The CREST trial showed benefit driven by patients with carcinoma in situ (CIS), while the Potomac trial showed a lack of benefit in the same subgroup. This stark inconsistency demonstrates that subgroup analyses, even for stratified factors, can be unreliable and are a weak basis for regulatory decisions or label restrictions.
Several panelists voted "yes" for approval not because of a compelling risk/benefit profile, but because they believe physicians and patients should have the "option" to choose the therapy. This reveals a philosophy where regulatory approval is seen as a gateway to choice, deferring the final, nuanced risk-benefit decision to the clinic.
When debating immunotherapy risks, clinicians separate manageable side effects from truly life-altering events. Hypothyroidism requiring a daily pill is deemed acceptable, whereas toxicities like diabetes or myocarditis (each ~1% risk) are viewed as major concerns that heavily weigh on the risk-benefit scale for early-stage disease.
An advisory panel split 50/50 on a two-year immunotherapy regimen but voted 7-to-1 for a one-year drug with similar efficacy. This reveals that for adjuvant therapies in non-metastatic cancer, halving the treatment duration and toxicity exposure can decisively shift the risk/benefit calculation in favor of approval.
In the CREST trial, the FDA's critique heavily emphasized an overall survival hazard ratio above one. Though statistically insignificant and based on immature data, this single figure created a powerful suggestion of potential harm that overshadowed the positive primary endpoint and likely contributed to the panel's divided vote.
The FDA's critique of both CREST and Potomac trials highlights that while event-free survival (EFS) endpoints were met, the lack of improvement in overall survival or prevention of muscle-invasive disease makes the risk/benefit profile questionable for an early-stage cancer, where treatment-related harm is a primary concern.
The CREST trial's positive primary endpoint, assessed by investigators in an open-label setting, was rendered negative upon review by a blinded independent committee. This highlights the critical risk of confirmation bias and the immense weight regulators place on blinded data to determine a drug's true efficacy, especially when endpoints are subjective.