The FDA’s Oncology Drugs Advisory Committee voted near-unanimously that the benefits of PD-1 inhibitors like Keytruda and Opdivo in PD-L1 low patients do not outweigh the risks.
The FDA’s Oncology Drugs Advisory Committee voted Thursday to recommend the regulator narrow the label of blockbuster PD-1 inhibitors Keytruda, from Merck, and Opdivo and Yervoy, from Bristol Myers Squibb, in stomach and esophageal cancers.
In the morning session, the panel of experts voted overwhelmingly—10-2, with one abstention—against the use of these drugs for the first-line treatment of advanced HER2-negative stomach cancer in patients with PD-L1 expression scores less than 1. Currently, Keytruda and Opdivo can be used in these patients regardless of their PD-L1 levels. Although the FDA will make the final call on label adjustments, many panelists indicated a belief that their votes would lead to this end.
Merck, BMS, and BeiGene—which is developing the PD-1 blocker Tevimbra (tislelizumab) for this indication—participated in the meeting, along with physicians, patients and FDA representatives.
Using data from studies supporting the approval of Keytruda and Opdivo for sttomach and esophageal cancers, as well as data submitted to support the potential approval of tislelizumab, the panelists discussed the value of PD-L1 expression as a predictive biomarker for determining potential benefit from immune checkpoint inhibitors (ICI) such as these, and the risk-benefit profile of using them in people whose cancers have low PD-L1 expression.
In briefing documents released ahead of the adcomm, FDA reviewers determined that high PD-L1 is likely a “predictive biomarker of treatment efficacy” in first-line gastric adenocarcinoma. There is “unclear” benefit in those with PD-L1 scores lower than 10, the reviewers wrote, and these drugs do not appear to have substantial benefit in patients with PD-L1 scores lower than 1. Thus, the broad use of PD-1 blockers “may not be in the best interest of patients with tumors with low PD-L1 expression.”
Most panelists agreed with this assessment. “Clearly, having a high PD-L1 ligand measurement is associated with greater efficacy,” psychiatrist James Randolph Hillard, a patient representative on the panel who has battled stomach cancer, said in Thursday’s adcomm. But, he added, “I don’t think there’s clear evidence for the null hypothesis, that there’s no chance” ICIs will be valuable in treating people with scores less than 1.
The FDA also noted in its briefing documents the potential risks of PD-1 inhibitors, including serious immune-related adverse events, and panelists agreed. Hanna Sanoff, gastrointestinal cancer section chief at the University of North Carolina at Chapel Hill, said during the adcomm that fatal toxicities have been seen with these drugs. “The folks we don’t have at the microphone are the folks who have passed away from getting PD-1 inhibitors.”
Merck and BMS both argued in their briefing document to maintain the status quo, with BMS writing that the current labeling leaves the decision in the hands of the treating physician “and increases the chance for patients who may potentially benefit . . . to be considered for treatment.”
However, Daniel Spratt, chairman and professor of radiation oncology at University Hospitals, Seidman Cancer Center and Case Western Reserve University, who voted no, pointed out that physicians see very few low PD-L1 patients. “The average doctor sees five of these a year, so I’m just not sure we want to let their doctor make this decision,” he said during the adcomm, noting the financial and toxicity impacts for these patients if they are given PD-1 inhibitors.
Hillard, who voted yes, noted that differing PD-L1 assays pose a challenge. “It’s clear that there’s some variability in terms of how this is assessed in different settings.”
After morning voting was completed, ODAC Director Richard Pazdur put out a call to the three pharma companies present for harmonization among PD-L1 assays. “It’s a rare opportunity that I get three drug companies in front of me simultaneously,” Pazdur quipped.
Merck, BMS and BeiGene all signaled their support—though Pazdur repeated his appeal for their cooperation more than once during the meeting.
Secondly, Pazdur asked if, in the event that the drugs’ labels are restricted to those with a PD-L1 score higher than 1, the companies would consider offering them free of charge in an expanded use program to patients with scores lower than 1. On this question, the companies were open but noncommittal.
Esophageal Cancer Vote
The afternoon session posed the same questions about the use of the same PD-1 inhibitors in esophageal squamous cell carcinoma, in addition to Yervoy. The panel voted 11-1 against, with one abstention, the use of PD-1 inhibitors in first-line unresectable or metastatic esophageal squamous cell carcinoma with low or no PD-L1 expression.
Randy Hawkins, assistant professor of internal medicine at Charles R. Drew University of Medicine & Science, said he had “some apprehension about removing something that was available but could not ignore the numbers.”
Throughout the day, committee members lauded the quality of the data presented. “Despite the small numbers, I think this is the best dataset we’re going to get,” Heidi McKean, GI medical oncology program director at Avera Cancer Institute, who voted with the majority, said following the afternoon vote.
The lone dissenting vote came from patient representative Dana Deighton. “I understand the numbers, and I understand the concerns . . . but as a patient, I also had no options and had to fight for anything I could find,” she said. “And I do believe in the Hail Mary passes.”
