Daniel M. Keller, PhD

July 11, 2011

July 11, 2011 (Prague, Czech Republic) — There was no difference in overall survival or nonfatal cardiovascular (CV) events overall between postdilution online hemodiafiltration (HDF) and high-flux hemodialysis (HD) in a large population-based study of patients with end-stage renal disease. This result comes from the first prospective randomized controlled study with adequate power to compare the effects of the 2 strategies.

However, HDF treatment with a substitution volume of more than 17.4 L resulted in better CV and overall survival, Fatih Kircelli, MD, from the division of nephrology at Ege University School of Medicine in Izmir, Turkey, told delegates here at the XLVIII European Renal Association-European Dialysis and Transplant Association Congress during a late-breaking clinical trials session.

Hemodialysis removes low-molecular-weight uremic waste products from the blood, mainly by diffusion across a membrane, but is limited in its ability to clear larger uremic toxins. In HDF, hemodialysis is combined with hemofiltration, in which there is a high rate of movement of solutes and water from blood to dialysate, by convection under the influence of positive hydrostatic pressure. A replacement fluid keeps the patient in fluid balance. The combination of the techniques results in good removal of both large- and small-molecule solutes from the blood.

The morbidity and mortality of patients receiving HD remain unacceptably high, Dr. Kircelli noted. Based on the supposed advantages of HDF — including better clearance of middle- and larger-sized and protein-bound uremic solutes, better β2-microglobulin removal, better biocompatibility, and reduced inflammatory markers — he and his colleagues undertook the Turkish HDF Study, in which the 2 renal replacement modalities were compared. The primary outcomes were a composite of all-cause mortality and new nonfatal CV events. A secondary outcome was CV mortality.

Patients at least 18 years of age were recruited from 10 hemodialysis centers. After screening, 782 patients were randomized to HDF (n = 391) or HD (n = 391) treatment between January 2007 and March 2008, with a planned follow-up of 24 months. All underwent thrice-weekly maintenance bicarbonate HD for 12 hours/week before randomization, and all had achieved a mean single-pool Kt/V of 1.2 or greater. Participants with a temporary catheter or insufficient vascular access (less than 250 mL/min) were excluded.

During the study, patients underwent HDF or HD for 240 minutes thrice weekly, with a blood flow rate of 250 to 400 mL/min and a target substitution volume of more than 15 L per session. The HDF and HD groups were well matched for age (56 years), sex, diabetes (33% to 36%), vascular access (95% arteriovenous fistula), CV history (25.7% to 27.2%), dialysis duration (about 58 months), body mass index, adequate blood pressure and anemia control, and interdialytic weight gain.

On follow-up, the blood flow rate for the HDF group was 318 ± 27 mL/min and for the HD group was 303 ± 32 mL/min. The mean substitution volume in both groups was 17.2 ± 1.2 L per session (range, 9.8 to 20.3 L), with 93% achieving more than 15 L per session.

Forty patients in the HDF group transferred to the HD group. In addition, 59 patients in the HDF group and 84 patients in the HD group transferred to another center.

The HDF group had significantly higher systolic blood pressure (129 vs 126 mm Hg; < .001), similar antihypertensive medication use, and higher interdialytic weight gain as a percent of body weight (3.5% vs 3.2%; = .01).

Higher Efficiency of HDF Led to Better Overall Outcomes

"There was no difference with regard to the primary outcome" between groups, Dr. Kircelli reported, with 61 patients in the HDF group and 73 in the HD group achieving the primary end point (= .25). Event-free survival was about 80% at 40 months (hazard ratio, 0.82; 95% confidence interval [CI], 0.59 to 1.16; = .28).

Similarly, overall mortality was 6.98 deaths in the HDF group and 8.80 deaths in the HD group per 100 patient-years of follow-up (= .19), and CV mortality was 4.30 and 5.96, respectively, per 100 patient-years (= .40); there were no differences in hospitalization rates or intradialytic hypotension.

"Online HDF patients had a higher urea reduction ratio and Kt/V values [both < .001], compared to [HD] patients, so we consider it to be associated with better dialysis adequacy and small solute removal," Dr. Kircelli said. The HDF patients required a lower dose of erythropoiesis-stimulating agent (2282 vs 2852 U/week; = .001).

In subgroup analyses, "we found no difference in diabetics, patients with cardiovascular disease history, or patients with serum albumin levels less than 4 [g/dL] regarding primary outcome, overall, and cardiovascular survival," he said.

When the patients were stratified by median substitution volume (high-efficiency group, above 17.4 L per session; low-efficiency group, at or below 17.4 L per session), there was no statistically significant difference in the primary outcome. "However, with regard to overall survival, high-efficiency online HDF was associated with better overall survival," Dr. Kircelli told the audience (= .03), "and they had better cardiovascular survival" (= .002).

The high-efficiency group had a 46% reduction in risk for overall mortality and a 71% risk reduction for CV mortality. More patients had diabetes, lower serum albumin levels, and higher hemoglobin levels in the low-efficiency group than in the high-efficiency group.

Dr. Kircelli concluded that postdilutional online HDF provides better overall and CV survival when performed in high-efficiency mode, and reaching higher convection volumes "seems to be an essential issue in postdilution online HDF" treatment.

Markus Ketteler, MD, professor of medicine and chief of the division of nephrology at Klinikum Coburg, in Germany, who was not involved in the study, told Medscape Medical News that a comment after the presentation by Wim Van Biesen, MD, from University Hospital Ghent in Belgium "was an extremely valid one." Dr. Van Biesen had asked Dr. Kircelli if patients were assigned to high- or low-efficiency HDF, to which Dr. Kircelli responded that they were stratified post hoc according to their achieved fluid substitution volumes. Dr. Van Biesen then remarked that "patients with the better vascular access had better survival. That's basically what the study tells us."

Dr. Ketteler explained that if the stratification was made "just on the capacity of the shunts or the fistulas to transport larger volumes, then it is very likely or quite likely that patients were selected who had a better circulation, [and] that was a selection bias of maybe the more healthy ones," possibly explaining the better outcomes seen in the high-efficiency HDF group.

Furthermore, because the efficiency of HDF was not a prespecified analysis and because patients were not randomized to a high- or low-volume group, he advised that "one should focus on the primary end point results, which showed no difference between the treatment regimens."

Dr. Ketteler suggested that a key issue is to analyze the groups' comorbidities. "If they match between high and low volume, that would be a strong argument for high-volume HDF," he said. "And finally, one could easily look at the volumes that were driven through the fistula during dialysis, the average blood flow that was obtained there, because that could be a surrogate of improved circulation [and] improved condition of the patient."

The study was supported by the European Nephrology and Dialysis Institute and Fresenius Medical Care. Dr. Kircelli and Dr. Ketteler have disclosed no relevant financial relationships.

LVIII European Renal Association-European Dialysis and Transplant Association (ERA-EDTA) Congress. Presented June 24, 2011.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: