Gemfibrozil treatment of hypertriglyceridemia: Improvement on fibrinolysis without change of insulin resistance

Gemfibrozil treatment of hypertriglyceridemia: Improvement on fibrinolysis without change of insulin resistance Jing-Ren Jeng, MD, a Chii-Yuan Jeng, MD, b Wayne Huey-Herng Sheu, MD, b May Meei-Shyuan Lee, PhD, c Shyuh-Huei Huang, MD, d and Shyh-Ming Shieh, MD e Taipei, Taiwan, Republic of China

The fibrinolytic and metabolic changes associated with gemfibrozil treatment of hypertriglyceridemia were evaluated in 16 patients with type IV hyperlipidemia by criteria of triglyceride levels >250 mg/dl and total cholesterol levels <220 mg/dl. The plasma triglyceride level was significantly lower (323 _+71 vs 189 +_57 mg/dl; p = 0.000) and high-density lipoprotein cholesterol level significantly higher (33.5 _+4.6 vs 38.0 _+6.7 mg/dl; p = 0.005) after 3 to 4 months of gemfibrozil treatment. However, the glucose and insulin metabolism measured by oral glucose challenge and insulin suppression tests showed no significant changes after gemfibrozil therapy. In contrast, plasma plasminogen activator inhibitor-1 antigen (36.9 + 12.4 vs 2Z3 + 11.4 ng/ml; p = 0.008) and activity (15.5 _+5.5 vs 11.8 _+ 3.0 IU/ml; p = 0.009) and tissue plasminogen activator antigen (13.2 _+4.0 vs 10.4 _+ 3.7 ng/ml; p = 0.007) were significantly depressed, and tissue plasimogen activator activity (0.57 +_0.31 vs 0.69 + 0.38 IU/ml; p = 0.015) was significantly elevated by gemfibrozil. The data indicate that lowering plasma triglyceride and raising high-density lipoprotein cholesterol levels by gemfibrozil treatment also improved the fibrinolytic system without changes of insulin resistance and glucose intolerance in patients with isolated hypertriglyceridemia. (Am Heart J 1997;134:565-71.)

It has been reported that gemfibrozil, a lipid-lowering fibric acid derivative, may decrease early cardiac events and cardiac mortality rates in patients with hypertriglyceridemia. The Helsinki Heart Study was a primary prevention trial to test the hypothesis that gemfibrozil may reduce the incidence of coronary heart disease (CHD) in middle-aged men with dyslipidemia. During the 5year trial, a 34% reduction of CHD events was seen despite the fact that the overall mortality rate was unchanged. 1 However, the difference in the incidence of CHD between the original gemfibrozil and placebo recipients decreased to 16% at the end of the 8 1/2 years of follow-up. 2 In the largest trial of patients with dyslipidemia and CHD for secondary prevention, the incidence of cardiac end points, with or without coronary artery bypass grafting, cardiac mortality rates, or

From the Divisions of °Cardiology, bEndocrinology and Metabolism, and dHematology, the Department of Medicine, Clinical Research Center, Tri-Service General Hospital, the Department of cPublic Health, National Defense Medical Center, eChung-Shan Hospital. Supported by a grant from the Institute of Biomedical Science, Academia Sinica, Taipei, Taiwan, Republic of China. Received far publication Dec. 23, 1996; accepted April 24, 199Z Reprint requests: Shyh-Ming Shieh, MD, Chung-Sban Hospital, No. 1I, Lane 12, Section 4, Jen-Ai Rd., Taipei, Taiwan, 106. Copyright © 1997 by Mosby- Year Book, Inc. 0002-8703/97/$5.00 + 0 4/1/82879

total mortality rates, showed no significant difference between gemfibrozil- and placebo-treated patients. 3 As with all fibrates, there is concern regarding the adverse events 4-8 of gemfibrozil, including dyspepsia, flatulence, abdominal pain, gallstones, headache, vertigo, athralgia, urticaria, impotence, and myositis, although the incidence is low. The greatest benefit of gemfibrozil noted in the Helsinki Heart Study was shown to be derived by the subgroup with an elevated ratio of low-density lipoprotein cholesterol (LDL-C)/high-density lipoprotein cholesterol (HDL-C) and a high triglyceride level. 9 Recent reanalysis of their data discovered that gemfibrozil reduced the risk of CHD mainly in overweight subjects, with additional risk factors known to contribute or predispose to the insulin resistance syndrome. I° It has been proposed that insulin resistance is the primary abnormality, with the increase in plasma triglyceride levels a consequence of the defect in insulin action and subsequent compensatory hyperinsulinemia. 11 However, some postulated that high plasma triglyceride levels might be a condition worsening tissue sensitivity to insulin, and thus hypertriglyceridemia, insulin resistance, and hyperinsulinemia may be related in a vicious cycle. 12 Although it was reported that insulin action and glucose metabolism could be improved by

566

Jeng et al.

Variable Triglycerides (mg/dl) Total cholesterol (mg/dl} HDL-C (mg/dl) Total cholesterol/HDL-C

Before

After

p Value

180 _+57 191 _+ 28 38.0 + 6.7 5.14 _+ 0.98

0.000 0.352 0.005 0.009

Before

After

p Value

105__+ 13 19__+ 18 470 _+ 102 322 + 167 220 + 64 68 _+ 15

104 + 13 18__-+8 4 6 4 -t- 134 316 + 138 217 + 60 70 __+12

0.757 0.205 0.233 0.260 0.277 0.274

323 194 33.5 5.88

_+ 71 -+ 20 + 4.6 + 1.01

Data are expressedas mean+ SD; p valueswere analyzed by paired t test. HDL, High-densilylipoprotein. Valuesare mean+ SD.

Variable Fasting plasma glucose (mg/dl) Fasting plasma insulin (~lU/ml) Integrated glucose response (mg/dl x hr) Integrated insulin response (,I.tlU/ml × hr) Steady-state plasma glucose (mg/dl) Steady-state plasma insulin (,t.tlU/ml)

Data are expressedas mean-+ SD; p valueswere analyzed by paired t test.

gemfibrozil treatment in patients with hypertriglyceridemia, ~3,14 others s h o w e d that gemfibrozil had no effect on insulin resistance. 15,16 In addition, it seems established that impaired fibrinolysis caused by elevated plasminogen activator inhibitor1 (PM-1) is present in patients with excess weight, 17 high waist/hip ratio 18, hypertriglyceridemia,19 high blood pressure, 2° and hyperinsulinemia, 2: which are hallmarks of the insulin resistance syndrome. 22 Among these subjects, interventions such as weight reduction, 23 hypocaloric diets, 24 physical exercise, 25 and treatment with mefformin 26 or doxazosin 27 to improve insulin resistance were found to be associated with decreases in PM-1 levels and amelioration of hypofibrinolysis. However, it has also been shown that insulin sensitivity is not an independent determinant of PM-1 activity,28 challenging the notion of PAl-1 as a component of the insulin resistance syndrome. The effect of gemfibrozil treatment on the fibrinolytic system in patients with hypertriglyceridemia has been studied in the literature, but the results were conflicting. Although some showed no improvement of fibrinolysis despite marked reduction of plasma triglyceride levels, 29-31 others demonstrated favorable changes of the fibrinolytic pattern with gemifibrozil treatment. 32-34 None of them concerned the effect of gemfibrozil on insulin resistance, which may be associated with its changes on the fibrinolytic system. It seems important to clarify whether the protection

from CHD events with gemfibrozii treatment is achieved through modulation of lipid levels only or combined effects on impaired fibrinolysis and insulin resistance. Therefore this study was initiated to investigate the fibrinolytic changes and their relationship to effects on insulin resistance and glucose intolerance by gemfibrozil treatment in patients with high triglyceride levels.

Methods Patients with type IV hyperlipidemia were recruited in this study from the clinic of the Tri-Service General Hospital. They were between 30 and 70 years of age, in general good health, and neither taking medicine nor following any diet aimed at improvement of lipid metabolism. Those with documente d CHD, blood pressure >160/95 mm Hg, and fasting plasma glucose levels >140 mg/dl were excluded. Blood was obtained after an overnight fast on two occasions for measurement of fasting plasma cholesterol and triglyceride concentration, and individuals who satisfied the criteria listed above and also had a 14-hour fasting triglyceride level >250 mg/dl and total cholesterol level <220 mg/dl were eligible for this study. They voiunteered to participate after being informed of our interest in evaluating the effects of gemfibrozil treatment on fibrinolysis and metabolism in people with hypertriglyceridemia. After 4 weeks of Washout phase with a weight-maintenance diet by dietitian instruction, 16 patients were enrolled in the study and admitted to the Clinic Research Center (CRC) for baseline measurements of fibrinolytic variables and glucose, insulin, and lipid

American Heart journal September 1997

Jeng etal.

Variable PAl-1 activity (IU/ml) PAl-1 antigen (ng/ml) tPA antigen (ng/ml) tPA activity (IU/ml)

Before 15.5 36.9 13.2 0.57

_+5.5 _+ 12.4 + 4.0 + 0.31

After

p Value

11.8 + 3.0 2Z3 _+ t l . 4 10.4 _+ 3.7 0.69_+0.38

0.009 0.008 0.007 0.015

Data are expressedas mean+ SD; p valueswere analyzedby paired t test.

metabolism. There were eight m e n and eight w o m e n with an age range of 51 + 10 years, body mass index of 25.8 + 2.8 k g / m 2, blood pressure of 129/84 + 12/9 m m Hg, and fasting plasma glucose level of 105 + 12 mg/dl. Six of the subjects were smokers. Their baseline plasma triglyceride levels ranged from 256 to 498 mg/dl and total cholesterol levels from 168 to 220 mg/dl. For fibrinolytic measurement, blood was drawn at about 9 AM after a 15-minute rest and an overnight fast to reduce interference by the diurnal variation 35 of PAI-1 and tissue plasminogen activator (TPA). Samples were collected in Stabilyte tubes (Biopool AB, Umefi, Sweden) containing citrate buffer (pH 4.5) to ensure stability of TPA activity36 without affecting the assay of PAI-t activity. 37 Plasma was separated within 1 hour by centrifugation for 20 minutes at 3000g and stored at -70 ° C until assay. Plasma PAI-1 and tPA antigens were measured quantitatively with enzyme-linked immunosorbent assays. 3s,39 Plasma TPA activity was determined with a parabolic rate assay based on fibrin stimulation of the TPA-catalyzed conversion of Glu-plasminogen to plasmin, which subsequently cleaves the chromogenic substrate. 4° The TPA activity was expressed in international units per milliliter by reference to the World Health Organization First International Standard and Control, Potter Bar, England. Plasma PAI-1 activity was determined with a two-stage, indirect enzymatic assay based on the addition of excess TPA (40 IU/ml) to the samples and measurement of the residual TPA activity. 41 One unit of PAI-1 activity was defined as the amount of PAI-1 that inhibits 1 IU of international TPA standard. The reagent kits for assay of TPA and PAI-1 antigens and activities were purchased from Biopool AB. For metabolic measurement, blood was obtained after an overnight fast for assay of fasting plasma glucose, insulin, total cholesterol, and triglycerid e concentrations. HDL-C concentration was measured after precipitation of apoliprotein B-containing lipoproteins by phosphotungstic acid and magnesium chloride reagent. In addition, plasma glucose and insulin concentrations were also determined 30, 60, 90, 120, and 180 minutes after the administration of a 75 gm oral glucose load. The integrated glucose and insulin responses to oral glucose load were calculated from the total area under curve of the plasma glucose-time and insulin-time plots, respectively. The integrated glucose response in healthy adults is 375 + 40 (270 to 454) mg x hours/dl x hours based

American Heart Journal Volume 134, Number 3

on the data pool of our laboratory. The fasting insulin level is normal for values <10 to 15 btU/dl, abnormal if >30 glU/dl, and supernormal between them. The ability of insulin to promote glucose uptake was estimated by a modification 16,2°,27 of the insulin suppression test originally described by Shen et al. 42 After an overnight fast, intravenous catheters were placed in each arm. Blood was sampled from one arm for measurement of plasma glucose and insulin concentration and the contralateral arm was used ' for the administration of test substances. Somatostatin (Bachem, Inc., Torrence, Calif.) was administered by Graseby infusion pump (Graseby Medical Ltd., Warlord, Herts, U.K.) at 350 btg/hr in a solution containing 0.5% h u m a n serum albumin to suppress endogenous insulin secretion. Simultaneously, insulin and glucose were infused at 25 mU/m2/min and 240 mg/m2/min, respectively. Blood was sampled and plasma glucose and insulin concentrations were determined every 30 minutes until 150 minutes into the study and then every 10 minutes until 180 minutes had elapsed. The latter four values obtained from 150 to 180 minutes were averaged to determine the steady-state plasma glucose (SSPG) and insulin (SSPI) concentrations achieved during the infusion. According to our data pool, those with SSPG <130 mg/dl are insulin sensitive, SSPG >190 mg/dl are insulin resistant, and 130 > SSPG > 190 mg/dl have supernormal insulin sensitivity. The minimally clinically significant difference of SSPG was considered to be 10 to 20 mg/dl. The sample size (N d) of study that needs to be used to detect the difference was estimated as (Zc~ + Zf~)2 (Jd2/Sd 2 = (1.96 + 1.28) 2 122/102 = 15, where Zc~ = 1.96 for two-sided 5% significance level (0~ = 0.05), Zf~ = 1.28 for 90% power (1-[3), o d = standard deviation of difference, and 8 d = minimal clinically significant difference. After completion of the baseline studies, patients were treated with gemfibrozil at a dose of 300 mg twice a day and were discharged from the CRC. At the end of 4 weeks of medication, plasma triglyceride levels were determined. Ten patients with plasma triglyceride levels <180 mg/dl were prescribed the same dosage for another 8 weeks. Six patients with plasma triglyceride levels >180 mg/dl were given 600 mg twice a day for another 12 weeks. Three months later, all patients were again admitted to the CRC, and the baseline studies were repeated. All data are expressed as mean + standard deviation. A

567

568

Jengetak

paired t test was used to evaluate the significance of differences between lipid levels, integrated glucose and insulin responses, sSPG and SSPI concentrations, and fibrinolytic variables before and after gemfibrozil treatments.

Results As expected, these 16 patients with type IV hyperlipidemia showed significantly lower HDL-C levels (34 + 5 vs 53 -+ 7 mg/dl), higher triglyceride (323 + 71 vs 94 -+ 42 mg/dD and fasting insulin levels (!9 -+ 8 vs 10 _+4 btIU/ml), greater integrated glucose (470 _+ 102 vs 374 + 40 mg x hr~dD and insulin (322 + 167 vs 172 + 100/.tIU × hr/mD responses to a 75 gm oral glucose challenge, and elevated SSPG (220 _+ 64 vs 120 _+ 40 mg/dD concentrations compared with the normal data from our laboratory. It indicated that these patients with dyslipidemia also have some degree of insulin resistance, hyperinsuiinemia, and glucose intolerance. In addition, they showed significantly higher PAI-1 activity (15:5 + 5.5 vs 8.6 _+5.4 iU/mD, PAId antigen (36.9 -+ 12.4 vs 13.2-+ 7.5 ng/mD, and TPA antigen (13.2 -+ 4.0 vs &8 -+ 3.6 ng/ml) and significantly lower TPA activity (0.57 -+ 0.31 vs 0.96 _+ 0.35 IU/mD compared with data from normal control subjects in our laboratory. Thus these patients had hypofibrinolysis as a result of increased basal PAI-1 levels before treatment. The data of fasting plasma lipid levels before and after gemfibrozil treatment in 16 patients with type IV hyperlipidemia are listed in Table I. It can be noted that plasma triglyceride levels decreased significantly (p = 0.000) and plasma HDL-C levels and the ratio of total cholesterol/HDL-C increased significantly (p = 0.005 and 0.009, respectively) after 3 to 4 months of gemfibrozil treatment. In contrast, the normal baseline data of total cholesterol concentrations were not influenced significantly by gemfibrozil. The effect of gemfibrozil treatment on glucose and insulin metabolism measured by oral glucose challenge and insulin suppression tests in 16 patients with isolated hypertriglyceridemia is shown in Table II. It is apparent that fasting plasma glucose and insulin levels, integrated glucose and insulin responses, and steadystate plasma glucose and insulin concentrations were not significantly changed after 3 to 4 months of treatment with gemfibrozil. Although the power calculation for SSPG is only 17% from Z[~ = N d 8d/(7 d) - Z~z = (4 x 3/12) - 1.96 = -0.96, the difference of 3 mg/dl in SSPG data after gemfibrozil treatment is of no clinical significance. It is meaningless to increase the sample size, N d = (act 4- Z~) 20"d2/~d2 = (1.96 + 1.28) 2 122/32 = 168, and

obtain 90% power to get a statistically significant but clinically insignificant result. Hence, our result indicated that gemfibrozil treatment did not affect glucose and insulin metabolism in patients with type IV hyperlipidemia. The changes of the fibrinolytic system associated with gemfibrozil treatment in 16 patients with type IV hyperlipidemia are demonstrated in Table III. It can be seen that plasma PAI-1 activity and antigen were lowered significantly (p = 0.009 and 0.008, respectively) after 3 to 4 months of treatment with gemfibrzoil in patients with isolated hypertriglyceridemia. On the other hand, plasma TPA antigen decreased significantly (p = 0.007) and plasma TPA acitivity increased significantly (p = 0.015) after gemfibrozil treatment.

Discussion This study, consistent with the combined data from other reports, 15,16,32-34 showed that fibrinolytic activity could be improved modestly but significantly by gemfibrozil treatment with insignificant changes of insulin and glucose metabolism in patients with hypertriglyceridemia w h o have some degree of insulin resistance and glucose intolerance. The decrease in plasma triglyceride levels and increase in HDL-C levels by gemfibrozil is not accompanied by reduction of insulin resistance but is associated with lowering of PAI-1 levels and an increase of TPA activity. The protection from CHD with genifibrozil treatment might be achieved through its favorable effects on hypofibrinolysis and dyslipidemia rather than effects on insulin resistance, hyperinsulinemia, and glucose intolerance. Furthermore, the results of this study may support the hypothesis that high plasma triglyceride levels are a consequence rather than a cause of insulin resistance and also showed that PAI-1 levels can be changed independently of insulin sensitivity. Because elevated PAI-1 levels increase risk of thrombus formation and myocardial infarction, gemfibrozil treatment of patients with high plasma triglyceride levels might reduce the risk of events of CHD regardless of whether insulin resistance is present. Although elevated PAI-1 levels with hypofibrinolysis have been considered a part of the insulin resistance syndrome, the mechanisms of its increase remain unclear. In vitro studies indicated that PAI-1 synthesis in hepatocytes and endothelial cells could be enhanced by insulin, 43 proinsulin, 44 insulin-like growth factor-I, 43 LDL, 45 modified LDL,46 and very low-density lipoproteins, 47 all of which are candidates for triggering PAI-1 elevation in vivo. It is probable, from our observation,

American Heart Journal September 1997

Jeng etal.

that the profibrinolytic action of gemfibrozil may be assigned in part to the reduction of hypertriglyceridemia because of no changes in insulin and glucose metabolism. Our data may also support the hypothesis that the elevated PAI-1 levels in plasma of hypertriglyceridemia are the result, at least in part, of an interaction of very low-density lipoproteins with endothelial or hepatic cells. Despite the fact that, ultimately, the precise mechanisms could not be determined from this study, it warrants attention that gemfibrozil may also have intrinsic profibronolytic action in vivo because its direct effects on the inhibition of PAI-1 synthesis in hepatocytes and endothelial cells can be demonstrated clearly by in vitro studies. 48,49 Avellone at al. 32 have reported in serial studies that gemfibrozil treatment in type IV hyperlipidemia without thromboembolic disease was associated with the reduction of euglobulin lysis time and basal levels of PAI-1 activity33 and TPA antigen 34 and an increased release of TPA antigen in response to venous occlusion.34 However, others reported that PAI-1 levels and TPA antigens were unchanged by gemfibrozil in patients after myocardial infarction 29 or with a history of venous thrombosis. 3° Recently Br6ijersen et al. 31 published data demonstrating that lowering the plasma triglyceride concentration with gemfibrozil did not change PAI-1 activity and TPA levels in type lib hyperlipidemia with coexisting cardiovascular diseases. In our data for type VI hyperlipidemia without cardiovascular disease, basal PAI-1 levels and TPA antigen could be reduced and basal TPA activity elevated after gemfibrozil treatment. The fibrinoiytic system is regulated by the balance between the levels of TPA and PAI-1, and an increase in basal PAI-1 has been proposed as a good marker of hypofibrinolysis. The reduction of PM-1 levels by gemfibrozil may be expected to improve fibrinolytic activity as shown by the increase of basal TPA activity. The reduction of basal TPA antigen, which has been inactivated by PAI-1, may suggest that the decreased deposit of TPA in endothelial cells caused by slow, continuous release from the vessel wall in hypertriglyceridemia could be improved by gemfibrozil therapy. It has been shown that insulin resistance, assessed by either insulin-induced whole-body glucose disposal 15 or insulin-mediated glucose uptake, 16 was not changed greatly by gemfibrozil treatment in nondiabetic patients with primary hypertriglyceridemia. Steiner 13 reported that gemfibrozil reduced plasma insulin response during oral glucose tolerance testing in relation to the lowering of plasma triglyceride levels despite the fact that

American Heart Journal Volume 134, Number 3

glucose response was unchanged. Fasting blood glucose levels were not changed or were reduced slightly by gemfibrozil in a trial with a large number of patients. 5° Studies for patients with non-insulin-dependent diabetes mellitus and dyslipidemia have also produced conflicting results but showed gemfibrozil to have no detrimental effects on glucoregulatory effect of insulin. Recently Avogaro et al. 14 employed the insulin tolerance test (Kitt) for estimation of insulin action and found that gemfibrozil reduced fasting and postprandial glucose levels as a result of improvement of insulin action in patients with hypertriglyceridemia with and without non-insulin-dependent diabetes mellitus. However, in our nondiabetic patients, after lowering plasma triglyceride levels with gemfibrozil, no significant changes in fasting plasma levels of glucose and insulin and degrees of glucose tolerance and insulin sensitivity could be observed. In conclusion, gemfibrozil treatment of patients with isolated hypertriglyceridemia led to a significant decrease in plasma triglyceride levels and a rise of HDL-C levels, which were associated with modest but significant improvement of hypofibrinolysis and insignificant changes in insulin resistance, hyperinsulinemia, and glucose intolerance. The results may support the hypotheses that high plasma triglyceride levels are a consequence rather than a cause of insulin resistance and the presence of a partial link between hypertriglyceridemia and impaired fibrinolysis as a result of elevated PAI-1 levels. In addition, it is also suggested that gemfibrozil, given its beneficial fibrinolytic and lipid-lowering effects, would seem to be a logically useful agent that might reduce the risk for events of CHD in patients with type IV hyperlipidemia. We thank Dr. H-C Kwaan, Professor of Medicine at Northwest University, for his suggestion and instruction about the fibrinolytic measurement and Miss S-C Lee for her technical assistance.

References 1. Frick MH, Ela O, Haapa K, et al. The Helsinki Heart Study: primary prevention trial with gemfibrozil in middle-aged men with dyslipidemia: safety of treatment, changes in risk factors, and incidence of coronary heart disease. N Engl j M ed 19871317:1237-45. 2. Heinonen OP, Hutlunen JK, Manninen V, et ak The Helsinki Heart Study: coronary heart disease incidence during an extended followup. j Intern Med 1994;235:41-9. 3. Frick MH, Heinanen OP, Hutlunen JK, et d. Efficacy of gemfibrozil in dyslipidaemic subjects with suspected heart disease: anaJysis study in the Helsinki Heart Study frame popuJation. Ann Med 1993;25:41-5.

569

$70

Jeng etal.

4. Knipscheer HC, Nurmohamed MT, van den Ende A, et al. Gernfibrozil treatment of the high trigJyceride-low high-density lipoprorein cholesterol trait in men with established atherascJerosis.J Intern Med 1994;236:377-84. 5. Stelmach WJ, Rush DR, Brucker PC, et ak Diet and exercise and gemfibrozi/therapy for the management of dysJipidemia: ACEN study, j Farn Pract ]993;36:401-8. 6. Frick MH, EIo MO, Haapa K, et ak He[sinki Heart Study: primary-prevention trial with gernfJbrozil in middle-aged men with dyslipidernia: safety of treatment, change in risk factors, and incidence of coronary heart disease. N Engl J Med 1987;317:1237-45. Z Bharani A. Sexual dysfunction after gemfibrozil. Br Med J 1992;305:693. 8. Magarian GL, Lucas [M. Gemfibrozil-induced myopathy. Arch Intern Med 1991;151:18734. 9. Manninen V, Tenkanen L, Koskinen P, et al. Joint effects of serum triglyceride and LDL cholesterol and HDL cholesterol concentrations of coronary heart disease risk in the He[sinki Heart Study: implication for treatment. Circulation 1992;85:37-45. 10. Tenkanen L, Mentt@ri M, Manninen V. Some coronary risk factors related to the insulin resistance syndrome and treatment with gemfibrazik experience from the Hdsinki Heart Study. Circulation 1995;92:177985. 11. OlefskyjM, Farquhar jW, Reaven GM. Reappraisal of the role of insulin in hypertriglyceridernia. Am j Med 1974;57:551-60. 12. Steiner G, Vranic M. Insulin and hypertrigJyceridernia: a vicious cycle with atherosclerafic potential. Int J Obes 1982;6(suppl):117-24. 13. Steiner G. Altering triglyceride concentrations changes insulin-glucose relationships in hypertriglyceridemic patients: double-blind study with gemfibrozil with implications for atherosclerosis. Diabetes Care 1991;]4:1077-81. 14. Avogaro A, Beltrarnello P, Marin R, et al. Insulin action and glucose metabolism are improved by gemfibrozil treatment in hypertriglyceridemic patients. Atherosclerosis 1995;113:117-24. 15. Sane T, Knudsen P, Vuorinen-Markkda H, Yki-Jarvinen H, Taskinen MR. Decreasing triglyceride by gemfibrozil therapy does not affect the glucoregulatory or antiJipolytic effect of insulin in nondiabetic subjects with mild hypertrJglyceridemia. Metabolism 1995;44:589-96. 16. Jeng CY, Sheu WHH, Fuh MMT, Shieh SM, Chen YDI, Reaven GM. Gemfibrozil treatment of endogenous hypertriglyceridemia: effect on insulin-mediated glucose disposal and plasma insulin concentrations. J Clin Endocrinol Metab 1996;81:2550-3.. 1Z Vague P,Juhan-Vague I, Aillaud MF, et al. Correlation between blood fibrinolytic activity, p[asrninogen activator inhibitor level, plasma insulin level, and relative body weight in normal and obese subjects. Metabolism 1986;35:250-3. 18. Landin K, Stigendahl L, Eriksson E, et al. Abdominal obesity is associated with an impaired fibrinolytic activity and elevated plasminogen activator inhibitor-]. Metabolism 1990;39:1044-8. 19. Asplund-Carlson A, Hamsten A, Wiman B, Carlson LA. Relationship between plasma plasminogen activator inhibitor-1 activity and VLDL triglyceride concentration, insulin levels and insulin sensitivity: studies in randomly selected normo-and hypertriglyceridaemJc men. Diabetologia ]993;36:817-25. 20. Jeng jR, Sheu WHH, jeng CY, Huang SH, Shieh SM. Impaired fibrinolysis and insulin resistance in patients with hypertension. Am J Hypertens 1996;9:484-90. 21. Juhan-Vague I, Roul C, Alessi MC, Ardissone JP, Helm M, Vague P. Increased plasminogen activator inhibitor activity in non-insu[in-depen-

dent diabetic patients: relationship with p[asma insulin. Thromb Haemost 1989;61:370-3. 22. Vague P, Raccah D, Scelles V. Hypofibronolysis and the insulin resistance syndrome. Int J Obes 1995;19(su pp[):S 1] -5. 23. Sundell IB, DaHIgren S, Tanby M, Lundin E, Stenling R, Nilsson TK. Reduction of elevated plasrninogen activator inhibitor levels during modest weight loss. Fibrindysis 1989;3:51-31 24. Mehrabian M, PeterJB, Barnard RJ, LusisAJ. Dietary regulation of fibrindytic factors. Atherosclerosis 1990;84:25-32. 25. Rosentha/M, Haskdl Wl, Solomon R, Widstrom A, Reaven G. Demonstration of a relationship belween the level of physical training and insulin stimulated glucose utilization in normal human. Diabetes 1983;32:408-11. 26. Vague P,Juhan-Vsague I, Alessi MC, Badier C, Valadier j. Metformin decreases the high plasminogen activatory inhibition capacity, plasma insulin and triglyceride levels in nandiabetic obese subjects. Thromb Haemost 1987;57:326 8. 2Z jeng JR, Sheu WHH, Jeng CY, Huang SH, Shieh SM Effect of doxazosin on fibrindysis in hypertensive patients with and without insulin resistance. Am Heart J 1996; 132:783-9. 28. Mykk@nen L, R6nnemaa T, MarniemiJ, Haffner SM, Bergrnan R, Laakso M. Insulin sensitivity is not an independent determinant of plasma plasminogen activator inhibitor-1 activity. Arierioscler Thromb 1994; 14:1264-71. 29. Andersen P, Smith P, Seljeflot [, Brataker S, Arnesen H. Effect of gemfibrozil on [ipids and haemostasis after myocardial infarction. Thrornb Haemost 1990;63:174-Z 30. Haire WD. Gemfibrozil predictably lowers triglycerides but does not significantly change plasminogen activator inhibitor activity in hypertriglyceridemic patients with a history of thrombosis. Thromb Res 1991 ;64:493-501. 31. Br6ijersen A, Eriksson M, Wiman B, Angelin B, Hjemdahl P. Gernfibrozil treatment of combined hyperlipoproteinernia: no improvement of fibrindysis despite marked reduction of plasma triglyceride levels. ArterJosder Thromb Vasc Bid 1996;16:511-6. 32. Avellone G, Di Garbo V, Panno AV, Cordova R, Lepore R, Strana A. Changes induced by gemfibrozil on lipidic, coagulative and fibrinolytic pattern in patients with type IV hyperlipoproteinemia. Int Angiol 1988;7:270-Z 33. Avellone G, Di Garbo V, Cordova R, Raneli G, D'e Sir'none R, Bompiani GD. Effect of gemfibrozil treatment on fibrinolysis system in patients with hypertriglyceridemia. Curr Ther Res 1992;52:338-48. 34. Avellone G, Di Ga!bo V, Cordova R, et al. Improvement of fibrinolysis and plasma lipoprotein levels induced by gemfibrozil in hypertriglyceridemia. Blood Coagul Fibrinolysis 1995;6:543 8. 35. Kluft C, Jie AFH, Rijken DC, Verheijen JH. Daytime fluctuations in blood of tissue-type plasminogen activator (t-PA) and its fast acting inhibitor (PAl-I).Thromb Haemost 1989;59:329-32. 36. RdmbyM, Sundell IB, Nilsson TK. Blood collection in strong acidic citrate anticoagulant used in a study of dietary influence on basal tPA activiV. Thromb Haernost 1989;62:917-22. 3Z Eriksson E, Tengbom L, Risberg B. The effect of various anticoagulant/antipJatelet mixtures on determination of plasminogen activator inhibitor, platelet proteins and hemostasis parameters. Thromb Haemost 1989;61:511-6. 38. RdmbyM, Bergsdorf N, Nilsson T, Meflbring G, Winblad B, Bucht G. Age dependence of tissue pIasminogen activator concentrations in plasma, as studied by an improved enzyme linked immunosorbent assay. Clin Chem 1986;32:2160-5.

American Heart Journal September 1997

Jeng etal.

39. Declerck PJ,Alessi MC, Verstreken M, Kruithaf EKO, Juhan-Vague I, Cdlen E. Measurement of plasminogen activator inhibitor-1 in biolog: ic fluids with a murine monoclOnal antibody based enzyme-linked irnmunosorbent assay. Blood 1988;71:220-5. 40. Wilman B, Mellbring G, R6nby M. Plasminogen activator release during venous stasis and exercise as determined by a new specific assay. Clin Chim Acta 1983;127:279-88. 41. Chimielewska J, Ranby M, Wiman B. Evidence for a rapid inhibitor for tissue plasminogen activator in plasma. Thromb Res 1983;3] :42736. 42. Shen SW, Reaven GM, Farquahar J. Comparison of impedance to insulin mediated glucose uptake in normal subjects and in subjects with latent diabetes, j Clin Invest 1970;49:2151-60. 43. Anfosso F, Alessi MC, Nalbone G, Chomiki N, Henry M, juhanVague I. Up-regulated expression of plasminogen activator inhibitor-1 in Hep G2 cel[s: interrelationship be~een insulin and insulin-like growth factor 1. Thromb Haemost 1995;73:288-74. 44. Nordl TK, Schneider DJ, Sobel BE. Augmentation of the synthesis of

American Heart journal Volume 134, Number 3

45. 46.

4Z 48.

49.

50.

plasminogen activator !nhibitor type-1 by precursors of insulin: a potential risk factor for vascular disease. Circulation 1994;89:321-30. Mussoni L, Madema P, Camera M, et al. Atherogenic lipoproleins and PAl-1 releases by endothelia cells. Fibrinolysis 1990;4(suppl):79-81. Latron Y, Chautan M, Anfossa F, et al. Stimulation effect of oxidized low density IJpoproteins on plasminogen activator inhibitor 1 synthesis by endothelial cells. Fibrinolysis 1993;7:171-6~ Mussoni L, Mannucci L, Sirlori M, et ak Hypertriglyceridemia and regulation of fibrJnolytic activity. Arteriosc[er Thramb 1992; 12:19-2Z ffujii S, Sobel BE. Direct effects of gemfibrozil on the fibrinolylic syslem: diminution of synthesis of plasminogen activator inhibitor type 1. Circulation 1992;85:1888-93. Fujii S, Sawa H, Sobd BE. Inhibition of endothelial cell expression of plasminogen activator inhibitor Vpe-1 by gemfibrozJl. Thromb Haemost 1993;70:642-Z Descovich GC, D'Addato S, Benassi B, et ak Effects of decreasing doses of gemfibrozil in middle-aged and elderly dyslJpidemic patients: a mu[ticenler study. Adv Ther 1993; 10:272-8Z

STI