Does NCT03851614 meet its primary endpoint at primary completion?

75% likely to miss · workos:user_01KSZTKKXBYCCV0QQHTN6WP9S8

target PARP + PD-L1 + VEGFR · moa checkpoint-inhibition+PARP+anti-angiogenic · failure mode egress-barrier-no-licensing-lever

1. premise In MMR-proficient colorectal cancer the tumour-specific CD8 response is characteristically built and parked at the egress/licensing step and carries little PD-1, so PD-(L)1 blockade has nothing to release; MMRp tumours respond to checkpoint blockade at under 5% versus near 50% in MMR-deficient disease doi:10.5281/zenodo.20415561 nct:NCT02563002 pmid:33264544
2. premise Single-cell data (Pelka et al., Cell 2021) place the structural difference between MMR-proficient and MMR-deficient colorectal cancer at the egress position, with the parked-state odds ratio 4.64, rather than at the synapse where PD-1 blockade acts
3. premise This regimen adds olaparib (PARP inhibition) and cediranib (VEGFR inhibition) to durvalumab; neither supplies the CD40-mediated dendritic-cell licensing handshake the parked cells await, and the population is unselected for the parked state
4. inference Acting downstream of and beside the actual bottleneck in an unenriched MMR-proficient population, the regimen leaves the rate-limiting licensing step untouched
5. prediction NCT03851614 does NOT meet its primary endpoint in the intended direction

Evidence: doi:10.5281/zenodo.20415561, nct:NCT02563002, pmid:33264544