But how did this one PHARE?

PHARE was a very similar French trial designed to prove the non-inferiority of disease-free survival (DFS) at 2 years (within a relative difference of 15%). Different from PERSEPHONE, this trial actually ran into a negative efficacy signal part-way through. As a result, around 3400 women were randomized. Again, roughly 55% had N0 disease and another 30% had 1-3 involved nodes. Again, most (58%) also has ER(+) tumors. Final results demonstrate that 6 months of trastuzumab could not be proven non-inferior to 12 months. While the DFS hazard ratio for the two trials was 1.08 (PHARE) and 1.07 (PERSEPHONE), they resulted in very different conclusions based on the approach to non-inferiority margin assumptions. Just to throw this out there, radiation doesn’t even appear in the heterogeneity analyses. We’ve seen these tiny differences in DFS before in the radiation world, which highlights how a few events here or there can influence sweeping conclusions. TBL: Among women with fairly early-stage and mostly ER(+) HER2(+) breast cancer receiving anthracycline-based chemo, 6 months was not non-inferior to 12 months of adjuvant trastuzumab . | Pivot, Lancet 2019