Session (CTI)

Issue Clinical Outcome Assessments (COAs) are central to patient-centered oncology trials and regulatory decision-making, yet many COA strategies fail to produce interpretable or label-enabling evidence. A key reason is that COAs are too often treated as downstream measurement tools rather than integral parts of trial design. This leads to fragmentation across endpoint selection, operational implementation, and statistical analysis, with decisions made in isolation from the research question, feasibility constraints, and analytic needs. The result is often missing, biased, or non-actionable data. These failures are not inevitable - they reflect avoidable design and implementation choices. Without a more integrated, end-to-end approach, oncology COA strategies will continue to underdeliver. Overview This panel will examine why oncology COA strategies fail, where they succeed, and how to improve them. It will begin by addressing a core weakness: poorly specified objectives and endpoints that undermine downstream decisions. An industry perspective will explore the tension between scientific ambition and operational feasibility, with lessons from global oncology programs on cross-functional alignment and development planning. An implementation perspective will show how burden, inconsistent training, and system-level issues can compromise data quality, while highlighting practical solutions. An FDA representative will offer a regulatory perspective reviewing existing FDA guidance and recent labeling examples to show how rigorously defined and collected COA-based endpoints can support oncology product labeling. The panel will open up for a 20-minute Q&A with the audience, welcoming attendees’ successes and failures to stress-test the strategies proposed. Together, the panel will move beyond diagnosis to challenge entrenched practices and offer a concrete path toward more rigorous, aligned, and decision-relevant COA strategies.