Introduction

Multivariable assessment of independent predictors of treatment failure after ventral onlay buccal mucosal graft urethroplasty (VO-BMGU) is commonly limited by small samples sizes. Our aim was to generate a prediction model for treatment failure in a large and homogeneous contemporary population using easily available patient and surgical characteristics.

Methods

491 men underwent one-stage VO-BMGU at our institution between 01/2009 and 12/2015. Treatment failure was defined as any postoperative instrumentation needed. First, we compared the distribution of patient (age, BMI) and surgical characteristics (previous treatments, site of stenosis, surgical volume, length of graft, and success of voiding trial 21 days post-surgery) between patients with treatment success und treatment failure. Secondly, we performed Cox regression analyses to identify independent predictors of treatment failure. In subgroup analyses, we identified predictors of treatment failure in 406 men undergoing VO-BMGU who had never received a urethroplasty before.

Results

At a median follow-up of 34.3 months (IQR 22.3-54.2 months), treatment success rates were 96.1% at three months, 92.3% at six months, 87.7% at 12 months, and 79.8% at 36 months. Overall, 98 (20%) patients suffered from treatment failure. In multivariable Cox regression analyses, age (HR=1.01; 95% CI=1.00-1.03; P=0.047), second (HR=1.02; 95% CI=1.02-2.82; P=0.043) and third tertile of length of graft (HR=1.74; 95% CI=1.03-2.94; P=0.038), as well as failure at voiding trial 21 days after surgery (HR=2.32; 95% CI=1.51-3.56; P<0.001) held true as significant independent predictors of treatment failure. In multivariable subgroup analyses, failure at voiding trial 21 days after surgery (HR=2.49; 95% CI=1.54-4.02), length of graft above the median (>4cm; HR=1.78; 95% CI=1.08-2.89), and penile stenosis (HR=2.00; 95% CI=1.17-3.43; all P≤0.022) were independent predictors of treatment failure. Previous treatment or surgical volume did not influence treatment failure significantly neither in main, nor in subgroup analyses (all P≥0.3)._x000D_

Conclusions

In a large contemporary cohort of 491 patients undergoing VO-BMGU we found age, length of buccal mucosal graft, and delayed suprapubic catheter removal due to failure at voiding trial 21 days after surgery highly predictive of treatment failure. Our model may help in patient counseling during the postoperative setting regarding the stricture recurrence risk and function as basis for future clinically practicable risk calculators.

Funding

None