Dosimetric Predictors of Cardiotoxicity in Thoracic Radiotherapy for Lung Cancer

Abstract Background Higher cardiac radiotherapy (RT) doses when treating lung cancer are associated with worse overall survival (OS), although the direct association between cardiac dose and early cardiotoxicity is poorly understood. We hypothesized that RT doses to the heart and cardiac substructures are associated with under-reported early cardiotoxicity and worse OS. Patients and Methods We conducted an institutional retrospective review of lung cancer patients treated with conventionally fractionated RT from 2010 to 2015. Collected data included pre-RT cardiac risk factors, post-RT cardiotoxicities, and dose-volume parameters for cardiac substructures. Univariate and multivariate analyses were performed to identify predictors of cardiotoxicity and OS. Results Seventy-six cases were evaluated with 1.2 years median follow-up. Cardiotoxicities included atrial arrhythmia (n = 5), pericardial effusion (n = 16), and valvular disease (n = 1). In univariate analysis, significant dose-volume predictors for cardiotoxicity included mean RT dose to structure of interest, volume of structure of interest receiving ≥30 Gy RT dose, and volume of structure of interest receiving ≥45 Gy RT dose (V45) to the atria, ventricles, and pericardium. Higher ventricular V45 was associated with post-RT cardiotoxicity in multivariate analysis (hazard ratio [HR], 1.50; P = .027). Cardiotoxicity occurrence was a highly significant predictor of OS in multivariate analysis (HR, 12.7; P P  = .450). Conclusion Early cardiac events were relatively common after lung cancer RT and associated with multiple cardiac dose-volume parameters. Occurrence of early cardiotoxicity was strongly associated with worse OS. In practice, early cardiotoxicity is under-reported, supporting the need for more detailed cardiac evaluations in high-risk patients to detect and address early cardiotoxicity.