The Phase 2
Problem
Most Phase 2 failures aren't biology failures. They're patient selection failures. And by the time you know, it's too late.
The most expensive word in clinical development isn't "failed." It's "enrichable." When a Phase 2 trial returns a null result, the post-hoc conversation almost always goes the same way: the biology was real, but we enrolled the wrong patients. A biomarker was available — in the literature, in your preclinical data, buried in a collaborator's dataset — and the trial wasn't designed to test it. The program is effectively over. The asset isn't.
The enrichment gap.
An analysis of Phase 2 oncology failures over the last decade shows that approximately 60–70% of programs that failed in unselected populations showed meaningful activity in biomarker-defined subgroups identified post-hoc. The signal was there. The trial wasn't designed to find it.
The biomarker problem isn't a scientific problem — it's an analytical and decision-making problem. The literature on biomarker associations with your mechanism almost always exists before you write your protocol. The question is whether your team has the infrastructure to synthesize it before you lock the design.
The biomarker wasn't discovered in the post-hoc analysis. It was already in the literature. What failed was the synthesis — connecting the mechanism evidence to the trial design before the protocol was locked.
The protocol lock problem.
Clinical program teams are making biomarker decisions under extraordinary time pressure. The IND is filed. The sites are ready. The board wants the readout timeline. Under that pressure, the question of whether to add a biomarker stratification — which requires additional assay validation, patient consent language, and potentially delays enrollment — often loses to expediency.
The decision calculus isn't wrong. It's made with incomplete information. The scientific case for the biomarker hypothesis, across the full literature and your internal data, is rarely synthesized at the moment of protocol design. It gets reviewed in pieces: the CMO knows one study, the translational lead knows another, the consultants produce a summary that's six months old by the time it arrives.
Protocol design is where programs are won or lost. Gauge Labs gives your clinical and translational teams a connected view of the biomarker evidence — before the protocol is locked.
Discuss your program