Atherosclerosis, 85 (1990) 193-202 Elsevier Scientific Publishers Ireland,
193 Ltd.
ATHERO
Postprandial lipemia, fenofibrate and coronary artery disease H.S. Simpson ‘, C.M. Williamson 2, T. Olivecrona 3, S. Pringle I, J. Maclean I, A.R. Lorimer I, F. Bonnefous 4, Y. Bogaievsky 4, C.J. Packard 2 and J. Shepherd 2 ’Department of Medical Cardiology and ’ Department of Biochemistry, Royal Infirmary Glasgow G4 OSF (U.K.). ’ Department of Physiological Chemistry,
University of Ume6, Limed S-901 87 (Sweden), and * Luboratoires
Fournier, Dijon (France)
(Received 11 May, 1990) (Revised, received 10 August, 1990) (Accepted 20 August, 1990)
Summary
This report describes the response of patients with severe coronary artery disease to a dynamic fat load test and monitors the change induced by fenofibrate therapy. The presence of disease was associated with prolonged and exaggerated hypertriglyceridemia following the meal and with lower basal HDL cholesterol and HDL subfraction masses. A further indicator of risk was the persistence of increased amounts of retinyl palmitate in the plasma of severely affected individuals 24 h after its ingestion with the meal. These observations are consistent with the proposal that the clearance of chylomicrons and their remnants is impaired in coronary atherosclerosis. Fenofibrate reduced alimentary lipemia following the fat load in both normo- and hypercholesterolemic subjects. This was associated with a 10% rise in plasma HDL cholesterol levels. The improvement in chylomicron catabolism probably derived from a 37% increase (P < 0.001) in lipoprotein lipase activity induced by fenofibrate. Hepatic lipase on the other had was only slightly affected by treatment. Key words: Chylomicrons; Atherosclerosis; lipases; Fat load test
Lipoprotein
Introduction
It is now widely accepted that aberrations of lipid and lipoprotein metabolism are involved in the initiation and progression of atherosclerosis [l]. Raised plasma cholesterol, or more specifi-
Correspondence to: Professor J. Shepherd, Department Biochemistry, Royal Infirmary, Glasgow G4 OSF, U.K.
0021-9150/90/$03.50
0 1990 Elsevier Scientific
Publishers
of
Ireland,
lipase; High density lipoproteins;
Post-heparin
tally, an increment in the cholesterol content of the low density lipoprotein (LDL) fraction confers an increased risk of coronary artery disease (CAD). Conversely, high density lipoprotein (HDL) cholesterol [2] or that subfraction known as HDL, is negatively correlated [3] with the disease, leading to the speculation that such lipoproteins are ‘antiatherogenic’. Although triglycerides have been implicated in atherogenesis, the nature of their involvement in Ltd.
194 Patients and methods
the process is the subject of dispute [4]. Prospective and cross-sectional epidemiological studies have often revealed the presence of hypertriglyceridemia in CAD sufferers, but the independent role of this parameter is not substantiated by multivariate statistical analysis [5]. Individuals with high plasma triglyceride levels commonly express low HDL cholesterol, and it is the latter which appears to play the dominant role in predisposing to atherosclerosis. The syndrome of high plasma triglxceride in the presence of low HDL cholesterol ‘and small, dense LDL is now being increasingly recognized as a combined risk marker [4,6]. This premise, however, is based on risk parameter measurements that are generally made on fasting patients. Under normal circumstances, the plasma concentrations of triglyceride and its carrier lipoproteins (chylomicrons and very low density lipoproteins (VLDL)) show a wide diurnal fluctuation [7] producing a variable lipemia which often persists for up to 8 h. It has been postulated that the latter may constitute an additional CAD risk [8] which is conventionally ignored in the design of most studies. On the basis of this argument we set out to compare post-prandial lipemia in patients with and without CAD and assess the effect of fenofibrate.
TABLE SERUM
The project was conducted in two parts. In the first, 52 subjects (Table 1) aged between 40 and 60 years (mean + 1.0 SD = 50 f 7 yrs) consented to undergo a fat tolerance test. All had been subjected to routine coronary angiography for assessment of CAD within the previous 6 months but were otherwise healthy. Hematological and biochemical screening showed no evidence of the presence of disorders likely to produce secondary hyperlipidemia. Diabetes mellitus, gout, renal or hepatic disease constituted specific exclusion criteria. All patients had been counselled about their smoking habits and on average consumed less than 10 cigarettes/day. The objective of the second study was to determine the efficacy of fenofibrate in improving chylomicron clearance in subjects with and without angiographic evidence of coronary artery disease. Three groups, each of eight subjects were chosen from those participating in study 1 (see Table 1). The first was drawn at random from the 18 individuals who had no evidence of CAD and had plasma cholesterol levels of less than 7.0 mmol/l. The patients in group 2 were selected from the 24 subjects positive for CAD and with plasma
1 LIPIDS
AND
LIPOPROTEINS
IN CAD-POSITIVE
AND
CAD-NEGATIVE
SUBJECTS
(STUDY
1)
Values are means f 1.0 SD. Group
CAD negative (n =18) CAD positive (n=34) Normocholesterolemic CAD positive (n=24)
Serum * cholesterol
Serum * * triglyceride
Cholesterol
(mmol/l)
(mmol/l)
(mmol/l)
6.11+ 0.96
2.09 f 0.94
0.96 + 0.36
3.98 kO.85
6.91 f 1.34
2.52 + 0.96
1.06kO.52
6.19&0.84
2.31+ 0.97
0.94 f 0.41
VLDL
HDL, mass
HDL, mass
(mg/dl)
(mg/dl)
1.18kO.27
46+26
249 f 42
4.89 f 1.22 a
0.98 k 0.24 b
31*14c
213+41
a
4.29 f 0.79
0.96 + 0.24 a
31*14c
209k36
a
LDL
HDL
+
+ Subjects from the CAD positive group Unpaired r-test, versus CAD negative. = PcO.01; b P < 0.02; c P < 0.05. * Multiply by 38.7 to convert to rng/dl. * * Multiply by 88.5 to convert to mg/dl.
who had a fasting
cholesterol
level
< 7.00 mrnol/l.
195 cholesterol levels below 7.0 mmol/l. Group 3 comprised 8 of the 10 subjects with CAD and plasma cholesterol values greater than 7.0 mrnol/l. The lipid parameters of these three groups are detailed in Table 2. Fat tolerance test
The protocol for the fat tolerance test was based on the method of Patsch et al. [9]. Retinyl palmitate was employed as a measure of chylomicron remnant metabolism in the study [lo] since it was the major species of esterified retinol found in the plasma of all patients and did not vary as a percentage of their total retinyl esters. After a 16 h overnight fast the subjects were given a meal consisting of 280 ml cream, 20 g sucrose, 20 g dried skimmed milk and 40 ml flavored syrup, made up to 500 ml with water and homogenised in a blender. The total energy content of the meal was 1480 kcal, coming from 11.4 g protein, 137 g fat (approximately 60% saturated) and 50.3 g carbohydrate. The meal contained 400 mg cholesterol and 300000 IU of retinyl palmitate (approx. 165 mg) were added as an oily solution before homogenisation. A fasting blood sample was taken prior to ingestion of the meal. Subsequent samples of blood were drawn at frequent intervals thereafter throughout the day. A vitamin A-free meal was permitted following collection of the 8 h specimen and thereafter the patients were fasted until a 24 h specimen was collected. The areas under the triglyceride and retinyl palmitate curves were calculated by integrating the values between 0 and 8 h following subtraction of the zero time (fasting) baseline value. Results are expressed in arbitrary units. ,:-: -0a k.&
E E
ss;;vi
fX8$?-ZS ,VVV%Z ,g444EE d P.nu
5s rr 9B g; 99 “. ‘c! Z% 0”s ST
h
?? ??
x
Laboratory
analyses
Cholesterol and triglyceride measurements were made by enzymatic procedures (Boehringer Mannheim, Mannheim, F.R.G.) on an Encore centrifugal analyser (Baker Instruments Cot-p, Allentown, PA 18001, U.S.A.). The laboratory is standardised to the Centers for Disease Control Lipid Standardisation program (CDC, Atlanta, Georgia). VLDL and chylomicrons were isolated by ultracentrifugation at a density of 1.006 kg/l [ll]. HDL cholesterol was measured as described by Warnick and Albers [12]. HDL subfractions
196 (HDL, and HDL,) were estimated by an analytical ultracentrifugation procedure [13]. Retinyl esters present in d < 1.006 kg/l lipoproteins were quantified as follows: 500 ~1 sample, to which was added 0.5 pg retinyl acetate internal standard, was extracted with diethyl ether containing 1% acetic acid. The extract was dried under N,, redissolved in 250 ~1 of isopropanol and analyzed for retinyl pahnitate by high performance liquid chromatography on a Brownlea Aquapore RP 300 reversed phase column using a gradient of 50-100% acetonitrile in water. The retinoids of interest were quantified by integration of their absorption peaks at 326 nm using a Shimadzu Chromatopac C-RlB chart recorder and integrator. Lipase determinations In study 2, measurements of hepatic (HL) and lipoprotein lipase (LpL) were made on postheparin plasma samples drawn after the 24 h fasting blood sample. These were frozen immediately (-70°C) and shipped on dry ice to UmeH, Sweden. Lipoprotein lipase was measured after immunoinhibition of hepatic lipase with immunoglobulins from a goat antiserum [14] to human HL. Antiserum (0.5 vol.) was added to the sample (10 ~1) and the mixture incubated for 2 h on ice before assay. The substrate was Intralipid into which [ 3H]oleic acid labeled triolein had been incorporated by sonication. Apo C-II was added to activate the enzyme and the reaction allowed to proceed for 30 min at 25°C. The assay is linear over this time, at least for normal subjects [15]. Hepatic lipase was measured by a similar procedure using as substrate [3H]oleic acid labeled triolein emulsified with gum arabic, and with 1 M NaCl in the medium to inhibit LpL [15]. CAD scoring CAD scoring was performed as outlined elsewhere [16]. Basically, all three major coronary vessels were divided into 3 segments, each of which was scored from 0 to 4 depending on the extent of atherosclerotic narrowing. On this basis we identified in the first study 18 CAD negative individuals (score O-l) and 34 CAD positive patients (score 3-17). Of the 18 CAD negative subjects one had been prescribed atenolol and one metoprolol. In the CAD positive group, 3 were
receiving atenolol, 5 metoprolol and one propranolol. For study 2, one subject in group 1 was prescribed atenolol and bendrofluazide together, 3 in group 2 were taking atenolol and one metopro101 while 3 in group 3 were prescribed atenolol alone. Statistical analysis Statistical analysis was performed using the Student’s t-test or, for non-gaussian parameters, the Mann Whitney and Wilcoxon tests.
Results
Study 1: postprandial lipemia and coronary artery disease Non lipid risk factors were equally represented in both the positive and CAD negative groups. In particular, there were no significant differences in age (CAD negative = 50 + 8 yrs; CAD positive = 50 f 8 yrs), blood pressure (CAD negative = 122 + 12/76 & 12 mm Hg; CAD positive = 130 h 15/83 f 13 mm Hg), Quetelet index (CAD negative = 25.6 & 4.0; CAD positive 26.8 f 3.0) and smoking. There were however substantial differences in the plasma lipid and lipoprotein distributions (Table 1). Cholesterol levels were elevated in the group with significant disease, due to an increment (P < 0.01) in LDL cholesterol. Conversely, HDL cholesterol, HDL, and HDL, were lower in the CAD positive individuals. Ten individuals in the CAD positive group had plasma cholesterol values so high that they might be considered to constitute an over-riding risk, obscuring the importance of other factors. When these subjects were excluded from the statistical analyses (Table l), the mean serum cholesterol in the CAD positive group fell to values similar to those in the CAD negative group. However the differences in HDL between the groups persisted. The responses of both groups to the fatty meal are illustrated in Fig. 1. The mean fasting d < 1.006 kg/l triglyceride value, which was similar in the control and CAD positive groups at baseline, became significantly higher in the CAD positive subjects at the peak of lipemia (i.e., between 4 and 8 h after the fat meal). This suggests that CAD positive patients metabolize intestinally derived
197
hours
+ p
193 Ltd.
ATHERO
Postprandial lipemia, fenofibrate and coronary artery disease H.S. Simpson ‘, C.M. Williamson 2, T. Olivecrona 3, S. Pringle I, J. Maclean I, A.R. Lorimer I, F. Bonnefous 4, Y. Bogaievsky 4, C.J. Packard 2 and J. Shepherd 2 ’Department of Medical Cardiology and ’ Department of Biochemistry, Royal Infirmary Glasgow G4 OSF (U.K.). ’ Department of Physiological Chemistry,
University of Ume6, Limed S-901 87 (Sweden), and * Luboratoires
Fournier, Dijon (France)
(Received 11 May, 1990) (Revised, received 10 August, 1990) (Accepted 20 August, 1990)
Summary
This report describes the response of patients with severe coronary artery disease to a dynamic fat load test and monitors the change induced by fenofibrate therapy. The presence of disease was associated with prolonged and exaggerated hypertriglyceridemia following the meal and with lower basal HDL cholesterol and HDL subfraction masses. A further indicator of risk was the persistence of increased amounts of retinyl palmitate in the plasma of severely affected individuals 24 h after its ingestion with the meal. These observations are consistent with the proposal that the clearance of chylomicrons and their remnants is impaired in coronary atherosclerosis. Fenofibrate reduced alimentary lipemia following the fat load in both normo- and hypercholesterolemic subjects. This was associated with a 10% rise in plasma HDL cholesterol levels. The improvement in chylomicron catabolism probably derived from a 37% increase (P < 0.001) in lipoprotein lipase activity induced by fenofibrate. Hepatic lipase on the other had was only slightly affected by treatment. Key words: Chylomicrons; Atherosclerosis; lipases; Fat load test
Lipoprotein
Introduction
It is now widely accepted that aberrations of lipid and lipoprotein metabolism are involved in the initiation and progression of atherosclerosis [l]. Raised plasma cholesterol, or more specifi-
Correspondence to: Professor J. Shepherd, Department Biochemistry, Royal Infirmary, Glasgow G4 OSF, U.K.
0021-9150/90/$03.50
0 1990 Elsevier Scientific
Publishers
of
Ireland,
lipase; High density lipoproteins;
Post-heparin
tally, an increment in the cholesterol content of the low density lipoprotein (LDL) fraction confers an increased risk of coronary artery disease (CAD). Conversely, high density lipoprotein (HDL) cholesterol [2] or that subfraction known as HDL, is negatively correlated [3] with the disease, leading to the speculation that such lipoproteins are ‘antiatherogenic’. Although triglycerides have been implicated in atherogenesis, the nature of their involvement in Ltd.
194 Patients and methods
the process is the subject of dispute [4]. Prospective and cross-sectional epidemiological studies have often revealed the presence of hypertriglyceridemia in CAD sufferers, but the independent role of this parameter is not substantiated by multivariate statistical analysis [5]. Individuals with high plasma triglyceride levels commonly express low HDL cholesterol, and it is the latter which appears to play the dominant role in predisposing to atherosclerosis. The syndrome of high plasma triglxceride in the presence of low HDL cholesterol ‘and small, dense LDL is now being increasingly recognized as a combined risk marker [4,6]. This premise, however, is based on risk parameter measurements that are generally made on fasting patients. Under normal circumstances, the plasma concentrations of triglyceride and its carrier lipoproteins (chylomicrons and very low density lipoproteins (VLDL)) show a wide diurnal fluctuation [7] producing a variable lipemia which often persists for up to 8 h. It has been postulated that the latter may constitute an additional CAD risk [8] which is conventionally ignored in the design of most studies. On the basis of this argument we set out to compare post-prandial lipemia in patients with and without CAD and assess the effect of fenofibrate.
TABLE SERUM
The project was conducted in two parts. In the first, 52 subjects (Table 1) aged between 40 and 60 years (mean + 1.0 SD = 50 f 7 yrs) consented to undergo a fat tolerance test. All had been subjected to routine coronary angiography for assessment of CAD within the previous 6 months but were otherwise healthy. Hematological and biochemical screening showed no evidence of the presence of disorders likely to produce secondary hyperlipidemia. Diabetes mellitus, gout, renal or hepatic disease constituted specific exclusion criteria. All patients had been counselled about their smoking habits and on average consumed less than 10 cigarettes/day. The objective of the second study was to determine the efficacy of fenofibrate in improving chylomicron clearance in subjects with and without angiographic evidence of coronary artery disease. Three groups, each of eight subjects were chosen from those participating in study 1 (see Table 1). The first was drawn at random from the 18 individuals who had no evidence of CAD and had plasma cholesterol levels of less than 7.0 mmol/l. The patients in group 2 were selected from the 24 subjects positive for CAD and with plasma
1 LIPIDS
AND
LIPOPROTEINS
IN CAD-POSITIVE
AND
CAD-NEGATIVE
SUBJECTS
(STUDY
1)
Values are means f 1.0 SD. Group
CAD negative (n =18) CAD positive (n=34) Normocholesterolemic CAD positive (n=24)
Serum * cholesterol
Serum * * triglyceride
Cholesterol
(mmol/l)
(mmol/l)
(mmol/l)
6.11+ 0.96
2.09 f 0.94
0.96 + 0.36
3.98 kO.85
6.91 f 1.34
2.52 + 0.96
1.06kO.52
6.19&0.84
2.31+ 0.97
0.94 f 0.41
VLDL
HDL, mass
HDL, mass
(mg/dl)
(mg/dl)
1.18kO.27
46+26
249 f 42
4.89 f 1.22 a
0.98 k 0.24 b
31*14c
213+41
a
4.29 f 0.79
0.96 + 0.24 a
31*14c
209k36
a
LDL
HDL
+
+ Subjects from the CAD positive group Unpaired r-test, versus CAD negative. = PcO.01; b P < 0.02; c P < 0.05. * Multiply by 38.7 to convert to rng/dl. * * Multiply by 88.5 to convert to mg/dl.
who had a fasting
cholesterol
level
< 7.00 mrnol/l.
195 cholesterol levels below 7.0 mmol/l. Group 3 comprised 8 of the 10 subjects with CAD and plasma cholesterol values greater than 7.0 mrnol/l. The lipid parameters of these three groups are detailed in Table 2. Fat tolerance test
The protocol for the fat tolerance test was based on the method of Patsch et al. [9]. Retinyl palmitate was employed as a measure of chylomicron remnant metabolism in the study [lo] since it was the major species of esterified retinol found in the plasma of all patients and did not vary as a percentage of their total retinyl esters. After a 16 h overnight fast the subjects were given a meal consisting of 280 ml cream, 20 g sucrose, 20 g dried skimmed milk and 40 ml flavored syrup, made up to 500 ml with water and homogenised in a blender. The total energy content of the meal was 1480 kcal, coming from 11.4 g protein, 137 g fat (approximately 60% saturated) and 50.3 g carbohydrate. The meal contained 400 mg cholesterol and 300000 IU of retinyl palmitate (approx. 165 mg) were added as an oily solution before homogenisation. A fasting blood sample was taken prior to ingestion of the meal. Subsequent samples of blood were drawn at frequent intervals thereafter throughout the day. A vitamin A-free meal was permitted following collection of the 8 h specimen and thereafter the patients were fasted until a 24 h specimen was collected. The areas under the triglyceride and retinyl palmitate curves were calculated by integrating the values between 0 and 8 h following subtraction of the zero time (fasting) baseline value. Results are expressed in arbitrary units. ,:-: -0a k.&
E E
ss;;vi
fX8$?-ZS ,VVV%Z ,g444EE d P.nu
5s rr 9B g; 99 “. ‘c! Z% 0”s ST
h
?? ??
x
Laboratory
analyses
Cholesterol and triglyceride measurements were made by enzymatic procedures (Boehringer Mannheim, Mannheim, F.R.G.) on an Encore centrifugal analyser (Baker Instruments Cot-p, Allentown, PA 18001, U.S.A.). The laboratory is standardised to the Centers for Disease Control Lipid Standardisation program (CDC, Atlanta, Georgia). VLDL and chylomicrons were isolated by ultracentrifugation at a density of 1.006 kg/l [ll]. HDL cholesterol was measured as described by Warnick and Albers [12]. HDL subfractions
196 (HDL, and HDL,) were estimated by an analytical ultracentrifugation procedure [13]. Retinyl esters present in d < 1.006 kg/l lipoproteins were quantified as follows: 500 ~1 sample, to which was added 0.5 pg retinyl acetate internal standard, was extracted with diethyl ether containing 1% acetic acid. The extract was dried under N,, redissolved in 250 ~1 of isopropanol and analyzed for retinyl pahnitate by high performance liquid chromatography on a Brownlea Aquapore RP 300 reversed phase column using a gradient of 50-100% acetonitrile in water. The retinoids of interest were quantified by integration of their absorption peaks at 326 nm using a Shimadzu Chromatopac C-RlB chart recorder and integrator. Lipase determinations In study 2, measurements of hepatic (HL) and lipoprotein lipase (LpL) were made on postheparin plasma samples drawn after the 24 h fasting blood sample. These were frozen immediately (-70°C) and shipped on dry ice to UmeH, Sweden. Lipoprotein lipase was measured after immunoinhibition of hepatic lipase with immunoglobulins from a goat antiserum [14] to human HL. Antiserum (0.5 vol.) was added to the sample (10 ~1) and the mixture incubated for 2 h on ice before assay. The substrate was Intralipid into which [ 3H]oleic acid labeled triolein had been incorporated by sonication. Apo C-II was added to activate the enzyme and the reaction allowed to proceed for 30 min at 25°C. The assay is linear over this time, at least for normal subjects [15]. Hepatic lipase was measured by a similar procedure using as substrate [3H]oleic acid labeled triolein emulsified with gum arabic, and with 1 M NaCl in the medium to inhibit LpL [15]. CAD scoring CAD scoring was performed as outlined elsewhere [16]. Basically, all three major coronary vessels were divided into 3 segments, each of which was scored from 0 to 4 depending on the extent of atherosclerotic narrowing. On this basis we identified in the first study 18 CAD negative individuals (score O-l) and 34 CAD positive patients (score 3-17). Of the 18 CAD negative subjects one had been prescribed atenolol and one metoprolol. In the CAD positive group, 3 were
receiving atenolol, 5 metoprolol and one propranolol. For study 2, one subject in group 1 was prescribed atenolol and bendrofluazide together, 3 in group 2 were taking atenolol and one metopro101 while 3 in group 3 were prescribed atenolol alone. Statistical analysis Statistical analysis was performed using the Student’s t-test or, for non-gaussian parameters, the Mann Whitney and Wilcoxon tests.
Results
Study 1: postprandial lipemia and coronary artery disease Non lipid risk factors were equally represented in both the positive and CAD negative groups. In particular, there were no significant differences in age (CAD negative = 50 + 8 yrs; CAD positive = 50 f 8 yrs), blood pressure (CAD negative = 122 + 12/76 & 12 mm Hg; CAD positive = 130 h 15/83 f 13 mm Hg), Quetelet index (CAD negative = 25.6 & 4.0; CAD positive 26.8 f 3.0) and smoking. There were however substantial differences in the plasma lipid and lipoprotein distributions (Table 1). Cholesterol levels were elevated in the group with significant disease, due to an increment (P < 0.01) in LDL cholesterol. Conversely, HDL cholesterol, HDL, and HDL, were lower in the CAD positive individuals. Ten individuals in the CAD positive group had plasma cholesterol values so high that they might be considered to constitute an over-riding risk, obscuring the importance of other factors. When these subjects were excluded from the statistical analyses (Table l), the mean serum cholesterol in the CAD positive group fell to values similar to those in the CAD negative group. However the differences in HDL between the groups persisted. The responses of both groups to the fatty meal are illustrated in Fig. 1. The mean fasting d < 1.006 kg/l triglyceride value, which was similar in the control and CAD positive groups at baseline, became significantly higher in the CAD positive subjects at the peak of lipemia (i.e., between 4 and 8 h after the fat meal). This suggests that CAD positive patients metabolize intestinally derived
197
hours
+ p
