Clinical implications of statin therapy in patients undergoing hemodialysis

Abstract

Purpose

The clinical implications regarding the use of statins in patients with end-stage renal disease (ESRD) undergoing hemodialysis are explored.

Summary

The majority of the evidence reviewed from randomized controlled trials and recent meta-analyses suggest that there is minimal to no benefit of statin therapy for reducing the risk of coronary heart disease (CHD), including cardiovascular events and mortality, for statin-naive patients undergoing hemodialysis. The Kidney Disease Outcomes Quality Initiative (KDOQI) 2003 dyslipidemia guidelines recommended that patients with ESRD receive a statin to reach a goal low-density-lipoprotein (LDL) cholesterol concentration of <100 mg/dL; however, there was no distinction between nondialysis and dialysis patients, and newer evidence has since been published. Although KDOQI released 2012 guidelines that recommended against the initiation of statins in dialysis patients due to the lack of evidence to support benefit, the guidelines were specific for diabetic dialysis patients. Clinicians should use their clinical judgment and weigh the risks and benefits from the available evidence when deciding whether to initiate statins in hemodialysis patients. A statin may be warranted for secondary prevention of cardiovascular events or in younger hemodialysis patients who have a longer life expectancy.

Conclusion

The available literature does not support the initiation of statins in hemodialysis patients who were not receiving statin therapy before requiring hemodialysis. At this time, there are no conclusive data to support discontinuation of statins in ESRD patients on hemodialysis receiving statins for either primary or secondary prevention of CHD.

Patients with chronic kidney disease (CKD) have a high risk of cardiovascular events and death, and these risks increase with progression of the disease.¹ In addition to the traditional risk factors for cardiovascular disease, several nontraditional risk factors have been identified for patients with CKD, including albuminuria, inflammation, anemia, calcium or phosphate imbalances, volume overload, and electrolyte disturbances.² The National Kidney Foundation Kidney Disease Outcomes Quality Initiative (KDOQI) considers CKD to be a coronary heart disease (CHD) risk equivalent, which translates to a 20% 10-year cumulative risk of CHD.³ To this end, the 2003 KDOQI dyslipidemia guidelines recommended a goal low-density-lipoprotein (LDL) cholesterol concentration of <100 mg/dL for patients with end-stage renal disease (ESRD) (i.e., stage 5 CKD, defined as a glomerular filtration rate of <15 mL/min/1.73 m² or requiring hemodialysis) for the primary prevention of CHD.³ However, these guidelines were based on extrapolated data from trials in the general population due to the lack of randomized trials in CKD patients.³ More recently, the 2012 KDOQI guidelines for diabetes and CKD recommended against the initiation of statins in patients with diabetes who are undergoing hemodialysis.⁴ The Kidney Disease Improving Global Outcomes (KDIGO) Lipid Work Group has released the KDIGO Clinical Practice Guideline for Lipid Management in Chronic Kidney Disease and recommends against starting statins in patients undergoing hemodialysis. It also recommends to continue statin therapy in patients who were taking a statin before the initiation of hemodialysis. Interestingly, statin therapy is not recommended to help reach a goal LDL cholesterol level.⁵

The risk of all-cause mortality for ESRD patients undergoing hemodialysis is up to eight times that of the general population.¹ Interestingly, the most common cause of death reported by the United States Renal Data System (USRDS) in ESRD patients from 2008 to 2010 was arrhythmia or cardiac arrest, not atherosclerotic events.¹ Also, within the first year of hemodialysis, the majority of deaths reported were due to noncardiovascular causes.¹ Despite these reports, 34% of hemodialysis patients received statins for primary prevention of CHD in 2010.¹ However, the evidence regarding the efficacy of statins for survival and cardiovascular events, as well as the safety of statin therapy for ESRD patients, has been conflicting. The Dialysis Outcomes and Practice Patterns Study, a prospective, observational study, evaluated approximately 7300 hemodialysis patients and found that those treated with statins had a 31% lower relative risk of death and a 23% lower risk of cardiac death compared with patients who were not prescribed statins.⁶ These observational results were confirmed by a meta-analysis showing that statins, when compared with placebo, significantly decreased fatal and nonfatal cardiovascular events despite the stage of CKD; however, the meta-analysis was limited by its inclusion of small and low-quality studies.⁷ A 2009 Cochrane review evaluated the effect of statins versus placebo in patients requiring hemodialysis.⁸ Although a nonsignificant reduction in mortality was found, the rate of nonfatal cardiovascular events was significantly reduced with statins. There have also been concerns regarding the safety of administering statins to patients with ESRD, as the decrease in renal excretion may increase the risk of myopathy or rhabdomyolysis, but this has not been confirmed in trials.⁴

Since the publication of the previously conflicting results, three meta-analyses further investigated the use of statins to reduce the rate of mortality and cardiovascular events in patients undergoing hemodialysis. We evaluated the recent meta-analyses and the three major trials that may have strongly affected the results to determine the clinical implications of the use of statin use in patients with ESRD on hemodialysis.

Randomized controlled trials

German diabetes and dialysis study

Wanner and colleagues⁹ conducted this prospective, randomized, placebo-controlled trial (also known as the 4D trial) to evaluate the effects of atorvastatin 20 mg daily on cardiovascular outcomes in adult patients with type 2 diabetes mellitus who were undergoing hemodialysis (n = 1255). There were no significant differences in baseline demographics, with a mean age of 65 years, approximately one-third of patients having a history of cardiovascular disease, and about 80% of patients receiving erythropoietin-stimulating agents (ESAs) in each group. The composite primary endpoint (death from cardiac causes, fatal stroke, nonfatal myocardial infarction [MI], or nonfatal stroke) occurred in 37% and 38% of patients in the atorvastatin and placebo groups, respectively, with no statistical difference between groups (relative risk [RR], 0.92; 95% confidence interval [CI], 0.77–1.10; p = 0.37). For all cardiac events combined (death from cardiac causes, nonfatal MI, percutaneous transluminal coronary angioplasty, coronary artery bypass graft, and other interventions for CHD), the events were significantly reduced in the atorvastatin group compared with placebo (p = 0.03), with a number needed to treat of 17.

A post hoc analysis of this trial was conducted to assess the benefit of atorvastatin in patients with higher baseline LDL cholesterol levels.¹⁰ Patients were stratified into four groups with the following LDL cholesterol ranges: <103, 104–122, 123–144, and >145 mg/dL. This analysis demonstrated that in patients with type 2 diabetes undergoing hemodialysis who had the highest LDL cholesterol concentrations (>145 mg/dL), there was a statistically significant reduction in cardiac death (hazard ratio [HR], 0.58; 95% CI, 0.34–0.99) and sudden cardiac death (HR, 0.48; 95% CI, 0.25–0.94). There were no significant differences for the lower three LDL quartiles.

The strengths of this trial included the large sample size, the use of a placebo group, and the long duration of follow-up (mean duration, 3.91 years).⁹ One of the limitations noted by the trial investigators was that approximately 15% of patients in the placebo group were started on non-study statins during the course of the trial, which could have confounded the results leading to fewer primary composite events in the placebo group, considering there were many patients receiving active drug. In addition, this study was conducted before the 2011 Food and Drug Administration (FDA) warning regarding the increased risk of cardiovascular events, including fatal and nonfatal strokes, in patients receiving ESAs when targeting near-normal hemoglobin levels.¹¹ Due to the timing of this trial, it is unknown if ESA use was a confounder. Also, the percentage of patients with a diagnosis of atrial fibrillation (AF) at baseline was not described, despite the fact that over 30% of patients in each group had a diagnosis of congestive heart failure, a known risk factor for AF. Studies have demonstrated that upward of approximately 30% of patients with congestive heart failure will develop AF.¹² AF is associated with a nearly fivefold increased risk of stroke,¹³ and the frequency of AF and the anticoagulation status of patients in the trial were not described. One interesting finding was the reduction in all cardiac events, suggesting that there may be a cost reduction for hemodialysis patients receiving statins through a reduction in health care costs related to cardiovascular events.

AURORA

AURORA (A Study to Evaluate the Use of Rosuvastatin in Subjects on Regular Hemodialysis: An Assessment of Survival and Cardiovascular Events) was a randomized, double-blind, placebo-controlled, multicenter trial that evaluated the effect of rosuvastatin 10 mg daily on cardiovascular outcomes in patients receiving hemodialysis (n = 2773).¹⁴ Baseline characteristics were similar between groups. The frequency of the primary composite outcome of nonfatal MI or stroke and cardiovascular death did not differ significantly between the rosuvastatin and placebo groups, occurring in 28.5% and 29.5% of patients, respectively (HR, 0.96; 95% CI, 0.8–1.11; p = 0.59). In addition, there was no significant difference between groups in any of the prespecified secondary outcomes, including all-cause mortality.

An important strength of this trial was its large sample size; however, the study and some limitations that may impede the extrapolation of its results to all hemodialysis patients. This study shares a limitation with the 4D trial conducted by Wanner et al.,⁹ in that the cardiovascular risk of ESAs was not known at the time AURORA was conducted. At baseline, patients had a mean hemoglobin concentration of 11.7 g/dL in both groups and more than 80% of patients were receiving ESAs. In addition, the overall baseline characteristics of the patients in this trial are not highly representative of the majority of patients with ESRD. Approximately 85% of patients in AURORA were Caucasian, with only about 20% of patients progressing to ESRD secondary to diabetes mellitus. These characteristics limit the generalizability of the results. Lastly, the investigators conducted an updated power calculation during the trial and the required number of primary events was not met to achieve power, leading to the risk of a Type II error. These limitations make it difficult to conclude that there is no benefit on major cardiovascular outcomes when initiating statin therapy in hemodialysis patients.

Study of Heart and Renal Protection

The Study of Heart and Renal Protection (SHARP) evaluated the effects of simvastatin 20 mg daily plus ezetimibe 10 mg daily on cardiovascular outcomes compared to placebo in 9270 CKD patients, including those undergoing dialysis.¹⁵ Baseline characteristics were similar between groups; however, 27% of patients were undergoing hemodialysis, and 5% were receiving peritoneal dialysis. There was a significant reduction in the primary composite outcome of major vascular events (cardiac death or nonfatal MI, stroke, or arterial revascularization excluding hemodialysis access), which occurred in 11.3% and 13.4% of all patients in the simvastatin plus ezetimibe and placebo groups, respectively (RR, 0.83; 95% CI, 0.74– 0.94; p = 0.0021; number needed to treat = 48). There was no significant reduction in major atherosclerotic events for dialysis patients; however, the trial was not powered to detect a reduction in this population (RR, 0.90; 95% CI, 0.75–1.08). A notable subgroup analysis was conducted to evaluate the relationship between the baseline LDL cholesterol level and the occurrence of the primary outcome. This analysis revealed a significant reduction in the primary outcome in all patients receiving simvastatin and ezetimibe with a baseline LDL concentration of >117 mg/dL (RR, 0.73; 95% CI, 0.60–0.88). This reduction is important to note, because this outcome was not evaluated in dialysis patients alone. However, hemodialysis patients comprised about one third of the initial SHARP population, and about 30% of the nonhemodialysis patients transitioned to hemodialysis during the study period.

One of the major limitations of this trial was the fact that two agents were combined and compared with placebo, making it impossible to determine which agent led to the effects seen in the trial. Further, the trial included both nondialysis and dialysis patients and had insufficient power to separately analyze the primary outcome in hemodialysis patients. Although 27% of patients were receiving hemodialysis at baseline, it is difficult to extrapolate these results to clinical practice, given the lack of sufficient statistical power. Also, the inclusion of peritoneal dialysis patients may have confounded the results, because the lipoprotein profile in these patients may be more atherogenic (increased LDL cholesterol and total cholesterol) than in hemodialysis patients.^16,17 It has been suggested that the lipid abnormalities in peritoneal dialysis patients may be caused by increased absorption of glucose from the peritoneal dialysis fluid, protein loss that stimulates hepatic very-low-density lipoprotein synthesis, and higher levels of apolipoprotein B.^18,19 However, the investigators reported subgroup analyses of the dialysis population, and the results were consistent, with no benefit found when separated by mode of dialysis (RR, 0.70; 95% CI, 0.46–1.08 with peritoneal dialysis versus RR, 0.95; 95% CI, 0.78–1.15 with hemodialysis).¹⁵

As seen with previous trials, SHARP was conducted before the FDA warnings regarding target hemoglobin concentration and ESA use, and there was no description of baseline or endpoint hemoglobin concentration in patients receiving ESAs. Prior statin use was also not described, thus limiting our interpretation of confounders. Given the lack of statistical power, it cannot be concluded from the trial results that initiating statin therapy in hemodialysis patients will improve cardiovascular outcomes. This does, however, generate the hypothesis that lipid-lowering therapy, including a statin, may be beneficial in hemodialysis patients, especially if therapy is initiated before the patient transitions to hemodialysis.

Meta-analyses

Palmer et al

Palmer et al.²⁰ conducted a meta-analysis to evaluate whether the safety and efficacy of statin therapy are associated with the stage of CKD. Eighty trials with a total of 51,099 patients with CKD were selected for inclusion. Among the trials included for evaluation, 20 included patients receiving dialysis. Of the 7,982 dialysis patients, approximately 88% were pooled from the 4D trial, AURORA, and SHARP. The overall median follow-up for all of the studies analyzed was six months (range, two months to 5.5 years). The overall median duration of follow-up for the dialysis trials only was not reported; however, for the individual trials, the median duration ranged between two months and 4.9 years. The mean age varied from 46 to 70 years for the dialysis patients included. Significant heterogeneity among trials was indicated by a p of <0.10, with an I² of >50% representing high heterogeneity.²⁰ A random-effects model was used to report treatment-effect estimates.

The all-cause mortality benefit of statins for dialysis patients was not significant compared with no treatment or placebo (RR, 0.96; 95% CI, 0.88–1.04). There were no significant differences found in the rates of cardiovascular mortality (RR, 0.94; 95% CI, 0.82–1.07) and major cardiovascular events (RR, 0.95; 95% CI, 0.87–1.03) for dialysis patients treated with statins. Also, there was no benefit of statin use for fatal or nonfatal stroke in dialysis patients compared with no treatment or placebo (RR, 1.30; 95% CI, 0.79–2.11).²⁰

There was no heterogeneity (I² = 0%) reported for trials including only dialysis patients; however, the results were analyzed using a random-effects model. A fixed-effects model may have been a more appropriate method to analyze these results, as it assumes that the treatment effect does not vary among studies. One strength of this meta-analysis was the separate analysis without the inclusion of SHARP due to dual simvastatin and ezetimibe therapy, which showed consistent results for not reducing major cardiovascular events for dialysis patients treated with statins compared with placebo (RR, 0.96; 95% CI, 0.85–1.08). As described with SHARP, one major limitation of this meta-analysis was the inclusion of both peritoneal dialysis and hemodialysis patients, which may have confounded the results due to the different lipoprotein abnormalities and response to statin therapies. A total of nine dialysis trials included peritoneal dialysis patients and contributed approximately 12% of the dialysis population analyzed. Unlike SHARP, the investigators did not conduct separate subgroup analyses. Other limitations of this meta-analysis include the lack of quality trials, as only a third of the trials analyzed had minimal risk for bias, and fewer than half of the trials reported adverse effects of statin therapy. The evaluation of the long-term efficacy of statins may not have been adequate due to the inclusion of studies with a short duration of follow-up (ranging from two months to 5.5 years). Lastly, the investigators were unable to evaluate the relationship between decreases in cholesterol levels and clinical outcomes due to insufficient power.²⁰

Hou et al

In this meta-analysis, Hou et al.²¹ aimed to examine effects of statin therapy on renal and cardiovascular outcomes in patients with CKD, including dialysis patients. Randomized controlled trials were included in which statin therapy was given for at least six months in patients with CKD. A total of 31 studies were analyzed, including 48,429 patients with CKD. Nine trials exclusively evaluated dialysis patients (4,458 patients), and 3 trials included a combination of nondialysis and dialysis patients (3,239 dialysis patients). Of the 7,697 dialysis patients included, over 90% were from the 4D trial, AURORA, and SHARP. Dialysis patients with a broad range of ages were included, with the mean age varying from 49 to 68 years. The median duration of follow-up was nine months (range, two months to four years). The primary outcome evaluated was major cardiovascular events, defined as fatal or nonfatal MI, fatal or nonfatal stroke, revascularization procedures, cardiovascular death, and heart failure.

Compared with placebo, statin therapy was shown to significantly reduce the risk of cardiovascular events in ESRD dialysis and nondialysis patients by 8% (RR, 0.92; 95% CI, 0.85–0.99; p < 0.031; I² = 0%). Heterogeneity in the effect of statin therapy on major cardiovascular events across trials was assessed by subgroup analysis and meta-regression. When ESRD patients were separated from ESRD patients not receiving dialysis, the reductions in risk were 7% (RR, 0.93; 95% CI, 0.86–1.00) and 18% (RR, 0.82; 95% CI, 0.60–1.11), respectively, and statistical significance was lost. Of note, the number of ESRD patients not undergoing dialysis included in this analysis was low (1,221 compared with 7,289 dialysis patients), and all of these patients were drawn from SHARP. Although the reduction in risk of major cardiovascular events in ESRD patients undergoing dialysis was not statistically significant, it correlated with an approximately 10% risk reduction per 1 mmol/L of LDL cholesterol lowered (RR, 0.90; 95% CI, 0.83–0.97; p = 0.010), with little heterogeneity among trials (I² = 2.6%, p = 0.392). The fact that the risk reduction per 1 mmol/L of LDL cholesterol decreased did not correspond to a significant risk reduction in major cardiovascular events in the dialysis population may further support the theory that risk in these patients is associated with nonatherosclerotic factors. For all other outcomes assessed (all-cause mortality, cardiovascular death, coronary events, and stroke), there were no differences reported for dialysis patients receiving statin therapy compared to placebo.

A strength of this meta-analysis was its examination of the effect of statins across all stages of CKD. As a result, the authors were able to demonstrate a decreasing trend in reduction of cardiovascular risk with statin use as CKD progressed to ESRD. Similar to the findings of Palmer et al.,²⁰ the results of this meta-analysis are limited by a lack of quality trials and the inclusion of many trials that were possibly of insufficient duration to assess long-term effects of statin therapy. Another possible limitation is the inclusion of peritoneal dialysis patients. Six of the included trials (746 patients) contained patients receiving peritoneal dialysis, comprising approximately 10% of the dialysis population analyzed. As noted above, these patients may receive greater benefit from statin therapy than do hemodialysis patients. Despite these limitations, Hou et al.²¹ concluded that a benefit of statins was seen across all stages of CKD, including patients with advanced kidney disease; however, the nonsignificant findings in cardiovascular risk reduction in the dialysis population indicates that this benefit cannot be assumed to extend to dialysis patients.

Palmer and colleagues

Unlike the previous meta-analyses, Palmer and colleagues²² further compared the benefits of statins with placebo, no treatment, or another statin in ESRD patients undergoing dialysis. The investigators evaluated 25 studies, which included a total of 8,289 dialysis patients, and over 80% of patients were from the 4D trial, AURORA, and SHARP. The median follow-up duration was 6 months (range, 2–59 months). The mean age of patients varied from 43.6 to 70 years. Heterogeneity among the trials was considered to be high if the I² value was 75% or greater. Both the random-effects model and the fixed-effects model were used to estimate treatment effects.

Consistent with the previous two meta-analyses reviewed, the investigators did not find a benefit for major cardiovascular events with the administration of a statin compared with placebo or no treatment in dialysis patients (RR, 0.95; 95% CI, 0.88–1.03). Also, there were no significant differences in all-cause mortality (RR, 0.96; 95% CI, 0.90–1.02) and cardiovascular mortality (RR, 0.94; 95% CI, 0.84–1.06) for dialysis patients treated with a statin. There was no evidence of heterogeneity in the outcomes (I² = 0%).

Strengths of this meta-analysis included the investigators’ restriction of the evaluation to the dialysis population and the use of appropriate models for data analysis. Also, when duration of follow-up was restricted to 6 months or greater, the results for all-cause mortality and cardiovascular mortality did not change. Consistent with the findings of the previous meta-analyses,^20,21 13 of the 25 trials included peritoneal dialysis patients, comprising over 10% of the dialysis patients analyzed. Heterogeneity was not identified in the analyses for total or LDL cholesterol concentration based on prespecified subgroup analyses, including dialysis type; however, a separate outcome analysis by dialysis type was not reported. With the exception of the three large randomized controlled trials (4D, AURORA, and SHARP), the other studies had a median study population of 42 patients and were identified to be at high risk of bias. The smaller studies had a short duration of follow-up and may not have been adequate to determine the mortality benefit of statins; however, the mortality outcome was mainly driven by the larger studies, which had a longer duration of treatment (three to five years).²²

Safety data

All of the randomized controlled trials reported adverse effects of statin therapy, and none found an increase in the rate of myopathy, rhabdomyolysis, elevated creatine kinase, or liver disease in the statin groups versus those receiving placebo.⁹, ¹⁴, ¹⁵ However, there was a higher discontinuation rate (p = 0.02) due to myopathy for patients in the combination simvastatin and ezetimibe group.¹⁵ It is important to note that 4D, AURORA, and SHARP investigated low-to-moderate intensity statin dosing, possibly owing to the low frequency of adverse drug reactions. The meta-analyses described limited conclusions regarding the adverse outcomes of statin therapy with placebo or no treatment, as adverse outcomes were only reported in fewer than half of the studies analyzed.^20,–22 However, for the studies in which safety outcomes were reported, no significant differences in safety were found.^20,–22

Of concern, the 4D trial reported a higher risk for fatal stroke in patients receiving atorvastatin versus placebo (RR, 2.03; 95% CI, 1.05–3.93; p = 0.04); however, the statistical significance may have been due to chance because of the small number of stroke events observed.⁹ A subgroup analysis in AURORA revealed a significant increase in hemorrhagic stroke in diabetic patients receiving rosuvastatin, yet this finding only serves to generate future hypotheses.¹⁴ While the results of the meta-analyses revealed a trend toward an increased risk of fatal or nonfatal stroke for dialysis patients receiving statin therapy, there was evidence of heterogeneity among two of the three meta-analyses included, and Palmer and colleagues²² concluded that the effect was uncertain due to limited data.

Discussion

Despite a reduction of >40% in baseline LDL cholesterol in both the 4D trial and AURORA, no significant reduction in fatal cardiovascular events was found with statin therapy compared with placebo in hemodialysis patients.^9,14 It is important to consider, however, that the 4D trial targeted diabetic patients and that 25% of the population evaluated in AURORA had diabetes.^9,14

SHARP investigators found a significant benefit for the primary composite outcome (including MI, stroke, and arterial revascularization procedures) for all CKD patients, regardless of dialysis status, receiving simvastatin plus ezetimibe.¹⁵ The study lacked power to determine the efficacy of dual therapy separately in dialysis and nondialysis patients. However, it is important to consider that one third of nondialysis patients transitioned to dialysis during the study, raising the question as to whether predialysis patients may benefit from statin therapy when initiated before hemodialysis. Another major limitation of SHARP was that its results cannot be directly compared to either the 4D trial or AURORA, because it evaluated the combination of simvastatin and ezetimibe, included peritoneal dialysis patients, and evaluated revascularization procedures as part of the primary composite outcome.

Among the recent meta-analyses reviewed, Hou et al.²¹ had differing conclusions regarding the efficacy of statin therapy compared with the other two meta-analyses. None of the meta-analyses found a significant benefit for cardiovascular events and mortality with the use of statin therapy in dialysis patients. However, all three meta-analyses reported a trend toward increased risk of stroke.^20,–22 Hou et al. concluded that there was still a significant benefit for reducing major cardiovascular events despite the stage of CKD. However, both dialysis and nondialysis patients with ESRD (stage 5 CKD) were included. The lack of statin benefit for dialysis patients reported from the meta-analyses had no heterogeneity among the evaluated trials for all-cause mortality and cardiovascular death. The majority of the trials included in the meta-analyses evaluated the use of statin monotherapy, with the exception of SHARP. A limitation of all three meta-analyses was that over 10% of the dialysis patients analyzed were receiving peritoneal dialysis. There is evidence to suggest that statin therapy may be more efficacious for peritoneal dialysis patients due to higher levels of apolipoprotein B and more atherogenic lipid abnormalities including elevated LDL levels and total cholesterol compared to hemodialysis patients.^16,–19 It would have strengthened the meta-analyses if separate analyses were reported for both hemodialysis and peritoneal dialysis patients.

Another possible limitation of all the trials is the fact that they were designed and conducted before evidence demonstrating an increased risk of cardiovascular events in patients receiving ESAs when targeting near-normal hemoglobin levels. At baseline, patients’ mean hemoglobin concentration was approximately 11 g/dL, and over 80% of patients in both the 4D trial and AURORA were being treated with ESAs.^9,14 FDA does not recommend a specific hemoglobin concentration target due to the lack of evidence; however, the agency suggests that hemoglobin concentrations should not exceed 11 g/dL with the use of ESAs.¹¹ The TREAT trial, published in 2009, demonstrated a twofold increased risk of stroke in CKD patients with type 2 diabetes receiving darbepoetin alfa with a target hemoglobin concentration of 13 g/dL.²³ It is important to consider that the use of ESAs in these trials could have confounded the results regarding major cardiovascular events, as we are unaware of significant differences between study groups for ESA dose or hemoglobin concentration for patients at the end of the study.

The mean age of hemodialysis patients in the 4D trial, AURORA, and SHARP was 62–66 years. The mean durations of hemodialysis were about eight months and 3.5 years in the 4D trial and AURORA, respectively. These factors are important to consider, given that the life expectancy for patients on hemodialysis varies depending on patient age. According to the USRDS 2007 annual report, hemodialysis patients age 60–64 years have a life expectancy of 4.5 years, compared with 3.8 years for hemodialysis patients age 65–69 years.²⁴ In AURORA, in which patients had been receiving dialysis for a mean of 3.5 years, the overall life expectancy from the start of the trial would be expected to be approximately 1 year based on the mean age of participants. This should urge clinicians to consider the results of these trials with caution when dealing with younger patients who may benefit from statin therapy due to longer exposure if there is a longer life expectancy. This point should also be considered in elderly patients, as it has been demonstrated that patients older than 75 years undergoing hemodialysis have an average life expectancy of <2.7 years, and any benefit of statin therapy may not be seen due to the short length of exposure.²⁴

Low-to-moderate intensity statin therapy appears to be relatively safe to administer to hemodialysis patients without an increased risk of myopathy, rhabdomyolysis, or hepatic impairment compared with placebo; however, when simvastatin was given with ezetimibe, there was a higher discontinuation rate due to myopathy.⁹, ¹⁴, ¹⁵ Although there were trends toward higher rates of stroke observed in the meta-analyses, due to the low event rates, these findings could have been due to chance.^20,–22

It is unclear if statin therapy has a specific role in primary or secondary prevention, as the 4D trial and AURORA both included patients with and without a history of cardiovascular disease, whereas SHARP excluded patients with a definite history of MI or revascularization procedures. It is possible that the efficacy of statins may be different based on the baseline LDL cholesterol level before treatment. Although the mean baseline LDL cholesterol concentration for the 4D trial was 125 mg/dL in the atorvastatin group, the post hoc analysis revealed that patients in the atorvastatin group who had the highest LDL cholesterol concentration (>145 mg/dL) experienced significant reductions in cardiac death.¹⁰ A similar benefit was found in SHARP in patients with a baseline LDL cholesterol concentration of >117 mg/dL; however, this relationship was not found in AURORA.^14,15 Further studies are needed to address whether statins should be discontinued once patients require hemodialysis or started in patients undergoing dialysis who are younger or who have atherosclerotic cardiovascular disease.

The majority of the evidence reviewed from randomized controlled trials and recent meta-analyses suggest that there is minimal to no benefit of statin therapy for cardiovascular events and mortality for statin-naïve patients on hemodialysis. The KDOQI 2003 dyslipidemia guidelines recommended that patients with ESRD (stage 5 CKD) receive a statin to reach a goal LDL cholesterol concentration of <100 mg/dL; however, there was no distinction between nondialysis and dialysis patients, and newer evidence has since been published. Although KDOQI released 2012 guidelines that recommended against the initiation of statins in dialysis patients due to the lack of evidence to support benefit, the guidelines were specific for diabetic dialysis patients. The recent KDIGO dyslipidemia guidelines concur with the 2012 KDOQI guidelines and recommend against the initiation of statins in patients undergoing hemodialysis. However, they suggest to continue statins in patients already taking them despite the initiation of hemodialysis until further evidence clarifies the issue.⁵ They do not make the distinction between hemodialysis and peritoneal dialysis patients. Clinicians should use their clinical judgment and weigh the risks and benefits from the available evidence when deciding whether to initiate statins in hemodialysis patients. A statin may be warranted when the estimated 10-year incidence of coronary death or nonfatal myocardial infarction is greater than 10% or in younger hemodialysis patients who have a longer life expectancy. Until there are conclusive data to support discontinuing statins in CKD patients progressing to ESRD and requiring hemodialysis, it is recommended to continue statins for these patients.

Conclusion

The available literature does not support the initiation of statins in hemodialysis patients who were not receiving statin therapy before requiring hemodialysis. At this time, there are no conclusive data to support discontinuation of statins in ESRD patients on hemodialysis receiving statins for either primary or secondary prevention of CHD.

Footnotes

References