Tom says perioperative EV-pembrolizumab produced hazard ratios of 0.53 and 0.60 versus Niagara's 0.68 and 0.75, with path CR jumping to 55%. Rob still expects Niagara to remain relevant until EV is licensed and reimbursed in many countries. Access barriers, not just efficacy, shape near-term standard of care.
Niagara's control arm had 60% two-year event-free survival versus about 70% in the negative trials. The discussion says this makes Niagara the highest-risk population, where more events and greater risk can let an immune-therapy benefit show up. That may explain more than drug differences alone.
Rob notes the nivolumab trial presented circulating tumor DNA data and more patients in the nivolumab arm had detectable ctDNA at baseline, a strong poor prognostic factor. He says this is not a full explanation but illustrates unseen imbalances that can appear despite randomization. Such imbalances can swing large trials.
Tom argues platinum chemotherapy, including gemcitabine-cisplatin, is inhibitory to immunotherapy, and he thinks most benefit in these perioperative trials comes from the adjuvant period. He notes survival curves separate mainly after chemotherapy stops, when some patients finally receive immune therapy alone. That timing suggests antagonism, not synergy, during chemo.
Tom says he sees no big role for adjuvant immune therapy because waiting for neoadjuvant chemotherapy and cystectomy means many patients never reach that phase. He prefers early immune therapy, as in Niagara, because some patients are cured with neoadjuvant treatment. This sequencing argument favors perioperative IO over postoperative-only IO.
Tom says he is sure that rerunning Niagara, Energize, and the pembrolizumab trial would not produce the same two negatives and one positive. He argues randomized phase 3 trials are imperfect and vulnerable to noise, chance, and unseen imbalances. That should temper how confidently we interpret any single result.
Rob points out all three chemo-IO trials raised pathological complete response by roughly 8-10%, yet only Niagara improved survival. He says path CR is not clinically meaningful by itself and has never been proven as a surrogate for event-free or overall survival. So similar activity signals cannot explain divergent outcomes.
