Immune Response to MVA-BN and Prior Smallpox Vaccination in People With HIV or at Risk for HIV Acquisition

Abstract

Little is known about the serologic response elicited by the Modified Vaccinia Ankara-Bavarian Nordic (MVA-BN) vaccination (JYNNEOS) against mpox in elderly individuals with or without HIV. In this study, we measured levels of antibody against orthopoxvirus in selected participants (n = 114) with HIV or at risk for HIV acquisition from the Multicenter AIDS Cohort Study/Women's Interagency HIV Study Combined Cohort Study, who had no prior reported history of mpox. Participants reported MVA-BN vaccination history via questionnaires. The median age was 64 years, 24.6% were female, and 46% were HIV seropositive. Twenty out of 114 participants received at least 1 dose of MVA-BN. We found that MVA-BN induced sustained IgG levels regardless of HIV status, even up to 1 year. Birth before 1973 was correlated with higher IgG. MVA-BN-unvaccinated individuals with HIV had lower IgG than those without HIV. Although limited by small sample size, our study is among the first to assess anti-orthopoxvirus antibodies specifically in a vulnerable, older population and stratified by HIV status.

Monkeypox virus (MPXV) is an orthopoxvirus that causes mpox, a clinical illness similar to smallpox. In 2022, a global outbreak of mpox was caused by Clade IIb MPXV and disproportionately affected men who reported having sex with men [1]. As of 31 December 2024, a total of 34 490 cases of mpox were reported to the U.S. Centers for Disease Control and Prevention (CDC) [2]. Even though most cases were mild and self-limited, severe or even fatal cases occurred in people with HIV (PWH) [3]. Routine childhood smallpox vaccination was terminated in 1972 after the disease was eradicated in the United States [4], and few civilians received smallpox vaccination until the 2022 mpox outbreak, when the Modified Vaccinia Ankara-Bavarian Nordic (MVA-BN) vaccine (JYNNEOS) was made available through the national stockpile. Although exact correlates of protection following MVA-BN are not well-characterized, there is evidence that MVA-BN protects against mpox [5]. Breakthrough infections after MVA-BN are rare, and in most cases, the symptoms are milder [6, 7].

Few studies have focused on long-term immunogenicity and effectiveness in PWH. Prior efforts found that in the short term (1 month after immunizations), MVA-BN is immunogenic, but HIV infection appears to negatively affect neutralizing antibody response [8]. However, existing data on long-term immune response to MVA-BN raise concerns, as MPXV-IgG levels after either 1 or 2 doses of MVA-BN were reported to decline almost to prevaccination levels 12 months after immunization [9]. In that study, however, most participants were young, living without HIV (only 9% were living with HIV), and otherwise had no previous smallpox vaccination exposure [9]. Moreover, little is known about the association between HIV, previous smallpox vaccination, and long-term vaccine immunogenicity in regard to aging populations. To this end, we leveraged the Multicenter AIDS Cohort Study (MACS)/Women's Interagency HIV Study (WIHS) Combined Cohort Study (MWCCS), a multicenter prospective cohort study of men and women with HIV or at risk for HIV acquisition [10], to evaluate serologic response to either historic smallpox vaccination or MVA-BN vaccination during the 2022 mpox outbreak.

METHODS

Participants

Participants enrolled in the MWCCS cohort who had self-reported mpox and smallpox vaccination information were included in this study. All participants signed informed consent, and the parent study was approved by each site's Institutional Review Board (IRB). Additional inclusion criteria were

1. Age ≥18 years and all sexes

2. Available HIV-related information

Serum samples were obtained at each MWCC scheduled visit and cryopreserved at the central repository at the Johns Hopkins University Data Analysis and Coordination Center. For participants vaccinated with MVA-BN, we analyzed anti-orthopoxvirus IgG levels from available serum samples from prevaccination, <6 months after the first or final dose, and over 6 months after the first or final dose. For participants not vaccinated with MVA-BN during this outbreak, available serum samples from pre-outbreak (before 2022), 6 months within the outbreak (1 June 2022 to 1 December 2022), and over 6 months after the outbreak started (after 1 December 2022) were used.

Anti-orthopoxvirus-Binding IgG Measurement

Anti-orthopoxvirus-binding IgG measurement was reported in our previous studies. Briefly, serum samples were tested for orthopoxvirus IgG at 1:100 dilution using MVA-BN as antigen. We used optical density (OD)–cutoff value (COV) (OD-COV) derived from ELISA to determine anti-orthopoxvirus IgG levels [11, 12]. Results below 0 were censored to 0 and considered seronegative.

Statistical Analysis

Categorical variables were summarized using number and percentage and compared using either χ² test or Fisher’s exact test. Continuous variables were summarized using median and interquartile range and compared using nonparametric methods. For 2 independent groups, we used the Wilcoxon rank sum test, and for paired samples (longitudinal measurements), we used the Wilcoxon signed rank test. P values from multiple comparisons were adjusted using the Benjamini–Hochberg method. For longitudinal repeated measurements, we used the generalized estimating equation (GEE) to estimate variables associated with IgG levels with autoregressive correlation structure (“geepack” package, version 1.3.12). The first dose of MVA-BN was set as day 0 (index date). In longitudinal GEE analysis, time points earlier than 12 months prior to the first MVA-BN vaccination (for vaccinated individuals) or earlier than 1 June 2021 (for unvaccinated individuals) were set to −12 months relative to the index date. Of note, vaccination dose was set as a time-varying variable in the GEE analyses. In the case that a participant received 2 doses of MVA-BN, and the second sampling time point fell between 2 doses (only 2 participants had this case scenario), the number of vaccinations was counted as 1 dose for the first 2 time points and as 2 doses for the third time point. All statistical analyses were performed using R (version 4.3.1).

RESULTS

Of the 114 MWCCS participants included in the study (median age of 64 years; Supplementary Table 1), 75% were male and 89% were born before 1973. Among them, 52 (46%) were HIV seropositive, with an HIV-1 viral suppression rate of 88.5% (HIV-1 viral load <50 copies/mL). Of all participants, 20 received MVA-BN, and among these participants, 15 (75%) received 2 doses. In the analysis including all participants, the MVA-BN-vaccinated group showed higher anti-orthopoxvirus IgG levels at time point 3 (6 months after the first dose of MVA-BN or 6 months beyond the 2022 outbreak, adjusted P < .0001, Figure 1A; Supplementary Table 2) relative to the group that had not received MVA-BN. Since most vaccinated participants received 2 doses of MVA-BN, and the second dose of MVA-BN may explain the persistent increase in the IgG levels at time point 3, we performed a sensitivity analysis including only those whose time point definition remained unchanged based on either the first or final dose of MVA-BN. In addition, only those with all 3 time points available were included (Figure 1B). MVA-BN vaccinations remained associated with higher IgG levels at time point 3 (adjusted P = .0003) and in this subgroup analysis, also at time point 2 (adjusted P = .004, Figure 1B). In a multivariate GEE model including all participants, 2 doses of MVA-BN and birth before 1973 were positively correlated with IgG levels, while HIV seropositivity was negatively correlated with IgG levels (Figure 1C). Longitudinally, the MVA-BN-vaccinated group showed a persistent increase in IgG levels in both whole group analysis (time point 3 vs 1, adjusted P = .013, Figure 1A) and sensitivity analysis (time point 3 vs 1, adjusted P = .004, Figure 1B). In MVA-BN vaccines, the number of vaccine doses received and HIV serostatus were not associated with IgG levels, while birth before 1973 (those who likely had childhood smallpox vaccination) was significantly associated with higher IgG levels (Supplementary Figure 1; Supplementary Table 3). In those who had not received MVA-BN, IgG levels were stably low, with a seropositivity rate at approximately 80% (Figure 1A). HIV seropositivity was negatively associated with IgG levels, while birth before 1973 was again significantly associated with higher IgG levels (Supplementary Figure 2; Supplementary Table 4). No participants reported ever having a positive mpox test result.

Figure 1.

Anti-orthopoxvirus IgG levels in MVA-BN-vaccinated group (MVA-BN) and those without MVA-BN-vaccinated group (none). A, IgG levels in the full cohort, with the first dose MVA-BN as day 0. Details on the time interval between each time point and vaccinations were shown in Supplementary Table 2. B, IgG levels in a subgroup analysis, with the final dose of MVA-BN as day 0. In this subgroup analysis, only those with samples from all 3 time points available, and whose time point definition remained the same using either the first or final dose of MVA-BN as day 0. In (A) and (B), time 1: before MVA-BN vaccination (vaccinated group) or before the 2022 outbreak (unvaccinated groups). Time 2: <6 m of MVA-BN vaccination (vaccinated group) or within 6 m of the 2022 outbreak (designated as 1 June 2022). Time 3: ≥6 m after MVA-BN vaccination (vaccinated group) or over 6 m since the 2022 outbreak. P values between groups were calculated using the Wilcoxon rank sum test and P values within groups were calculated using the Wilcoxon singed rank test. Final P values were adjusted for multiple comparisons using the Benjamini–Hochberg method. Nonsignificant P values were not shown. C, Longitudinal IgG levels in the full cohort, with the first dose of MVA-BN as day 0. Multivariate generalized estimating equation (GEE) results were attached in the right panel. Sex was also adjusted for but not listed (P = .5).

DISCUSSION

Our results showed that MVA-BN vaccination was associated with stable anti-orthopoxvirus-binding IgG levels in elderly participants living with HIV or at risk for HIV acquisition. Of note, among older participants who probably received childhood smallpox vaccination using a live, replicating vaccine (eg, Dryvax, given before 1973), there was a detectable level of humoral immunity against mpox in the majority (approximately 80%). Importantly, birth before 1973 was associated with higher levels of IgG in both MVA-BN-vaccinated and MVA-BN-unvaccinated groups, which is consistent with recent findings [11, 13]. We also confirmed that HIV seropositivity was associated with lower anti-orthopoxvirus levels [8, 14]; however, this negative association was mostly seen in the MVA-BN-unvaccinated group.

Our study was limited by a small sample size, especially in the MVA-BN vaccination group. We acknowledge that only binding IgG levels were measured in our study, and we were not able to measure memory B-cell [15] and T-cell levels [8], which would reflect memory adaptive immune response to mpox re-exposure. In addition, we did not measure the neutralizing antibody levels, which would have a better correlation with protection. Based on our previous studies using the same platform, binding IgG levels were positively correlated with neutralizing antibody levels [11]. Another caveat was that vaccination status was by self-report, which could introduce recall bias; in addition, the MVA-BN-vaccinated and MVA-BN-unvaccinated groups may differ in unmeasured ways, including health-seeking behaviors, access to care, or perceived risks. Finally, most of the MVA-BN vaccines were non-Hispanic, Caucasian males living with HIV or at risk for HIV acquisition and were not representative of the Black and Hispanic communities disproportionately affected by mpox during the 2022 outbreak in the United States [16]. Additional studies inclusive of these representative populations are needed to better understand pre-existing immunity and immunity following MVA-BN vaccination and subsequent impact on protection against mpox.

In summary, we demonstrate that MVA-BN provides comparable humoral immunogenicity regardless of HIV serostatus and significantly increases anti-orthopoxvirus IgG levels that are sustained for 6–12 months postvaccination. In MVA-BN-unvaccinated PWH who previously received smallpox vaccinations, antibody levels were lower than among people without HIV. These findings may guide vaccination recommendations for those living with HIV, and especially for those of older age who received smallpox vaccinations in childhood.

Supplementary Data

Supplementary materials are available at Open Forum Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author.