Cancer patterns and association with mortality and renal outcomes in non-dialysis dependent chronic kidney disease: a matched cohort study.

backgroundCancer in patients with chronic kidney disease (CKD) is an added burden to their overall morbidity and mortality. Cancer can be a cause or an effect of CKD. In CKD patients, a better understanding of cancer distribution and associations can aid in the proper planning of renal replacement therapy (RRT) and in the choice of chemotherapeutic agents, many of which are precluded in more advanced CKD. This study aims to investigate the distribution and the association of cancer with mortality, renal progression and RRT assignment in a non-dialysis dependent CKD cohort, few studies have investigated this in the past.

methodsThe study was carried out on 2952 patients registered in the Salford Kidney Study (SKS) between October 2002 and December 2016. A comparative analysis was performed between 339 patients with a history of cancer (previous and current) and 2613 patients without cancer at recruitment. A propensity score matched cohort of 337 patients was derived from each group and used for analysis. Cox-regression models and Kaplan-Meier estimates were used to compare the association of cancer with mortality and end-stage renal disease (ESRD) outcomes. Linear regression analysis was applied to generate the annual rate of decline in estimated glomerular filtration rate (delta eGFR).

resultsOf our cohort, 13.3% had a history of cancer at recruitment and the annual rate of de novo cancers in the non-cancer patients was 1.6%. Urogenital cancers including kidney and bladder, and prostate and testicle in males, ovary and uterus in females, were the most prevalent cancers (46%), as expected from the anatomical or physiological roles of these organs and relationship to nephrology. Over a median follow-up of 48 months, 1084 (36.7%) of patients died. All-cause mortality was higher in the previous and current cancer group (49.6% vs 35%, p < 0.001), primarily because of cancer-specific mortality. Multivariate Cox regression analysis showed a strong association of cancer with all-cause mortality (HR:1.41; 95%CI: 1.12-1.78; p = 0.004). There was no difference between the groups regarding reaching end-stage renal disease (26% in both groups) or the rate of decline in eGFR (- 0.97 for cancer vs - 0.93 mL/min/year for non-cancer, p = 0.93). RRT uptake was similar between the groups (17.2% vs 19.3%, p = 0.49).

conclusionsCancer status proved to be an added burden and an independent risk factor for all-cause mortality but not for renal progression. CKD patients with a previous or current history of cancer should be assessed on a case by case basis in planning for renal replacement therapy options, and the presence of cancer should not be a limitation for RRT provision including transplantation.