Long-term cardiac mortality after radiotherapy of breast cancer--application of the relative seriality model.

Effects on the heart constitute a potentially significant and serious clinical problem in primary radiation therapy of early breast cancer. Increased cardiac mortality among irradiated patients may offset the potential benefit in terms of a reduced risk of recurrence or of death from breast cancer. Clinical data on long-term cardiac mortality among breast cancer patients included in two randomized trials (the Stockholm and Oslo studies) of radiation therapy as an adjunct to primary surgery were analysed using the relative seriality model of radiation response. Five different radiation therapy techniques were used in the trials. The original treatment plans were recalculated on a group of model patients using a three-dimensional treatment planning system. A mean dose-volume histogram (DVH) was calculated for each treatment technique. Both heart and myocardium, i.e. excluding circulating blood within the heart, were separately investigated as risk organs. Model parameters, (D50, i.e. the dose giving 50% complication probability; gamma, i.e. the maximum relative slope of the dose-response curve; s, describing the organ relative seriality) were determined by a chi 2 fitting of the calculated probability of excess cardiac mortality, based on the DVHs, to the incidence data. Computed complication probabilities for each treatment technique were modelled within the 95% confidence interval (CI) of the clinical incidence data. It was shown that the relative seriality model, assuming a homogeneous radiation sensitivity within the volume of the heart/myocardium can be used to describe the incidence data. A small dependence on the volume was found. The results do not, however, exclude the possibility that more sensitive structures within the myocardium are the main target for radiation.