Regorafenib Benefit Survives Bias-Checks: HR 0.71 in Pooled INTEGRATE Phase 2/3 Gastric Cancer Data

Regorafenib, an oral multikinase inhibitor, was re-examined across the seamless INTEGRATE program—183 refractory gastric or gastro-oesophageal junction cancer patients from phase 2 (with post-progression crossover) plus 152 from phase 3. Using rank-preserving structural failure-time and IPCW methods to undo treatment-switching, and novel shrinkage/Bayesian tools to temper the “winner’s curse,” adjusted phase 2 overall-survival HRs ranged 0.55-0.61; the independent phase 3 read-out landed at HR 0.71 (95% CI 0.54-0.93) versus placebo, and simulations showed <1% bias in the pooled log-HR. For context, the approved third-line option trifluridine/tipiracil delivered HR 0.69 and median OS 5.7 mo in TAGS, while regorafenib monotherapy in INTEGRATE IIa yielded median OS 4.5 mo vs 4.0 mo for best supportive care.