| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Departments of Endocrinology and Biochemistry (O.J., R.B., J.O.-L.), Hospital Sant Pau, and Department of Biochemistry and Molecular Biology, Universitat Autònoma de Barcelona (J.O.-L.), 08025 Barcelona, Spain
Address all correspondence and requests for reprints to: Dr. Antonio Pérez, Department of Endocrinology, Hospital Sant Pau, S. Antoni M. Claret 167, 08025 Barcelona, Spain. E-mail: aperez{at}hsp.santpau.es.
| Abstract |
|---|
|
|
|---|
| Introduction |
|---|
|
|
|---|
This study was undertaken to compare the efficacies of atorvastatin and gemfibrozil and the combination of both drugs in patients with type 2 diabetes and LDLc above the recommended goal in the absence of severe hypertriglyceridemia.
| Subjects and Methods |
|---|
|
|
|---|
The study was designed as a randomized, open, cross-over study, comparing the effects of atorvastatin and gemfibrozil on the different components of diabetic dyslipidemia. All patients received either atorvastatin or gemfibrozil for 12 wk, followed by the alternative treatment. If they did not suffer clinically or biologically significant side effects to any of the drugs, they received 12 wk of combined treatment thereafter. Each treatment period was separated from the next by a 4-wk wash-out period. Measurement of liver enzymes, creatine kinase, and lipid concentrations and recording of side effects were performed 6 wk into each treatment period to assess safety and decide dose titration. If any of the following was not achieved. atorvastatin was increased from 10 to 20 mg and gemfibrozil from 900 to 1200 mg/d: LDLc less than 100 mg/dl (2.6 mmol/liter), triglyceride less than 150 mg/dl (1.7 mmol/liter), and apoB less than 1.0 g/liter, and in the absence of significant side effects. Combination treatment consisted of 10 mg atorvastatin and 900 mg gemfibrozil for 12 wk without titration. The study protocol was accepted by the local ethics committee, and the study itself was performed between May 1999 and May 2001.
Patients
To be included in the study patients had to fulfill the following criteria: men and women with type 2 diabetes, aged 3575 yr, no treatment known to interfere with lipid metabolism (nonselective ß-blockers, high dose diuretics, systemic steroids, lipid-lowering drugs) in the month preceding inclusion in the study, plasma LDLc greater than 100 mg/dl (2.6 mmol/liter), and triglycerides less than 400 mg/dl (4.51 mmol/liter). Patients were excluded if they were pregnant, if no reliable contraceptive method was used, or if they displayed serum creatinine more than 1.7 mg/dl (150 µmol/liter), hepatic dysfunction (transaminases >1.5 times upper normal limit at inclusion), creatine kinase more than 3 times the upper normal limit, or acute or chronic disorders that might interfere with compliance. All patients signed informed written consent before being included in the study. Randomization was made following a computer-generated table.
Physical examination, including anthropometric parameters and blood pressure, was performed at baseline and at each visit. Blood samples were drawn at baseline and 6 and 12 wk after the administration of each treatment.
Laboratory measurements
Total cholesterol and triglycerides were measured by enzymatic methods, and HDLc was measured by a direct method (Roche, Indianapolis, IN). LDLc was calculated using Friedewalds formula (13) when triglycerides did not exceed 300 mg/dl (3.45 mmol/liter). Otherwise, ultracentrifugation was performed, and LDLc was estimated in the infranatant after separating the fraction with a density less than 1.006 kg/liter. ApoB and apoAI were measured using an immunoturbidimetric method (Roche) calibrated against World Health Organization/International Federation of Clinical Chemistry Reference Standard SP3-07 for apoB and SP1-01 for apoAI (14). LDL size was determined by electrophoresis on gradient (216%) polyacrylamide gels, cast in the laboratory, according to the method described by Nichols with modifications (15). A volume of 10 µl of plasma samples was applied on lanes in a final concentration of 10% sucrose, stained with Sudan Black [prepared in the laboratory using ethylene glycol and 0.1% (wt/vol) Sudan Black (Sigma-Aldrich Corp., St. Louis, MO)]. Electrophoresis was performed in a refrigerated cell for a prerun of 60 min at 120 V, followed by 30 min at 20 V, 30 min at 70 V, and 16 h at 100 V. A pool containing sera with four LDL fractions whose diameters (22.9 ± 0.7, 24.5 ± 0.6, 26.2 ± 0.5, and 28.4 ± 0.9 nm) had been previously assessed by electron microscopy was used as a control. The gels were scanned (Gel-DOC 2000, Bio-Rad Laboratories, Inc., Hercules, CA), and migration distances (from the top of the gel to the most prominent band) were measured. The predominant LDL particle diameter of each sample was calculated from a calibration line using the four standards of known diameter. LDL particle subclasses were classified as predominantly small LDL or phenotype B (diameter <25.5 nm) and nonsmall LDL (phenotype A, diameter >25.5 nm) (16). Both intra- and intergel imprecisions were less than 1%. Hemoglobin A1c was measured by ion exchange HPLC (Variant, Bio-Rad Laboratories, Inc., Hercules, CA; normal range, 4.65.8%; mean, 5.1 ± 0.3%).
Efficacy and safety
Changes in lipid and apoprotein concentrations and LDL size were assessed for each treatment and then compared. In addition, achievement of therapeutic goals was evaluated and compared among treatments. The therapeutic goals considered were those recommended by international guidelines (4, 17). Because normal apoB concentrations vary between populations and because no internationally accepted goal is available, we calculated the equivalent to an LDLc of 100 mg/dl (2.6 mmol/liter) following a previously published equation based on data from a nondiabetic normolipidemic population (apoB = 0.8 g/liter) (2). Side effects were registered as were increases in liver enzymes and creatine kinase concentrations.
Statistical analysis
Analysis was performed using the SAS 8.2 statistical package (SAS Institute, Inc., Cary, NC). Continuous variables were expressed as the mean ± SD or median (interval) depending on their normality distribution, and qualitative data are expressed as percentages. Comparison of changes in lipid and lipoprotein concentrations with each treatment was performed using ANOVA. Frequencies between groups were compared using
2. Tests were bilateral, and P < 0.05 was considered significant.
| Results |
|---|
|
|
|---|
A total of 46 patients were initially included in the study. Two dropped out immediately after randomization (before starting treatment): 1 was excluded because the second determination of LDLc was less than 100 mg/dl (2.6 mmol/liter) and the other for personal reasons. A total of 22 started atorvastatin, and 22 started gemfibrozil. One patient dropped out during the study because of unspecific side effects to gemfibrozil. Two patients participated in the comparison between both drugs, but dropped out thereafter due to personal and family health problems unrelated to the study treatment. Thus, 43 patients were available for single drug comparison, and 41 were also available for combination therapy. Their baseline features are displayed in Tables 1
and 2
.
|
|
Drug dose was increased from 10 to 20 mg/d at 6 wk in 45% of the patients when treated with atorvastatin (mean final dose, 14.5 mg/d) and from 900 to 1200 mg/d in 88% of them when treated with gemfibrozil (mean final dose, 1161 mg/d; P < 0.0001). Baseline and posttreatment body mass indexes, hemoglobin A1c, and lipid components are displayed in Table 2
. The changes in lipid and apoprotein concentrations with each drug and their combination are shown in Fig. 1
.
|
Mean LDL size increased only after treatment with gemfibrozil (either alone or combined with atorvastatin; see Table 2
). In the subgroup of 18 patients (40%) who displayed an LDL phenotype B at baseline, the 1.45% increase in LDL size obtained with gemfibrozil was significantly different (P < 0.05) from the absence of change seen after atorvastatin and combination therapy. A shift toward phenotype A was seen in 12%, 47%, and 29% of the patients with phenotype B treated with atorvastatin, gemfibrozil, or combination therapy, respectively (P = 0.07 for atorvastatin vs. gemfibrozil).
After 12-wk treatment, the percentage of patients achieving the targets of LDLc less than 100 mg/dl (2.6 mmol/liter), non-HDLc less than 130 mg/dl (3.36 mmol/liter), triglyceride less than 150 mg/dl (1.7 mmol/liter), and apoB less than 0.8 g/liter were 60%, 65%, 70%, and 26% with atorvastatin, 5%, 16%, 78%, and 5% with gemfibrozil (P < 0.0005 for LDLc and non-HDLc; P < 0.005 for apoB vs. atorvastatin), and 51%, 54%, 78%, and 20% with combination therapy (P < 0.01 for LDLc; P < 0.05 for apoB vs. gemfibrozil). All of the targets were achieved in only 23%, 5%, and 15% of the patients treated with atorvastatin, gemfibrozil, or combined therapy, respectively. In the subgroup with baseline LDLc below 130 mg/dl (3.36 mmol/liter; n = 9), they were reached in 67%, 11%, and 38% of the patients (P < 0.02 for atorvastatin vs. gemfibrozil). In the subgroup with baseline triglyceride above 150 mg/dl (1.7 mmol/liter; n = 17), none of the patients achieved LDLc concentrations below 100 mg/dl (2.6 mmol/liter), non-HDLc below 130 mg/dl (3.36 mmol/liter), or apoB below 0.8 g/liter with gemfibrozil treatment. With atorvastatin, LDLc concentrations less than 100 mg/dl (2.6 mmol/liter) were reached in 53% (P < 0.0001 vs. gemfibrozil), non-HDLc less than 130 mg/dl (3.36 mmol/liter) in 47% (P = 0.0002 vs. gemfibrozil), and apoB less than 0.8 g/liter in 6% (P > 0.05 vs. gemfibrozil); with combination therapy, they were reached in 41% (P = 0.002 vs. gemfibrozil), 47% (P = 0.0002 vs. gemfibrozil), and 12% (P = 0.09 vs. gemfibrozil), respectively. In this same subgroup, a triglyceride level below 150 mg/dl (1.7 mmol/liter) was achieved in a similar proportion of patients during all treatment periods (35%, 59%, and 59% with atorvastatin, gemfibrozil, and combined therapy, respectively; P > 0.05).
Side effects
No new cardiovascular events were suffered by the patients during the study period. Gastrointestinal manifestations, including abdominal discomfort, constipation, loose stools, and nausea, were referred by 6, 11, and 8 of the patients during treatment with atorvastatin, gemfibrozil, and combination therapy, respectively. Drowsiness was suffered by 1 patient taking atorvastatin, and nightmares by another. Muscle aches with normal creatine kinase were reported by 1 patient during treatment with atorvastatin, 2 during gemfibrozil, and none during combination therapy. Only 1 patient dropped out of the study due to unspecific side effects during treatment with gemfibrozil.
Slight increases in liver enzymes were observed in five, two, and two of the patients treated with atorvastatin, gemfibrozil, and combination therapy, respectively, and in creatine kinase in two, three, and four patients, respectively. Nevertheless, no patient experienced an elevation in liver enzymes greater than 3 times the upper normal limit, or an elevation in creatine kinase greater than 10 times the upper normal limit, and no patient was excluded for these reasons.
| Discussion |
|---|
|
|
|---|
According to the recently released Adult Treatment Panel III guidelines of the National Cholesterol Education Program, diabetes is considered a coronary heart disease equivalent, because subjects with diabetes have as high a risk of dying of myocardial infarction as nondiabetic subjects who have already suffered a coronary event (17, 18). For this high risk group, the LDLc target is less than 100 mg/dl (2.6 mmol/liter), which is supported by clinical evidence showing that LDLc is atherogenic (19, 20) and that lowering its concentrations slows the progression of coronary artery disease (5, 21, 22, 23). However, the most typical pattern of lipid abnormalities shown by diabetic patients comprises moderate hypertriglyceridemia, low HDLc, a high proportion of small dense LDL particles, and increased apoB (a crude marker of the number of atherogenic particles) (2, 3, 24). All of these disorders are atherogenic, and results from two clinical trials suggest that the reduction in triglyceride and the increase in HDLc are also linked to a reduction in coronary heart disease (6, 7). On the other hand, both apoB and non-HDLc (a surrogate of the former) (25), seem to be better predictors of outcome than LDLc in several studies (21, 26, 27, 28). Taken together, these findings suggest that whenever a lipid disorder is detected in type 2 diabetic patients, it should be corrected (7). When National Cholesterol Education Program goals are not achieved by lifestyle intervention and improvement of glycemic control, as was the case for the patients included in this study, lipid-lowering drugs should be prescribed. The two main drug groups used in the management of diabetic dyslipidemia are statins and fibrates. Their indications are straightforward when either hypercholesterolemia or hypertriglyceridemia is present, and the treatment of mixed hyperlipidemia has previously been assessed (11, 12). However, optimal treatment in patients with only moderately increased LDLc and normal or moderately increased triglyceride concentrations has, to our knowledge, not been evaluated previously. In the present study atorvastatin was more potent than gemfibrozil in the reduction of LDLc, apoB, and non-HDLc and more effective in achieving treatment goals in all of the subgroups analyzed, including the patients with low risk LDLc concentrations (<130 mg/dl; 3.36 mmol/liter) and those with triglyceride above 150 mg/dl (1.7 mmol/liter). Gemfibrozil was more effective in the reduction of triglyceride and, in agreement with previous studies, improved LDL particle size (8, 12, 22, 29, 30, 31, 32, 33, 34, 35). The lack of triglyceride reduction with atorvastatin could be due to the near-normal baseline triglyceride concentrations and the dose used (36). Despite that the increase in HDLc with gemfibrozil depends on baseline triglyceride concentrations and the ability of fibrates to reduce triglyceride concentration, in the present study HDLc increased similarly with atorvastatin and gemfibrozil, in accordance with the 47% increase obtained in large clinical trials (5, 6, 22, 23, 37). Thus, our results support statins as first line drugs for diabetic patients with LDLc more than 100 mg/dl (2.6 mmol/liter) and triglyceride levels less than 400 mg/dl (4.51 mmol/liter).
The idea of using the combination of a statin and a fibrate is very attractive as a way of improving not only LDLc, but also the other components of diabetic dyslipidemia. As has previously been shown in patients with combined hyperlipidemia (11, 38, 39), in the present study the combination of low doses of a statin and a fibrate was not only highly effective in reducing LDLc, but was also more effective than statin alone in the reduction of triglyceride concentrations and increasing LDL particle size. Thus, it might be the most appropriate treatment, especially in those patients with triglycerides above 150 mg/dl (1.7 mmol/liter). The low efficacy in achieving apoB goals is worthy of note. Although the different cut-off values used for titration and efficacy assessment might have influenced our results, other studies show that apoB is a better predictor of cardiovascular events than LDLc and even non-HDLc, but that equivalent apoB goals are less frequently reached (40). The latter, which is consistent with our findings, might be explained by the fact that type 2 diabetic patients often show hyper-apoB despite normal LDLc concentrations (2). The fact that the combined effect was slightly less than would be expected from the addition of the effects of each single treatment was probably related to the dose administered, which was not titrated, as the individual drugs were. Indeed, titration of the combined therapy would be expected to achieve treatment goals more effectively. Although combination therapy with statins and fibrates increases the risk of myopathy, data from previous, larger studies (12, 39) show that combined treatment is generally well tolerated and safe. Thus, excluding patients with high risk of myopathy, a good therapeutic response can be achieved with an appropriate safety profile.
In conclusion, our data support the idea that in the pharmacological treatment of type 2 diabetic patients with low or moderate risk LDLc concentrations and normal or moderately increased triglyceride concentrations, statins are the choice. However, the addition of a fibrate may be necessary to correct the multiple abnormalities comprising diabetic dyslipidemia. Combined therapy, probably at higher a dose than that used in the present study, may be the most appropriate treatment in these patients, especially when triglyceride concentrations are above treatment goals. However, data from ongoing trials will hopefully solve this controversy.
| Acknowledgments |
|---|
| Footnotes |
|---|
Present address for O.J., R.B., and J.O.-L.: Department of Endocrinology, Hospital Sant Pau, S. Antoni M. Claret 167, Barcelona 08025, Spain.
Present address for A.M.W.: Department of Biochemistry, Hospital Sant Pau, S. Antoni M. Claret 167, Barcelona 08025, Spain.
Abbreviations: apoB, Apolipoprotein B; HDLc, high density lipoprotein cholesterol; LDLc, low density lipoprotein cholesterol.
Received January 30, 2003.
Accepted April 1, 2003.
| 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.-F. Landrier, C. Thomas, J. Grober, H. Duez, F. Percevault, M. Souidi, C. Linard, B. Staels, and P. Besnard Statin Induction of Liver Fatty Acid-Binding Protein (L-FABP) Gene Expression Is Peroxisome Proliferator-activated Receptor-{alpha}-dependent J. Biol. Chem., October 29, 2004; 279(44): 45512 - 45518. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. M. Huss and D. P. Kelly Nuclear Receptor Signaling and Cardiac Energetics Circ. Res., September 17, 2004; 95(6): 568 - 578. [Abstract] [Full Text] [PDF] |
||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |