5
Hazard Characterization

The Food Safety and Inspection Service (FSIS) draft risk assessment Hazard Characterization chapter describes a method to estimate the number of symptomatic infections resulting from the consumption of cooked ground beef contaminated with Escherichia coli O157:H7. This is considered to be a type of “dose-response assessment.” The principal concepts contained in the chapter are these:

  • generating an adjusted estimate of the annual disease burden of symptomatic E. coli O157:H7 infections that is derived by using FoodNet data and making corrections for underdiagnosis and underreporting;

  • estimating the proportion of all E. coli O157:H7 cases that are due to ground-beef exposure (ingestion);

  • deriving the dose-response function for E. coli O157:H7 by using data from three sources: the estimated annual number of symptomatic E. coli O157:H7 infections due to ground-beef exposure, the estimated number of contaminated ground-beef servings (from the exposure assessment), and the upper- and lower-bound dose-response curves based on surrogate pathogens; and

  • validating the derived dose-response function for E. coli O157:H7 by comparison with data from an outbreak associated with ground beef on which clinical, epidemiologic, and bacteriologic (isolation of pathogen from uncooked hamburger patties) data were available.



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 71
Escherichia Coli O157:H7 In Ground Beef: Review of a Draft Risk Assessment 5 Hazard Characterization The Food Safety and Inspection Service (FSIS) draft risk assessment Hazard Characterization chapter describes a method to estimate the number of symptomatic infections resulting from the consumption of cooked ground beef contaminated with Escherichia coli O157:H7. This is considered to be a type of “dose-response assessment.” The principal concepts contained in the chapter are these: generating an adjusted estimate of the annual disease burden of symptomatic E. coli O157:H7 infections that is derived by using FoodNet data and making corrections for underdiagnosis and underreporting; estimating the proportion of all E. coli O157:H7 cases that are due to ground-beef exposure (ingestion); deriving the dose-response function for E. coli O157:H7 by using data from three sources: the estimated annual number of symptomatic E. coli O157:H7 infections due to ground-beef exposure, the estimated number of contaminated ground-beef servings (from the exposure assessment), and the upper- and lower-bound dose-response curves based on surrogate pathogens; and validating the derived dose-response function for E. coli O157:H7 by comparison with data from an outbreak associated with ground beef on which clinical, epidemiologic, and bacteriologic (isolation of pathogen from uncooked hamburger patties) data were available.

OCR for page 71
Escherichia Coli O157:H7 In Ground Beef: Review of a Draft Risk Assessment REVIEW OF THE HAZARD CHARACTERIZATION CHAPTER The draft’s discussions of the baseline number of E. coli O157:H7 infections and adjustments for underdiagnosis and underreporting are scientifically sound. The logic followed is clear, and the epidemiologic data are used in a reasonable and plausible way. However, by focusing solely on the O157:H7 serotype of enterohemorrhagic E. coli and on direct contamination, the draft underestimates the overall burden of disease caused by this category of pathogen and the benefit that could derive from interventions. O157:H7 as One Member of the Enterohemorrhagic E. coli Category of E. coli Pathogen Strains of the O157:H7 serotype of Escherichia coli isolated since the early 1980s typically carry a set of virulence factors encoded by chromosomal, plasmid, and phage genes that allow them to cause a spectrum of clinical illness in humans ranging from watery diarrhea and hemorrhagic colitis to hemolytic uremic syndrome (HUS) and thrombotic thrombocytopenic purpura (TTP). It is the last two severe clinical syndromes, particularly HUS, that make O157:H7 an important public health problem in the United States because they may result in hospitalization, chronic disease, and death. The virulence properties that allow O157:H7 to cause hemorrhagic colitis, HUS, and TTP are common to a category of diarrheogenic E. coli often called enterohemorrhagic E. coli, or EHEC. It is important to recognize that a number of serotypes of E. coli other than O157:H7 also possess these properties. The common virulence factors carried by EHEC include a chromosomal pathogenicity island that encodes proteins allowing the bacteria to cause attaching and effacing lesions of the intestinal mucosa, an approximately 60 megadalton plasmid that encodes attachment factors and an enterohemolyin, and bacteriophages that encode Shiga toxins 1, 2, or both. That array of virulence properties stably carried by some E. coli strains makes them “EHEC” and renders them capable of causing the severe diseases that stimulate the demand for interventions. It should be emphasized that the vast majority of E. coli strains associated with HUS and hemorrhagic colitis carry the full array of virulence traits. Most of, although not all, those strains are in a known set of O:H serotypes of which O157:H7 is the most common. Others include O111:H8, O111:NM, O26:H11, O145:H25, and O103:H21. In contrast, E. coli strains that produce only Shiga toxin but do not have other virulence properties

OCR for page 71
Escherichia Coli O157:H7 In Ground Beef: Review of a Draft Risk Assessment are occasionally recovered from stool cultures of healthy people or people with mild diarrhea. Only uncommonly are such strains incriminated in association with HUS. The Disease Burden of Non-O157:H7 Enterohemorrhagic E. coli O157:H7 is undoubtedly the most common EHEC serotype in the United States and Canada (different E. coli serotypes predominate in other parts of the world). Nevertheless, the true prevalence of other EHEC serotypes in the United States and their contribution to the EHEC disease burden have not been well studied. One reason is that most bacteriologic surveillance for EHEC is geared specifically to the detection of O157:H7. Early studies in Canada that incriminated Shiga toxin-producing (referred to at that time as Vero-toxin producing) E. coli as a cause of HUS showed an association with multiple serotypes in addition to O157:H7 (Karmali et al., 1985). A nationwide Centers for Disease Control and Prevention study of patients with HUS showed that among pediatric cases, 85% of the patients that yielded EHEC isolates in stool cultures had O157:H7 and 15% had other serotypes (Banatvala et al., 2001). Analyses of several outbreaks of colitis and HUS that used appropriate detection techniques have clearly demonstrated that non-O157:H7 EHEC exists in the United States and is responsible for disease (McCarthy et al., 2001). Moreover, surveillance data from other countries—such as Chile (Cordovez et al., 1992; Ojeda et al., 1995; Prado et al., 1997; Rios et al., 1999), Australia (Elliott et al., 2001), the United Kingdom (Kleanthous et al., 1990), Germany (Beutin et al., 1998; Verweyen et al., 1999), and Italy (Caprioli et al., 1994)—clearly show the importance of EHEC O:H serotypes in addition to O157:H7. The Hazard Identification chapter of the FSIS draft risk assessment indicates that because E. coli O157:H7 is the most important serotype in the United States from a public-health standpoint and because there is a paucity of epidemiologic data on non-O157 serotypes, the risk assessment is limited to E. coli O157:H7. The committee acknowledges that decision but points out its implication: whatever risk to US public health the risk assessment attributes to O157:H7 as a ground-beef contaminant, it is an underestimate of the overall risk because other EHEC serotypes also cause disease. Because non-O157:H7 serotypes contribute to the EHEC disease burden—particularly as a cause of HUS—the committee suggests that the decision to exclude these serotypes be revisited. If the final risk assessment is limited to O157:H7, the committee recommends that the decision and its implications for the model be explicitly discussed in the Hazard Characterization chapter.

OCR for page 71
Escherichia Coli O157:H7 In Ground Beef: Review of a Draft Risk Assessment Estimating the Number of E. coli O157:H7 Illnesses Due to Contaminated Ground Beef (Etiologic Fraction) The draft’s analysis of the etiologic fraction is likely to underestimate the proportion of illness due to ground beef because the only mode of transmission considered is direct consumption (ingestion) of ground beef. However, contaminated raw ground beef may be epidemiologically important and lead to clinical infections even if the meat that is ultimately ingested is properly cooked and harbors no living O157:H7 or other EHEC organisms. The reason is that inadequate cooking, mistakes in food handling, or poor hygiene in the kitchen may lead to cross contamination of other food vehicles that may be eaten raw (salads, for example) or after little heating. Because the inoculum of EHEC necessary to cause disease is believed to be low, such cross contamination may be epidemiologically important (Buchanan and Doyle, 1997). The contribution of this mode of transmission could be diminished by future interventions that decrease the degree of contamination of ground beef. In contrast, interventions that aim to ensure the proper cooking of ground beef would not affect the cases of EHEC that result from compromised handling or hygiene practices and resulting cross contamination. The committee thus wishes to reiterate the comment offered in its review of the Preparation Module: it suggests that consideration be given to factoring in cross contamination as an additional step. If that is not possible, it recommends that the final risk assessment highlight more clearly the role of cross contamination in E. coli O157:H7 infection and emphasize the limitations in the model engendered by a decision to not factor it in. Deriving the Dose-Response Function for EHEC O157:H7 The E. coli O157:H7 dose-response function for the FSIS draft risk assessment was derived by applying data from three sources: the estimated annual number of symptomatic E. coli O157:H7 infections resulting from ground-beef consumption, the estimated annual number of contaminated ground-beef servings, and the lower- and upper-bound dose-response curves derived from dose-response data from experimental challenges of volunteers with Shigella dysenteriae 1 and enteropathogenic E. coli (EPEC). The first two estimates are generated directly from relevant data. In contrast, in an indirect approach, S. dysenteriae 1 and EPEC dose-response

OCR for page 71
Escherichia Coli O157:H7 In Ground Beef: Review of a Draft Risk Assessment data are used to create upper and lower brackets within which the EHEC dose-response relationship is presumed to lie. Shigella Dose-Response Relationship as Upper Limit of Bracket The FSIS draft states (p. 115): E. coli O157:H7 may be most similar to Shigella spp. with regard to transmission and infectivity; however, Shigella spp. are invasive pathogens that multiply within host epithelial cells, whereas E. coli O157:H7 does not. Both are transmitted by food, although humans are the reservoir of Shigella spp. contamination of food and water. The probability of infection with low doses of Shigella spp. is thought to be high. Most of those statements are correct. Although Shigella may indeed be transmitted by contaminated food and water vehicles, and outbreaks due to contaminated food vehicles have been described, in fact the vast body of accumulated epidemiologic data indicates that Shigella are most often transmitted through direct person-to-person contact by means of fecally contaminated hands or fomites.1 Thus, transmission of Shigella correlates with the level of personal hygiene rather than sanitation or food hygiene. Populations that manifest compromised personal hygiene are at high risk of transmission of Shigella even in industrialized countries. That explains why Shigella (particularly S. sonnei) poses a health problem in day-care centers and in custodial institutions that house mentally impaired or psychotic patients. Most outbreaks of shigellosis exhibit a protracted epidemiologic curve characteristic of person-to-person propagation rather than the abrupt pattern characteristic of point-source food-vehicle contamination. S. dysenteriae 1—the Shiga bacillus—is unique among the roughly 40 Shigella serotypes and subtypes because of the severity of clinical disease that it causes, including HUS as an uncommon complication (Khin et al., 1987; Raghupathy et al., 1978; Rahaman and Greenough, 1978); its elaboration of Shiga toxin 1 (Keusch et al., 1982; Strockbine et al., 1988); and its ability to cause pandemics that extend for years over wide geographic areas (Gangarosa et al., 1970; Mata et al., 1970; Rahaman et al., 1975). The main mode of transmission during Shiga dysentery pandemics is person-to-person spread (Ebright et al., 1984; Gangarosa et al., 1970). Those epidemiologic observations suggest that minute inocula are capable of causing shigellosis and that Shigella may be relatively resistant 1   A fomite is any inanimate object via which pathogenic organisms may be transferred (a knife, for example). A fomite does not support the growth of the pathogen.

OCR for page 71
Escherichia Coli O157:H7 In Ground Beef: Review of a Draft Risk Assessment to the effects of the gastric acid barrier that constitutes a potent defense against many other bacterial enteropathogens. Many other bacterial enteropathogens require relatively large inocula to cause clinical illness and typically require transmission via food vehicles to allow them to pass through the gastric barrier. Results of volunteer studies with multiple Shigella serotypes show that small inocula can cause notable attack rates (DuPont et al., 1989). Moreover, Shigella can cause diarrheal illness in volunteers when administered without a buffer like NaHCO3 (which neutralizes gastric acid). Older data are available from volunteer studies with S. dysenteriae 1 and S. flexneri 2a, whereas more recent dose-response data on experimental challenges with S. sonnei and S. flexneri 2a are available. The most extensive recent data come from challenges with S. flexneri 2a. Although Shigella species administered without buffer can cause diarrheal illness, administering Shigella in 150 ml of water containing 2.0 grams of NaHCO3 buffer results in higher attack rates and a more consistent clinical illness pattern (discussed further below). Accumulated epidemiologic data show that EHEC O157:H7 is most often transmitted by ingestion of a contaminated food vehicle (Griffin and Tauxe, 1991). Nevertheless, EHEC can, like Shigella, be transmitted directly person to person, particularly in young children and in the elderly (Carter et al., 1987; Ostroff et al., 1990; Pavia et al., 1990; Spika et al., 1986). Thus, as in the case of Shigella, there are reports of transmission of EHEC by direct contact within day-care centers and institutions for the elderly. In vitro studies show that EHEC (Duffy et al., 2000; Koodie and Dhople, 2001; Lin et al., 1996), like Shigella (Gorden and Small, 1993; Small et al., 1994), exhibit an unusual degree of acid resistance among bacterial enteropathogens. The fact that ground beef and some other common food vehicles responsible for transmission of EHEC O157:H7 involve cooking means that the ingested inocula, like those of Shigella, can be quite small. However, the fact that the organisms are ingested in a food vehicle undoubtedly offers the surviving EHEC a degree of protection against the gastric defense barrier. It may thus be the case that the most appropriate dose-response data to use from Shigella challenges are those involving administration of Shigella with buffer. Taken together, the above comments strongly support the relevance of the decision to use dose-response data from Shigella for the upper limit of the bracket. The data further argue that the EHEC dose-response function is likely to be very close to that of Shigella.2 Arguably, it may 2   The draft report does note, on p. 119, that the dose-response function “more clearly approximates that estimated for Shigella dysenteriae than for EPEC.”

OCR for page 71
Escherichia Coli O157:H7 In Ground Beef: Review of a Draft Risk Assessment be most appropriate to use dose-response data from experimental challenges with Shigella administered with buffer. Vehicle of Transmission and Mode of Ingestion Affect the Dose-Response Curve of Bacterial Enteropathogens In attempting to derive a dose-response function for EHEC by using dose-response data from other bacterial enteropathogens, such as Shigella and EPEC, the draft’s authors focus only on dose. They do not address the precise context in which the dose (inoculum) is ingested. In fact, such context is fundamental when considering the epidemiologic relevance of dose-response data. That is best illustrated by using dose-response data from Vibrio cholerae O1, because this example constitutes an extreme. Consider the following observations: When fasting healthy adult US volunteers ingested 106 colonyforming units (CFU) of V. cholerae O1 suspended in water without either buffer or food, neither infection nor diarrhea ensued (Cash et al., 1974; Levine et al., 1981). When fasting healthy adult US volunteers ingested 106 CFU of V. cholerae O1 with NaHCO3 buffer (which neutralizes gastric acid), about 90% became infected and 90% developed cholera diarrhea (Levine et al., 1979a, 1981, 1988; Tacket et al., 1995a). When fasting healthy adult US volunteers ingested 106 CFU of V. cholerae O1 with food (a quasi-Bengali meal), the attack rate for cholera diarrhea (about 90%), the infection rate, and the clinical severity were identical with those observed when the same dose was administered with NaHCO3 buffer (Levine et al., 1981). When fasting healthy adult US volunteers were given much lower doses of V. cholerae O1 (as low as 103 CFU) with NaHCO3 buffer, attack rates for diarrhea remained high (67%), but its severity diminished (Levine et al., 1981). Those results emphasize that both the dose and the context in which the bacteria are ingested are important determinants of disease. The same dose may be innocuous or cause cholera in 90% of subjects, depending on how the inoculum is ingested. Although the phenomenon is less prominent with some other bacterial enteropathogens, it is nevertheless a factor, even with Shigella. Effect of Mode of Ingestion on S. Flexneri 2a Attack Rate Because the upper bound of the presumed E. coli O157:H7 dose-re

OCR for page 71
Escherichia Coli O157:H7 In Ground Beef: Review of a Draft Risk Assessment sponse function that is set by the S. dysenteriae 1 data is critical, it is important to examine the data critically. Two constraints can be cited with respect to the S. dysenteriae 1 dose-response data that were used to craft an upper limit of the bracket. First, the studies cited in the draft were carried out more than 30 years ago (Levine et al., 1973), and there were some shortcomings in the clinical methods used at that time; in the ensuing decades, methods used in challenge studies have become more rigorous. Second, it is now recognized from challenge studies with S. flexneri 2a that the dose-response relationship of Shigella can be substantially influenced by how the challenge inoculum is administered to the volunteers in the experimental challenge studies. Those observations come from multiple challenge studies with S. flexneri 2a that involve challenge inocula prepared by the same laboratory (the Center for Vaccine Development of the University of Maryland School of Medicine). Clinical supervision of the studies that generated eight of the nine datasets was provided by one institution (the Center for Vaccine Development; the Walter Reed Army Institute of Research provided clinical supervision of the remaining trial). Five challenge studies were carried out in which inocula containing 103 CFU of S. flexneri 2a were fed in 45 ml of skim milk to immunologically naive healthy adult community volunteers; the overall clinical attack rate was 48% (24 of 50), with a range of 33% to 58% for attack rates in individual challenge studies. In contrast, three challenges were undertaken in which 103 CFU suspended in 150 ml of water containing 2.0 gram of NaHCO3 (to neutralize gastric acid) were fed to groups of immunologically naive volunteers; these three challenges resulted in an overall attack rate of 88% (29 of 33), with individual study attack rates of 86%, 86%, and 92%. The difference in overall response to the same dose administered by two methods is highly significant (p < 0.01). The one challenge study in which volunteers ingested 102 CFU with NaHCO3 led to an attack rate of 43% (three of seven)—similar to the attack rate encountered when 103 CFU were administered without buffer in 45 ml of skim milk. The results of the clinical trials are summarized in Table 5-1. The above data clearly demonstrate the effect of mode of administration of a Shigella inoculum on clinical response, but only two data points are available to construct the dose-response curve. It is notable that the dose-response curve from modern challenge studies with S. flexneri 2a administered with buffer is similar to the S. dysenteriae 1 dose-response curve based on the early challenges that administered that serotype without buffer. That suggests that the upper limit of the bracket, as constructed, is valid even though the effect of buffering was not factored in. It is unlikely that EHEC would elicit a higher attack rate than S. dysenteriae 1 at the same dose. The committee suggests that—in order to strengthen

OCR for page 71
Escherichia Coli O157:H7 In Ground Beef: Review of a Draft Risk Assessment TABLE 5-1 Attack rates with different doses (CFU) of S. flexneri 2a administered without or with buffer Challenge Inoculum (CFU) Attack Rate When Administered Without Buffera Reference Attack Rate When Administered with Bufferb Reference 102   3/7 (43%)   103 58% (7/12) Kotloff et al., 1992 92% (11/12) Kotloff et al., 1995a 103 33% (3/9) Kotloff et al., 1992 86% (12/14) Kotloff et al., 1995b 103 45% (5/11) Tacket et al., 1992 86% (6/7) Coster et al., 1999 103 57% (4/7) Tacket et al., 1992     103 45% (5/11) Tacket et al., 1992 aGiven in 45 ml of skim milk. bSuspended in 150 ml of water containing 2.0 g of NaHCO3.

OCR for page 71
Escherichia Coli O157:H7 In Ground Beef: Review of a Draft Risk Assessment the scientific foundation for the decision to use dose-response data for S. dysenteriae 1 to construct the upper bracket—the final risk assessment discuss how the mode of ingestion affects expected attack rates. Extrapolation of Dose-Response Data to High-Risk Age Groups Whatever dose-response data (from studies with Shigella or other bacterial enteropathogens) are used as surrogates to help to estimate the dose-response relationship for O157:H7 and other EHEC, it must be remembered that the data are derived from experimental challenge studies in healthy adults. One must extrapolate the data to assess their relevance to the dose-response relationship for toddlers, preschool children, and the elderly, the age groups that suffer the highest incidence of severe clinical outcomes. For Shigella, the dose-response data derived from adults appear to be compatible with epidemiologic patterns of endemic shigellosis in which peak incidence rates are observed in children 1–4 years old. Comparable data for the elderly are lacking. EPEC Dose-Response as Lower Limit of Bracket The use of EPEC dose-response data as the lower limit is reasonable but somewhat more problematic than the use of Shigella data to set the upper bracket. One argument in favor of using EPEC data is that EPEC, like EHEC, contain the chromosomal locus that encodes genes involved in attaching to and effacing intestinal mucosa. However, epidemiologic data do not support the relevance of this model. Few data incriminate EPEC as a cause of outbreaks of diarrhea in older children or adults (Levine, 1987; Levine and Edelman, 1984). Rather, in the wild, EPEC are pathogenic in very young infants. Indeed, in developing countries, the pathogen can be incriminated only within the first 6 months of life, when a substantially higher rate of isolation of EPEC is found in cases with diarrhea than in nondiarrheal controls (Levine et al., 1993). Beyond that age group, the isolation rates are equal. When EPEC are fed to adult volunteers, moderate to high attack rates of diarrheal illness ensue (Bieber et al., 1998; Donnenberg et al., 1993; Levine et al., 1978). However, the inocula required tend to be rather large ≥ 108 logs) and the bacteria must be fed with buffer to protect them from gastric acid (Levine et al., 1978). Moreover, the incubation period is extraordinarily short, and the diarrheal illness tends to be short-lived (although severe cholera-like purging was induced by one strain at high dosage) (Levine et al., 1978). EPEC is indeed likely to be less pathogenic than EHEC with respect to the inoculum required to induce a clinical re

OCR for page 71
Escherichia Coli O157:H7 In Ground Beef: Review of a Draft Risk Assessment sponse. However, EPEC challenge of adults is an artificial system not usually found in nature. Role of Host Factors in Clinical Response to Challenge with Bacterial Enteropathogens It is obvious in experimental challenge studies that different healthy adults may respond differently to ingestion of identical inocula of a bacterial enteropathogen. Prior immunity or nonspecific innate immune mechanisms can partly explain the differences. However, other host factors that represent genetic susceptibilities (or protective factors) may also play an important role in the clinical response. The extreme susceptibility of persons of blood group O to cholera gravis and the role of diminished gastric acid production in the development of severe cholera are examples (Levine et al., 1979b; Nalin et al., 1978; Tacket et al., 1995b). Dose-Response Curves of Other Possible Bacterial Enteropathogens That Might Serve as Alternatives to Set Lower Limit of Bracket Dose-response studies of other bacterial enteropathogens have been carried out in healthy volunteers, including enterotoxigenic E. coli (ETEC), enteroinvasive E. coli (EIEC), enteroaggregative E. coli (EAggEC), diffusely adherent E. coli (DAEC), Campylobacter jejuni, Salmonella enterica serovar Typhi, and V. cholerae O1, O139, and non-O1/ non-O139. Some of these fall in between the dose-response of Shigella and EPEC and, arguably, might serve as a more rational source for a lower limit to bracket the presumed EHEC dose-response relationship. Table 5-2 summarizes epidemiologic, pathogenetic, and clinical characteristics of the enteropathogens. The committee believes that the EPEC dose-response relationship is a conservative choice for the lower limit and suggests that—if the bounding approach continues to be used in the final risk assessment— consideration be given to alternatives like these that might reflect the pathogenicity of EHEC better. Uncertainty in Cases and Exposure Distribution The FSIS draft risk assessment properly notes that “uncertainty about the E. coli O157:H7 dose-response function extends almost across the full range enveloped by the lower and upper bound curves” (p. 119). This is an important statement and is likely to be correct for all the reasons mentioned in the other comments. Overall, the draft chapter’s authors did an elegant job in generating an EHEC dose-response function. According to

OCR for page 71
Escherichia Coli O157:H7 In Ground Beef: Review of a Draft Risk Assessment TABLE 5-2 Epidemiologic, pathogenetic, and clinical characteristics of EHEC-like enteropathogens Pathogen Reservoir Foodborne? Epithelial-Cell Invasiveness? Natural Pathogen for Adults and Children? References ETEC (human) Humans Yes No Yes DuPont et al., 1971; Levine et al., 1977; Levine et al., 1979c EIEC Humans Yes Yes Yes DuPont et al., 1971 EAggEC Humans ? No ? Nataro et al., 1995 DAEC ? ? No ? Tacket et al., 1990 C. jejuni Animals Yes Yes Yes Black et al., 1988 S. Typhi Humans Yes Yes Yes Hornick et al., 1970 V. cholerae O1 Environment Yes (seafood; water) No Yes Levine et al., 1981; Levine et al., 1979b V. cholerae O139 Environment Yes No Yes Tacket et al., 1995a V. cholerae non-O1/non-O139 Environment Yes (seafood) Some Yes Morris et al., 1990

OCR for page 71
Escherichia Coli O157:H7 In Ground Beef: Review of a Draft Risk Assessment Figure 4-5 of the draft, the 50th percentile derived dose-response curve predicts an ingested dose of ~4.8 logs (~63,000 organisms) of EHEC will result in clinical illness in 50% of subjects. The dose of S. dysenteriae 1 that is expected to cause a 50% attack rate is ~2.9 logs (~740 organisms). Thus, the FSIS draft model suggests that the dose-response curve for EHEC is somewhat to the right of that of Shigella (requires more organisms). This model suggests that up to 2 logs more EHEC must be ingested to result in the same attack rate as Shigella. Severe Clinical Outcomes and Sensitive Populations The FSIS draft offers two observations regarding severe clinical outcomes and sensitive populations that the committee would like to highlight. It asserts that “estimating the clinical outcomes of symptomatic infection is essential for future cost-benefit analyses of intervention options” (p. 121). That is an important point. As previously stated, it is the propensity for EHEC to cause severe illness, chronic disease, and death—particularly in young children—that makes it an important public-health problem and stimulates the demand for interventions. Thus, the risk assessment should focus primarily on HUS (for simplicity, TTP may be considered a variant of HUS seen in adults) as a clinical outcome. If O157:H7 and other EHEC caused only gastroenteritis, they would be in the same category, as a public-health problem, as Campylobacter jejuni and nontyphoidal Salmonella enterica and might well have a less prominent public profile. By extension, it may be strongly argued that whatever interventions are contemplated, both public health authorities and the general public will expect them to significantly diminish the burden of HUS and other serious outcomes. The draft also notes that “the reason why children have the highest reported incidence of E. coli O157:H7 infection is not known” (p. 123). It proceeds to offer several possible explanations, including differences in health-care patterns (health care may be more likely to be sought for children), exposure differences, and biologic differences. The draft offers a fair and honest statement of the lack of data to explain the observed age differences in clinical expression. The lack of data, however, should not be an obstacle to evaluating the disproportionate risk that children—and the elderly (who are omitted from the chapter’s discussion of sensitive subpopulations)—manifest for severe clinical outcomes. This issue and the committee’s recommendations regarding it are addressed in the review of the draft risk characterization—the next chapter of this review.

OCR for page 71
Escherichia Coli O157:H7 In Ground Beef: Review of a Draft Risk Assessment Validation of E. coli O157:H7 Dose-Response Function Using Outbreak Data Using data from a 1992–1993 hamburger-associated outbreak that included clinical, epidemiologic, and microbiologic information, an analysis presented in the FSIS draft estimates that the 45-gram contaminated hamburgers consumed during the outbreak harbored a median of about 96 CFU of E. coli O157:H7 before cooking. That is similar to the data of Tuttle et al. (1999), who calculated that the median dose of O157:H7 in hamburger patties associated with a large outbreak was 67.5 CFU before cooking. Assuming that even inadequate cooking results in a diminution of the inoculum, the dose ingested must indeed be quite small. Although not mentioned in the text, this is further evidence that the dose required to cause EHEC disease is similar to the low doses of Shigella that can cause disease. The concentration of pediatric cases in the outbreak also under-scores the importance of focusing attention on this population in the risk characterization. SUMMARY REMARKS Overall, the FSIS draft risk assessment’s authors did an excellent job with the hazard characterization, given the limitations of data and gaps in data. Their model is elegant, and they use a logical progression of steps in this chapter. One might argue with the use of EPEC dose-response data to serve as the lower limit of a presumed EHEC dose-response function. In fact, epidemiologic and microbiologic data suggest that the true EHEC dose-response function is likely to resemble that of Shigella. The failure to account for non-O157:H7 as a cause of hemorrhagic colitis and HUS underestimates the overall burden of EHEC disease in the United States and the benefits that may derive from future interventions. The true burden of severe EHEC disease probably is 15–20% greater than estimates based on O157:H7 alone. REFERENCES Banatvala N, Griffin PM, Greene KD, Barrett TJ, Bibb WF, Green JH, Wells JG. 2001. The United States National Prospective Hemolytic Uremic Syndrome Study: Microbiologic, serologic, clinical, and epidemiologic findings. Journal of Infectious Diseases 183:1063–1070. Beutin L, Zimmermann S, Gleier K. 1998. Human infections with Shiga toxin-producing Escherichia coli other than serogroup O157 in Germany. Emerging Infectious Diseases 4:635– 639. Bieber D, Ramer SW, Wu CY, Murray WJ, Tobe T, Fernandez R, Schoolnik GK. 1998. Type IV pili, transient bacterial aggregates, and virulence of enteropathogenic Escherichia coli. Science 280:2114–2118. Buchanan RL, Doyle MP. 1997. Foodborne disease significance of Escherichia coli O157:H7 and other enterohemorrhagic E. coli. Food Technology 51:69–75.

OCR for page 71
Escherichia Coli O157:H7 In Ground Beef: Review of a Draft Risk Assessment Black RE, Levine MM, Clements ML, Hughes TP, Blaser MJ. 1988. Experimental Campylobacter jejuni infection in humans. Journal of Infectious Diseases 157:472–479. Caprioli A, Luzzi I, Rosmini F, Resti C, Edefonti A, Perfumo F, Farina C, Goglio A, Gianviti A, Rizzoni G. 1994. Community-wide outbreak of hemolytic-uremic syndrome associated with non-O157 verocytotoxin-producing Escherichia coli. Journal of Infectious Diseases 169:208–211. Carter AO, Borczyk AA, Carlson JA, Harvey B, Hockin JC, Karmali MA, Krishnan C, Korn DA, Lior H. 1987. A severe outbreak of Escherichia coli O157:H7-associated hemorrhagic colitis in a nursing home. New England Journal of Medicine 317:1496–1500. Cash RA, Music SI, Libonati JP, Snyder MJ, Wenzel RP, Hornick RB. 1974. Response of man to infection with Vibrio cholerae. I. Clinical, serologic, and bacteriologic responses to a known inoculum. Journal of Infectious Diseases 129:45–52. Cordovez A, Prado V, Maggi L, Cordero J, Martinez J, Misraji A, Rios R, Soza G, Ojeda A, Levine MM. 1992. Enterohemorrhagic Escherichia coli associated with hemolytic-uremic syndrome in Chilean children. Journal of Clinical Microbiology 30:2153–2157. Coster TS, Hoge CW, VanDeVerg LL, Hartman AB, Oaks EV, Venkatesan MM, Cohen D, Robin G, Fontaine-Thompson A, Sansonetti PJ, Hale TL. 1999. Vaccination against shigellosis with attenuated Shigella flexneri 2a strain SC602. Infection and Immunity 67:3437–3443. Donnenberg MS, Tacket CO, James SP, Losonsky G, Nataro JP, Wasserman SS, Kaper JB, Levine MM. 1993. Role of the eaeA gene in experimental enteropathogenic Escherichia coli infection. Journal of Clinical Investigation 92:1412–1417. Duffy LL, Grau FH, Vanderlinde PB. 2000. Acid resistance of enterohaemorrhagic and generic Escherichia coli associated with foodborne disease and meat. International Journal of Food Microbiology 60:83–89. DuPont HL, Formal SB, Hornick RB, Snyder MJ, Libonati JP, Sheahan DG, LaBrec EH, Kalas JP . 1971. Pathogenesis of Escherichia coli diarrhea. New England Journal of Medicine 285: 1–9. DuPont HL, Levine MM, Hornick RB, Formal SB. 1989. Inoculum size in shigellosis and implications for expected mode of transmission. Journal of Infectious Diseases 159:1126–1128. Ebright JR, Moore EC, Sanborn WR, Schaberg D, Kyle J, Ishida K. 1984. Epidemic Shiga bacillus dysentery in Central Africa. American Journal of Tropical Medicine and Hygiene 33:1192–1197. Elliott EJ, Robins-Browne RM, O’Loughlin EV, Bennett-Wood V, Bourke J, Henning P, Hogg GG, Knight J, Powell H, Redmond D. 2001. Nationwide study of haemolytic uraemic syndrome: Clinical, microbiological, and epidemiological features. Archives of Disease in Childhood 85:125–131. Gangarosa EJ, Perera DR, Mata LJ, Mendizabal-Morris C, Guzman G, Reller LB. 1970. Epidemic Shiga bacillus dysentery in Central America. II. Epidemiologic studies in 1969. Journal of Infectious Diseases 122:181–190. Gorden J, Small PLC. 1993. Acid resistance in enteric bacteria. Infection and Immunity 61:364– 367. Griffin PM, Tauxe RV. 1991. The epidemiology of infections caused by Escherichia coli O157:H7, other enterohemorrhagic E. coli, and the associated hemolytic uremic syndrome. Epidemiologic Reviews 13:60–98. Hornick RB, Greisman SE, Woodward TE, DuPont HL, Dawkins AT, Snyder MJ. 1970. Typhoid fever; pathogenesis and immunologic control. New England Journal of Medicine 283:686–691, 739–746.

OCR for page 71
Escherichia Coli O157:H7 In Ground Beef: Review of a Draft Risk Assessment Karmali MA, Petric M, Lim C, Fleming PC, Arbus GS, Lior H. 1985. The association between idiopathic hemolytic uremic syndrome and infection by verotoxin-producing Escherichia coli. Journal of Infectious Diseases 151:775–782. Keusch GT, Donohue-Rolfe A, Jacewicz M. 1982. Shigella toxin(s): Description and role in diarrhea and dysentery. Pharmacology & Therapeutics 15:403–438. Khin MU, Myo K, Tin A, Myo MA, Soe SA, Thane-Oke KM, Khin TN. 1987. Clinical features, including haemolytic-uraemic syndrome, in Shigella dysenteriae type 1 infection in children of Rangoon. Journal of Diarrhoeal Diseases Research 5:175–177. Kleanthous H, Smith HR, Scotland SM, Gross RJ, Rowe B, Taylor CM, Milford DV. 1990. Haemolytic uraemic syndromes in the British Isles, 1985-8: Association with verocytotoxin producing Escherichia coli. Part 2: Microbiological aspects. Archives of Disease in Childhood 65:722–727. Koodie L, Dhople AM. 2001. Acid tolerance of Escherichia coli O157:H7 and its survival in apple juice. Microbios 104:167–175. Kotloff KL, Herrington DA, Hale TL, Newland JW, Van de Verg L, Cogan JP, Snoy PJ, Sadoff JC, Formal SB, Levine MM. 1992. Safety, immunogenicity, and efficacy in monkeys and humans of invasive Escherichia coli K-12 hybrid vaccine candidates expressing Shigella flexneri 2a somatic antigen. Infection and Immunity 60:2218–2224. Kotloff KL, Losonsky GA, Nataro JP, Wasserman SS, Hale TL, Taylor DN, Newland JW, Sadoff JC, Formal SB, Levine MM. 1995a. Evaluation of the safety, immunogenicity and efficacy in healthy adults of four doses of live oral hybrid Escherichia coli-Shigella flexneri 2a vaccine strain EcSf2a-2. Vaccine 13:495–502. Kotloff KL, Nataro JP, Losonsky GA, Wasserman SS, Hale TL, Taylor DN, Sadoff JC, Levine MM. 1995b. A modified Shigella volunteer challenge model in which the inoculum is administered with bicarbonate buffer: Clinical experience and implications for Shigella infectivity. Vaccine 13:1488–1494. Levine MM. 1987. Escherichia coli that cause diarrhea: enterotoxigenic, enteropathogenic, enteroinvasive, enterohemorrhagic, and enteroadherent. Journal of Infectious Diseases 155:377–389. Levine MM, Edelman R. 1984. Enteropathogenic Escherichia coli of classic serotypes associated with infant diarrhea: epidemiology and pathogenesis. Epidemiologic Reviews 6:31–51. Levine MM, Dupont HL, Formal SB, Hornick RB, Takeuchi A, Gangarosa EJ, Snyder MJ, Libonati JP. 1973. Pathogenesis of Shigella dysenteriae 1 (Shiga) dysentery. Journal of Infectious Diseases 127(3):261–270. Levine MM, Caplan ES, Waterman D, Cash RA, Hornick RB, Snyder MJ. 1977. Diarrhea caused by Escherichia coli that produce only heat-stable enterotoxin. Infection and Immunity 17:78–82. Levine MM, Bergquist EJ, Nalin DR, Waterman DH, Hornick RB, Young CR, Sotman S, Rowe B. 1978. Escherichia coli strains that cause diarrhoea but do not produce heatlabile or heat-stable enterotoxins and are non-invasive. Lancet 1:1119–1122. Levine MM, Nalin DR, Craig JP, Hoover D, Bergquist EJ, Waterman D, Holley HP, Hornick RB, Pierce NP, Libonati JP. 1979a. Immunity of cholera in man: Relative role of antibacterial versus antitoxic immunity. Transactions of the Royal Society of Tropical Medicine and Hygiene 73:3–9. Levine MM, Nalin DR, Rennels MB, Hornick RB, Sotman S, Van Blerk G, Hughes TP, O’Donnell S, Barua D. 1979b. Genetic susceptibility to cholera. Annals of Human Biology 6(4):369–374. Levine MM, Nalin DR, Hoover DL, Bergquist EJ, Hornick RB, Young CR. 1979c. Immunity to enterotoxigenic Escherichia coli. Infection and Immunity 23:729–736.

OCR for page 71
Escherichia Coli O157:H7 In Ground Beef: Review of a Draft Risk Assessment Levine MM, Black RE, Clements ML, Nalin DR, Cisneros L, Finkelstein RA. 1981. Volunteer studies in development of vaccines against cholera and enterotoxigenic Escherichia coli: A review. In: Holme T, Holmgren J, Merson MH, Mollby R, eds. Acute Enteric Infections in Children: New Prospects for Treatment and Prevention. Amsterdam: Elsevier/ North-Holland Biomedical Press, pp. 443–459. Levine MM, Kaper JB, Herrington D, Ketley J, Losonsky G, Tacket CO, Tall B, Cryz S. 1988. Safety, immunogenicity, and efficacy of recombinant live oral cholera vaccines, CVD 103 and CVD 103-HgR. Lancet 2:467–470. Levine MM, Ferreccio C, Prado V, Cayazzo M, Abrego P, Martinez J, Maggi L, Baldini M, Martin W, Maneval D, Kay B, Guers L, Lior H, Wasserman SS, Nataro JP. 1993. Epidemiologic studies of Escherichia coli infections in a low socioeconomic level periurban community in Santiago, Chile. American Journal of Epidemiology 138:849–869. Lin J, Smith MP, Chapin KC, Baik HS, Bennett GN, Foster JW. 1996. Mechanisms of acid resistance in enterohemorrhagic Escherichia coli. Applied and Environmental Microbiology 62:3094–3100. Mata L, Gangarosa E, Caceres A, Perera D, Mejicanos M. 1970. Epidemic Shiga bacilllus dysentery in Central America. I. Etiologic investigations in Guatemala, 1969. Journal of Infectious Diseases 122:170–180. McCarthy TA, Barrett NL, Hadler JL, Salsbury B, Howard RT, Dingman DW, Brinkman CD, Bibb WF, Cartter ML. 2001. Hemolytic-uremic syndrome and Escherichia coli O121 at a lake in Connecticut, 1999. Pediatrics 108:E59. Morris JG Jr, Takeda T, Tall BD, Losonsky GA, Bhattacharya SK, Forrest BD, Kay BA, Nishibuchi M. 1990. Experimental non-01 group1 Vibrio cholerae gastroenteritis in humans. Journal of Clinical Investigation 85:697–705. Nalin DR, Levine RJ, Levine MM, Hoover D, Bergquist E, McLaughlin J, Libonati J, Alam J, Hornick RB. 1978. Cholera, non-vibrio cholera, and stomach acid. Lancet 2:856–859. Nataro JP, Deng Y, Cookson S, Cravioto A, Savarino SJ, Guers LD, Levine MM, Tacket CO. 1995. Heterogeneity of enteroaggregative Escherichia coli virulence demonstrated in volunteers. Journal of Infectious Diseases 171:465–468. Ojeda A, Prado V, Martinez J, Arellano C, Borczyk A, Johnson W, Lior H, Levine MM. 1995. Sorbitol-negative phenotype among enterohemorrhagic Escherichia coli strains of different serotypes and from different sources. Journal of Clinical Microbiology 33:2199– 2201. Ostroff SM, Griffin PM, Tauxe RV, Shipman LD, Greene KD, Wells JG, Lewis JH, Blake PA, Kobayashi JM. 1990. A statewide outbreak of Escherichia coli O157:H7 infections in Washington State. American Journal of Epidemiology 132:239–247. Pavia AT, Nichols CR, Green DP, Tauxe RV, Mottice S, Greene KD, Wells JG, Siegler RL, Brewer ED, Hannon D. 1990. Hemolytic-uremic syndrome during an outbreak of Escherichia coli O157:H7 infections in institutions for mentally retarded persons: Clinical and epidemiologic observations. Jornal de Pediatria 116:544–551. Prado V, Martinez J, Arellano C, Levine MM. 1997. [Temporal variation of genotypes and serotypes of enterohemorrhagic E. coli isolated from Chilean children with intestinal infections or hemolytic uremic syndrome]. Revista Medica de Chile 125:291–297. Raghupathy P, Date A, Shastry JC, Sudarsanam A, Jadhav M. 1978. Haemolytic-uraemic syndrome complicating Shigella dysentery in south Indian children. British Medical Journal 1(6126):1518–1521. Rahaman MM, Greenough WB III. 1978. Shigellosis and haemolytic uraemic syndrome. Lancet 1(8072):1051. Rahaman MM, Khan MM, Aziz KMS, Islam MS, Kibriya AK. 1975. An outbreak of dysentery caused by Shigella dysenteriae type 1 on a Coral Island in the Bay of Bengal. Journal of Infectious Diseases 132(1):15–19.

OCR for page 71
Escherichia Coli O157:H7 In Ground Beef: Review of a Draft Risk Assessment Rios M, Prado V, Trucksis M, Arellano C, Borie C, Alexandre M, Fica A, Levine MM. 1999. Clonal diversity of Chilean isolates of enterohemorrhagic Escherichia coli from patients with hemolytic-uremic syndrome, asymptomatic subjects, animal reservoirs, and food products. Journal of Clinical Microbiology 37:778–781. Small P, Blankenhorn D, Welty D, Zinser E, Slonczewski JL. 1994. Acid and base resistance in Escherichia coli and Shigella flexneri: Role of rpoS and growth pH. Journal of Bacteriology 176:1729–1737. Spika JS, Parsons JE, Nordenberg D, Wells JG, Gunn RA, Blake PA. 1986. Hemolytic uremic syndrome and diarrhea associated with Escherichia coli O157:H7 in a day care center. Jornal de Pediatria 109:287–291. Strockbine NA, Jackson MP, Sung LM, Holmes RK, O’Brien AD. 1988. Cloning and sequencing of the genes for Shiga toxin from Shigella dysenteriae type 1. Journal of Bacteriology 170:1116–1122. Tacket CO, Moseley SL, Kay B, Losonsky G, Levine MM. 1990. Challenge studies in volunteers using Escherichia coli strains with diffuse adherence to HEp-2 cells. Journal of Infectious Diseases 162:550–552. Tacket CO, Binion SB, Bostwick E, Losonsky GA, Roy MJ, Edelman R. 1992. Efficacy of bovine milk immunoglobulin concentrate in preventing illness after Shigella flexneri challenge. American Journal of Tropical Medicine and Hygiene 47:276–283. Tacket CO, Losonsky G, Nataro JP, Comstock L, Michalski J, Edelman R, Kaper JB, Levine MM. 1995a. Initial clinical studies of CVD 112 Vibrio cholerae O139 live oral vaccine: safety and efficacy against experimental challenge. Journal of Infectious Diseases 172:883–886. Tacket CO, Losonsky G, Nataro JP, Wasserman SS, Cryz SJ, Edelman R, Levine MM. 1995b. Extension of the volunteer challenge model to study South American cholera in a population of volunteers predominantly with blood group antigen O. Transactions of the Royal Society of Tropical Medicine and Hygiene 89:75–77. Tuttle J, Gomez T, Doyle MP, Wells JG, Zhao T, Tauxe RV, Griffin PM. 1999. Lessons from a large outbreak of Escherichia coli O157:H7 infections: Insights into the infectious dose and method of widespread contamination of hamburger patties. Epidemiology and Infection 122:185–192. Verweyen HM, Karch H, Allerberger F, Zimmerhackl LB. 1999. Enterohemorrhagic Escherichia coli (EHEC) in pediatric hemolytic-uremic syndrome: A prospective study in Germany and Austria. Infection 27:341–347.