Portal de Periódicos da CAPES

Prevalence of anti‐platelet factor 4 antibodies in healthy vaccinees with adenoviral vector vaccines—A systematic review and meta‐analysis

2023; Wiley; Volume: 200; Issue: 6 Linguagem: Inglês

10.1111/bjh.18635

1365-2141

Bianca Clerici, Mariangela Scavone, Simone Birocchi, Chiara Aloise, Benedetta Peroni, A Negrini, Claudia Ghali, Giovanni Casazza, Gian Marco Podda,

Intramuscular injections and effects

Cases of thrombocytopenia and thrombosis have been reported following vaccination with COVID-19 adenoviral vector vaccines, Ad26.COV.2 and ChAdOx1 nCoV-19.1-3 The condition, termed vaccine-induced immune thrombotic thrombocytopenia (VITT),4 is caused by antibodies against platelet factor 4 (PF4) bound to platelets.5 VITT strongly mimics auto-immune heparin-induced thrombocytopenia (aHIT).6 Thus, the HIT diagnostic algorithm, based on anti-PF4 enzyme-linked immunosorbent assays (ELISAs) as screening tests and on platelet activation assays (PAAs) as confirmation tests, has been applied to VITT.1, 2, 4 PF4-ELISAs are enzyme immunoassays designed to detect serum immunoglobulin G (IgG) or IgG/A/M anti-PF4 antibodies. PAAs can ascertain whether detected anti-PF4 antibodies display platelet-activating properties by exploring the capacity of patient serum to induce platelet aggregation.7 Since the very first descriptions of VITT, it has been proposed to optimize PAAs by adding exogenous PF4 (PF4-PAAs).1 Anti-PF4 antibodies have been reported in as many as 6.6% of healthy blood donors8 and in up to 50% of inpatients exposed to heparin.9-11 Whether the prevalence of anti-PF4 antibodies in healthy subjects with adenoviral vector vaccines is comparable to that of unvaccinated individuals is currently unknown. Here we report the results of a systematic review and meta-analysis of the prevalence of anti-PF4 antibodies in these subjects. PubMed and Embase were searched by Chiara Aloise and Benedetta Peroni with a pre-defined research string (see Data S1). Our research was restricted to articles written in English and published from 1 January 2020, until 31 July 2022. We evaluated: (i) VITT case reports and case series providing details on performed laboratory tests; (ii) studies investigating anti-PF4 antibodies in vaccinees with an adenoviral vector vaccine with no signs or symptoms of VITT; and (iii) laboratory studies investigating novel diagnostic tests and/or comparing different available diagnostic tests. We included only studies on vaccinees and laboratory studies that included vaccinees, and excluded studies or single patients of eligible studies if: (i) details on test methods were absent, scarce or unclear; or (ii) test results were unclear. Eligibility assessment was performed at title and/or abstract level by Chiara Aloise and Benedetta Peroni, who independently assessed each study. Disagreements were solved by consensus or by a third reviewer (Simone Birocchi). Performed diagnostic tests were categorized according to brand (PF4-ELISAs) and type (PAAs and PF4-PAAs). We only recorded results of diagnostic tests performed at least five times. For each study, we calculated the prevalence of anti-PF4 antibodies with their 95% confidence intervals (CI) using the exact binomial method. Further details on statistical analysis can be found in Data S1. We retrieved 2157 citations. After the exclusion of duplicates (n = 886) and ineligible studies (n = 1264), seven studies were included (Figure S1, full references in Data S1). Six out of seven eligible studies were observational clinical studies designed to determine the frequency of anti-PF4 antibodies in healthy vaccinees. Healthy vaccinees served as controls in a VITT case series by Althaus et al.12 The diagnostic tests performed in each study are shown in Table S1. The seven retrieved studies provided results of 3149 PF4-ELISAs, 22 PAAs and 16 PF4-PAAs, performed on a total of 3149 healthy vaccinees. Anti-PF4 antibodies were assayed with a PF4-ELISA in all eligible studies. In five out of seven eligible studies, PAAs and/or PF4-PAAs followed positive PF4-ELISAs. Table 1 shows PF4-ELISA results. Overall, the prevalence of anti-PF4 antibodies with PF4-ELISAs was relatively low (3.7%, 95% CI 2.1%–5.6%), and the optical densities (ODs) of all but one positive result were below 2.0. Table 2 shows PAA and PF4-PAA results. The prevalence of platelet-activating anti-PF4 antibodies detected by PAAs was 0% (95% CI, 0%–15.4%), with wide confidence intervals reflecting the paucity of tested subjects (n = 22). The prevalence of platelet-activating anti-PF4 antibodies detected by PF4-PAAs [n = 16, all tested with heparin-induced platelet activation (HIPA)] was 0% (95% CI, 0%–20.6%). Our systematic review and meta-analysis shows that the prevalence estimates of anti-PF4 antibodies detected by PF4-ELISA in healthy vaccinees (3.7%, 95% CI 2.1%–5.6%) is low and comparable to the reported frequency in healthy blood donors (6.6%, 95% CI 5.8%–7.4%).8 The positivity of potentially pathogenic auto-antibodies is a common finding in healthy subjects. In the setting of VITT, PAAs can help identify pathogenic anti-PF4 antibodies. In our systematic review, none of the anti-PF4 antibodies detected by ELISA displayed platelet-activating properties. In addition, whereas most VITT cases have PF4-ELISA ODs >2.0,4 ODs in healthy vaccinees were below 2.0 in 91 out of 92 positive results (99.0%). Despite the low number of tested subjects, our results point towards high PAA and PF4-PAA specificity, which supports their use as confirmation tests, where feasible. It is likely that PF4-ELISA positive results merely represent the consequence of seroconversion following vaccination or, perhaps more likely, reflect a pre-existing condition, considering that their prevalence is similar to that of blood donors.8 The calculated prevalence was lower than that in acutely ill patients exposed to heparin in whom HIT had been ruled out, which reaches 16.9% in patients with sepsis11 and 50% during or after cardiac surgery.9 Our findings are important in light of the uncertain duration of the COVID-19 pandemic, with many developing countries using mainly adenoviral vector vaccines due to their favourable storage profile,13 and of the prospective employment of this vaccine technology for communicable diseases other than COVID-19. Laboratory test results from healthy vaccinees point to a high specificity of VITT diagnostic tests and might be useful for the future implementation of evidence-based surveillance programs. This study has some limitations. Firstly, due to the small rate of seroconversion in healthy vaccinees, few PAAs and PF4-PAAs were performed, resulting in wide estimate CIs. In addition, the possibility of false negatives results obtained with PAAs and PF4-PAAs cannot be excluded. Moreover, in observational studies designed to determine the frequency of anti-PF4 antibodies in healthy vaccinees there was no declared clinicopathologic definition for the identification of VITT cases with the exception of the study by Uzun et al.14 In conclusion, our results show that the prevalence of anti-PF4 antibodies detected by PF4-ELISAs in healthy vaccinees is low and comparable to that of the general unvaccinated population; detected anti-PF4 antibodies did not activate platelets. Thus, our data support current VITT diagnostic strategies. Gian Marco Podda and Simone Birocchi designed the study, supervised data collection and analysis and critically revised the manuscript. Bianca Clerici wrote the manuscript and participated in study design, data extraction and analysis. Mariangela Scavone contributed to manuscript writing, data extraction and analysis and critically reviewed the intellectual content. Benedetta Peroni and Chiara Aloise performed the literature search under the supervision of Simone Birocchi. Benedetta Peroni, Chiara Aloise, Claudia Ghali and Alessandra Negrini performed data extraction and contributed to data analysis. Tables and Figures were prepared by Bianca Clerici, Mariangela Scavone, Benedetta Peroni, Chiara Aloise and Alessandra Negrini. Giovanni Casazza supervised the statistical approach and critically reviewed the manuscript. We thank professor Marco Cattaneo for critically reviewing the manuscript. This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors. Nothing to declare. Figure S1. Table S1. Data S1. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.