Twenty patients treated with DIBH-RT after breast-conserving surgery were included. Retrospectively, heart registrations were performed for cone-beam computed tomography (CBCT) to planning CT. Further, breast-surface registrations were performed for a surface, captured concurrently with CBCT, to planning CT. The resulting setup errors were compared with linear regression analysis. Furthermore, geometric uncertainties of the heart (systematic [危] and random [蟽]) were estimated relative to the surface registration. Based on these uncertainties planning organ at risk volume (PRV) margins for the heart were calculated: 1.3危 鈭?#xA0;0.5蟽.
Moderate correlation between surface and heart setup errors was found: R2 = 0.64, 0.37, 0.53 in left-right (LR), cranio-caudal (CC), and in anterior-posterior (AP) direction, respectively. When surface imaging would be used for monitoring, the geometric uncertainties of the heart (cm) are [危 = 0.14, 蟽 = 0.14]; [危 = 0.66, 蟽 = 0.38]; [危 = 0.27, 蟽 = 0.19] in LR; CC; AP. This results in PRV margins of 0.11; 0.67; 0.25 cm in LR; CC; AP.
When DIBH-RT after breast-conserving surgery is guided by the breast-surface position then PRV margins should be used to take into account the heart-position variability relative to the breast-surface.