Introduction: Programmed death-ligand 1 (PD-L1) expression as measured by immunohistochemistry (IHC) has been employed to predict the efficacy of anti-PD-1/PD-L1 therapy. Nevertheless, heterogeneous PD-L1 expression represents a challenge for the selection of patients for anti-PD-1/PD-L1 therapy.
Methods: PD-L1 expression using clone 22C3 in 76 resected non-small-cell lung cancer and paired nodal metastases was assessed and classified according to the proportion of immunostained tumour cells using cutoff values of 1%, 5%, and 50%.
Results: The concordance rates for PD-L1 expression between primary and metastatic lymph nodes (N1) at these cutoff values were 67.7% (21/31) (Kappa value: 0.455, p<0.000), 60.0% (15/25) (Kappa value: 0.668, p<0.000), and 62.5% (5/8) (Kappa value: 0.497, p<0.000). In 36 paired N1 lymph nodes and N2 lymph nodes, 54.5% (6/11) (Kappa value: 0.625, p<0.000) of cases of PD-L1 expression were coincident at cutoffs of 1%. If stratified by adenocarcinoma and squamous cell carcinoma, 87.5% (14/16) (Kappa value: 0.830, p<0.000) of cases at the 1% cutoff and 46.7% (7/15) (Kappa value: 0.324, p<0.000) of cases at the 1% cutoff were coincident.
Conclusion: The results of this study demonstrate that the concordance of PD-L1 expression between primary tumour and nodal metastases is low in non-small-cell lung cancer but is high in adenocarcinoma. Our results also suggest that PD-L1 expression in either lymph nodes or tumour tissues does not predict survival. PD-L1 detection in metastatic lymph nodes is not a suitable replacement for PD-L1 detection in the primary lesion.
Keywords: programmed death ligand 1, 22C3, lymph node metastasis, non-small-cell lung cancer, heterogeneity