Residual macrovascular risk in 2013: what have we learned?

Fibrates (peroxisome proliferator-activated receptor-α [PPARα] agonists) represent one option, especially in the context of high-risk individuals with insulin-resistant conditions [44]. In subgroup analyses of the FIELD (Fenofibrate Intervention and Event Lowering in Diabetes) study, type 2 diabetes patients with marked atherogenic dyslipidaemia, defined as triglycerides ≥204 mg/dL (2.3 mmol/L) and low HDL cholesterol, obtained greater benefit from fenofibrate treatment than those without this dyslipidaemia. In this group, which comprised 19% of the total study population, there was a 27% relative reduction in CV risk (versus 11% in all patients) [45]. Subsequently, in the ACCORD (Action to Control Cardiovascular Risk in Diabetes) Lipid Trial (n = 5,518), a pre-defined subgroup analysis indicated substantial benefit associated with combining fenofibrate with background simvastatin in type 2 diabetes patients with marked atherogenic dyslipidaemia defined by baseline triglycerides in the upper third of the population (≥204 mg/dL or 2.3 mmol/L) and baseline HDL cholesterol levels in the lower third (≤34 mg/dL or 0.9 mmol/L). This group represented about 17% of the overall study population. In these patients, there was ~30% relative reduction in CV events versus no benefit in 83% of the study population without this dyslipidaemia (Table 1) [46]. Furthermore, a meta-analysis of subgroups with similar lipid criteria for atherogenic dyslipidaemia in the major fibrate trials confirmed this benefit on residual CV risk (Figure 3) [47]. Fibrate treatment was associated with a 35% relative reduction in CV risk in individuals with atherogenic dyslipidaemia versus 6% in individuals without this dyslipidaemia. A subsequent meta-analysis of fibrate trials (n = 45,058) indicated that the reduction in CV risk associated with this therapy class is predominantly due to prevention of coronary events [48]. The observed benefits of triglyceride-lowering on residual CV risk may relate to the effects of endogenous pathways for PPARα activation on atherosclerosis. Lipolysis in certain situations can generate PPAR ligands and also limit some known inflammatory responses [49]. Indeed, recent genome-wide association studies show that gain of function variants of the lipoprotein lipase gene, coding for the enzyme which hydrolyses the triglyceride core of plasma chylomicrons and VLDL to release free fatty acids, are associated with a decrease in plasma triglycerides and coronary risk [50].

An ancillary study of the ACCORD Lipid trial (supported by the R³i), which evaluated effects on postprandial lipaemia, provides a possible explanation as to why the benefit of fibrates is specific to patients with atherogenic dyslipidaemia [51]. Compared with the simvastatin monotherapy group, treatment with the combination of fenofibrate plus simvastatin led to significant reductions in postprandial exposure to triglycerides and apoB48, indicating a decrease in the accumulation of atherogenic, intestinally-derived TRL remnants. This effect of fenofibrate was selective to patients with elevated fasting triglycerides at baseline. Added to this, there is also evidence of benefit with fenofibrate on diabetes-related microvascular complications, including diabetic retinopathy and nephropathy [52‐55].

Despite this favourable profile of macro- and microvascular benefits, the efficacy of the current fibrates may be limited by uncertainties relating to the optimal level of PPARα-agonism, agonist-specific biologic responses, and the side effects of current synthetic agonists. Of relevance for patients already on statin therapy, is the increased risk of muscle-related symptoms with the combination of a statin and certain fibrates, specifically gemfibrozil. This does not appear to be an issue with fenofibrate, probably due to the lack of interference of fenofibrate on statin pharmacokinetics [8, 56]. Indeed, there was no adverse signal for myopathy in the ACCORD Lipid trial [44]. Other commonly reported adverse effects include increases in homocysteine, creatinine and liver enzymes, markers of CVD, renal disease and hepatic dysfunction, respectively [46, 54]. In the ACCORD Lipid study, 48% of statin-treated patients who received the full dose of fenofibrate (160 mg/day) for at least 30 days before the 4-month follow-up showed >20% increase from baseline in serum creatinine (versus 9% in the placebo group) at this visit [57]. Both the ACCORD Lipid and the FIELD studies subsequently showed that increases in serum creatinine associated with fenofibrate were reversible, with levels returning to those observed in the placebo group about 6–8 weeks after cessation of treatment, and did not appear to detrimentally influence risk for CVD or deterioration in renal function [58‐60]. In the FIELD study, despite increasing creatinine levels, fenofibrate treatment was associated with less renal function decline, as shown by a slower rate of loss of estimated glomerular filtration rate over 5 years compared with placebo [59]. However, in high-risk older patients with impaired renal function receiving multiple treatments likely to affect renal haemodynamics, this may be a relevant consideration.

Recent trials with niacin (nicotinic acid) have been disappointing (Table 1). The AIM-HIGH (Atherothrombosis Intervention in Metabolic Syndrome With Low HDL/High Triglycerides: Impact on Global Health Outcomes) trial (n = 3,414) [61] evaluated the effect of extended-release (ER) niacin (1.5-2 g/day) on residual CV risk in patients with CVD and optimally treated with a statin but with residual atherogenic dyslipidaemia (median HDL cholesterol 35 mg/dL or 0.91 mmol/L and median fasting triglycerides 161 mg/dL or 1.82 mmol/L). The study was terminated 18 months earlier than planned due to futility. A number of factors may explain the lack of benefit. First, because of limited financial support, the small study cohort required a very large effect size of 25% benefit for statistical significance. Second, the trial design was flawed by the inclusion of a low dose of niacin (50 mg/capsule) in the placebo, which may have contributed to the 12% increase in HDL cholesterol plasma concentration in the control group. Third, there was imbalance in concomitant LDL lowering therapy between the two groups; in the placebo group, 75% of patients received simvastatin 40 mg/day or higher and 21% also received add-on ezetimibe. Despite these limitations, a subgroup analysis of AIM-HIGH in individuals with triglycerides >200 mg/dL (2.3 mmol/L) and HDL cholesterol <32 mg/dL (0.83 mmol/L) showed a 36% relative reduction (p = 0.032) in the primary composite end point, consistent with findings from the fibrate meta-analysis [47, 62].

More recently, the much larger HPS2-THRIVE (Heart Protection Study 2-Treatment of HDL to Reduce the Incidence of Vascular Events) (n = 25,673) [63] failed to show a benefit on clinical outcomes with a different niacin formulation (combined with laropiprant, which attenuates the niacin flushing response). However, it should be borne in mind that at baseline the HPS2-THRIVE patient population had very well controlled LDL cholesterol levels (mean 63 mg/dL or 1.64 mmol/L) and was outside of the thresholds used to define atherogenic dyslipidaemia (mean HDL cholesterol was 44 mg/dL or 1.14 mmol/L, and triglycerides were 125 mg/dL or 1.43 mmol/L). Consequently the trial did not test the effect of niacin on residual CV risk due to atherogenic dyslipidaemia. Moreover, there were safety issues with niacin/laropiprant, notably significant increases in diabetes complications, new-onset diabetes, infections, and gastrointestinal, musculoskeletal, bleeding and skin adverse events, leading to subsequent world-wide withdrawal of this therapy [64]. Niacin remains a therapeutic option in North and South America, but is not an option in Europe following suspension of niacin/laropiprant by the European Medicines Agency.

With respect to omega-3 fatty acids, there were positive findings from JELIS (Japan Eicosapentaenoic acid Lipid Intervention Study) (n = 18,645) with the combination of omega-3 fatty acids (1.8 g/day of eicosapentaenoic acid [EPA]) plus statin versus statin alone, both overall and in patients with the metabolic syndrome (Table 1) [65]. A post hoc analysis of JELIS evaluated the effect of EPA treatment in individuals with atherogenic dyslipidaemia as defined by triglycerides ≥150 mg/dL (1.7 mmol/L) and HDL cholesterol <40 mg/dL (1.0 mmol/L), representing 5% of the total study population. In this dyslipidaemic group, EPA treatment reduced the risk of coronary artery disease by 53% (p = 0.043), compared with individuals without this dyslipidaemia [66].

In contrast, the Alpha-Omega trial (n = 4,837) failed to show a benefit on major CV events with the addition of n−3 fatty acids (margarine containing EPA plus docosahexaenoic acid with a targeted additional daily intake of 400 mg) in patients with a previous MI receiving best evidence-based antihypertensive, antithrombotic, and lipid-modifying therapy (mainly statins) [67]. More recently, the ORIGIN (Outcome Reduction with an Initial Glargine Intervention) trial [68] used a 2x2 factorial design to investigate the effect of treatment with omega-3 fatty acids (1 g capsule containing ≥900 mg of ethyl esters of n-3 fatty acids) versus placebo (one arm) and insulin glargine versus standard care (second arm) in subjects with or at high risk for diabetes and at increased CV risk (n = 12,536). Treatment with n-3 fatty acids had no significant effect on the primary outcome (CV death) or any of the secondary or additional outcomes. It has been suggested that the lack of effect may relate to the selected dose, and/or extensive use of concomitant cardioprotective treatments. Furthermore, a recent meta-analysis of trials of omega-3 supplementation (n = 68,680 subjects in 20 trials, mean dose 1.51 g/day) failed to show any significant association with all-cause mortality, cardiac death, sudden death, MI, or stroke [69]. There is also some evidence from recent trials to suggest a possible link between high plasma concentrations of omega-3 fatty acids and increased risk of prostate cancer [70, 71].

A new formulation of omega-3 fatty acids (AMR101, ≥96% EPA ethyl ester) is being evaluated in the REDUCE-IT trial in statin-treated patients with elevated trigycerides (>1.7 mmol/L or 150 mg/dL) and at least one CV risk factor [72]. However, results are not expected until 2017.

Combination therapy with statin plus ezetimibe may represent an alternative approach to the management of residual CV risk [10]. In a meta-analysis of 27 trials (>21,000 patients), there was incremental lowering of LDL cholesterol (by 15%), non-HDL cholesterol (by 13%) and triglycerides (by 5%), and raising of HDL cholesterol (by 1.6%) with the combination of ezetimibe plus statin compared with statin alone [73]. However, so far there are limited outcomes data to support this strategy. Although there was significant reduction in major atherosclerotic events (coronary death, MI, ischaemic stroke, or any revascularisation procedure) in patients with advanced renal disease in SHARP (Study of Heart and Renal Protection), SEAS (Simvastatin and Ezetimibe in Aortic Stenosis) failed to show significant clinical benefit in patients with aortic stenosis [74, 75]. Additionally, in the ENHANCE (Ezetimibe and Simvastatin in Hypercholesterolemia Enhances Atherosclerosis Regression) trial, treatment with the combination of simvastatin plus ezetimibe did not result in a significant difference in changes in intima-media thickness compared with simvastatin alone in patients with familial hypercholesterolaemia. This was despite significant reductions from baseline in LDL cholesterol by 55% and triglycerides by 30%, and an increase in HDL cholesterol by 10% in the combination therapy group [76].

It may be that treatment with ezetimibe for 2 years is insufficient to differentiate significant incremental clinical benefit, against a background of lipid-modifying and pleiotropic effects of statin therapy. In support, while partial ileal bypass surgery in POSCH (Program on the Surgical Control of the Hyperlipidemia) and statin treatment in 4S (Scandinavian Simvastatin Survival Study) produced similar LDL cholesterol lowering (37.7% versus 35%), the separation of curves for CV outcomes of treated versus non-treated subjects was evident after about 4.5 years in POSCH but only 1.5 years in 4S [77, 78]. However, the possibility of pleiotropic effects with ezetimibe cannot be discounted, on the basis of recent findings [79]. Results from IMPROVE-IT (IMProved Reduction of Outcomes: Vytorin Efficacy International Trial) [80] comparing statin-ezetimibe combination therapy versus statin alone against a background of best evidence-based treatment in the acute coronary syndrome (ACS) setting are critical to resolve this issue.