Interpreting PD-L1 Subgroups and the Case for the Triplet in HER2-Positive GEA

Drs Kim and Pereira discuss one of the most unexpected findings of HERIZON-GEA-01. Unlike prior first-line PD-1 inhibitor trials in gastroesophageal cancer, benefit with the triplet did not track with increasing PD-L1 expression and appeared especially pronounced in PD-L1-negative tumors. Dr. Kim explores possible explanations, including a potential immunogenic effect of zanidatamab and changes in PD-L1 expression over time, and notes that the findings expand immunotherapy options for patients previously excluded based on KEYNOTE-811. Dr. Pereira highlights the limitations of PD-L1 as a predictive biomarker, including variability across biopsy sites and pathology laboratories. The physicians then consider whether tislelizumab is needed, given the similar progression-free survival between the zanidatamab-containing arms. They point to improved disease control, longer duration of response, and a more favorable tail of the survival curve with the triplet. Both favor the triplet for most patients without contraindications to immunotherapy.