Combined error of patient positioning variability and prostate motion uncertainty in 3D conformal radiotherapy of localized prostate cancer

Purpose: To measure the patient positioning and prostate motion variability and to estimate its influence on the calculated 3D dose distribution in 3D conformal radiotherapy of patients with localized prostate carcinoma. Methods and Materials: Patient positioning variability was determined retrospectively by comparing 54 orthogonal simulator films with 125 corresponding portal films from 27 patients. Prostate motion variability was determined by 107 computed tomography (CT) examinations with a CT simulator in 28 patients during radiotherapy. Results: In each observed direction, the patient positioning variability and prostate motion showed a normal distribution. This observation enabled the calculation of a combined error of both components. The standard deviation (1 SD) of the patient positioning error in three directions ranged from 3.1 to 5.4 mm; the prostate motion variability was significantly greater in the anterior-posterior direction (1 SD = 2.8 mm) than in the mediolateral direction (1 SD = 1.4 mm). The 1 SD of the estimated combined error was in the anterior-posterior direction 6.1 mm and in mediolateral direction 3.6 mm. Conclusion: The range of patient positioning variability and prostate motion were statistically predictable under the patient setup conditions used. Dose-volume histograms demonstrating the influence of the combined error of both components on the calculated dose distribution are presented.