Programmed Death-Ligand 1 Immunohistochemistry Assay Comparison Studies in NSCLC: Characterization of the 73-10 Assay.

Abstract

Several programmed death-ligand 1 (PD-L1) immunohistochemistry (IHC) assays have been developed independently within clinical programs for therapeutic anti-programmed cell death protein 1 (anti-PD-1) or PD-L1 antibodies, necessitating assessment of assay comparability. We characterized the Dako PD-L1 IHC 73-10 assay used in clinical trials of avelumab (anti-PD-L1) or bintrafusp alfa (M7824; bifunctional immunotherapy) and compared it with the Dako PD-L1 IHC 22C3 pharmDx assay, an approved companion diagnostic for pembrolizumab monotherapy in patients with advanced NSCLC. Formalin-fixed, paraffin-embedded NSCLC tumor samples from a commercial source and from the JAVELIN Solid Tumor phase 1 trial of avelumab (NCT01772004) were stained using the 73-10 and 22C3 IHC assays with a standard protocol. Both assays displayed expected PD-L1 staining patterns. In 148 commercial NSCLC samples, the 73-10 assay stained greater than or equal to 1%, greater than or equal to 50%, and greater than or equal to 80% of tumor cells as PD-L1+ in 64.2%, 36.5%, and 23.6% of the samples, respectively, whereas the 22C3 assay stained 20.3% of the samples as greater than or equal to 50% PD-L1+. In 83 NSCLC clinical trial samples, the 73-10 assay stained 79.5% and 31.3% of the samples as greater than or equal to 1% and greater than or equal to 80% PD-L1+, respectively, whereas the 22C3 assay stained 59.0% and 21.7% as greater than or equal to 1% and greater than or equal to 50% PD-L1+, respectively. Efficacy of avelumab was similar in the subgroups classified with the 73-10 and 22C3 assays using greater than or equal to 80% and greater than or equal to 50% PD-L1+ cutoffs, with objective response rates of 26.9% and 33.3%, respectively. The 73-10 assay demonstrated high sensitivity for PD-L1 staining, and staining was comparable between the greater than or equal to 80% cutoff of the 73-10 assay and greater than or equal to 50% cutoff of the 22C3 assay.