| アブストラクト | OBJECTIVES: Posterior reversible encephalopathy syndrome (PRES) is a rare but potentially severe complication of VEGF-pathway inhibition. A recent FAERS analysis supported PRES as a probable class effect of angiogenesis inhibitors but found no association with continuous VEGFR affinity, did not formally test a categorical VEGFR-2 selectivity contrast, and could not characterise time to onset. Complementing that work, we compared disproportionality signals across 22 VEGFi/VEGFRi agents and examined whether signal ranking tracked VEGFR-2 selectivity, systemic exposure, and published case-report volume. METHODS: In this observational pharmacovigilance study we analysed FAERS reports (2010-Q1 to 2026-Q1) via the openFDA API. PRES was ascertained by MedDRA Preferred Terms. Four disproportionality metrics (ROR, PRR, BCPNN-IC with IC025, EBGM) were computed for each agent; robust signals were confirmed by Bonferroni and Benjamini-Hochberg correction. We characterised the drug-start-to-FDA-receipt reporting interval (a reporting-delay measure, not true clinical time-to-onset), explored the correlation between published VEGFR-2 IC50 and log-ROR, and performed a targeted PubMed scan against the case-report literature. RESULTS: Under formal multiplicity correction (Bonferroni and Benjamini-Hochberg), 13 of 16 evaluable agents produced robust elevated PRES signals (ROR up to 16.68). Signal intensity separated cleanly by VEGFR-2 selectivity: every highly selective VEGFR-TKI (tivozanib, lenvatinib, fruquintinib, axitinib; ROR 9.74-16.68) ranked above every multi-kinase agent (ROR 2.00-6.05) (Mann-Whitney p = 0.005; Spearman rho = 0.84, p = 0.001). By contrast, the continuous IC50-ROR correlation was non-significant primarily (rho = -0.38, p = 0.25) and exclusion-sensitive, hence hypothesis-generating only. Intravitreal anti-VEGF agents produced no positive signal, an internal negative control. Nintedanib yielded an inverse signal (ROR 0.34), likely indication channelling. Tivozanib and ramucirumab (ROR 7.56) were under-represented in the PubMed case-report literature (FAERS-PubMed rho = 0.09, p = 0.78). CONCLUSION: PRES disproportionality varied substantially across VEGFi/VEGFRi agents, with intravitreal agents serving as an internal negative control consistent with a systemic-exposure mechanism. IC50-based analyses were small-sample and hypothesis-generating only. Tivozanib and ramucirumab emerged as under-represented signals meriting confirmation; selective VEGFR-TKI recipients may warrant close blood-pressure monitoring and a low threshold for neuroimaging. |
|---|
