-
PDF
- Split View
-
Views
-
Cite
Cite
Charlotte Martin, Eric Florence, Cristina Domingo, Marc Delforge, Stéphane De Wit, Nicolas Dauby, Seroconversion and antibody persistence after yellow fever vaccination in people living with HIV: impact of baseline HIV viral load and yellow fever seropositivity, Journal of Travel Medicine, Volume 29, Issue 8, December 2022, taac024, https://doi.org/10.1093/jtm/taac024
- Share Icon Share
Abstract
Data on seroconversion rates following yellow fever (YF) vaccine and effect of revaccination in people living with HIV (PLWH) are scarce. We aimed at determining key factors for seroconversion after YF vaccine in PLWH and the role of preexisting neutralizing antibodies (NAbs) at vaccination.
A retrospective cross-sectional study at several timepoints in two Belgian AIDS Reference Center. For each individual, plasma samples from three timepoints were selected: Timepoint 0 (TP0) in the year before administration of the YF vaccine, Timepoint 1 (TP1) in the year following the YF vaccine, Timepoint 2 (TP2) >1 year after the YF vaccine. Plasma samples were analysed for YF NAbs by plaque reduction neutralization test. The primary endpoint was the number of patients with protective levels of NAbs ≥ 1/10. A boosted immune response was defined as a 4-fold increase in serologic titres following revaccination.
Of the 160 PLWH included, protective levels of NAbs were present in 36%, 87% and 72% of subjects at baseline, at a median of 12 months and a median of 96 months after YF vaccination, respectively. Among vaccine recipients negative for YF NAbs at baseline (n = 102), 83% seroconverted. PLWH with undetectable HIV viral load (VL) at baseline were more likely to seroconvert (P < 0·01). A booster response was observed in only 17% of subjects with baseline seropositivity (n = 10 out of 58). In multivariate analysis, undetectable HIV VL at vaccination and baseline YF seropositivity were associated with persistent levels of protective NAbs at a median of 8 years after YF vaccination.
Undetectable HIV VL at baseline is associated with high rates of seroconversion. YF seropositivity before revaccination is associated with low rates of booster effect but a higher chance of long term persistent NAbs response, suggesting a benefit of revaccination in PLWH.
Introduction
International travel has increased significantly over the last two decades. People living with HIV (PLWH) are no exception to the rule. They have a better quality of life and therefore travel more often than before the advent of combined antiretroviral therapy.1 A significant proportion of PLWH regularly visit friends and relatives (VFR) or return to their country of origin,2 which may be an endemic country for yellow fever (YF) (sub-Saharan Africa, Central or Latin America). Many YF endemic countries have experienced recent outbreaks or resurgence of cases.3 PLWH wishing to travel to these areas will, therefore, have to be vaccinated against YF, either on administrative or epidemiological grounds.
The live-attenuated YF-17D vaccine is strongly immunogenic in healthy recipients.4 One month after vaccination, primary YF immunization leads to the production of neutralizing antibodies (NAbs) in >98% of immunocompetent individuals.5
Different factors could impact the rate of seroconversion in immunocompetent vaccine recipients, among others the intensity of YF endemicity of the region,6,7 the co-administration with other vaccines (i.e. measles) and the age at YF vaccine administration.8,9 Although WHO lifted the requirement for YF booster every 10 years in 2013 for reasons of rational use of vaccine doses and chronic stock-outs, it is more and more admitted that NAbs levels wane with time.7,10,11 However, the relationship between the presence of circulating NAbs (and their level) and protection from YF disease remains unclear.4,11 Revaccination may protect against this NAbs decrease.7,12,13 As described for most live vaccines,14 response to YF revaccination in healthy adults is inversely related to the level of neutralizing antibodies present before revaccination.13
HIV infection has been correlated to both poorer humoral and cellular response to vaccination,15,16 in terms of seroconversion rates as well as of persistence of vaccine-induced antibodies. There is little data on seroconversion after YF vaccine in PLWH. We recently performed a systematic review of the literature with a meta-analysis on 253 patients showing that, in non-endemic countries, PLWH with undetectable HIV viral load at vaccination present a 97% seroconversion rate after YF vaccination.17 Data about revaccination in PLWH were however very limited as none of the studies included in our systematic review were initially designed to explore the question of the impact of revaccination on persistence of humoral immunity after YF vaccination.17
Here we aim to assess the seroconversion rate after YF vaccine in PLWH, to determine key factors for seroconversion and the role of preexisting NAbs at vaccination in their long-term persistence.
Materials and methods
Study design and participants
Two Belgian HIV Reference centers (Centre Hospitalier Universitaire Saint-Pierre, Brussels and Instituut voor Tropische Geneeskunde Antwerp) selected a group of patients with the following inclusion criteria: regular follow-up for HIV infection, history of documented YF timepoints described below.
The design of the study is shown on Figure 1. For each individual, plasma samples from three timepoints were collected: Timepoint 0 (TP0) in the year that preceded the YF vaccine, Timepoint 1 (TP1) in the year following the YF vaccine, Timepoint 2 (TP2) >1 year after the YF vaccine.

Study Design and number of samples at each timepoint of the study
The following elements were retrieved from hospital respective databases: demographics, HIV infection history (date of HIV diagnosis, HIV mode of transmission), immunological parameters across time (nadir CD4 + T cell count, CD4+ T cell count, CD4/CD8 ratio), HIV viral load (VL), antiretroviral therapy (ART) history.
Informed consent was provided by all participants and the institutional ethical committees of both CHU Saint-Pierre and ITG Antwerp approved the study (AK/14–09-79/4408 and IRB/AB/ac/109).
Procedures
To evaluate the presence of neutralizing antibodies in the serum samples, a PRNT90 assay was used.18 Briefly, sera were analysed in duplicate using 2-fold dilutions (from 1:5 to 1:320) and challenged against 25 PFU YF virus (strain 17D). The lysis plaques on a monolayer of PS infected cells were counted and the 90% plaque reduction neutralization titres (PRNT) was calculated. Serum from a healthy vaccine recipients with known titre and a negative serum were included as controls in all assay runs. A neutralization antibody titre of 1:10 or higher was considered a surrogate of protection and therefore we classified samples as seropositive when they were reactive at a titre of 1:10 or higher,13 and seronegative if they were ineffective at the initial dilution in the assay (1:5).
Seroconversion was defined as a change from seronegative to a NABs titre ≥1:10 in the PRNT. A boosted immune response was defined as a 4-fold increase in serologic titres.14 Undetectable HIV VL was defined as plasma HIV RNA below 200 copies/ml.
Statistics
We used descriptive statistics to summarize the characteristics of our population: median and interquartile ranges (IQRs) for continuous data; and frequencies and percentages for categorical data. Hypothesis tests for differences between groups were performed using non-parametric Wilcoxon–Mann–Whitney and Kruskal–Wallis tests for continuous variables, and Fisher exact tests for categorical variables. The analysis of the predictive factors of seroconversion and persistence of NAbs was done using a logistic regression model, given the presence of an antibody titre≥1:10 as threshold. The dependent variable was categorized as negative (<1:10) vs positive (≥1:10). For the analysis of the predictive factors of persistence of NABs, only PLWH with positive neutralizing antibody titres at TP1 and TP2 were selected. In a complementary analysis of seroconversion, we introduce an additional category for high neutralizing antibody titres defined as ≥40.19 The latter analysis was done with ordinal logistic regression. We used SAS statistical software (version 9·4 SAS institute, Cary NC, USA). All P-values were two-sided and considered significant if P < 0·05.
Role of the funding source
We thank the Association Vésale, International Society of Travel Medicine and Gilead for the research grants awarded for this project. These sources had no role in the design of the study, the collection, analysis and interpretation of data, in the writing of the report, and in the decision to submit the paper for publication.
Results
Plasma samples from 160 patients were available for TP0, 159 for TP1 and 158 for TP2 for a total of 477 samples, with 158 patients having all three timepoints samples available. At the time of vaccination, median age was 39 years (IQR 34–46), median CD4+ T cell count was 539/mm3 (IQR 405–717) and 69% (110/160) had undetectable HIV VL. Two patients (1%) had CD4 T-cell count below 200/mm3 at vaccination. Patients’ characteristics are summarized on Table 1. Median time between documented YF vaccine and TP1 and TP2 was 12 months (IQR 11–13) and 96 months (IQR 77–114), respectively. Overall, 36% (58/160) plasma samples were positive for NAbs before vaccination (TP0), 87% (139/159) at TP1 and 72% (114/158) at TP2. No patient reported a previous YF infection.
. | Overall (n = 160) . | Negative NAbs at TP0 (n = 102) . | Positive NAbs at TP0 (n = 58) . | P value . |
---|---|---|---|---|
Social demographics | ||||
Age, years | 39 (34–46) | 38 (33–43) | 44 (36–51) | 0.001 |
Women | 78 (49) | 52 (51) | 26 (45) | 0.51 |
Black race | 112 (70) | 70 (69) | 42 (73) | 0.72 |
HIV mode of transmission | ||||
Heterosexual | 115 (72) | 70 (69) | 45 (78) | <0.05 |
Homo/bisexual | 32 (21) | 26 (26) | 6 (10) | |
HIV and YF vaccine factors | ||||
Nadir CD4+ T cell <200/mm3 | 73 (46) | 45 (45) | 28 (48) | 0.7 |
CD4+ T cell count at YF vaccination (/mm3) | 539 (405–717) | 509 (424–689) | 554 (385–718) | 0.8 |
Delay between HIV diagnosis and YF vaccine, months | 75 (46–137) | 78 (47–120) | 75 (40–152) | 0.58 |
Undetectable HIV viral load at YF vaccination | 110 (69) | 65 (64) | 44 (76) | 0.16 |
Treatment duration at YF vaccination, years | 4.4 (0.9–7.5) | 4.2 (0.9–7.5) | 4.8 (1–7.6) | 0.7 |
Delay between YF vaccine and TP2, years | 8 (6.4–9.5) | 8.1 (6.4–9.4) | 7.9 (6.3–9.5) | 0.8 |
. | Overall (n = 160) . | Negative NAbs at TP0 (n = 102) . | Positive NAbs at TP0 (n = 58) . | P value . |
---|---|---|---|---|
Social demographics | ||||
Age, years | 39 (34–46) | 38 (33–43) | 44 (36–51) | 0.001 |
Women | 78 (49) | 52 (51) | 26 (45) | 0.51 |
Black race | 112 (70) | 70 (69) | 42 (73) | 0.72 |
HIV mode of transmission | ||||
Heterosexual | 115 (72) | 70 (69) | 45 (78) | <0.05 |
Homo/bisexual | 32 (21) | 26 (26) | 6 (10) | |
HIV and YF vaccine factors | ||||
Nadir CD4+ T cell <200/mm3 | 73 (46) | 45 (45) | 28 (48) | 0.7 |
CD4+ T cell count at YF vaccination (/mm3) | 539 (405–717) | 509 (424–689) | 554 (385–718) | 0.8 |
Delay between HIV diagnosis and YF vaccine, months | 75 (46–137) | 78 (47–120) | 75 (40–152) | 0.58 |
Undetectable HIV viral load at YF vaccination | 110 (69) | 65 (64) | 44 (76) | 0.16 |
Treatment duration at YF vaccination, years | 4.4 (0.9–7.5) | 4.2 (0.9–7.5) | 4.8 (1–7.6) | 0.7 |
Delay between YF vaccine and TP2, years | 8 (6.4–9.5) | 8.1 (6.4–9.4) | 7.9 (6.3–9.5) | 0.8 |
Data are median (IQR) or n (%) unless otherwise stated.
NAbs: Yellow Fever neutralizing antibodies, TP: timepoint.
. | Overall (n = 160) . | Negative NAbs at TP0 (n = 102) . | Positive NAbs at TP0 (n = 58) . | P value . |
---|---|---|---|---|
Social demographics | ||||
Age, years | 39 (34–46) | 38 (33–43) | 44 (36–51) | 0.001 |
Women | 78 (49) | 52 (51) | 26 (45) | 0.51 |
Black race | 112 (70) | 70 (69) | 42 (73) | 0.72 |
HIV mode of transmission | ||||
Heterosexual | 115 (72) | 70 (69) | 45 (78) | <0.05 |
Homo/bisexual | 32 (21) | 26 (26) | 6 (10) | |
HIV and YF vaccine factors | ||||
Nadir CD4+ T cell <200/mm3 | 73 (46) | 45 (45) | 28 (48) | 0.7 |
CD4+ T cell count at YF vaccination (/mm3) | 539 (405–717) | 509 (424–689) | 554 (385–718) | 0.8 |
Delay between HIV diagnosis and YF vaccine, months | 75 (46–137) | 78 (47–120) | 75 (40–152) | 0.58 |
Undetectable HIV viral load at YF vaccination | 110 (69) | 65 (64) | 44 (76) | 0.16 |
Treatment duration at YF vaccination, years | 4.4 (0.9–7.5) | 4.2 (0.9–7.5) | 4.8 (1–7.6) | 0.7 |
Delay between YF vaccine and TP2, years | 8 (6.4–9.5) | 8.1 (6.4–9.4) | 7.9 (6.3–9.5) | 0.8 |
. | Overall (n = 160) . | Negative NAbs at TP0 (n = 102) . | Positive NAbs at TP0 (n = 58) . | P value . |
---|---|---|---|---|
Social demographics | ||||
Age, years | 39 (34–46) | 38 (33–43) | 44 (36–51) | 0.001 |
Women | 78 (49) | 52 (51) | 26 (45) | 0.51 |
Black race | 112 (70) | 70 (69) | 42 (73) | 0.72 |
HIV mode of transmission | ||||
Heterosexual | 115 (72) | 70 (69) | 45 (78) | <0.05 |
Homo/bisexual | 32 (21) | 26 (26) | 6 (10) | |
HIV and YF vaccine factors | ||||
Nadir CD4+ T cell <200/mm3 | 73 (46) | 45 (45) | 28 (48) | 0.7 |
CD4+ T cell count at YF vaccination (/mm3) | 539 (405–717) | 509 (424–689) | 554 (385–718) | 0.8 |
Delay between HIV diagnosis and YF vaccine, months | 75 (46–137) | 78 (47–120) | 75 (40–152) | 0.58 |
Undetectable HIV viral load at YF vaccination | 110 (69) | 65 (64) | 44 (76) | 0.16 |
Treatment duration at YF vaccination, years | 4.4 (0.9–7.5) | 4.2 (0.9–7.5) | 4.8 (1–7.6) | 0.7 |
Delay between YF vaccine and TP2, years | 8 (6.4–9.5) | 8.1 (6.4–9.4) | 7.9 (6.3–9.5) | 0.8 |
Data are median (IQR) or n (%) unless otherwise stated.
NAbs: Yellow Fever neutralizing antibodies, TP: timepoint.
Impact of undetectable HIV viral load on seroconversion at TP1
Overall, seroconversion rate for vaccine recipients with negative NAb titres at baseline was 83% (84/101). Predictive factors for seroconversion are summarized in Table 2. Age and gender were not associated with seroconversion while subjects with undetectable HIV VL and/or under antiretroviral therapy at TP0 were more likely to seroconvert (P < 0.01) in univariate analysis. A multivariable logistic regression model confirmed that undetectable HIV VL at vaccination was the only factor independently and significantly associated with seroconversion (aOR 5.64, 95%CI 1.77–18.03). Undetectable HIV VL at vaccination and being heterosexually HIV infected were also significantly associated with higher (≥1/40) antibody levels after vaccination (P < 0.01 and P < 0.05, respectively) (multivariate analysis—data not shown).
Univariate and multivariate analysis of predictive factors of seroconversion after YF vaccine in baseline YF seronegative PLWH (n = 101)
Characteristic . | Efficient seroconversion at TP1 (N = 84) n/N (%) . | No seroconversion at TP1 (N = 17) n/N (%) . | P value . | OR (95% CI) . | aOR (95%CI) . |
---|---|---|---|---|---|
Age > 40 at vaccination | 32 (38) | 7 (41) | 0.79 | 1.14 (0.39–3.29) | |
Black ethnicity | 59 (70) | 10 (59) | 0.40 | 0.61 (0.21–1.77) | |
Homo/Bisexual HIV transmission | 19 (23) | 6 (35) | 0.36 | 0.54 (0.18–1.67) | |
Nadir CD4+ T cell count<200/mm3 | 37 (45) | 7 (41) | 1 | 0.87 (0.30–2.51) | |
ART > 5 y at TP0 | 33 (40) | 7 (41) | 1 | 0.96 (0.33–2.78) | |
HIV infection >5 y at TP0 | 50 (61) | 9 (53) | 0.59 | 0.72 (0.25–2.06) | |
CD4+ T cell count >350/mm3 at TP0 | 71 (86) | 15 (88) | 1.0 | 0.79 (0.16–3.9) | |
CD4/CD45 ≥ 25% at TP0 | 43 (52) | 7 (41) | 0.6 | 1.54 (0.53–4.42) | |
CD4/CD8 ratio ≥ 1 at TP0 | 12 (14) | 0 (0) | 0.21 | Infinity | |
HIV viral load <200 cop/ml at TP0 | 60 (71) | 5 (29) | <0.01 | 6 (1.91–18.87) | 5.64 (1.77–18.03) |
ART at TP0 | 71 (85) | 9 (53) | <0.01 | 4·85 (1.58–14.9) |
Characteristic . | Efficient seroconversion at TP1 (N = 84) n/N (%) . | No seroconversion at TP1 (N = 17) n/N (%) . | P value . | OR (95% CI) . | aOR (95%CI) . |
---|---|---|---|---|---|
Age > 40 at vaccination | 32 (38) | 7 (41) | 0.79 | 1.14 (0.39–3.29) | |
Black ethnicity | 59 (70) | 10 (59) | 0.40 | 0.61 (0.21–1.77) | |
Homo/Bisexual HIV transmission | 19 (23) | 6 (35) | 0.36 | 0.54 (0.18–1.67) | |
Nadir CD4+ T cell count<200/mm3 | 37 (45) | 7 (41) | 1 | 0.87 (0.30–2.51) | |
ART > 5 y at TP0 | 33 (40) | 7 (41) | 1 | 0.96 (0.33–2.78) | |
HIV infection >5 y at TP0 | 50 (61) | 9 (53) | 0.59 | 0.72 (0.25–2.06) | |
CD4+ T cell count >350/mm3 at TP0 | 71 (86) | 15 (88) | 1.0 | 0.79 (0.16–3.9) | |
CD4/CD45 ≥ 25% at TP0 | 43 (52) | 7 (41) | 0.6 | 1.54 (0.53–4.42) | |
CD4/CD8 ratio ≥ 1 at TP0 | 12 (14) | 0 (0) | 0.21 | Infinity | |
HIV viral load <200 cop/ml at TP0 | 60 (71) | 5 (29) | <0.01 | 6 (1.91–18.87) | 5.64 (1.77–18.03) |
ART at TP0 | 71 (85) | 9 (53) | <0.01 | 4·85 (1.58–14.9) |
P < 0·05 means statistically significant/ART: antiretroviral therapy/y: years.
Univariate and multivariate analysis of predictive factors of seroconversion after YF vaccine in baseline YF seronegative PLWH (n = 101)
Characteristic . | Efficient seroconversion at TP1 (N = 84) n/N (%) . | No seroconversion at TP1 (N = 17) n/N (%) . | P value . | OR (95% CI) . | aOR (95%CI) . |
---|---|---|---|---|---|
Age > 40 at vaccination | 32 (38) | 7 (41) | 0.79 | 1.14 (0.39–3.29) | |
Black ethnicity | 59 (70) | 10 (59) | 0.40 | 0.61 (0.21–1.77) | |
Homo/Bisexual HIV transmission | 19 (23) | 6 (35) | 0.36 | 0.54 (0.18–1.67) | |
Nadir CD4+ T cell count<200/mm3 | 37 (45) | 7 (41) | 1 | 0.87 (0.30–2.51) | |
ART > 5 y at TP0 | 33 (40) | 7 (41) | 1 | 0.96 (0.33–2.78) | |
HIV infection >5 y at TP0 | 50 (61) | 9 (53) | 0.59 | 0.72 (0.25–2.06) | |
CD4+ T cell count >350/mm3 at TP0 | 71 (86) | 15 (88) | 1.0 | 0.79 (0.16–3.9) | |
CD4/CD45 ≥ 25% at TP0 | 43 (52) | 7 (41) | 0.6 | 1.54 (0.53–4.42) | |
CD4/CD8 ratio ≥ 1 at TP0 | 12 (14) | 0 (0) | 0.21 | Infinity | |
HIV viral load <200 cop/ml at TP0 | 60 (71) | 5 (29) | <0.01 | 6 (1.91–18.87) | 5.64 (1.77–18.03) |
ART at TP0 | 71 (85) | 9 (53) | <0.01 | 4·85 (1.58–14.9) |
Characteristic . | Efficient seroconversion at TP1 (N = 84) n/N (%) . | No seroconversion at TP1 (N = 17) n/N (%) . | P value . | OR (95% CI) . | aOR (95%CI) . |
---|---|---|---|---|---|
Age > 40 at vaccination | 32 (38) | 7 (41) | 0.79 | 1.14 (0.39–3.29) | |
Black ethnicity | 59 (70) | 10 (59) | 0.40 | 0.61 (0.21–1.77) | |
Homo/Bisexual HIV transmission | 19 (23) | 6 (35) | 0.36 | 0.54 (0.18–1.67) | |
Nadir CD4+ T cell count<200/mm3 | 37 (45) | 7 (41) | 1 | 0.87 (0.30–2.51) | |
ART > 5 y at TP0 | 33 (40) | 7 (41) | 1 | 0.96 (0.33–2.78) | |
HIV infection >5 y at TP0 | 50 (61) | 9 (53) | 0.59 | 0.72 (0.25–2.06) | |
CD4+ T cell count >350/mm3 at TP0 | 71 (86) | 15 (88) | 1.0 | 0.79 (0.16–3.9) | |
CD4/CD45 ≥ 25% at TP0 | 43 (52) | 7 (41) | 0.6 | 1.54 (0.53–4.42) | |
CD4/CD8 ratio ≥ 1 at TP0 | 12 (14) | 0 (0) | 0.21 | Infinity | |
HIV viral load <200 cop/ml at TP0 | 60 (71) | 5 (29) | <0.01 | 6 (1.91–18.87) | 5.64 (1.77–18.03) |
ART at TP0 | 71 (85) | 9 (53) | <0.01 | 4·85 (1.58–14.9) |
P < 0·05 means statistically significant/ART: antiretroviral therapy/y: years.
Impact of vaccination in PLWH with positive YF NAbs at baseline
In vaccines with positive NAbs at baseline, we observed a booster response in 17% (10/58) of individuals at TP1 as defined by a 4-fold increase in NAb titres. Predictive factors of boosted immune response are summarized in Table 3. The only factor associated with a booster response in uni- and multivariate logistic regression was Nab titres <1/20 before immunization (P = 0.0114; aOR 6.75, 95%CI 1.54–29.62). VL undetectability had no impact on the development of a boosted immune response.
Predictive factors for persistence of NAbs at TP2
At TP2 (median 8 years after YF vaccination, IQR 6.3–9.5), 72% (114/158) vaccines showed protective titres of NAbs. Predictive factors for persistence of NAbs at TP2 are summarized on Table 4. In patients with positive Nab titres at baseline, 91% (53/58) were still positive at TP2 whereas in patients with negative Nab titres at baseline, only 61% (61/100) maintained a protective humoral immunity (P < 0.001). Subjects with undetectable VL at the time of immunization and longer duration of ART at the time of immunization were also more likely to maintain the presence of NAb titres at TP2 (P < 0.0001 and P < 0.001, respectively). A multivariable logistic regression model confirmed that factors significantly associated with increased likelihood of maintaining the presence of NAbs at TP2 included undetectable HIV VL (aOR 5.89, 95%CI 2.4–14.48) and positive NAb titres at immunization (aOR 6.06, 95%CI 1.81–20.25).
Discussion
In this study, we observed waning of the humoral response as YF NAb titres decreased over time, with 87% of PLWH carrying protective levels in the first year and 72% 8 years after YF vaccination. We confirmed that HIV VL undetectability is a critical factor for seroconversion after YF vaccine and for humoral response persistence with time in PLWH. Being YF seropositive before vaccination was also strongly associated with the long-term persistence of neutralizing antibodies against YF after vaccination, suggesting a benefit of revaccination in this population.
Independently of CD4 T-cell count, undetectable HIV VL at vaccination, as an indirect marker of decreased immune activation, has been pointed out as a crucial element to ensure seroconversion after immunization with different vaccines, including YF vaccine.15,16,20–24 Provided that the HIV VL is controlled, seroconversion rate after YF vaccination has been described to be close to 97% in non-endemic areas.17 Accordingly, in our study, undetectable HIV VL at vaccination was the onlyfactor statistically associated with seroconversion in the multivariate analysis (P < 0.01; aOR 5.64 95%CI 1.77–18.03). This raises the question of the systematic revaccination of a person who was initially vaccinated while his HIV viral load was not controlled.
Univariate and multivariate analysis of predictive factors of boosted immune response after YF vaccine in baseline YF seropositive PLWH (n = 58)
Characteristic . | Efficient boosted response at TP1 (N = 10) n/N (%) . | No boosted response at TP1 (N = 48) n/N (%) . | P value . | OR (95% CI) . | aOR (95%CI) . |
---|---|---|---|---|---|
Age > 40 at vaccination | 6 (60) | 31 (65) | 1 | 1.22 (0.3–4.9) | |
Black ethnicity | 10 (100) | 32 (67) | 0.06 | 0 (0–0.8) | |
Homo/bisexual HIV transmission | 0 (0) | 6 (14) | 0.58 | 0 (0–2.86) | |
Baseline NAbs < 1/20 | 6 (60) | 10 (21) | 0.02 | 5.7 (1.34–24.16) | 6.75 (1.54–29.62) |
Nadir CD4+ T cell count < 200/mm3 | 5 (50) | 23 (48) | 1 | 0.92 (0.19–4.58) | |
ART > 5 y at TP0 | 4 (40) | 24 (50) | 0.73 | 0.67 (0.17–2.67) | |
HIV infection >5 y at TP0 | 6 (60) | 31 (65) | 1.0 | 1.22 (0.3–4.91) | |
CD4+ T cell count >350/mm3 at TP0 | 8 (80) | 41 (85) | 0.65 | 0.68 (0.12–3.9) | |
CD4/CD45 ≥ 25% at TP0 | 6 (60) | 28 (58) | 1 | 1.07 (0.27–4.3) | |
CD4/CD8 ratio ≥ 1 at TP0 | 2 (20) | 10 (21) | 1 | 0.95 (0.17–5.19) | |
HIV viral load <200 cop/ml at TP0 | 9 (90) | 35 (73) | 0.42 | 3.34 (0.38–29.04) | |
ART at TP0 | 9 (90) | 38 (79) | 0.67 | 2.37 (0.27–20.96) |
Characteristic . | Efficient boosted response at TP1 (N = 10) n/N (%) . | No boosted response at TP1 (N = 48) n/N (%) . | P value . | OR (95% CI) . | aOR (95%CI) . |
---|---|---|---|---|---|
Age > 40 at vaccination | 6 (60) | 31 (65) | 1 | 1.22 (0.3–4.9) | |
Black ethnicity | 10 (100) | 32 (67) | 0.06 | 0 (0–0.8) | |
Homo/bisexual HIV transmission | 0 (0) | 6 (14) | 0.58 | 0 (0–2.86) | |
Baseline NAbs < 1/20 | 6 (60) | 10 (21) | 0.02 | 5.7 (1.34–24.16) | 6.75 (1.54–29.62) |
Nadir CD4+ T cell count < 200/mm3 | 5 (50) | 23 (48) | 1 | 0.92 (0.19–4.58) | |
ART > 5 y at TP0 | 4 (40) | 24 (50) | 0.73 | 0.67 (0.17–2.67) | |
HIV infection >5 y at TP0 | 6 (60) | 31 (65) | 1.0 | 1.22 (0.3–4.91) | |
CD4+ T cell count >350/mm3 at TP0 | 8 (80) | 41 (85) | 0.65 | 0.68 (0.12–3.9) | |
CD4/CD45 ≥ 25% at TP0 | 6 (60) | 28 (58) | 1 | 1.07 (0.27–4.3) | |
CD4/CD8 ratio ≥ 1 at TP0 | 2 (20) | 10 (21) | 1 | 0.95 (0.17–5.19) | |
HIV viral load <200 cop/ml at TP0 | 9 (90) | 35 (73) | 0.42 | 3.34 (0.38–29.04) | |
ART at TP0 | 9 (90) | 38 (79) | 0.67 | 2.37 (0.27–20.96) |
P < 0.05 means statistically significant/ART: antiretroviral therapy/y: years.
Univariate and multivariate analysis of predictive factors of boosted immune response after YF vaccine in baseline YF seropositive PLWH (n = 58)
Characteristic . | Efficient boosted response at TP1 (N = 10) n/N (%) . | No boosted response at TP1 (N = 48) n/N (%) . | P value . | OR (95% CI) . | aOR (95%CI) . |
---|---|---|---|---|---|
Age > 40 at vaccination | 6 (60) | 31 (65) | 1 | 1.22 (0.3–4.9) | |
Black ethnicity | 10 (100) | 32 (67) | 0.06 | 0 (0–0.8) | |
Homo/bisexual HIV transmission | 0 (0) | 6 (14) | 0.58 | 0 (0–2.86) | |
Baseline NAbs < 1/20 | 6 (60) | 10 (21) | 0.02 | 5.7 (1.34–24.16) | 6.75 (1.54–29.62) |
Nadir CD4+ T cell count < 200/mm3 | 5 (50) | 23 (48) | 1 | 0.92 (0.19–4.58) | |
ART > 5 y at TP0 | 4 (40) | 24 (50) | 0.73 | 0.67 (0.17–2.67) | |
HIV infection >5 y at TP0 | 6 (60) | 31 (65) | 1.0 | 1.22 (0.3–4.91) | |
CD4+ T cell count >350/mm3 at TP0 | 8 (80) | 41 (85) | 0.65 | 0.68 (0.12–3.9) | |
CD4/CD45 ≥ 25% at TP0 | 6 (60) | 28 (58) | 1 | 1.07 (0.27–4.3) | |
CD4/CD8 ratio ≥ 1 at TP0 | 2 (20) | 10 (21) | 1 | 0.95 (0.17–5.19) | |
HIV viral load <200 cop/ml at TP0 | 9 (90) | 35 (73) | 0.42 | 3.34 (0.38–29.04) | |
ART at TP0 | 9 (90) | 38 (79) | 0.67 | 2.37 (0.27–20.96) |
Characteristic . | Efficient boosted response at TP1 (N = 10) n/N (%) . | No boosted response at TP1 (N = 48) n/N (%) . | P value . | OR (95% CI) . | aOR (95%CI) . |
---|---|---|---|---|---|
Age > 40 at vaccination | 6 (60) | 31 (65) | 1 | 1.22 (0.3–4.9) | |
Black ethnicity | 10 (100) | 32 (67) | 0.06 | 0 (0–0.8) | |
Homo/bisexual HIV transmission | 0 (0) | 6 (14) | 0.58 | 0 (0–2.86) | |
Baseline NAbs < 1/20 | 6 (60) | 10 (21) | 0.02 | 5.7 (1.34–24.16) | 6.75 (1.54–29.62) |
Nadir CD4+ T cell count < 200/mm3 | 5 (50) | 23 (48) | 1 | 0.92 (0.19–4.58) | |
ART > 5 y at TP0 | 4 (40) | 24 (50) | 0.73 | 0.67 (0.17–2.67) | |
HIV infection >5 y at TP0 | 6 (60) | 31 (65) | 1.0 | 1.22 (0.3–4.91) | |
CD4+ T cell count >350/mm3 at TP0 | 8 (80) | 41 (85) | 0.65 | 0.68 (0.12–3.9) | |
CD4/CD45 ≥ 25% at TP0 | 6 (60) | 28 (58) | 1 | 1.07 (0.27–4.3) | |
CD4/CD8 ratio ≥ 1 at TP0 | 2 (20) | 10 (21) | 1 | 0.95 (0.17–5.19) | |
HIV viral load <200 cop/ml at TP0 | 9 (90) | 35 (73) | 0.42 | 3.34 (0.38–29.04) | |
ART at TP0 | 9 (90) | 38 (79) | 0.67 | 2.37 (0.27–20.96) |
P < 0.05 means statistically significant/ART: antiretroviral therapy/y: years.
Univariate and multivariate analysis of predictive factors for maintaining protective levels of NAbs at Timepoint 2 (TP2) (median 8 years after vaccination) (n = 137)
Characteristic . | Protective levels of NAbs at TP2 (N = 107) n/N (%) . | NAbs below protective levels at TP2 (N = 30) n/N (%) . | P value . | OR (CI 95%) . | aOR (CI 95%) . |
---|---|---|---|---|---|
Age > 40 at vaccination | 51 (48) | 14 (47) | 1 | 0.96 (0.43–2.16) | |
Black ethnicity | 78 (73) | 19 (63) | 0.36 | 0.64 (0.27–1.51) | |
Homo/bisexual HIV transmission | 15 (15) | 9 (31) | 0.06 | 0.38 (0.15–0.99) | |
Baseline NAbs>1/10 | 50 (47) | 5 (17) | 0.003 | 4.39 (1.56–12.32) | 6.06 (1.81–20.25) |
Nadir CD4+ T cell count<200/mm3 | 53 (50) | 9 (30) | 0.06 | 0.43 (0.18–1.02) | |
ART > 5 y at TP0 | 49 (46) | 8 (28) | 0.09 | 2.26 (0.92–5.55) | |
HIV infection >5 y at TP0 | 67 (63) | 15 (52) | 0.29 | 0.62 (0.27–1.43) | |
CD4+ T cell count >350/mm3 at TP0 | 90 (85) | 26 (87) | 1 | 0.87 (0.27–2.81) | |
CD4/CD45 ≥ 25% at TP0 | 62 (58) | 12 (40) | 0.1 | 2.11 (0.93–4.23) | |
CD4/CD8 ratio ≥ 1 at TP0 | 22 (21) | 1 (3) | 0.03 | 7.51 (0.97–58.18) | |
HIV viral load <200 cop/ml at TP0 | 88 (82) | 12 (40) | <0.0001 | 6.95 (2.87–16.8) | 5.89 (2.4–14.48) |
ART at TP0 | 95 (89) | 19 (63) | 0.02 | 4.58 (1.76–11.91) | |
Delay between vaccination and TP2 (y), mean (SD) | 8 (2) | 8.9 (2.4) | 0.06 |
Characteristic . | Protective levels of NAbs at TP2 (N = 107) n/N (%) . | NAbs below protective levels at TP2 (N = 30) n/N (%) . | P value . | OR (CI 95%) . | aOR (CI 95%) . |
---|---|---|---|---|---|
Age > 40 at vaccination | 51 (48) | 14 (47) | 1 | 0.96 (0.43–2.16) | |
Black ethnicity | 78 (73) | 19 (63) | 0.36 | 0.64 (0.27–1.51) | |
Homo/bisexual HIV transmission | 15 (15) | 9 (31) | 0.06 | 0.38 (0.15–0.99) | |
Baseline NAbs>1/10 | 50 (47) | 5 (17) | 0.003 | 4.39 (1.56–12.32) | 6.06 (1.81–20.25) |
Nadir CD4+ T cell count<200/mm3 | 53 (50) | 9 (30) | 0.06 | 0.43 (0.18–1.02) | |
ART > 5 y at TP0 | 49 (46) | 8 (28) | 0.09 | 2.26 (0.92–5.55) | |
HIV infection >5 y at TP0 | 67 (63) | 15 (52) | 0.29 | 0.62 (0.27–1.43) | |
CD4+ T cell count >350/mm3 at TP0 | 90 (85) | 26 (87) | 1 | 0.87 (0.27–2.81) | |
CD4/CD45 ≥ 25% at TP0 | 62 (58) | 12 (40) | 0.1 | 2.11 (0.93–4.23) | |
CD4/CD8 ratio ≥ 1 at TP0 | 22 (21) | 1 (3) | 0.03 | 7.51 (0.97–58.18) | |
HIV viral load <200 cop/ml at TP0 | 88 (82) | 12 (40) | <0.0001 | 6.95 (2.87–16.8) | 5.89 (2.4–14.48) |
ART at TP0 | 95 (89) | 19 (63) | 0.02 | 4.58 (1.76–11.91) | |
Delay between vaccination and TP2 (y), mean (SD) | 8 (2) | 8.9 (2.4) | 0.06 |
P < 0.05 means statistically significant/ART: antiretroviral therapy/y: years.
For the analysis of the predictive factors of persistence of NAbs, only PLWH with positive NAbs at TP1 AND TP2 were selected.
Univariate and multivariate analysis of predictive factors for maintaining protective levels of NAbs at Timepoint 2 (TP2) (median 8 years after vaccination) (n = 137)
Characteristic . | Protective levels of NAbs at TP2 (N = 107) n/N (%) . | NAbs below protective levels at TP2 (N = 30) n/N (%) . | P value . | OR (CI 95%) . | aOR (CI 95%) . |
---|---|---|---|---|---|
Age > 40 at vaccination | 51 (48) | 14 (47) | 1 | 0.96 (0.43–2.16) | |
Black ethnicity | 78 (73) | 19 (63) | 0.36 | 0.64 (0.27–1.51) | |
Homo/bisexual HIV transmission | 15 (15) | 9 (31) | 0.06 | 0.38 (0.15–0.99) | |
Baseline NAbs>1/10 | 50 (47) | 5 (17) | 0.003 | 4.39 (1.56–12.32) | 6.06 (1.81–20.25) |
Nadir CD4+ T cell count<200/mm3 | 53 (50) | 9 (30) | 0.06 | 0.43 (0.18–1.02) | |
ART > 5 y at TP0 | 49 (46) | 8 (28) | 0.09 | 2.26 (0.92–5.55) | |
HIV infection >5 y at TP0 | 67 (63) | 15 (52) | 0.29 | 0.62 (0.27–1.43) | |
CD4+ T cell count >350/mm3 at TP0 | 90 (85) | 26 (87) | 1 | 0.87 (0.27–2.81) | |
CD4/CD45 ≥ 25% at TP0 | 62 (58) | 12 (40) | 0.1 | 2.11 (0.93–4.23) | |
CD4/CD8 ratio ≥ 1 at TP0 | 22 (21) | 1 (3) | 0.03 | 7.51 (0.97–58.18) | |
HIV viral load <200 cop/ml at TP0 | 88 (82) | 12 (40) | <0.0001 | 6.95 (2.87–16.8) | 5.89 (2.4–14.48) |
ART at TP0 | 95 (89) | 19 (63) | 0.02 | 4.58 (1.76–11.91) | |
Delay between vaccination and TP2 (y), mean (SD) | 8 (2) | 8.9 (2.4) | 0.06 |
Characteristic . | Protective levels of NAbs at TP2 (N = 107) n/N (%) . | NAbs below protective levels at TP2 (N = 30) n/N (%) . | P value . | OR (CI 95%) . | aOR (CI 95%) . |
---|---|---|---|---|---|
Age > 40 at vaccination | 51 (48) | 14 (47) | 1 | 0.96 (0.43–2.16) | |
Black ethnicity | 78 (73) | 19 (63) | 0.36 | 0.64 (0.27–1.51) | |
Homo/bisexual HIV transmission | 15 (15) | 9 (31) | 0.06 | 0.38 (0.15–0.99) | |
Baseline NAbs>1/10 | 50 (47) | 5 (17) | 0.003 | 4.39 (1.56–12.32) | 6.06 (1.81–20.25) |
Nadir CD4+ T cell count<200/mm3 | 53 (50) | 9 (30) | 0.06 | 0.43 (0.18–1.02) | |
ART > 5 y at TP0 | 49 (46) | 8 (28) | 0.09 | 2.26 (0.92–5.55) | |
HIV infection >5 y at TP0 | 67 (63) | 15 (52) | 0.29 | 0.62 (0.27–1.43) | |
CD4+ T cell count >350/mm3 at TP0 | 90 (85) | 26 (87) | 1 | 0.87 (0.27–2.81) | |
CD4/CD45 ≥ 25% at TP0 | 62 (58) | 12 (40) | 0.1 | 2.11 (0.93–4.23) | |
CD4/CD8 ratio ≥ 1 at TP0 | 22 (21) | 1 (3) | 0.03 | 7.51 (0.97–58.18) | |
HIV viral load <200 cop/ml at TP0 | 88 (82) | 12 (40) | <0.0001 | 6.95 (2.87–16.8) | 5.89 (2.4–14.48) |
ART at TP0 | 95 (89) | 19 (63) | 0.02 | 4.58 (1.76–11.91) | |
Delay between vaccination and TP2 (y), mean (SD) | 8 (2) | 8.9 (2.4) | 0.06 |
P < 0.05 means statistically significant/ART: antiretroviral therapy/y: years.
For the analysis of the predictive factors of persistence of NAbs, only PLWH with positive NAbs at TP1 AND TP2 were selected.
In immunocompetent vaccine recipients who have lost their neutralizing antibodies after vaccination and in absence of revaccination, it is still unclear whether there is a NAbs ‘boostability’ (or anamnestic immune response) that is effective and rapid enough to be protective in case of exposure to wild YF.11 In healthy live attenuated vaccine recipients, an inverse correlation between the level of Nab titres prior to vaccination and the increase in NAbs after revaccination has been demonstrated in multiples studies,10,12,14,18,25,26 suggesting that if primary vaccination was effective, revaccination will not be of benefit at least in the following weeks.10,11 The underlying hypothesis is that the live attenuated virus administrated at revaccination is dampened by circulating NAbs, possibly also by specific cellular memory response, preventing a boost to the immune response.4,13,14,17 In contrast, the efficacy of YF revaccination is well-documented when the pre-booster level of NAbs is low or negative.10,12 We observed an infrequent booster response (17%), as described in healthy vaccinees.19 Low pre-existing NAbs titre was the only factor independently associated with a booster response to revaccination, which, to our knowledge, has not previously been reported in PLWH.17 Importantly, our data highlight the ability of PLWH to respond immunologically to revaccination.
Revaccination can also be used to extend vaccine protection. Live attenuated vaccines usually provide sustained antibody response in healthy recipients. However, data in healthy YF vaccine recipients show a slow decline in NAbs and of the levels of effector memory CD4+ and CD8+ T cells >10 years after primary vaccination. Revaccination ensures long-term persistence of circulating NAbs and cell-mediated memory.7,12,27 Overall, almost three-quarters (72%) of our cohort of PLWH were still carriers of protective titres of Nabs a median of 8 years after YF vaccination. As a comparison, Lindsey et al. reported that 94% and 82% of a cohort of travellers were still carriers of Nab titres within 10 years and >10 years after vaccination, respectively.12
Long-term serological memory is dependent on continuous antibody production from long-lived plasma cells residing in bone marrow. Conditions such as very young age or immunosenescence are known to limit the induction and persistence of long-lived plasma cells, leading to the need for an additional booster of vaccines such as Measles Mumps Rubella (MMR) vaccine (and possibly of YF vaccine) in persons vaccinated during infancy.10,14,20 HIV infection alters the environment needed for maintenance of bone marrow long-lived plasma cells and infects T follicular helper (Tfh) cells that play a central role in production of long-lasting NAbs which may also lead to the need for an additional antigenic booster compared to the general population.4,28,29 In our study, a highly protective factor for maintaining protective titres of NAbs in uni- and multivariate analysis was the presence of a protective humoral immunity before revaccination (either through unreported and undocumented vaccination, previous YF infection or cross-immunity to other flaviviruses) suggesting that revaccination was beneficial in maintaining humoral protection against YF for several years. The consequence could be a schedule of at least two YF vaccinations for PLWH (within an interval of 5–10 years), even in PLWH with an undetectable HIV VL at vaccination.
Undetectable HIV VL at vaccination was also a predictive factor for persistence of humoral protection. The importance of a controlled HIV VL in maintaining long-term vaccine immunity has been highlighted many times, including for the YF vaccine.22,24,30,31 Veit et al. reported greater persistence of neutralizing antibodies in PLWH whose VL had been undetectable at the time of vaccination at the two timepoints studied (5 years: 86% vs 99% and 10 years: 75% vs 100%, respectively).24
Our study has several limitations, including usual limitations of a retrospective study. First, we had no data about the precise number of YF vaccines received before the study. Secondly, the median delay of 1 year from vaccination for the TP1 analysis does not allow a full picture of the short-term immune response to vaccination. This may have led to an underestimation of the rate of seroconversion or boosted immune response. Thirdly, the median delay after documented YF vaccine was 8 years. It is therefore possible that we have underestimated the loss of protection that occurs >10 years after vaccination. Longitudinal follow-up of larger cohorts of PLWH are needed to answer these questions.
In conclusion, our study demonstrates that seroconversion rate after YF vaccine is excellent in case of controlled HIV VL and that at least one revaccination could be necessary to ensure long-term presence of circulating Nabs. Our findings in this longitudinal study allow to add to the data needed for the development of a specific vaccination schedule for PLWH.
Authors’ contribution
CM drafted the initial manuscript and did the initial literature search and conceptualization. EF provided data from his HIV Reference Center, reviewed, and edited the manuscript and provided substantial comments. CDC did the laboratory testing. MD performed the majority of the statistical analyses. ND conceptualized, reviewed and edited the manuscript and provided substantial comments. CM, CDC and ND made the decision to submit for publication.
Acknowledgments
BREACH is the Belgium Research on AIDS and HIV Consortium. It is the result of a collaborative effort of the Belgian HIV Reference Centers (HRC) and Laboratories (HRL), scientific research groups and interest organizations in the field of HIV infection and AIDS, aiming to maximize synergies between its members. It has enabled collaboration between HIV reference centers and the sharing of data on, among others, patients vaccinated against YF.
We thank the Biostatistics platform of ULB for their precious advice in the methodology to be applied in our analyses.
N.D. is a post doctorate clinical master specialist of the F.R.S-FNRS.
We thank Dr Mariana Andrade who provided editorial assistance. No commercial funding was received for this purpose.
Funding
International Society of Travel Medicine (ISTM), Association Vésale, Gilead.
Conflict of interest: None declared.