Reporting of Acute Inflammatory Neuropathies with COVID-19 Vaccines: Subgroup Disproportionality Analyses in VigiBase

Since marketing authorization, cases of neuralgic amyotrophy (NA), facial paralysis/Bell’s palsy (FP/BP), and Guillain-Barré syndrome (GBS) were reported with COVID-19 vaccines of different technologies. This study aimed to assess whether NA, FP/BP, and GBS were more frequently reported in VigiBase with COVID-19 vaccines (of any technologies) than with other viral vaccines, over the full database and across potential risk groups by sex and age. The reporting odds ratio (ROR) with 95% confidence interval (95% CI) was used as the measure of disproportionality and subgroup disproportionality analyses were performed by sex and age. Out of 808,906 safety reports with COVID-19 vaccines, 57 (0.01%) reported NA, 3320 (0.4%) FP/BP, and 632 (0.1%) GBS. There were not signals of disproportionate reporting for NA and GBS with COVID-19 vaccines against other viral vaccines. FP/BP was disproportionately more frequently reported with COVID-19 vaccines than with other viral vaccines over the full database (ROR 1.12, 95%CI 1.07–1.17), in males (ROR 1.65, 95%CI 1.54–1.78) and in age subgroups 65–74 years (ROR 1.21, 95%CI 1.05–1.39) and ≥75 years (ROR 1.84, 95%CI 1.52–2.22). Albeit not proving causation, these findings might support clinicians in decision-making for patients potentially at risk for developing an acute inflammatory neuropathy with COVID-19 vaccines.

In pharmacovigilance, disproportionality analysis by subgroups represents a useful methodological approach to uncover risk groups for ADRs [13]. Notably, in large international spontaneous databases, such as VigiBase, subgroup disproportionality analysis performs better than crude analysis (in the full database) and stratified analysis, in both sensitivity and precision [14].
The present study performed in VigiBase aimed to assess whether NA, FP/BP, and GBS were disproportionately more frequently reported with COVID-19 vaccines (of any technologies) than with other viral vaccines (overall and restricted to influenza vaccines), over the full database and across potential risk subgroups by sex and age.

Materials and Methods
De-duplicated safety reports of COVID-19 vaccine-related NA, FP/BP, and GBS, collected in VigiBase as of 16 May 2021, were retrieved. COVID-19 vaccines were selected based on the WHO drug standardized drug grouping "Vaccines for COVID-19" that includes COVID-19 vaccines of any technologies; NA, FP/BP, and GBS ADRs were selected based on the correspondent preferred terms according to the Medical Dictionary for Regulatory Activities (MedDRA, version 24.0, https://www.meddra.org/ accessed on 13 September 2021).
The baseline characteristics of COVID-19 vaccine-related safety reports of NA, FP/BP, and GBS included patient sex and age, reporter qualification, suspected COVID-19 vaccine, time-to-onset, and outcome of the acute inflammatory neuropathy of interest.
In disproportionality analyses, either safety reports concerning NA, FP/BP, and GBS, or other ADRs, reported with any other viral vaccines (of the anatomical therapeutic chemical group, ATC, J07B-excluding COVID-19 vaccines), were used as comparators groups. The reporting odds ratio (ROR) with the 95% confidence interval (95% CI) was used as measure of disproportionality and computed when ≥5 safety reports with the acute inflammatory neuropathy of interest were present. Signals of disproportionate reporting were characterized by 95% CI lower bound >1.
Subgroup disproportionality analyses were performed by sex and age (<18 years, 18-44 years, 45-64 years, 65-74 years, ≥75 years). Safety reports concerning NA, FP/BP, and GBS, or other ADRs, reported with influenza vaccines (ATC J07BB) were used as comparator groups in sensitivity disproportionality analyses performed both on the full database and across risk subgroups. In subgroup disproportionality analyses, RORs were calculated within each group separately and a signal was counted if the score from any subgroups met signal criteria. Safety reports with missing values for the two variables used to define subgroups were excluded from the analyses.
Data management and analysis were carried out by Microsoft Excel (2010, Microsoft Corporation, Washington, USA) and GraphPad Prism 9 (GraphPad Software Inc., San Diego, California, USA).
According to the Swiss Human Research Act 810.30 (status as of 26 May 2021), ethical approval by the local Ethical Committee was not required for this study as it involved anonymized health-related data.

Discussion
By querying VigiBase, this study found that the reporting frequency of NA and GBS with COVID-19 vaccines of different technologies was similar to that of other viral vaccines, and neither patient sex nor age were risk factors. Conversely, a signal of disproportionate reporting was observed for FP/BP with COVID-19 vaccines in the analysis on the full database. Subgroup disproportionality analyses highlighted male sex and advanced patient age (≥65 years) as risk factors for the reporting of FP/BP with COVID-19 vaccines against other viral vaccines. Notably, the safety signals for COVID-19 vaccine-related FP/BP detected in advanced age subgroups were confirmed in the sensitivity analysis against influenza vaccines.
The reporting frequency of FP/BP with COVID-19 vaccines in VigiBase was uncommon, those of NA and GBS were rare, reflecting the background incidence rates of these three acute inflammatory neuropathies [1][2][3]. Notably, most of the safety reports of NA, FP/BP and GBS were reported with the Pfizer BioNTech COVID-19 vaccine, the COVID-19 Vaccine Moderna, and the COVID-19 Vaccine AstraZeneca, whilst less frequently were associated with the Janssen COVID-19 vaccine, for which, as for the COVID-19 Vaccine AstraZeneca, rather an association with unusual blood clot formation was reported [15].
Because of the new technology, the use by a large number of individuals in a brief period, and the recommendation of an additional dose at least for immunocompromised patients, concerns about the safety of mRNA COVID-19 vaccines have emerged. Cases of FP/BP were reported with mRNA COVID-19 vaccines but have also occurred with COVID-19 vaccines of different technologies [6][7][8]. Indeed, the biological mechanism linking FP/BP onset and COVID-19 vaccines is likely related to immune activation and independent on vaccine technology [16].
Therefore, by broadening the spectrum of safety reports of interest to those with COVID-19 vaccines (of all technologies) suspected of being associated with FP/BP onset, the present study controverts the one previously performed in VigiBase by Renoud et al., which found that the reporting rate of FP/BP after mRNA COVID-19 vaccines was not higher than that observed with other viral vaccines [12]. In that study, authors used stratification to control confounding factors (including sex and age) and, assuming the absence of risk variation across strata, provided a single combined (not significant) disproportionality measure [12]. Nevertheless, ignoring the diversity within a dataset may result in signals being masked. When a confounding factor acts as an effect modifier, subgroup disproportionality analysis can highlight risk variation across subgroups by computing a disproportionately measure within each subgroup [13]. Consistently, the present study uncovered that reporting of FP/BP with COVID-19 vaccines was disproportionately more frequent in males and in patients aged ≥65 years. As passive surveillance system, VigiBase suffers of underreporting and selective reporting, lack of clinical details, missing and/or inaccurate data. Safety reports reflect suspicions without being necessarily indicative of a causal relationship between the suspected drug/vaccine and the reported adverse event. Moreover, lacking information on the number of patients exposed to a certain drug/vaccine, VigiBase does not allow defining the incidence of adverse events.

Conclusions
Subgroup disproportionality analyses in VigiBase on safety reports of NA, FP/BP, and GBS with COVID-19 vaccines did not detect any signals for NA and GBS, whereas highlighted that FP/BP was disproportionately more frequently reported with COVID-19 vaccines than with other viral vaccines. Moreover, male sex and advanced age (≥65 years) were risk factors for increased reporting of FP/BP with COVID-19 vaccines.
Awareness of these findings might support clinicians in decision making when confronted with patients who are potentially at risk for developing an acute inflammatory neuropathy following COVID-19 vaccine administration.