Reproducibility of heart and thoracic wall position in repeated deep inspiration breath holds for radiotherapy of left-sided breast cancer patients

Background: Deep inspiration breath hold (DIBH) for radiotherapy of left-sided breast cancer patientscan effectively move the heart away from the target and reduce the heart dose compared to treatmentsin free breathing. This study aims to investigate the positional reproducibility of heart edge(HE) and thoracic wall (TW) during repeated DIBHs. Material and methods: At three occasions, 11 left-sided breast cancer patients were CT imaged during6 minutes of repeated DIBHs with 60 cine CT series. The series were evenly distributed over threebed positions and for each bed position, the heart edge associated maximum heart distance (MHD)and thoracic wall-associated maximum lung distance (MLD) from a reference line were retrospectivelyanalyzed. The high temporal resolution of the CT series enabled intrinsic heart movements to beresolved from breath hold variations. A body surface laser scanning system continuously extracted thethorax height and displayed it in a pair of goggles for patient feedback. To check for ‘fake-breathing’movements, e.g. that the patient lifts its back from the couch to reach DIBH, the couch-to-spine distancewas also measured in all CT series. Results: The analysis was done for 1432 cine CTs captured during 292 breath holds. The DIBH movedthe heart on average 15mm in medial direction compared with free breathing. For the three bed positionsstudied, the mean value of the max range, across all patients, was between 11–13mm for theMHD and 4–8mm for the MLD. The MHD variation due to breath hold variation was twice as large asthe MHD variation due to intrinsic heart movement. The couch-to-spine distance varied less than3mm for all fractions, i.e., no fake-breathing was discovered. Conclusions: The heart edge and thoracic wall reproducibility was high in relation to the medial heartdisplacement induced by the DIBH.