Inclusion of heart substructures in the optimization process of volumetric modulated arc therapy techniques may reduce the risk of heart disease in Hodgkin's lymphoma patients

Background and purpose: Radiotherapy is an effective treatment for Hodgkin's lymphoma (HL), but increases the risk of long term complications as cardiac events and second cancers. This study aimed to reduce the risk of cardiovascular events through an optimization of the dose distribution on heart substructures in mediastinal HL patients with the adoption of different volumetric modulated arc therapy (VMAT) techniques, while maintaining the same risk of second cancer induction on lungs and breasts. Materials and methods: Thirty patients (15 males and 15 females, 15 bulky lesions) treated between 2012 and 2017 at our institution were selected. Disease extent was mediastinum plus neck (n = 10), mediastinum plus unilateral axilla (n = 10) and mediastinum alone (n = 10). Lungs, breasts, whole heart and sub-structures (coronary arteries, valves and chambers) were contoured as organs at risk and included in the optimization process. A “first-generation” multi-arc butterfly VMAT (B-VMAT) planning solution was compared to a full-arc butterfly VMAT (FaB-VMAT) approach, consisting of a full arc plus a non-coplanar arc. Lifetime attributable risk (LAR) of second breast and lung cancer and relative risk (RR) of coronary artery disease (CAD) and chronic heart failure (CHF) were estimated. Results: FaB-VMAT resulted in lower mean dose to whole heart (7.6 vs 6.9 Gy, p = 0.003), all coronary arteries (16.1 vs 13.5 Gy, p < 0.001), left ventricle (4.2 vs 3.4 Gy, p = 0.007) and in lower V20Gy to the lungs (15% vs 14%, p = 0.008). A significant lower RR for CAD and CHF was observed for FaB-VMAT. The risk of second breast and lung cancer was comparable between the two solutions, with the exception of female patients with mediastinal bulky involvement, where B-VMAT resulted in lower mean dose (2.8 vs 3.5 Gy, p = 0.03) and V4Gy (22% vs 16%, 0.04) to breasts, with a significant reduction in LAR (p = 0.03). Conclusions: FaB-VMAT significantly decreased the RR for CAD and CHF compared to B-VMAT, with almost the same overall risk of lung and breast cancer induction. These results are influenced by the different anatomical presentations, supporting the need for an individualized approach