Inter-assay reliability of programmed cell death-ligand 1 in head and neck squamous cell carcinoma

•PD-L1 22C3 is used to select head and neck patients for pembrolizumab treatment.•There is a clinical demand for PD-L1 immunohistochemistry laboratory developed tests.•Concordance between the laboratory developed test and pharmDx assay is almost perfect.•Interobserver agreement is substantial and comparable between assays.•The 22C3 laboratory developed test is non-inferior to the 22C3 pharmDx assay. The programmed cell death-ligand 1 (PD-L1) 22C3 pharmDx assay is used as a companion diagnostic test to select head and neck squamous cell carcinoma (HNSCC) patients that may benefit from treatment with the checkpoint inhibitor pembrolizumab. Because the Dako platform is not universally available, we studied the performance of a 22C3 laboratory developed test (LDT) performed on a Ventana BenchMark Ultra compared to the 22C3 pharmDx assay. Serial sections from tissue micro arrays (TMAs) containing tumour tissue from 97 HNSCC patients were stained with the 22C3 pharmDx assay and 22C3 LDT. All TMA cores were scored by three dedicated head and neck pathologists for PD-L1 expression. Substantial interobserver agreement was reported for both the standardized 22C3 pharmDx assay and the 22C3 LDT (respectively Fleiss’ κ 0.62, 95% CI 0.57–0.67 and 0.63, 95% CI 0.58–0.68). Concordance between the assays was almost perfect on core and patient level (respectively Weighted κ 0.84, 95% CI 0.79–0.89 and 0.84, 95% CI 0.75–0.92). Intratumor heterogeneity between the cores per patient case was similar in both assays. After validation a 22C3 LDT is non-inferior to the standardized 22C3 pharmDx assay and can be safely used to select HNSCC patients for pembrolizumab treatment.