HDR Prostate Brachytherapy Based on Registered TRUS and In-Room Cone Beam CT Images

Abstract

To present a high-dose-rate (HDR) brachytherapy procedure for prostate cancer using transrectal ultrasound (TRUS) to contour the regions of interest and registered in-room cone-beam computed tomography (CBCT) images for needle reconstruction. To characterize the registration uncertainties between the two imaging modalities and to explore the possibility of performing the procedure solely on TRUS. Patients were treated with a combined TRUS-CBCT HDR brachytherapy procedure. For 100 patients dosimetric results were analyzed. For 40 patients registration uncertainties were examined by determining differences in fiducial marker positions on TRUS and registered CBCT. The accuracy of needle reconstruction on TRUS was investigated by determining the position differences of needle tips on TRUS and CBCT. The dosimetric impact of re-registration, and needle reconstruction on TRUS was studied for 8 patients. The average prostate V100 was 97.8% (range, 89.3%-99.7%), urethra D10 was 116.3% (range, 106.8.3%-126.8%) and rectum D1cc was 66.4% (range,52.3%-78.3%) of the prescribed dose. For 85% of the patients registration inaccuracies were within 3 mm. Large differences were found between needle tips on TRUS and CBCT, especially in cranial-caudal direction, with a maximum of 10.4 mm. Re-registration resulted in a maximum V100 reduction of 0.9%, whereas needle reconstruction on TRUS gave a maximum reduction of 9.4%. HDR prostate brachytherapy based on TRUS combined with CBCT is an accurate method. Registration uncertainties, and consequently dosimetric inaccuracies, are small compared with the uncertainties of performing the procedure solely based on static TRUS images. CBCT imaging is a requisite in our current procedure.