J Am Soc Nephrol 12: 341–348, 2001 Comparative Effects of Cerivastatin and Fenofibrate on the Atherogenic Lipoprotein Phenotype in Proteinuric Renal Disease CHRISTOPHER J. DEIGHAN,*† MURIEL J. CASLAKE,† MICHAEL MCCONNELL,† J. MICHAEL BOULTON-JONES,* and CHRISTOPHER J. PACKARD† *Renal Unit and †Department of Pathological Biochemistry, Glasgow Royal Infirmary, Glasgow, United Kingdom. Abstract. Patients with nephrotic-range proteinuria have im- centration (49%), RLP-C (35%), and RLP-TG (44%; all P Ͻ paired clearance of triglyceride-rich lipoproteins. This results 0.01). Fenofibrate also reduced %LDLIII from 60 to 33% (P Ͻ in the atherogenic lipoprotein phenotype (mild hypertriglycer- 0.01). HDL-C (19%, P Ͻ 0.01) increased with fenofibrate idemia, low high-density lipoproteins [HDL], and excess treatment; LDL-C and total LDL were unchanged. The reduc- small, dense low-density lipoproteins [LDLIII]). Excess rem- tion in LDLIII concentration and RLP-C with fenofibrate treat- nant lipoproteins (RLP) are linked to hypertriglyceridemia and ment correlated with plasma triglyceride reduction (LDLIII r2 may contribute to the atherogenicity of nephrotic dyslipidemia. ϭ 67%, P Ͻ 0.001; RLP cholesterol r2 ϭ 58%, P Ͻ 0.005). A randomized crossover study compared the effects of a statin Serum creatinine increased with fenofibrate treatment (14%, P (cerivastatin) and a fibrate (fenofibrate) on LDLIII and RLP in Ͻ 0.01); however, creatinine clearance was unchanged. LD- 12 patients with nephrotic-range proteinuria. Cerivastatin re- LIII concentration was 187 Ϯ 85 mg/dl after cerivastatin duced cholesterol (21%, P Ͻ 0.01), triglyceride (14%, P Ͻ treatment and 133 Ϯ 95 mg/dl after fenofibrate treatment. 0.05), LDL cholesterol (LDL-C; 23%, P Ͻ 0.01), total LDL Cerivastatin and fenofibrate reduce LDLIII concentration in (18%, P Ͻ 0.01), and LDLIII concentration (27% P Ͻ 0.01). nephrotic-range proteinuria. However, atherogenic concentra- %LDLIII, RLP-C, and RLP triglyceride (RLP-TG) were un- tions of LDLIII remain prevalent after either treatment. Feno- changed. Plasma LDLIII reduction with cerivastatin treatment fibrate but not cerivastatin reduces remnant lipoproteins. The correlated with LDL-C reduction (r2 ϭ 34%, P Ͻ 0.05). two treatments seem to reduce LDLIII by different mecha- Fenofibrate lowered cholesterol (19%), triglyceride (41%), nisms, suggesting a potential role for combination therapy to very low-density lipoprotein cholesterol (52%), LDLIII con- optimize lowering of LDLIII and RLP. Patients with proteinuria have an increased incidence of car- and retrospective studies have demonstrated a link between diovascular disease (1). Those with nephrotic-range proteinuria LDLIII and increased risk of coronary artery disease (4,5), and experience a 5.5-fold increase in the relative risk of myocardial a plasma level of LDLIII Ͼ100 mg/dl confers a sevenfold infarction, even after correction for hypertension and smoking increase in the risk of myocardial infarction (6). We recently (2). More than two thirds of patients with nephrotic syndrome demonstrated that patients with nephrotic-range proteinuria have at least two risk factors for the development of athero- possess atherogenic quantities of LDLIII (7). Very low-density sclerosis (3), and the dyslipidemia that is prevalent in this lipoprotein (VLDL) can similarly be divided into large, light population is likely to contribute to the increased incidence of VLDL1 and smaller, denser VLDL2. VLDL1 and VLDL2 are cardiovascular disease. thought to be regulated independently. VLDL1 is increased in Lipoproteins exhibit heterogeneity. Low-density lipopro- patients with raised triglycerides (6), whereas excess VLDL2 is teins (LDL) can be divided into large, light LDLI (d 1.025 to found in patients with increased LDL cholesterol (LDL-C) (6). 1.034 g/ml) and LDLII (d 1.034 to 1.044 g/ml), and small, In nephrotic-range proteinuria, VLDL1 clearance is delayed dense LDL (LDLIII; d 1.044 to 1.063 g/ml). LDLIII is con- and VLDL2 production is increased (8). As a result, both sidered to be the most atherogenic LDL species. Prospective subfractions are found in increased quantities (8). Pharmacological lipid lowering with either hepatic 3-meth- ylglutaryl coenzyme A reductase inhibitors (statins) or fibrates Received June 29, 1999. Accepted June 18, 2000. reduces cardiovascular morbidity and mortality (9–11). How- Correspondence to Dr. Christopher J. Deighan, Consultant Physician, Re- ever, the majority of patients with coronary artery disease do nal Unit, Walton Building, Glasgow Royal Infirmary, Castle Street, Glas- not have hypercholesterolemia, and a large proportion have a gow, G4 0SF, UK. Phone: 0141-211-5049; Fax: 0141-211-4843; E-mail: [email protected] pattern of mild hypertriglyceridemia, low HDL, and excess LDLIII (12), known as the atherogenic lipoprotein phenotype 1046-6673/1202-0341 Journal of the American Society of Nephrology (6). In the normal populations, statins reduce LDLIII by low- Copyright © 2001 by the American Society of Nephrology ering all three LDL subfractions in concert (13). Fibrates 342 Journal of the American Society of Nephrology J Am Soc Nephrol 12: 341–348, 2001 promote a shift in LDL size toward larger, lighter particles, lesterol ester, phospholipid, and protein contents of all lipoproteins with or without a change in total LDL (14,15). Recent evidence were assayed, and lipoprotein concentrations were calculated as the suggests that remnants of triglyceride-rich lipoproteins may sum of these components. Three LDL subfractions (LDLI, d 1.025 to also be implicated in the pathogenesis of atherosclerosis (16). 1.034 g/ml; LDLII, d 1.034 to 1.044 g/ml; and LDLIII, d 1.044 to Elevated lipoprotein remnants (RLP) are found in patients with 1.063 g/ml) were isolated from fresh plasma by nonequilibrium den- sity gradient ultracentrifugation as described previously (21). The raised plasma triglyceride (17) and therefore may contribute to individual subfraction areas were quantified and expressed as percent- the increased cardiovascular risk of the atherogenic lipoprotein age of total LDL. The lipoprotein concentration of LDL was used to phenotype. generate individual subfraction concentrations. The cholesterol and In addition to hypercholesterolemia, raised plasma triglyc- triglyceride content of remnant lipoproteins (RLP-C, RLP-TG) was eride is well recognized in nephrotic-range proteinuria (18), measured using a diagnostic reagent kit from Japan Immunoresearch with reports of 67% (8) and 71% (3) of patients having a Laboratories (JIMRO; Takasaki, Japan). RLP were separated from plasma triglyceride greater than 2 mmol/L and 2.3 mmol/L, other lipoproteins by an immunoaffinity gel containing monoclonal respectively. Statins have been shown to reduce cholesterol antibodies to human apoB100 and human apoA1 (22). The cholesterol and LDL-C in patients with nephrotic dyslipidemia (8); how- and triglyceride contents of the unbound fraction were measured using ever, this cholesterol-lowering effect is also accompanied by enzymatic spectrophotometric assays. Apolipoproteins B, CII, CIII, plasma triglyceride reduction and increased receptor-mediated and E were analyzed from ethylenediaminetetraacetate plasma using kits purchased from Wako Pure Chemical Industries (Osaka, Japan). LDL clearance (19). LDLIII formation is a consequence of mild increases in plasma triglyceride (6), and therefore the Statistical Analyses triglyceride-lowering effect of statins may have a beneficial Statistical analysis was performed using MINITAB 11 for Win- effect not only on LDLIII concentration but also on the relative dows (Minitab Inc., State College, PA) and SPSS statistical package. proportions of each LDL subfraction and on remnant lipopro- Results are shown as means Ϯ SD. Baseline and end point data were teins. The aim of this study, therefore, was to compare the compared using repeated measures ANOVA, with the two treatment ability of a statin (cerivastatin) and a fibrate (fenofibrate) to modalities further compared using paired t test if the ANOVA showed lower levels of these atherogenic lipoproteins in patients with a significant difference. Pre- and posttreatment data were compared nephrotic-range proteinuria. A secondary aim was to assess using paired t test. Regression analysis identified significant changes in VLDL subfractions and plasma apolipoproteins to correlations. help identify mechanisms that underlie any changes. Results Materials and Methods Baseline Data Patients The diagnoses of the study population were IgA nephropathy Twelve patients (10 male, 2 female) were recruited from the (four), membranous nephropathy (three), minimal change ne- Glasgow Royal Infirmary Renal Unit. Inclusion criteria were (1) phropathy (two), mesangiocapillary glomerulonephritis (two) and biopsy-proven glomerular disease, (2) urinary albumin Ͼ1.5 g/24 h focal and segmental glomerulosclerosis (one). Baseline data ob- (equivalent to 3 g proteinuria/24 h), (3) serum creatinine Ͻ250 tained before treatment are shown in Tables 1 to 3. Values at the ␮mol/L, (4) cholesterol Ͼ6.5 mmol/L, and (5) triglyceride Ͼ1.5 beginning of each treatment phase (baseline and washout) did not mmol/L. Exclusion criteria were (1) significant deterioration in renal differ for any parameter. The median age of the study patients was function in the previous 6 mo and (2) having diseases or receiving 52 yr (range, 18 to 66). Average BP was 143/87 mmHg (nine treatment that influenced lipid profile (specifically, diabetes