| A clinically reliable predictive assay based on normal-tissue radiosensitivity may lead to improved tumour control through individualised dose prescriptions. In-vitro fibroblast radiosensitivity has been shown, in several studies, to correlate with late radiation morbidity. The aim of this study was to investigate some of the cellular mechanisms underlying the normal-tissue response. In this study, seventeen primary fibroblast strains were established by enzymatic disaggregation of skin biopsies obtained from patients. These comprised seven who experienced acute tissue reactions to radiotherapy, four patients with a normal response and six non-cancer volunteers. An AT cell line was included as a positive control for radiosensitivity. In-vitro radiosensitivity was measured using a clonogenic assay at both high (HDR: 1.6 Gymin-1) and low dose rate (LDR: 0.01 Gymin-1). The radiation parameter HDR SF2 was the most sensitive in discriminating the seven sensitive patients from the remaining ten normal patients (range 0.11-0.19 sensitive patients compared with 0.17-0.34 control patients: p<0.0001). Neither the use of an internal control or LDR radiation protocol increased this discrimination. Pulsed-field gel electrophoresis (PFGE) was used to measure the level of initial and residual double-strand breaks following irradiation. No correlation was found between HDR SF2 and initial DNA damage. However, a strong correlation was found between clonogenic survival and both residual DNA damage (measured over 10-70 Gy, allowing 4 h repair, correlation coefficient: 0.90, <0.0001) and the ratio of residual/initial DNA damage, with the sensitive cell lines generally showing a higher level of residual DNA damage. Cell-cycle delays were found in all 18 cell strains in response to 2 Gy irradiation, but were not found to discriminate between sensitive and normal patients. Associated studies found no mutations of the ATM gene in the five radiosensitive patients studied. However, a coding sequence alteration was found in the XRCC1 gene in one of the radiosensitive patients. These findings indicate that a DNA repair defect may be partly responsible for the extreme reactions to radiotherapy observed in a small percentage of patients and that with further modifications, an assay based on measurement of residual DNA damage may form the basis of a predictive test for radiosensitivity. |
