PRAGUE—For patients with congestive heart failure (CHF) who initiate maintenance dialysis, those starting on peritoneal dialysis (PD) have a significantly greater death risk than those starting on hemodialysis (HD), according to a French study.
About one third of patients begin dialysis with a pre-existing CHF. Their median survival is less than three years.
Medical student Florence Sens of the Department of Nephrology at Lyon-Sud University Hospital in Pierre Bénite, Rhone-Alpes, France, and colleagues analyzed data from the French Renal Epidemiology and Information Network (REIN), a regional and national network that includes all dialysis patients.
The researchers compared survival for all incident dialysis patients who started planned chronic dialysis from 2002 to 2008 with a history of CHF. A total of 4,401 patients were included: 3,468 on HD and 933 on PD. Dialysis modality that patients were using on day 90 after starting dialysis determined which group they were in. The main analysis considered survival from first dialysis session until death, renal transplantation, or the close of the study at the end of 2008, whichever occurred first.
The PD population was about three years older and had about 4% more patients in New York Heart Association (NYHA) classes 3 and 4 heart failure. Type 2 diabetes, peripheral vascular disease, stage III and IV peripheral vascular disease and cigarette smoking were less common among PD patients. At first dialysis session, the rate of patients starting dialysis with an estimated glomerular filtration rate of 15 mL/min/1.73m² or higher was twofold greater in the PD group than the HD group (27.8 vs. 13.1%).
Over four years of follow-up, and after adjusting for multiple confounders, the PD patients had a significant 47% increased risk of death as compared with HD patients, Sens reported at the 48th Congress of the European Renal Association-European Dialysis and Transplant Association.
Other variables significantly associated with death were age, NYHA stage 3/4 versus 1/2 CHF, central venous catheter use at dialysis initiation, peripheral vascular disease, liver cirrhosis, and behavioral disturbances. In subgroup analyses that stratified patients by eGFR, NYHA CHF class, gender, and age and diabetes status, the results were the same as for the entire study population.
The study has several limitations, the first being the reproducibility of the CHF diagnosis. A validation study was performed by the researchers, that showed a misclassification rate of less than 3%. Second, the study excluded patients with unplanned first dialysis session to avoid a selection bias, with 93% of them starting on HD. Third, the registry did not provide data on all possible confounders. “So we have to be very cautious with interpretation of these results,” Sens advised.
This study highlights the need for further study to determine the pathophysiology underlying the findings as well as for a randomized controlled trial to confirm the results, she said.
Hydration status a factor
Karel Leunissen, MD, Professor of Nephrology at Maastricht University Hospital in Maastricht, Netherlands, who moderated the session at which Sens presented study findings, raised the issue of hydration status, which he said could be a large confounder in the study because the clinical diagnosis of CHF greatly depends on the hydration status of the patient.
He told Renal & Urology News that “it’s very important that you fix the hydration status of these patients in such a way that you are in the [ascending part] of the Starling curve and not in the descending line.” Relieving fluid overload increases stroke volume and cardiac output and leads to improved kidney function, he explained.
By this reasoning, PD patients should do at least as well as HD patients because PD removes fluid constantly, he said, unlike thrice weekly HD, where fluid overload may occur in the interdialytic period.
Therefore, Dr. Leunissen questioned the pathophysiological basis of better survival observed with HD as seen in the French study. Sens postulated that fluid removal is less predictable and may have been inadequate in PD patients. The other factor that could have influenced results is the difference in medical monitoring related to home care for the PD patients.