| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
In-Depth Reviews |
Department of Medicine, University of Maryland School of Medicine, Baltimore, Maryland
Correspondence: Dr. Joseph M. Nogueira, N3W143, Division of Nephrology, University of Maryland School of Medicine, 22 S. Greene Street, Baltimore, MD 21201. Phone: 410-328-5720; Fax: 410-328-5685; E-mail: jnogueir{at}medicine.umaryland.edu
 |
Abstract |
|---|
|
Introduction |
|---|
3-Hydroxyl-3-methylglutaryl CoA reductase inhibitors (statins) and other lipid-lowering drugs (LLD) have been clearly demonstrated to lessen the morbidity and mortality of CVD in the general population; therefore, there is hope that the use of these drugs may likewise help patients with CKD (In addition, it has been theorized that dyslipidemia may contribute to the progression of CKD, such that LLD may prove to slow CKD progression.) Nevertheless, given differences in epidemiology and pathophysiology of CVD in CKD, there are reasons to question whether the cardiovascular benefits of LLD that are observed in the general population will also be realized in the CKD population. It is widely acknowledged that the current medical literature provides insufficient data to guide the use of LLD in the CKD population, in large part because renal insufficiency has been an exclusion criterion for most of the randomized, controlled studies that have addressed this issue. Although consensus-based guidelines on the treatment of dyslipidemias in individuals with CKD and with kidney transplants have been published by the National Kidney Foundation (NKF) Kidney Disease Outcomes Quality Initiative (K/DOQI) Work Groups, the nephrology and cardiology communities recognize the urgent need for high-quality clinical studies on this issue, and this has fueled the publication of subgroup analyses of previous large studies in the general population and the launching of large-scale clinical trials in the CKD population, as summarized in Table 1. To provide the reader with a better appreciation of the important issues that are germane to management of dyslipidemia in CKD, this article (1) briefly summarizes the literature on the benefits of LLD in the general population, (2) demonstrates the limited involvement of individuals with CKD in the studies in the general population, (3) discusses the epidemiologic and pathophysiologic reasons that the data that are derived from and the guidelines that are directed at the non-CKD population may not be applicable to the CKD population, (4) reviews the emerging literature on the management of dyslipidemia in patients with CKD (with emphasis on studies that have been published since the K/DOQI guidelines were put forth), (5) discusses issues of safety of LLD in the CKD population, and (6) provides updated recommendations for management pending the completion of important ongoing trials.
|
|
Risk of Hyperlipidemia and the Benefits of LLD in the General Population |
|---|
On the basis of the vast body of literature available in the general population, the National Cholesterol Education Program (NCEP) has periodically provided rigorously derived, consensus-based, and widely accepted recommendations on the treatment of hyperlipidemia, and the most recent full report was published in 2001 as the NCEP Adult Treatment Panel III (NCEP ATP-III) guidelines (40). This report outlined a method of assessing an individual's risk for coronary heart disease, and then it proposed specific lifestyle and pharmacologic treatment guidelines based primarily on the CHD risk profile and the fasting LDL cholesterol level. Since the writing of the NCEP AT-III guidelines, several important studies, such as the HPS, have been published. To incorporate the findings of the subsequent trials, the NCEP issued a report in 2004 addressing the implications of these trials for the NCEP AT-III guidelines (41). Although the recommended thresholds for drug treatment and the LDL goals for the various groups did not change, they decided to footnote the algorithm to indicate that an LDL cholesterol goal of <70 mg/dl in very high-risk patients and <100 mg/dl in moderately high-risk individuals (two or more risk factors and a 10-yr risk of 10 to 20%) are reasonable therapeutic options on the basis of available clinical trial evidence. They also suggested that when LDL-lowering drug therapy is used in high- or moderately high-risk individuals, the intensity of therapy should be sufficient to achieve at least a 30 to 40% reduction in LDL cholesterol levels.
|
Limited Data from Trials in the General Population on Effects of Lipid Lowering in Patients with Renal Insufficiency |
|---|
1.3 mg/dl for the Helsinki Heart Study (10), serum creatinine >1.75 mg/dl for West of Scotland Coronary Prevention Study (WOSCOPS) (12,42), serum creatinine 2 mg/dl for Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial (ALLHAT) (43), "renal disease" for the Long-Term Intervention with Pravastatin in Ischaemic Disease (LIPID) study (25), and creatinine 1.5 mg/dl or nephrotic syndrome for Bezafibrate Infarction Prevention (BIP) study (22). The HPS (38) excluded individuals with "severe renal disease or evidence of substantially impaired renal function," such that only 375 individuals with serum creatinines from 110 to 200 µmol/L (1.25 to 2.28 mg/dl) for women and from 130 to 200 µmol/L (1.48 to 2.28 mg/dl) for men were included. Although the study found a modest benefit in the subgroup of individuals without diabetes and with renal insufficiency (see Table 1), the subgroup of individuals with diabetes and with renal insufficiency in the HPS study had a 95% confidence interval (CI) of the primary event rate ratio that crossed the line of unity, indicating a lack of statistically significant benefit. Also detailed in Table 1, subgroup analyses of the likewise massive Anglo-Scandinavian Cardiac Outcomes Trial–Lipid Lowering Arm (ASCOT-LLA) showed a lower rate of the primary end point of nonfatal myocardial infarction (MI) and fatal CHD in those patients with relatively mild renal dysfunction (11). The Pravastatin Pooling Project (44) is the largest available data set involving statin treatment in individuals with CKD. It used a database that combined patient-level results from three randomized trials of pravastatin versus placebo in the general population: WOSCOPS (12), Cholesterol and Recurrent Events (CARE) study (26), and LIPID study (25). In the database, 4491 (22.8%) of 19,700 patients had "moderate" CKD, as defined by an estimated GFR of 30 to 60 ml/min per 1.73 m2 body surface area. The primary outcome was time to MI, coronary death, or percutaneous/surgical coronary revascularization. Moderate CKD was independently associated with an increased risk for the primary outcome, with an adjusted hazard ratio (HR) of 1.26 (95% CI 1.07 to 1.49) compared with those with normal renal function. Pravastatin significantly reduced the incidence of the primary outcome in the patients with moderate CKD (HR 0.77; 95% CI 0.68 to 0.86), and this was similar to the effect of pravastatin on the primary outcome in patients with normal kidney function. The total mortality rate in those with moderate CKD also was lower in those who were treated with pravastatin (adjusted HR 0.86; 95% CI 0.74 to 1.00; P = 0.045). The authors concluded that pravastatin reduces cardiovascular event rates in people who have moderate CKD and are at high risk for CHD or have known CAD. It should be noted that CARE and LIPID patients all had CHD, such that the majority of patients in the Pravastatin Pooling Project—3310 (73.7%) of 4491—had known coronary disease. A benefit was demonstrated in primary and secondary outcomes in patients with known coronary disease at baseline, and statistical tests for interaction revealed no significant differential effect in patients with and without symptomatic coronary disease at baseline. Although the authors did not demonstrate a statistically lower relative risk (RR) for study outcomes in the subgroup of patients with this degree of CKD and without coronary disease (HR for primary outcome 0.82; 95% CI 0.52 to 1.31; NS), they did conclude that the statistical homogeneity among the groups with and without coronary disease at baseline (P > 0.5 for interactions) suggests that pravastatin benefits these patients irrespective of the presence or absence of known coronary disease at baseline. Although the Pravastatin Pooling Project strongly suggests that patients with GFR between 30 and 60 ml/min per 1.73 m2 and known CHD will benefit from statins, the lack of statistically better outcomes in those without known coronary disease renders less forceful its conclusions of benefit in primary prevention. Also, it provides no guidance for treatment of individuals with more advanced renal failure.
|
Unique Epidemiology and Pathophysiology of Cardiovascular Disease in CKD |
|---|
Without doubt, the cardiovascular mortality of hemodialysis patients is extremely high relative to the general population. Foley et al. (4) demonstrated in the late 1990s that this risk is several-fold higher for older individuals, and they showed an extremely chilling risk of approximately 500-fold greater for patients in their 20s. At first glance, these data would seem to make it obvious that such patients should be treated with statins. However, it should be noted that the definition of CVD mortality for this study incorporated deaths that would be caused by disorders in which dyslipidemia may not be a major pathogenic factor and on which the effect of LLD would be dubious, such as those caused by cardiomyopathy, cardiac arrest, pulmonary edema, and arrhythmias (1). For example, left ventricular hypertrophy is extremely common in patients with CKD (71–73) as well as ESRD (74–78), is a very strong predictor of mortality in this setting (76,78–80), and is probably largely unrelated to dyslipidemia.
Further complicating our understanding of CVD in CKD is that the epidemiologic associations of cholesterol level with mortality may be different in this setting. In contrast to the graded positive association in the general population, the relationship between total cholesterol levels and all-cause mortality among approximately 12,000 hemodialysis patients (81) demonstrated a somewhat U-shaped association, as shown in Figure 1. It shows that patients with very low cholesterol levels have a marked increase in morality and that the risk is fairly flat across the range of normal and moderately high cholesterol levels. It is not until very high levels of cholesterol that the mortality seems to increase. Of course, low cholesterol levels may be a marker of malnutrition, inflammation, and other chronic illnesses in this population rather than an independent risk factor, and clearly this may explain the high mortality at very low cholesterol levels. However, the lack of a graded increase in CVD risk in the normal and high-normal range (as is seen in the general population) argues for a less robust and less direct relationship between increasing cholesterol and mortality in the dialysis population. It should also be noted that three subsequent reports have documented a positive association with CVD outcomes. First, Tschope et al. (82) prospectively followed all consecutive patients who had diabetes and were newly admitted to 28 German dialysis centers between January 1985 and October 1987, and they found that patients who had diabetes and subsequently died from MI had significantly higher median total and LDL cholesterol levels than survivors. Second, Nishizawa et al. (83) showed that higher non-HDL cholesterol was a significant and independent predictor of cardiovascular mortality in hemodialysis in a cohort of 525 Japanese hemodialysis patients. Third, Block et al. (84) recently reported that total cholesterol was directly related to the risk for hospitalization for CVD in a large hemodialysis database. In sum, elevated cholesterol levels seem to be predictive of cardiovascular morbidity and mortality in the CKD population, but the relationship seems to be less direct than in the general population, especially in hemodialysis patients, in whom low cholesterol levels seem to identify a group at high risk for mortality.
|
75 ml/min and with a mean creatinine clearance of 30.3 ± 18 ml/min) that only 35.6% of those with CVD and known hyperlipidemia were administered statins. Levin et al. (52) also reported that among a cohort of 313 patients with mild to moderate CKD (mean creatinine clearance 36 ml/min) and with a 46% incidence of CVD, only 13.5% were treated with an LLD; furthermore, only 25 (17.7%) of 143 in this cohort with known CVD were receiving such drugs. It is not clear whether these low rates of treatment with LLD are the result of therapeutic nihilism, concern for risk of adverse effects, problems in health care delivery, or others, but this does indicate that during the late 1990s, nephrologists and patients "voted with their feet" against widespread use of these agents in this high-risk population.
Pathophysiology of CVD in CKD
The character of CVD in CKD is distinct in large part because of differences in pathophysiology of cardiac and vascular disease in this setting. The list of potential pathophysiologic contributors to the CVD epidemic in CKD, especially in dialysis patients, is legion. Table 2 lists many of these, and a few are discussed in some detail. Clearly, these differences in pathophysiology may alter the effects of dyslipidemia and the risks/benefits of LLD in the CKD populations.
|
|
Intermediate-density lipoprotein (IDL) cholesterol is another of the remnant lipoproteins that may be included in the non-HDL cholesterol fraction, and it has been reported to be an important atherogenic lipid component in the general population (95,96). It has been correlated with increased aortic pulse wave velocity in hemodialysis patients (97). Serum IDL cholesterol levels have been reported to be higher in patients with renal failure (98,99). Statins seem to lower IDL cholesterol in the general population (100,101) and in dialysis patients (102,103). Whether reductions in IDL cholesterol are associated with reductions in mortality is not known, but some have suggested that IDL cholesterol may be an important lipid component to target in individuals with renal failure (103).
Also, lipoprotein(a), a modified form of LDL, is a highly atherogenic lipoprotein particle that may be a relatively important player in CVD pathogenesis in the CKD population. It structurally resembles plasminogen and interferes with fibrinolysis, and it binds to macrophages and promotes foam cell formation and the deposition of cholesterol in atherosclerotic plaques (104,105). An elevated level predicted vascular events (106) and was found to be an independent risk factor for CVD in hemodialysis patients (107). The levels of lipoprotein(a) are especially high in peritoneal dialysis patients (89) and seem to decrease after kidney transplantation (90).
The increased levels of oxidative stress and inflammation that are seen in CKD may cause oxidative conversion from the native LDL to the more atherogenic oxidized LDL molecule (108). Statins seem to have antioxidant properties and anti-inflammatory effects (109), and they may ameliorate atherosclerotic processes by preventing oxidation of LDL cholesterol (110).
Multifaceted Vascular Disease. To understand vascular pathology, one must appreciate the contributions of both of the two important types or arterial disease: atherosclerosis (thickening of the intimal layer of arteries with lipid-rich deposits and plaques) and arteriosclerosis (diffuse thickening and loss of elasticity of arterial walls). Both seem to be problematic in patients with CKD, and there may be important differences in the pathophysiology of these processes as compared with the general population.
Atherosclerosis seems to be more common in individuals with CKD (111) and ESRD (112–116). The atherosclerotic lesions in ESRD have a distinct morphology, with a more calcified appearance (as opposed to fibroatheromatous) and relatively increased medial thickness (as opposed to intimal) as compared with the general population (117). It has been hypothesized that calcification in the setting of CKD may contribute to the atherosclerotic process, but to what degree this calcification may have an impact on the stability of the plaques is not clear (118,119). As is the case in the general population, dyslipidemia probably plays a critical role in the atherosclerotic process in CKD (120).
Arteriosclerosis also seems to be an extremely important component of CVD that affects patients with CKD. Large artery distensibility is reduced in individuals with CKD (121) and ESRD (122), and the arterial stiffness seems to predict cardiovascular and all-cause mortality in patients with ESRD (123–126). Vascular calcification (VC), especially in the arterial media, is probably a major contributor to arteriosclerosis in CKD (127–130). Numerous autopsy and radiologic studies in the past several decades have demonstrated high rates of calcification of peripheral and central arteries (131–140), and coronary artery calcification is even common and progressive in young adult hemodialysis patients (141). VC is probably primarily due to abnormal mineral metabolism that is seen in renal failure (141–145), perhaps through a complex process involving osseous metaplasia of the vascular wall components (118,119,146). Associations of calcium/phosphorus/parathyroid hormone dysregulation with CVD and concerns that coronary artery calcification may beget coronary artery occlusive disease (147–150) argue for an important role of VC in the CVD epidemic in ESRD (84,151).
Precisely how VC interacts with lipids in the pathogenesis of vascular disease in CKD is unclear, but this issue is critical in understanding the effects of LLD in this population. A positive relationship between VC and serum total and LDL cholesterol and triglyceride levels was shown in the general population (149,152,153). A small Japanese study found that rapid progression of coronary artery calcification was associated with high triglyceride and low HDL cholesterol concentrations in long-term hemodialysis patients (154). Another study likewise found an association of low HDL cholesterol with VC (155). Two studies reported lower total cholesterol levels in dialysis patients with VC (141,156). Another study of 44 peritoneal dialysis patients did not note an association with lipid parameters (157).
Endothelial Dysfunction. Finally, endothelial dysfunction may also importantly contribute to the development of CVD in patients with renal insufficiency (158). Impaired endothelial vasodilation seems to be related to oxidative stress (159) and reduced levels of nitric oxide (160), likely as a result if accumulation in renal failure of an inhibitor of nitric oxide synthase, asymmetrical dimethylarginine (161–166). In addition, reduced levels of circulating endothelial progenitor cells and increased numbers of mature endothelial cells have been noted in renal failure and have been hypothesized to reflect ongoing endothelial injury and perhaps poor repair function (167–173).
In sum, we speculate that the pathophysiology of CVD in CKD involves many pathways that are interrelated and contribute in complex ways to the epidemic of CVD in this population. Although the discussion of this issue could continue for many more pages, the bottom line is that the relative contributions of these various insults in an individual patient with CKD may be much different than those seen in the general population and that these complex differences render very questionable the extrapolation of data on benefits of lipid lowering that are derived in nonuremic individuals to individuals with CKD. Also, the pleiotropic effects of statins on cell signaling, inflammation, cellular proliferation, atherosclerotic plaque stabilization, and endothelial dysfunction (174–179) may be different than those seen in the general population and may be important in modulating the effects of statins in the CKD population.
|
Studies of LLD in the CKD Population |
|---|
In this effort to assess mortality effects of statins in this population, Seliger et al. (86) used a large national database, the US Renal Data System Dialysis Morbidity and Mortality Study Wave 2, to assess the association between statin use and mortality in incident dialysis patients. They found that statin use was associated with a 32% lower risk for total mortality (95% CI 0.54 to 0.87) and that this effect was independent of other parameters that are known to have an impact on mortality, including diabetes, age, previous CVD, and smoking. They also found a similarly reduced risk for cardiovascular-specific mortality (RR 0.64; 95% CI 0.45 to 0.91). Although this study suffered from the limitations that are inherent in a retrospective database review, the mortality effects of statins were noted to be similar to those that were observed in large, randomized, controlled trials in the general population; were independent of other effects of other known risk factors; and were thus interpreted to suggest that statins may be effective in improving survival in dialysis patients.
Assessment of Lescol in Renal Transplant Study
The Assessment of Lescol in Renal Transplant (ALERT) study (186) was a multicenter, randomized, double-blind, placebo-controlled trial that involved 2102 renal transplant recipients with total cholesterol levels of 154 to 346 mg/dl and compared treatment of fluvastatin with placebo. Patients were not excluded when they had known coronary heart disease as long as they had not had an MI in the preceding 6 mo. Of course, the patients were on transplant immunosuppressive drugs. The primary end point was the occurrence of a major adverse cardiac event, defined as cardiac death, nonfatal MI, or coronary intervention procedure. After a mean follow-up of 5.1 yr, fluvastatin lowered LDL cholesterol concentrations by 32%. Risk reduction with fluvastatin for the primary end point was not significant, with a risk ratio of 0.83 (95% CI 0.64 to 1.06; P = 0.139). There were fewer combined cardiac deaths or nonfatal MI in the fluvastatin group than in the placebo group (70 versus 104; risk ratio 0.65; 95% CI 0.48 to 0.88; P = 0.005). Coronary intervention procedures and other secondary end points did not differ significantly between groups. Also, no effect on renal allograft survival or renal function could be demonstrated in secondary analysis of the data (187). The authors of the study commented that because the observed cardiac event rate was lower than predicted at the outset, the trial had insufficient power to detect a significant reduction in the chosen primary end point. Relevant to the CKD population, the mean creatinine clearance in the ALERT trial was 60 ml/min, and approximately half of the patients (n = 1017) in the ALERT study had significant renal insufficiency at inclusion (serum creatinine at least 1.57 mg/dl). Although the study could not show a clear benefit of statins in this population, it is important to note that it did not identify any substantial safety concerns in this population either.
Die Deutsche Diabetes Dialyse Studie (4-D) Study
The 4-D study (188) was a multicenter, randomized, double-blind, prospective study of 1255 patients who had type 2 diabetes, were receiving maintenance hemodialysis, and were randomly assigned to receive 20 mg/d atorvastatin or matching placebo. The primary end point was a composite of death from cardiac causes, nonfatal MI, and stroke. Secondary end points included death from all causes and all cardiac and cerebrovascular events combined. During a median follow-up period of 4 yr, 469 (37%) patients reached the primary end point, 226 of whom were assigned to atorvastatin and 243 to placebo (RR 0.92; 95% CI 0.77 to 1.10; P = 0.37). Atorvastatin had no significant effect on the individual components of the primary end point, except that the RR of fatal stroke among those who received the drug was 2.03 (95% CI 1.05 to 3.93; P = 0.04). Atorvastatin reduced the rate of all combined cardiac events (RR 0.82; 95% CI 0.68 to 0.99; P = 0.03) but not all cerebrovascular events combined (RR 1.12; 95% CI 0.81 to 1.55; P = 0.49) or total mortality (RR 0.93; 95% CI 0.79 to 1.08; P = 0.33). Importantly, no cases of rhabdomyolysis or severe liver disease were detected, further alleviating safety concerns of statins in this population. The finding in this adequately powered study of lack of benefit in this population at extremely high cardiovascular risk demonstrates the danger in applying literature of the general population to those with ESRD, in whom so many other pathogenic pathways to CVD coexist with dyslipidemia. Nevertheless, given the compelling literature in the general population to support treatment of patients with diabetes with statins and given the marginally significant reduction of all cardiac events (a secondary outcome) in patients who were on statin in this trial, the negative results from this single trial are not sufficiently robust to establish nontreatment of such patients as the only acceptable standard of care. Furthermore, it is certainly very plausible that treatment of patients with diabetes with statins earlier in the course of CKD would be beneficial. Finally, it also should be noted that a 20-mg dose of atorvastatin was used without dosage escalation and that a previous study suggested that an 80-mg dose of the drug is more effective than a 10-mg dose in patients with stable CAD (24), such that it is possible that benefit could have been realized had a larger dosage of statin been used.
Prevention of Renal and Vascular End-Stage Disease Intervention Trial
Prevention of Renal and Vascular End-Stage Disease Intervention Trial (PREVEND IT) involved individuals at the other (early) end of the kidney disease spectrum and likewise was unable to demonstrate a benefit of statins (189). It was a single-center (in the Netherlands), double-blind, randomized, placebo-controlled trial with a 2 x 2 factorial design involving fosinopril or placebo and pravastatin 40 mg or placebo. Patients were 28 to 75 yr of age and were found to have microalbuminuria on screening, and they were excluded when the creatinine clearance was <60% of the normal age-adjusted value. A total of 864 patients were randomly assigned, the mean follow-up was 46 mo, and the primary end point was cardiovascular mortality and hospitalization for cardiovascular morbidity. Although fosinopril produced a significant reduction in albuminuria and was associated with a trend in decreasing cardiovascular events (HR 0.60; 95% CI 0.33 to 1.10; P = 0.098), 4 yr of treatment with pravastatin did not result in a significant reduction in urinary albumin excretion or cardiovascular events (HR 0.87; 95% CI 0.49 to 1.57; P = 0.649).
United Kingdom Heart and Renal Protection Studies
Two pilot studies have been completed in preparation for the Study of Heart and Renal Protection (SHARP), a large-scale, randomized trial of LLD in CKD to be discussed in the next section. The main aims of these studies were to establish the biochemical efficacy and to gain more information about the safety of cholesterol-lowering drugs in patients with CKD. In the first UK Heart and Renal Protection study (UK-HARP-I) (190), 448 patients with CKD were randomly assigned to simvastatin 20 mg/d or to placebo (along with aspirin 100 mg/d versus placebo in a 2 x 2 factorial design) for 1 yr. At the start, 242 patients were predialysis, 73 were receiving maintenance dialysis, and 133 had a functioning transplant. Simvastatin lowered LDL cholesterol by 26%, similar to the lipid-lowering effect in patients without CKD. The simvastatin-allocated group did not have a significantly increased risk for abnormal liver transaminases, elevated creatinine kinase, or serious myopathy. Then, because a more substantial lowering of cholesterol was expected to be needed to produce a demonstrable reduction in cardiac events, the second HARP pilot study (UK-HARP-II) was designed to assess the tolerability, safety, and biochemical efficacy of ezetimibe in combination with simvastatin (20 mg/d). Among the 203 patients who had CKD and were randomly assigned to ezetimibe/simvastatin versus simvastatin, proportional reductions in LDL cholesterol similar to those achieved in the general population were observed, and the combination was well tolerated (191).
Study of Heart and Renal Protection (SHARP)
SHARP is an investigator-initiated, industry-sponsored trial designed to compare ezetimibe/simvastatin versus placebo in approximately 9000 patients with CKD around the world, approximately 6000 of whom were intended to be predialysis and 3000 of whom were to be undergoing dialysis (192) Patients with previous MI or coronary revascularization and those with a functioning kidney transplant are excluded. With a plan to continue the treatment for at least 4 yr, the primary aim of the study is to assess the effects of LDL cholesterol lowering on the time to a first "major vascular event," defined as nonfatal MI or cardiac death, nonfatal or fatal stroke, or revascularization. Secondary aims will include assessments of the effects of the combination drug on progression to ESRD, various causes of death, major cardiac events, stroke, and hospitalization for angina. Of course, subgroup analyses will be conducted as well. The study commenced in 2003, and the recruitment phase was completed in the summer 2006. The landmark study is the largest nephrology trial ever undertaken, and it has the power to establish whether the drugs will protect patients with CKD from a first major vascular event. Its highly anticipated results should greatly clarify the indications for LLD in patients with CKD.
A Study to Evaluate the Use of Rosuvastatin in Subjects on Regular Hemodialysis: An Assessment of Survival and Cardiovascular Events (AURORA) Study
The AURORA study is a placebo-controlled, double-blind study of rosuvastatin versus placebo in approximately 3000 hemodialysis patients from age 50 to 80 yr, irrespective of previous CVD and irrespective of baseline lipid levels. The primary study end point is the time to a major cardiovascular event (first occurrence of cardiovascular death, nonfatal MI, or nonfatal stroke). Secondary end points include all-cause mortality, major cardiovascular event-free survival time, time to cardiovascular death, time to noncardiovascular death, cardiovascular interventions, tolerability of treatment, and health economic costs per life-year saved. Study medication will be given until 620 patients have experienced a major cardiovascular event. By early 2005, more than 2750 patients had been randomly assigned (193,194). When published, the results of this study should help to clarify the role of statins in hemodialysis patients.
|
Safety of LLD in the CKD Population |
|---|
Fibric acid derivatives may cause myositis and rhabdomyolysis, and it has been suggested that adjusting the dosage of these medications on the basis of renal function and exercising caution with potentially interactive combinations (e.g., with statins) may minimize risk for toxicity (195,196). Of note, three cases of reversible acute renal allograft dysfunction have been reported in patients who were treated with fenofibrate (197). Bile acid sequestrants have been reported to interfere with cyclosporine absorption, but this problem can be overcome if the drugs are administered at different times of the day. The effect of bile acid sequestrants on the absorption of other immunosuppressives is not known. It should also be remembered that bile acid sequestrants may increase triglyceride levels and therefore should be avoided in patients with marked hypertriglyceridemia, which is relatively common in dialysis patients (195,196). Nicotinic acid is partially (approximately one third) renally cleared, and it is recommended that the dosage be decreased to 50% in patients with GFR <15 ml/min (195). We are not aware of any data that would indicate special safety concerns regarding the use of nicotinic acid in patients who have CKD or are on immunosuppressive drugs.
The issue of statin-induced myositis and rhabdomyolysis has been a significant concern in this population, and the scores of fatalities as a result of rhabdomyolysis that were attributed to cerivastatin (Baycol, Bayer Corporation, Leverkusen, Germany), which ultimately caused its removal from the market in 2001, have weighed heavily in the risk–benefit debate of this class of drugs. Statin-induced rhabdomyolysis in patients with renal failure is the subject of a relatively recent review by Sica and Gehr (198), and the reader is referred to this article for a more thorough treatise on the subject. High levels of statins have been implicated as the major factor that predisposes to myotoxicity; therefore, an understanding of the clearance and metabolism of the drugs is important. Except for pravastatin, all statins are extensively metabolized by the cytochrome P450 (CYP) enzymes. Most of the statins, including atorvastatin, lovastatin, and simvastatin, are metabolized primarily by the CYP3A4 isoenzyme, and they are therefore susceptible to interactions with the myriad of drugs that have an impact on this enzyme. Among the important drug interactions are cyclosporin A, tacrolimus, azole antifungals, macrolide antibiotics, fibrates, nicotinic acid derivatives, protease inhibitors, and warfarin. Fluvastatin, however, is metabolized by the CYP2C9 pathway, and far fewer commonly prescribed drugs inhibit this pathway. Therefore, the potential for drug–drug interactions theoretically should be less with fluvastatin and pravastatin. The statins are primarily removed by hepatic metabolism, and the proportion of renal clearance is low except for pravastatin at 20 to 60% (199) and lovastatin at 30% renal excretion (198). All of the marketed statins are highly protein bound (>90%), except pravastatin, which is estimated to be 50% bound; therefore, removal with hemodialysis treatments should be minimal. Atorvastatin has a negligible renal excretion rate of <2%, pharmacokinetic studies have showed no effect of renal dysfunction (although only three patients with stage 5 CKD were studied) (200), and the package insert states that dosage adjustments are not necessary in patients with renal insufficiency. As discussed by Sica and Gehr (198), the limited pharmacokinetic studies have not indicated that dosage adjustments of other statins are necessary in renal failure, but the manufacturers have generally recommended lower starting dosages and cautious titration in patients with renal failure. Such recommendations seem prudent given evidence that the state of renal insufficiency has been associated with downregulation of the CYP enzyme system and hepatic metabolism of the drug (201–203), given the potential alterations in protein binding of these highly protein-bound drugs as a result of uremic toxins, and given the high rate of polypharmacy in this population. As discussed previously, the UK-HARP studies did not indicate higher rates of rhabdomyolysis in patients who had CKD and were treated with simvastatin or ezetimibe/simvastatin (190,192). Likewise, the ALERT study did not show higher rates of muscle toxicity in kidney transplant recipients who were on immunosuppressive drugs (187), and no cases of rhabdomyolysis were seen during treatment with atorvastatin in the 4-D study (188). In practice, we believe that it is advisable to measure baseline creatinine kinase (CK) levels before starting statin therapy, and then patients should be instructed to report muscle symptoms and be queried about muscle symptoms on follow-up. Although routine monitoring of CK levels is probably not indicated, repeat CK measurements should be obtained and compared with baseline levels if symptoms develop. The reader is referred to the American College of Cardiology/American Heart Association/National Heart, Lung, and Blood Institute Clinical Advisory on Statins for specific recommendations for managing abnormalities (204).
Although elevated levels of hepatic transaminases occur in 0.5 to 2.0% of patients who are treated with statins in the general population, it remains controversial whether statins cause significant hepatotoxicity (93,204). Several randomized, controlled trials have reported no significant difference in the incidence of persistently elevated aminotransferases between those who were treated with statins and those who were treated with placebo (9,21,38), and statins were noted not to worsen liver injury in patients with chronic transaminase elevations as a result of hepatitis B or C (93,204). Hepatotoxicity of statins could be dosage and/or level related; therefore, the issues of drug clearance and drug–drug interactions may also be applicable in this regard. To our knowledge, however, no study has demonstrated higher risk for statin-related hepatotoxicity in the CKD population; and the UK-HARP studies, the ALERT study, and the 4-D study did not indicate evidence of statin-related hepatotoxicity. Although many experts argue that routine monitoring of transaminases may be unnecessary, it is probably reasonable to check the levels initially, at 12 wk after starting therapy, and then periodically thereafter, as recommended by the US Food and Drug Administration and the American College of Cardiology/American Heart Association/National Heart, Lung, and Blood Institute Clinical Advisory on Statins (204).
Cyclosporine and possibly tacrolimus increase levels of most statins, likely by competing for hepatic CYP enzymes that metabolize the drugs. As discussed previously, pravastatin does not rely on the CYP system for metabolism, and lovastatin is metabolized by a different isozyme (CYP2C9 rather than CYP34A) than what is affected by the calcineurin inhibitors (CNI). Despite the theoretical lack of interactions, cyclosporine has been associated with a 3.5-fold increase in fluvastatin area under the curve (AUC) in heart transplant patients (205), and cyclosporine produced a five-fold increase in pravastatin AUC in kidney transplant patients (206). Because of these findings of increased levels of non–CYP34A-metabolized statins in patients who are on cyclosporine, different mechanisms of interactions are likely at play, and emerging evidence suggests that cyclosporine-mediated inhibition of transport proteins may be involved (207). Organic anion-transporting polypeptide 1B1 (OATP1B1) is a liver-specific uptake transporter of various substrates, which include pravastatin and cerivastatin (207–209). Studies have demonstrated that elevations of cerivastatin when used in combination with cyclosporine are largely due to cyclosporine-induced inhibition of hepatic uptake by this OATP1B1 transporter (209,210), and a similar inhibition of OATP1B1 by gemfibrozil may be an important mechanism of its interaction with this statin (211,212). Also, P-glycoprotein transporter is an energy-dependent multidrug efflux pump that serves to protect against toxic compound accumulation by transporting such substances, which include statins, into the intestinal lumen, bile, and urine. Cyclosporine but not tacrolimus has been shown to be a potent inhibitor of P-glycoprotein (213,214), and its transport activity seems to be reduced in rats with chronic renal failure (215). Clearly, this inhibition of hepatic uptake and enteric efflux of statins in cyclosporine-treated patients could increase risk for toxic accumulation.
Nevertheless, the available evidence suggests that statins can be used safely with CNI if the statin dosage is reduced and if treatment with other potentially interacting agents is avoided or monitored carefully (195). The reader is referred to the K/DOQI Clinical Practice Guidelines for Managing Dyslipidemias in Kidney Transplant Patients (195) for a more detailed discussion of the safety of LLD in the kidney transplant population. Although fluvastatin was safely used in kidney transplant patients in the ALERT study, no comparative trials of statins in this population have been done to our knowledge. Therefore, firm conclusions on the relative safety of the various statins in this setting cannot be made with confidence.
Limited evidence is accumulating that ezetimibe can be used safely in patients with CKD and in renal transplant recipients. As discussed previously, the UK-HARP-II study team reported that the ezetimibe/simvastatin combination was safe in patients with CKD (192). A noncontrolled study of 40 stable kidney transplant recipients, most of whom were on CNI, found the drug to be effective and safe as monotherapy or in combination with a statin (216). Three subsequent small studies likewise demonstrated effective lipid lowering without significant toxicity in renal transplant patients (217–219). Cyclosporine administration yielded a 3.4-fold greater AUC of ezetimibe in renal transplant patients (220), and there is a case report of supratherapeutic LDL cholesterol reduction in a cyclosporine-treated heart transplant patient who was on ezetimibe 10 mg/d (221). Therefore, a starting dosage of 5 mg/d, cautious dose titration, and close monitoring for adverse effects have been recommended (221) and are probably prudent when ezetimibe is co-administered with a CNI.
|
Treatment Guidelines |
|---|
The following are specific recommendations for treatment of dyslipidemias in adults with stage 5 CKD:
500 mg/dl (5.65 mmol/L) that cannot be corrected by removing an underlying cause, treatment with therapeutic lifestyle changes and a triglyceride-lowering agent should be considered (Grade C recommendation) 100 mg/dl (2.59 mmol/L), treatment should be considered to reduce LDL to <100 mg/dl (<2.59 mmol/L) (Grade B recommendation) 200 mg/dl (2.26 mmol/L), and non-HDL cholesterol (total cholesterol minus HDL) 130 mg/dl (3.36 mmol/L), treatment should be considered to reduce non-HDL cholesterol to <130 mg/dl (<3.36 mmol/L) (Grade C recommendation) Detailed rationales for each recommendation are provided in the text of the document. As discussed, specific recommendations targeting non-HDL cholesterol were provided in large part to address the probable contribution of "remnant lipoproteins" (VLDL and IDL) in patients with hypertriglyceridemia and ESRD. The article suggested that a statin be used first in the absence of contraindications, and it encouraged starting with a low dosage and carefully titrating while monitoring for toxicity. It suggested that second-line agents could include bile acid sequestrants (if triglyceride level <400 mg/dl), nicotinic acid, or possibly sevelamer hydrochloride.
Shortly after the publication of the ALERT study, the NKF K/DOQI convened a work group to draft guidelines for managing dyslipidemia in kidney transplant patients, and the group published their clinical practice guidelines in 2004 (195). The work group concluded that the incidence of CVD is very high in kidney transplant patients and that the available evidence suggests that the 10-yr cumulative risk for CHD is at least 20%. They stated that kidney transplant patients should be considered to have a "risk equivalent" for CHD in the NCEP ATP-III schema for management. Although they admitted that additional studies are needed in patients with CKD and kidney transplant patients, the work group adopted an approach that closely paralleled that of the NCEP ATP-III; but, as with the NKF K/DOQI guidelines for patients with CKD, it also addressed the issue of severe hypertriglyceridemia and remnant lipoproteins as measured with non-HDL cholesterol. The group offered exactly the same guidelines for transplant patients as those listed for patients with stage 5 CKD. The group stated that the existing data support the conclusion that statins can be safely used in combination with CNI (cyclosporine and tacrolimus) if the dosage is reduced and other agents that are metabolized by the CYP system (e.g., macrolide antibiotics, azole antifungal agents, calcium channel blockers, fibrates, nicotinic acid) are avoided. The group recommended avoiding ezetimibe in kidney transplant patients until safety data were available. As discussed, subsequent data have indicated that cyclosporine may increase AUC of ezetimibe but that cautious use of ezetimibe in renal transplant is probably effective and safe.
In general, we believe that the current literature supports the recommendations of K/DOQI in regard to individuals with CKD as well as renal transplant recipients. We offer the following caveats on the basis of our review of the current literature:
|
Conclusion |
|---|
|
Disclosures |
|---|
|
Footnotes |
|---|
|
References |
|---|
Top
Abstract
Introduction