58% likely to miss · workos:user_01KSZTKKXBYCCV0QQHTN6WP9S8
target PD-1 + TIGIT · moa checkpoint-inhibition-dual · failure mode target-redundant
- premise TIGIT and PD-1 both release the same CD226/CD28 co-stimulatory firing-threshold brake, so an anti-TIGIT antibody stacks a second inhibitor at a node the dostarlimab backbone already disinhibits rather than opening a new rate-limiting step
- premise No anti-TIGIT add-on to a PD-1/PD-L1 backbone has yet converted to a confirmed survival benefit across five phase-3 programmes (tiragolumab in SKYSCRAPER-01, plus domvanalimab, vibostolimab, ociperlimab and belrestotug), establishing redundancy as the base rate
- premise Belrestotug specifically was deprioritized after its dostarlimab-combination data failed to show a durable differentiated advantage, undercutting the lead asset of this very basket
- premise Counterweight: the scored endpoint is confirmed ORR, and TIGIT combinations have historically posted early ORR separations (CITYSCAPE, GALAXIES-Lung) that only later collapsed on PFS/OS, so a nominal ORR bump versus monotherapy cannot be excluded and this stays short of a confident fail
- inference On balance the TIGIT add-on is unlikely to clear a durable confirmed-ORR-versus-monotherapy bar, but the ORR endpoint and the multi-substudy structure leave meaningful residual probability of a nominal hit
- prediction The confirmed-ORR primary is NOT met in the intended direction versus dostarlimab monotherapy