Decisions to delay reporting positive interim results, as seen in LITESPARK 011 and other major trials, are often driven by the Independent Data Monitoring Committee (IDMC), not investigators. This highlights the IDMC's power in managing trial conduct, especially when co-primary endpoints like Overall Survival are immature and require longer follow-up.

When a billion-dollar drug trial fails, society learns nothing from the operational process. The detailed documentation of regulatory interactions, manufacturing, and trial design—the "lab notes" of clinical development—is locked away as a trade secret and effectively destroyed, preventing collective industry learning.

The FDA receives raw and cleaned datasets from sponsors, not just summary reports. Their internal teams conduct independent analyses, which can lead to findings or data presentations in the official drug label that differ from or expand upon what's in the published paper.

We possess millions of data points on interventions, but they are useless to AI models because they're trapped in thousands of disparate EMRs in varied formats. The challenge is not generating more data, but solving the human incentive and alignment problems required to create unified data registries.

Advancing circulating tumor DNA (ctDNA) as a surrogate endpoint is stalled because the necessary large-scale, prospective validation studies are too expensive for any single company. The path forward requires a massive public-private partnership to fund research and establish standards, otherwise progress will remain incremental.

While PCWG4 advocates for using Patient-Reported Outcomes (PROs), it doesn't mandate specific analysis methods. This flexibility creates a risk where researchers can explore numerous permutations of the data post-hoc to find a result that supports their desired conclusion, whether positive or negative.

The traditional drug-centric trial model is failing. The next evolution is trials designed to validate the *decision-making process* itself, using platforms to assign the best therapy to heterogeneous patient groups, rather than testing one drug on a narrow population.

Academics with novel research questions can collaborate with the FDA. However, due to the confidential nature of sponsor data, all analyses are performed internally by FDA statisticians. External partners provide clinical insight and interpretation on summarized, non-confidential outputs.

Gossamer's Phase 3 drug for PAH failed after being designed around a promising subgroup identified in a post-hoc analysis of a less-than-stellar Phase 2 trial. This outcome serves as a cautionary tale for clinical development, highlighting the high risk of basing expensive pivotal studies on retrospective data mining rather than robust, pre-specified endpoints.