Transient Autoreactive PF4 and Antiphospholipid Antibodies in COVID-19 Vaccine Recipients

Abstract

Vaccine-induced immune thrombotic thrombocytopenia (VITT) is a rare autoimmune condition associated with recombinant adenovirus (rAV)-based COVID-19 vaccines. It is thought to arise from autoantibodies targeting platelet factor 4 (aPF4), triggered by vaccine-induced inflammation and the formation of neo-antigenic complexes between PF4 and the rAV vector. To investigate the specific induction of aPF4 by rAV-based vaccines, we examined sera from rAV vaccine recipients (AZD1222, AD26.COV2.S) and messenger RNA (mRNA) based (mRNA-1273, BNT162b2) COVID-19 vaccine recipients. We compared the antibody fold change (FC) for aPF4 and for antiphospholipid antibodies (aPL) of rAV to mRNA vaccine recipients. We combined two biobanks of Dutch healthcare workers and matched rAV-vaccinated individuals to mRNA-vaccinated controls, based on age, sex and prior history of COVID-19 (AZD1222: 37, Ad26.COV2.S: 35, mRNA-1273: 47, BNT162b2: 26). We found no significant differences in aPF4 FCs after the first (0.99 vs. 1.08, mean difference (MD) = -0.11 (95% CI -0.23 to 0.057)) and second doses of AZD1222 (0.99 vs. 1.10, MD = -0.11 (95% CI -0.31 to 0.10)) and after a single dose of Ad26.COV2.S compared to mRNA-based vaccines (1.01 vs. 0.99, MD = 0.026 (95% CI -0.13 to 0.18)). The mean FCs for the aPL in rAV-based vaccine recipients were similar to those in mRNA-based vaccines. No correlation was observed between post-vaccination aPF4 levels and vaccine type (mean aPF difference -0.070 (95% CI -0.14 to 0.002) mRNA vs. rAV). In summary, our study indicates that rAV and mRNA-based COVID-19 vaccines do not substantially elevate aPF4 levels in healthy individuals.

Keywords: COVID-19 vaccines; autoantibodies; platelet factor 4; thrombocytopenia; thrombosis.

Figure 1

Flow diagram of participant selection for analysis. Abbreviations: EMC: Erasmus Medical Center; AUMC: Amsterdam University Medical Centers; rAV: recombinant AdenoVirus; mRNA: messenger RNA.

Figure 2

Longitudinal assessment of autoantibody responses to COVID-19 vaccines of aPF4IgG (A–E), aβ2GP (F–J), aCLIgM (K–O) and aCLIgG (P–T), in the combined rAV and mRNA groups (A,F,K,P), AZD1222 (B,G,L,Q), Ad26.COV2.S (C,H,M,R), mRNA-1273 (D,I,N,S) and BNT162b2 (E,J,O,T). Colored symbols indicate individual values, black lines are means and whiskers indicate 95% confidence intervals. The dotted lines indicate the ELISA cut-offs according to the manufacturer. Abbreviations: rAV: recombinant AdenoVirus; mRNA: messenger RNA; aPF4: Autoantibodies against Platelet Factor 4; IgG: Immunoglobulin G; Ig: Immunoglobulin; aβ2GP: anti-beta2-Glycoprotein; aCL: anti-CardioLipin; IgM: Immunoglobulin M. Mixed-modeling was used with Geisser-Greenhouse correction for violation of sphericity. Šídák’s multiple comparisons test was used for primary endpoints and uncorrected Fisher’s Least Significant Difference for secondary endpoints. ns = not significant, * = p < 0.05, ** = p < 0.01, **** = p < 0.0001.

Figure 3

Correlations between aPF O.D values and several covariates. Forrest plot for mean differences with a 95% confidence interval for the correlation between aPF4 O.D. values and several covariates. The vaccine is the comparison of rAV-based versus mRNA-based vaccines. COVID-19 history is the comparison of yes versus no. The study site is the comparison between Amsterdam University Medical Centers vs Erasmus Medical Center. Age is expressed per 10 years from 0 years. The time between vaccine and sampling is expressed per 10 days from the day of the first COVID-19 vaccination (T1). Abbreviations: rAV: recombinant AdenoVirus; mRNA: messenger RNA; EMC: Erasmus Medical Center; AUMC: Amsterdam University Medical Center.