Clinical implication of programmed cell death-1 ligand-1 expression in tonsillar squamous cell carcinoma in association with intratumoral heterogeneity, human papillomavirus, and epithelial-to-mesenchymal transition.

Programmed cell death-1 ligand-1 (PD-L1), essential for immune evasion, is a potential candidate for pathogenesis and therapeutic target of human papillomavirus (HPV)-positive tonsillar squamous cell carcinomas (TSCCs). MET/hepatocyte growth factor signaling and transcription factors involved in epithelial-to-mesenchymal transition (EMT) upregulate PD-L1, which can contribute to clinical outcome. Intratumoral heterogeneity of PD-L1 expression is of clinical importance in selection bias due to false-negative patient enrollment. However, the clinicopathological features, prognostic value, and coexpressed molecules of PD-L1 remain unclear in TSCCs. PD-L1 expression was evaluated via immunohistochemistry using a specific monoclonal antibody (SP142) between whole-tissue and tissue microarray (TMA) sections of 79 tumors (5% cutoff value with weak staining). Expressions of EMT markers (TWIST1, Snail, and SNIP1) and MET/hepatocyte growth factor were also analyzed. Staining of the TMA sections showed 78.5% concordance rate to the whole section. PD-L1 positivity and its intratumoral heterogeneity were 29.1% and 15.2% of TSCCs by whole section, respectively. PD-L1 positivity was prevalent in females, HPV-positive tumors without base of tongue invasion, and SNIP1-overexpressed tumors. SNIP1 overexpression, unmethylated TWIST1, smoking, and poorly differentiated tumors were predictive for PD-L1 overexpression. PD-L1 positivity was a favorable independent prognostic factor. Subgroup analyses according to the coexpression of PD-L1 with HPV, SNIP1, or unmethylated TWIST1 indicated the best clinical outcome than any other subgroups. In conclusion, intratumoral heterogeneity of PD-L1 expression was frequent, warranting a caution in punching TMA cores. A combined analysis of PD-L1 with EMT and HPV may define a characteristic subset of patients and prognostic group.