Effect of priming interval on reactogenicity, peak immunological response, and waning after homologous and heterologous COVID-19 vaccine schedules: exploratory analyses of Com-COV, a randomised control trial

Abstract

Background: Priming COVID-19 vaccine schedules have been deployed at variable intervals globally, which might influence immune persistence and the relative importance of third-dose booster programmes. Here, we report exploratory analyses from the Com-COV trial, assessing the effect of 4-week versus 12-week priming intervals on reactogenicity and the persistence of immune response up to 6 months after homologous and heterologous priming schedules using the vaccines BNT162b2 (tozinameran, Pfizer/BioNTech) and ChAdOx1 nCoV-19 (AstraZeneca).

Methods: Com-COV was a participant-masked, randomised immunogenicity trial. For these exploratory analyses, we used the trial's general cohort, in which adults aged 50 years or older were randomly assigned to four homologous and four heterologous vaccine schedules using BNT162b2 and ChAdOx1 nCoV-19 with 4-week or 12-week priming intervals (eight groups in total). Immunogenicity analyses were done on the intention-to-treat (ITT) population, comprising participants with no evidence of SARS-CoV-2 infection at baseline or for the trial duration, to assess the effect of priming interval on humoral and cellular immune response 28 days and 6 months post-second dose, in addition to the effects on reactogenicity and safety. The Com-COV trial is registered with the ISRCTN registry, 69254139 (EudraCT 2020-005085-33).

Findings: Between Feb 11 and 26, 2021, 730 participants were randomly assigned in the general cohort, with 77-89 per group in the ITT analysis. At 28 days and 6 months post-second dose, the geometric mean concentration of anti-SARS-CoV-2 spike IgG was significantly higher in the 12-week interval groups than in the 4-week groups for homologous schedules. In heterologous schedule groups, we observed a significant difference between intervals only for the BNT162b2-ChAdOx1 nCoV-19 group at 28 days. Pseudotyped virus neutralisation titres were significantly higher in all 12-week interval groups versus 4-week groups, 28 days post-second dose, with geometric mean ratios of 1·4 (95% CI 1·1-1·8) for homologous BNT162b2, 1·5 (1·2-1·9) for ChAdOx1 nCoV-19-BNT162b2, 1·6 (1·3-2·1) for BNT162b2-ChAdOx1 nCoV-19, and 2·4 (1·7-3·2) for homologous ChAdOx1 nCoV-19. At 6 months post-second dose, anti-spike IgG geometric mean concentrations fell to 0·17-0·24 of the 28-day post-second dose value across all eight study groups, with only homologous BNT162b2 showing a slightly slower decay for the 12-week versus 4-week interval in the adjusted analysis. The rank order of schedules by humoral response was unaffected by interval, with homologous BNT162b2 remaining the most immunogenic by antibody response. T-cell responses were reduced in all 12-week priming intervals compared with their 4-week counterparts. 12-week schedules for homologous BNT162b2 and ChAdOx1 nCoV-19-BNT162b2 were up to 80% less reactogenic than 4-week schedules.

Interpretation: These data support flexibility in priming interval in all studied COVID-19 vaccine schedules. Longer priming intervals might result in lower reactogenicity in schedules with BNT162b2 as a second dose and higher humoral immunogenicity in homologous schedules, but overall lower T-cell responses across all schedules. Future vaccines using these novel platforms might benefit from schedules with long intervals.

Funding: UK Vaccine Taskforce and National Institute for Health and Care Research.

Figure 1

Immune responses between 4-week and 12-week intervals at 28 days and 6 months post-second dose in the general cohort Data presented are the geometric means and 95% CIs. Fold changes were calculated by dividing the immune response at 6 months post-second dose by that at 28 days post-second dose. GMRs between schedules with 4-week and 12-week intervals were adjusted for study site and paracetamol usage in the first 24 h post-vaccination (yes or no) for the 28-day data; 6-month visit time (in days) was further adjusted for the 6-month data and fold change. The vertical line is the line of no difference between 4-week and 12-week interval groups. aGMR=adjusted geometric mean ratio. ELU=ELISA units. GMR=geometric mean ratio. NT₅₀=50% neutralisation titre. PBMCs=peripheral blood mononuclear cells. SFC=spot-forming cell. WT=wild type.

Figure 2

Kinetics of immune response over time with all schedules normalised by time of second dose in the seronegative general cohort Figure shows results for anti-spike IgG titres (A) and T-cell ELISpot counts (B). Day 0 refers to time of second dose. Datapoints are geometric mean concentrations, with whiskers showing the 95% CIs. Numbers of participants per timepoint are shown in the appendix (p 16). ELU=exponential linear unit. PBMCs=peripheral blood mononuclear cells. SFC=spot-forming cell.

Figure 3

Immune responses between prophylactic and reactive paracetamol use at 28 days, 3 months, and 6 months post-second dose in the 12-week interval groups Data shown are geometric means (95% CI) in the intention-to-treat population. GMRs and two-sided 95% CIs were adjusted for study site, age, gender, and pre-boost immunogenicity (anti-spike IgG for humoral responses and ELISpot for cellular response). Numbers of participants per timepoint are shown in the appendix (p 16). aGMR=adjusted geometric mean ratio. ELU=ELISA units. FRNT₅₀=50% focal reduction neutralisation titre. GMR=geometric mean ratio. NT₅₀=50% neutralisation titre. PBMCs=peripheral blood mononuclear cells. SFC=spot-forming cell. WT=wild type.

Figure 4

Forest plots of solicited adverse events in days 0–7 post-second dose by vaccine schedule comparing 12-week interval groups to 4-week interval groups in the general cohort Models adjusted for paracetamol recommendation (reactive or prophylactic use). Models with no adjusted OR were non-estimable due to few events in that study group. The vertical line is the line of no difference between 4-week and 12-week interval groups. NE=not estimable. OR=odds ratio.