Background
Head and neck squamous cell carcinoma (HNSCC) negative for Human Papillomavirus (HPV) has remained a difficult to treat entity, whereas tumors positive for HPV are characterized by radiosensitivity and favorable patient outcome. On the cellular level, radiosensitivity is largely governed by the tumor cells` ability to repair radiation-induced DNA double-strand breaks (DSBs), but no biomarker is established that could guide clinical decision making. Therefore, we tested the impact of the expression levels of ATM, the central kinase of the DNA damage response as well as DNA-PKcs and Ku80, two major factors in the main DSB repair pathway non-homologous end joining (NHEJ).
Conclusions
Our data suggest that HPV-negative HNSCC with particularly low Ku80 expression represent a highly radiosensitive subpopulation. Confirmation in independent cohorts is required.
Methods
A tissue microarray of a single center HNSCC cohort was stained for ATM, DNA-PKcs and Ku80 and the expression scored based on staining intensity and the percentages of tumor cells stained. Scores were correlated with clinicopathological parameters and survival.
Results
Samples from 427 HNSCC patients yielded interpretable stainings and were scored following an established algorithm. The majority of tumors showed strong expression of both NHEJ factors, whereas the expression of ATM varied more. The expression scores of ATM and DNA-PKcs were not associated with patient survival. For HPV-negative HNSCC, the minority of tumors without strong Ku80 expression trended towards superior survival when treatment included radiotherapy. Focusing stronger on staining intensity to define the subgroup with lowest and therefore potentially insufficient expression levels in the HPV-negative subgroup, we observed significantly better overall survival for patients treated with radiotherapy but not with surgery alone. Conclusions: Our data suggest that HPV-negative HNSCC with particularly low Ku80 expression represent a highly radiosensitive subpopulation. Confirmation in independent cohorts is required.