Percutaneous nephrostomy as a marker of clinical vulnerability in non-metastatic muscle-invasive bladder cancer: prognostic and infectious implications.

Background: Percutaneous nephrostomy (PCN) is frequently employed to relieve malignant ureteral obstruction in patients with muscle-invasive bladder cancer (MIBC), yet its prognostic and infectious implications in non-metastatic disease remain poorly characterized. This study aimed to evaluate the association between PCN at diagnosis and clinical outcomes in patients with non-metastatic MIBC. Methods: We retrospectively analyzed 199 patients with non-metastatic MIBC treated between January 2019 and March 2025 at two centers in Türkiye. Patients were grouped by PCN status at diagnosis. The primary endpoint was overall survival (OS); secondary endpoints were metastasis-free survival (MFS) and infection-related outcomes. Survival was analyzed using Kaplan-Meier and multivariable Cox regression, including ECOG performance status and treatment modality as covariates. Logistic regression was used to assess infectious outcomes. Collinearity was evaluated using variance inflation factor analysis. Results: PCN was present in 27.1% of patients. Median OS was shorter in the PCN group compared with those without PCN (24 vs. 41 months; p = 0.004); however, PCN did not retain independent prognostic significance in the multivariable model after adjustment for ECOG performance status, CRP, N stage, and treatment modality (HR 0.433; 95% CI 0.175-1.068; p = 0.069). ECOG performance status (HR 1.819; p = 0.039), CRP (HR 1.005; p = 0.01), and N stage (HR 1.724; p = 0.023) were independent predictors of OS. A subgroup analysis within the hydronephrosis-positive population showed no significant difference in OS between PCN and non-PCN patients (26 vs. 29 months; p = 0.785). Treatment modality was the sole independent predictor of MFS (HR 0.594; p = 0.008). PCN was a strong and independent predictor of positive urine cultures (OR 2.685; 95% CI 1.055-6.837; p = 0.038) and infection-related hospitalizations (OR 13.995; 95% CI 4.923-39.785; p < 0.001). Conclusions: In patients with non-metastatic MIBC, PCN was not an independent predictor of OS after comprehensive adjustment, suggesting that the survival disadvantage observed in unadjusted analysis reflects baseline clinical vulnerability rather than a direct effect of the procedure. PCN was, however, independently and strongly associated with infectious morbidity. These findings position PCN as a marker of clinical frailty that warrants careful multidisciplinary evaluation before placement and proactive management of infectious complications thereafter.