- Email: [email protected]
day) in patients with mixed hyperlipidemia
Effects of Simvastatin (40 and 80 mg/day) in Patients With Mixed Hyperlipidemia Evan Stein, MD, PhD, Diane Plotkin, PhD, Harold Bays, MD, Michael Davidson, MD, Carlos Dujovne, MD, Stanley Korenman, MD, Michael Stepanavage, BS, and Michele Mercuri, MD, PhD Mixed hyperlipidemia is characterized by both elevated total cholesterol and triglycerides. It is estimated to account for 10% to 20% of patients with dyslipidemia. This study assessed the lipid-altering efficacy and tolerability of simvastatin 40 and 80 mg/day as monotherapy. One hundred thirty patients (62 women [48%], 24 [16%] with type 2 diabetes mellitus, mean age 53 years) with mixed hyperlipidemia (baseline low-density lipoprotein [LDL] cholesterol 156 mg/dl [mean], and triglycerides 391 mg/dl [median) were randomized in a multicenter, double-masked, placebo-controlled, 3-period, 22-week, balanced crossover study, and received placebo, and simvastatin 40 and 80 mg/day each for 6 weeks. Compared with placebo, simvastatin produced significant (p <0.01) and dose-dependent changes in all lipid and lipoprotein parameters (LDL cholesterol 2.1%, ⴚ28.9%, and ⴚ35.5%; triglycerides ⴚ3.5%, ⴚ27.8%, and
ⴚ33.0%; high-density lipoprotein cholesterol 3.3%, 13.1%, and 15.7%; apolipoprotein B 3.8%, ⴚ23.1%, and ⴚ30.6%; and apolipoprotein A-I 4.0%, 8.2%, and 10.5% with placebo, and simvastatin 40 and 80 mg/ day, respectively). The changes were consistent in patients with diabetes mellitus. One patient taking simvastatin 80 mg/day had an asymptomatic and reversible increase in hepatic transaminases 3 times above the upper limit of normal. Simvastatin 40 and 80 mg/day is effective in patients with mixed hyperlipidemia across the entire lipid and lipoprotein profile. The reductions in LDL cholesterol and triglycerides are large, significant, and dose dependent. The increase in high-density lipoprotein cholesterol was greater than that observed in patients with hypercholesterolemia, and appears dose dependent. 䊚2000 by Excerpta Medica, Inc. (Am J Cardiol 2000;86:406 – 411)
ixed hyperlipidemia is characterized by both elevated total cholesterol and triglycerides due M to the accumulation of low-density lipoprotein (LDL)
counseling improve the plasma levels of all lipoprotein fractions, but are unsuccessful in optimizing or maintaining desirable levels of cholesterol and triglyceride over time. In patients with primary elevations of LDL cholesterol, hydroxymethylglutaryl coenzyme A (HMG-CoA) reductase inhibitors reduce triglycerides by approximately 10% to 15%, and increase highdensity lipoprotein (HDL) cholesterol by 5% to 7%. Most previous studies of mixed hyperlipidemia included subjects with relatively low triglycerides and have also focused on comparing the efficacy of lower doses of HMG-CoA reductase inhibitors with fibric acid derivatives. Furthermore, these studies have lacked a placebo control group.6 –9 The present study determines in a large, placebo-controlled randomized trial the efficacy of simvastatin, 40 and 80 mg/day, on the lipid and lipoprotein profile of patients with only mixed hyperlipidemia.
and very low density lipoprotein (VLDL) lipoproteins. Mixed hyperlipidemia is traditionally referred to as Fredrickson type IIb dyslipidemia,1 and represents a constellation of metabolic disorders that produces a similar lipoprotein profile with variable and different expressions of LDL cholesterol and triglyceride elevations. Mixed hyperlipidemia accounts for 10% to 20% of patients with dyslipidemia,2 whereas familial combined hyperlipidemia is among the most common phenotypic disorders in humans (0.5% to 1%) and is frequently observed in adult survivors of myocardial infarction.3 The lipoprotein profile of mixed hyperlipidemia is highly atherogenic, and mixed hyperlipidemia is often found in patients with obesity, renal disease, or type 2 diabetes.4,5 In most hyperlipidemic patients, lifestyle changes, weight loss, and dietary From the Medical Research Laboratory, Highland Heights, Kentucky; Endocrinology and Metabolism, and Biostatistics, Merck Research Laboratories, Rahway, New Jersey; Louisville Metabolic & Atherosclerosis Research Center, Louisville, Kentucky; Chicago Center for Clinical Research, Chicago, Illinois; Kansas Foundation for Clinical Pharmacology-Radiant Research, Kansas City, Kansas; and University of California at Los Angeles, Los Angeles, California. Manuscript received October 5, 1999; revised manuscript received and accepted March 1, 2000. Address for reprints: Diane Plotkin, PhD, Endocrinology and Metabolism, Merck Research Laboratories, 126 Lincoln Avenue, Rahway, New Jersey, 06065– 0900.
406
©2000 by Excerpta Medica, Inc. All rights reserved. The American Journal of Cardiology Vol. 86 August 15, 2000
METHODS
Patients: Institutional review board approvals were obtained for all centers, and all patients signed a written informed consent. Eligible patients (aged 21 to 70 years) were men, postmenopausal women, and premenopausal women highly unlikely to conceive. Patients taking a lipid-lowering diet but not lipidlowering medication for at least 6 weeks (8 weeks for fibric acid derivatives) with LDL cholesterol levels ⱖ130 mg/dl (3.4 mmol/L), and triglyceride levels between 300 and 700 mg/dl (3.9 and 7.9 mmol/L) on 0002-9149/00/$–see front matter PII S0002-9149(00)00955-3
TABLE I Patient Baseline Demographic and Clinical Characteristics Patients (n ⫽ 130) Mean age (years ⫾ SD) ⱕ40 years ⱖ60 years Men Women White Non-white Cardiovascular disease† Systemic hypertension Type 2 diabetes mellitus*
53 ⫾ 10.3 13 (10.0) 21 (16.2) 68 (52.3) 62 (47.7) 121 (93.1) 9 (6.9) 53 (40.8) 34 (26.2) 24 (18.5)
*Defined as history of type 2 diabetes mellitus or at least 2 fasting serum glucose values ⱖ126 mg/dl. † Defined as a history of angina, myocardial infarction, coronary artery bypass graft, angioplasty.
TABLE II Baseline Lipid Parameters of Randomized Patients (mg/dl, except for ratios) Lipid Parameters
Mean
Median
SD
LDL cholesterol Total cholesterol VLDL cholesterol Non-HDL cholesterol Apolipoprotein B LDL apolipoprotein B Triglyceride HDL cholesterol Apolipoprotein A-I LDL cholesterol/HDL cholesterol* Total cholesterol/HDL cholesterol* Lipoprotein A-I Lipoprotein A-I:A-II Lipoprotein(a)
156 281 86 242 169 142 407 39 143 4.2 7.6 42 101 24
149 277 77 233 167 143 391 38 138 4.0 7.3 40 99 8.0
32 43 42 43 30 31 130 8.8 27.0 1.1 1.9 12 18 33
*n ⫽ 130 patients.
repeat measurement were eligible for inclusion. Patients were classified as having type 2 diabetes mellitus if they either had a clinical diagnosis of type 2 diabetes at study entry, or they had fasting serum glucose levels ⱖ126 mg/dl on at least 2 occasions before randomization, and hemoglobin A1C levels ⬍10%.10 Patients with types I or V hyperlipidemia or homozygous familial hypercholesterolemia were excluded. Patients with serum creatinine ⬎1.8 mg/dl, and patients with type 2 diabetes ⬎1.4 mg/dl, liver transaminases ⬎15% above the upper limit of normal, active liver disease, or unexplained creatine kinase ⬎50% over the upper limit of normal were also excluded. Study design: This was a multicenter, randomized, double-masked, placebo-controlled, 3-period, complete block, balanced, crossover study with a 4-week diet/placebo run-in period before randomization. One hundred thirty eligible patients were randomized to 1 of 6 unique treatment sequences (three 6-week treatment periods) of placebo and simvastatin 40 and 80 mg/day (Zocor, Merck & Co., Inc., Whitehouse Station, New Jersey) given in different orders. There were no washout periods between treatment periods.
From week ⫺4 through study end, patients were instructed to consume a lipid-lowering diet (calories derived from 50% carbohydrate, 20% protein, and 30% fat; a polyunsaturated:saturated fatty acid ratio of 1; and cholesterol content ⬍300 mg/day), and to restrain their alcohol consumption not to exceed 7 drinks/week. Study end points: The primary efficacy end point was the percent change in LDL cholesterol relative to baseline. Secondary end points included percent change in total cholesterol, VLDL cholesterol, HDL cholesterol, triglycerides, non-HDL cholesterol, apolipoprotein B, apolipoprotein A-I, LDL apolipoprotein B, LDL cholesterol/HDL cholesterol, total cholesterol/HDL cholesterol, lipoprotein A-I only particles, lipoprotein A-I:A-II, and lipoprotein(a), all relative to baseline values. Safety and tolerability were evaluated by assessing clinical and laboratory adverse event reports and laboratory determinations. Alanine aminotransferase or aspartate aminotransferase elevations ⱖ3 times, or creatine kinase ⱖ10 times the upper limit of normal were considered clinically relevant, and persistent elevations resulted in patient withdrawal from the study. Laboratory assessment: All blood samples were obtained in the fasting state (10 to 12 hours), airfreighted overnight at 4°C, to the central laboratory (Medical Research Laboratories, Highland Heights, Kentucky) for analysis. Throughout the study, the laboratory participated in and remained certified by the National Heart, Lung, and Blood Institute, Centers for Disease Control Part III program.11 All lipid, lipoprotein, and apolipoprotein fractions were analyzed on ethylenediaminetetraacetic acid–treated plasma as previously described.12,13 LDL and VLDL cholesterol levels were determined by ultracentrifugation.14 LDL cholesterol was obtained by subtracting HDL cholesterol from the density ⱖ1.006 g/ml fraction cholesterol. VLDL cholesterol levels were obtained by subtracting the density ⬎1.006 g/ml cholesterol from the total cholesterol. A single laboratory measurement was obtained at the end of each 6-week treatment period. Statistical analyses: For efficacy end points, the percent change from baseline was analyzed. Betweengroup comparisons were performed using a linear model analysis of variance, with factors for treatment, center, patient (center), and period. Carryover effect was tested and removed from the analysis of variance model due to nonsignificance (p ⬎0.050). Because of the relatively small number of patients with type 2 diabetes and observed variability, robust nonparametric summary statistics were used in the efficacy analyses. The study was designed based on a minimum of 96 evaluable patients completing the study. This sample size corresponded to a detectable difference between treatments of 4.9% in LDL cholesterol percent change from baseline (90% power, ␣ ⫽ 0.050, 2-sided test). All statistical tests were 2-tailed with ⌱ ⫽ 0.05. The p values were rounded to 3 decimal places, with p ⱕ0.05 considered statistically significant.
PREVENTIVE CARDIOLOGY/SIMVASTATIN IN MIXED HYPERLIPIDEMIA
407
TABLE III Treatment Effects of Placebo and Simvastatin 40 and 80 mg/day on Fasting Levels of Apolipoprotein B–Containing Fractions (all evaluable patients) Baseline (SD)*
Week 6 (SD)*
% ⌬†
95% CI‡
⌬ vs 40 mg (p value)
⌬ vs 80 mg (p value)
⬍0.001
⬍0.001 ⬍0.001
Non-HDL Cholesterol Placebo 40 mg 80 mg
240 241 242
242.0 163 146
0.8 ⫺31.6 ⫺39.1
⫺0.9, 3.0 ⫺33.6, ⫺29.8 ⫺41.4, ⫺37.7
sistent with Fredrickson type III dyslipidemia. All results are presented as means (or medians) and SDs. The 95% confidence intervals are shown where appropriate. All analyses used the intention-to-treat approach, i.e., patients with a baseline measurement and at least 1 on-treatment measurement that was carried forward. Changes in lipid and lipoprotein levels: Tables III to V summarize the
Apolipoprotein B
lipid and lipoprotein efficacy results obtained after 6 weeks of treatment. Mean changes from baseline for LDL cholesterol levels were 2%, ⫺29%, and ⫺36% for the placebo, LDL Cholesterol and simvastatin 40- and 80-mg/day Placebo 156 159 2.1 1.1, 5.9 ⬍0.001 ⬍0.001 treatments, respectively, and were 40 mg 156 111 ⫺28.9 ⫺30.3, ⫺25.4 ⬍0.001 significant (p ⬍0.001) in the simva80 mg 156 99 ⫺35.5 ⫺37.2, ⫺32.5 statin subjects. Simvastatin treatment reduced all apolipoprotein B– conLDL Apolipoprotein taining fractions, attaining reductions Placebo 141 147 3.9 1.5, 9.0 ⬍0.001 ⬍0.001 in VLDL cholesterol of 44.6% (me40 mg 142 112 ⫺21.1 ⫺23.6, ⫺16.1 0.002 dian) and in non-HDL cholesterol of 80 mg 142 99 ⫺28.9 ⫺31.3, ⫺23.9 39.1% with the 80-mg dose. Triglyceride measurements, as expected, VLDL Cholesterol displayed considerable variability in Placebo 77 73 ⫺3.5 ⫺9.1, 2.1 ⬍0.001 ⬍0.001 percent change from baseline for all 40 mg 76 48 ⫺36.9 ⫺42.2, ⫺31.7 ⬍0.001 treatments, and thus a more robust 80 mg 77 44 ⫺44.6 ⫺49.5, ⫺39.6 rank analysis was used. The median *Baseline values and percent change are median values for VLDL cholesterol, triglycerides, and baseline level for triglyceride-evalulipoprotein(a), and mean values for the other parameters. able patients was 389 mg/dl, with † Percent change from baseline. median percent changes from base‡ 95% confidence interval (CI) for the mean or median as reported. line of ⫺3.5%, ⫺27.8%, and ⫺33.0% observed for the placebo, and simvastatin 40-, and 80-mg/day RESULTS treatments. Significant pairwise differences were obPatients: The first patient entered the study in May served for each of the treatments (p ⬍0.05). A signif1997 and the last patient completed the study in March icant HDL cholesterol dose-response relation was ev1998. ident across the placebo to simvastatin 80-mg/day During the recruitment period, 226 potential sub- doses, i.e., each of the pairwise comparisons of treatjects were screened and 130 patients (121 Cauca- ment means was significant (p ⬍0.050). Mean HDL sians), who met the eligibility requirements, were cholesterol baseline values were 39.0 mg/dl, and mean randomized. Of these, 118 patients (91%) completed percent changes in HDL cholesterol from baseline for the study, with 6 (4.6%), 3 (2.3%), and 3 (2.3%) placebo, and simvastatin 40- and 80-mg/day treatpatients in the placebo and simvastatin 40- and 80- ments were 3.3%, 13.1%, and 15.7%, respectively. mg/day groups, respectively, withdrawing from the Apolipoprotein A– containing fractions showed study. Four subjects discontinued because of an adchanges consistent with those observed for HDL choverse experience: 2 during period 1, one on simvastatin due to a rash and one on placebo with prostate lesterol. Mean percent changes from baseline for apocarcinoma; 1 during period 2 on placebo with an lipoprotein A-I were 4.0%, 8.2%, and 10.5% for the allergic reaction, and 1 during period 3 on placebo placebo, and simvastatin 40- and 80-mg treatments, with back pain. Eight subjects withdrew consent for a respectively. Similar changes were observed for livariety of personal reasons: 7 during period 1, 3 on poprotein A-I and A-I:A-II particles. Concordant with placebo, 3 on simvastatin 80 mg, and 1 on simvastatin the changes observed in the cholesterol fractions, pro4 mg. The eighth patient was in period 2 on placebo. gressively increasing percent reductions across the Baseline demographic characteristics for randomized placebo to simvastatin 80-mg/day treatments were obpatients are presented in Table I, and mean (or me- served for LDL cholesterol/HDL cholesterol and total dian) baseline lipid values are presented in Table II. cholesterol/HDL cholesterol (Table V). Again, pairTwenty-four of the 130 patients had type 2 diabetes. wise differences for each of the treatment comparisons Eight randomized patients were homozygous for the were significant (p ⬍0.001), demonstrating a signifiapolipoprotein E2 allele (apoE2/2), a genotype con- cant dose-response relation. Placebo 40 mg 80 mg
169 169 169
175 130 117
3.8 ⫺23.1 ⫺30.6
2.8, 6.4 ⫺24.3, ⫺20.8 ⫺32.1, ⫺28.6
408 THE AMERICAN JOURNAL OF CARDIOLOGY姞
⬍0.001
VOL. 86
⬍0.001 ⬍0.001
AUGUST 15, 2000
TABLE IV Treatment Effects of Placebo and Simvastatin 40 and 80 mg/day on Fasting Levels of Apolipoprotein A–Containing Fractions Baseline (SD)*
Week 6 (SD)*
% ⌬†
95% CI‡
⌬ vs 40 mg (p value)
⌬ vs 80 mg (p value)
upper limit of normal) and aspartate aminotransferase (⬎3 times the upper limit of normal).
DISCUSSION
HDL Cholesterol
This study set out to determine if monotherapy with simvastatin at Placebo 39 40 3.3 2.5, 6.2 ⬍0.001 ⬍0.001 higher doses of 40 and 80 mg/day 40 mg 39 44 13.1 12.3, 16.0 0.033 was effective in the treatment of 80 mg 39 45 15.7 15.0, 18.7 mixed hyperlipidemia. The hypotheses were tested by using a placeboApolipoprotein A-I controlled, crossover study in which Placebo 144 149 4.0 2.7, 5.8 ⬍0.001 ⬍0.001 all subjects received the 3 treatments. 40 mg 144 154 8.2 6.9, 10.0 0.030 Mixed hyperlipidemia was carefully 80 mg 144 158 10.5 9.2, 12.2 defined as LDL cholesterol ⬎130 mg/dl and triglyceride levels beLipoprotein A-I tween 300 and 700 mg/dl. These criPlacebo 42.4 43.1 2.9 0.4, 5.7 0.004 ⬍0.001 teria insured selection of patients 40 mg 42.5 45.2 8.3 5.6, 10.9 0.057 with a dyslipidemic syndrome char80 mg 42.5 46.8 11.8 9.0, 14.3 acterized by elevated cholesterol (i.e., apolipoprotein B– containing liLipoprotein A-I:A-II poproteins, or LDL cholesterol and Placebo 100 105 5.0 3.2, 7.2 ⬍0.001 ⬍0.001 VLDL cholesterol), elevated triglyc40 mg 101 109 9.1 7.7, 11.6 0.242 erides, and relatively low HDL cho80 mg 101 111 10.7 9.2, 13.1 lesterol levels. The results clearly *Baseline values and percent change are median values for VLDL cholesterol, triglycerides, and demonstrate that the higher doses of lipoprotein(a), and mean values for the other parameters. simvastatin (40 and 80 mg/day) † Percent change from baseline. alone significantly altered the lipid ‡ 95% confidence interval (CI) for the mean or median as reported. (total cholesterol and triglyceride), lipoprotein (LDL, VLDL, HDL), and Patients with type 2 diabetes mellitus: A subgroup of apolipoprotein (apolipoprotein B and AI) profiles of 24 patients with type 2 diabetes was analyzed sepa- patients with mixed hyperlipidemia. Effects on LDL cholesterol and other apolipoprotein rately and compared with patients without diabetes for the key lipid parameters. Patients with type 2 diabetes B– containing lipoproteins: Apolipoprotein B– containhad a median age of 59 years, and 13 were men (54%) ing lipoproteins, and LDL cholesterol in particular, are and 11 were women (46%). The treatment-by-sub- established primary targets of cardiovascular prevengroup effects (diabetes vs nondiabetes) were exam- tion and remain the main therapeutic targets in pained for significance for each of the key lipid param- tients at risk for coronary heart disease.15 Although eters (Table VI), and none were significant (p ⬎0.100) HMG-CoA reductase inhibitors have been used in (i.e., the treatment effects were consistent between millions of patients worldwide, and studied in thousands of hypercholesterolemic patients in controlled diabetic and nondiabetic patients). Tolerability: The percentage of patients reporting clinical studies, less is known about their effects in clinical adverse events attributed by the investigators patients with mixed hyperlipidemia. In this study, to drug treatment was comparable during the placebo LDL cholesterol changes were large and dose depen(4 patients, 3.2%), 40-mg simvastatin (8 patients, dent, with the expected 6% dose-dependent additional 6.5%), and 80-mg simvastatin (5 patients, 4.0%) treat- efficacy achieved by doubling the dose.16 Similar to ment periods. During the study there were 4 serious previous observations with atorvastatin in patients nonfatal, non-drug–related clinical adverse events (1 with elevated triglyceride,17,18 LDL cholesterol reduc[prostatic neoplasm] during placebo, 1 [chest pain] in tions with simvastatin in patients with mixed hyperthe simvastatin 40-mg/day group, and 2 [urolithiasis, lipidemia were less than those seen in patients with right arm pain] in the simvastatin 80-mg/day treat- hypercholesterolemia and relatively normal triglycerment). No patient was withdrawn because of a drug- ides in whom simvastatin 40 and 80 mg/day reduced related adverse event in the placebo or simvastatin LDL cholesterol by 40% and 47%, respectively.19 40-mg/day groups, and 1 patient was withdrawn dur- This reduced LDL cholesterol response in subjects ing the 80-mg/day treatment because of a rash. There with mixed hyperlipidemia, in whom circulating levwere no cases of myopathy, and the incidence of els of VLDL, LDL, and often intermediate-density myalgia was 1%, 2%, and 1% for the placebo, and lipoprotein or remnants are increased, is consistent simvastatin 40- and 80-mg/day treatments, respec- with increased competition for the LDL receptor by tively, with none considered drug related by the in- the more triglyceride-rich apolipoprotein B– containvestigators. After 6 weeks of treatment with simvasta- ing lipoproteins, especially those enriched in apolitin 80-mg/day, 1 patient experienced a transient ele- poprotein E. In patients with mixed hyperlipidemia, vation in both alanine aminotransferase (⬎4 times the particularly those with the familial form, type 2 diaPREVENTIVE CARDIOLOGY/SIMVASTATIN IN MIXED HYPERLIPIDEMIA
409
observations in patients with primary hypercholesterolemia, for whom reported increases in HDL cholesterol Baseline Week 6 ⌬ vs 40 mg ⌬ vs 80 mg are typically more modest in the 4% † ‡ (SD)* (SD)* %⌬ 95% CI (p value) (p value) to 6% range.19,21 The large increases Triglycerides in HDL in patients with mixed hyperlipidemia were supported by asPlacebo 389 376 ⫺3.5 ⫺9.4, 2.4 ⬍0.001 ⬍0.001 sociated increases in total apoli40 mg 389 278 ⫺27.8 ⫺32.2, ⫺23.3 0.007 poprotein A-I (8.2% and 10.5% on 80 mg 391 263 ⫺33.0 ⫺37.6, ⫺28.4 40 and 80 mg/day, respectively, vs Total Cholesterol 4.0% on placebo). These increases in apolipoprotein A-I were concomitant Placebo 279 282 1.3 0.0, 3.2 ⬍0.001 ⬍0.001 with marked reductions in triglycer40 mg 279 207 ⫺25.3 ⫺26.9, ⫺23.6 ⬍0.001 ide-rich lipoproteins, also known to 80 mg 280 191 ⫺31.4 ⫺33.2, ⫺30.0 contain increased concentrations of apolipoprotein A-I. These observaLDL/HDL Cholesterol tions indicate that the increases in Placebo 4.1 4.1 ⫺0.8 ⫺2.4, 2.2 ⬍0.001 ⬍0.001 apolipoprotein A-I concentration re40 mg 4.2 2.7 ⫺36.5 ⫺38.5, ⫺33.9 ⬍0.001 flected increases in the number of 80 mg 4.2 2.3 ⫺43.7 ⫺46.1, ⫺41.5 HDL cholesterol particles rather than an enrichment in the cholesterol conTotal/HDL Cholesterol tents of HDL. This finding was subPlacebo 7.5 7.5 ⫺1.2 ⫺3.6, 0.5 ⬍0.001 ⬍0.001 stantiated by a high, statistically sig40 mg 7.5 5.0 ⫺33.3 ⫺36.0, ⫺31.9 ⬍0.001 nificant increase (p ⬍0.005) in li80 mg 7.6 4.5 ⫺40.0 ⫺43.0, ⫺38.9 poprotein A-I during treatment with simvastatin. Lipoprotein(a) Effects on triglycerides: Although Placebo 8.0 9.0 1.0 0.2, 1.8 0.319 0.140 an independent role for triglycerides 40 mg 8.0 8.0 1.0 0.3, 1.7 0.629 in promoting risk for coronary heart 80 mg 8.0 8.0 1.0 0.2, 1.8 disease is still controversial, many *Baseline values and % change are median values for VLDL cholesterol, triglycerides, and lipoproepidemiologists regard triglyceride tein(a), and mean values for the other parameters. elevations as an important medical † Percent change from baseline. problem,22 and physicians regard it ‡ 95% confidence interval (CI) for the mean or median as reported. as a challenging therapeutic dilemma, especially in patients with mixed hyperlipidemia. This study betes or dysbetalipoproteinemia, increased levels of demonstrated significant, large, and dose-dependent triglyceride-rich lipoprotein combined with elevated median reductions in triglyceride levels of 27.8% and LDL is believed to be more atherogenic than in sub- 33% from baseline after 6 weeks of treatment with jects with isolated LDL elevation. In these conditions, simvastatin 40 and 80 mg/day, respectively. The balthe particles are smaller and more cholesterol-en- anced crossover design and inclusion of a placebo riched than in patients with isolated hypertriglyceri- treatment period increases the strength of these results demia.4 In the present study, VLDL cholesterol levels because each subject serves as their own control in the were substantially reduced by 36.9% and 44.6% with assessment of triglyceride dose dependence. Additionsimvastatin 40 and 80 mg/day, respectively. Similar ally, determination of any triglyceride regression to reductions were observed for non-HDL cholesterol. the mean effect was assessed by the placebo treatment Non-HDL cholesterol has been proposed as a more response. These data are consistent with findings by reliable parameter to assess coronary heart disease risk Bruckert et al,6 who analyzed the comparative efficacy and treatment efficacy, especially in mixed hyperlip- of simvastatin (mean dose about 30 mg/day) and sevidemia because it includes all of the apolipoprotein eral fibric acid derivatives and suggested that simvaB– containing or atherogenic lipoproteins.20 Reduc- statin produced a larger percent reduction in triglyctions in the ratio between total apolipoprotein B and erides when baseline levels were high. In a pooled non-HDL cholesterol were also comparable, indicat- analysis of nearly 2,700 patients who participated in ing that changes were achieved mainly through a studies with a number of different HMG-CoA reducreduction in the number of particles, and not through tase inhibitors, Stein et al23 reported a large and dosea redistribution of cholesterol between different dependent effect of HMG-CoA reductase inhibitors in atherogenic lipoprotein classes. patients with mixed hyperlipidemia. Stein et al also Effects on HDL cholesterol: Of particular clinical in- described a consistent 1:1 ratio for triglyceride and terest was the observation of large and dose-related LDL cholesterol reductions in subjects with mixed increases in HDL cholesterol with simvastatin therapy hyperlipidemia, a finding confirmed in this prospec(3.3%, 13.1%, and 15.7% for placebo, and simvastatin tive study. The present study, in which the median 40 and 80 mg, respectively). This is in contrast to triglyceride level of all patients randomized at baseTABLE V Treatment Effects of Placebo and Simvastatin 40 and 80 mg/day on Fasting Levels of Other Lipid End Points
410 THE AMERICAN JOURNAL OF CARDIOLOGY姞
VOL. 86
AUGUST 15, 2000
TABLE VI Treatment Effects of Placebo and Simvastatin 40 and 80 mg/day on Testing Levels of Lipids and Lipoproteins in Patients With and Without Type 2 Diabetes* Group
Treatment
LDL-C†
VLDL-C†
TG†
Non-HDL-C†
HDL-C†
Type 2 Diabetes
Placebo S 40 mg S 80 mg
5.0 (13.3) ⫺28.7 (23.7) ⫺39.5 (8.9)
⫺1.1 (39.4) ⫺40.2 (34.2) ⫺49.2 (46.2)
⫺1.6 (38.3) ⫺27.5 (27.6) ⫺33.0 (46.6)
2.6 (11.9) ⫺33.4 (16.0) ⫺42.7 (17.8)
1.1 (9.2) 9.0 (15.0) 13.0 (13.3)
No Diabetes
Placebo S 40 mg S 80 mg
0.9 (16.9) ⫺31.1 (19.3) ⫺35.8 (17.8)
5.1 (31.0) ⫺34.4 (30.5) ⫺43.1 (25.7)
⫺4.0 (32.9) ⫺28.3 (24.7) ⫺33.4 (23.6)
.8 (14.7) ⫺32.8 (17.3) ⫺38.0 (13.8)
3.2 (12.8) 12.3 (15.8) 15.7 (17.6)
*Values are expressed as median (SD) percent change from baseline at week 6. † Treatment by subgroup (diabetic vs nondiabetic) interaction for each of the treatments was not significant (p ⬎0.100 for all lipid parameters). Pairwise comparisons for each of the treatments were significant, p ⬍0.050. C ⫽ cholesterol; S ⫽ simvastatin; TG ⫽ triglycerides.
line was 389 mg/dl, confirms and extends the efficacy of simvastatin up to a dose of 80 mg/day in patients with mixed hyperlipidemia. This study also confirms and extends data reported by Stein et al7 in which simvastatin 10 mg/day reduced triglycerides 20% in a prospective and randomized but open-label study of 56 patients with more modest mixed hyperlipidemia (median triglyceride 231 mg/dl). Thus, simvastatin 40 and 80 mg/day produces large and significant dosedependent reductions in apolipoprotein B– containing lipoproteins and significant increases in in apolipoprotein AI– containing lipoproteins. Acknowledgment: We express sincere gratitude to the patients who volunteered to participate in the study, the clinic study coordinators (Leah Adams, Brian Gebhart, Glenn Lamkin, Leslie Votaw, Sherrie Viosca, Nancy Burgin, Erving London, Kevin Curtis, Ann Berenbaum), the Merck Medical Research Associates (Glorilyn Nelson-Plaza, Theresa Sheehan, Eric Olson, Dennis Watkins, Emma Gonzalez, Allan Rees, Jacqueline Smith, Maureen MacDonald, and Van J. Medina), and Medical Program Coordinators (Patricia Betz-Schiff, Hae Sook Lee, Laura O’Grady). Special thanks go to Drs. Jonathan Isaacsohn (Metabolic and Atherosclerosis Center, Cincinnati, Ohio), Donald Hunninghake (University of Minnesota, Minneapolis, Minnesota), David Robertson (Emory University, Atlanta, GA), Leonard Keilson (Center for Lipids and Cardiovascular Health, Portland, Maine), and Stuart Weiss (Endocrine and Metabolic Clinic, San Diego, California) for participating in the study and providing their professional assistance.
1. Fredrickson DS, Lees RS. Editorial. A system for phenotyping hyperlipopro-
teinemia. Circulation 1965;321:327. 2. The Lipid Research Clinics Program Epidemiology Committee. Plasma lipid
distributions in selected in North American population. The Lipid Research Clinics Program Prevalence Study. Circulation 1979 60:427– 439. 3. Brunzell J, Schrott H, Motulsky A, Bierman E. Myocardial infarction in the familial forms of hypertriglyceridemia. Metabolism 1976;25:313–320. 4. Grundy S. Lipids, nutrition, and coronary heart disease. In. Fuster V, Ross R, Topol E, eds. Atherosclerosis and Coronary Artery Disease. Vol. 1. Philadelphia/ New York: Lipincott-Raven Publishers, 1996:45– 68. 5. Grundy SM. Hypertriglyceridemia, atherogenic dyslipidemia, and the metabolic syndrome. Am J Cardiol 1998;81:18b-25b.
6. Bruckert E, De Gennes JL, Malbecq W, Baigts F. Comparison of the efficacy
of simvastatin and standard fibrate therapy in the treatment of primary hypercholesterolemia and combined hyperlipidemia. Clin Cardiol 1995;18:621– 629. 7. Stein EA, Davidson MH, Dujovne CA, Hunninghake DB, Goldberg RB, Illingworth DR, Knopp RH, Miller VT, Frost P, Isaacsohn JL, Mitchel YB, Melino MR, Shapiro D, Tobert JA. Efficacy and tolerability of low-dose simvastatin and niacin, alone and in combination, in patients with combined hyperlipidemia: a prospective trial. J Cardovasc Pharmacol Ther 1996;1:107–116. 8. Knopp RH. Drug treatment of lipid disorders. N Engl J Med 1999;341:498 – 511. 9. Tobert JA. Efficacy and long-term adverse effect pattern of lovastatin. Am J Cardiol 1988;62:28J-34J. 10. American Diabetes Association. Management of dyslipidemia in adults with diabetes. Diabetes Care 1998;21:179 –182. 11. Myers GL, Cooper GR, Winn CL, Smith SJ. The Centers for Disease Control-National Heart, Lung, and Blood Institute Lipid Standardization Program: an approach to accurate and precise lipid measurements. Clin Lab Med 1989;9:105–135. 12. Steiner P, Freidel J, Bremner W, Stein E. Standardization of micromethods for plasma cholesterol, triglyceride and HDL-cholesterol with the lipid clinics’ methodology. J Clin Chem 1981;19:850 – 851. 13. Warnick G, Albers J. A comprehensive evaluation of the heparin manganese precipitation procedure for estimating high-density lipoprotein cholesterol. J Lipid Res 1978;19:65–76. 14. Program LRC. Manual of Laboratory Operations: Lipid and Lipoprotein Analysis. Washington, DC: US Dept. of Health Education and Welfare. NIH publication no. 75-628, 1982. 15. Expert Panel on Detection. Evaluation and Treatment of High Blood Cholesterol in Adults. Summary of the second report of the National Cholesterol Education Program (NCEP) expert panel on detection, evaluation, and treatment of high blood cholesterol in adults (Adult Treatment Panel II). JAMA 1993;269: 3015–3023. 16. Davidson MH, Stein EA, Dujovne CA, Hunninghake DB, Weiss SR, Knopp RH, Illingworth DR, Mitchel YB, Melino MR, Zupkis RV, Dobrinska MR, Amin RD, Tobert JA. The efficacy and six-week tolerability of simvastatin 80 and 160 mg/day. Am J Cardiol 1997;79:38 – 42. 17. Bakker-Arkema RG, Davidson MH, Goldstein RJ, Davignon J, Isaacsohn JL, Weiss SR, Keilson LM, Brown WV, Miller VT, Shurzinske LJ, Black DM. Efficacy and safety of a new HMG-CoA reductase inhibitor, atorvastatin, in patients with hypertriglyceridemia. JAMA 1996;275:128 –133. 18. Alaupovic P, Heinonen T, Shurzinske L, Black DM. Effect of a new HMG-CoA reductase inhibitor, atorvastatin, on lipids, apolipoproteins and lipoprotein particles in patients with elevated serum cholesterol and triglyceride levels. Atherosclerosis 1997;133:123–133. 19. Stein E, Davidson M, Dobs A, Schrott H, Dujovne C, Bays H, Weiss S, Melino M, Stepanavage M, Mitchel Y. Efficacy and safety of simvastatin 80 mg/day in hypercholesterolemic patients. Am J Cardiol 1998;82:311–316. 20. Frost PH, Havel RJ. Rationale for use of non-high-density lipoprotein cholesterol rather than low-density lipoprotein cholesterol as a tool for lipoprotein cholesterol screening and assessment of risk and therapy. Am J Cardiol 1998; 81:26B-31B. 21. Ose L, Kastelein J, Scott R, Stein E, Campodonico S, Escobar J, Tate A, Corsetti L, Shahane A, Mitchel Y, Mercuri M. Efficacy and six-month safety of simvastatin 80 mg/day: results from the Worldwide Simvastatin Expanded Dose Program (WSEDP). Nutr Metab Cardiovasc Dis 1998;8:143–151. 22. Austin MA, Hokanson JE, Edwards KL. Hypertriglyceridemia as a cardiovascular risk factor. Am J Cardiol 1998;81:7b-12b. 23. Stein EA, Lane M, Laskarzewski P. Comparison of statins in hypertriglyceridemia. Am J Cardiol 1998;81:66B-69B.
PREVENTIVE CARDIOLOGY/SIMVASTATIN IN MIXED HYPERLIPIDEMIA
411
ⴚ33.0%; high-density lipoprotein cholesterol 3.3%, 13.1%, and 15.7%; apolipoprotein B 3.8%, ⴚ23.1%, and ⴚ30.6%; and apolipoprotein A-I 4.0%, 8.2%, and 10.5% with placebo, and simvastatin 40 and 80 mg/ day, respectively). The changes were consistent in patients with diabetes mellitus. One patient taking simvastatin 80 mg/day had an asymptomatic and reversible increase in hepatic transaminases 3 times above the upper limit of normal. Simvastatin 40 and 80 mg/day is effective in patients with mixed hyperlipidemia across the entire lipid and lipoprotein profile. The reductions in LDL cholesterol and triglycerides are large, significant, and dose dependent. The increase in high-density lipoprotein cholesterol was greater than that observed in patients with hypercholesterolemia, and appears dose dependent. 䊚2000 by Excerpta Medica, Inc. (Am J Cardiol 2000;86:406 – 411)
ixed hyperlipidemia is characterized by both elevated total cholesterol and triglycerides due M to the accumulation of low-density lipoprotein (LDL)
counseling improve the plasma levels of all lipoprotein fractions, but are unsuccessful in optimizing or maintaining desirable levels of cholesterol and triglyceride over time. In patients with primary elevations of LDL cholesterol, hydroxymethylglutaryl coenzyme A (HMG-CoA) reductase inhibitors reduce triglycerides by approximately 10% to 15%, and increase highdensity lipoprotein (HDL) cholesterol by 5% to 7%. Most previous studies of mixed hyperlipidemia included subjects with relatively low triglycerides and have also focused on comparing the efficacy of lower doses of HMG-CoA reductase inhibitors with fibric acid derivatives. Furthermore, these studies have lacked a placebo control group.6 –9 The present study determines in a large, placebo-controlled randomized trial the efficacy of simvastatin, 40 and 80 mg/day, on the lipid and lipoprotein profile of patients with only mixed hyperlipidemia.
and very low density lipoprotein (VLDL) lipoproteins. Mixed hyperlipidemia is traditionally referred to as Fredrickson type IIb dyslipidemia,1 and represents a constellation of metabolic disorders that produces a similar lipoprotein profile with variable and different expressions of LDL cholesterol and triglyceride elevations. Mixed hyperlipidemia accounts for 10% to 20% of patients with dyslipidemia,2 whereas familial combined hyperlipidemia is among the most common phenotypic disorders in humans (0.5% to 1%) and is frequently observed in adult survivors of myocardial infarction.3 The lipoprotein profile of mixed hyperlipidemia is highly atherogenic, and mixed hyperlipidemia is often found in patients with obesity, renal disease, or type 2 diabetes.4,5 In most hyperlipidemic patients, lifestyle changes, weight loss, and dietary From the Medical Research Laboratory, Highland Heights, Kentucky; Endocrinology and Metabolism, and Biostatistics, Merck Research Laboratories, Rahway, New Jersey; Louisville Metabolic & Atherosclerosis Research Center, Louisville, Kentucky; Chicago Center for Clinical Research, Chicago, Illinois; Kansas Foundation for Clinical Pharmacology-Radiant Research, Kansas City, Kansas; and University of California at Los Angeles, Los Angeles, California. Manuscript received October 5, 1999; revised manuscript received and accepted March 1, 2000. Address for reprints: Diane Plotkin, PhD, Endocrinology and Metabolism, Merck Research Laboratories, 126 Lincoln Avenue, Rahway, New Jersey, 06065– 0900.
406
©2000 by Excerpta Medica, Inc. All rights reserved. The American Journal of Cardiology Vol. 86 August 15, 2000
METHODS
Patients: Institutional review board approvals were obtained for all centers, and all patients signed a written informed consent. Eligible patients (aged 21 to 70 years) were men, postmenopausal women, and premenopausal women highly unlikely to conceive. Patients taking a lipid-lowering diet but not lipidlowering medication for at least 6 weeks (8 weeks for fibric acid derivatives) with LDL cholesterol levels ⱖ130 mg/dl (3.4 mmol/L), and triglyceride levels between 300 and 700 mg/dl (3.9 and 7.9 mmol/L) on 0002-9149/00/$–see front matter PII S0002-9149(00)00955-3
TABLE I Patient Baseline Demographic and Clinical Characteristics Patients (n ⫽ 130) Mean age (years ⫾ SD) ⱕ40 years ⱖ60 years Men Women White Non-white Cardiovascular disease† Systemic hypertension Type 2 diabetes mellitus*
53 ⫾ 10.3 13 (10.0) 21 (16.2) 68 (52.3) 62 (47.7) 121 (93.1) 9 (6.9) 53 (40.8) 34 (26.2) 24 (18.5)
*Defined as history of type 2 diabetes mellitus or at least 2 fasting serum glucose values ⱖ126 mg/dl. † Defined as a history of angina, myocardial infarction, coronary artery bypass graft, angioplasty.
TABLE II Baseline Lipid Parameters of Randomized Patients (mg/dl, except for ratios) Lipid Parameters
Mean
Median
SD
LDL cholesterol Total cholesterol VLDL cholesterol Non-HDL cholesterol Apolipoprotein B LDL apolipoprotein B Triglyceride HDL cholesterol Apolipoprotein A-I LDL cholesterol/HDL cholesterol* Total cholesterol/HDL cholesterol* Lipoprotein A-I Lipoprotein A-I:A-II Lipoprotein(a)
156 281 86 242 169 142 407 39 143 4.2 7.6 42 101 24
149 277 77 233 167 143 391 38 138 4.0 7.3 40 99 8.0
32 43 42 43 30 31 130 8.8 27.0 1.1 1.9 12 18 33
*n ⫽ 130 patients.
repeat measurement were eligible for inclusion. Patients were classified as having type 2 diabetes mellitus if they either had a clinical diagnosis of type 2 diabetes at study entry, or they had fasting serum glucose levels ⱖ126 mg/dl on at least 2 occasions before randomization, and hemoglobin A1C levels ⬍10%.10 Patients with types I or V hyperlipidemia or homozygous familial hypercholesterolemia were excluded. Patients with serum creatinine ⬎1.8 mg/dl, and patients with type 2 diabetes ⬎1.4 mg/dl, liver transaminases ⬎15% above the upper limit of normal, active liver disease, or unexplained creatine kinase ⬎50% over the upper limit of normal were also excluded. Study design: This was a multicenter, randomized, double-masked, placebo-controlled, 3-period, complete block, balanced, crossover study with a 4-week diet/placebo run-in period before randomization. One hundred thirty eligible patients were randomized to 1 of 6 unique treatment sequences (three 6-week treatment periods) of placebo and simvastatin 40 and 80 mg/day (Zocor, Merck & Co., Inc., Whitehouse Station, New Jersey) given in different orders. There were no washout periods between treatment periods.
From week ⫺4 through study end, patients were instructed to consume a lipid-lowering diet (calories derived from 50% carbohydrate, 20% protein, and 30% fat; a polyunsaturated:saturated fatty acid ratio of 1; and cholesterol content ⬍300 mg/day), and to restrain their alcohol consumption not to exceed 7 drinks/week. Study end points: The primary efficacy end point was the percent change in LDL cholesterol relative to baseline. Secondary end points included percent change in total cholesterol, VLDL cholesterol, HDL cholesterol, triglycerides, non-HDL cholesterol, apolipoprotein B, apolipoprotein A-I, LDL apolipoprotein B, LDL cholesterol/HDL cholesterol, total cholesterol/HDL cholesterol, lipoprotein A-I only particles, lipoprotein A-I:A-II, and lipoprotein(a), all relative to baseline values. Safety and tolerability were evaluated by assessing clinical and laboratory adverse event reports and laboratory determinations. Alanine aminotransferase or aspartate aminotransferase elevations ⱖ3 times, or creatine kinase ⱖ10 times the upper limit of normal were considered clinically relevant, and persistent elevations resulted in patient withdrawal from the study. Laboratory assessment: All blood samples were obtained in the fasting state (10 to 12 hours), airfreighted overnight at 4°C, to the central laboratory (Medical Research Laboratories, Highland Heights, Kentucky) for analysis. Throughout the study, the laboratory participated in and remained certified by the National Heart, Lung, and Blood Institute, Centers for Disease Control Part III program.11 All lipid, lipoprotein, and apolipoprotein fractions were analyzed on ethylenediaminetetraacetic acid–treated plasma as previously described.12,13 LDL and VLDL cholesterol levels were determined by ultracentrifugation.14 LDL cholesterol was obtained by subtracting HDL cholesterol from the density ⱖ1.006 g/ml fraction cholesterol. VLDL cholesterol levels were obtained by subtracting the density ⬎1.006 g/ml cholesterol from the total cholesterol. A single laboratory measurement was obtained at the end of each 6-week treatment period. Statistical analyses: For efficacy end points, the percent change from baseline was analyzed. Betweengroup comparisons were performed using a linear model analysis of variance, with factors for treatment, center, patient (center), and period. Carryover effect was tested and removed from the analysis of variance model due to nonsignificance (p ⬎0.050). Because of the relatively small number of patients with type 2 diabetes and observed variability, robust nonparametric summary statistics were used in the efficacy analyses. The study was designed based on a minimum of 96 evaluable patients completing the study. This sample size corresponded to a detectable difference between treatments of 4.9% in LDL cholesterol percent change from baseline (90% power, ␣ ⫽ 0.050, 2-sided test). All statistical tests were 2-tailed with ⌱ ⫽ 0.05. The p values were rounded to 3 decimal places, with p ⱕ0.05 considered statistically significant.
PREVENTIVE CARDIOLOGY/SIMVASTATIN IN MIXED HYPERLIPIDEMIA
407
TABLE III Treatment Effects of Placebo and Simvastatin 40 and 80 mg/day on Fasting Levels of Apolipoprotein B–Containing Fractions (all evaluable patients) Baseline (SD)*
Week 6 (SD)*
% ⌬†
95% CI‡
⌬ vs 40 mg (p value)
⌬ vs 80 mg (p value)
⬍0.001
⬍0.001 ⬍0.001
Non-HDL Cholesterol Placebo 40 mg 80 mg
240 241 242
242.0 163 146
0.8 ⫺31.6 ⫺39.1
⫺0.9, 3.0 ⫺33.6, ⫺29.8 ⫺41.4, ⫺37.7
sistent with Fredrickson type III dyslipidemia. All results are presented as means (or medians) and SDs. The 95% confidence intervals are shown where appropriate. All analyses used the intention-to-treat approach, i.e., patients with a baseline measurement and at least 1 on-treatment measurement that was carried forward. Changes in lipid and lipoprotein levels: Tables III to V summarize the
Apolipoprotein B
lipid and lipoprotein efficacy results obtained after 6 weeks of treatment. Mean changes from baseline for LDL cholesterol levels were 2%, ⫺29%, and ⫺36% for the placebo, LDL Cholesterol and simvastatin 40- and 80-mg/day Placebo 156 159 2.1 1.1, 5.9 ⬍0.001 ⬍0.001 treatments, respectively, and were 40 mg 156 111 ⫺28.9 ⫺30.3, ⫺25.4 ⬍0.001 significant (p ⬍0.001) in the simva80 mg 156 99 ⫺35.5 ⫺37.2, ⫺32.5 statin subjects. Simvastatin treatment reduced all apolipoprotein B– conLDL Apolipoprotein taining fractions, attaining reductions Placebo 141 147 3.9 1.5, 9.0 ⬍0.001 ⬍0.001 in VLDL cholesterol of 44.6% (me40 mg 142 112 ⫺21.1 ⫺23.6, ⫺16.1 0.002 dian) and in non-HDL cholesterol of 80 mg 142 99 ⫺28.9 ⫺31.3, ⫺23.9 39.1% with the 80-mg dose. Triglyceride measurements, as expected, VLDL Cholesterol displayed considerable variability in Placebo 77 73 ⫺3.5 ⫺9.1, 2.1 ⬍0.001 ⬍0.001 percent change from baseline for all 40 mg 76 48 ⫺36.9 ⫺42.2, ⫺31.7 ⬍0.001 treatments, and thus a more robust 80 mg 77 44 ⫺44.6 ⫺49.5, ⫺39.6 rank analysis was used. The median *Baseline values and percent change are median values for VLDL cholesterol, triglycerides, and baseline level for triglyceride-evalulipoprotein(a), and mean values for the other parameters. able patients was 389 mg/dl, with † Percent change from baseline. median percent changes from base‡ 95% confidence interval (CI) for the mean or median as reported. line of ⫺3.5%, ⫺27.8%, and ⫺33.0% observed for the placebo, and simvastatin 40-, and 80-mg/day RESULTS treatments. Significant pairwise differences were obPatients: The first patient entered the study in May served for each of the treatments (p ⬍0.05). A signif1997 and the last patient completed the study in March icant HDL cholesterol dose-response relation was ev1998. ident across the placebo to simvastatin 80-mg/day During the recruitment period, 226 potential sub- doses, i.e., each of the pairwise comparisons of treatjects were screened and 130 patients (121 Cauca- ment means was significant (p ⬍0.050). Mean HDL sians), who met the eligibility requirements, were cholesterol baseline values were 39.0 mg/dl, and mean randomized. Of these, 118 patients (91%) completed percent changes in HDL cholesterol from baseline for the study, with 6 (4.6%), 3 (2.3%), and 3 (2.3%) placebo, and simvastatin 40- and 80-mg/day treatpatients in the placebo and simvastatin 40- and 80- ments were 3.3%, 13.1%, and 15.7%, respectively. mg/day groups, respectively, withdrawing from the Apolipoprotein A– containing fractions showed study. Four subjects discontinued because of an adchanges consistent with those observed for HDL choverse experience: 2 during period 1, one on simvastatin due to a rash and one on placebo with prostate lesterol. Mean percent changes from baseline for apocarcinoma; 1 during period 2 on placebo with an lipoprotein A-I were 4.0%, 8.2%, and 10.5% for the allergic reaction, and 1 during period 3 on placebo placebo, and simvastatin 40- and 80-mg treatments, with back pain. Eight subjects withdrew consent for a respectively. Similar changes were observed for livariety of personal reasons: 7 during period 1, 3 on poprotein A-I and A-I:A-II particles. Concordant with placebo, 3 on simvastatin 80 mg, and 1 on simvastatin the changes observed in the cholesterol fractions, pro4 mg. The eighth patient was in period 2 on placebo. gressively increasing percent reductions across the Baseline demographic characteristics for randomized placebo to simvastatin 80-mg/day treatments were obpatients are presented in Table I, and mean (or me- served for LDL cholesterol/HDL cholesterol and total dian) baseline lipid values are presented in Table II. cholesterol/HDL cholesterol (Table V). Again, pairTwenty-four of the 130 patients had type 2 diabetes. wise differences for each of the treatment comparisons Eight randomized patients were homozygous for the were significant (p ⬍0.001), demonstrating a signifiapolipoprotein E2 allele (apoE2/2), a genotype con- cant dose-response relation. Placebo 40 mg 80 mg
169 169 169
175 130 117
3.8 ⫺23.1 ⫺30.6
2.8, 6.4 ⫺24.3, ⫺20.8 ⫺32.1, ⫺28.6
408 THE AMERICAN JOURNAL OF CARDIOLOGY姞
⬍0.001
VOL. 86
⬍0.001 ⬍0.001
AUGUST 15, 2000
TABLE IV Treatment Effects of Placebo and Simvastatin 40 and 80 mg/day on Fasting Levels of Apolipoprotein A–Containing Fractions Baseline (SD)*
Week 6 (SD)*
% ⌬†
95% CI‡
⌬ vs 40 mg (p value)
⌬ vs 80 mg (p value)
upper limit of normal) and aspartate aminotransferase (⬎3 times the upper limit of normal).
DISCUSSION
HDL Cholesterol
This study set out to determine if monotherapy with simvastatin at Placebo 39 40 3.3 2.5, 6.2 ⬍0.001 ⬍0.001 higher doses of 40 and 80 mg/day 40 mg 39 44 13.1 12.3, 16.0 0.033 was effective in the treatment of 80 mg 39 45 15.7 15.0, 18.7 mixed hyperlipidemia. The hypotheses were tested by using a placeboApolipoprotein A-I controlled, crossover study in which Placebo 144 149 4.0 2.7, 5.8 ⬍0.001 ⬍0.001 all subjects received the 3 treatments. 40 mg 144 154 8.2 6.9, 10.0 0.030 Mixed hyperlipidemia was carefully 80 mg 144 158 10.5 9.2, 12.2 defined as LDL cholesterol ⬎130 mg/dl and triglyceride levels beLipoprotein A-I tween 300 and 700 mg/dl. These criPlacebo 42.4 43.1 2.9 0.4, 5.7 0.004 ⬍0.001 teria insured selection of patients 40 mg 42.5 45.2 8.3 5.6, 10.9 0.057 with a dyslipidemic syndrome char80 mg 42.5 46.8 11.8 9.0, 14.3 acterized by elevated cholesterol (i.e., apolipoprotein B– containing liLipoprotein A-I:A-II poproteins, or LDL cholesterol and Placebo 100 105 5.0 3.2, 7.2 ⬍0.001 ⬍0.001 VLDL cholesterol), elevated triglyc40 mg 101 109 9.1 7.7, 11.6 0.242 erides, and relatively low HDL cho80 mg 101 111 10.7 9.2, 13.1 lesterol levels. The results clearly *Baseline values and percent change are median values for VLDL cholesterol, triglycerides, and demonstrate that the higher doses of lipoprotein(a), and mean values for the other parameters. simvastatin (40 and 80 mg/day) † Percent change from baseline. alone significantly altered the lipid ‡ 95% confidence interval (CI) for the mean or median as reported. (total cholesterol and triglyceride), lipoprotein (LDL, VLDL, HDL), and Patients with type 2 diabetes mellitus: A subgroup of apolipoprotein (apolipoprotein B and AI) profiles of 24 patients with type 2 diabetes was analyzed sepa- patients with mixed hyperlipidemia. Effects on LDL cholesterol and other apolipoprotein rately and compared with patients without diabetes for the key lipid parameters. Patients with type 2 diabetes B– containing lipoproteins: Apolipoprotein B– containhad a median age of 59 years, and 13 were men (54%) ing lipoproteins, and LDL cholesterol in particular, are and 11 were women (46%). The treatment-by-sub- established primary targets of cardiovascular prevengroup effects (diabetes vs nondiabetes) were exam- tion and remain the main therapeutic targets in pained for significance for each of the key lipid param- tients at risk for coronary heart disease.15 Although eters (Table VI), and none were significant (p ⬎0.100) HMG-CoA reductase inhibitors have been used in (i.e., the treatment effects were consistent between millions of patients worldwide, and studied in thousands of hypercholesterolemic patients in controlled diabetic and nondiabetic patients). Tolerability: The percentage of patients reporting clinical studies, less is known about their effects in clinical adverse events attributed by the investigators patients with mixed hyperlipidemia. In this study, to drug treatment was comparable during the placebo LDL cholesterol changes were large and dose depen(4 patients, 3.2%), 40-mg simvastatin (8 patients, dent, with the expected 6% dose-dependent additional 6.5%), and 80-mg simvastatin (5 patients, 4.0%) treat- efficacy achieved by doubling the dose.16 Similar to ment periods. During the study there were 4 serious previous observations with atorvastatin in patients nonfatal, non-drug–related clinical adverse events (1 with elevated triglyceride,17,18 LDL cholesterol reduc[prostatic neoplasm] during placebo, 1 [chest pain] in tions with simvastatin in patients with mixed hyperthe simvastatin 40-mg/day group, and 2 [urolithiasis, lipidemia were less than those seen in patients with right arm pain] in the simvastatin 80-mg/day treat- hypercholesterolemia and relatively normal triglycerment). No patient was withdrawn because of a drug- ides in whom simvastatin 40 and 80 mg/day reduced related adverse event in the placebo or simvastatin LDL cholesterol by 40% and 47%, respectively.19 40-mg/day groups, and 1 patient was withdrawn dur- This reduced LDL cholesterol response in subjects ing the 80-mg/day treatment because of a rash. There with mixed hyperlipidemia, in whom circulating levwere no cases of myopathy, and the incidence of els of VLDL, LDL, and often intermediate-density myalgia was 1%, 2%, and 1% for the placebo, and lipoprotein or remnants are increased, is consistent simvastatin 40- and 80-mg/day treatments, respec- with increased competition for the LDL receptor by tively, with none considered drug related by the in- the more triglyceride-rich apolipoprotein B– containvestigators. After 6 weeks of treatment with simvasta- ing lipoproteins, especially those enriched in apolitin 80-mg/day, 1 patient experienced a transient ele- poprotein E. In patients with mixed hyperlipidemia, vation in both alanine aminotransferase (⬎4 times the particularly those with the familial form, type 2 diaPREVENTIVE CARDIOLOGY/SIMVASTATIN IN MIXED HYPERLIPIDEMIA
409
observations in patients with primary hypercholesterolemia, for whom reported increases in HDL cholesterol Baseline Week 6 ⌬ vs 40 mg ⌬ vs 80 mg are typically more modest in the 4% † ‡ (SD)* (SD)* %⌬ 95% CI (p value) (p value) to 6% range.19,21 The large increases Triglycerides in HDL in patients with mixed hyperlipidemia were supported by asPlacebo 389 376 ⫺3.5 ⫺9.4, 2.4 ⬍0.001 ⬍0.001 sociated increases in total apoli40 mg 389 278 ⫺27.8 ⫺32.2, ⫺23.3 0.007 poprotein A-I (8.2% and 10.5% on 80 mg 391 263 ⫺33.0 ⫺37.6, ⫺28.4 40 and 80 mg/day, respectively, vs Total Cholesterol 4.0% on placebo). These increases in apolipoprotein A-I were concomitant Placebo 279 282 1.3 0.0, 3.2 ⬍0.001 ⬍0.001 with marked reductions in triglycer40 mg 279 207 ⫺25.3 ⫺26.9, ⫺23.6 ⬍0.001 ide-rich lipoproteins, also known to 80 mg 280 191 ⫺31.4 ⫺33.2, ⫺30.0 contain increased concentrations of apolipoprotein A-I. These observaLDL/HDL Cholesterol tions indicate that the increases in Placebo 4.1 4.1 ⫺0.8 ⫺2.4, 2.2 ⬍0.001 ⬍0.001 apolipoprotein A-I concentration re40 mg 4.2 2.7 ⫺36.5 ⫺38.5, ⫺33.9 ⬍0.001 flected increases in the number of 80 mg 4.2 2.3 ⫺43.7 ⫺46.1, ⫺41.5 HDL cholesterol particles rather than an enrichment in the cholesterol conTotal/HDL Cholesterol tents of HDL. This finding was subPlacebo 7.5 7.5 ⫺1.2 ⫺3.6, 0.5 ⬍0.001 ⬍0.001 stantiated by a high, statistically sig40 mg 7.5 5.0 ⫺33.3 ⫺36.0, ⫺31.9 ⬍0.001 nificant increase (p ⬍0.005) in li80 mg 7.6 4.5 ⫺40.0 ⫺43.0, ⫺38.9 poprotein A-I during treatment with simvastatin. Lipoprotein(a) Effects on triglycerides: Although Placebo 8.0 9.0 1.0 0.2, 1.8 0.319 0.140 an independent role for triglycerides 40 mg 8.0 8.0 1.0 0.3, 1.7 0.629 in promoting risk for coronary heart 80 mg 8.0 8.0 1.0 0.2, 1.8 disease is still controversial, many *Baseline values and % change are median values for VLDL cholesterol, triglycerides, and lipoproepidemiologists regard triglyceride tein(a), and mean values for the other parameters. elevations as an important medical † Percent change from baseline. problem,22 and physicians regard it ‡ 95% confidence interval (CI) for the mean or median as reported. as a challenging therapeutic dilemma, especially in patients with mixed hyperlipidemia. This study betes or dysbetalipoproteinemia, increased levels of demonstrated significant, large, and dose-dependent triglyceride-rich lipoprotein combined with elevated median reductions in triglyceride levels of 27.8% and LDL is believed to be more atherogenic than in sub- 33% from baseline after 6 weeks of treatment with jects with isolated LDL elevation. In these conditions, simvastatin 40 and 80 mg/day, respectively. The balthe particles are smaller and more cholesterol-en- anced crossover design and inclusion of a placebo riched than in patients with isolated hypertriglyceri- treatment period increases the strength of these results demia.4 In the present study, VLDL cholesterol levels because each subject serves as their own control in the were substantially reduced by 36.9% and 44.6% with assessment of triglyceride dose dependence. Additionsimvastatin 40 and 80 mg/day, respectively. Similar ally, determination of any triglyceride regression to reductions were observed for non-HDL cholesterol. the mean effect was assessed by the placebo treatment Non-HDL cholesterol has been proposed as a more response. These data are consistent with findings by reliable parameter to assess coronary heart disease risk Bruckert et al,6 who analyzed the comparative efficacy and treatment efficacy, especially in mixed hyperlip- of simvastatin (mean dose about 30 mg/day) and sevidemia because it includes all of the apolipoprotein eral fibric acid derivatives and suggested that simvaB– containing or atherogenic lipoproteins.20 Reduc- statin produced a larger percent reduction in triglyctions in the ratio between total apolipoprotein B and erides when baseline levels were high. In a pooled non-HDL cholesterol were also comparable, indicat- analysis of nearly 2,700 patients who participated in ing that changes were achieved mainly through a studies with a number of different HMG-CoA reducreduction in the number of particles, and not through tase inhibitors, Stein et al23 reported a large and dosea redistribution of cholesterol between different dependent effect of HMG-CoA reductase inhibitors in atherogenic lipoprotein classes. patients with mixed hyperlipidemia. Stein et al also Effects on HDL cholesterol: Of particular clinical in- described a consistent 1:1 ratio for triglyceride and terest was the observation of large and dose-related LDL cholesterol reductions in subjects with mixed increases in HDL cholesterol with simvastatin therapy hyperlipidemia, a finding confirmed in this prospec(3.3%, 13.1%, and 15.7% for placebo, and simvastatin tive study. The present study, in which the median 40 and 80 mg, respectively). This is in contrast to triglyceride level of all patients randomized at baseTABLE V Treatment Effects of Placebo and Simvastatin 40 and 80 mg/day on Fasting Levels of Other Lipid End Points
410 THE AMERICAN JOURNAL OF CARDIOLOGY姞
VOL. 86
AUGUST 15, 2000
TABLE VI Treatment Effects of Placebo and Simvastatin 40 and 80 mg/day on Testing Levels of Lipids and Lipoproteins in Patients With and Without Type 2 Diabetes* Group
Treatment
LDL-C†
VLDL-C†
TG†
Non-HDL-C†
HDL-C†
Type 2 Diabetes
Placebo S 40 mg S 80 mg
5.0 (13.3) ⫺28.7 (23.7) ⫺39.5 (8.9)
⫺1.1 (39.4) ⫺40.2 (34.2) ⫺49.2 (46.2)
⫺1.6 (38.3) ⫺27.5 (27.6) ⫺33.0 (46.6)
2.6 (11.9) ⫺33.4 (16.0) ⫺42.7 (17.8)
1.1 (9.2) 9.0 (15.0) 13.0 (13.3)
No Diabetes
Placebo S 40 mg S 80 mg
0.9 (16.9) ⫺31.1 (19.3) ⫺35.8 (17.8)
5.1 (31.0) ⫺34.4 (30.5) ⫺43.1 (25.7)
⫺4.0 (32.9) ⫺28.3 (24.7) ⫺33.4 (23.6)
.8 (14.7) ⫺32.8 (17.3) ⫺38.0 (13.8)
3.2 (12.8) 12.3 (15.8) 15.7 (17.6)
*Values are expressed as median (SD) percent change from baseline at week 6. † Treatment by subgroup (diabetic vs nondiabetic) interaction for each of the treatments was not significant (p ⬎0.100 for all lipid parameters). Pairwise comparisons for each of the treatments were significant, p ⬍0.050. C ⫽ cholesterol; S ⫽ simvastatin; TG ⫽ triglycerides.
line was 389 mg/dl, confirms and extends the efficacy of simvastatin up to a dose of 80 mg/day in patients with mixed hyperlipidemia. This study also confirms and extends data reported by Stein et al7 in which simvastatin 10 mg/day reduced triglycerides 20% in a prospective and randomized but open-label study of 56 patients with more modest mixed hyperlipidemia (median triglyceride 231 mg/dl). Thus, simvastatin 40 and 80 mg/day produces large and significant dosedependent reductions in apolipoprotein B– containing lipoproteins and significant increases in in apolipoprotein AI– containing lipoproteins. Acknowledgment: We express sincere gratitude to the patients who volunteered to participate in the study, the clinic study coordinators (Leah Adams, Brian Gebhart, Glenn Lamkin, Leslie Votaw, Sherrie Viosca, Nancy Burgin, Erving London, Kevin Curtis, Ann Berenbaum), the Merck Medical Research Associates (Glorilyn Nelson-Plaza, Theresa Sheehan, Eric Olson, Dennis Watkins, Emma Gonzalez, Allan Rees, Jacqueline Smith, Maureen MacDonald, and Van J. Medina), and Medical Program Coordinators (Patricia Betz-Schiff, Hae Sook Lee, Laura O’Grady). Special thanks go to Drs. Jonathan Isaacsohn (Metabolic and Atherosclerosis Center, Cincinnati, Ohio), Donald Hunninghake (University of Minnesota, Minneapolis, Minnesota), David Robertson (Emory University, Atlanta, GA), Leonard Keilson (Center for Lipids and Cardiovascular Health, Portland, Maine), and Stuart Weiss (Endocrine and Metabolic Clinic, San Diego, California) for participating in the study and providing their professional assistance.
1. Fredrickson DS, Lees RS. Editorial. A system for phenotyping hyperlipopro-
teinemia. Circulation 1965;321:327. 2. The Lipid Research Clinics Program Epidemiology Committee. Plasma lipid
distributions in selected in North American population. The Lipid Research Clinics Program Prevalence Study. Circulation 1979 60:427– 439. 3. Brunzell J, Schrott H, Motulsky A, Bierman E. Myocardial infarction in the familial forms of hypertriglyceridemia. Metabolism 1976;25:313–320. 4. Grundy S. Lipids, nutrition, and coronary heart disease. In. Fuster V, Ross R, Topol E, eds. Atherosclerosis and Coronary Artery Disease. Vol. 1. Philadelphia/ New York: Lipincott-Raven Publishers, 1996:45– 68. 5. Grundy SM. Hypertriglyceridemia, atherogenic dyslipidemia, and the metabolic syndrome. Am J Cardiol 1998;81:18b-25b.
6. Bruckert E, De Gennes JL, Malbecq W, Baigts F. Comparison of the efficacy
of simvastatin and standard fibrate therapy in the treatment of primary hypercholesterolemia and combined hyperlipidemia. Clin Cardiol 1995;18:621– 629. 7. Stein EA, Davidson MH, Dujovne CA, Hunninghake DB, Goldberg RB, Illingworth DR, Knopp RH, Miller VT, Frost P, Isaacsohn JL, Mitchel YB, Melino MR, Shapiro D, Tobert JA. Efficacy and tolerability of low-dose simvastatin and niacin, alone and in combination, in patients with combined hyperlipidemia: a prospective trial. J Cardovasc Pharmacol Ther 1996;1:107–116. 8. Knopp RH. Drug treatment of lipid disorders. N Engl J Med 1999;341:498 – 511. 9. Tobert JA. Efficacy and long-term adverse effect pattern of lovastatin. Am J Cardiol 1988;62:28J-34J. 10. American Diabetes Association. Management of dyslipidemia in adults with diabetes. Diabetes Care 1998;21:179 –182. 11. Myers GL, Cooper GR, Winn CL, Smith SJ. The Centers for Disease Control-National Heart, Lung, and Blood Institute Lipid Standardization Program: an approach to accurate and precise lipid measurements. Clin Lab Med 1989;9:105–135. 12. Steiner P, Freidel J, Bremner W, Stein E. Standardization of micromethods for plasma cholesterol, triglyceride and HDL-cholesterol with the lipid clinics’ methodology. J Clin Chem 1981;19:850 – 851. 13. Warnick G, Albers J. A comprehensive evaluation of the heparin manganese precipitation procedure for estimating high-density lipoprotein cholesterol. J Lipid Res 1978;19:65–76. 14. Program LRC. Manual of Laboratory Operations: Lipid and Lipoprotein Analysis. Washington, DC: US Dept. of Health Education and Welfare. NIH publication no. 75-628, 1982. 15. Expert Panel on Detection. Evaluation and Treatment of High Blood Cholesterol in Adults. Summary of the second report of the National Cholesterol Education Program (NCEP) expert panel on detection, evaluation, and treatment of high blood cholesterol in adults (Adult Treatment Panel II). JAMA 1993;269: 3015–3023. 16. Davidson MH, Stein EA, Dujovne CA, Hunninghake DB, Weiss SR, Knopp RH, Illingworth DR, Mitchel YB, Melino MR, Zupkis RV, Dobrinska MR, Amin RD, Tobert JA. The efficacy and six-week tolerability of simvastatin 80 and 160 mg/day. Am J Cardiol 1997;79:38 – 42. 17. Bakker-Arkema RG, Davidson MH, Goldstein RJ, Davignon J, Isaacsohn JL, Weiss SR, Keilson LM, Brown WV, Miller VT, Shurzinske LJ, Black DM. Efficacy and safety of a new HMG-CoA reductase inhibitor, atorvastatin, in patients with hypertriglyceridemia. JAMA 1996;275:128 –133. 18. Alaupovic P, Heinonen T, Shurzinske L, Black DM. Effect of a new HMG-CoA reductase inhibitor, atorvastatin, on lipids, apolipoproteins and lipoprotein particles in patients with elevated serum cholesterol and triglyceride levels. Atherosclerosis 1997;133:123–133. 19. Stein E, Davidson M, Dobs A, Schrott H, Dujovne C, Bays H, Weiss S, Melino M, Stepanavage M, Mitchel Y. Efficacy and safety of simvastatin 80 mg/day in hypercholesterolemic patients. Am J Cardiol 1998;82:311–316. 20. Frost PH, Havel RJ. Rationale for use of non-high-density lipoprotein cholesterol rather than low-density lipoprotein cholesterol as a tool for lipoprotein cholesterol screening and assessment of risk and therapy. Am J Cardiol 1998; 81:26B-31B. 21. Ose L, Kastelein J, Scott R, Stein E, Campodonico S, Escobar J, Tate A, Corsetti L, Shahane A, Mitchel Y, Mercuri M. Efficacy and six-month safety of simvastatin 80 mg/day: results from the Worldwide Simvastatin Expanded Dose Program (WSEDP). Nutr Metab Cardiovasc Dis 1998;8:143–151. 22. Austin MA, Hokanson JE, Edwards KL. Hypertriglyceridemia as a cardiovascular risk factor. Am J Cardiol 1998;81:7b-12b. 23. Stein EA, Lane M, Laskarzewski P. Comparison of statins in hypertriglyceridemia. Am J Cardiol 1998;81:66B-69B.
PREVENTIVE CARDIOLOGY/SIMVASTATIN IN MIXED HYPERLIPIDEMIA
411