Can LDL cholesterol be too low? Possible risks of extremely low levels.

Following the continuous accumulation of evidence supporting the beneficial role of reducing low-density lipoprotein cholesterol (LDL-C) levels in the treatment and prevention of atherosclerotic cardiovascular disease and its complications, therapeutic possibilities now exist to lower LDL-C to very low levels, similar to or even lower than those seen in newborns and nonhuman species. In addition to the important task of evaluating potential side effects of such treatments, the question arises whether extremely low LDL-C levels per se may provoke adverse effects in humans. In this review, we summarize information from studies of human cellular and organ physiology, phenotypic characterization of rare genetic diseases of lipid metabolism, and experience from clinical trials. Specifically, we emphasize the importance of the robustness of the regulatory systems that maintain balanced fluxes and levels of cholesterol at both cellular and organismal levels. Even at extremely low LDL-C levels, critical capacities of steroid hormone and bile acid production are preserved, and the presence of a cholesterol blood-brain barrier protects cells in the central nervous system. Apparent relationships sometimes reported between less pronounced low LDL-C levels and disease states such as cancer, depression, infectious disease and others can generally be explained as secondary phenomena. Drug-related side effects including an increased propensity for development of type 2 diabetes occur during statin treatment, whilst further evaluation of more potent LDL-lowering treatments such as PCSK9 inhibitors is needed. Experience from the recently reported and ongoing large event-driven trials are of great interest, and further evaluation including careful analysis of cognitive functions will be important.

HDL-C levels modify the association between C-reactive protein and coronary artery calcium score.

BACKGROUNDS AND AIMS: C-reactive protein (CRP) levels predict incident and recurrent cardiovascular disease (CVD) events; however, associations between CRP and pre-clinical atherosclerosis is less certain. Since high concentrations of high-density lipoprotein cholesterol (HDL-C) are inversely associated with CVD risk, we investigated whether HDL-C modified the association between CRP concentration and measures of preclinical atherosclerosis. METHODS AND RESULTS: Data were analyzed from a Korean occupational cohort of 12,030 male subjects who underwent a cardiac computed tomography (CT) estimation of coronary artery calcification (CAC) score and an assessment of CVD risk factors. Logistic regression was used to describe associations between CRP and measures of pre-clinical atherosclerosis, such as CAC scores >0. As many as 1351 (11.2%) participants had a CAC score>0. CRP was stratified into 3 groups based on clinical category: <1 mg/L, 1 to <2 mg/L, and ≥ 2 mg/dL. In the bottom CRP group, 907/8697 (10.4%) of subjects had a CAC score >0, compared with 242/1943 (12.5%) in the middle group and 202/1396 (14.5%) in the top CRP group (p < 0.0001). After adjustment for multiple CVD risk factors, there was a positive association between CRP and CAC score>0 (OR between top and bottom CRP groups, 1.41 [1.04, 1.90], p = 0.027) in the lowest HDL-C quartile but not in the highest HDL-C (OR between top and bottom CRP group, 0.80 [0.46, 1.39], p = 0.425). CONCLUSION: The association between CRP concentration and CAC score differed according to HDL-C levels.

Prevalence of low LDL-cholesterol levels and elevated high-sensitivity C-reactive protein levels in apparently healthy Korean adults.

BACKGROUND AND AIMS: The Justification for the Use of Statins in Primary Prevention: An Intervention Trial Evaluating Rosuvastatin (JUPITER) reported reduced cardiovascular and all-cause mortality in patients with elevated C-reactive protein (CRP) and low LDL-cholesterol (LDL-C) levels treated with statins. The aims of this study were to determine the proportion of "JUPITER-eligible" Korean adults and to describe their characteristics. METHODS AND RESULTS: As many as 15,154 subjects with serum LDL-C levels <130 mg/dL were selected among 28,851 middle-aged participants (men ≥ 50 years, women ≥ 60 years) who participated in a routine health check-up program. Among the participants with LDL-C less than 130 mg/dL, only 15% had CRP levels ≥2.0 mg/L (7.9% of original participants). Subjects were divided into four groups according to CRP levels (<0.5, ≥0.5 - <1.0, ≥1.0 - <2.0, and ≥2.0 mg/L). Mean HDL-C and apolipoprotein A1 levels decreased significantly as the mean CRP values increased. The insulin and homeostasis model of insulin resistance was significantly different according to CRP quartile. The number of subjects with metabolic syndrome and its components increased significantly as the mean CRP values increased. CONCLUSION: In this Asian population, few individuals with low LDL-C levels had CRP levels ≥2.0 mg/L. Elevated CRP levels were associated with components of atherogenic dyslipidemia and insulin resistance. Additional clinical trials should be designed and performed in different ethnic groups, as different CRP cut-off levels may be required in different ethnic groups.

Factors associated with time to initiation of a PCSK9 inhibitor after hospital discharge for acute myocardial infarction.

BACKGROUND: Proprotein convertase subtilisin/kexin type 9 inhibitors (PCSK9i) lower atherosclerotic cardiovascular disease (ASCVD) event risk. OBJECTIVE: Analyze patient characteristics associated with time to PCSK9i initiation following an acute myocardial infarction (AMI). METHODS: We analyzed characteristics of patients ≥21 years of age in the Marketscan or Medicare databases who initiated a PCSK9i 0-89 days, 90-179 days, or 180-365 days after an AMI between July 2015 and December 2018 (n=1,705). We estimated the cumulative incidence of recurrent ASCVD events before PCSK9i initiation. RESULTS: Overall, 42%, 25%, and 33% of patients who initiated a PCSK9i did so 0-89 days, 90-179 days, and 180-365 days following AMI hospital discharge, respectively. Taking ezetimibe prior to AMI hospitalization and initiating ezetimibe within 30 days after AMI hospital discharge were each associated with a higher likelihood of PCSK9i initiation in the 0-89 days versus 180-365 days post-discharge (adjusted odds ratio [OR] 1.83, 95% confidence interval [95%CI] 1.35-2.49 and 1.76, 95%CI 1.11-2.80, respectively). Statin use before and statin initiation within 30 days after AMI hospitalization were associated with a lower likelihood of PCSK9i initiation 0-89 days versus 180-365 days post-discharge (adjusted OR 0.64, 95%CI 0.49-0.84 and 0.39, 95%CI 0.28-0.54, respectively). Overall, 8.0%, 10.5%, and 12.5% of patients had an ASCVD event at 90, 180, and 365 days following AMI hospital discharge and before initiating a PCSK9i, respectively. CONCLUSION: Among patients initiating a PCSK9i after AMI, a low proportion did so within 89 days of hospital discharge. Many patients had a recurrent ASCVD event before treatment initiation.

Myocardial injury: the acute phase response and lipoprotein metabolism.

Myocardial infarction and other types of tissue injury generate changes in plasma proteins known as the acute phase response. Variations in lipid and lipoprotein levels after acute myocardial infarction are manifest within 24 to 48 h after the onset of chest pain. Maximal postinfarction reductions in total cholesterol occur at days 4 to 5 with levels 47% below baseline; low and high density lipoprotein cholesterol fractions decrease to their nadir on day 7 to concentrations that are 48% and 32% below baseline, respectively. Triglyceride levels increase after acute myocardial infarction to a maximal level that is 58% above baseline on day 7. These alterations in lipid and lipoprotein levels generally stabilize by 2 months after the acute event. Screening for dyslipidemias in survivors of myocardial infarction requires clinical decision-making based on accurate and reliable measurements. The clinician must be familiar with characteristic changes in acute phase lipids and lipoproteins to ensure that patients receive appropriate, potentially life-saving therapy.

Low levels of high-density lipoprotein cholesterol (hypoalphalipoproteinemia). An approach to management.

Clinical management of dyslipidemias has focused primarily on the low-density lipoprotein cholesterol (LDL-C) fraction; however, lipid disorders accompanied by low levels of high-density lipoprotein cholesterol (HDL-C) (hypoalphalipoproteinemia) are common, particularly among subjects with the diagnosis of coronary artery disease prior to age 55 years. The therapeutic objectives for high-risk subjects with dyslipidemias is directed initially toward reduction of the LDL-C fraction; thereafter, aggressive efforts aimed at raising the HDL-C fraction may be warranted. Strategies for raising the HDL-C fraction start with hygienic measures that include aerobic exercise, weight loss, smoking cessation, withdrawal of agents secondarily lowering HDL-C, and estrogen replacement. Pharmacotherapy selected according to the dyslipidemia that accompanies the HDL-C disorder is indicated for subjects who manifest premature coronary artery disease or who have a familial history of coronary artery disease and hypoalphalipoproteinemia.

Beyond low-density lipoprotein cholesterol. A perspective on low high-density lipoprotein disorders and Lp(a) lipoprotein excess.

Evidence supports the involvement of 2 common dyslipidemias---low high-density lipoprotein disorders and Lp(a) lipoprotein excess--in coronary heart disease. Until clinical trials determine whether specific therapeutic interventions can prevent the occurrence and recurrence of coronary heart disease in patients with these dyslipidemias, the implementation of cholesterol-lowering guidelines can provide a reasonable way to manage low high-density lipoprotein disorders and to identify specific categories of patients who may be at particularly high risk for premature coronary heart disease. Empiric treatment guidelines are suggested for low high-density lipoprotein disorders and Lp(a) lipoprotein excess in order to foster further discussion and validation by clinical trial data.

Lipoprotein Lp(a) excess and coronary heart disease.

Lipoprotein Lp(a) excess has been identified as a powerful predictor of premature atherosclerotic vascular disease in several large, prospective studies. Lipoprotein Lp(a) levels modulate the risk of coronary heart disease in patients with hypercholesterolemia, and lipoprotein Lp(a) excess is commonly detected in men and women with premature coronary atherosclerosis. Lipoprotein Lp(a) contributes to atherothrombotic risk by multiple mechanisms that include impaired fibrinolysis, increased cholesterol deposition in the arterial wall, and enhanced oxidation of low density lipoprotein cholesterol. Although low density lipoprotein cholesterol reduction is the primary intervention in patients with lipoprotein Lp(a) excess, specific therapy to lower lipoprotein Lp(a) may be indicated for patients with premature coronary atherosclerosis, a strong family history of premature atherosclerosis, or refractory hypercholesterolemia. In consideration of the high prevalence of lipoprotein Lp(a) excess in patients with premature coronary heart disease and the intricate role of lipoprotein Lp(a) in atherothrombosis, this review provides an evidence-based approach to the screening and treatment of patients with lipoprotein Lp(a) excess.

Postprandial changes in plasma and serum viscosity and plasma lipids and lipoproteins after an acute test meal.

The influence of a fat-rich test meal on postprandial changes in plasma viscosity and serum viscosity was assessed in 12 normolipidemic adults. After a 12-14-h fast, volunteers (five men and seven women aged 23-50 y) were challenged with a test milk shake containing 50 g fat/m2 body surface area (BSA). Plasma viscosity, serum viscosity, and plasma lipids and lipoproteins were assessed at 0, 2, 3, 4, and 6 h. Viscosity values were determined by using a Mettler Contraves LS-40 rotational microviscometer. Postprandial changes in the study variables were assessed by area under the curve and included triacylglycerols (2.02 mmol/L), plasma viscosity (-0.10 mPa.s), and serum viscosity (-0.01 mPa.s). Peak plasma triacylglycerol concentrations were significantly greater than those observed at baseline (P = 0.0022). There were no significant changes in any other variable when fasting and peak values were compared. Peak plasma viscosity increased in three and decreased in two subjects with no changes in the remaining seven subjects. Changes in peak plasma viscosity ranged from -7% to 7% with similar changes for serum viscosity, from -8% to 10%, and a slightly greater range for plasma fibrinogen, -16% to 10%. In this cohort of normotriacylglycerolemic subjects, there were no significant postprandial changes in plasma viscosity or serum viscosity.

Low high density lipoprotein levels are associated with an elevated blood viscosity.

Low levels of high density lipoprotein (HDL) have been inversely correlated with blood viscosity and plasma viscosity; however, the contribution of concomitant hypertriglyceridemia may confound this association. This study evaluated the relationship between blood viscosity and HDL cholesterol in 70 subjects with fasting levels of total cholesterol <5.2 mmol/l (200 mg/dl) and triglycerides <2.3 mmol/l (200 mg/dl). Viscosity (mPa x s) was measured at 37 degrees C with a coaxial cylinder microviscometer. HDL cholesterol was inversely associated with corrected blood viscosity at 100 s(-1) (beta = -0.49, P<0.00005) and 20 s(-1) (beta = -0.38, P = 0.001) but not at 1 s(-1) (beta = -0.05, P = 0.69) using stepwise multivariate analyses. Low HDL levels are associated with an elevated blood viscosity, and this rheological abnormality may contribute to cardiovascular risk in subjects with isolated low HDL levels.

Effects of lipids and lipoproteins on thrombosis and rheology.

Atherosclerotic plaque rupture and erosions precipitate thrombus formation and may lead to an acute ischemic syndrome. Lipids and lipoproteins modulate the expression and/or function of thrombotic, fibrinolytic and rheologic factors, and thereby influence hemostasis and potential tissue damage resulting from vascular injury. Triglyceride-enriched lipoproteins are accompanied by elevations in factor VII clotting activity, plasminogen activator inhibitor (PAI-1) and viscosity of blood and plasma. Low density lipoprotein (LDL) promotes platelet activation and tissue factor expression and LDL levels correlate with levels of vitamin K dependent coagulation factors and fibrinogen. Conversely, LDL inhibits tissue factor pathway inhibitor (TFPI) which limits activation of the extrinsic coagulation pathway. High density lipoprotein (HDL) has anti-atherothrombotic properties that result from inhibition of platelet and erythrocyte aggregation, reduced blood viscosity and suppression of tissue factor activity and PAI-1 activity and antigen levels. The effects of lipids and lipoproteins on hemostasis and rheology may have important implications for the clinical sequelae following plaque disruption and erosion.

Treatment of severe hypertriglyceridemia lowers plasma viscosity.

Elevated plasma viscosity is a predictor of atherosclerotic vascular disease and is a potential mechanism by which hypertriglyceridemia increases cardiovascular risk. Previous studies of plasma viscosity reduction in hypertriglyceridemic patients used medications that lowered both triglyceride and fibrinogen levels. Because fibrinogen is a major determinant of viscosity, it is unclear whether triglyceride reduction alone is sufficient to reduce plasma viscosity. The purpose of this study was to determine whether triglyceride-lowering therapy reduces plasma viscosity. This was a prospective study of 24 adult patients with severe hypertriglyceridemia (> or = 5.67 mmol/l). Fasting lipid, total serum protein, fibrinogen, plasma viscosity and serum viscosity levels were measured before and after therapy with 1200 mg/d of gemfibrozil. Triglyceride levels decreased by 70% (P < 0.001). Mean plasma and serum viscosity levels decreased by 0.082 mPa/s (P = 0.003) and 0.086 mPa/s (P = 0.013), respectively. Fibrinogen levels did not change significantly. Triglyceride-lowering therapy reduced plasma and serum viscosity without changes in fibrinogen levels. Since serum samples are deplete of fibrinogen, the serum viscosity reduction observed is corroborative evidence for an independent effect of triglyceride-lowering therapy on plasma viscosity. This observation provides a physiological rationale for triglyceride-lowering therapy in patients at risk for atherosclerotic vascular disease, the chylomicronemia syndrome and pancreatitis.

Variability in fibrinogen measurements: an obstacle to cardiovascular risk stratification.

The clinical utility of fibrinogen measurement has been limited by large intraindividual variability. Several approaches that have been shown to improve the repeatability of fibrinogen include acquisition of samples at the same time of day, standardized sample procurement techniques, and multiple replicate sampling. This study employed established pre-analytical and analytical techniques known to reduce fibrinogen variability, including the acquisition of three replicate samples, each analyzed in duplicate, to evaluate the impact of intraindividual variability in fibrinogen measurement at baseline and 3 months on cardiovascular risk in 60 healthy subjects. Classification accuracy was evaluated by the ability to categorize subjects into tertiles of fibrinogen. Only 55% (33/60) of the subjects were correctly assigned to the appropriate fibrinogen tertile. Fibrinogen measurements varied by more than 10% in 45% of subjects and by 5% in 80% of subjects. Intraindividual variability in fibrinogen measurement with a functional assay limits cardiovascular risk assessment even when three replicates are averaged.

Comparative study of HMG-CoA reductase inhibitors on fibrinogen.

Statins have a variable response on fibrinogen, and these changes may have implications on cardiovascular events. In this randomized placebo-controlled crossover study, we evaluated whether changes in fibrinogen levels were different between atorvastatin and other statin-treated patients. Adult coronary heart disease (CHD) patients aged 39-83 years with LDL cholesterol levels > or = 130 mg/dl were randomized to atorvastatin 80 mg (n = 84) and one of the following statins: fluvastatin 80 mg (n = 23), lovastatin 80 mg (n = 20), pravastatin 40 mg (n = 22) or simvastatin 40 mg (n = 20) each for 12 weeks in either order. Fibrinogen was analyzed by an automated method of Clauss. Three independently acquired samples were obtained within 1 min of tourniquet application, and each specimen was measured in duplicate. Statistical analyses were performed using a mixed model repeated measures analysis of variance procedure with SAS version 6.12. There were no significant changes in fibrinogen between treatment groups. This study evaluated changes in fibrinogen with established pre-analytical and analytical procedures known to minimize variability in fibrinogen measurement, and we did not observe any differences in fibrinogen levels in the treatment groups.

Oral 17beta-estradiol and medroxyprogesterone acetate therapy in postmenopausal women increases HDL particle size.

Menopause is accompanied by changes in lipoprotein particles that include an increase in density of low density lipoproteins (LDL) and high density lipoproteins (HDL) particles. The effect of 3 months of oral hormone replacement therapy (HRT) on lipoprotein particle size in postmenopausal women who were randomized to (1) estrogen replacement therapy (ERT) alone (either 17beta-estradiol (1 mg) or conjugated equine estrogens (CEE) (0.625 mg); (2) combination therapy (17beta-estradiol plus medroxyprogesterone acetate (MPA) or CEE plus MPA); and (3) placebo were examined. Lipoprotein subclass concentrations and particle size were quantified by nuclear magnetic resonance spectroscopy (NMR). Combination HRT resulted in significant (P=0.002) increases in HDL particle size as compared with those on placebo formulations or ERT alone. Women assigned to combined HRT had lower concentrations of smaller HDL particles after 3 months (P=0.005) and higher concentrations of larger HDL particles (P=0.02), whereas women assigned to ERT or placebo experienced non-significant changes. In summary, combined HRT increases HDL particle size by altering concentrations of the smallest and largest HDL subspecies.

Elevated blood viscosity in systemic lupus erythematosus.

OBJECTIVES: It has been proposed that elevated blood viscosity contributes to atherothrombotic and thromboembolic processes. We evaluated whether there is increased blood viscosity in systemic lupus erythematosus (SLE) that might contribute to cardiovascular complications and reduced tissue perfusion. METHODS: Blood viscosity profiles were evaluated in SLE patients to determine whether rheologic disturbances contribute to the cardiovascular risk profile. Blood viscosity profiles were evaluated in 27 patients with SLE and 46 age- and gender-matched controls. Blood viscosity was measured at 37 degrees C and shear rates of 1 s(-1) and 100 s(-1), then corrected to the average hematocrit of the SLE patients. RESULTS: Corrected blood viscosity values were higher in SLE patients than in controls at 100 s(-1) (P =.002). Positive correlations were found between the Systemic Lupus International Collaborating Clinics/American College of Rheumatology Damage Index for SLE, which quantifies damage to 12 organ systems and fibrinogen (rho =.39; P =.042) and plasma viscosity (rho =.38; P =.049). CONCLUSIONS: Our data indicate that blood viscosity values at a standard hematocrit are elevated in SLE patients. Further investigations are needed to evaluate whether the increased blood viscosity values in SLE patients contribute to cardiovascular complications and tissue ischemia. CLINICAL RELEVANCE: Because blood viscosity values correlate with the clinical severity of SLE, blood viscosity may contribute to the cardiovascular complications and reduced tissue perfusion in SLE patients. Semin Arthritis Rheum 31:52-57.

Pluripotential mechanisms of cardioprotection with HMG-CoA reductase inhibitor therapy.

Treatment with hydroxymethylglutaryl coenzyme A (HMG-CoA) reductase inhibitors has been accompanied by a reduced risk of cardiovascular events. Rapid onset of clinical benefit and weak correlations between plasma low density lipoprotein-cholesterol levels and coronary lumen change or cardiovascular events indicates that nonlipid mechanisms are involved in this beneficial effects with HMG-CoA reductase inhibitors. Furthermore, more rapid onset of clinical benefit with HMG-CoA reductase inhibitors in patients with acute coronary syndromes or acute myocardial infarction than in those with stable coronary heart disease suggest that HMG-CoA reductase inhibitors facilitate repair of ruptured or ulcerated atherosclerotic plaque, facilitate plaque stabilization and/or reduce thrombus formation on ruptured plaques. Treatment with HMG-CoA reductase inhibitors improved endothelial dysfunction in patients with hypercholesterolemia and this improvement in endothelial function was not correlated with reduction in total serum cholesterol levels. Similarly, reduction in endothelial pre-proendothelin mRNA expression and endothelin synthesis and blood pressure lowering with HMG-CoA reductase inhibitors occurred independent of lipid-lowering. Finally, HMG-CoA reductase inhibitors increased endothelial nitric oxide levels i.e. upregulated endothelial nitric oxide synthetase expression via post-transcriptional mechanisms and prevented its down-regulation by oxidized LDL-C. HMG-CoA reductase inhibitors have been shown to modulate the immune response by inhibiting activation of immune-competent cells such as macrophages, and antigen presentation to macrophages by T cells. Treatment with HMG-CoA reductase inhibitors can reduce expression, production and circulating levels of chemokines (monocyte chemoattractant protein-1) and proinflammatory cytokines [tumor necrosis factoralpha, interleukin (IL)-6 and IL-1beta]. HMG-CoA reductase inhibitors reduced inflammation in human atheroma: significantly fewer macrophages and T cells, less oxidized LDL-C and higher collagen content. In addition, treatment with HMG-CoA reductase inhibitor led to decreased cell death within the atheroma. Treatment with these agents also reduced expression of inducible cellular adhesion molecules, decreased secretion of metalloproteinases by macrophages, reduced vascular smooth muscle cell apoptosis. Lastly, HMG-CoA reductase inhibitors appear to have important effects on the thrombogenesis: reduced expression of tissue factor production and activity; increased production of tissue factor package inhibitor; decreased platelet thrombus formation and improved fibrinolysis as a result of lowered plasminogen activator inhibitor-1 levels. As the pluripotential cardioprotective mechanisms of HMG-CoA reductase inhibitors are further elucidated, it is envisaged that treatment with HMG-CoA reductase inhibitors will be initiated earlier and more frequently in patients with hypercholesterolemia.

Hypertriglyceridemia and coronaryheart disease risk.

Several angiographic-based trials have established that triglyceride remnant particles are associated with progression of coronary stenosis, and a recent prospective study has shown that triglycerides are associated with coronary heart disease (CHD), even after adjustment for high-density lipoprotein cholesterol. The importance of plasma triglycerides in CHD risk must be interpreted within the context of other plasma lipoproteins, as well as the fasting and postprandial triglyceride levels. Recommendations for triglyceride lowering are empiric because of the lack of randomized clinical trial data that evaluate triglyceride-lowering therapies and CHD prevention. This article presents a clinical approach to the management of hypertriglyceridemia.

Long-term safety of pravastatin-gemfibrozil therapy in mixed hyperlipidemia.

BACKGROUND: Combined HMG-CoA reductase inhibitor and fibric acid derivative therapy is often necessary for the effective reduction of concentrations of low-density lipoprotein (LDL) cholesterol and triglycerides in patients with mixed hyperlipidemia; however, the potential risk of myopathy has limited the use of these agents. HYPOTHESIS: This study evaluated long-term safety and efficacy of combined pravastatin and gemfibrozil therapy. METHODS: Eighty-three patients with hyperlipidemia were treated with combined pravastatin and gemfibrozil therapy for a median of 44 months (range 9-78 months). Plasma lipids, serum liver function tests, creatinine, and creatinine kinase (CK) levels were measured every 3 to 4 months. RESULTS: One patient developed myalgia with a normal CK level after 4 months of combination therapy. Three patients had transient elevations in CK levels that ranged from 3 to 5 times the upper limits of "normal" and that returned to normal upon repeat testing. Liver function tests did not change significantly from baseline. In a subset of 26 previously untreated patients, combined pravastatin (mean daily dose 22 mg) and gemfibrozil (mean daily dose 1,154 mg) therapy lowered total cholesterol by 25% (p < 0.001), triglycerides by 53% (p = 0.0001), LDL cholesterol by 14% (p = 0.24), and increased high-density lipoprotein (HDL) cholesterol by 20% (p = 0.012). CONCLUSION: Pravastatin and gemfibrozil therapy is safe and efficacious in patients with mixed hyperlipidemia. The long-term safety results are consistent with other reports on follow-up of shorter duration.