Prognostic factors for tumour response and skin damage to combined radiotherapy and hyperthermia in superficial recurrent breast carcinomas

Prognostic factors for complete tumour response and acute skin damage to combined hyperthermia and radiotherapy were analysed in material of patients with breast cancer, recurrent in previously irradiated areas. Radiotherapy was given daily to a total absorbed dose of 30.0 Gy in 2 weeks or 34.5 Gy in 3 weeks. The first radiotherapy schedule was combined with heat twice weekly, a total of four heat treatments (schedule A). The second radiotherapy schedule was combined with heat either once or twice a week resulting in a total of three (schedule B) or six (schedule C) heat treatments. Heat was induced with microwaves (2450, 915 or 434 MHz) via external applicators and always given after the radiotherapy fraction. The complete response (CR) rate in evaluable patients was 71% (49/69). There was no significant difference in CR rate between the three different hyperthermia schedules. The CR rates were 74% (14/19), 65% (15/23) and 74% (20/27) for schedules A, B and C respectively. The only factor predicting CR, evaluated both uni- and multivariately, was the CRE-value for the present radiotherapy dose (p = 0.02). If only tumours treated with 915 MHz were taken into account, however, then the highest minimum temperature at a given heat session predicted complete response (p = 0.03). This was true also in a multivariate analysis of this subgroup of tumours. A Kaplan-Meier analysis (log rank test) showed no significant difference in duration of CR between the different treatment schedules. Cox's proportional hazards method revealed three significant factors: tumour size (negatively correlated, p = 0.007), the time interval between the diagnosis of the primary tumour and the present treatment (p = 0.02) and the average temperature (0.03). Maximum acute skin reactions in the treatment field were scored according to an ordinal scale of 0-8, modified after WHO 1979. Twenty-six treatment areas (32%) expressed more severe skin damage (score > or = 5) in terms of desquamation with blisters (14%) and necrosis or ulceration (19%). Factors correlated with skin damage were the size of the lesion area (p = 0.011), the highest average maximum temperature during a given heat session (p = 0.03) and the fractionation schedule of hyperthermia (p = 0.05). The extent of previous radiotherapy absorbed dose, previous surgery in the treated area or previous chemotherapy had no significant influence on the acute skin reactions