Non-alcoholic fatty liver disease (NAFLD) includes a range of chronic conditions characterized by excessive hepatic lipid accumulation, defined by the presence of steatosis in >5% of hepatocytes, in the absence of significant alcohol consumption or other causes of liver injury. This entry presents a comprehensive review of current experimental approaches to pharmacological treatment of NAFLD with a special focus on recent evidence from clinical trials.
In addition to their primary mechanism of action, pioglitazone and other thiazolidinediones are known to interact with the mitochondrial pyruvate carrier (MPC) to suppress pyruvate transport into the mitochondrial matrix. Since this direct, non-genomic effect is considered essential for the inhibition of hepatic gluconeogenesis by PPARγ ligands, novel MPC-inhibiting thiazolidinedione derivatives with minimal affinity towards PPARγ have been synthesized.
Azemiglitazone (MSDK-0602K) reduced liver enzyme levels in NASH/fibrosis patients regardless of T2DM presence [105], and was characterized by a markedly improved safety profile [106]. However, the drug did not demonstrate significant effects on any of the histological endpoints [105]. A phase 3 trial to evaluate azemiglitazone in diabetic or prediabetic patients with NAFLD/NASH has been initiated (NCT03970031). An alternative approach to minimizing the side effects of PPARγ agonists is represented by PXL065, the deuterium-stabilized R-isomer of pioglitazone [107], which is currently being assessed in a phase 2 RCT (NCT04321343).
Incretins are a small family of intestinal L-cell-derived peptide hormones that includes glucagon-like peptide 1 (GLP1), glucose-dependent insulinotropic polypeptide (GIP), and oxyntomodulin. The incretin axis mediates the physiological response to hyperglycaemia and couples glucose intake with pancreatic secretion [108]. GLP1 receptors (GLP1R) are expressed in the β-cells, hepatocytes, white adipose tissue, brain, and skeletal muscle [109]. GLP1R activation induces insulin secretion, decreases insulin resistance, inhibits glucagon release and lipogenesis, and suppresses appetite and gastrointestinal motility. Additionally, hepatic GLP1R stimulate FA β-oxidation, inhibit profibrogenic signaling pathways, and exert a mild anti-inflammatory effect by indirectly reducing CRP, proinflammatory cytokine, and chemokine production [110].
At the moment, semaglutide is the only GLP1R agonist being developed for the treatment of NASH in nondiabetic subjects. In the recently completed 72-week phase 2 trial, semaglutide treatment (0.4 mg/d) led to NASH resolution with no worsening of fibrosis in 59% vs. 17% in the placebo group, which is considered the highest response rate that a drug has ever achieved in a NASH trial up to now [111][112]. Currently, semaglutide is being evaluated in several phase 2 trials as monotherapy (NCT03884075, NCT04216589) as well as in combinations with empagliflozin (NCT04639414), and cilofexor/firsocostat (NCT04971785). A 5-year-long phase 3 study designed to include 1200 patients with precirrhotic NASH has been initiated (NCT04822181).
Exenatide, lixisenatide, liraglutide, and dulaglutide have all demonstrated significant antisteatotic activity, and (with the exception of lixisenatide) improved intrahepatic cholestasis in T2DM/NAFLD patients. Amelioration of cytolysis markers was reported for exenatide [113][114] and dulaglutide [115][116], and lixisenatide was effective against liver fibrosis and inflammation [117]. Liraglutide reduced hepatitis activity and liver fibrosis as well as attenuated HCB in NASH patients regardless of the presence of T2DM, as determined by the phase 2 LEAN study [118]. A recent meta-analysis by Ghosal et al., including 8 RCT and over 600 patients, found that GLP1R agonists in general improve liver function and histology by improving glycaemia, reducing body weight and hepatic fat content, which in turn might be beneficial for hepatic inflammation in NAFLD concomitant with T2DM [119].
While both GIP and GLP1 are potent insulin secretagogues, GIP has a more robust, dose-dependent secretory pattern, and appears to make a greater contribution than GLP1 to prandial insulin secretion in healthy subjects, while in T2M its activity is depleted. In addition, GIP, but not GLP1, stimulates glucagon secretion by the α-cells at low glycaemia under physiological conditions, and in a glucose-independent fashion in T2M [108]. GIP receptors (GIPR) upregulate lipogenesis, FA esterification and TAG accumulation in adipocytes, and inhibits prandial lipid absorption. Supraphysiological levels of GIP have been associated with increased systemic inflammatory response, and are considered a risk factor for the development of NASH [120].
Tirzepatide (LY3298176) is a synthetic injectable dual GLP1/GIP peptide agonist currently researched for NAFLD treatment. In T2DM subjects with NASH, tirzepatide (1–15 mg/week for 26 weeks) effectively reduced ALT, AST, cytokeratin 18, Pro-C3, and increased adiponectin levels [121]. Additionally, tirzepatide treatment led to greater improvements in liver fat content compared to titrated insulin degludec in T2DM, according to the phase 3 SURPASS-3 RCT results [122]. A phase 2 study, designated SYNERGY-NASH, has been initiated to evaluate the efficacy of tirzepatide in nondiabetic subjects with NASH (NCT04166773).
Oxyntomodulin shares sequence similarity with both and GLP1 and glucagon, and activates GLP1R and glucagon receptors (GCGR) under physiological conditions. Simultaneous GLP1R activation prevents hyperglycaemic response characteristic of glucagon, at the same time potentiating its catabolic effects and greatly intensifying hepatic glycolysis, glycogenolysis, and lipolysis [123]. Weight reduction, anorexigenic and hypoglycaemic effects have been linked to GLP1 activation, while GCGR activation is thought to contribute primarily to hepatic steatosis attenuation and improved mitochondrial respiration. The clinical utility of oxyntomodulin itself is limited by a short circulatory half-life due to rapid renal clearance and degradation by dipeptidyl peptidase 4 (DPP4) [124].
In contrast to the native hormone, synthetic oxyntomodulin mimetics are resistant to proteolytic cleavage and have prolonged pharmacological action. Cotadutide (MEDI0382) (100–300 µg/d) caused substantial improvements in liver enzyme levels and markers of liver fibrosis in concomitant obesity, T2DM, and NASH in a phase 2b study that included over 800 subjects [125]. Efinopegdutide (HM12525A, MK-6024), a PEGylated long-acting peptide agent, has demonstrated promising antihyperlipidaemic, antisteatotic, and anti-inflammatory activity in mice and hamsters [126], and is going to be evaluated in a phase 2 RCT in NASH with semaglutide as an active comparator (NCT04944992). Other dual GLP1/GCCR agonists intended for use in NAFLD include pemvidutide (ALT-801) (NCT05006885), danuglipron (PF-06882961) (in combination with ervogastat) [127], BI 456906 (NCT04771273), and HM14320 (a glucagon-containing combination) [128]. Finally, a novel triple GLP1R/GCGR/GIPR agonist, HM15211, induced significant reductions in liver steatosis, fibrosis, and inflammation in mice [129]; a phase 2 clinical trial is ongoing (NCT04505436).
GLP2, usually not considered an incretin, is prevalent in the gastrointestinal tract, where it promotes lipid absorption, regulates intestinal motility, mucosal morphology, function and integrity of the intestine [130]. Teduglutide, a selective GLP2R agonist, reduced liver steatosis and disease activity scores in rats, possibly by restoring normal intestinal permeability [131].
DPP4 inhibitors represent a group of indirect incretin mimetics as they prevent the proteolytic cleavage of GLP1, GIP, and oxyntomodulin. To threre is e best of our knowledge, no DPP4 inhibitors are yet in the global pipeline for liver disease. However, a number of small-scale clinical trials have evaluated their potential efficacy in NAFLD in the presence or absence of concomitant T2DM. Among this group, only sitagliptin (100 mg/d) was found effective against hepatic steatosis and HCB irrespective of T2DM in a 1-year open-label RCT [132]. Vildagliptin (100 mg/d) [133], saxagliptin (5 mg/d) [134], omarigliptin (25 mg/week) [135], and teneligliptin (20 mg/d) [136] improved liver function and some non-invasive markers of NAFLD, and alogliptin (25 mg/d) was only moderately effective against NASH over 12 months of treatment in T2DM/NAFLD patients [137]. A recent meta-analysis by dos Santos et al. found the existing evidence for DPP4 inhibitors in NAFLD to be of poor quality and altogether not supportive of their clinical effectiveness [138]. Evogliptin [139], anagliptin [140][141], trelagliptin [142], gemigliptin [143], and linagliptin [144] have demonstrated beneficial effects in experimental rodent models, but their clinical value remains to be explored in future trials.
Sodium/glucose cotransporter (SGLT) 2 inhibitors are a relatively novel class of oral antidiabetic agents that increase urinary glucose excretion by inhibiting glucose reabsorption by SGLT2 in the proximal tubules. Several trials have demonstrated the improvement of cardiovascular and renal outcomes by treatment with compounds of this class, namely, empagliflozin, canagliflozin, and dapagliflozin [145]. SGLT2 inhibitors are known for their multiple metabolic effects that are notably relevant to NAFLD pathophysiology, including the general shift towards increased ketogenesis, gluconeogenesis, glycogenolysis, and FA β-oxidation. They inhibit leptin production by adipocytes, leading to decreased food intake, increase adiponectin levels, provide mild insulin sensitization, suppress HSC activation and fibrogenesis. Additionally, SGLT2 inhibitors may indirectly suppress sympathetic innervation and increase the vagal tone, thereby preventing the activation of Kupffer cells and the associated inflammatory processes [146][147].
Recent evidence mostly supports the efficacy of the majority of SGLT2 inhibitors for improving liver dysfunction, steatosis and fibrosis in NAFLD concomitant with T2DM. Among this group, only dapagliflozin, empagliflozin, and canagliflozin treatment was associated with beneficial effects in nondiabetic NAFLD patients. Dapagliflozin (10 mg) significantly reduced ALT, AST, and GGT levels, according to a retrospective study [148], while empagliflozin (10 mg/d) also attenuated liver steatosis and liver stiffness, indicative of potential antifibrotic activity, in a small-scale RCT [149]. The phase 3 DEAN trial to evaluate dapagliflozin in biopsy-confirmed NASH patients has been initiated (NCT03723252). Canagliflozin (100 mg/d) improved liver enzyme levels and FIB-4 index values in an open-label, uncontrolled pilot study [150]. Additionally, empagliflozin had a beneficial effect on cognitive functions and reduced anxiety in an experimental NAFLD model [151].
Ipragliflozin (50 mg/d for 72 weeks) ameliorated liver fibrosis and enhanced NASH resolution [152], remogliflozin etabonate (50–1000 mg/d for 12 weeks) reduced FIB-4 and NAFLD-fibrosis scores [153], and ertugliflozin (5 or 15 mg/d for 52 weeks) reduced liver transaminase levels [156] in TD2M patients with different stages of NASH. Several pilot studies in T2DM/NAFLD subjects confirmed the antisteatotic properties of luseogliflozin and tofogliflozin [155][156][157].
The SGLT1 subtype plays a relatively smaller (10–20% and up to 40% when SGLT2 are blocked) role in the renal glucose reabsorption, but is more abundant in the small intestine along with the heart and lungs. Intestinal SGLT1 (iSGLT1) inhibition leads to substantially reduced glucose and galactose absorption from the intestinal lumen, and increased incretin (GLP1, peptide YY) release by enteroendocrine cells [158]. Currently, licogliflozin (LIK066) is the only SGLT1/2 inhibitor being evaluated in NASH independent of T2DM presence, alone and in combination with the FXR agonist tropifexor, in the ongoing phase 2 ELIVATE study (NCT04065841). Previously, licogliflozin (150 mg/d) reduced ALT, ALT, GGT levels and liver fat content over 12 weeks compared to placebo [159]. A novel compound, SGL5213, has been identified as a selective iSGLT1 inhibitor, and has demonstrated insulin-sensitizing, anti-inflammatory, and antifibrotic activity in a murine model of NAFLD [160].
α-Glucosidase is a carbohydrate hydrolase located in the brush border of the small intestine that catalyzes the breakdown of dietary starch and disaccharides to yield glucose. α-Glucosidase inhibitors slow down carbohydrate digestion and absorption, thereby reducing postprandial hyperglycaemia. However, they are characterized by only modest overall antidiabetic activity, and are not too often used in clinical practice [161]. Despite some scientific interest concerning the use of this class of drugs for the treatment of liver diseases, data regarding their efficacy for NAFLD remain scarce. Acarbose (100 mg/d) improved AST, ALT levels and lipid profiles, albeit to a lesser extent than ezetimibe, in a 10-week small-scale RCT in non-diabetic NASH patients [162]. Miglitol treatment (150 mg/d for 12 months) was associated with significant improvements in steatosis, lobular and portal inflammation, and NAS scores, while fibrosis and hepatocyte ballooning remained unchanged [163]. Finally, voglibose prevented hepatic steatosis in obese rats, but was slightly inferior to empagliflozin [164].
Lately, cell-based therapy has emerged as a feasible alternative for the treatment of different stages of NAFLD. In particular, experimental evidence supports the use of bone marrow- [176], umbilical cord- [177], and compact bone-derived mesenchymal stromal cells [178] as well as hepatocytes derived by differentiating induced pluripotent stem cells [179]. The crosstalk between hepatic stem cells and their possible therapeutic application for NAFLD are discussed in detail in a recent review by Overi et al. [180].
HepaStem® is a first-in-class allogeneic stem cell therapy product containing human adult liver-derived progenitor cells with potential indications including cirrhotic and precirrhotic NASH as well as acute-on-chronic liver failure (ACLF). HepaStem® cells, obtained from healthy donors, are expected to modulate the inflammatory response and inhibit HSC activation, thereby reducing liver fibrosis. A small-scale phase 2a RCT found HepaStem® to be safe and well tolerated, and indicated potential efficacy for ACLF and/or decompensated liver cirrhosis [181].
Fraudulent, or abnormal fatty acids, represented by bempedoic acid (ETC-1002) and gemcabene (CI-1027), are molecules with structures similar to those of oleic or linolenic acid that regulate metabolic pathways in the liver, resulting in enhanced FA catabolism. After conversion into its active form, bempedoic acid acts as false substrate and inhibits hepatic adenosine triphosphate citrate (pro-S)-lyase, an enzyme upstream of 3-hydroxy-3-methylglutaryl-CoA reductase in the cholesterol synthesis pathway. This links fraudulent fatty acids to statins, whose possible beneficial effects for NAFLD are reviewed elsewhere [182].
Bempedoic acid is approved in the USA and EU as monotherapy and as a fixed dose combination with ezetimibe for the treatment of hypercholesterolaemia [183]. In a high-fat diet-induced murine model of NASH, it caused significant reductions in ALT and AST levels, hepatic TAG accumulation, proinflammatory and profibrotic gene expression, resulting in improved NAFLD activity and liver fibrosis by histological analysis [184].
Gemcabene (PD-72953), a structurally optimized derivative of bempedoic acid, forms a CoA conjugate that inhibits ACC, and reduces apolipoprotein C-III expression [183]. In a mouse model of NASH/HCC, it diminished micro- and macrovesicular liver steatosis, HCB, inflammatory infiltration, and fibrosis, which corresponded to downregulated proinflammatory, lipogenesis, and profibrogenic marker expression [183]. Gemcabene was being developed for the treatment of paediatric NAFLD, but was discontinued and repurposed for another indication after a lack of efficacy was demonstrated in a phase 2a proof-of-concept study (NCT03436420).
Tesamorelin (TH9507) is a growth hormone (GH) releasing hormone analogue that is thought to stimulate lipolysis via increasing endogenous GH levels while maintaining feedback inhibition and limiting toxicity compared to native GH. Tesamorelin reduced liver fat content and visceral fat in a preliminary study in antiretroviral-treated patients with human immunodeficiency virus (HIV)-associated lipodystrophy [185]. A phase 2 trial to evaluate the effects of tesamorelin on liver steatosis and cardiovascular risk in obese NASH
patients is recruiting (NCT03375788), and a phase 3 study in the general population with NAFLD including a HIV cohort has been planned [186].
Berberine ursodeoxycholate (BUDCA, HTD1801) is an ionic salt of the isoquinoline alkaloid berberine and ursodeoxycholic acid (UDCA). According to a meta-analysis by Wei et al., berberine can significantly improve liver function, lipid profiles, and glycaemic control in patients with NAFLD [187] due to adenosine monophosphate-activated protein kinase (AMPK) activation, stimulation of glycolysis, and, possibly, inhibition of α-glucosidase [188]. UDCA, in turn, is a bile acid long used for the treatment of NASH and chronic cholestatic diseases, whose hepatoprotective effects are confirmed by several systematic reviews and meta-analyses [189][190]. In a phase 2 proof-of-concept RCT in T2DM patients with presumed NASH, BUDCA reduced liver enzyme levels and liver steatosis by MRI-PDFF [191].
Miricorilant (CORT 118335) is an investigational glucocorticoid receptor agonist/antagonist and a mineralocorticoid receptor antagonist currently in development for NASH and antipsychotic-induced weight gain. Results of a phase 2a study in NASH patients demonstrated that miricorilant (600 mg/d) effectively ameliorated liver steatosis by radiological measures. However, miricorilant treatment was associated by transient yet significant increases in serum transaminases [192], and the trial was subsequently put on hold due to safety concerns (NCT03823703).
Nitazoxanide (Alinia®) is an FDA-approved broad-spectrum thiazolide antiprotozoal and antiparasitic agent, lately reported to be a potent AMPK activator and inhibitor of HSC activation. In experimental studies in mice, nitazoxanide (100 mg/kg/d) attenuated dyslipidaemia, liver steatosis [193], fibrosis, inflammation, and HCB, demonstrating synergistic effects with the pan-PPAR agonist elafibranor [194]. Moreover, the anti-anaerobic activity of nitazoxanide may determine its use in preventing the recurrence of hepatic encephalopathy as a viable alternative to rifaximin (NCT04161053) [195].
Pirfenidone (Esbriet®) is a pyridine derivative with antifibrotic, anti-inflammatory, and antioxidant properties, the precise mechanisms of which are still unclear. In the liver, pirfenidone may decrease fibronectin, TGFβ, collagen production and attenuate fibrogenesis, hepatocyte necrosis, and necroinflammation [196][197]. In the phase 2 PROMETEO study, pirfenidone (1200 mg/d) markedly reduced transaminase levels and advanced liver fibrosis of predominantly nonalcoholic aetiology [198].
The modern pipeline for NAFLD includes a plethora of drug candidates with diverse and innovative mechanisms of action. Other investigational drugs with potential therapeutic value in NAFLD include antileukotriene agents [199], GPCR modulators [200], anti-IL mABs [201], IL22 axis modulators [202], purinergic receptor agonists [203], antioxidants [204], antisense oligonucleotides [205], multitarget epigenetic regulators [206], and many more. An overview of the current drug development pipeline for NAFLD is given in Figure 1. Given the exceptionally complex pathophysiology and the multifaceted nature of this disease, NAFLD pharmacotherapy can be expected to remain a priority for biomedical research in the nearest future.
Figure 1. An overview of the current drug development pipeline for non-alcoholic fatty liver disease.
FASN, fatty acid synthase; DGAT, diglyceride acyltransferase; FXR, farnesoid X receptor; FGF, fibroblast growth factor; TLR4, toll-like receptor 4; LOXL2, lysyl oxidase-like protein 2; ATX, autotaxin; SGLT, sodium/glucose contransporter; MPC, mitochondrial pyruvate carrier; PPAR, peroxisome proliferator-activated receptor; THRβ, thyroid hormone receptor β; PUFA, polyunsaturated fatty acid; ACC, acetyl-CoA carboxylase; BUDCA, berberine ursodeoxycholate; LMS, leucine-metforminsildenafil; LMSC, liver-derived mesenchymal stromal cells.