Longitudinal cellular and humoral immune responses after triple BNT162b2 and fourth full-dose mRNA-1273 vaccination in haemodialysis patients

Abstract

Haemodialysis patients respond poorly to vaccination and continue to be at-risk for severe COVID-19. Therefore, dialysis patients were among the first for which a fourth COVID-19 vaccination was recommended. However, targeted information on how to best maintain immune protection after SARS-CoV-2 vaccinations in at-risk groups for severe COVID-19 remains limited. We provide, to the best of our knowledge, for the first time longitudinal vaccination response data in dialysis patients and controls after a triple BNT162b2 vaccination and in the latter after a subsequent fourth full-dose of mRNA-1273. We analysed systemic and mucosal humoral IgG responses against the receptor-binding domain (RBD) and ACE2-binding inhibition towards variants of concern including Omicron and Delta with multiplex-based immunoassays. In addition, we assessed Spike S1-specific T-cell responses by interferon γ release assay. After triple BNT162b2 vaccination, anti-RBD B.1 IgG and ACE2 binding inhibition reached peak levels in dialysis patients, but remained inferior compared to controls. Whilst we detected B.1-specific ACE2 binding inhibition in 84% of dialysis patients after three BNT162b2 doses, binding inhibition towards the Omicron variant was only detectable in 38% of samples and declining to 16% before the fourth vaccination. By using mRNA-1273 as fourth dose, humoral immunity against all SARS-CoV-2 variants tested was strongly augmented with 80% of dialysis patients having Omicron-specific ACE2 binding inhibition. Modest declines in T-cell responses in dialysis patients and controls after the second vaccination were restored by the third BNT162b2 dose and significantly increased by the fourth vaccination. Our data support current advice for a four-dose COVID-19 immunisation scheme for at-risk individuals such as haemodialysis patients. We conclude that administration of a fourth full-dose of mRNA-1273 as part of a mixed mRNA vaccination scheme to boost immunity and to prevent severe COVID-19 could also be beneficial in other immune impaired individuals. Additionally, strategic application of such mixed vaccine regimens may be an immediate response against SARS-CoV-2 variants with increased immune evasion potential.

Keywords: COVID-19; Omicron variant of concern; dialysis; immunocompromised; longitudinal response; mRNA vaccination; mixed mRNA vaccination; protective immunity.

Figure 1

Participant recruitment scheme for longitudinal vaccination response analysis in haemodialysis patients after triple BNT162b2 and fourth full-dose mRNA-1273. Patients on haemodialysis (n = 50) and healthcare workers as controls (n = 33) were triple-vaccinated with BNT162b2 (green syringe) followed by a 100 µg (full) dose of mRNA-1273 (blue syringe) for dialysed individuals only. Sampling and vaccination schedule is given in days and weeks as indicated.

Figure 2

Humoral immune response in haemodialysis patients after a triple vaccination with BNT162b2. IgG response (A) and ACE2 binding inhibition (B) towards the SARS-CoV-2 B.1 RBD isolate were measured in plasma from haemodialysis patients (blue circles, n = 50) and controls (orange circles, n=33) using MULTICOV-AB (A) or an ACE2-RBD competition assay (B) after double or triple vaccination with BNT162b2 at the indicated time points. Data is displayed as normalised median fluorescence intensity (MFI) signal (A) for IgG binding or as % ACE2 binding inhibition where 100% indicates maximum inhibition and 0% no inhibition (B). Samples with an ACE2 binding inhibition of less than 20% (dashed line) are classified as non-responders. Boxes represent the median, 25th and 75th percentiles, whiskers show the largest and smallest non-outlier values. Outliers were determined by 1.5 times IQR. Mean sampling time in days after two-dose BNT162b2 vaccination as Δt is displayed on the x-axis. Statistical significance was calculated by two-sided Mann-Whitney-U test. P-values for relevant comparisons are given above the sample groups. Significance was defined as p < 0.05. Response data from dialysed individuals from day 21 and day 113 after the second BNT162b2 dose were already published before as part of Strengert et al. (16) and Dulovic et al. (17).

Figure 3

Longitudinal humoral immune response in haemodialysis patients after a triple vaccination with BNT162b2 and a fourth full-dose of mRNA-1273. IgG response (A) and ACE2 binding inhibition (B–D) towards the SARS-CoV-2 RBD of B.1 (A, B), δ (C) and Ο BA.1 (D) isolates were measured in plasma from haemodialysis patients (n = 50) using MULTICOV-AB (A) or an ACE2-RBD competition assay (B–D) after immunisation with a triple dose of BNT162b2 (green syringe) and a fourth full-dose of mRNA-1273 (blue syringe). Data is displayed as normalised median fluorescence intensity (MFI) signal for IgG binding (A) or as % ACE2 binding inhibition where 100% indicates maximum inhibition and 0% no inhibition (B–D). Samples with an ACE2 binding inhibition of less than 20% (dashed line) are classified as non-responders (B–D). Interconnecting lines represent samples from the same individual. Sampling time points in days after the standard complete two-dose BNT162b2 vaccination is stated below the graph. Statistical significance was calculated by two-sided paired Wilcoxon rank test. Significance was defined as p < 0.05. All p-values for relevant comparisons are listed in Table S3 .

Figure 4

Impact of triple vaccination with BNT162b2 and a fourth full-dose of mRNA-1273 on cellular immune response in haemodialysis patients. (A) Whole blood from longitudinally-sampled vaccinated haemodialysis patients (n = 50) was ex vivo stimulated using a SARS-CoV-2 Spike S1-specific peptide pool. Supernatant fractions were then analysed by interferon γ release assay (IGRA). Interconnecting lines represent samples from the same individual. Sampling time points in days after two-dose BNT162b2 vaccination is displayed on the x-axis. Statistical significance was calculated by two-sided paired Wilcoxon rank test. Significance was defined as p<0.05. All p-values for relevant comparisons are listed in Table S3 . Response data from dialysed individuals from day 21 after the second BNT162b2 dose were already published before as part of Strengert et al. (16) and Dulovic et al. (17). (B) T-cell responses assessed by IGRA and B-cell responses assessed by ACE2-RBD competition assay towards the RBD of B.1, Delta, Omicron BA.1 isolates were plotted for correlation analysis (B). Correlation was calculated using Spearman’s coefficient rho. P values are marked as * <0.05, ** <0.01, *** <0.001. Dashed lines indicated the respective responder thresholds for IGRA (IFNγ release >200 mIU/mL) and RBD-ACE2 binding assay (20%). Responder rates (%) for both cellular and humoral response are shown in the upper right quadrant.