Introduction
Nonalcoholic fatty liver disease (NAFLD) describes a collection of steatotic liver conditions. Nonalcoholic steatohepatitis (NASH), first coined by Ludwig et al.1 in 1980 is the inflammatory subtype of NAFLD involving liver steatosis, inflammation, and hepatocyte ballooning often along with fibrosis.2 Even though NASH has become a global epidemic,3 it receives much less attention from the public, healthcare professionals, and policymakers than other metabolic diseases such as diabetes and cancer. Recent consensus and recommendations from a Delphi study may raise public awareness of the disease and provide a solid foundation for a comprehensive public health response to NAFLD.4 Nevertheless, there is no approved drug for NASH treatment in the clinic. Considerable efforts have been vested in NASH drug development; the arduous road continues.
Heterogeneous nature of NASH
Approximately 25% of the world population develops NAFLD, and 20% progress to NASH (Fig. 1). NASH is a heterogeneous disease, and its phenotypic manifestation likely reflects the interactions of different primary drivers and coexisting disease modifiers. Environmental factors such as diet and microbiota, genetic determinants such as PNPLA3 and TM6SF2 mutations as well as epigenetic modifications such as methylation and acetylation can all contribute to NASH. Comorbidities, e.g., type 2 diabetes, dyslipidemia, and gout may exacerbate the disease and thus should be evaluated in NASH patients. Other types of liver conditions such as autoimmune hepatitis, HBV/HCV infection, and Wilson’s disease differ from NASH and should be excluded from NASH clinical trials. Unfortunately, present trial recruitment is mainly based on histologic grading and staging, and the response rates to investigational agents are often low.5
Clinical trials
Currently, 1282 NASH-related trials are registered at the ClinicalTrials.gov website (https://clinicaltrials.gov ), 537 are drug intervention trials including 289 phase II trials, 75 phase III trials, and 67 phase IV trials. Of all registered trials, 68 (5.6%) have published results in peer-reviewed journals (Supplementary Table 1). This percentage is significantly lower than those of type 2 diabetes (18%) and obesity trials (89%). The vast majority (94%) of NASH trials are in the adult population (Supplementary Table 1), and most involve Caucasians (66%) followed by Hispanics (43%), Asians (22%), and Africans (21%).
Trials using surrogate biomarkers or histological outcome endpoints can lead to regulatory agency conditional or full approval of drugs. Some investigators argue whether a biopsy-based evaluation is an appropriate efficacy endpoint because imprecise histological staging may disproportionally impact the active arm relative to the placebo. Many clinicians and investigators do consider liver biopsy and histological evaluation as the reference standard for NASH. Thirty-three of 68 trials (49%) have histological outcome measures (Supplementary Table 1). Thirty-nine trials evaluate drug effects on noncirrhotic NASH (F1-F3), four trials are on cirrhotic NASH (F4) and four trials include both (F1–F4). Of three known histological assessment systems, the National Institute of Diabetes and Digestive and Kidney Diseases NASH Clinical Research Network (commonly referred to as CRN) system, the Steatosis-Activity-Fibrosis system, and the Goodman classification, the CRN system have been unanimously adopted in all 33 publications. Frequently used noninvasive measures (Supplementary Table 1) include biometric markers (body weight, body mass index, and hepatic fat), liver function markers (alanine aminotransferase, aspartate aminotransferase, and gamma-glutamyl transferase), lipid markers (plasma cholesterol, triacylglycerol, high-density lipid cholesterol, and low-density lipid cholesterol), glycemic markers (fasting plasma glucose, insulin, homeostatic model assessment for insulin resistance and glycated hemoglobin). However, common sets of noninvasive clinical measures for steatosis, inflammation, and fibrosis remain to be established, and biomarkers for hepatocyte ballooning are absent.
Therapeutic drugs and targets
Approximately 216 drugs are being or have been evaluated for the treatment of NASH (Fig. 2 and Table 1). Most drugs target metabolic pathways; some aim at inflammation or fibrosis. The vast majority (>80%) are small-molecule drugs. Biological drugs include fibroblast growth factor 21 (FGF21), FGF19, glucagon-like peptide 1 (GLP-1), glucagon, glucose-dependent insulinotropic polypeptide (GIP or gastric inhibitory polypeptide), or a combination thereof (Fig. 2).
Table 1Additional targets and drugs
Drug | Target | Company |
---|
Cysteamine | / | Raptor Pharma |
FIA586 | / | Novartis |
LY3849891 | / | Eli Lilly |
MG-1 | / | / |
NC101 (undefined) | / | / |
NRL972 (cholyl lysyl fluorescein) | / | Norgine |
TRO19622 (Olesoxime) | / | Hoffmann-La Roche |
XZP-5610 | / | Xuanzhu Biotech |
XZP-6019 | / | Xuanzhu Biotech |
ZSP0678 | / | Raynovent |
TEV-45478 | / | Teva Pharma |
FM101 | A3AR | Future Medicine |
Namodenoson | A3AR | Can-Fite BioPharma |
PBF-1650 | A3AR | Palobiopharma |
Aceon (Perindopril) | ACE | / |
CF102 | ADORA3 | Can-Fite BioPharma |
WY-8678 (Guanabenz acetate) | ADRA2A | Pfizer |
Cozaar (Losartan) | AGTR | Organon |
Nitazoxanide | Antiparasitic | Genfit |
LPCN 1144 (testosterone) | AR | Lipocine |
SHP626 (Volixibat) | ASBT | Shire |
BLD-0409 (Cudetaxestat) | ATX | Blade Therapeutics |
PLN-1474 | avb1 | Pilant |
IDL2965 | avb3 | Indalo |
Dasatinib (Sprycel) | BCR-ABL | BMS |
CP-945598 | CB1R | STEMCELL Technologies |
JNJ-2463 (Nimacimab) | CB1R | Bird Rock Bio |
Proglumide | CCKA/BR | AdvaCare Pharma |
CM-101 | CCL24 | Chemomab |
RYI-018 | CNR1 | Bird Rock Bio |
SNP-630 | CYP450 | Sinew Pharma |
Singulair (Montelukast) | CYSLTR | Merck |
MN-001 (Tipelukast) | CYSLTR/PDE/5-LOX | MediciNova |
DUR-928 | Epigenetic | Durect |
Estradiol | ER | / |
NGM282 (Aldafermin) | FGFR4-KLB | NGM |
GB1211 | Galectin-3 | Galecto Biotech |
GR-MD-02 (Belapectin) | Galectin-3 | Galectin Therapeutics |
NGM395 | GFRAL | NGM |
Tesamorelin (GHRH) | GH | Theratechnologies |
LUM-201 (Ibutamoren) | GHSR | Lumos Pharma |
Crestor (Rosuvastatin) | HMGCR | Astra Zeneca |
BMS-986263 | HSP47 | BMS |
GM-60106 | HTR2A | JD Bioscience |
Elobixibat | IBAT | Albireo |
HPN-01 | IKK | Hepanova |
CC-90001 | JNK | Celgene |
PF-06835919 | KHK | Pfizer |
Metreleptin | LEPR | Amylin Pharma |
IMM-124E | LPS | Immuron |
Oltipraz (Dithiolethiones) | LXRa | PharmaKing |
BMS-963272 | MGAT2 | BMS |
Metformin | mGPD | / |
LB-P6 | Microbiome | LISCure Biosciences |
RG-125 (AZD4076) | miR-103/107 | Regulus |
MSDC-0602K | MPC | Cirius Therapeutics |
IdB 1016 (Siliphos) | NF-κB | Indena |
GRI0621 | NK cell | GRI Bio |
DFV890 | NLPR3 | Novartis |
SGM-1019 | P2X7R | Second Genome |
CER209 | P2Y13R | Abionyx |
ZSP1601 | PDE | Raynovent |
ASP9831 | PDE4 | Astellas Pharma |
CRV-431 (Rencofilstat) | PPI | Hepion Pharma |
Rifampicin | PXR | / |
Denosumab | RANKL | Amgen |
TB-840 | RORa | Therasid Bioscience |
Hydroxytyrosol | ROS | / |
IDDF2019-ABS-0026 (Metadoxine) | ROS | Micro Labs |
Vitamin E | ROS | / |
AGN-242266 | RXR | Abbvie |
Aramchol | SCD1 | Galmed |
BI 685509 | sGC | Boehringer Ingelheim |
Idebex (Idebenone) | Shc | ABCO Lab |
NS-0200 | SIRT1/AMPK | NuSirt Biopharma |
Amlexanox | TBK1/IKKε | / |
ZED1227 | TG2 | Zedira |
Pradaxa (Dabigatran) | Thrombin | Boehringer Ingelheim |
JKB-122 | TLR4 | TaiwanJ Pharma |
Trental (Pentoxifylline) | TNFα | / |
Vitamin D | VDR | / |
Uloric (Febuxostat) | XDH | Takeda |
Major therapeutic targets include lipid metabolic enzymes acetyl-CoA carboxylase (ACC)6 and fatty acid synthase (FASN),7 which affect liver steatosis; farnesoid X receptor (FXR)8 which interferes with bile acid signaling; nuclear receptors peroxisome proliferator-activated receptor (PPAR),9 and thyroid hormone receptor beta (THRb)10 as well as antidiabetic-peptide receptors GLP-1 receptor (GLP1R),11 glucagon receptor (GCGR)12 and GIP receptor (GIPR),13 which alter glucose and lipid metabolism; FGF21 and FGF19 receptors, which have pleiotropic effects on liver steatosis, inflammation and fibrosis;14 chemokine (C-C motif) ligand 2 and 5 (CCL2/5)15 and amine oxidase copper containing 3 (AOC3),16 which influence inflammation; lysyl oxidase-like 2 (LOXL2),17 which involves in fibrosis; and caspases, which are key players in apoptosis (Fig. 2). Despite hundreds of potential drugs and a plethora of targets, there is no clear winner.
Seeking alpha
Although no consensus on noninvasive measures for NASH has been reached, histological evaluations of steatosis, inflammation, hepatocyte morphology, and fibrosis are well accepted. We have appraised treatment efficacy in the 33 trials with histological outcome measures (Supplementary Table 1, trials labeled in red) from seven perspectives, i.e. effects on liver steatosis, hepatocyte ballooning, inflammation, fibrosis (F1-F3), cirrhosis (F4), treatment duration, and safety (treatment-emergent adverse events). The first five parameters are NASH-specific and the last two are important clinical considerations for all medications.
FGF21 (NCT03976401), FGF19 (NCT03912532), obeticholic acid (NCT01265498, T02548351), pioglitazone (NCT00994682, NCT00063622, NCT00062764, CT00227110), semaglutide (NCT02970942), and resmetirom (NCT02912260) had significant effects on NASH patients (Table 2). It seems that FGF21 was the most efficacious, followed by resmetirom, pioglitazone, FGF19, semaglutide, and obeticholic acid (Table 2 and Fig. 3). FGF21 had a greater antifibrosis activity and shorter treatment duration than other drugs.
Table 2Outcomes of major clinical trials
Trial | Phase | Drug | Steatosis (Score) | Ballooning (Score) | Inflammation (Score) | Fibrosis F1-F3 (Score) | Cirrhosis (Score) | Tx Duration (Score) | Safety (Score) | Total score |
---|
NCT03976401 | II | FGF21 | 17/17 (1) | 13/14 (0.93) | 13/14 (0.93) | 10/13 (0.77) | 7/12 (0.58) | 12 (1) | No increase (1) | 6.21 |
NCT03912532 | II | FGF19 | 34/43 (0.79) | 21/43 (0.49) | 0/43 (0) | 7/43 (0.16) | ND (0) | 24 (0.8) | Grade 1 (0.8) | 3.04 |
NCT01265498 | III | Obeticholic acid | 62/102 (0.61) | 47/102 (0.46) | 54/102 (0.53) | 36/102 (0.35) | ND (0) | 72 (0.4) | Grade 3 (0.4) | 2.75 |
NCT02548351 | III | | 127/308 (0.41) | 108/308 (0.35) | 136/308 (0.44) | 84/308 (0.27) | ND (0) | 72 (0.4) | Grade 3 (0.4) | 2.27 |
| | | | | | | | | | 2.51 |
NCT00994682 | IV | Pioglitazone | 35/50 (0.70) | 25/50 (0.50) | 25/50 (0.50) | 20/50 (0.40) | NC (0) | 72 (0.4) | No increase (1) | 3.10 |
NCT00063622 | III | | 55/80 (0.69) | 35/80 (0.44) | 48/60 (0.6) | 35/80 (0.44) | ND (0) | 96 (0.2) | Weight gain (0.8) | 3.17 |
NCT00062764 | II | | 6/18 (0.33) | 6/18 (0.33) | 6/18 (0.33) | 6/18 (0.33) | ND (0) | 48 (0.6) | Weight gain (0.8) | 2.72 |
NCT00227110 | IV | | 17/26 (0.65) | 14/26 (0.54) | 17/26 (0.65) | 12/26 (0.46) | ND (0) | 24 (0.8) | No increase (1) | 4.10 |
| | | | | | | | | | 3.27 |
NCT02970942 | II | Semaglutide | 33/56 (0.59) | 33/56 (0.59) | 33/56 (0.59) | 0/56 (0) | ND (0) | 72 (0.4) | Grade 2 (0.6) | 2.77 |
NCT02912260 | II | Resmetirom | 41/73 (0.60) | 41/73 (0.60) | 41/73 (0.60) | 21/73 (0.29) | ND (0) | 12 (1) | Grade 2 (0.6) | 3.69 |
Obeticholic acid improved steatosis, ballooning, inflammation, and fibrosis but had significant side effects, i.e. pruritus (aka itching).18,19 The FXR agonist cilofexor20,21 and EDP-30522 treatment were also associated with pruritus. It is unclear whether the itching side effect is obeticholic acid-specific or associated with bile acid metabolism in general. Compared with other PPAR agonists lanifibranor, rosiglitazone,and elafibranor, pioglitazone had a greater effect on steatosis, hepatocyte ballooning, inflammation, and fibrosis.23–26 Glitazones for type 2 diabetes therapy have been linked to serious side effects such as fluid retention, congestive heart failure, weight gain, bone loss, and increased risk of bladder cancer.27–29 Interestingly, pioglitazone had a safe profile in all four trials (NCT00994682, NCT00063622, NCT00062764, CT00227110). Side effects of longer treatment duration remain unclear. GLP-1 analog semaglutide30 and exenatide31 reduced liver steatosis, ballooning, and inflammation, but not fibrosis. It is noteworthy that obeticholic acid and pioglitazone data are from phase III trials and other drug results are from phase II trials. Resmetirom was effective in a phase II trial32 and positive phase III results have recently been announced by Madrigal (https://www.madrigalpharma.com/ ).
FGF21 phase II trial results are encouraging. AKERO-001 (an FGF21 analog) had excellent effectiveness on liver steatosis, ballooning, inflammation, and fibrosis with a short duration and mild adverse drug reactions.33 It was also effective in patients with cirrhosis (https://ir.akerotx.com/news-releases/news-release-details/akero-announces-positive-histological-improvements-cirrhotic ). Among patients with PNPLA3I148M mutation, 8/18 (44% I/M carriers) and 5/9 (56% M/M carriers) achieved a ≥ 4-point improvement in NAS and 10/18 (56% I/M carriers) and 7/9 (78% M/M carriers) had a ≥ 1 stage improvement in fibrosis (kyale@akerotx.com). FGF19, a close relative, was less effective than FGF21, possibly owing to a significant elevation of total plasma cholesterol and low-density lipid cholesterol.34 Therefore, the FGF21 analog appears to be a strong therapeutic candidate for NASH.
There may still be room for improvement of FGF21 analogs. Pegbelfermin and BIO89-100 are FGF21 peptides stabilized by polyethylene glycol modification, whereas AKERO-001 is stabilized by FGF21-IgG1 Fc fusion. FGF21-Fc fusion protein seems to be more stable and thus more effective than PEG-modified ones. However, AKERO-001 is expressed in prokaryotic cells, and our experiment suggests that eukaryotic-expressed FGF21-Fc is significantly more stable and consequently more efficacious than the prokaryotic-expressed proteins. In addition, multitarget fusion peptides are emerging; for example, GLP1-FGF21 (DDKJ Biomedicals, HEC Pharm), GLP1-GCG-FGF21 (Doer Biologics), and GLP1-GIP-FGF21 (DDKJ Biomedicals). Preclinical evidence indicated that dual targeting molecule GLP1-FGF21 was more effective than FGF21 alone.35 GLP1-GIP combination has been reported to reduce GLP-1-caused gastrointestinal problems,36 therefore, the combination of GLP1-GIP-FGF21 may have a significant effect on both glucose and lipid metabolism while minimizing gastrointestinal discomfort. Are we seeing the light at the end of the tunnel? One should cautiously wait for the outcome of the FGF21 Phase III clinical trial.
Supporting information
Supplementary Table 1
Clinical outcome measures.
(XLSX)
Abbreviations
- FGF21:
fibroblast growth factor 21
- NASH:
nonalcoholic steatohepatitis
Declarations
Funding
None to declare.
Conflict of interest
YQC holds shares of DDKJ Biomedicals.