Validating a Simple Urethra Surrogate Model to Facilitate Dosimetric Analysis to Predict Genitourinary Toxicity.

PURPOSE/OBJECTIVE(S): The urethra may be a critical structure in prostate radiotherapy planning as some studies have shown that higher urethral dose correlates with worse genitourinary (GU) toxicity. Identifying the urethra requires an MRI planning scan or foley catheter insertion at CT planning. Most surrogates have been developed and validated against the urethra identified by a foley catheter. However, the urethral position can shift with catheter placement. We, therefore, aim to validate a simple urethra surrogate model against MRI-defined urethra. The surrogate model can be used to correlate urethra dose-volume parameters (DVP) with late GU toxicity and to apply urethral constraints in those with a CT-only based workflow.

MATERIALS/METHODS: Thirty-nine MRI-defined urethras from patients in the PACE-C trial were assessed to determine the average position of the urethra in the midline sagittal prostate plane along the ¼ gland, midgland, ¾ gland, apex and 3mm below apex. Using these average positions, a Python script was developed, which places a 10mm diameter circle in the 1/4 gland, midgland, ¾ gland, apex and 3mm below the apex. The observer manually contours a 10mm circle at the prostate base (prostate-bladder neck interface) to infer the urethra position and interpolates the contours. The urethra surrogate model was compared against 20 MRI-defined urethras (within the treatment PTV) in patients treated with 36.25Gy in 5 fractions as part of the PACE-B trial. To assess the surrogate's geometric performance, the Dice similarity coefficient (DSC), Hausdorff distance (HD), mean surface distance to agreement (MDTA) and the percentage of MRI-defined urethra outside the surrogate (UOS) were calculated. The surrogate model's dosimetric performance was assessed by comparing the mean D99, D98, average dose, D50, D2 and D1 using a paired t-test. The D(n) is the dose (Gy) to (n)% of the urethra.

RESULTS: The median results were: DSC 0.36 (IQR 0.28-0.42), HD 0.88cm (IQR, 0.70-1.04), MDTA 0.24cm (IQR, 0.21-0.28), UOS 29% (IQR, 17-52%). When comparing DVP between the MRI-defined urethra and surrogate urethra, the mean D99, D98 and D95 as 38.8Gy vs 39.1Gy (p = 0.17), 39.3Gy vs 39.5Gy (p = 0.23), 40.1Gy vs 40.4Gy (p = 0.21), respectively. The mean D50, average dose, D2 and D1 was 41.8Gy vs 41.9Gy (p = 0.03), 41.7 vs 41.8Gy (p = 0.04), 42.9Gy vs 43.0Gy (p = 0.05) and 43.0Gy vs 43.1Gy (p = 0.03), respectively.

CONCLUSION: While there were geometric differences between the surrogate urethra and MRI-defined urethra, there was no statistically significant difference between most urethral dose-volume parameters (D99, D98, D95, and D1). Similarly, the actual differences in urethra DVP were not clinically significant. This surrogate model could be validated in a larger cohort and then used to estimate the urethra position on CT planning scans for dosimetric analysis in those without an MRI planning scan or urinary catheter.