Journal analysis: phase 2 combination missed its primary endpoint overall and hit it in a 41% subgroup
The oncology replacement is being bought with an upfront payment, and the first replacement readout missed
The replacement franchise is being assembled by purchase rather than discovery, which means the diligence happened on someone else's timeline and the downside is already priced into milestone structures rather than into share count. A 41 percent biomarker-defined subgroup is both the reason a phase 3 could still be worth running and the ceiling on the revenue that phase 3 can address, so the same fact is doing both jobs. The paragraph certifications matter because they start a 30-month clock against the 2031 and 2032 formulation patents rather than the 2043 dosing claims, which means the exclusivity the purchases are meant to protect is layered rather than solid.
A peer-reviewed analysis published this week of Pfizer's phase 2 study of a combination regimen in advanced breast cancer found the primary endpoint was missed in the intention-to-treat population, with 71 events reached against a prespecified target of 78. The regimen did meet the primary endpoint in a biomarker-defined subgroup representing 41 percent of enrolled patients. The authors, from two academic centres, wrote that the overall result was driven by a subgroup with no measurable target expression and that the study was not designed to establish the subgroup effect. Pfizer funded the trial and two of the six authors disclose consultancy fees from the company. The paper states that a phase 3 decision is expected in the first half of 2027 and that the sponsor has not disclosed whether the study will proceed as designed. An accompanying editorial argues that the subgroup result should not be read as a rescue and that the confirmatory design should follow the biomarker rather than the population.
A 71 versus 78 event shortfall in the overall population is a miss by design margin, and a 41 percent biomarker subgroup that hit the endpoint is the only path to a phase 3 that is worth running. The sponsor's plan is the open question, and it is described as undisclosed.
The positive number lives in four people out of ten, and the paper that says so also says the study was never built to prove it. An editorial arguing against a rescue was scheduled alongside the analysis. That is not an accident of the journal.
- Primary endpoint missed in the intention-to-treat population at 71 events against 78 prespecified
- Endpoint met in a biomarker-defined subgroup representing 41 percent of enrolled patients
- Phase 3 decision expected in the first half of 2027 with the sponsor's plan not disclosed
- Pfizer funded the trial and two of six authors disclose consultancy fees