The patient positioning in pelvic radiotherapy (RT) should be decided based on both reproducibility and on which position that yields the lowest radiation dose to the organs at risk (OAR), and thereby less side effects to patients. The present randomized study aimed to evaluate the influence of patient positioning on setup reproducibility and dose distribution to OAR in rectal cancer patients.

Ninety-one patients were randomized into receiving RT in either supine or prone position. The recruitment period was from 2005 to 2008. Position deviations were derived from electronic portal image registrations, and setup errors were defined as deviations between the expected and the actual position of bony landmarks. Setup deviations were expressed into three table shift values (∆x, ∆y, ∆z) from which the deviation vector [Formula: see text] were calculated. The estimated lengths of [Formula: see text] defined the main outcome and were compared between prone and supine positions using linear mixed model statistics. The mean volume of each 5 Gy increments between 5 and 45 Gy was calculated for the small bowel and the total bowel, and the dose volumes were compared between prone and supine position.

Data from 83 patients was evaluable. The mean [Formula: see text] was 5.8 mm in supine position and 7.1 mm in prone position (p = 0.024), hence the reproducibility was significantly superior in supine position. However, the difference was marginal and may have borderline clinical importance. The irradiated volumes of the small bowel and the total bowel were largest in the supine position for all dose levels, but none of those were significantly different. The patient positioning in RT of rectal cancer patients may therefore be decided based on other factors such as the most comfortable position for the patients.