Statin Lowers Cardiac Events in Diabetics on Hemodialysis With High Cholesterol

From Medscape Medical News
Statin Lowers Cardiac Events in Diabetics on Hemodialysis With High Cholesterol Levels

Daniel M. Keller, PhD

July 8, 2010 (Munich, Germany) — From a post hoc analysis of the German Diabetes Dialysis Study (4D) of atorvastatin and low-density-lipoprotein (LDL) cholesterol, German researchers report that atorvastatin significantly decreased fatal and nonfatal cardiac events and all-cause mortality in patients with type 2 diabetes on hemodialysis who had the highest levels of LDL cholesterol in the study population.

Only 3 patients in this highest LDL cholesterol group would need to be treated for 5 years to prevent 1 major adverse cardiac event (MACE).

The 4D study was conducted to see if the LDL cholesterol level at baseline predicted cardiovascular (CV) events and all-cause mortality in this patient group. In addition, the trial investigated whether the effect of atorvastatin on clinical outcomes depended on LDL cholesterol level at baseline. Hemodialysis patients are at markedly increased risk for CV events, and statin drugs have been shown to reduce the rate of CV events in many patient populations in large series of prospective studies. Lipid-lowering appears to lower CV risk in the early stages of chronic kidney disease and in renal transplant recipients.

"However, the significance of lipids to cardiovascular disease in hemodialysis is only poorly understood so far," according to Winfried März, MD, from Synlab in Heidelberg, Germany, and the Medical University of Graz, Austria.

Reporting for the 4D investigators here at the XLVII European Renal Association-European Dialysis and Transplant Association Congress, Dr. März said that in general in statin trials, the relative and absolute benefits of statins are greater in patients with very high baseline LDL cholesterol levels. But 2 prospective randomized placebo-controlled trials in patients with end-stage renal disease, 4D and AURORA, had basically negative outcomes.

4D involved 1255 men and women with type 2 diabetes on hemodialysis with LDL cholesterol levels of 80 to 190 mg/dL and triglyceride levels below 1000 mg/dL who were randomly assigned to atorvastatin 20 mg/day or placebo. The primary combined end point was CV mortality and nonfatal myocardial infarction and stroke. The trial continued until 424 primary end point events had occurred.

AURORA involved 2776 patients (27% with diabetes mellitus) on hemodialysis or hemofiltration treated with rosuvastatin 10 mg/day or placebo. The end points were the same as in 4D.

Dr. März noted that it was interesting that the amount of LDL cholesterol reduction was the same in both trials (a median of 42% in 4D and a mean of 43% in AURORA). The baseline LDL cholesterol level was 120 mg/dL in 4D but only 100 mg/dL in AURORA.

4D and AURORA Outcomes Similar

"The overall primary results of 4D and AURORA were almost identical," Dr. März told the audience during a late-breaking trial session. There was an 8% reduction in the incidence rate of the primary end point in 4D and a 4% reduction in it in AURORA for the active drug compared with placebo over 6 years of follow-up, neither of which was statistically significant (P = .37 and P = .59, respectively).

Dr. März concluded that overall there was no effect of cholesterol-lowering with statins in these hemodialysis patients in the 2 trials.

He said it is known from meta-analyses that there is an approximately 25% decrease in CV events for each 39 mg/dL (1 mmol/L) reduction in LDL cholesterol levels. Furthermore, the relative risk reduction achieved by a given absolute reduction in LDL cholesterol is uniform across all strata of baseline values. Finally, absolute reductions in LDL cholesterol produced by a given dose of a statin are larger at higher than at lower baseline LDL cholesterol levels.

"So we set up the hypothesis that the cardiovascular benefit of atorvastatin in 4D was greater at higher LDL cholesterol baseline levels than at lower ones," Dr. März said. "Second, we were interested to see the effect of LDL cholesterol as a predictor of cardiovascular events." These hypotheses formed the basis of the post hoc analysis in the 4D trial. The investigators stratified the 4D population into quartiles according to baseline LDL cholesterol levels.

Results of Post Hoc Analysis

The 2 end points that approached statistical significance for atorvastatin vs placebo with increasing baseline LDL cholesterol levels were the primary composite end point (P = .061) and all-cause death (P = .079). "So there was a slight trend toward decreased risk of vascular end points in relation to LDL cholesterol," Dr. März reported.

In all the quartiles, the relative reduction in LDL cholesterol was approximately 35%. So the absolute reduction was higher in the quartile with the highest baseline LDL cholesterol level (53.1 mg/dL) than in the quartile with the lowest (35.3 mg/dL). This LDL cholesterol reduction in highest quartile resulted in a 35% reduction in the primary composite end point for atorvastatin vs placebo at 6 years (hazard ratio

, 0.65; confidence interval, 0.44 - 0.96; P = .032).

Similarly, all secondary end points for the highest quartile were significant at P < .05 when atorvastatin was compared with placebo over 6 years: cardiac deaths, nonfatal myocardial infarction, sudden cardiac death, and all cardiac events combined (HR, 0.58, 0.50, 0.48, and 0.54, respectively).

"With regard to total mortality, we also saw a statistically significant reduction amounting to approximately 30%. So it's not only the vascular end points. It's also total mortality, which is significantly reduced with treatment," Dr. März concluded.

To calculate a number needed to treat (NNT), he normalized the 4D results to a treatment duration of 5 years and used MACE (nonfatal myocardial infarction, cardiac death, revascularization, etc.) — a common vascular end point in statin trials. In previous statin trials, the NNT has ranged from 3 to 10 over 5 years to prevent 1 event.

"Just looking at the primary composite end point of 4D, we draw the conclusion that you have to treat 4 patients on dialysis with diabetes and an LDL cholesterol level above 145 mg/dL [higherst quartile] with atorvastatin for 5 years," Dr. März calculated.

"If you go to deaths from all causes, you can prevent 1 death in 5 years by treating 6 patients." To prevent 1 MACE of any form over 5 years, the NNT is 3, "the lowest one that we've ever seen in any statin trial," he said.

"So my conclusion in regard to these data is that statin treatment should be considered in dialysis patients with type 2 diabetes with appropriately elevated LDL cholesterol levels, which means LDL cholesterol of more than 145 mg/dL (or 3.8 mmol/L)," Dr. März advised. He speculated that it would even be good to set the threshold for treatment at 130 mg/dL, an LDL cholesterol level between the highest and second-highest quartiles.

After Dr. März's presentation of the 4D post hoc analysis, one of the AURORA investigators in the audience commented that his trial group found very similar results on outcomes when they looked at the effect of rosuvastatin administered to their study subjects with the highest baseline LDL cholesterol levels.

Session moderator Johannes Mann, MD, head of nephrology at the Schwabing General Hospital in Munich, Germany, took some issue with Dr. März's recommendation to treat patients with atorvastatin based on the results of this post hoc analysis of the 4D trial data. Dr. Mann was not involved in the 4D study.

"It's difficult to make recommendations on a subgroup analysis that was not prespecified," Dr. Mann said. "I would also interpret the study somewhat differently, in a way saying that the higher the LDL cholesterol, the bigger the benefit from the drug. But where it starts and where it ends is difficult to say," he told Medscape Medical News. "Just to pick one group and not show the trend is not the right way to analyze it. Furthermore, before making recommendations, I would really want to see a meta-analysis of the same subgroups of the [other] trials that have been done."

The 4D study was supported by Pfizer Inc. None of the members of the steering committee, the end point adjudication committee, the data and safety monitoring committee, or any of the investigators received honoraria for participating in the study. The post hoc analysis was supported by Pfizer with an unrestricted grant to Dr. März solely covering the expenses for data management and statistics. Dr. März did not receive honoraria for the presentation or any other honoraria, but reports being a member of the speakers' bureau and a consultant with Pfizer Inc., but not on the 4D study. Dr. Mann has disclosed no relevant financial relationships.

XLVII European Renal Association-European Dialysis and Transplant Association (ERA-EDTA) Congress. Presented June 26, 2010.

Medscape Medical News © 2010 WebMD, LLC