Cancer Mortality in People Receiving Dialysis for Kidney Failure: An Australian and New Zealand Cohort Study, 1980-2013

Rationale & Objective: Cancer is a significant cause of morbidity in the population with kidney failure; however, cancer mortality in people undergoing dialysis has not been well described. We sought to compare cancer mortality in people on dialysis for kidney failure with cancer mortality in the general population. Study Design: A retrospective cohort study using linked health-administrative and dialysis registry data. Setting & Participants: All people receiving dialysis represented in the Australian and New Zealand Dialysis and Transplantation Registry, 1980-2013. Exposure: Dialysis; hemodialysis (HD) and peritoneal dialysis (PD). Outcome: Death and underlying cause of death ascertained using health administrative data and classified using International Classification of Diseases, Tenth Revision, Australian Modification (ICD-10-AM) codes. Analytical Approach: Indirect standardization on age at death, sex, year, and country to estimate standardized mortality ratios (SMR). Results: Over 269,598 person years of observation, 34,100 deaths occurred among 59,648 people on dialysis, including 3,677 cancer deaths. The relative risk of all-site cancer death in dialysis was twice (SMR, 2.4 [95% CI, 2.33- 2.49]) that of the general population and highest for oral and pharynx cancers (SMR, 24.3 [95% CI, 18.0-31.5]) and multiple myeloma (SMR, 22.5 [95% CI, 20.3-23.9]). Women on dialysis had a significantly higher risk of all-site cancer mortality (SMR, 2.7 [95% CI, 2.59-2.89]) compared with men (SMR, 2.3 [95% CI, 2.17-2.36]) (P < 0.001). People on HD (SMR, 2.2 [95% CI, 2.11-2.30]) experienced greater excess deaths from all-site cancer compared with people on PD (SMR, 1.3 [95% CI, 1.23-1.44]). Excess deaths have gradually decreased over time for all-site, multiple myeloma, and kidney cancers (P < 0.001) but have not kept up with improvements in the general population. By contrast, among people receiving dialysis, excess deaths increased for colorectal and lung cancers (P < 0.001). Limitations: Confirmation of cancer diagnoses and population incidence data were not available; inability to exclude pre-existing cancers. Conclusions: People on dialysis experience excess all-site and site-specific cancer mortality compared with the general population. Mortality differs by modality type, age, and sex. Understanding the role of kidney failure and other morbidities in the treatment of cancer is important for shared decision-making regarding cancer treatments and identifying potential approaches to improve outcomes.