Abstract
Objective IgA vasculitis (IgAV) can occur after vaccination. We aimed to assess a potential safety signal on the association between coronavirus disease 2019 (COVID-19) vaccines and IgAV.
Methods Cases of IgAV involving COVID-19 vaccines were retrieved in VigiBase. Disproportionate reporting was assessed using the Bayesian information component (IC) with all other drugs and vaccines as control groups.
Results Three hundred thirty patients with de novo IgAV from 24 countries were included, mostly from the United States (193/330, 58%). Fifty percent (163/328) were female and median age was 32 years (IQR 15-59), of which 33% (84/254) were young (1-17 yrs). Median time to onset of IgAV was 7 days (IQR 2-16; n = 256) and 85% (280/330) of patients were vaccinated with mRNA vaccines. Seriousness was reported in 188/324 (58%) cases. Sixty-five percent (95/147) recovered and 1% (2/147) died. A positive rechallenge was reported for 3 of 4 patients (75%). A total of 996 cases of IgAV were identified with other vaccines. There was a small significant increase in IgAV reporting with COVID-19 vaccines compared with all other drugs (IC 0.22, 95% credible interval [CrI] 0.04 to 0.35). No disproportionality signal was found between COVID-19 vaccines and other vaccines (IC −1.42, 95% CrI −1.60 to −1.28). There was no significant difference between mRNA vaccines and viral vector COVID-19 vaccines. Men and children had a significant overreporting of IgAV compared with women and adults, respectively.
Conclusion This study provides reassuring results regarding the safety of COVID-19 vaccines in the occurrence of IgAV compared to other vaccines.
IgA vasculitis (IgAV) is an immune complex vasculitis1 affecting mainly small-size vessels.2 IgAV can occur after respiratory or digestive mucosa infections but can also be induced by vaccines such as the influenza vaccine or diphtheria tetanus pertussis vaccine.3-5
The coronavirus disease 2019 (COVID-19) pandemic triggered a worldwide vaccination campaign using mainly a new mRNA-based technology,6 followed by several reports of related autoimmune manifestations.7,8 Cases of vasculitis, including newly developed or relapsed IgAV, have also been reported.8-11 Although these could be incidental findings, they cannot be ruled out. The World Health Organization (WHO) pharmacovigilance database, VigiBase, gathers individual case safety reports (ICSRs) of adverse drug reactions (ADRs) reported through national pharmacovigilance systems from over 150 countries worldwide since 1968. Using VigiBase, we aimed to assess a potential safety signal on the association between COVID-19 vaccines and IgAV.
METHODS
Population. In this observational disproportionality study, all ICSRs reported as de novo IgAV cases involving COVID-19 vaccines (elasomeran, tozinameran, ChAdOx1 nCoV-19, and JNJ-Ad26.COV2.S) until June 1, 2022, were retrieved in VigiBase using the previously validated Medical Dictionary for Regulatory Activities (MedDRA) Lowest Level Terms (LLTs).12 We included all patients corresponding to the following MedDRA LLTs: rheumatic purpura, Henoch-Shönlein purpura, Henoch-Schönlein, Henoch-Schönlein purpura, Schönlein-Henoch purpura, vasculitis Henoch-Schönlein like, IgA vasculitis, and IgA-associated vasculitis. De novo IgAV cases were identified by excluding cases with an underlying worsening of IgAV co-reported using the MedDRA Preferred Terms (PTs) “condition aggravated” and “concomitant disease aggravated.”
For each ICSR, extracted data included administrative information (date, country, reporter qualification), patient demographics (age, sex), exposure characteristics (drug name, rechallenge), and IgAV features (time to onset, seriousness, outcome, positive rechallenge [ie, recurrence of IgAV after an additional COVID-19 vaccine]). The time to onset was computed as the time in days between the last available shot date and the date of IgAV onset.
As per the Jardé law in France regarding research involving human participants, no ethical review or informed consent was required because an existing anonymized database was used.
Statistical analysis. Disproportionate reporting was assessed using the Bayesian information component (IC), which displays the best sensitivity and specificity among disproportionality methods,13 with all other drugs and all other vaccines (ie, ATC J07A-X) as control groups. Only drugs characterized as suspect or interacting according to the WHO (ie, the drugs that had the most likely causal relationship with the ADR) definition were considered. Pharmacovigilance disproportionality analysis is considered a relevant tool for identifying new drug ADR-associated potential safety signals, using spontaneous reports as material.14 The concept of disproportionality is based on the comparison of the reporting proportions between the study drug (eg, COVID-19 vaccines) and the comparator (eg, all other drugs, all other vaccines) in a spontaneous reporting database. In other words, this study design is intended to highlight the existence or absence of an increase (disproportionality) in the number of observed cases as compared to the expected number of cases, where the overall reporting rate of the ADR of interest for the comparator is used as a proxy for the background occurrence of the ADR.15 The Table describes the contingency table of the disproportionality study and the formulas for estimating the IC and its 95% credible interval (CrI). A signal of disproportionate reporting of IgAV was defined as an IC with a 95% CrI lower boundary that exceeded 0.16
Contingency table for the case-noncase analysis and operative definitions of the IC used as measure of disproportionality.
Several sensitivity analyses were performed, including (1) restriction to IgAV cases using solely the drug with the higher imputability score to calculate the IC, (2) restriction to IgAV cases that occurred within 30 days after vaccine administration to increase the likelihood that signals were true ADRs, and (3) exclusion of vaccine-associated ICSRs with reactogenicity events (identified using the MedDRA PTs “headache” and “pyrexia”) to limit the risk of a competition bias between events that may mask a true safety signal.14
The statistical analysis was performed with R, version 4.1.3 (R Foundation).
RESULTS
Of the 31,738,658 ICSRs in VigiBase retrieved until June 1, 2022, 3,712,164 were related to COVID-19 vaccines, including 330 de novo IgAV cases from 24 countries, of which the 3 largest contributors were the United States (193/330, 58%), the United Kingdom (31/330, 9%), and France (29/330, 9%). Among these patients, 50% were female (163/328) and the median age was 32 years (IQR 15-59), of which 17% (42/254) were children (1-12 yrs), 17% (42/254) were adolescent (12-17 yrs), 48% (122/254) were adult (18-65 yrs), and 19% (48/254) were elderly (> 65 yrs). The median time from vaccination to onset of IgAV was 7 days (IQR 2-16; n = 256) and 85% (280/330) of patients were vaccinated with mRNA vaccines.
Seriousness as defined by the WHO criteria12 was reported in 188/324 (58%) cases. Of the 147 cases where disease outcome was available at the time of reporting, 95 recovered (65%), of which 1 had sequelae, and 2 (1%) died. A positive rechallenge was reported for 3 of 4 patients (75%) who received a new dose of COVID-19 vaccine.
A total of 2093 cases of IgAV were identified with other drugs than COVID-19 vaccines, including 996 with other vaccines (Figure). There was a small significant increase in IgAV cases reporting with COVID-19 vaccines compared with all other drugs (IC 0.22, 95% CrI 0.04 to 0.35). No disproportionality signal was found when COVID-19 vaccines were compared with other vaccines (IC −1.42, 95% CrI −1.60 to −1.28). The results were consistent in all sensitivity analyses, in which no significant overreporting of IgAV was found compared with other vaccines (Figure).
Forest plot of the IC values of COVID-19 vaccine-associated de novo IgAV vs all other drugs and all other vaccines and subgroup analyses by vaccine type, age, and sex. COVID-19 vaccines were mRNA vaccines (elasomeran, tozinameran) or viral vector vaccines (ChAdOx1 nCoV-19, Ad26.COV2.S). The definition of IgAV corresponds to the following MedDRA LLTs: Rheumatic purpura, Henoch-Shönlein purpura, Henoch-Schönlein, Henoch-Schönlein purpura, Schönlein-Henoch purpura, vasculitis Henoch-Schönlein like, IgA vasculitis, and IgA-associated vasculitis. De novo IgAV cases were identified by excluding cases with an underlying worsening of IgAV co-reported using the following MedDRA PTs: condition aggravated and concomitant disease aggravated. The excluded reactogenicity events correspond to the 2 most reported for COVID-19 vaccines: headache and pyrexia (MedDRA PTs). COVID-19: coronavirus disease 2019; CrI: credibility interval; IC: information component; IgAV: IgA vasculitis; LLT: Lowest Level Term; MedDRA: Medical Dictionary for Drug Regulatory Activities; PT: Preferred Term; TTO: time to onset.
In our subgroup analyses, no significant difference was found in IgAV reporting between mRNA vaccines and with viral vector COVID-19 vaccines even after stratification on age. Among patients exposed to COVID-19 vaccines, men and children had a significant overreporting of IgAV compared with women and adults, respectively (Figure).
DISCUSSION
Altogether, our results showed that COVID-19 vaccines are associated with a slight but significant overreporting of IgAV, but not greater than that observed with other vaccines. Of note, both sex ratio and mean age of IgAV following COVID-19 were roughly comparable to those induced by viral conditions or drugs.12,17 Despite the small number of cases considered, the existence of positive rechallenge with IgAV relapse in this study, and the demographic characteristics of the reported cases, support the hypothesis of a triggering effect of COVID-19 vaccines. To explain this possible effect, one of our hypotheses is a vasculitis-driven hypoglycosylated IgA-mediated response following vaccination. Indeed, IgA responses have been observed after BnT162B2 and ChAdOx1 vaccination.18,19 Moreover, IgA is known to play a pivotal role during the onset of IgAV forming immune complexes, which further activate the immune system.20
We also confirmed the results of a recent study that compared the reporting of systemic vasculitis (including relapses) associated with COVID-19 mRNA vaccines vs influenza vaccines,21 extending it to other vaccines. Our results should be viewed in light of the known limitations of the spontaneous reported cases. First, underreporting is frequent and, therefore, the selective nature of spontaneous reporting may result in a lack of representativeness. Second, some important data were missing (eg, outcomes) and cases do not contain sufficient clinical data to ascertain the IgAV diagnosis by external review with the use of this database. This may have affected the results of the disproportionality analysis under the assumption of differential reporting quality between COVID-19 vaccines and the other drugs used as comparators. Nevertheless, to increase the likelihood that the potential signals were true ADRs with consistent results, we included in sensitivity analyses only patients who developed IgAV within 30 days of COVID-19 vaccination and only ICSRs where vaccines were the sole suspected drugs in the occurrence of IgAV. Further studies are needed both to confirm or refute these results on the association between COVID-19 vaccination and IgAV using other data sources. Comparing a cohort of IgAV induced by COVID-19 vaccines to IgAV without any previous vaccination could have been interesting, but precise clinical data of IgAV following COVID-19 vaccination are lacking to perform such comparison.
COVID-19 vaccines, like other vaccines, could serve as a trigger for the expression of a first episode of IgAV in predisposed patients. This study, however, provides reassuring results regarding the safety of COVID-19 vaccines in the occurrence of IgAV compared to other vaccines.
Footnotes
Y. Ramdani and B. Largeau contributed equally to this work.
The authors declare no conflicts of interest relevant to this article.
- Accepted for publication October 27, 2022.
- Copyright © 2023 by the Journal of Rheumatology
This is an Open Access article, which permits use, distribution, and reproduction, without modification, provided the original article is correctly cited and is not used for commercial purposes.
REFERENCES
DATA AVAILABILITY
We thank VigiBase for giving us access to the data. The data supplied to VigiBase come from a variety of sources, and the likelihood of a causal relationship is not the same in all reports. The information does not represent the opinions of the Uppsala Monitoring Centre or the WHO. The data underlying this article are available in VigiBase at https://www.who-umc.org/vigibase/vigibase/, and can be accessed with a personal identifier, as part of the network of pharmacovigilance centers and the WHO Programme for International Drug Monitoring.
