Biological predictors of radiosensitivity in head and neck squamous cell carcinoma

Abstract

Objectives

The aim of this study is to investigate the influence of prognostic biomarkers on radiosensitivity and survival of advanced head and neck squamous cell carcinomas treated by primary (chemo)radiation.

Material and methods

The clinicopathological data and immunohistochemical staining of p16, c-Met, survivin, PD-1, and PD-L1 of 82 primarily (chemo)irradiated patients with head and neck squamous cell carcinoma were analyzed. Associations with local and locoregional radiation response, overall survival (OS), disease-free (DFS), and disease-specific survival (DSS) were assessed.

Results

Complete tumor response was associated with increased patient age (p = 0.007), N0-status (p = 0.022), M0-status (p = 0.007), and p16-positivity (p = 0.022). High PD-L1 was associated with M0-status (p = 0.026) and indicated tumor response to irradiation (p = 0.057); survivin expression showed higher rates of response failure (p = 0.073). Low PD-1 was associated with increased T-stage (p = 0.029) and local recurrence (p = 0.014). High PD-1 was strongly correlated with PD-L1-positive tumor infiltrating lymphocytes (p < 0.001). Low PD-L1 showed a significant correlation with high c-Met expression (p = 0.01). Significant predictors for unfavorable univariate survival were incomplete tumor response (DSS, p < 0.001), single radiotherapy (DSS, p = 0.002), M1-status (DSS, p < 0.001), decreased radiation dose (DSS, p = 0.014), high survivin (DSS, p = 0.045), and high c-Met (OS, p < 0.05). Survivin and c-Met also showed prognostic significance in multivariate survival analysis.

Conclusions

P16 and PD-L1 indicate radiosensitivity, whereas survivin and c-Met implicate radioresistance in primarily (chemo)irradiated head and neck squamous cell carcinomas. The role of the PD-1/PD-L1 immune checkpoints in radiation response and survival merits further investigation.

Clinical relevance

The findings may improve patient-specific therapy according to individual tumor characteristics.

http://ift.tt/2mDBwMh