Merck & Co.’s Keytruda has risen from the ashes with a positive stomach cancer trial readout. But there’s some reconciliation to do with a past failure that led to the withdrawal of an accelerated approval.
In a new phase 3 trial, Keytruda plus chemotherapy significantly extended patients’ lives over chemo alone in newly diagnosed HER2-negative gastric or gastroesophageal junction cancers that are advanced or metastatic. The Keytruda combo cut the risk of death by 22%, according to data presented Thursday at an ESMO virtual plenary session.
The KEYNOTE-859 trial win sets Keytruda up for an expanded indication to officially challenge Bristol Myers Squibb’s Opdivo, which in 2021 became the first FDA-approved immunotherapy in front-line stomach cancer.
The two PD-1 drugs showed relatively similar results though in slightly different trial populations. In the CheckMate-649 trial, adding Opdivo to chemotherapy reduced the risk of death by 20% in patients with HER2-negative gastric cancer, gastroesophageal junction cancer and esophageal adenocarcinoma. Both Keytruda’s and Opdivo’s data cover patients regardless of their PD-L1 expression status.
But before Keytruda can claim its place in HER2-negative stomach cancer, it needs to resolve what on the surface looks like contradictory findings from two phase 3 trials. Back in 2019, the phase 3 KEYNOTE-062 trial found that Keytruda plus chemo failed to significantly prolong patients’ lives compared with chemo alone in patients with PD-L1-positive gastric and gastroesophageal cancers.
At first glance, it’s baffling how Keytruda could have succeeded in a PD-L1 all-comers trial but fail in a PD-L1-positive study, because the drug is assumed to work better in PD-L1-expressing tumors. But as Merck’s vice president of clinical research, Scot Ebbinghaus, M.D., sees it, the two Keytruda trials actually offered a similar message: that adding chemotherapy to Keytruda could confer a treatment effect in stomach cancer.
In the previous KEYNOTE-062 trial in PD-L1-positive disease, Keytruda was not only tested in combination with chemotherapy but also as a monotherapy. Each Keytruda arm was compared to chemo alone, and the trial was designed to prove superiority for the combo cohort and noninferiority for the single agent. Merck found success with single-agent Keytruda but not with the combo, although adding chemo to Keytruda did provide better results compared to the monotherapy.
For the Keytruda single-agent analysis, those on the Merck PD-1 inhibitor lived a median 10.6 months, versus 11.1 months for those who received chemo. That translated into a 9% death risk reduction in favor of Keytruda. The Keytruda-chemo combo, on the other hand, helped PD-L1-positive patients live a median 12.5 months, lowering the risk of death 15% over chemo alone. But that improvement narrowly missed the statistical significance bar for superiority versus chemo.
Because of the KEYNOTE-062 flop and Opdivo’s FDA approval, Merck in 2021 withdrew a Keytruda accelerated approval in third-line stomach cancer. That regulatory nod was based on tumor response data from a separate single-arm study .
Ebbinghaus admitted that Merck overestimated Keytruda’s efficacy when designing the KEYNOTE-062 trial and that “the magnitude of the benefit wasn’t going to be as forceful as what it was in lung cancer.” That trial had about 250 patients in each arm, so it wasn’t powered enough to show statistical significance for a relatively modest treatment effect.
This time around, the KEYNOTE-859 study enrolled about 750 patients for each arm. The Keytruda-chemo combo increased patient’s life expectancy by 1.4 months—the same as in KEYNOTE-062—to a median 12.9 months. But one key difference is that the new trial included patients who are PD-L1-negative. And in that subgroup of patients, Keytruda’s performance was, as expected, more mutated.
A subgroup analysis shared at ESMO plenary showed that Keytruda and chemo only managed to reduce the risk of death by 8% in patients without PD-L1 expression. This raises the possibility of a restricted FDA approval in PD-L1-positive disease alone.
But Opdivo’s precedent might give Merck some peace of mind. The FDA approved Opdivo and chemo in stomach cancer regardless of PD-L1 status even though the regimen also only delivered an 8% death reduction in PD-L1-negative patients.
Ebbinghaus acknowledged that Keytruda did have a greater magnitude of benefit in patients with higher levels of PD-L1 expression. But he argued that all patients benefited from the drug.
“An alternative point of view, which is, why leave patients behind,” Ebbinghaus said. “The efficacy may be somewhat more modest, but it’s still positive.” Stomach cancer is the fifth most common cancer in the world and the fourth most common cause of cancer death, Ebbinghaus noted.
Before the positive readout in HER2-negative patients, Keytruda, used in tandem with Roche’s Herceptin, was approved in 2021 to treat HER2-positive gastric or gastroesophageal junction adenocarcinoma, a smaller indication. At that time, SVB Securities analyst Daina Graybosch, Ph.D., suspected that doctors will use Keytruda off label in HER2-negative cases, too.