Giardia Human Health Criteria Document


              United States             Office of Water      EPA-823-R-002
              Environmental Protection      4304            August 1998
              Agency
SEPA   GIARDIA: HUMAN HEALTH
               CRITERIA DOCUMENT

-------
                              ACKNOWLEDGMENTS

       This document was prepared for the U.S. Environmental Protection Agency, Office of
Ground Water and Drinking Water (OGWDW) by the Office of Science and Technology (OST)
under contract with Gunther F. Craun and Associates. Overall planning and management for the
preparation of this document was provided by Latisha Parker, of OST.

       EPA also recognizes the following external peer reviews for their excellent review and
valuable comments on the draft document: Carrie Hancock Ph.D., Walter Jakubowski M.S., and
Joan Rose Ph.D.

-------
TABLE OF CONTENTS
I. SUMMARY 1-1

II. GENERAL INFORMATION AND PROPERTIES
A. History and Taxonomy II-1
B. Life Cycle of Giardia 11-2
1. Excystation II-3
a. In vivo Excystation II-3
b. In vitro Excystation II-3
2. Encystation II-5
a. In vivo Encystation II-5
b. In vitro Encystation II-6
C. Morphological Features II-8
1. Trophozite II-9
2. Cyst II-9
D. Species Transmission 11-11
1. Direct Transmission Between Humans II-11
2. Transmission Between Animals and Humans 11-12
3. Transmission Between Animals 11-14
4. Summary of Cross-Species Transmission 11-14
E. Species Concept in the Genus Giardia 11-16
1. Filice's Concept 11-17
2. Grant and Woo's Concept II-l7
3. Other Concepts for the Speciation of Giardia II-l 8
F. Summary 11-19

HI. OCCURRENCE
A. Worldwide Distribution III-l
1. Distribution in Animal Populations III-l
2. Distribution in Human Populations III-2
B. Occurrence in Water III-4
1. Wastewaters III-4
2. Surface Waters III-7
3. Groundwaters 111-23
C. Occurrence in Soil 111-24
D. Occurrence in Air 111-25
E. Occurrence on Surfaces 111-25
F. Occurrence in Food 111-26
G. Disease Outbreaks and Endemic Risks 111-27
1. Outbreaks Associated with Drinking Water 111-27
a. Drinking Water Outbreaks in the United States 111-29
b. Waterborne Outbreaks in Canada 111-45
c. Waterborne Outbreaks in Europe III46

-------
2. Outbreaks Associated with Recreational Water 111-47
3. Outbreaks Associated with Other Water Sources III49
4. Endemic Waterborne Giardiasis III49
a. Drinking Water 111-49
b. Water Recreation and Other Water Sources 111-53
5. Foodborne Outbreaks 111-53
6. Travelers 111-56
7. Day-Care Centers 111-57
8. Sensitive Populations 111-60
H. Environmental Factors Affecting the Survival of Giardia cysts 111-61
1. Effects of Water Temperature on Giardia Cyst Survival 111-61
2. Other Factors that Affect Giardia Cyst Survival 111-63
I. Summary 111-64
1. Occurrence 111-64
2. Prevalence, Outbreaks, and Endemic Risks 111-66

IV. HEALTH EFFECT SIN ANIMALS
A. Symptomatology IV-1
B. Therapy IV-2
C. Epidemiological Data IV-3
D. Summary IV-5

V. HEALTH EFFECTS IN HUMANS
A. Symptoms and Clinical Features V-l
B. Epidemiology V-3
C. Clinical Laboratory Findings and Therapeutic Management V-6
1. Clinical Findings V-6
2. Therapeutic Treatment and Management V-7
D. Mechanism of Action V-13
E. Immunity V-l 6
1. Epidemiological Data Supporting Acquired Immunity V-l6
2. Breast Milk and Breast feeding Reduces Risk of Giardiasis V-18
3. Increased Giardiasis Risk in Immunosuppressed Populations V-19
4. Measuring Epidemic and Endemic Infections in Humans V-21
a. Anti-Giardia Antibodies in Sera V-22
b. Anti-Giardia Antibodies in Saliva V-24
c. Anti-Giardia Antibodies in Intestinal Secretions V-25
5. Mechanisms of Protection V-25
6. Summary of Evidence for Immunity V-26
F. Nonspecific Defenses Against Human Giardia V-21
G. Variation in Pathogenicity V-29
H. Chronic Conditions V-30
I. Summary V-31

-------
VI.    GIARDIA RISK ASSESSMENT
      A.     Risk Assessment Paradigms                                        VI-1
      B.     Health Effects                                                    VI-3
      C.     Dose-response Modeling                                           VT-6
      D.     Exposure Assessment                                             VI-8
      E.     Risk Characterization                                             VI-10
      F.     Risk Management and Federal Regulations                           VI-15

VII.   ANALYSIS AND TREATMENT OF GIARDIA
      A.     Analysis in Water                                                VII-1
             1.     Detection  and Identification Methods                          VII-1
             2.     Determination of Viability                                   VII-19
      B     Detection in Biological Samples                                    VII-24
      C.     Water Treatment Practices                                         VII-30
             1.     Filtration                                                  VII-31
                   a. Conventional and Direct Filtration                          VII-37
                   b. Slow Sand and Diatomaceous Filtration                     VII-41
                   c. Membrane and Other Filters                               VII-43
             2.     Disinfection                                               VII-44
      D.     Summary                                                        VII-52
             1.     Analysis                                                  VII-52
             2.     Water Treatment                                           VII-58
             3.
VHI.  RECOMMENDATIONS FOR RESEARCH                                VIII-1

IX.   REFERENCES                                                         IX-1

-------
I. SUMMARY

This document updates information in the U.S. Environmental Protection Agency's
(EPA) Drinking Water Criteria Document on Giardia (ICAIR, 1984) and is intended to serve as
an addendum to that report. Where appropriate, relevant information from the 1984 document is
summarized in each chapter of the addendum. For more a more detailed description of
information published before 1985, readers are referred to the 1984 document.

This chapter presents a summary of the information contained in Chapters II through VII
Each of these chapters also contain a more extensive summary section. Chapter VIII contains a
discussion of research recommendations and Chapter IX lists references.

E. Chapter II. General Information and Properties

Giardia is a protozoan parasite that has been identified as a important cause of
waterborne illness. The parasite is transmitted via the fecal-oral route of exposure, and both
endemic and epidemic giardiasis can occur. Ingestion of contaminated water is only one source
of infection, and the relative importance of waterborne transmission among other risk factors will
vary from place to place depending on general sanitation practices. In the United States,
contaminated water has caused a number of outbreaks and illnesses but is not likely the most
important mode of transmission. Giardia is a common cause of illness in travelers and is
frequently spread directly from person to person, especially among children or among persons in
areas with poor sanitation and hygiene. Although all age groups are affected, the highest
incidence is in children. Breast-fed infants under 6 months of age are not likely to be infected.

Large waterborne outbreaks have been reported, and illness has been associated with
ingestion of water from unfiltered surface water sources, shallow wells, and during water
recreational activities. Contaminated ice used in beverages and food and the person to person
1-6

-------
transmission in day-care centers have caused smaller outbreaks. Although infected children in
day-care centers are frequently asymptomatic, they may transmit giardiasis to other children, care
givers, and family members.

During 1965 to 1996, 133 waterborne outbreaks and almost 28,000 cases of giardiasis
have been reported in the United States, primarily in unfiltered surface water systems. Giardia
has been the most commonly identified pathogen in waterborne outbreaks reported in the United
States since 1971. Ten (8%) of these outbreaks were associated with the use of individual
drinking water systems or non-potable water sources, and 108 (81%) outbreaks were associated
with public water systems; 14 (11%) outbreaks were associated with accidental ingestionof
water during recreation. Unfiltered surface water systems were responsible for 56% of the
reported waterborne giardiasis outbreaks in the United States. Communities with unfiltered
surface water systems have experienced a waterborne outbreak rate that is eight times greater
than communities where surface water is both filtered and disinfected. Epidemiological studies
of endemic giardiasis have also reported high risks among persons using unfiltered surface water.
Based on these data, the 155 million people who continue to use unfiltered surface water in the
United States are at a higher risk for waterborne giardiasis than those who drink filtered surface
water.

Organisms in the genus Giardia are binucleate, flagellated protozoan parasites which
exist in trophozoite and cyst forms. While numerous species of Giardia have been described in a
variety of mammals and in lower vertebrates, there is no general agreement on the criteria which
define species in this genus. Criteria used to date include: host specificity; body size and shape,
and the morphology of a microtubular organelle, the median body; and biochemical, molecular,
and genetic techniques, such as the polymerase chain reaction (PCR) for DNA-based detection
and identification. The median body is an organelle that appears to be unique to Giardia
trophozoites. In this document, Giardia responsible for human infections will be found referred
to variously as G. duodenatis, G. intestinalis, or G. lamblia reflecting use by those authors cited.
1-7

-------
In the Giardia life cycle, the trophozoites divide by binary fission, attach to the brush
border of the small intestinal epithelium, detach for unknown reasons, then become rounded and
elaborate a cyst wall. The viable, environmentally-resistant cyst is excreted in the feces, moves
passively through the environment, primarily aquatic, and may be transmitted to another
vertebrate host if ingested. Following ingestion, the excystation process is initiated by conditions
in the stomach and completed once the excysting trophozoites pass into the less acidic conditions
of the small intestine where the trophozoites attach to the small intestinal epithelium.
Encystment is initiated by exposure of the trophozoites to bile (exact components unknown) in
the upper bowel and continues in the lower small intestine where the trophozoite rounds up and
secretes cyst wall components which move into encystment vesicles to begin the process of cyst
wall formation.

Recent, carefully controlled studies indicate that cross-species transmission of Giardia
can occur. Experimental human and animal infection studies offer evidence that rats, mice, dogs,
cats, beaver, muskrat, gerbils, and mule deer are capable of harboring Giardia that can infect
humans. The role of these animals as a source of human infection, however, remains
controversial. Of all of these animals, the beaver and muskrat are the most likely candidate
mammals that may serve as a source of infection or reservoir of Giardia for waterborne
outbreaks among humans. Both aquatic mammals can be infected with isolates of Giardia from
humans, but each has also been shown to harbor strains of Giardia that are phenotypically
distinct from those found in humans. G. mictoti, a species distinct from that in humans, has been
found in muskrat. It is possible that the beaver harbors two types of Giardia. One type may be
highly adapted to this animal and is rarely if ever transmitted to humans. Theother type maybe
one acquired by the beaver from human sources, which can multiply in the beaver and in turn br
transmitted via water back to humans. Thus, while Giardia that are indistinguishable from those
that infect humans are widespread throughout the Animal Kingdom, current evidence remains
insufficient regarding their ability to be transferred to humans.

-------
F. Chapter III. Occurrence

4. Prevalence and Waterborne Risks

In the United States, Giardia is the most frequently identified etiologic agent causing
waterborne outbreaks and the most frequently identified parasite in stool specimens submitted for
ova and parasites (4.0% up to 12% depending on the year and state). The prevalence of human
infection ranges world-wide between 2 and 5% in industrialized countries and 20 to 40% in
developing countries.

High risk groups forgiardiasis include infants and young children, travelers to developing
countries, the immunocompromised, homosexuals who practice oral-anal intercourse, and
persons who consume untreated water from lakes, streams, and shallow wells. Waterborne
outbreaks are more common in the United States and Canada than Europe, and this may be due
to the larger number of unfiltered surface water systems in North America. Populations in
communities with unfiltered surface water or groundwater that has been contaminated by surface
water or sewage are at high risk of infection.

Several small foodborne outbreaks of giardiasis have been associated with the
contamination of ice and foods by infected food service workers. Restaurant-associated
transmission of Giardia does not appear to be a significant public health problem. Outbreaks
have occurred in day-care populations and prevalence of Giardia infection is relatively high in
these populations; however, risk factors for the introduction, spread, and persistence of Giardia
in child day-care centers are not completely understood.

In the United States, waterborne outbreaks of giardiasis have been reported primarily in
unfiltered, chlorinated surface water systems. In most outbreaks, disinfection was found to be
1-9

-------
inadequate; chlorine concentrations and contact times were insufficient. In a few outbreaks
disinfection was interrupted. In many outbreaks neither the turbidity limit nor the coliform limit
was exceeded, but outbreaks have also occurred when turbidity levels were increased. Outbreaks
have generally occurred in areas of low water temperature where water disinfection may be less
effective and Giardia cysts can survive for longer periods of time. Outbreaks have occurred in
ground water systems emphasizing the need to protect these sources from sewage and surface
water contamination. Vulnerable ground water sources that cannot be protected from these
sources of contamination should be considered to be at the same high risk of contamination as
surface water sources. Filtration may be required for some groundwater sources to reduce
waterborne risks. Outbreaks have also occurred in filtered water supplies, and these emphasize
the need for proper chemical pretreatment and the importance of good design, installation,
maintenance, and operation of treatment fadlities. Because 10% of the waterborne outbreaks of
giardiasis occurred as a result of distribution system contamination, adequate precautions should
also be taken to protect treated water quality during storage and delivery.

Endemic risks of waterborne giardiasis are high among persons who consume untreated
water. In the United States, Canada, and New Zealand, endemic risks are higher among
populations that use unfiltered surface water compared to those that use filtered surface water.

2. Environmental Occurrence of Giardia

Interpretation of occurrence data is dependent upon methods used to detect and quantify
the cysts. Methods used to date generally provide little or no information on viability, infectivity,
or species identification when Giardia cysts are detected in environmental samples. Quantitative
data may not be reliable due to low efficiency and precision of methods.

Giardia cysts are distributed worldwide in surface waters, even those of excellent quality.
Cysts have been found in surface waters from the Arctic to the tropics. All surface waters
1-10

-------
probably always contain Giardia cysts at some level. Whether or not they are detected is
dependent upon the methods used to collect and analyze water samples. Cyst levels that have
been reported generally are on the order of 103"4/!, in raw sewage; 101"2/!, in secondary treated
wastewaters, and 10°/L or less in surface waters. Generally, there is no correlation of cyst levels
in water with bacterial indicator organisms. Cysts occur in surface waters throughout all months
of the year. Occasionally, seasonal variations are reported but these may be site or region
specific. When they are reported in North America, the levels are generally higher in the late
summer, fall and early winter.

Longitudinal studies using high frequency sampling indicate spikes in cyst levels that
might be missed by monitoring programs using low frequency sampling schedules. Cyst levels
are generally higher in rivers or streams influenced by agricultural (e.g., cattle or dairy farming)
or residential (e.g., sewage outfall) activities. Municipal wastewaters likely always contain
Giardia cysts at some level.

In the United States, levels of Giardia usually reported in water are somewhat lower than
Cryptosporidium levels. In other countries, e.g., Canada, widespread water surveys have found
higher levels of Giardia than Cryptosporidium.

National, regional, state or local surveys for the occurrence of Giardia in water may not
be representative of levels for a specific watershed. Sources of contamination and factors
affecting the transport and survival of cysts need to be determined for each watershed. It should
not be assumed that contamination levels of sources will remain constant. They may fluctuate
significantly due to poorly defined factors including weather events, agricultural practices and
treatment plant (wastewater and drinking water) infrastructure and operational practices. The
first-flush run-off from storm events will significantly affect source water cyst occurrence.
1-11

-------
No published reports on the occurrence of Giardia in soil or air were found. One study
reported the occurrence of cysts on stainless steel and Formica® surfaces in day care centers.
Data are sparse to non-existent on quantitative levels of cysts in or on foods. In 26 waterborne
outbreaks associated with drinking water, levels of Giardia cysts ranging from <1/100L to
580,000/lOOL were detected from tap or treated water or the water source in unfiltered systems.

The viability and longevity of Giardia cysts in the environment is significantly affected
by temperature-as the temperature increases, survivability decreases. A small fraction of cysts
can withstand a single freeze-thaw cycle. Cysts subjected to repeated freeze-thaws as might
occur in the environment are likely inactivated but still will be detected with present methods.

Cyst inactivation in municipal wastewater treatment plant sludge is temperature-
dependent. There is a factor or factors in swine manure slurry that results in more rapid
degradation of cysts under field conditions. A bacterium that is capable of killing Giardia cysts
has been isolated from a fresh water stream.

C. Chapter IV. Health Effects in Animals

In animal species (e.g., cats and dogs) whose Giardia infections have been studied in
detail, the resultant effects resemble those seen in humans. Infected calves also have been
observed to have diarrhea and mucus. Mortality appears to be significant in some animals, e.g.,
chinchillas and budgerigars, but it is rare in humans. Giardia infection may occur in animals of
any age but is more likely to occur, and to be symptomatic, in young animals. Many, if not most,
animals infected with Giardia exhibit no symptoms. These animals do, however, serve as
sources of infection for other animals. Symptomatic infection in animals that require therapy
usually respond to the same agents, with the same caveats, used in treating human infections.

D. Chapter V. Health Effects in Humans
1-12

-------
There is a wide clinical spectrum of giardiasis which ranges from asymptomatic infection
and acute self-limiting diarrhea to persistent chronic diarrhea, which sometimes fails to respond
to therapy. Asymptomatic infection is most common. Symptoms of giardiasis include: diarrhea,
steatorrhea, abdominal cramps, bloating, flatulence, pale greasy and malodorous stools, weight
loss, and vomiting. Severe disease may result in malabsorption or growth retardation but rarely
death. Chronic giardiasis appears to be infrequent, but when it occurs, may persist for years.

As with all diarrheas, fluid replacement is an important aspect of treatment; anti-giardial
drugs are also important in the management of the giardiasis. Chemotherapeutic agents used for
treatment of giardiasis include metronidazole, tinidazole, quinacrine, furazoli done, albendazole,
and ornidazole. Various doses and treatment periods are recommended for each drug. The drugs
may have different effectiveness in their ability to clear Giardia, and side-effects should be
considered. Drug resistance and relapses may occur. Paromomycin has been used to treat
giardiasis in pregnant women, but the cure rate may be low.

Progress has been made in understanding the biology of Giardia. However, the
mechanisms by which Giardia produces diarrhea and malabsorption and the key immunologic
determinants for clearance of acute infection and development of protective immunity remain
poorly understood. Data on the nature of human immune response to giardiasis are somewhat
limited, but there are indications that both humoral and cellular responses are present. Most
subjects infected with Giardia produce detectable levels of anti-parasite antibodies. However,
the role of specific antibody to Giardia in determining the host's clinical response to infection
has not been delineated.

-------
capable of distinguishing asymptomatic from symptomatic infection. The presence of anti-
Giardia antibodies in serum may indicate either past or present infection with Giardia, whereas
the presence of Giardia antigen in stool specimens indicates current infection.

The epidemiology of giardiasis is complicated by an apparent genetic heterogeneity in
this species. Differences in virulence, pathogenicity, infectivity, growth, drug sensitivity, and
antigenicity have been reported. In endemic areas where extensive heterogeneity exists, mixed
infections with more than one genotype may occur.

E. Chapter VI. Risk Assessment

Current risk assessment models have been used to estimate the risk of waterborne
Giardia infection in the United States. Based on levels of Giardia cysts found in treated
drinking water in the United States, the annual risks of Giardia infection are estimated to be 20 x
10~4 (20 waterborne Giardia infections per 10,000 persons annually) and may be as high as 250 x
10"4 (250 waterborne Giardia infections per 10,000 persons annually). These point estimates of
risk from drinking water exposures are 10 to 100 times greater than the annual risk suggested that
drinking water systems should attempt to maintain (10~4 or one waterborne Giardia infection per
10,000 persons). However, it is difficult to ascertain the level of accuracy that these risk
estimates represent, since no comparable risk estimates are available from epidemiological
studies and the risks do not account for viability, speciation, or analytical sensitivity and
specificity.

F. Chapter VII. Analysis and Treatment

1. Collection and Analysis of Environmental and Clinical Samples
1-14

-------
The absence of a practical cultural method for Giardia in environmental samples, and the
probability that one could not be developed, led to the development of microscopic examination
assay methods. Large-volume sample collection methods were developed using filtration
through microporous cartridge media.

Collecting large volume samples of raw source water resulted in many eluates containing
a significant amount of parti culates that had been retained on the filters. Initially, flotation
clarification techniques used zinc sulfate solutions; subsequently, other compounds including
sucrose, Percoll, and Percoll-sucrose were evaluated and incorporated into the method. The
development of fluorescent antibodies for Giardia revolutionized the detection step which had
previously been dependent upon examining concentrates with non-selective iodine staining. A
combination method was also developed whereby a single sample could be simultaneously
assayed for Giardia cysts and Cryptosporidium oocysts.

The original Giardia method was developed to assist in waterborne outbreak
investigations. It subsequently was adapted to different applications by those with a need to
study drinking water treatment effectiveness, occurrence and distribution of cysts in the
environment, or the fate and transport of cysts. In the absence of regulatory requirements to
monitor for Giardia, there was no official standardized method. However, voluntary efforts
through groups such as Standard Methods for the Examination of Water and Wastewater and
ASTM resulted in consensus reference or proposed methods that could be used as a baseline and
modified as needed for particular applications.

-------
recovery. While retention of cysts and oocysts on the sampling filter was improved by higher
turbidities in the water being sampled, the greater quantity of material obtained in the sample
pellets presented difficulties in the flotation purification and microscopic assay steps. The nature
of the turbidity (e.g., organic or inorganic, particle size, etc.) was more important than the total
amount in causing detection and identification problems. For example, algae could make
clarification and detection more difficult in certain types of water and at certain times of the year.

The fluorescent antibody assay, while improving detection of cysts, necessitated
developing a new definition for identifying cysts. Presumptive cysts were defined by size, shape
and apple green fluorescence under specified conditions of reagent type and use and microscope
configuration. Confirmed cysts met the presumptive criteria and had defined internal structures
characteristic of the genus. These definitions created confusion for interpreting results,
especially by persons not familiar with the methodology. Results were often ignored if no
confirmed cysts were identified. The presumptive designation included all objects that might be
Giardia cysts. The confirmed designation was applied to those presumptive cysts that could
definitely be identified as Giardia. The remaining objects might or might not be Giardia
because interferences, e.g., cross-reactions or degradation of internal structures are known to
occur. Some cysts in a known, purified preparation of Giardia will not meet the criteria for
confirmation. The presumptive/confirmed terminology was replaced with total counts and
counts with internal structures in the Standard Methods and Information Collection Rule (ICR)
methods. Another limitation of fluorescent antibody identification is that it is only specific to the
genus level. While antibodies with various specificities have been developed, the application
and interpretation of results with them is complicated by uncertainty in defining species within
the genus, and in identifying those species that might have public health significance.

-------
techniques have yet to realize their full potential. Problems have been encountered wilh
reproducibility of the assays and with inhibition of the PCR reaction in environmental samples.

The advent of the ICR, and the necessity for developing defined data quality objectives
for that monitoring effort, resulted in the collection of performance evaluation data that
underscored the low precision of the method in unapproved laboratories. With the promulgation
of the ICR, for the first time a process was implemented in the United States for approving and
conducting continual performance evaluation of analysts and laboratories that wished to do
environmental protozoa analyses. Until that time, adherence to specific methodological
protocols, or performance of recommended quality assurance/quality control procedures, was
strictly voluntary. Maintaining or developing a similar process after completion of the ICR may
help to ensure the reliability of data obtained through continued monitoring efforts.

Increased awareness of method limitations has also spurred development of alternative
methods and procedures. In the area of sample collection, sampling 10 L volumes instead of 100
L or more for raw waters is being investigated. Processing the entire concentrate for a 10 L
sample may be preferable to processing an undefined portion of a 100 L sample. This may
improve the detection limit helping laboratories and drinking water treatment utilities better
interpret results. Collecting smaller sample volumes also results in fewer particulates to cause
interferences in the detection assay and makes it easier to apply alternate separation technology
such as immunomagnetic techniques (instead of flotation separation where cyst recovery is low
or erratic). Also, the use of membrane filters with defined porosity (instead of yarn-wound filters
with nominal porosities) for sample collection can improve recoveries. For the assay portion of
the methodology, much of the tedium and fatigue associated with examining concentrates may be
relieved by using techniques such as flow cytometry and cell sorting.

Dependence upon non-cultural methods for the detection and identification of Giardia in
environmental samples has rendered determining the public health significance of positive
1-17

-------
findings problematical. Determining the viability or infectivity potential of small numbers of
cysts detected with non-cultural methods has been difficult or impossible to do. A detected cyst
may be eitherviable or non-viable. If the organism is alive, it maybe capable of causing
infections or, if it has been injured, it may not be infectious. While viability determinations
might not be necessary for some applications, such as waterborne outbreak investigations or
determining the effectiveness of a treatment process to physically remove cysts, they are very
important in assessing disinfection effectiveness and developing risk assessments upon which to
base treatment requirements or drinking water regulations.

Procedures used to determine viability have included dye staining, morphological criteria,
in vitro excystation, animal infectivity, and nucleic acid-based assays. Traditional dye staining
methods (e.g., with eosin) were found not to correlate with in vitro excystation or animal
infectivity. Subsequent research produced dyes that enter the viable cyst, e.g., fluorescein
diacetate (FDA) and those that are excluded from the viable cyst while they can enter non-viable
cysts, e.g.,propidium iodide (PI). Work that has been done with PI to date indicates that cysts
stained with this compound are not viable. However, cysts that do not take up the stain maybe
either viable or non-viable, and whether or not inactivated cysts stain depends in part on how
they were inactivated.

At least with G. muris, morphological criteria have been shown to correlate with PI
staining and animal infectivity. Clearly defined internal characteristics and the absence of a
peritrophic space are indicative of non-viable cysts. In vitro excystation also works well with G.
muris but it is erratic with G. lamblia cysts. Another problem is that while excystation may be a
good measure of viability for determining disinfectant effectiveness where large numbers of cysts
are used in an experimental design, the procedures are not practical for application to the small
numbers of cysts likely to be detected in water samples.
1-18

-------
Dye staining, morphological criteria and in vitro excystation may be adequate indicators
of viability for some applications but they could be conservative in estimating the potential for
infection. Animal infectivity has commonly been used in experiments to determine disinfectant
efficacy. However, it has seldom been used to evaluate the health significance of environmental
Giardia isolates because of costs and difficulties with interpreting results from some animal
models (i.e., poor specificity).

Nucleic acid-based viability assays have focused on the detection of mRNA by RT-PCR
techniques using either the giardin gene or an HSP gene. Amplification of the HSP gene has not
proven reliable and there is some question about the survival and longevity of mRNA when the
organism is inactivated by different techniques. Besides practical problems relating to the
sensitivity and application of PCR techniques to environmental samples, the question of how
viability determined by these techniques relates to infectivity remains to be resolved.

For diagnosis of giardiasis in either humans or animals, stools continue to remain the
specimen of choice. In humans, the majority of infections can be detected by stool examination,
but in some instances, examination of duodenal or intestinal fluids (by aspiration, biopsy or
string test) or the use of radiological procedures may be necessary. Fresh stools can be used to
prepare wet mounts that are examined by conventional light microscopy for the presence of cysts
or trophozoites.

Fresh, frozen or preserved stools can be examined using traditional dye staining
techniques or with increasingly popular immunofluorescence assays. A variety of commercially-
available fluorescent antibody kits that target cysts or antigens are available. Evaluation of these
kits indicates that they have a high degree of sensitivity and specificity. They may require less
time to perform and produce results with a single stool sample equivalent to fresh stool and dye
staining techniques that require multiple stool examinations. The use of flow cytometry with
1-19

-------
immunofluorescence reagents may allow a greater number of human or animal specimens to be
examined in a given time period with less operator fatigue.

Sero-diagnosis is still not a useful technique in the clinical setting due to the inability to
distinguish between present and prior infections. However, serologic testing may have value in
conducting epidemiological studies. Secretory antibody has been detected in a small study of
saliva specimens from patients infected with Giardia, but the potential for developing tests that
could be useful for either diagnostic or epidemiologic purposes remains to be determined. Also,
the development and application of gene probe techniques (e.g., PCR) for clinical diagnostic
purposes has thus far proved challenging due to inhibitory substances in feces and resulting
problems with sensitivity and specificity.

5. Water Treatment

-------
consider multiple barriers for the protection and treatment of both surface and ground water
sources: a combination of watershed protection for surface waters, well-head and aquifer
protection for ground water sources, water filtration, disinfection, and protection of the integrity
of the distribution system. Use of all of these barriers affords the most effective means for
assuring the microbial safety of public water supplies.

It is impossible and morally unacceptable to eliminate wild animals from a watershed,
but their affect on source water quality can be reduced. The strict control of contamination from
farming, domestic animals, and human sewage discharges can also reduce contamination of
source waters. Wells and springs should be protected from the influence of surface water and
sewage discharges from septic tanks and municipal wastewaters. While watershed management
practices can reduce the potential for contamination, they cannot eliminate it. To effectively
protect against the waterborne transmission of Giardia, adequate water treatment is also required.
For surface water sources and groundwater sources under the influence of surface water, both
disinfection and filtration are recommended. Filtration exceptions may be granted where water
sources meet criteria of EPA's Surface Water Treatment Rule (SWTR); however, if water
sources are also subject to contamination with Cryptosporidium, it should be remembered that
disinfection levels used to inactivate Giardia cysts may not be sufficient to inactivate
Cryptosporidium oocysts.

Filtration technologies commonly used by water supplies can be designed and operated to
remove 99% or more of Giardia cysts. Conventional and direct filtration, when operated under
appropriate coagulation conditions, can remove 99.9% to 99.99% of Giardia cysts. The highest
removal rates occurred in pilot plants and water utilities that optimized coagulation and achieved
very low finished water turbidities (0.1- 0.3 nephelometric turbidity units). Cyst removal was
poor in filtration plants where coagulation was not optimized even though the turbidity of filtered
water was low.
1-21

-------
Conventional and direct filtration facilities should include chemical pretreatment to
provide adequate coagulation. In some source waters, sedimentation may be needed to
effectively remove cysts. Removals similar to or better than conventional and direct filtration
effectiveness have been found for slow-sand and diatomaceous earth filtration, but operational
and other factors are important in maintaining high removals of cysts by these filters. Low water
temperatures may adversely affect the efficiency of slow sand filters.

Membrane filtration is promising for some water systems, but care must be exercised
when selecting the type and effective size of the membrane. IfGiardia cyst removal is desired,
the effective size of the membrane should be rated to remove at least 99.9% of cysts or cyst-sized
particles. It should be remembered that membranes that are effective for removing Giardia cysts
may not be effective for removing other protozoa of a smaller size, such as Cryptosporidium,
Cyclospora, or microsporidia. High levels of cysts are found in filtered backwash water, and this
potential source of contamination should be considered before this water is discharged to the
environment or recycled back to the beginning of the water treatment plant.

Disinfectants can also achieve 99% or greater inactivation of Giardia cysts, but the
effectiveness of a chemical disinfectant may be affected by factors including water temperature
and pH, applied and residual disinfectant concentration and contact time, particles which may
shield cysts from contact with the disinfectant, and organic matter which may cause disinfectant
demand. Filtration can make disinfection more effective by reducing the disinfectant demand
and removing particles that may interfere with disinfection effectiveness.

The EPA regulates disinfectants and disinfection by-products, and this limits the
concentration and contact time of any chemical disinfectant that can be applied. When lower
concentrations of a disinfectant are required to meet disinfection and disinfection by-product
limits, both filtration and disinfection may be necessary.
1-22

-------
Disinfection employed by the water industry can inactivate Giardia cysts; however, cysts
can be resistant to low doses of chlorine and chloramines, and there are differences between the
inactivation efficiencies of the various disinfectants. The reported effectiveness of inactivation
by the typically utilized water disinfectants, in decreasing order of efficiency, is as follows:
ozone, mixed oxidants, chlorine dioxide, iodine, free chlorine, and chloramines. Under current
operating conditions and with current designs, ultraviolet irradiation does not appear to be useful
for disinfection of Giardia cysts. Ct (disinfectant concentration and contact time) values are
available to compare disinfectants, and values are recommended for various conditions of water
temperature and pH. Applied and residual concentrations, as well as how the disinfectant is
applied, are important to consider. For example, ozone-peroxide is less effective than ozone and
preformed chloramines are less effective than chloramines that are not preformed. Since Ct
values are based on results of laboratory studies in demand-free water, caution is recommended
in extrapolating these data to natural waters and beyond the experimental conditions. If source
waters are heavily contaminated with Giardia cysts, disinfection alone may not be sufficient to
protect against waterborne infection. Even though disinfection is adequate to inactivate 99.9% of
Giardia cysts in heavily contaminated source waters, sufficient numbers of cysts may survive to
cause infection in a fraction of the population.
1-2'-

-------
II. GENERAL INFORMATION AND PROPERTIES

B. History and Taxonomy

Organisms in the genus Giardia are protozoa that are parasitic in the intestinal tract of
humans and a wide variety of vertebrates. Dobell (1932) persuasively argued that Giardia was
the first protozoan to be described; the description was recorded by Antony van Leeuwenhoek in
1681 after examining his own diarrheic stool with one of his simple microscopes.

The available evidence supports the idea thatGiardia can cause human intestinal disease
as well as disease in some lower animals. Giardia has not always been considered a human
pathogen because asymptomatic human infections are common. If exposure to an organism fails
to consistently result in symptoms or disease, it is difficult to satisfy Koch's postulates which
define the causal relationship between a microorganism and a specific disease (Last, 1995). Until
the second quarter of this century, most physicians believed Giardia to be a harmless intestinal
commensal, primarily because the parasite was identified in many persons who did not have
symptoms of disease.

The introduction of the drug quinacrine as a means of eliminating Giardia infection
played a role in the recognition of Giardia as a pathogen. The administration of quinacrine
resulted in the simultaneous disappearance of these protozoa as well as the host's intestinal
symptoms, lending credence to the pathogenic nature of the Giardia (Brumpt, 1937; Galli-
Valerio, 1937). A more recent study of infection in human volunteers confirms Koch's postulates
for Giardia (Nash et al, 1987).

We now know that this genus contains a variety of related organisms parasitic in the
intestinal tract of many vertebrate species. While historically, various genus names have been
applied, Giardia is now widely accepted as the genus name for this group. How to designate
ii-1

-------
species in this genus has been a question debated for at least a half century. Until the 1950s, it
was common to assign Giardia species names on the basis of a belief in strict host specificity and
the importance of protozoan body dimensions. This resulted in more than 40 Giardia "species"
being described (1CAIR, 1984). We now know that Giardia dimensions are not a reliable sole
criterion for Giardia speciation and that all Giardia are not strictly host specific. Thus, the
earlier species criteria have been called into serious question. While recent enzyme, DNA, RNA
and morphologic studies have provided data important to the resolution of this question, there is
no general agreement as to Giardia species criteria.

Tibayrenc (1994), in considering whether Giardia is a complex of several species or not,
concluded: "Due to the fact that Giardia is probably a donal organism, the biological concept of
species cannot be used to address the question." Reflecting the opinions of the contributors to
the scientific literature, the Giardia responsible for human infections will be found in this
document referred to variously asG. duodenalis, G. intestinalis, or G. lamblia.

B. Life Cycle of Giardia

Information on the life cycle of Giardia is important for characterizing the two stages of
the genus, and the steps which occur when the organism changes from one stage to the other. In
addition, information on the life cycle is of interest in understanding the health effects,
transmission of the cysts, development of symptoms of giardiasis, and identification of the
methods of in vitro cultivation to study the survival of the stages, and for producing cysts for
conducting experimental studies. The life cycle of Giardia was described in the previous
document (ICAIR, 1984) and recently reviewed by Marshall et al. (1997).

Giardia have a simple life cycle (ICAIR, 1984) in which the flagellated, binucleate
trophozoites reside in the upper two thirds of the host's small intestine (the duodenum and
jejunum), where they attach, by means of a ventral adhesive disk, to the brush border of the
II-2

-------
epithelium, and reproduce by binary fission. The trophozoite-to-cyst transformation takes place
in the intestinal tract except in cases where the encystment process may not be completed within
the intestines of the host. Sometimes the transition from trophozoite to cyst fails to be initiated
within the small intestine, or to be completed following transit through the intestinal tract and
excretion with the feces (Schaefer, 1990). This may happen in situations when there is rapid
intestinal transit and the residence time is not sufficient for cyst formation, especially in cases of
severe diarrhea. Thus, diarrheic stools from patients with giardiasis may frequently contain
trophozoites (1C AIR, 1984). The life cycle continues with excretion of the cyst, followed by the
subsequent ingestion of the cyst by a suitable vertebrate host (ICAIR, 1984).

1. Excystation

The passage of viable cysts through the stomach initiates the process of excystation,
whereby the trophozoite emerges from the encysted stage. Excystation is completed in the small
intestine (ICAIR, 1984). At the time of excystation, a quadrinucleate trophozoite in the process
of division emerges from the cyst wall, and promptly completes the division process, yielding
two binucleate trophozoites (Bingham and Meyer, 1979).

a. In vivo Excystation

Gillin et al. (1988) presented information on the conditions for in vivo excystation of
Giardia cysts. It is crucial to the life cycle that the cysts not excyst and trophozoites emerge in
the stomach, because the trophozoites would be killed by the gastric acids. These stomach acids
and the resulting low pH play a role in triggering excystation, but the trophozoites will not
emerge until a neutral pH is encountered in the upper portions of the small intestine.

b. In vitro Excystation
II-3

-------
The in vitro excystation ofGiardia cysts from human feces was successfully induced by a
combination of filtration and purification, employing centrifugation and layering of the cysts on
0.85 M sucrose (ICAIR, 1984). Exposure of purified cysts to synthetic gastric juices at pH 1.6
induced excystation to occur within 30 minutes. When the individual components of the gastric
juices were tested, only the hydrochloric acid (HC1) was required to induce excystation. Because
exposure ofGiardia cysts to variety of acids at pH 2.0 induced significantly higher percentages
of excystation than water controls, it was concluded it is the hydrogen ion, rather than the
specific counter ion, was necessary to induce excystation ofGiardia (ICAIR, 1984)

Other techniques have been developed to improve both the cyst purification and
excystation methods. Sauch (1984) reported that a Percoll solution at pH 7 and 1.08 g/cm3 gave
consistently higher cyst recoveries than either the sucrose or zinc sulfate methods. To improve
the purification of the recovered cysts, cyst suspensions were centrifuged at different
sedimentation velocities at unit gravity in Percoll density gradients from 1.01 to 1.03 g/cm3.
Consistently high levels ofGiardia excystation ranging from 40 to 95% were found with a
procedure that requires a low-pH induction step (involving three separate solutions) and an
excystation step (ICAIR, 1984). Sauch (1988) conducted excystation trials with a modification
of the previous procedure. Trypsin, serum, or bile salts were replaced by peptone, and
excystation was observed in most tests, indicating that neither trypsin, serum, nor bile salts is
required for excystation. However, the mean percentage of observed excystation ranged from
1% to 96%, indicating there were sources of substantial variability that remained to be identified.

Schaefer (1990) reviewed studies which described the excystation process. In the
process, the caudal flagella and distal ends of the other flagella extend outside the cyst wall and
begin to move slowly. Within 5 to 10 minutes of emergence, the flagella beat rapidly pulling
and/or breaking the trophozoite out through the cyst wall. This appears to result in atearing of
the cyst wall, which further facilitates the release of the trophozoite from the cyst.
II-4

-------
2. Encystation

Encystation is the process whereby the trophozoite transforms into the cyst stage of the
life cycle. During the intestinal portion of the Giardia life cycle, some trophozoites detach from
the brush border for various reasons and enter the fecal stream. The process of encystment
begins in the small intestine with the trophozoites becoming rounded and elaborating a cyst wall;
the resultant cysts are then excreted with the feces (1C AIR, 1984).

a. In vivo Encystation

Danciger and Lopez (1975) observed three patterns of Giardia cyst excretion or
production in 15 infected children over a 1-to 3-month period. In children labeled as "high cyst
excreters," large numbers of cysts were present in nearly all stool spedmens. In contrast, "low
cyst excreters" had detectable levels of cysts in only 40% of the stool specimens. A third group,
"mixed excreters," had 1-to 3-week periods of high cyst excretion, alternating with periods of
low cyst excretion. No correlation was found between the numbers of cysts produced and the
consistency of the stool or frequency of defecation. Furthermore, the use of purgatives failed to
increase cyst production.

Grant and Woo (1979) observed cyclic cyst excretion by laboratory rats and mice
experimentally infected with Giardia simoni and Giardia muris, respectively. This cyclic cyst
excretion exhibited periods of 7-8 days between peaks and was also observed in captive deer
mice and meadow voles that had been naturally infected with Giardia. Craft (1982) reported that
when rats were infected with G. lamblia from humans, they exhibited a similar cyst excretion
pattern to that observed in humans by Danciger and Lopez - 12% of the rats excreted large
numbers of cysts continuously, 80% excreted variable numbers of cysts intermittently, and no
cysts were excreted by 8% of the infected rats.
II-5

-------
b. In vitro Encystation

Gillin et al. (1987) conducted in vitro studies of the encystation and expression of cyst
antigens by G. lamblia. Cultured trophozoites tested in the presence of bile salts showed the
importance of bile salts to the process of encystation. Exposure to primary bile salts yielded
greater levels of encystation than exposures to secondary bile salts. In addition, the cultured
trophozoites exhibited more than a 20-fold increase in the numbers of oval, refractile cells that
reacted strongly to anti-cyst antibodies. The refractile cells also showed higher levels of
expression of major cyst antigens.

Schupp et al. (1988) described the morphology ofGiardia encystation in vitro. In several
strains grown axenically, light microscopy examination revealed an identical morphology with
Giardia cysts isolated from fecal samples. These morphological comparisons were based on
characteristic size and shape of the cysts and the presence of 2 to 4 nuclei. The cysts grown in
vitro were found to exhibit a similar positive immunoreactivity for the cyst wall.

The conditions required to trigger the encystment process in vitro have also been
described in detail. Qllin et al. (1989) characterized the roles of bile, lactic acid, and pH in the
completion of the life cycle of G. lamblia under in vitro test conditions. Bile and alkaline
conditions, such as those found in the lower small intestine, induced high levels of encystation.
In addition, lactic acid, which is a major product of bacterial metabolism within the colon, was
found to have stimulated the encystation process. The cysts produced in these tests exhibited
greater than 90% viability based on uptake of fluorogenic dyes and exclusion of propidium
iodide (PI), two conditions associated with the viability determination. See Section VII, Chapter
VII for a discussion of the studies concerning cyst viability. Gillin et al. (1989) reported that this
was the first quantitative demonstration of the in vitro complete life cycle for G. lamblia.
II-6

-------
Campbell and Faubert (1994) evaluated both the in vitro and in vivo encystation of G.
lamblia. Examinations of the intestines of the gerbil indicated that trophozoites and encysting
trophozoites were found in the three equal sections of the small intestine; much lower numbers
were found in the colon. Conversely, cysts were only found in the lower two sections of the
small intestine and the colon. In the in vitro tests with four strains of Giardia, there were
significant differences in the production of encysting trophozoites and cysts, but in vivo tests in
gerbils did not reveal similar differences. The encysting trophozoites were characterized as
having the presence of encystati on-specific vesicles (ESV), and there maybe cyst antigens.
Campbell and Faubert (1994) concluded that these encysting trophozoites represent a transient
population of cells which appeared during both in vitro and in vivo encystation, that the relative
differences observed in encystation among the strains during in vitro testing were not reflected in
vivo, and that passage through the gerbil during one cycle of encystati on/ex cystati on can result in
disparate test results during in vitro encystation testing.

Erlandsen et al. (1990a) analyzed for the presence of these cyst wall antigens by field
emission scanning electron microscopy, and located their presence in the filaments associated
with the outer portions of intact cysts and on the developing cyst wall filaments in encysting
trophozoites. With polyclonal and monoclonal antibodies specific for the cyst wall antigens,
there was strong labeling observed on the filamentous cyst wall, but no labeling on the
membranous portion.

McCaffery and Gillin (1994) conducted studies on protein transport during the processes
of growth and encystation. The endomembrane system has proteins present during growth and
encystation, but the ESV are novel secretory vesicles, that serve to transport cyst antigens to the
nascent encysting wall. These results suggest that Giardia, which is a primitive parasite, has
evolved various complex structures for protein transport to the cell wall.
II-7

-------
McCaffery etal. (1994) further studied these protein transport mechanisms. There is a
progression of the types of antigenic chemicals in these vesicles, and that there maybe
indications of the initiation of a stage in the antigenic switching which is differentiation-driven.
At this stage, while the cyst wall is being laid down, the antigen might no longer be produced or
transported to the wall, but may be taken back into the cell. This process might facilitate immune
evasion by the Giardia, both by providing a covering over the trophozoite surface and by
initiating the antigenic switching, which appears to provide the trophozoite with increased
resistance to host-mediated defenses.

C. Morphological Features

Members of the genus Giardia are flagellated protozoan parasites belonging to the
phylum Sarcomastigophora, class Zoomastigophorasida, order Diplomonadida, and family
Hexamitidae. All organisms in this genus are parasites which occur in trophozoite and cyst
forms (1CAIR, 1984).

The parasite adaptations promoting cyst survival in the external environment, and
trophozoite infectiveness and persistence in the mammalian small intestine, each contribute to
being key virulence properties for this parasite to cause symptomatic disease (Aley and Qllin,
1995). However, the actual properties of thetrophozoites that cause the diarrhea, such as toxins
or conventional virulence factors (if such exist), have not yet been identified.

The application of modem biochemical techniques has resulted in rapid advances in our
understanding of a number of areas of Giardia metabolism (Mendis and Schofield, 1994; Paget
et al, 1989, 1993). These studies have shown that the metabolism of these organisms is far more
complex, and that they have the ability to use a far wider range of substrates, than was originally
believed.

-------
       1.  Trophozoite

       Trophozoites of the genus Giardia inhabit the upper small intestine of the vertebrate host
(ICAIR, 1984). According to Meloni et al. (1995), the trophozoites are vegetative in that they are
"capable of growing" and "function in processes such as growth and nutrition and not in sexual
reproduction." Giardia appear to reproduce only asexually; sexual reproduction has not yet been
reported for this protozoan genus.

       The trophozoites are parasitic on the wall of the small intestine, but the pathophysiology
of infection by the trophozoites is poorly understood. The major anatomic change being blunting
of the villi in the small intestine.  The trophozoites are not invasive of the epithelial cell layer and
can survive only within the small intestine (Aley and Gillin, 1995).

       The pyriform bodies of trophozoites of the genus Giardia range from 9 to 21 jimlong, 5
to 15 |im wide, and 2 to 4 jim thick (ICAIR, 1984).  Trophozoites are identified by the presence
of two morphologically indistinguishable anterior nuclei, eight flagella, two central axonemes,
microtubular median bodies, and a ventral adhesive disk.  A pair of staining structures (median
bodies) lie dorsal  to the axonemes and are tipped dorsoventrally and anterioposteriorly so that the
right tip is more dorsal and anterior (ICAIR, 1984).  These are found in every species of Giardia
described.  Median bodies consist of random arrangements of microtubules that lack an origin or
insertion into any other structure and may play a supporting function in the posterior portion of
the trophozoite behind the striated (ventral) disk (ICAIR, 1984).
       2.  Cyst
       Giardia cysts are typically ovoid, and measure from 10 to 15 jim in length, and from 7 to
10 |im in width, with the cyst wall being approximately 0.3 |im thick (ICAIR, 1984). Newly
                                           II-9

-------
formed cysts contain two morphologically indistinguishable nuclei. Each nucleus in the cyst
undergoes a single further division, so that mature cysts contain four nuclei.

Filice (1952) stated that the median bodies of the trophozoites were rarely, if ever, seen in
cysts, but Sheffield and Bjorvatn (1977) found a group of randomly arranged microtubules near
the flagellar axonemes in cysts that could be median bodies. They also observed that the
microtubules were less compact than those observed by Friend (1966) in the trophozoite,
possibly accounting for the apparent absence of median bodies in cysts when viewed with
visible-light microscopy. Gillin et al. (1989) reported that a median body is visible in what they
designate as Type I cyst when viewed in relief with Nomarski differential interference contrast
optics. Gillin et al. (1989) described these Type I G. lamblia cysts as water resistant, oval
shaped, smooth, and refractile, with cyst wall, axostyle, and median body visible in relief by
Nomarski differential interference contrast optics.

Jarroll et al. (1989) demonstrated that a substantial component of the purified cyst wall
(PCW) is comprised of a polymer of galactosamine or N-acetylgalactosamine, with a structure as
yet undescribed, but with a function comparable to chitin, providing both physical strength and
resistance to chemicals. These polysaccharides and putative proteins of the PCW may also play a
role in excystment, functioning in some way in signal transduction and recognizing the proper
chemical stimuli leading to the excystment of the cyst and release of trophozoites, following
passage through the host stomach and when entering the small intestine (Aley and Gillin, 1995).

Cysts, unlike the trophozoites, are not motile, and must be protected from wide variations
of pH, temperature, and osmolarity. Cysts also provide protection from hypotonic lysis, since it
has been observed that trophozoites, which are shed during extreme cases of diarrhea without
going through encystment, will readily disintegrate. One of the document authors (Meyer) found
that when laboratory cultured G. lamblia trophozoites are removed from culture medium and
placed in tap water, the trophozoites become enlarged and undergo lysis within 30 minutes.
11-10

-------
These results are consistent with the fact that no osmoregulatory systems have been reported in
the trophozoites of these or other cyst-forming parasitic protozoa.

On the other hand, Giardia cysts may remain viable in water for long periods of time
under typical environmental conditions. For example, after being stored in water for 77 days at
8 C, the encysted forms were found to be viable by dye testing (Bingham et al., 1979). The
encysted forms were also found to be capable of excystation in mouse infectivity tests after
periods of 28 days storage in water with longer survivals observed at lower temperatures of less
than 10 C (deRegnier et al, 1989). See Section H.3., Chapter IE for further details of these
studies.

The metabolism of Giardia cysts has been studied. Paget et al (1989, 1993) have
compared oxygen uptake in Giardia cysts and trophozoites from both human and mouse sources.
Since the oxygen uptake of cysts has been shown to be 10 to 20 percent of that of trophozoites,
one can conclude that these protozoan forms are not dormant in the sense that bacterial
endospores are. Rather, Giardia cysts continue to metabolize, but at a rate that is much less than
that of their trophozoite counterparts.

D. Species Transmission

1. Direct Transmission Between Humans

Early studies demonstrated that Giardia cysts, ingested either in capsules or in water,
were capable of excysting and proliferating in the challenged host; most infections disappeared
spontaneously, different patterns of infection were observed, depending on the donor source of
cysts, and some infected persons failed to shed cysts for a long period (1CAIR, 1984). See
Section A, Chapter Vlfor further discussions of these studies. More recently, Nash et al.(1987)
reported successful infection of human volunteers from inoculations of cultured Giardia
ii-n

-------
trophozoites that had been isolated from two infected persons. Five volunteers each received
50,000 trophozoites of one isolate; another group of five each received 50,000 trophozoites of the
other cultured isolate. All volunteers in the group receiving the first isolate became infected;
three developed symptoms of giardiasis. None of the five volunteers in the group receiving the
second isolate became infected. This result is consistent with earlier observation (ICAIR, 1984)
that strain differences may exist between Giardia isolated from different human hosts, and it
suggests that failure of an isolate to cause infection in a particular host is insufficient evidence
for assuming that it cannot be infective for that species host. Nash et al. (1987) proposed that
these results fulfilled Koch's postulates for Giardia.

2. Transmission Between Animals and Humans

Research on the cross-species transmission of Giardia is important to identify sources
and reservoirs of infection. That is, can human-isolated cysts cause infection in animals and can
animal-isolated cysts cause giardiasis among humans? It is difficult to interpret the results of
early studies of cross-species transmission; the viability of the source cysts was either unknown
or assessed by eosin staining and, thus, failure of transmission could be due to feeding non-viable
cysts, rather than an indicator of species specificity. Research prior to 1985 (ICAIR, 1984)
suggest cross-species transmission can occur in some instances. In studies with Giardia cysts
from humans, attempts were unsuccessful to infect hamsters, domestic rabbits, laboratory mice,
deer mice, cattle, wapiti, mule deer, white-tailed deer, black bear, and domestic sheep. Success
was reported in infecting laboratory rats, gerbils, guinea pigs, beavers, dogs, raccoons, bighorn
sheep, and pronghorns with human-source Giardia cysts (ICAIR, 1984). It was also reported
that two of the three human volunteers became cyst-positive after ingestion of Giardia cysts from
a beaver, but when Giardia cysts were collected from beavers and fed to laboratory mice, rats,
guinea pigs, and hamsters, none of these animals became infected (ICAIR, 1984). When cysts
from the same beavers were given to four beagle puppies, all became cyst-positive, but these
results cannot be interpreted because a control animal was also found positive. Giardia cysts
11-12

-------
from a naturally-infected mule deer failed to infect two beagle puppies. Studies using Giardia
cysts from asymptomatic and symptomatic human donors, as well as axenically cultured
trophozoites originally isolated from a human, concluded: infection with G. lamblia is not
restricted to humans; the trophozoite stage is also infectious; and household pets, particularly
dogs, should be considered as a possible source of infection for humans (1C AIR, 1984).

Erlandsen et al.(1988a) studied the question, can Giardia cysts isolated from humans
cause infections in animals? They tested the ability of human-source Giardia to infect beavers
and muskrats. These investigators first showed that their cysts, from symptomatic human
donors, were viable. Inoculation of 5 x 105 G. lamblia cysts resulted in infection in 75% of
beavers. In some experiments, fewer than 50 cysts were sufficient to infect the beaver. In
contrast, muskrats could only be infected with human-source Giardia when the dose was equal
to or greater than 1.2 x 105. As a result of these studies, the authors concluded that the beaver and
muskrat must be considered possible intermediate reservoirs for Giardia that infect humans.
However, at that time, they were not able to assign to these animals a maj or role in the
epidemiology of waterborne giardiasis in humans.

A polymerase chain reaction (PCR)-based method for genotyping G. duodenalis isolates
using a polymorphic region near the 5' end of the small subunit (SSU) ribosomal RNA gene was
described by Hopkins et al. (1997). Analysis was performed using Giardia cysts purified directly
from feces. Isolates were collected from humans and dogs living in isolated Aboriginal
communities in Australia where Giardia infections are highly endemic. This is the first report of
the genetic characterization of Giardia from dogs and humans living in the same locality.
Comparison of the SSU-rRNA sequences from 13 human and 9 dog isolates revealed four
different genetic groups. Groups 1 and 2 contained all of the human isolates, whereas groups 3
and 4 consisted entirely of Giardia recovered from dogs. These results suggest that zoonotic
transmission of Giardia infections between humans and dogs does not occur frequently in these
communities. The dog-associated SSU-rRNA sequences have not been reported before,
11-1:

-------
suggesting a possible new G. duodenalis subgroup. A genetic basis for the differences observed
between the groups was supported by sequence analysis of nine in vitro cultured isolates that
were placed into the same genetic groups established by enzyme electrophoresis.

3. Transmission Between Animals

Information from studies using Giardia- free mice (1CAIR, 1984) demonstrated that G.
simoni, G. muris and G. peromysci were host-specific while G. microti and G. mesocricetus were
not. Pathogen-free mice were also successfully infected with Giardia from hamsters, but
Giardia cysts obtained from parakeet feces and stored for 1-3 days were unable to infect mice or
canaries (1CAIR, 1984).

Cross transmission studies in which beavers and muskrats were fed Giardia cysts from
muskrats, beavers and mice were conducted by Erlandsen et al.(1988b). Beavers did not become
infected when inoculated with cysts of G. ondatras (source: muskrats) or G. muris (source:
mice). Five of eight (62%) muskrats became infected when administered Giardia cysts of beaver
origin.

4. Summary of Cross-Species Transmission

Many early Giardia transfer studies were poorly controlled, but more recent carefully
controlled studies indicate that cross-species transmission of Giardia can occur. Experimental
human and animal infection studies offer increasing evidence that some lower animals,
particularly fur-bearing water mammals, are capable of harboring Giardia that can also infect
humans (Isaac-Renton, 1994). While Giardia that are indistinguishable from those
that infect humans are widespread throughout the Animal Kingdom, current evidence remains
insufficient regarding their ability to be transferred to humans.
11-14

-------
Experimental infection studies suggest that rats, mice, dogs, cats, beaver, muskrat,
gerbils, and mule deer are capable of harboring Giardia that may also infect humans. The role in
nature of these animals as a source of human infection, however, remains controversial. Of all of
these animals, existing evidence suggests that the beaver and the muskrat are the most likely
candidate mammals to serve as a source or reservoir of giardiasis and possible cause of some
outbreaks in humans. Both of these aquatic mammals can be infected with isolates of Giardia
from humans. However, each has also been shown to harbor strains of Giardia that are
phenotypically distinct from those found in humans. G. mictoti, a species distinct from that in
humans, has been found in muskrat (van Keulen et al., 1998). It is possible that the beaver
harbors two types of Giardia. One type may be highly adapted to this animal and is rarely if ever
transmitted to humans. The other type maybe one acquired by the beaver from human sources,
which can multiply in the beaver and in turn br transmitted via water back to humans. The
argument supporting the complicity of the beaver and muskrat in human giardiasis and
minimizing the role of other animals is as follows: In North America, epidemic and endemic
giardiasis is frequently transmitted by contaminated water. See Section G, Chapter III for a
further discussion of these studies. To deposit sufficient numbers of cysts that can infect large
numbers of humans in a short time arguably is best accomplished by Giardia-mfected animals
which, by nature, defecate in fresh water. While cyst-bearing feces of rats, mice, dogs, cats and
deer may occasionally reach drinking water, these animals do not, as beavers and muskrats do, by
nature defecate in water. Beavers have been implicated as a possible source of contamination in
several waterborne outbreaks (Craun, 1990). Thus, while Giardia that are indistinguishable from
those that infect humans are widespread throughout the Animal Kingdom, current evidence
remains insufficient regarding their ability to be transferred to humans.

To conclusively determine whether human giardiasis can be acquired by zoonotic routes
and whether the ultimate source was human or a lower animal will require carefully controlled
feeding studies and more detailed investigation of waterborne outbreaks that includes: systematic
collection of Giardia cysts (1) from infected humans, (2) from animals suspected of
11-15

-------
transmission, and (3) from environmental samples, and their characterization by molecular
approaches such as zymodeme or karyotype identification. Studies taking this approach are in
progress in British Columbia, and have strengthened the evidence that suggests a role for beaver
in the spread of giardiasis to humans (Isaac-Renton, 1994).

Buret et al. (1990) postulated that domestic ruminants may be a reservoir for human
infection. A study ofGiardia infection of ruminants found that cyst output and clinical signs
resembled human disease and that the Giardia from infected ruminants was morphologically and
antigenically similar to humans. Giardia trophozoites from sheep were successfully cultured in
TYI-S-33 medium; cytosolic, cytoskeletal, and membrane fractions were found to exhibit protein
profiles similar to human isolates. Immunoblotting indicated that sera from infected sheep
recognized human Giardia, and sera from human patients with giardiasis recognized Giardia
from sheep. In both cases, recognition involved antigenic proteins of similar molecular weight.
A pilot study of experimentally infected eastern barred bandicoots (Perameles gunnii) in
Tasmania suggested their susceptibility to infection with Giardia from a human source (Bettiol et
al., 1997).

E. Species Concepts in the Genus Giardia

In the past, there was no general agreement regarding the characteristics which define
species in the genus Giardia. Characteristics used previously include host specificity,
morphology, and variations in the shape of the median bodies (1CAIR, 1984). While Giardia
size and shape may vary somewhat with the organisms collected from the respective host species,
Giardia isolated from different host species may also be morphologically indistinguishable, and
additional characteristics should be included in determining the speciation ofGiardia (1C AIR,
1984). Based on work with G. ardeae, Erlandsen et al (1990b) stated that median body structure
alone should no longer be considered adequate for classification at the species level. Their
11-16

-------
axenic culture work was used to derive trophozoites, and the species description was based on a
variety of morphological criteria and on chromosomal migration patterns.

1. Filice's Concept

In 1952, Filice concluded that the use of differences in body dimensions ofGiardia and
host specificity were untrustworthy criteria for distinguishing between species and suggested that
Giardia morphological groups based primarily on structural differences. It was recognized,
however, that physiologically distinct species may exist among those that appear similar
morphologically. Filice proposed recognizing three Giardia groups:
(1) G. duodenalis, with a single or double median body which somewhat resembles the
claw of a claw hammer. These organisms have been isolated from humans, other mammals
(including rodents), birds, and reptiles.
(2) G. muris, with two small rounded median bodies in the center of the organism.
Rodents, birds, and reptiles have been shown to be hosts for this morphological type ofGiardia.
(3) G. agilis, with long, teardrop-shaped median bodies. While the adhesive disk of the
other two proposed species is on the order of half the trophozoite body length, the G. agilis
adhesive disk is only approximately one-fifth the body length of these organisms. Organisms of
the G. agilis -type have only been described from amphibian hosts.

Filice also suggested assigning a non-taxonomic status, such as race, to allow
incorporation of the Giardia that were considered to be distinct species on the basis of size or
host specificity alone.

2. Grant and Woo's Concept

Grant and Woo (1978a, 1978b) and Erlandsen et al. (1990b) questioned the concept of
Giardia speciation based only on median body structure. Grant and Woo felt that species of
11-17

-------
Giardia should be defined using a combination of morphological, morphometric, and host-
specificity criteria. They were able to distinguish five species of Giardia within small mammals
in Ontario, Canada: G. muris, G. mesocricetus, G. simoni, G. microti, and G. peromysci. The
five species were divided into two types (1CAIR, 1984). In type I, trophozoites have elongated
nuclei in the posterior region of the sucking disk, the sucking disk occupies a large portion of the
body, basal bodies are anterior to the nuclei, and the median bodies are round or oval and are
located near the center of the body. Type I includes G. muris and G. mesocricetus. In type II,
trophozoites are reported to be longer than wide, the sucking disk is in the anterior half of the
body, the nuclei are in the central region of the disk and the median bodies are comma- or claw
hammer-shaped. Type II trophozoites include G. simoni, G. microti, and G. peromysci. Cross-
transmission studies using Giardia-free mice and rats were used to examine the host specificity
of the two types of Giardia that were discerned morphologically. G. simoni, G. muris, and G.
peromysci were host specific, but G. microti and G. mesocricetus were not. Grant and Woo were
not able to distinguish any further characterization of these species within each type based on
morphological observations but subsequently found statistically significant differences among
the major dimensions of some trophozoites within and between types I and II. They also
described a variety of problems dealing with Giardia speciation based on morphometrics
including differences in size, general morphology, density of cytoplasmic staining, and relative
proportions of cytoplasmic organelles of Giardia trophozoites under the influences of prefixation
drying times.

3. Other Concepts for the Speciation of Giardia

With the advent of new biochemical, molecular, and genetic techniques, there has been
considerable activity in characterizing Giardia speciation. Meloni et al. (1995) described the
isoenzyme electrophoresis analysis studies (Baveja et al., 1986; Andrews et al., 1989; Homan et
al., 1992), and recombinant DNA probe characterizations (Homan et al., 1992; van Keulen et al.,
1992, 1993) which have been conducted with Giardia to attempt to characterize the speciation.
11-18

-------
Based on the host-specificity ofGiardia, the genus was characterized in 1926 as having
over 40 species (ICAIR, 1984). In 1952, Filice proposed 3 groups based on its morphology.
However, these present taxonomic groupings do not reflect the genetic and phenotypic
heterogeneity within the species G. lamblia, and afford little information on which to estimate
host specificity, infectivity, or virulence.

In general, organisms which are primarily clonal and reproduce by asexual reproduction
are characterized by distinctive population structures, showing excess heterozygosity, association
between independent genes (linkage disequilibrium), and a greater proportion of their genetic
variation being distributed between as opposed to within, their populations (Meloni et al., 1995).
Moreover, these types of clonal organisms (e.g., those with asexual reproduction) present
problems in taxonomic characterization, because some of the typical biological species concepts
do not apply to them. These factors contribute to the past difficulties in clarifying the correct
species designation for the grouping(s) of interest from the perspective of human giardiasis.

Meloni et al.(1995) utilized enzyme electrophoresis techniques on 97 isolates of G.
duodenalis collected from humans, cats, cattle, sheep, dogs, goat, beaver, and rats in Australia.
The intent was to characterize the mode of reproduction, population structure, taxonomy, and
zoonotic potential. From these enzyme electrophoresis studies, it was possible to identify 47
groupings of enzyme patterns (called zymodemes) based on their cluster patterns with each other.
These zymodemes imply that the parasitic organisms have similar genetic structures, but the
clonal lineages imply they are evolutionarily independent, and mean that the actual mode of
reproduction cannot be inferred with confidence (and may not be strictly asexual). The
information presented by Meloni et al.(1995) does not provide information on which to support
the species concept of either Filice or Grant and Woo; rather, the evidence shows the complexity
of attempting a specifically-clarified approach to presenting a species designation. Thus, at
present, there does not exist a completely satisfying designation of the actual number of species
11-19

-------
with the genus Giardia, and further research is needed in the areas of molecular biochemistry
and genetic biology.

F. Summary

All of the organisms in the genus Giardia are binucleate, flagellated protozoan parasites
which cause infection by attaching to the wall of the small intestine in the upper gastrointestinal
tract of humans and other vertebrates. The parasites exist in trophozoite and cyst forms.

Giardia have been reported in a variety of mammals and in lower vertebrates. While
numerous species of Giardia have been described, there is no general agreement on those criteria
which define species in this genus. Criteria used to date include: host specificity; body size and
shape, and the morphology of a microtubular organelle, the median body; and biochemical,
molecular, and genetic techniques, such as the PCR for DNA-based detection and identification.
The median body is an organelle that appears to be unique to Giardia trophozoites.

In the Giardia life cycle, the trophozoites divide by binary fission, attach to the brush
border of the small intestinal epithelium, detach for unknown reasons, then become rounded and
elaborate a cyst wall. The environmentaly-resistant cyst is excreted in the feces, and the
transmission to a new host is accomplished by ingestion of viable cysts. The excystation process
is initiated by conditions in the stomach, and is only completed once the excysting trophozoites
pass into the less acidic conditions of the small intestine, where the trophozoites promptly attach
to small intestinal epithelium.

Excystation is induced by exposure to low pH (as exists in the stomach), and has been
induced in vitro. Encystment is initiated by exposure of trophozoites to bile (exact components
unknown) in the upper bowel and continues as the excysting trophozoites pass to the lower small
intestine, where the trophozoite rounds up and secretes cyst wall components which move in
11-20

-------
vesicles to the cell wall to begin the process of cyst wall formation. This cyst wall protects the
cysts when they pass out of the host with the feces. At that time, the cyst moves through the
environment, primarily aquatic, and can possibly be transmitted to another vertebrate host.

Although many early studies were poorly controlled, more recent studies suggest that
cross-species transmission of Giardia can occur. Experimental human and animal infection
studies and information from waterborne outbreak investigations offer increasing evidence that
beaver, muskrat, rats, mice, dogs, cats, gerbils and mule deer can be infected experimentally and
may harbor human Giardia; in addition, humans appear to have been infected by G. lamblia
cysts isolated from beaver and mule deer. Furthermore, it seems that these experimental
infections may be established by direct-feeding with doses of either cysts or cultured
trophozoites. Mice and rats, experimentally infected with G. lamblia, appear to produce a
relatively low number of cysts. The necessity for meticulous attention to the use of Giardia -free
animals is discussed. To determine the Giardia-free status in experimental animals, reliance on
stool examination alone is not sufficient, because necropsy analyses have shown that
trophozoites may be present in the intestinal tract in a higher percentage of the population than is
revealed by relying solely on fecal analyses.

Of all of the animal species suspected of being a significant zoonotic source of human
giardiasis, the evidence presently available suggests that the beaver and muskrat are the most
likely candidates. The role of these animals as a source of human infection, however, remains
controversial. Both aquatic mammals can be infected with isolates of Giardia from humans, but
each has also been shown to harbor strains of Giardia that are phenotypically distinct from those
found in humans. It is possible that the beaver harbors two types of Giardia. One type may be
highly adapted to this animal and is rarely if ever transmitted to humans. The other type may be
one acquired by the beaver from human sources, which can multiply in the beaver and in turn br
transmitted via water back to humans.
11-21

-------
11-22

-------
III.    OCCURRENCE

G.     Worldwide Distribution

       1. Distribution in Animal Populations

       Organisms in the genus Giardia have been reported as intestinal inhabitants within a
variety of mammals, birds, reptiles, amphibians, and fishes (1C AIR, 1984). Thus, Giardia should
be considered among the most widely occurring of the intestinal protozoan parasites. Early
workers assumed Giardia identified in animals was host specific.  Giardia from some animals
exhibit an apparent high degree of host specificity; other isolates may infect more than one host
(Grant and Woo, 1978b; Davies and Hibler, 1979; Erlandsen et al., 1988a,b; Ey et al., 1997).

       Giardia is a common protozoan parasite of farm animals and occurs with greater
prevalence in young animals. Buret et al. (1990) found 18% of sheep and 10% of cattle infected
with Giardia; a higher prevalence was found in lambs (36%) and calves (28%).  Olson et al.
(1997a) reported the following prevalence in farm livestock in Canada: cattle (29%), sheep
(38%), pigs (9%), and horses (20%).  A high prevalence of Giardia infection has been found in
dogs (77%); a lower prevalence (3-11%) has been found in cats (Bemrick 1961; Kirkpatrick,
1986). A high prevalence of Giardia (>90%) has been reported for wading birds and muskrats
(Erlandsen, 1994; Erlandsen et al., 1990b,c). Giardia have also been foundin beaver (7-16%),
voles, mice, shrews, native marsupials, ringed seals, and llamas.  See Chapter IV, Section C for
further information on the prevalence of Giardia in animals.

       Surveys of Giardia in animal populations have often relied on detecting cysts in fecal
samples. In a survey of beaver and muskrat populations from the northeastern United States and
Minnesota, Erlandsen et al.(1990c) compared the analysis of fecal samples with the detection of
internal trophozoites at necropsy. Beaver infection with Giardia was 9.2% (n = 662) by analysis
                                          m-i

-------
of cysts in fecal samples from kill-trapped samples and 13.7% (n= 302) by examination for
intestinal trophozoites in live-trapped animals.  For muskrat, the differences were even greater,
with the prevalence oi Giardia 36.6% (n = 790) by detection in fecal samples from kill-trapped
animals and 95.9% (n = 219) by examination of the intestinal contents. These data suggest that
surveys based on fecal analyses may under-report the actual numbers of infected animals.

       2.  Distribution in Human Populations

       Giardiasis is the most commonly reported intestinal protozoan infection worldwide. The
World Health Organization estimates 200  million people are infected each year (Swarbricket al.
1997). Human infections with Giardia have been reported in all of the major climatic regions
from the tropics to the arctic (1CAIR, 1984). All age groups are affected, but children are more
frequently infected than adults (Benenson, 1995).  An analysis of analyzed randomly collected
stool specimens in two counties of Washington State found that 7.1% of 515 healthy 1- to 3-
year-old children were positive for Giardia cysts (1CAIR, 1984). In the United States, United
Kingdom, and Mexico, endemic infection most commonly occurs during July to October and
among children under five years of age and adults aged 25-39 years of age (Benenson, 1995).
The prevalence of stool positivity may range from 1-40%, depending on the geographic area and
age group surveyed; prevalence is higher in areas with poor sanitation and institutions with
children not toilet trained (Benenson, 1995). The prevalence of infection can be as high as 35%
among children attending child care centers with attack rates in outbreaks of  50% or more
(Adam, 1991; Hall, 1994; Ortega and Adam, 1997; Steketee et al.,  1989). Prevalence rates vary
from 2-5% in developed countries; in developing countries, prevalence can be as high as 20-40%
(Farthing, 1996).  All of a birth cohort of 45 Guatemalan children had giardiasis before age four
(Farthing, 1996) and 40% of Peruvian children were infected by six months of age (Ortega and
Adam, 1997;  Miotti et al.,  1986). See Chapter III, Section F for further discussion.

       The Centers for Disease Control and Prevention (CDC) do not require notification or
                                           m-2

-------
reporting of cases of giardiasis. Localities, states, and U.S. territories conduct their own disease
surveillance and voluntarily report cases of giardiasis to the CDC. Forty-three states, the District
of Columbia, and three U.S. territories have mandatory reporting requirements for giardiasis
(Chorba et al., 1989). Giardiasis has been reported in the District of Columbia and the following
40 states and 4 territories: Alaska, Arizona, Arkansas, California, Colorado, Connecticut,
Delaware, District of Columbia, Florida, Guam, Hawaii, Idaho, Illinois, Indiana, Iowa, Kansas,
Kentucky, Maine, Maryland, Minnesota, Mississippi, Missouri, Montana, New Hampshire, New
Jersey, New Mexico, New York, Ohio, Oklahoma, Oregon, Pacific Trust Territory,
Pennsylvania, Puerto Rico, South Carolina, South Dakota, Tennessee, Texas, Utah, Vermont,
Virgin Islands, Virginia, Washington, West Virginia, Wisconsin (ICAIR, 1984).

In a non-random survey of 332,312 stool specimens from patients submitted for
parasitological evaluation in 1978 by 53 state and territorial public health laboratories (ICAIR,
1984), the CDC found that Giardia was the most commonly identified parasite in the United
States. The percentage of Giardia-poshive stools ranged from 1.1% of specimens in Virginia to
greater than 8% of specimens in Arizona, California, and Washington (ICAIR, 1984). Giardia
was also the most frequently identified parasite in laboratory surveys conducted in 1976 and
1977 in the United States (Kappus et al., 1994) and in 1979 in Canada (Gyorkos, 1983). In 1984,
12.5% of 1,710 public health clinic patients seen in Oregon were found positive for Giardia
(Skeels et al., 1986). National surveys were also conducted in 1987 and 1991 with 49 states
participating (Kappus et al., 1994). G. lamblia continued to be the most frequently identified
parasite. It was found in 7.2% of 216,675 specimens examined in 1987 and 5.6% of 178,786
specimens examined in 1991, an increase from the 3.8% to 4.0% average found in the 1976,
1977, and 1978 surveys. Forty states reported an increased identification in specimens submitted
to the laboratories. Seasonally, Giardia identifications increased in the summer and fall,
especially in the Midwest. States reporting high percentages of Giardia identification for both
1987 and 1991 were located in the Midwest or Northwest. In 39 states and eight of the nine
regions, Giardia was the most frequently identified parasite every month of the survey periods.
m-3

-------
An analysis of hospital discharge data from the National Center for Health Statistics
(Lengerich et al., 1994) estimated that 4,600 persons were hospitalized with giardiasis annually
in the United States from 1979 to 1988 — an incidence of 2.0 hospitalizations per 100,000
persons per year. The estimated hospitalization rates were highest for children less than 5-years-
old (4.6 per 100,000 per year) and women of child-bearing age, 25- to 34-years-old (3.5 per
100,000 per year). Although children younger than 5 years of age also had a high rate of
hospitalization in Scotland, women of child-bearing age were no more likely to be hospitalized
with giardiasis than men of the same age group (Robertson, 1996). Among residents of
Michigan from 1983 to 1987, the average incidence of hospitalization was 1.4 per 100,000
persons per year (Lengerich et al., 1994). Hospital admissions for giardiasis account for only
one-tenth of all cases of giardiasis reported by physicians and only a fraction of symptomatic and
asymptomatic cases are seen by physicians (Lengerich et al., 1994).

B. Occurrence in Water

Giardia cysts are ubiquitous in surface waters of all qualities. Because Giardia infections
are widespread in human and animal populations, contamination of the environment is inevitable
and cysts have been detected in even the most pristine of surface waters. The limitations of the
detection methodology with respect to efficiency of recovery and viability or infectivity of the
detected organisms should be borne in mind when evaluating the significance of occurrence data.
See Chapter VII, Section A for a further discussion.

1. Wastewaters

Wastewaters that are discharged into rivers and streams are sources of Giardia in surface
waters. Sykora et al. (1991) studied the occurrence of Giardia in the wastewaters and sludges of
11 cities across the continental United States. In examining monthly samples from each site,
m-4

-------
they found that all of the raw sewage samples were positive for cysts at levels ranging from 4
cysts/L to 14,000 cysts/L. The geometric mean level of cysts at each site ranged from 642/L at a
Pennsylvania treatment plant to 3,375/L at a California plant. A seasonal distribution was noted
with cyst levels reported highest during late summer through early winter. While all of the raw
sewage samples were positive, about 48% of the secondary effluents contained cysts at levels
ranging from 1 to 44/L. About 80% of the sludges were positive at levels ranging from 70 to
30,000 cysts/L. It should be recognized that the cyst counts in this study can be considered
minimal because the analyses used the iodine staining method which has been shown to produce
significantly lower cyst counts than an immunofluorescence assay (IFA) applied to
environmental samples (LeChevallier et al., 1990).

Investigators in Scotland examined eight raw sewage samples collected from November,
1990 to January, 1991 (Smith et al., 1994). Cyst levels ranged from 38 to 44 cysts/L, but the
method used to examine the samples was not specified. Seven effluent samples were also
examined with cyst levels ranging from 0.8 to 5.9/L. A subsequent investigation in Scotland
examined raw sewage samples collected bimonthly for one year from six sewage treatment plants
(Robertson et al.,1995). All of the samples were positive for cysts at levels ranging from 102 to
43,907/L.

Roach et al. (1993) examined raw and treated sewage samples collected from five sites in
the Yukon. Forty four samples of raw sewage were examined, and all were positive for Giardia
cysts at levels ranging from 26 to 3,022 cysts/L. Five samples of treated sewage from two
locations were positive at cyst levels ranging from 2 to 3,511 cysts/L. Daily monitoring of raw
sewage at one site suggested an increasing level of giardiasis in the community during the
summer months that was possibly related to episodic occurrence of cysts in the water supply in
the spring and early summer (Roach et al., 1993).
Ill-5

-------
Enriquez et al. (1995) used IF A to examine 130 wastewater samples from two activated
sludge treatment plants and a water reclamation plant for Giardia over a three year period. The
geometric mean level of cysts in secondary effluents from each of two activated sludge treatment
plants was 8.3 and 6.6 cysts/40 L. The geometric mean cyst level in the tertiary effluent from the
water reclamation plant was 3 cysts/40 L. No seasonal variation in cyst levels was observed.

Grimason et al. (1996) reported finding Giardia cysts in 37% of raw sewage samples
collected at a plant in Kenya and in 100% of similar samples collected at a plant in southern
France. These treatment plants processed municipal wastewater using stabilization pond
systems. Cyst levels ranged from 1,000 to 25,000 cysts/L in the Kenyan sewage and from 230 to
25,000 cysts/Lin the French sewage. Cysts were detected in all final pond effluents at the
Kenyan plant and in 44% of those sampled at the French plant. Calculations indicated that pond
retention periods of 25 to 40 days would remove more than 99% of the cysts but were not
sufficient to assure cyst-free effluents. Wiandt et al. (1995) evaluated seasonal differences in
cyst levels in the final effluent at the same French plant. They were unable to detect cysts in the
final effluent in samples collected during the spring and summer but did find cysts at levels of
0.1 to 2.5/L during the wintertime.

In evaluating FA and polymerase chain reaction (PCR) methods for detecting cysts,
Mayer and Palmer (1996) examined 11 samples of raw sewage, 11 samples of primary effluent
and 10 samples of secondary effluent at a large metropolitan treatment plant in California. They
reported levels of 1.3 x 104, 2 .6 x 103 and 1.1 x 101 cysts/L in the primary influent, primary
effluent and secondary effluent samples, respectively.

Hirata and Hashimoto (1997) reported a survey on the occurrence of Giardia and
Cryptosporidium in nine activated sludge sewage treatment plants located in the Kanto area of
Japan. Grab samples were centrifuged, the pellets were purified by flotation and the cysts were
detected and enumerated using an immunofluorescence assay. Viability was not determined.
m-6

-------
The limit of detection of their methods was about 16 cysts/L for raw sewage and primary
effluents and 2 cysts/L for secondary and final effluents.  They found 95% (112/118) of the
samples positive for cysts: 100% (29/29) of the raw sewage samples (geometric mean = 1,500/L;
range = 130 to 7,900/L); 100% (37/37) of the primary effluent samples (geometric mean =
1,100/L; range = 150 to 6,600/L); 86% (30/35) of the secondary effluent samples (geometric
mean = 18; range = 2 to 310), and 82% (27/33) of the final effluent samples (geometric mean =
14/L; range, 4 to 130/L).  The average removal of cysts by the activated sludge treatment process
was 1.8 Iog10. The level oi Giardia cysts was significantly correlated with coliforms, Escherichia
coli,  Clostridium perfringens, and turbidity. Turbidity was also correlated with removal of cysts
by the treatment process.

       Examination of Giardia in raw sewage has been suggested as an alternative method of
assessing the prevalence of Giardia infection and detecting possible outbreaks in communities,
but only one research study  has examined the relationship between cysts in sewage and illness in
the community. A correlation was found between raw sewage cyst levels and reported cases of
giardiasis in 11 communities in the United States (Jakubowski et al.,  1991).

       2.  Surface Waters

       Numerous studies have been directed at determining levels of Giardia in drinking waters
and the surface waters that serve as the source waters for drinking waters to help  assess the risks
of waterborne Giardia infection in populations using these waters. In the United States, studies
have been conducted to collect water samples over large geographical areas and within individual
drainage basins; studies have also been conducted in Canada, the Virgin Islands, Europe, and
Asia (Table III-l). Authors of original studies reported various units, but in presenting the results
of these studies, data have been converted and presented here for consistency as "cysts/100 L."
                                           m-7

-------
The largest single data base of water samples examined for Giardia was developed and
compiled by Hibler(1988). He analyzed 4,423 water samples from 301 municipal sites in 28
states between 1979 and 1986 using a zinc sulfate flotation/iodine staining method to detect cysts
(APHA-AWWA-WPCF, 1981). He reported his results as percentages of various types of
samples found positive for cysts and did not include quantitative data on cyst levels. He found
34% of the municipal sites (102/301) were sometimes positive for cysts, with 26% (512/1,968) of
raw water samples and 11% (267/2,732) of finished water samples demonstrating cysts. In order
of decreasing prevalence, the positivity rates for source water types were 28% (346/1,218) for
creeks; 26% (212/828) for rivers, 10% (193/1,983) for lakes, 19% (16/84) for springs, and 3%
(2/63) for wells.

Rose et al. (1988a) conducted a biweekly survey of a watershed in the western United
States for a one-year period to sample for the presence of Giardia cysts and Cryptosporidium
oocysts. Giardia cysts were detected in 31% (12/39) of samples collected from a lake receiving
sewage effluents and a river beginning at the lake and running through an area where there were
a number of cattle pastures. Higher Giardia cyst levels were found in the river downstream from
the cattle pastures: amean level of 22 cysts/100 L, and a range from 0 to 625 cysts/100 L. The
geometric mean Giardia cyst levels detected in the lake was 8 cysts/100 Lwith a range of Oto
222 cysts/100 L. A seasonal variation was found. Geometric means for the summer, fall, winter,
and spring were 35/100 L, 31/100 L, 0.7/100 L, and 0.1/100 L, respectively. Giardia cyst levels
were significantly correlated with Cryptosporidium oocyst levels, but no significant correlations
were observed between cyst levels and either total or fecal coliforms or turbidity.

Ongerth (1989a) conducted a study to assess Giardia cyst levels in three pristine
watersheds in Washington state. A membrane filter sampling method was utilized with 40 L
samples collected in the field and returned to the laboratory. Cyst recovery efficiencies were
measured (average = 21.8% ± 6%; range = 5-44%). Reported cyst levels were calculated based
on the percentage recovery measured in the positive controls, as well as the more routinely
m-s

-------
applied calculation using the percentage of the microscope slide counted. Giardia cysts were
found in 94 (43%) of the 222 samples. Based on the recovery efficiencies and sample volume,
the cyst levels ranged from 10 to 520 cysts/100 L. The median levels were extrapolated for the
three rivers, with the Green, Cedar, and Tolt Rivers being 6/100 L, 4/100 L, and 0.3/100 L,
respectively; distributions of cyst values were reported to be lognormal, but the cyst level (based
on mean values) and its variability (based on the slopes of the distributions) differed among the
rivers. Samples were also collected in tributaries to each of the three rivers, and the levels were
generally lower than found in the main-stem sites for each river. In addition, samples in each
river were collected above and below impoundments, and levels were generally higher below the
impoundments than above, although the differences were not statistically significant. No
statistically significant seasonal variations were observed over the nine-month sampling period;
levels varied in each river by less than a factor of 2. Ongerth (1989a) concluded that Giardia
cysts, even though present at low levels, appear to be continuously present in these relatively
pristine rivers.

Rose et al. (1991a) detected Giardia cysts in 16% of 257 water samples from 17 states at
an average level of 3 cysts/100 L. The arithmetic average of Giardia cysts ranged from <1 to
140 cysts/100 L, depending on the nature of the waters being sampled. The geometric mean was
4 cysts/100 Lfor all rivers. 11 cysts/100 L (maximum = 625/100 L) for polluted rivers, and 0.35
cysts/100 L (maximum = 12/100 L) for pristine rivers. For lakes, the geometric mean was
similar to rivers, 3 cysts/100 L. The geometric mean was 6.5 cysts/100 L (maximum = 156/100
L) for polluted lakes and 0.5 cysts/100 L (maximum = 7/100 L) for pristine lakes. Giardia cysts
were not detected in any of the drinking water samples. Levels of Giardia cysts were not found
to be correlated with turbidity, total coliform or fecal coliform water quality indicators. Samples
collected in the fall were more likely to be positive than samples collected during other seasons
and cyst levels tended to be higher in summer/fall samples than in winter/spring samples.
m-9

-------
       LeChevallier et al. (199la) sampled the source waters at 66 surface water treatment plants
in 14 states and one Canadian Province.  Giardia cysts were found in 81.2% (69/85) of raw water
samples with levels ranging from 4 to 6,600 cysts/100 L; the geometric mean of positive samples
was 277 cysts/100 L. About 13% of the 618 Giardia cysts observed in raw water samples
demonstrated morphological characteristics suggesting that these cysts may be viable.  Higher
cyst densities were associated with watersheds receiving decreasing protection.  The authors
noted that sources receiving industrial (urban) pollution contained 10 times more Giardia cysts
than protected watersheds. Unlike the Rose etal. (1988a, 1991a) results, LeChevallier et al.
(199la) found a significant relationship between Giardia cyst densities and total coliforms or
fecal coliforms, postulating that this difference might have been due to the types of water
samples analyzed in each study. Rose et al. (1988a,  1991a) examined relatively pristine waters
while LeChevallier et al. (1991a) examined a  variety of source waters, including those with high
levels of pollution as demonstrated by bacterial indicator counts. LeChevallier et al. (1991a)
concluded that as levels of pollution increase as indicated by higher bacterial indicator counts and
turbidities there is an increasing probability of the presence of Giardia cysts at higher densities.
                                           Ill-10

-------
Table III-l. Reported Data for Giardia in Surface Waters

Sample

River (downstream from
cattle pastures)
Lake (received sewage
effluents; upstream of
cattle pastures)
Pristine rivers in WA:
Green
Cedar
Toll
17 states in western US:
Polluted Rivers
Pristine Rivers
Polluted Lakes
Pristine Lakes
66 water treatment plants
in 14 states and 1
Canadian province:
Raw surface waters
Filtered drinking waters

No. of
Samples
(n)
19

222

257
38
59
24
34

85
83

Positive
Samples
(%)
21

16
26
7
33
12

81
17

Detection
Limit
(cysts/100 L)

Recovery
Efficiency
(%)

Mean Level of
Cysts"
(cysts/100 L)
22

6 (median)
4(median)
0.3(median)
o
3
n
0.35
6.5
0.5

277
4.45

Range of Cyst
Levels
(cysts/100 L)
0-625

0-222

10-520

-------

Sample

Continuation of
LeChevallieret al. 1991a
Raw surface waters:
1988-1993

1991-1993
Three water treatment
plants in Quebec, Canada
Raw water
Settled water
Filtered water
Finished water
Yukon Surface waters
Pristine
Drinking water intake
Ottawa, Ontario rivers

Lakes/rivers

WI waters: Streams

Lakes

No. of
Samples
(n)

347

262

17
74
34
31
22
42
41

147

210

179

Positive
Samples
(%)

94
66
o
3
o
3
32
17
78

31(48 in
winter;
25-29
other)
5(11 in
spring)
Detection
Limit
(cysts/100 L)

Recovery
Efficiency
(%)

42(<1NTU)
50(150
NTU)

Mean Level of
Cystsa
(cysts/100 L)

200

7-1,376
0.14-0.4
0.01-<.2
0.2

210

38(median)

1.2(2.6 in
spring,
medians)
Range of Cyst
Levels
(cysts/100 L)

2-4,380

<2-2,800
<1-15
<0. 1-0.1
<0. 1-0.7

<1-1.4
-------

Sample

86 sites in British
Columbia, Canada:
Raw water
Finished water
2 community supplies:
BMID Raw

Reservoir settled
Chlorinated
Tap
VID
Raw
Chlorinated
Treatment plant in
Germany:
Raw water
Backwash water
PA rivers:
Allegheny R.
Youghiogeny R.
Dairy farm stream
Settled water
Filtered water
Backwash water
No. of
Samples
(n)

153
91

70
75
77
7

53
53

12
50
24
24
22
23
24
22
Positive
Samples
(%)

64
59

100
99
77
71

100
98

83
84
63
54
55
8
0
8
Detection
Limit
(cysts/100 L)

Recovery
Efficiency
(%)

Mean Level of
Cystsa
(cysts/100 L)

2.9
2.1

229
95
31
9

30
14

(medians)
24.5
22.3-55
34
118
82
29

59
Range of Cyst
Levels
(cysts/100 L)

7-2,215
12-626
0.3-371
1.5-18.5

8-114
2-73

2-103
3-374
12-421
44-526
13-1,527
12-70

15-237

Viability
(%)
34C

11°

Reference

Isaac-Renton et
al.,1996

Karanis et al.,
1996a

States et al.,
1997

111-13
-------

Sample

NJ reservoirs:
Inlet
Outlet
PA filter plants:
Raw water intakes
Finished water
2 CA watersheds
Sampling method:
First flush (5 L, storm)
Filter (100 L , non-storm)
Grab (4 L, storm, non-
first flush)
NY reservoirs:
Catskill
Delaware
Malcolm Brooks
Selangor, Malaysia, 2
plants:
Raw water
Treated water
6 surface water treatment
plants in Germany:
Raw water
Treated water
Drinking water
No. of
Samples
(n)

60
60

148
148

20
87
21

20
20

105
150
47
Positive
Samples
(%)

13
15

23
0

60
29
19

36
29
46

90
0

64
20
15
Detection
Limit
(cysts/100 L)

2.4
6.2

Recovery
Efficiency
(%)
39

Mean Level of
Cystsa
(cysts/100 L)

1.9
6.1

0.23

1.2
0.7
1.3

88. 2( ?)
2.9( ?)
3.8( ?)
Range of Cyst
Levels
(cysts/100 L)

0.7-2.4
1.2-107

0.04-5.7

25-16,666
2-119
42-2,428

9.3(max.)
8.2(max.)
23.4(max.)

100-2,140

1,314 (max.)
19.2 (max.)
16.8 (max.)

Viability
(%)
20b

Reference

LeChevallier et
al.,1997

Consonery et
al.,1997

Stewart et al.,
1997

Okun et al.,
1997

Ahmad et al.,
1997

Karanis et al.,
1998

Geometric mean unless otherwise specified.
b By morphological criteria
0 By mouse infectivity
111-14
-------
LeChevallier et al. (1991b) also conducted analyses of treated water samples and reported
that cysts were detected in 17% of the 83 finished drinking water samples collected from the
same 66 surface water treatment plants. A viable type morphology was demonstrated by 13.3%
of the 46 Giardia cysts that were examined from these samples. The authors explained that this
does not mean the cysts were alive, but instead, that cysts which were distorted or shrunken were
probably dead. The communities served by the treatment systems in which the cysts were found
did not experience any apparent outbreaks of giardiasis, but the disease and waterborne outbreak
surveillance activities were not described for these communities. The investigators indicated that
24% of the treatment plants examined would not meet the risk level specified in the Surface
Water Treatment Rule (SWTR), and under cold water conditions (0.5°C) 46% of the plants
would not meet the risk level (U.S. EPA, 1989a). See Chapter VI for a discussion of the
recommended Giardia infection risk level.

In a subsequent study, LeChevallier and Norton (1995) updated their previous studies
from 1991, and presented data from an extensive monitoring program of drinking water source
waters. The levels of Giardia cysts were determined in a total of 347 surface water samples
collected from 1988 through 1993 at 72 treatment plants in 15 states and 2 Canadian provinces.
Giardia cysts were detected in 53.9% of the samples. This compares to 81.2% of samples
positive in their previous study. The authors dismissed method differences as possible
explanations for the decreased prevalence and suggested that occurrence levels may fluctuate due
to unknown causes or that there may have been an actual decline in levels over the four year
period examined. For the 1991 to 1993 sampling period, Giardia cysts were detected in 118
(45.0%) of 262 raw water samples. The geometric mean of positive samples was 200 cysts/100
L, ranging from 2 to 4,380 cysts/100 L. Based on microscopic examination, 14.6% (50/343) of
the cysts had potentially viable morphology. Giardia cysts were detected in 4.6% (12/262) of
filtered plant effluent samples at levels ranging from 0.98 to 9 cysts/100 L (mean = 2.6 cysts/100
L) during the 1991 to 1993 period. Because microscopic examination suggested that the majority
III-15
-------
of the organisms were dead, the authors concluded that there was little reason to believe that any
of the communities served by these plants was at risk for a waterborne outbreak of giardiasis.

Roach et al. (1993) analyzed for Giardia cysts in samples collected in Yukon surface
waters in the Canadian North. Pristine streams were sampled, and cysts found in 32% (7/22) of
the samples. A sample taken through the ice at Lake Laberge contained 7.5 cysts/100 L, and this
was the highest level found in the surface waters included in the study. The drinking water
intakes for two communities, Whitehorse and Dawson, were also sampled, and cysts were
detected in 17% (7/42) of the Whitehorse samples, but in none of the ten Dawson samples. Six
of the seven positive Whitehorse samples contained about 0.2 cysts/100 L, and one contained
about 1.4/100 L.

Payment and Franco (1993) sampled raw, settled, filtered and finished water from three
drinking water treatment plants in the Montreal, Canada, area. Two of the plants used the same
river for a source and the third used another river. For the raw water samples, 94% (16/17) were
positive at levels ranging up to 2,800 cysts/100 L. Settled water samples were positive 66%
(49/74) of the time at cyst levels up to 15/100 L, but only one (3%) of 34 filtered water samples
was positive at 0.1 cyst/100 L. Similarly, 3% of finished water samples (1/31) were positive for
Giardia (0.7 cysts/100 L).

Chauret et al. (1995) examined raw water samples from two Ottawa, Canada watersheds
characterized as "relatively pristine". They examined 41 raw water samples collected during the
summer months at 15 sites along the Ottawa River (23 samples) and Rideau River (18 samples)
watersheds. Giardia cysts were detected in 78% of samples at levels ranging from 1 to 52
cysts/100 L. An additional 12 raw water samples were collected at the intakes to two drinking
water treatment plants located on the Ottawa River. They found 83% of these samples positive
for cysts at levels up to 25/100 L, but none of an unspecified number of treated water samples
were positive. No significant correlation was found between the presence of Giardia and any of
III-16
-------
the microbial indicator examined (fecal and total coliforms, fecal streptococci, Aeromonas sp.,
Pseudomonas aeruginosa, Clostridium perfringens, algae and coliphages).

In another examination of pristine watersheds, Ongerth et al. (1995) collected duplicate
samples from five locations on each of two rivers in the Olympic Mountains, Washington. They
found greater numbers of Giardia cysts in water samples from the more heavily used watershed
compared to the adjacent watershed that experienced lesser human use. The authors reported a
relationship between the numbers of cysts in the water, the prevalence of cysts in selected animal
species, and the extent of human use of the watershed area. The numbers of cysts found in the
water samples ranged from 0.2/100 L to 3/100 L. Based on the water samples they analyzed,
they calculated that a median cyst level of 1 cyst/20 L can be expected in relatively pristine
waters.

In an investigation of Wisconsin waters, Archer et al. (1995) found Giardia at least once
in all 18 streams examined and in 30.9% (65/210) of all stream samples. The highest level in a
stream was 2,601 cysts/100 L from a site characterized as pristine. However, there was no
relationship between land use types and presence or absence of Giardia on a State-wide basis.
Although almost half of all samples collected in the winter were positive compared to about 25%
of samples collected during other seasons, the authors indicated that seasonal differences were
only weakly statistically significant (P=0.08). Giardia were not detected in any of 17 well water
samples collected from six wells. Cysts were found in 5% (9/179) drinking water intake samples
at levels ranging to 125 cysts/100 L(median, 1.2/100 L). None of 28 drinking water intake
samples collected during the summer were positive for cysts.

Norton et al. (1995) examined 147 samples from 15 sites in New Jersey source waters for
Giardia. These sites constituted 45% of the water supply for New Jersey. Sampling was
conducted over a one-year period during 10 sampling events. The mean cyst level in the 23% of
samples that were positive was 210/100 L; the range was 40 to 630/100 L.
III-17
-------
Isaac-Renton etal.(1996) surveyed 86 sites throughout British Columbia, Canada for
Giardia over a 12-month period. Although none of sites were downstream of large urban sewage
discharges, they were not considered pristine because they were open to public access and many
had agricultural activities in the vicinity. Of 153 raw water samples, 64% were positive for cysts
at geometric mean levels of 2.9/100 L. These positive samples came from 69% of the sites. No
seasonal variation was observed in raw water cyst levels. The 91 chlorinated drinking water
samples examined were positive 59% of the time at a geometric mean level of 2.1 cysts/100 L.
Thirty four percent (45/133) of cyst-positive samples inoculated into gerbils produced infection,
and the frequency of infectivity was higher when raw water concentrates were inoculated as
compared to treated water concentrates.

Isaac-Renton etal. (1996) also studied two community drinking water supplies in detail
over a 2-year period. One community, the Black Mountain Irrigation District (BMID), was
selected because it had previously been determined to have Giardia cysts more frequently and at
higher levels than most other sites examined. The second community, the Vernon Irrigation
District (VID), was selected because it frequently contained Giardia cysts but at lower levels
than the BMID supply. For BMID, 100% (70/70) of raw water samples were positive at cyst
levels ranging from 7 to 2,215/100 L; the geometric mean was 229/100 L. In samples collected
after reservoir settling, 99% (74/75) were positive with levels ranging from 12 to 626/100 L; the
geometric mean was 95/100 L. Chlorinated water samples were positive 77% (59/77) of the time
with levels ranging from 0.3 to 371/100 L;the geometric mean was 31/100 L. Seventy one
percent (5/7) of tap water samples were positive at levels from 1.5 to 18.5 cysts/lOOL. A
seasonal trend was noted with peak levels occurring in late autumn and early winter. Eleven
percent (11/125) of cyst-positive concentrates were infective for gerbils. Infectivity of
concentrates was less for settled water than for raw water and none of the chlorinated
concentrates inoculated into gerbils produced infections. For the VID, 100% (53/53) of raw
water samples were positive at 8 to 114 cysts/100 L; the geometric mean was 30/100 L.
Chlorinated water samples produced 98% (53/54) positives containing 2 to 73 cysts/100 L; the
III-18
-------
geometric mean was 14/100 L. No clear seasonal trend was observed in the VID samples. None
of 66 samples positive for cysts and inoculated into gerbils produced infections. In both
communities, cyst levels that spiked, i.e., rose to considerably higher values on a particular
sampling date than was indicated by the trend, were observed. Ong et al. (1996) further
evaluated sources of contamination of the watersheds in these two communities and found
significantly higher levels of cysts in raw water from a creek downstream of a cattle ranch as
compared to upstream samples in the BMID watershed. Differences were found in the
percentage of cattle infected on both watersheds and the authors indicated the importance of good
watershed management practices to providing best-quality drinking water supplies.

Wallis et al. (1996) analyzed 1,760 raw water, treated water, and raw sewage samples
from 72 cities across Canada for Giardia cysts. Fifty-eight of the municipalities treated their
drinking water by chlorination alone. The authors found Giardia cysts in 73% of the raw sewage
samples, 21% of the raw water samples, and 18% of the treated water samples. Water samples
from 74% (53/72) of the municipalities investigated were positive for cysts at least once.
Detailed quantitative data were not supplied but the authors indicated that most of the water
samples contained fewer than 2 cysts/100 L. The highest level of cysts encountered in a water
sample was 230 cysts/100 L found in a community experiencing an outbreak of waterborne
giardiasis. Sewage samples contained up to 88,000 cysts/Lbut the mode was under 1,000/L.
Cysts infective for gerbils were detected in 2.2% (5/223) of raw water samples and in 7.6%
(6/79) of treated water samples. Giardia cysts were recovered more frequently in late winter-
early spring and fall and in higher levels than in other seasons, although the cysts were recovered
in all seasons. By biotyping and karyotyping analyses, the cysts recovered were genetically
similar to those recovered in other areas of the world. These authors reported that in Canada
Cryptosporidium oocysts were less common than Giardia cysts. Based on monitoring data from
waterborne outbreak investigations, they proposed an "action level" of three to five cysts/100 L.
Ill-19
-------
States et al. (1997) conducted sampling in the vicinity of Pittsburgh, PA, in the Allegheny
and Youghiogheny Rivers. Samples were collected monthly for two years. The occurrence of
positive samples and the geometric mean values of detected Giardia cysts were, respectively, as
follows: in the Allegheny River, 63% and 34 cysts/100 L; in the Youghiogheny River, 54% and
118 cysts/100 L. Sources of cysts in the Allegheny River were identified as combined sewer
overflows (CSOs), a stream running through a dairy farm, and a sewage treatment plant.
Samples were collected from CSOs during five storm events. All samples were positive and the
cyst levels ranged from 3,750 to 114,000/100 L; the geometric mean was 28,681/100 L. In the
stream running through the dairy farm, 55% (12/22) samples were positive at cyst levels of 13 to
1,527/100 L (geometric mean = 82/100 L). Effluents from the sewage treatment plant were
positive 83% (20/24) of the time with densities of 102 to 4,614 cysts/100 L (geometric mean =
664 cysts/100 L). Filter backwash samples from the drinking water treatment plant were positive
8% (3/22) of the time atlevels of 15 to 237 cysts/100 L (geometric mean = 59/100 L).

LeChevallier et al. (1997) studied the potential for contamination of six open finished
water reservoirs in New Jersey. Ten samples of water from the influent and 10 samples of the
effluent were collected over a 12 month period at each of the six reservoirs. Five of the
reservoirs received finished water from conventionally filtered water plants, and the sixth
received water from a high quality unfiltered surface water system. All of the reservoirs had
fences but holes in the fences and evidence of human activity was evident at some of the sites.
Thirteen percent (8/60) of the influent samples were positive for cysts at geometric mean levels
of 1.9/100 L (median = 1.6). Of the effluent samples 15% (9/60) were positive at a geometric
mean of 6.1/100 L (median = 6.0). The authors indicated that most cysts were probably non-
viable based on morphological considerations. Twenty percent (9/45) of the cysts had potentially
viable morphology; the remainder were empty shells and were probably non-viable. The
differences in Giardia cyst levels between the influent and effluent samples were not statistically
significant. The authors assumed that the differences in levels were most likely due to
indigenous animal activity in the vicinity of the reservoirs.
111-20
-------
Stewart et al. (1997) studied three sample collection methods for storm events in two
California watersheds. Filter samples (100 L), first-flush samples (5 L) and grab samples (4 L)
were collected. The first-flush samples had the highest positivity rate, 60% (12/20) and also had
the highest cyst levels ranging from 25 to 16,666/100 L). Filter samples had an intermediate rate,
29% (25/87) with the lowest cyst levels (2 to 119/100 L). Only 19% (4/21) of the grab samples
had cysts, but the cysts (42-2,428) were at intermediate levels. There was a seasonal effect with
45% of 31 filter samples positive during October to January sampling while only 20% of 56
samples were positive during the February to July collection period. The authors concluded that
peak levels oi Giardia occur following storm events, especially the first storm of the season. An
unattended first-flush sampling device was effective and minimized logistical sampling
problems.

Okun et al. (1997) summarized the results of the New York City Department of
Environmental Protection monitoring for Giardia between June, 1992 and January, 1995 of three
of the City's drinking water reservoirs (Catskill, Delaware and Malcolm Brooks). The City's
supply is not filtered and the source water was sampled prior to chlorination. Positivity rates of
36%, 29% and 46% were found for the Catskill, Delaware and Malcolm Brooks reservoirs ,
respectively. Cysts were found at mean levels of 1.2/100 L (maximum = 9.3/100 L), 0.7/100 L
(maximum = 8.2/100 L) and 1.3/100 L (maximum = 23.4/100 L) atthe reservoirs, respectively.

Raw and finished water samples from surface water treatment plants in Pennsylvania
were examined intensively by Consonery et al. (1997) for Giardia between 1994 and 1996 as
part of Pennsylvania's Filter Plant Performance Evaluation (EPPE). Cyst levels ranging from 0.4
to 5.7 cysts/100 L (mean = 0.23/100 L) were found in 23% (34/148) of the raw water samples.
No cysts were detected in finished water samples during the last two years of theFPPE program.

Crockett and Haas (1997), in evaluating Philadelphia's watershed for Giardia and
Cryptosporidium, concluded that information on watershed characteristics was necessary in order
111-21
-------
to adequately interpret monitoring data on occurrence of the organisms. By identifying land uses
associated with protozoa sources and using factors like runoff during wet weather to set
priorities, it should be possible to identify the type of pollution (point or non-point), its general
location (immediate or upper regions of the watershed), and whether it occurred daily or only
during wet weather.

Crabtree et al. (1996) reported the analyses of water samples from various cisterns in the
U.S. Virgin Islands. Over a one-year sampling period, a total 45 samples were analyzed for
Giardia cysts. The reported average level ofGiardia was 1.09 cysts/100 L, with a range from <1
to 3.79 cysts/100 L. Of the samples analyzed, 26% were positive for cysts. The cisterns positive
for Giardia ranged from 18% in January 1993 to 54% in July 1992. These cisterns used roof
catchment systems and are obviously open to the air to collect the rain water. They were also
described as having a "dark and moist interior," so most maybe somewhat covered to reduce the
effects of evaporation. However, the investigators described the likelihood that droppings from
birds, rodents, and other animals may fall onto the collection areas, and that frogs may also enter
and/or live in the cisterns. This represents an unusual type of "surface water," but it is indicative
of the potential for the presence ofGiardia in another type of drinking water source that is used
in areas of the Caribbean, as well as other parts of the United States, where rainfall amounts are
low and surface water supplies are scarce.

Occurrence data for surface waters have also been reported from European and Asian
countries. Karanis et al. (1996a) examined raw river water and filter backwash waters from a
treatment plant in Germany. For the raw water samples, 83% (10/12) were positive for Giardia
cysts at levels of 2 to 103 cysts/100 L (median = 24.5). Similarly, 84% (42/50) of backwash
water samples were positive at densities ranging from 3 to 374 cysts/100 L; the median was 22.3
to 55.1, depending upon the method used to process the samples and the sampling depth in the
sedimentation basins used to hold the backwash water.
111-22
-------
Ahmad et al. (1997) investigated raw and treated waters from two drinking water
treatment plants in Selangor, Malaysia. The samples were collected on 10 separate occasions
between July, 1994 and January, 1995 and examined for Giardia and Cryptosporidium. Both
treatment plants used the same river as a source and provided conventional treatment. Ninety
percent (18/20) of the raw water samples were positive for Giardia at levels of 100 to 2,140
cysts/100 L. The two negative samples were collected during periods of high turbidity which
may have affected detection of cysts. The authors did not indicate the detection limit for the
method used. Cryptosporidium oocysts were not detected in the raw water samples and no
Giardia or Cryptosporidium were detected in the treated water samples. No correlation was
found between Giardia cyst levels in the raw water and fecal coliforms or physical parameters.

Karanis et al (1998) summarized the results of examining raw water, intermediate steps in
the treatment process and drinking water samples from six surface water treatment plants in
Germany. Sixty-four percent (67/105) of the raw water samples were positive for Giardia cysts
at an average level of 88.2/100 L (maximum = 1,314/100 L). Twenty percent (30/150) of the
samples examined after intermediate steps in the treatment process (e.g., flocculation, filtration)
were positive, and the average level was 2.86 cysts/100 L (maximum = 16.8/100 L). About 15%
(7/47) of the drinking water samples were positive. These authors had previously reported 84%
of backwash samples positive for Giardia with a maximum of 374 cysts/100 L (Karanis et al.,
1996a). In the most recent study, a single sample of backwash water from an activated carbon
filter was examined and found to contain 3,428 cysts/100 L.

Ho and Tarn (1998) examined samples over a 13-month period from two Hong Kong
rivers for Giardia, E. coli, and Cryptosporidium. Although oocysts were seldom detected in both
rivers, high levels of Giardia cysts were found: up to 46,880 cysts/100 L in one river and more
than 10,000 cysts in the other. The highest cyst levels were found at sampling sites located near
the more densely populated areas. No relationship was found between cyst levels and E. coli
levels in either river. The authors suggested that Giardia might be used as an indicator of
111-23
-------
sewage treatment plant efficacy and that consideration should be given to including a
parasitological indicator of water quality in local river waters.

2. Groundwaters

Hibler (1988) conducted analyses of drinking water samples from various municipal
systems, including some from groundwater sources. Giardia cysts were found in 19% (16/84) of
springs sampled, in 3% (2/63) of wells, and 19% (5/16) of infiltration galleries. His report did
not include quantitative data on the cyst levels detected. The author indicated that one of the
positive wells was "... essentially an infiltration gallery of the creek ..." which was about 25 feet
from the well. The other was a deep well that had been contaminated by priming with
contaminated river water.

Lee (1993) reported the contamination of two wells in Pennsylvania with G. lamblia by
surface streams less than 100 feet from the wells. Information about particulate analyses, water
quality analyses, well construction, stream flow, and aquifer characteristics helped demonstrate
the contamination source. G. lamblia was recovered from all samples collected from the wells.

Hancock et al. (1997) conducted a study involving 463 groundwater samples from 199
sites in 23 states in the United States. Information on aquifer type, geologic setting or well
construction details was not submitted with the samples. Giardia cysts were found in 6%
(12/199) of the sites: 14% (5/35) of the springs, 1% (2/149) of the vertical wells, 36% (4/11) of
the horizontal wells, and 25% (1/4) of the infiltration galleries. Of the total of 463 samples
analyzed, 23 (5%) were found to contain Giardia cysts, with the mean levels in cyst-positive
samples being 8/100 L and the median being 2/100 L. The range was 0.1 to 120 cysts/100 L.
These water samples were collected by drinking water utilities as part of their routine sampling.
Withdrawal rates from the ground water are high and these wells are flushed rapidly. These data
111-24
-------
suggest that groundwaters, including some types of springs, and especially groundwater supplies
under the influence of surface water, should not be assumed to be free of Giardia cysts.

n Occurrence in Soil

No published reports indicating the detection of Giardia cysts in soil were found. The
wide distribution of cysts in human and animal populations (Chapter III, Section A) indicates that
soil is being contaminated with Giardia through fecal deposition, irrigation and sewage treatment
practices. However, no data are available on cyst levels in soil, survival in soil, or transport
through soil matrices. A progress report from the New York State Water Resources Institute
Center for the Environment (Cornell University Whole Farm Planning Scientific Support Group,
1993) discussed development and evaluation of a provisional soil sampling protocol and
detection method for Cryptosporidiumparvum oocysts and Giardia cysts. The methods were
intended to be used to evaluate protozoan prevalence and transport on demonstration dairy farms
within a watershed. The investigators concluded that their soil smear/fluorescent antibody
method was a rapid method for the detection of oocysts in soil if the numbers of organisms were
above the detection limit (about 5 x 103 oocysts/g of soil, wet weight). The authors also
suggested that a higher sensitivity automated counting method may be needed for routine
counting of oocysts in soil and sediment. No detection limits were indicated for Giardia cysts,
and no data were presented on occurrence of either cysts or oocysts in the farm soil samples
examined.

II Occurrence in Air

No data were found indicating that Giardia cysts are released into the air and are
transported via the airborne route.

II Occurrence on Surfaces
111-25
-------
Cody et al. (1994) developed and evaluated a method for recovering Giardia cysts from
environmental surfaces, and then field tested the method in six commercial child day-care
centers. The method was capable of recovering spiked cysts from Formica® surfaces when they
were inoculated with 10 to!90 cysts on a surface area of 50 cm2 or with 10 to 20 cysts/400 cm2.
Cysts were also recovered from stainless steel surfaces inoculated with 20-186 cysts/400 cm2, but
mean recoveries were lower than from Formica® and false negatives were higher. Cysts were
not recovered from wood and fiberglass surfaces spiked with 190 cysts/400 cm2. In the field test,
cysts were detected on surfaces in two of the six day-care centers where samples were collected.
A total of 53 chairs and tables were examined and two fiberglass chairs (6%) and one Formica®
table (2%) surface were positive for Giardia cysts. The authors indicated that this was the first
reported successful method for examining environmental surfaces for disease-causing protozoan
organisms.

n Occurrence in Food

There is a lack of quantitative data on the occurrence of Giardia cysts in foods. Although
foodborne giardiasis outbreaks (Section F.4, Chapter IE) have involved fish, sandwiches,
vegetables, fruit and noodle salad, the source of cyst contamination of the food has generally
been epidemiologically associated with infected food handlers. In one instance the food had been
prepared in the home of women who had a diapered child and a pet rabbit, both positive for G.
lamblia. Sheep tripe soup was considered to be the vehicle of an outbreak of giardiasis affecting
two Turkish families (Karabiber and Aktas, 1991). It was suggested that Giardia cysts in deep
layers of the tripe were protected from heat inactivation during preparation of the soup.

Barnard and Jackson (1984) discussed methods for detecting Giardia cysts in foods and
reviewed outbreaks of foodborne giardiasis. They indicated there are no standardized methods to
examine foods for Giardia but described four techniques originally developed for clinical
111-26
-------
specimens that have been adapted to foods. One of these techniques (sedimentation/zinc sulfate
flotation) was used by Italian investigators to isolate Giardia cysts from lettuce. In that 1968
study, 75% of 64 heads of lettuce collected at random from four markets in Rome, Italy, were
found positive for cysts. Barnard and Jackson (1984) indicated this may have been the first
reported finding of Giardia in vegetables. Oliveira and Germano (1992) found Giardia sp. In
4% of lettuce and 10% endive sampled from vegetables traded in S. Paulo, Brazil. Bier (1991)
found recovery of seeded Giardia cysts from fruits and vegetables using a method employed by
the Food and Drug Administration. Improvements are needed in both sampling and analysis.

Rabbani and Islam (1994) found that foodborne outbreaks of giardiasis have been
suspected and suggested as early as 1922 and that water, vegetables and other foods were
reported contaminated with cysts. They indicated that eating raw or undercooked food because
of taste considerations or to conserve heat-sensitive nutrients might increase the risk of spreading
Giardia through food.

Payer et al. (1998) examined 360 oysters (Crassostrea virginicd) collected from six sites
in the Chesapeake Bay during May, June, August and September, 1997. Although presumptive
Cryptosporidium oocysts were identified in either hemocytes or gill washings from oysters
collected at all six sites, Giardia cysts were not found in any of the samples. The authors
indicated that cysts were detected in positive control specimens with the methods used and
suggested that failure to detect them in the oysters from the six sites may have been because cysts
were not present in the water, not removed by the oysters if they were present, were present
below detectable levels or because detection was masked by unknown factors in thehemolymph
or gill washings.

U Disease Outbreaks and Endemic Risks
111-21
-------
Waterborne giardiasishas been associated with the ingestion of contaminated water from
public and private water systems, from untreated and non-potable water sources, and during
water recreation and other water-related activities. Both visitors and residents have been affected
in outbreaks. In addition to outbreaks, endemic waterborne disease has been reported.
Outbreaks have also been traced to ice used for beverages and foods contaminated during their
preparation and handling. Person to person transmission has been documented among travelers
and in day-care settings, and a high risk of giardiasis has been associated with oral-anal sex
among male homosexuals. The relative importance of waterborne giardiasis compared to other
modes of transmission of giardiasis has not been well studied, but it has been estimated that
perhaps up to 60% of all cases may be waterborne (Bennett et al., 1987).

1. Outbreaks Associated with Drinking Water

The waterborne transmission of Giardia was suggested as early as 1946 by an outbreak of
amebiasis caused by sewage contamination of the water supply in a Tokyo apartment building;
Giardia was isolated from 86% of the occupants who had negative stools for E. histolytica and
experienced diarrhea with abdominal discomfort (Craun, 1990). Waterborne outbreaks of
giardiasis were not reported in the United States until 1965, most likely because the
pathogenicity of Giardia was still being debated. However, it appears in retrospect that a large
outbreak of 50,000 cases of illness in Portland, Oregon, in 1954-55 may have been caused by
Giardia and may possibly have been associated with drinking water (Veazie, 1979). In this
outbreak, an unusual prevalence of G. lamblia cysts was found in the stools of patients,
especially among those with a chronic illness of 14.8 days average duration characterized by
abdominal discomfort, diarrhea, loss of appetite, nausea, and weight loss. The first well
documented waterborne outbreak of giardiasis in the United States was recognized and
investigated primarily because a physician had developed characteristic symptoms of giardiasis
after returning from a ski holiday at Aspen, Colorado, in 1965 (Craun, 1990). Fluorescent and
111-28
-------
detergent tracers placed in Aspen's sewage system were detected in two of the city's wells, and
Giardia cysts were isolated from the sewage leaking from sewer mains near the wells.

Although waterborne outbreaks of giardiasis have also been reported in Europe, these
outbreaks have been more frequently reported in the United States (Hunter, 1997). Information
about waterborne outbreaks of giardiasis reported in the United States during 1965-1977 was
previously been summarized (1CAIR, 1984). Giardia., the most commonly identified cause of
waterborne outbreaks during this period of time, continues to be the most commonly identified
cause of outbreaks of waterborne disease in the United States (Craun, 1986; 1990; Moore et al.,
1993; 1994; Kramer et al., 1996). Statistics (Craun and Calderon, 1997) published on the causes
of waterborne outbreaks in the United States during the period 1971 to 1994 show 740
waterborne outbreaks were reported in community (37%), non-community (40%), individual
(11%) water systems, and during recreational activities such as swimming or backpacking when
contaminated water is ingested (12%). An etiologic agent was identified in 53% of these
reported waterborne outbreaks (Table in-2). Outbreaks were caused by protozoa (20%),bacteria
(15%), chemicals (10%), or viruses (8%). G. lamblia was the most frequently identified
etiologic agent of waterborne outbreaks (17%) during this period.

To be defined as a waterborne outbreak and included in these statistics, epidemiological
evidence must implicate water as the probable source of illness. Information about water system
deficiencies is usually available, but information about coliform bacteria or Giardia in source
waters or tap water is not always available. Epidemiological data have been weighted more
heavily than water quality data in defining waterborne outbreaks (Kramer et al., 1996).
Table III-2. Summary of Waterborne Outbreaks Reported in the U.S.A., 1971-94*
Etiology
Unidentified
Percent of Outbreaks
47% 1
111-29
-------
Giardia
Bacterial
Chemical
Viral
Other Parasitic
Total
17%
15%
10%
8%
3%
100%
*Includes outbreaks associated with accidental
ingestion during water recreation and consumption
of non-potable water.
a. Drinking Water Outbreaks in the United States 1965-1996

From 1965 to 1996, 118 outbreaks of giardiasis and 26,305 cases of illness (data
compiled by Craun for this document) were associated with the consumption of contaminated
drinking water from public and individual water systems in the United States (Table III-3).
Cases of illness in these outbreaks include laboratory confirmed cases and persons with
symptoms compatible with giardiasis. No deaths were associated with these outbreaks. Most
outbreaks (70%) and cases of illness (88%) occurred in community water systems; 22% of the
outbreaks and 12% of the cases occurred in non-community water systems. Thirteen additional
outbreaks during this period were associated with water recreational and other water-associated
activities (discussed in Section G. 2, Chapter in). Waterborne outbreaks were reported in 27
states; five states reported only water recreational-associated outbreaks.

Colorado reported 45 outbreaks associated with drinking water. Pennsylvania reported
nine outbreaks, the second highest number. Waterborne outbreak statistics do not provide the
actual incidence of waterborne outbreaks or disease and are largely a reflection of surveillance
activities of local and state health agencies during the various time periods (Craun, 1986; 1996).
Many factors influence the degree to which outbreaks are recognized, investigated, and reported
in any single year, including interest in the problem and the capabilities for recognition and
investigation at the state and local level (Frost et al., 1996; Berkelman et al., 1994). Improved
111-30
-------
surveillance activities have resulted in increased reporting of waterborne outbreaks (Craun, 1986;
Foster, 1990; Craun, 1990; Harter et al., 1985; Hopkins et al., 1985). For example, during a three
year period of intensive waterborne disease surveillance in Colorado from 1980 to 1983, 18
waterborne outbreaks were reported compared to only six during the previous three year period
when a passive surveillance program was in effect (Hopkins et al., 1985). Improved surveillance
activities were felt to be responsible for several states reporting large numbers of outbreaks
during certain time periods (Craun, 1986). Estimates suggest that in the United States one-half to
one-third, or even as few as 10%, of waterborne outbreaks may be detected, investigated, and
reported (Craun, 1986; 1996).

Information was recently made available for waterborne outbreaks reported during 1995-
1996 (Levy, in press 1998). An outbreak of 10 cases of giardiasis was associated with drinking
untreated surface water in Alaska, and 1449 cases were associated with inadequate filtration of a
surface water source in a community of 20,000 persons in New York. The outbreak in New
York occurred during December 1995 to February 1996 (Hopkins et al., 1998). Heavy rains
occurred before the outbreak, and turbidity in filtered water exceeded regulatory limits before
and during the outbreak; there was no interruption of chlorination. In 1996, an outbreak of
giardiasis was associated with a contaminated wading pool in Florida.
111-31
-------
Table III-3. Drinking-Water Related Giardiasis Outbreaks Reported in the U.S.A., 1965 to 1996
Year

1965
1966-8
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
Community Water
Systems
Outbreaks
1
0
0
1
0
1
2
2
0
1
2
2
5
7
8
9
17
3
1
4
2
2
o
3
i
0
2
2
3
1
0
Cases
123
0
0
34
0
12
52
4,878
0
600
950
5,130
3,789
1,724
265
497
2,216
463
703
251
633
262
380
123
0
95
27
358
1449
0
Non-community Water
Systems
Outbreaks
0
0
1
0
0
3
1
1
0
2
2
1
2
0
2
2
0
1
2
1
0
0
1
2
2
0
0
0
0
0
Cases
0
0
19
0
0
112
16
18
0
39
62
23
2,120
0
39
60
0
400
38
23
0
0
152
42
28
0
0
0
0
0
Individual Systems &
Non-potable Water
Outbreaks
0
0
0
0
0
0
1
1
1
0
0
1
0
1
1
1
1
1
0
0
0
0
0
0
0
0
0
0
1
0
Cases
0
0
0
0
0
0
5
34
9
0
0
18
0
6
7
4
4
3
0
0
0
0
0
0
0
0
0
0
10
0
Total
Outbreaks
1
0
1
1
0
3
4
4
1
3
4
4
7
8
11
12
18
5
3
5
2
2
4
3
2
2
2
3
2
0
Cases
123
0
19
34
0
124
73
4,930
9
639
1,012
5,171
5,909
1,730
311
561
2,220
866
741
274
633
262
532
165
28
95
27
358
1459
0
111-32
-------
Total 82 25,014 26 3,191 10 100 118 28,305
Waterborne outbreaks of giardiasis have occurred primarily in surface water systems.
Statistics available for the period 1971 to 1994 allow for a comparison of etiological agents that
were associated with untreated or inadequately treated surface water and groundwater source.
An etiologic agent was identified in most outbreaks (66%) caused by inadequately treated surface
water but in only 38% of outbreaks caused by inadequately treated groundwater (Table III-4).
Most (78%) outbreaks of known etiology in inadequately treated surface water systems were
caused by Giardia. In untreated or inadequately treated groundwater systems, Giardia caused
12% of the outbreaks ofknown etiology.
Table III-4. Etiology of Waterborne Outbreaks Caused by Contamination of Untreated Water
Sources, Inadequate or Interrupted Disinfection, and Ineffective Filtration, U.S.A., 1971-94*
Surface water systems
Etiologic Agent
Giardia
Undetermined
Bacterial
Viral
Chemical
Other Protozoa
Total
Outbreaks
81
53
10
6
4
3
157
Cases of Illness
22,171
20,791
6,108
2,049
268
416,240
467,627
Hospitalized Cases
28
31
30
0
0
4,400
4,489
Groundwater systems
Etiologic Agent
Undetermined
Bacteria
Viral
Chemical
Giardia
Other Protozoa
Total
Outbreaks
212
47
35
26
16
6
342
Cases of Illness
49,351
8,578
8,686
978
336
3,573
71,502
Hospitalized Cases
142
481
85
19
7
8
742
111-33
-------
*Outbreaksin surface water and groundwater systems do not include those caused by distribution/storage
contamination, recreational activities, or miscellaneous and unknown causes (Unpublished data compiled
by G. Craun, 1998).
Craun (1996) computed outbreak rates for community systems using either surface or
groundwater sources. Only outbreaks caused by source contamination and treatment
inadequacies were considered. These statistically stable rates were used to compare outbreak
risks among the different types of water sources and treatment. Community water systems that
filter and disinfect surface water experienced 6.3 outbreaks per 1000 systems (95% C.1=4.2-9.1).
Community water systems that used surface water sources with disinfection as the only treatment
experienced an outbreak rate of 52.8 per 1000 systems (95% C.I.=40.6-67.6), eight times the rate
of outbreaks in filtered surface water systems. Outbreaks that contributed to the high outbreak
rates in unfiltered community water systems were caused primarily by G. lamblia.

Of the 127 waterborne outbreaks of giardiasis reported during 1971-94, 109 were
attributed to drinking water sources and had adequate information to describe the water system
deficiency (Table ni-5).

Table III-5. Causes of Waterborne Outbreaks of Giardiasis, USA, 1971-94
|| Water Source, Treatment, or Deficiency I Outbreaks l|
Surface Water Source:
Untreated
Chlorination Only
Filtered (includes outbreaks where filtration was by-passed)
Ground Water Source:
Untreated
Chlorination Only
Filtration
Contamination of Distribution System or Storage
Use of Water not Intended for Drinking or Ingestion during Water
Recreation or Other Water Activities
Insufficient Information
Totals
13
51
17
8
7
1
12
14
4
127
111-34
-------
Eighty-one (74%) of these outbreaks were caused by inadequate treatment of surface water with
the majority occurring in surface water systems that were chlorinated but not filtered. Eighteen
(14%) outbreaks occurred in filtered surface water systems when filtration deficiencies were
noted or pretreatment or filtration was by-passed. Almost 15% of giardiasis outbreaks occurred
in water systems where groundwater sources were inadequately protected and/or treated; 11% of
the drinking water-associated outbreaks of giardiasis were attributed to distribution system
deficiencies.

Table III-6 presents basic information about all waterborne outbreaks of giardiasis
reported in the United States since 1965: state and beginning month of each outbreak, type of
water system, deficiencies causing the outbreak, and information available about bacterial water
quality, turbidity, and qualitative and quantitative information aboutGiardia cysts in source or
treated water. Outbreaks described in the scientific literature since the ICAIR document are also
discussed in this section.

A large outbreak of 703 confirmed and 3800 estimated cases of giardiasis occurred during
November 1985 to January 1986 in Pittsfield, Massachusetts, where chlorination was the only
treatment of surface water (Kent et al., 1988). Illness peaked two weeks after an auxiliary
surface water reservoir was placed into service, causing increased complaints of turbid water. An
epidemiological study found that giardiasis was higher among residents of areas supplied by the
auxiliary reservoir. Giardia cysts were detected in water samples from both the auxiliary
reservoir and the two other surface water reservoirs. Beavers and muskrats in the vicinity of the
reservoirs were found to be infected with Giardia and may have contributed to the contamination
of the reservoirs. The chlorinator at the auxiliary reservoir also malfunctioned during the entire
month of November, but it is not clear that the malfunction contributed to the outbreak, since the
chlorine concentrations and contact times which were normally provided were insufficient to
inactivate 99.9% of Giardia cysts in this water (Craun, 1990). Five days after the episode of
turbid water, greater than 5 coliforms/100 mL were found in five of seventeen water samples
111-35
-------
from the distribution system. Check samples were taken daily for the next seven days, but
apparently no attempt was made at this time to determine the cause of contamination or repair
and adjust the chlorinator, as coliforms continued to be found in the check samples.
111-36
-------
Table III-6. Outbreaks of Giardiasis Associated with Drinking Water Systems, USA, 1965-96
Location
Date
Cases
Water
System8
Deficiency
Water Quality
Giardia
Found
Aspen, CO
Lookout
Mountain, CO
Idyllwild, CA
Campground,
Boulder, CO
Resort,
High Co, CO
Camp,UT
Park Co, CO
Grand Co, CO
Farm, TN
Essex Center, VT
Danville Green,
VT
Grand Co, CO
Meriden,NH
Dec. 1965
Aug. 1969
May 1970
May 1972
May 1972
Sept. 1972
Dec. 1972
July 1973
Aug. 1973
Nov. 1973
Dec. 1973
June 1974
June 1974
123
19
34
28
24
60
12
16
5
32
20
18
78
C
NC
C
NC
NC
NC
C
NC
I
C
C
NC
C
Sewage contamination of wells; also used was
an unfiltered, disinfected stream source
Unknown
Surface water source with filtration and
disinfection; filters used intermittently
Unfiltered, chlorinated surface water;
interrupted disinfection
Unfiltered, chlorinated surface water;
interrupted disinfection
Untreated surface water
Septic tank seepage into wells; no treatment
Untreated river water
Seepage from a pit privy contaminated cistern
Unfiltered, chlorinated surface water; pond
Unfiltered, chlorinated surface water; pond
Unfiltered, chlorinated surface water
Unfiltered, chlorinated surface water; river
Coliform contamination of
water distribution system
Unknown
Positive tap water sample
during outbreak (>16
coliforms/100 mL); negative
samples before outbreak
Unknown
Unknown
Unknown
Unknown
Unknown
Unknown
Low coliform count after
chlorination
Septic tank drainage into
pond; no chlorine residual
Unknown
Unknown

111-37
-------
Table III-6. (Cont.) Outbreaks of Giardiasis Associated with Drinking Water Systems, USA, 1965-96
Location
Date
Cases
Water
System8
Deficiency
Water Quality
Giardia
Foundb
Campers, UT
Rome, NY
Campers, ID
Grand Co., CO
Camas , WA
Camp, Estes Park,
CO
Berlin, NH
Resort, Wasatch
Co.,UT
Hotel, Glacier
Park,MT
W. Sulfur Springs,
MT
Vail, CO
Sept. 1974
Nov. 1974
Sept. 1975
Feb. 1976
May 1976
June 1976
April 1977
June 1977
July 1977
July 1977
Mar. 1978
34
4,800
9
12
600
27
750
7
55
200
5,000
I
C
I
NC
C
NC
C
NC
NC
C
C
Untreated river water
Unfiltered, chlorinated surface water; lake
Untreated surface water
Untreated river water
Wells and river water treated by pressure
filtration and chlorination; deficiencies in
operation of filters and disinfection
Unfiltered, chlorinated surface water; lake
River water sources with conventional and
pressure filtration and chlorination;
deficiencies in construction of rapid sand filter
and operation of pressure filter
Untreated well water under the influence of
surface water (river)
Use of untreated surface water
Unfiltered, chlorinated surface water
Inadequate filtration of river water
Fecal coliforms42/100 mL
527 tap samples Nov-June
negative for coliforms; Feb
samples: 20, 20,30, and 40
coliforms/100 mL
Unknown
23 coliforms/100 mL
Coliforms found but treated
water met coliform MCL;
turbidity was 42 NTU.
Water source contained <0.5
fecal coliforms/100 mL.
Routine distribution system
samples were consistently
negative and met MCL.
14-110 coliforms/100 mL;
turbidity=1.4 NTU.
Unknown
Unknown
14 coliforms/100 mL,MCL
not exceeded; turbid. =3 NTU

Yes

Yes
Yes
Yes

Yes
111-38
-------
Table III-6. (Cont.) Outbreaks of Giardiasis Associated with Drinking Water Systems, USA, 1965-96
Location
Date
Cases
Water
System8
Deficiency
Water Quality
Giardia
Foundb
Picnic, UT
Camp,WA
Town, NY
CA
Town,NH
Campground, AZ
Town, CO
Bradford, PA
Town, OR
Town, OR
Town, WA
Town, WA
Apart. Bid., PA
Backpackers, W A
Red Lodge, MT
Town, OR
Town, Alaska
July 1978
Aug. 1978
Nov. 1978
Feb. 1979
April 1979
June 1979
June 1979
Sept. 1979
Oct. 1979
Dec. 1979
Jan. 1980
Mar. 1980
Mar. 1980
Apr. 1980
June 1980
June 1980
Sept. 1980
18
23
130
120
50
2,000
53
3,500
66
120
79
578
15
6
780
63
189
I
NC
C
c
C
NC
C
c
c
NC
C
c
c
I
c
c
c
Untreated irrigation water
Untreated surface water
Unfiltered, chlorinated surface water
Unfiltered, chlorinated surface water
Unfiltered, chlorinated surface water
Distribution system deficiency; well water
Inadequate filtration of river water
Unfiltered, chlorinated surface water
Unfiltered, chlorinated surface water
Unfiltered, chlorinated river water
Untreated surface water
Inadequate filtration of river water
Treatment deficiency; spring
Untreated stream water
Unfiltered, chlorinated stream water
Unfiltered, chlorinated stream water
Distribution system; cross-connection
Not Reported
Not Reported
Not Reported
Not Reported
Not Reported
9 fecal coliforms/100 mL
Not Reported
High coliforms count;
turbidity =10NTU
Not Reported
Not Reported
Heavy rains; high turbidity
Not Reported
Not Reported
Not Reported
Not Reported
No coliforms detected
>2400 fecal coliforms/100
mL

Yes

Yes
Yes
Yes

111-39
-------
Table III-6. (Cont.) Outbreaks of Giardiasis Associated with Drinking Water Systems, USA, 1965-96
Location
Date
Cases
Water
System8
Deficiency
Water Quality
Giardia
Foundb
WA
Town, CO
Town, CO
Town, CO
CO
Motel, WI
Town, VT
Hikers, CO
Factory, FL
Ski area, CO
Town, CO
Town, CO
Ski resort, CO
Ski resort, CO
Ski resort, CO
Ski resort, CO
Home,VA
CO
Sept. 1980
June 1981
July 1981
Aug. 1981
Sept. 1981
Sept. 1981
Oct. 1981
Oct. 1981
Oct. 1981
Nov. 1981
Dec. 1981
Dec. 1981
Jan. 1982
Feb. 1982
Mar. 1982
April 1982
April 1982
July 1982
20
8
30
110
32
25
22
7
7
38
14
18
10
17
4
8
4
72
C
C
C
C
C
C
C
I
NC
C
C
C
C
C
C
C
I
C
Distribution system; cross-connection
Inadequate filtration of stream water
Unfiltered, chlorinated stream water
Treatment deficiency; creek
Treatment deficiency; creek
Insufficient data; well water
Untreated surface water, spring
Untreated river water
Water not intended for drinking
Inadequate filtration; stream
Interrupted chlorination; unfiltered spring
Inadequate filtration; stream
Inadequate filtration; stream
Inadequate filtration; stream
Inadequate filtration; stream
Insufficient data
Seepage of sewage into untreated well
Unfiltered, chlorinated stream water
Not Reported
Not Reported
Not Reported
Not Reported
No coliforms detected
Not Reported
Not Reported
Not Reported
Not Reported
Fecal coliforms found
2-27 fecal coliforms/100 mL
Not Reported
Not Reported
Not Reported
Not Reported
Not Reported
Not Reported
Not Reported

Yes

111-40
-------
Table III-6. (Cont.) Outbreaks of Giardiasis Associated with Drinking Water Systems, USA, 1965-96
Location
Date
Cases
Water
System8
Deficiency
Water Quality
Giardia
Foundb
Ranch resort, CO
Reno,NV
Ski resort, CO
Community, OR
Community, NH
Community, VT
Community, CO
Community, CO
Community, CO
Community, UT
Trailer park, FL
CO
Community, CO
Community, NH
Household, VA
Community, MT
Community, NM
Community, UT
Aug. 1982
Aug. 1982
Sept. 1982
Sept. 1982
Oct. 1982
Oct. 1982
Jan. 1983
Jan. 1983
Jan. 1983
Jan. 1983
Mar. 1983
May 1983
May 1983
May 1983
June 1983
July 1983
Aug. 1983
Aug. 1983
28
342
32
9
13
22
4
11
17
41
o
3
11
10
7
4
100
100
1,272
NC
C
NC
C
C
C
C
C
C
C
C
C
C
C
I
C
C
C
Treatment deficiency; river
Unfiltered, chlorinated surface water
Interrupted chlorination of stream water
Unfiltered, chlorinated surface water
Insufficient data
Untreated ground water; springs
Untreated surface water, river
Untreated surface water, river
Inadequate filtration of river water
Contamination of water main under repair
Inadequate chlorination; wells
Inadequate filtration of lake water
Filtration bypassed; stream
Unfiltered, chlorinated surface water
Untreated well water
Unfiltered, chlorinated surface water
Unfiltered, chlorinated surface water
Contamination of broken water main
Not Reported
No coliforms detected
16-69 coliforms/100 mL
Not Reported
Not Reported
2 coliforms/100 mL
No coliforms detected
No coliforms detected
No coliforms detected
No coliforms detected
Not Reported
No coliforms detected
No coliforms detected
Not Reported
9.2 fecal coliforms/100 mL
Not Reported
Not Reported
Coliforms in 3 samples

0. 26/100 L
Yes

Yes

111-41
-------
Table III-6. (Cont.) Outbreaks of Giardiasis Associated with Drinking Water Systems, USA, 1965-96
Location
Date
Cases
Water
System8
Deficiency
Water Quality
Giardia
Foundb
Community, CO
16 commun., PA
Community, PA
CO
Community, ID
Community, ID
Community, PA
Ski resort, CO
Community, OR
Camp, AK
Community, AK
Factory, NY
Resort, VA
Pittsfield, MA
Trailer park, VT
Prison, CA
Resort, CO
Oct. 1983
Oct. 1983
Oct. 1983
Nov 1983
Nov. 1983
Nov. 1983
Feb. 1984
Mar. 1984
July 1984
Sept. 1984
Oct. 1984
Feb. 1985
April 1985
Nov. 1985
Jan. 1986
Apr. 1986
Aug. 1986
11
366
135
13
44
71
298
400
42
o
3
123
6
32
703
68
127
23
C
C
C
C
C
C
C
NC
C
I
C
NC
NC
C
C
C
NC
Coss-connection
Unfiltered, chlorinated surface water
Unfiltered, chlorinated surface water
Inadequate filtration of surface water
Unfiltered, chlorinated surface water; reservoir
Unfiltered, chlorinated surface water; river
Inadequate filtration of surface water; river
Unfiltered, chlorinated surface water; pond
Inadequate filtration of surface water; river
Untreated surface water
Unfiltered, chlorinated surface water; reservoir
Distribution system; cross-connection
Inadequate chlorination; spring
Interrupted chlorination; unfiltered water from
reservoir
Unfiltered, chlorinated surface water; river
Leaking, broken water lines
Interrupted chlorination; well
106 coliforms/100 mL
Not Reported
Not Reported
Not Reported
36 coliforms/100 mL
No coliforms detected
Not Reported
Not Reported
Not Reported
Not Reported
Not Reported
24 coliforms/100 mL
Not Reported
>5 & 8-41 coliforms/100 mL;
period of high turbidity
Water source
>20,000 coliforms/100 mL
Not Reported
12 coliforms/100 mL

Yes
Yes

265/100 L
0. 02/100 L

0.02-
0. 07/100 L
Yes

111-42
-------
Table III-6. (Cont.) Outbreaks of Giardiasis Associated with Drinking Water Systems, USA, 1965-96
Location
Date
Cases
Water
System8
Deficiency
Water Quality
Giardia
Foundb
Community, ME
City, NY
Community, PA
Community, CT
Resort, CO
Community, PA
Community, CO
Community, NY
Prison, NY
Community, NY
Lodge, AK
Resort, VT
Community, CO
Park,CA
Park, PA
Trailer park, CO
Community; NV
Trailer park, PA
Nov. 1986
Nov. 1986
April 1987
July 1987
Feb. 1988
July 1988
Feb. 1989
April 1989
June 1989
July 1989
Mar. 1990
Mar. 1990
Aug. 1990
July 1991
Sept. 1991
Mar. 1992
Mar. 1992
Jan. 1993
12
44
513
120
90
172
19
308
152
53
18
24
123
15
13
15
80
20
C
C
C
C
C
C
C
C
NC
C
NC
NC
C
NC
NC
C
C
C
Unfiltered, chlorinated surface water; river
Unfiltered, chlorinated surface water; lake
Unfiltered, chlorinated surface water; river
Distribution system; cross-connection
Unfiltered, chlorinated surface water; river
Unfiltered, chlorinated surface water; lake
Unfiltered, chlorinated surface water; river
Unfiltered, chlorinated surface water; reservoir
Unfiltered, chlorinated surface water; reservoir
Unfiltered, chlorinated surface water; lake
Untreated surface water, river
Unfiltered, chlorinated surface water; lake
Inadequate chlorination of spring; surface
water influence
Cross-connection with raw water; spring
Inadequate chlorination of well
Untreated ground water; well
Unfiltered, chlorinated surface water; lake
Inadequate filtration of well
1 of 16 samples positive
Not Reported
Turbidity >1 NTU
80-800 coliforms/100 mL
Not Reported
Not Reported
No coliforms detected
Not Reported
Not Reported
Not Reported
Not Reported
No coliforms detected
35-200 coliforms/100 mL
No coliforms detected
No coliforms detected
1 of 2 samples positive
No coliforms detected
Not Reported

0. 1-0.3/100 L

<0. 01/100 L

0. 02/100 L

<0. 01/100 L

Yes

50/100 L

111-43
-------
Table III-6. (Cont.) Outbreaks of Giardiasis Associated with Drinking Water Systems, USA, 1965-96
Location Date Cases Water Deficiency Water Quality Giardia
System8 Foundb
Subdivision, SD
Prison, TN
Community, NH
Community, NH
AK
City, NY
Sept. 1993
Mar. 1994
May 1994
May 1994
Aug. 1995
Dec. 1995
7
304
18
36
10
1449
C
C
C
C
I
C
Untreated well in fissured rock
Cross-connection
Unfiltered, chlorinated surface water; reservoir
Unfiltered surface water; lake
Unfiltered surface water
Filtered surface water; lake
Fecal, total coliforms detect.
No coliforms detected
No coliforms detected
Coliforms detected

Turbidity exceeded MCL
32/100 L
5807 L
Yes
No

No
*Data compiled by G. Craun, 1998.
a C - Community Water System;NC - Non-community Water System; I - Individual Water System Including Personal Use of Non-potable Water Sources
b Yes/No - Giardia Cysts Found in Source Water or Distribution System; blank -No Samples Collected or Reported
111-44
-------
An outbreak of 342 confirmed cases of giardiasis in Pennsylvania during December 1983
and January 1984 is important because it occurred in a surface water system with conventional
filtration that routinely met water quality standards (Akin and Jakubowski, 1986). Operational
deficiencies were found. An unusually high demand for water had left an insufficient volume of
water for filter backwashing, and turbidity breakthrough occurred in the filters because of longer
filter runs. A free chlorine residual of 1.0 to 1.3 mg/1 was maintained, and the system met
coliform standards; however, turbidity increased to 2.80 NTU (a weekly average). Hopkins and
Juranek (1991) reported an outbreak among university students and staff on a geology field
course in Colorado in June 1983 where the risk of stool positivity was strongly related to the
amount of untreated water consumed

Several outbreaks were caused by contamination of water mains through cross-
connections, damage of mains, and repair of mains. The largest outbreak of this type, 2000 cases
at a private campground in Arizona, occurred when sewage-contaminated water entered the
drinking water system through a direct cross-connection between the potable water system and a
pipe carrying sewage effluent for irrigation (Starko et al., 1986; Craun, 1986). In Tooele, Utah,
1272 cases occurred when contamination entered a water transmission line which had been
damaged by mud slides and flooding due to heavy rains; routine water samples were positive for
coliforms prior to this outbreak. (MMWR, 1983). Contaminated water during the repair of a
water main was identified as the cause of 41 cases of giardiasis in another outbreak which
occurred in Utah (Craun, 1986).

A waterborne giardiasis outbreak at Aspen Highlands, Colorado, in November 1981
affected a small number of people but is important because a clear dose-response relationship
was found for water consumption and clinical illness and it offers evidence of acquired immunity
to Giardia (Istre et al., 1984). An attack rate of 42% was found among persons who drank six or
more glasses of water per day. Residents who lived in the area for more than two years had a
111-45
-------
lower attack rate for illness than short-term residents. Giardia cysts were isolated from raw and
treated water samples, and beavers were suspected as the source of contamination.

Vogt et al. (1984) investigated an outbreak in a small Vermont community in December
1981. A serological survey found people who drank the town water had a significantly higher
antibody titer to G. lamblia than those who had not. G. lamblia was identified from only four
cases. The community's water supply source was an unprotected, unfiltered, and unchlorinated
spring. Birkhead et al. (1989) investigated an outbreak of gjardiasis that affected 37 (30%)
residents of a trailer park in rural Vermont in 1986. An increased risk of disease was associated
with increased water consumption from the trailer park's water system, a chlorinated, unfiltered
surface water supply. Giardia cysts were found in the water. Convalescent sera from 24 ill
residents and from 20 nonresidents were tested by enzyme-linked immunosorbent assay (ELIS A)
for immunoglobulin G (IgG), IgM, and IgA antibodies to G. lamblia. Higher IgA and IgG
antibody levels were detected in the ill residents compared to nonresident controls. Mne ill
residents had a higher median level of IgA antibody but not of IgG or IgM than 15 healthy
residents. In addition, IgA antibody levels for G. lamblia were higher in those who consumed tap
water than in those who did not. The authors suggested that elevated IgA antibody to G. lamblia
may effectively determine exposure to cyst contaminated water and subsequent illness during
waterborne outbreaks.

b. Waterborne Outbreaks in Canada

Allison (1984) discussed a giardiasis outbreak in a Saskatchewan municipality in 1982
where water may have been the source of infection and the need for better disease surveillance.
Several waterborne outbreaks were reported in British Columbia where most municipal water
systems are not filtered and, in some instances, not disinfected (Isaac-Renton, 1994a). In
Canada, suspected waterborne outbreaks occurred at Banff and Edmonton, Alberta (Wallis et al.,
1986). Over 800 people were infected during the Edmonton outbreak. In the Banff outbreak,
111-46
-------
infected beavers were found in the creek which supplied the town with water (Wilson et al.,
1982). Three documented waterborne outbreaks of giardiasis occurred within five years in
British Columbia (Isaac-Renton et al., 1987); one was reported at 100 Mile House (Wallis et al.,
1986). All water systems used surface water without filtration, and Giardia-positive beavers
were found in the water sources. An estimated 3,125 cases of giardiasis occurred in awaterborne
outbreak in Penticton, British Columbia from June to August 1986 (Moorehead et al., 1990).
The city obtained water from a creek, lake and well. The creek water was very turbid, and
although no Giardia cysts were isolated from water, beavers trapped in the creek were stool
positive. Two muskrats trapped from the lake were also positive.

Isaac-Renton etal. (1993) investigated an outbreak of waterborne giardiasis inCreston,
British Columbia, where 124 laboratory-confirmed cases were identified during an eight-week
period. This was the second outbreak in this town (Isaac-Renton et al., 1994b). A previous
waterborne outbreak of giardiasis had occurred five years earlier but had not caused the small
rural town to initiate treatment of the drinking water nor change their water source. An isolate of
the outbreak-associated Giardia cysts was obtained from the contaminated drinking water, and
the antigen from this strain was utilized in the serological testing. Sera from symptomatic and
asymptomatic residents were tested byELISA which positively identified 84% of the 124
laboratory-confirmed cases. There was greater success in identifying elevated anti-Giardia IgG
levels compared with identifications of elevated levels of IgA or IgM. Residents who had been
infected during the first outbreak were significantly less likely to be infected during the second
outbreak suggesting an acquired immunity to giardiasis that may last for at least five years. Of
54 laboratory-confirmed cases from the first outbreak, only 4% were infected during the second
outbreak, whereas of 57 residents who had moved to the town after the first outbreak, 68% were
infected.

c. Waterborne Outbreaks in Europe
111-47
-------
The first recognized waterborne outbreak of giardiasis in the United Kingdom occurred in
1985; 108 laboratory-confirmed cases were reported. (Jephcottet al., 1986). Neither coliforms
nor Giardia were detected in water samples collected from the distribution system of the filtered
and disinfected water system. The epidemiologic investigation found a strong association
between illness and consumption of water during a time when water mains were repaired. It was
suspected that contamination occurred either during the repairs or from backsiphonage from
pressure changes associated with the repair. Neringer et al. (1987) reported the first recognized
outbreak of giardiasis in Western Europe. The outbreak occurred in Mjovik, Sweden, and had
multiple etiologies. During the first few days after sewage contamination of the village well,
76% of the population became ill with gastroenteritis with at least 56 cases of giardiasis
occurring several weeks later. An unusual outbreak affected four people in Scotland in June
1990 when a roof-top storage tank was deliberately contaminated with human feces (Ramsay and
Marsh, 1990). Giardia cysts were found in water from the tank.

One of the largest waterborne giardiasis outbreaks in Europe affected more than 3000
individuals at a ski resort in Sweden during Christmas holidays in 1986 after the overflow of
sewage into the drinking water system (Ljungstrom and Castor, 1992). Serum samples were
collected from 352 exposed and 428 healthy, unexposed persons. Of the unexposed persons 10%
had either IgG or IgA, or both antibodies. Of those exposed, 41% had anti-Giardia antibodies.
IgG was present in 60% and IgA in 28% of those who had Giardia in their stool. In those who
were stool negative, 20% and 4% had IgG and IgA. More sera were positive (66% versus 49%)
when collection was delayed for three weeks after onset of illness. These results emphasize the
importance of identifying infected asymptomatic persons who are not Giardia-positive stool
microscopy. In outbreak investigations, the selection of cases and controls has been based on
symptoms or stool positivity, and this may have caused misclassification bias.

2. Outbreaks Associated with Recreational Water
111-48
-------
Because of its low infectious dose, Giardia can be transmitted via the accidental ingestion
of relatively small volumes of contaminated water while swimming. Fourteen swimming-
associated outbreaks and 637 cases of giardiasis were reported in the United States during 1982-
1996 and are summarized in Table III-7 (unpublished data compiled by Craun, 1998).

Table III-7 Swimming-Associated Giardiasis Outbreaks Reported 1982-96, U.S.A.
State
Washington
Illinois
New Jersey
Maryland
Maryland
Georgia
Georgia
Maryland
Washington
Maryland
New Jersey
Washington
Indiana
Florida
Year
1982
1985
1985
1987
1988
1991
1991
1991
1991
1993
1993
1993
1994
1996
Cases
78
15
9
266
34
9
7
14
4
12
43
6
80
60
Location
swimming pool
swimming pool
indoor pool
swimming pool
swimming pool
swimming pool
swimming pool
pool
swimming lake
swimming lake
swimming pond
river
swimming pool
wading pool
Additional Information

fecal contamination

inadequate chlorination

day-care center
day-care center
park; fecal contamination
wild animals near lake

met water quality limits

filter malfunction

Relatively few recreational water outbreaks of giardiasis have been described in the
literature. In an infant and toddler swim class in Washington State, 71 participants were found to
have Giardia- positive stools (Harteret al., 1984). Investigation found high turbidity and low
chlorine levels were an occasional problem at the pool, and fecal accidents were often reported
during the class. Nine cases of giardiasis were identified in people who had been swimming at a
pool in New Jersey during one day in September 1985 when a handicapped child had a fecal
111-49
-------
accident (Porter et al., 1988). The handicapped child and eight others in a group of twenty were
found stool positive. At ahotel in Manitoba, Canada, 59 of 107 guests were reported ill with
giardiasis (Greensmilh et al., 1988). An association was found between illness and using the
water slide on one of three days and swallowing pool water.

3. Outbreaks Associated with Non-potable Water

Outbreaks and illness have also been associated with consumption of untreated water
while camping, backpacking, and hiking, from individual water supplies, and from other sources
Craun, 1996). For example, an outbreak of 12 cases of giardiasis was reported in 1982 among a
group of New York City police and fire fighter divers in the Hudson River. Nine outbreaks and
116 cases of giardiasis were reported in the United States due to use of contaminated water from
non-potable and individual water systems. These outbreaks are the least likely to be detected and
investigated and thus, do not reflect the incidence of outbreaks among persons using untreated
surface water. However, they illustrate the high risk associated with consumption of untreated
water. Two outbreaks were associated with use of contaminated, untreated well water at homes
without public water supplies, and one outbreak occurred on a farm after contamination of cistern
water from a pit privy. An outbreak of 18 cases occurred in Utah in 1978 when persons on a
picnic mistakenly drank from a tap that provided irrigation water. Five outbreaks were
associated with the use of untreated surface water during hiking, backpacking, or camping.

4. Endemic Waterborne Giardiasis

a. Drinking Water

Because of the low infectious dose and ubiquitous nature of the sources of cysts in many
drinking water supplies, it is possible that sporadic cases of waterborne infection might occur in
marginally treated water systems, and that these cases of giardiasis would not be recognized as an
111-50
-------
outbreak. Epidemiologic studies, however, must be specifically designed to assess endemic
waterborne disease risks. Craun (1997) evaluated epidemiological studies where endemic
giardiasis was found to be associated with the consumption of untreated or inadequately treated
drinking water. Studies in Colorado (Wright et al., 1977), Minnesota (Weiss et al., 1977),
Washington (Harter etal., 1982; Frost et al, 1983), New Hampshire (Chute et al., 1985, 1987;
Dennis et al. 1993), Utah (Laxter, 1985), and Vermont (Birkhead and Vogt, 1989) have
suggested that consumption of untreated drinking water may be an important cause of endemic
infection and illness in the United States. A 1973 survey of 256 Colorado residents having
Giardia-positive stools, when compared to 256 controls matched by age, gender, race, and place
of residence, showed a higher proportion of cases among those who visited Colorado mountains
(69% vs. 47%), camped overnight (38% vs. 18%), and drank untreated mountain water (50% vs.
17%). A 1975 survey of 78 Minnesota residents with Giardiet-positive stools and no history of
recent foreign travel found that 63% had consumed untreated water during the period of study.
Unfortunately, an appropriate control group was not included for comparison. In a case-control
study of 349 Washington State residents having Giardia-positive stools during July 1978 to
March 1980 and 349 controls matched by age and gender, Frost et al. (1996) found consumption
of untreated water, nursery school exposure for children, and foreign travel to developing
countries to be associated with higher risk of acquiring Giardia infection. A survey of intestinal
parasites conducted in two Washington counties found a 7% Giardia prevalence among 518
children, one to three years of age (Harter et al., 1983). No difference in the prevalence of
infection was found for source (surface or well) of drinking water, but a higher risk was
associated with use of unfiltered surface water. Ten of 175 (7%) children residing in homes
using unfiltered surface water were found to be infected with Giardia compared with only one of
37 (3%) children residing in a home using filtered surface water. An increased prevalence of
infection was also found in children who had a history of drinking untreated surface water from
streams or lakes during recreational activities.
111-51
-------
Results of a case-control study of 171 giardiasis patients in New Hampshire during
January, 1977 to June, 1984 and 684 controls matched by age and gender found an increased risk
of acquiring giardiasis was associated with travel outside the United States, family member
diagnosed giardiasis, family member in a day-care program, camping and use of a shallow well
or surface water for individual, household water supply (Chute et al., 1985, 1987). Persons who
used shallow well or surface water sources for their household water supply had twice the risk of
giardiasis compared to persons who used any other water source, either drilled well or municipal.
A higher risk of giardiasis was associated with the household use of shallow well or surface
water sources compared to use of municipal water sources. Significant risk factors found in a
study of 273 cases and 375 matched controls during 1984 and 1985 in New Hampshire (Dennis
et al. 1993) were: drinking water from shallow wells, contact with a person in day care, and
swimming in a lake, pond, stream or river. A survey of 383 Utah National Guard members
showed that 15% had symptoms suggestive of giardiasis and that the guardsmen were at risk of
contracting giardiasis by drinking contaminated water during field exercises in the state (Laxter,
1985); 62% of the men who had symptoms drank untreated water from lakes, streams, and a
cattle watering trough.

Birkhead and Vogt (1989) studied risk factors among 1211 cases of laboratory-confirmed
giardiasis that were not associated with outbreaks in Vermont and found increased relative risks
(JAR) of giardiasis for persons using municipal surface water systems without filtration (RR=1.9)
or well water (RR=1.8) and persons using private water systems (RR=2.2). The average annual
incidence rate of giardiasis was found to lowest in populations using municipal surface water
systems with filtration (15.1/100,000) compared populations using municipal surface water
systems without filtration (28.6/100,000), populations using municipal well water
(26.8/100,000), persons using private water systems (32.8/100,000).

Fraser and Cooke (1991) conducted an epidemiological study in Dunedin, New Zealand,
to investigate the risk of endemic giardiasis in a part of the city where water was mechanically
111-52
-------
microstrained through a 23 jim screen, chlorinated and fluoridated. A three-fold increased
relative risk for giardiasis was found for persons using this water compared to persons in another
part of the city where water was treated by coagulation/flocculation and direct dual media
filtration (anthracite and sand).

A case-control study in Vancouver, a city with a population of 1.4 million served by an
unfiltered but chlorinated supply, found no increased risks of giardiasis associated with drinking
water (Mathias et al., 1992). Risk factors for giardiasis were travel abroad and travel elsewhere
within British Columbia. Another case-control study elsewhere in British Columbia, however,
found that drinking water was the most important risk factor for laboratory-confirmed giardiasis
(Isaac-Renton and Philon, 1992). Persons who drank unchlorinated and unfiltered surface water
were at a much higher risk compared to those who drank well-water. There was little difference
in giardiasis risk, however, for people who drank either chlorinated, unfiltered surface water and
unchlorinated, unfiltered surface water. For persons who traveled to rural areas of British
Columbia, drinking local tap-water was identified as a risk factor for giardiasis (Isaac-Renton and
Philon, 1992).

In a year-long longitudinal study, Isaac-Renton et al. (1996) assessed Giardia cyst levels
and parasite viability in the drinking water of two British Columbia communities From 69% of
the sample locations, 64% of source water samples were found to be cyst-positive. In one
community, 77% of water samples of treated water were cyst-positive; in the other, 98% of
samples were cyst-positive. In the overall survey and surveys for each community, decreased
Giardia cyst levels and decreased viability, based on infectivity testing in the Mongolian gerbil
model, were found in chlorinated water samples compared with their respective source water
samples. Assays in which the gerbils were inoculated orally by gavage found 0% infectivity for
the cysts. Low-level endemic waterborne transmission, however, was suggested by results of a
health survey. Compared to a nearby community that obtained water from a protected watershed,
111-53
-------
both communities had an increased seroprevalence and prevalence of laboratory-confirmed cases.

The consistency of the findings in these studies strongly suggests that the risk of
giardiasis is high among persons who consume untreated water and that in the United States,
Canada, and New Zealand, endemic risks ofwaterborne giardiasis are high among populations
that consume unfiltered surface water. However, in some developing countries the endemic
waterborne giardiasis risk may not be as significant. Esrey et al. (1989) investigated the
relationship between the presence of Giardia infection in pre-school children and drinking water
in Lesotho, South Africa. Results of this study found that in this population personal hygiene
and person to person transmission were more important than waterborne transmission.

b. Water Recreation and Other Water Sources

Epidemiological studies of endemic waterborne giardiasis have also identified
swimming-associated risks. Engaging in recreational water activities was found to be a risk
factor for giardiasis among travelers to rural areas of British Columbia (Isaac-Renton and Philon,
1992). In a case-control epidemiological study in New Hampshire (Dennis et al., 1993) found
that swimming in a lake, pond, stream or river was among the several important risk factors for
giardiasis. A survey (Harter et al., 1982) in Washington also found an increased prevalence of
infection among children who had a history of drinking untreated surface water from streams or
lakes during recreational activities. In a study of endemic giardiasis in the Canterbury area of
New Zealand, contact with sewage and travel abroad were the most significant risk factors for
giardiasis (Hunter, 1998). A case-control study of 74 cases and 108 matched controls from July
1992 to May 1993 in the counties of Somerset and Avon in England found that drinking
potentially contaminated water and swimming were significant risk factors (Gray et al., 1994).

5. Foodborne Outbreaks
111-54
-------
Outbreaks can occur when food becomes contaminated, but few well-documented
foodborne outbreaks of giardiasis have been reported. Rabbani and Islam (1994) subsequently
reviewed foodborne giardiasis and concluded that it is rare in developed countries but that food
can be an important vehicle of transmission in areas where food hygiene is poor and a significant
proportion of the population is infected. They found that foodborne outbreaks of giardiasis have
been suspected and suggested as early as 1922.

Epidemiological investigation of a giardiasis outbreak among employees of a school in
Minnesota in 1979 found two food items statistically associated with illness, home-canned
salmon and cream cheese dip (ICAIR, 1984). The wife of one employee had diapered her 12-
month-old grandson just prior to preparing the salmon and touched the salmon by hand before it
was given to the employees. Although she was free ofGiardia symptoms, the grandson's stools
were positive for Giardia. Peterson et al. (1988) described a foodborne outbreak among 16
attendees at a picnic in rural Connecticut. Food was prepared and brought to the picnic by seven
family groups and a neighbor who did not attend. This neighbor prepared a cold noodle salad
which was implicated as the vehicle of infection. The salad preparer, who was symptomatic one
day after making the salad, could have been excreting Giardia cysts the day she made the salad
and could have contaminated it while mixing with her hands.

In 1986, a giardiasis outbreak occurred in a nursing home in Minnesota; 73 residents or
employees of the home and 15 children participating in day care at the home became ill (White et
al., 1989). The mean age of resident cases was 80 years and their mean duration of diarrhea was
16 days. Epidemiological investigation implicated both foodborne and person to person
transmission. An association was found between sandwich consumption and illness in nursing
home staff, and there was a significant lack of illness among residents who consumed only a
pureed diet that was cooked before serving. An outbreak of 10 cases of giardiasis occurred
among 25 persons attending a family party in New Jersey in 1986 (Porter et al., 1990).
Epidemiological evidence implicated fruit salad that had been prepared in the home of women
111-55
-------
who had a diapered child and a pet rabbit, both positive for G. lamblia. Nine cases had onset of
symptoms 6 to 12 days following the party.

The first reported common-source outbreak of giardiasis traced to a commercial food
establishment occurred at a restaurant in central Washington State in 1990 (Quick et al., 1992).
Twenty-seven of 36 (75%) persons who had attended a meeting and eaten at the restuarant
became ill. Onset of illness occurred 2-19 days after the meeting and illness lasted from 7 to 28
days (median =18 days). One patient was hospitalized. No single food or beverage was
statistically associated with illness, but 26 of the 27 ill persons drank ice water. Although
contaminated water was felt to be an unlikely source of infection, the ice could have become
contaminated during handling at the restaurant. The restaurant was served by a community water
system that obtained water from 18 untreated wells; routine water samples were negative for
coliform bacteria. Ice for beverages had been served by an employee who had an asymptomatic
Giardia infection and an employee who had a Giardia-infected child still in diapers. Either food
handler could have transferred Giardia cysts from their hands to the beverage ice either directly
or via the ice scoop. The plausibility of ice as the vehicle of infection is supported by an earlier
outbreak in Canada in which ice was suspected and by evidence that surface contamination may
be the only mechanism by which ice can serve as a vehicle for Giardia transmission (Quick et
al., 1992). A similar mode of transmission was suggested in two previous foodborne outbreaks
of giardiasis (ICAIR, 1984; Porter etal, 1990). Mintz et al. (1993) describe an outbreak of
giardiasis associated with an insurance company cafeteria. A case-control study of 26 sick
employees and 162 well employees suggested the probable vehicle of infection was raw, sliced
vegetables served in the employee cafeteria. The sliced vegetables had been prepared by a food
handler who was infected with G. lamblia. The median duration of diarrhea in this outbreak was
35 days.

Foodborne outbreaks of giardiasis have been reported in the United States; small
outbreaks have occurred because ice used for beverages and foods were accidently contaminated
111-56
-------
by food service workers. In five reported foodborne outbreaks, cold foods such as salmon, raw
vegetables, noodle salad, fruit salad, and sandwiches were implicated as vehicles of infection.
Ensuring good hygienic practices among food service workers, including hand washing, washing
vegetables, and using gloves and clean utensils, are important in preventing foodborne outbreaks
of giardiasis (Quick et al., 1992; Mintz et al., 1993).

Although giardiasis outbreaks may often be unrecognized because of the long incubation
period and large number of asymptomatic infections, restaurant-associated transmission of G.
lamblia does not appear to be a significant public health problem (Quick et al., 1992). Reports of
parasitological screening studies of food service workers in Panama and Algeria have revealed
high rates of asymptomatic infection but no evidence of outbreaks (Quick et al., 1992). Eating
raw or undercooked food because of taste considerations or to conserve heat-sensitive nutrients
might increase the risk of spreading Giardia through food (Rabbani and Islam, 1994). Sheep
tripe soup was considered to be the vehicle of an outbreak of giardiasis affecting two Turkish
families (Karabiber and Aktas, 1991). It was suggested that Giardia cysts in deep layers of the
tripe were protected from heat inactivation during preparation of the soup.

6. Travelers

Travelers' diarrhea can be caused by exposure to a number of bacterial, viral, and
proptozoan pahogens, including Giardia (ICAIR, 1984). Although giardiasis probably accounts
for less than 5% of traverlers' diarrhea, high attack rates have been reported in Europeans and
North Americians traveling to certain areas of the world (Farthing, 1996). Reports of giardiasis
in travelers first appeared in 1970 as outbreaks occurred among travelers to the Soviet Union,
especially St. Petersburg (ICAIR, 1984). In the 1970's, outbreaks of giardiasis were reported
among a group of travelers to Portugal and a group of children and adults on a Mediterranean
cruise ship (ICAIR, 1984). Expatriates in endemic areas may also be at high risk. The incidence
of giardiasis in expatriates in Bangladesh was found to more common than among new-comers
111-57
-------
and children less than 10 years old (Rabbani and Islam, 1994). Epidemiological studies have
also reported an increased risk of giardiasis among visitors to the Colorado mountains and hikers,
backpackers, and campers in other areas who drink untreated or inadequately treated water from
lakes and streams (1CAIR, 1984; Farthing, 1994).

A case-control study of 74 cases and 108 matched controls from July 1992 to May 1993
in Somerset and Avon, England, found that travel to developing countries was among several
important risk factor for giardiasis (Gray et al., 1994). In a case-control study of 349 Washington
State residents having Giardia-poshivQ stools during July 1978 to March 1980 and 349 controls
matched by age and gender, Frost et al. (1983) found that foreign travel to developing countries
was associated with higher risk of acquiring infection for adults. A case-control study of 171
giardiasis patients in New Hampshire during January, 1977 to June, 1984 and 684 controls
matched by age and gender also showed an increased risk of acquiring giardiasis associated with
travel outside the United States (Chute et al., 1985, 1987). A case-control study of giardiasis in
Vancouver (Mathias etal., 1992) found that travel abroad and travel within British Columbia
were important risk factors. Isaac-Renton and Philon (1992) found higher risks of giardiasis
among persons who traveled to rural areas of British Columbia and drank tap water or engaged in
recreational water activities. A study of endemic giardiasis in Canterbury, New Zealand, travel
abroad was a significant risk factor (Hunter, 1998).

7. Day-Care Centers

Outbreaks of giardiasis and a high prevalence of infection have been reported in settings
where infants and young children in diapers are collectively cared for. Not only is the spread of
Giardia likely within the care center but secondary transmission to family members is also likely.
The occurrence and transmission of giardiasis in the child day care setting was recently reviewed
by Thompson (1994). Thompson (1994) found prevalence rates of Giardia infection in Australia
to range between 2% and 46% and to be highest among children 1-5 years of age who attended
111-58
-------
preschool care. In three day-care centers in Atlanta, Georgia, the infection rate ranged from 29%
to 54% compared to 2% among children not attending day-care centers; in two day-care facilities
in metropolitan Toronto, Canada, infection rates of 17% and 39% were reported (ICAIR, 1984).
The most susceptible were children 1 to 3 years old. Infections were reported to have been
spread to as many as 23% of the children's household contacts (ICAIR, 1984). Secondary spread
was found to be important within families having Giardia-positive children between 1- and 3-
years-old; 10% of 47 family members studied were also Giardia-positive.

A two-year prospective study of diarrheal illness in children up to 36 months of age in 22
day care centers in Maricopa County, Arizona, identified 465 sporadic cases and 170
outbreak-associated cases of diarrhea (Bartlett et al., 1991). Giardia was significantly more
common in toddlers than in infants and was found in 19% of asymptomatic child contacts of
symptomatic infected children. In the second year, the study included children of the same age in
30 day care homes and 102 households not using day care. The seasonal pattern of diarrhea,
frequency of pathogen isolation, and relative frequency of individual pathogens were similar in
each setting. G. lamblia and rotavirus were the most commonly isolated enteropathogens.

In 1989 and 1990, a survey of stool specimens from 292 diapered children attending 17
randomly selected day-care centers in Fulton County, Georgia, found that 21 (7%) children in 7
centers were infected with Giardia (Addiss et al., 1991) Infected children ranged in age from 3
to 30 months, and only 57% of Giardia-positive children had symptoms. In 1986 the prevalence
of infection in these same centers had been higher, 11% (Addiss et al., 1991). Both prevalence
rates, however, are lower than the 16% to 38% infection reported in other studies for children
attending non-outbreak centers (Addiss et al., 1991). The percentage of day-care centers with
one or more infected children (41%) was also less than the 66% to 85% that had been reported in
other studies (Addiss et al., 1991). In addition to person to person spread, a possible role was
suggested for fomites in the transmission of Giardia in day-care centers. Laboratory studies
found that Giardia cysts survive for less than 24 hours on dry environmental surfaces, and unless
111-59
-------
surfaces are continuously being contaminated, a child's risk of exposure from to Giardia from
fomites is limited to the day of contamination (Addiss et al., 1991). Another survey of 80 of231
children 2 to 3 years of age in six commercial day care centers found that 13 (16%) were infected
with Giardia and that five family members from four different families were positive (Cody et
al., 1994). Only seven of the infected children (54%) had diarrhea. No care-givers were found to
be Giardia-poshive. Oretega and Adam (1997) reported that no seasonal pattern has been
observed for Giardia infection in day-care situations; however, Rodriguez-Hernandez et al.
(1996) observed a higher frequency of giardiasis in the autumn season in a study of eight day care
centers in Salamanca, Spain. G. intestinalis was identified in 25% of the children studied.

An epidemiological study of endemic cases of giardiasis not associated with outbreaks
reported from 1983 to 1986 in Vermont found that person to person transmission in child-care
facilities was important in the transmission of non-outbreak cases of giardiasis.(Birkhead and
Vogt, 1989). Children aged one to four years had the highest incidence rate for symptomatic
infection of any age group, and child-care attendees had an incidence rate 50% greater than non-
attendees (300.0/100,000 versus 194.7/100,000). Hatter et al. (1982) found no differences in
prevalence of infection between children who normally attended day-care centers and those who
did not and found no correlation between the socioeconomic status of the families and the
presence of Giardia. An important risk factor identified by Harter et al. (1982) was having two
or more siblings between the ages of 3- and 10-years-dd.

Epidemiological investigation of a foodborne giardiasis outbreak in a Minnesota nursing
home also suggested person to person transmission, as illness in residents was associated with
physical contact with children at the day care facility through an adopted grandparent program
(White et al., 1989). Steketee et al. (1989) described three outbreaks of giardiasis that occurred
over a 19-month period in a Wisconsin facility that cared for a daily average of 115 children aged
1 month to 6 years. Estimated attack rates in the three outbreaks were: 47%, 17%, 37% for
children; 35%, 13%, 9% for staff; and 18%, 9%, 5% for household contacts. Infections recurred
111-60
-------
in outbreak proportions even though a variety of control measures were instituted, including
pharmacological treatment with a cure rate of >90%, better case identification, follow-up testing
of stools, and improved personal and environmental hygiene practices. Bartlett et al.(1991)
found Giardia infection in almost 11% of 6761 new infants and toddlers tested for admission to
31 day care centers. A prospective randomized trial comparing three strategies for control of
Giardia in infant-toddler day care centers found that more strict intervention (exclusion and
treatment of both symptomatic and asymptomatic infected children) did not result in better
control of infections; an initial Giardia prevalence of 18-22% in the three intervention groups
was reduced to 7-8% in each group at 6 months intervention (Bartlett et al., 1991).

Steketee et al. (1989) found that attack rates were highest among the ambulatory children
in diapers, children who attended the center for 40 or more hours per week, and children who had
been infected in the respective previous outbreak. This latter finding suggests that giardiasis
infection may not provide immunity for subsequent re-infections in this age group. Acquired
immunity to giardiasis in adult populations has been suggested by epidemiologic studies, and
there is experimental evidence of acquired immunity in mice; however, it is possible that
immunity may not be established in young children or the immune response may be reduced by
early drugtherapy (Steketee et al., 1989).

8. Sensitive Populations

Immunodeficiency with varying degrees of hypogammaglobulinemia or
agammaglobulinema predisposes to the acquisition of giardiasis and is the most commonly
reported form of immunodeficiency associated with chronic giardiasis (Farthing, 1996).
Giardiasis is more prevalent in homosexual men both with and without human
immunodeficiency virus (HIV) infection (Farthing, 1996). In a selected New York City
population examined forparasitological diseases by the same laboratory using the same
procedures, 18.3% of 126 homosexual males were found to be cyst-positive for Giardia
111-61
-------
compared to a 2.1% positivity among the other 5,885 patients (ICAIR, 1984). Giardiasis can be
transmitted by some sexual activities, particularly among male homosexuals who practice oral-
anal sex (ICAIR, 1984; Turner, 1985; Farthing, 1996). Chronic giardiasis is not a major clinical
problem in persons with HIV infection or acquired immunodeficiency syndrome (ADDS) patients
(Farthing, 1996). It is not clear why the intracellular protozoa (Cryptosporidium parvum,
microsporidia, and Cyclospord) produce severe chronic diarrhea in AIDS patients but the effect
of giardiasis is relatively mild (Farthing, 1996).

H. Environmental Factors Affecting the Survival of Giardia Cysts

1. Effects of Water Temperature on Giardia Cyst Survival

Temperature is a significant factor in the survival of Giardia cysts. Information reviewed
in the Giardia criteria document (ICAIR, 1984) indicated that cysts suspended in tap water could
survive more than two months when held at temperatures of 8°C, about 26 days at 21°C, and
about 6 days at 37°C. Fewer than 1% of the cysts survived freezing at -13°C for 14 days and
raising the temperature of cyst suspensions to boiling immediately inactivated 1he cysts. These
data had been developed using G. lamblia cysts and excystation as the indicator of viability.

Using excystation with G. muris cysts, Schaefer et al. (1984) determined that the cysts
were inactivated after freezing in distilled water at -20°C and then thawed. The length of time
the cysts remained frozen was not specified. These investigators also found that the thermal
death point (the lowest temperature at which the organisms are inactivated in 10 minutes) for G.
muris cysts was 54°C. Tests (deRegnier et al., 1989) have also been conducted tests on the
viability of G. muris cysts in fecal pellets when stored in different types of waters (distilled, lake,
river and tap). The cyst viability measurements were based on exclusion of a fluorogenic dye
propidium iodide (PI), mouse infectivity, and cyst morphology by Nomarski microscopy. In
tests with cysts stored in distilled water in the refrigerator (5° to 7°C), viability by PI was 83% to
111-62
-------
90% at 7 days, 13% to 25% at 28 days, and less than 1% at 56 days. Viability by mouse
infectivity testing was 100% at 7 days, 17% to 100% at 28 days, and 0% at 56 days. In storage
tests in both natural lake and river waters, these investigators reported that only decreased water
temperature was correlated with survival. They observed longer viability at lower temperatures
with cysts stored at <10°C remaining viable for 2 to 3 months. Other water quality parameters
that were tested and for which no correlation was found were: pH, dissolved oxygen, turbidity,
color, hardness, ammonia, nitrate and phosphorous.

Cysts were also exposed to tap water (deRegnier et al., 1989). Cysts in fecal pellets were
placed in glass vials containing Minneapolis city tap water, and the vials were suspended in
flowing tap water (20° to 28°C). The viability at 7 days was <2%, based on PI dye exclusion,
and 0-17% viable, based on mouse infectivity testing. At 14 days, no viable cysts were detected
by either PI or mouse infectivity. The authors were surprised by the loss of viability when the
cysts were exposed to tap water for as little as 3 days as compared to survival of cysts in
unprocessed river water. They indicated that the factors responsible for the cysticidal effect had
not been determined but that the effect was most likely due to residual chloramine.

Although the viability was not determined, Erlandsen et al. (1990d) studied the effects of
freeze-thaw cycles on the recovery of G. muris and G. lamblia cysts. G muris cysts at levels
ranging from 104 to lOVmL were suspended in an unspecified medium and frozen at -16°C and
thawed at room temperature (approximately 20°C) through either one or three cycles. After one
freeze-thaw cycle there was no detectable loss of cysts in preparations with high levels of cysts.
However, there was about a 40% to 60% loss of cysts from preparations with low levels of cysts.
After three freeze-thaw cycles, the loss in the high level preparations was about 22% to 27% and
in the low level preparations, about 70% to 80%. These investigators also indicated that while
the cysts were detectable with IF A staining, only 10% or less were recognizable using bright field
microscopy. G. lamblia cysts were easily detected with IFA but immunostaining was variable
and this was attributed to freeze-thaw damage.
111-63
-------
2. Other Factors that Affect Giardia Cyst Survival

Since soil surface disposal of mixed human and animal wastes is a possibility, Deng and
Cliver (1992) studied the survival of G. lamblia cysts in mixed human wastes (septic tank
effluent or STE) and swine wastes (swine manure slurry or SMS). They used PI exclusion as an
indicator of cyst viability. Under field conditions, the degradation rates of the cysts suspended in
either buffer or STE were similar as determined by D values (the time in days for a 90%
reduction in the number of cysts). The D values for viable cysts in buffer and STE were 16.9 and
17.6, respectively. However, mixing STE and SMS greatly increased the rate of degradation
producing a viable cyst D value of 3.7. The temperature significantly influenced the rate of
degradation in mixed wastes. The D values forviable cysts at 5, 15 and 25°C were 129.9, 26.2
and 4.1, respectively. The authors reported that they did not determine the mode of degradation
but they postulated that bacteria may have been involved. They indicated that STE and SMS are
both rich in bacteria and that electron micrographs showed bacteria adhered to the cyst walls.

Land application of municipal wastewater treatment plant sludges is a disposal option that
prompted Van Praagh et al. (1993) to study the inactivation of Giardia muris cysts in anaerobic
digester sludge. Cysts were seeded into sludge in the laboratory, anaerobic conditions were
simulated in air-tight containers, and samples were taken at various exposure intervals. Cyst
inactivation was determined based on original and final cyst levels, and the fraction of the
original cysts which exhibited excystation. There were 99.9% cyst inactivations at 15.1 days,
20.5 hours, and 10.7 minutes with exposures at 21.5°, 37°, and 50°C, respectively. Casson et al.
(1990) sampled activate sludge and trickling filter effluents at a Maryland wastewater plant
finding a geometric mean of 4 cysts/L and 11 cysts/L, respectively. The plant influent contained
a geometric mean of 137 cysts/L. Cysts were concentrated in suspended solids and in the sludge.

Johnson et al. (1997) investigated the survival of G. muris cysts and other enteric
pathogens in marine waters in Hawaii. The viability of the cysts, as determined by excystation,
111-64
-------
was reduced by 99.9% in only 3 hours when the cysts were suspended in marine waters and
exposed to sunlight. However, when the cysts were kept in the dark, 77 hours were required to
obtain a 99.9% reduction. In two marine waters with different salinities (28 and 35 mg/L), cysts
survived longer in the lower salinity water. However, since different waters were used, the
potential effect of factors other than salinity cannot be ruled out.

Rodgers et al. (1998) have isolated a bacterium from a Kentucky stream that can kill G.
lamblia cysts. The bacterium has been characterized as a Gram negative, aerobic rod and it
produces a yellow pigment not of the flexirubin type. The organism, designated Sun4, produces
a spreading colony morphology on low nutrient agar although true gliding motility has not been
observed. Using ribosomal RNA sequencing and phylogenetic analysis, the organism has been
identified as a Flavobacterium most closely related to F. columnar e. Static cultures, as opposed
to shaken cultures, are more effective in killing cysts and calcium must be present in order for the
bacterium to grow and to kill cysts. The intact bacterial cells must be present for cyst
inactivation to occur as evidenced by the ineffectiveness of cell-free extracts. The authors
suggest that Sun4 or other bacteria might be used as biological control agents for Giardia cysts in
drinking water.

I. Summary

1. Occurence

Interpretation of the occurrence data for Giardia in water and other environmental
samples is dependent upon methods used to detect and quantify the cysts. Methods used to date
generally provide little or no information on viability/infectivity or species identification of
organisms to assist in assessing the epidemiological significance of positive findings of cysts in
environmental samples. Quantitative data may not be reliable due to low efficiency and
precision of methods.
111-65
-------
Giardia cysts are distributed worldwide in surface waters of all qualities. Cysts have
been found in surface waters from the Arctic to the tropics. All municipal wastewaters and
surface waters likely always contain Giardia cysts at some level. Whether or not they are
detected is dependent upon the methods used. Cyst densities that have been reported generally
are on the order of 103~4/L in raw sewage; lO^/L in secondary treated wastewaters, and 10°/L or
less in surface waters. Generally, there is no correlation of cyst densities in water with bacterial
indicator organisms. Cysts occur in surface waters throughout all months of the year.
Occasionally, seasonal variations are reported but these may be site or region specific. When
they are reported in North America, the levels are generally higher in the late summer, fall and
early winter.

Longitudinal studies using high frequency sampling indicate spikes in cyst levels that
might be missed by monitoring programs using low frequency sampling schedules. Cyst levels
are generally higher in rivers or streams impacted by agricultural (e.g., cattle or dairy farming) or
residential (e.g., sewage outfall) activities.

Levels of Giardia are usually reported to be somewhat lower than Cryptosporidium
densities in U.S. waters. In other countries, e.g., Canada, widespread surveys have produced the
opposite results.

National, regional, state or local surveys for occurrence of cysts in water may not be
predictive of what will be found in a specific watershed. Sources of contamination and factors
affecting transport and survival of cysts need to be determined for each watershed.
Contamination levels of sources may fluctuate significantly due to poorly defined factors
including weather events, agricultural practices and treatment plant (wastewater and drinking
water) infrastructure and practices. The first-flush waters from storm events have been found to
significantly affect source water cyst occurrence.
111-66
-------
No published reports on the occurrence of Giardia in soil or air were found. One study
reported the occurrence of cysts on stainless steel and Formica® surfaces in day care centers. A
number of foodborne outbreaks have been reported but data are sparse to non-existent on
quantitative levels of cysts in or on foods.

The viability and longevity of Giardia cysts in the environment is significantly affected
by temperature-as the temperature increases, the survivability decreases. A small fraction of
cysts can withstand a single freeze-thaw cycle. Cysts subjected to repeated freeze-thaws as might
occur in the environment are likely inactivated but still will be detected with present methods.

Cyst inactivation in municipal wastewater treatment plant sludges is temperature-
dependent. There is a factor or factors in swine manure slurry that results in more rapid
degradation of cysts under field conditions. A bacterium has been isolated from a fresh water
stream that is capable of killing Giardia cysts.

2. Prevalence, Outbreaks, and Endemic Risks

Giardiasis affects all age groups. High risk groups for giardiasis include infants and
young children, travelers to developing countries, the immunocompromised, homosexuals who
practice oral-anal intercourse, and persons who consume untreated water from lakes, streams,
and shallow wells. Populations in communities with unfiltered surface water or groundwater that
has been contaminated by surface water or sewage are also at high risk of infection. Waterborne
outbreaks are more common in the United States and Canada than Europe, and this may be due
to the larger number of unfiltered surface water systems in North America. Waterborne
outbreaks of giardiasis can occur when disinfection is interrupted, disinfection levels are
inadequate, disinfection contact time is low, or turbidity levels are increased, especially in areas
where water temperatures are low. In low water temperatures, water disinfection may be less
111-67
-------
effective and Giardia cysts survive for longer periods of time. In many waterborne outbreaks of
giardiasis in the United States, neither the turbidity limit nor the coliform limit was exceeded.

Outbreaks of giardiasis in ground water systems emphasize the need to protect these
sources from sewage and surface water contamination. Vulnerable ground water sources that
cannot be protected should be considered to be at the same high risk of contamination as surface
water sources.

Outbreaks of giardiasis that have occurred in filtered water supplies emphasize the need
for proper chemical pretreatment and the importance of good design, installation, maintenance,
and operation of treatment facilities. Ten percent of the waterborne outbreaks of giardiasis in the
United States occurred as a result of contamination in the distribution system, and adequate
precautions should be taken to protect treated water quality during storage and delivery.

Endemic risks of waterborne giardiasis are higher among persons who consume untreated
water, and in the United States, Canada, and New Zealand, higher endemic risks have been
identified among populations that consume unfiltered surface water.

Several small foodborne outbreaks of giardiasis have been associated with the
contamination of ice used for beverages and foods by infected food service workers. Outbreaks
have occurred in day-care populations, and the prevalence of Giardia infection is relatively high
in these populations. However, risk factors for the introduction, spread, and persistence of
Giardia in child day-care centers are not well understood.

IV. HEALTH EFFECTS IN ANIMALS

Giardia has been reported to infect virtually all vertebrate animals, including higher
mammals (humans and other primates), domestic mammals (cats and dogs; cattle and sheep),
IV-l
-------
wildlife (beavers and muskrats), and other animals (birds, reptiles and amphibians). Calves may
be jointly infected in a natural farm setting with both Giardia and Cryptosporidium (Xiao et al.,
1993).

H. Symptomatology

Several reports, supported by light micrographs, have appeared in the literature,
suggesting that Giardia trophozoites penetrate the mucosal cells of the intestine, as well as
various extra intestinal tissues in rodents (ICAIR, 1984). The signs of symptomatic giardiasis in
animals include diarrhea, steatorrhea (i.e., excessive discharge of fat in the feces), anorexia,
weight loss, and vomiting (ICAIR, 1984), and in general, are similar to symptoms observed in
humans. See Chapter V for a discussion of human health effects.

The signs and clinical picture of symptomatic giardiasis in primates observed at the
Kansas City Zoo were similar to those seen in their human attendants, who also contracted the
disease (ICAIR, 1984). Although no fatalities were reported, all patients suffered from loose
stools, diarrhea, and vomiting. Since all infected apes and monkeys received chemotherapy, it is
not known whether the giardiasis in these animals was a self-limiting disease.

In a study of calves naturally infected with Giardia, all infected animals were noted to
have intermittent diarrhea, and mucus was seen in many fecal samples (Ruest et al., 1997). In
mice inoculated with Giardia cysts, impaired weight gain and diarrhea were observed (ICAIR,
1984). The infection was observed to clear spontaneously, with most animals appearing to be
free of giardiasis symptoms within 21 to 28 days. In parakeets, giardiasis has been associated
with diarrhea, decreased intake of food and water, debility, and high mortality ranging from 20 to
50% (ICAIR, 1984). Diarrhea, anorexia, and occasionally, cessation of fecal elimination were
observed in chinchillas infected with Giardia (ICAIR, 1984). Not all of the chinchillas infected
became symptomatic, but of the four that had symptoms, three died.
IV-2
-------
Household pets have also been observed to show signs oi Giardia infection including
weight loss, mucoid and soft stools, and the presence of split and unsplit fats in the feces (ICAIR,
1984). Dogs less than one-year-old appear to be more likely to have symptomatic giardiasis than
older dogs (ICAIR, 1984). Signs in these canines include diarrhea with mucus and fats,
listlessness, and anorexia. In puppies, severe giardiasis may result in complications ranging from
growth retardation to death (ICAIR, 1984).

B. Therapy

At present, no drugs are approved for treating giardiasis in animals. The benzimidazoles,
albendazole and fenbendazole, have been shown to clear Giardia cysts from the feces of infected
dogs (Barr et al., 1993; Zajac et al., 1998). Because albendazole is suspected ofbeing
teratogenic, it should not be given to pregnant animals.

Fenbendazole can be used safely to treat giardiasis in dogs, including pregnant and
lactating animals (Barr et al., 1994). Other drugs, including quinacrine hydrochloride and
metronidazole, have been used with varying degrees of success to treat giardiasis in dogs
(Zimmer and Burrington, 1986). Olson et al. (1997c) found that immunization of puppies
provided protection against giardiasis. Twenty puppies were vaccinated with a
trophozoite-derived Giardia vaccine and boosted on day 21; 10 control puppies received only
saline. Both groups were experimentally infected on day 35 with 1 x 106 Giardia duodenalis
trophozoites by intraduodenal injection.
Giardiasis in cats can be treated with albendazole, but it should not be administered to
pregnant animals. Metronidazole or furazolidone may also be used for Giardia infections in cats
(Kirkpatrick, 1986; Patton 1998).
IV-3
-------
McAllister et al. (1996) found that the number of Giardia cysts shed in feces of growing
lambs was not affected by salinomycin in their diet but did decline with time. Although a
beneficial effect of 10 mg/L salinomycin on lamb performance was seen, the development of
natural resistance made it difficult to attribute this response to the control of giardiasis.

Giardia cyst excretion in naturally infected calves was shown to be reduced or eliminated
after treatment with albendazole or fenbendazole (Xiao et al., 1996; O'Handley et al., 1997).
Calves may be also treated with quinacrine hydrochloride, ipronidazole or dimetridazole.
Giardia infection in horses can be cured with metronidazole. Finally, giardiasis in large animals
and birds may also be treated orally with fenbendazole (Patton, 1998).

C. Epi demi ol ogi cal D ata

Epidemiological data (Erlandsen et al., 1988a, b) show that Giardia infection in animals:
(a) is spread via the fecal-oral route;
(b) occurs worldwide, in most animal species;
(c) is more often than not asymptomatic;
(d) is primarily a disease of the young (suggesting a role for immunity in these
infections);
(e) is much more likely to spread within a host species than from one host species to
another.

Some information is available about the prevalence of Giardia in lower animals. In
general, the prevalence data are based on the examination of animals selected with no regard for
their symptoms. The prevalence of Giardia infection in beaver was found to be 7% to 16% in
different parts of the United States, and in muskrats the prevalence was greater than 95%
(Erlandsen et al., 1990c). Giardia infection was found in 153 (77%) of 200 dogs and 9 (3%) of
300 cats tested in Minnesota (Bemrick 1961). Similar prevalences were reported in Spain
IV-4
-------
(Lopez-Brea, 1982) and Japan (Miyamoto and Kutsume, 1978; Asama et al., 1991; Arashima et
al., 1990). Kirkpatrick (1986) reported the prevalence of Giardia infection in cats to range from
1 to 11 percent in the United States. In Washington, Pacha et al. (1987) found that 469 (65%) of
722 of the fecal samples collected from a variety of animals including voles, mice and shrews
were positive for Giardia.

The results of recent studies underscore the fact that Giardia is a common protozoan
parasite of farm animals (including cattle, sheep, pigs, and horses) and occurs with greater
prevalence in young, than in adult, animals (Buret et al., 1990; Olson et al., 1997a). Buret et al.
(1990) found the prevalence of Giardia infection was 18% in sheep and 10% in cattle and was
significantly higher in lambs and calves (36% and 28%, respectively). Olson et al. (1997a) also
found Giardia to be a prevalent infection in farm livestock; 104 cattle, 89 sheep, 236 pigs and 35
horses were sampled from 15 different locations in Canada. Giardia were present in cattle and
sheep in all six sites sampled with a prevalence of 29% and 38%, respectively; the prevalence
was greater in calves and lambs. All horse sampling locations were positive for Giardia with
20% of animals infected. Giardia was identified in four of six hog operations with a prevalence
of 9%.

Bettiol et al. (1997) found Giardia in 21% of 295 Tasmanian native marsupials screened
over a three-year period. After isolating immunologically-confirmed human-infective Giardia
from two Australian marsupials, the northern brown bandicoot (Isoodon macrouris) and the red-
necked pademelon (Thylogale thetis), Buckley et al. (1997) suggested that the potential exists for
the waterborne transmission of human-infective Giardia in pristine watersheds of Australia even
though humans and domestic livestock are excluded. Buckley also noted the isolation of Giardia
from Australian bushtail possums.

Olson et al. (1997b) identified Giardia in the intestinal contents of three of 15 ringed
seals (Phoca hispidd) slaughtered by Inuit hunters in the western arctic region of Canada.
IV-5
-------
Giardia has also been identified in llamas (Rings and Rings, 1996) and a captive population of
marmosets (Kalishman et al., 1996). Finally, wading birds (blue herons, egrets, green herons and
black crowned night herons) have been reported to have Giardia prevalence rates greater than
90% (Erlandsen, 1994; Erlandsen et al., 1990b). McRoberts et al. (1996) describe the
morphological and molecular characteristics of Giardia isolated from a straw-necked ibis in
Australia.

D. Summary

In general, the symptoms seen in lower animals resemble those seen in humans. Many, if
not most, animals with Giardia infection exhibit no symptoms. These animals do, however,
serve as sources of infection for other animals. In those animal species (e.g., cats and dogs)
whose Giardia infections have been studied in detail, the epidemiology is similar to humans.
That is, Giardia infections may occur in animals of any age, but they are more likely to occur,
and to be symptomatic, in young animals. Symptomatic infections in animals that require
therapy usually respond to the same agents, with the same caveats, used in treating human
infections. Mortality is rare in humans but appears to be significant in some animals, e.g.,
chinchillas and budgerigars.
IV-6
-------
V. HEALTH EFFECTS IN HUMANS

A. Symptoms and Clinical Features

Giardia infection is often asymptomatic. Asymptomatic cases may represent as many as
50% to 75% of infected persons (Mintz et al., 1993). In a study at the Swiss Tropical Institute,
only 27% of 158 patients who had Giardia cysts in their feces exhibited symptoms (ICAIR,
1984). Infection may also be associated with a variety of intestinal symptoms including chronic
diarrhea, steatorrhea, abdominal cramps, bloating, flatulence, pale greasy and malodorous stools,
and weight loss (ICAIR, 1984; Benenson, 1995). Nausea or vomiting may also occur (Hopkins
and Juranek, 1991). Fever is occasionally present at the beginning of the infection (Ortega and
Adams, 1997). Blood is not present in stools unless it is due to anal irritation from the diarrhea
(Wolfe, 1990). Malabsorption of fats or of fat-soluble vitamins may also occur (Benenson,
1995). For example, subnormal fractional absorptions of folate and vitamin B12 were found in
one-sixth and one-third, respectively, of 29 Swedish children, age 8 months to 13.5 years, with
chronic giardiasis (Casterline et al., 1997).

Giardia trophozoites principally infect the small intestine. There is usually no extra-
intestinal invasion, but reactive arthritis may occur (Shaw and Stevens, 1987). In severe
giardiasis, duodenum and jejunal mucosal cells may be damaged (Benenson, 1995). Cases of
severe, reversible impairment in pancreatic function have also been reported (Carroccio et al.,
1997; Nakano et al., 1995). Uveitis and urticaria have been observed in several patients with
giardiasis but may have been coincidental (ICAIR, 1984). Inflamation of the synovial
membranes of major joints has also been seen in children with giardiasis, but following anti-
giardial chemotherapy, intestinal and synovial symptoms were abated (ICAIR, 1984).

Infection is frequently self-limited, but persons with AIDS may have more serious and
prolonged infection (Benenson, 1995). Immunocompromised persons, especially those with
acquired immune deficiency syndrome (AIDS) and achlorhydria maybe more susceptible to
v-i
-------
symptomatic infection (1CAIR, 1984). Immunodeficiency with varying degrees of
hypogammaglobulinemia or agammaglobulinema is the most commonly reported form of
immunodeficiency associated with chronic giardiasis (Farthing, 1996). Giardiasisis one of the
few potentially treatable causes of diarrhea in persons with ADDS, and although Giardia infection
is not as prevalent as other pathogens in ADDS patients, it is important that the infection be
accurately diagnosed (Hewan-Lowe, 1997). Co-infection with Giardia lamblia and
Enterocytozoon bieneusi was detected by endoscopically obtained small intestine biopsies from a
patient with AIDS and chronic diarrhea who had repeated negative stool examinations for ova
and parasites (Hewan-Lowe, 1997).

Deaths due to giardiasis are rare; CDC reported that giardiasis had caused only four
deaths in the United States in 1982 (1CAIR, 1984). An estimated 4,600 persons were
hospitalized with giardiasis annually in the United States from 1979 to 1988 with a median
length of hospital stay of 4 days (Lengerich et al., 1994). Volume depletion or dehydration was
the most frequently listed co-diagnosis on admission, and almost 19% of the children younger
than 5 years of age who had severe giardiasis also were diagnosed with failure to thrive
(Lengerich et al., 1994). In Scotland, the median length of stay in the hospital for giardiasis was
significantly longer for persons older than 70 (11 days compared to 3 days) than for other age
groups (Robertson, 1996). Dehydration did not occur as frequently with giardiasis in Scotland,
either because of Giardia strain differences or because rehydration treatments are more widely
self-administered in Scotland. Some 11% of the children who were hospitalized for giardiasis in
Scotland were also found lacking in expected normal physiological development (Robertson,
1996).

The duration of acute clinical illness may vary greatly. In some patients, symptoms last
for only 3 or 4 days, while in others the symptoms last for months. Generally, patients
commonly resolve their infections spontaneously, and acute disease lasts from 1 to 4 weeks
(ICAIR, 1984). In some patients, the acute stage may last for months (Wolfe, 1990). In
V-2
-------
untreated patients, the median duration of illness is six weeks with symptoms lasting less than
one week (Adam, 1991).

The period of communicability lasts for the entire duration of infection, and during
infection, the shedding of cysts can be intermittent (Benenson, 1995). Although persons with
asymptomatic Giardia infection are not likely to seek medical treatment and be diagnosed, they
can serve as unidentified carriers of infection. Carrier infections may last for months or years
(1CAIR, 1984). Asymptomatic Giardia infection for children may be epidemiologically
significant (1C AIR, 1984). Infected children in day-care centers are frequently asymptomatic but
can transmit the infection to other children, care givers, and family members (Ortega and Adam,
1997). In a longitudinal study, almost 15% of 82 children in a day-care center excreted cysts for
a mean of six months (Turner, 1985).

When giardiasis is suspected, it is advisable to confirm that Giardia is the cause of the
illness. For patients with chronic diarrhea, upper abdominal cramps, and "frothy" stools, the
examination of up to three concentrated stool specimens are recommended (Donwitz et al.,
1995; Conboy 1997). The collection of three stools has a sensitivity of 60% to 85% for detecting
Giardia cysts (Donwitz et al., 1995). ELISA for determining Giardia antigen in stool (sensitivity
92%; specificity 98%) has largely supplanted intestinal biopsies, wet preparation, and the
duodenal string test (Donwitz et al., 1995). Benenson (1995) reports that test kits are
commercially available to detect the Giardia antigen in the stool. Howard et al. (1995) recently
detected Giardia in biopsies of the colon and terminal ileum and suggested that physicians may
wish to perform colonic or ileal biopsies when the clinical symptoms suggest giardiasis and the
more routinely-performed duodenal aspirates or biopsies have been found to be negative.
Diagnostic tests are more completely described in Chapter VII, Section B.

B. Epidemiology
V-3
-------
Based on data from infected travelers to the U.S.S.R., the reported mean time period
between infection and the onset of acute disease was 12 to 15 days, but the time ranged from 1 to
75 days (1CAIR, 1984). Ortega and Adams (1997) report the incubation period for symptomatic
persons is one to two weeks but may vary from 1-45 days. Benenson (1995) reports that the
incubation period is usually 3 to 25 days or longer, with a median of 7 to 10 days. In a
prospective epidemiological study, Jokipii et al. (1985) found thatthe incubation period for
giardiasis may typically be in the range of 12 to 19 days. In human volunteers inoculated with G.
lamblia trophozoites by intestinal intubation, Nash et al. (1987) found that and that diarrhea or
loose stools appeared within 7.25 (± 2.99) days of inoculation. In human volunteers fed human-
source Giardia cysts, the incubation period of giardiasis (based on cyst detection in the feces)
ranged from 9 to 22 days with a mean of 13.1 days (Rendtorff, 1954a, b; 1979).

In the United States and Scotland, more severe cases (i.e., hospitalized) of giardiasis
seemed to occur primarily in children under the age of five (Lengerich et al, 1994; Robertson,
1996). Infants and young children may have increased susceptibility to giardiasis because of
behavioral factors that increase exposure and immunological factors (Robertson, 1996). In
Scotland, marked differences were found in the age distribution of hospitalized cases of
cryptosporidiosis and giardiasis (Robertson, 1996). The median age for hospitalization of
giardiasis was 30 years, whereas, the median age for cryptosporidiosis was 5 years, and the
proportion of hospitalized cases in children under five was greater for cryptosporidiosis (49%)
than giardiasis (28%). Robertson (1996) suggested this difference between severity of illness
between these two protozoa may be because the development of protective immunity to Giardia
infection is more prolonged than it is for Cryptosporidium infection. Development of protective
immunity for Giardia infection has been considered to be relatively lengthy and does not
necessarily result from a single infection (Farthing, 1994). The variation of antigenic profiles
between Giardia isolates and its antigenic complexity also suggest there may also be more
immunological sub-types of Giardia, and immunity may be specific for the particular sub-types
V-4
-------
(Robertson, 1996; Rabbani and Islam, 1994). Immunity is discussed further in Section E of this
chapter.

Giardiasis is transmitted via the fecal oral route of exposure, and both endemic and
epidemic transmission are important. Although all age groups are affected, the highest incidence
is in children (Benenson, 1995). Infants under 6 months of age who are breast-fed are not likely
to be infected (Rabbani and Islam, 1994). It is a common cause of illness in travelers and often
spread directly from person to person, especially among children or persons living in areas with
poor sanitation and hygiene. Waterborne outbreaks have been reported, and some have resulted
in a large number of cases of illness. Endemic waterborne giardiasis has been associated with
drinking unfiltered surface water or shallow wells and swimming. Smaller outbreaks have
resulted from contaminated food and person to person transmission in day-care centers. Oral-
anal sexual activities among homosexuals has also been described as a risk factor (Turner, 1985).

Although people living in urban and rural areas may have different levels of risk of
Giardia infection, both are at high risk of infection. In Zimbabwe, the annual incidence of the
disease in urban children was 22%, compared to 12% for rural children (Rabbani and Islam,
1994). High population density in urban areas, overcrowding, poverty, and poor sanitation of the
urban slum areas, especially in third world countries, contribute to the high rate of infection.
Like other gastrointestinal infections, giardiasis is very common in populations living in poverty
and with poor sanitation, and a high level of fecal contamination of the environment. Mason et
al., (1986) indicates that even providing piped, clean drinking water to developing countries may
not significantly reduce the incidence of giardiasis. Although contaminated drinking water is a
likely source of exposure, the variety of other exposures including personal hygiene, food
hygiene, and environmental factors may overwhelm the beneficial effect of clean drinking water.
Further studies are required to understand the definite role of socio-environmental factors
contributing to giardiasis.
V-5
-------
C. Clinical Laboratory Findings and Therapeutic Management

1. Clinical Laboratory Findings

Clinical data suggest Giardia cysts are highly infective for humans. In a controlled,
clinical study of male volunteers who were fed human-source Giardia cysts contained in gelatin
capsules, a dosage often cysts was found to produce human infection, as determined by
observing presence of Giardia in fecal smears (1C AIR, 1984). Eight dosage levels ranging from
1 cyst to 1 x 106 cysts per capsule were studied. Since cyst viability was not determined before
being fed to volunteers, the failure to elicit infection in the five men treated with a dose of only
one cyst may have been due to dosing with inactive cysts (1CAIR, 1984).

Nash et al. (1987) inoculated by intestinal intubation human volunteers with trophozoites
of two distinct human isolates of G. lamblia, termed GS/M and Isr. Groups of five volunteers
received 50,000 trophozoites of either isolate. None of the volunteers receiving Isr became
infected, but all of the group inoculated with GS/M became infected. Of five volunteers
inoculated and infected with GS/M, 3 became ill, with 2 showing diarrhea and other signs typical
of giardiasis. These data suggest there are strain variations for G. lamblia and confirm in a
controlled clinical setting that infected persons can exhibit a range of symptoms in addition to
being asymptomatic.

Although one species of Giardia is believed to infect humans, the epidemiology of
giardiasis is complicated by apparent genetic heterogeneity in this species (Thompson et al.,
1993; Erlandsen, 1994; Nash et al., 1987). Differences in virulence, pathogenicity, infectivity,
growth, drug sensitivity, and antigenicity have been reported (Thompson et al., 1996). Genetic
diversity in the species of Giardia believed to infect humans is extensive with some clones
widely distributed and others localized, especially in areas where endemic infection is high
(Thompson et al., 1996). In these endemic areas where extensive heterogeneity of Giardia
V-6
-------
exists, mixed infections with more than one genotype may occur and interference with clonal
competition may have an important influence on the genetic variation (Thompson et al., 1996).
Upcroft et al. (1995) conducted a long term study of the biology and genetics ofGiardia after
establishing in vitro and in vivo Giardia cultures in mice from 1829 duodenal aspirates collected
over an eleven year period from children who were being treated for diarrhea and failure to thrive
and in whom stool examinations were negative. Based on biochemical characteristics of
electrophoretic karyotype, RFLP analysis and rDNA hybridization studies, at least two distinct
varieties, or demes, of Giardia were found to have infected the population of the South East
Queensland area of Australia. From 1983 to 1991 only one variety was documented, but in 1991
a new variety was seen with a predominance of this deme beginning in 1993. Since all of the
stocks were derived from children with similar chronic symptoms it appears that at least two
demes of Giardia were pathogenic in the South East Queensland area of Australia.

Thompson et al. (1996) has suggested that the regular suboptimal application of
chemotherapeutic regimes may be a contributing factor to the persistence of genetic
heterogeneity and that this, combined with the variable sensitivity ofGiardia to these drugs, may
inhibit competitive interactions between clones of Giardia. Competitive interactions studied in
vitro found that competition occurred between genetically distinct isolates of G. duodenalis and
that exposure to metronidazole has differential effects on growth of the clones; however,
although these are necessary conditions, they are not sufficient to support suggestions that
genetic heterogeneity is due to regular suboptimal drug treatments (Thompson et al., 1996)

2. Therapeutic Treatment and Management

As with all diarrheas, fluid replacement is an important aspect of treatment. Anti-giardial
drugs are also important in the management of the giardiasis. Chemotherapeutic agents used for
treatment of giardiasis are listed in Table V-l (Benenson, 1995; Rabbani and Islam, 1994; Bulut
et al., 1996; Freeman et al., 1997; Farthing, 1996; Adam, 1991). The drugs may have different
V-7
-------
effectiveness in their ability to clear Giardia and various doses and treatment periods are
recommended for each drug. Important implications for the management of patients include
problems of drug resistance and relapses that may occur (Benenson, 1995) and cross-resistance
and treatment failures that can occur in the absence of resistance (Upcroft et al., 1990).

After inducing albendazole resistance in three different Giardia cultures following
growth in successively increasing amounts of drug, Upcroft et al. (1996a) found that the cultures
were resistant to concentrations of albendazole against Giardia in vitro at 4-10 times normally
lethal concentrations. Albendazole-resistant Giardia were also cross-resistant to parbendazole
(Upcroft et al., 1996a). Recovering a metronidazole-resistant strain of Giardia from a patient,
Butcher et al. (1994) felt that an unsuccessful course metronidazole treatment for chronic
giardiasis may have resulted in the selection of the resistant strain of the parasite. Quinacrine
resistance was studied by Upcroft et al. (1996b). Quinacrine was found to be rapidly absorbed by
sensitive trophozoites but actively excluded from resistant trophozoites. Upcroft et al. (1990)
reviewed the evidence for drug resistance in giardiasis and suggested markers, such as DNA
probes, be developed to provide methods for monitoring the spread of drug resistant Giardia in
populations. Biochemical studies were also undertaken to determine the basis for this resistance
(Upcroft et al., 1990). Metronidazole and furazolidone, which produce toxic radicals within the
cell, have different biochemical mechanisms of action. At the molecular level, resistance to
metronidazole is associated with DNA changes.

Metronidazole or tinidazole has been the drug of choice for giardiasis probably because
the treatment period is short and compliance good (Farthing, 1996; Benenson, 1995). Quinacrine
and furazolidone have also been commonly used (Freeman et al., 1997). Freeman etal. (1997)
reports that metronidazole is not approved for therapy in Giardia infection in many countries,
and Paget et al.(1989) has found Giardia cysts are resistant to metronidazole. Rabbani and Islam
(1994) report that in the United States, metronidazole is not approved by the Food and Drug
Administration (FDA) for treatment of giardiasis but is approved for amebiasis. Farthing (1996)
-------
and Benenson (1995) note that tinidazole is also not approved by the FDA for giardiasis.
Furazolidone i s reportedly the only drug approved by the FDA for treatment of giardiasis (Ortega
and Adam, 1997).
Table V-l Chemotherapeutic Agents for Giardiasis
Drug
Quinacrine
Metro nidazole
Tinidazole
Furazolidone
Paromomycin
Albendazole
Ornidazole
Duration of Treatment
5-10 days
2-14 days; single dose adults
7 days; single dose adults & children
7-10 days
5-7 days
5 days
single dose children
Efficacy
>90%
>90%
>95%
>80%
Low may be <50%
>90%
>90%
*Adapted from Benenson, 1995; Rabbani and Islam, 1994; Bulut et al., 1996; Freeman et al., 1997;
Farthing, 1996;Adam, 1991.
Studies have evaluated the effectiveness of metronidazole and compared its effectiveness
to other medications and in combination with quinacrine or d-propranolol, an adrenocepter
antagonist that appears to inhibit the mobility and growth of G. lamblia (Freeman et al., 1997).
A large single or repeated dose of 2.0 g or 0.25 g three times daily for 7 days is reported effective
for adults; dosages of 5.0 to 7.5 mg/kg three times daily for seven days are effective for children
(Freeman et al., 1997). Ellis et al. (1993) noted that G. intestinalis is often refractory to
treatment with metronidazole. Some patients who fail to respond to a single dose of
metronidazole have responded to a second therapy of 3 or 7 days duration (Freeman et al., 1997).
A combined formulation of diloxanide furoate and metronidazole was successfully used to clear
the parasite from all of their patients with giardiasis (Qureshi et al., 1997). Metronidazole
V-9
-------
appears to have fewer side effects than furazolidone and quinacrine, but nausea, metallic taste,
and headache may occur (Turner, 1985). Metronidazole and furazolidone have been found
mutagenic and carcinogenic in animal experiments (Turner, 1985).

Tinidazole, a chemical relative of metronidazole, is reported to be equally or more
effective than metronidazole (Freeman et al., 1997) and has fewer side-effects (Rabbani and
Islam, 1994). A single dose of 2.0 gin adults (0.5 or 1.0 gin children) has been used with
success (Rabbani and Mam, 1994; Nahmias et al., 1991). Mepacrine or quinacrine is still used in
some parts of the world and has proved effective where metronidazole and tinidazole have been
unsuccessful (Upcroftet al., 1995; Farthing 1996); however, it is not available in a number of
countries, and side-effects including the risk of psychoses have been reported (Boreham, 1991).
Mepacrine may be a useful alternative to metronidazole in the United States, but its use often
leads to yellow staining of the skin and conjunctivae (Rabbani and Islam, 1994). Albendazole, a
benzimidazole derivative, has also been shown effective in vitro at concentrations 30 times lower
than metronidazole (Rabbani and Islam, 1994). Misra et al. (1995) found that albendazole is as
effective as metronidazole for treating gjardiasis in children and does not produce the anorexia
that is often seen with metronidazole treatment. It was found to be almost as effective as
metronidazole in treating Bangladeshi children; 95% of those infected and treated with
albendazole cleared the parasite compared to 97% clearance with metronidazole (Hall and Nahar,
1993). In addition, albendazole is less expensive and has fewer side-effects than metronidazole
(Bulut et al., 1996). Dutta et al. (1994) found albendazole as effective as metronidazole in a
study of 150 Indian children aged 2-10 years randomized to receive either a single dose of 400
mg of albendazole suspension, or 22.5 mg/kg/day of metronidazole in 3 doses for 5 consecutive
days. Two days after completion of therapy, 97% of children in both treatment groups were
Giardia free. Side effects were noted in 3 children in the albendazole group, but in 20 children
in the metronidazole group. Pungpak et al. (1996) found that a seven day course of albendazole
was effective with no serious side effects among 63 children and 15 adults in Thailand. Another
v-io
-------
benzimidazole derivative, mebendazole has been found effective in treating giardiasis (Bulut et
al., 1996; Adam, 1991).

In a randomized clinical trial of 48 infected children, Bulut et al. (1996) found that
ornidazole was 100% effective in clearing Giardia compared to 93% effectiveness for
metronidazole and less than 60% for mebendazole (Bulut et al., 1996). Ornidazole at 40 mg/kg
was administered as a single dose with only minor side-effects in three children (urticaria,
vertigo, nausea). Bassily et al. (1970) found that furazolidone treatment cleared the parasite in
80% of infected Egyptian adults; quinacrine and metronidazole were 100% and 95% effective,
respectively. Furazolidone is widely used for children in the United States partially because it is
available in pediatric suspension (Farthing, 1996). A Giardia clearance rate of 92% was reported
in children treated with furazolidone (Craft et al., 1981). Furazolidone is well tolerated by most
patients, but may cause a reaction if taken with alcohol and may induce hemolysis in patients
with glucose-6-phosphate dehydrogenase deficiency (Rabbani and Islam, 1994). While
furazolidone is an effective treatment for giardiasis, Elliott et al. (1998) warn of the potent
cumulative inhibition of monoamine oxidase associated with the prolonged use of the drug with
potential effects on blood pressure and mood disorder; interactions with antidepressant drugs and
foods rich in tyramine should also be considered.

Turner (1985) recommended treatment of giardiasis be avoided in pregnancy unless
symptoms cannot be controlled by conservative measures. Because it is poorly absorbed,
paromomycin has been used to treat giardiasis in pregnant women in whom other drugs are
contraindicated; the cure rate is variable and may be as low as 55% (Rabbani and Islam, 1994;
Farthing, 1997).

Pearce et al. (1996) compared two published methods for assessing thein-vitro drug
sensitivity of Giardia duodenalis to metronidazole or albendazole: inhibition of adherence and
the 3H-thymidine incorporation assay which radiometrically measures nucleic acid synthesis.
v-n
-------
Because of the different modes of action of metronidazole and albendazole on Giardia,
measuring the inhibition of adherence appears to be a more accurate indicator of trophozoite
viability than measuring 3H-thymidine incorporation. This finding emphasizes the importance of
considering the biochemical mechanisms of action when evaluating drug sensitivity. Feeling that
differences in enzyme characteristics between the parasite and host may lead to development of
future chemotherapeutics for giardiasis, Swarbrick et al. (1997) sequenced the cytidine
triphosphate synthetase genes from three diverse strains of G. duodenalis and found that they
varied significantly from each other. Boreham et al. (1987) tested two stocks of G. intestinalis
by the 3H-thymidine uptake assay to determine their sensitivity to metronidazole, tinidazole,
furazolidone and quinacrine. Each stock was composed of different populations of organisms
and not homogeneous with respect to drug sensitivity, and this may, in part, account for
treatment failures in giardiasis patients.

Pentamidine and 38 analogs of pentamidine were screened for in vitro activity against G.
lamblia (Bell et al. 1991). All compounds were active against G. lamblia as measured by the
3H-thymidine incorporation assay, but anti-giardial activity varied widely. The activity of the
most potent anti-giardial agent, l,3-di(4-amidino-2-methoxyphenoxy)propane compared
favorably with furazolidone, metronidazole, quinacrine, and tinidazole. Gordts et al. (1985)
evaluated the in vitro susceptibility 25 Giardia lamblia isolates to six commonly used
antiprotozoal drugs; tinidazole was the most active drug. Metronidazole was equally active on
all but one isolate, and furazolidone was the most active nonimidazole compound tested. More
than 50% of the isolates were very susceptible to paromomycin, pyrimethamine, and
chloroquine. Crouch et al. (1986) evaluated the in vitro sensitivity of G. lamblia to 23
chemotherapeutic agents; tinidazole, metronidazole, and furazolidone were found to have strong
inhibitory effects on both growth and adherence, while mepacrine had a strong effect on growth
only. Three drugs (mefloquine, doxycycline and rifampin) not previously used in giardiasis were
also found to have significant in vitro activity and deserve consideration for clinical evaluation of
efficacy. Farbey et al. (1995) examined 12 isolates of G. duodenalis from Caucasian hosts in the
V-12
-------
Perth metropolitan area, 16 isolates from Aborigines in Western Australia, and a reference isolate
P1C10 for their in vitro drug sensitivity to metronidazole, benzimidazole, and albendazole.
Metronidazole showed the most resistance. In addition, it was found that isolates of Giardia
obtained from Aboriginal hosts were significantly less sensitive to albendazole than those
obtained from Caucasians. In vitro growth of G. lamblia was found to be highly sensitive to
certain anthelmintic benzimidazoles (Edlind et al., 1990). Albendazole and mebendazole were
30- to 50-fold more active than metronidazole and 4- to 40-fold more active than quinacrine.
Thiabendazole was less active. Since lack of intestinal absorption makes mebendazole an
attractive anti-giardial agent, its in vitro activity was further characterized. Lethal activity was
observed at a concentration fivefold lower than necessary for metronidazole. Attachment of
cells to the culture tube was rapidly disrupted by mebendazole treatment, and the characteristic
cell structure was grossly distorted. Azithromycin was found to produce significant growth
inhibition of G. intestinalis in vitro (Crouch et al., 1990). Crouch et al. (1990) also found that
the dyadic combinations of azithromycin-furazolidone, doxycycline-mefloquine,
doxycycline-tinidazole and mefloquine-tinidazole were synergistic for inhibition of adherence of
G. intestinalis in vitro, suggesting combinations may be worthy of consideration for
chemotherapy of recalcitrant giardiasis. Meloni et al. (1990) compared the effects of albendazole
against G. duodenalis in vitro with those of tinidazole and metronidazole, finding it to have
superior potency. Trophozoite morphology, adherence and viability were markedly affected by
albendazole, to a far greater extent than by either metronidazole or tinidazole.

Ponce-Macotela et al. (1994) evaluated in vitro anti-giardial activity of 14 species of
plants in Mexico as anti-diarrheics and/or anti-parasitics. Trophozoites of G. duodenalis were
incubated with plant extracts and their viability was ascertained. In vitro anti-giardial effects
were seen in nine species; Justicia spicigera (muicle), Lipia beriandieri (oregano), andPsidium
guajava (guava) were found to be superior to tinidazole.

D. Mechanism of Action
V-13
-------
Giardia cysts survive in the environment, and when ingested pass through the stomach
where the acid environment triggers excystation, which usually takes place in the duodenum.
The trophozoites attach to the duodenal or proximal jejunal mucosa, probably via contraction of
the ventral disk, and replicate by repeated binary division (Marshall et al., 1997). Attachment of
trophozoites is essential for colonization of the small intestine and a prerequisite for Giardia-
induced enterocyte damage (Katelaris et al., 1995). A predominant role for mechanical
attachment via cytoskeletal mechanisms is suggested by in vitro studies in cultured human
intestinal cells, but Katelaris et al. (1995) feel that lectin associated binding may also have a role
in vivo. Abnormal structural changes may occur in the mucosa but are usually reversed after
treatment (Hall, 1994).

In severe giardiasis, duodenum and jejunal mucosal cells maybe damaged (Benenson,
1995). The severity of the diarrhea has been positively correlated with the severity of changes in
the villus histology (1CAIR, 1984). The more severe the villus atrophy, the more severe the
diarrhea. After anti-giardial chemotherapy, recovery of the villus architecture occurred as
diarrhea disappeared. Structural changes often involve a flattening of the mucosal surface and a
change in the ratio of the length of crypts and villi; this could be responsible for malabsorption
(Hall, 1994). The relationship between clinical disease and structural change in the mucosa is
not always consistent, but the presence of inflammatory cells in the lamina propria seems to be
common (Hall, 1994). A case of giardiasis was reported in a female college student whose
symptoms persisted for 5 months and no villus atrophy was noted, but marked round cell
infiltration occurred in the lamina propria (1CAIR, 1984). Because coeliac disease, a result of an
immune response to gluten, leads to an enteropathy and inflammation similar to that seen in
giardiasis, Hall (1994) suggests that diarrhea and other symptoms of giardiasis result from an
inflammatory response to infection. It has been proposed that a toxic excretory or secretory
product could be responsible for diarrhea, but the presence of a toxin has not been found nor is
there evidence showingthat Giardia is toxigenic (Hall, 1994; ICAIR, 1984). Other proposed
V-14
-------
mechanisms for diarrhea include disruption of the brush border or immunopathologic processes
(Ortega and Adam, 1997). Studies of Mongolian gerbils infected orogastrically with trophozoites
suggest that an altered gastrointestinal transit and smooth muscle contractility maybe involved in
the pathophysiology of giardiasis (Deselliers et al., 1997).

Trophozoites do not normally penetrate the intestinal epithelium in humans, but evidence
of mucosal invasion has been seen in patients who had diarrhea and large numbers of
trophozoites in the lumen; similar invasion has not been seen in asymptomatic persons (Ferguson
et al., 1990). An ultrastructural study of mouse infection found mucosal invasion only in areas
where necrosis or mechanical trauma was present, and reports of mucosal invasion in humans are
suspect if specimens were obtained by forceps biopsy or peroral suction (Ferguson et al., 1990).
Chavez et al. (1995) found that all strains of G. lamblia recovered from children with
symptomatic and asymptomatic giardiasis produced focal regions of microvilli depletion in vitro
(MDCK epithelial cells), but none of the isolate strains were invasive.

Several mechanisms have been suggested to account for nutrient absorption
abnormalities, including mechanical blockage of mucosal surfaces, functional mucosal changes
brought on by invading trophozoites, mucosal damage from inflammation even in the absence of
actual invasion, associated bacterial overgrowth, deconjugation of bile acids, and interference
with lipolysis (Hall, 1994; 1C AIR, 1984). However, experimental evidence for only the latter
mechanism was found in the pre-1984 literature and the results were equivocal (ICAIR, 1984).
Laboratory experiments suggested Giardia may interfere with the active transport of glucose and
glycine, but not with the passive transport of potassium, and that the defect in the active transport
mechanism might be due to structural damage of the substrate carrier or an alteration in cell
maturation due to Giardia (ICAIR, 1984).

As noted in Sections A and C of this chapter, clinical effects range from the
asymptomatic carrier state to severe malabsorption syndrome. Similarly, histopathological
V-15
-------
changes in the affected mucosa maybe minimal, or there maybe significant enteropathy with
enterocyte damage, villus atrophy, and crypt hyperplasia (Ferguson et al., 1990). Reasons for this
variation in host susceptibility are poorly understood (Ferguson et al., 1990). Hall (1994)
concluded that the mechanism by which Giardia causes disease and effects nutrition is unclear.

E. Immunity

Data supporting the role of acquired immunity to giardiasis comes from studies of both
humans and animals. Results of epidemiological studies, studies of immunosuppressed human
populations, animal experiments where artificial immunosuppression was induced, and studies of
the immune and physiological reactions to Giardia infection of both humans and animals are all
consistent with the development of protective immunity to illness from prior Giardia infections.
These studies indicate that prior infections induce an immune response, and persons with an
immune response from prior Giardia infection(s) have a reduced risk of illness when a
subsequent Giardia infection occurs. Although the immune response may not result in reduced
risk of subsequent Giardia infections, subsequent infections are more likely to be asymptomatic.

1. Epidemiological Data Supporting Acquired Immunity

The epidemiology of giardiasis in developed countries indicates that persons episodically
exposed to Giardia cysts are more likely to suffer symptoms and illness. These persons include
travelers to certain locales, backpackers, expatriates, case-contacts, and persons exposed during
waterborne outbreaks (Janoff and Smith, 1990). In contrast, residents with recurrent exposure to
Giardia are commonly asymptomatic (Janoff and Smith, 1990).

High rates of exposure to and high rates of carriage of Giardia are associated with low
rates of symptomatic illness, and persons infected with Giardia in developing countries are
usually asymptomatic (Oilman et al., 1988; Zaki et al., 1986; Walia, et al., 1986; Nacapunchai et
V-16
-------
al., 1986). In developing countries, exposure to Giardiabegins early in life (Zaki et al., 1986;
Oilman et al., 1985), high rates ofGiardia carriage are present in all age groups (Zaki et al.,
1986; Janoff et al.,1990). Asymptomatic reinfection occurs at ahigh rate (Oilman et al., 1988).
For example, in Egypt Giardia is frequently detected in healthy subjects. Immunity to Giardia
may be particularly important in recurrently exposed persons who, though often infected, are
infrequently ill. In Bangladesh, symptoms are reported to occur in 86% of infants newly infected
with Giardia but in only 4% of infected mothers (Janoff and Smith, 1990). In Thailand, most
school children and adults, repeatedly exposed to Giardia, are asymptomatic (Chavalittamrong et
al., 1978; Waliaetal., 1986).

Studies have also suggested that persons who travel to Giardia endemic areas from
relatively non-endemic areas carry a high risk of developing giardiasis. Speelman and
Ljungstrom (1986) reported that the annual incidence of giardiasis was 12% among 251
expatriates in Bangladesh with 37% of infected persons developing diarrhea. Giardiasis was
more common among the newcomers and children less than 10 years old. Epidemiological
studies also suggest that new settlers or travelers visiting areas where Giardia is endemic are
more likely to develop symptomatic illness, possibly related to the lack of immunity from prior
infections (Rabbani and Islam, 1994). Age which may be a measure of prior exposure or
infection has also been related to a reduced risk of giardiasis. If increasing age is related to the
likelihood of prior Giardia infection and persons develop immunity, then older people would be
expected to be less susceptible to illness or severe illness from infection. The attack rate for
giardiasis has been found to decline with increasing age supporting this hypothesis (Farthing
1989).

Since infection occurs at an early age and reinfection is common in developing countries,
it has been argued that a vaccine to Giardia, if developed, would have only limited application in
these countries (Janoff and Smith 1990). Immunity to giardiasis likely develops at a early age,
possibly before a vaccine could be effectively administered (Oilman, 1988). Based on this
V-17
-------
argument a vaccine would be most helpful for persons infrequently or episodically exposed to
Giardia, such as travelers and military personnel (Janoff and Smith, 1990).

Persons in developing countries are at a low risk of symptomatic giardiasis and also have
high levels of parasite-specific antibody (Oilman et al., 1985; Nacapunchai et al., 1986; Miotti et
al., Janoff et al., 1988a). Healthy homosexual men also have a high frequency of asymptomatic
infection in association with increased levels of Giardia-specific antibody (Janoff et al., 1988a).

Specific evidence for acquired immunity to giardiasis has been found in several
epidemiological studies in developed countries. In Colorado, visitors were found more likely to
experience symptomatic giardiasis than long-term residents, and residents who lived in the area
for more than two years had a lower attack rate for illness during a waterborne outbreak than
short-term residents (Wright, 1977; Istre 1984). The presumption is that long-term residents
were repeatedly exposed to Giardia cysts through drinking water and perhaps other sources, and
they developed an immunity which protected them, when reinfected, from illness or less severe
illness. Visitors, with fewer prior exposures to Giardia, did not develop this immunity and were,
therefore, more likely to become ill when infected. Another study found that persons with prior
diagnosed giardiasis were at a lower risk of giardiasis during a subsequent exposure. In a
community that experienced two waterborne giardiasis outbreaks separated by a five year period,
individuals infected during the first outbreakwere at significantly lower risk during the second
outbreak (Isaac-Renton, 1994; Isaac-Renton et al., 1994). See Section F, Chapter III for a further
discussion of these outbreaks.

2. Breast Milk and Breast Feeding Reduces the Risk of Giardiasis

The role of breast-feeding in preventing enteric infections in young children is now been
well recognized. However, in these studies it is often difficult to distinguish the role of
antibodies in the breast milk versus other anti-parasite effects of milk. Secretory IgA (slgA) to
V-18
-------
Giardia has been demonstrated in the breast milk from women in Bangladesh (73%) and Mexico
(77%), indicating a high incidence of the disease in these populations (Islam et al, 1983; Miotti
et al., 1985). Infants in these areas may thus acquire antibodies byway ofbreast milk (Islam,
1983, Miotti, 1985). An experiment in which children were exclusively fed breast milk resulted
in a very low rate of Giardia infection, and the infections which occurred were mostly
asymptomatic (Rabbani and Islam, 1994). This may explain why giardiasis is uncommon in
infants younger than 6 months, during which period most children are usually breast-fed.

The role ofbreast milk in the prevention of giardiasis has been investigated in both
animals and humans. Suckling mice are protected against G. muris when they are fed milk from
immune mothers containing specific IgA antibodies to Giardia. It has also been shown that
Giardia is rapidly killed by exposure to normal human milk in vitro. However, this killing effect
is not mediated through antibodies; rather, the effect is related to the exposure of the organisms
to an enzyme, bile salt-stimulated lipase. Similar observations are reported by Reiner et al.
(1986) who showed that normal human milk kills Giardia trophozoites in vitro and that this
effect is mediated by the release of free-fatty acids from milk triglycerides by the action of the
bile-stimulated lipase on human milk. Thus, from the public health point of view, children who
are not breast-fed are at a higher risk of developing Giardia infections, particularly in the
endemic areas with high levels of fecal contamination of the environment, and are most at risk of
severe consequences of giardiasis.

3. Increased Giardiasis Risks in Immunosuppressed Populations

Investigators have found that athymic mice exposed to Giardia are at an elevated risk of
suffering from chronic infections whereas exposed immunocompetent mice appear to clear the
infection and become immune to reinfection (Roberts-Thomson et al. 1976). Implantation of the
thymic tissue in alhymic mice reduced the number of mice chronically infected with Giardia
(Roberts-Thomson and Mitchell, 1978). Other studies have indicated that most strains of mice
V-19
-------
appear to be resistant to reinfection after clearing a primary infection whereas T-deficient mice
have prolonged carriage and fail to produce local antibody (Janoff and Smith, 1990).

Based on evidence from the studies of G. muris infection in athymic mice, the role of T
lymphocytes in the host response to Giardia was investigated in normal mice. During infection
with G. muris, the number of Peyer's patch lymphocytes may double, although the ratio of T
helper to T suppressor lymphocytes (>5:1) does not change (Clark and Holberton, 1986). When
these mice are depleted of T helper cells they develop chronic infection, whereas mice depleted
of T suppressor cells or deficient in natural killer cells clear their infection (Heyworth et al.,
1986, 1987). This indicates that T helper cells may play an important role in the ability of the
murine host to clear Giardia. Recently, T helper cells were reported to contribute, possibly as
'switch' cells, to the sequential change in Peyer's patch B cells from IgM-bearing during the first
week of infection to IgA-bearing B cells during the second week of infection (Clark and
Holberton, 1986). Although intriguing, the role of T helper cells in the local immune response to
G. muris remains speculative at this time. (Janoff and Smith, 1990).

At the level of the intestinal mucosa, the host response to Giardia is initiated by the
presentation of parasite antigen to T lymphocytes. In vitro studies of murine, rabbit and human
macrophages confirm that macrophages are capable of phagocytosing Giardia trophozoites. The
ability of the parasite to invade the intestinal mucosa, and the presence of macrophages in the
lamina propria, would facilitate contact between the parasite and antigen-presenting cells. In
addition to antigen presentation, effector cell function by monocytes-macrophages contributes to
host defense mechanisms against many pathogens. The ability of resident macrophages and local
blood monocytes in the lamina propria to kill Giardia could provide an important defense
mechanism against invading trophozoites. In this regard, three reports indicate that human
mononuclear phagocytes are capable of killing Giardia, possibly by the products of oxidative
metabolism (Hill and Pearson, 1987). Aggarwal and Nash (1986) did not confirm these findings.
V-20
-------
Results of animal studies conducted to understand the importance of immunological
responses to infection indicated that when anti-Giardia antibodies are transferred to
immunoglobulin deficient mice the risk of developing chronic infection is not reduced (Snyder et
al., 1985). The authors concluded that immune responses in the intestine tract are required for
protection against illness. In vitro work (Goka et al 1986, Nash et al., 1987 Janoff etal., 1988b)
found that anti-Giardia IgM antibodies were capable of sensitizing Giardia for complement
lysis. These findings support the role of secretory anti-Giardia antibodies in clearing infection.

The findings of a high risk of chronic infection in immunocompromised animals are
consistent with findings from human studies. People with hypogammaglobulinemia are at high
risk of chronic infection (Janoff and Smith, 1990). Children with chronic diarrhea and giardiasis
have an increased incidence of hypogammaglobulinemia (Perlmutter et al., 1985). The presence
of circulating anti-Giardia antibody may not be sufficient protect immunocompromised persons
from infection, and one study (Smith et al., 1982) reported recurrent giardiasis in some human
patients with high liters of anti-trophozoite antibodies.

Cellular immune response to infection occurs in the intestinal mucosa. Varying degrees
of histological change following infection have been observed. The accumulation of
inflammatory cells in the small intestinal mucosa of infected persons suggests that cellular
response may be an important component of the host response to infection. As previously
indicated, T cells appear to play a significant role in clearing infection. Human peripheral blood
mononuclear cells were spontaneously cytotoxic for G. lamblia trophozoites, and the observed
cytotoxicity was a host-defense mechanism directed against extracellular protozoa in general
(ICAIR, 1984). Where there was both giardiasis and malabsorption, the lymphocyte counts were
significantly higher than those in the controls, as well as in those patients with giardiasis and
normal absorption. In most cases, the intraepithelial lymphocyte counts declined after giardiasis
treatment.
V-21
-------
4. Measuring Epidemic and Endemic Infections in Humans

A variety of techniques have been developed for improving the detection oi Giardia cysts
in stool specimens. These include an ELKA assay which improves the microbiologist's ability to
distinguish cysts from other similar sized particles. Since Giardia is a common intestinal
infection among children, stool surveys are a feasible method for estimating the prevalence of
infection. Approximately 7% of children in diapers were found to be infected in one survey
(Frost etal., 1983).

a. Anti-Giardia Antibodies in Sera

A variety of methodologies have been employed for detection of anti-Giardia antibodies
in serum, including immunodiffusion, hemagglutination, immunofluorescence, and ELISA
(ICAIR, 1984; Sullivan et al., 1987). Use of serum antibody tests to identify infections have
several limitations. Serological assays have a reduced clinical value since the serological
response may take several weeks to appear. There is also no agreement on whether serological
tests can distinguish an active infection from a recently cleared infection, and it is unclear how to
interpret the results of serological tests performed on chronically infected persons.

Serological tests have advantages for epidemiological studies. Sera tend to be easier to
obtain than are stool samples. Also, if the serological response is longer lived than the infection,
it may be possible to better distinguish populations with a low versus high prevalence of
infection. Unfortunately, the current serological methodologies have significant limitations.
Most of the ELISA tests are based on serological reaction to all proteins in the parasite. Since
Giardia is a complex organism with many potential human antigens, it is likely that much of the
response measured by the ELISA test is non-specific, resulting from serological reactions to
antigens which are shared by Giardia and a variety of other organisms. Western blot tests, which
look for serological response to antigens separated by weight, are likely to be more specific, but
V-22
-------
there is currently only limited experience in using Western blots for serological surveys of anti-
Giardia antibodies in human populations.

The pre-1984 literature on detection of anti-Giardia antibodies in serum has been
reviewed (1CAIR, 1984). Elevated levels of serum immunoglobulin IgG, IgM and IgA were
reported in patients withgiardiasis (ICAIR, 1984). There seemed to be a consensus that levels of
serum IgE and probably IgD in giardiasis patients may not differ from those of normal controls.
Unfortunately, various studies have not reported consistent findings of elevated immunoglobulins
G, A and M, due in part to the limitations of the assay used and in part to different populations
surveyed. In theory, the early immune response to Giardia infection was believed to be restricted
to IgM, followed later by IgA and IgG (ICAIR, 1984), however not all serological studies have
found elevated levels of IgM in recently infected persons (Birkhead et al., 1989).

Nash et al. (1987) reported specific anti-Giardia serum IgM responses in 100% of human
volunteers experimentally infected with human-source Giardia. Such responses were observed
within two weeks of infection in 70% of acute episodes, but were low in chronically infected and
rechallenged individuals. The IgM detected by the ELISA procedure may have limited
diagnostic usefulness in chronic or repeated infections.

Problems associated with attempts to diagnose giardiasis by testing a single serum sample
for total human globulin IgG antibodies have been described. Although some authors believe
that serologic procedures are reliable for retrospective diagnosis of symptomatic patients, they
are unreliable in asymptomatic patients due to nonspecific antibody liters caused by intestinal
parasites other than Giardia. Jokipii et al. (1988) concluded that their ELISA procedure, which
employed Giardia cysts as antigen, was insufficiently sensitive and specific because the anti-
Giardia antibody liters in the patient population were indistinguishable from liters in healthy
controls.
V-23
-------
Haralabidis (1984) demonstrated that sera from patients with microscopically-proven
Giardia infection reacted by ELK A with a variety of parasite antigens indicating either co-
infection or a cross-reaction between anti-Giardia antibodies and non-Giardia parasite antigens
(a lack of specificity). Jokipii et al. (1988) concluded that sera from most people contains
antibodies which cross-react with Giardia cysts, and that these antibodies maybe induced by
immunogens other than Giardia.

In 1986 a method was published for detect!on of relatively short-lived IgM antibodies.
Results might be more clinically relevant, in that they would more likely indicate the cause of a
patient's present symptoms. Goka et al. (1986) concluded that serum IgG responses were not
helpful in distinguishing active from past Giardia infection because they were relatively long-
lived.

Taylor and Wenman (1987) and Heyworth and Pappo (1990) reported that convalescent
sera of most patients with giardiasis contain antibodies directed against a specific 30/31 kDa G.
duodenalis antigen. Serologic assays using this purified antigen might prove to be more sensitive
and specific than assays employing a mixture of antigens, such as crude trophozoite or cyst
preparations. Other authors have suggested that high molecular weight antigens (155 and 170
kDa) or a 57 kDa antigen are predictive of infection (Char, 1991). These approaches have
considerable merit, since cross-reactions are likely reduced in an assay which only examines
serological response to selected antigen groups rather than to the entire parasite. Additional
research is needed to confirm these findings and apply these approaches in population studies,
especially during outbreaks. This may result in a highly sensitive and specific marker of
infection which could be used to compare endemic levels of infection in various communities.

b. Anti-Giardia Antibodies in Saliva
V-24
-------
Results of an immunofluorescence procedure to detect Giardia-specific antibodies in
saliva were compared to serum liters. Results suggest that saliva may be a more suitable
specimen than serum for detection of anti-Giardia antibodies in patients with giardiasis but more
studies are needed to confirm these results (and Pickering, 1990). Saliva tests have an intuitive
appeal since they can be applied to studies of children and do not require drawing blood.
Rosales-Borjas et al. (1998) studied the secretory immune response during natural Giardia
infections in 24 patients and were able to demonstrate a secretory IgA response in their saliva
that was not present in healthy individuals.

c. Anti-Giardia Antibodies in Intestinal Secretions

Nash et al. (1987) described humoral and intestinal fluid IgA responses which were
detected at the same time as intestinal IgA responses; however only 50% of the patients
demonstrated significant rises in Giardia-specific intestinal IgA antibodies and the presence of
these intestinal antibodies did not prevent Giardia infection in these patients.

5. Mechanisms of Protection

The role of humoral antibodies in protecting against Giardia infections is poorly
understood. Although it has been reported that anti-trophozoite antibodies including IgG, IgM
and IgA, are produced in the body in response to Giardia infections (Miotti et al., 1985; Goka et
al., 1986), it is not known whether these antibodies are immunologically relevant to clinical
disease. In an epidemiologic study in Bangladesh, Islam et al. (1983) reported that 45% of
asymptomatic mothers who had serum IgG antibodies against Giardia excreted the cysts in their
feces. Similar findings in Peru were reported by Gilman et al. (1985) who observed that 39% of
persons with serum IgG antibodies excreted the organisms. These observations suggest
development of a partial immunity which protects against the disease but not against infection.
In Bangladesh, young children especially lacked large amounts of antibodies in their serum and
V-25
-------
the first infections were usually symptomatic (Islam et al., 1983). However, conflicting evidence
was reported byMiotti et al. (1986) in Peruvian children.

Regarding the role of antibody in asymptomatic infections, antibody levels may either
predict a protective effect or serve as a marker of infection. High levels of serum antibody to
Giardia are observed in developing countries where endemic levels of Giardia infection are
high. These countries also have very low levels of symptomatic giardiasis (Janoff and Smith
1990). In animal studies, anti-Giardia secretory IgA levels are predictive of the clearance of
infection, and serum IgA levels appear to predict secretory IgA levels (Conley and Delacroix,
1987). This suggests that serum IgA levels should be predictive of clearance of infections.
Evidence for a relationship between serum IgA levels and infection in human populations is
provided by Nash (1987) from his experimental infection of 15 healthy volunteers. High levels
of infection were found with one strain and seroconversion for IgG, IgA and IgM among persons
with confirmed infections (Nash 1987). Rechallenge and infection resulted in less illness than
the initial challenge, but the number of individuals involved in the rechallenge was very small.

6. Summary of Evidence for Immunity

There is variability in the humoral response to Giardia infection. Some patients with
symptomatic infections fail to develop sufficiently high antibody levels for results to be called
positive. (Engelkirk and Pickering, 1990). In some patients, levels of anti-Giardia IgG
antibodies remain elevated long after the infection appears to have been eradicated. No sero-
diagnostic procedure has been reported that is capable of distinguishing asymptomatic from
symptomatic infection. Despite these shortcomings, serologic assays have proven to be of some
value in epidemiologic investigation of giardiasis (Miotti et al., 1986). The presence ofanti-
Giardia antibodies may indicate either past or present infection with Giardia, whereas the
presence of Giardia antigen in stool specimens indicates current infection. The specificity of
antigen detection assays maybe significantly improved by assays based on certain antigens
V-26
-------
groups (30/31 kDa, 57 kDa and high molecular weight antigens). Additional research is needed
to evaluate these markers under both controlled and field conditions.

F. Nonspecific Defenses Against Human Giardia

The severity of symptoms and the duration of Giardia infection are extremely variable.
This was reflected in a study of experimental human infections with a single Giardia strain (Nash
et al., 1987) where the severity and duration of giardiasis bore no apparent relationship to the
magnitude of the serum or secretory antibody responses (Nash et al., 1987). Gillin et al. (1990)
feel the variability in the severity of giardiasis symptoms may be due in part to trophozoite
interactions with non-immune elements of intestinal milieu.

The upper human small intestine colonized by Giardia is a complex and ever-changing
environment, which is normally inhabited by relatively types of few microbes (Gillin et al.,
1990). Giardia trophozoites in the intestinal lumen are exposed to fluctuating pH and
concentrations of bile, nutrients, digestive enzymes and their products. The intestinal tract has a
unique system of defenses which is less well understood than circulating defenses. Secretory
defenses may be produced locally, secreted into the duodenum, or ingested—as by breast-fed
babies. This system has both immune (secretory antibody) and non-immune components
(lysozyme, lactoferrin, Upases, many of which are found in both breast milk and intestinal fluid
(Gillin et al., 1990). Certain intestinal fluid factors might be toxic to the parasites, while other
factors such as mucus might protect them (Gillin et al., 1990)

Gillin et al., 1990 and others (Hernell et al., 1986; Rohrer et al., 1986) have shown that
Giardia trophozoites are killed in vitro by normal human milk and by human intestinal fluid (Das
et al., 1988). Products of lipolysis such as unsaturated free fatty acids (FFA) (Gillin et al., 1990;
Reiner et al., 1986; Hernell et al., 1986; Rohrer et al., 1986), lysophospholipids or
monoglycerides generated by Upases in human milk (Reiner et al., 1986) or intestinal fluid such
V-27
-------
as mucus and bile salts which bind lipolytic products protect trophozoites from killing by human
milk (Zenian and Gillin, 1987) and intestinal fluid (Das et al., 1988), as well as free fatty acids.
(Reiner et al., 1986). Bile also contains major growth factors for Giardia. This may help explain
how Giardia specifically colonizes the human small intestine.

Symptomatic and asymptomatic infections may result from the interaction of several
different aspects of the host-parasite biology of giardiasis. Intestinal conditions that are
conducive to growth of Giardia may vary between individuals or may be different in some
populations. Some patients with giardiasis and malabsorption show bacterial colonization of the
small bowel, which may predispose Giardia to express virulence factors (Janoff and Smith,
1990).

Because cellular immune responses to Giardia occur at the level of the intestinal mucosa,
it is important to appreciate the spectrum of mucosal inflammatory changes that may accompany
giardiasis. In the mouse, trophozoites colonize the proximal small intestine, attaching to the
mucosa preferentially near the bases of villi, at the edges ofPeyer's patch follicles, and less
frequently on villous tips. Although Giardia may betaken up by the specialized membranous 'M'
cells overlying Peyer's patches, trophozoites appear not to attach preferentially to these cells,
whose function is to transport luminal antigens and microorganisms to the underlying lymphoid
cells. Infrequently, trophozoites may invade the mucosal epithelium and penetrate the lamina
propria, possibly coming into direct contact with lymphoid cells. However, the frequency of
tissue invasion is unknown, because intestinal biopsies are not routinely performed in persons
with giardiasis.

Although the mucosal histology of the proximal small intestine of persons infected with
Giardia is frequently normal, varying degrees of histological change have been observed during
giardiasis. The histological changes include infiltration of polymorphonuclear leukocytes into
the epithelium and mononuclear leukocytes into the lamina propria, development of shortened
V-28
-------
villi, loss of the brush bonder, damage to epithelial cells, and an increase in epithelial cell mitosis.
Rarely, infection with Giardia may be associated with total villous atrophy, dense mononuclear
cell infiltration, and flattening of the epithelial cells, changes that resemble celiac sprue. These
findings indicate thatthe parasite is capable of eliciting an inflammatory cell response of varying
degree in infected subjects.

The accumulation of inflammatory cells in the small intestinal mucosa of infected persons
suggests that cellular responses may contribute to the host response to Giardia. The presence of
recurrent giardiasis in some patients with high liters of anti-trophozoite antibodies supports the
notion that circulating antibodies alone are not protective against the parasite. Investigations
using an experimental animal model of giardiasis provide evidence that lymphoid cells,
particularly T lymphocytes, contribute to host responses to the parasite. In athymic mice, which
are deficient in both circulating T lymphocytes and Peyer's patch helper T lymphocytes
(Heyworth et al., 1985), inoculation with G. muris cysts results in chronic infection with large
numbers of trophozoites. In contrast, immunocompetentmice clear the parasite and may develop
resistance to reinfection, implicating a role for T lymphocytes in the pathogenesis of chronic G.
muris infection.

G. Variation in Pathogenicity

Variability in the clinical response of the host to infection with Giardia may also be due,
in part, to differences in the pathogenicily of various strains of the parasite. Variation in the host
response to different strains has been identified in both humans (Nashet al., 1987) and animals
(Aggarwal and Nash, 1987) but strain differences only partly explain the variable host response
to the parasite. (Janoff and Smith, 1990). The variable pathogenicity among strains of Giardia
may also reflect differences in certain parasite antigens. In vitro demonstration of antigenic
variation, both spontaneous and in response to immunologic selection, suggests a possible
V-29
-------
mechanism by which Giardia survives chronically in the intestine, despite the presence of an
active immune response (Jannoff and Smith, 1990).

H. Chronic Conditions

Chronic patients often present with recurrent, persistent, brief episodes of loose, foul-
smelling stools which maybe yellowish and frothy in appearance, frequently accompanied by
distension, foul flatus, anorexia, nausea, and uneasiness in the epigastrium (1CAIR, 1984). With
chronic giardiasis, there is increased frequency of constipation and upper gastrointestinal tract
complaint (1CAIR, 1984). In some cases, these symptoms may persist for years, and in the
majority of cases, the parasite and symptoms disappear spontaneously (1CAIR, 1984). Among 65
giardiasis cases encountered in an urban private practice outpatient setting, the mean duration of
symptoms was reported to be 1.9 years, and of 38 (58%) patients who exhibited chronic
symptoms for six months or longer, the mean duration of symptoms was 3.3 years (1CAIR,
1984). Hoskins et al. (1CAIR, 1984) reported one patient who had a sprue-like syndrome with
IgA deficiency, vitamin B12 malabsorption, and recurrent diarrhea which persisted for 25 years.
The episodes consisted of 15 to 20 foul-smelling stools daily, anorexia, vomiting, nausea, and
abdominal cramps. Histologjcal examination revealed abnormal villus architecture.

There is usually no extra-intestinal invasion when Giardia trophozoites infect the small
intestine, but reactive arthritis may occur, and in severe giardiasis duodenum and jejunal mucosal
cells may be damaged (Benenson, 1995). Some symptomatic patients suffer from chronic
diarrhea, steatorrhea, and malabsorption of fats or of fat-soluble vitamins. A small amount of fat
in feces is not unusual, but when daily losses are greater than 7 grams, the condition is classified
as steatorrhea (Hall, 1994). Prolonged malabsorption of fat and its excretion in stools could lead
to a significant loss of potential dietary energy, especially as a result of chronic infection. This
will be of greater consequence for young children since they have greater requirements for energy
than adults and have small stomachs. The energy density of the diet and its efficient absorption
V-30
-------
are important, and persistent malabsorption of fat due to Giardia could lead to protein-energy
malnutrition (Hall, 1994). A loss of appetite is a commonly reported symptom, and nausea or
vomiting, abdominal cramps and bloating may occur (Hopkins and Juranek, 1991). All of these
symptoms are likely to contribute to reduced food intake (Hall, 1994). Studies have also shown
malabsorption of micro nutrients, especially vitamins A and B12, in infected persons (Hall, 1994).

The association between Giardia infection and ocular changes has previously been
described, but until recently, no large scale studies had been conducted. Corsi etal. (1998)
evaluated ocular manifestations in 141 Italian children with current and past giardiasis and 300
children without giardiasis. Salt and pepper retinal changes were diagnosed in 20% of the
children with giardiasis (mean age was 4.7 years) and in none of the children without giardiasis.
These findings suggest that asymptomatic, non-progressive retinal lesions may be common in
young children with giardiasis, and the risk did not seem to be related to the severity of infection,
its duration, or use of metronidazole but may reflect a genetic predisposition (Corsi et al., 1998).

Lactose intolerance is common during active infection and may persist following anti-
Giardia treatment, especially in ethnic groups predisposed to lactase deficiency (Wolfe, 1990).

II. Summary

Progress has been made in understanding the biology of Giardia. However, there is still
no adequate explanation for the wide clinical spectrum of giardiasis which ranges from
asymptomatic infection to acute self-limiting diarrhea to more persistent chronic diarrhea, which
sometimes fails to respond to therapy; the mechanisms by which Giardia produces diarrhea and
malabsorption and the key immunologic determinants for clearance of acute infection and
development of protective immunity also remain poorly understood.
V-31
-------
Symptomatic giardiasis, although less common than asymptomatic infection, occurs
frequently and results in diarrhea, flatulence, abdominal pain, weight loss, and vomiting. Severe
disease may result in malabsorption or growth retardation, but rarely death. Diarrheal symptoms
have been related to abnormal villus architecture. Chronic giardiasis appears to be infrequent,
but when it occurs, may persist for years. The precise mechanism of giardiasis responsible for its
pathology and symptomatology in humans remains unknown at present.

As with all diarrheas, fluid replacement is an important aspect of treatment; anti-giardial
drugs are also important in the management of giardiasis. Chemotherapeutic agents used for
treatment of giardiasis include metronidazole, tinidazole, quinacrine, furazoli done, albendazole,
and ornidazole. Various doses and treatment periods are recommended for each drug. The drugs
may have different effectiveness in their ability to clear Giardia; drug resistance and relapses
may occur, and the drugs have side-effects that should be considered. Paromomycin has been
used to treat giardiasis in pregnant women, but the cure rate may be low.

Data on the nature of human immune response to giardiasis are somewhat limited, but
there are indications that both humoral and cellular responses are present. Most subjects infected
with Giardia produce detectable levels of anti-parasite antibodies. However, the role of specific
antibody to Giardia in determining the host's clinical response to infection has not been
delineated. Is the presence of specific antibody more frequently associated with symptomatic
Giardia infections or asymptomatic infections (Janoff and Smith, 1990)?

When Giardia organisms enter the small intestine, local factors such as bile salts,
intestinal mucus, or the presence of other organisms enhance or inhibit their initial growth.
Intestinal enzymes, which may induce the expression of surface lectins that mediate mucosal
adherence, facilitate colonization (Lev et al., 1986). Once infected, the individual may become
ill, depending on the virulence of the strain or the presence of pre-existing immunity. In the
absence of prior immunity, interaction of trophozoites with intestinal macrophages may initiate
V-32
-------
an immune response. This nonspecific interaction may eliminate the parasite via cytotoxic
mechanisms, or macrophages may process trophozoite antigens leading to the induction of a
specific antibody response. These antibodies may be directly cytotoxic for the parasite or
promote antibody-dependent cytotoxicity with granulocytes, and possibly monocytes. This
integrated immune response results in clearance of the organism and resolution of symptoms.
However, this sequence of events does not explain the chronic, asymptomatic excretion of the
parasite observed in recurrently exposed persons.

Although one species of Giardia is believed to infect humans, the epidemiology of
giardiasis is complicated by an apparent genetic heterogeneity in this species. Differences in
virulence, pathogenicity, infectivity, growth, drug sensitivity, and antigenicity have been
reported. In endemic areas where extensive heterogeneity exists, mixed infections with more
than one genotype may occur.
V-33
-------
VI.
GIARDIA RISK ASSESSMENT
A. Risk Assessment Paradigms

Risk assessment is the qualitative or quantitative characterization and estimation of
potential adverse health effects associated with exposure of individuals or populations to hazards
(materials or situations, physical, chemical and or microbial agents) and is a component of risk
analysis. Risk analysis also includes risk management and risk communication (Table VI-1).
This section will address risk assessment and to some extent risk management.
Table VI-1 Components ofRisk Analysis
Risk Assessment
Risk Management
Risk Communication
The qualitative or quantitative characterization and estimation ofpotential
adverse health effects associated with exposure of individuals or populations to
hazards (materials or situations; physical, chemical and or microbial agents).
The process for controlling risks, weighing alternatives, selecting appropriate
action, taking into account risk assessment, values, engineering, economics, legal
and political issues.
The communication of risks to managers, stakeholders, public officials, and the
public, includes public perception and ability to exchange scientific information.
Source: Hoppin, 1993

Before 1991, risk assessment methods following the National Academy paradigm (NRC,
1983) of hazard identification, dose-response, exposure assessment, and risk characterization
were only used on a limited scale for assessing risks of waterborne pathogens (Haas, 1983; Rose
et al., 1991b; Rose and Gerba, 1991; Regli et al., 1991). The National Academy of Sciences has
suggested that risk assessment and risk management be kept separate (NRC, 1983), but in reality
the integration of risk management and risk assessment is seen as a necessary requirement in the
development of a workable framework. Regulatory agencies are now attempting to develop the
best approach for undertaking and using microbial risk assessment for policies that will improve
water quality, food safety and public health. The EPA (1989a) first used risk assessment based
VI-1
-------
on dose-response models for the development of the Surface Water Treatment Rule (SWTR).
The National Institute of Public Health and Environmental Protection in the Netherlands has also
used formal risk assessment procedures for waterborne microorganisms (Teunis et al., 1994).

A national committee established by the EPA in 1995 developed a framework for
pathogen risk assessment describing how to conduct a risk assessment, the type of data needed,
and the available information (ILSI et al., 1996). The framework attempts to integrate risk
management and microbial risk assessment. The committee was composed of a multi-
disciplinary group of scientists from the fields of epidemiology, medicine, microbiology, water
treatment, food safety, chemical risk assessment and public policy. The initial step in the
framework is problem formulation. Articulated in this step is the information needed by
managers to make decisions including the types of risks and priorities that are to be addressed.

The risk assessment itself is defined by an analysis involving the characterization of both
exposure and health. This leads to an estimate of risk and risk management options. The
analysis phase of the risk assessment considers:
1. human health effects (symptomatic and asymptomatic infection, severity and duration
of illness, medical care and hospitalization, mortality, host immune status, susceptible
populations);
2. dose-response modeling based on clinical and epidemiological data;
3. exposure analysis (vehicle of infection, amount and route of exposure and whether it is
acute or recurring, demographics and other characteristics of persons exposed);
4. occurrence assessment (analytical methods, quantitative measures of the pathogen in
the vehicle of exposure and its frequency, spatial and temporal variation, regrowth, die-off, and
transport).

Haas (1983) first estimated quantitative risks associated with microbiologically contaminated
drinking water using dose-response data from exposure of humans in clinical experiments. Several
VI-2
-------
mathematical models were evaluated to determine the model that best described the probability of
infection from existing dose-response data. For waterborne viruses, Haas (1983) found that a beta-
Poisson model best described the probability of infection, and this model was used to estimate
annual and lifetime risks for infection, clinical disease, and mortality associated with hypothetical
levels of viruses in drinking water.

Rose et al. (1991b) have used an exponential model to evaluate risks of Giardia infection
from estimated exposures to Giardia in drinking water. Drinking water exposures were obtained
from survey data describing the occurrence of Giardia in polluted and pristine water sources and
considering average removals and inactivation of cysts with various types of water treatment. The
same approach was used in the development of the SWTR where performance-based standards for
the control of Giardia were developed to meet the EPA's recommended public health goal of no
more than one Giardia infection per 10,000 persons from drinking water exposures (U.S. EPA,
1989a). The EPA felt that this goal could be maintained by achieving 99.9% reductions of Giardia
cysts through filtration and disinfection in all water systems.

B. Health Effects

The anticipated health effects are defined by the specific microbial agent, host characteristics,
the spectrum of symptoms, and pathology associated with the infection. The types of clinical
outcomes may include asymptomatic infection, acute disease, chronic disease, hospitalization, or
death, and outcomes may differ for sensitive subgroups. The pathogenicity and virulence of the
microorganism itself is of interest as well as the full spectrum of human disease which it can cause.
The host's response to the microorganism in regard to immunity is also of concern. Information
from endemic epidemiological studies, epidemic disease or outbreak investigations, case-studies,
hospitalization studies and other clinical is needed to complete this step in the risk assessment.
VI-3
-------
The transmission of disease and the vehicle of infection is often microbial specific (i.e.
rabies, malaria, influenza), but this is not the case with Giardia, which is transmitted by the fecal-
oral route. Water is only one of several modes of transmission.

The human health effects and epidemiology of giardiasis are described in detail in
Chapters III and V. Information about the prevalence of infection and the importance of
contaminated drinking water is found in Chapters in and V. An examination of the waterborne
outbreak and prevalence data can provide information about the relative importance of
waterborne transmission of Giardia compared to other fecal-oral pathogens that can also be
transmitted by contaminated water. Based on these data, Giardia is a waterborne pathogen of
primary importance in the United States. Giardia is the most frequently identified etiologic
agent causing waterborne outbreaks in public water systems in the United States, and it is the
most frequently identified parasite in national surveys of stool specimens where its prevalence
ranges from 4.0% up to 12% depending on the year and state. Outbreaks summarized in Chapter
III, Section G (Table III-7) show that Giardia is also transmitted by accidental ingestion during
swimming and other water recreational activities. Foodborne outbreaks of giardiasis have also
occurred but are much less frequently reported than waterborne outbreaks. Other important
transmission routes and risk factors include person to person transmission, travel to developing
countries, homosexual activities, and day-care center use.

The quantitative description of the health effects include the types, severity, and duration
of the illness. Illness factors are summarized in Table VI-2, and further information about them
can be found in more detail in other chapters of this document. Associated health care costs and
costs associated with days lost from work have been quantified, but these data are limited. In 18
reported waterborne outbreaks of giardiasis, hospitalization data were reported — 60 persons
were hospitalized among a total of 13,239 cases for a 0.5% ratio. Using estimates of 4,600
hospitalized cases or2 hospitalizations per 100,000 persons per year (Lengerich et al., 1994) and
the computed hospitalization ratio (0.5%), as many as 593,400 cases of giardiasis are estimated
VI-4
-------
to occur annually. Bennett et al. (1987) has suggested that 60% of giardiasis in the United States
is waterborne and estimated that 120,000 cases of waterborne giardiasis may occur each year,
however, this estimate is not based on epidemiological data. Mortality associated with Giardia
infections is rare (Bennett et al., 1987). Evidence in both animals and humans suggests that the
immunocompromised may be at a greater risk of acquiring a chronic giardiasis infection with
chronic diarrhea (Chapter V). An immune response is found after infection and may confer
protection; epidemiological studies (Isaac-Renton, 1994; Isaac-Renton et al, 1994 ) suggest that
immunity for Giardia may last for five years.
Table VI-2 Selected Health Factors Considered in Assessing Waterborne Giardia Risks
Health Effect
Symptoms
Asymptomatic infection
Waterborne outbreaks in
U.S.
Prevalence of infection
Prevalence of waterborne
illness
Immune status
Sensitive populations
Severity
Costs associated with
illness
Quantitative Assessment
Diarrhea with loose, foul-smelling stools that are greasy, frothy or bulky;
abdominal cramps, bloating, nausea, decreased appetite; malaise and
weight loss in the majority of patients.
Illness lasts in untreated individuals on average about 1 week with
infection lasting 6 weeks.
Chronic diarrhea can last an average of 1 .9 years.
Chronic outcomes include failure to thrive, urticaria, and reactive arthritis.
25% to 75%
Associated with 32% of all drinking water outbreaks, from 1971 to 1994,
and 70% of all recreational outbreaks from 1991 to 1994 relative to other
known etiologic agents
Detection of the cyst varies world-wide between 2 and 5% in diarrheal
stools in industrialized countries.
Found in 4% to 12% of all diarrheic stools examined for parasites in U.S.
593,400 cases annually based on hospitalized cases and hospitalization
ratio. 60% of cases may be waterborne; estimated 120,000-356,000
waterborne cases annually but estimate not from epidemiological studies.
Immunity clears infection and may provide some protection up to 5 years.
10% of population with IgA deficiencies may suffer chronic infections.
Immunocompromised greater sensitivity to chronic outcomes , however
this has not been quantified.
Annually 2 hospitalizations/100,000. Hospitalization ratio = 0.5%.
Mortality = 0.0001%
One week of illness estimated to cost $250 to $500. This does not include
treatment costs for those with chronic cases.
VI-5
-------
Methods for diagnosis
Available, routinely in use only when requested that specimen be sent to
ova and parasite laboratory. Can be difficult to diagnose (3 stool
submissions), chronic infections detected through intestinal biopsies.
C. Dose-Response Modeling

Dose-response studies provide information which can be used to mathematically
characterize the relationship between the administered dose and the probability of infection or
disease in an exposed population. Natural routes of exposure are used in clinical studies; that is,
the direct ingestion by volunteers in a controlled experiment. Both disease and infection are
measured. Epidemiological studies can also provide similar information, but the dose or
exposure may not be well known. Methods used to determine the number of microorganisms in
a given dose are those routinely used in the laboratory (e.g., direct microscopic counts of Giardia
cysts similar to those used to detect the cysts in environment samples). However, for Giardia
studies, this means non-viable cysts viewed microscopically will be included in estimating the
administered dose.

One of the more controversial areas surrounding microbial modeling is the assumption
that a single organism (one Giardia cyst) can initiate a human infection. The early literature
suggested that many microorganisms were needed to act cooperatively to overcome host
defenses in order to initiate an infection in humans (Blaser and Newman, 1982). More recent
data support the "independent-action" or "single-organism" hypothesis, which is based partially
on a phenomenon observed in laboratory studies - that given proper conditions a single
bacterium, virus, or protozoan can reproduce to cause infection (Rubin, 1987). Although each
microorganism alone is capable of initiating the infection, infection may still require exposure to
more than one organism. More must be ingested because of the relatively small probability that
a single microorganism will successfully evade host defenses (Rubin, 1987). This is analogous
to another biological phenomenon, that of spermatozoa and fertilization (Rubin, 1987).
However, once a single cyst passes the host defenses, it is presumed to be able to colonize and
VI-6
-------
infect the host. Immunity at the cellular and humoral level may play a critical role in
determination of which individuals may develop infection and more severe disease.
The Giardia risk assessment model used by Rose et al. (1991b) to assess waterborne risks
was based on limited dose-response studies in the 1950s (1C AIR, 1984) which of course were not
done with modeling in mind. The Giardia risk model is described in terms of the probability of
infection
where r is the fraction of microorganisms that are ingested which survive to initiate infection
(this is organism specific) and N is the daily exposure assuming consumption of 2 liters of
drinking water per day. The parameter for Giardia was r=0.0198 (0.009798-0.03582, 95%
confidence interval). The assumption of 2 liters per day is conservative, as it is an overestimate
of water consumption.
Table VI-3. Dose-Response Assessment for Human and Animal Hosts (Haas et al., 1998)
Infectious Agent
G. lam blia
G. lam blia
G. lam blia
G. lam blia
G. muris
Host
Humans
Gerbils
Musk rats
Humans
Mice
Value of r
0.0199
0.0019
0.000004
0.0698
0.56
Probability of Infection*
2.0xlO"2
2.0xlO"3
2.7xlO"6
9.3xlO"2
4.3X10'1
*Probablity of infection for exposure to 1 organism except 9.3x10" where probability of infection is for
exposure to 1 glass of water (an estimated 2 cysts). During one Giardia outbreak, a dose-response
relationship between glasses of water consumed per day and infection suggested that each glass of water
may have contained 2 cysts.
Table VI-3 shows risk estimates for mathematical models developed in a variety of
species and hosts. These data indicate that dose-response data for the muskrat, mouse, or gerbil
VI-7
-------
are not appropriate for use in estimating human risks. For G. lamblia, the animal models
predicted a lower risk than did the human dose-response data; for G. muris however, the risk
predicted was greater. Schaefer et al. (1991) found that infective dose levels of G. lamblia for
the Mongolian gerbil were much higher than those found for human volunteers and for G. muris
in the murine model. The importance of host specificity and the danger in extrapolation of
animal data to humans for infectivity and dose-response modeling is supported by the
comparisons in Table VI-3.

D. Exposure assessment

In risk assessment, exposure assessment is an attempt to determine the size and nature of
the population exposed and the route, levels and distribution of the microorganisms and the
duration of the exposure. The description of exposure includes not only the occurrence of
Giardia in water but how often the microorganisms are found. Exposure assessment depends on
adequate methods for recovery, detection, quantification, sensitivity, specificity, virulence and
viability, as well as studies and models addressing transport and fate through the environment.
Often the amount of contaminant in the medium associated with the direct exposure (i.e. drinking
water or food) is not known and must be estimated from other data bases. Therefore, knowledge
is needed about the ecology of these microorganisms, sources, transport and fate including
inactivation rates and survival or resistance to environmental factors (temperatures, humidity,
sunlight etc) and movement through soil, air and water. Finally, because the current methods for
monitoring microorganisms in environmental samples often do not have the necessary sensitivity
to detect actual levels in treated drinking water, raw water must be analyzed, and additional data
are needed on the inactivation and removal of microorganisms through treatment processes.
These data can then be used to estimate levels in treated drinking water.

Water quality monitoring data are needed, and the analytical methods used to obtain these
data will greatly influence the estimate of exposure. In the case of Giardia, all cysts detected are
-------
assumed to be viable and of the type that infect humans, and this overestimates the risk.
However, the risk may also be underestimated due to the inefficiency in recovery and detection
of cysts. Studies by Clancy et al. (1994) found that the greatest problem was specificity as
opposed to sensitivity and that the level of cysts may be underestimated by almost ten-fold. In
order to meet the current safely goal of no more than one Giardia infection per 10,000 persons
from drinking water exposures, the analytical method must be capable of detecting one cyst in
150,000 L. Because detection of cysts at this level is difficult with the current methodology,
source water levels must be monitored and exposures estimated based on presumed water
treatment reductions. The many factors that influence the exposure estimate for Giardia are
described in Chapter El and summarized in Table VI-4. A summary of the water occurrence data
is shown in Table VI-5.
Table VI-4. Exposure Factors associated with Assessment of Waterborne Giardia Risks
Exposure Factor
Transmission
Environmental
sources
Survival
potential
Regrowth
Occurrence in
raw water
supplies
Summary Information
Fecal-oral, waterborne (drinking and recreational) transmissions appears to be more
important than other routes; suggested to be responsible for 60% of all cases.
Levels found in:
Sewage (treated wastewater, secondary effluent) - range of 0.2 to 130 cysts/L & average
of 0.88 cysts/L; discharges of 1 MGD would put 3 million cysts into the waters each
day) (Rose et al., 1991b).
Animals - prevalence in cattle 29%, sheep 38%, pigs 9% and horses 20% (Chap. IV).
Storm waters - average 1100/L (Gibson et al., 1998)
Time for 90% inactivation is temperature dependent. Data suggest 1 month is required
to achieve 90% reductions at ambient temperatures (Chapter III). Adequate survival
model for temperatures not available. Travel time should be delineated from source to
exposure.
None
Levels ranging from 0.005 to 44/L
Range of Averages: 0.002to 2.2 /L (Table IH-1)
VI-9
-------
Resistance to
water treatment
Environmental
transport
Availability of
methods
Removals by sand filtration ranged from 1.1 to 5.1 Iog10'slow sand filtration removals
ranged from 1.2 to 4 Iog10; diatomaceous earth removals ranged from 2 to 3 Iog10;
microfiltration removals ranged from 6 to 7 Iog10 ( Table VII-1).
Coliform bacteria inadequate to evaluate cyst inactivation (45% of all giardiasis
waterborne outbreaks where coliform data were reported had no coliform detection).
Turbidity and particles appear to be appropriate for evaluating filtration effectiveness but
not for occasional spike of contamination or when coagulation not optimized.
Inactivation by disinfection can be estimated from Ct (Table VII-2). 90% to 99%
reductions depending on disinfectant, application and water characteristics, applied
throughout in U.S. facilities.
No transport models available.
Methods for detection are available (10 to 20% recoveries); variability in assessing
levels; no methods for identification of environmental sources (animal versus human), or
viability (however 11% viable by animal infectivity) (Chapter VII).
In full-scale seeded experiments, conventional filtration treatment used in drinking water
reduced cysts by an average of 3.3 Iog10. Direct filtration of these organisms showed slightly
improved performance reducing and cysts by an average of 3.9 Iog10 (Nieminski & Ongerth,
1995). Removal efficiencies are expressed as either percent removal (e.g., 99%) or more
typically in terms of logarithmic (base 10) removal of cysts. For example, a 1 Iog10 removal
indicates a 90 percent reduction in densities; 2 Iog10 removal means that 99 percent of cysts are
removed; 3 Iog10 removal means that 99.9 percent are removed.

However, monitoring data from actual water plants show that these reductions probably
overestimate the effectiveness of the filtration barrier. Data from two surveys of finished
drinking water was used to estimate Giardia risks: LeChevallier, et al. (1991b) found 17% of
treated water samples to be for Giardia (cysts ranged from 0.2 to 64 cysts/1 OOL) and
LeChevallier andNorton (1995) found Giardia in 4.6% of the samples (an average of 2.6
cysts/100 L and a range of 0.98 -9.0 cysts/100 L).
Table VI-5. The Occurrence of Giardia in Various Waters
Type of Water
Percent
Range of Cyst Levels (Cysts per Liter)
VI-10
-------
Untreated Wastewater
Activated Sludge Effluent
Filtered Secondary Effluent
Surface Water
Groundwater
Treated Drinking Water
Combined Sewer Overflows
100
83
75
45
6
17
100
642-3,375
0.14-23
<0.01-0.2
O.02-44
0.1-120
0.29-64
90-2,830
Source: Rose et al., 1991b; LeChevallier etal., 1991a, b; LeChevallier and Norton, 1995; Hancock et al, 1998;
Gibson et al., 1998.
E. Risk Characterization

Risk characterization is an integration of the previously described three steps in order to
estimate the magnitude of the public health problem and to understand the variability and uncertainty
of the hazard. This encompasses the spectrum of health outcomes, the confidence limits surrounding
the dose-response model, the distribution of the occurrence of the microorganism, and the
distribution of exposure. The occurrence and exposure can be further delineated by distributions
surrounding the method recovery and survival (water treatment reduction) distributions.

The estimates of daily risk of infection associated with the averages and ranges of
Giardia cysts found in drinking water are shown in Table VI-6. Annual risks of Giardia
infection from drinking water, including asymptomatic infections, averaged approximately 20 x
10~4 (20 infections per 10,000 people annually) and were as high as 250 x 10"4 (250 infections per
10,000 people annually. These annual risk estimates are presented as point estimates without
confidence limits and do not account for Giardia speciation and viability or analytical sensitivity
and specificity. Although they have these limitations, they do suggest that the annual risk of
infection due to current levels of Giardia in treated drinking water may be greater than the
recommended annual risk of Giardia infection that drinking water systems should attempt to
maintain. Point estimates of computed risk are 10 to 100 times the recommended risk level of no
more than one Giardia infection per 10,000 persons from drinking water exposures.
VI-11
-------
Table VI-6. Risk Estimates for Waterborne Giardia Infection
Exposure
Average (1991 survey)
Range(1991 survey)
Average(1995 survey)
Range(1995 survey)
Cysts Levels (Cysts per Liter)
0.04
0.003 to 0.64
0.026
0.0098 to 0.09
Daily Risk of Infection from 2 L/day
1.8x 10"3
0.1 to 25 x 10"3
l.Ox 10"3
0.39 to 3.6 x 10"3
*LeChevallier etal., 1991b; LeChevallierand Norton, 1995.
It is difficult to ascertain from Giardia surveillance statistics the accuracy that these risks
levels represent because most infections would be asymptomatic and persons that do develop
symptoms are not likely to seek medical care and undergo laboratory diagnosis. Estimates of
endemic waterborne infection risks may be able to be obtained from epidemiological studies
using serological techniques, but such studies have not yet been conducted.

Teunis et al. (1997) recently completed a comprehensive risk assessment of both
Cryptosporidium and Giardia using monte carlo analysis and the distributions rather than single
estimates for the following parameters: levels of oocysts and cysts (average <1/1000 L),
analytical method recovery effectiveness (<2%), viability of the recovered oocysts (30%) and
cysts (15%), removal of protozoa during water treatment based on Clostridium spores (2.8 Iog10),
the daily consumption of tap water (0.15 L/day), and dose-response r values (Cryptosporidium =
0.00467, Giardia = 0.01982). The cumulative estimate for an annual risk of waterborne infection
ranged from 10"5 to 10"4 for Cryptosporidium or Giardia and from 10"4 to 10"3 from exposure to
both organisms. The data used to develop the parameters utilized by Teunis et al. (1997) were
specific to the Netherlands with exception of the viability and the dose-response models, and
using a similar approach in other geographical areas should result indifferent annual risks
worldwide based the occurrence of protozoa in water, water treatment practices, and water
consumption in the area evaluated. For example, in the United States a higher estimate for the
VI-12
-------
annual risk of waterborne Giardia infection is expected because of the higher occurrence and
exposure to these protozoa in drinking water. A comparison of the estimated risks of waterborne
Giardia infection from the Netherlands and the United States computed using the different
mathematical models shows that risks in the United States are higher (e.g., 200 to 2500 times
greater). This may be due to both higher drinking water exposures and limitations of the model
used to compute the risk estimates (e.g., lack of consideration of analytical recoveries and
viability in the model).

A marriage of risk assessment methods with epidemiological models that describe the
transmission of disease through a population has been suggested as an appropriate approach for
examining population risks (Eisenberg et al., 1996). This would take into account factors such as
incubation time (time from exposure to infection and illness), immunity (protective as well as
impaired) and secondary non-waterborne transmission within the population, in addition to water
exposure and the dose-response relationship. However, by considering these factors, the
mathematical models can require as many as 13 model parameters with accompanying
information about their distributions, and much of this information is currently not well
understood. Therefore many assumptions must be made in order to use these models. The
simplicity of the Rose et al. (1991b) model make it easy to use, but it also has limitations due to
assumptions made. The complexities of the models proposed Eisenberg et al. (1996) may make
them more difficult to use, and the more assumptions needed for the additional parameters in the
model may make the results more difficult to evaluate. However, these models do attempt to
include all of the relevant information that may be needed to estimate waterborne risks and used
in combination with a sensitivity analysis should help identify the parameters that may have the
greatest effect on the risk estimate. A comparison of waterborne Giardia infection risks for the
United States computed from each of the models (Rose et al., 1991b; Eisenberg et al., 1996;
Teunis et al., 1997) would help risk managers in their interpretation of the risk assessments.

F. Risk Management and Federal Regulations
VI-13
-------
As technological advancements increase in microbial detection and more knowledge
becomes available about risks of multiple microbial exposures in drinking water, directors of
municipalities and health officials will be called upon to ensure that the public is adequately
protected from epidemic and endemic waterbome disease. Risk assessment is a tool that can be
used by health officials and water utility managers to help interpret water quality surveys, assist
in defining the adequacy of drinking water treatment, and communicate possible health risks.

Currently, the EPA regulations (U.S. EPA, 1989a, b; 1994a, b, c; 1996) that specifically
address Giardia in potable supplies are found within the SWTR, ICR, and the proposed
Enhanced Surface Water Treatment Rule (ESWTR). The Clean Water Act also regulates point
and non-point discharges that may contain fecal material into receiving waters, with fishable and
swimmable being the goals, but there are no specific limits for Giardia in discharges to
recreational areas or to receiving waters that are used as sources for drinking water. Coliform
bacteria are used as the indicator of fecal contamination of discharges to source and recreational
waters. This approach has a serious limitation in that coliform bacteria are much more
susceptible to disinfection than cysts. Effluent discharges may contain low levels of coliform
bacteria because they are disinfected, however, levels of Giardia may be high. Watershed
programs for protection of drinking water sources and emphasis in the future on the safety of
recreational sites will mean that occurrence databases from point and non-point discharges will
need to be better defined.

The SWTR requires filtration and disinfection of all surface water supplies and
groundwater directly impacted by surface water. Because monitoring for waterborne pathogens
was deemed impractical, the rule developed a series of treatment requirements for surface and
groundwater under the influence of surface water. These requirements specified a minimum
removal or inactivation of 3 Iog10 for Giardia and 4 Iog10 for viruses in water treatment provided;
water treatment levels could be increased for poor source water quality. The rule also lowered
VI-14
-------
the acceptable limit for turbidity in finished drinking water from a monthly average of 1.0 NTU
to a level not to exceed 0.5 NTU in 95% of 4-hour measurements. The requirements for meeting
these limits went into effect in June 29, 1993. Based on the current risk analysis, most water
utilities would be required to provide more than the minimum 3 Iog10 treatment for Giardia
specified by the SWTR in order to meet the safety goal of no more than one Giardia infection
per 10,000 persons from drinking water exposures.

With the development of regulations to limit the levels of disinfectants and disinfectant
by-products (D/DBP), the EPA recognized the possibility that efforts to reduce DBF levels could
inadvertently increase the risk from microbial agents. To ensure that implementation of the
D/DBP Rule (U.S. EPAb), did not increase microbial risk, the EPA considered it necessary to
examine the health and economic implications of various approaches to DBF regulation, to
compare microbial risk from Giardia infection to cancer risk for several DBF control scenarios,
and to review the adequacy of the existing SWTR.

The data generated by the ICR will be used to help formulate the final draft of the
ESWTR and it is likely that the final rule will be subject to many modifications. Water treatment
plant performance may require greater reliability and removal of Giardia, and alternative
treatments such as membrane filtration and use of ozone may be considered. Cost and benefit
assessments will be required, and risk assessments will be used to evaluate the benefit side of
producing better drinking water quality. The rule will likely be developed to protect against
waterborne risks of Cryptosporidium. Cryptosporidiosis has a high mortality in
immunocompromised persons and the infection is not treatable. Also, Cryptosporidium is
smaller than Giardia and may be more difficult to removal by conventional filtration and much
more resistant to disinfection. However, Giardia cyst levels in sewage and wastewaters are very
high, the waterborne exposure to Giardia is likely to be greater, and the dose-response model
suggests it is more infectious. Thus, it is important to consider waterborne risk assessments for
both Giardia and Cryptosporidium.
VI-15
-------
VI-16
-------
VH. ANALYSIS AND TREATMENT OF GIARDIA

In order to make informed decisions about drinking water treatment and regulations for
Giardia, reliable data are required on the occurrence and distribution of the organism in the
environment and in human and animal populations. There is also a need to assure that drinking
water treatment regimes for removing or inactivating cysts are adequate. Interpreting available
data on the occurrence, distribution and treatment effectiveness for Giardia in water supplies
requires an understanding of the capabilities and the limitations of the methods used to collect
the data. Classical cultural techniques for microorganisms, such as are used for bacteria and
many of the enteric viruses, are not applicable to detecting, identifying and enumerating Giardia.
This Chapter will review methods for detecting cysts in water, clinical diagnosis procedures, and
effectiveness of drinking water treatment.

A. Analysis in Water

1. Detection and Identification Methods

The previous Giardia Criteria Document (1CAIR, 1984) reviewed an available method
for detecting cysts in water. Significant steps in the method included sample collection and
sample processing. Sample collection was accomplished by filtering large volumes of water
(about 100 gallons or 380 L were recommended as the sample size) through microporous filters
constructed of tough fibers (Orion or polypropylene). Sample processing involved procedures
(filter extraction, sedimentation, centrifugation, flotation) to recover the cysts from the filter and
separate them from interfering debris and other organisms. The final step in sample processing
was detection and identification using microscopic examination of iodine stained material. The
development of a cultural technique was not considered likely, and although a viability
determination might be accomplished by animal feeding, this could only be done by laboratories
equipped with the necessary isolation and animal handling facilities.
VII-1
-------
Jakubowski (1984) reported the results of a Giardia method workshop convened to
address issues such as a "reference method", cyst identification, viability and suitable
applications for the methods. The reference method (also known as the zinc sulfate flotation
method or the zinc sulfate/Lugol's iodine method) was essentially as described in ICAIR (1984),
but many of the steps within the method had not been verified under controlled laboratory
conditions and were based on experience and professional judgement of the participants. For
Giardia cyst identification, the consensus was that "Suspect organisms should possess the proper
size and shape and at least two internal characteristics (nuclei, axostylar rods, median bodies)."
A minority opinion was that only one internal characteristic should be required if the organism
was of the correct size and shape, but the majority felt thatthis could result in misidentification.
Suspect objects that met all requirements for identification except the possession of two internal
characteristics should be reported as "Giardia cyst-like". It was recommended that the
examination of sample concentrates and identification of cysts be conducted by competent, well-
trained individuals who have a demonstrated proficiency for diagnosing intestinal protozoa.

For determining cyst viability, the workshop participants noted that animal testing was
problematical, excystation testing of cysts detected in water samples was questionable, and dye
staining methods (at that time, eosin exclusion) did not correlate with in vitro excystation.
Participants considered the current methods for viability suitable only for use in outbreak
investigations or special research applications but not for routine monitoring or surveillance. In
addition, a consensus recommendation was that viability determinations were not necessary for
water samples collected in outbreak investigations where epidemiological data implicated the
drinking water. Since G. lamblia cysts were only obtainable at the time from human stools,
workshop participants encouraged the identification of a suitable laboratory animal host and
encouraged the investigation of immunological methods for detection and identification.

Sauch (1985) developed an indirect immunofluorescence assay for detecting and
identifying Giardia cysts in water samples. The assay used a polyclonal primary antibody
VII-2
-------
developed in rabbits against whole G. lamblia cysts obtained from an asymptomatic donor. The
secondary antibody was goat anti-rabbit IgG conjugated to fluorescein isothiocyanate (FITC).
The assay was evaluated against a variety of human and non-human (beaver, dog, mice, vole,
muskrat, gerbil) source cysts and bright apple green fluorescence was produced by cysts from all
of these sources when the specimens were illuminated with UV light. Sauch (1985) indicated
that the assay allowed rapid location of cysts even in samples that contained significant quantities
of other microorganisms and debris. When fluorescing cysts of the appropriate size and shape
(presumptive Giardid) were located, phase-contrast microscopy was used to examine them for
internal characteristics in order to confirm their identity.

In applying the assay to water samples which were collected in outbreak investigations
and processed using the reference method (except that Percoll-sucrose was used for flotation
instead of just sucrose or zinc sulfate), Sauch (1985) was able to detect and confirm Giardia
cysts in 58% (14/24) of raw water samples and in 27% (6/22) of distribution system samples.
She pointed out that failure to confirm an object did not mean it was not Giardia since internal
morphology often cannot be discerned in some known Giardia cysts even though they fluoresce
brightly. In the water samples she examined, 13% (3/24) of the raw water and 40% (9/22) of the
distribution system samples contained objects that fluoresced and were the right size and shape
but could not be confirmed by internal morphology. As recommended by the consensus
reference method, these water samples were reported to contain cyst-like objects.

In developing a method to recover Cryptosporidium from water, Musial et al. (1987)
based the sample collection procedure on the Giardia reference method procedure using 1 jim
nominal porosity polypropylene filter cartridges. Although they did not evaluate their method
with Giardia cysts, they did find that using 0.1% Tween 80 as the eluent instead of just water
resulted in higher recoveries of Cryptosporidium oocysts. They also found that adding Tween 80
and sodium dodecyl sulfate detergents to the centrifuged pellet from the filter elution and
VII-3
-------
sonicating resulted in the highest recovery of oocysts in sediment-containing preparations.
Sonication did not help to recover particles from the filter material itself.

Rose et al. (1988b) evaluated a method similar to that of Musial et al. (1987) for detecting
both Giardia and Cryptosporidium in water. They tested six gradient solutions for recovering
Giardia cysts. Although recoveries ranged from a low of 40% for zinc sulfate (specific gravity,
1.18) to 77% for Percoll-sucrose (specific gravity, 1.09) there was no statistically significant
difference in these results. They noted that potassium citrate (specific gravity, 1.16) and 4/5
Sheather's (specific gravity 1.24) gave cleaner preparations when used with environmental
samples. They did not find any statistically significant increase in recoveries with the use of
sonication. For detection in a direct fluorescence assay, they used 5 jim porosity cellulose nitrate
membrane filters and monoclonal antibodies conjugated to FITC.

Stibbs et al. (1988) investigated mouse monoclonal antibodies developed against cysts of
G. muris, G. simoni and G. lamblia. The G muris antibodies reacted with homologous cysts and
with rat source cysts but not with cysts from beaver, dog, human, muskrat or vole. The G. simoni
antibodies reacted only with rat and cow cysts. The G. lamblia antibodies reacted with all
human, beaver, dog and rat source cysts but not with G. muris or cysts from muskrat or vole.
The authors suggested that systematic differences occur in cyst surface antigens and that it might
be possible to develop a monoclonal antibody-based typing scheme for strain and animal source
identification.

Januschka et al. (1988) tested the reactivity of antibodies produced against cysts of G.
lamblia and G. muris with cysts of Spironucleus muris. Spironucleus sp. share morphological
characteristics including axonemes, nuclei and peritrophic space with Giardia sp. And they are
both in the same family, Hexamitadae. Januschka et al. (1988) found that all of the antibodies
tested reacted with both Giardia and Spironucleus cysts, indicating that these antibodies were not
genus specific. Spironucleus cysts are somewhat smaller than Giardia cysts, averaging about 5 x
VII-4
-------
8 jim, but they share morphological characteristics and are widely distributed in animals, and the
authors advised caution in confusing Spironucleus and Giardia cysts when examining water
samples.

Payment et al. (1989) reported studies on development of a method to concentrate G.
lamblia cysts, Legionellapneumophila, Clostridium perfringens, human enteric viruses and
coliphages from large volumes of water using a single filtration. They used electronegative
fiberglass cartridge filters of 3 jim and 1 |im nominal porosity in series to collect conditioned
water (pH 3.5, 0.001 M aluminum chloride) at a flow rate of 10-40 L/min. Filters were eluted by
backflushing with 1.5% beef extract, pH 9.75 containing 0.5% Tween 80. After adjusting thepH
to 7.2 with 1.2 N HC1, the eluate was centrifuged and the pellet was further processed for
Giardia by re-suspending a portion on a discontinuous sucrose density gradient and centrifuging.
The interfaces were collected and examined by phase microscopy or immunofluorescence. In
seeded studies with about 10,000 cysts, the efficiency of retention by the filters was >99%, the
recovery after elution was 71% and the final recovery after flotation and centrifugation was 52%.

As the immunofluorescence technique for cyst detection became more widely used, the
number and variety of antibodies available increased. Rose et al. (1989) evaluated four
antibodies for detecting Giardia in environmental samples. In preliminary trials, one of the
antibodies that was developed against trophozoites would not react with G. lamblia or G. muris
and was not studied further. The remaining three antibodies had been developed against cysts
and were of the IgG type; two used an indirect fluorescence procedure and one used a direct
procedure, and two were monoclonal antibodies and one was polyclonal. The monoclonal direct
antibody did not react with G. muris cysts but the other two did, and the polyclonal antibody
detected more G. muris cysts than did the monoclonal. G. lamblia cyst counts with one of the
monoclonal antibodies and with the polyclonal antibody were evaluated in three storage media
(deionized water, 2.5% potassium dichromate and 3.7% formaldehyde). Cyst counts were 29%
higher when they were stained with the polyclonal antibody. There were no significant
VII-5
-------
differences in cyst counts among the three storage media. After 20 to 22 weeks storage at 4°C,
there was no significant association between cyst counts and either the formaldehyde or
potassium dichromate storage media. However, cyst counts were reduced 67% in the deionized
water and those cysts remaining had diminished fluorescence and distorted shape.

Laboratory studies revealed differences among the antibodies with regard to species
identification and number of cysts detected, but when applied to environmental wastewater
samples, there were no significant differences in cyst counts even though one of the monoclonals
consistently produced higher maximum counts. The precision of determining cyst counts on
membrane filters was evaluated with four replicates from three cyst preparations stained with a
monoclonal antibody. The coefficient of variation ranged from 7.89 to 11.62 and averaged 9.22
at cyst densities between 24 and 70/filter. Rose et al. (1989) also tested the reaction of antibodies
with cysts exposed for 20 minutes at room temperature to sodium hypochlorite at concentrations
of 5 to 20,000 mg/L. At exposures up to 50 mg/L, the counts remained stable with all three
antibodies. With the monoclonal antibodies, Giardia cyst counts were decreased by 35% to 57%
at exposures of 500 mg/L, and no cysts were detected at exposures higher than 5,000 mg/L. With
the polyclonal antibody, cyst counts remained stable at sodium hypochlorite concentrations up to
500 mg/L and decreased 66% to 83% at higher exposures. Sauch and Berman (1991) also
investigated the effect of chlorine, at levels that would be expected to encompass the range used
in drinking waters, on the morphology and fluorescence of cysts. They used exposure
temperatures of 5° and 15°C, residual chlorine lev els from 1 to 11 mg/L, and an exposure time of
48 hours. The majority of the cysts lost their internal morphology but were still detectable by
immunofluorescence. They concluded that if chlorinated (>1 mg/L) water samples are to be
examined for cysts, the majority of them will lose morphology necessary for identification if they
are collected, transported, and stored for more than four hours without neutralizing the chlorine.

Concerning species identification as previously indicated, differences were observed in
specificity of the antibodies for G. lamblia and G. muris. However, Rose et al. (1989) concluded
VII-6
-------
that until questions of species definition in Giardia are resolved "...it is unlikely that any single
antibody will identify cysts in water which only pose a health risk to humans." The authors
indicated that while immunofluorescence has increased the ability to detect cysts in water,
underestimation of densities in samples is a significant problem. This is due to a combination of
factors including the length of time the organisms have been in the environment, the type of
antibody used and the amount of debris in the samples. They also expressed concern about the
lack of information on the viability of detected organisms.

LeChevallier et al. (1990) compared the zinc sulfate flotation/Lugol's iodine reference
method to the immunofluorescence (IFA)/Percoll-sucrose method for detecting Giardia cysts.
They used a monoclonal antibody (Meridian Diagnostics, Cincinnati, OH) in an indirect assay
adapted from Sauch (1985). When tap water concentrates were spiked with 550 to 5,145 Giardia
cysts, recovery by the zinc sulfate method averaged 5.9% (range = 1.6% to 13.5%). With the
IFA method, recovery averaged 74.1% (range = 37.1% to 92.7%). In comparing both methods
for the detection of Giardia cysts in natural raw water samples from rivers, they found on
average about 12 times (range = 1.5 to >40) more cysts with the IFA method. They concluded
that the Meridian antibodies were convenient and reliable regents for the water utility laboratory
and that the IFA/Percoll-sucrose method is more efficient in detecting parasites in water than the
zinc sulfate flotation method.

Abbaszadegan et al. (1991) developed and evaluated a gene probe method for detecting
Giardia cysts in water samples. Their method used a 265-base pair (bp) cDNA probe from the
small subunit rRNA of G. lamblia. They evaluated 6 protocols for extracting nucleic acids from
cysts and found they could detect approximately 1 to 5 cysts (determined by dilution) by a dot
blot hybridization assay when they used glass beads and proteinase K to disruptthe cysts.
Specificity testing against 15 other microorganisms including parasites, bacteria and yeast
produced negative results with the probe. The assay was specific at the genus level since it
detected Giardia from mice, birds, beaver and humans, although with different sensitivities. The
VII-7
-------
authors suggested that specificity at the genus level might be more desirable for evaluating
drinking water treatment efficiency, and that species level specificity would be desirable for
evaluating health hazards. A limited number of water and sewage samples were examined by the
IF A method and by the gene probe method. Positive results were obtained by both methods in
two samples and in a third, only by the IF A method. An advantage of the IF A method is that it
quantitatively detects cysts. Gene probe methods can be made semi-quantitative by using quantal
assays, but intact cysts might not be detected.

Mahbubani et al. (1991) reported the development of PCR amplification techniques for
detecting Giardia cysts. DNA was released from cysts by heating and a 171-bp region of the
giardin gene was amplified by PCR. The giardin gene was selected as a target because this is a
potentially unique protein found in the ventral disk of Giardia (Crossley et al, 1985). The
amplified product was detected with gel electrophoresis, ethidium bromide staining and Southern
hybridization of radio-labeled gene probes. The specificity was tested against 19 Giardia isolates
(including human beaver, muskrat, cat, mouse and bird strains) and 24 isolates of other
microorganisms (including protozoa, bacteria, algae and yeast). A 171-bp PCR product was only
produced by the Giardia isolates. Using single cysts recovered with a micromanipulator, these
investigators demonstrated that the assay was sensitive enough to detect one cyst.

Mahbubani et al. (1992) also developed a PCR and gene probe method for differentiating
G. duodenalis from other Giardia species. Human, muskrat, beaver, bird, cat and mouse source
cysts all produced a 171-bp PCR amplification product with primers GGL639-658 and GGR789-
809. Confirmation of specificity for all Giardia species tested and none of the non-Giardia
species was obtained through Southern blot analysis with a radio-labeled probe (GGP751-756).
When cysts were spiked into 100 mL river water and potable water samples at densities of less
than 1/mL, they were detectable by PCR. They were also detectable by PCR when spiked in 4-
liter river water samples when they were present at equal to or greater than 0.25 cysts/mL. PCR
VII- £
-------
did not detect cysts in concentrates from 400 L river samples even when spiked at cyst densities
of 105/sample.

In 1992, the ASTM published a proposed test method for detecting and enumerating
Giardia cysts and Cryptosporidium oocysts in low turbidity ground and surface waters. Low
turbidity was defined as equal to or less than 1 nephelometric turbidity unit (NTU). The method
was designated 'proposed' because the precision and accuracy of the method had not yet been
determined in accordance with ASTM requirements. The method involves collecting a
recommended minimum sample volume of 100 gal (380 L) by filtration, at flow rates up to 4
L/min, through a 1 m nominal porosity polypropylene cartridge microporous filter. The filter is
eluted with a detergent solution which is then centrifuged to recover particulates. The pelleted
material is subjected to a flotation purification procedure using Percoll-sucrose solution (specific
gravity 1.1). The purified material is applied to a membrane filter in a concentration that will
result in a depth not exceeding a monolayer. After staining the material with indirect fluorescent
antibody assay reagents, the membrane is mounted on a glass slide and subsequently examined
using a combination of epifluorescence and phase or differential interference contrast
microscopy. Results are reported as presumptive or confirmed cysts or oocysts/100 L using
specific criteria for immunofluorescence, size, shape and internal morphological characteristics
to define the categories. Known interferences with the detection and identification of cysts and
oocysts were listed as including turbidity due to inorganic and organic material, autofluorescing
or nonspecific fluorescing organisms and debris, chemical compounds and freezing samples,
eluates, or concentrates. The method will not identify the species of protozoa detected or the
species of the host of origin nor will it provide information on the viability or infectivity of the
cysts. The proposed method was intended for use with raw and treated drinking waters, and it
was suggested the method may be useful in identifying contamination sources and in evaluating
water treatment effectiveness.
VII-9
-------
Turbidity continued to pose challenging problems for investigators attempting to
determine cyst levels in many source water supplies, primarily rivers. Bifulco and Schaefer
(1993) explored the application of an immunomagnetic procedure for selectively recovering cysts
from water sample concentrates. They developed an indirect antibody-magnetite method using a
mouse IgG anti-Giardia antibody as the primary antibody. The secondary labeling reagent was
an anti-mouse IgG antibody-coated colloidal magnetite particle with an average size of 40 nm.
They chose the small size magnetite particles to prevent potential interference with microscopic
visualization of the cysts. Cysts labeled with the magnetite reagent were concentrated using
high-gradient magnetic separation. The mean recovery of cysts from water samples with various
turbidities (70 to 6,400 NTU) was 82%; 90% of cysts seeded into buffer were recovered. Cyst
recoveries were highest at turbidities below 600 NTU. Bifulco and Schaefer (1993) felt this
method had the potential for being linked with IFA detection methods since the cysts were
already coated with anti-Giardia antibody.

Erlandsen et al. (1994) proposed a molecular approach using fluorochrome-rDNA probes
to differentiate species and detect Giardia by in situ hybridization. Carboxymethylindocyamine
dyes were conjugated to oligomeric probes (17-22 mer) to the 16S-like rRNA of G. lamblia, G.
muris, and G. ardeae. Alternatively, the oligomeric probes were labeled by incorporating a
fluorescent marker (e.g., fluorescein) to the 5' end of the oligomer. They were able to
specifically identify G. lamblia and G. muris cysts in the same sample using specific rDNA
probes each conjugated to a fluorochrome producing a different color when examined with a
confocal microscope having a dual krypton-argon laser. Also, using a combination of the
fluorochrome-rDNA conjugates and antibodies to the cyst wall, they could identify G. lamblia
cysts in fecal samples and in samples from a sewage lagoon. The G. lamblia-specific probe did
not hybridize to G. duodenalis-type organisms from the parakeet, great blue heron, and the vole.
This latter finding led Erlandsen et al. (1994) to strongly suggest that Giardia from these animals
is most likely different from G. lamblia and that G. duodenalis does not have any continued
value as a species name for Giardia.
VII-10
-------
Clancy et al. (1994) indicated that the proposed ASTM (1992) IFA method was
considered to be the method of choice for detecting Giardia and Cryptosporidium for a planned
nationwide monitoring effort. They conducted a blinded survey of 12 laboratories performing
protozoa analyses with environmental samples, i.e., the laboratories were not informed of the
study so they should not have given special treatment to the samples. Known private,
government and university laboratories providing this type of analysis were contacted; sampling
equipment was obtained and evaluated; filters were spiked with known quantities of sediment,
cysts, oocysts, or algae; the filters were returned to the laboratories for analysis, and the results
from each laboratory were evaluated. Filters were spiked with about 387 formalin-preserved G.
lamblia cysts or with about 500 Oocystis minuta algal cells. Of the 12 laboratories, 11 submitted
reports, but only seven laboratories found Giardia cysts. One laboratory reported cysts present
without providing quantitative data; the remaining laboratories reported recoveries ranging from
less than 1% to 22%. Only three laboratories discussed the differences between presumptive and
confirmed cysts; a fourth mentioned that internal features were not found in the cysts. Of the
reports received for water samples containing algae, two laboratories incorrectly identified these
as Giardia cysts. Clancy et al. (1994) concluded that not all laboratories were strictly following
the ASTM analytical method, even though they said they were and that the majority of
laboratories needed to improve in one or more of the following areas: client response, quality of
sampling equipment, sampling directions, turnaround time for result reporting, quality of data
and report format.

In the 18th edition supplement to Standard Methods for the Examination of Water and
Wastewater (APHA-AWWA-WEF, 1994) and in the 19th edition (APHA-AWWA-WEF, 1995)
the previously described zinc sulfate flotation method was replaced with a method basically the
same as the ASTM (1992) IFA method. Significant differences in the method included reporting
requirements and a limit for the turbidity of water to be sampled. A limit of 1 NTU or less is
required by ASTM; no turbidity limit is in Standard Methods. Reporting of results differed from
VII-11
-------
the ASTM method in that the presumptive/confirmed categories were replaced with total
count/count with internal structure categories. In addition, the Standard Methods version was
published as a proposed method. It was noted that the method had been developed by a
consensus panel and not standardized and that modifications might be needed depending upon
local water quality conditions, equipment availability and analyst experience. The difficulty of
interpreting positive and negative findings was also discussed.

In 1994, the EPA proposed the Information Collection Rule (ICR) which required water
systems serving populations of 10,000 people and using either surface water or groundwater
under the influence of surface water to monitor for Giardia and Cryptosporidium (U.S. EPA,
1994a). Monitoring was to be conducted by the ASTM method as modified by expert workshops
and performance evaluation studies. The principal differences between the ASTM and ICR
methods included: no turbidity limitation for the ICR monitoring; use of differential interference
contrast (DIG) or Hoffman modulation optics for visualizing internal structures in cysts rather
than phase contrast microscopy as specified by ASTM; use of goat serum as a blocking agent to
minimize nonspecific immunofluorescence in the ICR method; stringent positive and negative
quality control requirements in the ICR monitoring rather than the less stringent, non-mandatory
ASTM recommendations; mandatory requirements for the use of filters to collect primary water
samples and immunofluorescence reagents; and differences in porosity of the assay membrane.
There are also differences in the reporting of results; ICR uses the total count terminology
whereas ASTM uses the presumptive/confirmed terminology. The final ICR method was
published along with the regulation (U.S. EPA, 1996).

In reporting the development and application of flow cytometric methods for detecting
cysts and oocysts in water, Vesey et al. (1994) cited deficiencies in the IFA method including the
inefficiency of the method, requirement for skilled analysts, tediousness of the examination due
to the frequent necessity for spending hours examining a single sample, and interferences from
non-target particles in the sample. The investigators compared direct microscopy of concentrates
VII-12
-------
to flow cytometry for detecting cysts and oocysts in sewage effluents, reservoir and river water
samples, and dam water and tank water samples. The sewage effluents were concentrated by
centrifugation; the reservoir and river water samples (100 L) were concentrated by tangential
flow filtration, and the dam water and tank water samples (10 L) were concentrated by calcium
carbonate flocculation. A direct antibody technique was used for the microscopy and the flow
cytometry. Two flow cytometers were used, either a Coulter EPICS V or a Coulter Elite (Coulter
Corp., Hialeah, FL). Preparations from environmental samples were stained with monoclonal
antibody in suspension and analyzed directly without any washing steps. All particles
demonstrating the fluorescence and light scatter characteristics of cysts and oocysts were
examined by epifluorescence microscopy after flow cytometry cell sorting (FCCS) onto glass
microscope slides. In six river and reservoir samples spiked with 77 Giardia cysts, recoveries by
direct microscopy ranged from 80% to 91%; by FCCS/microscopy recoveries ranged from 92%
to 115% with each of the six values being higher than that obtained by direct microscopy. In
seven unspiked samples from sewage effluents, dam water and tank water, cyst counts by direct
microscopy ranged from 1 to 420 (mean =75.7) whereas by FCCS/microscopy the range was 4
to 611 (mean = 121.0) with each sample being significantly higher by the latter technique. Vesey
et al. (1994) concluded that the FCCS technique simplified and improved analysis of water
samples for cysts and oocysts with high recoveries and without operator fatigue problems.

Ho et al. (1995) developed a technique for recovering Giardia cysts from river waters and
from other environmental waters that depended upon concentrating 1he samples by the
flocculation-Percoll/sucrose gradient method followed by immunofluorescence staining with
FITC labeled anti-Giardia monoclonal antibodies. The procedure had a cyst recovery rate of
61% compared to zero recovery for a membrane filter method and 4% for a method using
polypropylene filter cartridges.

LeChevallier et al. (1995) conducted tests to evaluate the sources of losses of Giardia
cysts and Cryptosporidium oocysts in the IF A method. For Cryptosporidium, major losses were
VII-13
-------
found to occur in the centrifugation and clarification (flotation) steps. For Giardia major losses
occurred only in the centrifugation step. With a swinging-bucket rotor, Giardia cyst losses
decreased from 25% to 11% when the relative centrifugal force (RCF) increased from 1,040 to
6,700 xg. With a fixed-angle rotor, cyst loss decreased from 40% to 24% when the RCF
increased from 800 to 5,000 xg.

Nieminski et al. (1995) evaluated two different approaches to the IFA method; one was
the ASTM procedure for low-turbidity water samples, and the other procedure involved sampling
less water (40 L)by filtration through a large diameter (142- or 293-mm) membrane filter (2.0
m porosity), followed by concentration on a Percoll-Percoll step-gradient. The goal was to find
the best method for evaluating the effectiveness of water treatment processes. In their laboratory,
the ASTM method resulted in a 12% average recovery of cysts while the membrane filter method
recovered 49%from water samples spiked with 1,000 cysts/16 L. Evaluation of spiked river
water, filtered water and flocculation basin water having turbidities of 5 NTU, 0.5 NTU and 20
NTU, respectively, indicated that both methods more efficiently recovered Giardia cysts
compared to Cryptosporidium oocysts and that the sampling step resulted in the highest loss of
both protozoa. The membrane method was less time-consuming and cheaper but did not allow
determining presumptive and confirmed cysts and oocysts. Therefore, the potential for falsely
identifying a cyst or oocyst would be increased when cross-reacting algae were present. The
investigators suggested investigation of a hybrid method that combined the most efficient steps
of both methods.

Rodgers et al. (1995) studied the potential for algal interference when using the IFA
method. They tested 54 algal species for cross-reactivity with the antibodies used in the assay.
Twenty-four species exhibited some degree of fluorescence with the ASTM antibody reagents
while two species, Navicula minima and Synechococcus elongatus, produced a bright apple green
fluorescence. Adding goat serum to the assay mixture blocked the fluorescence of most
nontarget organisms while decreasing background fluorescence on the membrane. The
VII-14
-------
fluorescence ofGiardia cysts and Cryptosporidium oocysts and the identification of internal
structures was not compromised by the addition of goat serum.

Fredericksen et al. (1995) compared the IF A membrane technique with a proposed well-
slide modification. The results indicated that the well-slide method was superior to the
membrane technique. Danielson et al. (1995) found FCCS to be more accurate and sensitive
than the IFA method, however, attempts to determine viability using dyes and flow cytometry
were inconclusive. They used a dual color dye system, FITC (green) and phycoerythrin (red), to
label cysts and indicated that this enhanced the sorting of the cysts.

Bielec et al. (1996) recently proposed a modified procedure for the collection and
recovery ofGiardia cysts from diverse water sources. The method involves the use of a
compact, battery-operated, portable field filtration apparatus. Cyst recoveries from seeded
samples ranged from 60% to 80% (mean = 66.1%). The design of the sampling unit also permits
spiking to allow the introduction of Giardia cysts during the filtration step which allows
developing estimates of recovery during the sampling procedure. Charles et al. (1995) described
an alternative bench-scale spiking procedure for use in the laboratory.

Shepherd and Wyn-Jones (1996) studied cartridge filtration, membrane filtration and
flocculation for simultaneous recovery of cysts and oocysts. They evaluated different procedures
to optimize the density of the organisms. In both tap water and river water, cyst recovery was
highest when organism suspensions were centrifuged at 5,000 xg as compared to 1,500 xg.
Also, an antibody staining technique where the cysts were stained in suspension was found
superior to a technique that used direct staining on a multiwell slide. The calcium carbonate
flocculation method gave the highest recoveries in seeded river water and tap water samples for
both cysts and oocysts. Examination of a variety of 142-mm membrane types and porosities
indicated that a cellulose nitrate filter with a porosity of 3.0 jim produced the best cyst
recoveries.
VII-15
-------
LeChevallier et al. (1997) presented results on the occurrence of cysts and oocysts in open
reservoirs and discussed various methodological considerations. They noted that a 100%
variation in recovery rates with the ICR method is not unexpected because of the number of steps
where losses of organisms can occur. Before examining water sample concentrates in their
laboratory, analysts examined positive control fecal samples comparing characteristics of
environmental isolates were compared with the control organisms. As a result, they indicated
that the definition of Giardia and Cryptosporidium is primarily an operational definition where
the identification of environmental isolates depends on agreement of the morphological
characteristics with those in positive controls. As a quality control measure, all analysts in their
laboratory were required to count the number of cysts and oocysts in a positive control on a
weekly basis. Variation of more than 30% from the mean of all counts was an action level that
triggered a review of microscopic procedures and a discussion of method technical details.

Examination of the results from the reservoir sampling revealed that, although sample
volumes of the influent and effluent waters were the same (1,579 L), the equivalent volumes for
each type of sample varied by about 250% (75 L for influent samples and 32 L for effluent
samples). The pellet volumes were not very different but the big difference in amount examined
was probably due to a difference in the nature of the material in the pellet. The effluent samples
had higher amounts of algal cells that could interfere with cyst and oocyst detection. While there
is controversy over whether internal morphological characteristics can be affected by sampling or
analytical procedures, the authors indicated that they have never observed this with spiked
samples examined in their laboratory. In reviewing recovery efficiencies, LeChevallier et al.
(1997) cited the EPA data from performance evaluation (PE) studies among 10 laboratories that
indicated a 30% recovery efficiency for Giardia cysts with a 77% coefficient of variation;
recovery in their laboratory was similar with 34.5% recovery and 73% coefficient of variation.
LeChevallier et al. (1997) also indicated that the EPA studies involved spiked filters whereas
recovery efficiencies in their laboratory included the sample collection step. Jakubowski et al.
(1996) summarized two EPA round-robin PE studies that indicated 25 to 44% mean recovery
VII-16
-------
efficiencies for Giardia with 58 to 93% coefficients of variation. The EPA also conducted a
field spiking/sampling study during 1995 at seven water utilities (ARCTECH 1996). These field
spiking and laboratory analyses for Giardia cysts showed recovery rates of 36.7% for raw waters,
and 26.4% for finished waters, similar to those in the PE studies.

Stewart et al. (1997) reported the development of a device to collect "first flush" water in
rugged terrain during storm events. Accessing these sites can be problematical during storm
events and the devices were constructed to function unattended. They consist of a 5 L
polyethylene contained fitted with an adjustable PVC valve. There are four 1/8-in openings in
the valve through which water enters. The samplers were installed on the upstream side of
existing creek weir structures. Each weir had a gauge that adjusted the sampler valve to a
predetermined flow. Samplers were retrieved within 24 hours of a storm event and returned to
the laboratory for processing. These samplers proved superior to 100 L filter samples or to 4 L
grab samples in terms of ease of use, positivity of samples, and levels of cysts detected.

Hoffman et al. (1997) compared IFA and FCCS for detecting cysts and oocysts in 262
fresh or archived (preserved and stored) samples. The FCCS procedure was not performed
concurrently with the IFA procedure but may have been carried out as long as 16 months after
the IFA results were obtained. With this set of samples they found little difference in detection
of cysts with either method: 116/262 samples were positive by IFA and 114/262 were positive by
FCCS. They did find a difference when they examined the results taking archive storage time
into consideration. For 132 samples that were analyzed by FCCS within 5 months of the IFA
analysis, 70 were positive for cysts with 43 detected by IFA and 55 detected by FCCS. This
difference was significant (p=0.04) but there was no statistically significant difference between
the methods with samples that had been stored for 1, 10 or 16 months. There was also no
difference in results for Giardia cysts with 26 samples purchased from a private laboratory and
analyzed simultaneously by IFA and FCCS.
VII-17
-------
lonas et al. (1997) developed a PCR method using random amplification of polymorphic
DNA that could differentiate G. muris from G. intestinalis. They developed primers specific for
G muris and suggested that they could be used in combination with primers for G. intestinalis to
determine if drinking water is safe for human consumption.

Rochelle et al. (1997) evaluated four pairs of PCR primers and probes previously reported
for Giardia while investigating the effects of primer annealing temperature, magnesium
concentration, specificity, sensitivity and PCR additives on the assays. The primers were
directed to heat-shock protein (HSP), giardin, and small-subunit rRNA genes, and reported
sensitivities ranged from <1 to one cyst. Only one primer pair was specific for G. lamblia and
the other three were specific at the genus level. After preliminary evaluation and due to
inefficient amplification, one primer pair (the one that targeted the small-subunit rRNA) was
eliminated from further consideration. A test of multiplex PCR for Giardia and
Cryptosporidium resulted in a sensitivity level of one cyst and one oocyst in purified preparations
and 50 cysts and oocysts in seeded water samples. The authors concluded that PCR has the
potential to be effective for detecting cysts and oocysts in water but that primer evaluation and
optimization was necessary for obtaining suitable sensitivity and specificity. The
Cryptosporidium primers with the greatest sensitivity and specificity were not compatible with
any of the Giardia primers for multiplex PCR. An ideal combination of sensitivity, specificity
and compatibility with multiplex PCR was not demonstrated by any of the available primers.

Kaucner and Stinear (1998) described a reverse transcription-PCR (RT-PCR) assay for
direct analysis of primary sample concentrates from large volume water samples to
simultaneously detect viable oocysts and cysts. The assay incorporated an internal positive
control (PC) to assess the efficiency of mRNA isolation and the potential for RT-PCR
inhibition. The method used the fiberglass filter sampling and elution procedure recommended
by Payment et al. (1989), selective capture of mRNA witholigo (dT)25 magnetic beads, and
detection of viable cysts by RT-PCR. The IPC hybridized to the magnetic beads and allowed
VII-18
-------
monitoring the process from mRNA capture through RT-PCR on a per sample basis. The intent
was to develop a multiplex PCR but sensitivity of detection decreased 10-fold in this system, so
separate PCRs were used for Giardia and Cryptosporidium. Also, these investigators initially
tried to use the HSP primers reported by Abbaszadegan et al. (1997) but found this RT-PCR to
be unreliable. Instead, they used the giardin mRNA primers indicated by Mahbubani et al.
(1991).

Kaucner and Stinear (1998) found positive signals in spiked 100 jiL packed pellet
concentrates from river and creek samples at all cyst levels tested (2 to 340 cysts/100 jiL). They
compared RT-PCR and an IFA technique for detecting cysts in 29 environmental water samples
ranging in volume from 20 to 1,500 L. Cysts were detected in24% of the samples by IFA and in
69% of the samples by RT-PCR. Cysts were detected in treated sewage effluents, river water and
creek water but not in treated drinking water. The authors considered the combination of a large
volume sampling method with RT-PCR to be a significant advance in protozoan pathogen
monitoring. They did indicate that a limitation of the method is that it is not quantitative. In
addition, the giardin primer set used was genus specific, not species specific.

2. Determination of Viability

Although one of the workshop recommendations reported by Jakubowski (1984)
indicated that viability determinations were not necessary in certain outbreak investigations, the
viability or infectivity of Giardia cysts found in drinking water at other times is important to
determining their public health significance and the need for regulations.

Hoff et al. (1985) compared animal infectivity and excystation for quantitatively
determining the viability of G. muris cysts. With their mouse model, they found that 1 to 15
cysts constituted an infectious dose. For cysts exposed to chlorine, they concluded that in vitro
VII-19
-------
excystation was an adequate indicator of G. muris infectivity and that it could be used to study
disinfectant effect on viability.

Schupp and Erlandsen (1987a) evaluated a method for determining viability based on
staining ofGiardia cysts with the fluorogenic dyes fluorescein diacetate (FDA) and PI Cysts
stained with FDA fluoresce green and those stained with PI fluoresce red when subjected to light
of the appropriate wavelength by epifluorescence microscopy. These investigators determined
that G. muris cysts that took up the PI would not cause infections in mice when they were
inoculated with either 5,000 or 50,000 cysts. However, cysts stained with FDA readily produced
infections in mice. They concluded that FDA-positive cysts are viable and that Pi-positive cysts
are nonviable as determined by animal infectivity.

Schupp and Erlandsen (1987b) also compared FDA/PI staining to DIG, phase, or bright
field microscopy for determining viability. Using G. muris cysts, they determined that cysts
incorporating FDA were morphologically identical and had (1) a clearly defined cyst wall, (2) a
distinct space between the cyst wall and the cytoplasm and (3) polar flagella. When examined
under DIG, the cytoplasm had a hyaline appearance that made detection of the internal
characteristics difficult. In PI stained cysts, the internal characteristics could be seen and a space
was not observed between the cyst wall and the cytoplasm. They indicated that further testing
was necessary in order to determine the reliability of these criteria for G. lamblia.

Schaefer (1988) reviewed methods used to determine Giardia cyst viability and
concluded that excystation was reliable and reproducible for G. muris cysts but not for G.
lamblia cysts. For both species ofGiardia, the factors found to promote excystation included
low pH, the presence of carbon dioxide, a temperature of about 37°C and a final neutralization at
pH 7.0. G. muris cysts did not require a maturation period before producing maximum
excystation, but G. lamblia did. Schaefer (1988) indicated that the fluorescent dye staining and
morphological criteria methods for determining viability were new and might require validation
VII-20
-------
but in a subsequent review, Schaefer (1990) concluded that fluorogenic dyes correlated well with
excystation in G. muris but that data on G. duodenalis were inconclusive. Schaefer (1988) also
noted disadvantages to animal infectivity for determining viability. Only some Giardia isolates
from humans will infect gerbils, and those that infect gerbils may not produce cysts requiring
necropsy to verify infection. Infectivity is dose dependent, and the failure to produce infection
with a given inoculum of cysts does not mean that the cysts were not alive or not infective for
humans. Other disadvantages include high cost and the need for maintaining approved
laboratory animal facilities.

Sauch et al. (1991) evaluated PI with G. lamblia cysts for compatibility with the
previously developed IF A test (Sauch, 1985). They found that the PI stain was compatible with
this method which uses a fluorescein label for detecting cysts in water samples. They also
studied PI as an indicator of viability in G. muris cysts exposed to heat, chlorine, chloramine and
a quaternary ammonium compound. There was a correlation between PI and excystation for
cysts inactivated with either heat or the quaternary ammonium compound. In both cases, the
percentage of cysts stained with PI tended to lag the percentage of intact cysts. This could
indicate that the cysts were injured or damaged to the point where they could not excyst while the
cyst wall was still intact and prevented entry of the PI. There was no correlation between
excystation and PI staining with chlorine or chloramine-exposed cysts. For the chlorine results,
the authors speculated that this strong oxidizing agent may have masked or destroyed the sites
within the organism needed for the PI to be detected. With chloramine, the authors suggested
that the lack of correlation may have been due to decreased permeability in the cyst wall from
exposure to this disinfectant. The authors stated that any tests proposed for determining the
viability of Giardia cysts should be evaluated with cysts inactivated by agents relevant to water
supply disinfection.

Mahbubani et al. (1991) developed a PCR test for distinguishing live from dead cysts
using giardin mRNA as the target. Cysts of G. muris and G. lamblia were killed by freezing,
VII-21
-------
heating and chloramine exposure. The PCR test was compared to excystation and animal
infectivity. G. muris cysts that had been killed by any of the three methods did not excyst and
did not produce infections in animals. A positive PCR signal for the 171-bp diagnostic region
was obtained indicating that the target DNA was preserved in dead cysts. However, when the
giardin mRNA was used as the target, a positive signal was produced in cysts killed by heat or
monochloramine but was not was produced in cysts after freezing. G. lamblia cysts produced
similar results. Live cysts that were induced to initiate excystation produced greater amounts of
RNA that could be measured spectrophotometrically, thus allowing dead cysts to be
distinguished from live cysts. There was no increase in RNA when cysts killed by any of the
three methods were induced to excyst.

Linquist (1995) described the use of a dual fluorochrome method forthe detection of
Giardia lamblia cysts and the determination of their viability. The use of the IF A method
involves light microscopy but requires switching the optical components from fluorescence to
DIG in order to verify the internal morphological characteristics for confirmation of the Giardia
cysts. Linquist (1995) reported on the use of a dye combination, FDA and Texas Red ®
hydrazide (TRH), with the FDA being taken up by viable cysts, and the TRH being actively
excluded. When each of these two dyes were used, together or separately, in combination with a
fluorescent antibody specific for G. lamblia cysts, the dual labeling permitted identification of
the internal morphological structures and good observations without the need to convert to DIG
optics. However, the procedure was accomplished with confocal microscopy, a technology
currently found in research laboratories and not in laboratories performing routine analyses.

As part of a study on the prevalence and characterization of Giardia cysts isolated from
Canadian drinking waters, Wallis et al. (1996) compared PI exclusion and gerbil infectivity as
indicators of viability. When tested by dye exclusion, the average viability of 167 drinking water
and sewage samples that were found positive for cysts was 24.6% (n=127) and 38.9% (n=40),
respectively. Gerbils were inoculated with cysts from 33 of these samples and 8 (24%) infections
VII-22
-------
were produced. The authors did not note any clear relationship between dose and infection.
Also, based on failure to infect gerbils with many samples containing viable cysts, they stated
that the similarity of viabilities by PI exclusion and gerbil infectivity was coincidental. There are
factors other than the number and viability of inoculated cysts that determine the probability of
infection with some isolates. They found that 11/194 (57%) isolates from water samples infected
gerbils even when no Giardia cysts were detected in the sample. Wallis et al. (1996) concluded
that cysts isolated from water and sewage have viabilities much less than 100% and the gerbil
model is not always accurate because some strains do not infect gerbils. They also indicated that
the PI technique is much easier to use and works well with the indirect IFA when PI is added
during the last 5 minutes of incubation with the primary antibody. Satisfactory results were not
obtained with direct antibody kits or with using PI after the FITC conjugate has been added.

Abbaszadegan et al. (1997) developed a PCR viability assay for Giardia that was based
on heat-shock protein (HSP). They used an hsp-70-\ike gene that coded for HSP and was
specific to G. lamblia to develop the assay. The basis of the assay is that a cell cannot produce
HSP mRNA when exposed to elevated temperatures unless it is viable. They also developed a
presence/absence test for Giardia based on amplification of total DNA or RNA from lysed cysts.
The PCR viability procedure involved heating cyst-containing samples to 42°C for 15 min;
freeze-thaw the preparation through 5 cycles; use magnetic beads to isolate the mRNA; heat the
sample to denature proteins, then go through PCR amplification and detection of a 163-bp
product. Giardia DNA or the corresponding RNA from lysed cysts was detectable by the HSP
primer set. The test was sensitive at levels of one cyst but is not quantitative and does not
determine viability. Using the same primers, an amplification product was detectable in heat-
shock induced cyst at a sensitivity level of 10 cysts. The product was not produced in live cysts
that had not been heat induced, nor in dead cysts killed by heating. The authors suggested that
PCR was an attractive method for detecting and determining the viability of Giardia cysts, and
that the method had advantages of speed, cost, sensitivity and specificity.
VII-23
-------
As indicated previously, Kaucner and Stinear (1998) used RT-PCR to detect viable
Giardia cysts in environmental water samples. The primer set they used was that of Mahbubani
et al. (1991) which detected giardin mRNA. Mahbubani et al. (1991) had used this primer set to
differentiate live from dead cysts by inducing excystation in the cysts and determining whether
an increase had occurred in the mRNA signal after induction. Mahbubani et al. (1991) found that
heat killed cysts and cysts killed by monochloramine produced a positive signal for giardin
mRNA, so just detecting it would not indicate viability. However, in contrast, Kaucner and
Stinear (1998) found that heat inactivated cysts did not produce or maintain mRNA. They
suggested that the discrepancy mightbe related to experimental differences such as the time
between heat treatment and mRNA extraction. They did not address the issue of a positive signal
produced by monochloramine treated cysts.

B. Detection in Biological Samples

The following methods commonly used for diagnosis of giardiasis were summarized in
the previous Giardia Criteria Document (ICAIR, 1984): (1) direct microscopic examination of
fecal smears for cysts or trophozoites, (2) identification of motile trophozoites in specimens from
the upper intestines, (3) intestinal biopsies or (4) gastrointestinal radiology using barium.
However, it was pointed out that duodenal aspirations, biopsies and radiological techniques were
not suitable for routine screening of human or animal populations and there were no clinically
accepted serological methods for diagnosis of giardiasis.

Wolfe (1984) indicated that stool examination should be the initial diagnostic procedure
of choice and that it should be performed by an experienced technician who was using
appropriate collection and laboratory procedures. When formed stools are examined, the cyst
form is more likely to be present in fresh, unpreserved specimens. Loose or watery stools should
be immediately examined in a wet smear of a just-passed specimen or preserved using one of
several formalin-containing preservatives or commercially available kits. If parasites are
VII-24
-------
numerous, direct smears may be adequate for examination but concentration of the sample using
formol-ether or zinc sulfate flotation may be necessary for light infections. Lugol's iodine may
be used to stain wet smear preparations.

Riggs et al. (1983) were the first to report development of an immunofluorescence assay
for Giardia cysts in fecal specimens. They developed high titer sera to G. lamblia cysts and
conjugated that to fluorescein isothiocyanate. There was no cross-reactivity when tested against
the following protozoan parasites: lodamoeba butschlii, Dientamoeba fragilis, Entamoeba
histolytica, Entamoeba coll and Endolimax nana. In stool samples, the conjugate did cross-react
with cysts of Chilomastix mesnili which fluoresced as intensely as the Giardia cysts. However,
the authors indicated that the smaller Chilomastix cysts could be easily differentiated.

Ungar et al. (1984) developed an ELISA specifically for the detection of G. lamblia in
fecal samples. ELISA methods are rapid, accurate, inexpensive, and require a lower degree of
technical training than microscopical analyses of fixed slides. The assay was capable of
detecting between 37 to 375 cultured trophozoites and 12.5 to 125 cysts purified from human
stool. The sensitivity of the test was found to be 92%; ELK A produced 36 positives in 39
specimens known to be positive for Giardia by direct examination of formalin-fixed stool
specimens or by biopsy. A high specificity (98%) was also found; only 3 of 128 stool specimens
from patients without demonstrable giardiasis were positive.

Erlandsen et al. (1990c) compared two methods for determining the prevalence of
Giardia in beaver and muskrat populations, the detection of cysts in fecal samples of kill-trapped
animals and examination of mucosal scrapings from live-trapped animals. Examining the
intestinal contents resulted in significantly higher prevalence rates of infection in these animals.
Based on detection of cysts in fecal specimens, the prevalence of infection in muskrats and
beavers was 36.6% and 9.2%, respectively. By examining intestinal contents for trophozoites,
the prevalence of infection in muskrats and beavers increased to 95.9% and 13.7%, respectively.
VII-25
-------
The results indicated the superiority of the mucosal scraping technique for determining the
prevalence of infection in these animal populations. Erlandsen et al. (1990d) also studied the
recovery of cysts from animal tissues and fecal samples subjected to cycles of freezing and
thawing. They found that cysts might not be detected in specimens of this type if only bright
field microscopy was used for the examination. However, the cysts were detectable by
immunofluorescence even though the cyst walls had been distorted.

Wolfe (1992) again reviewed procedures for diagnosis of giardiasis and described a
number of possible problems that could prevent identification of Giardia in clinical specimens.
These included: (1) medication administration - could cause organism distortion, low numbers
of organisms or mask their presence; (2) diagnostic procedures - may cause organism distortion;
(3) radiographic examination — barium may cause organism distortion and mask their presence;
(4) intermittent shedding — numbers of organisms in stool fluctuate widely; (5) specimen
collection — trophozoites may disintegrate if fixative not used; (6) laboratory techniques — using
concentration techniques and preparing permanent stained fecal smears are mandatory; (7)
specimen examination — requires trained personnel, and (8) interpretation of results — failure to
obtain additional specimen types when needed, e.g., duodenal biopsies. He also summarized the
sensitivity and specificity of 11 rapid detection immunoassays for Giardia including enzyme
immunoassays (EIA), counterimmunoelectrophoresis, and indirect immunofluorescence. One of
these was directed at detecting cysts; the others all targeted specific antigens. The reported
sensitivity of these assays ranged from 88% to 98%; specificity ranged from 87% to 100%.

Xiao and Herd (1993) developed and evaluated a quantitative direct fluorescent antibody
(DFA) assay. For the purpose of uniformity in this section, DFA will be used in place of FA,
DIF or other descriptors that the authors might have used in their published articles to indicate an
immunofluorescence assay employing a direct antibody. Their test had a theoretical sensitivity of
100 Giardia cysts/g of feces and they compared recovery rates with this test to those obtained
with sucrose gradient and zinc sulfate flotation methods. Their procedure used the commercially
VII-26
-------
available MERIFLUOR® Cryptosporidium/Giardia antibody kit (Meridian Diagnostics,
Cincinnati, OH). In calf feces seeded with 1,000, 10,000 and 100,000 cysts/g the recovery rates
were 76.4%, 96.9% and 89.6%, respectively. In contrast, the recovery rates using sucrose
gradient flotation were 20.5%, 51.2% and 42.9%, respectively. Zinc sulfate flotation detected
36.4% of infections when the cyst level was equal to or less than 1,000 cysts/g.

The DFA assay was compared to routine stains (saline and iodine-stained wet
preparations; chlorazol black E and Kinyoun acid-fast stained permanent smears) for detecting
Giardia in 2,696 fresh human fecal specimens examined in a hospital clinical laboratory (Alles et
al., 1995). These investigators also used the MERIFLUOR® antibody kit. The DFA assay
produced a significantly increased detect!on rate; the sensitivity by routine examination was 66%
compared to 99% DFA. A limitation of the DFA assay is the requirement of an epifluorescence
microscope that many hospital clinical laboratories might not have.

Stazzone et al. (1996) used MERIFLUOR® antibody staining to retrospectively evaluate
human stool specimens processed by a laboratory in Egypt. The laboratory was reporting
abnormally low identification rates for Giardia and Cryptosporidium using conventional dye
staining methods (trichrome for Giardia cysts). Antibody staining was shown to be almost 3
times more sensitive for detecting cysts than conventional trichrome staining. There was no
significant difference in the results whether fresh or frozen stools (stored at equal to or less than
70°C) were used. This latter finding led the authors to suggest that the immunofluorescence
method could be used for retrospective quality control on frozen specimens when fresh
specimens were not available.

Karanis et al. (1996b)used a different antibody kit, Giardia-CEL® (Cellabs P.L., Sydney,
Australia), to compare phase contrast microscopy and the DFA assay for detecting Giardia cysts
in cattle and wild rodent feces. Of 40 cattle fecal specimens examined, 31(78%) were positive by
DFA and only 17 (55%) were positive by phase contrast microscopy. All of the detected cysts
VII-27
-------
were identified as belonging to the G. duodenalis group. Giardia was found in 103/216 (48%)
fecal specimens collected from wild rodents. In 97 (94%)of the samples, DFA was positive;
phase microscopy detected 57 (55%) positives. Both methods were positive for 51 samples.
Only 19/103 wild rodent specimens found positive contained cysts belonging to the G.
duodenalis group. The authors concluded that, while DFA was more sensitive for detecting
cysts, phase microscopy was required to differentiate at the species level.

Hassan et al. (1995) detected G. lamblia antigens in stool samples before and after patient
treatment with a double antibody sandwich ELISA method. The sensitivity of the assay was 98%
with a high specificity. The authors found a direct relationship between the levels of antigens in
stool samples and the numbers of cysts detected.

Dixon et al. (1997) compared conventional microscopy, immunofluorescence microscopy
and flow cytometry for detecting Giardia cysts in beaver fecal samples. They examined 94
formalin preserved beaver fecal specimens that had been concentrated by sucrose flotation. For
the conventional microscopy, one drop of floated suspension was placed on a slide and examined
by scanning at 400X magnification. Immunofluorescence microscopy was performed using 20
jiL of floated fecal suspension air dried on a microscope slide, stained with Giardia-CEL®
antibody (Cellabs, distributed by Wellmark Diagnostics Ltd., Guelph, Ontario) and scanned at
400X magnification. Flow cytometry used 200 jiL of floated fecal suspension stained with the
same antibody and processed in a FACSCAN(Becton Dickinson, Mississauga, Ontario) flow
cytometer. Of the 94 specimens, 7 were positive by conventional microscopy, 9 by
immunofluorescence microscopy and 14 by flow cytometry. Those found positive by flow
cytometry were verified by cell sorting and examination under immunofluorescence microscopy.
The authors concluded that flow cytometry was more rapid than the other techniques and allowed
for screening a larger number of samples within a given time period. They suggested using this
technique for prevalence studies in animals and for screening human clinical specimens during
outbreak situations.
VII-28
-------
Mank et al. (1997) compared microscopy and EIA for examining human stools for
Giardia. Their results indicated that EIA techniques may be about as sensitive using a single
stool sample as microscopy is when using two sequential stool samples.

Garcia and Shimizu (1997) evaluated commercial diagnostic kits for detecting Giardia:
two were DFA kits and 5 were EIA kits. One of the DFA kits (MERIFLUOR®) was used as the
reference method. The kits were tested against 100 formalin preserved stool specimens that were
found positive using the reference method and an additional 50 specimens that were negative
with that method. The other DFA kit was the TechLab Giardia/Crypto IF kit (TechLab,
Blacksburg, VA). The sensitivity and specificity of the TechLab kit was 100% when compared
to the reference method. However, the authors indicated that the fluorescence intensity of the
TechLab reagents was one level lower than that of the reference method. In the evaluation of the
EIA test, 92/100 specimens were positive by all 5 kits, and with the negative samples, all 5 kits
correctly identified 50/50. All kits performed within expected levels as stated in the
manufacturer's documentation for sensitivity and specificity. The authors suggested that the
decision on which kit to use was up to each laboratory and would depend on factors other than
sensitivity and specificity, such as cost or ease of use.

Aldeen et al. (1998) also examined EIA kits for detecting Giardia in fecal specimens. In
addition to sensitivity and specificity, they compared other factors such as ease of use, cost and
processing time. They evaluated nine kits but only four of these were still available at the time
the article was published. They found sensitivities ranging from 88% to 100% and specificities
of 99% to 100%. Total hands-on time to run one specimen ranged from 1 to 2.25 minutes. The
cost per test generally ranged between five and six dollars. The authors suggested replacing three
exams for ova and parasites with a single EIA when the likely clinical diagnosis is giardiasis.
VII-29
-------
Ortega and Adam (1997) prepared a state-of-the-art clinical article in which they
concluded that the use of serodiagnosis (i.e., using patient serum samples instead of stool
samples) to detect giardiasis was not useful because the available methods cannot distinguish
between present and previous infections. Adam (1991) had earlier concluded that serologic
testing was useful for epidemiological studies but not for diagnosis of individual patients because
of sensitivity and specificity problems. Ortega and Adam (1997) also indicated that fecal
specimens have inhibitors that reduce the sensitivity of PCR for diagnosing giardiasis. Marshall
et al. (1997), in a review of waterborne protozoan pathogens, suggested that nucleic acid-based
detection methods are challenging due to difficulty in lysing the cysts. However, they indicated
that the large amount of DNA and the presence of inhibitory substances in clinical specimens
were more important. There has been only one published report on the application of PCR to
human samples and that investigator experienced some false-negative and false-positive results
when comparing PCR to a microscopic reference method (Weiss, 1995).

Rosales-Borjas et al. (1998) studied the secretory immune response during natural
Giardia infections in humans by examining saliva samples. For antigen, they used the
membrane-rich protein fraction of cultured trophozoites. They were able to demonstrate a
secretory IgA response in the saliva of infected patients that was not present in healthy
individuals. They indicated their results were from only 24 patients, but if substantiated, they
could have significance for the isolation of important protective or diagnostic G. lamblia
antigens.

C. Water Treatment Practices

Information from laboratory, pilot plant, and full scale treatment plant studies
demonstrate that Giardia cysts can be effectively removed or inactivated by commonly used
water filtration technologies and disinfectants. A combination of water filtration and disinfection
operated under optimum conditions can protect against waterborne transmission of giardiasis.
VII-30
-------
In this document, removal and inactivation efficiencies are expressed as a percent
removal or inactivation (e.g., 99%) or in terms of the logarithmic (base 10) removal or
inactivation ofGiardia cysts. For example, a 1 Iog10 removal or inactivation indicates a 90
percent reduction in cyst levels; a 2 Iog10 removal means that 99 percent of cysts are removed; 3
Iog10 removal means that 99.9 percent are removed. In some instances, removal is shown as a
greater than value which is calculated when no cysts are detected in the filtered water; the
calculations are based on Giardia cyst analytical detection limits for the methodology used..

1. Filtration

The filtration technologies most frequently used to remove microbial contaminants and
particles that cause turbidity from water sources are: conventional filtration, direct filtration,
slow-sand filtration, diatomaceous earth (DE) filtration, and membrane filtration. Conventional
filtration refers to 1he use of rapid-rate filters that are composed of granular material, either all
sand or dual/mixed media (e.g., anthracite-sand, anthracite-sand-garnet, activated carbon-sand)
preceded by chemical coagulation, flocculation, and sedimentation; filtration is followed by
disinfection. An important process in conventional filtration is flocculation which allows the
suspended particles in the water to form into a larger mass. Sedimentation allows the heavier
floe-particle masses to settle before the water is filtered. In some instances, flotation rather than
sedimentation is employed to reduce suspended particles. Giardia cysts and other pathogens
may become enmeshed in the floe-particle masses. Direct or in-line filtration with dual or mixed
media is often used for water sources with low turbidity and uniform water quality. This type
filtration does not employ the sedimentation or flotation process. After the addition of coagulant
or filter-aid chemical(s), sufficient time is allowed for mixing and coagulation, usually in the
water pipes, before the water is filtered Higher filtration rates are often used in direct filtration.
VII-31
-------
During the filtration process, particles in suspension enter a deep bed of the granular
material, which contains a random network of interconnected pores. The water flow is laminar in
these deep-bed granular filters, and in such a flow and pore structure, Giardia cysts will not be
significantly removed by straining. Rather, physico-chemical interactions govern the attachment
of suspended particle masses and cysts to the surface of sand and other filter media. Chemical
coagulation by aluminum or ferric salts, or the addition of poly electrolytes, ensures effective
filtration of very small size particles and microorganisms including Giardia cysts. Coagulant
chemicals hydrolyze and form hydroxide precipitates in water, providing positive potentials
which interact with and attract particles and cysts in the water. These particle masses are more
easily attracted to filter media and removed principally due to their larger size and electro-charge.
Most rapid-granular filters operate by gravity flow but are sometimes operated under pressure.
Pressure and gravity filtration employ the same process, and the operating principles are
identical. Coagulation is necessary for effective removal of Giardia cysts by either gravity or
pressure sand filters.

Al-Ani et al. (1986) demonstrated the importance of coagulation and optimum dosage of
coagulant chemicals in pilot- and full-scale treatment plants operated in both conventional and
direct filtration modes. Effective coagulation adequate to reduce turbidity from 0.5 to 0.1 NTU
was capable of removing 95% to 99.9% of G. lamblia cysts. Filtration efficiency was similar for
conventional and direct filtration. However, when no chemical coagulation was used, removal of
G. lamblia cysts was very poor, ranging from 0 to 50%. Similar poor removals were obtained
when ineffective coagulants or improper dosages were used.

Filtration by rapid granular filter media is not effective for 100% removal of cysts. Not
all Giardia cysts will be removed and some cysts that are removed may become dislodged. This
must be recognized and steps should be taken to optimize the filtration process and to monitor
each filter to detect changes in water quality and it effectiveness. An increase in turbidity or
particles in the filter effluent may indicate filter breakthrough (i.e., the ineffective removal of
VII-32
-------
cysts or the release of previously removed cysts). This may occur towards the end of a filter run
when the filter efficiency is poor, during the restarting of a filter after it has been cleaned, when
filtration flow rates are increased, when coagulation dosages are inadequate, or when the water
source becomes more contaminated. Filter breakthrough should be closely monitored. Ongerth
(1990) found that major deficiencies in the operation of three small water plants with either
conventional filtration, direct filtration, or DE filtration caused Giardia cyst removals to range
from about 40% to 99%. These deficiencies included poor optimization of chemical addition and
on-off filter cycles without backwashing. For the sand filters, filtering to waste after
backwashing was recommended, as the initial period of filtration immediately after backwashing
indicated the potential for passage of cysts.
Bellamy et al. (1993) reviewed various treatment plant performance factors that may
affect the removal of Giardia cysts. Especially important are: adequate rapid mix-coagulation
with appropriate chemical coagulants; appropriate flocculation times and sufficient volumes and
baffling within the sedimentation tanks; adequate depths for filtration and use of multiple media
filters (e.g., sand and anthracite). Where the turbidity of raw water is low and direct filtration is
used, it was recommended that source waters be monitored for episodes of higher turbidity and
contamination with Giardia cysts. To maintain their high flow rates and removal effectiveness,
filters must be frequently backwashedto remove material that has been retained in the filters.
Because clean filters are less efficient immediately after the backwash process, operational
procedures should include provisions to slowly increase the flow rates to the filters and there
should be sufficient time for ripening of the filter prior to being placed back online with
continuous measurement of turbidity. For a short period of time after being restarted, the filtered
water should be discharged to waste.

Sand in slow-sand filters has a much smaller effective size than that used in conventional
or direct filtration and the filtration rates are greatly reduced. Removal is accomplished by
VII-33
-------
physical-chemical and biological mechanisms within the top layers of sand (Weber-Shirk and
Dick, 1997). Straining of particles is the dominant mechanism within the filter cakes, and inter-
particle attraction is responsible for particle removal in the filter bed. Biologically mediated
particle removal was observed for particles smaller than 2 jim. Generally, no pre-treatment is
used with slow-sand filters, but some may be preceded by coagulation, sedimentation, or
roughing filters which remove large size particles that may clog the filter necessitating more
frequent cleaning. DE filtration is also used for the direct treatment of surface waters with
relatively low levels of turbidity. Water is filtered through a precoat cake of DE filter media that
has been deposited on a support membrane; additional body feed of DE is continuously added to
the raw water to maintain the filter cake permeability. Pressure-driven membrane filtration
processes used for municipal water treatment are categorized by the effective size of the
membranes (i.e., what is the largest partide, colloid, or molecule that can pass through the
membrane). The four categories of membranes, in order of increasing removal effectiveness of
micron-size contaminants are: microfiltration, ultrafiltration, nanofiltration, reverse osmosis.
Membrane systems may also require pretreatment to remove material that can clog or foul the
membranes.

The effectiveness of Giardia cyst removal by various filter technologies has been
evaluated using a challenge of G. muris or lamblia cysts or beads of a similar size in laboratory-,
pilot-, and full-scale test conditions. Field studies of Giardia removal have also been conducted
to evaluate the effectiveness of operating filtration plants. The ability of particle counters to
accurately size Giardia cysts has been investigated, and particle counting methods have been
used to quantify removal efficiencies of water filters. Results have shown that particle removal
can be indicative of microorganism removal, although particle counters may tend to undersize the
organisms (O'Shaughnessy et al., 1997).

In order for the results of pilot-scale studies to be meaningful, pilot plants must be
properly designed to reflect the conditions of a specific treatment process(es) in an actual
VII-34
-------
treatment plant, and the influence of various operational factors must be understood and
controlled to the extent possible (McTigue and MacPhee, 1997). When results of pilot- and full
scale systems are compared, similar cyst removals are often not found. Even the results of pilot-
scale studies may differ when conducted in areas where water quality differs. Many factors may
influence the removal effectiveness observed in these studies, and the interpretation of their
results depends on a thorough evaluation of the study design, operating conditions, and water
quality characteristics. Important considerations include: the levels of cysts used in seeding
studies and how the cysts are added to the system; cyst source, age and preparation; water quality
characteristics, especially water temperature, pH, turbidity; water system demand or flow rates;
coagulant chemicals and dosages, and methods of sample collection and analysis of cyst levels in
raw and filtered water. To determine removal effectiveness, the level of cysts in the raw water is
compared with the level in filtered water. In some studies, the level of cysts in raw water was
calculated based on the number of seeded cysts added to the raw water while in other studies the
level of cysts is measured from samples of raw water collected after cysts are added. Calculated
versus analytically measured cyst levels in raw water often provide a different measure of seeded
cysts that will be then compared with an analytically measured level in the filtered water.
Different removal effectiveness can also be due to the operational dynamics and hydraulics
between pilot- and full-scale plants and among pilot plants in different areas. Giardia removal
efficiencies from pilot- and full-scale filtration studies are summarized in Table VII-1. Brief
descriptions are provided for the studies reported in the Table.
VII-35
-------
Table VII-1. Summary of Removal Effectiveness of Various Filtration Processes
Type of Filtration
Conventional
Conventional
Conventional
Conventional
Conventional
Conventional
Conventional
Conventional
Conventional
Direct
Direct
Direct
Direct
Package plant
Slow Sand
Slow Sand
Slow Sand
Slow Sand
DE
DE
DE
Experimental
Design
Pilot-scale
Pilot-scale
Pilot-scale
Full-scale
Field
Field
Field
Field
Field
Pilot-scale
Pilot-scale
Pilot-scale
Full-Scale
Full-scale
Pilot-scale
Pilot-scale
Pilot-scale
Field
Pilot-scale
Pilot-scale
Pilot-scale
Removal Iog10
3.4-5.1
3.4
l.l->3
3.3
2-2.5
>2.2->2.8
>5
1.5-1.7
1.5
3.1-3.6
1.5-4.8
3.3
3.9
3
>3-4
>3-4
2.8->4
1-2
>2->3
^^ o
> 5
>3
References
Pataniaet al, 1995
Nieminski & Ongerth, 1995
Logsdon etal., 1985
Nieminski & OngerUi, 1995
LeChevallieret al., 1991b
LeChevallier& Norton, 1992
Payment and Franco, 1993
Stateset al. 1995, 1997
Kelleyetal, 1995
Ongerth & Pecoraro, 1995
Pataniaet al., 1995
Nieminski& Ongerth, 1995
Nieminski & OngerHi, 1995
Horn et al., 1988
Bellamy et al., 1985; Jakubowski, 1990
Jakubowski, 1990
Schuleret al., 1991
Fogel etal., 1993
Logsdon etal., 1981; Jakubowski, 1990
Lange etal., 1986
Schuleret al., 1991
VII-36
-------
| Microfiltration
Pilot-scale
6-7
Jacangelo et al.
1995 |
a. Conventional and Direct Filtration

Logsdon et al. (1985) conducted pilot-scale studies to evaluate sedimentation efficiency
and removal of G. muris cysts with various types of media (granular activated carbon, sand,
coarse anthracite, and dual-media). In waters with turbidities of 27 to 32 NTU, sedimentation of
alum-coagulated water resulted in 65% to 83% removals of Giardia cysts; in waters with
turbidities of 7.5 to 15 NTU, sedimentation of water coagulated with alum and a slightly anionic
polymer resulted in 79% to 93% removals of Giardia cysts. In evaluations that compared
removals among different filter media, coarse anthracite did not perform as well as the other
types of filter media when only alum was used; its performance was improved by use of the
polymer. Logsdon et al. (1985) noted that cyst levels were higher during the initial phase of the
filter run. This emphasizes the need to provide for filter ripening after a backwashed filter is
placed back into service, and for a short period of time after being restarted, the filtered water
should be discharged to waste. These studies showed that 3 Iog10 of Giardia cysts can be
removed and decreased removals are associated with increased turbidity indicating filter
breakthrough.

Patania et al. (1995) conducted pilot-scale studies of conventional filtration of waters
with turbidities between 0.2 and 13 NTU and Giardia cyst levels between 10 and 200/L. With
treatment optimized for turbidity removal, Giardia cyst removal ranged from 3.4 to 5.1 Iog10
during stable filter operation. Although the median turbidity and particle removals were only 1.4
and 2 Iog10, respectively, the median Giardia cyst removal was 4.2 Iog10. A filter effluent
turbidity of 0.1 NTU or less resulted in the most effective Giardia cyst removal. Giardia cyst
removal was 0.2 to 1.8 Iogs10 higher during conventional treatment, which included
sedimentation, compared to direct filtration. Giardia cyst removal was reduced by up to 1 Iog10
VII-37
-------
when the filter effluent turbidity increased from 0.1 to 0.3 NTU. Giardia cyst removal was
generally 0.4 to 0.5 Iogs10 lower during filter maturation or ripening after it had been
backwashed.

Nieminski and Ongerth (1995) evaluated both direct and conventional filtration in pilot-
and full-scale water treatment plants in studies using G. lamblia cysts that had been inactivated
by heat and formalin. The pilot plant was operated with a filter loading rate of 5.75 gpm/sq. ft.
(gallons per minute per square foot of filter surface area) at 0.5 gpm. The full-scale plant was
operated with a filter loading rate of 4.8 gpm/sq. ft. at 600 gpm. Turbidities in the source water
for the full-scale plant varied from 2.5 NTU to 11 NTU during the spring and were as high as 28
NTU during August. The source water for the pilot-scale plant typically had turbidities of 4
NTU. In the pilot-scale studies, G. lamblia cyst removals averaged 3.4 and 3.3 Iog10 for
conventional and direct filtration, respectively, when filtered water turbidities were between 0.1-
0.2 NTU. When the full-scale plant achieved similar filtered water turbidities, G. lamblia cyst
removals averaged3.3 Iog10 for conventional filtration and 3.9 Iog10 for direct filtration
Differences in the performance of direct filtration and conventional treatment in the full-scale
plant were attributed primarily to different source water quality. These studies also showed that
removals of cyst-sized particles and turbidity are useful indicators of cyst removal effectiveness.

Ongerth and Pecoraro (1995) evaluated the removal of Giardia cysts obtained from
infected animal fecal material in a very low turbidity source water (0.33 to 0.58 NTU). The 1-
gpm pilot plant used multimedia filters operated in direct filtration at a loading rate of 5 gpm/sq.
ft; alum coagulation was used, and a filter maturation or ripening period was allowed. With
optimal coagulation, 3.1 to 3.6-log10 removals of Giardia cysts were obtained. In one test run,
where coagulation was intentionally suboptimal, cyst removal was only 1.3 Iog10 even though the
filtered water turbidity was less than 0.5 NTU This emphasizes the importance of maintaining
optimum coagulation for effective Giardia cyst removal.
VII-38
-------
A packaged dual-stage filtration system for small water systems was evaluated by Horn et
al (1988). In two Colorado river waters, G. lamblia removals ranged from <1 to 2 Iog10 in a
water with turbidity of 4 NTU to >3 Iog10 in a water with turbidity of <1 NTU.

Surveys of municipal water supplies have also provided data to evaluate the effectiveness
of water filtration. Since not all water treatment plants are operated in the most efficient manner,
surveys ofGiardia occurrence in water samples from full-scale water treatment should provide
more realistic information about actual removal than pilot-plant studies where operation is highly
controlled. Rose et al. (1991a) analyzed 257 water samples collected from water sources and
potable water from 17 states in the United States. Giardia cysts were found in 16% of the
surface waters at an average level of 3 cysts per 100 L. Although Cryptosporidium oocysts were
found in 4 (14%) of 28 treated drinking water samples from systems using conventional and
direct filtration, no Giardia cysts were detected in any of these samples. Although these results
suggest that the facilities sampled were effectively removing Giardia cysts, the levels of cysts in
source waters were relatively low, and sampling was limited. Chauret et al. (1995) reported data
for the removal ofGiardia cysts from the raw drinking water in the water treatment plants in
Ottawa, Ontario. No cysts were detected in any treated water samples from the treatment plants,
even though cysts were detected in 83% of the raw water samples at the plant intakes.

LeChevallier and Norton (1992) evaluated Giardia occurrence in source and filtered
waters at three locations with high (1.8-120 NTU), moderate (3.5-75 NTU), and low (0.4-25
NTU) turbidity. The geometric mean number of cysts detected in raw water at each location was
2.9, 5.8, and 9.1 cysts per L, respectively. The detection ofGiardia in treated water samples
depended primarily on the number of cysts in the raw water, and reported removals at each
location were >2.3, >2.8, and >2.2 Iog10, respectively. In a more extensive monitoring of 347
surface water samples collected between 1988 and 1993, LeChevallier and Norton (1995) found
Giardia present in 54% of the samples collected. In a survey conducted during 1991 to 1993,
water samples were collected from 72 surface water plants in 15 states and 2 Canadian provinces
VII-39
-------
(LeChevallier and Norton, 1995). Giardia cysts were detected in filtered water on 12 occasions
(4.6% of 262 samples). When the Giardia cysts were detected, there was an average of 2.6
cysts/100 L (range = 0.98 to 9.0 cysts/100 L). Earlier, LeChevallier et al. (1991a, b) had
conducted a survey of the occurrence of Giardia cysts and evaluated removal efficiencies for
Giardia in 66 surface water treatment plants in 14 States and 1 Canadian province. Most of these
water systems achieved between 2 and 2.5 Iog10 removals for Giardia. Giardia cysts were
detected in 17% of the 83 filtered water effluents that were sampled. The geometric mean for the
positive samples was 4.45 cysts per 100 L with a range of 0.29 to 64 cysts per 100 L. Giardia
was frequently found in filtered water at facilities with poor quality source waters. For treatment
plants with Giardia-poshive samples, an average of 2.14 Iog10 removal was found. For Giardia-
negative plants, >2.45 Iog10 removals were calculated based on the analytical detection limits of
the methodology. Water treatment plants studied used sand, granular activated carbon (GAC),
dual media or mixed media filtration systems. Effluent samples from dual media and mixed
media filtration plants were more likely to be negative for Giardia cysts, while effluent samples
from the GAC and rapid sand filter type plants were more likely to be positive (LeChevallier et
al., 1991b).

In two conventional water filtration plants, Kelley et al. (1995) observed a 1.5 Iog10 mean
removal of Giardia cysts. Raw water turbidities ranged from <0.1 to 60 NTU at one location and
<0.1 to 101 NTU at the other. The authors suggested that the low removal was due to poor
coagulation. Even though the coagulation process was not optimized and cyst removal was poor,
the finished water turbidity was less than 0.5 NTU.

Three Montreal area water treatment facilities that used conventional filtration with ozone
disinfection were sampled for a number of pathogens and indicator microorganisms at various
locations (raw, settled, filtered water) by Payment and Franco (1993). Giardia cysts were
detected in 80%-100% of raw water samples with geometric mean levels of cysts at each of the
three locations 7.23, 336, and 1376 per 100 L, respectively. Giardia cysts were detected in only
VII-40
-------
one sample of filtered water — at the treatment plant with the highest level of cysts. Based on a
geometric mean of 0.1 Giardia cysts per 100 Lin filtered water at this plant, a removal of 5.2
Iog10 cysts was calculated. At the other two plants >5 Iog10 removals of Giardia cysts were
calculated based on analytical detection limits. Sedimentation at the plants was shown to remove
2.7-2.9 Iog10 of Giardia cysts.

States et al. (1995, 1997) reported that the Pittsburgh Water Treatment Plant effectively
removed Giardia cysts from the Allegheny River source. Based on the relatively low arithmetic
mean of Giardia in the Allegheny river source, removal of cysts was calculated to be 1.7 Iog10;
the removal based on the geometric mean was 1.5 Iog10. Although no cysts were detected in
filtered water, filter backwash water samples showed positive occurrences of Giardia cysts on
13% of the sampling occasions and in 8% of the water samples, with arithmetic and geometric
means of 16.8 and 58.6 cysts/100 L of filter backwash water, respectively. This emphasizes the
importance of backwash water as a source of contamination whether it is disposed or recycled to
the influent of the treatment plant. Proper management including treatment, equalization of flow,
and monitoring is required when backwash water is recycled (Cornwell and Lee, 1994).

In summary, studies indicate that conventional and direct filtration, when operated under
appropriate coagulation conditions, can removes to 4 Iog10 of Giardia cysts. The highest
removal rates occurred in pilot plants and water systems where coagulation was optimized and
low filtered water turbidities (0.1- 0.3 NTU) were achieved. In plants where coagulation was not
optimized, cyst removal was poor even when low turbidities were achieved in filtered water.
High levels of cysts are found in filtered backwash water, and this source of contamination
should be considered before backwash water is discharged or recycled.

b. Slow Sand and Diatomaceous Earth filtration
VII-41
-------
Jakubowski (1990) reported data from various studies which indicated that both DE and
slow-sand filtration are effective in removing Giardia cysts from water. A well designed and
operated plant using slow-sand or DE filtration is capable of removing at least 3 Iog10 of Giardia.
Removal is less efficient for slow-sand filters at near freezing temperatures (Fogel et al. 1993;
Schuler et al., 1991). Cleasby et al. (1984) found that when source water quality is high, slow
sand filtration outperforms direct filtration with alum or cationic polymer as a coagulant. Direct
filtration also required substantially more operational skill and attention. Slow sand filtration is
particularly appropriate for small water treatment systems where there may be limited operating
personnel present on-site for continuous monitoring of filter efficiency. Riesenberget al. (1995)
found a slow sand filter in Camptonville, California, satisfactorily maintained filtered water
turbidity levels of <1.0 NTU despite stream turbidities of >30 NTU.

Bellamy et al. (1985) reported the results of pilot-scale studies which showed slow sand
filtration was capable of removing virtually 100% of Giardia cysts as the sand bed matures. At
hydraulic loading rates of 0.04 to 0.4 m/h, human-source Giardia cyst removals were uniformly
high and averaged > 3 to 4 Iog10. Fogel et al. (1993) reported on the efficiency of removal for a
slow sand filtration system in British Columbia where no detectable cysts were found in 34 of 35
filter effluent samples. One sample contained a single cyst (11 cysts per 100 L). Based on this
limited sampling, the slow sand filter was reported to be able to remove an average of only 93%
of the Giardia cysts found in the raw water. Schuler et al. (1991) reported data from pilot-scale
studies of slow sand and DE filtration of water with turbidities ranging from 0.1 to 5.8 NTU.
Results indicated that both types of filters were able to remove greater than 3 Iog10 of G. muris
cysts; however, only 2-3 Iog10 removals could be achieved in the slow sand filter during the
winter months and removal efficiency of the DE filter was decreased during a malfunction that
caused the filter cake to crack.

Logsdon et al. (1981) evaluated the removal of 9- m-diameter radioactive microspheres
and G. muris cysts byDE filters. DE filtration consistently removed >2 Iog10 of microspheres
VII-42
-------
and cysts and frequently achieved >3 Iog10 removal. Effective filtration was dependent on DE
precoat thickness up to 1.0 kg/m2 precoat of diatomite. Effluent turbidity was not found to be an
effective indicator of DE filtration efficiency, and thus, reliance of effective removal depends
solely on proper operation. Subsequent studies with human-source Giardia cysts confirmed that
DE filtration could remove >2 Iog10 of cysts (Jakubowski, 1990). Pilot plant studies by Lange et
al. (1986) showed that virtually 100% (qualified by detection limits) of G. lamblia cysts were
removed by DE filtration. Water temperature did not affect DE filter performance but finer size
DE and lower filtration rates resulted in higher removal of bacteria and turbidity. DE filtration is
effective for Giardia cyst removal; however, the raw water must be of low turbidity and good
microbial quality, and the DE filter must be operated properly (Logsdon, 1988).

c. Membrane and Other Filters

At least 2 Iog10 removal of Giardia cysts should be possible with various types of point-
of-use/point-of-entry (POU/POE) systems employing such devices as cartridges containing
materials such as yarn-wound fibers, ceramics, paper, or other types of filtration media of an
appropriate effective size (Jakubowski, 1990). However, to adequately protect against
waterborne disease, systems should be selected based on their capability to remove 3 Iog10
Giardia cysts or cyst-sized particles, and it should be remembered that smaller-sized protozoa
may not be removed by systems that can effectively remove Giardia cysts. Jacangelo et al.
(1995) studied the removal of G. muris by two hollow fiber microfiltration membranes, one
spiral wound fiber microfiltration membrane, two hollow fiber ultrafiltration membranes, and
one tubular ceramic ultrafiltration membrane. Nominal pore sizes of the microfiltration
membranes were 0.1 to 0.2 jim, and nominal molecular weight cutoffs of the ultrafiltration
membranes were 100,000 to 500,000 daltons. The membranes were found to achieve from 4.6 to
>5.2 Iog10 removals of cysts under bench-scale worst case operating conditions, and >6.4 Iog10
removals of cysts under pilot plant normal operating condition. All of the hollow-fiber
membranes removed G. muris cysts to less than detectable levels; no cysts were detected as long
VII-43
-------
as the membrane remained intact. Physical straining of cysts appeared to be the primary
mechanism of filtration. Earlier studies by Jacangelo et al. (1991) also found good removal of G.
muris cysts by low-pressure hollow fiber ultrafiltration membranes; 4.1 to 5.0 Iog10 removals
were obtained from four different source waters — two from northern California with mean
turbidities of 0.5 and 9 NTU and two from Boise, Idaho with mean turbidities of 0.5 and 4.9
NTU. A pilot study to determine the feasibility of reclaiming municipal wastewater found an 8-
10 log removal ofGiardiaby ultrafiltration/nanofiltration membranes (Madireddi et al., 1997).

2. Disinfection

Jakubowski (1990), Hoff (1986) and Jarroll (1988) reviewed the effectiveness of
disinfectants to inactivate Giardia cysts. These reviews considered only studies where in vitro
excystation or animal infectivity was used to assess cyst viability or infectivity because these are
more sensitive indicators than other methods, such as the ability of cysts to exclude vital stains
(Bingham et al., 1979). Hoff et al. (1985) compared animal infectivity and excystation as
endpoints for determining the efficacy of disinfection of Giardia muris cysts; viability was
assessed before and after exposure to free residual chlorine. Substantial inactivation of cysts was
observed by both mouse infectivity and excystation after exposure to an initial free chlorine
residual of 1 mg/L at pH 7.0, at 5°C, and Hoff et al. (1985) concluded that in vitro excystation
was an adequate indication of G. muris cyst infectivity. Studies by Rice et al. (1982) found that
G. muris cysts were more resistant to inactivation by chlorine than human-source cysts and thus,
should provide a conservative indication of disinfection effectiveness. Hoff et al. (1985) also
found that Giardia muris cysts were more resistant than human-source cysts to free chlorine.
Human-source Giardia cysts for chlorination studies can be obtained from either asymptomatic
or symptomatic persons, since Rice et al. (1982) found that they have similar resistance to
chlorine. Information from a number of early disinfection studies indicated that G. muris were
VII-44
-------
among the most resistant waterborne microorganisms to chlorine and other disinfectants
(Jakubowski, 1990).

Results of studies by Jarroll et al. (1981) are a reminder that chlorine does not always
result in 100% inactivation of Giardia cysts. Jarroll et al. (1981) studied chlorine inactivation of
human-source Giardia cysts at water temperatures of 5°C, 15°C, and 25°C, water pH of 6, 7, and
8, chlorine contact times of 10, 30, and 60 minutes, and chlorine concentrations from 1 to 8
mg/L. The inactivation of cysts by chlorine was found to be less effective at higher pH values,
and lower water temperatures. Less than 30% of cysts were inactivated at water temperatures of
5°C and exposures to 2 mg/L chlorine for 30 minutes contact time at pH 8. Further, at water
temperatures of 5°C, exposures to 1 mg/L chlorine for 10 minutes contact time at pH 8 less than
45% of cysts were inactivated. At the time of this study, many unfiltered surface water systems
in the United States used similar contact times and chlorine concentrations for these water
temperatures and pH values suggesting that chlorination was inadequate in many water systems,
especially those that reported waterborne outbreaks. At 25°C, exposure to 1.5 mg/L chlorine for
10 minutes killed all cysts at pH 6, 7, and 8. At 15°C, 100% mortality required exposure to 2.5
mg chlorine/L for 10 minutes at pH 6; however, at pH 7 and 8, small numbers of cysts (less than
0.8%) remained viable after 30 minutes while no cysts were viable after 60 minutes. At 5°C,
exposures to 1 mg/L chlorine for 60 minutes did not kill 100% of the cysts at any pH tested,
while 2 mg/L resulted in 100% mortality of the cysts after 60 minutes at pH 6 and 7 but not at pH
8. A chlorine concentration of 4 mg/L also caused 100% mortality at all three pH values after 60
minutes but not after 30 minutes. Chlorine concentration of 8 mg/L killed 100% of the Giardia
cysts at pH 6 and 7 after contact for 10 minutes but required 30 minutes exposure at pH 8.

Hoff (1986) calculated Ct values for 99% inactivation of Giardia cysts by chlorine using
published and unpublished data (Table VII-2). Ct is the product of the concentration (C) of a
disinfectant (mg/L) and its contact time in minutes (t). A low Ct value indicates more effective
disinfection. For example at 15°C, a Ct value of 20 (pH 6) is almost twice as effective as a Ct
VII-45
-------
value of 37 (pH 8). In general, the effectiveness of chlorination was found to increase
considerably at higher water temperatures and at lower pH values. The most pronounced pH
effect on chlorination of human-source Giardia cysts was seen at lower water temperatures.
Jakubowski (1990) summarized extensive investigations reported in 1987 by Hibler et al.
conducted of the inactivation of human-source Giardia cysts by chlorine at various water
temperatures, pH, contact times, and concentrations (Table VII-3). Infectivity of Mongolian
gerbils was the end-point studied. Experimental water temperatures were selected based on the
temperatures of water sources in areas where most outbreaks were being reported. Similar to the
findings of Hoff (1986) and others, chlorination was found to be less effective at lower water
temperatures and pH. Also noted were erratic results in experiments with chlorine
concentrations above 2.5 mg/L and suggested that Ct values calculated with high chlorine
concentrations may not be reliable (Jakubowski, 1990).
Table VII-2. Ct Values for 99% Inactivation of Giardia cysts by Free Chlorine
Water Temp.
3°C

5°C
5°C

15°C

25°C

pH
6.5
7.5
7
6
7
8
6
7
8
5
7
9
6
7
8
Chlorine (mg/L)
0.24-1.1
0.24-1.0
0.41-2.73
1.0-8.0
2.0-8.0
2.0-8.0
2.5-3.0
2.5-3.0
2.5-3.0
4.4-13
2.9-7.1
11.6-72.6
1.5
1.5
1.5
Time (min)
37-297
150-770
236-467
6-84
7-152
57-164
7
6-18
7-21
4-16
4-16
3-16
<6
<7
<8
Mean Ct
68
140
360
65-75
97-118
110-142
20
32
37
66
29
206
<9
<10
<12
Source of Cysts
G. muris
G. muris
G. muris
human
human
human
human
human
human
G. muris
G. muris
G. muris
human
human
human
"Adapted irom Ho it (1986) and Jakubowski (1990).
VII-46
-------
Jarroll (1988) reviewed the inactivation data available for chlorine, chloramines, chlorine
dioxide, ozone, ultraviolet (UV) irradiation, and iodine. Ct values were presented for the
chemical disinfectants considering the effects of water temperature, pH, disinfectant
concentration, contact time for each disinfectant. Ozone, chlorine dioxide, free chlorine, iodine,
and chloramines, listed in descending order of effectiveness, were judged to be effective for
Giardia cysts. Jarroll (1988) cautioned that, while G. muris had in every disinfectant tested up to
that time been more resistant than G. lamblia cysts, the kinetics of the pH effect on chlorine
disinfection had recently been found by Leahy et al. (1987) to be different between G. muris and
G. lamblia cysts and contrary to earlier work.
Table VII-3. Ct Values for 99.9 to 99.99% Free Chlorine Inactivation
of Human-Source Giardia cysts (Mongolian Gerbil Infectivity Assay)
Water Temp.
0.5°C
2.5°C
5.0°C
pH
6-8
6-8
6-8
Mean Ct Values
185-342
142-268
146-280
*Adapted from Hibler et al. (1987) and Jakubowski (1990).
Rubin et al. (1989) evaluated the inactivation of human-source Giardia cysts by free
chlorine using Mongolian gerbils; Ct values were found to be higher than previously reported
with lower Ct values at higher pH levels. At 15°C the Ct ranged from 5-62 at pH 9 compared to
Ct values of 139-182 at pH 5. Jakubowski (1990) reported that Rubin also found that Ct values
for G. muris cyst inactivation by preformed monochloramine were substantially higher than those
for chlorine at pH 7 and 5°C. Meyer et al. (1988) found lower Ct values than Rubin for
inactivation of G. muris cysts by chloramines as they are being formed, but Ct values were still
VII-47
-------
found to be much greater than those for chlorine. Jarroll (1988) found that G. muris cysts were
more resistant to chloramines at lower pH values and that preformed chloramines were less
effective than chloramines that are not preformed. Hofmann and Andrews (1995) reported data
from disinfection experiments that indicated inactivation of Giardia cysts was more efficient at
pH 6.5 than 8.5, at 25°C than at 5°C, and that chlorine is more effective than chloramines. These
pH values were selected to be representative of typical water sources in Ontario, Canada, with
water temperatures representing winter and summer conditions.

Finch et al. (1995) summarized the Leahy's (Master of Science thesis, Ohio State
university, 1985) evaluation of chlorine dioxide inactivation of G. muris cysts using in vitro
excystation as an indicator of viability. Chlorine dioxide was an order of magnitude more
effective than free chlorine at 25°C and two orders of magnitude more effective at pH 9. In
contrast to findings with chlorine, chlorine dioxide effectiveness increased at higher pH values
(Jakubowski, 1990). At 25°C, the Ct value for chlorine dioxide ranged from 4.9-6.9 at pH 5
compared to a Ct of 1.7-3.0 at pH 9 (Leahy, 1985). In a pilot-scale study of G. muris
inactivation, a Ct value of 12 was reported for 99.9% inactivation at pH 8 and 8°C; viability was
determined by animal infectivity and in vitro excystation with similar results for each (Finch et
al., 1995).

Giardia cysts are readily inactivated by ozone (Wickramanayake et al. 1984a, b, 1985;
Wolfe et al., 1989; Labatiuk et al., 1991; Finch et al., 1993; Owens et al., 1994).
Wickramanayake et al. (1984a, b, 1985) found ozone to be more effective than chlorine for
inactivation of either human-source Giardia or G. muris cysts and less affected by water
temperatures. G. muris was slightly more sensitive to ozone at pH 5 than at pH 7, but was nearly
one and one-half times more resistant at pH 9 (Wickramanayake et al., 1984b). Finch et al.
(1993) found that the resistance of G. lamblia to ozone was not significantly different from that
of G. muris at 22°C and contact times of 2 and 5 minutes. Viability was assessed by the
C3H/HeN mouse and Mongolian gerbil models for G. lamblia and G. muris, respectively. The
VII-48
-------
Ct value for 99.9% inactivation of G lamblia by ozone was found to be 2.4 times greater than the
recommended Ct value in the SWTR Guidance Manual (U.S. EPA, 1989).

After comparing animal infectivity, excystation, and fluorogenic dye as measures of cyst
inactivation by ozone, Labatiuk et al. (1991) concluded there were no significant differences
among the three methods for inactivations up to 99.9%; however, only the C3H/HeN mouse
model had the sensitivity to detect inactivations greater than 99.9%. Labatiuk et al. (1992) also
found that water temperature, pH, and applied/residual ozone dose were important factors
affecting inactivation of G. muris cysts. Contact times of up to 2 minutes had a significant effect
in demand-free buffered water, but contact times up to 5 minutes were required for inactivation
in natural waters suggesting caution in applying results of laboratory disinfection studies to
natural waters. It was also found to be more difficult to achieve 99 or 99.9% inactivation of cysts
in natural waters at 22°C than 5°C. Haas and Heller (1989) studied the experimental data of
Hibler (1988) for the inactivation of Giardia by free chlorine and recommended caution against
extrapolation of inactivation results outside the range of experimental conditions. Haas et al.
(1996) also cautioned that inactivation data obtained in laboratory studies using buffered
demand-free water do not appear to be adequate for predicting inactivation in actual waters to be
disinfected. In earlier studies, Haas et al. (1994) determined that Giardia inactivation can be
achieved with free chlorine, monochloramine, and ozone in buffered demand-free water, as well
as in waters from two rivers, at dosages similar to those that might be employed in water
treatment facilities, providing the concentrations are high enough.

Hydrogen peroxide is sometimes added to ozone for oxidation of organic compounds.
Bench scale studies of the inactivation of G. muris with ozone and ozone-hydrogen peroxide
showed significantly greater inactivations in the presence of an ozone residual, leading Labatiuk
et al. (1994) to conclude that the design of ozone disinfection processes should maintain an
ozone residual for disinfection before the addition of hydrogen peroxide.
VII-49
-------
An on-site disinfection process for small communities uses electrolysis of a sodium
chloride solution to produce a mixture of oxygen and chlorine species (mixed oxidant gases or
MOGOD). Witt and Reiff (1997) presented data showing the effectiveness of MOGOD; Ct
values were comparable to those for ozone and chlorine dioxide (Table VII-5).

Rice and Hoff (1981) evaluated UV irradiation for inactivation of human-source Giardia
cysts. They found a reduction of less than 90% of Giardia cysts at the maximum dose tested,
63,000 W-s/cm2 of UV irradiation at a wavelength of 254 nm. Both human-source Giardia and
G. muris cysts are significantly more resistant to ultraviolet irradiation thanE1. coll and Yersinia
sp. (Rice and Hoff, 1981; Jakubowski, 1990; Karanis et al., 1992). Karanis et al. (1992) noted
that UV disinfection is not reliable because commercial use a dose range of 250-350 J/m2; a 2
Iog10 reduction of G. lamblia cysts required 1800 J/m2. A new generation of UV irradiation
devices with improved disinfection capabilities are currently being evaluated for their ability to
inactivate Cryptosporidium oocysts; however, no studies have been reported for Giardia.

Clark et al. (1989) and Clark (1990) described the development of a model based on first-
order kinetics to relate Ct values from inactivation data to free chlorine concentration, pH, and
temperature for use by water utilities in meeting provisions of the EPA's SWTR that requires
99.9% reduction of Giardia cysts from surface water sources. The Ct values of the SWTR
Guidance Manual (U.S. EPA, 1989) were based on a number of simplifying assumptions, such as
the Chick-Watson relationship for microbial inactivation and extrapolated values from data that
had been obtained from laboratory studies in buffered demand free water. Haas et al. (1996)
cautioned that the type of source water may be significantly affect predictions for microbial
inactivation. Factors other than disinfectant demand appear to influence inactivation of
organisms in natural waters, inactivation data obtained in laboratory studies using buffered
demand-free water do not appear to be adequate predictors of inactivation in actual waters to be
disinfected, and the use of pH in adjusting Ct values is insufficient to characterize the effect of
water quality on disinfection performance (Haas et al., 1996). Haas and Joffe (1994) proposed an
VII-50
-------
approximation to the exact solution of Horn's model for microbial disinfection kinetics to
describe inactivation in batch systems subject to disinfection decay/demand. The approximate
and exact solutions were found comparable for G. muris cysts exposed to preformed
monochloramine concentrations of 1 and 2 mg/Lat pH 6.9 and 18°C in water obtained from the
Bull Run Reservoir.
Table VII-5. Effectiveness of Water Disinfectants for 99% Inactivation ofGiardia Cysts
Disinfectant
Ozone
Ozone
Ozone
Ozone
MOGOD
MOGOD
Chlorine Dioxide
Chlorine Dioxide
Free Chlorine
Free Chlorine
Free Chlorine
Free Chlorine
Free Chlorine
Chloramine
Chloramine
Preformed Chloramine
Preformed Chloramine
Temp
25°C
5°C
25°C
5°C
20°C
3-5°C
25°C
5°C
25°C
5°C
25°C
15°C
5°C
18°C
3°C
15°C
5°C
pH
7
7
7
7
6-7.5
6-7.5
7
7
7
7
7
7
7
7
7
7
8-9
Ct
0.3
1.9
0.2
0.5
o
J
6-10
5
11
26-45
360-1012
<15
120-236
90-170
144-246
425-556
825-902
1400
Cysts
G. muris
G. muris
Human
Human
Human
Human
G. muris
G. muris
G. muris
G. muris
Human
Human
Human
G. muris
G. muris
G. muris
G. muris
Reference
Wickramanayake et al., 1984b
Wickramanayake et al., 1984b
Wickramanayake et al., 1984a
Wickramanayake et al., 1984a
Witt &Reiff, 1996
Witt&Reiff, 1996
Jarroll, 1988
Jarroll, 1988
Leahy et al., 1987; Jakubowski,
1990
Leahy et al., 1987; Jakubowski,
1990
Jarroll et al., 1981 ; Jakubowski,
1990
Rubin etal., 1989
Jarroll et al., 1981 ; Rice et al.,
1982; Jakubowski, 1990
Jarroll, 1988
Jarroll, 1988
Jarroll, 1988
Witt &Reiff, 1996
VII-51
-------
Because of the high risk of waterbornegiardiasis among campers, hikers, backpackers,
and travelers to developing countries, personal water purification methods employing iodine or
chlorine disinfection were reviewed by Jarroll (1988). At water temperatures of 20°C, all of the
six disinfection methods tested by Jarroll (1988) were completely effective against G. lamblia
cysts when prescribed procedures were employed for cloudy and clear water. However, at 3°C
several methods failed to completely inactivate cysts suggesting that procedures (residual
concentrations and contact times) be revised for low water temperatures. Ongerth et al. (1989b)
found that for seven disinfection methods tested, iodine-based methods were more effective than
chlorine-based methods; however, no chemical treatment achieved 99.9% cyst inactivation after
30 minutes. Heating water to at least 70°C for 10 minutes was found to be an acceptable
alternative (Ongerth et al., 1989b).

IV. Summary

1. Analysis

The absence of a practical cultural method for Giardia in environmental samples, and the
probability that one could not be developed, led to the development of microscopic examination
assay methods. Since Giardia does not reproduce in the environment once it leaves the host,
large-volume sample collection methods were developed using filtration through microporous
cartridge media. Collecting large volume samples of raw source water resulted in many eluates
containing a significant amount of participates that had been retained on the filters. Initially,
flotation clarification techniques used zinc sulfate solutions; subsequently, other compounds
including sucrose, Percoll, and Percoll-sucrose were evaluated and incorporated into the method.
The development of fluorescent antibodies for Giardia revolutionized the detection step which
had previously been dependent upon examining concentrates with none-selective iodine staining.
A combination method was also developed whereby a single sample could be simultaneously
assayed for Giardia and Cryptosporidium.
VII-52
-------
The original Giardia method was developed to assist in waterborne outbreak
investigations. It subsequently was adapted to different applications by those with a need to
determine drinking water treatment effectiveness, or occurrence and distribution of cysts in the
environment, or to study fate and transport of cysts. In the absence of regulatory requirements to
monitor for Giardia, there was no official standardized method. However, voluntary efforts
through groups such as Standard Methods for the Examination of Water and Wastewater and
ASTM resulted in consensus reference or proposed methods that could be used as a baseline and
modified as needed for particular applications.

The availability of consensus methods resulted in evaluation studies of all steps involved
in the methodology including sampling, elution, flotation clarification, and microscopic assay.
The sample collection and elution steps were found to account for significant losses of cysts. In
addition, aspects of flotation clarification, especially the specific gravity of the gradient solution
and the relative centrifugal force used to spin samples, were found to significantly affect
recovery. While retention of cysts and oocysts on the sampling filter was improved by higher
turbidities in the water being sampled, the greater quantity of material obtained in the sample
pellets presented difficulties in the flotation purification and microscopic assay steps. The nature
of the turbidity (e.g., organic or inorganic, particle size, etc.) was more important than the total
amount in causing detection and identification problems. For example, algae could make
clarification and detection more difficult in certain types of water and at certain times of the year.

The fluorescent antibody assay, while improving detection of cysts, necessitated
developing a new definition for identifying cysts. Presumptive cysts were defined by size, shape
and apple green fluorescence under specified conditions of reagent type and use and microscope
configuration; confirmed cysts were those meeting the presumptive criteria plus having defined
internal characteristics. This terminology created confusion for interpreting results, especially by
those not familiar with the methodology, who often ignored the results if no confirmed cysts
were found. The presumptive designation included all objects that might be Giardia cysts. The
VII-53
-------
confirmed designation was applied to those presumptive cysts that could definitely be identified
as Giardia. The remaining objects might or might not be Giardia because interferences, e.g.,
cross-reactions, were known to occur. Some cysts in a known, purified preparation of Giardia
will not meet the criteria for confirmation. The presumptive/confirmed terminology was
replaced with total counts/counts with internal structures in the Standard Methods and ICR
methods. Another limitation of fluorescent antibody identification is that it may only be specific
to the family level. While antibodies with various specificities have been developed, the
application and interpretation of results with them is complicated by uncertainty in defining
species within the genus, and in identifying those species that might have public health
significance.

Nucleic acid-based detection and identification techniques have been developed. While
they have the potential to specifically detect those species that may be important in human
infection, and they have demonstrated sensitivity down to one cyst in purified preparations, they
have yet to realize their full potential. Problems have been encountered with reproducibility of
the assays and with inhibition of the PCR reaction in environmental samples.

The advent of the ICR, and the necessity for developing defined data quality objectives
for that monitoring effort, resulted in performance evaluation data that underscored the low
precision of the method in unapproved laboratories. With the promulgation of the ICR, for the
first time a process was implemented in the United States for approving and conducting continual
performance evaluation of analysts and laboratories that wished to do environmental protozoa
analyses. Until that time, adherence to specific methodological protocols, or performance of
recommended quality assurance/quality control procedures, was strictly voluntary. Maintaining a
similar process after completion of the ICR may help to ensure the reliability of data obtained
through continued monitoring efforts.
VII-54
-------
Increased awareness of method limitations has also spurred development of alternative
methods and procedures. In the area of sample collection, sampling 10 L volumes instead of 100
L or more for raw waters is being investigated. Processing the entire concentrate for a 10 L
sample may be preferable to processing an undefined portion of a 100 L sample in terms of the
detection limit and it could help the laboratory and drinking water treatment utilities interpret
their results. Collecting smaller sample volumes means fewer particulates to cause interferences
in the detection assay, and it makes it easier to apply alternate separation technology (instead of
flotation separation where cyst recovery is low or erratic), such as immunomagnetic techniques.
Also, the use of membrane filters with defined porosity (instead of yarn-wound filters with
nominal porosities) for sample collection can result in better recoveries. For the assay portion of
the methodology, much of the tedium and fatigue associated with examining concentrates may be
relieved by using techniques such as flow cytometry and cell sorting.

Dependence upon non-cultural methods for the detection and identification of Giardia in
environmental samples has rendered determining the public health significance of positive
findings problematical. Determining the viability or infectivity potential of small numbers of
cysts detected with non-cultural methods has been difficult or impossible to do. A detected cyst
may be either viable (alive) or non-viable (dead). Viable cysts are not necessarily infectious. If
the organism is alive but has been injured, it may not be infectious. While viability
determinations might not be necessary for some applications, such as waterborne outbreak
investigations or determining the effectiveness of a treatment process to physically remove cysts,
they are very important in developing risk assessments upon which to base treatment
requirements or drinking water regulations.

Procedures used to determine viability have included dye staining, morphological criteria,
in vitro excystation, animal infectivity, and nucleic acid-based assays. Traditional dye staining
methods (e.g., with eosin) were found not to correlate with in vitro excystation or animal
infectivity. Subsequent research produced dyes that enter the viable cyst (e.g., FDA) and those
VII-55
-------
that are excluded from the viable cyst while they can enter non-viable cysts (e.g., PI). Work that
has been done with PI to date indicates that cysts stained with this compound are not viable.
However, cysts that do not take up the stain may be eilher viable or non-viable, and whether or
not inactivated cysts stain may depend in part on how they were inactivated.

Dye staining, morphological criteria, and in vitro excystation may be adequate indicators
of viability for some applications but could be conservative in estimating the potential for
infection. Animal infectivity has commonly been used in experiments to determine disinfectant
efficacy. However, it has seldom been used to evaluate the health significance of environmental
isolates. An exception would be the Canadian studies described by Wallis et al. (1996). They
concluded that gerbil animal infectivity can be inaccurate because some strains do not infect
gerbils. There is also the possibility that infectious cysts below the infectious dose might be
present and not detected. Other problems with this method are the cost and the necessity for
maintaining approved laboratory animal facilities.

Nucleic acid-based viability assays have focused on the detection of mRNA by RT-PCR
techniques using either the giardin gene or an HSP gene. Amplification of the HSP gene has not
proven reliable and there is some question about the survival and longevity of mRNA when the
organism is inactivated by different techniques. Besides practical problems relating to the
VII-56
-------
sensitivity and application of PCR techniques to environmental samples, the question of how
viability determined by these techniques relates to infectivity remains to be resolved.

Fresh, frozen, or preserved stools can be examined using traditional dye staining
techniques or with increasingly popular immunofluorescence assays. A variety of commercially-
available fluorescent antibody kits that target cysts or antigens are available. Evaluation of these
kits indicates that they have a high degree of sensitivity and specificity. They may require less
time to perform and produce results with a single stool sample equivalent to fresh stool and dye
staining techniques that require multiple stool examinations. The use of flow cytometry with
immunofluorescence reagents may allow a greater number of human or animal specimens to be
examined in a given time period with less operator fatigue.

For surveys of giardiasis in animal populations, examination of intestinal scrapings from
live-trapped animals may prove more fruitful than examination of feces from kill-trapped
animals. With either human or animal specimens that have been frozen and thawed before
examination, immunofluorescence assays are more likely to detect cysts than is examination by
conventional microscopy. This may allow samples to be archived and subsequently re-examined
for a variety of purposes, including quality control. One author concluded that phase microscopy
had an advantage over immunofluorescence assays in that phase microscopy allowed some
differentiation to the species level of cysts found in wild rodent populations.
VII-57
-------
Serodiagnosis is still not a useful technique in the clinical environment due to the
inability to distinguish between present and prior infections. However, serologic testing may
have value in conducting epidemiological studies. Secretory antibody has been detected in a
small study of saliva specimens from patients infected with Giardia. The potential for
developing tests that could be useful for either diagnostic or epidemiologic purposes based on
this finding remains to be determined. Also, the development and application of gene probe
techniques (e.g., PCR) for clinical diagnostic purposes has thus far proved challenging due to
inhibitory substances in feces and resulting problems with sensitivity and specificity.

2. Water Treatment

Information obtained during the past 20 years from laboratory, pilot plant, and full scale
treatment plant studies show that Giardia cysts can be effectively removed and inactivated by a
combination of filtration and disinfection. Because of Giardia''$ low infectious dose, the wide-
spread occurrence of the infection in humans and a variety of animals, and the relative resistance
of Giardia cysts to environmental conditions and water disinfectants, it is important to consider
multiple barriers for 1he protection and treatment of both surface and ground water sources: a
combination of watershed protection for surface waters and well-head/aquifer protection for
ground water sources, water filtration, disinfection, and protection of the integrity of the
distribution system. Use of all of these barriers affords the most effective means for assuring the
microbial safety of public water supplies (Geldreich and Craun, 1996).

It is impossible and morally unacceptable to eliminate animal sources of contamination
from a watershed. However, sources of contamination from farming and domestic animals can
be controlled and sources of contamination from wild animals and their impact on surface water
sources can be reduced. Control of human sewage discharges and appropriate wastewater
treatment will also help reduce contamination of source waters. The majority of waterborne
giardiasis outbreaks have occurred in unfiltered surface water systems, but outbreaks have also
VII-58
-------
occurred in groundwater systems impacted by surface water or sewage discharges. Wells and
springs should be protected from the influence of surface water and sewage discharges from
septic tanks and municipal wastewaters. While these controls can reduce the potential for
contamination of source waters, they cannot eliminate it. To effectively protect against the
waterborne transmission of Giardia, water treatment barriers are required. For surface water
sources and groundwater sources under the influence of surface water, disinfect!on and filtration
are also recommended. Filtration exceptions maybe granted where water sources meet criteria
of the EPA's SWTR; however, if water sources are subject to contamination with
Cryptosporidium, it should be remembered that disinfection levels used to inactivate Giardia
cysts may not be sufficient to inactivate Cryptosporidium oocysts (Craun, 1997) and filtration
may be necessary.

Filtration technologies commonly used by water utilities can be designed and operated to
remove 99% or more of Giardia cysts. Studies indicate that conventional and direct filtration,
when operated under appropriate coagulation conditions, can removes to 4 logs Iog10 of Giardia
cysts. The highest removal rates occurred in pilot plants and water systems that optimized
coagulation and achieved very lowfinished water turbidities (0.1- 0.3 NTU).

Disinfection is an important part of the multiple barrier concept of water treatment.
Chemical disinfectants can also reduce cyst densities, but the effectiveness of disinfection can be
affected by water temperature and pH, applied and residual disinfectant concentration and contact
time, particles that can shield cysts from contact with the chemical, and organic matter that can
cause disinfectant demand. Filtration can make disinfection more effective by reducing the
disinfectant demand and removing particles that may interfere with disinfection effectiveness.

Disinfection can also achieve 99% or greater inactivation of Giardia cysts: however, the
EPA regulates disinfectants and disinfection by-products and this may limit the concentration and
contact time of any chemical disinfectant that can be applied. When lower concentrations of a
VII-59
-------
disinfectant are required to meet disinfection and disinfection by-product limits, both filtration
and disinfection may be necessary so that sufficient levels ofGiardia cysts are removed to
prevent the waterbome transmission of giardiasis.

The majority of waterborne giardiasis outbreaks have occurred in unfiltered surface water
systems or unfiltered ground water systems impacted by surface water or sewage discharges.
This emphasizes the need for water filtration. Outbreaks have also occurred in filtered water
systems, and this shows the need for good design, optimization of the filtration process(es), and
frequent monitoring of treatment effectiveness. Both conventional and direct filtration facilities
should be designed with proper chemical pretreatment to provide adequate coagulation. In some
source waters, sedimentation may be needed to effectively remove cysts. In water filtration
plants where coagulation was not optimized, cyst removal was poor even though relatively low
turbidities were achieved in filtered water. High levels of cysts are found in filtered backwash
water, and this potential source of contamination should be considered before this water is
discharged to the environment or recycled back to the beginning of the water treatment plant.

Chemical disinfection employed by the water industry can inactivate Giardia cysts;
however, Giardia can be resistant to low doses of chlorine and chloramines, and there are
differences between the inactivation efficiencies of the various disinfectants. The reported
effectiveness of inactivation by the typically utilized disinfectants, in decreasing order of
efficiency, is follows: ozone, MOGOD, chlorine dioxide, iodine, free chlorine, and chloramines.
UV irradiation does not appear to be useful under current operating conditions. Water quality
including temperature and pH are important factors to consider when selecting a disinfectant and
its concentration and contact time. The turbidity and disinfectant demand of the water to be
disinfected are also important concerns. In waters of high turbidity the effectiveness of a
disinfectant maybe greatly reduced. Ct values are available to compare disinfectants, and values
are recommended for various conditions of temperature and pH. Applied and residual
concentrations are important to consider as is how disinfectants are added. For example, ozone-
VII-60
-------
peroxide is less effective than ozone and preformed chloramines are less effective than
chloramines that are not preformed. Information is available to provide guidance in selecting Ct
values, but it must be remembered that source water quality is also important. Since published
Ct values are based on results of laboratory studies in demand-free water, caution is
recommended in extrapolating of these data to natural waters and beyond the experimental
conditions. It should also be remembered that if source waters are heavily contaminated with
Giardia cysts, disinfection alone may not be sufficient to protect against waterbome infection.
Disinfection maybe adequate to inactivate 99.9% of the cysts, but the remaining cysts may still
be at infectious dose levels.
VII-61
-------
VIE. RECOMMENDATIONS FOR ADDITIONAL RESEARCH

F. Chapter II. General Information and Properties

Additional information is required to answer questions about Giardia species and
zoonotic routes of transmission. The characterization of Giardia by molecular approaches, such
as zymodeme or karyotype identification, can be useful in this regard. To conclusively determine
whether human giardiasis can be acquired by zoonotic routes and whether the ultimate source
was human or a lower animal will require carefully controlled animal feeding studies and more
comprehensive epidemiological investigations. Epidemiological investigations of outbreaks and
endemic risks, especially waterborne, should include the systematic collection of Giardia cysts
from infected humans, from animals suspected of transmission, and from environmental samples
and their characterization by molecular approaches.

G. Chapter III. Occurrence

A significant database has been developed on the occurrence and distribution of Giardia
in a variety of waters including municipal wastewater, surface water, and groundwater. Another
large database of occurrence information is being developed through water monitoring required
by the ICR. However, the sources of contamination on specific watersheds and the factors
affecting fate and transport of cysts are not as well characterized. For example, giardiasis is more
common in immature animals, and studies should be conducted to determine water quality
changes associated with the reduction and relocation of suspected animal sources, such as beaver.
As beaver begin to repopulate the watershed, their average age distribution will be much younger
with perhaps an accompanying larger prevalence of infection and greater contribution to source
water contamination.
vm-1
-------
There is no published information on the occurrence of Giardia in soils and sediments,
probably due to the difficulty in examining this type of sample for cysts. Methods are needed for
detecting, identifying and enumerating cysts in soils and sediments. After suitable methods are
developed and evaluated, they should be applied in laboratory and field studies to determine the
persistence of Giardia in these media.

Additional information is also needed on the occurrence and survival of Giardia on
surfaces and the potential for transmission by fomites. This information can assist in identifying
and controlling risks of Giardia infection among children in day-care settings.

A variety of foods have been associated with giardiasis outbreaks, and it appears that the
foods in these outbreaks were locally contaminated by infected food handlers. There are few
quantitative data, however, on the occurrence of Giardia on various fruits, vegetables, and foods
that are usually consumed without cooking. This may be related to the difficulty in recovering
cysts from foods. Research is needed on methods to quantitatively detect these organisms and
determine their survivability on foods. With the increased globalization of our food supply, more
surveillance of domestic and imported foods should be conducted in order to develop data for use
in risk assessments and to ensure against outbreaks. Recent outbreaks caused by another
protozoan, Cyclospora, have been associated with contaminated fresh raspberries imported from
Guatemala, and this problem underscores the need for additional knowledge about the occurrence
and survival of Giardia and other protozoa on various foods.

Temperature has been demonstrated to be an important factor in Giardia survival. The
effects of freezing have been determined in water and clinical specimens at low laboratory
freezer temperatures. However, recent work has demonstrated that significant fractions of
Cryptosporidium oocysts can survive temperatures just below freezing for relatively long time
periods. Similar data, reflecting conditions likely to be encountered in the environment, should
be developed for Giardia.
VIII-2
-------
Information on occurrence of cysts in estuarine environments, and the factors in such
environments that affect their survival, are limited and additional research should be conducted
in this area.

The recent report that an isolate of Flavobacterium can kill Giardia prompted the
suggestion that biological control of the organism may be possible. The identification of other
organisms that could affect cyst survival may be fruitful and should be further explored.

H. Chapter IV. Health Effects in Animals

Since giardiasis is more common in immature animals and since growth retardation may
be a consequence of this disease, improved diagnosis and treatment of animal giardiasis is a
desirable goal. Incorrect diagnoses of Giardia infection may occur in animals when cysts are
confused with yeasts or missed altogether in light infections. In this regard, it is important for
veterinarians to be able to correctly process animal fecal specimens for microscopic examination.

Consideration should be given to the development of vaccines to prevent giardiasis in
animals. The first successful Giardia vaccine, if one is developed, will probably be used in
humans. When it is shown that such a vaccine is feasible, it should be a small step to develop
similar vaccines for animals.

D. Chapter V. Health Effects in Humans

Many questions related to the host-parasite biology of Giardia remain. Following similar
exposure, is the intestinal environment in persons who develop clinical giardiasis different from
that of persons who do not? What is the role of adherence and invasion in determining the
establishment and clinical course of Giardia infections? Do trophozoites mutate in vivo, and is
this mutation in response to antigen-specific antibody? What are the functionally important
VIII-3
-------
antigens? How are animal models relevant to understanding Giardia infections in man? Are
intestinal phagocytic cells functionally active in the lumen of the intestine? How does Giardia
cause illness, and can a vaccine be developed? Additional research is needed to help answer all
of these questions.

It is questionable whether a vaccine can be developed, as this is already difficult enough
for the blood dwelling parasites. Even if one can be developed, its use will likely be limited.
Additional research should be conducted on the treatment of the disease. Although current drugs
have been found effective, resistance has been observed for certain strains/genotypes. An
alternative is to treat giardiasis with drugs aimed at the metabolism of Giardia. Research on its
unique metabolism and differences with that of its warm-blooded hosts might be suggest a way
to interfere with its life cycle.

Additional epidemiological studies are needed to better determine the prevalence of
giardiasis and Giardia infection, sources of infection, quantitative estimates of risks associated
with waterborne and other exposures, and the role of protective immunity. In order to conduct
meaningful epidemiological studies, one of the highest priorities is the development of a
sensitive and specific assay for anti-Giardia antibodies. The assay must be well characterized
with information on the duration of serological response for each of the markers of interest. If an
assay is available, population-based studies of endemic levels of Giardia infection could be done
around the world. Serological studies can not only help in evaluating the efficacy of various
water treatment systems in reducing risks but also in identifying other routes of infection.

There is variability in the humoral response to Giardia infection. Some patients with
symptomatic infections fail to develop sufficiently high antibody levels for results to be called
positive. In some patients, levels of anti-Giardia IgG antibodies remain elevated long after the
infection appears to have been eradicated. No sero-diagnostic procedure has been reported that is
capable of distinguishing asymptomatic from symptomatic infection, and additional work is
VIII-4
-------
needed in this area. It is possible that existing sera from experimental studies can help evaluate
some of the proposed Western blot serological assays. The specificity of antigen detection assays
may be significantly improved by assays based on certain antigens groups (30/31 kDa, 57 kDa
and high molecular weight antigens). Additional work is needed to evaluate these markers under
both controlled and field conditions.

Additional research is needed on the suitability of using saliva for detect!on of anti-
Giardia antibodies in patients with giardiasis. Saliva tests have an intuitive appeal since they can
be applied to studies of children and do not require drawing blood.

E. Chapter VI. Risk Assessment

To improve risk assessments, better epidemiological information is needed about the
risks of endemic waterborne giardiasis, role of immunity, and potential for secondary
transmission among families where primary cases are waterborne. This will require clinical and
specially designed epidemiological studies that include serological analyses as previously
described. Additional research is needed to better describe the role of protective immunity in
symptomatic illness and how long this might last. Information is also needed to better describe
the risks of chronic diarrhea, malabsorption, and other chronic effects associated with exposure
to Giardia.

Additional epidemiological studies are needed to provide better quantitative information
about the endemic waterborne risks of Giardia infection and giardiasis for populations using
unfiltered surface water, filtered surface water, and unfiltered groundwater sources. Risks from
well designed epidemiological studies can then be compared with current estimates of risk from
mathematical models used for risk assessment purposes. Epidemiologically determined risks can
also be used to evaluate and revise, if necessary, the EPA's currently recommended annual risk
VIII-5
-------
ofGiardia infection that drinking water systems should attempt to maintain (one waterborne
Giardia infection per 10,000 persons).

More complete information about exposures to Giardia cysts is needed to improve
estimates from currently used models. This includes studies of the occurrence, transport, and fate
of cysts in water sources, storm waters, reservoirs, animal waste ponds and lagoons, and septic
tank effluents. Improved collection procedures and analytical methods are needed for detecting
cysts in water and food samples as well as an inexpensive method to determine the viability of
cysts detected. This will improve the exposure assessments used to estimate waterborne risk
Serological epidemiological studies can also assist with providing better exposure information
for risk assessment purposes.

A major concern is the interpretation of the currently estimated waterborne risk of
Giardia infection in the United States. The mathematical model used to estimate these risks is
simple to use but has limitations. A recent model developed in the Netherlands considers several
of these limitations and should be applied to water exposures in the United States. This will
allow a comparison of risks with those estimated from the more simple model. Another recently
developed risk assessment model considers population risks and attempts to include all of the
relevant information that may be needed to estimate waterborne risks. This model should be
used in combination with a sensitivity analysis to identify the parameters that may have the
greatest effect on risk estimates. Waterborne Giardia infection risks should be estimated for the
United States with each of the models.

F. Chapter VTI. Analysis and Treatment

Combination analytical methods were developed for Giardia cysts and Cryptosporidium
oocysts in the same sample because early studies suggested that acceptable recoveries of both
could be obtained. Performance evaluation studies have shown that the methods generally have
VIII-6
-------
low recovery and poor precision. Both recovery and precision are better for Giardia than for
Cryptosporidium and consideration should be given to developing methods specific to each. A
draft method recently proposed by the EPA (Method 1622) is recommended only for oocysts at
the present time. The method protocol includes 10 L sample volumes, cartridge membrane
filtration, immunomagnetic separation and microscopic examination with or without flow
cytometric cell sorting. Work is in progress to apply this methodology to Giardia detection.
More research needs to be done on practical environmental sample methods for determining the
species of cyst detected and whether or not they are viable or infective. Research is needed on
the appropriate sample volumes for raw and treated waters and on whether substituting 10 L
sample volumes for 100 L will result in improved method recovery and precision. Also, the
effect of collecting larger sample volumes with methods designed for smaller volumes should be
evaluated.

The availability of protocols or guidelines listing minimum requirements for comparing
different procedures or methods for cyst detection could assist investigators in producing the
required data for demonstrating acceptability or equivalency of modifications to approved
methods. Consideration should also be given to developing a mechanism for analyst and
laboratory approval processes that might be established to continue the certification program
initiated with the ICR.

With the recent increased emphasis on studies of the effectiveness of water disinfection
and filtration to inactivate and remove Cryptosporidium oocysts, less attention has been paid to
Giardia. The assumption is made that if a disinfectant is sufficient to inactivate
Cryptosporidium that it will also be effective for Giardia . A similar reasoning is applied to
water filtration technologies. A more critical approach should betaken because these two
protozoa have different life cycles and biology. It may not be appropriate to assume that all
disinfection and filtration effectiveness studies conducted for Cryptosporidium will be effective
for Giardia. Whether a technology is equally effective will depend on the mechanism of
VIII-7
-------
inactivation and mechanism of filtration. For example, if the mechanism for removal is primarily
physical straining, as in slow-sand, diatomaceous earth, or membrane filtration technologies, then
Giardia cysts should be removed at least as effectively as Cryptosporidium oocysts, since the
cysts are larger. However, in conventional and direct granular filtration the optimum operating
conditions for removal of Cryptosporidium oocysts may not be the same as for Giardia cyst
removal. Additional research is needed to belter define optimum operating parameters and
coagulant chemicals that are effective for the simultaneous removal of both cysts and oocysts by
these filtration processes. In regard to disinfection, research is needed on the inactivation of
Giardia cysts by the newer UV processes and other electrotechnologies since these technologies
are now being evaluated for Cryptosporidium. Also, the potential forphotoreactivation of UV
inactivated cysts should be examined as well as the potential for damage repair when chemical
disinfectants are used.

Since the information on effectiveness of chemical disinfectants is based on results of
laboratory studies in demand-free water, additional studies should be conducted to compare the
effectiveness of disinfectants under representative conditions in natural waters.

In response to the recent report that an isolate of Flavobacterium can kill Giardia,
research should be encouraged on the development and application of biological control agents
for wastewater and drinking water treatment.
VIII-8
-------
IX. REFERENCES

Abbaszadegan, M., Gerba, C. P., and Rose, J. B. 1991. Detection of Giardia cysts with a cDNA
probe and application to water samples. Appl. Environ. Microbiol., 57(4):927-931.

Abbaszadegan, M., Huber, M.S., Gerba, C. P., and Pepper, I. L. 1997. Detection of viable
Giardia cysts by amplification of heat shock-induced mRNA. Appl. Environ. Microbiol.,
63(l):324-328.

Adam, R.D. 1991. The biology of Giardia. Microbiological Reviews, 55(4):706-732.

Addiss, D.G., Stewart, J.M, Finton, R.J., Wahlquist, S.P., Williams, R.M., Dickerson, J.W.,
Spencer, H.C., and Juranek, D.D. 1991. Giardia lamblia and Cryptosporidium infections in child
day-care centers in Fulton County, Georgia. Pediatr. Infect. Dis. J., 10:907-1011.

Ahmad, R.A., Lee, E., Tan, I. T. L., and Mohamad-Kamel, A. G. 1997. Occurrence of Giardia
cysts and Cryptosporidium oocysts in raw and treated water from two water treatment plants in
Selangor, Malaysia. Water Research, 31(12):3132-3136.

Aggarwal, A. and Nash, T.E. 1986. Lack of cellular cytotoxicity by human mononuclear cells to
Giardia. J. Immunology, 136(9):3486-8.

Aggarwal, A. and Nash, T.E. 1987. Comparison of two antigenically distinct Giardia lamblia
isolates in gerbils. Am. J. Trop. Med. Hyg, 36:325-332.

Akin, E.W. and Jakubowski, W. 1986. Drinking water transmissions of giardiasis in the United
States. Water Sci. Tech., 18(10):219-226.
IX-1
-------
Al-Ani, M.Y., Hendricks, D.W., Logsdon, G.S., and Hi bier, C.P. 1986. Removing Giardia cysts
from low turbidity waters by rapid rate filtration. J. Am. Water Works Assoc., 78(5):66-73.

Aldeen, W.E., Carroll, K.,Robison, A., Morrison, M., and Hale, D. 1998. Comparison of nine
commercially available enzyme-linked immunosorbent assays for detection of Giardia lamblia in
fecal specimens. J. Clin. Microbiol., 36(5): 1338-1340.

Aley, S.B. and Gillin, F.D. 1995. Specialized surface adaptations of Giardia lamblia. Infect.
Agents Dis., 4(3): 161-166.

Alles, A.J., Waldron, M.A., Sierra, L.S., and Mattia, A.R. 1995. Prospective comparison of direct
immunofluorescence and conventional staining methods for detection of Giardia and
Cryptosporidium spp. in human fecal specimens. J. Clin. Microbiol., 33(6): 1632-1634.

Allison, D. 1984. Giardiasis- a recent investigation. Canadian Journal ofPublic Health. Vol. 75.
318-320.

Andrews R.H., Adams, M., Boreham, P.F.L., Mayrhofer, G.,Meloni, B.P. 1989. Giardia
intestinalis: Electrophoretic Evidence for a Species Complex. International Journal for
Parasitology, 19:183-190.

APHA-AWWA-WEF. 1994. 9711 Pathogenic Protozoa. In: Standard Methods for the
Examination of Water and Wastewater, 18th ed. Suppl., Eaton, A. D., Clesceri, L. S., and
Greenberg, A. E., eds., Washington, DC: American Public Health Association, pp. 64-70.
IX-2
-------
APHA-AWWA-WEF. 1995. 9711 Pathogenic Protozoa. In: Standard Methods for the
Examination of Water and Wastewater, 19th ed., Eaton, A. D., Clesceri, L. S., and Greenberg, A.
E., eds., Washington, D.C.: American Public Health Association, pp. 9/110-9/117.

APHA-AWWA-WPCF. 1981. 912 H. Pathogenic Protozoa. In: Standard Methods for the
Examination of Water and Wastewater, 15th ed., Greenberg, A. E., Connors, J. J., and Jenkins,
D., eds., Washington, DC: American Public Health Association, pp. 842-847.

Arashima, Y., Iguchi, K., Kubo, N., Kumasaka, K., Okuyama, K., Kawano, K., Harado, M.,
Shimabukuro, H., Saitoh, T., Isa, H. 1990. Studies on the giardiasis as the zoonosis.
Kansenshogaku-Zasshi, 64:295-298.

Archer, J.R., Ball, J.R., Standridge, J.H., Greb, S.R., Rasmussen, P.W., Masterson, J. P., and
Boushon, L. 1995. Cryptosporidium spp. oocyst andGiardia spp. cyst occurrence, concentrations
and distribution in Wisconsin waters. WR420-95, Madison, WI, Wisconsin Dept. of Natural
Resources, 81 pp.

ARCTECH. 1996. Draft Report: Field Recovery ofGiardia Cysts and Cryptosporidium
Oocysts. Contract No. 68-C5-3909, Work Assignment B-03, U.S. EPA, Office of Water,
Washington, DC.

Asama, R., Hokari, S., Murasugi, E., Arashima, Y., Kubo, N., Kawano, K. 1991. II. Giardiasis in
dogs and cats. Kansenshogaku-Zasshi, 65: 157-161.

ASTM. 1991. P229 proposed test method for Giardia cysts and Cryptosporidium oocysts in low
turbidity water by a fluorescent antibody procedure. In: Annual Book of ASTM Standards,
IX-3
-------
Section 11, Water and Environmental Technology, Vol. 11.02, Philadelphia, PA: American
Society for Testing and Materials, pp. 899-909.

Barnard, RJ. and Jackson, GJ. 1984. Giardia lamblia. The Transfer of Human Infections by
Foods. In: Giardia and Giardiasis, Erlandsen, S. L. and Meyer, E. A., eds., Plenum Press, New
York, pp. 365-378.

Barr, S.C., Bowman, DD., Heller, R.L. 1993. Efficacy of albendazole against giardiasis in dogs.
Am. J. Vet. Res., 54: 926-928.

Barr, S.C., Bowman, DD., Heller, R.L. 1994. Efficacy of fenbendazole against giardiasis in
dogs. Am. J. Vet. Res., 55: 988-990.

Bartlett, A.V., Englender, S.J., Jarvis, B.A.,Ludwig, L., Carlson, J.F., and Topping J.P. 1991.
Controlled trial of Giardia lamblia: control strategies in day care centers. Am. J. Public Health,
81(6):1001-1006.

Bassily, S.Z., Farid, J.W., Mikhail, et al. 1970. The treatment of Giardia lamblia infections with
mepacrine, metronidazole and furazolidone. J. Trop.Med. Hyg., 73:15.

Baveja, U.K., Jyoti, A.S., Kaur, M., Agarwal, D.S., Anand, B.S., Nanda, R 1986. Isoenzyme
studies of Giardia lamblia isolated from human cases. Aust. J. Exp. Biol. Med. Sci., 64:119-126.

Bell, A., Gusparini, R., Meeds, D., Mathias, R.G., Farley, J.D. 1993. A swimming pool-
associated outbreak of cryptosporidiosis in British Columbia. Can. J. PublicHealth, 84(5):334-
337.
IX-4
-------
Bell, C.A., Cory,M., Fairley, T.A., Hall, I.E., Tidwell, R.R. 1991. Structure-activity
relationships of pentamidine analogs against Giardia lamblia and correlation of antigiardial
activity with DNA-binding affinity. Antimicrobial Agents & Chemotherapy, 35(6): 1099-107.

Bellamy, W.D., Cleasby, J.L., Logsdon, G.S., and Alle, MJ. 1993. Assessing treatment plant
performance. J. Am. Waterworks Assoc., 85(12):34-38.

Bemrick W. J. 1961. A note on the incidence of three species of Giardia in Minnesota. J.
Parasitol., 47:87-89.

Benenson A.S. 1995. Giardiasis. Control of Communicable Disease in Man. Sixteenth Edition,
American Public Health Association, Washington, DC .

Bennett, J.V., Holmberg, S.D., Rogers, M.F., Solomon, S.L. 1987. In: Closing the Gap: The
Burden of Unnecessary Illness, Amler, R.W. and Dull, H.B., eds., Oxford Univ. Press, New
York, 102-114.

Berkelman, R.L., Bryan, R.T., Osterholm, M.T., Leduc, J.W., Hughes, J.M. 1994. Infectious
disease surveillance: a crumbling foundation. Science, 264:368.

Bertolucci, G. C., Gilli, G., Carraro, E., Giacosa, D., and Puppo, M. 1998. Influence of raw water
storage on Giardia, Cryptosporidiumandnematodes. Water Sci. Tech. 37(2):261-267.

Bielec, L., Boisvert, T.C., and Jackson, S.G. 1996. Modified procedure for recovery of Giardia cysts
from diverse water sources. Lett. Appl. Micro. 22:21-25.
IX-5
-------
Bier, J. 1991. Isolation of parasites on fruits and vegetables. Southeast Asian Journal of Tropical
Medicine and Public Health, 22: 144-145.

Bifulco, J.M. and Schaefer, F.W. 1993. Antibody-magnetite method for selective concentration of
Giardia lamblia cysts from water samples. Appl. Environ. Microbiol., 59:772-776.

Bingham, A.K., Jarroll, E.L., Meyer, E.A., Radulescu, S. 1979. Giardia sp.: physical factors of
excystation in vitro and excystation vs eosin exclusion as determinants of viability. Exp. Parasitol.,
47:284-291.

Bingham, A.K. and Meyer, E.A. 1979. Giardia excystation can be induced in vitro in acidic
solutions. Nature, 277(5694):301-302.

Birkhead, G. and Vogt, R.L. 1989 Epidemiological surveillance for endemic Giardia lamblia
infection in Vermont. Amer. J. Epidemiol., 129:762.

Birkhead, G., Janoff, E.N, Vogt,R.L., and Smith, P.O. 1989. Elevated levels of immunoglobulin A
to Giardia lamblia during a waterborne outbreak of giardiasis. Journal of Clinical microbiology.
27(8):1707-1710.

Blaser, M.J. and L.S. Newman. 1982. A review of human salmonellosis: I. infective dose. Rev.
of Infect. Dis., 4:1096-1106.

Boreham, P.P., Phillips, R.E, Shepherd, R.W. 1987. Heterogeneity in the responses of clones of
Giardia intestinalis to anti-giardial drugs.
IX-6
-------
BrumptL. 1937. Traitement experimental de lalambliase. C. R. Seances Soc. Biol. 124:1040-
1042.

Bulut, B.U., Gulnar, S.B., and Aysev, D. 1996. Alternative treatment protocols in giardiasis: a
pilot study. Scand J Infect Dis 28:493-495.

Butcher, P.O., Cevallos, A.M., Carnaby, S., Alstead, E.M., Swarbrick, E.T., Farthing, M.J.G..
1994. Phenotypic and genotypic variation in Giardia lamblia isolates during chronic infection.
Gut, 35:51-54.

Campbell, J.D. and Faubert, G.M. 1994. Comparative studies on Giardia lamblia encystation in
vitro and in vivo. Jour. Parasitol., 80:36-44.

Carroccio, A., Montalto, G., lacono, G., Ippolito, S., Soresi, M, Notarbartolo, A. 1997.
Secondary impairment of pancreatic function as a cause of severe malabsorption in intestinal
giardiasis: a case report. Am J. Trop. Med. Hyg., 56(6): 599-602.

Casson, L.W., Sorber, C.A., Sykora, J.L, Gavaghan, P.O., Shapiro, M.A., Jakubowski, W. 1990.
Giardia in wastewater—effect of treatment. J. Water Poll. Control Feder., 62(5):670-675.

Casterline, I.E., Allen, L.H., Ruel, M.T. 1997. Vitamin B-12 deficiency is very prevalent in
lactating Guatemalan women and their infants at three months postpartum. The Journal of
Nutrition, 127:1966-72.

Char, S., Shetty, N., Narasimha, M., Elliott, E., Macaden, R., and Farthing, M.J.G. 1991. Serum
antibody response in children with Giardia lamblia infection and identifection of an
immunodominant 57-kilodalton antigen. Parasite Immunology, 13. 329-337.
IX-7
-------
Charles, G., Morgan, J. M., MacPhee, M., Kim, M., and Fredericksen, D. 1995. Bench-scale
parasite spiking - An alternative to pilot-scale Giardia and Cryptosporidium spiking
investigations. In: Water Quality Technology Conference, American Waterworks Association,
Denver, CO, pp. 37-47.

Chauret, C., Armstrong,N., Fisher, J., Sharma, R., Springthorpe, S., and Sattar, S. 1995.
Correlating Cryptosporidium and Giardia with microbial indicators. J. Am. Waterworks Assoc.,
87(ll):76-84.
Chavalittamrong, B., Charoenvidhya, S.,Tuchinda, P., Suntornpoch, V., Chearskul, S. 1978.
Prevalence of Giardia lambliain children attending an out-patient department of Sirirj Hospital.
SE Asian J. Trop. Med. Public Health, 9:51-54.

Chavez, B., Gonzalez-Mariscal, L., Cedillo-Rivera, R.,and Martinez-Palomo, A 1995. Giardia
lamblia: in vitro cytopathic effect of human isolates. Experimental Parasitology, 80:133-138.

Chorba, T.L., Berkelman, R.L., Safford, S.K., Gibbs, N.P., and Hull, H.F. 1989. Mandatory
reporting of infectious diseases by clinicians. J. Am. Water Works Assoc., 262(21):3018-3026.

Chute, C., Smith, R., and Baron, J. 1985. Risk factors for endemic giardiasis, Am. J. Epidmiol.,
122:515.

Chute, C.G., Smith, R.P., and Baron, J.A. , 1987. Risk factors for endemic giardiasis, Am. J.
Public Health, 77 (5): 585-87.
-------
Clancy, J.L., Gollnitz, W.D., and Tabib, Z. 1994. Commercial labs: how accurate are they? J.
Am. Water Works Assoc., 86(5):89-97.

Clark, J.T. and Holberton, D.V. 1986. Plasma membrane isolated from Giardia lamblia.
identification of membrane proteins. European J. Cell Biology, 42(2):200-6.

Clark, R.M. Read, E.J., Hoff, J.C. 1989. Analysis of inactivation of Giardia lambliaby chlorine.
Journal of Environmental Engineering, 115(1):80-90.

Clark, R.M. 1990. Modeling inactivation of Giardia lamblia. Journal of Environmental
Engineering, 116(5)837-853.

Cleasby, J.L., Hilmoe, D.J., and Dimitracopoulos, C.J. 1984. Slow sand and direct in-line
filtration of a surface water. J. Am. Water Works Assoc. 44-55.

Cody, M. M., Sottnek, H. M., and O'Leary, V. S. 1994. Recovery of Giardia lamblia cysts from
chairs and tables in child day-care centers. Pediatrics, 94(6 Pt 2): 1006-1008.

Conboy, G. 1997. Giardia. Can. Veterinary J., 38(4):245-7.

Conley, M.E., and Delacroix, D.L. 1987. Intravascular and mucosal immunoglobulin A: Two
separate but related systems of immune defense? Annals of Internal Medicine, 106:892-899.

Consonery, P. J., Greenfield, D. N., and Lee, J. J. 1997. Evaluating and optimizing surface water
treatment plants: how good is good enough? Pennsylvania Department of Environmental
Protection, Harrisburg, PA.
IX-9
-------
Cornell University Whole Farm Planning Scientific Support Group. 1993. Cornell University
First Quarter Progress Report For the Period January 26-April 26, 1993; Phase I Watershed
Agricultural Program. New York State Water Resources Institute Center for the Environment,
Ithaca, New York, 103 pp.

Corsi, A., Nucci, C., Knafelz, D., Bulgarini, D., Di lorio, L., et al. 1998. Ocular changes
associated with Giardia lamblia infection in children. British J. Ophthalmology, 82:59-62.

Crabtree, K. D., Ruskin,R. H., Shaw, S. B., and Rose, J. B. 1996. The detection of
Cryptosporidium oocysts and Giardia cysts in cistern water in the U.S. Virgin Islands. Water
Research, 30(1):208-216.

Craft, J.C. 1982. Experimental infection with Giardia lambliam rats. J. Infect. Dis., 145(4): 495-
498.

Craft, J.C., Murphy, T., Nelson, J.D. 1981. Furazolidone and quinacrine. Am. J. Dis. Children,
135:164-166.

Craun, G.F. 1986. Statistics of waterborne outbreaks in the U.S. (1920-1980), In: Waterborne
Diseases in the United States, Craun, G.F., ed., CRC Press, Inc., Boca Raton, FL, pp. 73-160.

Craun, G.F. 1990. Waterborne giardiasis. In: Human Parasitic Diseases, Vol. 3, Giardiasis, E. A.
Meyer, ed., Elsevier Science Publishers, Amsterdam, pp. 267-293.

Craun, G.F. 1996. Waterborne disease in the United States. In: Water Quality in Latin America:
Balancing the Microbial and Chemical Risks in Drinking Water Disinfection, Craun, G.F., ed.,
ILSI Press, Washington, DC, pp. 55-77.
IX-10
-------
Craun, G.F., and Calderon, R.L., 1997. Microbial risks in groundwater systems epidemiology of
waterborne outbreaks. In: Under the Microscope Examining Microbes in Groundwater, AWWA
Research Foundation, Amer. Water Works Assoc., Denver CO, pp. 9-20.

Crockett, C.S. and Haas, C.N. 1997. Understanding protozoa in your watershed. J. Am. Water
Works Assoc., 89(9):62-73.

Crossley, R., Holberton D.V. 1985. Assembly of 2.5 nm filaments fromgiardin, a protein
associated with cytoskeleton microtubules mGiardia. J. Cell Sci., 78:205-231.

Crouch, A. A., Seow, W.K., Thong, Y.H. 1986. Effect of twenty-three chemotherapeutic agents
on the adherence and growth of Giardia lamblia in vitro. Transactions Royal Society of Tropical
Medicine and Hygiene, 80(6):893-6.

Crouch, A.A., Seow, W.K., Whitman, L.M., Thong, Y.H. 1990. Sensitivity in vitro of Giardia
intestinalis to dyadic combinations of azithromycin, doxycycline, mefloquine, tinidazole and
furazolidone. Transactions Royal Society Tropical Medicine and Hygiene, 84(2):246-8.

Danciger, M., Lopez, M. 1975. Numbers of Giardia in the feces of infected children. Am. J.
Trop. Med. Hyg., 24:237-242.

Danielson, R. E., Cooper, R. C., and Riggs, J. L. 1995. Giardia and Cryptosporidium analysis: a
comparison of microscopic and flow cytometric techniques. In: Proceedings 1995 Water Quality
Technology Conference. American Water Works Association, Denver, CO, pp. 1673-1685.
IX-11
-------
Das, S., Reiner, D.S., Zenian, 1, Hogan, D.L., Koss, M.A., Wang, C-S., Gillin, F.D. 1988.
Killing ofGiardia lamblia trophozoites by human intestinal fluid in vitro. J. Infect. Dis.,
157:1257-1260.

Davies, R.B. andHibler, C.P. 1979. Animal reservoirs and cross-species transmission of
Giardia. In: Jakubowski, W. andHoff, J.C., eds., Waterborne Transmission of Giardiasis.
Cincinnati: U.S. Environmental Protection Agency (EPA-600/9-79-001), pp. 104-126.

Deng, M. Y. and Cliver, D. O. 1992. Degradation ofGiardia lamblia cysts in mixed human and
swine wastes. Appl. Environ. Microbiol., 58(8):2368-2374.

Dennis, D.T., Smith, R.P., Welch, J.J., Chute, C.G., Anderson, B., Herndon, J.L., von Reyn, C.F.
1993. Endemic gjardiasis in New Hampshire: a case-control study of environmental risks. J.
Infect. Dis., 167(6): 1391-5.

deRegnier, D.P., Cole, L., Schupp, D.G., and Erlandsen, S.L. 1989. Viability ofGiardia cysts
suspended in lake, river, and tap water. Appl. Environ. Microbiol., 55(5): 1223-1229.

Deselliers, L.P., Tan, D.T.M., Scott, R.B., Olson, M.E. 1997. Effects of Giardia lamblia
infection on gastrointestinal transit and contractility in Mongolian gerbils. Digestive Diseases
and Sciences, 42(12):2411-19.

Dixon, B. R., Parenteau, M., Martineau, C., and Fournier, J. 1997. A comparison of conventional
microscopy, immunofluorescence microscopy and flow cytometry in the detection ofGiardia
lamblia cysts in beaver fecal samples. J. Immunol. Methods, 202(l):27-33.

Dobell, C. 1920. The discovery of the intestinal protozoa of man. Proc. R. Soc. Med., 13:1-15.
IX-12
-------
Donowitz, M., Kokke, F.T., Saidi, R. 1995. Evaluation of patients with chronic diarrhea. New
England J. Medicine, 332 (ll):725-9.

Dutta, A.K., Phadke, M.A., Bagade, A.C., Joshi, V., Gazder, A., Biswas, T.K., Gill, H.H., Jagota,
S.C. 1994. A randomized multicentre study to compare the safety and efficacy of albendazole
and metronidazole in the treatment of giardiasisin children. Indian Journal of Pediatrics,
61(6):689-93.

Edlind, T.D., Hang, T.L., Chakraborty, P.R. 1990. Activity of the anthelmintic benzimidazoles
against Giardia lamblia in vitro. Journal of Infectious Diseases. 162(6): 1408-11.

Eisenberg, J.N., Seto, E.Y.W., Olivieri, A.W. and Spear, R.C. 1996. Quantifying Water Pathogen
Risk in an Epidemiological Framework. Risk Analysis, 16:549-563.

Elliott, A.M., Klaus, B.D., North, D.S., Martin, H.P. 1998. Furazolidone-induced mood disorder
during the treatment of refractory giardiasis in a patient with AIDS. Clinical Infectious diseases,
26:1015.

Ellis, J.E., Wingfield, J.M., Cole,D., Boreham, P.F.L., Lloyd, D 1993. Oxygen affinities of
metronidazole-resistant and-sensitive stocks of Giardia intestinalis. Int. J. Parasitology, 23 (1):
35-39.

Enriquez, V., Rose, J. B., Enriquez, C. E., and Gerba, C. P. 1995. Occurrence of
Cryptosporidium and Giardia in secondary and tertiary wastewater effluents. In: Protozoan
Parasites and Water. Betts, W. B., Casemore, D., Fricker, C. R., Smith, H, and Watkins, J., eds.,
The Royal Society of Chemistry, Thomas Graham House, Science Park, Cambridge, Great
Britain, pp. 84-86.
ix-1:
-------
Engelkirk, P.G. and Pickering, L.K. 1990. Detection ofGiardiaby immunologicalmethods. In:
Human Parasitic Diseases, Vol. 3, Giardiasis, E.A Meyer, ed.,Elsevier Science Publishers,
Amsterdam, pp. 187-198.

Erlandsen, S.L. 1994. Biotic transmission-Is giardiasis a zoonosis? In: Giardia. From Molecules
to Disease, R.C.A. Thompson, J.A. Reynoldson, and AJ. Lymbery, eds., CAB
INTERNATIONAL, Wallingford, U.K., pp. 83-97.

Erlandsen, S.L. andBemrick, WJ. 1988a. Waterborne giardiasis: sources of Giardia cysts and
evidence pertaining to their implications in human infection. In: Advances in Giardia Research,
Wallis, P. M. and Hammond, B. R., eds., University of Calgary Press, Calgary, pp. 227-236.

Erlandsen, S.L., Sherlock, L.A., Januschka, M., Schupp, D.G., Schaeffer, F.W., Jakubowski, W.,
and Bemrick, WJ. 1988b. Cross-species transmission of Giardia spp: inoculation of beavers and
muskrats with cysts of human, beaver, mouse, and muskrat origin.

Erlandsen, S.L., Bemrick, W.J., Schupp, D.E., Shields, J.M., Jarroll, E.L., Sauch, J.F., and
Pawley, J.B. 1990a. High-resolution immunogold localization of Giardia cyst wall antigens
using field emission SEM with secondary andbackscatter electron imaging. The Journal of
Histochemistry and Cytochemistry. 38(5):625-632.

Erlandsen, S.L., Bemrick, W.J., Wells, C.L., Feely, D.F., Knudson, L., Campbell, S.R., Keulan,
H.V., and Jarroll, E.L. 1990b. Axenic culture and characterization of Giardia ardeae from the
great blue heron. J. Parasitol. 76(5):717-724.

Erlandsen, S. L., Sherlock, L. A., Bemrick, W. J., Ghobrial, H., and Jakubowski, W. 1990c.
Prevalence of Giardia spp. in beaver and muskrat populations in northeastern states and
IX-14
-------
Minnesota: detection of intestinal trophozoites at necropsy provides greater sensitivity than
detection of cysts in fecal samples. Appl. Environ. Microbiol., 56(l):31-36.

Erlandsen, S. L., Sherlock, L. A., and Bemrick, W. J. 1990d. The detection of Giardia muris and
Giardia lamblia cysts by immunofluorescence in animal tissues and fecal samples subjected to
cycles of freezing and thawing. J. Parasitol., 76(2):267-271.

Erlandsen, S.L., van Keulen, H., Brelje, T., Gurien, A., Jakubowski, W., Schaefer, F.W., Wallis,
P.M., Feely, D.E, and Jarroll, E.L. 1994. Molecular approach to the speciation and detection of
Giardia: fluorochrome-rDNA probes for identification of Giardia lamblia, Giardia muris, and
Giardia ardeae in laboratory and environmental samples by in situ hybridization. In: Giardia:
From Molecules to Disease, R.C.A. Thompson, J.A. Reynoldson, and AJ. Lymbery, eds., CAB
INTERNATIONAL, Wallingford, U.K., pp. 64-65.

Esrey, S.A., Collett, J., Miliotis, M.M., Koornhof, J., Makhale, P. 1989. The risk of infection
from Giardia lamblia due to drinking water supply, use of water, and latrines among preschool
children in rural Lesotho. Int. J. Epidemiol., 18(l):248-253.

Farbey, M.D., Reynoldson, J.A., Thompson, R.C. 1995. In vitro drug susceptibility of 29 isolates
of Giardia duodenalis from humans as assessed by an adhesion assay. International Journal for
Parasitology, 25(5):593-9.

Farthing, M.J.G. 1989. Host-parasite interactions in human giardiasis. Quarterly Journal of
Medicine, 70(263): 191-204.
IX-15
-------
Farthing, M.J.G. 1994. Giardiasis as a disease. In: Giardia: From Molecules to Disease, R.C.A.
Thompson, J.A. Reynoldson, and AJ. Lymbery, eds., CAB INTERNATIONAL, Wallingford,
U.K., pp. 15-37.

Farthing, M.J.G. 1996. Giardiasis. Gastroenterology Clinics of North America, 25(3):493-515.

Payer, R., Graczyk, T.D., Lewis, E.J., Trout, J.M., and Farley, C.A. 1998. Survival of infectious
Cryptosporidium parvum in seawater and eastern oysters (Crassostrea virginica) in the
Chesapeake Bay. Appl. Environ. Microbiol., 64(3): 1070-1074.

Ferguson, A., Gillon, J., Munro, G. 1990. Pathology and pathogenesis of the intestinal mucosal
damage in giardiasis. In: Human Parasitic Diseases, Vol. 3, Giardiasis, E.A. Meyer, ed., Elsevier
Science Publishers, Amsterdam, pp. 155-173.

Filice, F.P. 1952. Studies on the cytology and life history of a Giardia from the laboratory rat. U.
Calif. Publ. Zool. 57(2):53-146.

Finch, G.R., Black, E.K., Labatiuk, C.W., Gyurek, L., and Belosevic, M. 1993. Comparison f
Giardia lamblia and Giardia muris cyst inactivation by ozone. Applied and Environmental
Microbiology, 59(11):3674-3680.

Fogel, D., Isaac-Renton, J., Guasparini, R., Moorehead, W., and Ongerth, J. 1993. Removing
Giardia and Cryptosporidium by slow sand filtration. Am. Water Works Assoc., 86(11):77-83.

Foster, L.R. 1990. Surveillance for waterborne illness and disease reporting: state and local
responsibilities, in Craun, G.F. ed., Methods for the Investigation and Prevention of Waterborne
Disease Outbreaks, U.S. Environmental Protection Agency, Cincinnati, OH, p. 39-43.
IX-16
-------
Fraser, G.G. and Cooke, K.R. 1991. Endemic Giardiasis and municipal water supply. Am. J.
Public Health, 81: 760.
Fredericksen, D.W., Tabib, Z., Boutros, S., Cullen, W., Dittmer, R., McCarthy, L., and McCuin,
R. 1995. Examination of the slide technique vs membrane technique to recover Giardia lamblia
cysts and Cryptosporidium parvum oocysts. In: Proceedings 1995 Water Quality Technology
Conference. American Waterworks Association, Denver, CO, pp. 837-845.

Freeman, C.D., Klutman,N.E., and Lamp, K.C. 1997. Metronidazole. Drugs, 54(5):679-708.

Friend, D.S. 1966. The fine structure of Giardia muris. J. Cell Biol., 29(2):317-332.

Frost, F., Hartner, L., Plan, B., Fukutaki, K., Holman, B. 1983. Giardiasis in Washington State.
Project Summary. U.S. Environmental Protection Agency (EPA-600/S1-82-016), Research
Triangle Park, NC.

Frost, F., Craun, G., Calderon, R., 1996. Waterborne disease surveillance: what is it and is it
necessary? J. Am. Waterworks Assoc., 88(9):66-75.

Gahndi, B.M, Buch, P., Sharma, M.P., Irshad, M., and Samantray, S.C. 1989. Lancet, 2 (Sept.
16:685.

Galli-Valerio, B. 1937. La lambliase et son traitement par 1'atebrine. Schweiz. Med.
Wochenschr., 67:1181-1182.
IX-17
-------
Garcia, L.S. and Shimizu, R.Y. 1997. Evaluation of nine immunoassay kits (enzyme
immunoassay and direct fluorescence) for detection of Giardia lamblia and Cryptosporidium
parvum in human fecal specimens. J. Clin. Microbiol., 35(6): 1526-1529.

Geldreich, E.E. and Craun, G.F. 1996. Multiple barriers for protection and treatment of drinking
water supplies: a proven method for preventing transmission of waterborne disease. In: Water
Quality in Latin America: Balancing the Microbial and Chemical Risks in Drinking Water
Disinfection, Craun, G.F., ed., ILSI Press, Washington, DC, pp. 1-5.

Gibson, C.J., Stadterman, K.L., States, S. and Sykora, J. 1998. Combined Sewer Overflows: A
Source of Cryptosporidium and Giardia. In:

Gillin, F.D., Reiner, D.S., Gault,MJ., Douglas, H., Das, S., Wunderlich, A., and Sauch, J. 1987.
Encystation and expression of cyst antigens by Giardia lamblia in vitro. Science. 235:1040-1043.

Gillin, F.D. 1988. Panel discussion on excystation and encystation. In: Advances in Giardia
Research, Wallis, P. M. and Hammond, B. R, eds., University of Calgary Press, Calgary, p. 273.

Gillin, F.D., Boucher, S.E., Rossi, S.S., and Reiner, D.S. 1989. Giardia lamblia: the roles of bile,
lactic acid, and ph in the completion of the life cycle in vitro. Experimental Parasitology, 69:164-
174.

Gillin, F.D., Das, S., Reiner, D.S. 1990. Nonspecific defenses against human Giardia. In: Human
Parasitic Diseases, Vol. 3, Giardiasis, E.A. Meyer, ed., Elsevier Science Publishers, Amsterdam, pp.
199-213.
IX-18
-------
Oilman, R.H., Brown, K.H., Visvesvara, G.S., Mondal, G., Greenberg, B., Sack, R.B., Brandt, F.,
and Khan, M.U. 1985. Epidemiology and serology of Giardia lamblia in a developing country:
Bangladesh. Transactions Royal Society of Tropical Medicine and Hygiene, 79: 469-473.

Gilman, R.H., Marquis, G.S., Miranda, E., Vestegui, M., Martinez, H. 1988. Rapid reinfection by
Giardia lamblia after treatment in a hyperendemic third world community. Lancet, 343-345.

Goka, A.K.J., Rolston, D.D.K., Mathan, V.I., and Farthing, M.J.G. 1986. Diagnosis of giardiasisby
specific IgM antibody enzyme-linked immunosorbent assay. Lancet, 2:184-186.

Gordts, B., Hemelhof, W., Asselman, C., Butzler, J.P. 1985. In vitro susceptibilities of25 Giardia
lamblia isolates of human origin to six commonly used antiprotozoal agents. Antimicrobial Agents
and Chemotherapy, 28(3):378-80.

Grant, D.R. and Woo, P.T.K. 1978a. Comparative studies of Giardia spp. in small mammals in
southern Ontario. I. Prevalence and identity of the parasites with a taxonomic discussion of the
genus. Can. J. Zool., 56:1348-1359.

Grant, D.R. and Woo, P.T.K. 1978b. Comparative studies of Giardia spp. in small mammals in
southern Ontario, n. Host specificity and infectivily of stored cysts. Can. J. Zool., 56:1360-66.

Grant, D.R. and Woo, P.T.K. 1979. Comparative studies of Giardia spp. in small mammals in
southern Ontario. HI. Duration and cyst production in natural and experimental infections. Can. J.
Zool., 57:307-313.

Gray, S.F., Gunnel, D.J., Peters, TJ. 1994. Risk factors for giardiasis: a case-control studyin Avon
and Somerset. Epidem. Infect, 113(1):95-102.
IX-19
-------
Greensmith, C.T., Stanwick, R.S., Elliot, B.E., Fast, M.V. 1988. Giardiasis associated with the use
of a water slide. Pediatric Infectious Dis. J., 7(2):91-91.

Grimason, A.M., Wiandt, S., Baleux, B, Thitai, W.N., Bontoux, J., and Smith, H.V. 1996.
Occurrence and removal of Giardia sp. cysts by Kenyan and French waste stabilization pond
systems. Water Sci. Tech., 33(7):83-89.

Gyorkos, T. 1983. Estimation of parasite prevalence based onsubmissions to provincial laboratories.
Can. J. Public Health, 74(7)281-284.

Haas, C.N. 1983. Estimation of risk due to low doses of microorganisms: A comparison of
alternative methodologies. American Journal Epidemiology, 118 (4):573-582.

Haas, C.N. and Heller, B. 1990. Kinetics of inactivation of Giardia lambliaby free chlorine. Water
Res., 24(2):233-238.

Haas, C.N. and Joffe, J. 1994. Disinfect! on under dynamic conditions: modification of Horn's model
for decay. Environ. Sci. Technol., 28(7): 1367-1369.

Haas, C.N, Rose, J.B., Gerba, C.,. Regli, S. 1993. Risk Assessment of Virus in Drinking Water.
Risk Analysis, 13:545-552.

Haas, C.N., Hornberger, J.C., Anmangandla, U., Heath, M., Jacangelo, J.G. 1994. A volumetric
method for assessing Giardia inactivation. J. Am. Water Works Assoc., 86(2): 115-120.
IX-20
-------
Haas, C.N., Joffe, J., Anmangandla, U., Jacangelo, J.G., Heath, M. 1996. Water quality and
disinfection kinetics. J. Am. Waterworks Assoc., 89(3): 95-103.

Haas, C.N, J.B. Rose, C.P. Gerba, (eds) 1998. Quantitative Microbial Risk Assessment, John
Wiley and Sons, New York, NY.

Hall, A. 1994. Giardia infections: epidemiology and nutritional consequences. In: Giardia: From
Molecules to Disease, R.CA. Thompson, J.A. Reynoldson, and AJ. Lymbery, eds., CAB
INTERNATIONAL, Wallingford, U.K., pp. 251-279.

Hall, A. and Nahar, Q. 1993. Albendazole as a treatment for infections with Giardia duodenalis
in children in Bangladesh. Trans. Royal Soc. Trop. Medicine Hyg., 87:84-86.

Hancock, C.M., Rose, J.B., and Callahan, M. 1997. Cryptoposidium and Giardia in U.S.
groundwater. J. Am. Waterworks Assoc., 90(3)58-61.

Haralabidis, S.T. 1984. Immunodiagnosis of giardiasis by ELISA and studies on cross-reactivity
between the anti-Giardia lamblia antibodies and some parasitic antigens and fractions. Ann.
Trop. Med. Parasitol., 78,295-300.

Hassan, M.M., Farghaly, A.M., Darwish, R. A., Shoukrany, N., el-Hayawan, I. A., and Nassar,
A.K. 1995. Detection of Giardia antigen in stool samples before and after treatment. J. Egypt.
Soc. Parasitol., 25(1): 175-182.

Harter, L., Frost, P., Jakubowski, W. 1982. Giardia prevalence among l-to-3-year-old children in
two Washington State counties. Am. J. Public Health, 72(4)386-388.
IX-21
-------
Harter, L., Frost, F., Gruenenfelder, G., et al. 1984. Giardiasis in an infant and toddler swim
class. Am. J. Public Health, 74:155-156.

Harter L., Frost, F., Vogt, R., Little, A.A., Hopkins, R., Gaspard, B., Lippy, B.C. 1985. A three
state study of waterbome disease surveillance techniques. Am. J. Public Health, 75: 1327.

Hernell, O., Ward, H., Blackberg, L., Pereira, M.E.A. 1986. Killing of Giardia lambliaby
human milk. Infect. Immun., 47:619-622.

Hewan-Lowe, K., Furlong, B., Sims, M. 1997. Coinfection with Giardia lamblia and
Enterocytozoon bieneusi in a patient with acquired immunodeficiency syndrome and chronic
diarrhea. Arch. Pathol. Lab. Med., 121(^:417-422.

Heyworth, M.F and Pappo, J. 1990. Recognition of a 30,000 mw antigen of Giardia muris
trophozoitesby intestinal IgA from Giardia-infected mice. Immunology, 70:535-539.

Heyworth, M.F., Owen, R.L, Jones, A.L 1985. Comparison of leukocytes obtained from the
intestinal lumen of Giardia-infecied immunocompetent mice and nude mice. Gastroenterology,
89:1360-65.

Heyworth, M.F., Kung, I.E., Eriksson, E.G. 1986. Clearance of Giardia muris infection in mice
deficient in natural killer cells. Infect. Immun., 54:903-904.

Heyworth, M.F., Carlson, J.R., Ermak, T.H. 1987. Clearance of Giardia muris infection requires
helper/inducer T lymphocytes. J. Exp. Med., 165:1743-1798.
IX-22
-------
Hibler, C. P. 1988. Analysis of municipal water samples for cysts ofGiardia. In: Advances in
Giardia Research, Wallis, P. M. and Hammond, B. R., eds., University of Calgary Press,
Calgary, pp. 237-245.

Hill, D.R. and Pearson, R.D. 1987. Ingestion ofGiardia lamblia trophozoites by human
mononuclear phagocytes. Infect. Immun., 55:3155-3161.

Hirata, T. and Hashimoto, A. 1997. A field survey on occurrence ofGiardia cysts and
Cryptosporidium oocysts in sewage treatment plants. In: 1997 International Symposium on
Waterborne Cryptosporidiosis Proceedings. Fricker, C. R., Clancy, J. L., and Rochelle, P. A.,
eds., American Water Works Association, Denver, CO, pp. 183-193.

Ho, B.S.W., Tarn, T.-Y., Hutton, P., and Yam, W. C. 1995. Detection and enumeration of
Giardia cysts in river waters of Hong Kong by flocculation-percoll/sucrose gradient
immunofluorescence method. Water Sci. Tech., 31(5-6):431-434.

Ho, B.S.W., Tarn, T.-Y. 1998. Giardia and Cryptosporidium in sewage contaminated river
waters. Water Research, 32(9): 2860-64.

Hoff, J.C. 1986 Inactivation of microbial agents by chemical disinfection. U.S. Environmental
Protection Agency (EPA-600/2-86-067), Cincinnati, OH.

Hoff, J.C., Rice, E.W., and Schaefer, F.W. 1985. Comparison of animal infectivity and
excy station as measures ofGiardia muris cyst inactivation by chlorine. Appl. Environ.
Microbiol., 50(4): 1115-1117.
-------
Hofmann R. and Andrews, R.C. 1995. Factors influencing disinfection by-product formation
when disinfecting to Giardia., virus, and Cryptosporidium inactivation requirements. Proc. 1995
Water Qual. Technol. Conf, New Orleans, Am. Waterworks Assoc., Denver, CO, pp. 483-501.

Hoffman, R.M., Standridge, J.H., Prieve, A.F., Cucunato, J.C., and Bernhardt, M. 1997. Using
flow cytometry to detect protozoa. J. Am. Water Works Assoc., 89(9): 104-111.

Homan, W.L., van Enckevort, F.H.J., Limper, L., van Eys, G.J.J.M., Schoone, G.J., Kasprzak,
W., Majewska, A.C., van Knapen, F. 1992. Comparison of Giardia isolates from different
laboratories by isoenzyme analysis and recombinant DNA probes. Parasitology Research,
78:316-323.

Hopkins, D.P., Brady, G., Drabkin, P., Ackman, D., Morse, D., Burke, M. 1995. An outbreak of
giardiasis associated with a filtered water supply, New York. NY Depart, of Health, Albany, NY.

Hopkins, R.M., Meloni, B.P., Groth, D.M., Wetherall, J.D. 1997. Ribosomal RNA sequencing
reveals differences between the genotypes of Giardia isolates recovered from humans and dogs
living in the same locality. Jour. Parasitol., 83(1):44-51.

Hopkins, R.S. and Juranek, D.D 1991. Acute giardiasis: an improved clinical case definition for
epidemiologic studies. Am. J. Epidemiol., 133:402-407.

Hopkins, R.S., Shillam, P., Gaspard, B., Eisnach, L, and Karlin, RJ. 1985. Waterborne disease
in Colorado: three years' surveillance and 18 outbreaks. Am. J. Public Health, 75(3):254-257.

Hopping, J. 1993. Risk Assessment in the Federal Government: Questions and Answers. Center
for Risk Analysis, Harvard School of Public Health, Boston, MA ACER Life Systems, I. 1994.
IX-24
-------
Final Draft for the Drinking Water Criteria Document on Giardia. Contract No. 68-01-6750,
Prepared for U.S. EPA, Office of Drinking Water.

Horn, J.B., Hendricks, D.W., Scanalan, J.M., Rozelle, L.T., and Trnka, W.C. 1988. Removing
Giardia cysts and other particles from low turbidity waters using dual stage filtration. J. Am.
Waterworks Assoc., 80(2):68-77.

Hoskins, L.C., Winawer, S.J., Broitman, S.A., Gottlieb, L.S., Zamcheck, N. 1967. Clinical
giardiasis and intestinal malabsorption. Gastroenterol., 53(2):265-279.

Howard, L.H., Fink, D.S., Lubin, J., Robinson, MJ. 1995. Giardiasis diagnosed by biopsy of the
colon and terminal ileum: Unusual sites for a common pathogen. Am. J. Gastroenterol., 90:1011-
1013.

Hunter, P. 1997. Giardiasis. Waterborne Disease Epidemiology and Ecology. Wiley, New York,
69-73.

ICAIR, Life Systems, Inc., 1984, Criteria Document on Giardia, U.S. EPA, Washington, DC.

ILSI, International Life Sciences Institute, 1996. A conceptual framework to assess the risks of
human disease following exposure to pathogens. Risk Analysis, 16:841-848.

lonas, G, Farrant, K J., McLenachan, P.A., Clarke, J.K., and Brown, T.J. 1997. Molecular (PCR)
differentiation of Giardia muris and Giardia intestinalis. Int. J. Environ. Hlth. Res., 7:63-69.
IX-25
-------
Isaac-Renton, J. L. 1994. Giardiasis in British Columbia: Studies in an area of high endemicity in
Canada. In: Giardia: From Molecules to Disease, R.C.A. Thompson, J.A. Reynoldson, and AJ.
Lymbery, eds., CAB INTERNATIONAL, Wallingford, U.K., pp. 123-'124.

Isaac-Renton, J.L. and Philion, JJ. 1992. Factors associated with acquiring giardiasis in British
Columbia residents. Can. J. Public Health, 83(2):155-158.

Isaac-Renton, J.L, Moricz, M.M., and Proctor, E.M. 1987. A Giardia survey of fur-bearing
water mammals in British Columbia, Canada. Journal of Environmental Health, 50(2):80-82.

Isaac-Renton, J.L, Lewis, L.F., Ong, C.S.L., andNulsen, M.F. 1994. A second community
outbreak of waterbome giardiasis in Canada and serological investigation of patients.
Transaction Royal Society of Tropical Medicine and Hygiene, 88:395-399.

Isaac-Renton, J., Moorehead, W., and Ross, A. 1996. Longitudinal studies of Giardia
contamination in two community drinking water supplies: cyst levels, parasite viability, and
health impact. Appl. Environ. Microbiol., 62(l):47-54.

Islam, A., Stoll, B.J., Ljungstrom, I Biswas, J., Nazrul, H., Huldt, G. 1983. Giardia lamblia
infections in a cohort of Bangladeshi mothers and infants followed for one year. J. Paediatrics,
103:996-1000.

Istre, G.R., Dunlop, T.S., Gaspard, G.B., and Hopkins, R.S. 1984. Waterborne giardiasis at a
mountain resort: evidence for acquired immunity. Am. J. Public Health, 74(6):602-604.
IX-26
-------
Jacangelo, J.G., Laine, J-M., Cams, K.E., Cummings, E.W., Mallevialle, J. 1991. Low-pressure
membrane filtration for removing Giardia and microbial indicators. J. Am. Water Works Assoc.,
83(9):97-106.

Jacangelo, J.G., Adham, S.S., and Laine, J.M. 1995. Mechanism of Cryptosporidium, Giardia,
and MS2 virus removal by MF and UF. J. Am. Water Works Assoc., 87(9)107-121.

Jakubowski, W. 1984. Detection of Giardia Cysts in Drinking Water: State of the Art. In:
Giardia and Giardiasis, Erlandsen, S.L. and Meyer, E.A., eds., Plenum Press, New York, pp.
263-286.

Jakubowski, W. 1990. The control of Giardia in water supplies. In: Human Parasitic Diseases,
Vol. 3, Giardiasis, E.A. Meyer, ed., Elsevier Science Publishers, Amsterdam, pp. 335-356.

Jakubowski, W., Sykora, J.L., Sorber, C.A., Casson, L.W., and Gavaghan, P.O. 1991.
Determining giardiasis prevalence by examination of sewage. Water Sci. Tech., 24(2): 173-178.

Jakubowski, W., Boutros, S., Faber, W., Payer, R., Ghiorse, W., LeChevallier, M. W., Rose, J.
B., Schaub, S., Singh, A, and Stewart, M. H. 1996. Environmental methods for
Cryptosporidium. J. Am. Waterworks Assoc., 88(9)107-121.

Janoff, E.N., Smith, P.D., andBlaser, M.J. 1988a. Acute antibody responses to Giardia lamblia
are depressed in patients with aids. J. Infect. Diseases, 157(4)798-804.

Janoff, E.N., Douglas, J.M., Jr., Gabriel, M., Blaser, M.J., et al. 1988b. Class-specific antibody
response to pneumoncoccal capsular polysaccharides in men with muman immunodeficiency
virus-1 infection. J. Infect. Diseases, 158:983-990
IX-27
-------
Janoff, E.N., Taylor, D.N., Echeverria, P., Globe, M., Blaser, MJ. 1990. Giardia lamblia-
specific antibodies in healthy U.S. and Thai children. West. G. Med.

Janoff, E.N and Smith, P.D. 1990. The role of immunity in Giardia infections.In: Human
Parasitic Diseases, Vol. 3, Giardiasis, E.A Meyer, ed., Elsevier Science Publishers, Amsterdam,
pp.215-233.

Januschka, M.M., Erlansen, S.L., Bemrick, W.J., Schupp, D.G., Feely, D.E. 1988. A comparison
of Giardia microti and Spironucleus muris cysts in the vole: an immunocytochemical, light, and
electron microscopic study. J. Parasit, 74(3):452-58.

Jarroll, E.L. 1988. Effect of disinfectants on Giardia cysts. CRC Critical Reviews in
Environmental Control, 18(1)1-28.

Jarroll, E.L., Bingham, A.K., Meyer, E.A. 1981. Effect of chlorine on Giardia lamblia cyst
viability. Appl. Environ. Microbiol., 41(2):483-487.

Jarroll, E.L., Manning P., Lindmark, D.G., Coggins, J.R., Erlandsen, S.L. 1989. Giardia cyst
wall-specific carbohydrate: evidence for the presence ofgalactosamine. Mol. Biochem.
Parasitol., 32:121-132.

Jephcott, A.E., Begg, N.T., and Baker, I A. 1986. Outbreak of giardiasis associated with mains
water in the United Kingdom. Lancet, i (March 29):730-732.

Johnson, D.C., Enriquez, C.E., Pepper, I.L., Davis, T.L., Gerba, C.P., and Rose, J.B. 1997.
Survival of Giardia, Cryptosporidium, poliovirus and Salmonella in marine waters. Water
Sci.Tech., 35(ll/12):261-268.
IX-28
-------
Jokipii, A.M.M., Hemila, M., and Jokipii, L., 1985. Prospective study of acquisition of
Cryptosporidium., Giardia lamblia, and gastrointestinal illness. Lancet, 2: 487-489.

Jokopii, L., Miettinen, A., Jokipii, A.M. 1988. Antibodies to cysts of Giardia lamblia in primary
giardiasis and in the absence of giardiasis. J. Clin. Microbiol. 26:121-125.

Kappus, K.D., Lundgren, R.G., Juranek, D.D., Roberts, J.M., and Spencer, H.C. 1994. Intestinal
parasitism in the United States: update on a continuing problem. Am. J. Trop. Med. Hyg.,
50(6):705-713.

Karabiber, N. and Aktas, F. 1991. Foodborne giardiasis [letter; comment]. Lancet
337(8737):376-377.

Karanis, P., Maier, W.A., Seitz, H. M., Schoenen, D. 1992. UV sensitivity of protozoan
parasites. J. Water SRT-Aqua, 41(2):95-100.

Karanis, P., Schoenen, D., and Seitz, H. M. 1996a. Giardia and Cryptosporidium in backwash
water from rapid sand filters used for drinking water production. Zentralbl. Bakteriol., 284:107-
114.

Karanis, P., Opiela, K., Al-Arousi, M., and Seitz, H. M. 1996b. A comparison of phase contrast
microscopy and an immunofluorescence test for the detection of Giardia spp. in faecal
specimens from cattle and wild rodents. Trans. R. Soc. Trop. Med. Hyg., 90(3):250-251.

Karanis, P., Schoenen, D., and Seitz, H.M. 1998. Distribution and removal of Giardia and
Cryptosporidium in water supplies in Germany. Water Sci. Tech., 37(2):9-18.
IX-29
-------
Katelaris, P.H., Naeem, A., Farthing, M.J.G. 1995. Attachment of Giardia lamblia trophozoites
to a cultured human intestinal cell line. Gut, 37:512-18.

Kaucner, C. and Stinear, T. 1998. Sensitive and rapid detection of viable Giardia cysts and
Cryptosporidiumparvum oocysts in large-volume water samples with wound fiberglass cartridge
filters and reverse transcription-PCR. Appl. Environ. Microbiol., 64(5): 1743-1749.

Kelley, M.B., Warrier, P.K., Brokaw, J.K., Barrett, K.L., and Kromisar, SJ. 1995. A study of
two U.S. army installation drinking water sources and treatment systems for the removal of
Giardia and Cryptosporidium.

Kent, G.P., Greenspan, J.R., Herndn, J.L., Mofenson, L.M., Harris, J.S., Eng, T.R., and Waskin,
H.A. 1988. Epidemic giardiasis caused by a contaminated public water supply. Am. J. Public
Health, 78(2): 139-143.

Kirkpatrick, C.E. 1986. Feline giardiasis: a review. J. Small Anim. Pract, 27: 69-80

Knotts, D. 1986. Prevention of waterborne disease for remote back country and developing
nations traveler. In: Proceedings of the International Symposium on Water-Related Health Issues,
C.L. Tate, ed., Am. Water Resources Assoc., Bethesda, MD, pp. 39-44.

Kramer, M.H., Herwaldt, B.L., Craun, G.F., Calderon, R.L., and Juranek, D.D. 1996. Waterborne
disease: 1993 and 1994. J. Am. Waterworks Assoc., 88(3)66-80.

Kramer, M.H., Herwaldt, B.L., Craun, G.F., Calderon, R.L., Juranek, D.D. 1996. Surveillance for
Waterborne-Disease Outbreaks-United States, 1993-1994. Morbidity and Mortality Weekly
Report, Vol45(SS-l):l-31.
IX-30
-------
Kulda, J. and Nohynkova, E. 1978. Flagellates of the Human Intestine and of Intestines of Other
Species. In: Kreier J.P., ed., Protozoa of Veterinary and Medical Interest, Academic Press, New
York, London, Vol. n, pp. 69-104.

Labatiuk, C.W., Schaefer, F.W., III, Finch, G.R., Belosevic, M., 1991. Comparison of animal
infectivity, excystation, and fluorogenic dye as measures oi Giardia muris cyst inactivation by
ozone. Appl. Environ. Microbiol., 57:3187-3192.

Labatuik, C.W., Belosevic, M., and Finch, G.R. 1994. Inactivation ofGiardia muris using ozone
and ozone hydrogen peroxide. Ozone Science and Engineering. 16:67-78.

Lange, K.P., Bellamy, W.D., Hendricks, Logsdon, G.S. 1986. Diatomaceous earth filtration of
Giardia cysts and other sub stances. J. Am. Water Works Assoc., 78(l):76-84.

Lapham, S., Hopkins, R., White, M., Blair, J., Bissell, R., and Simpson, G. 1987. A prospective
study of giardiasis and water supplies in Colorado. Am. J. Public Health, 77(3): 354-355.

Last, J.M., A Dictionary of Epidemiology, Third Edition. Oxford University Press, New York,
1995.

Laxer, M.A. 1985. Potential exposure of Utah army national guard personnel to giardiasis during
field training exercise: a preliminary survey. Military Medicine, 150:23-26.

Leahy, J.G., Rubin, A.J., and Sproul, OJ. 1987. Inactivation of Giardia muris cysts by free
chlorine. Applied Environmental Microbiology, 53(7):1448-1453.
IX-31
-------
LeChevallier, M.W. and Norton, W.D. 1992. Examining relationships between particle counts
and Giardia, Cryptosporidium, turbidity. J. Am. Waterworks Assoc., 84(12):54.

LeChevallier, M.W. and Norton, W.D. 1995. Giardia and Cryptosporidium in raw and finished
drinking water, J. Am. Water Works Assoc., 87(9):54-68.

LeChevallier, M.W., Trok, T.M., Burns, M.O., and Lee, R.G. 1990. Comparison of the zinc
sulfate and immunofluorescence techniques for detecting Giardia and Cryptosporidium. J. Am.
Water Works Assoc., 82(9):75-82.

LeChevallier, M. W., Norton, W. D., and Lee, R. G. 1991a. Occurrence of Giardia and
Cryptosporidium spp. in surface water supplies [published erratum appears in Appl Environ
Microbiol 1992 Feb; 58(2)780]. Appl. Environ. Microbiol., 57(9):2610-2616.

LeChevallier, M.W., Norton, W.D., and Lee, R.G. 1991b. Giardia and Cryptosporidium spp. in
filtered drinking water supplies. Appl. Environ. Microbiol., 57(9):2617-2621.

LeChevallier, M.W., Norton, W.D., Siegel, I.E., and Abbaszadegan, M. 1995. Evaluation of the
immunofluorescence procedure for detection of Giardia cysts and Cryptosporidium oocysts in
water. Appl. Environ. Microbiol., 61:690-697.

LeChevallier, M.W., Norton, W.D., and Atherholt, T.B. 1997. Protozoa in open reservoirs. J.
Am. Water Works Assoc., 89(9)84-96.

Lee, J. J. Jr. 1993. Contamination of a well by the direct influence of surface water. AWWA
Annual Conference, San Antonio, American Waterworks Association, Denver, CO,
IX-32
-------
Lengerich, E.J., Addiss, D.G., and Juranek, D.D. 1994. Severe giardiasis in the United States.
Clinical Infectious Diseases, 18:760-763.

Lev, B., Ward, H., Keuscg, G.T., Pereira, M.E.A. 1986. Lectin activation in Giardia lambliato
mammalian intestinal cells. Gut, 29:795-801.

Levy, D.A., Bens, M.S., Craun, G.F., Calderon, R.L., Herwaldt, B.L. In press, 1998. Surveillance
for Waterborne-Disease Outbreaks-United States, 1995-1996. Morbidity and Mortality Weekly
Report, Vol45(SS-l):l-31.

Linquist, H.D.A. 1995. A dual fluorochrome method for detection and viability of Giardia
lamblia cysts. Mol. Biochem. Parasitol., 31:170-175.

Logsdon, G., Symons, J., Hoye, R., and Arozarena, M. 1981. Alternative filtration methods for
removal of Giardia cysts and cyst models. J. Am. Water Works Assoc., 73(2): 111-118.

Logsdon, G.S., Thurman, V.C., Frindt, E.S., Stoecker, J.G. 1985. Evaluating sedimentation and
various filter media for removal of Giardia cysts. J. Am. Water Works Assoc., 77(2):61-66.

Lopez-Brea, M. 1982. Giardia lamblia: incidence in man and dogs. Trans. Roy. Soc. Trop. Med.
Hyg., 76: 565.

Ljungstrom, I. and Caster, B. 1992. Immune response to Giardia lamblia in a waterborne
outbreak in Sweden. J. Medical Microbiology, 36:347-52.
-------
MMWR. 1983. Outbreak of diarrheal illness associated with a natural disaster Utah. Morbidity
Mortality Weekly Report, 32:662-664.

Mahbubani, M.H., Bej, AK., Perlin, M., Schaefer, F.W., Jakubowski, W., and Atlas, R.M. 1991.
Detection ofGiardia cysts by using the polymerase chain reaction and distinguishing live from
dead cysts. Appl. Environ. Microbiol., 57(12):3456-3461.

Mahbubani, M.H., Bej, A.K., Perlin, M.H., Schaefer, F.W., Jakubowski, W., and Atlas, R.M.
1992. Differentiation of Giardia duodenalis from other Giardia spp. by using polymerase chain
reaction and gene probes. J. Clin. Microbiol., 30(l):74-78.

Mank, T.G., Zaat, J.O.,Deelder, A.M., van Eijk, J.T., and Polderman, A.M. 1997. Sensitivity of
microscopy versus enzyme immunoassay in the laboratory diagnosis of giardiasis. Eur. J. Clin.
Microbiol. Infect. Dis. 16(8):615-619.

Marshall, M.M., Naumovitz, D., Ortega, Y., and Sterling, C.R. 1997. Waterborne protozoan
pathogens. Clin. Microbiol. Rev., 10(l):67-85.

Mathias, R.G, Riben, P.D., Osei, W.D. Lack of an association between endemic giardiasis and a
drinking water source. Can. J. Public Health, 83(5):382-384.

Mayer, C.L. and Palmer, CJ. 1996. Evaluation of PCR, nested PCR, and fluorescent antibodies
for detection ofGiardia and Cryptosporidium species in wastewater. Appl. Environ. Microbiol.,
62(6):2081-2085.

McCaffery, J.M. and Gillin, F.D. 1994. Giardia lamblia. ultrastructural basis of protein transport
during growth and encystation. Experimental Parasitology, 79:220-235.
IX-34
-------
McCaffery, J.M.,Faubert, G.M., Gillin, RD. 1994. Giardia lamblia. traffic of a trophozoite
variant surface protein and a major cyst wall epitope during growth, encystation, and antigenic
switching. Exper. Parasitol., 79:236-249.

McTigue, N.E. and MacPhee, MJ. 1996. Engineering pilot plant studies for assessing pathogen
removal: treatment design and operational considerations. Proc. 1996 Water Quality Technology
Conference, Boston, MA, Am. Water Work Assoc., Denver, CO.

Meloni, B.P., Thompson, R.C., Reynoldson, J.A., Seville, P. 1990. Albendazole: a more effective
antigiardial agent in vitro than metronidazole or tinidazole. Transactions Royal Society Tropical
Medicine and Hygiene, 84(3):375-9.

Meloni, B.P., Lymbery, A.J., and Thompson, R.C.A. 1995. Genetic characterization of isolates of
Giardia duodenalis by enzyme electrophoresis: implications for reproductive biology, population
structure, taxonomy, and epidemiology. J. Parasitol., 81(3):368-383.

Mendis, A.H.W. and Schofield, P. J. 1994. Discussants report: Giardia Biochemistry. In:
Giardia: From Molecules to Disease, R.C.A. Thompson, J.A. Reynoldson, and AJ. Lymbery,
eds., CAB INTERNATIONAL, Wallingford,U.K., pp. 205-213.

Meyer, E.A., dicker, J., Bingham, A.K., Edwards, R. 1988. Inactivation of G. muris cysts by
chloramines. Water Res. Bull., 25:235-340.

Mintz, E.D., Hudson-Wragg, M., Mshar, P., Cartter,M.L., Hadler, J.L 1993. Foodborne
giardiasis in a corporate office setting. Jour. Infect. Dis., 167(l):250-253.
IX-35
-------
Miotti, P.O., Oilman, R.H., Pickering, L.K., Ruiz-Palacios, G., Park, H.S., and Yolken, R.H.
1985. Prevalence of serum and milk antibodies to Giardia lamblia in different populations of
lactating women. J. Infect. Dis., 152:1026-1031.

Miotti, P.O., Oilman, R.H., Santosham, M., Ryder, R.W., and Yolken, R.H. 1986. Age-related
rate of seropositivity of antibody to Giardia lamblia in four diverse populations. Journal of
Clinical Microbiology, 24(6):972-975.

Misra, P.K., Kumar, A., Agarwal, V., Jagota, S.C. 1995. A comparative clinical trail of
albendazole versus metronidazole in giardiasis. Indian Peadiatrics, 32:291-294.

Miyamoto, K., Kutsume, H. 1978. Studies on zoonosesin Hokkaido, Japan. 1. An
epidemiological survey of protozoan and helminthic infections of stray dogs in Kamikawa
district. Japanese J. Parasitol., 27: 369-374.

Moody, A.H., Ridley, D.S., Tomkins, A.M., and Wright, S.G. The specificity of serum antibodies
to Giardia lamblia and to enterobacteria in gastrointestinal disease. Transactions Royal Society
Tropical Medicine and Hygiene, 76(5):630-632.

Moore, A.C., Herwaldt, B.L., Craun, G.F., Calderon, R.L., Highsmith, A.K., Juranek, D.D. 1993.
Surveillance for Waterborne Disease Outbreaks-United States, 1991-1992. Morbidity and
Mortality Weekly Report, Vol. 42(SS-5):l-22.

Moore, A.C., Herwaldt, BL., Craun, G.F., Calderon, R.L., Highsmith, A.K., and Juranek, D.D.
1994. Waterborne-disease in the United States, 1991 and 1992. J. Am. Waterworks Assoc.,
86(2):87.
IX-36
-------
Moorehead, W.P., Guasparini, R., Dovovan, C.A., Cottle, R., Baytalan, G. 1990. Giardiasis
outbreak from a chlorinated community water supply. Can, J. Public Health, 81(5):358-62.

Musial, C.E., Arrowood, M.J., Sterling, C.R., and Gerba, C.P. 1987. Detection of
Cryptosporidium in water by using polypropylene cartridge filters. Appl. Environ. Microbiol.,
53:687-692.

Nacapunchai, D., Tepmongkol, M., Tharavanij, S., Thammapalerd, N., Subchareon, A. 1986. A
comparative study four methods for detecting antibody in asymptomatic giardiasis. SE Asian J.
Trop. Med. Public Health, 17(1):96-100.

Nahnias, J., Greenburg, Z., Djerrasi, L., Gildai, L. 1991. Mass treatment of intestinal parasites
among Ethiopian immigrants. Israel J. Med. Sciences, 27:278-283.

Nakano, L, Miyahara, T., Ito, T., Migita. Y., Nawata, H. 1995. Giardiasis in pancreas. Lancet,
345:524-5 (February 25).

Nash, T.E., Herrington, D.A., Losonsky, G.A., Levine, M.M. 1987. Experimental human
infections with Giardia lamblia. J. Infect. Dis., 156:974-984.

National Research Council (NRC). 1983. Risk Assessment in the Federal Government:
Managing the Process. NAS-NRC Committee on the Institutional Means for Assessment of
Risks to Public Health, National Academy Press, Washington, D.C..

Nayak, N., Ganguly, N.K., Walia, B.N.S., Wahi, V., Kanwar, S.S., and Mahajan,R.C. 1987.
Specific secretory IgA in the milk of Giardia lamblia-infected and uninfected women. J. Infect.
Dis. 155(4):724-727.
IX-37
-------
Neringer, R., Anderson, Y., and Eitrem, R. 1987. A water-borne outbreak of giardiasis in
Sweden. Scand. J. Infect Dis. 19:85-90.

Nieminski, E.G., and Ongerth, I.E. 1995. Removing Giardia and Cryptosporidium by
Conventional and Direct Filtration. J. Am. Waterworks Assoc., 87:96-106.

Nieminski, E.G., Schaefer, F.W., and Ongerth, I.E. 1995. Comparison of two methods for
detection of Giardia cysts and Cryptosporidium oocysts in water. Appl. Environ. Microbiol.
61:1714-1719.

Norton, W.D., LeChevallier, M.W., Marino, A.D., andKelleher, D.L. 1995. Survey of source
water for Giardia cysts and Cryptosporidium oocysts. Prepared for the New Jersey Department
of Environmental Protection, June, 1995, American Waterworks Service Company, Inc.,
Voorhees, NJ, 82 pp.

Okun, D., Craun, G.F., Edzwald, J., Gilbert, J., and Rose, J. 1997. New York City: to filter or not
to filter? J. Am. Waterworks Assoc, 89(3):62-74.

O'Handley, R.M., Olsen, M.E., McAllister, T.A., Morck, D.W., Jelinski, M., Royan, G., Cheng,
K.J., 1997. Efficacy of fenbendazule for treatment of g/'ardiasis in calves. Am. J. Vet. Res., 58:
384-388.

Olson, M.E., Thorlakson, C.,Deselliers, L., Morck, D.W., McAllister, T.A. 1997. Giardia and
Cryptosporidium in Canadian farm animals. Vet. Parasitol., 68(4):375-81.
IX-38
-------
Oliveira, C.A.F. De and Germano, P.M.L. 1992. Study of the occurrence of intestinal parasites
on vegetables commercially traded in the metropolitan area of S. Paulo, SP-Brazil II.-research
into protozoan cysts. Rev. Saude Publ., 26(5):332-5.

Ong, C., Moorehead, W., Ross, A., and Isaac-Renton, J. 1996. Studies ofGiardia spp. and
Cryptosporidium spp. in two adjacent watersheds. Appl. Environ. Microbiol., 62(8):2798-2805.

Ongerth, I.E. 1989a. Giardia cyst concentrations in river water. J. Am. Water Works Assoc.,
81(9):81-86.

Ongerth, J.E, Johnson, R.L., Macdonald, S.C., Frost, F., and Stibbs, H.H. 1989. Backcountry
water treatment to prevent giardiasis. Am. J. Public Health, 79(12): 1633-1637.

Ongerth, J.E. 1990. Evaluation of treatment for removing Giardia cysts. J. Am. Water Works
Assoc., 82(6):85-96.

Ongerth, J.E., Hunter, G.D., and DeWalle, F.B. 1995. Watershed use andGiardia cyst presence.
Water Research, 29(5): 1295-1299.

Ongerth, J.E. and Pecoraro, J.P. 1995. Removing Cryptosporidium using multimedia filters. J.
Am. Waterworks Assoc., 87(12):83-89.

O'Shaughnessy, P.T., Barsotti, M.G., Fay, J.W., Tighe, S.W. 1997. Evaluating particle counters.
J. Am. Waterworks Assoc., 89(12):60-70.
IX-39
-------
Ortega, Y.R. and Adam, R.D. 1997. Giardia: overview and update. Clinical Infectious Diseases,
25:545-50.

Owens, J.H., Miltner, R.J., Schaefer, F.W. in, Rice, E.W. 1994. Pilot-scale ozone inactivation of
Cryptosporidium and Giardia. Proc. 1994 Water Qual. Technol. Conf, San Francisco, Am.
Waterworks Assoc., Denver, CO, pp. 1319-1324.

Pacha, R.E., Clark, G.W., Williams, E.A., Carter, A.M., Scheffelmaier, J.J., Debusschere, P.
1987. Small rodents and other mammals associated with mountain meadows as reservoirs of
Giardia spp. and Campylobacter spp. Appl. Environ. Microbiol., 53(7): 1574-1579.

Paget, T.A., Jarroll, EL, Manning, P., Lindmark, D.G. and Lloyd, D. 1989. Respiration in the
cysts and trophozoites of Giardia muris. Journal of General Microbiology, 135(1): 145-154.

Paget, T.A., Manning, P., and Jarroll, EL. 1993. Oxygen uptake in cysts and trophozoites of
Giardia lamblia. Journal of Eukaryotic Microbiology, 40(3):246-250, 1993.

Patania, N., Jacangelo, J.G., Cummings, L., Wilczak, A. 1995. Removal of Cryptosporidium,
Giardia, and particles by granular media filtration in pilot studies, AWWA Annual Conference,
Anaheim, CA, Am. Waterworks Assoc., Denver, CO.

Patton, S. 1998. Giardiasis. In: Aiello, S., ed., The Merck Veterinary Manual, Merck and
Co.,Whitehouse Station, NJ, pp. 146-147.

Payment, P., Berube, A., Perreault, D., Armon, R., and Trudel, M. 1989. Concentration of
Giardia lamblia cysts, Legionellapneumophila, Clostridium perfringens, human enteric viruses,
IX-40
-------
and coliphages from large volumes of drinking water, using a single filtration. Can. J. Microbiol.,
35(10):932-935.

Payment, P. and Franco, E. 1993. Clostridium perfringens and somatic coliphages as indicators
of the efficiency of drinking water treatment for viruses and protozoan cysts. Appl. Environ.
Microbiol., 59:2418-2424.

Pearce, D.A., Reynoldson, J.A., Thompson, R.C. 1996. A comparison of two methods for
assessing drug sensitivity in Giardia duodenalis. Applied Parasitology, 37(2): 111-6.

Perlmutter, D.H., Leichtner, A.M., Goldman, H., Winter, H.S. 1985. Chronic diarhea associated
with hypogammaglobulinemia and enteropathy in infants and children. Dig. Dis. Sci., 30:1149-
1155.

Peterson, L., Carter, M., and Hadler, J. 1988. A food-borne outbreak of Giardia lamblia. The
Journal Infectious Diseases, 157(4): 846-848.

Ponce-Macotela, M., Navarro-Alegria, I, Martinez-Gordillo, M.N., Alvarez-Chacon, R. 1994. In
vitro effect against Giardia of 14 plant extracts [Abstract]. Revista de Investigacion Clinica,
46(5):343-7.

Porter, J.D., Ragazzani, HP., Buchanon, J.D., Waskin, H.A., Juranek, D.D., and Parkin, W.E.
1988. Giardia transmission in a swimming pool. Am. J. Public Health, 78(6):659-662.

Pungpak, S., Singhasivanon, V., Bunnag, D., Radomyos, B., Nibaddhasopon, P., Harinasuta,
K.T. 1996. Albendazole as a treatment for Giardia infection. Annals of Tropical Medicine and
Parasitology, 90(5):563-5.
IX-41
-------
Quick, R., Paugh, K., Addiss, D., Kobasyashi, J., and Baron, R. 1992. Restaurant associated
outbreak of giardiasis. J. Infect. Dis., 166:673-676.

Qureshi, H., Ali, A., Baqai, R., Ahmed, W. 1997. J. Jut. Medical Res., 25:167-170

Rabbani, G.H. and A. Mam. 1994. Giardiasis in humans: Populations most at risk and prospects
for control. In: Giardia: From Molecules to Disease, R.C.A. Thompson, J.A. Reynoldson, and
AJ. Lymbery, eds., CAB INTERNATIONAL, Wallingford, U.K., pp. 83-97.

Ramsey, C.N. and Marsh, J. 1990. Giardiasis due to deliberate contamination of water supply.
Lancet. 336:880-881.

Regli, S., Rose, J.B., Haas, C.N., and Gerba, C.P. 1991. Modeling the risk from Giardia and
viruses in drinking water. J. Am. Water Works Assoc., 83:11:76-84.

Reiner, D.S., Wang, C-S, Gllin, F.D. 1986. Human milk kills Giardia lambliaby generating
toxic lipolytic products. J. Infect. Dis., 154:825-832.

Rendtorff, R.C. 1954a. The experimental transmission of human intestinal protozoan parasites.
II. Giardia lamblia cysts given in capsules. Am. J. Hyg., 59:209-220.

Rendtorff, R.C. and Holt, CJ. 1954b. The experimental transmission of human intestinal
protozoan parasites. IV. Attempts to transmitEndamoeba coli and Giardia lamblia cysts by
water. Am. J. Hyg, 60:327-338.
IX-42
-------
Rendtorff, R.C. 1979. The experimental transmission of Giardia lamblia among volunteer
subjects. In: Jakubowski W., Hoff, J.C., eds. Waterbome Transmission of Giardiasis. Cincinnati:
U.S. Environmental Protection Agency, pp. 64-81. EPA-600/9-79-001.

Reiner, D.S. and Gillin, F.D. 1992. Human secretory and serum antibodies recognize
environmentally induced antigens of Giardia lamblia. Infect. Immunity, 60(2):637-642.

Rice, E.W. and Hoff, J.C. 1981. Inactivation of Giardia lamblia cysts by ultraviolet irradiation.
Applied and Environmental Microbiology, 42(3):546-547.

Rice, E.W., Hoff, J.C., Schaefer, F.W. 1982. Inactivation of Giardia cysts by chlorine. Appl.
Environ. Microbiol., 43(1):250-251.

Riesenberg, F., Walters, B., Steele, A., and Ryder, R. 1995. Slow sand filters for a small water
system. J. Am. Waterworks Assoc., 87(ll):48-56.

Riggs, J. L., Dupuis, K., Nakamura, K., and Spath, D. 1983. Detection of Giardia lambliaby
immunofluorescence. Appl. Environ. Microbiol., 45(2):698-700.

Roach, P.O., Olson, M.E., Whitley, G., and Wallis, P.M. 1993. Waterborne Giardia cysts and
Cryptosporidium oocysts in the Yukon, Canada. Appl. Environ. Microbiol., 59(l):67-73.

Robertson, LJ. 1996. Severe giardiasis andcryptosporidiosis in Scotland, UK. Epidemiol.
Infect, 117:551-561.
IX-43
-------
Robertson, L.J., Smith, H.V., and Paton, C.A. 1995. Occurrence of Giardia cysts and
Cryptosporidium oocysts in sewage influents in six sewage treatment plants in Scotland and
prevalence of cryptosporidiosis and giardiasis diagnosed in the community served by those
plants. In: Protozoan Parasites and Water. Belts, W. B., Casemore, D., Flicker, C. R., Smith, H.,
and Watkins, J., eds., The Royal Society of Chemistry, Thomas Graham House, Science Park,
Cambridge, Great Britain, pp. 47-49.

Roberts-Thomson, 1C. and Anders, R.F. 1981. Serum antibodies in adults with giardiasis.
Gastroenterol., 80(5):1262.

Roberts-Thomson, 1C. and Mitchell, G.F. 1978. Giardiasis in mice. I Prolonged infections in
certain mouse strains and hypothymic (nude) mice. Gastroenterol., 75:42-46.

Roberts-Thomson 1C, Stevens DP, Mahmoud AAF, Warren KS. 1976. Giardiasis in the mouse:
An animal model. Gastroenterol. 71:57-61.

Rochelle, P. A., DeLeon, R., Stewart, M. H., and Wolfe, R. L. 1997. Comparison of primers and
optimization of PCR conditions for detection of Cryptosporidium parvum and Giardia lamblia in
water. Appl. Environ. Microbiol. 63(1): 106-114.

Rodgers, M., Flanigan, D., Pfaller, S., and Kinkle, B. 1998. Isolation and characterization of a
Flavobacterium strain virulent against Giardia lamblia cysts. ASM Abstracts in Microbial
Ecology, No. N-141, p. 389.

Rodgers, M.R., Flanigan, D., and Jakubowski, W. 1995. Identification of algae which interfere
with the detection of Giardia cysts and Cryptosporidium oocysts and a method for alleviating
this interference. Appl. Environ. Microbiol., 61:3759-3763.
IX-44
-------
Rodriguez-Hernandez, J., Canut-Blasco, A., Martin-Sanchez, A.M. 1996. Seasonal prevalences
of Cryptosporidium and Giardia infections in children attending day care centers in Salamanca
(Spain) studied for a period of 15 months. European J. Epidemiol., 12:291-295.

Rohrer, L. Winterhaulter, K.H., Eckert, J. Kohler, P. 1986. Killing of Giardia lambliaby human
milk is mediated bu unsaturated fatty acids. Antimicrob. Agents Chemother., 30:254-257.

Rosales-Borjas, D.M., Diaz-Rivadeneyra, J., Dona-Leyva, A., Zambrano-Villa, S.A., Mascaro,
C., Osuna, A., and Ortiz-Ortiz, L. 1998. Secretory immune response to membrane antigens
during Giardia lamblia infection in humans. Infect. Immun., 66(2):756-759.

Rose, J.B., Darbin, H., and Gerba, C.P. 1988a. Correlations of the protozoa, Cryptosporidium
and Giardia, with water quality variables in a watershed. Water Sci.Tech., 20:271-276.

Rose, J.B., Kayed, D., Madore, M.S., Gerba, C.P., Arrowood, M.J., Sterling, C.R., and Riggs,
J.L. 1988b. Methods for the recovery of Giardia and Cryptosporidium from environmental
waters and their comparative occurrence. In: Advances in Giardia Research, Wallis, P. M. and
Hammond, B. R., eds., University of Calgary Press, Calgary, pp. 205-209.

Rose, J.B., Landeen, L.K., Riley, K.R., and Gerba, C.P. 1989. Evaluation of immunofluorescence
techniques for detection of Cryptosporidium oocysts and Giardia cysts from environmental
samples. Appl. Environ. Microbiol., 55(12):3189-3196.

Rose, J.B., Gerba, C.P., and Jakubowski, W. 1991a. Survey of potable water supplies for
Cryptosporidium and Giardia. Environmental Science and Technology, 25:1393-1400.
IX-45
-------
Rose J.B., Haas C.N., andRegli S. 1991b. Risk assessment and control of waterborne giardiasis.
Am. J. Public Health, 81(6): 709-13.

Rose, J.B. and Gerba, C.P. 1991. Use of risk assessment for development ofmicrobial standards.
Water Sci. Technol. 24:29-34.

Rubin, L.G. 1987. Bacterial colonization and infection resulting from multiplication of a single
organism. Rev. Infect. Dis., 9:(3) 488-493.

Rubin, A.J., Evers, D.P., Eyman, C.M., Jarroll, E.L. 1989. Inactivation of gerbil-cuhured Giardia
lamblia cysts by free chlorine. Appl. Environ. Microbiol., 55(10):2592-2594.

Ruest, N., Couture, Y., Faubert, G.M., Girard, C. 1997. Morphological changes in the jejunum
of calves naturally infected with Giardia spp. and Cryptosporidium spp. Vet. Parasitol., 69(3-4):
177-186.

Sauch, J.F. 1984. Purification of Giardia muris cysts by velocity sedimentation. Applied and
Environmental Microbiology, 48(2):454-455.

Sauch, J.F. 1985. Use of immunofluorescence and phase-contrast microscopy for detection and
identification of Giardia cysts in water samples. Appl. Environ. Microbiol., 50:1434-1438.

Sauch, J.F. 1988. A new method for excystation of Giardia. In: Advances in Giardia Research,
Wallis, P. M. and Hammond, B. R., eds., University of Calgary Press, Calgary, pp. 261-264.

Sauch, J.F. and Berman, D. 1991. Immunofluorescence and morphology of Giardia lamblia cysts
exposed to chlorine. Appl. Environ. Microbiol., 57(5):1573-1575.
IX-46
-------
Sauch, J.F., Flanigan, D., Galvin, M.L., Berman, D., and Jakubowski, W. 1991. Propidium
iodide as an indicator of Giardia cyst viability. Appl. Environ. Microbiol., 57(11):3243-3247.

Schaefer, F.W. 1988. A review of methods that are used to determine Giardia cyst viability. In:
Advances in Giardia Research, Wallis, P. M. and Hammond, B. R., eds., University of Calgary
Press, Calgary, pp. 249-254.

Schaefer, F.W. 1990. Methods for excystation of Giardia. In: Human Parasitic Diseases, Vol. 3,
Giardiasis, Meyer, E. A., ed., Elsevier, Amsterdam,, pp. 111-136.

Schaefer, F.W., Rice, E.W., and Hoff, J.C. 1984. Factors promoting in vitro excystation of
Giardia muris cysts. Trans. R. Soc. Trop. Med. Hyg., 78:795-800.

Schuler, P.F., Ghosh, M.M., Gopalan, P. 1991. Slow sand and diatomaceous earth filtration of
cysts and other particulates. Water Res., 25(8):995-1005.

Schupp, D.G., Januschka, M.M., and Erlandsen, J.L. 1988. Morphology of Giardia encystation
in vitro. In: Advances in Giardia Research, Wallis, P. M. and Hammond, B. R., eds., University
of Calgary Press, Calgary, pp. 29-32.

Schupp, D.G. and Erlandsen, S.L. 1987a. A new method to determine Giardia cyst viability:
correlation of fluorescein diacetate andpropidium iodide staining with animal infectivity. Appl.
Environ. Microbiol., 53(4)704-707.

Schupp, D.G. and Erlandsen, S.L. 1987b. Determination of Giardia muris cyst viability by
differential interference contrast, phase, or brightfield microscopy. Jour. Parasitol., 73:723-729.
IX-47
-------
Shaw, R.A. and Stevens, M.B. 1987. Reactive arthritis associated with giardiasis. JAMA, 258
(19):2734 (Nov. 20).

Sheffield, H.G. and Bjorvatn B. 1977. infrastructure of the cyst ofGiardia lamblia. Am. Soc.
Trop. Med. Hyg., 26(1):23-30.

Shepherd, K.M. and Wyn-Jones, A.P. 1996. An evaluation of methods for the simultaneous
detection of Cryptosporidium oocysts andGiardia cysts from water. Appl. Environ. Microbiol.,
62(4): 1317-1322.

Skeels, M.R., Sokolpw, R., Hubbard, C.V., and Foster, L.R. 1986. Screening for coinfection with
Cryptosporidium and Giardia in Oregon public health clinic patients. Am. J. Public Health,
76(3):270-273.

Smith, P.O., Gillin, F.D., Brown, W., and Nash, T. 1981. IgG antibody to Giardia lamblia
detected by enzyme-linked immunosorbent assay. Gastroenterology, 80(6): 1476-1480.

Smith, P.O., Gillin, F.D., Kaushal, N.A., and Nash, T. 1982. Antigenic analysis of Giardia
lamblia from Afghanistan, Puerto Rico, Ecuador, andOregan. Infect. Immun., 36:714-719.

Smith, H.V., Robertson, L.J., Reay, D., and Young, C.J. 1994. Occurrence ofGiardia cysts in
raw and treated sewage in relation to the prevalence of giardiasis in a Scottish community. In:
Giardia: From Molecules to Disease, R.C.A. Thompson, J. A. Reynoldson, and A.J. Lymbery,
eds., CAB INTERNATIONAL, Wallingford,U.K., pp. 128-129.

Speelman, P. and Ljungstrom, I. 1986. Protozoa! enteric infections among expatriates in
Bangladesh. Am. J. Trop. Med. Hyg., 35:1140-1145.
IX-48
-------
Stager, S. and Muller, N. 1997. Giardia lamblia infections in b-cell deficient transgenic mice.
Infection and Immunity, 65(9):3944-3946.

Stager,S., Gottstein, B., Sager, H., Jungi, T.W., and Muller, N. 1998. Influence of antibodies in
mother's milk on antigenic variation of Giardia lamblia in the murine mother offspring model of
infection. Infection ans Immunity, 66(4): 1287-1292.

Starko, K.M., Lippy, E.G. Domingguez, L.B., et al. 1986. Campers diarrhea outbreak traced to
water-sewage link. Public Health Reports, 101:527-531.

States, S., Sykora, I, Stadterman, K., Wright, D., Baldizar, 1, Conley, L. 1995. Sources,
occurrence, and drinking water treatment removal of Cryptosporidium and Giardia in the
Allegheny river. Proc. 1995 Water Qual. Technol. Conf, New Orleans, Am. Water Works
Assoc., Denver, CO, pp. 1587-1601.

States, S., Stadterman, K., Ammon, L., Vogel, P., Baldizar, J., Wright, D., Conley, L., and
Sykora, J.L. 1997. Protozoa in river water: sources, occurrence, and treatment. J. Am. Water
Works Assoc., 89(9):74-83.

Stazzone, A.M., Slaats, S., Mortagy, A., Kleinosky, M., Diab, A., Mourad, A., Hebert, A.,
Merrell, B.R., Watson, R.R., and Murphy, J.R. 1996. Frequency of Giardia and Cryptosporidium
infections in Egyptian children as determined by conventional and immunofluorescence methods.
Pediatr. Infect. Dis. J., 15(11): 1044-1046.

Steketee, R.W., Reid, S., Cheng, T., Stoebig, J.S., Harrington, R.G., and Davis, J.P. 1989.
Recurrent outbreaks of giardiasis in a child day care, Wisconsin. Am. J. Public Health,
79(40):485-490.
IX-49
-------
Stewart, M.H., Ferguson, D.M., DeLeon, R., and Taylor, W.D. 1997. Monitoring program to
determine pathogen occurrence in relationship to storm events and watershed conditions. In:
Proceedings 1997 Water Quality Technology Conference, American Water Works Association,
Denver, CO,

Stibbs, H.H., Riley,E.T., Stockard, I, Riggs, J.L., Wallis, P.M., and Isaac-Renton, J. 1988.
Immunofluorescence differentiation between various animal and human source Giardia cysts
using monoclonal antibodies. In: Advances in Giardia Research. Wallis, P.M. and Hammond,
B.R., eds., University of Calgary Press, Calgary, pp. 159-163.

Sullivan, R., Linneman, C.C., Clark, C.S., and Walzer, P.D. 1987. Seroepidemiologic study of
giardiasis patients and high risk groups in a Midwestern city in the United States. Am. J. Public
Health, 77(8): 960-963.

Swarbrick, A., Lim, R.L., Upcroft, J.A., Stewart, T.S. 1997. Nucleotide variation in the cytidine
triphosphate synthetase gene of Giardia duodenalis. Journal of Eukaryotic Microbiology,
44(6):531-4.

Sykora, J.L., Sorber, C.A., Jakubowski, W., Casson, L.W., Gavaghan, P.D., Shapiro, M.A., and
Schott, MJ. 1991. Distribution of Giardia cysts in wastewater. Water Sci.Tech., 24(2): 187-192.

Taylor, G.D. and Wenman, W.W. 1987. Human immune response to Giardia lamblia infection.
J. Infect. Dis., 155(1): 137-140.

Teunis, P.F.M., Havelaar, A.H., and Medema, GJ. 1994. A Literature survey on the assessment
of microbiological risk for drinking water. Report 734301006, National Institute of Public Health
and Environmental Protection, Bilthoven, The Netherlands.
IX-50
-------
Teunis, P.P.M., Medema, G. J., Kruidenier, L., Havelaar, A.H. 1997. Assessment of the risk of
infection by Cryptosporidium and Giardia in drinking water from a surface water source. Water
Research, 31(6): 1333-1346.

Thompson, R.C.A., Reynoldson, A.J., and Mendis, A.H. 1993. Giardia and giardiasis. Advances
in Parasitology, 32: 71-160.

Thompson, S.C. 1994. Giardia lamblia in children and the child care setting: a review of the
literature. J. Pediar. Child Health, 30:202-209.

Thompson, S.C., Lymbery, A.J., Pearce, D.A., Finn, K.C., Reynoldson, J.A., and Meloni, B.P.
1996. Giardia duodenalis: exposure to metronidazole inhibits competitive interactions between
isolates of the parasite in vitro. J. Parasitol., 82(4):679-683.

Tibayrenc, M. 1994. How many species of Giardia are there? In: Giardia: From Molecules to
Disease. Thompson, R.C.A., Reynoldson, J.A., and Lymbery, A.J., eds. CAB
INTERNATIONAL, Wallingford, pp. 41-48.

Turner, J.A. 1985. Giardiasis and infections withDientamoebafragilis. Pediatric Clinics of
North America, 32(4):865-79.

Ungar, B.L., Yolken, R.H., Nash, T.E., and Quinn, T.C. 1984. Enzyme-linked immunosorbent
assay for the detection of Giardia lamblia in fecal specimens. J. Infect. Dis., 149(l):90-97.

Upcroft, J.A., Upcroft, P., Boreham, P.F. 1990. Drug resistance in Giardia intestinalis.
International Journal for Parasitology, 20(4):489-96.
IX-51
-------
Upcroft, J.A., Boreham,P.F., Campbell, R.W., Shepherd, R.W., Upcroft, P. 1995. Biological and
genetic analysis of a longitudinal collection ofGiardia samples derived from humans. Acta
Tropica, 60(l):35-46.

Upcroft, J., Mitchell, R., Chen, N., Upcroft, P. 1996a Albendazole resistance in Giardia is
correlated with cytoskeletal changes but not with a mutation at amino acid 200 in beta-tubulin.
Microbial Drug Resistance, 2(3):303-8.

Upcroft, J.A., Campbell, R.W., Upcroft, P. 1996b. Quinacrine-resistant Giardia duodenalis.
Parasitology, 112 (Pt 3):309-13.

U.S. EPA. 1989a. Guidance Manual for Compliance with Filtration and Disinfection
Requirements for Public Water Systems using Surface Water Sources. United States
Environmental Protection Agency, (EPA Report No. 570/9-89-018), Washington, D.C.

U.S. EPA. 1989b. National primary drinking water regulations; filtration and disinfection;
turbidity; Giardia lamblia, viruses, Legionella, and heterotrophic bacteria, Federal Register
54(124), 27486-27541.

U.S. EPA. 1994a. Monitoring requirements for public drinking water supplies; proposed rule.
Federal Register 59:6332-6444.

U.S. EPA. 1994b. National primary drinking water regulations; disinfectants and disinfection
byproducts; proposed rule, Federal Register, 59(145), 38668-28829, 1994.

U.S. EPA. 1994c. National Primary Drinking Water Regulations: Enhanced surface water
treatment requirements; proposed rule., Federal Register, 59(159), 38832-38858, 1994.
IX-52
-------
U.S. EPA. 1996. National Primary Drinking Water Regulations: Monitoring Requirements for
Public Drinking Water Supplies; Final Rule. Federal Register 61:24354-24388.

van Keulen, H., Gutell,R.R., Campbell, S.R., Erlandsen, S.L., Jarroll, E.L. 1992. The nucleotide
sequence of the entire ribosomal DNA operon and the structure of the large subunit rRNA of
Giardiamuris. Jour. Molec. Evol., 35:318-328.

van Keulen, H., Gutell, R.R., Gates, M.A., Campbell, S.R., Erlandsen, S.R., Jarroll, E.L., Kulda,
J., Meyer, E.A. 1993. Unique phylogenetic position of Diplomonadida based on the complete
small subunit ribosomal RNA sequence of Giardia ardeae, G. muris, G. duodenalis and
Hexamita sp. FASEB Jour., 7:223-231.

Van Praagh, A.D., Gavaghan, P.O., and Sykora, J.L. 1993. Giardia muris cyst inactivation in
anaerobic digester sludge. Water Sci.Tech. 27:105-109.

Veazie, L., Brownlee, I, Sears, H.J. 1979. An outbreak of gastroenteritis associated with
Giardia lamblia. In: Jakubowski W., Hoff, J.C., eds., Waterborne Transmission of Giardiasis,
U.S. Environmental Protection Agency (EPA-600/9-79-001), Cincinnati, OH, pp. 174-192.

Vesey, G., Hutton, P., Champion, A., Ashbolt, N., Williams, K. L., Warton, A., and Veal, D.
1994. Application of flow cytometric methods for the routine detection of Cryptosporidium and
Giardia in water. Cytometry, 16(1): 1-6.

Walia, B.N., Ganguly, N.K., Mahajan, R.C., Kimar, D., Madan, I.J., Gambhir, S.K., Kanwar,
S.S. 1986. Morbidity in preschool Giardia cyst excretors. Trop Geograp. Med., 38(4):367-70.
IX-5'-
-------
Wallis, P.M., Buchanan-Mappin, J.M., Faubert, G.M., and Belosevic, M. 1984. Reservoirs of
Giardia spp in Southwestern Alberta. Journal of Wildlife Diseases, 20(4):279-283.

Wallis, P.M., Zammuto, R.M., and Buchanan-Mappin, J.M. 1986. Cysts of Giardia spp in
mammals and surface waters in Southwestern Alberta. Journal of Wildlife Diseases, 22(1): 115-
118.

Wallis, P.M., Erlandsen, S.L., Isaac-Renton, J.L., Olson, M.E., Robertson, W.J., and van Keulen,
H. 1996. Prevalence of Giardia cysts and Cryptosporidium oocysts and characterization of
Giardia spp. isolated from drinking water in Canada. Appl. Environ. Microbiol., 62(8):2789-
2797.

Weiss, H.B., Winegar, DA., Levy, B.S., and Washburn, J.W. 1977. Giardiasis in Minnesota,
1971-1975. Minn. Med., 60:815-820.

Weiss, J.B. 1995. DNA probes and PCR for diagnosis of parasitic infections. Clin. Microbiol.
Rev., 8:113-130.
Weber-Shirk, ML. and Dick, RI. 1997. Physical-chemical mechanisms in slow sand filters. J.
Am. Water Works Assoc., 89 (l):87-99.

White, K.E., Hedberg, C.W., Edmonson, L.M., Jones, D.B.W., Osterholm, M.T., and
MacDonald, K.L. 1989. Journal Infectious Disease, 160(2):298-304.

Wiandt, S., Baleux, B., Casellas, C., and Bontoux, J. 1995. Occurrence of Giardia sp. cysts
during wastewater treatment by a stabilization pond in the south of France. Water Sci.Tech.,
31(12):257-265.
IX-54
-------
Wickramanayake, G.B., Rubin, A.J., and Sproul, OJ. 1984a. Inactivation of Giardia lamblia
cysts with ozone. Applied and Environmental Microbiology, 48(3):671-672.

Wickramanayake, G.B., Rubin, A.J., and Sproul, OJ. 1984b. Inactivation ofNaegleria and
Giardia cysts in waterby ozonation. J. Water Poll. Control Fed., 56(8):983-988.

Wickramanayake, G.B., Rubin, A.J., Sproul, OJ. 1985. Effects of ozone and storage on Giardia
cysts. J. Am. Waterworks Assoc., 77(8):74-77.

Wilson, H.P.S., Stauffer, S.J., and Walker, T.S. 1982. Waterborne giardiasis outbreak-Alberta.
Canada Diseases Weekly Report, 8(20):97-98.

Witt, V.M. and Reiff, P.M. 1997. Water disinfection technologies for small communities and
rural areas. In: Water Quality in Latin America: Balancing theMicrobial and Chemical Risks in
Drinking Water Disinfection. Craun, G.F., ed., ILSI Press, Washington, DC, pp. 55-77.

Wolfe, R.L., Stewart, M.H., Scott, K.N., and McGuire, MJ. 1989. Inactivation of Giardia muris
and indicator organisms seeded in surface water supplies by peroxone and ozone. Environ. Sci.
Techol., 23(6): 744-745.

Wolfe, M.S. 1984. Symptomatology, diagnosis and treatment. In: Erlandsen S.L., Meyer E.A.,
eds., Giardia and Giardiasis. Plenum Press, New York, pp. 147-162.

Wolfe, M,.S. 1990. Clinical symptoms and diagnosis by traditional methods. In: Human
Parasitic Diseases, Vol. 3, Giardiasis, E.A Meyer, ed., Elsevier Science Publishers, Amsterdam,
pp. 175-186.

Wolfe, M.S. 1992. Giardiasis. Clin. Microbiol. Rev. 5(1):93-100.

Wright, R.A., Spenser, H.C., Brodsky, R.E., and Vernon, T.M. 1977. Giardiasis in Colorado: an
epidemiological study. Am. J. Epidemiol., 105:330-336.
IX-55
-------
Xiao, L. and Herd, R.P. 1993. Quantitation of Giardia cysts and Cryptosporidium oocysts in
fecal samples by direct immunofluorescence assay. J. Clin. Microbiol., 31(11):2944-2946.

Xiao, L., Saeed, K. and Herd, R.P. 1996. Efficacy of albendazole and fenbendazole against
Giardia infection in cattle. Veternary Parasitology, 61:165-170.

Zajac, A.M., LaBranche, B.S., Donoghue, A.R., Chu, T.C. 1998. Efficacy of fenbendazole in the
treatment of experimental Giardia infection in dogs. Am. J. Vet. Res., 59(l):61-63.

Zaki, A.M., DuPont, H.L., El Alamy, M.A., Arfat, R.R., et al. 1986. The detection of entero-
pathogens in acute diarrhea in a family cohort population in rural Egypt. Am. J. Trop. Med. Hyg.,
35:1013-22.

Zenian, AJ. and Gillin, ED. 1987. Intestinal mucus protects Giardia lamblia from killing by
human milk. J. Protozol., 34, 22-26.

Zimmer, J.F. and Burrington, D.B. 1986. Comparison of four protocols for the treatment of
canine giardiasis, J. Am. Anim. Hosp. Assoc., 22: 168-172.
IX-56
-------