UNDERPREDICTION OF HUMAN SKIN ERYTHEMA AT LOW-DOSES PER FRACTION BY THE LINEAR-QUADRATIC MODEL

Background and purpose. The erythematous response of human skin to rad iotherapy has proven useful for testing the predictions of the linear quadratic (LQ) model in terms of fractionation sensitivity and repair half time. No formal investigation of the response of human skin to do ses less than 2 Gy per fraction has occurred. This study aims to test the validity of the LQ model for human skin at doses ranging from 0.4 to 5.2 Gy per fraction. Materials and methods. Complete erythema react ion profiles were obtained using reflectance spectrophotometry in two patient populations: 65 patients treated palliatively with 5, 10, 12 a nd 20 daily treatment fractions (varying thicknesses of bolus, various body sites) and 52 patients undergoing prostatic irradiation for loca lised carcinoma of the prostate (no bolus, 30-32 fractions). Results a nd conclusions. Gender, age, site and prior sun exposure influence pre - and post-treatment erythema values independently of dose administere d. Out-of-field effects were also noted. The linear quadratic model si gnificantly underpredicted peak erythema values at doses less than 1.5 Gy per fraction. This suggests that either the conventional linear qu adratic model does not apply for low doses per fraction in human skin or that erythema is not exclusively initiated by radiation damage to t he basal layer. The data are potentially explained by an induced repai r model.