Modeling set-up error by daily MVCT for prostate adjuvant treatment delivered in 20 fractions: Implications for the assessment of the optimal correction strategies.

Abstract

To optimise image-guided radiotherapy (IGRT) approaches in set-up error correction for hypofractionated adjuvant treatment of prostate cancer using helical tomotherapy (HT). HT unit provides megavoltage - CT (MVCT) imaging capability for daily patient set-up. The intrinsic uncertainty of the available image registration techniques was first investigated on a pelvic phantom. To optimise different set-up correction strategies, MVCT data of 37 patients treated with 58 Gy delivered in 20 fractions were analyzed. 620 MVCT images were globally available for the analysis. Daily on-line set-up correction without any action level was applied for 27/37 patients. The residual and the effective set-up error were assessed. For 10/37 patients an off-line correction strategy with an action level was applied. Phantom measurements showed an intrinsic uncertainty in MVCT imaging and registration around 0.6 mm (+/-0.5 mm) for each of the three main axes. The minimum value of the residual systematic error was found at 7th-10th session, with values between 0.7 and 1.1 mm (1 SD); a systematic residual error <2 mm for all directions was in any case found at 3rd-4th session. This result was also confirmed by the effective systematic set-up error analysis, with a minimum value within 1.5mm nearly at the 4th or 5th fraction. Although the minimum systematic residual error is reached at 7th-10th treatment session, the effective systematic set-up error analysis confirmed that an off-line correction at the 4th fraction without any action level may be a robust compromise with a large sparing of time and resources compared to on-line correction.