Abstract
We evaluated immune responses following bivalent oral cholera vaccination (Shanchol [Shantha Biotechnics]; BivWC) in a cohort of 25 human immunodeficiency virus (HIV)–infected adults in Haiti. Compared with adults without HIV infection, vaccination in HIV-infected individuals resulted in lower vibriocidal responses against Vibrio cholerae O1, and there was a positive relationship between the CD4+ T-cell count and vibriocidal responses following vaccination. Nevertheless, seroconversion occurred at a rate of 65% against the Ogawa serotype and 74% against the Inaba serotype in adults with HIV infection. These results suggest that the vaccine retains substantial immunogenicity in adults with HIV infection and may benefit this population by protecting against cholera.
Keywords: Cholera vaccine, HIV, Shanchol (BivWC), immunogenicity
Despite increasing geographic overlap between the global pandemics of cholera and human immunodeficiency virus (HIV) infection, little information is available on how HIV infection influences the susceptibility to Vibrio cholerae and the immune response to oral cholera vaccines. A study of the 2005 cholera outbreak in Mozambique suggested a higher attack rate among HIV-infected individuals than among HIV-uninfected individuals [1], but the study was limited by the small number of enrollees who agreed to voluntary HIV testing. A 2010 study in Port au Prince, Haiti, demonstrated an 11% prevalence of HIV infection in a cohort of patients presenting with cholera, compared with a prevalence of <2% in the community [2]. Although these findings are limited in scope, both suggest a potential association between HIV infection and vulnerability to cholera and highlight the need for a better understanding of the effectiveness of cholera prevention efforts, such as oral cholera vaccination, in individuals with HIV infection.
There are currently 2 licensed cholera vaccines; both are orally administered killed whole-cell vaccines. One vaccine contains both the Inaba and Ogawa serotypes of V. cholerae O1 along with recombinant cholera toxin B subunit (WC-rBS), and it is marketed as Dukoral (Crucell). In a case-control study conducted in 2004 in Biera, Mozambique, the WC-rBS vaccine was associated with 78% protection overall, despite an estimated 20%–30% prevalence of HIV infection in this community [3]. A newer bivalent oral cholera vaccine contains V. cholerae serogroups O1 and O139 but lacks the cholera toxin B subunit (BivWC), and it is marketed as Shanchol (Shantha Biotechnics). BivWC is currently more affordable and easier to administer than WC-rBS and may be associated with longer-lasting immunity against cholera [4]. As part of comprehensive cholera control efforts in Haiti, the Haitian Ministry of Health and its partners are rolling out the BivWC vaccine to targeted populations. An assessment of a previously licensed live attenuated oral cholera vaccine, CVD103HgR, found that HIV-infected individuals had a significant but lower rise in vibriocidal antibody titer after vaccination [5]. However, an assessment specifically examining the immunogenicity and efficacy of the currently licensed oral cholera vaccines in individuals with HIV infection has not been reported.
In this study, we evaluated immune responses following immunization with BivWC in a cohort of HIV-infected adults in Haiti. We evaluated vibriocidal antibody responses—the best characterized immunologic correlate of protection against cholera—as well as immunoglobulin A (IgA) responses to the O antigen–specific polysaccharide (OSP), a surrogate of the mucosal immune response against the major protective antigen of V. cholerae. We compared the immunogenicity of the vaccine in individuals with and those without known HIV infection and evaluated the effect of HIV-associated immune suppression, as assessed by CD4+ T-cell count, on vaccine responses.
METHODS
Study Participants
The study was conducted in St. Marc, Haiti, in 2013. Adults with HIV infection and healthy adults with no known history of HIV infection were eligible to receive BivWC and to participate in a study of immune responses. Adults with HIV infection were identified through a well-established HIV clinical program run by Partners In Health and were eligible for the study if they had a CD4+ T-cell count measured within the last 6 months. The immunogenicity of the vaccine in adults with HIV infection was compared with that in a cohort of adults without known HIV infection, which has been previously described [6]. Exclusion criteria for both the HIV-infected adults and the adults without known infection included pregnancy, serious chronic illness, prior receipt of any oral cholera vaccine, and/or a history of hospitalization for cholera since its introduction into Haiti in October 2010. This study was approved by the institutional review board of Partners HealthCare and the Haitian National Bioethics Committee. Written informed consent was obtained from all participants.
Vaccine Administration and Specimen Collection
The BivWC vaccine was stored at 2°C–8°C prior to administration. Two doses of vaccine were given 14 days apart, consistent with the manufacturer's protocol. Each 1.5-mL dose was administered orally from a single dose vial. Participants were monitored after vaccination for 30 minutes and were asked to return if they felt ill after receipt of the vaccine. Physicians monitored participants for adverse events. Venous blood specimens were obtained prior to immunization and 7 days after each dose of vaccine (on days 0, 7, and 21).
Laboratory Procedures
Serum was stored at −80°C and shipped on dry ice to Massachusetts General Hospital for the performance of immunological analyses. Vibriocidal antibody assays were performed using target strains of V. cholerae O1 Inaba (strain T19479) and V. cholerae O1 Ogawa (strain X25049), which were incubated in the presence of inactivated serum and exogenous guinea pig complement as previously described [6]. Vibriocidal titers were defined as the reciprocal of the highest dilution of serum resulting in a 50% reduction in optical density (595 nm) as compared to control wells without serum. Seroconversion after vaccination was defined as a ≥4-fold increase from the baseline vibriocidal titer. OSP responses were measured using a previously described enzyme-linked immunosorbent assay [6, 7].
Statistical Analyses
Antibody titers were log2 transformed, and the normalized data were used for statistical analyses. Immunologic results were expressed as geometric mean titers and compared by a paired t test for within-group comparisons and by the Kruskal–Wallis analysis of variance and/or Student t test for between-group comparisons. A result was considered statistically significant if the 2-tailed P value was <.05.
RESULTS
Study Enrollment and Participation
Table 1 shows the demographic characteristics and immune responses of the 25 adult participants with HIV infection and the 25 adults without known HIV infection. Participants with HIV infection had a median CD4+ T-cell count of 433 cells/mm3 (interquartile range [IQR], 344–574 cells/mm3). Of the 25 participants with HIV infection, 23 were receiving antiretroviral therapy: 22 were receiving a dual-nucleoside reverse transcriptase inhibitor (NRTI) plus nonnucleoside reverse transcriptase inhibitor regimen, and 1 was receiving a dual NRTI and boosted protease inhibitor regimen. The 2 study participants not receiving antiretroviral therapy had CD4+ T-cell counts of >500 cells/mm3. Twenty-three participants received both doses of BivWC and completed the 3-week observation period. Two subjects received both doses of vaccine but withdrew prior to blood sample collection on day 21. There were no reported adverse events related to vaccination.
Table 1.
Demographic and Clinical Characteristics and Oral Cholera Vaccine Responses Among Adults Who Were or Were Not Known to Have Human Immunodeficiency Virus (HIV) Infection
| Characteristic | HIV Infection (n = 25) | No HIV Infection (n = 25) | P Valuesa |
|---|---|---|---|
| Age, y, mean ± SD | 37 ± 12 | 33 ± 11 | .23 |
| Female sex, no. (%) | 15 (60) | 20 (80) | .12 |
| Blood type, no. (%) | |||
| O | 14 (56) | 11 (44) | .40 |
| Non-O | 11 (44) | 14 (56) | .40 |
| CD4+ T-cell count, cells/μL, median (IQR) | 433 (344–574) | … | |
| HIV-associated immunodeficiency classificationb | |||
| Severe | 4 | … | |
| Advanced | 4 | … | |
| Mild | 8 | … | |
| Not significant | 9 | … | |
| Serum sample obtained, % | |||
| Day 0 | 25 (100) | 25 (100) | 1.00 |
| Day 7 | 25 (100) | 23 (92) | .49 |
| Day 21 | 23 (92) | 22 (88) | 1.00 |
| Vibriocidal response | |||
| Inaba serotype | |||
| GMT (95% CI) | |||
| Day 0 | 11 (6–17) | 11 (6–17) | 1.00 |
| Day 7 | 78 (35–171) | 186 (105–329) | .07 |
| Day 21 | 80 (42–154) | 219 (122–396) | .02 |
| Seroconversion, % (95% CI) | 74 (52–90) | 91 (71–99) | .24 |
| Ogawa serotype | |||
| GMT (95% CI) | |||
| Day 0 | 11 (7–20) | 14 (8–23) | .64 |
| Day 7 | 62 (31–125) | 142 (83–243) | .06 |
| Day 21 | 80 (44–147) | 181 (123–267) | .02 |
| Seroconversion, % (95% CI) | 65 (43–84) | 91 (71–99) | .07 |
| OSP IgA response | |||
| Inaba serotype | |||
| ELISA units (95% CI) | |||
| Day 0 | 72 (56–92) | 38 (28–52) | <.01 |
| Day 7 | 134 (87–207) | 81 (51–128) | .10 |
| Day 21 | 113 (73–174) | 79 (49–127) | .26 |
| Seroconversion, % | 30 | 50 | .18 |
| Ogawa serotype | |||
| ELISA units (95% CI) | |||
| Day 0 | 51 (37–70) | 29 (22–40) | .01 |
| Day 7 | 107 (66–174) | 79 (46–134) | .38 |
| Day 21 | 100 (66–150) | 72 (45–116) | .29 |
| Seroconversion, % | 57 | 59 | .86 |
Abbreviations: CI, confidence interval; ELISA, enzyme-linked immunosorbent assay; GMT, geometric mean titer; IgA, immunoglobulin A; IQR, interquartile range; OSP, O antigen–specific polysaccharide.
a Two-tailed P values of <.05 are considered statistically significant.
b Data indicate World Health Organization–based classification of the immunological status of individuals infected with HIV for >5 years. Severe denotes a CD4+ T-cell count of <200 cells/mm3, advanced denotes a count of 200–349 cells/mm3, mild denotes a count of 350–499 cells/mm3, and not significant denotes a count of >500 cells/mm3.
Immunologic Responses to BivWC in Adults With HIV Infection Versus Adults Without Known HIV Infection
Table 1 and Figure 1A and 1B show the immunologic responses in the adults with HIV infection, compared with adults without known HIV infection. Relative to adults without known HIV infection, HIV-infected adults tended to have lower vibriocidal antibody titers after 1 dose of BivWC and had significantly lower antibody titers after 2 doses of BivWC at day 21. Individuals with HIV infection had higher baseline OSP IgA titers than individuals without HIV infection, but there was no difference in the OSP IgA titers after vaccination. Seroconversion occurred at a rate of 65% (95% confidence interval [CI], 43%–84%) against the Ogawa serotype and 74% (95% CI, 52%–90%) against the Inaba serotype in adults with HIV infection, compared with a 91% seroconversion rate (95% CI, 71%–99%) for both serotypes observed in the cohort without HIV infection.
Figure 1.
Geometric mean titer (+standard error of the mean) of vibriocidal responses at day 0 and 7 days after each immunization (day 7 and day 21). A and B, Responses to Vibrio cholerae O1 Ogawa serotype (A) and Inaba serotype (B) in adults with and those without human immunodeficiency virus (HIV) infection. C and D, Responses to Ogawa serotype (C) and Inaba serotype (D), by baseline CD4+ T-cell count (bound by quartiles). *P < .05.
Comparison of Vibriocidal Responses, by CD4+ T-Cell Count
Figure 1C and 1D shows that there was significant variation in the vibriocidal responses in HIV-infected adults based on the CD4+ T-cell count. Those in the lowest quartile of CD4+ T-cell counts (<343 cells/mm3) had markedly diminished responses to the vaccine, while those with CD4+ T-cell counts higher than the median had similar responses as the non–HIV-infected adults.
DISCUSSION
In this study, we found that Haitian adults with HIV infection had diminished vibriocidal antibody responses to the BivWC oral cholera vaccine. These diminished vibriocidal immune responses were primarily seen in individuals in the lowest CD4+ T-cell count quartile. However, the overall seroconversion rates were 65% for the Ogawa serotype and 74% for the Inaba serotype; thus, the majority of HIV-infected individuals were still vaccine responders. In addition, the OSP IgA responses seen in HIV-infected individuals were similar to those in other healthy adults, suggesting a significant mucosal response to the main protective antigen of the vaccine.
Our findings on cholera vaccination are consistent with observations of immune responses to other vaccines in HIV-infected individuals. For example, systemic vaccines, such as pneumococcal vaccines, and mucosal vaccines, such as oral polio vaccine, produce lower responses in individuals with HIV infection [8, 9]. However, oral cholera vaccine is unlike other mucosal vaccines in that it is a killed whole-cell bacterial vaccine, and there is no intracellular infection, as there is in attenuated viral mucosal vaccines (for which CD4+ T-cell-mediated responses play a clear role). The fact that HIV infection diminishes responses to these different types of vaccines underscores the breadth of the immune deficits associated with HIV infection.
Limitations of our study are that the comparison cohort was not screened for HIV infection; however, the low HIV prevalence among adults aged 15–49 years of age in the region (2.3%) [10] and provider-initiated routine HIV testing at the study site likely limited the inclusion of HIV-infected individuals in the comparison group. Another limitation may be confounding from unidentified factors that differ between the HIV-infected and comparison cohort, such as nutritional status. An important consideration in evaluating these results is that, while serum vibriocidal antibodies are the best characterized marker of immunity to cholera and the historical immunologic benchmark for cholera vaccine evaluation, the vibriocidal antibodies are not likely to directly mediate protection from cholera, and there is no threshold titer at which 100% protection is achieved [11]. Therefore, this immunogenicity study does not definitively address the question of whether oral cholera vaccine efficacy in HIV-infected individuals differs from that in uninfected individuals. In addition, the median CD4+ T-cell count of HIV-infected subjects in our study was relatively high, which was likely the result of the subjects being engaged in HIV-associated care at the clinic study site. For this reason, we were unable to comment on the immunogenicity of the vaccine in advanced HIV infection (ie, CD4+ T-cell count, <50 cells/mm3). Last, our sample size was based on logistical feasibility and therefore was likely underpowered to detect the small differences in conversion rates of the magnitude that we observed.
In summary, while our study demonstrates decreased immune responses to oral cholera vaccine, predominantly in those with low CD4+ T-cell counts, there remains a substantial vaccine response rate in this population. In assessing the impact of this finding, it is important to consider that the decreased immunologic benefit to oral cholera vaccination may be offset by an increased risk of cholera in an HIV-infected population. Future studies of vaccination should estimate both the risk of cholera and the protective efficacy of cholera vaccines in HIV-infected individuals. Additional studies may also be useful to evaluate immunogenicity at very low CD4+ T-cell counts and to address whether administration of additional doses of vaccine to individuals with known HIV infection may offset a decreased immune response, as this strategy has proved beneficial for other vaccines [12].
Notes
Acknowledgments. We thank the study participants and the Zanmi Lasante staff in Haiti that supported the study.
Financial support. This work was supported by the National Institute of Allergy and Infectious Diseases, National Institutes of Health (NIH; R01 AI099243 to J. B. H. and L. C. I.; R01 AI106878 to E. T. R., P. K., and F. Q.; U01 AI AI058935 to E. T. R. and F. Q.; and R01 AI10355); an NIH Career Development Award (K08 AI089721 to R. C. C.); and Massachusetts General Hospital (Massachusetts General Hospital physician scientist development award to R. C. C.).
Potential conflicts of interest. All authors: No reported conflicts.
All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.
References
- 1.von Seidlein L, Wang XY, Macuamule A, et al. Is HIV infection associated with an increased risk for cholera? Findings from a case-control study in Mozambique. Trop Med Int Health 2008; 13:683–8. [DOI] [PubMed] [Google Scholar]
- 2.Valcin CL, Severe K, Riche CT, et al. Predictors of disease severity in patients admitted to a cholera treatment center in urban Haiti. Am J Trop Med Hyg 2013; 89:625–32. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Lucas ME, Deen JL, von Seidlein L, et al. Effectiveness of mass oral cholera vaccination in Beira, Mozambique. N Engl J Med 2005; 352:757–67. [DOI] [PubMed] [Google Scholar]
- 4.Bhattacharya SK, Sur D, Ali M, et al. 5 year efficacy of a bivalent killed whole-cell oral cholera vaccine in Kolkata, India: a cluster-randomised, double-blind, placebo-controlled trial. Lancet Infect Dis 2013; 13:1050–6. [DOI] [PubMed] [Google Scholar]
- 5.Perry RT, Plowe CV, Koumare B, et al. A single dose of live oral cholera vaccine CVD 103-HgR is safe and immunogenic in HIV-infected and HIV-noninfected adults in Mali. Bull World Health Organ 1998; 76:63–71. [PMC free article] [PubMed] [Google Scholar]
- 6.Charles RC, Hilaire IJ, Mayo-Smith LM, et al. Immunogenicity of a killed bivalent (O1 and O139) whole cell oral cholera vaccine, Shanchol, in Haiti. PLoS Negl Trop Dis 2014; 8:e2828. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Uddin T, Aktar A, Xu P, et al. Immune responses to O-specific polysaccharide and lipopolysaccharide of Vibrio cholerae O1 Ogawa in adult Bangladeshi recipients of an oral killed cholera vaccine and comparison to responses in patients with cholera. Am J Trop Med Hyg 2014; 90:873–81. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Troy SB, Musingwini G, Halpern MS, et al. Vaccine poliovirus shedding and immune response to oral polio vaccine in HIV-infected and -uninfected Zimbabwean infants. J Infect Dis 2013; 208:672–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Bliss SJ, O'Brien KL, Janoff EN, et al. The evidence for using conjugate vaccines to protect HIV-infected children against pneumococcal disease. Lancet Infect Dis 2008; 8:67–80. [DOI] [PubMed] [Google Scholar]
- 10.Cayemittes M, Fatuma Busangu M, de Dieu Bizimana J, Barrère B, Sévère B, Cayemittes et Emmanuel Charles V. Enquête Mortalité, Morbidité et Utilisation des Services, Haïti, 2012. Calverton, MD: MSPP, IHE et ICF International, 2013. [Google Scholar]
- 11.Saha D, LaRocque RC, Khan AI, et al. Incomplete correlation of serum vibriocidal antibody titer with protection from Vibrio cholerae infection in urban Bangladesh. J Infect Dis 2004; 189:2318–22. [DOI] [PubMed] [Google Scholar]
- 12.Cohen C, von Mollendorf C, de Gouveia L, et al. Effectiveness of seven-valent pneumococcal conjugate vaccine (PCV-7) against invasive pneumococcal disease in HIV-infected and -uninfected children in South Africa: a matched case-control study. Clin Infect Dis 2014; 59:808–18. [DOI] [PMC free article] [PubMed] [Google Scholar]