It is well known that atherosclerosis has a complex pathogenesis that involves local factors in the vessel wall such as vascular inflammation and also systemic factors such as dyslipidemia and insulin resistance.1) Peroxisome proliferator-activated receptor (PPAR) agonists have been regarded as one of the promising antiatherogenic agents to control both local and systemic atherogenic factors.2) Experimental animal and in vitro studies and some clinical studies too have demonstrated that PPAR agonists limit vascular inflammation and improve insulin resistance and the lipid profiles.3-5) Fibric acid derivatives are known as PPAR-α agonists, and they are already being clinically used for lowering lipid levels. Previous clinical trials of fibrates have demonstrated their beneficial effects on cardiovascular clinical outcomes, and particularly among patients suffering with metabolic syndrome and diabetes.6-10) However, the recent large randomized clinical studies11)12) that have evaluated the more potent PPAR-α agonist have questioned the beneficial effects of PPAR-α agonists, and these studies have presented a variety of debatable issues to clinicians who have prescribed PPAR-α agonists for such diseases.

This review 1) describes the PPAR biology and the mechanism of action of PPAR-α agonists and 2) it summarizes the results of clinical trials of PPAR agonists on atherosclerotic diseases, with a special focus on their safety and efficacy.

Issemann and Green13) discovered that clofibrate activated an orphan nuclear receptor, which they named the PPAR.

This name was based on the observation that these agents induce the proliferation of peroxisomes, a cell organelle, in rodents. The PPAR family consists of 3 members, namely, PPAR-α, PPAR-γ and PPAR-β/δ, which all share approximately 60-80% homology in their ligand-binding domains (LBD) and DNA-binding domains (DBD). Each subtype of PPAR has a distinct tissue distribution, target gene, individual encoding gene and physiological action (Table 1).

PPARs have 5 domains14) (Fig. 1A): 1) the LBD, to which the specific PPAR agonist binds; 2) the N-terminal AF2 domain, which can be activated by phosphorylation through the mitogen-activated protein (MAP) kinase sites via ligand-independent activation; 3) the DBD, which interacts with specific PPAR response elements (PPRE) in the promoter region of the PPAR-activated target genes; 4) the C-terminal AF2 domain, which, in response to ligand binding, undergoes a permissive conformational change that's required for transcriptional activation; 5) the hinge domain, whose function is not well understood, but whose structure is very flexible and it may be crucial for efficient binding of the DBD.

PPAR activation via the binding of LBD to specific ligands leads to its heterodimerization with the retinoid X receptor (RXR), and it undergoes a conformational change in the AF2 domain, which facilitates the release of co-repressors and the recruitment of coactivators(Fig. 1B).15) In contrast to the positive transcriptional regulation of PPAR as described above, PPAR activation can also repress the transcription of the target genes by unknown mechanisms.16) This repressive action of PPARs is the major mechanism that underlies its anti-inflammatory action.

As described above, transcriptional regulation by PPAR agonists requires multiple levels of control as follows: PPAR ligand with synthetic ligands and natural ligands, corepressors and coactivators, homology of the LBD between each subtype of PPAR, and many known and unknown target genes in multiple organs. In fact, synthetic agonists can activate only a specific subtype of PPAR. For example, fibrates bind PPAR-α, which controls lipid metabolism, and thiazolinediones bind PPAR-γ, which controls hyperglycemia. However the effects of synthetic PPAR agonists in vivo differ from those of PPAR itself because the synthetic agonists may be greatly influenced by natural ligands in vivo, and these natural ligands are as yet unknown.5) Corepressors and coactivators form a large, diverse family with members such as nuclear corespressor,17) PPAR-binding protein,18) PPAR-γ coactivator19) and cAMP response element-binding protein.20) PPAR activation by synthetic ligands may modulate a large number of genes, and some of which produce unknown effects. Particularly, the dual or broad PPAR agonists may be more dangerous.21) PPAR activation may be beneficial in one organ, but harmful in another organ. The clinical application of PPAR agonists for therapeutic targets may require a more comprehensive understanding about PPAR biology and this must be carefully approached.

PPAR-α is metabolically active in the liver, heart, kidney, skeletal muscle and brown fat.22)23) It is also present in all vascular cells, including endothelial cells, smooth muscle cells and monocytes/macrophages.24-26) The effects of PPAR-α include hypolipidemic action, an anti-inflammatory effect on the vascular wall and metabolic effects on the myocardium.

Several large clinical studies have investigated the potential cardioprotective effects of PPAR-α agonists, and particularly the fibrate derivatives. These studies have exhibited a wide range of results (Table 2):

In the LOCAT66) (Lopid Coronary Angiography Trial), the DAIS67) (Diabetes Atherosclerosis Intervention Study) and other clinical studies,68) fibrates have demonstrated a reduction in the rates of progression of coronary atheromas as measured by quantitative angiography. However, the cardioprotective effects of PPAR-α agonists were not consistent in the individual clinical trials.

When comparing the results of 2 initial large-scale trials of clofibrate, i.e., the CDP trial62) and the WHO cooperative trial,60) the benefit for reducing CHD events was not significant in the former, but it was significant in the latter (relative risk reduction: 20%, p<0.05). However, both studies showed a significant increase in the total mortality (p<0.05). In the BIP trial,7) bezafibrate treatment showed an 11% relative reduction in the risk of CHD events, but this was not statistically significant (p=0.26). However, post hoc analysis revealed a significant reduction in the risk of CHD events (relative risk reduction: 22%, p=0.02) in those patients with higher baseline TG levels (≥200 mg/dL). The LEADER trial63) showed no beneficial effect of bezafibrate treatment on cardiovascular disease except for a reduction in the risk of nonfatal MI in young men. All the patients of the LEADER trial were also treated with statins, and the outcome results were compounded by the effects of statins. In contrast to the disappointing results of the clofibrate and bezafibrate trials, the gemfibrozil trials (the HHS and VA-HIT) demonstrated a statistically significant reduction in the incidence of cardiovascular events. The VA-HIT results helped to stimulate considerable anticipation for the results of the FIELD study11) that used fenofibrate for treating first or recurrent cardiovascular events in patients with type 2 diabetes. However, the results of the FIELD study showed a statistically nonsignificant difference between the treatment and placebo groups with respect to reducing the primary endpoints, including CHD death and non-fatal MI, and there was a slightly increase of the total and cardiovascular mortality. In the "primary prevention" group, fenofibrate treatment resulted in a cardioprotective effect only against the secondary endpoints such as nonfatal MI and revasularization procedures. This major discrepancy between the results of the VA-HIT and FIELD has led to several debates and it has raised many questions.

First, the cardioprotective effects of gemfibrozil in the HHS and VA-HIT were greater than those of other fibrates in the WHO cooperative trail, the BIP trial and the FIELD study. Some researchers suggest that gemfibrozil has less severe adverse effects than do the other fibrates. Gemfibrozil does not increase the levels of homocysteine and creatinine to the same extent that has been observed with using the other fibrates.69)70) Indeed, in the VA-HIT study, the benefit of gemfibrozil due to the increase in HDL-C could account for only 20% of the CHD reduction in terms of the lipid changes.71) This result suggests that gemfibrozil itself has other beneficial effects that extend beyond its lipid-modulating effect.

Second, the cohort of the FIELD study had different characteristics from those subjects of the VA-HIT. The baseline HDL-C levels of the FIELD cohort were relatively high (mean HDL-C: 42 mg/dL); this might have masked the cardioprotective effects of fenofibrate. The VA-HIT was designed to evaluate the effect of gemfibrozil in patients who were not using statin. However, the control group of the FIELD study had an asymmetrical higher statin drop-in due to the LDL-C-lowering effect of fenofibrate, which might have also masked the beneficial effects of fenofibrate. In contrast, some might argue that the beneficial effect in the FIELD study might be attributed to the modest LDL-C low-ering effect of fenofibrate.72)

Third, some suggest that the potency of PPAR-α agonist might have had an influence on the cardiac outcome results. PPAR is an extremely critical transcriptional regulator, and a number of factors can affect it. In view of the mechanism of action of PPAR agonists, it may be very difficult to get clinically beneficial effects by using only a synthetic agent, which may result in beneficial and/or unwanted effects. Theoretically, the final net effect of both the beneficial and unwanted effects must be considered to be more important. Gemfibrozil is less potent than fenofibrate. The result of the VA-HIT suggests that more potent PPAR activation may not be essential for achieving clinical benefits. Furthermore, LY518674, which is a potent PPAR-α agonist, showed no better results in achieving an improved lipid profile than did fenofibrate or statin monotherapy.12)

Fourth, we do not have a clear answer as to whether combination therapy with statin and fibrate may lead to a beneficial cardiac outcome. Combination therapy may achieve an optimal improvement in the lipid profile and offer greater cardiovascular risk reduction; however, it might increase the risk of rhabdomyolysis.73) When the Action to Control Cardiometabolic Risk in Diabetes (ACCORD) study is completed, good evidence regarding combination therapy will become available. In the ACCORD study design, fenofibrate is added to baseline 20-40 mg simvastatin administration in 5,800 patients in a 2×2 design trial with ultratight glycemic control versus the usual glycemic control.

The safety issues of PPAR agonists have recently entered the spotlight since Nissen et al.74) argued about the potentially cardiotoxic effects of rosiglitazone, a PPAR-γ agonist that's widely used as a hypoglycemic agent. Fortunately, none of the failed PPAR agonists that showed severe side effects were pure PPAR-α or PPAR-α-preferential dual agonists when they tested on the human PPAR isoforms.75) Indeed, for most of the failed PPAR agonists, their apparent affinity for PPAR-γ is higher than their affinity for PPAR-α. Furthermore, because the PPAR-α agonists, i.e., fibrates, have been used for 40 years, they are generally considered as safe drugs with only a few side effects. However, the safety issues of PPAR-α agonists include increased total mortality, muscle toxicity such as myopathy or rhabdomyopathy, increased levels of plasma creatinine and homocysteine, and lithogenicity. We will briefly discuss here the unwanted effects of PPAR-α agonists based on the results of clinical trials.

Considering the benefits and safety concerns, the PPAR-α agonists (fibrates) that are currently in use may definitely be beneficial for decreasing the rate of clinical cardiovascular events in a specific population, and particularly among those patients suffering with metabolic syndrome and atherogenic dyslipidemia. Large-scale long-term clinical trials of PPAR-α agonists have shown that these drugs very rarely have side effects; therefore, clinicians consider fibrates to be safe drugs. However, it should be noted that long-term treatment with most fibrates, unlike that with statins, has not been shown to decrease total mortality; in fact, it has been demonstrated to increase the total mortality in some instances. Because PPAR is a very critical and important gene regulator in the cell and its natural controlling factors are yet unknown, the actions of PPAR-α activation may include not only the known beneficial effects, such as the hypolipidemic and anti-inflammatory actions, but also the unknown harmful effects. Therefore, clinicians should bear in mind the risk/benefit ratio of a PPAR-α agonist in each particular case before prescribing it in clinical practice. In the future, intensive research on this interesting therapeutic target, PPAR, will lead to developing safer and more effective PPAR-α agonists.