- Email: [email protected]
Effects of fluvastatin (XU 62-320), an HMG-CoA reductase inhibitor, on the distribution and composition of low density lipoprotein subspecies in humans
Arherosclerosis, 87 (1991) 147-157 61 1991 Elsevier Scientific Publishers ADONIS 002191509100093V
ATHERO
147 Ireland,
Ltd. 0021-9150/91/$03.50
04613
Effects of fluvastatin (XU 62-320), an HMG-CoA reductase inhibitor, on the distribution and composition of low density lipoprotein subspecies in humans Jiana Yuan, Michael Y. Tsai, Janet Hegland and Donald B. Hunninghake Depuriment of Laboratory Medicine and Heart Disease Prevention Clinic, University of Minnesota Medical School, Minneapolis. MN (U.S.A.) (Received 18 July, 1990) (Revised, received 22 October, 1990) (Accepted 30 November, 1990)
Summary
We studied the effect of fluvastatin (XU 62-320) a new HMG-CoA reductase inhibitor, on the distribution of low density lipoprotein (LDL) subspecies and composition in humans. As expected, fluvastatin significantly lowered serum LDL levels (25% after 6 weeks of therapy). In addition, treatment with fluvastatin changed the LDL subspecies. In the group treated with fluvastatin, 38.5% of the individuals showed changes in the shape of LDL absorbance profile obtained from density gradient ultracentrifugation and 54% of the group showed changes in the electrophoretic mobility of the LDL bands. Of those showing changes in electrophoretic mobility, the majority (78%) shifted to slightly larger, less dense LDL after drug therapy. However, the LDL-cholesterol/ape B ratio changes were relatively small in all fluvastatin-treated individuals including the group with changes in eletrophoretic mobility, confirming that HMG-CoA reductase inhibitor causes relatively small and subtle changes in the distribution of LDL subspecies.
Key words: Fluvastatin; HMG-CoA Electrophoresis
reductase
Correspondence to: Dr. Michael Y. Tsai, Box 198 UMHC, Laboratory Medicine and Pathology, University of Minnesota Medical School, 420 Delaware Street, S.E. Minneapolis, MN 55455, U.S.A.
inhibitor;
LDL
subspecies;
Ultracentrifugation;
Introduction
The levels of low density lipoprotein-cholesterol (LDL-C) are now well recognized to be causally
148 related to the development of coronary heart disease [l-3]. The recent availability of 3-hydroxy-3methylglutaryl CoA (HMG-CoA) reductase inhibitors offers an effective means of reducing the levels of serum LDL in patients with hypercholesterolemia [4,5]. HMG-CoA reductase inhibitors are thought to decrease primarily LDLcholesterol by increasing the catabolism of LDL via increasing the number of available LDL receptors [4]. LDL is heterogeneous [6-81, and at least seven subspecies have been identified with density gradient ultracentrifugation and gradient gel electrophoresis [7,X]. The various subclasses of LDL have different compositions, with associated differences in the characteristics of metabolism [9]. These observations suggested to us that it would be important to examine whether the reduction of serum LDL concentration by an HMG-CoA reductase inhibitor was associated with changes in the distribution and composition of LDL subclasses. In the current investigation, we examined the effect of fluvastatin, a new synthetic HMGCoA reductase inhibitor [lo], on the distribution and composition of LDL subclasses in patients with hypercholesterolemia. Patients and methods
Study population Patients were volunteers at the University of Minnesota Heart Disease Prevention Clinic participating in a multicenter, double-blind randomized, placebo-controlled study intended to characterize dose response, safety and efficacy of fluvastatin (Sandoz Corp.). Thirty-six consecutive patients were studied initially. All patients entering the study must have an LDL-cholesterol level > 160 mg/dl (4.1 mmol/l) at each of the first 3 biweekly visits in conjunction with plasma triglyceride levels of < 300 mg/dl (3.4 mmol/l). No patient had secondary hypercholesterolemia, or was taking drugs with known effects on lipid metabolism. All had been on National Cholesterol Education Program Step One Diet for at least 4 weeks before and throughout the study. Patients with a myocardial infarction, coronary artery bypass, or angioplasty during the 6 months preceding the study, and patients with congestive heart failure or unstable angina were excluded.
Study protocol Treatment. Patients were withdrawn from all other lipid-lowering therapies at least 10 weeks and from probucol at least 1 year before the start of the study. All patients received placebo for 6 weeks (weeks -6 to 0), in a single-blind fashion, followed by random assignment to either placebo or fluvastatin groups at total daily doses of 5,15,20, 30 or 40 mg per day for 6 weeks. Samples from two patients receiving 5 mg per day were excluded from the final analysis of this study. Patients were seen at intervals of 2 weeks, blood samples were taken for lipid and lipoprotein analysis. In the present study, patients included in the final analysis were simply grouped into placebo and fluvastatin-treated. Eight patients (age 57 k 17 years, 6 men and 2 women) received placebo, 26 patients (age 56 f 13 years, 18 men and 8 women) received fluvastatin at a mean total daily dose of 23.1 mg. Blood samples obtained at 0 and 6 week visit were chosen as baseline and end point of therapy, respectively. Maximum effect of HMG CoA reductase inhibitors on lowering LDL-C was generally observed around 3-4 weeks of treatment. Preliminary results showed that this is also true for fluvastatin (data not shown). Lipid and apolipoprotein determination on plasma samples were done in a central laboratory participating in the CDC-NHLBL lipid standardization program as follows: Cholesterol and triglycerides were determined by enzymatic methods on a Hitachi 705 analyzer. HDL-cholesterol was measured in the supematant after plasma was precipitated with heparin-manganese. LDL-cholesterol was calculated as described by Friedwald et al. [ll]. Apolipoprotein B (apo B) was determined by an enzyme-linked immunoassay using a Behring plate reader. Fractionation of LDL. A single spin density gradient ultracentrifugation procedure modified from the methods of Nilsson et al. [12] and Kelly and Kruski [13] was used to subfractionate the plasma lipoproteins. In brief, plasma was adjusted to a density of 1.300 g/ml by adding solid NaBr. A discontinuous gradient was formed by pipetting 1 ml of this plasma followed by layers of NaBr solutions of different concentrations: 1.6 ml of 1.210 g/ml, 4.2 ml of 1.063 g/ml, 3.7 ml of 1.019
149 g/ml, and 1.2 ml of 1.006 g/ml into a 14 x 89 mm centrifuge tube (Beckman polyallomer centrifuge tube). Samples were centrifuged in a Beckman SW-41 rotor at 36 000 rpm for 24 h at 14O C in a Beckman Model L5-75 ultracentrifuge (Beckman Instrument, Palo Alto, CA 94304). At the end of centrifugation each centrifuge tube containing the separated lipoproteins was pierced at the bottom and a dense solution consisting of 1.40 g/ml NaBr was pumped in at a constant rate of 0.72 ml/mm. The effluent emerging from the top of the gradient tube was continuously monitored at 280 nm, and recorded on a strip chart recorder followed by collection of 0.36-ml fractions with the use of a fraction collector. To determine the densities in the individual fractions, one or more tubes not containing serum was included in each run, and the salt densities of the individual fractions determined with a Bausch and Lomb refractometer. Analysis of LDL. Under the conditions described above, up to 14 fractions were collected within the LDL densities of 1.019-1.063 g/ml. Most of the LDL (> 95%) were contained in up to 9 fractions in the density range of 1.025-1.060 g/ml. In the first 22 individuals, these fractions were pooled and total cholesterol and apo B were determined. While the study was in progress, we decided also to study the effect of fluvastatin treatment on the composition of the subfractions of LDL. Thus in the last 12 individuals, some of the 9 fractions in the density range of 1.025-1.060 g/ml were pooled to form a total of 6 fractions with the following density ranges: 1.025-1.032, 1.033-1.038, 1.039-1.045, 1.046-1.049, 1.0501.054 and 1.055-1.060 g/ml. Eleven of the 12 turned out to be fluvastatin-treated and one placebo-treated. Total cholesterol, free cholesterol, triglyceride and apo B were determined in these 6 fractions. Cholesterol and triglyceride were determined by enzymatic methods using reagents from Boehringer Mannheim (for cholesterol) and Abbott Laboratory (for triglyceride) on an ABA-100 analyzer (Abbott Laboratories, Irving, TK 75015). Free cholesterol and cholesterol ester were determined by enzymatic determination of cholesterol in the presence and absence of cholesterol esterase and mass of cholesterol ester calculated as
1.67 x mg of (total was determined by a Beckman Array reagents (Beckman
- free cholesterol). LDL-apo B a nephelometric method using Protein System and Beckman Instrument, Brea, CA 92621). Gradient gel electrophoresis of LDL. Gradient gel electrophoresis was used to separate LDL subfractions by a modified technique of McNamara et al. [7]. We used non-denaturing polyacrylamide gels of 3-16% forming a gel of 1.5 x 90 x 140 mm in size containing 10 lanes. Electrophoresis was carried out with an HSI electrophoresis unit (Model SE 500) connected to a cooling unit (Lauda). The apparatus was filled with fresh buffer
TABLE
7
EFFECT OF FLUVASTATIN ON SERUM LIPIDS LIPOPROTEINS IN HYPERCHOLESTEROLEMIA Values represent the mean f SD. Numbers cate values in range. Baseline (mg/dl)
in parentheses
AND
indi-
Week 6 percent
mg/dI
change
(Range) Cholesterol Placebo (n=8) XU 62-320 (n=26) LDL-Cholesterol Placebo (n=8) XU 62-320 (n = 26) Apo B Placebo (n = 8) XU 62-320 (n=26) Triglyceride Placebo (n=8) XU 62-320 (n=26) HDL-cholesterol Placebo (n=8) XU 62-320 (n = 26) * ** ” ’
P P P P
272 + 28 (246-335) 269 + 66 (238-372)
274*31 (226-308) 226 f 34.4 * * (180-290)
+0.8*9.1 (-11 to+15) -19*6.2++ (-33 to -3)
197+23 (171-248) 206 zt 35 (166-307)
196&24 (162-235) 155+31** (113-240)
-1.1*8.8 (-14to +12) - 24.7 f 8.9 ++ (-41 to +l)
162 f 22 (146-212) 165 * 39 (103-238)
159.5 f 23 (136-209) 127+31 ** (96-172)
-1.4f6.7 (-13 to +7) -22.2*14+ ( - 48 to - 1)
154*54 (92-245) 134+43 (61-247)
146 f 94 (72-333) 110+37 * (27-210)
- 9.8 f 29.7 (-43 to +36) - 12.7 + 34.9 (-52 to +72)
43*9 (31-55) 47+11 (34-75)
47*13 (33-57) 50+13 (35-94)
+9.5*13 (-3 to +32) *5.7*9.2 (-13 to +25)
*
c 0.05, paired t-test, week 6 versus baseline i 0.001, paired t-test, week 6 versus baseline -c 0.0001 XU 62-320 versus placebo c 0.0003, XU 62-320 versus placebo
150 TABLE 2 EFFECT OF FLUVASTATIN AND PLACEBO ON DENSITY, ABSORBANCE AND ELECTROPHORETIC CHOLESTEROL/APO B RATIO
MOBILITY, LDL
Numbers indicate number of patients. Subjects
Main peak density ’ changes denser d
Placebo (n =8)
Fluvastatin (n=26)
1 (3.8%)
LDL absorbance profile b
Main band ’ Change of electrophoretic mobility
change f
no change
increase
less dense e
no change
(G.58)
I (87.5%)
(1:.5%)
(8:.5%)
(lZ.48)
21 (80.8%)
10 b (38.5%)
$F.5%)
decrease 1 (12.5%) 1.82 g 1.91
I (1:g) 1.86 f 1.10 1.7kO.06
&%) 1.72kO.20 1.79kO.22
no change 7
(87.5%) 1.84+0.33 1.78 + 0.23 12 (46%) 1.75 +0.18 1.75 f 0.23
a Main peak is the peak with the highest absorbance of LDL. b LDL absorbance profile-plasma was subjected to ultracentrifugation for 24 h and effluent from centrifuge tube containing LDL monitored at 280 nm. ’ Main band is the most prominent and densest band on gradient gel electrophoresis. d Density increase 2 0.003 mg/mL as compared to baseline. e Density decrease 2 0.003 mg/mL as compared to baseline. f Change in shape of LDL profile other than height. g LDL-cholesterol/ape B ratios. Top before therapy. Lower after therapy. h 8 of the 10 had polydisperse, 2 of the 10 monodisperse LDL before treatment. All 10 showed polydisperse LDL after treatment. i 6 had polydisperse and 10 monodisperse LDL.
(90 mM Tris, 80 mM boric acid, 2.5 mM EDTA, pH 8.4) precooled at 4O C. Gels were equilibrated for 20 min at 120 V at 8” C, followed by the addition of 12-16 ml of plasma/40% sucrose solution (3 : 1, v/v) to each lane. After pre-electrophoresis of sample at 70 V for 20 min the runs were carried out at a constant voltage of 325 V for 9 h. Gels were stained with Sudan black B for 20 h, followed by destaining in a 50% Cellosolve for 22-28 h. After destaining, gels were stored in 25% Cellosolve and photographed after 2-3 days. We identified LDL bands as described by McNamara [8], such that the main band (defined as the most prominent and darkest band) in gradient gels was assigned a band number according to the peak density of LDL in ultracentrifugation. Thus LDL-1 migrated in the density of 1.019-1.033, LDL-2 and LDL-3 in the range of 1.033-1.038, LDL-4 and LDL-5, 1.038-1.050, and LDL-6 and LDL-7, 1.050-1.063 g/ml. A pooled plasma from 2 individuals, one with LDL-3 and LDL-4, another with
LDL4 and LDL-6, were then used as reference standard. This standard was spotted onto 2 lanes in each gel. A sample was designated to have polydisperse LDL if it either showed two or more distinct bands on electrophoresis or a marked asymmetry or the presence of more than one peak in the absorbance profile obtained from density gradient ultracentrifugation. One of us (J.Y.) interpreted the absorbance profiles and electrophoretograms of participants before and after treatment, and was blinded with respect to the treatment individuals received. Statistical analysis. The various parameters measured were entered into a Macintosh personal computer. Statistical analyses were performed with the Statworks software. Paired I-test was used to assess the significance of differences in lipid parameters before and after treatment within the drug and placebo treated groups (Tables 1, 3). Student’s C-test was used to assess the significance of the differences between the drug and placebo-
151 treated groups (Table 1). Chi-square was used to determine the significance of change in LDL subspecies after drug therapy (Table 2).
Results Table 1 shows the effect of fluvastatin on plasma lipids and lipoproteins. In the group treated with fluvastatin there was a significant reduction after therapy in total cholesterol (19%) LDL-cholesterol (25’%), and apo B (22%), a small but significant reduction of triglyceride and a small, but significant increase in HDL-cholesterol compared to baseline value. Compared to the placebo group, however, only the reductions in total cholesterol, LDL-cholesterol and apo B in the fluvastatin-treated group were statistically significant. Since this study was designed to look specifically at the changes in LDL subspecies caused by fluvastatin and not to discriminate effects of the different doses used, all patients on fluvastatin were grouped together for analysis. Table 2 summarizes the results of changes in the main peak density and shape of LDL absorbance profile obtained from density gradient ultracentrifugation of plasma, and mobility of the main LDL band upon gradient gel electrophoresis of samples obtained before and after drug or placebo treatment. One of the eight patients in the placebo-treated group and 5 of the 26 patients in the fluvastatin treated group showed changes in the density of main LDL peak after density gradient centrifugation. All individuals treated with fluvastatin showed a reduction in LDL peak height. Ten of the 26 in the fluvastatin-treated group showed a disproportionate reduction of different LDL subspecies, as evidenced by a change in the shape of the LDL profiles (Table 2). Of the 10 individuals, 8 had polydisperse and 2 had monodisperse LDL profiles before treatment. All 10 showed polydisperse LDL after treatment. Of the 16 individuals treated with fluvastatin who did not show apparent change in the shape of their LDL profiles (other than height), 10 had monodisperse and 6 had polydisperse LDL (see Table 2, footnote h, i). All samples that showed a change in the shape
of LDL profiles after treatment also showed expected changes in the mobility of the main LDL band upon gradient gel electrophoresis (Table 2). Four of the 16 samples with no change in shape of LDL profile also showed changes of LDL bands upon electrophoresis. Thus, 14 of the 26 individuals in the fluvastatin-treated group and 1 of 8 in the placebo-treated group showed changes in the mobility of their main LDL bands after gradient gel electrophoresis. (Xi-square analysis, however, showed that the number of individuals with changed electrophoretic mobility after fluvastatin versus placebo treatment did not reach statistical (P < 0.09). significance Of the 14 in fluvastatin-treated group, 3 showed increased mobility of LDL bands (decreased molecular weight), and 11 of the drug-treated group and 1 in the placebo group showed decreased mobility of LDL bands (increased molecular weight). The total cholesterol/ape B ratios of LDL showed small and statistically insignificant changes. Nevertheless, of the two groups that showed changes in LDL mobility by electrophoresis, the changes were in the expected direction for most samples (2 of the 3 showed decreased cholesterol/ape B ratio in the group with increased electrophoretic mobility and 8 of 11 showed increased cholesterol/ape B ratio in the group with decreased electrophoretic mobility). Examples of how fluvastatin treatment affect plasma LDL absorbance profiles are shown in Fig. 1. Profiles were chosen from 5 fluvastatin and 1 placebo-treated individuals before and after 6 weeks treatment to illustrate changes in absorbance profiles. Peak heights of LDL in all 5 fluvastatin-treated individuals were significantly reduced after therapy. Four of the 5 samples (Nos. 1, 2, 3, 4) represented those with disproportionate reduction of the LDL subspecies, while one (No. 5) of the 5 showed proportionate reduction. Disproportionate reduction of the subspecies can result in different shape changes. In sample 1, an extra shoulder with density lower than the main peak (1.042 g/ml) appeared after therapy. In sample 2, a shoulder with density lower than the main peak (1.044 g/ml) present before therapy disappeared after drug therapy. In samples 3 and 4, the peaks of the LDL profile after drug therapy became broader, most likely due to a preferential
DenSity
1
iu
iu Ab8orbmoe
Abrorbmce
G
ii
DensHy
,
Abrorbmce
w
P
I
N
P
AbSOrb6nCe
cn
Den6lty
Denrlty
AbsOrb6nCe
AbSOrb6llCe
153 +
Fig. 1. LDL absorbance profiles obtained from density gradiant ultracentrifugation from six individuals. (A) before treatment, (B) after treatment: l-5, fluvastatin-treated. Disproportionate reduction of LDL subspecies in individuals 1-4 resulted in shape change, while the shape (other than height) of the LDL absorbance profile in individual 5 remained the same after fluvastatin treatment, implying proportional reduction of all LDL subspecies. 6, an individual treated with placebo showing no change of either height or shape.
1
2
3
4 5
6
7 8
9 10
Fig. 2. Gradient gel electrophoresis of plasma from individuals labelled Nos. 1-4 in Fig. 1. Lanes 5 and 10 represent pooled plasma containing LDL bands 3, 4, and 6. Plasma from individuals Nos. l-4 before drug therapy were spotted on Lanes 1, 3, 6, 8 and after 6 weeks of therapy on Lanes 2, 4, 7, 9 respectively.
reduction of the more dense LDL sent in the main peak. Fig. 2 illustrates that changes profiles were always accompanied electrophoretic patterns. Thus the patterns of samples from individuals
originally
pre-
in absorbance by changes in electrophoretic Nos. l-4 as
illustrated in Fig. 1 are shown in Fig. 2. In sample 1, LDL-6, the only band before treatment (lane 1) was replaced by a mixture of LDL-5 and LDL-6 (lane 2). Sample 2 was composed of LDL-4 and LDL-5, with LDL4 as the predominant band before treatment (lane 3). After treatment, the relatively larger, less dense LDL-4 disappeared and was replaced by a mixture of LDL-5 and LDL-6 (lane 4). In sample 3 and 4 pretreatment samples were characterized by a main band of LDL-6 with lesser quantities of LDL-5 (lane 6,8). Post-treatment pattern showed mixtures of LDL-5 and LDL-6 with LDL-5 being the darker of the two bands (lane 7,9). Thus the changes in electrophoretic patterns were always in the same direction as those observed in LDL absorbance profiles. In order to determine whether fluvastatin therapy affects the chemical composition of LDL subspecies, we determined cholesterol and apo B contents in 6 fractions in the density range of 1.0251.060 g/ml in the last 12 individuals participating
Fig. 3. Effect of fluvastatin on percent of change of apo B (panel A) and cholesterol (panel B) in six LDL fractions (I-VI) of the following density ranges. I, 1.025-1.032; II, 1:033-1.038; III, 1.039-1.045; IV, 1.046-1.049; V, 1.050-1.054; VI, 1.055-1.060 g/ml. The results were from 5 individuals who did not show a change in the shape of LDL absorbance profile or mobility of the main LDL individual No. 5 depicted in Fig. 1. band upon electrophoresis after 6 weeks of fluvastatin therapy. 0 -0,
B
Fig. 4. Effect of fluvastatin on percent change of Fig. 3. The results were from 6 individuals that LDL band upon electrophoresis after 6 weeks ?? -m,
apo B (panel A) and cholesterol (panel B) in 6 LDL fractions (I-VI) as described in showed changes in the shape of LDL absorbance profile and mobility of the main No. 1; of drug therapy. Four of these were also depicted in Fig. 1 (0 -0, No. 2; x x,No.3;rA, No. 4).
in this study. Eleven of the 12 were treated with fluvastatin in this group. Fig. 3 shows the percentage change of apo B and cholesterol in the 6 LDL fractions of 5 individuals in whom drug treatment reduced the peak height of LDL profiles not accompanied by other shape change. Similar results from 6 individuals in whom drug therapy probably caused preferential removal of certain subspecies of LDL, resulting in a change in the shape of LDL profile after therapy are shown in Fig. 4. In Fig. 3, the percentage of cholesterol or apo B reduction in the 6 fractions is similar in each individual, confirming the lack of significant preferential removal of any subspecies. In Fig. 4, percentage of reduction of either cholesterol or apo B in the individual fractions varied from one to the other, confirming the preferential removal of selected subspecies resulting in a change of the shape of the LDL profile after drug therapy. While fluvastatin seemed to cause preferential removal of LDL subspecies in a portion of the
treated individuals, it did not change the composition of the individual LDL subspecies measured by the cholesterol/ape B and cholesteryl ester/ cholesterol ratios. Table 3 showed that the cholesterol/ apo B and cholesteryl ester/ cholesterol ratios in the individual fractions were unchanged before and after therapy in both the groups with proportionate and disproportionate reduction of LDL subspecies. Discussion
In the present report, we documented changes of LDL subspecies in patients with hypercholesterolemia after they had been treated with an HMG-CoA reductase inhibitor for 6 weeks. Previous studies showed that LDL exists as seven subspecies by density gradient ultracentrifugation and gradient gel electrophoresis [7-91. In this study, LDL heterogeneity was studied by density gradient ultracentrifugation and gradient gel electro-
155
156 phoresis before and after placebo or drug therapy. After treatment with fluvastatin, plasma LDL in 42% of the patients became larger and less dense, while in 12% of the patients changed to smaller, more dense LDL as judged by the mobility of the main LDL band on electrophoresis. We know of no previous published studies comparing LDL subspecies in sera of individuals before and after cholesterol-lowering drug therapies documented by electrophoretic and density gradient ultracentrifugation techniques. Witztum et al. [14-171 studied the effects of bile acid sequestrant therapies on LDL composition. These investigators showed that colestipol and cholestyramine therapies changed LDL composition in men and guinea pigs. In men, colestipol was shown to reduce significantly LDLcholesterol and apo B [15]. However, the reduction in apo B level was less than LDL-C, resulting in a decrease in the LDL-C/ape B ratio post therapy [15]. From these results and electron microscopic measurement of LDL particle sizes, the authors suggested that colestipol therapy caused preferential removal of larger, less dense LDL, resulting in protein-enriched, smaller, more dense LDL post-therapy [15,17]. Similar results had been observed in LDL of guinea pigs treated with cholestyramine [16]. The results of the current study showed that fluvastatin, an HMG-CoA reductase inhibitor, reduced LDL-C without significantly changing the total cholesterol/ape B ratio in LDL. Recently, Bergland et al. [18] showed lovastatin slightly affected the composition of LDL in guinea pigs. Our results, obtained in humans using a different HMG CoA reductase inhibitor, basically agree with theirs. While the compositions of LDL seemed little changed, small changes in LDL can be detected by the sensitive method of gradient gel electrophoresis in over 50% of the treated patients. Lesser percentages (38.5%) showed changes in the shape (other than height) of the LDL profile, obtained from density gradient ultracentrifugation. But as illustrated in Figs. 1 and 2, changes in the LDL profile from density gradient ultracentrifugation were always accompanied by electrophoretic mobility of LDL-bands in the same direction. Relatively few samples (19.2%) showed a significant change in main peak density, confirming that changes in LDL after HMG-CoA
reductase inhibitor therapy consisted mainly of small, subtle changes. In summary, fluvastatin, in lowering serum LDL concentration, also changed the LDL subspecies in more than 50% of the patients. However, in the present study with the relatively small placebo-control group, the increased change of LDL subspecies in the fluvastatin-treated group as compared to the placebo group did not reach statistical significance (P < 0.09). While the fluvastatin-treated group clearly showed a trend of increased change of LDL subspecies than the placebo-controlled group and this comparison will most likely become statistically significant in a larger group, the finding also underscored the relatively small changes in LDL subspecies and the lack of LDL subspecies change in almost 50% of the individuals treated with HMG-CoA reductase inhibitor. Unlike bile acid sequestrant therapies, HMG-CoA reductase inhibitor, therapy did not result in the preferential removal of larger, less dense LDL particles. Of those patients that showed changes in LDL subspecies after drug therapy, LDL in more individuals (78%) became larger and less dense. However, the tendency to shift towards less dense LDL is not statistically significant. Further studies of the effects of other therapies or combined therapies on LDL subspecies should facilitate our understanding at the molecular level. References 1 Consensus Conference. Lowering blood cholesterol to prevent heart disease. JAMA, 253 (1985) 2080. 2 Lipid Research Clinics, The Lipid Research Clinics Coronary Primary Prevention Trials Results. JAMA, 251 (1984) 351. 3 Reports of the National Cholesterol Education Program Expert Panel on detection, evaluation and treatment of high blood cholesterol in adults. Arch Intern. Med., 148 (1988) 36. 4 Grundy, S.M., HMG-CoA reductase inhibitors for treatment of hypercholesterolemia. N. Engl. J. Med., 319 (1988) 24. 5 Hunninghake, D.B., Clinical trials of lovastatin and simvastatin versus cholestyramine. Atheroscler. Rev., 18 (1988) 133. 6 Lindgren, F.T., Elliot, A. and Gofman, J.W., The ultracentrifugal characterization and isolation of human blood lipids and lipoproteins and applications to the study of atherosclerosis. J. Phys. Colloid Chem., 55 (1951) 80.
157 Krauss. R.M. and Burke, D.J., Identification of multiple subclasses of plasma low density lipoproteins in normal human. J. Lipid Res., 23 (1982) 97. McNamara, J.R., Campos, H., Ordovas, J.M., et al., Effect of gender, age, and lipid status on low density lipoprotein subfraction distribution. Arteriosclerosis, 7 (1987) 483. Teng, B., Sniderman, A.D., Soutar, A.K. and Thompson, G.R., Metabolic basis of hyperapobeta-lipoproteinemia: turnover of apolipoprotein B in low density lipoprotein and its precursors and subfractions compared with normal and familial hypercholesterolemia. J. Clin. Invest., 77 (1986) 663. 10 Mantell, G., Lipid lowering drugs in atherosclerosis-the HMG-CoA reductase inhibitors. Clin. Exp. Hypertens. Theory Pratt., All (1989) 927. W.T., Levy, RI. and Frederickson, D.S., Esti11 Friedwald, mation of the concentration of low-density lipoprotein cholesterol in plasma without use of the preparative ultracentrifuge. Clin. Chem., 18 (1972) 499. V., Edelstein, C. and Scanu, 12 Nilsson, J., Mannickarouttu, A.M., An improved detection system applied to the study of serum lipoproteins after single-step density gradient ultracentrifugation. Anal. Biochem., 110 (1981) 342.
13 Kelley, J.L. and Kruski, A.W., Density gradient ultracentrifugation of serum lipoproteins in a swinging bucket rotor. Atherosclerosis, 24 (1976) 170. 14 Witztum, J.L., Schonfeld, G. and Wiedman, S.W.. The effects of colestipol on the metabolism of very-low-density lipoprotein in man. J. Lab. Clin. Med., 88 (1976) 1008. 15 Witztum, J.L.. Schonfeld, G., Weidman. SW. et al.. Bile sequestrant therapy alters the composition of low-density and high-density lipoproteins. Metabolism, 28 (1979) 221. 16 Witztum, J.L., Young, S.G., Elam. R.L. et al., Cholestyramine-induced changes in low density lipoprotein composition and metabolism. I. Studies in the guinea pig J. Lipid Res., 26 (1985) 92. 17 Young, SC., Witztum, J.L., Carew, T.E. et al., Colestipolinduced changes in LDL composition and metabolism. II. Studies in humans. J. Lipid Res., 30 (1989) 225. 18 Berglund. L., Sharkey, M.F., Elam, R.L. and Witztum, J.L.. Effects of lovastatin therapy on guinea pig low density lipoprotein composition and metabolism. J. Lipid Res., 30 (1989) 1591.
ATHERO
147 Ireland,
Ltd. 0021-9150/91/$03.50
04613
Effects of fluvastatin (XU 62-320), an HMG-CoA reductase inhibitor, on the distribution and composition of low density lipoprotein subspecies in humans Jiana Yuan, Michael Y. Tsai, Janet Hegland and Donald B. Hunninghake Depuriment of Laboratory Medicine and Heart Disease Prevention Clinic, University of Minnesota Medical School, Minneapolis. MN (U.S.A.) (Received 18 July, 1990) (Revised, received 22 October, 1990) (Accepted 30 November, 1990)
Summary
We studied the effect of fluvastatin (XU 62-320) a new HMG-CoA reductase inhibitor, on the distribution of low density lipoprotein (LDL) subspecies and composition in humans. As expected, fluvastatin significantly lowered serum LDL levels (25% after 6 weeks of therapy). In addition, treatment with fluvastatin changed the LDL subspecies. In the group treated with fluvastatin, 38.5% of the individuals showed changes in the shape of LDL absorbance profile obtained from density gradient ultracentrifugation and 54% of the group showed changes in the electrophoretic mobility of the LDL bands. Of those showing changes in electrophoretic mobility, the majority (78%) shifted to slightly larger, less dense LDL after drug therapy. However, the LDL-cholesterol/ape B ratio changes were relatively small in all fluvastatin-treated individuals including the group with changes in eletrophoretic mobility, confirming that HMG-CoA reductase inhibitor causes relatively small and subtle changes in the distribution of LDL subspecies.
Key words: Fluvastatin; HMG-CoA Electrophoresis
reductase
Correspondence to: Dr. Michael Y. Tsai, Box 198 UMHC, Laboratory Medicine and Pathology, University of Minnesota Medical School, 420 Delaware Street, S.E. Minneapolis, MN 55455, U.S.A.
inhibitor;
LDL
subspecies;
Ultracentrifugation;
Introduction
The levels of low density lipoprotein-cholesterol (LDL-C) are now well recognized to be causally
148 related to the development of coronary heart disease [l-3]. The recent availability of 3-hydroxy-3methylglutaryl CoA (HMG-CoA) reductase inhibitors offers an effective means of reducing the levels of serum LDL in patients with hypercholesterolemia [4,5]. HMG-CoA reductase inhibitors are thought to decrease primarily LDLcholesterol by increasing the catabolism of LDL via increasing the number of available LDL receptors [4]. LDL is heterogeneous [6-81, and at least seven subspecies have been identified with density gradient ultracentrifugation and gradient gel electrophoresis [7,X]. The various subclasses of LDL have different compositions, with associated differences in the characteristics of metabolism [9]. These observations suggested to us that it would be important to examine whether the reduction of serum LDL concentration by an HMG-CoA reductase inhibitor was associated with changes in the distribution and composition of LDL subclasses. In the current investigation, we examined the effect of fluvastatin, a new synthetic HMGCoA reductase inhibitor [lo], on the distribution and composition of LDL subclasses in patients with hypercholesterolemia. Patients and methods
Study population Patients were volunteers at the University of Minnesota Heart Disease Prevention Clinic participating in a multicenter, double-blind randomized, placebo-controlled study intended to characterize dose response, safety and efficacy of fluvastatin (Sandoz Corp.). Thirty-six consecutive patients were studied initially. All patients entering the study must have an LDL-cholesterol level > 160 mg/dl (4.1 mmol/l) at each of the first 3 biweekly visits in conjunction with plasma triglyceride levels of < 300 mg/dl (3.4 mmol/l). No patient had secondary hypercholesterolemia, or was taking drugs with known effects on lipid metabolism. All had been on National Cholesterol Education Program Step One Diet for at least 4 weeks before and throughout the study. Patients with a myocardial infarction, coronary artery bypass, or angioplasty during the 6 months preceding the study, and patients with congestive heart failure or unstable angina were excluded.
Study protocol Treatment. Patients were withdrawn from all other lipid-lowering therapies at least 10 weeks and from probucol at least 1 year before the start of the study. All patients received placebo for 6 weeks (weeks -6 to 0), in a single-blind fashion, followed by random assignment to either placebo or fluvastatin groups at total daily doses of 5,15,20, 30 or 40 mg per day for 6 weeks. Samples from two patients receiving 5 mg per day were excluded from the final analysis of this study. Patients were seen at intervals of 2 weeks, blood samples were taken for lipid and lipoprotein analysis. In the present study, patients included in the final analysis were simply grouped into placebo and fluvastatin-treated. Eight patients (age 57 k 17 years, 6 men and 2 women) received placebo, 26 patients (age 56 f 13 years, 18 men and 8 women) received fluvastatin at a mean total daily dose of 23.1 mg. Blood samples obtained at 0 and 6 week visit were chosen as baseline and end point of therapy, respectively. Maximum effect of HMG CoA reductase inhibitors on lowering LDL-C was generally observed around 3-4 weeks of treatment. Preliminary results showed that this is also true for fluvastatin (data not shown). Lipid and apolipoprotein determination on plasma samples were done in a central laboratory participating in the CDC-NHLBL lipid standardization program as follows: Cholesterol and triglycerides were determined by enzymatic methods on a Hitachi 705 analyzer. HDL-cholesterol was measured in the supematant after plasma was precipitated with heparin-manganese. LDL-cholesterol was calculated as described by Friedwald et al. [ll]. Apolipoprotein B (apo B) was determined by an enzyme-linked immunoassay using a Behring plate reader. Fractionation of LDL. A single spin density gradient ultracentrifugation procedure modified from the methods of Nilsson et al. [12] and Kelly and Kruski [13] was used to subfractionate the plasma lipoproteins. In brief, plasma was adjusted to a density of 1.300 g/ml by adding solid NaBr. A discontinuous gradient was formed by pipetting 1 ml of this plasma followed by layers of NaBr solutions of different concentrations: 1.6 ml of 1.210 g/ml, 4.2 ml of 1.063 g/ml, 3.7 ml of 1.019
149 g/ml, and 1.2 ml of 1.006 g/ml into a 14 x 89 mm centrifuge tube (Beckman polyallomer centrifuge tube). Samples were centrifuged in a Beckman SW-41 rotor at 36 000 rpm for 24 h at 14O C in a Beckman Model L5-75 ultracentrifuge (Beckman Instrument, Palo Alto, CA 94304). At the end of centrifugation each centrifuge tube containing the separated lipoproteins was pierced at the bottom and a dense solution consisting of 1.40 g/ml NaBr was pumped in at a constant rate of 0.72 ml/mm. The effluent emerging from the top of the gradient tube was continuously monitored at 280 nm, and recorded on a strip chart recorder followed by collection of 0.36-ml fractions with the use of a fraction collector. To determine the densities in the individual fractions, one or more tubes not containing serum was included in each run, and the salt densities of the individual fractions determined with a Bausch and Lomb refractometer. Analysis of LDL. Under the conditions described above, up to 14 fractions were collected within the LDL densities of 1.019-1.063 g/ml. Most of the LDL (> 95%) were contained in up to 9 fractions in the density range of 1.025-1.060 g/ml. In the first 22 individuals, these fractions were pooled and total cholesterol and apo B were determined. While the study was in progress, we decided also to study the effect of fluvastatin treatment on the composition of the subfractions of LDL. Thus in the last 12 individuals, some of the 9 fractions in the density range of 1.025-1.060 g/ml were pooled to form a total of 6 fractions with the following density ranges: 1.025-1.032, 1.033-1.038, 1.039-1.045, 1.046-1.049, 1.0501.054 and 1.055-1.060 g/ml. Eleven of the 12 turned out to be fluvastatin-treated and one placebo-treated. Total cholesterol, free cholesterol, triglyceride and apo B were determined in these 6 fractions. Cholesterol and triglyceride were determined by enzymatic methods using reagents from Boehringer Mannheim (for cholesterol) and Abbott Laboratory (for triglyceride) on an ABA-100 analyzer (Abbott Laboratories, Irving, TK 75015). Free cholesterol and cholesterol ester were determined by enzymatic determination of cholesterol in the presence and absence of cholesterol esterase and mass of cholesterol ester calculated as
1.67 x mg of (total was determined by a Beckman Array reagents (Beckman
- free cholesterol). LDL-apo B a nephelometric method using Protein System and Beckman Instrument, Brea, CA 92621). Gradient gel electrophoresis of LDL. Gradient gel electrophoresis was used to separate LDL subfractions by a modified technique of McNamara et al. [7]. We used non-denaturing polyacrylamide gels of 3-16% forming a gel of 1.5 x 90 x 140 mm in size containing 10 lanes. Electrophoresis was carried out with an HSI electrophoresis unit (Model SE 500) connected to a cooling unit (Lauda). The apparatus was filled with fresh buffer
TABLE
7
EFFECT OF FLUVASTATIN ON SERUM LIPIDS LIPOPROTEINS IN HYPERCHOLESTEROLEMIA Values represent the mean f SD. Numbers cate values in range. Baseline (mg/dl)
in parentheses
AND
indi-
Week 6 percent
mg/dI
change
(Range) Cholesterol Placebo (n=8) XU 62-320 (n=26) LDL-Cholesterol Placebo (n=8) XU 62-320 (n = 26) Apo B Placebo (n = 8) XU 62-320 (n=26) Triglyceride Placebo (n=8) XU 62-320 (n=26) HDL-cholesterol Placebo (n=8) XU 62-320 (n = 26) * ** ” ’
P P P P
272 + 28 (246-335) 269 + 66 (238-372)
274*31 (226-308) 226 f 34.4 * * (180-290)
+0.8*9.1 (-11 to+15) -19*6.2++ (-33 to -3)
197+23 (171-248) 206 zt 35 (166-307)
196&24 (162-235) 155+31** (113-240)
-1.1*8.8 (-14to +12) - 24.7 f 8.9 ++ (-41 to +l)
162 f 22 (146-212) 165 * 39 (103-238)
159.5 f 23 (136-209) 127+31 ** (96-172)
-1.4f6.7 (-13 to +7) -22.2*14+ ( - 48 to - 1)
154*54 (92-245) 134+43 (61-247)
146 f 94 (72-333) 110+37 * (27-210)
- 9.8 f 29.7 (-43 to +36) - 12.7 + 34.9 (-52 to +72)
43*9 (31-55) 47+11 (34-75)
47*13 (33-57) 50+13 (35-94)
+9.5*13 (-3 to +32) *5.7*9.2 (-13 to +25)
*
c 0.05, paired t-test, week 6 versus baseline i 0.001, paired t-test, week 6 versus baseline -c 0.0001 XU 62-320 versus placebo c 0.0003, XU 62-320 versus placebo
150 TABLE 2 EFFECT OF FLUVASTATIN AND PLACEBO ON DENSITY, ABSORBANCE AND ELECTROPHORETIC CHOLESTEROL/APO B RATIO
MOBILITY, LDL
Numbers indicate number of patients. Subjects
Main peak density ’ changes denser d
Placebo (n =8)
Fluvastatin (n=26)
1 (3.8%)
LDL absorbance profile b
Main band ’ Change of electrophoretic mobility
change f
no change
increase
less dense e
no change
(G.58)
I (87.5%)
(1:.5%)
(8:.5%)
(lZ.48)
21 (80.8%)
10 b (38.5%)
$F.5%)
decrease 1 (12.5%) 1.82 g 1.91
I (1:g) 1.86 f 1.10 1.7kO.06
&%) 1.72kO.20 1.79kO.22
no change 7
(87.5%) 1.84+0.33 1.78 + 0.23 12 (46%) 1.75 +0.18 1.75 f 0.23
a Main peak is the peak with the highest absorbance of LDL. b LDL absorbance profile-plasma was subjected to ultracentrifugation for 24 h and effluent from centrifuge tube containing LDL monitored at 280 nm. ’ Main band is the most prominent and densest band on gradient gel electrophoresis. d Density increase 2 0.003 mg/mL as compared to baseline. e Density decrease 2 0.003 mg/mL as compared to baseline. f Change in shape of LDL profile other than height. g LDL-cholesterol/ape B ratios. Top before therapy. Lower after therapy. h 8 of the 10 had polydisperse, 2 of the 10 monodisperse LDL before treatment. All 10 showed polydisperse LDL after treatment. i 6 had polydisperse and 10 monodisperse LDL.
(90 mM Tris, 80 mM boric acid, 2.5 mM EDTA, pH 8.4) precooled at 4O C. Gels were equilibrated for 20 min at 120 V at 8” C, followed by the addition of 12-16 ml of plasma/40% sucrose solution (3 : 1, v/v) to each lane. After pre-electrophoresis of sample at 70 V for 20 min the runs were carried out at a constant voltage of 325 V for 9 h. Gels were stained with Sudan black B for 20 h, followed by destaining in a 50% Cellosolve for 22-28 h. After destaining, gels were stored in 25% Cellosolve and photographed after 2-3 days. We identified LDL bands as described by McNamara [8], such that the main band (defined as the most prominent and darkest band) in gradient gels was assigned a band number according to the peak density of LDL in ultracentrifugation. Thus LDL-1 migrated in the density of 1.019-1.033, LDL-2 and LDL-3 in the range of 1.033-1.038, LDL-4 and LDL-5, 1.038-1.050, and LDL-6 and LDL-7, 1.050-1.063 g/ml. A pooled plasma from 2 individuals, one with LDL-3 and LDL-4, another with
LDL4 and LDL-6, were then used as reference standard. This standard was spotted onto 2 lanes in each gel. A sample was designated to have polydisperse LDL if it either showed two or more distinct bands on electrophoresis or a marked asymmetry or the presence of more than one peak in the absorbance profile obtained from density gradient ultracentrifugation. One of us (J.Y.) interpreted the absorbance profiles and electrophoretograms of participants before and after treatment, and was blinded with respect to the treatment individuals received. Statistical analysis. The various parameters measured were entered into a Macintosh personal computer. Statistical analyses were performed with the Statworks software. Paired I-test was used to assess the significance of differences in lipid parameters before and after treatment within the drug and placebo treated groups (Tables 1, 3). Student’s C-test was used to assess the significance of the differences between the drug and placebo-
151 treated groups (Table 1). Chi-square was used to determine the significance of change in LDL subspecies after drug therapy (Table 2).
Results Table 1 shows the effect of fluvastatin on plasma lipids and lipoproteins. In the group treated with fluvastatin there was a significant reduction after therapy in total cholesterol (19%) LDL-cholesterol (25’%), and apo B (22%), a small but significant reduction of triglyceride and a small, but significant increase in HDL-cholesterol compared to baseline value. Compared to the placebo group, however, only the reductions in total cholesterol, LDL-cholesterol and apo B in the fluvastatin-treated group were statistically significant. Since this study was designed to look specifically at the changes in LDL subspecies caused by fluvastatin and not to discriminate effects of the different doses used, all patients on fluvastatin were grouped together for analysis. Table 2 summarizes the results of changes in the main peak density and shape of LDL absorbance profile obtained from density gradient ultracentrifugation of plasma, and mobility of the main LDL band upon gradient gel electrophoresis of samples obtained before and after drug or placebo treatment. One of the eight patients in the placebo-treated group and 5 of the 26 patients in the fluvastatin treated group showed changes in the density of main LDL peak after density gradient centrifugation. All individuals treated with fluvastatin showed a reduction in LDL peak height. Ten of the 26 in the fluvastatin-treated group showed a disproportionate reduction of different LDL subspecies, as evidenced by a change in the shape of the LDL profiles (Table 2). Of the 10 individuals, 8 had polydisperse and 2 had monodisperse LDL profiles before treatment. All 10 showed polydisperse LDL after treatment. Of the 16 individuals treated with fluvastatin who did not show apparent change in the shape of their LDL profiles (other than height), 10 had monodisperse and 6 had polydisperse LDL (see Table 2, footnote h, i). All samples that showed a change in the shape
of LDL profiles after treatment also showed expected changes in the mobility of the main LDL band upon gradient gel electrophoresis (Table 2). Four of the 16 samples with no change in shape of LDL profile also showed changes of LDL bands upon electrophoresis. Thus, 14 of the 26 individuals in the fluvastatin-treated group and 1 of 8 in the placebo-treated group showed changes in the mobility of their main LDL bands after gradient gel electrophoresis. (Xi-square analysis, however, showed that the number of individuals with changed electrophoretic mobility after fluvastatin versus placebo treatment did not reach statistical (P < 0.09). significance Of the 14 in fluvastatin-treated group, 3 showed increased mobility of LDL bands (decreased molecular weight), and 11 of the drug-treated group and 1 in the placebo group showed decreased mobility of LDL bands (increased molecular weight). The total cholesterol/ape B ratios of LDL showed small and statistically insignificant changes. Nevertheless, of the two groups that showed changes in LDL mobility by electrophoresis, the changes were in the expected direction for most samples (2 of the 3 showed decreased cholesterol/ape B ratio in the group with increased electrophoretic mobility and 8 of 11 showed increased cholesterol/ape B ratio in the group with decreased electrophoretic mobility). Examples of how fluvastatin treatment affect plasma LDL absorbance profiles are shown in Fig. 1. Profiles were chosen from 5 fluvastatin and 1 placebo-treated individuals before and after 6 weeks treatment to illustrate changes in absorbance profiles. Peak heights of LDL in all 5 fluvastatin-treated individuals were significantly reduced after therapy. Four of the 5 samples (Nos. 1, 2, 3, 4) represented those with disproportionate reduction of the LDL subspecies, while one (No. 5) of the 5 showed proportionate reduction. Disproportionate reduction of the subspecies can result in different shape changes. In sample 1, an extra shoulder with density lower than the main peak (1.042 g/ml) appeared after therapy. In sample 2, a shoulder with density lower than the main peak (1.044 g/ml) present before therapy disappeared after drug therapy. In samples 3 and 4, the peaks of the LDL profile after drug therapy became broader, most likely due to a preferential
DenSity
1
iu
iu Ab8orbmoe
Abrorbmce
G
ii
DensHy
,
Abrorbmce
w
P
I
N
P
AbSOrb6nCe
cn
Den6lty
Denrlty
AbsOrb6nCe
AbSOrb6llCe
153 +
Fig. 1. LDL absorbance profiles obtained from density gradiant ultracentrifugation from six individuals. (A) before treatment, (B) after treatment: l-5, fluvastatin-treated. Disproportionate reduction of LDL subspecies in individuals 1-4 resulted in shape change, while the shape (other than height) of the LDL absorbance profile in individual 5 remained the same after fluvastatin treatment, implying proportional reduction of all LDL subspecies. 6, an individual treated with placebo showing no change of either height or shape.
1
2
3
4 5
6
7 8
9 10
Fig. 2. Gradient gel electrophoresis of plasma from individuals labelled Nos. 1-4 in Fig. 1. Lanes 5 and 10 represent pooled plasma containing LDL bands 3, 4, and 6. Plasma from individuals Nos. l-4 before drug therapy were spotted on Lanes 1, 3, 6, 8 and after 6 weeks of therapy on Lanes 2, 4, 7, 9 respectively.
reduction of the more dense LDL sent in the main peak. Fig. 2 illustrates that changes profiles were always accompanied electrophoretic patterns. Thus the patterns of samples from individuals
originally
pre-
in absorbance by changes in electrophoretic Nos. l-4 as
illustrated in Fig. 1 are shown in Fig. 2. In sample 1, LDL-6, the only band before treatment (lane 1) was replaced by a mixture of LDL-5 and LDL-6 (lane 2). Sample 2 was composed of LDL-4 and LDL-5, with LDL4 as the predominant band before treatment (lane 3). After treatment, the relatively larger, less dense LDL-4 disappeared and was replaced by a mixture of LDL-5 and LDL-6 (lane 4). In sample 3 and 4 pretreatment samples were characterized by a main band of LDL-6 with lesser quantities of LDL-5 (lane 6,8). Post-treatment pattern showed mixtures of LDL-5 and LDL-6 with LDL-5 being the darker of the two bands (lane 7,9). Thus the changes in electrophoretic patterns were always in the same direction as those observed in LDL absorbance profiles. In order to determine whether fluvastatin therapy affects the chemical composition of LDL subspecies, we determined cholesterol and apo B contents in 6 fractions in the density range of 1.0251.060 g/ml in the last 12 individuals participating
Fig. 3. Effect of fluvastatin on percent of change of apo B (panel A) and cholesterol (panel B) in six LDL fractions (I-VI) of the following density ranges. I, 1.025-1.032; II, 1:033-1.038; III, 1.039-1.045; IV, 1.046-1.049; V, 1.050-1.054; VI, 1.055-1.060 g/ml. The results were from 5 individuals who did not show a change in the shape of LDL absorbance profile or mobility of the main LDL individual No. 5 depicted in Fig. 1. band upon electrophoresis after 6 weeks of fluvastatin therapy. 0 -0,
B
Fig. 4. Effect of fluvastatin on percent change of Fig. 3. The results were from 6 individuals that LDL band upon electrophoresis after 6 weeks ?? -m,
apo B (panel A) and cholesterol (panel B) in 6 LDL fractions (I-VI) as described in showed changes in the shape of LDL absorbance profile and mobility of the main No. 1; of drug therapy. Four of these were also depicted in Fig. 1 (0 -0, No. 2; x x,No.3;rA, No. 4).
in this study. Eleven of the 12 were treated with fluvastatin in this group. Fig. 3 shows the percentage change of apo B and cholesterol in the 6 LDL fractions of 5 individuals in whom drug treatment reduced the peak height of LDL profiles not accompanied by other shape change. Similar results from 6 individuals in whom drug therapy probably caused preferential removal of certain subspecies of LDL, resulting in a change in the shape of LDL profile after therapy are shown in Fig. 4. In Fig. 3, the percentage of cholesterol or apo B reduction in the 6 fractions is similar in each individual, confirming the lack of significant preferential removal of any subspecies. In Fig. 4, percentage of reduction of either cholesterol or apo B in the individual fractions varied from one to the other, confirming the preferential removal of selected subspecies resulting in a change of the shape of the LDL profile after drug therapy. While fluvastatin seemed to cause preferential removal of LDL subspecies in a portion of the
treated individuals, it did not change the composition of the individual LDL subspecies measured by the cholesterol/ape B and cholesteryl ester/ cholesterol ratios. Table 3 showed that the cholesterol/ apo B and cholesteryl ester/ cholesterol ratios in the individual fractions were unchanged before and after therapy in both the groups with proportionate and disproportionate reduction of LDL subspecies. Discussion
In the present report, we documented changes of LDL subspecies in patients with hypercholesterolemia after they had been treated with an HMG-CoA reductase inhibitor for 6 weeks. Previous studies showed that LDL exists as seven subspecies by density gradient ultracentrifugation and gradient gel electrophoresis [7-91. In this study, LDL heterogeneity was studied by density gradient ultracentrifugation and gradient gel electro-
155
156 phoresis before and after placebo or drug therapy. After treatment with fluvastatin, plasma LDL in 42% of the patients became larger and less dense, while in 12% of the patients changed to smaller, more dense LDL as judged by the mobility of the main LDL band on electrophoresis. We know of no previous published studies comparing LDL subspecies in sera of individuals before and after cholesterol-lowering drug therapies documented by electrophoretic and density gradient ultracentrifugation techniques. Witztum et al. [14-171 studied the effects of bile acid sequestrant therapies on LDL composition. These investigators showed that colestipol and cholestyramine therapies changed LDL composition in men and guinea pigs. In men, colestipol was shown to reduce significantly LDLcholesterol and apo B [15]. However, the reduction in apo B level was less than LDL-C, resulting in a decrease in the LDL-C/ape B ratio post therapy [15]. From these results and electron microscopic measurement of LDL particle sizes, the authors suggested that colestipol therapy caused preferential removal of larger, less dense LDL, resulting in protein-enriched, smaller, more dense LDL post-therapy [15,17]. Similar results had been observed in LDL of guinea pigs treated with cholestyramine [16]. The results of the current study showed that fluvastatin, an HMG-CoA reductase inhibitor, reduced LDL-C without significantly changing the total cholesterol/ape B ratio in LDL. Recently, Bergland et al. [18] showed lovastatin slightly affected the composition of LDL in guinea pigs. Our results, obtained in humans using a different HMG CoA reductase inhibitor, basically agree with theirs. While the compositions of LDL seemed little changed, small changes in LDL can be detected by the sensitive method of gradient gel electrophoresis in over 50% of the treated patients. Lesser percentages (38.5%) showed changes in the shape (other than height) of the LDL profile, obtained from density gradient ultracentrifugation. But as illustrated in Figs. 1 and 2, changes in the LDL profile from density gradient ultracentrifugation were always accompanied by electrophoretic mobility of LDL-bands in the same direction. Relatively few samples (19.2%) showed a significant change in main peak density, confirming that changes in LDL after HMG-CoA
reductase inhibitor therapy consisted mainly of small, subtle changes. In summary, fluvastatin, in lowering serum LDL concentration, also changed the LDL subspecies in more than 50% of the patients. However, in the present study with the relatively small placebo-control group, the increased change of LDL subspecies in the fluvastatin-treated group as compared to the placebo group did not reach statistical significance (P < 0.09). While the fluvastatin-treated group clearly showed a trend of increased change of LDL subspecies than the placebo-controlled group and this comparison will most likely become statistically significant in a larger group, the finding also underscored the relatively small changes in LDL subspecies and the lack of LDL subspecies change in almost 50% of the individuals treated with HMG-CoA reductase inhibitor. Unlike bile acid sequestrant therapies, HMG-CoA reductase inhibitor, therapy did not result in the preferential removal of larger, less dense LDL particles. Of those patients that showed changes in LDL subspecies after drug therapy, LDL in more individuals (78%) became larger and less dense. However, the tendency to shift towards less dense LDL is not statistically significant. Further studies of the effects of other therapies or combined therapies on LDL subspecies should facilitate our understanding at the molecular level. References 1 Consensus Conference. Lowering blood cholesterol to prevent heart disease. JAMA, 253 (1985) 2080. 2 Lipid Research Clinics, The Lipid Research Clinics Coronary Primary Prevention Trials Results. JAMA, 251 (1984) 351. 3 Reports of the National Cholesterol Education Program Expert Panel on detection, evaluation and treatment of high blood cholesterol in adults. Arch Intern. Med., 148 (1988) 36. 4 Grundy, S.M., HMG-CoA reductase inhibitors for treatment of hypercholesterolemia. N. Engl. J. Med., 319 (1988) 24. 5 Hunninghake, D.B., Clinical trials of lovastatin and simvastatin versus cholestyramine. Atheroscler. Rev., 18 (1988) 133. 6 Lindgren, F.T., Elliot, A. and Gofman, J.W., The ultracentrifugal characterization and isolation of human blood lipids and lipoproteins and applications to the study of atherosclerosis. J. Phys. Colloid Chem., 55 (1951) 80.
157 Krauss. R.M. and Burke, D.J., Identification of multiple subclasses of plasma low density lipoproteins in normal human. J. Lipid Res., 23 (1982) 97. McNamara, J.R., Campos, H., Ordovas, J.M., et al., Effect of gender, age, and lipid status on low density lipoprotein subfraction distribution. Arteriosclerosis, 7 (1987) 483. Teng, B., Sniderman, A.D., Soutar, A.K. and Thompson, G.R., Metabolic basis of hyperapobeta-lipoproteinemia: turnover of apolipoprotein B in low density lipoprotein and its precursors and subfractions compared with normal and familial hypercholesterolemia. J. Clin. Invest., 77 (1986) 663. 10 Mantell, G., Lipid lowering drugs in atherosclerosis-the HMG-CoA reductase inhibitors. Clin. Exp. Hypertens. Theory Pratt., All (1989) 927. W.T., Levy, RI. and Frederickson, D.S., Esti11 Friedwald, mation of the concentration of low-density lipoprotein cholesterol in plasma without use of the preparative ultracentrifuge. Clin. Chem., 18 (1972) 499. V., Edelstein, C. and Scanu, 12 Nilsson, J., Mannickarouttu, A.M., An improved detection system applied to the study of serum lipoproteins after single-step density gradient ultracentrifugation. Anal. Biochem., 110 (1981) 342.
13 Kelley, J.L. and Kruski, A.W., Density gradient ultracentrifugation of serum lipoproteins in a swinging bucket rotor. Atherosclerosis, 24 (1976) 170. 14 Witztum, J.L., Schonfeld, G. and Wiedman, S.W.. The effects of colestipol on the metabolism of very-low-density lipoprotein in man. J. Lab. Clin. Med., 88 (1976) 1008. 15 Witztum, J.L.. Schonfeld, G., Weidman. SW. et al.. Bile sequestrant therapy alters the composition of low-density and high-density lipoproteins. Metabolism, 28 (1979) 221. 16 Witztum, J.L., Young, S.G., Elam. R.L. et al., Cholestyramine-induced changes in low density lipoprotein composition and metabolism. I. Studies in the guinea pig J. Lipid Res., 26 (1985) 92. 17 Young, SC., Witztum, J.L., Carew, T.E. et al., Colestipolinduced changes in LDL composition and metabolism. II. Studies in humans. J. Lipid Res., 30 (1989) 225. 18 Berglund. L., Sharkey, M.F., Elam, R.L. and Witztum, J.L.. Effects of lovastatin therapy on guinea pig low density lipoprotein composition and metabolism. J. Lipid Res., 30 (1989) 1591.