Association of radiation dose to cardiac substructures with major ischaemic events following breast cancer radiotherapy

Abstract Background and Aims Patients with left-sided breast cancer receive a higher mean heart dose (MHD) after radiotherapy, with subsequent risk of ischaemic heart disease. However, the optimum dosimetric predictor among cardiac substructures has not yet been determined. Methods and results This study retrospectively reviewed 2158 women with breast cancer receiving adjuvant radiotherapy. The primary endpoint was a major ischaemic event. The dose–volume parameters of each delineated cardiac substructure were calculated. The risk factors for major ischaemic events and the association between MHD and major ischaemic events were analysed by Cox regression. The optimum dose–volume predictors among cardiac substructures were explored in multivariable models by comparing performance metrics of each model. At a median follow-up of 7.9 years (interquartile range 5.6–10.8 years), 89 patients developed major ischaemic events. The cumulative incidence rate of major ischaemic events was significantly higher in left-sided disease (P = 0.044). Overall, MHD increased the risk of major ischaemic events by 6.2% per Gy (hazard ratio 1.062, 95% confidence interval 1.01–1.12; P = 0.012). The model containing the volume of the left ventricle receiving 25 Gy (LV V25) with the cut-point of 4% presented with the best goodness of fit and discrimination performance in left-sided breast cancer. Age, chronic kidney disease, and hyperlipidaemia were also significant risk factors. Conclusion Risk of major ischaemic events exist in the era of modern radiotherapy. LV V25 ≥ 4% appeared to be the optimum parameter and was superior to MHD in predicting major ischaemic events. This dose constraint could aid in achieving better heart protection in breast cancer radiotherapy, though a further validation study is warranted.