Impact of preoperative body mass index on nephroureterectomy outcomes at an Australian tertiary hospital.

Abstract

Introduction & Objectives: The impact of Body Mass Index (BMI) on nephroureterectomy (NU) outcomes in upper tract urothelial carcinomas (UTUC) remains debated, with limited Australian data available. Herein, we evaluated the relationship between BMI, postoperative complications, and oncological outcomes within an Australian cohort. Method(s): This retrospective study included patients who underwent NU for histologically confirmed UTUC at an Australian tertiary hospital from 2010 to 2022. Patient-, tumour-, and treatment-related variables and postoperative complications (length of stay, ICU admission, 30/90-day readmission, and Clavien-Dindo >= III complications) were extracted. Group comparison's used Fisher's exact test, Welch's t test, and Mann-Whitney Utest. Overall survival (OS), cancerspecific survival (CSS), and localrecurrence free survival (LRFS) were analysed using univariate and multivariate cox models adjusting for age, ASA status (>=3), and tumour grade (as events per variable permitted). Proportional hazards were assessed using Schoenfeld residuals. Kaplan-Meier (KM) curves and logrank tests provided unadjusted milestones. Median follow up was estimated by reverse KM method. Result(s): Among 103 eligible patients, 75 were non-obese (BMI < 30 kg/m2) while 28 patients were obese (BMI >= 30 kg/m2). Median follow-up time was 7.8 years (95% CI: 5.8-8.9). Baseline characterises, including surgical technique, were comparable between groups. On average, operative time was significantly longer in the obese group (347 min vs 293 min, P = 0.002); while postoperative complication rates did not differ. Overall, there were 36 deaths, 16 cancer-specific deaths, and 41 localrecurrence events. On multivariate analysis adjusting for covariates, obesity was independently associated with poorer oncological outcomes (see Table 1). Schoenfeld tests showed no evidence of non-proportional hazards for BMI across all end points (all P > 0.10). Furthermore, KM milestones at 1/3/5-years were consistently lower in obese groups compared to their non-obese counterparts (see Table 2). Conclusion(s): In this Australian cohort, obesity was associated with longer operative time and significantly worse OS, CSS and LRFS, without increased postoperative complication rates. These findings warrant confirmation in larger, multi-centre cohorts with sufficient power for more extensive covariate adjustment and stronger risk-stratification guidance.