Postmarketing reporting profile of mirikizumab in ulcerative colitis: a disproportionality analysis of FAERS reports.

Abstract

Purpose: Mirikizumab was approved for ulcerative colitis (UC) in 2023, and its early postmarketing reporting pattern remains incompletely characterized. We described indication-restricted reports in the US Food and Drug Administration Adverse Event Reporting System (FAERS) and evaluated signals of disproportionate reporting (SDRs), while distinguishing established reactions from sparse, hypothesis-generating observations.

Methods: Standardized FAERS data were queried for 2023Q1 through 2025Q4. Reports containing an exact mirikizumab or OMVOH name mapping as the primary-suspect drug and a UC indication term were included. System organ class (SOC) frequencies were counted once per report; preferred-term (PT) disproportionality used 299,281 UC reaction records as the disease-specific background. The reporting odds ratio (ROR) was the primary measure; proportional reporting ratio, empirical Bayes geometric mean, and information component were supportive measures. Bonferroni correction covered 338 PTs. Exploratory report-level subgroup comparisons and time-to-onset (TTO) analyses were also performed.

Findings: The cohort contained 408 reports; 402 (98.5%) were submitted in 2024 to 2025, 377 (92.4%) were from consumers, and 223 (54.7%) were classified as serious. The leading SOCs were general disorders and administration site conditions (138 reports) and gastrointestinal disorders (110). Injection site pain had the largest count and ROR among the displayed PTs (n = 46; ROR = 29.47; 95% CI, 21.66-40.10). Six of the 15 displayed PTs had Bonferroni-adjusted P < 0.05: injection site pain, maternal exposure during pregnancy, injection site bruising, pulmonary embolism, spontaneous abortion, and flushing. Most nonadministration-site findings were based on fewer than 10 reports. No age- or sex-based PT difference survived multiplicity correction; injection site pain and arthralgia differed by report seriousness. TTO was evaluable in 83/408 reports (20.3%); the median was 34 days (IQR, 12-68.5), and 71/83 (85.5%) occurred within 90 days.

Implications: Injection site pain, arthralgia, headache, and rash were concordant with labeled or trial-observed reactions. Venous thromboembolic and pregnancy-related findings were sparse and confounded; maternal exposure during pregnancy is an exposure circumstance, not an adverse event. These SDRs cannot establish causality, incidence, or comparative safety. The TTO pattern applies only to reports with complete dates and a positive interval. The findings warrant continued surveillance and validation in clinically adjudicated, denominator-based studies.