|Home | About | Journals | Submit | Contact Us | Français|
Aggressive reduction of LDL cholesterol by the use of statins is a cornerstone of preventive cardiovascular care, but additional therapies to prevent atherosclerosis and its clinical sequelae are still needed. Thyromimetics selective for the liver or the thyroid hormone receptor isoform β1 constitute a novel approach to treat dyslipidemia. In preclinical studies, selective thyromimetics were clearly shown to reduce plasma cholesterol and to protect from atherosclerosis through the upregulation of hepatic LDL receptor and through the promotion of the so-called reverse cholesterol transport. Importantly, there is first evidence from on-going clinical trials that selective thyromimetics may reduce plasma cholesterol also in humans.
Atherosclerosis still represents the leading cause of death and morbidity among adults in the developed countries. Aggressive reduction of low density lipoprotein (LDL) cholesterol is a cornerstone of preventive cardiovascular care , but additional therapeutic approaches to reduce atherogenesis are still needed. The aim of this review is to discuss the current resurgence of thyromimetics as hypolipidemic drugs, and the underlying mechanisms conferring their lipid-lowering action including increased LDL clearance and the promotion of reverse transport of cholesterol from extrahepatic tissues, eg atherosclerotic plaque macrophages, back to the liver for fecal excretion.
It has been known since 1930 that hyperthyroidism is associated with reduced plasma cholesterol levels , and since then many efforts were made to exploit the ability of thyroid hormones (TH) to lower cholesterol [3, 4]. In initial studies in the 1950s, desiccated thyroid was administered to a small group of patients, all of whom responded with a fall in cholesterol . A low dosage of dessicated thyroid led to a significant reduction in plasma cholesterol, “escape” occurred after 20-30 weeks of treatment . Patients treated with high doses of desiccated thyroid were not refractory to treatment, but a large number presented with tachycardia, angina pectoris, diarrhea, weight loss and insomnia, in brief, with overt hyperthyroidism . Thus, studies with thyroid preparations were stopped at that time, and the search for synthetic analogs began.
A large number of TH analogs were synthesized and tested in experimental animal models for their lipid-lowering activity (summarized in ). Among all these analogs tested in animal studies, dextrothyroxine (D-T4) appeared to have the highest specific cholesterol-lowering action, without showing concomitant deleterious effects on the heart. In the late 1960s, a large clinical trial of D-T4 therapy was conducted, as part of The Coronary Drug Project by the National Institutes of Health, which aimed to answer the question as to whether cholesterol reduction may prevent coronary heart disease . The study was terminated after average follow-up of 36 months due to a higher proportion of deaths in the D-T4 treated group, although this difference did not reach statistical significance. However, the design and performance of this study may have not been sufficient to elucidate the lipid-lowering effect of D-T4 in humans. First, subsequent investigations revealed the preparations used in the D-T4 study to be contaminated with as much as 0.5% levothyroxine, the enantiomer of D-T4, equivalent to a dosage of 30 μg per day, which may have been the only active metabolite of the study . Secondly, the deaths occurred in patients already carrying high cardiovascular morbidity at the initiation of the study, including angina pectoris, congestive heart failure, and tachycardia. After exclusion of high risk patients, the overall survival in the D-T4-treated group was greater than in the controls . The unfavourable recruitment of patients together with the accidental employment of preparations contaminated with the enantiomer of D- T4 led to the discontinuation of clinical studies with TH analogs in the 1970s.
With the introduction into clinical practice of 3-hydroxy-3-methyglutaryl coenzyme A reductase (HMG CoA reductase) inhibitors, usually known as ‘statins’, to lower plasma cholesterol in the mid 1980s, efforts on the development of TH analogs slowed. It was in this time period, however, that the first novel ‘selective’ compounds mimicking the cholesterol-lowering actions of TH were developed.
The first described selective thyromimetic compound was 3,5-dibromo-3′-pyridazinone-L-thyronine (L-94901), which showed half of the binding affinity of T3 to hepatic TRs, but only minor affinity (1.3% of T3) to cardiac TRs [12, 13]. This selective thyromimetic has been reported to lower plasma cholesterol levels in experimental animals at doses that do not exhibit cardiotoxic side effects. Very recently, another organ-selective compound with lipid-lowering properties was described, namely N-(4-[3-[(4-fluorophenyl)hydroxymethyl]-4-hydroxyphenoxy]-3,5-dimethylphenyl) malonamic acid sodium (T-0681, former KAT-681) [14, 15] (Figure 1). GC-1 (Sobetirome; 3,5-dimethyl-4[(4′-hydroxy-3′-isopropylbenzyl)-phenoxy] acetic acid), an organ- and TRβ1-selective compound, is discussed below.
In the late 1980s, different TR isoforms (TRα, TRβ) were cloned [16, 17], and their tissue-specific expression characterized (reviewed in [18, 19]). Both, TRα and TRα occur in two isoforms, respectively . Expression of TRβ2 was found to be restricted to the brain and anterior pituitary gland where it plays a key role in mediating negative feedback of TH on the thyroid hypothalamic axis (THA). TRα2 is a nonhormone binding receptor acting as negative regulator. TRα1 and TRβ1 are functional receptors which bind T3 with high affinity. Both were found to be expressed in virtually all tissues, with TRβ1 being predominantly expressed in liver (80% of T3-binding in this organ). Evidence that tachycardia is mediated by the TRα1 came from studies in TRα knockout mice, which displayed a slow pulse rate that could not be increased by administration of even large doses of T3 [21, 22]. Furthermore, experiments with TRβ knockout mice suggested that the effect of T3 on plasma cholesterol is mediated through TRβ1 . These findings led to the design of isoform-specific, TRβ1-selective thyromimetics, such as KB-141 (3,5-dichloro-4[(4-hydroxy-3-isopropylphenoxy)-phenyl] acetic acid), and compounds currently tested in clinical trials such as GC-1 (Sobetirome), KB2115 (Eprotirome; 3-[3,5-dibromo-4-[4-hydroxy-3-(1-methylethyl)-phenoxy]-phenyl]-amino]-3-oxopropanoic acid), and MB07811 ((2R, 4S)-4-(3-chlorophenyl)-2-((3,5-dimethyl-4-(4′-hydroxy-3′-isopropylbenzyl)phenoxy)methyl)-2-oxido-(1,3,2)-dioxaphosphonane) (Figure 1). Previously developed substances CGS23425 (N-[3,5-dimethyl-4-(4′-hydroxy-3′-isopropylphenoxy)-phenyl] oxamic acid), DITPA (3,5-diiodothyropropionic acid) were found to be TRβ1-selective as well.
The primary mechanism for TRβ1-selectivity of GC-1 is related to the presence of the oxo-acetate at position 1. This group forms enhanced polar interactions with conserved arginine residues in a hydrophilic part of the THRβ pocket . KB141 exploits a similar mechanism of selectivity, although there are also weaker contributions of thyronine ring substituents. TR-isoform selectivity arises from the fact that the region that rearranges to accommodate the phenyl extension group seems to be more flexible in TRβ than in TRa .
Organ-selectivity of thyromimetics is thought to be related to high rates of liver first-pass uptake, as well as differences in cellular uptake and retention mechanism . Of note, GC-1 and KB2115 combine both, organ- and TRβ1-selectivity, which may enhance hepatic targeting. A further interesting approach to enhance hepatic targeting has been the development of the liver-selective prodrug MB07811 that is activated within hepatocytes by enzymatic cleavage and distributes poorly into other tissues .
All of the above mentioned thyromimetic compounds were demonstrated in preclinical animal studies to markedly lower plasma cholesterol without deleterious effects on the heart [13-15, 25-29]. This effect is thought to be mainly due to increased LDL-cholesterol plasma clearance through increased LDL receptor (LDLr) expression in liver, similarly as described for TH action [30-32]. Of note, recent mechanistic studies in mice showed LDLr expression to be crucial for the effect of selective thyromimetics on lipid metabolism, as LDLr KO mice did not respond to treatment with MB07811,  or T-0681 (xxx, unpublished data). These data suggest that patients with a homozygous mutation of the LDLr, i.e. familial hypercholesteremia, may not benefit from a treatment with thyromimetic compounds.
In different animal models, including mice, rats and rabbits, thyromimetics also decreased plasma triglycerides (TG), which are naturally associated with apolipoprotein B (apoB)-containing lipoproteins such as LDL and very low density lipoproteins (VLDL) (summarized in , ). Accordingly, it is probable that a reduction in plasma TG simply reflects the increased uptake of LDL particles into liver cells; on the other hand, as suggested by Baxter and Webb , the inhibition of hepatic transcription factor sterol regulatory element-binding protein 1 (SREBP1) may result in reduced VLDL assembly. Accordingly, hypercholesterolemic NZW rabbits treated with T-0681 displayed a reduction of LDL and VLDL cholesterol, as well as of apoB mass , which probably resulted from enhanced hepatic clearance, reduced secretion into plasma or the combination of both. Similarly to the animal studies, both KB2115 and GC-1 reportedly lowered plasma LDL cholesterol by 40% and significantly reduced plasma triglycerides also in humans [33-35]
The concept of reverse cholesterol transport (RCT) describes the process by which extrahepatic (peripheral) cholesterol is returned to the liver via high density lipoproteins (HDL) for excretion in the bile and ultimately the feces . Most non-hepatic cells, eg macrophages, accumulate cholesterol through uptake of lipoproteins and de novo synthesis and are unable to catabolize it. Excess unesterified cholesterol is toxic to cells and initiates local inflammation; accordingly, hyperlipidemia leads to migration of macrophages into the arterial wall, which represents a crucial step of atherosclerosis development. Novel therapeutic strategies aim to promote the reverse transport of cholesterol from such atherogenic macrophages back to the liver, and/or to promote hepatobiliary flux of excess cholesterol to disrupt the vicious circle taking place in the vasculature.
GC-1 and T-0681 were reported to stimulate the expression of key players of RCT in mice and rabbits [14, 15, 37]. In mice, both compounds markedly induced the expression of the HDL-receptor SR-BI in the liver, stimulated the activity of cholesterol 7α-hydroxylase (CYP7A1), the rate-limiting enzyme of bile acid synthesis, and induced the expression of hepatic ABCG5 and ABCG8 (ABCG5/G8), which promote biliary cholesterol secretion. As a consequence, treated animals displayed increased turnover of plasma HDL cholesterol, and increased fecal excretion of bile acids and cholesterol. Mice treated with T-0681 displayed reduced intestinal absorption of dietary sterols, most likely due to competition with sterols of biliary origin . Very recently, Parini and coworkers demonstrated KB495, a novel liver-selctive TRβ1 activator to induce hepatic CYP7A1 expression in mice, which was associated with increased fecal bile acid and neutral sterol excretion . Baxter and colleagues reported that human subjects treated with KB2115 displayed an increase in plasma 7α-hydroxy-4-cholesten-3-one, a surrogate marker of bile acid synthesis . Thus, the promotion of bile acid synthesis represents an inherent pharmacological principle of selective thyromimetics.
The hypothesis of promotion of RCT by thyromimetics was tested by measuring RCT from macrophages to feces in mice treated with T-0681 , according to the method developed by Rader and coworkers . 48 h after intraperitoneal injection of cholesterol-loaded, [3H]-labeled J774 macrophages, T-0681 treated animals displayed a significant increase of both, fecal [3H]-bile acids and [3H]-cholesterol . Thus, employment of thyromimetics may promote the reverse transport of cholesterol from atherosclerotic plaque macrophages to the liver for fecal excretion.
RCT in humans is different from that found in rodents in that cholesterol from plaque macrophages can be transported to the liver either directly via HDL particles, or – after transfer to VLDL and LDL mediated by cholesteryl ester transfer protein (CETP) – via apoB-containing lipoproteins . Accordingly, CETP-transgenic mice require hepatic expression of LDLr to counterbalance accumulation of apoB-containing lipoproteins in plasma . In line with this finding, adenoviral overexpression of SR-BI in rabbits, naturally expressing CETP in plasma, led to accumulation of VLDL and LDL cholesterol . Thus, hepatic stimulation of SR-BI expression necessitates a concomitant, appropriate clearance of apoB-lipoproteins to maintain RCT in CETP-expressing species like humans.
In summary, thyromimetics represent a rational approach for the promotion of RCT in humans, as they lead to simultaneous upregulation of hepatic SR-BI and LDLr directing excessive cholesterol from the periphery to the liver, and to simultaneous activation of hepatic CYP7A1 and ABCG5/G8 directing cholesterol from the liver to the feces. The mechanisms conferring the lipid-lowering action of selective thyromimetics are summarized in Figure 2. However, extrapolation of mechanistic data from animal studies to the situation in humans should be done with caution, as most of the animals used to study thyromimetics were mice, which in contrast to humans transport plasma cholesterol mainly in HDL particles, do not express CETP in their plasma, and do not develop atherosclerosis.
The first published evidence that accelerated clearance of LDL cholesterol and promotion of RCT by a thyromimetic may constitute a powerful approach to prevent the development of atherosclerosis came from studies in cholesterol-fed NZW rabbits treated with T-0681 [14, 15]. T-0681 reportedly decreased plasma cholesterol by 60%, triglyceride levels by more than 80%, it increased hepatic expression of SR-BI and LDLr, whereas no effect on plasma CETP activity was found. Lipid staining of rabbit aortas revealed an 80% decrease in atherosclerotic lesion area, when compared to placebo-treated controls. In summary, current data strongly suggest selective thyromimetic compounds to have great clinical potential as agents to treat hyperlipidemia and to protect from atherosclerosis and its clinical sequelae.
There are three main untoward effects of selective thyromimetics to be mentioned; relative hypothyroidism, elevation of liver enzymes, and cardiotoxictiy.
Ideally, a selective TR agonist would cause modest increase in metabolic rate without tachycardia but would not reduce thyroid stimulating hormone (TSH) and/or T4, as observed with most of the selective analogs at therapeutic doses (reviewed in ). A reduction in TSH and subsequently T4 may cause a paradoxical hypothyroidism in some tissues. However, so far there are no data available on the consequences of such a relative hypothyroidism. To overcome the inhibition of the thyroid hypothalamic axis (THA), two potential approaches are worth of note: first, in rats, enhanced liver targeting with intrahepatic activation of the prodrug MB07811 and subsequent fast biliary clearance allowed for a reduction in circulating plasma levels of the active compound. As a consequence, MB07811 reduced plasma cholesterol and triglycerides at doses devoid of any negative effects on the THA . Second, combination of thyromimetics with other hypolipidemic drugs such as statins may allow for a reduction of the employed dosage, thus minimizing the inhibition of the THA .
T-0681 and KB2115 were reported to induce an elevation of transaminases, both in normolipidemic NZW rabbits and in humans [15, 33]. In rabbits, this effect was shown to be dose-dependent; liver enzymes returned to normal levels after cessation of treatment . On one hand, the organ-selective action of these compounds constitutes the prerequisite for their lipid-lowering effects, on the other hand this liver-selectivity may lead to accumulation in hepatocytes, and lead to the observed toxic damage. Thus, strategies to accelerate biliary clearance of thyromimetic compounds may help to overcome the observed hepatitis.
Another possible untoward effect is tachycardia. However, the use of TRβ1-selective thyromimetics is likely to be safe, as GC-1 was reported to have a relative selectivity for cholesterol-lowering versus tachycardia of 18-fold , when compared to T3. KB-141 was shown to be approximately 27-fold more selective for cholesterol-lowering when normalized for T3 . For both compounds, tachycardia was observed exclusively at highest dosages, which may be explained by a loss of selectivity in such an oversaturated setting.
Neither KB2115 (Eprotirome) nor GC-1 (Sobetirome) elicit harmful effects follow ing 2 weeks of dosing . There were no changes in heart rate or electrocardiogram and echocardiography readings at any dose of compound. As mentioned above, mild increases of serum liver enzymes were reported. KB2115 did affect circulating thyroid hormone levels. Total and free T4 levels fell, but remained in the lower normal range; at the higher dose, T3 levels were slightly reduced. TSH levels were unchanged, which suggests that reduced T4 levels in humans could be related to increased clearance following transcriptional induction of DIO1, the enzyme that converts T4 to T3 in the liver . Reductions in thyroid hormone levels were fully reversible at follow up. Furthermore, clinical scales for hyperthyroidism (the Hyperthyroid Symptoms Scale) and hypothyroidism (the Billewicz Scale) were unchanged .
As the here described potential side-effects were shown to occur in a dose-dependent fashion, dosing regimens in humans will need to be tightly controlled.
The first clinical trial with a thyromimetic from KaroBio (Eprotirome, KB2115, ) has been published in 2008 , results with GC-1 (Sobetirome) have been presented at a scientific meeting , and Metabasis announced the successful completion of a Phase Ib trial with MB07811 . To summarize these results, the tested thyromimetics significantly reduced plasma LDL cholesterol and TG levels by up to 40% in a dose-dependent fashion in both normolipidemic and hyperlipidemic subjects without severe adverse events .
As outlined by Baxter and Webb , a first randomized, placebo-controlled, double-blinded study of Eprotirome in 98 subjects with hypercholesterolaemia was conducted. Following a 4 week diet-lead-in period, patients were given placebo, 100 μg or 200 μg of KB2115 for 12 weeks and followed for 4 weeks after discontinuation of the drug. Plasma LDL–cholesterol levels were reduced at both doses of KB2115 by 30% compared with placebo, and returned to control levels following discontinuation.
A recent press release by KaroBio  suggested that KB2115 was just as effective in patients who are already on statin therapy as in those who are not, indicating that the two classes of drug could be used together ; this phase IIb study included 189 patients. The results show that eprotirome, in a dose dependent manner, significantly lowered LDL-cholesterol and triglycerides, when added to statins. This conclusion is corroborated by a recent study in primates where MB07811 and atorvastatin had additive effects on plasma cholesterol . As statins are the standard of clinical care, any new therapy should have adjunctive activity, when given in combination with statins , and selective thyromimetics appear to qualify for a future combination therapy.
Finally, a third phase II study with Eprotirome is ongoing where the compound is evaluated as add on to treatment with the cholesterol absorption inhibitor ezetimibe .
In conclusion, efforts spanning more than 50 years led to the development of selective thyromimetic compounds, a resurgent drug class for the treatment of dyslipidemia, and for prevention of atherosclerosis. Drugs of this type may be useful in treating patients who do not tolerate statins, or who do not respond to this type of medication. They also may be used to treat obesity, hepatic steatosis, congestive heart failure and type 2 diabetes (reviewed in ). Current studies do not provide sufficient data to respond to the central question as to whether selective thyromimetics will reduce cardiovascular morbidity and mortality in a long-term clinical application. However, we remain very enthusiastic about the resurgence of an old approach to treat dyslipidemia which was rendered possible by the translation of recent molecular discoveries into drug development.