Evaluation of intra- and inter-fraction motion in breast radiotherapy using electronic portal cine imaging

Breast irradiation is one of the most challenging problems in radiotherapy due to the complex shape of the target volume, proximity of radiation sensitive normal structures and breathing motion. It was the aim of the present study to use electronic portal imaging (EPI) during treatment to determine intra- and inter-fraction motion in patients undergoing radiotherapy and to correlate the magnitude of motion with patient specific parameters. EPI cine images were acquired from the medial tangential fields of twenty radiotherapy patients on a minimum of 5 days each over the course of their treatment. The treatments were administered using 10 MV X-rays and dynamic wedges on a Varian Clinac 2100CD linear accelerator. Depending on the incident dose and the angle of the wedge, between 4 and 16 images could be acquired in one session using an EPI device based on liquid ionization chambers (Varian). The border between lung and chest-wall could be easily detected in all images and quantitative measurements were taken for the amount of lung in the field and the distance of the breast tissue from the field edges. Inter-fraction variability was found to be about twice as large as intra-fraction variability. The largest variability was detected in cranio/caudal direction (intra-fraction: 1.3 +/- 0.4 mm; inter-fraction: 2.6 +/- 1.3 mm) while the lung involvement varied by 1.1 +/- 0.2 mm and 1.8 +/- 0.6 mm intra- and inter-fraction, respectively. This indicates that the effect of breathing motion on the amount of radiated lung was not of major concern in the patients studied. Of other patient specific parameters such as body weight, breast separation, field size and location of the target, only increasing age was significantly correlated with larger inter-fraction motion. Acquisition of EPI cine loops proved to be a quick and easy technique to establish the amount of patient movement during breast radiotherapy. The relatively small variability found in the present pilot study justifies considerations for more conformal dose delivery.