Contouring Differences between Imaging Modalities Utilized for High Dose Rate Brachytherapy of the Prostate
Abstract
Purpose
High dose rate (HDR) brachytherapy is a well-established modality for treating prostate cancer. Imaging used for treatment planning is typically transrectal ultrasound (TRUS) or CT. This study aims to compare the volumetric accuracy of, and differences between, TRUS and CT-based prostate contours for HDR brachytherapy using MRI as a baseline.
Methods
The prostate was contoured by a resident physician and reviewed by an attending physician in Eclipse (V16.1) for 10 patients on four different image types (when available); 1) HDR TRUS pre-needle insertion (N=5), 2) HDR TRUS post-needle insertion (N=8), 3) HDR CT post-needle insertion (N=10), and 4) post-HDR treatment MRI (N=9). The difference in prostate volume between pre- and post-needle insertion HDR TRUS, both TRUS versus MRI, and CT versus MRI were analyzed. Differences were averaged and compared to determine agreement with MRI volumes.
Results
The average prostate volumes were 23.2±14.0cc (range: 8.1-36.7), 22.0±10.9cc (range: 5.3-39.0), 44.2±18.1cc (range: 18.9-71.5), and 27.7±11.2cc (range: 16.0-47.6) for (1), (2), (3), and (4), respectively. Post-HDR needle insertion TRUS contours were, on average, 51±12% (range: 32-72%) smaller than their CT counterparts. Post-HDR needle insertion TRUS contours were, on average, 14±14% (range: 1-35%) smaller than pre-needle insertion. Compared to MRI, pre- and post-needle insertion HDR TRUS contours were 20±19% (range: 2-47%) and 23±21% (range: 1-65%) smaller on average respectively, whereas CT volumes were 50±18% (range: 24-84%) larger on average.
Conclusion
The larger deviation from MRI volumes that CT contouring produced suggests that TRUS provides a more accurate volume for prostate contouring in HDR brachytherapy. Future research will utilize image fusion to calculate Dice-Sørensen coefficients and adjust for modality-specific differences in institutional contouring protocols to evaluate spatial agreement and dosimetric impact. Other future studies will investigate differences in perceived target coverage depending on which imaging modality is used for planning given the same needle insertion.