The Clinicopathological and Molecular Associations of PD-L1 Expression in Non-small Cell Lung Cancer: Analysis of a Series of 10,005 Cases Tested with the 22C3 Assay

PD-L1 expression testing is mandatory prior to pembrolizumab prescription in non-small cell lung cancer. Our service offers PD-L1 testing using the PD-L1 IHC 22C3 pharmDx assay, in parallel with EGFR, ALK, ROS1 and (in some cases) KRAS testing. We correlate PD-L1 expression in 10,005 tumours with patient age and sex, with tumour histological subtypes, with the sampling modality and type of tissue, and with the presence of other molecular alterations. PD-L1 expression testing was performed using the aforementioned assay; tumour proportion scores (TPS) of 1 and 50% were taken as cut-offs for low and high positivity, respectively. EGFR testing was performed using the cobas (R) EGFR Mutation Test v2. ALK testing was performed using the VENTANA ALK (D5F3) CDx Assay. KRAS testing was performed using pyrosequencing. TPS <1% was seen in 44.4% of tumours, 1-49% in 25.0% and >= 50% in 30.6%. We identified no significant relationship with age. Female patients were slightly more likely to express PD-L1. Poorly-differentiated tumour histology and ALK translocation were significantly associated with PD-L1 expression. Rare EGFR mutations tended to be associated with PD-L1 expression. Pleural and nodal metastases were more likely to express PD-L1 than primary tumours, but biopsy and cytological specimens did not show different PD-L1 expression rates. Our data show that the means of acquiring a tumour sample (biopsy versus cytology) does not have a significant impact on PD-L1 expression. However, we found that certain metastatic sites were associated with significantly higher expression rates, which has substantial implications for selection of tissue for testing.