When Crossover Muddies the Waters: Selpercatinib Trial Exposes OS Interpretation Flaw
A landmark RET-fusion lung cancer trial shows strong PFS gains but a puzzling OS signal — revealing a critical flaw in how crossover trials are designed.
Summary
The LIBRETTO-431 trial tested selpercatinib, a targeted therapy for RET-fusion-positive non-small cell lung cancer, against standard chemotherapy with or without immunotherapy. While selpercatinib showed a large improvement in progression-free survival and an acceptable safety profile, the overall survival analysis produced a hazard ratio of 1.26 — actually favoring the chemotherapy arm. FDA reviewers argue this counterintuitive result stems from high crossover rates, where control-arm patients switched to selpercatinib after progression, and variable post-progression treatments that confound survival data. The paper calls for all randomized cancer trials to pre-specify analytical methods that account for crossover effects on overall survival, regardless of whether OS is the primary endpoint.
Detailed Summary
Overall survival is the gold standard endpoint in oncology trials because it captures both efficacy and safety in a single, unambiguous measure. Yet when trial designs allow patients in the control arm to cross over to the experimental drug after disease progression — a common and ethically motivated practice — interpreting OS becomes extraordinarily difficult. This paper from FDA's Center for Drug Evaluation and Research uses the LIBRETTO-431 trial as a case study to illustrate this challenge.
LIBRETTO-431 was a multinational, open-label, randomized trial comparing selpercatinib — a highly selective RET kinase inhibitor — against platinum-based chemotherapy plus pemetrexed, with or without pembrolizumab, in treatment-naïve patients with advanced RET fusion-positive non-small cell lung cancer. The trial was designed to evaluate whether targeted therapy could outperform standard chemoimmunotherapy in this genomically defined population.
Selpercatinib demonstrated a large and clinically meaningful improvement in progression-free survival, consistent with its mechanism of action. However, the OS analysis, though immature, returned a hazard ratio of 1.26 favoring the chemoimmunotherapy arm — a result that appears to contradict the PFS benefit and raises immediate questions about interpretation.
The FDA authors attribute this discordance primarily to high crossover rates from the control arm to selpercatinib after progression, combined with heterogeneous post-progression therapies that varied across patients and regions. These factors dilute the survival signal and make it nearly impossible to attribute OS differences to the initial treatment assignment alone.
The paper's core recommendation is that all randomized oncology trials — not just those with OS as the primary endpoint — must pre-specify data collection plans and statistical methods to account for crossover and post-progression treatment effects. Without such planning, OS data risk being uninterpretable, potentially misleading regulators, clinicians, and patients about a drug's true survival benefit.
Key Findings
- Selpercatinib showed large PFS improvement in RET fusion-positive NSCLC versus chemoimmunotherapy.
- Immature OS analysis showed HR of 1.26 favoring the chemotherapy control arm, contradicting PFS data.
- High crossover rates from control to selpercatinib after progression likely confounded OS results.
- Variable post-progression therapies across patients and regions further complicated survival interpretation.
- FDA authors call for pre-specified crossover adjustment methods in all randomized cancer trials.
Methodology
LIBRETTO-431 was an ex-US, open-label, randomized, active-controlled phase III trial comparing selpercatinib to platinum-pemetrexed chemotherapy with or without pembrolizumab in treatment-naïve advanced RET fusion-positive NSCLC. The OS analysis was immature at the time of reporting. This paper is an FDA regulatory commentary analyzing the trial's design and interpretive challenges rather than a primary trial report.
Study Limitations
Summary is based on the abstract only; full methodology, crossover rates, and statistical adjustment details are not available. The OS analysis was explicitly described as immature, limiting conclusions about long-term survival. The paper is a regulatory commentary, not a peer-reviewed independent analysis of the trial data.
Enjoyed this summary?
Get the latest longevity research delivered to your inbox every week.