Differential Kinetics of Immune Responses Elicited by Covid-19 Vaccines.

To the Editor: Previous studies have shown that the BNT162b2 (Pfizer–BioNTech), mRNA-1273 (Moderna), and Ad26.COV2.S (Johnson & Johnson–Janssen) vaccines provide robust protective efficacy against coronavirus disease 2019 (Covid-19). Here, we report comparative kinetics of humoral and cellular immune responses elicited by the two-dose BNT162b2 vaccine (in 31 participants), the two-dose mRNA-1273 vaccine (in 22 participants), and the one-dose Ad26.COV2.S vaccine (in 8 participants). We evaluated antibody and T-cell responses from peak immunity at 2 to 4 weeks after the second immunization in recipients of the messenger RNA (mRNA) vaccines or after the first immunization in recipients of the Ad26.COV2.S vaccine to 8 months (Table S1 in the Supplementary Appendix, available with the full text of this letter at NEJM.org).

At peak immunity, the BNT162b2 vaccine induced a high median live-virus neutralizing antibody titer (1789), a high median pseudovirus neutralizing antibody titer (700), and a high median binding antibody titer against the receptor-binding domain (RBD) (21,564). However, these titers declined sharply by 6 months after vaccination, as previously reported,[1,2] and they declined further by 8 months (Figure 1A through 1C, S1, and S2). By 8 months after BNT162b2 vaccination, the median live-virus neutralizing antibody titer (53), pseudovirus neutralizing antibody titer (160), and RBD-specific binding antibody titer (755) elicited by the vaccine were lower than the peak titers by a factor of 34, 4, and 29, respectively.

Figure 1

Kinetics of Humoral and Cellular Immune Responses Elicited by the BNT162b2, mRNA-1273, and Ad26.COV2.S Vaccines.

Shown are immune responses after vaccination with BNT162b2, mRNA-1273, and Ad26.COV2.S at peak immunity (2 to 4 weeks after the second dose in recipients of the messenger RNA vaccines or 4 weeks after one dose in recipients of the Ad26.COV2.S vaccine) and at 6 months, 8 months, or both after the first dose. Panel A shows the serum 50% inhibitory dilution (ID50) titers in the live-virus neutralizing antibody assay. Red bars indicate medians, dashed lines the limit of detection for each assay, and each dot a single participant. Panel B shows the serum dilution for 50% reduction (NT50) expressed in relative light units in the pseudovirus neutralizing antibody assay. Panel C shows the binding IgG antibody titers against the receptor-binding domain (RBD) in the serum enzyme-linked immunosorbent assay. Intracellular cytokine-staining assays were performed to measure the percentage of interferon-γ production in T cells; Panel D shows this percentage in CD4+ T cells, and Panel E shows this percentage in CD8+ T cells. Flow cytometric gating was performed to identify T cells (which are CD3+) rather than other CD4+- or CD8+-expressing immune cells. All assays were performed with the use of the SARS-CoV-2 WA1/2020 strain. The Ad26.COV2.S vaccine data in Panels B through E were published previously[3] and are included here for comparative purposes.

At peak immunity, the mRNA-1273 vaccine also elicited a high median live-virus neutralizing antibody titer (5848), pseudovirus neutralizing antibody titer (1569), and RBD-specific binding antibody titer (25,677). By 8 months after mRNA-1273 vaccination, the median live-virus neutralizing antibody titer was 133, the pseudovirus neutralizing antibody titer was 273, and the median RBD-specific binding antibody titer was 1546; these titers were lower than the peak titers by a factor of 44, 6, and 17, respectively.

The Ad26.COV2.S vaccine induced substantially lower median titers than the mRNA vaccines at peak immunity. At 4 weeks after single-shot Ad26.COV2.S immunization, the median live-virus neutralizing antibody titer was 146, the median pseudovirus neutralizing antibody titer was 391, and the median RBD-specific binding antibody titer was 1361; however, these titers remained relatively stable over 8 months.[3] At 8 months, the median live-virus neutralizing antibody titer was 629, the median pseudovirus neutralizing antibody titer was 185, and the median RBD-specific binding antibody titer was 843; these titers were similar to the titers at week 4. With all three vaccines, there were generally stable antibody-dependent cellular phagocytosis and antibody-dependent complement deposition responses (Fig. S3).

Recipients of the BNT162b2 and mRNA-1273 vaccines also had decreases in titers of live-virus neutralizing antibodies, pseudovirus neutralizing antibodies, and RBD- and spike protein (S)–specific binding antibody responses against severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) variants from peak immunity to 8 months; after Ad26.COV2.S vaccination, however, there were stable or in some cases increasing antibody titers against these variants (Figs. S4 and S5). At 8 months, the median pseudovirus neutralizing antibody titers against the SARS-CoV-2 B.1.617.2 (delta) variant were similar with the BNT162b2 vaccine (67), the mRNA-1273 vaccine (76), and the Ad26.COV2.S vaccine (107).

T-cell responses were assessed by CD4+ and CD8+ intracellular cytokine-staining assays that used pooled S peptides for stimulation (Figure 1D and 1E). At 8 months, the median CD8+ T-cell responses were 0.016% with the BNT162b2 vaccine, 0.017% with the mRNA-1273 vaccine, and 0.12% with the Ad26.COV2.S vaccine. With all three vaccines, T-cell responses showed broad cross-reactivity against SARS-CoV-2 variants (Fig. S6).

These data show differential kinetics of immune responses induced by the mRNA and Ad26.COV2.S vaccines over an 8-month follow-up period. As shown in previous studies,[1,2] the BNT162b2 and mRNA-1273 vaccines were characterized by high peak antibody responses that declined sharply by 6 months; these responses declined further by 8 months. Antibody titers in recipients of the mRNA-1273 vaccine were generally higher than those in recipients of the BNT162b2 vaccine. The Ad26.COV2.S vaccine induced lower initial antibody responses, but these responses were relatively stable over the 8-month follow-up period, with minimal-to-no evidence of decline.[3] These findings have important implications for waning vaccine immunity, although correlates of protection from SARS-CoV-2 are not yet defined.