Development of a predictive nomogram for treatment response to urethral botulinum toxin A injection in women with non-neurogenic dysfunctional voiding.

Abstract

Purpose: To identify predictors of treatment success after urethral botulinum toxin A (100 U) injection in women with video-urodynamic study (VUDS)-diagnosed non-neurogenic dysfunctional voiding (DV), and to develop a nomogram for individualized clinical prediction.

Methods: We retrospectively reviewed 398 women with VUDS-confirmed non-neurogenic DV treated with urethral botulinum toxin A at a tertiary center. Baseline demographics and urodynamic variables were compared between responders (Global Response Assessment [GRA] ≥ 2 at 3 months) and non-responders (GRA < 2). Predictive modeling used multivariate logistic regression with stepwise selection. The entire cohort was randomly split into a training set (70%) for model development and a test set (30%) for internal validation. Model discrimination was evaluated by area under the receiver operating characteristic curve (AUC), and a nomogram was constructed from independent predictors.

Results: At 3-month follow-up, 186 patients (46.7%) were classified as responders. Responders had significantly fewer prior Botox injections, lower bladder sensation thresholds, higher detrusor pressure at Qmax (Pdet), higher maximum flow rate (Qmax), greater voided volume, lower cystometric bladder capacity, higher voiding efficiency, and lower post-void residual (PVR). Multivariate analysis identified older age (OR: 1.025), higher Pdet (OR: 1.029), and lower PVR (OR: 0.989) as independent predictors of success. The model showed excellent discrimination in training (AUC = 0.847) and test (AUC = 0.871) sets. A nomogram was developed for individualized risk estimation.

Conclusion: Older age, higher Pdet, and lower PVR independently predict successful response to urethral botulinum toxin A in women with non-neurogenic DV. This nomogram may guide personalized counseling and treatment decisions.