Yonglong Su, Yirong Huang, Lili Pan, Honghong Fu, Xun Wang
Multiple vaccines showed disproportionate GBS reporting signals with delayed temporal patterns consistent with biological plausibility; because these derive from passive surveillance, they are hypothesis-generating and do not establish causality. Despite rare associations, absolute risks (1-9 per million doses) are far outweighed by disease prevention benefits, supporting current vaccination policies with ongoing surveillance.
BACKGROUND: Guillain-Barré syndrome (GBS) remains a critical vaccine safety concern. We conducted a comprehensive pharmacovigilance analysis of 21-year VAERS data to identify vaccine-GBS disproportionality signals and temporal patterns.
METHODS: We analyzed 9,755,435 VAERS reports (2004-2024) including 2,216 GBS cases using four disproportionality algorithms (ROR, PRR, IC, EBGM). Time-to-onset was compared between 1,534 GBS and 120,763 non-GBS events using Kaplan-Meier analysis.
RESULTS: Twenty-six vaccines met all four algorithm criteria, predominantly influenza formulations (19/26; RORs 2.4-8.5), validating historical surveillance observations. Novel signals emerged for both RSV vaccines (Arexvy ROR 3.01, Abrysvo ROR 6.82) and zoster vaccine (ROR 3.01), consistent with post-marketing estimates of 3-9 excess cases per million doses. COVID-19 vaccines showed no concordant signals despite established Janssen associations, reflecting denominator dilution from 7.89 million COVID-19 reports (81% of the database). GBS exhibited characteristic delayed onset-median 13 days (IQR 5-29) versus 0 days for non-GBS events (p < 0.001)-with 44.8% of cases occurring 8-30 days post-vaccination, the biologically plausible window for vaccine-triggered autoimmunity. Demographically, GBS reports showed male predominance (OR 2.4) and were overwhelmingly classified as serious outcomes compared with non-GBS reports (76.1% vs. 9.6%, p < 0.001); serious outcomes included death, life-threatening events, hospitalization (new or prolonged), or permanent disability.
CONCLUSION: Multiple vaccines showed disproportionate GBS reporting signals with delayed temporal patterns consistent with biological plausibility; because these derive from passive surveillance, they are hypothesis-generating and do not establish causality. Despite rare associations, absolute risks (1-9 per million doses) are far outweighed by disease prevention benefits, supporting current vaccination policies with ongoing surveillance.