Appendix D-2
The Prospects for Immunizing Against Escherichia coli
DISEASE DESCRIPTION
Escherichia coli strains that produce enterotoxins are important causes of diarrhea in developing countries. The illness resulting from infection with these organisms ranges from mild diarrhea to a dehydrating diarrheal illness. It is usually characterized by watery, nonbloody diarrhea lasting up to 7 days. Dehydration may result if stool losses are not replaced, and in severe cases the clinical features are similar to those of cholera (Dupont, 1982).
PATHOGEN DESCRIPTION
Information on the pathogenesis of E. coli diarrhea pertinent to vaccine development has recently been exhaustively reviewed by Levine et al. (1983). The information presented below summarizes that review.
Escherichia coli are gram-negative bacilli that are found in the normal intestinal flora of humans and many animals; but certain strains can cause enteric illness. Researchers first recognized that some E. coli produce enterotoxins in the late 1960s. Enterotoxigenic E. coli are now known to produce two plasmid-mediated enterotoxins: one is heat-labile (LT) and the other is heat-stable (ST).
Heat-labile toxin is a high molecular weight protein that resembles cholera toxin in structure, function, and mechanism of action. It is composed of one enzymatically active A subunit joined to five binding B subunits. The receptors for the B subunits found on enterocytes include GM1 ganglioside and a recently described glycoprotein. After binding to the enterocyte, the A subunit gains entrance to the cell and activates adenylate cyclase, leading to an accumulation of cyclic AMP. This in turn causes secretion by the crypt cells and decreased absorp-
The committee gratefully acknowledges the efforts of R.E.Black, who prepared major portions of this appendix, and the advice and assistance of C.C.J.Carpenter. The committee assumes full responsibility for all judgments and assumptions.
tion by villus tip cells, resulting in loss of electrolyte-rich fluid, which appears clinically as watery diarrhea.
E. coli heat-stable toxin is a small polypeptide that activates guanylate cyclase activity, leading to an accumulation of cyclic GMP. This alters the enterocyte membrane function, resulting in net secretion and diarrhea.
Enterotoxigenic E. coli have accessory virulence properties in addition to LT or ST that are important factors in their ability to cause disease. The best-characterized accessory virulence properties are colonization factors that allow the E. coli to adhere to specific receptors on enterocytes of the proximal small intestinal mucosa. These colonization factors have been identified as fimbriae found on the surface of bacteria. There are at least three distinct types of adhesion fimbriae detected in enterotoxigenic E. coli of human origin, including colonization factor antigen (CFA) I, CFA II, and E8775 fimbriae. Several other fimbriae that may serve as colonization factors of E. coli also have been described.
HOST IMMUNE RESPONSE
Persons with illness caused by enterotoxigenic E. coli develop both serum and intestinal secretory IgA (SIgA) antibody responses to the homologous O antigen. The serum antibody against the O antigen is predominantly in the IgM class and peaks approximately 10 days after the onset of illness. In addition, persons who have diarrhea due to LT-producing strains of E. coli manifest significant rises in serum antitoxin. Increases in the level of secretory IgA antitoxin in intestinal fluid also have been detected after infection with LT-producing E. coli. Persons infected and ill with E. coli that produce only ST do not appear to develop neutralizing or binding antitoxin to ST. After infection with strains producing colonization factor antigens, rises in serum IgG and intestinal secretory IgA antibodies have been demonstrated (Levine et al., 1983).
DISTRIBUTION OF DISEASE
Geographic Distribution of Disease
Enterotoxigenic E. coli have been shown to cause diarrhea worldwide, but seem to be far more common as a cause of diarrhea in developing countries (Dupont, 1982).
Disease Burden Estimates
The disease burden estimates for enterotoxigenic E. coli assuming the current level of intervention are shown in Table D-2.1. Table D-2.2 shows the disease burden estimates based on a scenario in which oral rehydration therapy prevents 50 percent of enterotoxigenic E. coli
TABLE D-2.1 Disease Burden: Enterotoxigenic E. coli
|
|
|
Under 5 Years |
5–14 Years |
15–59 Years |
60 Years and Over |
||||
Morbidity Category |
Description |
Condition |
Number of Cases |
Duration |
Number of Cases |
Duration |
Number of Cases |
Duration |
Number of Cases |
Duration |
A |
Moderate localized pain and/or mild systemic reaction, or impairment requiring minor change in normal activities, and associated with some restriction of work activity |
Mild diarrhea |
353,503,000 |
4 |
142,442,000 |
4 |
84,786,000 |
4 |
9,586,000 |
4 |
B |
Moderate pain and/or moderate impairment requiring moderate change in normal activities, e.g., housebound or in bed, and associated with temporary loss of ability to work |
Moderately severe diarrhea |
31,423,000 |
5 |
1,151,000 |
5 |
685,000 |
5 |
404,000 |
5 |
C |
Severe pain, severe short-term impairment, or hospitalization |
Severe diarrhea |
6,983,000 |
6 |
230,000 |
6 |
171,000 |
6 |
101,000 |
6 |
D |
Mild chronic disability (not requiring hospitalization, institutionalization, or other major limitation of normal activity, and resulting in minor limitation of ability to work) |
|
|
n.a. |
|
n.a. |
|
n.a. |
|
n.a. |
E |
Moderate to severe chronic disability (requiring hospitalization, special care, or other major limitation of normal activity, and seriously restricting ability to work) |
|
|
n.a. |
|
n.a. |
|
n.a. |
|
n.a. |
F |
Total impairment |
|
|
n.a. |
|
n.a. |
|
n.a. |
|
n.a. |
G |
Reproductive impairment resulting in infertility |
|
|
n.a. |
|
n.a. |
|
n.a. |
|
n.a. |
H |
Death |
|
698,000 |
n.a. |
46,000 |
n.a. |
26,000 |
n.a. |
5,000 |
n.a. |
TABLE D-2.2 Disease Burden: Enterotoxigenic E. coli. Assuming Increased Use of Oral Rehydration Therapy
|
|
|
Under 5 Years |
5–14 Years |
15–59 Years |
60 Years and Over |
||||
Morbidity Category |
Description |
Condition |
Number of Cases |
Duration |
Number of Cases |
Duration |
Number of Cases |
Duration |
Number of Cases |
Duration |
A |
Moderate localized pain and/or mild systemic reaction, or impairment requiring minor change in normal activities, and associated with some restriction of work activity |
Mild diarrhea |
353,503,000 |
4 |
142,442,000 |
4 |
84,786,000 |
4 |
9,586,000 |
4 |
B |
Moderate pain and/or moderate impairment requiring moderate change in normal activities, e.g., housebound or in bed, and associated with temporary loss of ability to work |
Moderately severe diarrhea |
31,423,000 |
5 |
1,151,000 |
5 |
685,000 |
5 |
404,000 |
5 |
C |
Severe pain, severe short-term impairment, or hospitalization |
Severe diarrhea |
6,983,000 |
6 |
230,000 |
6 |
171,000 |
6 |
101,000 |
6 |
D |
Mild chronic disability (not requiring hospitalization, institutionalization, or other major limitation of normal activity, and resulting in minor limitation of ability to work) |
|
|
n.a. |
|
n.a. |
|
n.a. |
|
n.a. |
E |
Moderate to severe chronic disability (requiring hospitalization, special care, or other major limitation of normal activity, and seriously restricting ability to work) |
|
|
n.a. |
|
n.a. |
|
n.a. |
|
n.a. |
F |
Total impairment |
|
|
n.a. |
|
n.a. |
|
n.a. |
|
n.a. |
G |
Reproductive impairment resulting in infertility |
|
|
n.a. |
|
n.a. |
|
n.a. |
|
n.a. |
H |
Death |
|
349,000 |
n.a. |
23,000 |
n.a. |
13,000 |
n.a. |
2,500 |
n.a. |
deaths. The derivation of both sets of estimates are discussed in Appendix C.
PROBABLE VACCINE TARGET POPULATION
In developing countries, the incidence of disease associated with enterotoxigenic E. coli appears to be highest in children during the first 2 years of life, when one or even two episodes per year per child have been noted. The incidence remains high for the first 5 years of life and moderately high for children 5 to 9 years old (Black, personal communication, 1984). Although older children and adults also suffer from E. coli diarrhea, partial immunity does appear to develop after childhood. Thus, the probable vaccine target population would be children within the first 6 months of life. The vaccine could be incorporated into the World Health Organization Expanded Program on Immunization (WHO-EPI) with delivery at the earliest current age of vaccine administration.
Travelers to developing countries, primarily adults, constitute a second potential vaccine target group.
Vaccine Preventable Illness*
Two considerations influence the estimation of vaccine preventable illness for vaccines against enterotoxigenic E. coli. First, knowledge is incomplete regarding which components should be included in a vaccine to cover all, or a high proportion of, possible natural challenges. Known adhesion determinants (which are relatively few in number) are encountered in about 65 to 75 percent of LT+/ST+ strains (which cause most of the severe disease). The percentage is lower (up to 25 percent) in strains that produce only LT or ST, but these strains probably cause most of the cases (Levine et al., 1983). Thus, it is difficult to estimate the coverage of strains that a vaccine could currently achieve. Work is currently under way to identify the adhesion determinants or other colonization factors present on strains lacking CFA I and CFA II or other known determinants. Thus, in a few years it may be possible to better identify the determinants that should be included in the vaccine. For the calculations, it is assumed that by the time of licensure a vaccine could probably be formulated with a reasonable number of purified components (CFA I and CFA II, adhesion factors, and possibly other components, e.g., toxoids) that would cover about 70 percent of strains. (It is recognized that this assumption is more optimistic than some researchers believe.)
* |
Vaccine preventable illness is defined as that portion of the disease burden that could be prevented by immunization of the entire target population (at the anticipated age of administration) with a hypothetical vaccine that is 100 percent effective (see Chapter 7). |
Second, a small amount of disease occurs at an early age, before completed immunization schedules can confer full protection. Taking into account these two factors, the proportion of illness that is vaccine preventable with a subunit vaccine is estimated to be about 60 percent.
In the case of a genetically attenuated strain, engineering a strain to carry a sufficient number of antigens to cover all strains might be difficult, so a somewhat lower proportion of the total disease burden is assumed to be potentially vaccine preventable (i.e., about 50 percent).
Efficacy (as predicted in Chapter 5, Table 5.1) in both cases is estimated against challenge by strains in the vaccine. In both cases the vaccine is assured to confer long-lasting immunity.
SUITABILITY FOR VACCINE CONTROL
Diarrhea due to enterotoxigenic E. coli would be suitable for vaccine control if relatively long-lasting protection could be elicited at an early age to carry children throughout the early childhood period when they are at greatest risk of dehydration and death. The prevalence of this and other diarrheal diseases in areas of developing countries where medical care services are rarely available suggests that a prevention approach is desirable.
Alternative Control Measures and Treatments
Death from acute diarrhea most often results from dehydration caused by unreplaced losses of body water and electrolytes. The treatment of diarrhea due to enterotoxigenic E. coli depends primarily on replacement of these deficits. The increased use of oral rehydration therapy (ORT) could have a considerable impact on the disease burden, especially deaths. However, the application of ORT depends on the availability of oral rehydration salt preparations and community education programs explaining how to use them. A scenario assuming increased ORT application is included in the calculations described in Chapter 7, based on the estimates in Table D-2.2 and discussed in Appendix C.
Adjunctive therapy with antibiotics also may reduce the duration and volume of diarrhea. Co-trimoxazole is the only drug that has been evaluated, and it has been studied mainly in adults with experimentally induced diarrhea or with travelers’ diarrhea.
Transmission of enterotoxigenic E. coli is thought to be primarily by water and food. Presumably, the disease could be prevented by avoidance of fecally contaminated water and attention to hygienic food handling techniques. The provision of clean water and improved sanitation in developing countries is desirable on many grounds but is unlikely to be a rapid solution to diarrhea prevention.
In addition, for travelers, prophylactic antibiotics have been utilized and appear to prevent enterotoxigenic E. coli diarrhea. However, because of the risks involved in taking antibiotics, chemo-
prophylaxis is currently recommended only for persons who cannot obtain safe food and water and who would be endangered or greatly inconvenienced if they were to get diarrhea while traveling. This particularly includes persons with serious underlying medical conditions, in whom diarrhea could present a difficult management problem.
PROSPECTS FOR VACCINE DEVELOPMENT
Current work in vaccine development against enterotoxigenic E. coli. diarrhea involves vaccines that stimulate antitoxic (antitoxin) or anti-adhesion immunity or both by means of killed antigens or attenuated strains. Recent developments have been reviewed by Levine et al. (1983). The most effective vaccines may contain antigens that stimulate both antitoxic and antibacterial immunity, producing a synergistic protective effect. It is believed that the critical site of immunity is the mucosal surface of the upper intestinal tract and that this site is protected mainly by secretory IgA antibody.
Enterotoxigenic E. coli also cause serious diarrhea and death in animals. Extensive veterinary research has focused on the development of vaccines against the organisms that produce these problems. In the veterinary studies, purified fimbrial vaccines protected newborn piglets and calves, which were suckled on immunized mothers, against death from diarrhea caused by challenge with enterotoxigenic E. coli bearing the homologous fimbriae. In addition, CFA I and CFA II fimbrial vaccines administered orally or enterally stimulated intestinal SIgA antibody to CFA and resulted in protective immunity in animal models. Studies in humans with purified CFA vaccines are beginning. If a prototype CFA vaccine is found to protect against E. coli with the homologous fimbriae, intensive research will be pursued to identify other colonization factors. Only a small number of pathogenic enterotoxigenic E. coli possess currently recognized colonization factors; other factors must be identified to ensure broad-spectrum protection by a polyvalent fimbrial vaccine.
Another approach to the development of a vaccine against enterotoxigenic E. coli involves the use of toxoids. Animal studies have shown that immunization with either B subunit or LT holotoxin elicits an immune response and a protective effect, and that holotoxin is the superior immunogen. The technology for large-scale production of B subunit has not yet been described, but the successful cloning of LT genes from a human pathogen into a high-copy plasmid vector indicates that it is possible.
Heat treatment of cholera enterotoxin results in a high molecular weight toxoid called procholeragenoid, which is comparable in immunogenicity to the parent toxin. This has been used to immunize pregnant sows. Piglets born to and suckled on the immunized sows were protected against diarrhea and death due to infection with enterotoxigenic E. coli, suggesting that procholeragenoid might serve as an oral vaccine to enhance protection against enterotoxigenic E. coli, as well as Vibrio cholerae.
Because a large portion of the enterotoxigenic E. coli that cause disease produce only ST, attempts have been made to immunize against this toxin, despite its poor immunogenicity. Encouraging results were obtained by conjugating ST to porcine IgG. Subsequently, a bivalent toxoid was prepared by cross-linking ST to LT. Testing of a bivalent toxoid consisting of a laboratory-synthesized ST conjugated to the B subunit also has begun. Both of these bivalent toxoids have demonstrated immunogenicity and protective effects in animal models.
Another approach toward prevention of enterotoxigenic E. coli diarrhea involves the use of attenuated strains of E. coli bearing critical antigens. These strains should be capable of colonizing the small intestine and stimulating an immune response, without causing adverse reactions. One live strain, a CFA II-positive, LT- and ST-negative variant of a previously enterotoxigenic strain, has been evaluated in humans. When given to volunteers, all excreted the strain, most had positive cultures of jejunal fluid, and most had serological responses. About 10 percent of the volunteers developed mild diarrhea, however, presumably as a consequence of colonization of the proximal small intestine.
With the advent of recombinant DNA technology, it is possible to construct an E. coli vaccine strain engineered to produce large quantities of multiple colonization factor antigens, B subunit, and perhaps an ST toxoid. Further work will be necessary to understand, and if possible eliminate, the mild diarrhea resulting from CFA-positive strains of E. coli.
A potential problem with all of the proposed vaccines is that there is no assurance they will provide long-term (up to 5 years) protection. This is an important goal, especially for vaccine use in developing countries.
REFERENCES
Black, R.E. 1984. Personal communication, Johns Hopkins University School of Hygiene and Public Health, Baltimore, Md.
Dupont, H.L. 1982. Escherichia coli diarrhea. Pp. 219–234 in Bacterial Infections of Human, A.S.Evans and H.A.Feldman, eds. New York: Plenum.
Levine, M.L., J.B.Kaper, R.E.Black, and M.L.Clements. 1983. New knowledge on pathogenesis of bacterial enteric infections as applied to vaccine development. Microbiol. Rev. 47:510–550.