(Circulation. 2001;103:1942.)
© 2001 American Heart Association, Inc.
Clinical Investigation and Reports |
From the Division of Clinical Chemistry (W.M., H.S., M.M.H., H.W.) and the Division of Sports Medicine, Department of Medicine (A.B., J.K., M.W.B.), Albert Ludwigs-University, Freiburg, Germany; and Novartis Pharma AG, Nürnberg, Germany (C.A.).
Correspondence to Winfried März, MD, Department of Medicine, Hugstetter Straße 55, D-79106 Freiburg, Germany. E-mail maerz{at}med1.ukl.uni-freiburg.de
| Abstract |
|---|
|
|
|---|
Methods and ResultsIn a multicenter, double-blind, randomized, placebo-controlled study, we determined the effect of the HMGRI fluvastatin on lipids, apolipoproteins, and LDL subfractions (by equilibrium density gradient ultracentrifugation). A total of 52 postmenopausal women with combined hyperlipidemia and increased dLDL were treated with either fluvastatin 40 mg/d (n=35) or placebo (n=17). After 12 weeks treatment, significant reductions (P<0.001) in total cholesterol (-19%), IDL cholesterol (-35%), LDL cholesterol (-23%), apolipoprotein B (-21%), and apolipoprotein B in dLDL (-42%) were apparent among fluvastatin recipients. No significant changes in triglycerides or HDL cholesterol were observed. The effect of fluvastatin on dLDL was correlated with baseline values. There was no consistent relationship, however, between the effect of fluvastatin on triglycerides and the decrease in dLDL.
ConclusionsFluvastatin lowers total and LDL cholesterol and the concentration of dLDL. This profile may contribute to an antiatherogenic effect for fluvastatin that is greater than expected on the basis of changes in lipids and apolipoproteins.
Key Words: fluvastatin coronary disease lipoproteins
| Introduction |
|---|
|
|
|---|
A link between dLDL and increased risk of coronary heart disease (CHD) was first proposed by Austin and coworkers.4 Subsequent case-control and prospective studies have shown that a predominance of dLDL increases the risk of CHD by up to 7-fold.5 6 7 8 9 10 11 12 Evidence for a role of dLDL in CHD also comes from angiographic trials. In the St Thomas Atherosclerosis Regression Study, patients showing regression of coronary atherosclerosis had low concentrations of dLDL during treatment.13 Dense LDLs are associated with other components of the metabolic syndrome, particularly elevated triglycerides and low HDL cholesterol (HDL-C). The independent contribution of dLDL to the risk of CHD has therefore been difficult to determine. Thus, LDL subclass profile remained predictive of CHD after adjustment for triglycerides or HDL in some9 11 but not all studies.6 7 8 10 12 Very recently,14 dLDLs were shown to cause endothelial dysfunction independent of LDL cholesterol (LDL-C), triglycerides, and HDL-C. Clearly, any therapeutic principle that lowers LDL-C and triglycerides and raises HDL-C might therefore be enhanced by a reduction in dLDL.
The effects of HMG-CoA reductase inhibitors (HMGRIs) on LDL subfractions have been studied by gradient gel electrophoresis (GGE) and ultracentrifugation.15 16 17 18 19 20 21 22 23 24 Providing a coarse estimate of the size distribution of LDL rather than the concentration of dLDL, GGE did not reveal increases of the average LDL diameter during HMGRI treatment.16 19 20 25 Some of the ultracentrifugation-based studies showed a decrease of the concentration of dLDL15 19 on HMGRI; others did not.18 19 22 23 24 It is thus controversial whether these agents lower dLDL. The HMGRI fluvastatin offers a wide spectrum of clinical benefits in various patient subgroups, including the elderly.26 We therefore examined the effect of short-term therapy with fluvastatin on dLDL levels in patients at increased risk of CHD, namely, postmenopausal women with an atherogenic lipoprotein profile.
| Methods |
|---|
|
|
|---|
Patients
A total of 52 postmenopausal women (
12 months
since last menstrual period, levels of follicle-stimulating hormone
>28 IU/L), 44 to 75 years old, with LDL-C >150 mg/dL, triglycerides
>120 mg/dL, and apolipoprotein (apo) B in dLDL (LDL-5+LDL-6) >25
mg/dL, participated. Major exclusion criteria were LDL-C
300 mg/dL;
triglycerides
500 mg/dL; acute myocardial infarction within 3 months
of study commencement; insulin-dependent diabetes mellitus or poorly
controlled noninsulin-dependent diabetes mellitus (glucose >150
mg/dL or HbA1c >8%); severe obesity; overt
liver disease; chronic renal failure; myopathy; alcohol or drug abuse;
several other significant diseases; known hypersensitivity to HMGRIs;
or use of other lipid-lowering therapy, immunosuppressants,
erythromycin and/or neomycin, ketoconazole, and hormone-replacement
therapy.
Patients commenced an 8-week run-in period, during which previous lipid-lowering therapy was discontinued. Dietary advice was provided according to the American Heart Association Step I diet, and patients were requested to maintain smoking habits, physical activity, and alcohol consumption. After the run-in phase, patients were randomized to receive either fluvastatin 40 mg every evening (n=35) or placebo (n=17) for 12 weeks. A randomization ratio of 2:1 was selected to reduce the number of patients receiving the potentially inferior treatment. At week 6, concomitant diseases, adverse events, and compliance (capsule counting) were recorded, and at week 12, laboratory assessments and physical examinations were repeated.
Laboratory Assessments
Two weeks before randomization (baseline) and after
12 weeks active treatment, fasting venous blood samples were drawn.
Serum samples were stored up to 1 week at 4°C before lipoprotein
subfractionation. Previous experiments indicated that the lipid and
lipoprotein measurements were not affected by storage at 4°C for 1
week. All laboratory assessments were performed centrally at the
Department of Medicine, University of Freiburg,
Germany.
Lipids and Apolipoproteins
Cholesterol, triglycerides, and phospholipids were
measured with enzymatic methods, calibrated with secondary standards
from Roche Diagnostics. ApoB was determined by kinetic nephelometry
(Beh-ring), standardized by reference to the Centers for Disease
Control standard.27 Lipid
and apoB measurements had coefficients of variation of
<5%.
Lipoproteins and Lipoprotein
Subfractions
Quantitative lipoprotein
electrophoresis28 was used
to estimate LDL-C before subfractionation at week -2. VLDL
(d<1.0063 kg/L), IDL
(1.0063<d<1.019 kg/L), LDL
(1.019<d<1.065 kg/L), and HDL
(1.065<d<1.21 kg/L) were
isolated by
ultracentrifugation.29 LDL
was subsequently fractionated into 6 density classes by equilibrium
density gradient
centrifugation.27 LDL
subfractions were quantified by measurement of lipids and apoB. The
coefficients of variation for the analysis of cholesterol, free
cholesterol, phospholipids, and apoB in LDL subfractions were <7%
(and <10% in the case of triglycerides). Radii of LDL subfractions
were calculated as
described.27
Statistical Methods
Contingency tables were analyzed by Fishers exact
test. Intraindividual changes between baseline and week 12 were
calculated and compared between treatment groups by Students
t test. Bivariate correlations
were analyzed by Pearsons correlation coefficients. A value of
P<0.05 was considered
significant.
| Results |
|---|
|
|
|---|
|
Patients were recruited from 10 centers, 1 of which was overrepresented (providing 11 and 6 patients in the fluvastatin and placebo groups, respectively). Exclusion of patients at this center from the statistical analyses did not affect the study findings.
Changes in Lipids, Lipoproteins, and
Apolipoproteins
After 12 weeks treatment, significant reductions in
total cholesterol, LDL-C, and total apoB were observed among
fluvastatin recipients compared with placebo
(Table 2
). Tri-glycerides and HDL-C did not change
significantly.
|
Major ApoB-Containing Lipoproteins
No significant changes in VLDL constituents were
observed
(Table 3
). More marked changes were apparent for IDL. With
the exception of triglycerides (decrease by 9%), all constituents of
IDL decreased by
35%, with apoB in IDL decreasing by almost 30%
(Table 3
). With regard to LDL, total cholesterol decreased
by 23%, and apoB in LDL decreased by 24%; similar changes were
observed for esterified and nonesterified cholesterol and phospholipid
content
(Table 3
). Relative to placebo, the effect of fluvastatin on
LDL triglycerides (-14%) was not statistically significant. No
significant changes in particle radius were apparent for VLDL, IDL, or
LDL during fluvastatin therapy
(Table 3
).
|
LDL Subfractions
Among LDL subfractions, the most marked changes were
observed in LDL-5 and LDL-6
(Figure 1
). Fluvastatin lowered lipids and apoB in LDL-5 by
40% to 45%
(Table 4
). In LDL-6, the respective reductions ranged from
36% to 42%, with the exception of triglycerides, which decreased by
25%
(Table 4
). For dLDL (LDL-5+LDL-6), relative changes were
39% to 44% for apoB, cholesterol, and phospholipids and 33% for
triglycerides
(Table 4
). Less marked (but statistically significant)
changes in the concentration of LDL-4 were also observed, whereas the
lightest subfractions (LDL-1 and LDL-2) remained unaffected
(Figure 1
).
|
|
Correlation Analysis
The decrease in LDL-6 apoB was most closely related to
the levels of LDL-6 apoB before treatment
(r=-0.878,
P<0.001), ie, the higher the
baseline LDL-6 apoB level, the greater the reduction evoked by
fluvastatin
(Figure 2
). Total triglycerides
(r=-0.403,
P=0.016) and VLDL apoB
(r=-0.589,
P<0.001) were weaker, albeit
significant, predictors of the change in LDL-6 apoB. Interestingly,
positive correlations were obtained between apoB in LDL-2 to LDL-4 and
the change in LDL-6 apoB, ie, the higher the apoB level in these
fractions, the smaller the change in LDL-6 apoB during fluvastatin
therapy
(Figure 2
).
|
Changes in Triglycerides and LDL
Subfractions
An inverse correlation between triglycerides and dLDL
was described
previously.5 6 30
We therefore examined whether the decrease in dLDL was associated with
changes in triglycerides. The overall correlation between change in
triglycerides and change in LDL-6 apoB was weak
(r=-0.101). Only in those
patients showing triglyceride reductions during fluvastatin therapy
(n=23; mean reduction 27%) did changes in LDL-6 apoB correlate
significantly with changes in tri-glycerides
(r=0.655,
P=0.001). The remaining
patients (n=12) showed a mean increase in triglycerides of 50%, but
experienced a substantial decrease in LDL-6 apoB as well (average of
-19.6 mg/dL, compared with -5.7 mg/dL in those patients showing a
decrease in triglycerides).
Safety and Tolerability
Fluvastatin was well tolerated, and there were no
safety concerns. Mean serum levels of total bilirubin, transaminases,
alkaline phosphatase,
-glutamyl transferase, and creatine
phosphokinase did not change.
| Discussion |
|---|
|
|
|---|
Our major finding was that fluvastatin decreased dLDL by
40%, an effect that was almost twice the change in total LDL-C
(-23%). Buoyant LDL subfractions did not change, whereas IDL (which
includes atherogenic remnants of triglyceride-rich lipoproteins)
decreased by
30%. In designing the study, we expected that
substantial reductions in dLDL would be seen in individuals exhibiting
high concentrations of dLDL at baseline. Hence, we selectively
recruited women with apoB in dLDL >25 mg/dL, which approximately
corresponds to the median in postmenopausal women (unpublished
observations). This assumption was confirmed in the present
study.
Our findings contrast with the general literature pertaining
to the effect of HMGRI on LDL subfractions. There are subtle
differences, however, between the present and previous studies. For
example, most studies derived LDL subfraction distribution from
measurements of LDL peak particle size by
GGE.16 17 19 20 24 25
Similarly, we were unable to detect a significant increase in the mean
size of whole LDL when particle radius was calculated according to a
previously published
algorithm.27 This is
surprising, given the substantial reduction in dLDL, but consistent
with theoretical considerations. Using the concentrations and particle
radii of all LDL subfractions, we calculated the expected radii for
whole LDL at baseline and at week 12 as weighted means of the
individual LDL subfractions. This yielded particle radii of 9.8 and 9.9
nm, respectively, at these time points, values close to those
determined empirically (9.7 and 9.9 nm, respectively;
Table 3
). Provided that there are no substantial changes in
the sizes of individual subfractions, it cannot therefore be expected
that the changes of LDL-5 and LDL-6 in the present study would
translate into more than a small change of the average size of the
whole LDL fraction. Other studies that used ultracentrifugation for LDL
subfractionation did not specifically select patients on the basis of
an atherogenic lipoprotein
profile.18 22 In
this context, it is noteworthy that 3 GGE-based studies that included
such patients were small (
12 patients) and may therefore have been
underpowered.16 19 25
The difference in findings between the present investigation and other
ultracentrifugation-based
studies15 18 21 22 23 36
may have methodical reasons. If ultracentrifugation is not performed
until complete equilibrium is reached (as in the present study),
separation may be driven not only by density but also by particle size.
Other explanations may relate to the study design. For example, most
studies that used ultracentrifugation failed to specifically select for
individuals with elevated dLDL at
baseline,15 18 21 22 23 36
and some lacked placebo
groups.15 18 22 23 36
The mechanisms by which fluvastatin selectively decreased the most atherogenic lipoprotein fractions, IDL and dLDL, remain elusive. One potential mechanism relates to the effect of HMGRIs on the release of triglyceride-rich particles in the liver, because an inverse correlation exists between dLDL and triglycerides.5 6 30 Surprisingly, however, there was no consistent relationship between changes in triglycerides and dLDL for the overall patient population, ruling out the possibility that alterations in VLDL metabolism (eg, reduced secretion of VLDL or high lipoprotein lipase activity) fully accounted for the changes in dLDL. The latter finding contrasts with those of Tilly-Kiesi,15 who reported decreases in dLDL on lovastatin only in those individuals who responded with decreases in triglycerides.
Treatment with fluvastatin stimulates the expression of hepatic LDL receptors,37 which also catabolize IDL.38 The most obvious explanation for the decrease in IDL in the present study is therefore that these lipoproteins are taken up by LDL receptors at an enhanced rate during fluvastatin therapy. The change in dLDL is less easily explained, because these particles are poor ligands of LDL receptors,2 and other, less evident, mechanisms might be active. One possible mechanism, an altered rate of transfer of cholesteryl esters from HDL to apoB-containing lipoproteins,39 was ruled out by the finding that activity of cholesteryl ester transfer protein was unchanged during fluvastatin therapy (not shown). Another potential mechanism focuses on hepatic lipase, which has been implicated in the generation of dLDL.40 41 42 In individuals showing high concentrations of triglyceride-rich lipoproteins, transfer of triglycerides (in exchange with cholesteryl esters) into LDL and HDL may occur; these triglycerides may then be hydrolyzed by hepatic lipase, leading to the formation of smaller, lipid-depleted particles.39 In the present study, significant correlations were found between the decreases in LDL-5 and LDL-6 and an increase of HDL-2b, the most buoyant HDL subfraction (not shown). Because buoyant HDL particles are the preferred substrate of hepatic lipase, this would accord with the suggestion of decreased hepatic lipase during fluvastatin therapy. This is also supported by Hoogerbrugge and Jansen,43 who found that atorvastatin lowers hepatic lipase.
The possibility that our results were confounded by
lifestyle changes and/or concomitant cardiovascular therapy (eg,
antihypertensive drugs) warrants discussion. For example, other authors
have shown that ß-adrenergic receptor antagonists raise both
IDL44 and
dLDL.8 In total, 7 patients
(fluvastatin, n=5; placebo, n=2) received ß-adrenergic receptor
antagonists, and no patient received
-adrenergic receptorblocking
agents. Thus, the proportion of patients receiving ß-adrenergic
receptor antagonists was small and comparable for each treatment group,
and no dosage adjustments were made during the study. Exclusion of
these patients from the statistical analyses had no effect on the
statistical significance of the findings, indicating that concomitant
ß-blocker therapy did not affect our results to a relevant extent.
Exclusion of diabetic individuals also did not affect the statistical
significance of our results
(P<0.001 for the change in
dLDL for all patients, P<0.001
for the nondiabetic patients). A similar conclusion can be drawn with
regard to the potential confounding effect of lifestyle changes. Thus,
although patients were not requested to formally record their dietary
habits and degree of exercise during the study (rather, patients were
advised to adhere to a lipid-modified diet and not change usual
exercise habits), there is no reason to assume that minor changes
occurred more frequently in the fluvastatin group than in those treated
with placebo. Moreover, there was no significant difference between the
changes in body mass index for either treatment group. ANOVA using the
change of dLDL as the dependent variable, treatment as the independent
variable, and the variation of body mass index as a covariable says
that changes in LDL subfractions remain significant after adjustment
for body mass index (data not shown).
In conclusion, fluvastatin produces a shift in LDL subfractions toward more buoyant, less atherogenic LDL particles in patients at increased risk of CHD. This profile may contribute to an antiatherogenic effect for fluvastatin that is greater than expected on the basis of changes in lipids and apolipoproteins.
| Acknowledgments |
|---|
| Footnotes |
|---|
Received September 6, 2000; revision received January 16, 2001; accepted January 22, 2001.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
M. Rizzo and K. Berneis Low-density lipoprotein size and cardiovascular risk assessment QJM, January 1, 2006; 99(1): 1 - 14. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. M Backes and C. A Gibson Effect of Lipid-Lowering Drug Therapy on Small-Dense Low-Density Lipoprotein Ann. Pharmacother., March 1, 2005; 39(3): 523 - 526. [Abstract] [Full Text] [PDF] |
||||
![]() |
F. M. Sacks and H. Campos Low-Density Lipoprotein Size and Cardiovascular Disease: A Reappraisal J. Clin. Endocrinol. Metab., October 1, 2003; 88(10): 4525 - 4532. [Full Text] [PDF] |
||||
![]() |
K. Winkler, T. Konrad, S. Fullert, I. Friedrich, R. Destani, M. W. Baumstark, K. Krebs, H. Wieland, and W. Marz Pioglitazone Reduces Atherogenic Dense LDL Particles in Nondiabetic Patients With Arterial Hypertension: A double-blind, placebo-controlled study Diabetes Care, September 1, 2003; 26(9): 2588 - 2594. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. A. Kreisberg and A. Oberman Medical Management of Hyperlipidemia/Dyslipidemia J. Clin. Endocrinol. Metab., June 1, 2003; 88(6): 2445 - 2461. [Full Text] [PDF] |
||||
![]() |
K. Winkler, C. Abletshauser, M. M. Hoffmann, I. Friedrich, M. W. Baumstark, H. Wieland, and W. Marz Effect of Fluvastatin Slow-Release on Low Density Lipoprotein (LDL) Subfractions in Patients with Type 2 Diabetes Mellitus: Baseline LDL Profile Determines Specific Mode of Action J. Clin. Endocrinol. Metab., December 1, 2002; 87(12): 5485 - 5490. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. M. Krauss, H. Campos, and F. M. Sacks Is the Size of Low-Density Lipoprotein Particles Related to the Risk of Coronary Heart Disease? JAMA, February 13, 2002; 287(6): 712 - 713. [Full Text] [PDF] |
||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Circulation Home | Subscriptions | Archives | Feedback | Authors | Help | AHA Journals Home | Search Copyright © 2001 American Heart Association, Inc. All rights reserved. Unauthorized use prohibited. |