Survival Probability in Patients Receiving Maintenance Dialysis versus Patients with Cancer

The burdens of end-stage renal disease (ESRD) may not be fully understood by the general public. In 2010, more than 2 million patients worldwide were being treated for ESRD with maintenance dialysis; the number is expected to increase to 5.4 million by 2030. The mortality rate among dialysis patients is high; approximately 55% of patients die within 5 years of dialysis initiation. The poor prognosis seems to be poorly understood, even among patients in the hemodialysis population; >90% overestimate their 5-year survival probability.

Noting that cancer is a well-known disease in the general population, Kyla L. Naylor, PhD, and colleagues conducted a population-based cohort study designed to compare the survival of patients on maintenance dialysis with that of patients with common cancers to enhance the understanding of the burdens of ESRD. Subanalyses included survival probability presented by age, trends in survival probability over time, and the hazard ratio (HR) of mortality adjusting for clinical characteristics. The primary outcome of interest was all-cause mortality. The researchers reported results of the study in the American Journal of Kidney Diseases [2019;73(6):765-776].

The study utilized data from linked administrative healthcare databases from Ontario, Canada, held at the ICES (Institute for Clinical Evaluation Sciences). Data from six linked databases were included. Patients initiating maintenance hemodialysis therapy were identified via the Canadian Organ Replacement Registry. Cancer patients were identified via the Ontario Cancer Registry. Data from the Registered Persons Database that captures demographic data and vital status for all Ontarians, the Canadian Institute for Health Information Discharge Abstract Database, the National Ambulatory Care Reporting System, and the Ontario Health Insurance Plan were also used in the study.

The maintenance dialysis subcohort included incident hemodialysis and peritoneal dialysis patients from January 1, 1997, to December 31, 2015. Exclusion criteria were <18 years of age or >105 years of age at time of dialysis therapy initiation, history of any cancer other than nonmelanoma skin cancer, previous receipt of an organ transplant (including kidney), or evidence of maintenance dialysis therapy >180 days prior to cohort entry. The index date for the maintenance dialysis subcohort was the date of maintenance dialysis therapy initiation.

Individuals in the cancer subcohort were diagnosed with cancer between January 1, 1997, to December 31, 2015. Included cancers were colorectal, lung, pancreatic, breast (women) and prostate (men). Exclusion criteria were age <18 years or >105 years at cancer diagnosis, history of cancer other than nonmelanoma skin cancer, and previous evidence of chronic kidney disease (dialysis and kidney transplantation). The index date for the cancer subcohort was the date of cancer diagnosis.

The final study cohort included 33,500 maintenance dialysis patients (women, n=13,587 [40.6%] and men, n=19,913 [59.4%]) and 532,452 patients with cancer (women, n=256, 938 [48.3%] and men, n=275,514 [51.7%]). For women, median age at initiation of dialysis therapy was 66 years and median age at diagnosis of cancer was 65 years. For men, median age at dialysis initiation was 65 years and median age at time of cancer diagnosis was 68 years.

Compared with cancer patients, patients on maintenance dialysis had a higher proportion of patients in the lowest income quintile (quintile 1). Dialysis patients had generally more comorbid conditions than patients in the cancer subcohort.

During a maximum of 20.3 years, in dialysis patients, total follow-up was 167,059 person years; in the cancer subcohort, total follow-up was 3,191,753 person-years. A total of 20,790 patients in the dialysis subcohort and 272,782 patients in the cancer subcohort died during the follow-up period.

Median survival in female dialysis patients was 5.0 years; in women with breast, lung, colorectal, and pancreatic cancer, median survival was 19.4, 0.8, 7.5, and 0.4 years, respectively (log rank P<.001). In men in the dialysis subcohort, median survival was 5.1 years; in men with prostate, lung, colorectal, and pancreatic cancer, median survival was 15.6, 0.6, 6.9, and 0.4 years, respectively (log rank P<.001).

In unadjusted analysis in women, dialysis had a worse 5-year survival probability (49.8%; 95% confidence interval [CI], 48.8%-50.7%) compared with breast (82.1%; 95% CI, 81.9%-82.4%) and colorectal (56.8%; 95% CI, 56.3%-57.2%) cancer. Women on maintenance dialysis had better 5-year survival probability than women with lung (19.7%; 95% CI, 19.4%-20.1%) and pancreatic cancer (9.4%; 95% CI, 8.9%-10.0%). In men, 5-year survival was worse in dialysis (50.8%; 95% CI, 50.1%-51.6%) compared with prostate (83.3%; 95% CI, 83.1%-83.5%) and colorectal (56.1%; 95% CI, 55.7%-56.5%) cancer; 5-year survival among dialysis patients was better than in men with lung (14.0%; 95% CI, 13.7%-14.3%) and pancreatic (9.1%; 95% CI, 8.5%-9.7%) cancer.

Rates were similar in analyses of 1- and 10-year survival probabilities. In both men and women, colorectal cancer had a lower 1-year survival probability compared with dialysis patients. Following adjustments for clinical characteristics, results were also similar for lung and pancreatic cancer in both men and women; dialysis patients had a higher rate of death at ≥4 years after diagnosis. Women and men ≥70 years of age with incident kidney failure treated with maintenance dialysis had unadjusted 10-year survival probabilities that were comparable to pancreatic and lung cancer.

The researchers cited some limitations to the study, including the possibility that the results are not generalizable to other countries and races; lack of data on cancer stage; and the possibility of unmeasured confounding.

The researchers said, “In conclusion, mortality is high in the maintenance dialysis population, with mortality being particularly high in the elderly dialysis population (aged ≥70 years). Many dialysis patients experienced a higher probability of death compared with several common cancers. The results of this study highlight the urgent need to fund, develop, and test interventions to improve survival in maintenance dialysis patients and include such patients in trials for other conditions (e.g., coronary heart disease). Furthermore, results highlight the need for advance care planning and can be used to facilitate this planning in elderly patients beginning treatment with maintenance dialysis.”

Takeaway Points

  1. Researchers in Canada conducted a study to compare survival of patients on maintenance dialysis with that of patients with common cancers.
  2. In women, patients on maintenance dialysis had worse unadjusted 5-year survival than breast and colorectal cancer, but better survival than patients with lung and pancreatic cancer.
  3. In men, unadjusted 5-year survival was worse in patients on dialysis compared with prostate and colorectal cancer, but better than in patients with lung and pancreatic cancer.