FormalPara Key Points

Fenofibrate should be considered a relatively safe and reliable supplemental lipid-lowering agent.

Available trials of fenofibrate have failed to show major benefit on clinical outcomes.

Extended administration of this drug may reveal beneficial effects in selected patients.

1 Introduction

Cardiovascular disease (CVD) represents and will represent the first cause of morbidity and mortality in Western countries even into the next decades, not even sparing developing countries [1,2,3,4]. Dyslipidemia, on the other hand, plays a primary role in mechanisms leading from atherosclerosis to ischemic heart disease [5]. Even after considering genetic causes, dyslipidemia is a potentially modifiable risk condition [6]; however, impressive progress has been made in primary and secondary prevention of CVD, both in terms of lifestyle modifications and drug therapy. Evidence from randomized trials has progressively led international guidelines to recommend a lower and lower threshold for low-density lipoprotein cholesterol (LDL-C) levels, aiming at the consequent reduction of cardiovascular (CV) risk, especially in high-risk populations, including patients with diabetes mellitus (DM) [7, 8]. However, atherogenic dyslipidemia, a laboratory condition characterized by low levels of HDL-C combined with high levels of triglycerides frequently found in DM patients and responsible for what is named the residual cardiovascular risk, [9], is often still under rated and under treated. For all these reasons, in selected patients, lower LDL-C levels do not represent the only target; further risk stratification and an appropriate lipid-lowering agent as add-on therapy are often required.

2 Definition and Epidemiology of Atherogenic Dyslipidemia

Reduction of total cholesterol (TC) and LDL-C generally improves CV risk. However, a full evaluation of lipid profile can show that in 10% of the general population and in 15% of statin-treated patients there are persisting lipid abnormalities [10], characterized by elevated triglyceride (TG) levels and decreased levels of high-density lipoprotein cholesterol (HDL-C) regardless of LDL-C value. Beside this qualitative indication, no accepted cut-off level exists because of intrinsic heterogeneity of ranges of normality, age and ethnicity in the studies [11]. In general, the following can be considered as indicative limits: fasting TG > 204 mg/dL or > 2.3 mmol/L, HDL-C < 40 mg/dL or < 1.0 mmol/L for men and < 50 mg/dL or < 1.3 mmol/L for women [12, 13]. This lipid pattern is known as combined or mixed dyslipidemia on the base of higher TG and concomitant lower HDL levels [14], and, although not considered the major component of the vascular plaque, very recent studies show that TG-rich lipoproteins facilitate plaque generation [15]; hence, the name of atherogenic dyslipidemia. Furthermore, since TG levels negatively correlate with LDL particle size, a possible additional explanation of atherogenic potential is mainly attributable to LDL particles of smaller size (the small dense LDL) [16].

This lipid impairment is associated with such dysmetabolic patterns as obesity, type 2 diabetes mellitus (T2DM) and metabolic syndrome [17]. Given that worldwide very large number of people feature these phenotypes (almost 200 million patients with T2DM and 300 million are overtly obese), atherogenic dyslipidemia is anything but rare and harmless, even more so if we think that 65% of mortality in patients with T2DM is caused by CVD [18]. The strict interplay of atherogenic dyslipidemia with T2DM accounts for the name of diabetic dyslipidemia. Among all the molecules, fenofibrate seems the best choice against atherogenic dyslipidemia, because of lack of interaction with principal statins compared to gemfibrozil (inhibitor of OATP1B1, enzyme principally involved in simvastatin clearance), along with safety issues (clofibrate was withdrawn from the market for this last reason) [19, 20].

Safety and efficacy regarding the combination of bezafibrate with statins, even if positive, are based only on the few experiences with smaller samples size and brief follow-up [21, 22]. For this reason, the most important clinical trials investigated fenofibrate in clinical and laboratory outcomes of diabetic patients. Fenofibrate pharmacology and its non-cardiovascular use go beyond the scope of this paper and are discussed elsewhere [11, 23, 24].

3 Fenofibrate: A Trilogy of Success, Disappointment, and Revival

3.1 The DAIS, the FIELD and the ACCORD Lipid Trials: The New Hope

Three large multinational randomized placebo-controlled trials were held between 1996 and 2010: the DAIS, the FIELD and the ACCORD Lipid trials. The DAIS (Diabetes Atherosclerosis Intervention Study) [25] was a randomized multinational angiographic study designed to demonstrate a slower coronary artery disease progression in patients treated with fenofibrate compared to placebo. A smaller decrease in lumen diameter (− 0.06 mm vs − 0.10 mm, p = 0.029), and a smaller increase in percentage diameter stenosis, (2.11 vs 3.65%, p = 0.02) were shown in fenofibrate recipients after 3 years. Despite significant results, the study was not powered to investigate clinical events, which however were slightly higher in the placebo group. The results were attributed to a substantial change in lipid abnormalities with TG reduction and HDL-C increase, as confirmed by a sub-analysis [26].

The FIELD (Fenofibrate Intervention and Event Lowering in Diabetes) [27] was a multinational, randomized controlled trial with 9795 participants aged 50–75 years with well-controlled T2DM, which aimed to evaluate the effect of fenofibrate 200 mg daily. After an average of 5 years’ follow-up, the hazard ratio (HR) for the primary composite endpoints for coronary events was not statistically significant [0.89, 95% confidence interval (CI), 0.75–1.05; p = 0.16], but a subgroup analysis showed a reduction in nonfatal myocardial infarction (HR = 0.76, 95% CI 0.62–0.94; p = 0.010) and also a 21% decrease in coronary revascularization (p = 0.003), while total mortality, total stroke and total CVD mortality were similar. In addition more significant effects were observed in terms of less albuminuria progression (2.6% more patients allocated fenofibrate than placebo regressing from or not progressing to albuminuria, p = 0.002), and less retinopathy needing laser treatment (5.2 vs 3.6%, p = 0.0003), this indicating that the major benefit regarded microvascular complications. This beneficial aspect was also confirmed by a nation-wide Japanese observational analysis, where fenofibrate, as add-on therapy, showed to decreased incidence of diabetic retinopathy [odds ratio (OR), 0.772; 95% CI 0.720–0.827; p < 0.001] [28].

However, the lack of significant difference was because a larger number of patients who were treated with placebo were then started on a statin; a post hoc analysis revealed that the efficacy was more notable in patients with higher risk profile (marked dyslipidemia and hypertension) [29].

The ACCORD (Action to Control Cardiovascular Risk in Diabetes) Lipid trial determined the effect of daily fenofibrate 200 mg plus simvastatin 20 or 40 mg compared to simvastatin alone in terms of cardiovascular events in a 5-year follow-up [30]. In spite of the expectations and lipidemic improvement, combined therapy was not superior to simple statin treatment in reducing the primary composite endpoint of major CV events (non-fatal myocardial infarction, non-fatal stroke and death from CV cause, HR = 0.92, 95% CI 0.79–1.08, p = 0.32). Neither secondary endpoints (primary endpoint plus revascularization or hospitalization, non-fatal myocardial infarction, stroke, all-cause mortality, fatal or nonfatal congestive heart failure) were significantly different (Table 1); however, beneficial effects were confirmed in microvascular complications (less renal disease and diabetic retinopathy progression) [30, 31]. In a pre-specified subgroup analysis, a relative-risk reduction of 31% was observed in patients with more severe atherogenic dyslipidemia (TG > 204 mg/dL, and HDL-C < 34 mg/dL) [32], although this difference was not statistically significant (p for interaction 0.06).

Table 1 Study characteristics

3.2 The SAFARI and the DIACOR Trials: The PPAR-α Strikes Back

The SAFARI trial was another milestone which showed a substantial reduction by 23.6% of TG levels, 5.4% in LDL-C and an 8.8% increase of HDL-C (p < 0.001 for all) by administering daily fenofibrate 160 mg combined with simvastatin 20 mg compared to monotherapy [33]. No drug-related serious adverse events were revealed in 4 months of treatment.

The DIACOR (Diabetes and Combined Lipid Therapy Regimen) study demonstrated greater reduction of TG in 12 weeks with combination therapy (fenofibrate 160 mg plus simvastatin 20 mg daily) compared with both monotherapies [34]: − 49 versus − 25% (simvastatin subgroup) and − 38% (fenofibrate subgroup) (p < 0.0001 and 0.07, respectively). As found in a sub-analysis, dual therapy also lowered small dense-LDL-C (sd-LDL-C) subset [35], and lipid pattern improvement with combination therapy was confirmed in other smaller studies [36, 37]. According to a meta-regression analysis based on these studies [11], fenofibrate 160 mg is the optimal dosage, offering an effective TC reduction (ranging from − 27 to − 43%), regardless of simvastatin dosage, with theoretical minor risk of collateral effects.

Interesting evidence also illustrated that even surrogate biomarkers of cardiovascular risk decreased, and the DIACOR showed that inflammatory molecules such as high-sensitivity C-reactive-protein (hs-CRP) and lipoprotein-phospholipase A2 (LP-PLA2) were reduced in all treated groups [34]. Daily fenofibrate association ameliorated endothelial function measured as flow mediated dilation and insulin sensitivity [38,39,40], diminishing atherogenic cytokine levels [IL-1b, IL-6, interferon-(IFN)-gamma and tumor necrosis factor (TNF)-α] in high risk patients [41, 42].

Consequently, after disappointing results from the ACCORD Lipid trial, it was only natural to conclude that fenofibrate could likely cause just a cosmetic effect on lipid profile, without any additional significant clinical cardiovascular benefit.

3.3 The ACCORDION and the EFECTL Studies: The Return of the Fenofibrate

In the 2017, the results of the ACCORDION study, a post-trial passive follow-up of the ACCORD Lipid trial, were published. The aim of the study was to verify whether fenofibrate extended add-on treatment reduced CVD risk compared to patients originally treated with placebo [43]. Once again, results showed that the primary composite outcome of fatal and nonfatal myocardial infarction and stroke were similar in the two groups (HR = 0.93, 95% CI 0.83–1.05; p = 0.25) after a median follow-up of 9.7 years, although a subgroup analysis further revealed that subjects with severe dyslipidemia (TG > 204 mg/dL and HDL-C < 34 mg/dL) did benefit to a greater extent (HR = 0.73, 95% CI 0.56–0.95, p = 0.05). Additionally, reduction in CV mortality almost reached statistical significance (HR = 0.84, 95% CI 0.69–1.01, p = 0.07), while an unexpected better response in men compared to women was found (HR = 0.84, 95% CI 0.73–0.96, vs HR = 1.30, 95% CI 1.10–1.68, p = 0.003). Given that only 4.3% of patients continued the peroxisome proliferator-activated receptor alpha (PPAR-α) after the original trial end, the authors of the ACCORDION study interpreted the results as a proof of the tardive legacy effect of fenofibrate.

The EFECTL (Effect of Fenofibrate and Ezetimibe Combination Treatment on Lipid) determined that 52 weeks of daily co-administration of ezetimibe 10 mg plus fenofibrate (160 mg or 200 mg in equivalent formulations) effectively reduced LDL-C by 29% and TC by 24% compared to single monotherapies (p < 0.001) [44]. TG decrease was significant only when compared to ezetimibe alone (− 45%) (p < 0.001), with the greatest percent reduction at week 12. Furthermore, an increase in LDL size (measured in nanometer) was found using fenofibrate in combination or alone compared to ezetimibe (+ 3.4%, p < 0.001); side effects were similar in the three groups.

4 Safety: A Phantom Menace

The most common adverse reactions to fenofibrate are hepato- and myotoxicity, either clinically silent or evident. The association with statins can increase the risk of serious adverse reaction for both pharmacodynamic and pharmacokinetic interaction.

Fibrate-related skeletal muscle damage is probably ascribed to reactive oxygen species derived from β-oxidation and mitochondrial dysfunction induced by fibrate, as shown in animal models treated with both potent and weak (fenofibrate) compounds [45, 46]; however, experimental doses are very much higher than those used in humans.

Similar origin for hepatotoxicity has been speculated, even if recent evidence suggests that PPAR-α agonist can directly increase alanine amino transferase (ALT) and aspartate aminotransferase (AST) gene expression and can also shift hepatic metabolism into a higher ALT/AST ratio [47]. As was observed in FIELD study, fibrates increase cholelithiasis risk, altering cholesterol biliary efflux, and accordingly pancreatitis is significantly more probable (0.5 vs 0.8%, p = 0.031) [29]. Indeed, gallbladder disease is a contraindication to the use of fibrates and, it is our opinion, that an abdominal ultrasound should be performed before starting the therapy.

Also, in all the aforementioned trials, serious adverse reactions were isolated cases: for instance, in the SAFARI trial no patient experienced severe abnormalities in liver function and clinical myopathy, and there were no cases of rhabdomyolysis. Asymptomatic creatine kinase (CPK) increased by ten times the upper limit of normality (ULN) in only one patient in the combination therapy group and in none in the simvastatin-only cohort—difference in proportion 0.2 (− 1.6, 1.4, 95% CI). No muscular adverse events (simple myalgia, stiffness, etc.) were associated with abnormal increases in CPK. However, there was a significant difference (p = 0.03) between treatment groups for ALT elevation > 3 times the ULN (9 vs 0 patients) without further serious clinical consequences (no case of rhabdomyolysis was recorded) [33]. In the ACCORD Lipid study (the trial with the longest follow-up and the highest number of patients enrolled), the incidence of myotoxicity (myopathy, myositis, and rhabdomyolysis) in the treatment group was equally balanced to placebo plus simvastatin recipients (0.1% for both) [31]. Although nephron-protection resulted in the long term, fibrate co-administration caused transient creatinine elevation (> 1.3 mg/dL for women; > 1.5 mg/dL for men), but no cases of acute renal failure were reported. Despite collateral homocysteine increase, no deep venous thrombosis or pulmonary embolism occurred, in contrast to data recorded in the FIELD study (0.7 vs 1.1%, p = 0.022) [29]. Additionally, absolute number of deaths was definitely comparable between the two groups (203 in the fenofibrate plus simvastatin group vs 221 in placebo). Overall risk of myotoxicity was very low in combination therapy, including asymptomatic and occasional CPK increment > 5 or > 10 times the ULN (usually between 0.3 and 2.2%) [31].

Minor drug adverse reactions, such as allergy, cataract, and interstitial lung disease, have been described in sporadic cases.

Finally, despite overall rarity of serious adverse events and the net beneficial effect of lipid-lowering medications in terms of primary and secondary cardiovascular prevention, physicians should be aware of safety aspects and frequently monitor patients on high dosage of statins and other drugs whose interferences may be harmful (almost exclusively for simvastatin) or with preexisting risk factors (chronic kidney disease, hepatic insufficiency, muscular disease, age, and hypothyroidism) [11, 48]. It is generally recommended to measure ALT and CPK at baseline and after some weeks of treatment or at the first suspected symptoms in healthy subjects. Food and Drug Administration (FDA) suggests testing renal function periodically in elderly patients or those with renal insufficiency [49]. Transaminases and CPK should be monitored periodically in any case; however, bilirubin is a more reliable indicator of damage, as National Lipid Association’s Liver Expert Panel states [50].

5 Commentary

Paradoxically, fenofibrate is a well-tolerated and safe molecule, able to effectively correct lipid abnormalities, improve endothelial function and balance systemic inflammation, but poor cardiovascular benefit was shown in randomized trials. Current European guidelines recommend the use of fenofibrate in cases of hypertriglyceridemia (> 200 mg/dL) (class IIb, level of evidence C) [7], while American recommendation is more supportive (grade A, best evidence level 1) [8], especially if HDL-C is < 40 mg/dL. Only post hoc subgroup analysis indicated that fenofibrate in association with other lipid-lowering agents reduces clinical events, but no effect was reported in terms of CV mortality. Nevertheless, a decrease by 16% of the residual risk of death due to CV disease, heart attack, or stroke in patients with established CVD was found in a Cochrane meta-analysis, which included both fenofibrate in monotherapy and on top of another lipid-modifying agent. A very scant effect was achieved in primary prevention [51], yet fenofibrate highly reduces microvascular complications, with lower diabetic nephropathy, retinopathy and limb amputation rates.

Where is the solution of the rebus? Perhaps, the most plausible answer is that we inappropriately tested the right drug in the wrong patients as the model of the precision medicine would instead recommend [52]. It is noteworthy that the FIELD study was conducted in a pre-guidelines age, when statin was not mandatory as it is nowadays, and furthermore the necessity of lipid-lowering therapy was even an exclusion criterion. Maybe combination therapy in selected patients is right way to use the right drug.

Moreover, in the ACCORDL study the vast majority of enrolled subjects did not suffer from atherogenic dyslipidemia (median TG levels 162 mg/dL), and fenofibrate was associated with a moderate-intensity statin. Thus, it now appears clear why the best results were observed in sub-population with true atherogenic dyslipidemia and elevated residual risk, that is the persisting risk despite the evidence-based standard of care.

Finally, despite the fact that improvement in lipid patterns occurs within a few months, the ACCORDION study showed that primary endpoint did not change after an additional 5 years of follow-up, and statistical significance was almost reached for cardiovascular mortality after less than ten years. Therefore, another consideration to be drawn is that, besides a more appropriate study population, maybe fenofibrate requires a longer period of time to produce its clinical effects.

Interesting, although not definitive, results from pharmacogenomics reinforce this concept [53,54,55]. In fact, sequencing the PPAR-α gene in 300 patients from the GOLDN study (Genetics of Lipid Lowering Drugs and Diet Network), 13 single nucleotide polymorphisms (SNPs)—as for instance rs4253793 and rs41332048—were associated with a very poor response after three weeks of treatment with fenofibrate (odds ratio = 6.46 for carriers of > 1 SNP; 95% CI 1.4–30.8, p = 0.02) [53]. In particular, lipoprotein lipase (LPL) mutation with gain-of-function (p.S447*) either in homo- and heterozygosis, was found to negatively correlate with the use of fenofibrate compared to placebo in terms of the cardiovascular events in 4414 participants of the ACCORD (RR 1.56, 95% CI 0.98–2.47, p = 0.01) [56].

Taken together, these results indicate that fenofibrate could still play a pivotal role in cardiovascular secondary prevention when used as add-on therapy, but perhaps, the more appropriate is the selected population—probably subjects with true atherogenic dyslipidemia—the more effective is the drug. Further studies with longer-term outcomes, stricter clinical inclusion criteria, hopefully with the benefit of genetic innovations, are warranted to confirm this hypothesis.

Pemafibrate, a promising third-generation molecule (named SPPARM, selective-PPAR-modulator-α), appears non-inferior and significantly safer compared to fenofibrate 200 mg daily [57, 58]. The PROMINENT trial (Pemafibrate to Reduce Cardiovascular Outcomes by Reducing Triglycerides in Patients with Diabetes) will test whether or not PPAR-α agonists are beneficial, exclusively enrolling patients with overt atherogenic dyslipidemia treated with pemafibrate [59].

6 Conclusions

Fenofibrate therapeutic efficacy is limited to specific populations, which should be actively pursued by clinicians as a prescription target. Future and better-designed trials are warranted to more definitely identify the place of fenofibrate in the pharmacological armamentarium.