A prognostic model for survival after palliative urinary diversion for malignant ureteric obstruction: a prospective study of 208 patients.

Abstract

To identify factors associated with survival after palliative urinary diversion (UD) for patients with malignant ureteric obstruction (MUO) and create a risk-stratification model for treatment decisions. We prospectively collected clinical and laboratory data for patients who underwent palliative UD by ureteric stenting or percutaneous nephrostomy (PCN) between 1 January 2009 and 1 November 2011 in two tertiary care university hospitals, with a minimum 6-month follow-up. Inclusion criteria were age >18 years and MUO confirmed by computed tomography, ultrasonography or magnetic resonance imaging. Factors related to poor prognosis were identified by Cox univariable and multivariable regression analyses, and a risk stratification model was created by Kaplan-Meier survival estimates at 1, 6 and 12 months, and log-rank tests. The median (range) survival was 144 (0-1084) days for the 208 patients included after UD (58 ureteric stenting, 150 PCN); 164 patients died, 44 (21.2%) during hospitalisation. Overall survival did not differ by UD type (P = 0.216). The number of events related to malignancy (≥4) and Eastern Cooperative Oncology Group (ECOG) index (≥2) were associated with short survival on multivariable analysis. These two risk factors were used to divide patients into three groups by survival type: favourable (no factors), intermediate (one factor) and unfavourable (two factors). The median survival at 1, 6, and 12 months was 94.4%, 57.3% and 44.9% in the favourable group; 78.0%, 36.3%, and 15.5% in the intermediate group; and 46.4%, 14.3%, and 7.1% in the unfavourable group (P < 0.001). Our stratification model may be useful to determine whether UD is indicated for patients with MUO.