- Email: [email protected]
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, Vol. 13, No. 5, pp. 4g5-W2, 1995 Copyright 0 1995 Elsevier Science Ltd Printed in Great Britain. All rights reserved m4-41oxI95 $10.00 + 0.00
0264-410X(94)00011-5
Evaluation of the safety, immunogenicity, and efficacy in healthy adults of four doses of live oral hybrid Escherichia coli-ShigellQ flexneri 2a vaccine strrainEcSf2a-2 Karen L. Kotloff*+‘“, Genevieve A. Losonsky*fs, James P. Nataro*+$, Steven S. Wasserman.*‘, Thomas L. Hale-, David N. Taylor-, John W. Newland-, Jerald C. Sadoff-, Samuel B. Formal-’ and Myron M. Levine*+’ In previous trials, live invasive Escherichia coli-Shigella flexneri 2a hybrid vaccine candidate EcS’a-2, administered to adult volunteers as 3 doses of ca. 2~10~ colony forming units (c.fu.) spaced over one week, induced fever andlor diarrhea in 11% of subjects and provided only limited protection (36% ef$cacy) against illness following challenge with virulent ;S. flexneri 2a. We sought to improve the clinical safety of this vaccine by administering a lower inoculum, and to enhance protective immunity by administering additional booster doses at 2 weeks. Twenty-one healthy adults were immunized with ca. 7x.10” c.Ju. of EcS’a-2 on days 0, 3, 14, and 17. The vaccine consistently colonized the intestine without causing serious adverse reactions; mild diarrhea developed in one subject and low grade fever in another. Vaccination elicited an antibody secreting cell (ASC) response to lipopolysaccharide (LPS) in all subjects, which was highest on da,y 7 and notably diminished thereafter on days 10, 16, 21, and 24, suggesting that active mucosal immunity developed rapidly. The magnitude of the response was modest (geometric mean peak = 16 IgA ASCIIO” peripheral blood mononuclear cells) and an IgG serological response to LPS was detected in only 19% of subjects. Following experimental challenge with virulent S. flexneri 2a administered with bicarbonate buffer, shigellosis (diarrhea, dysentery, or fever) developed in 10 of 16 vaccine recipients (63O/t) and in 12 of 14 unvaccinated controls (86OA), resulting in a vaccine efficacy of 27% (95% confidence limits -197, 82, p = 0.15, l-tailed). We conclude that four doses of EcSfla-2 spaced over 17 days at inocula that are welltolerated do not confe,r significant protection against illness following experimental challenge of healthy adult volunteers. Keywords: Phase I study; mucocsal immunity: Shigellu flexneri
Although streptomycin-dependent and T,,-ISTRATI attenuated Shigella flexneri 2a strains were used with some success as live oral vaccines in the 1960~‘-~, *Center for Vaccine Development, University of Maryland School of Medicine, Baltimore, MD 21201, USA. +Division of Infectious Diseases and Tropical Pediatrics, Department of Pediatrics, University of IMaryland School of Medicine, Baltimore, MD 21201, USA. I-Division of Geographic Medicine, Department of Medicine, Center for Vaccine Development, University of Maryland School of Medicine, Baltimore, MD 21201, USA. “Division of Communicable Diseases and Immunology, Walter Reed Army Institute of Research, Walter Reed Army Medical Center, Washington, DC. 20307-5100, USA. “University of Maryland School of Medicine, Center for Vaccine Development, 10 S. Pine Street, Baltimore, MD 21201, USA. “To whom correspondence should be addressed. (Received 11 July 1994; revised 26 September 1994; accepted 26 September 1994)
subsequent efforts to develop improved Shigella vaccines have failed to yield a promising candidate. More recently, vaccines have been constructed which invade and replicate within the intestinal mucosa, enabling delivery of Shigella antigens directly to the intestinal lymphoid tissue in an attempt to provide strong and durable immunologic stimulation, both local and systemic4. 5. We have previously reported the results of the safety, immunogenicity, and efficacy of an epithelial cell invasive live oral Escherichia coli-S. flexneri 2a hybrid vaccine candidate, EcSf2a-26. In construction of this vaccine, chromosomal genes encoding the group and type-specific antigens of S. jlexneri 2a and an S. JEexneri 5 invasiveness plasmid were introduced into E. coli K12. Thereafter a mutation in the aroD gene was selected, rendering the strain dependent on para-aminobenzoic acid for growth, a substrate not available in human
Vaccine
1995 Volume
13 Number
5
495
E. coli-S. flexneri 2a hybrid vaccine: K.L. Kotloffet
al.
tissues’. In previous trials, 3 doses of ca. 2~10~ c.f.u. of EcSf2a-2 spaced over one week were immunogenic, but only modestly protective (36% efficacy) against challenge with virulent S. Jlexneri 2a. Furthermore, several subjects experienced fever and/or diarrhea after vaccination. In this report, we describe the results of further studies conducted to assess an alternate dosing regimen designed to improve clinical safety by administering a lower inoculum and to maximize protective immunity by using a 4-dose regimen spaced over 17 days. METHODS Subjects
Adults of ages 1840 years were recruited from the Baltimore-Washington area. Eligible subjects were without significant abnormalities detected by medical history, physical examination, and a panel of screening clinical laboratory tests. Occupational food handlers, travelers to Shigellu-endemic areas, persons with known allergy to ciprofldxacin or sulfa drugs, and those with HLA B-27 haplotype (because of its association with post-infectious arthritis’) were excluded. Preparation of the vaccine inoculum
Details of vaccine construction and formulation are presented elsewhere6. 7. The lyophilized vaccine was stored at - 70°C in sealed vials containing 4x10” viable organisms. Vaccine stability was confirmed one week before vaccination using colony counts and slide agglutination with Shigeflu group B antisera (Difco Laboratories, Detroit, MI), and by demonstrating retention of HeLa cell invasiveness’ and Congo Red positivity in > 95% of colonies. The inoculum was prepared by rehydrating three ampules of vaccine, each with 5 ml of sterile water. The vials were maintained on wet ice for 30 min, then pooled and serially diluted with phosphate buffered saline (PBS) to achieve a 1:4 dilution. Colony counts of reconstituted vaccine were determined before and after vaccination and the results were averaged to estimate the actual inoculum administered. Preparation of the challenge strain
The challenge inoculum was prepared from a frozen master seed stock of S, Jexneri 2a strain 2457T” which was plated onto trypticase soy agar (TSA, Becton Dickinson Microbiology Systems, Cockeysville, MD) into which Congo Red dye (O.Ol%, Sigma Chemical Company, St. Louis, MO) was added. After incubation at 35°C for 18-24 h, several individual Congo Red colonies that characteristic exhibited Shigella morphology were confirmed as group B Shigella using slide agglutination. Several well-isolated Congo Red colonies were picked and suspended in sterile saline. The saline suspension was then used to inoculate TSA plates. Overnight growth at 35°C from the heavily seeded TSA plates was harvested into sterile PBS. The bacterial suspension was diluted with additional sterile PBS to produce a suspension with an optical density at 660 nm corresponding to the desired bacterial count per ml.
496
Vaccine 1995 Volume 13 Number 5
Clinical design of vaccine study
Twenty-one outpatient volunteers who fasted 90 min before and after inoculation received an oral dose of vaccine on days 0, 3, 14, and 17. To buffer gastric acid, subjects drank 120 ml of a solution containing 2 g of NaHCO, in 150 ml distilled water. One minute later, they ingested vaccine suspended in the remaining 30 ml of buffer. For a period of 24 days after the first dose of vaccine, the volunteers were evaluated once daily using a standardized interview that was conducted either in person or by telephone, to elicit the occurrence of adverse clinical reactions during the preceding 24 h, including diarrhea, vomiting, fever, hematechezia, cramps, anorexia, and malaise. They recorded the number and consistency of all stools passed, and their evening oral temperature. Signs and symptoms occurring within 72 h of each inoculation are reported herein. Clinical design of challenge study
One month after the fourth dose of vaccine (day 45 following the first dose), 16 vaccine recipients and 14 unvaccinated control subjects were admitted to the Research Isolation Ward of the Center for Vaccine Development. In the fasting state, they were challenged with 1~10~ c.f.u. virulent S. JEexneri 2a along with bicarbonate buffer, as described for the vaccine. For five days following challenge, subjects were evaluated for evidence of illness by physicians and nurses who were blinded to the volunteers’ status as vaccinees or controls. Temperature, pulse and blood pressure were measured every 6 h. All stools passed were examined for blood, weighed and graded on a l-5 scale for consistency”. Subjects received a 5-day course of ciprofloxacin (500 mg by mouth every 12 h), beginning five days after challenge, or earlier if deemed clinically necessary by the investigator. Definitions
Illness was defined as diarrhea, fever (oral temperature 2 lOO.O”F) or dysentery (gross blood in a loose stool). For the outpatient vaccination study, diarrhea was defined as three or more liquid stools in 24 h. In the challenge study, diarrhea was defined as two or more loose stools (grade 3-5) totalling 200 ml within 48 h or a single 300 ml grade 3-5 stool. Protective efficacy was determined in the challenge studies for diarrhea, dysentery, fever, and any illness as follows: [(attack rate of illness in controls - attack rate of illness in vaccinees)/attack rate of illness in controls]xlOO. Bacteriology
A perirectal swab following defecation was collected from each volunteer on days O-5, 7, 10, 14-19, 21, 24, and 28 following the first dose of vaccine to assess duration of vaccine shedding. The swab was immediately placed in Cary Blair transport medium (Becton Dickinson) and refrigerated for up to 24 h before transport to the laboratory. For participants in the inpatient challenge study, a swab of all stools passed during hospitalization and weekly for 3 weeks following discharge was obtained, inoculated into buffered glycerol saline, and maintained at 46C until arrival at the laboratory. A rectal swab was obtained from any
E. coli-S. flexneri 2% hybrid vaccine: K.L. Kotloff et al. compare vaccinees and controls because the question of interest was whether the vaccine reduced the rate of illness. Severity of signs and symptoms was compared by one-tailed Wilcoxon tests. To compare ASC responses, data were transformed to natural logs and analyzed by z-tests; geometric means for data including zeros were calculated by adding one to each datum prior The effects of vaccination to log-transformation’5. status and illness post-challenge upon post-challenge ASC and serologic responses were examined by analysis of variance (independent factors were vaccination status, illness and their interaction; log-transformed peak ASC or reciprocal immunoglobulin titer served as the dependent variable); each assay was analyzed separately. The correlation between peak geometric mean ASC response (log transformation) and protective efficacy (angular transformation) was examined by Pearson’s correlation coefficient. Resulting probabilities c 5% were considered statistically significant.
volunteer who failed to produce a stool in a required 24 h period and inoculated into Gram-negative (GN, Becton Dickinson) broth. Swabs and stools were inoculated directly onto MacConkey’s agar (Becton Dickinson) containing 500 pg ml-’ streptomycin for recovery of the streptomycinresistant vaccine strain; Salmonella-Shigella (Becton Dickinson) and MacCortkey’s agar without antibiotics were used for recovery of the Shigella challenge strain. Each specimen was also inoculated into GN enriehment broth and incubated overnight at 35°C before being plated onto enteric media. Lactose-negative colonies identified after 24 or 48 h incubation at 35°C on solid media were inoculated onto Triple Sugar Iron (TSI, Becton Dickinson) slants. Those producing an alkaline slant and an acid butt without gas were confirmed by slide agglutination. Measurement
of serum antibodies
To measure the serum antibody response to vaccination, serum specimens were collected on days 0, 7, 21, 28, 35, and 45 following the first dose of vaccine; to measure the response to challenge, serum was collected on days 0, 10, 21 and 28 following challenge with pathogenic Shigella (Table I). Serum antibodies (IgA, IgM, and IgG) to the invasiveness plasmid antigens (IPA) and LPS 0 antigens of S. JEexneri 2a were measured by an enzyme-linked immunosorbent assay (ELISA), as previously described”. 13. Four-fold rises in titer were considered significant. Detection
RESULTS Clinical response to vaccination
After each dose, between 3 and 5 of the 21 vaccinees experienced at least one constitutional symptom or sign of illness, which in every instance was judged to be mild by the volunteer. One subject reported diarrhea after the first dose (4 loose stools in 24 h) and again after the third dose (7 loose stools in 48 h). Another subject had low grade fever (I 100.2”F) after the first and third doses of vaccine.
of antibody secreting cells
To measure the antibody secreting cell (ASC) response to vaccination, peripheral blood mononuclear cells (PBMC) were collected on days 0, 7, 10, 16,21, and 24 days after the first dose of vaccine; the ASC response to challenge was measured on days 0 and 7 in relation to challenge (Table I). Circulating IgA, IgM, IgG ASC against the IPA and LPSl antigens were determined by count exceeding 3 ELISPOT14. A post-vaccination standard deviations above the geometric mean prevaccination or pre-challenge ASC was considered a positive ELISPOT response.
Vaccine excretion
Statistical
Immune response to vaccination
All subjects excreted the vaccine strain on at least one day. Excretion was detected 24 h after vaccination in 18 of 21 subjects (86%) after dose 1, in 14 of 21 subjects (67%) after dose 2, in 11 of 20 subjects (55%) after dose 3, and in 14 of 21 subjects (67%) after dose 4. Similarly, the duration of shedding (14 days) did not decline significantly with consecutive dosing. The HeLa cell invasive phenotype was found in 70-95% of vaccine strains randomly selected from volunteers’ stools.
analysis
Circulating IgA ASC specific for LPS was detected in all volunteers during the 24 days after the first dose of vaccine, and 90% had an IgG ASC response. However, the magnitude of the anti-LPS ASC response was significantly lower than the response elicited by challenge of unvaccinated controls (Table 2). The peak
Unless otherwise specified, two-tailed Fisher’s exact and x2 tests were used to compare proportions. A sample size for the challenge study was chosen that would allow detection at least 50% protective efficacy if the attack rate of illness in controls were 85% (p = 0.05, one-tailed); one-tailed Fisher’s exact tests were used to
Table 1
Schedule of inoculations and collection of specimens for measurement of immune response Day specimen collected following vaccination
Vaccination Serum antibody Antibody secreting cells Challenge Serum antibody Antibody secreting cells “Day 45 of vaccination
0
3
X
X
7
10
14
16
X
X
X
24
21
X X
26
35
45
X
X
XS
X
X
X
17
or challenge
X
X
X
Xe X
and day 0 of challenge
X
X
X
X X
are the same day
Vaccine
1995
Volume
13 Number
5
497
E. coli-S.
flexneri
2a hybrid
vaccine:
K.L. KotM
et al.
Table 2 immune responses to lipopolysaooharide (LPS) or invasiveness plasmid antigens (IPA) elicited by vaccination with 4 doses of EcSf2a-2 or by challenge of unvaccinated controls with virulent S. flexneti 2a Response to vaccination Geometric mean peak
with EcSf2a-2
No. (%)
Response to challenge
responders
Geometric mean peak
with S.
flexneri 2a
No. (%) responders
Antibody secreting cells (ASC)” Anti-LPS lgA lgG Anti-IPA
16” 5”
21/21 (100) 17/19 (90)
143” 24”
14/14 (100) 13/14 (93)
lgA jgG Serum antibodyb Anti-LPS
16 5
21/21 (100) 17/20 (85)
54 6
13114 (93) 11/13 (85)
lgA lgG Anti-IPA
121 53
6/21 (29)” 4/21 (19)
173 78
11114 (79)d 7114 (50)
lgA lgG
71” 2.51”
l/21 (5) 3/21 (14)O
220” 1073”
4/l 4 (29) 8/14 (57)“
“ASC response to vaccination was defined for each assay as a post-vaccination ASC count > 3 standard deviations above the geometric mean pre-vaccination ASC count. Accordingly, the following response levels were established: 2 2 ASC per lo6 peripheral blood mononuclear cells (PBMC) for anti-LPS IgG and IgA, t 1 ASC per lo6 PBMC for anti-IPA IgG and IgA “Serum antibody response defined as a fourfold rise in reciprocal titer following inoculation. Means are expressed as geometric mean of reciprocal titer %omparison of geometric mean peak response to a specific assay between vaccinees and challenge subjects, ~~0.05 dComparison of the proportion of responders in a specific assay between vaccinees and challenge subjects, ~~0.05
anti-LPS IgA ASC response to vaccination was observed 7 days after the first dose, decreasing markedly thereafter (Figure 2). Anti-IPA ASC responses were also seen in most volunteers (Table 2). Vaccination elicited a four-fold rise in serum IgG anti-LPS antibody in 4 subjects (19%), while 6 (29%) had a significant rise in IgA anti-LPS antibody (Table 2). Seroconversions of IgG and IgA anti-IPA were detected in 3(14%) and 1(5%) of subjects, respectively. IgA anti-LPS and IgG anti-IPA seroconversion was significantly less frequent following vaccination compared with challenge.
. \
.
\ 12
Clinicai evaluation of vaccine efficacy following wildtype challenge with S. jlexneui 2a
The overall attack rate for illness was 86% in controls and 63% in vaccinees (protective efficacy 27%, 95% confidence limits - 197, 82, p = 0.15, Table 3). There was a trend toward lower diarrhea1 volumes in vaccinees; however, there were no statistically significant differences between vaccinees and controls in
I8 t
24
t
Days after initial vaccination Figure 1 Anti-lipopolyaaccharide IgA antibody secreting cell (ASC) response expressed as number of cells per lo6 peripheral blood mononuclear cells (PBMC) to oral vaccination with live attenuated Shige//a flexnen’ 2a vaccine EcSf2a-2 in 4 spaced doses on days 0, 3, 14, and 17 (vaccination days indicated by arrows)
Table 3 Clinical and microbiological responses of vaccinated (four spaced doses of EcSf2a-2) and unvaccinated control volunteers following challenge with wild-type Shigela flexnen’2a Attack rate for: Group
Attack rate for any illness
Fever
Diarrhea
Dysentery
No. (%) with fecal shedding
Controls
12114 (86%) lot16 (63%) 27%
10114 (71%) 9116 (56%) 21%
9114 (64%) 8/l 6 (50%) 22%
9114 (64%) 9116 (56%) 13%
13114 (93%) 15116 (94%) 0
- 197,82 0.15
-169,77 0.31
- 155,76 0.34
-188,73 0.47
1.0
Vaccinees Efficacy” 95% confidence limits p value*
“Protective efficacy was determined controls] x 100 OFisher’s exact test, single tail
498
Vaccine
1995
Volume
as follows: [(attack rate of illness in controls
13 Number
5
-
attack rate of illness in vaccinees)/attack
rate of illness in
E. coli-S. flexneri 2a hy&id Table 4 Comparison of height of fever and total stool volume in ill EcSf2a-2 vaccine recipients and ill controls following challenge with wild-type Shigella flexneri 2a
Group
Mean peak fever
Controls Vaccinees” p value*
102.8”F 102.2”F 0.13
Diarrhea stool volume (Time from onset of first loose stool) (24 h)
(30 h)
999 ml 692 ml 0.17
1149 ml 930 ml 0.28
“One vaccinee received early treatment with ciprofloxacin 26 h after the onset of loose stools Wilcoxon test, single tail
beginning
the frequency or severity of these symptoms (Tables 3 and 4). Relationship between immune response to vaccination and protection against illness after challenge
Examination of individual ASC levels after vaccination revealed that the volunteer with the highest anti-LPS IgA ASC counts (333/106 PBMC) and antiIPA IgA ASC counts (1.30/106 PBMC) was protected against illness after challenge, whereas the next highest IgA responder (157 ASC/I O6PBMC against LPS and 51 ASC/106 PBMC against IPA) was not protected. A significant association between peak IgA anti-LPS ASC response to vaccination and resistance to illness following challenge could not be demonstrated statistically @ = 0.9). One of the 10 vaccinees (10%) who became ill after challenge had a fourfold rise in serum IgG anti-LPS after vaccination compared with 2 of 6 vaccinees (33%) who resisted illness @ = 0.52). Five of the unprotected (50%) and 1 of the protected volunteers (20%) had a fourfold rise in serum IgA anti-LPS after vaccination (p = 0.31). Immune response to challenge in vaccinated and unvaccinated subjects
Differences could not be detected between vaccinees and controls in the magnitude and frequency of ASC or serologic responses after challenge. However, subjects who became ill after challenge had significantly greater geometric mean peak ASC responses following challenge in each assay examined, as well as greater geometric mean peak reciprocal anti-LPS IgA and IgG serologic titers, compared to those who remained well (data not shown).
DISCUSSION Several lines of evidence suggest that shigellosis can be prevented with a live oral vaccine that directly stimulates the intestinal immune system. For one, intestinal immunity appears to be an important factor in controlling Shigella infection, as illustrated by the efficacy of passively transferred oral immunoglobulin in preventing shigellosis’6. ” and the possible role of mucosal antibody in limiting the duration of illness’8; in contrast, parenteral immunization with killed Shigella organisms failed to protect against illness despite high levels of vaccine-induced systemic antibody”, *“.
vaccine: /CL. Kofloffet
al.
Second, an initial episode of shigellosis confers serotypespecific immunity, as has been observed in children living in endemic areas2’.22 and in experimental infection of humans with S. sonnei23 and S. Jlexneri 2a24. Finally, noninvasive streptomycin-dependent Shigella strains utilized as live oral vaccines induce high level, but short-lived (< 1 year) protection against shigellosis as demonstrated in field trials conducted in Yugoslavia’. 25. The requirement for four large (> 10” c.f.u.) doses in the primary immunization schedule followed by annual boosters thereafter was considered a limitation of these vaccine?. It is of interest that oral administration of a noninvasive E. coli strain expressing S. JEexneri 2a 0 antigen was poorly immunogenic and nonprotective in experimental challenge studies in vo1unteers2’. To overcome the perceived drawbacks of the noninvasive oral vaccines, EcSf2a-2 was constructed with the following features: (1) epithelial cell invasiveness to enhance stimulation of the mucosal immune system; (2) 0 antigen expression to induce type-specific antibodies which have been correlated with resistance to shigellosis among Israeli soldiers serving in a field unitZ8; (3) limitation of intracellular multiplication due to a nutritional dependence on a substrate, paraaminobenzoic acid, not available in human tissues; (4) presence, in the E. coli carrier, of the ompT chromosomal genes degrading Shigella proteins necessary for intercellular spread29; and (5) absence of aerobactin and perhaps other critical Shigella virulence factors in the E. coli Carrie?‘. 3’. Initial volunteer studies suggested that immunization with ca. 2~10~ c.f.u. of EcSf2a-2 in 3 doses spaced over a week was reactogenic and could not induce significant protection against experimental challenge6. Accordingly, we sought to evaluate alternate dosing regimens. Findings from other studies suggested that a four dose schedule of EcSf2a-2 might enhance protective immunity. In the guinea pig keratoconjunctivitis model, vaccine efficacy was three times higher using a four dose regimen with vaccination on days 0, 2, 14 and 15 than with a three dose regimen with vaccination on days 0, 2, and 4’2. In this model, animals were immunized in the conjunctival sac with ea. 3.8~10~ c.f.u. of EcSf2a-2 and then challenged 14-28 days after the last dose of vaccine with a homologous wild-type Shigella strain (2457T). Furthermore, experience in humans with other oral attenuated bacterial vaccines, such as the typhoid vaccine Ty2 1a33 and the streptomycin-dependent attenuated Shigella vaccines”. 34 suggested that the level of protection correlates with the number of doses given. It was therefore hypothesized that at a slightly lower inoculum (7x108 c.f.u.) to ensure clinical safety, a four dose regimen of EcSf2a-2 would enhance vaccine efficacy. In general, the vaccine was well tolerated at this lower inoculum level. Several volunteers reported mild diarrhea, low grade fever, or mild constitutional complaints, but in the absence of a placebo control group, these symptoms could not be attributed to the vaccine in any individual case. In an expanded Phase II study conducted among active duty US military personnel, adverse reactions did not occur significantly more frequently among 205 recipients of four doses of EcSfZa-2 ranging from 2.3 to 9.0~10~ c.f.u. compared to 39 placebo recipients35.
Vaccine 1995 Volume 13 Number 5
499
E. coli-S. flexneri 2a hybrid vaccine: K.L. Kofloffet
al.
Patterns of vaccine excretion and ASC response to consecutive doses of vaccine can be examined to evaluate the ability of EcSf2a-2 to stimulate mucosal immunity. The vaccine consistently colonized the intestine. The highest colonization rate (87%) occurred after the first dose; however, significantly diminished colonization with consecutive doses, which might indicate vaccine-induced mucosal immunity”- 36, was not seen. Kantele has suggested that the ASC response following primary immunization is a method for assessing protective immunity conferred by live oral vaccination3’. This investigator demonstrated that three doses of Salmonella typhi Ty2la vaccine formulated as an enteric-coated capsule elicited a significantly lower geometric mean peak ASC response among volunteers compared with three doses of the vaccine formulated as a suspension. These patterns of ASC response agree with observations obtained in field trial that the liquid suspension provides the best efficacy3’. With our small sample sizes, we could not demonstrate a significant correlation between the ASC response to four doses of EcSf2a-2 Shigella vaccine and protective efficacy following challenge. However, it is notable that the primary response to vaccination was modest (geometric mean peak IgA LPS response 16 ASC/106 PBMC), particularly when compared to the response that accompanied the protective immunity elicited by experimental inoculation with virulent infection (143 ASC/lO” PBMC), and may have been insufficient to provide significant protection. Indeed, a pattern emerging from our recent trials suggests that the peak geometric mean anti-LPS IgA ASC response induced by vaccination or wild-type infection correlates with the ability of the immunizing strain to protect volunteers against shigellosis following experimental inoculation (Table 5 and K. Kotloff, submitted for publication). The challenge that remains for future Shigellu vaccine development is to achieve vigorous immunologic stimulation without causing reactogenicity.
We conclude that four doses of EcSI2a-2 in this trial did not enhance the protective efficacy of the vaccine elicited by three doses. However, a caveat in comparing the results of challenge studies involving recipients of three vs four doses of EcSf2a-2 must be considered. This study provided the first experience administering virulent Shigella in suspension with bicarbonate buffer instead of skimmed milk. This modification in the vehicle for inoculation of the challenge strain was intended to improve the study-to-study consistency of illness rates among unvaccinated controls. The ability of the modified model to demonstrate protective efficacy has been validated using homologous rechallenge studies of volunteers (K. Kotloff, submitted for publication). None the less, while the results of challenge studies using skimmed milk as a vehicle24 have been corroborated by field experience25, the level of protective efficacy of EcSf2a-2 as demonstrated in volunteers using the modified challenge model awaits comparison with the results of a vaccine field trial currently being conducted in Israel.
ACKNOWLEDGEMENTS This study was supported by contract No. DAMD-1 7-92-C-2058 from the U.S. Army Medical Materiel Development Command. We thank the volunteers who participated in these studies, Kathleen Palmer and Brenda Berger for assistance in patient recruitment and outpatient follow-up, Cathy Black, Ron Grochowski, and the nursing staff of the Center for Vaccine Development for providing clinical care, Linda Guers, Mardi Reymann, and Karen Taylor for technical assistance, and Drs. Carol Tacket and Robert Edelman for clinical support and valuable suggestions. Statement: informed consent was obtained from the volunteers and guidelines for human experimentation of the U.S. Department of Health and Human Services, the Human Volunteers Research Committee of the University of Maryland, and the Surgeon General’s Human Subjects Research Review Board of the
Table 5 Relationship between peak IgA antibody secreting cell (ASC) response to lipopolysaccharide following inoculation with wild-type Shigella flexneti 2a or with live oral EcSf2a-2 vaccine and protection against homologous wild-type challenge ASC response to stated immunizing regimen among subjects who later participated in efficacy trial Immunizing regimen
lnoculum
S. flexneri 2a’ EcSf2a-2” EcSf2a-2
1 x lo3 c.f.u. 2 x lo* c.f.u. 7 x lo8 c.f.u.
EcSf2a-2”
2 x lo0 c.f.u.
No. doses
Days of inoculation
Days ASC measured
1 3 4
0 0, 3, 6 0, 3, 14, 17
3
0, 7, 10
0, 7 0, 7, 10 0, 7, 10, 16, 21, 24 0, 7, 10
n
Peak geometric mean No. ASC per lo6 PBMC”
Eff icacy”,b
11 10 16
140 27 15
70%d 48% 27%
10
4
0%
PBMC = peripheral blood mononuclear cells; c.f.u. =colony forming units =V 0.986, p=O.Ol, 2-tailed, Pearson correlation coefficient, comparing peak geometric mean No. ASC/106 PBMC and level of protective efficacy *Protective efficacy against homologous challenge with wild-type S. flexneri 2a approximately one month after completion of immunizing regimen 70 assess the protective efficacy of wild-type infection, previously vaccinated (4 spaced doses of EcSf2a-2) and unvaccinated veterans of a S. flexneri 2a challenge study were rechallenged with wild-type S. flexneri 2a along with a group of immunologically naive control subjects (K. Kotloff, submitted for publication) dp
500
Vaccine 1995 Volume 13 Number 5
E. coli-S. flexneri 2a hybrid vaccine: K.L. Kofloffet
Department of the Army were followed in the conduct of the clinical research. Disclaimer: The views o:Fthe authors do not reflect the positions of the US Department of the Army or of the US Department of Defense.
18
19
REFERENCES 1
2
3
4
5
6
7
8 9
10
11
12
13
14
15
16
17
Meitert, T.E., Pencu, L., Ciudin, L. and Tonciu, M. Vaccine strain Shigella flexned T,-ISTR,ATI. Studies in animals and in volunteers. Antidysentery immunoprophylaxis and immunotherapy by five vaccine Vadizen (S.. flexnen’ T,,ISTRATI). Arch. Roum. P&ho/. EXD. Microbiiol. 1984. 43. 251-278 Mel, D., Gangarosa, E.J., Radovanovic, M.L., Arsic, B.L. and Litvinjenko, S. Studies on vaccination against bacillary dysentery. 6. Protection of children by oral immunization with streptomycin-dependent Shigella strains. Bull. WH.0. 1971, 45, 457-464 Mel, D.M., Arsic, B.L., Nikolic, B.D. and Radovanic, M.L. Studies on vaccination against bacillary dysentery. Bull. W.H.O. 1968, 39, 375-380. Formal, S.B., Baron, L.S.. Kopecko, D.J., Washington, O., Powell, C. and Life, C.A. Construction of a potential bivalent vaccine strain: introduction of Shigella sonnei form I antigen genes into the .galE Salmonella typhi Ty21 a typhoid vaccine strain. Infect. fmmun. 1981, 34, 746-754 Formal. S.B.. Hale T.L.. Kapfer. C.. Coaan. J.P.. Snov. P.J.. Chung,’ R. ei a/. Oral vaccination of ‘monkeys with an invasive Escherichia cofi K-l 2 hybrid expressing Shigella fiexneri 2a somatic antigen. Infect fmmun. 1984, 46, 465-469 Kotloff, K.L., Herrington, D.A., Hale, T.L., Newland, J.W., Van De Verg, L., Cogan, J.P. 1st al. Safety, immunogenicity, and efficacy in monkeys and humans of invasive Eschedchia cofi K12 hybrid vaccine candidai:es expressing Shigela flexnen 2a somatic antigen. lnfecf. lmmun. 1992, 80, 2218-2224 Newland, J.W., Hale, T.L. and Formal, S.B. Genotypic and phenotypic characterization of an aroD deletion-attenuated kschenchia colLShigel/a flexnen hybrid vaccine expressing S. ffexneti 2a somatic antiaen. Vaccine 1992. 10, 766-776 Keat, A. Reiter’s syndrome and reactive arthritis in perspective. N. Engl. J. Med. 1983, 26, 1606-l 614 Hale, T.L. and Formal, S.B. 1981. Protein synthesis in HeLa or Henle 407 cells infected with Shigela dysentenae 1, Shigela flexnen 2a, or Salmonella typhimurium W118. fnfect lmmun. 32, 137-l 44 DuPont, H.L., Levine, M.M., Hornick, R.B. and Formal, S.B. lnoculum size in shigellosis and implications for expected mode of transmission. J. infect Dis. 1989, 159, 1126-l 128 Levine, M.M., Black, R.E., Clements, M.L. Lanata, C., Sears, S., Honda, T. et a/. Evaluation in humans of attenuated Vibrio cholerae El Tor Ogawa strain Texas Star-SR as a live oral vaccine. Infeci Immun. 1984, 43, 515-522 Black, R.E., Levine, M.M., Clements, M.L., Losonsky, G., Herrington, D., Berman, S. and Formal, S. Prevention of shigellosis by a Salmoneih iyphi-ShigeMa sonnei bivalent vaccine. J. Infect. Dis. 1987, 155, 1260-1265 Oaks, E.V., Hale, T.L. arid Formal, S.B. Serum immune response to Shigella protein antigens in rhesus monkeys and humans infected with Shigelfa spp. infect. Immun. 1986, 53, 57-83 Van de Verg, L., Herrington, D.A., Murphy, J.R., Wasserman, S.S., Formal, S.B. and Levine, M.M. Specific immunoglobulin Asecreting cells in peripheral blood of humans following oral immunization with a bivalerlt Salmonella fyphi-Shigeffa sonnei vaccine or infection by pathogenic S. sonnei. Infect. Immun. 1990, 58, 2002-2004 Sokal, R.R. and Rohlf, F.J. Biometry The Principles and Practice of Statistics in Biological Research. W.H. Freeman and Co., San Francisco, 1981 Tacket, C.O., Binion, S.B., Bostwick, E., Losonsky, G., Roy, M.J. and Edelman, R. Efficacy of bovine milk immunoglobulin concentrate in preventing illness after Shigella flexnerichallenge. Am. J. Trap. Med. Hyg. 1992, 47, 276-283 Hayani, KC., Guerrero, M.L., Morrow, A.L., Gomez, H.F., Winsor. D.K.. Ruiz-Palaclos. G.M. and Clearv, T.G. Concentration of milk secr~atory immunoglobulin A- .against Shige//a virulence plasmid-associated antigens as a predictor of
20
21
22
23
24
25
26
27
28
29
30 31
32
33
34
35
al.
symptom status in Shigela-infected breast fed infants. J. Pediatr. 1992, 121, 852-856 Oberhelman, R.A., Kopecko, D.J., Salazar-Lindo, E., Gotuzzo, E., Buysse, J.M., Venkatesan, M.M. ef al. Prospective study of systemic and mucosal immune responses in dysenteric patients to specific ShigeMa invasion plasmid antigens and lipopoysaccharides. Infect. fmmun. 1991, 59, 2341-2350 Shaughnessy, H.J., Olsson, R.C., Bass, K., Friewer, F. and Levinson, S.O. Experimental human bacillaty dysentery. Polyvalent dysentery vaccine in its prevention. J.A.M.A. 1946, 132, 362-368 Higgins, A.R., Floyd, T.M. and Kader, M.A. Studies in shigellosis. Ill. A controlled evaluation of a monovalent vaccine in a highly endemic environment. Am. J. Trap. Med. Hyg. 1955, 4, 281-288 Ferreccio, C., Prado, V., Ojeda, A., Cayazzo, M., Abrego, P., Guers, L. and Levine, M.M. Epidemiologic patterns of acute diarrhea and endemic Shige//a infections in children in a poor periurban setting in Santiago, Chile. Am. J. Epidemiof 1991, 134, 614-627 Taylor, D.N., Bodhidatta, L. and Echeverria, P. Epidemiologic aspects of shigellosis and other causes of dysentery in Thailand. Rev. infect Dis. 1991, 3 (Suppl 4) 5226-5230 Herrington, D.A., Van De Verg, L., Formal, S.B., Hale, T.L., Tall, B.D., Cruz, S.J., Tramont, E.C. and Levine, M.M. Studies in volunteers to evaluate candidate Shige//a vaccines: further experience with a bivalent Salmonella fyphf-Shigella sonnei vaccine and protection conferred by previous Shige//a sonnei disease. Vaccine. 1990, 8, 353-357 DuPont, H.L., Hornick, R.B., Snyder, M.J., Libonati, J.P., Formal, S.B. and Gangarosa, E.J. Immunity in shigellosis II, Protection induced by oral live vaccine or primary infection. J. fnfect Dis. 1972, 125, 12-16 Mel, D.M., Terzin, A.L. and Vuksic, L. Studies on vaccination against bacillary dysentery Ill. Effective oral immunization against Shigella ffexneri 2a in a field trial. Bull. W.H.O. 1965, 32, 647-655 Ferreccio, C., Formal, S., Levine, M.M., Lindberg, A.A., Manning, P.A., Meitert, T. et al. Development of vaccines against shigellosis: memorandum from a WHO meeting. Bull. NH.0. 1987, 85, 17-25 Levine, M.M., Woodward, W.E., Formal, S.B., Gemski, P., DuPont, H.L., Hornick, R.B. and Snyder, M.J. Studies with a new generation of oral attenuated Shigeffa vaccine: Escherichia coli bearing surface antigens of Shigela flexnen. J. Infect. Dis. 1977, 136, 577-582 Cohen, D., Green, M.S., Block, C., Slepon, R. and Ofek, I. Prospective study of the association between serum antibodies to lipopolysaccharide 0 antigen and the attack rate of shigellosis. J. Cfin. Microbiio. 1991, 29, 386-389 Nakata, N., Tobe, T., Fukuda, I., Suzuki, T., Komatsu, K., Yoshikawa, M. and Sasakawa, C. The absence of a surface protease, OmpT, determines the intercellular spreading ability of Shigela: the relationship between the ompT and kcpA loci. Molec. Microbial. 1993, 9, 459-468 Hale, T.L. Genetic basis of virulence in Shigella species. Microbial. Rev. 1991, 55, 206-224 Sansonetti, P.J., Hale, L.T., Dammin, G.J., Kapfer, C., Collins, H. H. Jr. and Formal, S.B. Alterations in the pathogenicity of Escherichia co/i K-12 after transfer of plasmid and chromosomal genes from Shigella ffexneri. Infect. lmmun. 1983, 39, 1392-l 402 Hartman, A.B., Powell, C.J., Schultz, C.L., Oaks, E.V. and Eckels, K.H. Small animal model to measure efficacy and immunogenicity of Shigela vaccine strains. infect. Immun. 1991, 59, 4075-4083 Ferreccio, C., Levine, M.M., Rodriguez, H., Contreras, R. and the Chilean Typhoid Committee. Comparative efficacy of two, three, or four doses of Ty2la live oral typhoid vaccine in entericcoated capsules: a field trial in an endemic area. J. infect. Dis. 1989, 159, 766-769 Mel, D. M., Arsic, B. L., Radovanovic, M. L. and Litvinjenko, S. A. Live oral Shige//a vaccine: vaccination schedule and the effect of booster dose. Acta Microbiof. Acad. Sci. Hung. 1974, 21, 109-l 14 Taylor, D.N., Phillip, D.F., Zapor, M., Trofa, A, Van de Verg, L., Hartman, A., Bendiuk, N. et al. Outpatient studies of the safety and immunogenicity of an auxotrophic Eschenchia co/i K-
Vaccine 1995 Volume 13 Number 5
501
E. coli-S. flexneri 2a hybrid vaccine: K.L. Kotloff et al.
36
502
12-ShigeMa flexnen’ 2a hybrid vaccine candidate, EcSf2a-2. Vaccine 1994, 12, 565-568 Gilman, R.H., Hornick, R.B., Woodward, W.E., DuPont, H.L., Snyder, M.J., Levine, M.M. and Libonati, J.P. Evaluation of a UDP-glucose-4-epimeraseless mutant of salmonella fyphi as a live oral vaccine. J. infect. Dis. 1977, 138, 717-722
Vaccine 1995 Volume 13 Number 5
37
Kantele, A. Antibody-secreting cells in the evaluation of the immunogenicky of an oral vaccine. Vaccine 1990, 8, 321-326 38 Levine, M.M., Ferreccio, C., Cryz, S. and Ortiz, E. A randomized, controlled field trial comparing three doses of Ty2la live oral typhoid vaccine administered in enteric-coated capsule or liquid formulations. Lancet1990, 336, 891-894
0264-410X(94)00011-5
Evaluation of the safety, immunogenicity, and efficacy in healthy adults of four doses of live oral hybrid Escherichia coli-ShigellQ flexneri 2a vaccine strrainEcSf2a-2 Karen L. Kotloff*+‘“, Genevieve A. Losonsky*fs, James P. Nataro*+$, Steven S. Wasserman.*‘, Thomas L. Hale-, David N. Taylor-, John W. Newland-, Jerald C. Sadoff-, Samuel B. Formal-’ and Myron M. Levine*+’ In previous trials, live invasive Escherichia coli-Shigella flexneri 2a hybrid vaccine candidate EcS’a-2, administered to adult volunteers as 3 doses of ca. 2~10~ colony forming units (c.fu.) spaced over one week, induced fever andlor diarrhea in 11% of subjects and provided only limited protection (36% ef$cacy) against illness following challenge with virulent ;S. flexneri 2a. We sought to improve the clinical safety of this vaccine by administering a lower inoculum, and to enhance protective immunity by administering additional booster doses at 2 weeks. Twenty-one healthy adults were immunized with ca. 7x.10” c.Ju. of EcS’a-2 on days 0, 3, 14, and 17. The vaccine consistently colonized the intestine without causing serious adverse reactions; mild diarrhea developed in one subject and low grade fever in another. Vaccination elicited an antibody secreting cell (ASC) response to lipopolysaccharide (LPS) in all subjects, which was highest on da,y 7 and notably diminished thereafter on days 10, 16, 21, and 24, suggesting that active mucosal immunity developed rapidly. The magnitude of the response was modest (geometric mean peak = 16 IgA ASCIIO” peripheral blood mononuclear cells) and an IgG serological response to LPS was detected in only 19% of subjects. Following experimental challenge with virulent S. flexneri 2a administered with bicarbonate buffer, shigellosis (diarrhea, dysentery, or fever) developed in 10 of 16 vaccine recipients (63O/t) and in 12 of 14 unvaccinated controls (86OA), resulting in a vaccine efficacy of 27% (95% confidence limits -197, 82, p = 0.15, l-tailed). We conclude that four doses of EcSfla-2 spaced over 17 days at inocula that are welltolerated do not confe,r significant protection against illness following experimental challenge of healthy adult volunteers. Keywords: Phase I study; mucocsal immunity: Shigellu flexneri
Although streptomycin-dependent and T,,-ISTRATI attenuated Shigella flexneri 2a strains were used with some success as live oral vaccines in the 1960~‘-~, *Center for Vaccine Development, University of Maryland School of Medicine, Baltimore, MD 21201, USA. +Division of Infectious Diseases and Tropical Pediatrics, Department of Pediatrics, University of IMaryland School of Medicine, Baltimore, MD 21201, USA. I-Division of Geographic Medicine, Department of Medicine, Center for Vaccine Development, University of Maryland School of Medicine, Baltimore, MD 21201, USA. “Division of Communicable Diseases and Immunology, Walter Reed Army Institute of Research, Walter Reed Army Medical Center, Washington, DC. 20307-5100, USA. “University of Maryland School of Medicine, Center for Vaccine Development, 10 S. Pine Street, Baltimore, MD 21201, USA. “To whom correspondence should be addressed. (Received 11 July 1994; revised 26 September 1994; accepted 26 September 1994)
subsequent efforts to develop improved Shigella vaccines have failed to yield a promising candidate. More recently, vaccines have been constructed which invade and replicate within the intestinal mucosa, enabling delivery of Shigella antigens directly to the intestinal lymphoid tissue in an attempt to provide strong and durable immunologic stimulation, both local and systemic4. 5. We have previously reported the results of the safety, immunogenicity, and efficacy of an epithelial cell invasive live oral Escherichia coli-S. flexneri 2a hybrid vaccine candidate, EcSf2a-26. In construction of this vaccine, chromosomal genes encoding the group and type-specific antigens of S. jlexneri 2a and an S. JEexneri 5 invasiveness plasmid were introduced into E. coli K12. Thereafter a mutation in the aroD gene was selected, rendering the strain dependent on para-aminobenzoic acid for growth, a substrate not available in human
Vaccine
1995 Volume
13 Number
5
495
E. coli-S. flexneri 2a hybrid vaccine: K.L. Kotloffet
al.
tissues’. In previous trials, 3 doses of ca. 2~10~ c.f.u. of EcSf2a-2 spaced over one week were immunogenic, but only modestly protective (36% efficacy) against challenge with virulent S. Jlexneri 2a. Furthermore, several subjects experienced fever and/or diarrhea after vaccination. In this report, we describe the results of further studies conducted to assess an alternate dosing regimen designed to improve clinical safety by administering a lower inoculum and to maximize protective immunity by using a 4-dose regimen spaced over 17 days. METHODS Subjects
Adults of ages 1840 years were recruited from the Baltimore-Washington area. Eligible subjects were without significant abnormalities detected by medical history, physical examination, and a panel of screening clinical laboratory tests. Occupational food handlers, travelers to Shigellu-endemic areas, persons with known allergy to ciprofldxacin or sulfa drugs, and those with HLA B-27 haplotype (because of its association with post-infectious arthritis’) were excluded. Preparation of the vaccine inoculum
Details of vaccine construction and formulation are presented elsewhere6. 7. The lyophilized vaccine was stored at - 70°C in sealed vials containing 4x10” viable organisms. Vaccine stability was confirmed one week before vaccination using colony counts and slide agglutination with Shigeflu group B antisera (Difco Laboratories, Detroit, MI), and by demonstrating retention of HeLa cell invasiveness’ and Congo Red positivity in > 95% of colonies. The inoculum was prepared by rehydrating three ampules of vaccine, each with 5 ml of sterile water. The vials were maintained on wet ice for 30 min, then pooled and serially diluted with phosphate buffered saline (PBS) to achieve a 1:4 dilution. Colony counts of reconstituted vaccine were determined before and after vaccination and the results were averaged to estimate the actual inoculum administered. Preparation of the challenge strain
The challenge inoculum was prepared from a frozen master seed stock of S, Jexneri 2a strain 2457T” which was plated onto trypticase soy agar (TSA, Becton Dickinson Microbiology Systems, Cockeysville, MD) into which Congo Red dye (O.Ol%, Sigma Chemical Company, St. Louis, MO) was added. After incubation at 35°C for 18-24 h, several individual Congo Red colonies that characteristic exhibited Shigella morphology were confirmed as group B Shigella using slide agglutination. Several well-isolated Congo Red colonies were picked and suspended in sterile saline. The saline suspension was then used to inoculate TSA plates. Overnight growth at 35°C from the heavily seeded TSA plates was harvested into sterile PBS. The bacterial suspension was diluted with additional sterile PBS to produce a suspension with an optical density at 660 nm corresponding to the desired bacterial count per ml.
496
Vaccine 1995 Volume 13 Number 5
Clinical design of vaccine study
Twenty-one outpatient volunteers who fasted 90 min before and after inoculation received an oral dose of vaccine on days 0, 3, 14, and 17. To buffer gastric acid, subjects drank 120 ml of a solution containing 2 g of NaHCO, in 150 ml distilled water. One minute later, they ingested vaccine suspended in the remaining 30 ml of buffer. For a period of 24 days after the first dose of vaccine, the volunteers were evaluated once daily using a standardized interview that was conducted either in person or by telephone, to elicit the occurrence of adverse clinical reactions during the preceding 24 h, including diarrhea, vomiting, fever, hematechezia, cramps, anorexia, and malaise. They recorded the number and consistency of all stools passed, and their evening oral temperature. Signs and symptoms occurring within 72 h of each inoculation are reported herein. Clinical design of challenge study
One month after the fourth dose of vaccine (day 45 following the first dose), 16 vaccine recipients and 14 unvaccinated control subjects were admitted to the Research Isolation Ward of the Center for Vaccine Development. In the fasting state, they were challenged with 1~10~ c.f.u. virulent S. JEexneri 2a along with bicarbonate buffer, as described for the vaccine. For five days following challenge, subjects were evaluated for evidence of illness by physicians and nurses who were blinded to the volunteers’ status as vaccinees or controls. Temperature, pulse and blood pressure were measured every 6 h. All stools passed were examined for blood, weighed and graded on a l-5 scale for consistency”. Subjects received a 5-day course of ciprofloxacin (500 mg by mouth every 12 h), beginning five days after challenge, or earlier if deemed clinically necessary by the investigator. Definitions
Illness was defined as diarrhea, fever (oral temperature 2 lOO.O”F) or dysentery (gross blood in a loose stool). For the outpatient vaccination study, diarrhea was defined as three or more liquid stools in 24 h. In the challenge study, diarrhea was defined as two or more loose stools (grade 3-5) totalling 200 ml within 48 h or a single 300 ml grade 3-5 stool. Protective efficacy was determined in the challenge studies for diarrhea, dysentery, fever, and any illness as follows: [(attack rate of illness in controls - attack rate of illness in vaccinees)/attack rate of illness in controls]xlOO. Bacteriology
A perirectal swab following defecation was collected from each volunteer on days O-5, 7, 10, 14-19, 21, 24, and 28 following the first dose of vaccine to assess duration of vaccine shedding. The swab was immediately placed in Cary Blair transport medium (Becton Dickinson) and refrigerated for up to 24 h before transport to the laboratory. For participants in the inpatient challenge study, a swab of all stools passed during hospitalization and weekly for 3 weeks following discharge was obtained, inoculated into buffered glycerol saline, and maintained at 46C until arrival at the laboratory. A rectal swab was obtained from any
E. coli-S. flexneri 2% hybrid vaccine: K.L. Kotloff et al. compare vaccinees and controls because the question of interest was whether the vaccine reduced the rate of illness. Severity of signs and symptoms was compared by one-tailed Wilcoxon tests. To compare ASC responses, data were transformed to natural logs and analyzed by z-tests; geometric means for data including zeros were calculated by adding one to each datum prior The effects of vaccination to log-transformation’5. status and illness post-challenge upon post-challenge ASC and serologic responses were examined by analysis of variance (independent factors were vaccination status, illness and their interaction; log-transformed peak ASC or reciprocal immunoglobulin titer served as the dependent variable); each assay was analyzed separately. The correlation between peak geometric mean ASC response (log transformation) and protective efficacy (angular transformation) was examined by Pearson’s correlation coefficient. Resulting probabilities c 5% were considered statistically significant.
volunteer who failed to produce a stool in a required 24 h period and inoculated into Gram-negative (GN, Becton Dickinson) broth. Swabs and stools were inoculated directly onto MacConkey’s agar (Becton Dickinson) containing 500 pg ml-’ streptomycin for recovery of the streptomycinresistant vaccine strain; Salmonella-Shigella (Becton Dickinson) and MacCortkey’s agar without antibiotics were used for recovery of the Shigella challenge strain. Each specimen was also inoculated into GN enriehment broth and incubated overnight at 35°C before being plated onto enteric media. Lactose-negative colonies identified after 24 or 48 h incubation at 35°C on solid media were inoculated onto Triple Sugar Iron (TSI, Becton Dickinson) slants. Those producing an alkaline slant and an acid butt without gas were confirmed by slide agglutination. Measurement
of serum antibodies
To measure the serum antibody response to vaccination, serum specimens were collected on days 0, 7, 21, 28, 35, and 45 following the first dose of vaccine; to measure the response to challenge, serum was collected on days 0, 10, 21 and 28 following challenge with pathogenic Shigella (Table I). Serum antibodies (IgA, IgM, and IgG) to the invasiveness plasmid antigens (IPA) and LPS 0 antigens of S. JEexneri 2a were measured by an enzyme-linked immunosorbent assay (ELISA), as previously described”. 13. Four-fold rises in titer were considered significant. Detection
RESULTS Clinical response to vaccination
After each dose, between 3 and 5 of the 21 vaccinees experienced at least one constitutional symptom or sign of illness, which in every instance was judged to be mild by the volunteer. One subject reported diarrhea after the first dose (4 loose stools in 24 h) and again after the third dose (7 loose stools in 48 h). Another subject had low grade fever (I 100.2”F) after the first and third doses of vaccine.
of antibody secreting cells
To measure the antibody secreting cell (ASC) response to vaccination, peripheral blood mononuclear cells (PBMC) were collected on days 0, 7, 10, 16,21, and 24 days after the first dose of vaccine; the ASC response to challenge was measured on days 0 and 7 in relation to challenge (Table I). Circulating IgA, IgM, IgG ASC against the IPA and LPSl antigens were determined by count exceeding 3 ELISPOT14. A post-vaccination standard deviations above the geometric mean prevaccination or pre-challenge ASC was considered a positive ELISPOT response.
Vaccine excretion
Statistical
Immune response to vaccination
All subjects excreted the vaccine strain on at least one day. Excretion was detected 24 h after vaccination in 18 of 21 subjects (86%) after dose 1, in 14 of 21 subjects (67%) after dose 2, in 11 of 20 subjects (55%) after dose 3, and in 14 of 21 subjects (67%) after dose 4. Similarly, the duration of shedding (14 days) did not decline significantly with consecutive dosing. The HeLa cell invasive phenotype was found in 70-95% of vaccine strains randomly selected from volunteers’ stools.
analysis
Circulating IgA ASC specific for LPS was detected in all volunteers during the 24 days after the first dose of vaccine, and 90% had an IgG ASC response. However, the magnitude of the anti-LPS ASC response was significantly lower than the response elicited by challenge of unvaccinated controls (Table 2). The peak
Unless otherwise specified, two-tailed Fisher’s exact and x2 tests were used to compare proportions. A sample size for the challenge study was chosen that would allow detection at least 50% protective efficacy if the attack rate of illness in controls were 85% (p = 0.05, one-tailed); one-tailed Fisher’s exact tests were used to
Table 1
Schedule of inoculations and collection of specimens for measurement of immune response Day specimen collected following vaccination
Vaccination Serum antibody Antibody secreting cells Challenge Serum antibody Antibody secreting cells “Day 45 of vaccination
0
3
X
X
7
10
14
16
X
X
X
24
21
X X
26
35
45
X
X
XS
X
X
X
17
or challenge
X
X
X
Xe X
and day 0 of challenge
X
X
X
X X
are the same day
Vaccine
1995
Volume
13 Number
5
497
E. coli-S.
flexneri
2a hybrid
vaccine:
K.L. KotM
et al.
Table 2 immune responses to lipopolysaooharide (LPS) or invasiveness plasmid antigens (IPA) elicited by vaccination with 4 doses of EcSf2a-2 or by challenge of unvaccinated controls with virulent S. flexneti 2a Response to vaccination Geometric mean peak
with EcSf2a-2
No. (%)
Response to challenge
responders
Geometric mean peak
with S.
flexneri 2a
No. (%) responders
Antibody secreting cells (ASC)” Anti-LPS lgA lgG Anti-IPA
16” 5”
21/21 (100) 17/19 (90)
143” 24”
14/14 (100) 13/14 (93)
lgA jgG Serum antibodyb Anti-LPS
16 5
21/21 (100) 17/20 (85)
54 6
13114 (93) 11/13 (85)
lgA lgG Anti-IPA
121 53
6/21 (29)” 4/21 (19)
173 78
11114 (79)d 7114 (50)
lgA lgG
71” 2.51”
l/21 (5) 3/21 (14)O
220” 1073”
4/l 4 (29) 8/14 (57)“
“ASC response to vaccination was defined for each assay as a post-vaccination ASC count > 3 standard deviations above the geometric mean pre-vaccination ASC count. Accordingly, the following response levels were established: 2 2 ASC per lo6 peripheral blood mononuclear cells (PBMC) for anti-LPS IgG and IgA, t 1 ASC per lo6 PBMC for anti-IPA IgG and IgA “Serum antibody response defined as a fourfold rise in reciprocal titer following inoculation. Means are expressed as geometric mean of reciprocal titer %omparison of geometric mean peak response to a specific assay between vaccinees and challenge subjects, ~~0.05 dComparison of the proportion of responders in a specific assay between vaccinees and challenge subjects, ~~0.05
anti-LPS IgA ASC response to vaccination was observed 7 days after the first dose, decreasing markedly thereafter (Figure 2). Anti-IPA ASC responses were also seen in most volunteers (Table 2). Vaccination elicited a four-fold rise in serum IgG anti-LPS antibody in 4 subjects (19%), while 6 (29%) had a significant rise in IgA anti-LPS antibody (Table 2). Seroconversions of IgG and IgA anti-IPA were detected in 3(14%) and 1(5%) of subjects, respectively. IgA anti-LPS and IgG anti-IPA seroconversion was significantly less frequent following vaccination compared with challenge.
. \
.
\ 12
Clinicai evaluation of vaccine efficacy following wildtype challenge with S. jlexneui 2a
The overall attack rate for illness was 86% in controls and 63% in vaccinees (protective efficacy 27%, 95% confidence limits - 197, 82, p = 0.15, Table 3). There was a trend toward lower diarrhea1 volumes in vaccinees; however, there were no statistically significant differences between vaccinees and controls in
I8 t
24
t
Days after initial vaccination Figure 1 Anti-lipopolyaaccharide IgA antibody secreting cell (ASC) response expressed as number of cells per lo6 peripheral blood mononuclear cells (PBMC) to oral vaccination with live attenuated Shige//a flexnen’ 2a vaccine EcSf2a-2 in 4 spaced doses on days 0, 3, 14, and 17 (vaccination days indicated by arrows)
Table 3 Clinical and microbiological responses of vaccinated (four spaced doses of EcSf2a-2) and unvaccinated control volunteers following challenge with wild-type Shigela flexnen’2a Attack rate for: Group
Attack rate for any illness
Fever
Diarrhea
Dysentery
No. (%) with fecal shedding
Controls
12114 (86%) lot16 (63%) 27%
10114 (71%) 9116 (56%) 21%
9114 (64%) 8/l 6 (50%) 22%
9114 (64%) 9116 (56%) 13%
13114 (93%) 15116 (94%) 0
- 197,82 0.15
-169,77 0.31
- 155,76 0.34
-188,73 0.47
1.0
Vaccinees Efficacy” 95% confidence limits p value*
“Protective efficacy was determined controls] x 100 OFisher’s exact test, single tail
498
Vaccine
1995
Volume
as follows: [(attack rate of illness in controls
13 Number
5
-
attack rate of illness in vaccinees)/attack
rate of illness in
E. coli-S. flexneri 2a hy&id Table 4 Comparison of height of fever and total stool volume in ill EcSf2a-2 vaccine recipients and ill controls following challenge with wild-type Shigella flexneri 2a
Group
Mean peak fever
Controls Vaccinees” p value*
102.8”F 102.2”F 0.13
Diarrhea stool volume (Time from onset of first loose stool) (24 h)
(30 h)
999 ml 692 ml 0.17
1149 ml 930 ml 0.28
“One vaccinee received early treatment with ciprofloxacin 26 h after the onset of loose stools Wilcoxon test, single tail
beginning
the frequency or severity of these symptoms (Tables 3 and 4). Relationship between immune response to vaccination and protection against illness after challenge
Examination of individual ASC levels after vaccination revealed that the volunteer with the highest anti-LPS IgA ASC counts (333/106 PBMC) and antiIPA IgA ASC counts (1.30/106 PBMC) was protected against illness after challenge, whereas the next highest IgA responder (157 ASC/I O6PBMC against LPS and 51 ASC/106 PBMC against IPA) was not protected. A significant association between peak IgA anti-LPS ASC response to vaccination and resistance to illness following challenge could not be demonstrated statistically @ = 0.9). One of the 10 vaccinees (10%) who became ill after challenge had a fourfold rise in serum IgG anti-LPS after vaccination compared with 2 of 6 vaccinees (33%) who resisted illness @ = 0.52). Five of the unprotected (50%) and 1 of the protected volunteers (20%) had a fourfold rise in serum IgA anti-LPS after vaccination (p = 0.31). Immune response to challenge in vaccinated and unvaccinated subjects
Differences could not be detected between vaccinees and controls in the magnitude and frequency of ASC or serologic responses after challenge. However, subjects who became ill after challenge had significantly greater geometric mean peak ASC responses following challenge in each assay examined, as well as greater geometric mean peak reciprocal anti-LPS IgA and IgG serologic titers, compared to those who remained well (data not shown).
DISCUSSION Several lines of evidence suggest that shigellosis can be prevented with a live oral vaccine that directly stimulates the intestinal immune system. For one, intestinal immunity appears to be an important factor in controlling Shigella infection, as illustrated by the efficacy of passively transferred oral immunoglobulin in preventing shigellosis’6. ” and the possible role of mucosal antibody in limiting the duration of illness’8; in contrast, parenteral immunization with killed Shigella organisms failed to protect against illness despite high levels of vaccine-induced systemic antibody”, *“.
vaccine: /CL. Kofloffet
al.
Second, an initial episode of shigellosis confers serotypespecific immunity, as has been observed in children living in endemic areas2’.22 and in experimental infection of humans with S. sonnei23 and S. Jlexneri 2a24. Finally, noninvasive streptomycin-dependent Shigella strains utilized as live oral vaccines induce high level, but short-lived (< 1 year) protection against shigellosis as demonstrated in field trials conducted in Yugoslavia’. 25. The requirement for four large (> 10” c.f.u.) doses in the primary immunization schedule followed by annual boosters thereafter was considered a limitation of these vaccine?. It is of interest that oral administration of a noninvasive E. coli strain expressing S. JEexneri 2a 0 antigen was poorly immunogenic and nonprotective in experimental challenge studies in vo1unteers2’. To overcome the perceived drawbacks of the noninvasive oral vaccines, EcSf2a-2 was constructed with the following features: (1) epithelial cell invasiveness to enhance stimulation of the mucosal immune system; (2) 0 antigen expression to induce type-specific antibodies which have been correlated with resistance to shigellosis among Israeli soldiers serving in a field unitZ8; (3) limitation of intracellular multiplication due to a nutritional dependence on a substrate, paraaminobenzoic acid, not available in human tissues; (4) presence, in the E. coli carrier, of the ompT chromosomal genes degrading Shigella proteins necessary for intercellular spread29; and (5) absence of aerobactin and perhaps other critical Shigella virulence factors in the E. coli Carrie?‘. 3’. Initial volunteer studies suggested that immunization with ca. 2~10~ c.f.u. of EcSf2a-2 in 3 doses spaced over a week was reactogenic and could not induce significant protection against experimental challenge6. Accordingly, we sought to evaluate alternate dosing regimens. Findings from other studies suggested that a four dose schedule of EcSf2a-2 might enhance protective immunity. In the guinea pig keratoconjunctivitis model, vaccine efficacy was three times higher using a four dose regimen with vaccination on days 0, 2, 14 and 15 than with a three dose regimen with vaccination on days 0, 2, and 4’2. In this model, animals were immunized in the conjunctival sac with ea. 3.8~10~ c.f.u. of EcSf2a-2 and then challenged 14-28 days after the last dose of vaccine with a homologous wild-type Shigella strain (2457T). Furthermore, experience in humans with other oral attenuated bacterial vaccines, such as the typhoid vaccine Ty2 1a33 and the streptomycin-dependent attenuated Shigella vaccines”. 34 suggested that the level of protection correlates with the number of doses given. It was therefore hypothesized that at a slightly lower inoculum (7x108 c.f.u.) to ensure clinical safety, a four dose regimen of EcSf2a-2 would enhance vaccine efficacy. In general, the vaccine was well tolerated at this lower inoculum level. Several volunteers reported mild diarrhea, low grade fever, or mild constitutional complaints, but in the absence of a placebo control group, these symptoms could not be attributed to the vaccine in any individual case. In an expanded Phase II study conducted among active duty US military personnel, adverse reactions did not occur significantly more frequently among 205 recipients of four doses of EcSfZa-2 ranging from 2.3 to 9.0~10~ c.f.u. compared to 39 placebo recipients35.
Vaccine 1995 Volume 13 Number 5
499
E. coli-S. flexneri 2a hybrid vaccine: K.L. Kofloffet
al.
Patterns of vaccine excretion and ASC response to consecutive doses of vaccine can be examined to evaluate the ability of EcSf2a-2 to stimulate mucosal immunity. The vaccine consistently colonized the intestine. The highest colonization rate (87%) occurred after the first dose; however, significantly diminished colonization with consecutive doses, which might indicate vaccine-induced mucosal immunity”- 36, was not seen. Kantele has suggested that the ASC response following primary immunization is a method for assessing protective immunity conferred by live oral vaccination3’. This investigator demonstrated that three doses of Salmonella typhi Ty2la vaccine formulated as an enteric-coated capsule elicited a significantly lower geometric mean peak ASC response among volunteers compared with three doses of the vaccine formulated as a suspension. These patterns of ASC response agree with observations obtained in field trial that the liquid suspension provides the best efficacy3’. With our small sample sizes, we could not demonstrate a significant correlation between the ASC response to four doses of EcSf2a-2 Shigella vaccine and protective efficacy following challenge. However, it is notable that the primary response to vaccination was modest (geometric mean peak IgA LPS response 16 ASC/106 PBMC), particularly when compared to the response that accompanied the protective immunity elicited by experimental inoculation with virulent infection (143 ASC/lO” PBMC), and may have been insufficient to provide significant protection. Indeed, a pattern emerging from our recent trials suggests that the peak geometric mean anti-LPS IgA ASC response induced by vaccination or wild-type infection correlates with the ability of the immunizing strain to protect volunteers against shigellosis following experimental inoculation (Table 5 and K. Kotloff, submitted for publication). The challenge that remains for future Shigellu vaccine development is to achieve vigorous immunologic stimulation without causing reactogenicity.
We conclude that four doses of EcSI2a-2 in this trial did not enhance the protective efficacy of the vaccine elicited by three doses. However, a caveat in comparing the results of challenge studies involving recipients of three vs four doses of EcSf2a-2 must be considered. This study provided the first experience administering virulent Shigella in suspension with bicarbonate buffer instead of skimmed milk. This modification in the vehicle for inoculation of the challenge strain was intended to improve the study-to-study consistency of illness rates among unvaccinated controls. The ability of the modified model to demonstrate protective efficacy has been validated using homologous rechallenge studies of volunteers (K. Kotloff, submitted for publication). None the less, while the results of challenge studies using skimmed milk as a vehicle24 have been corroborated by field experience25, the level of protective efficacy of EcSf2a-2 as demonstrated in volunteers using the modified challenge model awaits comparison with the results of a vaccine field trial currently being conducted in Israel.
ACKNOWLEDGEMENTS This study was supported by contract No. DAMD-1 7-92-C-2058 from the U.S. Army Medical Materiel Development Command. We thank the volunteers who participated in these studies, Kathleen Palmer and Brenda Berger for assistance in patient recruitment and outpatient follow-up, Cathy Black, Ron Grochowski, and the nursing staff of the Center for Vaccine Development for providing clinical care, Linda Guers, Mardi Reymann, and Karen Taylor for technical assistance, and Drs. Carol Tacket and Robert Edelman for clinical support and valuable suggestions. Statement: informed consent was obtained from the volunteers and guidelines for human experimentation of the U.S. Department of Health and Human Services, the Human Volunteers Research Committee of the University of Maryland, and the Surgeon General’s Human Subjects Research Review Board of the
Table 5 Relationship between peak IgA antibody secreting cell (ASC) response to lipopolysaccharide following inoculation with wild-type Shigella flexneti 2a or with live oral EcSf2a-2 vaccine and protection against homologous wild-type challenge ASC response to stated immunizing regimen among subjects who later participated in efficacy trial Immunizing regimen
lnoculum
S. flexneri 2a’ EcSf2a-2” EcSf2a-2
1 x lo3 c.f.u. 2 x lo* c.f.u. 7 x lo8 c.f.u.
EcSf2a-2”
2 x lo0 c.f.u.
No. doses
Days of inoculation
Days ASC measured
1 3 4
0 0, 3, 6 0, 3, 14, 17
3
0, 7, 10
0, 7 0, 7, 10 0, 7, 10, 16, 21, 24 0, 7, 10
n
Peak geometric mean No. ASC per lo6 PBMC”
Eff icacy”,b
11 10 16
140 27 15
70%d 48% 27%
10
4
0%
PBMC = peripheral blood mononuclear cells; c.f.u. =colony forming units =V 0.986, p=O.Ol, 2-tailed, Pearson correlation coefficient, comparing peak geometric mean No. ASC/106 PBMC and level of protective efficacy *Protective efficacy against homologous challenge with wild-type S. flexneri 2a approximately one month after completion of immunizing regimen 70 assess the protective efficacy of wild-type infection, previously vaccinated (4 spaced doses of EcSf2a-2) and unvaccinated veterans of a S. flexneri 2a challenge study were rechallenged with wild-type S. flexneri 2a along with a group of immunologically naive control subjects (K. Kotloff, submitted for publication) dp
500
Vaccine 1995 Volume 13 Number 5
E. coli-S. flexneri 2a hybrid vaccine: K.L. Kofloffet
Department of the Army were followed in the conduct of the clinical research. Disclaimer: The views o:Fthe authors do not reflect the positions of the US Department of the Army or of the US Department of Defense.
18
19
REFERENCES 1
2
3
4
5
6
7
8 9
10
11
12
13
14
15
16
17
Meitert, T.E., Pencu, L., Ciudin, L. and Tonciu, M. Vaccine strain Shigella flexned T,-ISTR,ATI. Studies in animals and in volunteers. Antidysentery immunoprophylaxis and immunotherapy by five vaccine Vadizen (S.. flexnen’ T,,ISTRATI). Arch. Roum. P&ho/. EXD. Microbiiol. 1984. 43. 251-278 Mel, D., Gangarosa, E.J., Radovanovic, M.L., Arsic, B.L. and Litvinjenko, S. Studies on vaccination against bacillary dysentery. 6. Protection of children by oral immunization with streptomycin-dependent Shigella strains. Bull. WH.0. 1971, 45, 457-464 Mel, D.M., Arsic, B.L., Nikolic, B.D. and Radovanic, M.L. Studies on vaccination against bacillary dysentery. Bull. W.H.O. 1968, 39, 375-380. Formal, S.B., Baron, L.S.. Kopecko, D.J., Washington, O., Powell, C. and Life, C.A. Construction of a potential bivalent vaccine strain: introduction of Shigella sonnei form I antigen genes into the .galE Salmonella typhi Ty21 a typhoid vaccine strain. Infect. fmmun. 1981, 34, 746-754 Formal. S.B.. Hale T.L.. Kapfer. C.. Coaan. J.P.. Snov. P.J.. Chung,’ R. ei a/. Oral vaccination of ‘monkeys with an invasive Escherichia cofi K-l 2 hybrid expressing Shigella fiexneri 2a somatic antigen. Infect fmmun. 1984, 46, 465-469 Kotloff, K.L., Herrington, D.A., Hale, T.L., Newland, J.W., Van De Verg, L., Cogan, J.P. 1st al. Safety, immunogenicity, and efficacy in monkeys and humans of invasive Eschedchia cofi K12 hybrid vaccine candidai:es expressing Shigela flexnen 2a somatic antigen. lnfecf. lmmun. 1992, 80, 2218-2224 Newland, J.W., Hale, T.L. and Formal, S.B. Genotypic and phenotypic characterization of an aroD deletion-attenuated kschenchia colLShigel/a flexnen hybrid vaccine expressing S. ffexneti 2a somatic antiaen. Vaccine 1992. 10, 766-776 Keat, A. Reiter’s syndrome and reactive arthritis in perspective. N. Engl. J. Med. 1983, 26, 1606-l 614 Hale, T.L. and Formal, S.B. 1981. Protein synthesis in HeLa or Henle 407 cells infected with Shigela dysentenae 1, Shigela flexnen 2a, or Salmonella typhimurium W118. fnfect lmmun. 32, 137-l 44 DuPont, H.L., Levine, M.M., Hornick, R.B. and Formal, S.B. lnoculum size in shigellosis and implications for expected mode of transmission. J. infect Dis. 1989, 159, 1126-l 128 Levine, M.M., Black, R.E., Clements, M.L. Lanata, C., Sears, S., Honda, T. et a/. Evaluation in humans of attenuated Vibrio cholerae El Tor Ogawa strain Texas Star-SR as a live oral vaccine. Infeci Immun. 1984, 43, 515-522 Black, R.E., Levine, M.M., Clements, M.L., Losonsky, G., Herrington, D., Berman, S. and Formal, S. Prevention of shigellosis by a Salmoneih iyphi-ShigeMa sonnei bivalent vaccine. J. Infect. Dis. 1987, 155, 1260-1265 Oaks, E.V., Hale, T.L. arid Formal, S.B. Serum immune response to Shigella protein antigens in rhesus monkeys and humans infected with Shigelfa spp. infect. Immun. 1986, 53, 57-83 Van de Verg, L., Herrington, D.A., Murphy, J.R., Wasserman, S.S., Formal, S.B. and Levine, M.M. Specific immunoglobulin Asecreting cells in peripheral blood of humans following oral immunization with a bivalerlt Salmonella fyphi-Shigeffa sonnei vaccine or infection by pathogenic S. sonnei. Infect. Immun. 1990, 58, 2002-2004 Sokal, R.R. and Rohlf, F.J. Biometry The Principles and Practice of Statistics in Biological Research. W.H. Freeman and Co., San Francisco, 1981 Tacket, C.O., Binion, S.B., Bostwick, E., Losonsky, G., Roy, M.J. and Edelman, R. Efficacy of bovine milk immunoglobulin concentrate in preventing illness after Shigella flexnerichallenge. Am. J. Trap. Med. Hyg. 1992, 47, 276-283 Hayani, KC., Guerrero, M.L., Morrow, A.L., Gomez, H.F., Winsor. D.K.. Ruiz-Palaclos. G.M. and Clearv, T.G. Concentration of milk secr~atory immunoglobulin A- .against Shige//a virulence plasmid-associated antigens as a predictor of
20
21
22
23
24
25
26
27
28
29
30 31
32
33
34
35
al.
symptom status in Shigela-infected breast fed infants. J. Pediatr. 1992, 121, 852-856 Oberhelman, R.A., Kopecko, D.J., Salazar-Lindo, E., Gotuzzo, E., Buysse, J.M., Venkatesan, M.M. ef al. Prospective study of systemic and mucosal immune responses in dysenteric patients to specific ShigeMa invasion plasmid antigens and lipopoysaccharides. Infect. fmmun. 1991, 59, 2341-2350 Shaughnessy, H.J., Olsson, R.C., Bass, K., Friewer, F. and Levinson, S.O. Experimental human bacillaty dysentery. Polyvalent dysentery vaccine in its prevention. J.A.M.A. 1946, 132, 362-368 Higgins, A.R., Floyd, T.M. and Kader, M.A. Studies in shigellosis. Ill. A controlled evaluation of a monovalent vaccine in a highly endemic environment. Am. J. Trap. Med. Hyg. 1955, 4, 281-288 Ferreccio, C., Prado, V., Ojeda, A., Cayazzo, M., Abrego, P., Guers, L. and Levine, M.M. Epidemiologic patterns of acute diarrhea and endemic Shige//a infections in children in a poor periurban setting in Santiago, Chile. Am. J. Epidemiof 1991, 134, 614-627 Taylor, D.N., Bodhidatta, L. and Echeverria, P. Epidemiologic aspects of shigellosis and other causes of dysentery in Thailand. Rev. infect Dis. 1991, 3 (Suppl 4) 5226-5230 Herrington, D.A., Van De Verg, L., Formal, S.B., Hale, T.L., Tall, B.D., Cruz, S.J., Tramont, E.C. and Levine, M.M. Studies in volunteers to evaluate candidate Shige//a vaccines: further experience with a bivalent Salmonella fyphf-Shigella sonnei vaccine and protection conferred by previous Shige//a sonnei disease. Vaccine. 1990, 8, 353-357 DuPont, H.L., Hornick, R.B., Snyder, M.J., Libonati, J.P., Formal, S.B. and Gangarosa, E.J. Immunity in shigellosis II, Protection induced by oral live vaccine or primary infection. J. fnfect Dis. 1972, 125, 12-16 Mel, D.M., Terzin, A.L. and Vuksic, L. Studies on vaccination against bacillary dysentery Ill. Effective oral immunization against Shigella ffexneri 2a in a field trial. Bull. W.H.O. 1965, 32, 647-655 Ferreccio, C., Formal, S., Levine, M.M., Lindberg, A.A., Manning, P.A., Meitert, T. et al. Development of vaccines against shigellosis: memorandum from a WHO meeting. Bull. NH.0. 1987, 85, 17-25 Levine, M.M., Woodward, W.E., Formal, S.B., Gemski, P., DuPont, H.L., Hornick, R.B. and Snyder, M.J. Studies with a new generation of oral attenuated Shigeffa vaccine: Escherichia coli bearing surface antigens of Shigela flexnen. J. Infect. Dis. 1977, 136, 577-582 Cohen, D., Green, M.S., Block, C., Slepon, R. and Ofek, I. Prospective study of the association between serum antibodies to lipopolysaccharide 0 antigen and the attack rate of shigellosis. J. Cfin. Microbiio. 1991, 29, 386-389 Nakata, N., Tobe, T., Fukuda, I., Suzuki, T., Komatsu, K., Yoshikawa, M. and Sasakawa, C. The absence of a surface protease, OmpT, determines the intercellular spreading ability of Shigela: the relationship between the ompT and kcpA loci. Molec. Microbial. 1993, 9, 459-468 Hale, T.L. Genetic basis of virulence in Shigella species. Microbial. Rev. 1991, 55, 206-224 Sansonetti, P.J., Hale, L.T., Dammin, G.J., Kapfer, C., Collins, H. H. Jr. and Formal, S.B. Alterations in the pathogenicity of Escherichia co/i K-12 after transfer of plasmid and chromosomal genes from Shigella ffexneri. Infect. lmmun. 1983, 39, 1392-l 402 Hartman, A.B., Powell, C.J., Schultz, C.L., Oaks, E.V. and Eckels, K.H. Small animal model to measure efficacy and immunogenicity of Shigela vaccine strains. infect. Immun. 1991, 59, 4075-4083 Ferreccio, C., Levine, M.M., Rodriguez, H., Contreras, R. and the Chilean Typhoid Committee. Comparative efficacy of two, three, or four doses of Ty2la live oral typhoid vaccine in entericcoated capsules: a field trial in an endemic area. J. infect. Dis. 1989, 159, 766-769 Mel, D. M., Arsic, B. L., Radovanovic, M. L. and Litvinjenko, S. A. Live oral Shige//a vaccine: vaccination schedule and the effect of booster dose. Acta Microbiof. Acad. Sci. Hung. 1974, 21, 109-l 14 Taylor, D.N., Phillip, D.F., Zapor, M., Trofa, A, Van de Verg, L., Hartman, A., Bendiuk, N. et al. Outpatient studies of the safety and immunogenicity of an auxotrophic Eschenchia co/i K-
Vaccine 1995 Volume 13 Number 5
501
E. coli-S. flexneri 2a hybrid vaccine: K.L. Kotloff et al.
36
502
12-ShigeMa flexnen’ 2a hybrid vaccine candidate, EcSf2a-2. Vaccine 1994, 12, 565-568 Gilman, R.H., Hornick, R.B., Woodward, W.E., DuPont, H.L., Snyder, M.J., Levine, M.M. and Libonati, J.P. Evaluation of a UDP-glucose-4-epimeraseless mutant of salmonella fyphi as a live oral vaccine. J. infect. Dis. 1977, 138, 717-722
Vaccine 1995 Volume 13 Number 5
37
Kantele, A. Antibody-secreting cells in the evaluation of the immunogenicky of an oral vaccine. Vaccine 1990, 8, 321-326 38 Levine, M.M., Ferreccio, C., Cryz, S. and Ortiz, E. A randomized, controlled field trial comparing three doses of Ty2la live oral typhoid vaccine administered in enteric-coated capsule or liquid formulations. Lancet1990, 336, 891-894