Associations of urological malignancies with renal progression and mortality in advanced chronic kidney disease: a propensity-matched cohort study

Abstract

Background: Urological malignancy (UM) in patients with chronic kidney disease (CKD) is an added burden to their overall morbidity and mortality. UM is itself a common cause of CKD. Understanding the associations of UM with outcomes in advanced CKD can help in optimisation of the management of these patients. This study investigates the distribution and association of urological malignancy with outcomes (renal progression and mortality) in patients with advanced non-dialysis dependent CKD. Methods: The study was conducted in 2637 of 3115 patients recruited in the Salford Kidney Study between the years 2002 and 2016. A comparative analysis was performed between 160 patients with UM (at baseline and incident) and 2477 patients with no malignancy. Cox-regression models and Kaplan-Meir estimates were used to explore the association between the presence of UM with mortality and renal outcome. Linear regression analysis was used to calculate the rate of progression of CKD in the groups. A 1:3 propensity score matched cohort of 640 patients was generated and utilised in the above analyses. Results: 4.4% had a history of UM at baseline with the annual incident rate being 0.37%. The site of malignancy was the kidney in 40% with comparable numbers for prostatic malignancy (39%). 70% (111/160) of UM patients had a medical cause as their primary diagnosis for CKD. Over a median follow up of 4 years, 34% (905) patients died. In the matched sample, the proportion of deaths was similar between the groups (UM 44% versus no malignancy 48%, p = 0.36). 30% reached end-stage renal disease (ESRD) with no difference between the groups. In the Cox-regression model, UM did not prove to be a risk factor associated with either all-cause mortality (HR:1.03; CI: 0.79-1.35; p = 0.81) or reaching ESRD (HR:1.12; CI: 0.80-1.58; p = 0.49). The rate of decline in estimated glomerular filtration rate (eGFR) was similar between the groups (- 1.05 vs - 1.25 mL/min/1.73m2/year, p = 0.31). Conclusions: There was no correlation observed between UM and all-cause mortality or ESRD. Medical causes of CKD have a significant influence on the outcomes in patients with UM, whereas the UM did not. Hence, a coordinated approach with early liaison between the urology and nephrology teams is needed in the management of UM patients with CKD.