Prognostic factors for overall survival in malignant ureteral obstruction

To identify prognostic factors for overall survival (OS) in patients with malignant ureteral obstruction (MUO) from gynecologic malignancy (GM), with the goal of improving patient selection for urinary diversion.

Retrospective review of 126 patients with MUO from GM at two academic centers from 2011-2019. Factors related to OS identified by Cox regression proportional hazard model. In patients with incomplete survival data (n = 30), hospice was used as a surrogate for death. Multivariate models and receivers operating characteristics (ROC) curves were created for hemoglobin and albumin values.

Overall median survival was 6.2 months. On univariate analysis, age at diagnosis, Charlson Comorbidity Index (CCI) ≥ 8, advanced clinical stage, ascites, pleural effusion, albumin, and hemoglobin were associated with poor OS. OS was higher for those receiving ureteral stenting as compared with no intervention. There was no survival difference based on hydronephrosis grade, stent failure (SF), or creatinine at the time of intervention. On multivariate analysis, albumin < 2.85 g/dL and hemoglobin < 9.6 g/dL were predictive of poor OS.

OS in patients with MUO due to GM is poor. Several prognostic factors for poor survival including low serum albumin and hemoglobin were identified. Ureteral stenting was associated with improved OS compared to observation, but selection bias likely contributed to this result. Additional studies are needed to clarify this finding. These data can be utilized to counsel patients regarding outcomes after urinary diversion in the setting of MUO and perhaps avoid additional procedures in some of these patients who will not derive meaningful benefit.