Clinical efficacy and risk factors of minimally invasive surgery for failed pyeloplasty: a prospective multicenter cohort study

Background: Managing failed pyeloplasty is clinically challenging. This study aims to prospectively evaluate the efficacy of surgical interventions for failed pyeloplasty and identifies risk factors for secondary procedure failure. Methods: Patients undergoing secondary treatments for failed pyeloplasty were enrolled at four hospitals from May 2020. Postoperative follow-up and examinations were conducted regularly. Surgical failure was defined as persistent symptoms, progressive hydronephrosis, or worsening renal function requiring further intervention. Predictors of surgical failure were identified using LASSO and multivariable logistic regression. A nomogram was constructed based on independent risk factors, with performance assessed by ROC curves, calibration plots, and decision curve analysis (DCA). Results: A total of 124 patients were analyzed: 45 underwent balloon dilation, 46 redo pyeloplasty, and 33 autologous graft/flap ureteroplasty. Seventeen patients (13.7%) experienced surgical failure over a median follow-up of 24.4 months. Failure rates were highest after balloon dilation (22.2%), followed by graft/flap ureteroplasty (9.1%) and redo pyeloplasty (8.7%). Multivariable analysis identified three independent predictors: number of previous pyeloplasties (OR = 19.01, P = 0.012), preoperative DJ stent indwelling (OR = 6.21, P = 0.031), and ipsilateral renal parenchymal thickness (OR = 0.08, P = 0.021). The nomogram demonstrated good predictive performance (AUC = 0.841), strong calibration, and clinically meaningful net benefit on DCA. Conclusions: Number of previous pyeloplasties, preoperative DJ stent indwelling, and ipsilateral renal parenchymal thickness independently predict surgical failure after reoperation. The nomogram demonstrates good predictive performance; however, larger prospective studies are needed to validate these findings and further refine management strategies.