Short-Lived Immunity After 17DD Yellow Fever Single Dose Indicates That Booster Vaccination May Be Required to Guarantee Protective Immunity in Children

Abstract

The Yellow Fever (YF) vaccination is recommended for people living in endemic areas and represents the most effective strategy to reduce the risk of infection. Previous studies have warned that booster regimens should be considered to guarantee the long-term persistence of 17DD-YF-specific memory components in adults living in areas with YF-virus circulation. Considering the lower seroconversion rates observed in children (9-12 months of age) as compared to adults, this study was designed in order to access the duration of immunity in single-dose vaccinated children in a 10-years cross-sectional time-span. The levels of neutralizing antibodies (PRNT) and the phenotypic/functional memory status of T and B-cells were measured at a baseline, 30-45 days, 1, 2, 4, 7, and 10 years following primary vaccination. The results revealed that a single dose induced 85% of seropositivity at 30-45 days and a progressive time-dependent decrease was observed as early as 2 years and declines toward critical values (below 60%) at time-spans of ≥4-years. Moreover, short-lived YF-specific cellular immunity, mediated by memory T and B-cells was also observed after 4-years. Predicted probability and resultant memory analysis emphasize that correlates of protection (PRNT; effector memory CD8+ T-cells; non-classical memory B-cells) wane to critical values within ≥4-years after primary vaccination. Together, these results clearly demonstrate the decline of 17DD-YF-specific memory response along time in children primarily vaccinated at 9-12 months of age and support the need of booster regimen to guarantee the long-term persistence of memory components for children living in areas with high risk of YF transmission.

Keywords: 17DD vaccine; cellular memory; children; neutralizing antibodies; yellow fever.

Copyright © 2019 Campi-Azevedo, Reis, Peruhype-Magalhães, Coelho-dos-Reis, Antonelli, Fonseca, Costa-Pereira, Souza-Fagundes, Costa-Rocha, Mambrini, Lemos, Ribeiro, Caldas, Camacho, Maia, de Noronha, de Lima, Simões, Freire, Martins, Homma, Tauil, Vasconcelos, Romano, Domingues, Teixeira-Carvalho and Martins-Filho.

Figures

Figure 1

Neutralizing antibody titers after 17DD-YF primary vaccination in children. PRNT titers were measured in Ecteola-treated plasma samples (19) from non-vaccinated children at baseline NV(day 0)/(○, n = 47) and at different times after primary vaccination: PV(day 30–45)/(, n = 47), PV(year 1)/(, n = 141), PV(year 2)/(, n = 114), PV(year 4)/(, n = 128), PV(year 7)/(, n = 116), and PV(year 10)/(°, n = 127), as described previously by Simões et al. (20). (A) The PRNT levels were expressed in reverse of serum dilution. (B) Proportion of PRNT seropositivity (PRNT>1:10) were calculated for each group and the results expressed as seropositivity rates at baseline NV(day 0) [] and at different times after primary vaccination: PV(day 30–45) [], PV(year 1) [], PV(year 2) [], PV(year 4) [], PV(year 7) [], and PV(year 10) [], considering the serum dilution >1:10 as the cut-off (dashed line). (C) Correlation and logistic regression were employed to determine the wane of PRNT levels along time continuum and the results expressed as reverse of serum dilution and predicated probability, respectively. Statistical analysis was carried out as described in Methods. In all cases, significant differences at p < 0.05 were underscored by using letters “a,” “b,” “c,” and “d” for comparisons with NV(day 0), PV(day 30–45), PV(year 1), and PV(year 2), respectively and the p-values provide in the figure. Spearman correlation indices as well as Likelihood and Odds ratio are provided in the figure. Gray rectangle highlights the critical decrease of PRNT seropositivity rates ≥4 years after primary vaccination.

Figure 2

Phenotypic and functional memory biomarkers after 17DD-YF primary vaccination in children. The analysis of 17DD-YF-specific memory was measured upon in vitro 17DD-YF antigen recall as described previously by Campi-Azevedo et al. (8) for non-vaccinated children at baseline NV(day 0)/(, n = 47) and at different times after primary vaccination: PV(day 30–45)/(, n = 47), PV(year 1)/(, n = 141), PV(year 2)/(, n = 114), PV(year 4)/(, n = 128), PV(year 7)/(, n = 116), and PV(year 10)/(, n = 127). (A) Flow cytometric staining were used to quantify phenotypic features of T-cell memory subsets: Naïve T-cells/(NCD4;NCD8)/CD27+CD45RO−; early Effector Memory T-cells/(eEfCD4;eEfCD8)/CD27−CD45RO− Central Memory T-cells/(CMCD4;CMCD8)/CD27+CD45RO+ Effector Memory T-cells/(EMCD4;EMCD8)/CD27−CD45RO+ and B-cell memory subsets: Naïve B-cells/(NCD19)/CD27−IgD+; Non-classical Memory B-cells/(nCMCD19)/CD27+IgD+ and Classical Memory B-cells/(CMCD19)/CD27+IgD−. (B) Flow cytometric staining were also performed to quantity functional CD8+ T-cells producing TNF-α, IFN-γ and IL-5. The data were reported as median values ± inter-quartile range for 17DD-YF-stimulated Culture/non-stimulated Control Culture Index as described in Methods, highlighting the equivalence ratio by dashed line (Index = 1.0). Significant differences at p < 0.05 were underscored by using asterisk (*) to identify differences between NV(day 0) vs. PV(day 30–45) and intergroup differences identified by connecting lines.

Figure 3

Phenotypic and functional biomarker signatures after 17DD-YF primary vaccination in children. (A) Biomarker signatures of reference groups NV(day 0) [] and PV(day 30–45) [] were assembled to select biomarkers above the 75th percentile with proportions higher than the 25% in each group (white/black background rectangles). The selected biomarkers were underscored by asterisk (*) to identify differences between NV(day 0) vs. PV(day 30–45). (B) Venn diagram report was employed to identify the set of biomarkers selectively increased in [PV(day 30–45) vs. NV(day 0)]. The attributes EMCD4, EMCD8, nCMCD19, TNFCD8, and IFNCD8 were underscored as PV (day 30–45)-selective biomarkers. These attributes were tagged in bold underline format and employed for follow-up analysis overtime after 17DD-YF primary vaccination. Biomarkers with proportion higher than the 25% were underscored by asterisk (*) to identify differences between NV(day 0) vs. PV(day 30–45). (C) Overlaid signatures of selected biomarkers were assembled to identify changes in the 17DD-YF specific phenotypic and functional features at different times after primary vaccination: PV(day 30–45) [], PV(year 1) [], PV(year 2) [], PV(year 4) [], PV(year 7) [], and PV(year 10) []. Dashed rectangles underscore the critical decline of selected biomarkers overtime after primary vaccination with absence of EMCD8 ≥ 4 years after primary vaccination.

Figure 4

Descriptive analysis of selected biomarkers after 17DD-YF primary vaccination in children. (A) Prevalence of Biomarkers (PRNT>1:10, EMCD4, EMCD8, nCMCD19, TNFCD8, and IFNCD8) above the 75th percentile cut-off. Data are presented as proportion of subjects with biomarkers above the cut-off at baseline NV(day 0) [] and at different times after primary vaccination: PV(day 30–45) [], PV(year 1) [], PV(year 2) [], PV(year 4) [], PV(year 7) [], and PV(year 10) []. (B) Number of Biomarkers (PRNT>1:10, EMCD4, EMCD8, nCMCD19, TNFCD8, and IFNCD8) above the 75th Quartile cut-off. Data are presented as mean (min to max) number of biomarkers above the cut-off at baseline NV(day 0) [] and at different times after primary vaccination: PV(day 30–45) [], PV(year 1) [], PV(year 2) [], PV(year 4) [], PV(year 7) [], and PV(year 10) [] as well as PV(≤year 2) [], PV(≥year 4) []. Significant differences at p < 0.05 are underscored by using letters “a,” “b,” “c,” and “d” for comparisons with NV(day 0), PV(day 30–45), PV(year 1), and PV(year 2), respectively. Intergroup differences are identified by connecting lines. The p-values are provided in the figure.

Figure 5

Predictive capacity of biomarkers to monitor the 17DD-YF memory after primary vaccination in children. The Receiver Operating Characteristic (ROC) curves were used to estimate the capacity of time as a predictor of changes in biomarker levels to monitor the 17DD-YF memory after primary vaccination in children. Logistic and multinomial regression models were constructed to evaluate the association between time after vaccination and changes in the biomarker levels. Following, the fitted regression model was employed to calculate the predicted probabilities for each biomarker (A) isolated or (B) combined along time continuum. The Area Under the ROC Curves (AUC) were employed for comparative analysis of predictive capacity amongst biomarkers and the values provided in the figure. The gray background highlights the top three isolated biomarkers and the best combination of predictor biomarkers to monitor the 17DD-YF memory after primary vaccination in children.

Figure 6

Resultant memory (PRNT, EMCD8, or nCMCD19) after 17DD-YF Primary vaccination in children. The resultant memory status (PRNT, EMCD8, or nCMCD19) were assessed at individual level to define the overall proportion of subjects presenting None (), PRNT (), cellular memory “EMCD8 and/or nCMCD19” () or both attributes “PRNT and cellular memory” () at distinct time-points before/after primary 17DD-YF vaccination, including: NV(day 0), PV(≤year 2), PV(≥year 4) as well as PV(day 30–45), PV(year 1), PV(year 2), PV(year 4), PV(year 7), and PV(year 10). Significant differences (p < 0.05) of resultant memory status amongst study groups were assessed by Chi-square test and represented by letter “a” as compared to NV(day 0).