After a late-stage trial flop and FDA scrutiny, a conditional liver cancer nod for Merck’s Keytruda depended on a second confirmatory study—or so it seemed. Now that the test has come up positive, the indication’s future remains unclear.
Keytruda slashed the risk of death by 21% over placebo in hepatocellular carcinoma patients in Asia who had previously received Bayer’s Nexavar or chemotherapy. The Merck PD-1 inhibitor also beat placebo at shrinking tumors and preventing disease progression, according to data presented at the 2022 American Society of Clinical Oncology (ASCO) Gastrointestinal Cancers Symposium.
The results came from the phase 3 KEYNOTE-394 trial, which is meant to serve as the new confirmatory trial for Keytruda’s accelerated approval in post-Nexavar liver cancer; a previous phase 3 dubbed KEYNOTE-240 narrowly missed its mark. But now, the new results themselves—plus seemingly shifting FDA attitudes and a changing treatment landscape—are adding wrinkles to the indication’s validation.
First, a 21% death risk reduction is hardly a practice-changing showing that would get doctors excited, especially considering Keytruda only extended patients’ median life span by 1.6 months to 14.6 months. But still, Scot Ebbinghaus, M.D., vice president of oncology clinical research at Merck Research Laboratories, said Keytruda offers “an important advance” for patients, giving them a second-line option when they inevitably progress after front-line treatment.
RELATED: Merck grabs lifeline for problematic Keytruda liver cancer approval thanks to Asian trial win
Keytruda showed a similar 22% reduction in the risk of death in the global KEYNOTE-240 trial, which narrowly missed statistical significance. During a meeting in April on accelerated approvals that failed to deliver in confirmatory trials, an FDA expert panel voted unanimously to keep Keytruda’s conditional second-line liver cancer nod in place, pending the readout from this KEYNOTE-394 trial.
Despite the similarity in the numbers, the new Asian trial met statistical significance, thanks to a sample size that was expanded after Merck noted later lines of treatment had blurred the life extension analysis in KEYNOTE-240, Ebbinghaus said.
Ebbinghaus touted the similarity as a positive attribute: It shows Keytruda's consistent treatment effect across different populations. However, in the Asian subgroup of KEYNOTE-240, Keytruda delivered a remarkable 45% death risk reduction over placebo, with patients’ median survival at 13.8 months versus 8.3 months, respectively.
The difference between the two Asian groups is a “perplexing finding that we can’t completely rationalize or explain,” Ebbinghaus said. The way patients were managed and treated beyond the trial period might be one aspect of a complex issue, he said. KEYNOTE-394 is conducted primarily in China.
RELATED: Bristol Myers, yielding to FDA pressure, pulls Opdivo in post-Nexavar liver cancer
Then the question is, will the FDA be satisfied with Keytruda's meeting statistical significance in these new data? Or is a 20%-ish overall survival showing not good enough in this trial, just as it wasn't good enough before? Ebbinghaus tends to believe it’s the former.
The FDA “can’t use a negative study as a confirmatory study,” he said. “It just puts them in a very bad situation if they started using near-miss or negative studies.”
Because KEYNOTE-394 did come up positive, the FDA will at least accept the data for review, Ebbinghaus said, as they provide a “substantial basis” for the FDA’s appraisal of the accelerated approval.
Meanwhile, KEYNOTE-394’s Asia-only nature may add uncertainty to an FDA review. Richard Pazdur, M.D., director of the FDA’s Oncology Center of Excellence, recently dinged China-developed PD-1/L1 drug developers for seeking U.S. approvals based on trials run solely or predominantly in China.
Ebbinghaus argued this Asian trial is part of a larger Keytruda program that’s global in nature. In fact, KEYNOTE-394 was originally planned as part of KEYNOTE-240, but it was split out as a standalone after enrollment in China was delayed. Now, the two “nearly identical” trials—and the phase 2 KEYNOTE-224 study that got Keytruda its original nod—provide a look at the “totality of data,” he said.
Lastly, the liver treatment landscape with immuno-oncology agents is evolving as PD-1/L1 developers move their offerings to earlier treatment. Even before the FDA scrutiny, a combination of Roche’s PD-L1 inhibitor Tecentriq and VEGF inhibitor Avastin became the first immunotherapy regimen for previously untreated liver cancer patients in 2020.
Merck has argued that some patients may not get the Tecentriq combo in front-line treatment, leaving an opening for Keytruda as a second-line I-O option. But at ASCO GI, AstraZeneca is detailing a phase 3 win for its cocktail of PD-L1 inhibitor Imfinzi and investigational CTLA-4 inhibitor tremelimumab, which cut death risk by 22% over Nexavar. If approved, the AZ combo would become an immunotherapy-only treatment that could further expand I-O’s reach in the front line.
Roche’s Tecentriq-Abraxane use in metastatic triple-negative breast cancer could serve as a lesson for Merck. Roche pulled that indication off the market in August, even though the FDA advisory committee had backed the approval after a confirmatory trial failure. The reason? Keytruda had in July won an FDA go-ahead for both before and after surgery in high-risk, early-stage disease.
Merck has its own front-line combo coming. A phase 3 trial dubbed LEAP-002 of Keytruda and Eisai-partnered Lenvima in newly diagnosed liver cancer is expected to read out later this year. The FDA previously shot down an application there, blasting Merck for a premature filing based on tumor shrinkage data from an early-stage study. Plus, Ebbinghaus noted a triplet regimen of Keytruda, Lenvima and Merck’s experimental CTLA-4 agent quavonlimab is undergoing phase 2 testing.