Residual gammaH2AX foci in head and neck squamous cell carcinomas as predictors for tumour radiosensitivity: Evaluation in pre-clinical xenograft models and clinical specimens.

Abstract

Predictive biomarkers can be instrumental to treatment individualisation of cancer patients and improve therapy outcome. Residual γH2AX foci represent a promising biomarker to predict tumour radiosensitivity. In this pre-clinical study, the slope of the dose-response curve was evaluated for its predictive relevance in head and neck squamous cell carcinoma xenografts (HNSCC). Additionally, the feasibility of the translated assay was tested in a clinical setting in patient derived HNSCC samples, and associations between residual γH2AX foci and clinical parameters were analysed. Seven HNSCC xenografts models (FaDu, SAS, SKX, UT-SCC-5, UT-SCC-14, UT-SCC-45, XF354) were used. Tumour bearing NMRI nude mice were randomly distributed to five treatment arms (0-8 Gy). Residual γH2AX foci (24 h post irradiation) were counted by visual scoring in a micromilieu dependent manner (assessed with BrdU and pimonidazole). The local tumour control values measured as TCD₅₀ (tumour control dose 50%) have previously been published. Patient derived HNSCC biopsies were cultivated ex vivo for 24 h including 4 h of pimonidazole and BrdU treatment, subsequently irradiated with 0-8 Gy and fixed after 24 h. In the pre-clinical study, the dose-response curve slopes negatively correlated with the tumour control dose after fractionated irradiation (TCD_50,fx, R² = 0.63, p = 0.032) and after single dose irradiation under homogeneous hypoxia (TCD_50,SD,clamp, R² = 0.66, p = 0.027). The γH2AX assay in clinical HNSCC samples showed a dose-response relationship, with the values of the slopes ranging from 0.099 Gy^-1 to 0.920 Gy^-1 (coefficient of variation = 52.8%). Slopes derived from patients were in the same ranges as the sensitive, moderate and resistant models of the pre-clinical study. Statistical analysis revealed a significant negative correlation between the slope and the patients' age (R² = 0.65, p = 0.001). These results further support the promise of the slope of the residual γH2AX foci dose-response as a biomarker for radiosensitivity. In the clinical samples, the variation in the slopes reveals patients' specific repair capacities, which could hold potential value for treatment individualisation.