FDA Panel Votes to Limit Opdivo, Keytruda for Stomach, Esophageal Cancers Based on Biomarker

The FDA’s Oncologic Drugs Advisory Committee (ODAC) recently cast a decisive vote concerning the use of Merck’s Keytruda and Bristol Myers Squibb’s Opdivo as first-line treatments for stomach cancer patients who test negative for the PD-L1 biomarker. The committee voted 10-2, with one abstention, against the effectiveness of these immunotherapies for this group of patients.

Key discussions among the committee members, FDA reviewers, and sponsors emphasized that immune checkpoint inhibitors tend to show a higher benefit in patients with higher PD-L1 expression levels. Jeffrey Meyerhardt, chief clinical research officer at Dana-Farber Cancer Institute, supported the stance against using these therapies for patients with minimal or no PD-L1 expression. Meyerhardt pointed out that national guidelines already recommend not using these treatments on patients with very low PD-L1 levels.

The committee’s morning vote on the risk-benefit ratio could potentially lead to a label change by the FDA for Keytruda and Opdivo, specifically for patients with low PD-L1 expression in gastric cancer. This decision may also impact the labeling for BeiGene’s tislelizumab, should it receive approval for this indication. Presently, both Keytruda and Opdivo are approved as first-line treatments for stomach cancer, irrespective of PD-L1 status.

A major point of consensus was the predictive value of PD-L1 expression as a biomarker for selecting patients for these treatments. However, a significant challenge identified was the inconsistency in PD-L1 testing methods used by different companies, making it difficult to compare results across various clinical trials. In practical settings, not all patients are tested for PD-L1 expression before undergoing treatment.

Bristol Myers’ Vice President Ian Waxman informed the committee that physicians recognize the importance of PD-L1 testing, with less than 5% of patients treated with Opdivo for this indication being PD-L1 negative. Similarly, Merck’s Vice President Catherine Pietanza acknowledged the variability in real-world PD-L1 testing but highlighted that post-hoc subgroup analyses should not overshadow the positive outcomes from their Phase 3 trial.

The panel generally agreed that the available data justify revising the labels of these blockbuster drugs. FDA’s oncology chief, Rick Pazdur, concluded the morning session by urging pharmaceutical companies to standardize PD-L1 tests, highlighting the need for better collaboration.

In the afternoon, ODAC voted 11 to 1, with one abstention, against the risk-benefit ratio favoring the use of anti-PD-1 antibodies as a first-line treatment for unresectable or metastatic esophageal squamous cell carcinoma (ESCC) in patients with low or no PD-L1 expression. Although the panelists were not voting directly on label changes, many acknowledged that their “no” votes would likely result in such changes based on the best available data.

Panelist Christopher Lieu, professor of medicine at the University of Colorado, explained his “no” vote on ESCC, indicating that the dataset was quite small and represented a minority of patients who would be treated. While higher PD-L1 expression in ESCC is associated with greater overall survival benefits, as in stomach cancer, several panelists and both Merck and Bristol Myers noted that many ESCC patients are severely ill and require prompt treatment.

During earlier discussions, concerns were raised about the limited sample size of data for ESCC. Some panelists debated whether a “no” vote might hinder treatment for the approximately 90% of ESCC patients who express higher levels of PD-L1, due to the necessity for more extensive testing.