Details
The U.S. Food and Drug Administration (FDA) used an Active Postmarket Risk Identification and Analysis (ARIA) system to address a safety issue (serious infection) identified during clinical review of an application to market ustekinumab as a treatment for Crohn's disease (CD). FDA used an active-comparator new-user cohort design, data from six Sentinel Data Partners, a method developed in Sentinel for identifying serious infection, and ARIA analytic tools to estimate serious infection incidence in adult CD patients during treatment with ustekinumab or an active comparator (infliximab, adalimumab, or vedolizumab). Inverse probability of treatment weighting (IPTW) was used to control for baseline differences between treatment groups. Analyses compared ustekinumab and comparator cohorts with 15,490 and 51,503 patients, respectively. The primary composite outcome of serious infection or COVID-19 occurred in 747 ustekinumab patients (4.82 %; 44.1 per 1000 patient-years) during mean 400-day follow-up. IPTW proportional hazard (Cox) regression estimated risk in the ustekinumab cohort vs. active comparator with hazard ratio 0.88 (95% confidence interval 0.80–0.96). Findings were similar (a) according to line of therapy (use or non-use of an advanced inflammatory bowel disease therapy during a 183-day baseline period) and (b) before and after the start of the COVID-19 pandemic. Additional analyses excluded large magnitude risks from ustekinumab for major types of serious infection. In conclusion, IPTW analyses excluded large magnitude risks from ustekinumab for serious infection. This report describes the first ARIA study to use complex confounding adjustment to help resolve a safety issue identified during FDA review of a marketing application.