Regulators disagree over Keytruda’s gastric cancer nod

Keytruda’s US approval for HER2-negative gastric cancer will raise eyebrows: the regulator has greenlit the drug in all-comers even though the supporting Keynote-859 trial showed no benefit in patients whose tumours didn’t express PD-L1. The problem is amplified by the fact that in the EU, where Merck & Co is seeking a similar label, the CHMP has recommended approval only in PD-L1 ≥1% expressers. When Keynote-859, a trial in front-line HER2-negative gastric/gastroesophageal adenocarinoma, was presented at February’s ESMO virtual plenary, Keytruda plus chemo was shown to cut risk of death by 22% versus chemo. However, the benefit was clearly driven by PD-L1 expressers, as PD-L1-negatives yielded a 0.92 hazard ratio with confidence interval upper bound above 1.00. The discussant, Cambridge University’s Dr Elizabeth Smyth, said the latter patients “should not be treated” with checkpoint blockade, and the CHMP clearly agreed. Similarly telling is Keytruda’s corresponding approval in HER2-positives: in the EU a chemo/Herceptin triplet was greenlit in PD-L1 ≥1% expressers, while the FDA gave an all-comers label – only to restrict use to PD-L1-positives later. That said, Smith accepted that “PD-L1 can be an unreliable biomarker in gastroesophageal adenocarcinoma”.

Selected US approvals in gastric/oesophageal/GEJ cancer

Source: US prescribing information.