3287 - Comparison of Urethra Morphology on CT with Urinary Catheter and MRI without Urinary Catheter after Low Dose Rate Prostate Cancer Brachytherapy

I. Sidibe¹, P. Dionne^2,3, F. Bachand^1,2, W. Foster^1,2, A. G. Martin^1,2, M. C. Lavallee^2,3, E. Poulin^2,3, L. Archambault^2,3, L. Beaulieu^2,3, and E. Vigneault^1,2; ¹CHU de Québec – Université Laval, Radiation Oncology, Québec, Canada, Quebec, QC, Canada, ²CRCHU de Québec and Centre de recherche sur le cancer de l’Université Laval, Québec, Canada, Quebec, QC, Canada, ³Département de physique, de génie physique et d’optique, Université-Laval, Québec, Canada, Quebec, QC, Canada

Purpose/Objective(s): The precise delineation of the urethra holds importance in treatment planning for prostate brachytherapy. But, visualization of the prostatic urethra on computed tomography (CT) is less than optimal due to limited contrast. A Foley catheter can be used to enhance localization and visualization of the urethra on CT but its insertion is associated with discomfort and a risk of urinary infection. This study aims to evaluate the reproducibility of urethra contours on magnetic resonance imaging (MRI) in comparison to CT, with the ultimate objective to avoid the need for urinary catheter (UC) for the 1-month post implant CT after low dose rate (LDR) prostate brachytherapy. Materials/

Methods: In a prospective study, 25 patients were selected after LDR prostate brachytherapy. 21 patients underwent urinary catheterization during CT. Subsequently, the UC was removed for MRI. All examinations were imported into the treatment planning system for urethra delineation. Two senior radiation oncologists (RO) and one junior RO contoured the urethra on MRI. On CT the patient’s RO (which was not necessary one of the three RO implicated in the study) performed the urethra contour, which is used for clinical post-plan dosimetry. Urethra delineation was conducted on CT images and T2 space MRI (transverse and sagittal slices) using a 4 mm pearl diameter. The CT and MRI images have been fused to facilitate the comparison of contours. The contours of each observer on MRI were compared to those on CT to assess variability. Similarity indices, including Dice Coefficient (DC), Hausdorff distances (HD) and distance from the center were computed for each case and each RO.
Results: Comparing UC on MRI from all RO and CT, the mean Dice coefficient was 0.62 ± 0.02. When comparing each RO urethra contours to CT, the mean Dice coefficient values were 0.64 ± 0.15, 0.60 ± 0.15, and 0.62 ± 0.16 for RO1, RO2, and RO3, respectively, indicating a moderate similarity between CT and MRI delineation. Additionally, the mean HD95% was 3.14 mm ± 0.2 mm, mean HDmax was 4.29 mm ± 0.11 mm, and HDmean was 1.27 mm ± 0.07 mm, suggesting a significant difference between MRI urethra contouring and CT. The distance from the center was 2.07 mm ± 0.12 mm. When we visually comparing the contours, we noticed similarity at the base and apex, with significant difference in the middle portion of the prostate.
Conclusion: These findings suggest a difference in the reproducibility of urethra contours between CT and MRI, potentially attributed to lumen deformation induced by UC. This implies that the MRI-based urethra dose assessment is likely closer to the actual dose received by the patient in real-life compared to CT-based urethra dose assessment. In addition, since urethra contours are more representative on MRI, UC can be omitted on CT and contours can be made only on MRI when the examinations are available.