 1975, and became effective June 24, 1977. Second, EPA was
directed to contract with the National Academy of Sciences (NAS) for a
two-year investigation of all contaminants in drinking water that could be
harmful to man. This report was completed and a summary was published in
the Federal Register July 11, 1977. Third, EPA was directed to promulgate
more comprehensive revised drinking water regulations based upon the
NAS study, current available data and the results of the Agency's on-going
research program. These regulations are to be proposed for public comment
next year and will not become effective until 18 months after promulgation.
Besides these requirements, the Safe Drinking Water Act of 1974 provides
for:
-------
276 WATERBORNE GIARDIASIS/RESEARCH NEEDS
1. The development of national primary drinking water regulations
for contaminants that may adversely affect the public health which
will:
a. cover public water systems and protect health to the maximum
extent feasible;
b. include a Maximum Contaminant Level (MCL) for each
contaminant or a specified treatment technique if it is not
feasible to monitor the level of that contaminant;
c. contain criteria and procedures to assure compliance; and
d. may be enforced by the Administrator of the U.S.
Environmental Protection Agency if the States elect to not
enforce them.
2. The Administrator to take emergency action if a contaminant
presents an imminent or substantial hazard to the public.
3. The development of non-Federally enforceable secondary
regulations as guidance to the States.
4. The establishment of Federal State programs to improve
surveillance over drinking water quality and protect underground
sources of drinking water.
5. The authorization for Federal grants to assist the State surveillance
and enforcement provisions of the Act.
6. The authorization of variances and exemptions to primary
regulations.
In keeping with the precedent established in implementing the 1912
revision of the Interstate Quarantine Regulations, Congress envisioned that
the States would exercise primary enforcement responsibility (primacy)
over water supplies under their jurisdiction, with EPA assuming this task
only when a State was unable or unwilling to meet minimum requirements
contained in the Act and associated regulations. In order to attain primacy
under the Act, the States must: establish drinking water standards and
institute procedures for variances and exemptions at least as stringent as the
national regulations; adopt and implement an adequate enforcement
program; maintain records and submit reports as required by the
Administrator of the EPA; and, establish an emergency response plan and a
program for plan review.
National Interim Primary Drinking Water Regulations
The Interim Regulations establishes MCL's for coliform bacteria,
turbidity, ten inorganic chemicals, six pesticides, and radiological
contaminants (Table 1). As of June 24, 1977, water utilities are required to
conduct periodic monitoring at a prescribed frequency to ensure compliance
with the regulations and to notify the public if the standards are exceeded.
The Act also provides that the States will develop their own regulations that
are at least as stringent; some States have more stringent requirements than
those in the Interim Primary Regulations.
The Maximum Contaminant Levels for coliform bacteria and turbidity
are essentially those contained in the 1962 Public Health Service Standards.
-------
C.Hendricks
277
Table 1. The maximum contaminant levels for constituents in the
National Interim Primary Regulations
Constituent
Level
(mg/l unless specified)
Biological Parameters
Coliform bacteria
Turbidity
Inorganic Chemicals
Arsenic
Barium
Cadmium
Chromium
Lead
Mercury
Nitrate (as N)
Selenium
Silver
Fluoride
Organic Chemicals
Endrm
Lmdane
Methoxychlor
Toxaphene
2.4-D
2,4, 5-TP Silvex
Radionuclides
Radium 226 and 228 (combined)
Gross alpha particle activity
Gross beta particle activity
1 per 1 00 ml (mean)
1 NTU (Waiver to 5 NTU
possible)
0 05
1
0 010
0 05
0 05
0 002
10
0 01
0.05
1.4-2 4"
0 0002
0004
0 1
0 005
0.1
0 01
5 pCi/l
15 pCi/l
4 millirem/year
* Based on annual average air temperature.
The maximum contaminant level for coliform bacteria is 1 coliform/100 ml
of water, however a supply may substitute chlorine residual determination
for 75% of the required coliform analyses upon receiving state permission.
Supplies using this provision must maintain a 0.2 mg/l free chlorine
residual. The turbidity MCL is 1 Nephelometric turbidity unit (NTU) with
the exception that the State may allow up to 5 NTU if the water does not:
1. interfere with disinfection;
2. prevent maintenance of an effective disinfectant agent throughout
the distribution system; or
3. interfere with microbiological determinations.
It is generally agreed that these MCL's are sufficient to protect the public
from infectious disease transmitted by recent fecal contamination; however,
there are data to suggest that the coliform testing procedure may be not
-------
278 WATERBORNE GIARDIASIS/RESEARCH NEEDS
entirely effective in certain diseases of viral and protozoan etiologies. The
apparent problem with the indicator procedures is not whether these
particular disease producing organisms are found in fecal material, but
rather that a number of viruses and protozoans can survive in water for
longer periods than the coliform indicators. Some viruses and protozoans
also seem more resistant to disinfection than coliforms. For these reasons,
spurious positive coliform results must be fully evaluated before their
validity is doubted.
THE PHILOSOPHY OF STANDARD SETTING
It is profitable at this point to consider the decision-making process of
setting standards for microorganisms. The Safe Drinking Water Act is quite
explicit regarding means the Administrator of the EPA should use to
regulate contaminants. The regulatory means open to the Administrator
consist of either the establishment of an MCL or a mandatory treatment
technique or techniques. Although MCL's can be developed for specific
types of water systems (ground, surface, large or small), there is considerable
flexibility in specifying treatment for supplies where the population is at
risk. In both cases however EPA must demonstrate that: 1. there is a need
for regulation, and 2. the regulatory action is economically and
technologically feasible.
The determination of whether a contaminant should be regulated is often
a difficult and arduous task. The basic questions in determining human risk
are: 1. What is the toxic or infectious potential of the contaminant, and 2.
What is the probability of human exposure? Dr. D. V. Subrahmanyan (3),
while with the World Health Organization's Community Water Supply and
Sanitation Section, has suggested the following criteria on which to base
numerical limits:
1. Severity of adverse effects upon the population; irreversible or
chronic effects; those which have genetic, carcinogenic, or
teratogenic effects or which may be embryo toxic.
2. Persistence in the environment and resistance to environmental
degradation; accumulation in man or in the food chain.
3. Metabolic degradation or synthesis in biological systems which
may produce metabolites either more or less toxic than the parent
compound.
4. Ubiquity and abundance of the agent in man's environment
particularly those occurring naturally or produced by man.
5. Size, type and demographic characteristics of the population
exposed and magnitude of exposure.
6. Selective exposure of highly vulnerable groups of the population.
Regulatory Decision Making
A typical decision-making pathway for establishing primary drinking
water standards is given in Fig. 1-3. This pathway rather quickly provides
administrative guidance on when a constituent should be considered for
regulation, the scientific evidence required, the technical and economic
considerations to be made, and the social acceptability of the proposed
-------
C. Hendricks 279
standard. Although a similar decision pathway was used effectively by the
Office of Drinking Water in developing the Interim Primary Regulations,
these types of decision pathways lack flexibility in dealing with
contaminants that may pose substantial risks only at select water supplies. It
should be noted that the pathway presented here is highly stylized and
individual activities may occur simultaneously with other actions noted
earlier in the flow chart.
Preliminary Action
Regulatory action under the Safe Drinking Water Act is predicated upon
finding the contaminant in drinking water. Figure 1 describes the
preliminary action needed to determine whether a constituent should be
extensively considered for possible regulation. Information is gathered from
all available sources concerning the public health risk of a constituent and its
prevalence. Perhaps the major decision point at this stage in the evaluation
is to determine if the constituent can be controlled at the point of origin
before source water becomes contaminated. For example, when chlorine is
used in disinfection of plant sewage effluents it is effective in reducing or
eliminating those constituents. Obvious difficulties arise from non-point
sources such as wastes from home sewage treatment facilities or wastes from
animals and humans along a watershed. Therefore, consideration must be
given to other viable avenues open to the Agency for the control of
microorganisms such as the National Pollutant Discharge Elimination
System (NPDES) permit program.
Technical Evaluation
Decisions that must be made in setting standards regarding health risk of
a contaminant to the public, and the associated costs with both monitoring
and removal of a contaminant, is potentially the most controversial area.
Besides the capital, operation and maintenance costs, one must also
consider those costs to the public as a direct result of illness. It is important
to note that health standards in the Interim Primary Regulations are based
upon levels at which no adverse health effects are observed, with an
appropriate safety factor included. Although the Administrator has
authority to establish an MCL for a contaminant, it is possible that some
deleterious contaminants may not be regulated because the risk to the public
is not perceived to be great enough to warrant the costs. In addition to these
concepts, Figure 2 also shows the decision protocol for determining
monitoring and treatment capability as a means of controlling a
contaminant.
Administrative Action
If overall costs have been found acceptable and a tentative MCL can be
determined, a monitoring survey may be ordered to determine an estimate
of national risk and to gain additional cost information. Information about
analytical methods and treatment efficiency is also obtained in the survey.
Should the monitoring survey indicate that data for another regulated
constituent indicates that the two constituents respond similarly,
consideration will be given to a surrogate MCL (Fig. 3). The coliform test in
-------
280
WATERBORNE GIARDIASIS/RESEARCH NEEDS
I
I
§.
.r
•v
i
-------
C.Hendricks
281
I* 1-
Q cc
z
o
<
1
o
|
<
G4
2 *-
s*
23 '
2 oc 5- i*
11 s°
£ « !ii e
si I?
°i 25
* ^ ffl tJ
-ss gl
°s 1 =
!3* Si
iS 2£
c g Jg |=
ifili
slg»p
i ^ uj < G
- t lu * So
o £ en — »-(-><
,_
E K
O K '
2*
S'*
s<;
B M X
221"
S£g
°si
is:
siS
II? :
^ < z '
c <^x ;
z
=
5
<
f
E
C
£
3
3
' ° >
t * 1
C g
I
I
1
-------
282
WATERBORNE GIARDIASIS/RESEARC H NEEDS
CONSIDER
SURROGATE
MCL
5°iE*
^ Q U- t-
^ _J O <
£ w > 2
S|H£
Pol"
iZ 0- £ t~ _
COSTOFINS1
TREATMENT
EXCEED APPI
EXISTING TO
PRODUCTION
Pfi
O J Z
10 > T
|S|
s%*
Is"
OPERATION (
SHOULD NOT
CONSUMER'S
I
1-
0
a
S
X
SHOULD NOT
z
I ? -
S0
O <
K§
i<
NO ECONOMI
CONSTITUEN
ADVERSE HE
o
E
-------
C. Hendricks 283
determining the microbial quality of water is probably the best known
example of the use of a surrogate analytical procedure to warn of possible
risk of intestinal disease outbreaks.
Some indication of how the public perceives the need for a proposed
regulatory action must be obtained since social acceptability is an important
determinant in governing our behavior. Should the public view a particular
regulation as not acceptable, perhaps we have not followed through with
our responsibility to keep the public informed of the risks associated with a
particular contaminant. There are, however, instances where the public is
fully aware of a problem and is still willing to accept risks. Examples of this
include the use of tobacco and certain artificial sweeteners, the disregard by
some of the beneficial aspects of fluoride, and even disregard for the 55 mph
speed limit.
REGULATORY BASIS FOR THE CONTROL OF GIARDIA
AND OTHER INFECTIOUS MICROORGANISMS
Based upon the literature (5-7) the numerous reports of Giardia outbreaks
(4,8-11), and the information presented at this Symposium, some type of
regulatory action may be needed to eliminate or greatly reduce the risk of
waterborne giardiasis. There are several regulatory options that are
available for the control of organisms such as Giardia and these options
include the following:
1. Rely on the present coliform standard and/or encourage more use
of the chlorine residual substitution procedure.
2. Establish either an MCL or a surrogate for Giardia.
3. Establish a regulatory treatment technique for removal or
inactivation of Giardia in vulnerable supplies.
4. Require vulnerable supplies to be monitored.
5. Warn physicians to be alert.
6. Do nothing at this time and await developments.
7. Request the Administrator to take emergency action as in the case
of asbestos at Duluth, Minnesota.
While there is merit in some of the options, none are entirely satisfactory
by themselves. An MCL established for all public surface water supplies is
possible but there are serious economic drawbacks to it. Monitoring
procedures are available for Giardia, however routine monitoring would be
extremely expensive and require highly trained laboratory technical
personnel that appear to be out of reach for most utilities at this time (12,14).
Neither surrogate tests for Giardia nor increased reliance on coliform
bacteria indicator systems seems to be of any real value at this time.
There is merit in the option of requiring the installation of conventional
filtration for the control of Giardia. In addition to the obvious benefits
derived from the removal of cysts and ova of protozoans, the reduction in
hazards associated with virus, turbidity and certain chemicals could be
expected with that additional public health barrier(13). Small surface
supplies, however, appear to be at greatest risk from Giardia and the
additional construction, operation and maintenance costs associated with a
-------
284 WATERBORNE GIARDIASIS/RESEARCH NEEDS
mandatory treatment requirement applicable to all supplies could be
burdensome even though necessary. Options 4 and 5 seem sensible steps for
States to take under their own drinking water program. Action under the
Administrator's prerogatives could be effective should the situation become
more than presently perceived.
Infectious Disease
As stated previously, I believe some sort of regulatory action at the
Federal level for the control of giardiasis is appropriate. 1 do not believe a
generally applicable MCL for Giardia per se would be the most appropriate
approach since the monitoring frequency undoubtedly would be quite low
because of economic considerations. Instead, I prefer regulatory action that
would specify mandatory treatment if a supply, and, therefore the public,
was shown to be at risk from infectious disease. In this way, action could b»:
taken if the supply was vulnerable to a disease producing agent for which
treatment could be specified.
Clearly, in a proposal such as this, consideration must be given as to how
one determines whether a supply is at risk and an evaluation of the
likelihood that people may become ill. The relationship between
microorganisms and disease is fundamental, but I can suggest to you some
determining factors that could lead to the designation of supplies requiring
treatment or a modification of their present treatment scheme. These
include the following:
1. Has a supply been incriminated in an outbreak of waterborne
disease?
Important in this consideration would be whether the problem
resulted from source water, the treatment facility or the distribution
system since the nature of the problem would trigger specific
corrective treatment.
2. Have pathogenic organisms been recovered from the source water,
demonstrated on the watershed, found in the raw water intake, or in
finished water?
3. Has a State or Federal sanitary survey determined that a supply is
at risk? Is the supply properly operated?
4. Is the supply in continued violation of State and Federal drinking
water regulations for coliform bacteria and turbidity?
5. Is the supply in compliance with existing State and/or Federal
guidelines?
I am sure there are other items that could be added to the list and those
presented here could be sharpened considerably, but based upon these
considerations, research is needed in several areas. Specifically, these
recommendations were designed for Giardia, but they are more generally
applicable to research required for viruses and other protozoa:
1. Develop and define treatment techniques for the removal or
inactivation of Giardia cysts.
2. Develop a rapid, inexpensive monitoring procedure for Giardia
cysts.
-------
C. Hendrkks 285
3. Investigate the feasibility of an indicator system.
4. Perform studies to determine:
a. effect of the environment on the survival of cysts and
trophozoites;
b. mechanisms of disease transmission;
c. vulnerable human populations; and
d. how to establish populations at risk.
In conclusion, I think there are alternatives to developing MCL's and
mandatory treatment requirements applicable to all or even particular
classes of water. The development of rules specific for supplies shown to be
at risk would have an added advantage of being flexible enough to adapt to
local conditions while keeping local and state costs to a minimum. I would
encourage all of you to study this proposal and I look forward to your
comments as we move forward together in the implementation of the Safe
Drinking Water Act.
REFERENCES
1. Taylor, F. B. 1977. Drinking water standards - principals and history 1914-1976 J. New
England Water Works Assn. 91. 237-259
2. Office of Drinking Water. 1975. National interim primary drinking water regulations,
Report EPA-570/9-76-003, U.S. Environmental Protection Agency, Washington, D.C.
3. Subrahmanyan, D. V. 1975. Development and application of drinkingwater standards. In
Proc., Colloquium on Drinking Water Quality and Health, Water Research Centre,
Medmenham, United Kingdom.
4. Kirner, J. C., J. D. Littler, and L. A. Angelo. 1977. Giardiasis in Camas, Washington. In
Proceedings 97th Annual Conference, May 8-13, 1977, AWWA, Denver, Colorado.
5. Sobsey, M. D. 1978. Current microbiological concerns in drinkingwater. In Proceedings
Water Supply Engineering, Quality Treatment and Management, May 22-24, 1978,
Department of Environmental Sciences, School of Public Health, University of North
Carolina, Chapel Hill.
6. Wolfe, M. S. 1975. Giardiasis. J. Am. Medical Assn. 233: 1362-1365.
7 Kamarov, F. I., and V. K. Ilinich. 1976. Contemporarystateof the the problemof lamblia
pathogenicity. Klin. Med. 54:9-15
8. Center for Disease Control. 1975. Giardiasis - in residents of Rome, New York, and in U
S. travelers to the Soviet Union, p. 366, 371. In Morbidity and Mortality Weekly Report,
October 25, 1975, DHEW, Atlanta, Ga
9 Center for Disease Control. 1977. Food and waterborne disease outbreaks. U S
Department of Health, Education and Welfare, Atlanta, Ga.
10. Center for Disease Control. 1978. Giardiasis - Vail Colorado, p. 155. In Morbidity and
Mortality Report, May 12, 1978. DHEW, Atlanta, Ga.
11. Center for Disease Control. 1977. Waterborne giardiasis outbreaks Washington, New
Hampshire. Morbidity and Mortality Weekly Report, May 27, 1977, DHEW, Atlanta,
Ga.
12. Burke, J. A. 1977. The clinical and laboratory diagnosis of giardiasis, p. 373. In CRC
Critical Reviews in Clinical Laboratory Sciences, CRC Press, Cleveland, Ohio
13. Cotruvo, J. A. 1978. Regulatory aspects of potable water disinfection, p.817. In R. L.
Jolley, H. Gorchev, and D. H. Hamilton(eds.), Water chlorination-environmental impact
and health effects, Vol. 2, Ann Arbor Science, Ann Arbor, Michigan.
14. Jakubowski, W., T. H. Ericksen, and S. L. Chang. 1977. Detection, identification and
enumeration of Giardia cysts in water supplies. Proceedings AWWA Water Quality
Technology Conference, December 4-7, 1977, AWWA, Denver, Colorado.
Discussion
L. McCABE: You said one of the control options was to request the
Administrator to take emergency action. Could that be a selective option
exercised by him each time there is an outbreak?
-------
286 WATERBORNE GIARDIASIS/RESEARCH NEEDS
C. HENDRICKS: That is one of the options listed in the Safe Drinking
Water Act. As I would interpret it, in the case of an emergency where action
is needed rapidly without going through the complicated regulatory
procedure, the Administrator upon receiving a proper basis for a decision
could then move to take action that was recommended to him or action that
he may decide upon himself. I do not view it as a means of acting on every
outbreak that crops up. If, for example, suddenly there were a tremendous
epidemic involving deaths or very serious illness, then 1 suspect the
Administrator would move very rapidly. But I do not view it as a practical
means for controlling giardiasis unless the situation should deteriorate
substantially.
D. PRICE: The people of Florida are concerned because we have been
told, or it has been suggested, that by the year 2000 the southeastern United
States may be the only resource for fresh water for many parts of the United
States. Since the reduction of the chlorine content in water to 0.2 milligrams
is a reasonable amount, we are finding that throughout Florida now the
coliform counts are well over the basic limits, and no one seems to know
exactly what to do about this. I realize there is also a problem with
trihalomethanes with the higher levels of chlorine. I would appreciate any
comments you might make on this. It also seems to me that any safe drinking
water regulation must be tied to some extent to the volume of water used in
the particular community and to waste disposal methods and effluent
evaluations and to the type of water purification system that could be
employed under the circumstances. I also would like to know at what level
the decision really can be made that there is a hazardous situation.
C. HENDRICKS: First of all, the Safe Drinking Water Act is applicable
at the tap. We are concerned with source water and concerned with
treatment and distribution. However, the MCL's are applicable at the tap,
and we must rely upon other regulations to protect source water. According
to the Act, source water standards are to be considered when treatment
techniques are prescribed.
With respect to trihalomethanes, before September 1,1 would have been
glad to talk to you at length about them. However, I am legally bound not to
discuss those proposed regulations now that the comment period has closed.
G. ROBECK: I think it is fair to say there is no regulation that says
chlorine has to be limited to a residual of 0.2. That is a matter of option on
the part of the state agency with primacy. 1 do not think you could say that
the sudden increase is due to'that regulation. There may be some people who
want to taper off in the amount of chlorine. In Florida they used to advocate
0.5 parts free available chlorine and now they have reduced that. That may
be partly the cause, but I do not think the regulation per se is the cause.
-------
287
Panel Discussion
Current Research and Future Needs
Moderator: Gordon G. Robeck, USEPA
PANEL MEMBERS
Leland J. McCabe, USEPA
Myron G. Schultz, DHEW, CDC
Tom M. Vernon, Jr.,
Colorado Dept. Of Health
Kaz Kawata, Johns Hopkins University
G. ROBECK: In an effort to bring some focus to this Symposium, we
thought it would be advisable to have four representatives with different
perspectives share with us their ideas on what the research needs might be in
general, and particularly of course for the agency. We have, therefore, asked
federal representatives from EPA and from CDC, a state representative and
a university researcher to speak to us on how they see the research needs in
their areas. There will be opportunity for discussion from the floor.
Our first panelist is Leland McCabe who is the Director of the Field
Studies Division in the Health Effects Research Laboratory. He has been
responsible for directing a good share of the health effects research in the
USEPA on drinking water for many years and has been involved for a
decade in the general area of epidemiology as well as sanitary engineering.
He is well qualified to speak from the perspective of USEPA.
L. McCABE: It would seem that we have about four areas in which EPA
has research needs. I think there is clearly a public health problem that we
cannot tolerate. Certainly the outbreaks in Portland or Rome were of
sufficient magnitude to indicate that those outbreaks have to be stopped. So
I would put as our first research need to prevent the occurrence of
waterborne Giardia outbreaks, and some of this research is as mundane as
the data that were presented by Ed Lippy and then followed by Al Bingham,
like how long you have to boil the water to make sure you can stop an
outbreak. We had a lot of discussion about that, but we need definitive data
yet on that point. You apparently do not have to boil it for 5, 10 or 15
minutes. You may just have to get it to the boiling point to do the job.
If we are going to prevent these outbreaks from occurring, clearly there
will be larger utilities or state regulatory agencies that will want to move
ahead before we go through the decision matrix that Dr. Hend ricks outlined
for us. We should be able, from the research at EPA, to tell them what they
have to do to improve treatment. That generates another research need: how
to measure the efficacy of treatment.
-------
288 WATERBORNE GIARDIASIS/RESEARCH NEEDS
We have been doing some work in-house on how to get these things out of
the water and extramural work at Colorado State University on infectivity.
The new excystation work at the University of Oregon certainly holds
promise for determining the adequacy of disinfection. If this does not
mature very quickly, we may have to drop back to the mouse model or
something like that to measure the efficacy of treatment.
In the area of treatment, research has to be linked to EPA's current
problem that relates to the production of trihalomethanes from
chlorination. We are going to be promoting the use of shorter contact times
and lower residuals to prevent the formation of trihalomethanes. These of
course may be contraindicated from research on how to get rid of Giardia. It
is also possible that some of the alternate disinfectants might be much better
than we ever thought they would be.
There is a great difference in the type of research that is necessary for the
large utilities or for the individual home, and this will have to be considered.
The University of Oregon is going to pursue some of the issues on alternate
disinfection, so we will get data on how that would fit in since the EPA may
change its disinfection recommendations. Dr. Logsdon's data showed we
are moving along on how to do filtration. This will also be pursued by the
University of Washington through EPA funding.
The next area of research would be outbreak investigations. Water is a
very efficient transmitter of disease, even though the main endemic route
may be person to person. Contaminated water can obviously infect
thousands of people in a short period of time. Giardia presents a situation
where the source is usually contaminated, so if there is a deficiency it is not
like a cross-connection that may affect one block or something like that; it is
going to get the whole community if treatment is inadequate.
Certainly we have to do better workups on the outbreaks that do occur.
This of course is a primary responsibility of CDC, but EPA is always willing
to assist in defining the water situation more specifically. The state
regulatory agencies are going to have to take a greater interest in how to do
this. In some states EPA, because of the primacy rule, will be the regulatory
agency. We ran into an outbreak or two recently where there was a
dichotomy in the state concerning who worries about the water. The health
department is not that agency any more, so we are going to have to work out
some administrative mechanisms to be sure everybody is trying to do the
same thing.
We do plan within the next year to have some projects collaborating with
CDC on improving reporting of disease outbreaks. We will probably try this
out in a few model states with somebody on site in the state that will try to
stimulate the reporting of outbreaks so we can get a quicker response. In
50% of the outbreaks the etiology is unknown. We have to work on that
situation to understand the epidemiology of the causative agents. We want
to be alert to the fact that some disinfection practices will be changing
because of other regulations, and we will want to have a quick indication if
that is causing more outbreaks and stop them right away if this does occur.
-------
G. G. Robeck 289
Now, another decision to be made is if we should make everybody do
something about the problem under the Safe Drinking Water Act. There are
several areas of research, including defining the endemic situation,
determining the proportion of persons who are infected but are not sick, and
there are also many water supplies that are probably contaminated and do
not cause disease.
When I was with CDC, on one occasion we determined the effect of the
environment by treating everybody in the community and then measuring
their reinfection rate. That is a possibility here. You could clean everybody
up and then measure different qualities of water and how fast it takes for
reinfection to occur. Certainly the point made on the sewage monitoring is
one way we might get a feel for the endemic situation, and it also has a
relationship to the next item I will bring up.
As Dr. Hendricks has outlined, if there is a problem, we have to define the
scope of the problem and determine the most economical method of
treatment, then we must develop cost-benefit information. In doing
monitoring, as the regulation suggests, we have to have a suitable technique.
It does not have to be for the agent itself; we could establish something like
we do for pasteurization, a time-temperature relationship to see if the
treatment is working all right.
The fourth area involves EPA's effort to change the methods of waste
treatment in this country. We are very enamored at the moment with land
application, something that we deliberately did not do in the past. With the
cyst levels that were indicated yesterday in sewage, there should probably be
a big concern for spreading this on the land. It is going to have an ecological
impact, and we may start getting quite a few of the wildlife infected. We may
then have a geometric sort of progression between animal and man. The
nature of the research is split into two areas: how to get rid of the liquid
waste and how to get rid of the sludge. Both of them have a spray
component, so we get involved with an aerosol exposure. Nobody wants to
get a Giardia cyst down their throat because of aerosol from the spray
irrigation of waste water.
Of course, even if EPA's interests are totally taken care of in the research
and the regulatory sense with a good program, there is still going to be a
Giardia problem because the portion that is waterborne is only some
segment of the total. So I think that research in other areas by other agencies
may be needed with regard to this organism.
G. ROBECK: Our next panelist is Dr. Myron Schultz. You may recall
that he is the Director of the Parasitic Disease Division, Bureau of
Epidemiology, Center for Disease Control, DHEW in Atlanta. Dr. Schultz
will speak to us from that organizational perspective as well as his
professional perspective.
M. SCHULTZ: I would like to begin my comments by applauding EPA
for holding this conference. I think it is very timely and beneficial. 1 would
also like to applaud EPA for the degree of cooperation and collaboration
with CDC. In my own observation it has been a very salutary collabortion, a
-------
290 WATERBORNE GIARDIASIS/RESEARCH NEEDS
very smoothly functioning, good complementing of skills, and it is not
always typical of interagency collaboration that I have seen at other times
and places.
1 think one of the key concepts that I would like to leave with you is the
need for surveillance, and we are acting now to increase our surveillance for
waterborne diseases both unilaterally and collectively. Some philosopher
said that the proper study of mankind is man himself. Well, the proper study
of waterborne giardiasis is more and more the study of giardiasis until we
satisfy all the academic questions that have been raised. Mr. Craun
indicated that there have been 23 recognized outbreaks of giardiasis since
1966, and that is the most common cause of recognized waterborne disease
outbreaks. 1 think that this is just the tip of the iceberg, and there is probably
much more going on that we have not yet recognized. We should appreciate
that giardiasis does not cause death and does not commonly cause fever, so
these two indicators for an epidemic are missing, and the principal
symptom, diarrhea, is one that is not commonly reported even though it
may be a common symptom. We need to refine our surveillance techniques
to get a better fix on this, and 1 am sure once we do that we will find much
more giardiasis.
We have just established a unit within the Epidemiology Bureau of CDC
to investigate all waterborne infectious disease, including giardiasis, and it is
headed by Dr. Jim Hughes. He and his colleagues werejust here at EPA last
month to discuss protocols, and 1 think the sort of collaboration that has
existed in the past will be expanded. 1 think this is needed because there are a
lot of questions in epidemics that could be answered, and the place to answer
questions about Giardia is the epidemic rather than the endemic situation. I
think if the disease were only endemic we would not be here today. In the
epidemic we want to prevent further occurrence of cases and we want to
correct any problems in existing facilities that may be causing it. We want to
establish the traditional epidemiologic parameters of time, place and
person, and then we want to carry on special studies that would contribute
to our understanding of giardiasis.
The specifics of these remain to be elaborated and fixed in common
protocols between CDC and EPA, but they all deal with pathogenicity. The
other key concept 1 would like to mention this morning is pathogenesis—the
first concept being surveillance—because in all the talks during the past two
and a half days on various aspects of giardiasis the issue of pathogenesis
came up, and there were more questions raised than answered, as so often
occurs in new problems.
Just to mention a few, what is the significance of Giardia cysts in a water
supply? Does it mean that there is going to be an epidemic or that there was
an epidemic? In Rome, New York we had 5000 people infected. We found
one Giardia cyst in 250,000 gallons of water. Of course, this was somewhat
after the occurrence of the epidemic, and this emphasizes the point that we
should get in as promptly as we can. Another question in pathogenesis is
whether there are any coexistent factors that stimulate Giardia to become
pathogenic such as viral and other infections.
-------
G. G. Robeck 291
The role of domestic and wild animals in propagation of the disease needs
to be determined. There is a notion of beaver being a reservoir for man, but it
remains to be shown what species and strains go through beaver. Of course
we are prevented from infecting human volunteers in this day and age with
Giardia, which is in a way a hindrance to understanding pathogenesis. But
there are many questions on reservoirs of Giardia for man that could be
answered. What is the role of the domestic dog? Will it turn out to be another
story like toxoplasmosis? In that case the cat was found to have an
important role in the life cycle. Perhaps there is a stage that we are missing
here in Giardia in domestic animals.
Another question that intrigues me is, if this is a prevalent infection, as we
believe it is—in our surveillance work we found that 3.4% of stools
examined in state health department laboratories were positive for
Giardia—why is it that evidence of person-to-person transmission is
lacking? I made a point yesterday that in waterborne epidemics that we have
seen—Lenigrad, Rome, and other cities—there has been very little
secondary transmission on a person-to-person basis, and there is no
evidence for foodborne transmission, and we know that in children in
nurseries there is person-to-person transmission at quite a rapid rate, but in
adults we have seen very sparse evidence for this. I think only through
further epidemiologic work can we begin to answer these questions.
Then of course we have the whole human host-parasite relationship. A lot
of questions were raised during the conference on how it is manifested. Is it
increased virulence of some of the strains? Is it somehow a predisposition?
We know that there are a few factors like specific immunodeficiencies, but
there are probably many more that account for some people being infected
and other people not being infected. Also, what are the factors that account
for self-cure? It is not magic; it is some biologic phenomenon at work, and
we should research this. Most important of all, why do some people become
diseased and other people just become carriers? There are many other
questions to be asked but I think 1 will stop at this point.
G. ROBECK: I would like to go on and ask Dr. Tom Vernon, the Deputy
Director of the Colorado Department of Health, to discuss some of these
same issues from the State Health Department point of view. Dr. Vernon
was educated at Duke, got his medical degree at Harvard Medical School,
and has spent at least 3 years with CDC as an epidemiologist, so he has had a
variety of experience and training.
T. VERNON: With Colorado playing a leading actor role in the last
several days, it has not always been comfortable to be here. A number of
people have wondered why Colorado has taken a key role, and I think it is
useful, at least epidemiologically, to understand some background to the
Colorado situation.
As is very clear from the days of Dr. William Foege, now Director of CDC
and once an epidemiologist in the Colorado Department of Health, in his
studies of hepatitis, the attack rates were greater in the high altitude
mountain counties than they were in the plains counties, my supposition
-------
292 WATERBORNE GIARDIASIS/RESEARCH NEEDS
being that water treatment was less adequate. Our smaller towns in the past
have depended more on basic purity of the water and less on treatment
systems.
Many of us have been misled into thinking that the Coors beer slogan
"pure Rocky Mountain spring water" actually means pure Rocky Mountain
stream water. Backpackers tend to subscribe to this and drink the surface
water, and I think such people in other parts of the country as in the
Cascades and the Appalachians, do not drink stream water nearly as much.
So the exposure is clearly there in Colorado.
There has been a sensitization in Colorado toward Giardia since the
Aspen outbreak in 1966. We had an outbreak in '68, one in Golden in '69,
Wild Basin Lodge in '71, Winter Park in'71, Boulder was mentioned in'72,
the Grand Lake Lodge in '72 and '73, Lake City in '74, and Vail in '78.
Giardiasis has been of primary interest to the Colorado Department of
Health and our local health departments, and there is a great deal of
sensitization not only among the public but particularly among the
physicians in the state, so that endemic disease is also very widely
recognized.
The issue of the volume of people exposed is another point in the
Colorado history. Our population is now 2.6 million. It is larger by a wide
margin than any other Rocky Mountain state and is growing by 2.5% a year.
If you add to that large population, many of whom enjoy those backpacking
trips, the million or more tourists that we have, you can see why giardiasis is
a problem.
What are the research needs as we see them? We have told you that in
assuming primacy with the Safe Drinking Water Act, we have in our
regulations now a requirement for complete treatment of surface waters.
That is the phraseology used, and of course this will include for us more than
filtration. To meet this requirement additions must be made for 19
municipal water supplies, but that is not the main problem. We are
concerned about many smaller, non-municipal water supplies.
We have to think first about the source. What characterizes one surface
water source from another? What are the animal species present? We want
to continue the work that Hibler and Davies began. What detector systems
could we develop, or indicator systems, as Dr. Hendricks used the phrase
this morning, to tell us whether or not a given surface water is a problem?
What is the relationship of surface water to some so-called ground water
sources which are actually only surface water passed through poorly
filtering soils? What is the human habitation or recreational use of a given
watershed?
What are the seasonal changes in Giardia from the water source? What is
the relationship to flow? As you might guess, we have a tremendously
variable flow seasonally and also from year to year. The difference between
1977 and 1978 in total snow pack in the Rocky Mountains was tremendous
such that our flows through late 1977 were way down. We need to look at the
treatment systems. Everyone has addressed this as to effectiveness for our
-------
G. G. Robeck 293
municipalities, but we also need to check our non-municipal public water
supplies, the motels and lodges which take water out of the Rio Grande and
the Colorado Rivers and so forth. What are our choices of filtration media? I
was very well educated yesterday on diatomaceaous earth, and I think we
need to know more, a great deal more, about what works, what combination
of filtering media could be put together and what is the best means of
coagulation. What is the base line level of giardiasis we can expect in our
population which cannot be tracked to direct contamination of raw water?
And then of immediate importance and seemingly quite easy to determine,
what can we recommend as a disinfectant to our backpackers?
1 experienced the end of my Flagyl course this morning, left over from a
backpacking trip six weeks ago. I realized either my 2% tincture of iodine
did not work at all or I did not let it contact long enough, or I drank from
what appeared to be a particularly clear stream. At any rate, we are
recommending methods for disinfection to the backpacker in which we have
relatively little confidence.
G. ROBECK: We appreciate the personal as well as the official experience
and point of view. It brings the Symposium as I said initially, a little more
into focus.
To help us dwell a little more on the technology, we have with us Dr. Kaz
Kawata who comes to us from Johns Hopkins University where he is the
Acting Director of the Division of Environmental Health Engineering. He
received his education in various institutions in the U.S. and experience in
various places through the world. He was trained in Oregon as a civil
engineer, moved on to Minneapolis and Berkeley for more training and
received his Ph.D at Johns Hopkins. He comes to us with a great deal of
experience in environmental health in the tropics dealing with exotic
diseases, which to a certain extent some people think giardiasis to be.
K. KAWATA: 1 would like to take this assignment from the point of view
of the environmental health engineering consultant, because we work in the
area of environmental health engineering and there are some questions that
I would like to have answered. One of the main answers that I would like to
have is how big is the problem if there really is a problem, or if there is a
potential problem, what do we have to do as sanitary engineers and
environmental health engineers in terms of the treatment of water for the
consuming public, or what do we have to do, also, if there are that many
cysts being excreted? Chandler & Read (Introduction to Parasitology, John
Wiley & Sons, New York, 1961) indicate that one stool from a moderate
infection may produce as many as 300 million cysts. Now, if that many cysts
from infected people come down to a sewage treatment plant, then is there a
problem? What kind of treament is required not only for the liquid effluent
that is going out, but also for the sludge?
The American Public Health Association's Handbook on Control of
Communicable Diseases in Man(A.S. Benenson, 1975) gives a figure on the
prevalence of infection in the neighborhood of 1.5 to 20%. It is said that
prevalence is dependent upon the community and the population involved.
-------
294 WATERBORNE GIARDIASIS/RESEARCH NEEDS
If the infection rate is that high, then why isn't there more disease in the
community? The answer that was given during the symposium is that
perhaps this is an organism that infects many but causes disease in few. If
that is the case, what are the host factors that are involved; what is an
infective dose? In terms of treatment processes that we have to design, we
have to take into consideration the infective dose. Are the factors within the
etiologic agent itself? During this Symposium, the biological difference in
organisms from different donors was mentioned. We would like to know
what those important differences are. It is hard to believe that giardiasis is
not prevalent in other areas of the United States besides the mountain areas.
What about migrant labor camps?
It is also hard to believe that giardiasis is not foodborne. Perhaps a study
into foodborne transmission of giardiasis might be done in a country like
India. When I was assigned to India, we had to take our own toilet paper
there. It was very difficult in the early 1950's to buy toilet paper in India.
Many people in India do not use toilet paper; they wash their perianal
regions with their left hand. They take a little bit of water with them when
they go to defecate out in the field. We hired some of those people to be our
cooks, and we had giardiasis when we were in India. Chandler and Reed
stated that there are quite a number of invertebrates that harbor this
intestinal flagellate. They said that you can find them invertebrates ranging
from fish to man. Evidently there is much more to be done, much more to be
learned in terms of what animals may harbor the organisms that may infect
man.
Now, as to control, there is really very little information in the literature
for sanitary engineers or designers on treatment processes. We need to know
a lot more about the removal of the organism by coagulation, flocculation
and sedimentation, and of course also by filtration. In a rapid sand filter
plant the bulk of the work is done ahead of the filter, so we need to look at
those processes. As for filtration of unconditioned or preconditioned watej,
the designer needs to have data on the efficiency of cyst removal for varying
parameters of sand size and uniformity coefficient, temperature and
filtration rate. In doing this, we run into the problem of recovery and
enumeration that W. Jakubowski talked about in this conference. We
certainly should have a better method for enumeration and recovery.
At Johns Hopkins we have been involved in disinfection studies for a
number of years. It would be nice to have a good reliable in vitro cultivation
technique for Giardia so that we could harvest cysts from culture and then
subject them to all kinds of insults with disinfectants. Then, after hurling
these insults on the organism, we need to know how to tell a live cyst from a
dead cyst. Should we use excystation as proof of survival? If we are going to
use excystation, then we want to know the relationship between excystation
and viability. We also want to know the relationship between viability and
infectivity. We can also go to harvesting cysts from human donors, but
getting a good human donor is difficult. We at Hopkins have a pipe line to
Saudi Arabia, and we are importing human feces to do our tests. We also
-------
G. G. Robeck 295
need a good technique for cleaning the cysts so that we can do these tests in a
controlled manner.
Having taken care of those logistics, then we will want to evaluate a
variety of disinfectants on Giardia that infect man. Of course, it was
mentioned in the Symposium that there is a need to evaluate the coliform
group and other organisms as indicators for Giardia. Perhaps the military
might go back and evaluate water purification tablets for Giardia
inactivation. Then, having taken care of the problem under the clean system,
we can move into the dirty system and run similar tests in sewage and in raw
and digested sludges. We may also raise the nagging questions of the
possible health effects when wastewater and sludge reuses are
comtemplated.
There are certainly other areas of concern, including medical
management, that require better answers, but perhaps this will suffice for
now.
G. ROBECK: Thank you, Dr. Kawata, for your perspective. We now
have a few minutes for any further suggestions for research as seen from the
point of view of anyone in the audience. Please feel free to express either
your questions to the panelists or comments of your own about future needs
in research.
S. ERLANDSEN: 1 have a question for Mr. Davies. It is very interesting
to talk about the possiblity of infecting animals with G. lamblia. I wonder if
you have carried that work one step further and shown from these same
animals— and I guess the rat would be a very interesting model here since
supposedly one cannot transmit Giardia from rats into humans—if you
have those animals and they supposedly are carrying G. lamblia, can you
take those cysts and put them into a primate and grow them?
R. DAVIES: I do not know about putting them back into primates, as we
do not have any of those to work with. We are using SPF beagle puppies as a
human surrogate, and we feel that anything that will infect humans will
infect these dogs. We can take the rat Giardia and put it into beagle pups, but
that is about as far as we have gone.
I believe that beaver are an extremely good reservoir and a continuing
source of Giardia in watersheds, but I think we also should look at these as a
good surveillance animal. In three different watersheds in Colorado I have
gone up to a certain point and all of the beaver are infected near these beaver
dams or ponds, but upstream they are clean. One point was Ashcroft, where
I had reports of dog feces being washed into the river. Another place was on
Beaver Creek, where all of the beaver were infected near two summer homes,
but above that point none were infected. On Willow Creek, as long as the
river was alongside campgrounds, beaver were positive for Giardia. Once
there were no more campgrounds with heavy human use alongside that
river, beaver were once again negative. On the Frazier River beaver were
negative above the sewage plant and positive below the plant. Beaver may be
a good sentinel animal for surveillance work.
G. ROBECK: Do you find Giardia in the effluents from the sewage
discharges?
-------
296 WATERBORNE GIARDIASIS/RESEARC H NEEDS
R. DAVIES: We have not tried to examine that. The beaver ponds were
approximately 50 yd below the point where the effluent left the sewage
plant.
S. ERLANDSEN: For those of us who are not fortunate enough to live
close to beaver ponds but live in cities that have big storm drains and sewage
areas, if we are going to consider the rat as a possible intermediate for cyst
transmission, and rodents, muskrats and the like, how much contribution
do they make to the cyst content of sewage versus the human source itself?
G. ROBECK: Do any of the panelists have any comment on this?
L. McCABE: We have a project in Washington state that is trying three
techniques of surveillance of water supplies by filters, animal surveillance of
each watershed, and through state laboratory reports. So the point of using
animals as an indicator of how bad the watershed might be is being
evaluated.
D. PRICE: In response to Dr. Kawata's comments, we are working with
migrant communities in Florida. We are finding Giardia rates of up to 70%
in migrant childcare centers, and in the general population it is running
sornewhere between 12 and 17% and in day care centers in Lakeland and
Tampa, we are running about 8 or 10%. Overall, Florida is reporting a
Giardia stool positivity rate of about 3.8% in the total population.
I have a comment for Mr. McCabe regarding the use of sewage. This is
used in a number of places for golf courses and gardens, and one thing we are
finding is that Ascaris lumbricoides and Toxocara from dogs are viable for
years after this material has been passed unless it has been heated or is in a
place where it gets enough sunlight to destroy the eggs. We have not s"een
Giardia at all in any of this material out in the field.
F. FROST: Our data on animal surveys would support Mr. Davies'
findings. We have about 60 samples from protected watersheds, all of which
were negative animal samples. We have about a 33% positive rate in
muskrat, and we have beaver throughout the state with a 10% prevalency, so
I think our data are in support of the suggestion of how people are acquiring
it.
D. DEGIUSTl: I would like to know whether there are any ongoing
programs on examination of the domestic rat for Giardia. There are
resources available throughout the entire country, and I think that one of
the problems is coordination. For example, in Detroit, we are doing a survey
of Leptospira in rats. It never occurred to me to examine these same rats for
Giardia. If there are ongoing programs, I think we ought to have some effort
at coordination to utilize those resources.
G. ROBECK: Certainly we in EPA are always obligated to look to other
agencies, local, state or federal, to develop the background information that
we need to carry out our responsibilities and we are trying to do that from
existing sources, but we want to be able to know how to use our limited
resources to best advantage. Mr. McCable or Dr. Schultz, you might want
to respond to that as you see it now or in the future.
M. SCHULTZ: I do not know of any studies presently on Giardia in rats.
It would be interesting to study but then one would inevitably ask what does
it mean.
R. OWEN: When I was at CDC we were sensitized to the problem of
treating asymptomatic Shigella and Salmonella because of the potential for
-------
6. 6. Robeck 297
developing drug resistant organisms. I wonder if we treat asymptomatic
Giardia infections if we are not going to come up with the same problem.
M. SCHULTZ: I do not think there has been any evidence for resistance
developing against the anti-giardial drugs. Dr. Wolfe has had the most
experience. There is resistance in malaria protozoa, but I cannot think of
any for this.
M. WOLFE: Metronidazole has been used in amebiasis and for
Trichomonas for many years, and I do not know of any evidence of
resistance in people. Lower doses seem to be effective in Trichomonas. At
least analogizing to other protozoan parasites, there is no evidence of
resistance so far.
G. ROBECK: It would be fair to say that there was a lot of coordinating to
do for this conference. I think it has been reasonably successful thanks to
Walt Jakubowski and John Hoff, and to the various federal, state and local
agencies, that have made their technical and scientific contributions. The
Proceedings should serve as a good basis for the state of our knowledge and
where we need to go with our regulatory actions as well as research.
-------
298
REGISTRATION LIST
ALLISON, Betty K.
Environmental Monitoring & Support
Laboratory
Environmental Protection Agency
26 W. St. Clair Street
Cincinnati, Ohio 45268
AMUSO, Philip T.
Department of Health, E.R.C.
State of Florida
4000 West Buffalo Avenue
Tampa, Florida 33614
ANDERSON, Ferron L.
Professor of Zoology
Brigham Young University
197 W1DB
Provo, Utah 84602
ANDREWS, John S., Jr., M.D.
Geographic Medicine Fellow
University Hospitals of Cleveland
2065 Adelbert Road
Cleveland, Ohio 44106
ARIZUM1, Thomas E.
Department of Health
State of Hawaii
1250 Punchbowl Street
Honolulu, Hawaii 96813
ARLIAN, Larry
Department of Biological Sciences
Wright State University
Dayton, Ohio 45435
AROTIN, Barbara M.
Philadelphia Suburban Water Company
762 Lancaster Avenue
Bryn Mawr, Pennsylvania 19010
AROZARENA, Michael M.
Municipal Environmental Research
Laboratory
Environmental Protection Agency
26 W. St. Clair Street
Cincinnati, Ohio 45268
AVEDOVECH, Richard M.
Department of Environmental Quality
1712 S.W. llth Avenue
Portland, Oregon 97201
BA1RD, Justus, R.S., R.P.E.
Communicable Disease Division
Health Department
City of Houston
1115 N. MacGregor
Houston, Texas 77030
BALLWEG, Robert F.
Associate Professor of Biology
Long Island University
University Plaza
Brooklyn, New York 11201
BARNHART, Donald
Mercy Hospital
116 Dayton Street
Hamilton, Ohio 45011
BARTLETT, Marilyn S.
Assistant Professor
Indiana University
University Hospital #440N
Indianapolis, Indiana 46202
BENNETT, Dorothy L.
Illinois Environmental Protection
Agency
2200 Churchill Road
Springfield, Illinois 62706
BERGER, Paul
Environmental Protection Agency
WH-550
Waterside Mall Building
401 M Street, S.W.
Washington, D.C. 20460
BERRY, James W.
Fairfax County Water Authority
P. O. Box 151
Occoquan, Virginia 22125
BINGHAM, Alan K.
University of Oregon Health Sciences
Center
3181 S.W. Sam Jackson Park Road
Portland, Oregon 97201
BOX, Edith D , Sc.D.
University of Texas Medical Branch
Department of Microbiology
Galveston, Texas 77550
BRAK.NAGE, Brad
State Health Laboratory
3701 South 14th
Lincoln, Nebraska 68502
BREWER, William S.
Department of Biological Sciences
Wright State University
Dayton, Ohio 45435
-------
Registration List
299
BREZENSKI, Francis T.
Laboratory Director
Environmental Protection Agency
Woodbridge Avenue, Bldg. 209
Edison, New Jersey 08817
BRODINE, Charles E., M.D.
Assistant Medical Director for
Environmental Health & Preventive
Medicine
Medical Services, Dept. of State
2201 C. Street, N.W , Room 2906
Washington, D.C. 20520
BRUCE, Arthur
Connecticut State Department of Health
P. O. Box 1689
Hartford, Connecticut 06109
BURDICK, Robert L.
Onondaga County Health
421 Montgomery Street
Syracuse, New York 13201
CARL1N, Thomas P.
Philadelphia Water Department
Quality Control
4290 Ford Road
Philadelphia, Pennsylvania 19131
CHAPMAN, James
Water Treatment Supervisor
Seattle Water Department
1509 So. Spokane Street
Seattle, Washington 98144
CHAUSSEE, Dean R.
Water Supply Engineer
Environmental Protection Agency
1860 Lincoln Street
Denver, Colorado 80295
CHEHEY, Robert L
Supervisor, Microbiology
Department of Health & Welfare,
Lab Bureau
2220 Old Penitentiary Road
Boise, Idaho 83702
CLARK, Judith G.
University of Cincinnati, RWC
4285 Hutchinson Road
Cincinnati, Ohio 45211
CLARKE, Norman A., Dr.
Director, Laboratory Studies Division
Health Effects Research Laboratory
Environmental Protection Agency
26 W. St. Clair Street
Cincinnati, Ohio 45268
CLARKSON, Arthur W.
Sanitary Engineer
Water Quality Bureau
Department of Health &
Environmental Sciences
Helena, Montana 59601
COFFMAN, Robert L., MPH
Chief, Environmental Surveillance
& Control Services
Division of Health & Medical Services
Hathaway Building, Room 468
Cheyenne, Wyoming 82002
COLLINS, Charles E.
Georgia Environmental Protection
Division, Room 817
270 Washington Street
Atlanta, Georgia 30334
CONDER, George A.
Research Associate
Bngham Young University
123 W1DB
Provo, Utah 84057
CONNELL, Alastair M., M.D.
Professor of Medicine
University of Cincinnati
Medical Center
231 Bethesda Avenue
Cincinnati, Ohio 45267
COOPER, Karen
Division of Water Quality
Ohio Environmental Protection Agency
361 E. Broad Street
Columbus, Ohio 43215
CRAUN, Gunther F.
Epidemiology Branch
Health Effects Research Laboratory
Environmental Protection Agency
26 W. St. Clair Street
Cincinnati, Ohio 45268
CROFT, C. C., Sc.D.
Chief, Division of Public Health
Laboratory
Ohio Department of Health
P O. Box 2568
Columbus, Ohio 43216
CYPESS, Raymond H., DVM, Ph.D.
Department of Preventive Medicine
New York State College of Veterinary
Medicine
Cornell University
P O. Box 786
Ithaca, New York 14850
DAHLING, Daniel R.
Environmental Monitoring and Support
Laboratory
Environmental Protection Agency
26 W. St. Clair Street
Cincinnati, Ohio 45268
DAVIES, Robert B.
College of Veterinary Medicine
Colorado State University
Fort Collins, Colorado 80523
-------
300
WATERBORNE GIARDIASIS
DEGAETANO, Thomas J.
Sanitary Engineer
Environmental Protection Agency
26 Federal Plaza
New York City, New York 10007
DEG1USTI, Dianna
Wayne State University School of
Medicine
Department of Comparative Medicine
Detroit, Michigan 48201
DIAMOND, Louis S, Dr
Head, Section of Parasite Growth &
Differentiation
Laboratory of Parasitic Diseases
National Institutes of Health
Bethesda, Maryland 20014
DOUGHERTY, Ellen M.
State Parasitology Laboratory
Georgia Department of Human Resources
47 Trinity Avenue
Atlanta, Georgia 30334
DRAPER, Alfred
Wright State University
4984 Depay Street
Dayton, Ohio 45424
EDELL, Timm A., M.D.
Acting State Epidemiologist
Colorado Department of Health
4210 East llth Avenue
Denver, Colorado 80220
ENGEL, John
Professor of Civil Engineering
Ohio State University
Department of Civil Engineering
2070 Neil Avenue
Columbus, Ohio 43210
ENGESET, Joe
College of Engineering
University of Washington
369 Loew Hall
Seattle, Washington 98195
ERICKSEN, Theodore H.
Health Effects Research Laboratory
Environmental Protection Agency
26 W. St. Clair Street
Cincinnati, Ohio 45268
ERLANDSEN, Stanley L., Dr.
Department of Anatomy
University of Minnesota
Minneapolis, Minnesota 55455
FOX, Joseph Carl, Dr.
College of Veterinary Medicine
Department of Parasitology,
Microbiology & Public Health
Oklahoma State University
Stillwater, Oklahoma 74074
FR-EEMAN, Remo S., Dr.
Department of Microbiology &
Parasitology
Fitzgerald Building
University of Toronto
Toronto, Ontario MSSiAt
Canada
FROST, Floyd
Research Investigator
Department of Social & Health
Services
Environmental Health Programs
MS LD-ll
Olympia, Washington 98504
GARNER, R. John, Dr.
Director, Health Effects
Research Laboratory
Environmental Protection Agency
26 W St. Clair Street
Cincinnati, Ohio 45268
GELDREICH, Edwin E.
Research Microbiologist
Water Supply Research
Environmental Protection Agency
26 W. St. Clair Street
Cincinnati, Ohio 45268
GENTRY, Ralph E
Microbiologist
Environmental Protection Agency
College Station Road
Athens, Georgia 30605
G1LLIN, Frances D., Dr
Senior Staff Fellow
Laboratory of Parasitic Diseases
Building 5, Room 207 - NIH
Bethesda, Maryland 20014
GOLDBERG, Stephen J., M.D.
Assistant Professor of Medicine
University of Cincinnati Medical Center
231 Bethesda Avenue
Cincinnati, Ohio 45267
GOFF, David
1809 Park Road, N.W.
Washington, DC. 20010
GOLDSBY, Alice 1.
University of Wisconsin Green Bay
Green Bay, Wisconsin 54301
GOLTZ, James
Doctor of Veterinary Medicine
Department of Pathology, O.V.C.
University of Guelph
Guelph, Ontario N1G2W1
Canada
GOO, Velma
Microbiologist
Hawaii State Department of Health
P. O. Box 3378
Honolulu, Hawaii 96801
-------
Registration List
301
GRAHAM, Hugh
The Ministry of the Environment
Water Technology Section
135 St. Clair Avenue, W
Ontario, Canada
GRASSMICK, Robert
Associate Professor of Zoology
Department of Zoology
Miami University
Oxford, Ohio 45056
GRAY, Thomas
Technical Services Chief
Bureau of Water Supply Engineering
109 Governor Street
Richmond, Virginia 23219
GREENE, Lyford K., CPT MSC
Chief, Schistosomiasis Section
Department of Parasitology
Division of Expl. Therap.
WRAIR, WRAMC
Washington, D.C. 20012
GRE1NER, Donald J., P.E
Sanitary Engineer
Michigan Department of Public Health
3500 N. Logan Street
P. O, Box 30035
Lansing, Michigan 48909
GUTIERREZ, Yezid, M.D., Ph. D
Institute of Pathology
Case Western Reserve University
Cleveland, Ohio 44106
HABRAKEN, J. W
Operations Supervisor
Akron Water Plant
1570 Ravenna Road
Kent, Ohio 44240
HALPERN, Bernard, Dr.
Assistant Professor of Microbiology
Northwestern University Medical
School - 3-607
303 E. Chicago Avenue
Chicago, Illinois 60611
HEALY, George R , Dr.
Chief, General Parasitology Branch
Center for Disease Control
1600 Clifton Road
Atlanta, Georgia 30333
HENDRICKS, Charles W., Dr.
Environmental Protection Agency
WH-550
Waterside Mall Building
401 M Street, S.W.
Washington, D.C. 20460
HERLIHY, John J.
Laboratory Supervisor
Bridgeport Hydryloc Company
835 Mam Street
Bridgeport, Connecticut 06609
HEWLETT, Erik, Dr.
Assistant Professor of Medicine
Division of Geographic Medicine
Wearn Building
University Hospitals
Cleveland, Ohio 44106
HIBBERT, Larry E., Dr
Professor of Biology
Department of Biology
Ricks College
Rexburg, Idaho 83440
H1BLER, Charles P.
Director, Wild Animal Disease Center
College of Veterinary Medicine
Colorado State University
Fort Collins, Colorado 80523
HOFF, John C , Dr
Water Supply Division
Municipal Environmental Research
Laboratory
Environmental Protection Agency
26 W St. Clair Street
Cincinnati, Ohio 45268
HOPKINS, Richard S., M.D.
E1S Officer
CDC/Montana Department of Health
& Environmental Sciences
W F Cogswell Building
Helena, Montana 59601
HOWELL, Robert T., Dr.
Public Health Laboratory Administrator
Arkansas Department of Health
4815 W. Markham
Little Rock, Arkansas 72201
HOYE, Robert
1545 Northview
Cincinnati, Ohio 45223
HULDT, Gunnel, Dr.
ICarolmska Institute
Stockholm, Sweden
JACKSON, George J., Dr.
Food Parasitology Group
DHEW, FDA
200 C. Street, S.W.
Washington, D.C 20204
JAKUBOWSKI, Walter
Epidemiology Branch
Health Effects Research Laboratory
Environmental Protection Agency
26 W. St. Clair Street
Cincinnati, Ohio 45268
JARROLL, Edward L,., Jr., Dr.
Fellow in Microbiology
University of Oregon Health
Sciences Center
3181 S.W. Sam Jackson Park Road
Portland, Oregon 97201
-------
302
WATERBORNE GIARDIASIS
JENSEN, Michael
Environmental Sanitary Consultant
National Park Service
P. O. Box 25287
Denver, Colorado 80225
JOHNSON, Helen M.
Physical Science Technician,
Microbiology
Environmental Protection Agency
620 Central Avenue, Building 1
Alameda, California 94501
JONES, Wesley E., M.D.
Medical Epidemiologist
Parasitic Disease Division
Bureau of Epidemiology
Center for Disease Control
Atlanta, Georgia 30333
JURANEK, Dennis, Dr.
Parasitic Diseases Division
Bureau of Epidemiology
Center fo Disease Control
1600 Clifton Road, N.E.
Atlanta, Georgia 30333
KATAMAY, Michael
Laboratory Manager
Illinois Environmental Protection
Agency
2121 W. Taylor
Chicago, Illinois 60612
KAWATA, Kaz
Associate Professor
The Johns Hopkins University
615 N. Wolfe Street
Baltimore, Maryland 21205
KENNEDY, Harriet
Laboratory Director
Evanston-North Shore Health
Department, Box 870
1806 Maple Avenue
Evanston, Illinois 60204
KILLGORE, George E., Dr. P.H
Manager, Microbiology Branch
Kentucky Bureau for Health Services
275 E. Mam Street
Frankfort, Kentucky 40601
KISPERT, Edward C.
Supervisor of Treatment
Cincinnati Water Works
5651 Kellogg Avenue
Cincinnati, Ohio 45228
KNIPPER, John
Water Bacteriologist
Microbiology Unit
Chicago Water Purification Laboratory
1000 E. Ohio Street
Chicago, Illinois 60611
LEONARD, Steven D.
Senior Microbiologist
San Francisco Water Department
P. O. Box 367
Millbrae, California 94030
LEONG, Lawrence Y.C.
Laboratory Director
James M. Montgomery, Engineers
555 E. Walnut Street
Pasadena, California 91101
LEV1NE, Norman D
University of Illinois
College of Veterinary Medicine
Urbana, Illinois 61801
L1ECHTY, Bill
Special Projects Engineer
Water Supply and Waste Section
Mail Stop LD-11
Olympia, Washington 98504
LINDMARK, Donald G.
Rockefeller University
1230 York Avenue
New York, New York 10021
LIPPY, Edwin C.
Epidemiology Branch
Health Effects Research Laboratory
Environmental Protection Agency
26 W. St. Clair Street
Cincinnati, Ohio 45268
LOGSDON, Gary S., Dr.
Water Supply Division
Municipal Environmental Research
Laboratory
Environmental Protection Agency
26. W. St. Clair Street
Cincinnati, Ohio 45268
LUCAS, James B., Dr
Deputy Director
Health Effects Research Laboratory
Environmental Protection Agency
26 W. St Clair Street
Cincinnati, Ohio 45268
MACA1G, Terence
Chief Microbiologist
Vermont State Public Health Laboratory
115 Colchester Avenue
Burlington, Vermont 05401
MCCLANAHAN, Mark
345 Courtlant Street
Atlanta, Georgia 30308
MAC CABE, B. S.
Business Devlopment Manager
Carborundum Company
P. O. Box 1054
Niagara Falls, New York 14302
-------
Registration List
303
MANHART, John P.
Regional Microbiologist
Environmental Protection Agency
1860 Lincoln Street, Suite 900
Denver, Colorado 80295
MARDIN1, M. Mazen, M D.
Fellow in Digestive Diseases
University of Cincinnati, Medical
Center
231 Bethesda Avenue
Cincinnati, Ohio 45267
MAYNARD, Everett L
Senior Sanitary Engineer
U.S. Public Health Service
Environmental Protection Agency
Muschopauge Road
Jefferson, Massachusetts 01522
MAYO, Francis T.
Director, Municipal Environmental
Research Laboratory
Environmental Protection Agency
26 W. St. Clair Street
Cincinnati, Ohio 45268
MCCABE, Leland J.
Director, Field Studies Division
Health Effects Research Laboratory
Environmental Protection Agency
26 W. St. Clair Street
Cincinnati, Ohio 45268
MEYER, Ernest A., Dr.
Professor of Microbiology
University of Oregon Health Sciences
Center
3181 S.W. Sam Jackson Park Road
Portland, Oregon 97201
MILLER, Donald M.
Professor of Physiology
Department of Physiology
Southern Illinois University
Carbondale, Illinois 62901
MITCHELL, Bert W
Engineering Consultant
Public Health Service
P. O. Box 728
Santa Fe, New Mexico 87501
MORGAN. Stephen E
Microbiologist III
State Board of Health
1330 W. Michigan
Indianapolis, Indiana 46202
MORRISON, N. Don
Sanitarian Assistant
Tuolumne County Health Department
P. O. Box 1115
Pinecrest, California 95364
MOSER, Richard H.
Assistant Director, System Water
Quality
American Water Works Service Company
500 Grove Street
Haddon Heights, New Jersey 08035
MURPHY, Dennis J.
Environmental Health Office
Box 100
Vail, Colorado 81657
MURPHY, Tom
JFK Federal Building
Environmental Protection Agency
Boston, Massachusetts 02203
MYER, Donal G.
Southern Illinois University
Department of Biological Sciences
Edwardsville, Illinois 62026
MYERS, Betty June, Ph D.
Health Science Administrator
National Institutes of Health
DRG/TMP Westwood Building, Room 319
5333 Westbard Avenue
Bethesda, Maryland 20014
NASH, Theodore E
Medical Officer
Laboratory Parasitic Diseases
National Institutes of Health
Building 5, Room 114
Bethesda, Maryland 20014
NELSEN, Brian A.
Microbiologist
Denver Water Board
1600 W. 12th Avenue
Denver, Colorado 80254
NEMANIC, Paulina, Ph.D.
Cell Biologist
Veterans Administration Hospital
4150 Clement Street
San Francisco, California 94121
NEWTON, Walter L , Dr.
4709 Pickett Road
Fairfax, Virginia 22032
NICHOLS, Craig
Public Health Epidemiologist
Utah State Division of Health
150 West No. Temple
P. O. Box 2500
Salt Lake City, Utah 84110
ODA, Albert 1.
Chief, Laboratories Branch
Hawaii State Department of Health
P. O. Box 3378
Honolulu, Hawaii 96801
OLIVIERI, Vincent P., Dr.
The Johns Hopkins University
615 N. Wolfe Street
Baltimore, Maryland 21205
-------
304
WATERBORNE GIARDIASIS
OWEN, Robert L., M D.
Research Associate
Cell Biology Section (15IE)
Veterans Administration Hospital
San Francisco, California 94121
PAHREN, Herbert R
Health Effects Research Laboratory
Environmental Protection Agency
26 W. St Clair Street
Cincinnati, Ohio 45268
PARR, Thomas R.
Senior Chemist
Water & Gas Department Lab
8130 Congdon Boulevard
Duluth, Minnesota 55804
PAUL, Les
Regional Environmental Engineer
U.S Forest Service
324 25th Street
Ogden, Utah 84401
PETTY, Bill
Environmental Specialist
City of Houston
1220 Dallas
Houston, Texas 77002
PHIFER, Kenneth, Dr.
Parasitology Program Officer
National Institutes of Health
Westwood Building, Room 737
Bethesda, Maryland 20014
PIPES, Wesley O.
Professor of Biological Sciences
Drexel University
32nd & Chestnut
Philadelphia, Pennsylvania 19104
PLAN, Byron
Microbiologist
Department of Social & Health
Services
Environmental Health Programs MS LC
Olympia, Washington 98504
POWER, Joe Alan
Assistant Director
State Health Department
State Office Building
Montgomery, Alabama 36130
PRICE, Donald L., Ph. D.
Director, Multidisciphnary Services
and Research Institute
University of Sarasota
2080 Rmgling Boulevard
Sarasota, Florida 33577
RENDTORFF, Robert C, Dr.
University of Tennessee
Health Sciences Center
800 Madison Avenue
Memphis, Tennessee 38163
RES1, Louis A.
Microbiologist
Technical Support division
Environmental Protection Agency
5555 Ridge Avenue
Cincinnati, Ohio 45268
RICE, Eugene, W.
Municipal Environmental Research
Laboratory
Envnonmental Protection Agency
26 W St Clair Street
Cincinnati, Ohio 45268
ROBECK, Gordon G.
Director, Water Supply Research
Division
Municipal Environmental Research
Laboratory
Environmental Protection Agency
26 W. St Clair Street
Cincinnati, Ohio 45268
ROBINSON, Pamela A
Administrator, Office of Public
Water Supply
Ohio Environmental Protection Agency
361 E Broad Street, Box 1049
Columbus, Ohio 43216
SARNOSKI, Bernie
Chief, State Program Section -
Water Supply
Environmental Protection Agency
6th & Walnut Street, Curtis Bldg.
Philadelphia, Pennsylvania 19106
SCARPING, Pasquale V., Dr.
College of Engineering
639 Baldwin Hall
University of Cincinnati
Cincinnati, Ohio 45267
SCHAEFER, Frank W., Dr.
Health Effects Research Laboratory
Environmental Protection Agency
26. W. St. Clair Street
Cincinnati, Ohio 45268
SCHILL1NGER, John E.
Department of Environmental Ind. Health
School of Public Health
University of Michigan
Ann Arbor, Michigan 48189
SCHLAUTMAN, Yvonne
Assistant Chemist
Metropolitan Utilities District
1723 Harney Street
Omaha, Nebraska 68102
SCHULTZ, Myron G., D.V.M., M.D.
Director, Parasitic Diseases Division
Bureau of Epidemiology, Center for
Disease Control
Atlanta, Georgia 30333
-------
V°°'
-^
<\e*
-------
c
f
f
£i
J7
4/e
Pa,
>
^
0
**,
^
^e
S
•• OA c
^
S:?
%:^
fc
-------
Registration List
305
SEDAM, Jason K.
Principal Engineer
The Coca-Cola Company
P. O Drawer 1734
Atlanta, Georgia 30301
SHEFFIELD, H. G Dr
Project Review Branch, EAP
NIAID, NIH, Westwood Building
Bethesda, Maryland 20014
SIDNER, Richard A
University of Cincinnati
821 A-l Broder Science Complex
Cincinnati, Ohio 45221
SINCLAIR, Susanne P , M S
Microbiologist
Parasitic Diseases Division
Bureau of Epidemiology
Center for Disease Control
Atlanta, Georgia 30333
SMADES, Roger H
Section Chief
Colorado Department of Health
4210 E llth Avenue
Denver, Colorado 80228
SMITH, Arthur W
Laboratory Supervisor
City of Portland, Water Bureau
1800 S.W. 6th Avenue
Portland, Oregon 97201
SMITH, Merlin A.
Acting Director & Chief Environmental
Microbiology Section
Utah State Bureau of Laboratories
44 Medical Drive
Salt Lake City, Utah 84113
SOUKUP, Albert V , P.E
Environmental Sanitation Consultant
National Park Service
U S. Public Health Service
655 Parfet, P O. Box 25287
Denver, Colorado 80225
STARK.E, Wayne, P
Emery Industries, Inc
Building 53
4900 Este Avenue
Cincinnati, Ohio 45232
STETLER, Ronald
Microbiologist
Health Effects Research Laboratory
Environmental Protection Agency
26 West St. Clair Street
Cincinnati, Ohio 45268
STEVENS, David P., Dr.
Department of Medicine
University Hospitals
Case Western Reserve University
Cleveland, Ohio 44106
STUART, Bo
Forest Hydrology Technician
U S Forest Service
P O Box 1888
Jackson, Wymong 83001
SUG1, Mitsuto
Epidemiological Specialist
State Health Department
1250 Punchbowl Street
Honolulu, Hawaii 96813
SYMONS, James M., Dr.
Municipal Environmental Research
Laboratory
Environmental Protection Agency
26 W St. Clair Street
Cincinnati, Ohio 45268
TAYLOR, J. S
Florida Technical University
Orlando, Florida 32816
TOMBES, Averett S., Dr
Chairman, Biology Department
4400 University Drive
Fairfax, Virginia 22030
TOPPAN, W. Clough
Manager, Drinking Water Program
Division of Health Engineering
Department of Human Services
Augusta, Maine 04333
VERNON, Thomas M., M.D.
Deputy Director
Colorado Department of Health
4310 E. llth Avenue
Denver, Colorado 80220
WAKE, Ernest H.
Director, Water Quality
California-American Water Company
2602 Hoover Street
National City, California 92050
WHEAT, Bruce E.
Research Assistant
College of Veterinary Medicine
University of Illinois
Urbana, Illinois 61801
WIDDUS, Roy, Dr.
Staff Officer
National Academy of Sciences
2101 Constitution Avenue, N W.
Washington, D.C. 20418
WILLIAMS, H. Douglas
Environmental Assessment Staff
Environmental Protection Agency
26 West St. Clair Street
Cincinnati, Ohio 45268
WILLIAMS, Owen R.
Hydrologist
U. S. Forest Service
P. O. Box 1366
301 S. Howes
Fort Collins, Colorado 80522
-------
306
WATERBORNE GIARDIASIS
WITHERELL, Linden E.
P. E., Chief
Vermont Water Supply
Environmental Protection Agency
P. O. Box 35
Essex Junction, Vermont 05468
WlTTE, Ernest J., V.M.D., MPH
Chief, Veterinary Public Health
Department of Health
P. O Box 90
Harnsburg, Pennsylvania 17120
WOLFE, Martin S., Dr
Office of Medical Services
Department of State
Washington, D.C. 20520
WOOD, Helen
Health Effects Research Laboratory
Enviromental Protection Agency
26 West St. Clair Street
Cincinnati, Ohio 45268
WOODFORD, James
Graduate Assistant
Wright State University
Dayton, Ohio 45435
YOUNG, Carolyn C.
Bacteriologist
New York State Department of
Health
Division of Labs. & Research
Environmenjal Health Center
Tower Building-ESP-Room B765
Albany, New York 12201
ZIMMERMAN, Joel H.
Microbiology Supervisor
North Dakota State Department
of Health
Box 1618
Bismarck, North Dakota 58505
------- |