NEW YORK – The US Food and Drug Administration will ask experts in its Oncologic Drugs Advisory Committee (ODAC) on Thursday whether it should narrow two currently approved indications of checkpoint inhibitors — Merck's Keytruda (pembrolizumab) and Bristol Myers Squibb's Opdivo (nivolumab) — to patients with PD-L1-expressing gastric and gastroesophageal cancer.
The agency has already approved Keytruda and Opdivo when combined with chemotherapy as a first-line treatment for patients with HER2-negative unresectable or metastatic gastric or gastroesophageal adenocarcinoma regardless of PD-L1 expression levels. Keytruda plus chemo and Opdivo plus BMS's other checkpoint inhibitor Yervoy (ipilimumab) with chemo also have PD-L1 expression-agnostic approvals as a first-line treatment for unresectable or metastatic esophageal squamous cell carcinoma (ESCC).
The committee will also discuss BeiGene's submitted biologics license applications (BLA) seeking approval for its checkpoint inhibitor Tevimbra (tislelizumab) with chemo in these same first-line indications as Keytruda and Opdivo. The FDA is expected to decide in December whether to approve Tevimbra as a new gastric or gastroesophageal junction cancer treatment.
The FDA is asking its committee of experts to discuss the risks and benefits of adding these immune checkpoint inhibitors to chemo in these indications based on PD-L1 status and whether labeling should be amended for Keytruda and Opdivo so that patients are selected based on PD-L1 levels. The agency also said in briefing documents for the ODAC meeting that experts should consider how PD-L1 cutoffs will be determined for each drug and indication if they think current labels should be amended. Specifically, the agency is asking experts to discuss whether there will be one universal cutoff for all the checkpoint inhibitors in these indications, or if each drug and indication should have specific PD-L1 expression cutoffs based on data from clinical trials.
The ODAC will consider data on the activity of these checkpoint inhibitors from several Phase III studies in gastric and gastroesophageal cancer. For Opdivo, the committee will discuss data from the Phase III CheckMate-649 trial in gastric cancer and the CheckMate-648 trial in esophageal cancer. For Keytruda, they will weigh data from the Phase III KEYNOTE-859 study in gastric cancer and the KEYNOTE-590 trial in esophageal cancer. For Tevimbra, the experts will consider data from the RATIONALE-305 trial in gastric cancer and the RATIONALE-306 trial in esophageal cancer. In each of these studies, the sponsors analyzed their drugs' efficacies across different PD-L1 expression thresholds, but all ultimately concluded that the regimens benefited patients regardless of PD-L1 expression.
FDA reviewers conducted their own exploratory analyses of efficacy data by PD-L1 expression from the three Phase III trials in gastric cancer, finding that across all three studies those with the highest PD-L1 expression appeared to derive the most benefit from the checkpoint inhibitor-based regimen.
The CheckMate-649 and KEYNOTE-859 trials both used PD-L1 combined positive score (CPS) to measure expression. The agency found that the greatest overall survival benefit to Opdivo-chemo treatment in the CheckMate-649 study was among patients with a PD-L1 CPS greater than or equal to 10, while those with a PD-L1 CPS of less than 1 had only a marginal and uncertain improvement in survival.
In KEYNOTE-859, the FDA found that patients with a PD-L1 CPS of less than 1 demonstrated a marginal and uncertain estimated improvement in overall survival, while those with a PD-L1 CPS of less than 5 or a CPS of less than 10 had a "marginally greater, however still modest, improvement in overall survival," the FDA wrote. While patients with higher PD-L1 CPS tended to fare better on Keytruda-chemo, there was some uncertainty in the FDA's exploratory analysis.
Overall, among PD-L1-positive patients, the treatment effect appears to be primarily attributable to patients who have a CPS of at least 10, but the FDA was uncertain whether a CPS of 5 and above would be an acceptable cutoff using Agilent's PD-L1 IHC 22C3 assay. "Notably, patients in between the prespecified cutoffs of CPS [of at least] 1 and CPS [of at least] 10, had a point estimate consistent with an intermediate effect," the agency said in its review document. "There was no evidence to suggest clear detriment in overall survival at any PD-L1 CPS threshold assessed."
In the RATIONALE-305 trial, BeiGene used a different PD-L1 expression scoring method called tumor area positivity (TAP) score, defined as the percentage of PD-L1 positive tumor cells and immune cells divided by tumor area. The company used the Ventana PD-L1 IHC SP264 CDx assay for PD-L1 expression analysis. In this trial, the FDA's analysis also found that the greatest benefit on tislelizumab was among patients with a TAP score greater than or equal to 10.
The FDA conducted an analysis of the three trials in esophageal cancer and similarly found less benefit with the checkpoint inhibitor treatments among patients with lower PD-L1 expression scores.
The "FDA is concerned with the lack of benefit observed across patients with ESCC who have lower (or negative) PD-L1 scores, which would expose these patients to the incremental added toxicity of anti-PD-1 monoclonal antibodies, warranting a more contemporary discussion on the risk-benefit profile in a biomarker-selected patient population," the agency wrote in its briefing document.
BMS also provided its analysis of the data to the expert panel and argued that restricting the indications of these checkpoint inhibitors, including Opdivo, "would leave some potential responders untreated" due to a lack of access to PD-L1 testing, inconclusive testing results, and the complexity of PD-L1 scoring methods. BMS said these potential issues with PD-L1 testing necessitate oncologists exercise "flexibility" in determining how to treat patients, which comes with indications unrestricted by PD-L1 expression status.
BMS added that long-term follow-up data from both Phase III trials in gastric and esophageal cancer have continued to show consistent benefit for Opdivo-treated patients regardless of PD-L1 expression status. "Flexibility is especially important when making treatment decisions in the first-line setting since many patients do not go on to receive later-line therapy and, when they do, their choices are limited," BMS wrote in documents shared with the ODAC.
Merck also argued in its briefing document to the committee that the currently approved indications for Keytruda in HER2-negative gastric and gastroesophageal cancer should be retained. The firm noted that there is a high unmet medical need in these settings and asserted that the totality of clinical data on Keytruda in these cancers support keeping the indication as is "to ensure that all patients who may benefit from the addition of pembrolizumab to chemotherapy do not lose access to this foundational therapeutic option."
Merck estimated, based on the prevalence of PD-L1 expression seen in the KEYNOTE-859 study, that restricting Keytruda's gastric cancer indication to PD-L1 expressors could exclude between 22 percent and 65 percent of newly diagnosed gastric cancer patients. The firm conducted an analysis of electronic health record-derived data from Flatiron Health of patients with advanced or metastatic gastric and gastroesophageal junction cancer who initiated first-line treatment after the FDA approved immunotherapies in this setting and found that about 25 percent were not being tested for PD-L1 expression prior to starting treatment.
Merck has faced several challenges with its Keytruda approvals in gastric cancer. Last year, Merck asked the FDA to limit its accelerated approval of Keytruda in HER2-positive gastric cancer to only HER2-positive gastric or gastroesophageal cancer patients whose tumors express PD-L1 based on an updated analysis. However, at this ODAC meeting, the FDA is concerned with indications in which checkpoint inhibitors are combined with chemo, and this single-agent Keytruda indication is not being discussed.
In 2021, the FDA had asked the ODAC for advice on whether to rescind accelerated approval for a third-line indication of Keytruda in patients with advanced PD-L1-positive gastric cancer. In that case, the panel voted against maintaining the accelerated approval after seeing lackluster confirmatory data on overall survival.
BeiGene told the ODAC in a briefing document that it supports consistent labeling for Tevimbra and testing across the checkpoint inhibitors in this setting, "as it would help provide clarity among the medical community and would better support treatment decisions in clinical practice, along with harmonizing the use of PD-L1 testing, with these agents."
As a new competitor to established players Merck and BMS, BeiGene noted that results from the RATIONALE-305 trial indicated a more pronounced treatment benefit for the subgroups with a higher level of PD-L1 expression. The firm noted that the RATIONALE-305 study supports a favorable risk-benefit profile for patients with unresectable, advance gastric cancer with a PD-L1 TAP score greater than or equal to 5 percent. The RATIONALE-306 study, meanwhile, demonstrated the most favorable risk-benefit in esophageal cancer patients with a PD-L1 TAP score greater than or equal to 1 percent, the firm said.
At the upcoming meeting on Thursday, the ODAC will have to weigh the analysis from Merck, BMS, and BeiGene against the FDA's analysis of these checkpoint inhibitors' activity in gastric and gastroesophageal cancer and decide if these immunotherapies combined with chemo should be available to all-comers or to a subset of patients based on PD-L1 biomarkers.