Pharmacological Targeting of the Atherogenic Dyslipidemia Complex: The Next Frontier in CVD Prevention Beyond Lowering LDL Cholesterol

Epidemiological, genetic, and animal studies and randomized, controlled clinical trials support a central role for LDL cholesterol (LDL-C) in the development of atherosclerotic cardiovascular disease (CVD) events (1). LDL-C lowering has been the primary goal of dyslipidemia management, with statins as the treatment of choice for CVD prevention. Large-scale, randomized, clinical trials of LDL-lowering therapies have demonstrated significant reduction in CVD events over a wide range of baseline LDL-C levels (2,3). However, even with LDL-C levels lowered substantially or at treatment goals with statin therapy, CVD risks are not eliminated and there remains significant “residual risk.” Intensifying statin therapy may provide additional benefits (4,5); this approach, however, has limited potential, owing to tolerability, side effects, and finite efficacy. Further LDL-C lowering may also be achieved with the use of nonstatin agents, such as cholesterol absorption inhibitors and PCSK9 inhibitors, added to statin therapy. In patients with acute coronary syndrome, ezetimibe, when added to statin therapy, significantly improved CVD outcomes, in which improvement was proportional to LDL-C reduction (6). Additional benefits of PCSK9 inhibitors on CVD outcomes are anticipated, considering their efficacy in further lowering LDL-C levels when added to statin therapy (7). However, it remains to be established to what extent LDL-C can be safely reduced and whether earlier treatment or extreme LDL-C lowering alone is sufficient for CVD risk reduction.

The intent of this Perspective is to highlight an additional form of highly prevalent dyslipidemia, the atherogenic dyslipidemia complex, which represents a promising therapeutic target for further prevention of atherosclerotic CVD beyond LDL-C lowering. We fully acknowledge that lifestyle modification (e.g., body weight reduction in overweight or obese individuals and an increase in physical activity), if successfully implemented, remains the cornerstone of therapy for the atherogenic dyslipidemia complex. Other nonlipid CVD risk factor modifications, such as smoking cessation and antihypertensive, antiplatelet, and anti-inflammatory therapies, have also been shown to effectively reduce the risk of CVD. A review of nonlipid CVD risk factor modifications is beyond the scope of this Perspective, which focuses instead on pharmacological modulation of the atherogenic dyslipidemia complex.

The fourth quarter of the 20th century heralded the onset of a worldwide epidemic of obesity, metabolic syndrome, and type 2 diabetes (T2D). These conditions are associated with a cluster of lipid and lipoprotein abnormalities, the presence of which has increased in parallel with these conditions (8). Such lipid and lipoprotein abnormalities have been referred to as the “atherogenic lipoprotein phenotype” (9) or “atherogenic dyslipidemia” (10), which imply a causal relationship between the phenotype and atherosclerosis. As discussed below, atherogenicity can only be attributed to the collective lipid/lipoprotein abnormality as a whole complex, but not to a specific feature. Also, as discussed below, the features of this cluster of lipid/lipoprotein abnormalities intertwine in their metabolism converging on triglyceride (TG)-rich lipoproteins (TRLs). This atherogenic dyslipidemia complex is characterized by the presence of the following cluster of abnormalities: hypertriglyceridemia (HTG; indicative of an elevation of TRLs), low HDL cholesterol (HDL-C) levels, high small dense LDL (sdLDL) levels, elevated levels of remnant lipoproteins, and postprandial hyperlipidemia. There is currently no uniformly accepted definition of the atherogenic dyslipidemia complex or its component parts, nor would a strict definition necessarily add clarity, as was evidenced by a much criticized attempt to propose a working definition of the metabolic syndrome (11). Furthermore, certain characteristic abnormalities of the atherogenic dyslipidemia complex, such as postprandial hyperlipidemia, sdLDL particles, and remnant lipoproteins, have not typically been measured in large-scale population studies. As a result, the true prevalence of the complex is currently unknown. Nevertheless, because several features of the atherogenic dyslipidemia complex are integral components of the metabolic syndrome (12), and metabolic syndrome affects nearly 35% of all adults and 50% of those >60 years of age in the U.S. (13), the prevalence of the atherogenic dyslipidemia complex is expected to be alarmingly high.

The atherogenicity of the atherogenic dyslipidemia complex is supported by epidemiological studies. For instance, high TG levels and low HDL-C levels in the NHANES (National Health and Nutrition Examination Survey) III study (14) were each strongly and independently correlated with major CVD events. In the INTERHEART study (15), elevated apolipoprotein (apo)B/apoA-I ratio (reflective of apoB-containing lipoproteins and HDL, respectively) was found to be a major predicator of myocardial infarction in the general population around the world. The presence of high TG and low HDL-C levels, concurrent with high LDL-C levels, increased CVD risks in the Scandinavian Simvastatin Survival Study (16) and in the Helsinki Heart Study (17). Likewise, TG levels were strongly and independently associated with coronary heart disease risk in a meta-analysis of 29 prospective studies (18). Nonfasting TG level, which may reflect the atherogenic capacity of TRL remnants, is a stronger predictor of CVD events than fasting TG levels (19,20). Because most people are in the postprandial state for many hours each day, changes in postprandial lipid levels may have an important impact on atherosclerosis development.

Atherogenic properties have been assigned to atherogenic dyslipidemia complex components, as has been reviewed for TRL levels (21), remnant lipoprotein levels (22), postprandial lipemia (23), nonfasting TG levels (24), sdLDL particles (25), and the antiatherosclerotic properties of HDL (26). Epidemiological or observational approaches are not suited to unravel causality, hence the ongoing debate as to which individual component or components play a direct role in atherosclerosis. One should also bear in mind that the associated chronic inflammation, increased levels of prothrombotic factors, perturbed tissue factors, insulin resistance, and hyperglycemia play important roles in promoting atherosclerosis. The inability to disentangle individual components of the atherogenic dyslipidemia complex and clearly implicate direct roles of each in atherogenesis suggests that the phenotype is likely multidimensional and that continued efforts to reduce causality to a single component might be misguided. Indeed, individual components of the atherogenic dyslipidemia complex seldom exist in isolation; instead, they often present as a cluster because of their interrelated metabolism and etiology (see below). In addition, most therapies aimed primarily at correcting a certain lipid moiety, as well as lifestyle modification, improve multiple components of the atherogenic dyslipidemia complex simultaneously.

In contrast with elevated LDL-C or apoB levels, simple biomarkers to assess the atherogenicity of the atherogenic dyslipidemia complex are not well defined. Aside from plasma levels of TG, HDL-C, non-HDL-C, and ratios of these lipid parameters, various methodologies besides standard assays of plasma lipids are available for the assessment of atherogenic dyslipidemia complex components. Specialized analyses, such as gradient gel electrophoresis, vertical auto profile, nuclear magnetic resonance, or ion mobility, may yield information on cholesterol levels in lipoprotein subfractions to gain deeper insight into atherosclerotic risk (27). ApoB-48, the apolipoprotein present in chylomicrons and remnants, may serve as a surrogate marker of atherosclerosis in patients at risk (28). In addition, the postprandial lipid response to a high-fat meal may reveal useful information with regard to TG and TRL metabolism. In the atherogenic dyslipidemia complex, LDL particles are often smaller in size; therefore, the number of LDL particles may not be accurately reflected by LDL-C levels, especially in many patients with diabetic dyslipidemia in whom LDL-C levels are near normal. Non-HDL-C (29) and total apoB (30) levels are often increased along with the atherogenic dyslipidemia complex. Non-HDL-C includes cholesterol in all apoB-containing lipoproteins, including TRL (VLDL and chylomicron) and their remnants, IDL, LDL, and lipoprotein(a), thus reflecting the total atherogenic cholesterol burden. Consequently, non-HDL-C is strongly associated with coronary heart disease risk (29,31). Because much of the predictive risk of non-HDL-C is likely to be attributed to LDL-C, which is its major component, nonfasting, non-HDL-C may be an even stronger predictor of coronary heart disease, but this requires further analysis. Plasma total apoB level reflects the number of all apoB-containing lipoproteins. Both non-HDL-C and apoB may serve as alternate treatment targets (32). From a practical standpoint, non-HDL-C and apoB are easily measurable clinical parameters. Non-HDL-C has additional appeal because it is calculated from the two commonly measured lipid entities, as total cholesterol subtracted by HDL-C. Finally, it has been suggested that the cholesterol rather than the TG content of remnant lipoproteins, which can be estimated from nonfasting lipids, drives the atherogenicity of these particles (33).

In summary, there is convincing evidence that the atherogenic dyslipidemia complex as a whole is highly atherogenic, although it is not known which component/s of the complex are directly implicated in atherogenesis. There is currently no widely accepted clinical definition of the atherogenic dyslipidemia complex, and, hence, its true prevalence is not known, although it is predicted to be highly prevalent in populations that also have a high prevalence of the metabolic syndrome. There are a number of biomarkers of the atherogenic dyslipidemia complex, but there is no consensus as to which biomarkers best predict atherogenicity in this condition.

In individuals with the atherogenic dyslipidemia complex, the lipid phenotype primarily reflects the abnormal metabolism of TRL, as affected by both genetic and environmental factors, with secondary effects on HDL and LDL metabolism (Fig. 1). Net plasma TRL concentration reflects the balance between the production and clearance of the whole particle or various components of the particle. Elevated plasma TRL levels can therefore be a result of impaired clearance, increased production, or a combination of both (34). Increased TRL concentration in the circulation enhances lipid and apolipoprotein exchange with HDL, resulting in TG enrichment and subsequent accelerated catabolism of HDL particles (35). The formation of sdLDL particles is also closely related to increased TRL levels. In HTG, there is increased secretion of TG-enriched, larger VLDL particles, which are precursors of sdLDL, and increased delipidation of larger buoyant LDL to form sdLDL; furthermore, the liver directly secretes more LDL particles of smaller size in HTG (25,36,37).

The atherogenic dyslipidemia complex is frequently associated with reduced clearance capacity, impairing catabolism of VLDL, chylomicrons, and their remnants, including TG hydrolysis and particle removal (apoB catabolism) (reviewed in Lewis et al. [34] and Watts and Chan [38]). Such clearance defects may arise due to several abnormalities, including deficiency of lipoprotein lipase (LPL) or its key cofactors, such as apoC2; defects in cell surface receptors; and changes in lipoprotein composition (particularly, the apolipoprotein composition of TRL). Indeed, genetic variants in association with TG elevation that have been identified in genome-wide association studies (GWASs) mostly reflect defects in TG clearance, at least among identified genes encoding proteins with known functions (see below).

TRL overproduction is a prominent feature that coexists with clearance defects and exacerbates the severity of the atherogenic dyslipidemia complex. Insulin resistance is a major cause and is frequently associated with TRL overproduction by the liver. Under such metabolic conditions, adipose lipolysis is increased along with less effective trapping of fatty acids released into the capillary microcirculation by intravascular lipolysis of circulating TRLs (39). The resultant increase in free fatty acid flux to the liver provides excessive substrate for VLDL TG synthesis and drives increased VLDL secretion. In addition, insulin resistance leads to increased hepatic de novo lipogenesis and decreased fatty acid oxidation, further stimulating VLDL TG secretion (39–41). Diminished post-translational apoB-100 degradation in hepatocytes with increased biogenesis, stability, and secretion of nascent VLDL particles has been demonstrated in insulin resistance, resulting in increased production of VLDL by the liver (42). Although there are some differences in the regulation of VLDL and chylomicron particle production, similar mechanisms are also likely to be involved in enhancing chylomicron production in the intestine (43,44). Thus, the production of intestinally derived TRL apoB-48 is also increased in insulin resistance (45), obesity (46), and T2D (47,48), and chylomicron TG secretion is increased in metabolic syndrome (49).

Both genetic and environmental contributions are implicated in the etiology of HTG. Although rare monogenic forms of severe HTG manifesting as chylomicronemia exist (50), the majority of cases of HTG are typically mild to moderate and have a polygenic basis. Common genetic variants with small effect sizes, single nucleotide polymorphisms, such as those identified by large GWASs in >45 loci associated with variation in plasma TG concentrations (51), create a genetic background of susceptibility. Many of these loci have joint associations with other lipid phenotypes. In particular, the TG-raising allele at a GWAS locus is more often than not associated with lower HDL-C levels (51). Indeed, huge sample sizes and complicated statistical genetic models cannot fully dissociate such interrelationships (52). These genetic correlations confirm the connectivity of the biological pathways governing both TG and HDL levels, and help to explain the overall difficulty in dissecting out the effects of one lipid variable over the other, as discussed above.

On top of these common predisposing variants, rare heterozygous variants with large effect sizes may further exacerbate the phenotype and increase the severity of the clinical presentation (53). A few rare and many common genetic variants in the LPL gene have been identified and linked to the lipid abnormalities that characterize the atherogenic dyslipidemia complex (54,55). Among rare variants, numerous loss-of-function mutations in several genes that encode proteins that affect the functionality of LPL have been identified. Mutations in APOC2, GPIHBP1, LMF1, and APOA5 are associated with elevated TG, whereas mutations in ANGPTL3, ANGPTL4, and APOC3 are associated with reduced TG levels (53). Furthermore, noncoding RNAs (e.g., microRNAs) may regulate the expression of proteins involved in lipoprotein-cholesterol and TG metabolism (56), which adds to the expanded genetic contribution to dyslipidemia.

Whereas such genetic evidence points to defects involving TRL clearance mostly related to LPL, it is likely that TRL overproduction is also genetically determined; many of the novel GWAS loci associated with TG levels that encode products with unknown function may prove to have a role in the synthesis and secretion of TRL (51). Mutations in GPD1, which encodes glycerol-3-phosphate dehydrogenase 1, are associated with transient infantile HTG, likely through increased TG secretion (57). Besides modulating the clearance of TRL by LPL, apoC3 may be involved in TRL production, as suggested by the findings that apoC3 antisense reduced TG levels and TRL particle numbers in patients with null LPL mutation (58). Variants in GCKR encoding glucokinase regulatory protein are associated with HTG, possibly through affecting TG biosynthesis (59). In addition, some insulin resistance risk alleles are also associated with high TG and low HDL-C levels and increased risk of coronary heart disease (60). It is well established that insulin resistance underlies many features of metabolic defects in metabolic syndrome and T2D and that TRL production is known to be increased in insulin resistance.

At the opposite end of the TG spectrum is hypolipidemia. Loss-of-function mutations in APOB impair apoB trafficking and secretion or enhance its catabolism, causing hypobetalipoproteinemia with low plasma TG, apoB, and LDL-C levels (61) and reduced risk of atherosclerotic CVD (62). Loss-of-function mutations in ANGPTL3 also cause hypolipoproteinemia with low–apoB-containing lipoproteins and HDL-C (63), and loss-of-function mutations in other angiopoietin-like protein (ANGPTL) family members also are associated with low plasma TG levels (64,65). The prevalence of atherosclerotic CVD with ANGPTL mutations has not been clearly established. Loss-of-function mutations in MTTP cause abetalipoproteinemia, a condition too rare to assess atherosclerotic CVD prevalence.

Finally, many environmental (nongenetic) factors can exacerbate HTG and, in many cases, cause clinical HTG in otherwise dormant phenotypes in individuals with predisposing genetic backgrounds (53). These include lifestyle and acquired factors (e.g., adiposity, especially visceral obesity; physical inactivity; increased intake of saturated fats; and high carbohydrate levels, especially refined sugar or alcohol intake), medical conditions (e.g., insulin resistance, metabolic syndrome, untreated or poorly controlled diabetes), pregnancy, and certain medications (54). HTG, therefore, may be better depicted as a continuum of interactions between genetic and environmental contributions, where the cumulative effects of the polygenic susceptibility and multiple environmental burdens may each increase the risks for the development of HTG (34).

Fibrates are peroxisome proliferator–activated receptor α (PPARα) agonists that increase fatty acid oxidation in the liver and muscle and decrease hepatic lipogenesis. In addition, fibrates increase LPL activity and decrease the hepatic secretion and VLDL concentration of apoC3 (66). Kinetic studies (67) have indicated that fibrates lower VLDL apoB-100 levels through enhanced clearance and decreased production. Fibrates increase apoA-I and apoA-II transcription, enhance reverse cholesterol transport, and indirectly, through altered VLDL metabolism, increase HDL levels (66). Fibrates may also increase LDL particle removal, although this is a relatively minor effect. Collectively, fibrates are very effective TG-lowering therapies with modest HDL-raising effects. Fibrates are less effective at lowering LDL-C or apoB concentrations but may shift LDL particles from small to large size (68).

Despite these potentially beneficial effects in modifying the typical dyslipidemia that characterizes the atherogenic dyslipidemia complex, fibrate clinical trials have been generally disappointing with respect to CVD prevention. Early studies with clofibrate (the Coronary Drug Project and the WHO Clofibrate trial) conducted in the 1970s yielded conflicting results on CVD outcomes, possibly due to aspects of the population studied and inclusion criteria. More recent trials, including FIELD (Fenofibrate Intervention and Event Lowering in Diabetes) (fenofibrate in T2D patients) (69,70), BIP (Bezafibrate Infarction Prevention) (bezafibrate) (71), HHS (Helsinki Heart Study) (gemfibrozil) (17), ACCORD-Lipid (fenofibrate in statin-treated T2D patients) (72), and VA-HIT (Veterans Affairs HDL Intervention Trial) (gemfibrozil) (73) trials, have investigated the efficacy of several fibrates on CVD outcomes. Other than in the VA-HIT and HHS trials, the primary composite CVD end point was not significantly reduced by fibrate treatment. These results are mostly attributed to the failure of these trials to enroll the population with the typical dyslipidemia that is most likely to benefit from fibrate therapy (i.e., the atherogenic dyslipidemia complex). In contrast, the VA-HIT trial targeted individuals with low HDL-C levels; therefore, the study population was enriched with those who have features of the atherogenic dyslipidemia complex. Meta-analyses (74) of these fibrate outcomes trials have demonstrated significant CVD risk reduction in subgroups with HTG, but not in those with normotriglyceridemia. In a meta-analysis (75) of 18 prospective randomized controlled trials with fibrates, CVD benefits in the whole study population were very modest. However, in a meta-regression analysis of the fibrate trials, in subjects with baseline values of TGs >2 mmol/L, major CVD events were inversely associated with the magnitude of TG lowering (75). A recently published Cochrane Database systematic review (76) of 13 fibrate randomized controlled trials showed evidence for a significant 12% protective effect of fibrates compared with placebo for CVD but no reduction in all-cause mortality or death from vascular disease. In our opinion, these studies prove the futility of the use of fibrates in the prevention of CVD in those with normotriglyceridemia, but show small benefit in the population with the atherogenic dyslipidemia complex, even with background statin use. The efficacy of fibrates, alone or in combination with statins, on CVD outcomes needs to be further evaluated in trials with appropriate patient selection. Currently, fibrates are being used as adjunctive therapy to statins in high-risk patients with baseline HTG. More potent and selective PPARα agonists and PPARα/γ dual-receptor agonists are being developed (77). Such compounds may provide more effective TG lowering and may have additional insulin-sensitizing properties, although major questions remain regarding their overall safety.

Earlier studies, dating back to the 1970s, showed that niacin (nicotinic acid) reduced CVD (78). High-dose niacin increases HDL-C levels and lowers the levels of LDL-C, lipoprotein(a), TRL, and sdLDL particles (78). Immediate-release niacin is associated with the side effect of flushing, and its use for lipid control in T2D patients is discouraged because it may worsen insulin resistance and glycemic control. Slow-release forms of niacin reduce flushing but have increased risks of hepatotoxicity and reduced efficacy on HDL-C. Extended-release niacin also improves lipid profiles, without causing flushing or raising fasting glucose levels. The amide of niacin, nicotinamide, despite providing biochemical functions of niacin, does not provide benefits on lipids. The putative antiatherosclerotic effects of niacin were frequently attributed to its HDL-raising effect, despite the fact that it had multiple effects on plasma lipids and possible nonlipid antiatherosclerotic effects. Kinetic studies have indicated that extended-release niacin decreases the production of apo(a) and apoB-100–containing lipoproteins (79), increases the production of apoA-I, enhances the clearance of TRL apoB-100 and apoB-48 (80), and reduces the fractional catabolic rate of apoA-I (81).

Only until recently has the efficacy of niacin, when added to statin therapy, been examined in atherosclerosis and CVD. In the ARBITER 6-HALTS (Arterial Biology for the Investigation of the Treatment Effects of Reducing Cholesterol 6-HDL and LDL Treatment Strategies in Atherosclerosis) trial (82), extended-release niacin induced the regression of carotid intima-media thickness in patients receiving statin. Large, well-powered clinical trials have failed to demonstrate the CVD benefits of extended-release niacin when added to statin therapy. In the AIM-HIGH (Atherothrombosis Intervention in Metabolic Syndrome with low HDL/High Triglycerides: Impact on Global Health Outcomes) trial, extended-release niacin was added to therapy with simvastatin 40–80 mg/day, plus ezetimibe 10 mg/day if needed, to maintain LDL-C at levels between 40 and 80 mg/dL (between 1 and 2 mmol/L). Despite improvement in plasma HDL-C, TG, and LDL-C levels, there was no incremental clinical benefit derived from the addition of niacin to statin therapy (83). The HPS2-THRIVE trial, which used extended-release niacin and laropiprant (to prevent niacin-induced flushing), did not reduce the number of CVD events, but increased the risk of adverse events in patients with vascular disease and well-controlled LDL-C levels during the 3.9-year follow-up period (84). Both trials were prematurely terminated. We speculate that the effect of niacin on reducing apoB-containing lipoprotein levels is beneficial but is quantitatively small to negligible with background statin therapy, masking any putative antiatherosclerotic effect. Besides, if its HDL-raising effect is not accompanied by improved HDL function, its possible protection against atherosclerosis is questionable. Niacin is far less effective at lowering TRL levels than are the fibrates and is therefore less likely to be effective for treating the atherogenic dyslipidemia complex.

Cholesteryl ester transfer protein (CETP) mediates the heteroexchange of neutral lipids between the denser lipoproteins (HDL and LDL) and TRL (i.e., the transfer of TGs from TRL to HDL and LDL in exchange for cholesteryl ester from HDL and LDL to TRL). CETP-mediated exchange of neutral lipids between HDL and TRL is one of two major pathways in which cholesterol is transported from tissues back to the liver for excretion in bile and feces, namely, reverse cholesterol transport. The inhibition of CETP increases HDL-C concentration; however, this potentially advantageous effect may be offset by the impairment of reverse cholesterol transport due to HDL particle dysfunction. Thus far, clinical development of CETP inhibitors has been plagued by disappointing results. A clinical trial of torcetrapib added to atorvastatin (ILLUMINATE) was terminated prematurely because of an increased risk of mortality and morbidity in patients with high CVD risk, despite a 72% increase in HDL-C and a 25% reduction in LDL-C levels (85). The adverse effects of torcetrapib may have been due to off-target effects (86). In the Dal-OUTCOMES study (87), dalcetrapib did not reduce the risk of recurrent cardiovascular events in patients who had experienced a recent acute coronary syndrome event, despite increased HDL-C levels. It should be noted that dalcetrapib therapy resulted in a very modest reduction in TG levels and no effect on apoB. A trial with evacetrapib was recently discontinued because of futility in preventing CVD outcomes; although top line results were made public, a detailed peer review manuscript has not been published to date.

The aggregate of results thus far does not support CETP inhibition as an effective approach for CVD prevention. There may be several reasons for this. First, CETP inhibition per se may not be antiatherogenic (88). It has been increasingly suggested that HDL function instead of concentration may underlie its link to CVD benefits (26). Furthermore, these negative trials with CETP inhibitors studied patients whose HDL-C levels were on average in the normal range, and thus could not completely rule out the possibility that the raising of HDL-C levels provides CVD benefits. In our opinion, exclusively targeting this aspect of the atherogenic dyslipidemia complex (i.e., HDL alone) without ameliorating the accompanying abnormalities of elevated TRL and apoB-containing lipoprotein levels may not be an effective therapeutic strategy. More potent and highly selective CETP inhibitors are being developed. For instance, anacetrapib increases levels of HDL-C and decreases levels of both LDL-C and apoB-100 (89,90), and, when added to statin treatment, results in further reductions in LDL-C (91,92). Anacetrapib treatment also increases the LDL TG/cholesterol ratio and LDL size and increases LDL-apoB-100 clearance (93). CETP inhibitors that more potently lower atherogenic, apoB-containing lipoproteins, and that lack off-target adverse effects on blood pressure, may reduce the number of CVD events. Results from the ongoing CVD outcomes trial with anacetrapib (clinical trial reg. no. NCT01252953, clinicaltrials.gov) may help to address this issue.

In summary, we have provided some explanations and perspective with respect to the failure of the above three modalities of therapy, which all modify certain aspects of the atherogenic dyslipidemia complex. In our opinion, these apparent failures point to specific limitations of each particular agent or, in the case of fibrates, to pitfalls in the design of the clinical trials (i.e., patient selection). Their failure should not be interpreted as overall futility in attempting to modify the atherogenic dyslipidemia complex and should not prevent ongoing and future exploration of therapies to prevent atherosclerosis by non-LDL–lowering modification of dyslipidemia.

The atherogenic dyslipidemia complex presents a plethora of potential therapeutic targets to lower fasting and postprandial TRL levels, raise HDL-C levels, and shift LDL particle size to the larger, more buoyant, and potentially less atherogenic species. We will briefly discuss the potential of new therapeutic strategies that target what we believe to be the primary pathophysiological aspects of the atherogenic dyslipidemia complex: the clearance and secretion of hepatic and intestinally derived TRL particles, which will result in secondary beneficial effects on HDL particle concentration and LDL size.

One method of lowering circulating TRL concentration is to enhance LPL-mediated TG clearance. Among patients with LPL deficiency and chylomicronemia with severe HTG, ongoing evaluation of LPL gene therapy shows some promise (94). Although this highlights the important role of LPL, gene therapy is limited to cases of marked HTG due to homozygous LPL loss-of-function mutations. Several recent studies of nongene replacement therapies have provided some promise in improving LPL activity, as follows.

ApoC3 is an LPL inhibitor. It has also been shown to inhibit hepatic lipase activity, enhance VLDL secretion, and suppress TRL remnant clearance. Carriers of rare loss-of-function mutations in the APOC3 gene have low TG levels and reduced CVD risk (95,96). Antisense inhibition of apoC3 in preclinical models and in a phase I clinical trial of healthy subjects decreased plasma concentrations of apoC3 and TG (97). In three patients with familial chylomicronemia syndrome, APOC3 antisense RNA dramatically reduced TG levels and decreased the numbers of chylomicron and VLDL particles (58). Because familial chylomicronemia is caused by mutations in the gene encoding LPL or in the genes encoding proteins affecting LPL function, and the patients in the above study had homozygous or compound heterozygous-null LPL mutation, the reduction in TG concentration with APOC3 antisense treatment suggests that apoC3 may affect TRL metabolism through LPL-independent pathways. The findings on the efficacy of anti–APOC3 RNA-based strategies were expanded in a larger cohort of patients with fasting HTG, where APOC3 antisense as monotherapy or as an add-on to fibrates dose-dependently reduced plasma TG concentrations by up to 70% from baseline (98).

Several members of the ANGPTL family of proteins are found to regulate LPL. Rare loss-of-function mutations in ANGPTL3 are associated with lower levels of all lipids and lipoproteins, including TG, LDL-C, and HDL-C (63). The blockade of ANGPTL3 with antibody reduced plasma lipids, including TG, non-HDL-C, and HDL-C in mice and monkeys (99), which is likely associated with reduced VLDL TG production and increased clearance of apoB-containing lipoproteins (100). ANGPTL4 inhibits LPL and is a major regulator of LPL activity during fasting and exercise (101). Loss-of-function variants in ANGTPL4 are associated with reduced plasma TG levels and a lower risk of coronary artery disease (102–105). The inhibition of ANGPTL4 with antibody reduced serum TG levels in mice and monkeys (105). The hypolipidemic effects of the pharmacological inhibition of ANGPTL3 and ANGPTL4 in humans await further studies.

As discussed above, genetic studies have identified mutations in several genes encoding proteins that affect LPL function and activity. Therapeutic approaches targeting these genes or related proteins may yield effective TG reduction and CVD risk benefits in the future. In addition, the identification of noncoding microRNAs in the regulation of lipoprotein metabolism opens more opportunities for the development of novel therapeutics against atherosclerotic CVD.

Despite targeting the biogenesis of TRL, these therapies have major therapeutic effects in lowering LDL rather than TRL levels and therefore are not ideal agents for the treatment of the atherogenic dyslipidemia complex, although they could be considered for the treatment of very severe cases of genetic HTG not responding to other therapies (106,107). These therapies are currently approved in some countries for the treatment of homozygous familial hypercholesterolemia. Both of these costly therapies are associated with significant side effects. The apoB antisense drug (mipomersen) is administered by subcutaneous injection, and injection site reactions and flu symptoms frequently develop in treated patients. Both mipomersen and the oral microsomal triglyceride transfer protein inhibitor (lomitapide) can also cause hepatotoxicity (106,107). These therapies are currently reserved for the treatment of a very rare life-threatening genetic condition.

Pharmacological inhibition of enzymes involved in TG biosynthesis, namely, diacylglycerol acyltransferase (DGAT) and monoacylglycerol acyltransferase (MGAT), has shown promise for the treatment of obesity, diabetes, and dyslipidemia in animal models (108). MGAT and DGAT catalyze the final consecutive steps of TG biosynthesis. MGAT is mainly expressed in the gastrointestinal tract, whereas DGAT is expressed in multiple tissues, including the intestine and the liver. Isozymes with different functions and expression have been identified for both MGAT and DGAT. MGAT2 is exclusively expressed in the intestine, where the MGAT pathway accounts for ∼75% of TG synthesis (109), and MGAT plays an important role in dietary lipid absorption. DGAT1 is abundantly expressed in the small intestine (110). DGAT1 inhibition reduced postprandial TG plasma excursion in humans; however, this was accompanied by gastrointestinal side effects (111). Other DGAT1 inhibitor compounds are being developed, with the hope of satisfactory efficacy, safety, and tolerability (112).

Acetyl-CoA carboxylases (ACCs) catalyze the rate-limiting step of fatty acid biosynthesis (i.e., carboxylation of acetyl-CoA to malonyl Co-A). Two isozymes, ACC1 and ACC2, are expressed in different tissues. ACC1 is expressed in lipogenic tissues (the liver and adipose tissue) primarily catalyzing long-chain fatty acid biosynthesis, whereas ACC2 is expressed in oxidative tissues (the liver, skeletal muscle, and heart). Isozyme-nonselective ACC inhibitors have the potential of inhibiting lipogenesis and increasing fatty acid oxidation (113). A phase I clinical study (114) of one such compound demonstrated the inhibition of de novo lipogenesis and increased whole-body fatty acid oxidation. ACC inhibition may be useful in the treatment of T2D, HTG, and nonalcoholic steatohepatitis, but has not yet been examined in clinical trials.

The investigational agent ETC-1002 is a dual modulator of two hepatic enzymes; it inhibits ATP-citrate lyase and activates AMPK in the liver to inhibit sterol and fatty acid synthesis and to promote fatty acid oxidation (115). In patients with elevated LDL-C levels, ETC-1002 lowers LDL-C levels regardless of baseline TG levels (116). In addition, it lowers apoB-100, non-HDL-C, and LDL particle number, and improves factors related to cardiovascular health (e.g., inflammatory markers, blood pressure, and body weight). The lowering of LDL-C levels, as well as non-HDL-C and total cholesterol levels, was also observed in patients with T2D (117). In statin-intolerant patients with hypercholesterolemia, ETC-1002 decreased non-HDL-C, total cholesterol, and apoB levels without significantly affecting TG and HDL levels (118). Long-term studies are needed to assess its safety, tolerability, and efficacy as combination therapy compared with those of other lipid-lowering drugs.

Recent advances in the understanding of intestinal lipoprotein production offer novel therapeutic approaches in attenuating postprandial lipemia (reviewed in Dash et al. [44]). Among those, the incretin-based antidiabetic drugs exenatide (a glucagon-like peptide-1 receptor agonist) and sitagliptin (a dipeptidyl peptidase-4 inhibitor) suppress intestinal lipoprotein production in the short term in humans (119,120). Incretin-based antidiabetic drugs also attenuated postprandial TG excursion in clinical trials (reviewed in Xiao et al. [121]). Several recent CVD outcomes trials (122–124) have demonstrated the noninferiority of several such agents in T2D patients with established CVD or with increased CVD risks, and more trials with other agents are underway.

These nutraceuticals have been shown to provide cardiovascular health benefits and are recommended in treating marked HTG. Their mechanisms of action are multifaceted, including an antiarrhythmic effect, plasma TG reduction, blood pressure lowering, a decrease in platelet aggregation, improvement of vascular reactivity, and an anti-inflammation effect (125). It has been postulated that the hypolipidemic effect of n-3 fatty acids may be due to reduced hepatic lipogenesis, increased fatty acid oxidation, and suppressed VLDL secretion, although the exact mechanisms remain to be elucidated (126). The efficacy of n-3 fatty acids on reducing CVD outcomes is currently being evaluated in HTG patients receiving statin therapy (clinical trial reg. nos. NCT01492361 and NCT02104817, clinicaltrials.gov).

Several developments in the field have also indicated potentially novel mechanisms of lipid modification. For example, an FGF21 analog, originally developed as a glucose-lowering agent for the treatment of T2D, has demonstrated greater lipid-modifying effects in obese subjects with T2D, including lowering TG, total cholesterol, and LDL-C levels, and increasing HDL-C levels (127). The mechanism of FGF21 modulating lipid metabolism likely involves multiple organs, but its effects on TRL metabolism have not been examined. Recently, it has been shown (128,129) in rodents that the activation of brown adipose tissue enhances fatty acid uptake by brown adipose tissue and the clearance of remnant cholesterol by liver, leading to the lowering of TG and cholesterol levels, translation of which into therapeutic means in humans awaits further studies.

It is important to point out that lifestyle management, such as diet, exercise, and achievement of a healthy body weight, is an integral and key component of lipid management. Such approaches are difficult to implement and adhere to, but are highly effective in modulating, the atherogenic dyslipidemia complex. The above lifestyle changes are accompanied by improvement in insulin sensitivity and general metabolic status, which is expected to contribute to improved lipid profiles. Weight loss improves many components of the atherogenic dyslipidemia complex (130,131). Sustained weight loss also delays the onset of T2D, in which the atherogenic dyslipidemia complex is often a prevalent feature. Bariatric surgery, the most effective intervention in inducing sustained weight loss, also reduces TRL production (132,133). Nonpharmacological therapies thus far have not shown reductions in CVD outcomes.

Although much success has been achieved in LDL-C lowering and CVD reduction with statin-based therapies, residual CVD risk in statin-treated patients remains a significant challenge, especially with the ever increasing prevalence of obesity, diabetes, and metabolic syndrome. Recent advances in the management of dyslipidemia have provided promise in several aspects, including further lowering of LDL-C levels and CVD risk with intensified statin and nonstatin drugs, such as ezetimibe, and effective LDL-C lowering with PCSK9 inhibition. The atherogenic dyslipidemia complex contributes to the residual CVD risk and represents an expanded panel of therapeutic targets, especially in at-risk patients with near-normal LDL-C levels who are less responsive to statins or are statin intolerant. Among existing and currently approved therapies, fibrates should be reevaluated in those with the atherogenic dyslipidemia complex. Other emerging therapies, such as those directly targeting TRL metabolism, may provide novel opportunities for reducing CVD risk through the management of dyslipidemia. Targeting the atherogenic dyslipidemia complex is the next frontier in antiatherosclerotic therapy.

Funding. R.A.H. is supported by the Jacob J. Wolfe Distinguished Medical Research Chair, the Edith Schulich Vinet Canada Research Chair (Tier I) in Human Genetics, the Martha Blackburn Chair in Cardiovascular Research, and operating grants from the Canadian Institutes of Health Research (MOP-13430 and MOP-79533), the Heart and Stroke Foundation of Ontario (T6066 and 000353), and Genome Canada through Genome Quebec. G.F.L. is supported by the Sun Life Financial Chair in Diabetes, the Drucker Family Chair in Diabetes Research, an operating grants from the Canadian Institutes of Health Research (MOP-43839) and the Heart and Stroke Foundation of Ontario (000032).

Duality of Interest. No potential conflicts of interest relevant to this article were reported.