Durable immune responses after BNT162b2 vaccination in home-dwelling old adults

Abstract

Objectives: Elderly are an understudied, high-risk group vulnerable to severe COVID-19. We comprehensively analyzed the durability of humoral and cellular immune responses after BNT162b2 vaccination and SARS-CoV-2 infection in elderly and younger adults.

Methods: Home-dwelling old (n = 100, median 86 years) and younger adults (n = 449, median 38 years) were vaccinated with two doses of BNT162b2 vaccine at 3-week intervals and followed for 9-months. Vaccine-induced responses were compared to home-isolated COVID-19 patients (n = 183, median 47 years). Our analysis included neutralizing antibodies, spike-specific IgG, memory B-cells, IFN-γ and IL-2 secreting T-cells and sequencing of the T-cell receptor (TCR) repertoire.

Results: Spike-specific breadth and depth of the CD4⁺ and CD8⁺ TCR repertoires were significantly lower in the elderly after one and two vaccinations. Both vaccinations boosted IFN-γ and IL-2 secreting spike-specific T-cells responses, with 96 % of the elderly and 100 % of the younger adults responding after the second dose, although responses were not maintained at 9-months. In contrast, T-cell responses persisted up to 12-months in infected patients. Spike-specific memory B-cells were induced after the first dose in 87 % of the younger adults compared to 38 % of the elderly, which increased to 83 % after the second dose. Memory B-cells were maintained at 9-months post-vaccination in both vaccination groups. Neutralizing antibody titers were estimated to last for 1-year in younger adults but only 6-months in the older vaccinees. Interestingly, infected older patients (n = 15, median 75 years) had more durable neutralizing titers estimated to last 14-months, 8-months longer than the older vaccinees.

Conclusions: Vaccine-induced spike-specific IgG and neutralizing antibodies were consistently lower in the older than younger vaccinees. Overall, our data provide valuable insights into the kinetics of the humoral and cellular immune response in the elderly after SARS-CoV-2 vaccination or infection, highlighting the need for two doses, which can guide future vaccine design.Clinical trials.gov; NCT04706390.

Keywords: BNT162b2; Elderly; Memory B-cell; Neutralising antibody; SARS-CoV-2; T-cell.

Fig. 1

Study population flowchart. Number of vaccinees eligible for inclusion at baseline and sampling of serum and peripheral blood mononuclear cell (PBMC) for the participants during the study.

Fig. 2

Durability of spike-specific humoral immune responses after BNT162b2 vaccination. (A) Study design showing time points for vaccination and blood sampling in young adults (blue circles) aged 23–69 years (n = 449) and elderly persons (red triangles) aged 70–99 years (n = 100). Each symbol represents one individual. (B) Anti-spike serum IgG endpoint titers measured by ELISA. Data are presented as geometric mean titer (GMT) with 95 % confidence intervals (CIs). (C) Neutralizing antibody responses measured by microneutralization assay against the ancestral D614G strain in all elderly and a subgroup of adults including those that provided peripheral blood mononuclear cells (PBMCs) (n = 41). The neutralizing titer was defined as the reciprocal serum dilution resulting in 50 % neutralization. The data are presented as GMT with 95 % CIs. (D) Spike-specific memory B-cell responses were measured by ELISpot using PBMCs collected from a subgroup of vaccinated younger adults (n = 35) and elderly (n = 24). The frequency of spike-specific memory B-cells were defined as spot forming units (SFU) per 10⁶ PBMC. The data are presented as mean with 95 % CIs. Mixed-effects model with normalized outcome variables with fixed effects of sex, age group, presence of comorbidity, use of immunosuppressive medication and age-by-time interaction (except neutralization and memory B-cells), and individual repeated measures as a random factor. Significance of differences between pairs of group means was assessed by post-hoc tests. P values were only reported if they were significant at the 5 % level after Bonferroni correction. ****P < 0.0001, **P < 0.01, *P < 0.05. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 3

Spike-specific T-cell receptor sequencing after BNT162b2 vaccination. (A) The clonal breadth of the spike-specific T-cell receptor (TCR) repertoire measured by immunoSEQ in vaccinated adults (blue circles) aged 23–69 years (n = 449) and elderly persons (red triangles) aged 70–99 years (n = 100). Blood samples were tested at baseline, and 3 weeks after first and second vaccination (6 weeks post-first vaccination). Each symbol represents one individual. The proportion of unique spike-specific TCR sequences (breadth) for all T-cells (top), CD4⁺ T-cells (middle) and CD8⁺ T-cells (bottom). Data are presented as mean with 95 % confidence intervals. (B) The frequency of unique spike-specific TCR sequences (depth) for the total TCR repertoire (top), CD4⁺ T-cells (middle) and CD8⁺ T-cells (bottom). Mixed-effects model with normalized outcome variables with fixed effects of sex, age group, presence of comorbidity, use of immunosuppressive medication and age-by-time interaction (except CD8⁺ TCR depth), and individual repeated measures as a random factor. Significance of differences between pairs of group means was assessed by post-hoc tests. P values were only reported if they were significant at the 5 % level after Bonferroni correction. ****P < 0.0001, ***P < 0.001, **P < 0.01, *P < 0.05. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 4

Functional spike-specific T-cell responses after BNT162b2 vaccination. (A–C) The frequency of interferon-γ (IFN-γ) (A), interleukin-2 (IL-2) (B) and double cytokine-producing spike-specific T-cell responses (C) measured by fluorospot in vaccinated adults (n = 35, blue circles) and elderly (n = 24, red triangles). Each symbol represents one individual. T-cell frequencies are reported as spot-forming units (SFU) per 10⁶ peripheral blood mononuclear cells at baseline, and 3 weeks after first and second vaccination (6 weeks post-first vaccination) and 5- and 9-months post-first vaccination. Data are presented as mean with 95 % confidence intervals. (D) The proportion of T-cells producing single cytokines IFN-γ (green) or IL-2 (blue), or double-producing IFN-γ and IL-2 (gray) in adults (top row) and elderly (bottom row). The proportions are presented as percentage of single or double-producing T-cells of the total number of T-cells measured in the younger (n = 35) and older adults (n = 24). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 5

Spike-specific immune responses after vaccination and SARS-CoV-2 infection. (A) Spike-specific IgG titers, neutralizing titers and spike-specific memory B-cell responses were measured for vaccinees 3 and 6 week, and 5- and 9-months post-vaccination and infected subjects at 8 weeks, 6- and 12-months post-infection. Spike-specific IgG and neutralizing titers are presented as geometric means with 95 % confidence intervals (CIs). Spike-specific memory B-cell responses were measured by ELISpot using peripheral blood mononuclear cells (PBMCs). The frequency of spike-specific memory B-cells were defined as spot forming units (SFU) per 10⁶ PBMC. The data are presented as the mean with 95 % CIs. Statistical significance was assessed between the infected group (green) and vaccinated older (red) or younger adults (blue) using a mixed effects model for normalized outcome measures, adjusted for repeated-measure subject variance and demographic factors, and post-hoc tests with Bonferroni correction. Red stars represent the significance level between the older adults and the infected group and blue stars represent significance level between the younger adults and the infected group. IgG titer: younger adults n = 449, older adults n = 100, infected n = 198. Neutralizing titer: younger adults n = 43, older adults n = 100. Memory B-cells: younger adults n = 34, older adults n = 24, infected 6 months n = 10, infected 12 months n = 63. (B) Spike-specific T-cell responses measured as the frequency of interferon-γ (IFN-γ), interleukin-2 (IL-2) and double cytokine-producing spike-specific T-cells measured by FluoroSpot assay in vaccinees at 3- and 6-week, and 5- and 9-months post-vaccination (younger adults n = 35, older adults n = 24) and infected subjects at 6 and 12 months post-infection (6 months n = 38, 12 months n = 63). Data are reported as mean SFU/10⁶ PBMC with 95 % CIs. (C) Radar chart summarizing durable spike-specific immune responses in SARS-CoV-2 infected (left) and vaccinated younger adults (middle) and older adults (right). The data are presented as means on a log-axis. NT (neutralizing titer), memory B-cell (MBC). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 6

Age-specific antibody responses after vaccination and natural SARS-CoV-2 infection. (A–F) Spike-specific IgG (A–C) and neutralizing titers (D–F) in vaccinated individuals (n = 549) (dark blue circles) and infected individuals (n = 183) (green squares) were divided into 15-year age groups. Blood samples were collected 6-weeks after the first vaccination (3-weeks after the second vaccination), 5- and 9-months post-vaccination. Home-isolated SARS-CoV-2 infected individuals (n = 183) provided blood samples 8-weeks, 6- and 12-months post-acute infection. Data are presented as geometric mean titer (GMT) with 95 % confidence interval (CI) and each symbol represents one individual. The threshold for detectable neutralizing antibodies (<20) is indicated with a dotted line. Multiple linear regression analysis was used to test if vaccination or infection group predicted log normalized IgG or neutralizing titers, adjusted for repeated-measure subject variance and covariates sex, age, comorbidity, use of immunosuppressive medication and age-by-group interaction. P values were only reported if they were significant at the 5 % level after Bonferroni correction. ****P < 0.0001, ***P < 0.001, **P < 0.01, *P < 0.05. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)