Clinicopathological analysis of programmed death-ligand 1 testing in tumor cells of 325 patients with non-small cell lung cancer: Its predictive and potential prognostic value.

INTRODUCTION: Recent studies on check-point inhibitor therapy, which seems to improve the prognosis of patients with advanced non-small cell lung carcinoma increase the importance of immunohistochemical analyses of the programmed-death receptor and of its ligand, PD-L1 protein.

MATERIAL AND METHODS: In our study we present results of PD-L1 immunohistochemical tumor cell expression in a series of 325 lung carcinoma patients biopsies, using the clone 22C3 (and DAKO Link 48 immunostainer). Evaluation of the expression using tissue proportion scoring system allowed to distinguish negative cases (either 0 % or < 1 % of positive tumor cells) versus positive cases in the categories 1-9 %, 10-49 % and ≥ 50 % of positive tumor cells.

RESULTS: In association to histopathologic parameters we observed similar rates of positive expression in patients with adenocarcinoma types (47,8 % of all the cases) as well as with squamous cell carcinomas (44,4 %). Within these histological categories, the rates of positivity were similar also in patients with small versus large (resectional) biopsies. In the biopsies of patients with adenocarcinoma we identified differences in the PD-L1 protein expression associated with its histological subtype. In the cases with predominant lepidic pattern the PD-L1 positivity was present in 18,8 %, with predominant acinar or papillary pattern in 40,8 % and in cases with predominant solid or micropapillary component in 74,1 % of the cases resp. Keratinizing squamous cell carcinomas were positive in 38,5 % and non-keratinizing in 53,8 % of all the cases. The hiqhest incidence of an extensive posivity was observed in sarcomatoid carcinoma type.

DISCUSSION AND CONCLUSION: Immunohistochemically verified PD-L1 protein expression represents a broadly accepted predictive biomarker for immunotherapy of NSCLC patients. The indicated differences of the expression among various NSCLC types and subtypes require to be verified in larger cohorts of patients in relation with clinical parameters to demonstrate whether it could be plausible to use the PD-L1 expression in a role of a negative prognostic parameter.