Searching for an appropriate image-guided radiotherapy method in prostate cancer--implications for safety margin.

Abstract

The aim of the study was to compare the safety margin width using skin marks, pelvic skeleton-based targeting and adaptive protocol combining cone-beam computed tomography and kilovoltage image matching. A total of 434 consecutive patients were treated by image-guided radiotherapy from November 2008 to April 2012. An adaptive protocol combining cone-beam computed tomography and kilovoltage image matching with individualized safety margin calculation according to the Van Herk method was used in a total of 201 patients. The remaining 233 patients had their setup corrected using cone-beam computed tomography daily. Analysis of the 3,137 cone-beam computed tomography images (201 patients) revealed that the margins between the clinical target volume and planning target volume with skin marks registration should be 1.24 cm in the anteroposterior, 0.98 cm in the craniocaudal, and 1.03 cm in the laterolateral direction. Considering pelvic skeleton-based setup, values of the clinical target volume and planning target volume margins in the anteroposterior, craniocaudal and laterolateral axis were 0.79 cm, 0.41 cm, and 0.19 cm, respectively. In a group of 8,872 cone-beam computed tomography images (233 patients) using CBCT assessment, the calculated margins between clinical target volume and planning target volume with skin marks were 1.15 cm in anteroposterior, 1.06 in craniocaudal, and 1.19 in laterolateral directions. Considering the pelvic skeleton-based setup, the corresponding values were 0.74 cm, 0.51 cm, and 0.25 cm. With the adaptive technique, the margins of most patients in the anteroposterior, craniocaudal, and laterolateral axes were 6 mm, 6 mm, and 6 mm or 8 mm, 6 mm, and 6 mm, respectively. The adaptive protocol combining cone-beam computed tomography and kilovoltage image matching or daily cone-beam computed tomography allowed us to substantially reduce the safety margins compared with skin marks targeting.