Background and purpose: The existence of hereditary factors influencing the cellular response to ionising radiation has led to the hypothesis that the inter-patient variability of clinical radiation reactions may, at least in part, be attributable to an individual, or intrinsic, radiosensitivity. Considerable effort has been spent in the development of test systems that would determine individual radiosensitivity before or early during radiotherapy to possibly predict treatment outcome, but the results are still conflicting. The present explorative study was therefore aimed at the detection of associations between acute and late radiation effects, tumour control and in vitro radiosensitivity of primary normal tissue fibroblasts.
Patients and methods: Sixty-eight patients with squamous cell carcinoma of the head and neck (93% UICC stage IV) were treated with a simultaneous concomitant boost radiochemotherapy with Carboplatin as part of a prospective non-randomised multicenter study at the University of Heidelberg. Primary fibroblasts were obtained from skin biopsies prior to treatment from 25 unselected patients of this study and the SF2 was determined using the colony forming assay and high dose-rate irradiation. The median follow-up was 21 months (range 2.5-81 months).
Results: The locoregional control rate at three years was 32%. No significant association between acute (mucosa reaction grade 1 or 2 vs. grade 3 and 4), late radiation effects (subcutaneous fibrosis, osteonecrosis, larynx oedema), locoregional tumour control and SF2 of primary fibroblasts was found using Cox proportional hazards regression analysis, log-rank test and Mann-Whitney U-test. Although a steep dose-response relationship was observed for the radiation-induced severe larynx oedema, Cox proportional hazards regression analysis could not fully explain the occurrence of severe radiation-induced larynx oedema with the dose to the larynx (P = 0.09). In the subgroup of twenty-five patients, where the SF2 was determined, bivariate analysis revealed about the same non-significant influence of the dose to the larynx on the larynx oedema (P = 0.1) and no influence of the SF2 (P = 0.5).
Conclusions: In our study of patients with advanced cancer of the head and neck, neither the normal fibroblast SF2 nor the severity of acute radiation effects were able to predict late radiation effects or locoregional tumour control.