Importance of hyperlipidaemia and therapy in renal patients

Abnormalities in lipid metabolism can be detected in patients as early as renal function begins to decline (glomerular filtration rate (GFR) <50 ml/min). However, the type and severity of lipid abnormalities vary among different chronic renal insufficiency (CRI) patient populations [1]. The most characteristic feature of the CRD-associated dyslipidaemia is the accumulation of apolipoprotein-B-containing triglyceride-rich lipoprotein particles in the very low density (VLDL) and intermediate density range (IDL). This results in increased serum triglycerides in conjunction with low levels of high density lipoprotein (HDL) cholesterol. In the presence of type 2 diabetes mellitus, an accumulation of small dense LDL particles contributes to the so-called atherogenic dyslipidaemia [1]. A similar type of dyslipidaemia is seen in the general population. Atherogenic dyslipidaemia frequently occurs in patients with premature coronary heart disease and appears, in the absence of elevated LDL cholesterol, to be an independent atherogenic lipoprotein phenotype. Most patients with atherogenic dyslipidaemia are insulin resistant and many have also an elevated serum apolipoprotein B. Growing evidence suggests that all of the components of this type of dyslipidaemia (elevated VLDL remnants/IDL, small dense LDL, low HDL-cholesterol) are independently atherogenic (Figure 1). Together they represent a set of lipoprotein abnormalities that, besides elevated LDL-cholesterol, promote atherosclerosis. Lipoprotein subclasses and their associated risk of coronary artery disease. Plus and minus signs indicate the current medical understanding of the relations between levels of the different lipoprotein subclasses and risk of heart disease. Plus signs signify a positive association with disease (higher levels=higher risk). Larger plus signs signify especially high-risk subclasses. Minus signs signify a negative association with disease (higher levels=lower risk). Modified from reference 34. VLDL, very-low-density lipoprotein; IDL, intermediate-density lipoprotein; LDL, low-density lipoprotein; HDL, high-density lipoprotein. Abnormalities in the composition of lipoproteins are particularly common in patients with CRD. They may also play a role in the pathogenesis of cardiovascular disease (CVD). Oxidation and carbamoylation of lipoproteins or apolipoprotein B modification of lipoproteins by advanced glycation end-products, for example, may increase their atherogenicity. As yet, there have been no clinical trials to determine whether treating specific changes in lipoprotein composition or apolipoprotein modification modifies the incidence of CVD. However, it may be reasonable to consider that these abnormalities represent yet another risk factor for CVD in patients with CRD, and that patients with potentially atherogenic changes in lipoprotein composition should be viewed as candidates for aggressive management of CVD risk factors. It is likely that intensive reduction of lipoproteins in patients with compositional abnormalities will help reduce the risk of CVD. In CRD, both the type and severity of lipid abnormalities vary in relationship to the amount of urinary protein excretion. Additional secondary disorders such as latent diabetes mellitus, alcohol abuse, liver or thyroid disease and antihypertensive or immunosuppressive medication are often superimposed. Thus, dyslipidaemia in patients with kidney disease is often multifactorial in origin. The special report from the National Kidney Foundation Task Force on Cardiovascular Disease [2,3] in hyperlipidaemic patients with CRD identifies three renal target populations: patients with CRI and those undergoing haemodialysis (HD) or peritoneal dialysis (PD). Patients with nephrotic syndrome may not always have a decreased GFR. An alternative approach may, therefore, be to divide the CRI target population into two categories of patients; those with and without nephrotic syndrome, the latter defined as those with <3 g of total urinary protein excretion in 24 h. However, target lipid levels and management of lipid abnormalities are the same for proteinuric patients. In end-stage renal disease (ESRD) patients on HD, the classes and doses of lipid-lowering agents used to manage lipid abnormalities differ from those used in patients on PD. The prevalence, rationale for treatment and specific therapies were reviewed in respect to recommendations given by the second Joint Task Force of European and other Societies on coronary prevention [4,5] as well as the National Cholesterol Education Program (NCEP) guidelines [6]. Almost all patients with CRD have an absolute risk of coronary artery disease (CAD) of 20% over 10 years. They belong to the highest risk group and show an even higher prevalence of lipid abnormalities, than the general population, which may occur in the presence of a normal LDL-cholesterol. Terms such as ‘primary prevention’ are no longer appropriate in these patients, since the cumulative incidence of coronary risk factors usually exceeds those which can be assessed by coronary risk charts. Patients with microalbuminuria due to either type 1 or type 2 diabetes mellitus are considered as high-risk patients with already established vascular disease [7,8]. Microalbuminuric patients should in general be subject to secondary prevention measures. Using the high-cholesterol strategy, LDL-cholesterol is the major serum lipid component to focus on, although it does not reflect the entire risk of the dyslipidaemia profile in CRD patients. The target LDL-cholesterol level in high-risk patients or in patients with coronary heart disease according to the European Task Force and the NCEP guidelines is 115 mg/dl (3 mmol/l) and 100 mg/dl (2.6 mmol/l), respectively. In the general population, the primary strategy of every anti-hyperlipidaemic therapy is a dietary approach. However, proteinuric or ESRD patients on HD or PD need different dietary recommendations because these approaches are frequently insufficient to correct dyslipidaemia. In such patients, a pharmacological approach to lipid lowering is often necessary. The most effective pharmacological agents for reducing LDL-cholesterol are statins. However, other agents may be safe and equally effective in selected patients. No intervention trials have investigated whether treatment reduces CVD. In the absence of significant data, the European Task Force recommendations or the NCEP guidelines can be used to guide treatment of hyperlipidaemia in patients with renal disease [4–6]. The recommendations are specifically targeted to treat lipids for the prevention of CAD in the general population, but there is evidence that statins will also reduce the risk of stroke in coronary patients. In their guidelines published in 1994 [4], the Task Force for the first time proposed priorities for preventive cardiology in patients with other forms of dyslipidaemia. Therefore, it is probably reasonable to extrapolate the data obtained in subjects with normal kidney function (general population) to other groups of patients with impaired kidney function. Application of the Task Force recommendations may also help reduce the incidence of other types of CVD. Almost all (70–100%) patients with urinary protein excretion >3 g/24 h have elevated serum cholesterol levels (>240 mg/dl) or an abnormal profile of plasma lipids. The degree of hypercholesterolaemia (total cholesterol) correlates directly with proteinuria or urinary albumin excretion and inversely with the level of serum albumin [9,10]. Increased plasma levels of lipoprotein (a) (>30 mg/dl) have been found in >60% of these patients. A number of studies have compared lipoproteins in nephrotic patients with and without diabetes, or in patients with and without renal insufficiency. In non-uraemic patients suffering from a nephrotic syndrome, lipoproteins were similar in those with (n=21) and without (n=10) diabetes [11]. In contrast, a group of 11 nondiabetic, uraemic patients with nephrotic syndrome had a lipid profile that was more characteristic of uraemia than nephrotic syndrome per se [11]. In general, the prevalence of hyperlipidaemia appears to increase with worsening of the renal function. However, the level of renal dysfunction at which elevated lipid levels become more prevalent than in normal individuals is not well defined. Attman and Alaupovic [12] reported elevated triglyceride and VLDL-cholesterol levels in a group of patients with mild to moderate CRI (creatinine clearance >15 ml/min/1.73 m2 body surface area). Total cholesterol and LDL-cholesterol levels were the same as those in controls. Few investigators have examined the independent effects of diabetes and CRI on plasma lipids. Attman et al. [13] found that the lipid abnormalities in diabetic patients with CRI resembled those in nondiabetic patients with CRI. However, diabetic patients with CRI tended to have higher triglyceride and lower HDL-cholesterol levels than nondiabetic patients with renal insufficiency. Diabetes thus appeared to exacerbate the lipid abnormalities of CRI [14,15]. There are few studies examining the relationship between lipid abnormalities and CVD in CRI, and there are no prospective intervention trials investigating whether treatment reduces CVD. Epidemiological evidence suggests that, in patients with nephrotic syndrome, CVD is more prevalent than expected, and it is likely that the atherogenic lipoprotein profile which is associated with CAD in the general population is harmful to patients with nephrotic syndrome. Ordonez et al. [16] compared the prevalence of CAD in 142 patients with nephrotic syndrome with that in an age- and gender-matched control group. The relative risk of coronary death (adjusted for hypertension and smoking) was 2.8 (95% confidence interval (CI) 0.7–11.3). The relative risk for myocardial infarction was 5.5 (95% CI 1.6–18.3). The increased prevalence of CAD was similar when nephrotic patients with minimal change disease or systemic lupus erythematosus were excluded [16]. Jungers et al. [17] prospectively examined the relationship between serum lipids and the incidence of myocardial infarction in 147 patients with a creatinine clearance of 20–50 ml/min/1.73 m2. The incidence of myocardial infarction was approximately 2.5 times higher than in the general population. Compared with the 97 patients who did not develop myocardial infarction during follow up, the 41 patients with myocardial infarction had lower HDL-cholesterol levels and higher levels of triglycerides. LDL-cholesterol, apolipoprotein B, and lipoprotein (a) were not associated with CAD. In a multivariate analysis, independent correlates of myocardial infarction included: cigarette smoking, systolic blood pressure, fibrinogen and low HDL-cholesterol. Both histopathological and epidemiological data suggest that hyperlipidaemia may contribute to progression of renal disease. Lipoprotein deposits and foam cells have been reported in the glomeruli of patients with renal disease. Samuelsson et al. [18] conducted a prospective cohort study of 73 non-diabetic patients with CRI (baseline GFR 41±15 ml/min/1.73 m2 body surface area). Plasma total cholesterol, LDL-cholesterol and apolipoprotein B levels at baseline all correlated with the subsequent rate of decline of renal function. These associations were independent of blood pressure and urinary protein excretion. In a subset of patients with glomerulonephritis, triglycerides and HDL-cholesterol also correlated with the rate of decline of GFR. Similar results have been reported in cohort studies of patients with diabetic nephropathy. Several small pilot intervention trials have reported that lipid-lowering strategies may reduce the rate of renal disease progression. However, others have failed to confirm this. Threshold and target lipid levels from the European Task Force or the NCEP can be used to guide therapy for hyperlipidaemia in patients with CRI, proteinuria and nephrotic syndrome. Stringent diet modification using reinforcement protocols of a soy-based vegetarian diet and fish oil improve hyperlipidaemia and proteinuria [19]. A study of a group of Italian patients with nephrotic syndrome suggested that not only fat restriction itself, but also the quality of fats and proteins in the manipulated diets, are important for correction of hypercholesterolaemia and, possibly, also for urinary protein excretion [19]. However, additional administration of statins is almost always necessary for reducing total cholesterol and LDL-cholesterol. Long-term therapy for 2 years with 40 mg of simvastatin has proved to be safe and effective in 20 patients with nephrotic syndrome [20]. Fibric acid derivatives are effective in reducing triglyceride levels in patients with CRI. The appropriate dose depends on the level of renal function. An alternative lipid-lowering strategy is the reduction of proteinuria. Reduction in proteinuria induced by angiotensin-converting enzyme (ACE) inhibitors and salt depletion reduces lipids and lipoproteins and should be used as an adjuvant to standard lipid-lowering therapies [21,22]. Dependent on the methodology used for the evaluation, dyslipidaemia is present in >90% of HD patients. Triglyceride levels are elevated while the HDL-cholesterol concentration is low. LDL-cholesterol is usually normal or reduced. Total cholesterol levels may thus be high, normal or low. There are few observational studies examining the relationship between dyslipidaemia and CVD outcome in HD patients and only one prospective evaluation shows a positive correlation [23]. Data on prospective intervention trials investigating the effects of statin treatment on CVD outcome will not be available before 2002. Whether hypertriglyceridaemia is an independent risk factor for CVD in the general population is still a matter of debate. The evidence is stronger that a low HDL level is a risk factor for CVD. Serum levels of IDL-cholesterol are correlated with aortic atherosclerosis determined by aortic pulse wave velocity. Plasma levels of lipoprotein (a) and its low molecular weight apolipoprotein (a) isoform are independent predictors of CVD outcome [24–27]. As for other target populations, it is reasonable to use the European Task Force recommendations to guide therapy in HD patients. These guidelines are entirely based on the reduction in LDL-cholesterol, which frequently is not elevated in HD patients. The guidelines do not recommend treatment of elevated triglycerides or low HDL-cholesterol which are markers of increased CHD risk. Diet may not be an ideal approach to correct dyslipidaemia, although diet reduces total cholesterol and tends to reduce triglyceride levels. HD patients are already subject to specific food and fluid intake restrictions. The introduction of a diet restricted in saturated fatty acids might add nutritional difficulties and the patients may be at a greater risk of developing malnutrition. Diets rich in polyunsaturated fatty acids of fish-oil origin (omega-3) increase the removal of triglyceride-rich lipoprotein remnants and impressively reduce prostprandial lipoprotein levels. Eicosapentanoic acid reduces plasma levels of remnant lipoproteins and prevents in vivo peroxidation of LDL in dialysis patients [28]. Although elevated LDL-cholesterol levels are not common in many HD patients, statins are effective in reducing elevated LDL concentrations [29] and probably also IDL and small dense LDL particles. Fibric acid analogues are effective in lowering serum triglycerides and raising HDL cholesterol. The most common adverse effects of drug therapy are myositis and rhabdomyolysis. These adverse effects can, however, be minimized by adjusting the dose of the drugs according to the level of renal function. Gemfibrozil is the fibric acid analogue of choice in HD patients. It is well tolerated and the parent drug does not accumulate and is not toxic. Gemfibrozil has recently been shown to reduce cardiovascular mortality in the VA-HIT study, a large-scale secondary prevention study, by lowering triglycerides and raising HDL-cholesterol without affecting LDL-cholesterol levels [30]. Treatment of renal anaemia with erythropoietin in HD patients seems to have beneficial effects on plasma lipid concentrations [31]. Adjuvant therapy using low molecular weight heparin for anticoagulation during HD or the use of polysulfone high-flux dialysis attenuates hypertriglyceridaemia in some but not all patients. A multicentre randomized and controlled trial should be conducted to determine whether treatment of hypertriglyceridaemia and/or low HDL reduces the incidence of CVD. The population of HD patients may be ideally suited to study this question, given the high frequency of both lipoprotein abnormalities and CVD events. In addition, the logistical infrastructure needed to conduct such a study is already in place in many dialysis units. Hyperlipidaemia is found in approximately 50% of PD patients, but lipoprotein composition may be abnormal in almost all patients. Twenty-four per cent of a combined total of 299 patients had total cholesterol levels of > 240 mg/dl, and LDL-cholesterol levels were > 130 mg/dl in 45% of 83 patients. Fifty-one of 268 patients (19%) had HDL-cholesterol levels of < 35 mg/dl, whereas triglyceride levels were > 200 mg/dl in 59% of patients. Eighty-six of 232 patients (37%) had lipoprotein (a) levels of >30 mg/dl [2, 32]. Very few epidemiological investigations have examined the relationship between lipid abnormalities and CVD in PD patients. Webb and Brown [33] found that total LDL-cholesterol levels in patients with symptomatic CVD (n = 26) were higher than those in patients without symptomatic CVD (n = 26). There were no associations between triglycerides, LDL-cholesterol or lipoprotein (a) and CVD. In a cross-sectional study of 31 patients, the severity of CAD (based on standard clinical findings) correlated with total cholesterol, LDL-cholesterol, triglycerides and HDL-cholesterol in a multivariate analysis. It is reasonable to use the European Task Force Recommendations for treatment threshold and target lipid levels. Statins are the most effective of all therapies in PD patients with hypercholesterolaemia or combined hyperlipidaemia. Statins are usually safe and effective in lowering both LDL cholesterol and triglycerides in PD patients. Fibric acid analogues should be avoided whenever possible, with the exception of gemfibrozil. However, the use of gemfibrozil should be limited to a dose of 450–600 mg/day (the tablet strength differs between the USA and Europe). The issue of drug accumulation, especially of a major metabolite of gemfibrozil, is not definitively solved in PD patients. From few studies and occasional observations it appears that fibric acid analogues are more toxic to PD patients than to HD patients.