Creating CT- and MRI-guided urethra planning-risk-volumes for urethra-sparing prostate stereotactic body radiotherapy

Abstract

Dose-escalation in prostate cancer SBRT improves PSA nadir and negative post-treatment biopsy rates, but increases genitourinary toxicity. The definition of the urethra PRV (uPRV) is not standardized. A Foley-catheter based uPRV incorporates catheter-induced anatomic variability that is not reproduced during daily treatment without catheter, whereas anatomic uncertainty during MRI-CT registration is inherent to the MRI-based uPRV. We thus examined the urethra as delineated on CT and MRI to identify an optimal urethral PRV. The urethra was contoured on CT with catheter (uCT) and T2 MRI without catheter (uMRI) in 25 men with prostate cancer. Men received three intraprostatic gold fiducial markers for localization, and rectal spacer to reduce rectal toxicity. MRI was first registered to the CT (reg1) using automated rigid registration of fiducial markers; u_MRI in this position was labelled uMRI_1. MRI was rigidly registered to the CT again (reg2); this time manually to the posterior prostate/anterior rectal wall interface (a clinically relevant region of interest in delineating a hydrogel spacer for radiation planning without taking into account fiducial markers). The uMRI volume created from this manual MRI-CT registration was labeled uMRI_2. Urethral PRVs were created by uniform, and anterior-posterior (A-P) only, expansions in 1 mm increments around uCT. The proportion of uMRI_1, and uMRI_2 covered by each successive uPRV was measured, as was the decrement in the volume PTV-uPRV. Mean and standard deviation uCT and uMRI volumes were 1.72±0.58 and 1.01±0.51 cc, respectively. Automated MRI-CT registration (reg1) with fiducial markers resulted, on average, in 0.2 mm translational separation of each fiducial marker on MRI to its intended fiducial marker target on CT. The translation and rotation matrix differences between reg1 and reg2 (manual) were: -1.1±2.47, -2.87±6.17 and 0.08±4.68 mm; and -0.16±8.29, 0.75±3.17 and 1.42±4.01 degrees. The smallest average urethra contour that would encompass uCT and uMRI_1 was 2.34±0.75 cc; that for uCT, uMRI_1 and uMRI_2 was 2.87±1.04 cc. The likelihood of urethra coverage and mean percent PTV-uPRV for a given uPRV are shown in the table. Uniform 3 mm uPRV encompassed 88% of uCT and uMRI_1, and 83% of uCT, uMRI_1 and uMRI_2, versus 82% and 75% with A-P only uPRV. Uniform and A-P 3 mm uPRVs are created at the expense of 7% and 4% of PTV, respectively. Three-quarters of potential urethra variability on CT and MRI are included in an A-P 3 mm PRV around a catheter. Additional uniform, or A-P, uPRV expansions for urethral protection must be weighed against clinically relevant declines in PTV-uPRV.Abstract 4128; TablePRV (mm)123456% of:uCT, uMRI_1 included in uniform PRV808588919395uCT, uMRI_1 included in A-P PRV768082848687uCT, uMRI_1 and uMRI_2 included in uniform PRV717883889193uCT, uMRI_1 and uMRI_2 included in A-P PRV677275788183PTV decrease by uniform PRV457101215PTV decrease by A-P PRV344567 Open table in a new tab