Safety and efficacy of efruxifermin in metabolic dysfunction-associated steatohepatitis: A systematic review

Abstract

BACKGROUND

Efruxifermin (EFX), a fibroblast growth factor 21 analogue, has demonstrated the potential to improve liver fat and markers of liver injury, fibrosis, and key metabolic biomarkers in individuals with metabolic dysfunction-associated steatohepatitis (MASH) in phase 2 clinical trials.

AIM

To summarize the safety and effectiveness of EFX in managing MASH.

METHODS

Electronic databases and registries were systematically searched from their inception to May 15, 2025, for randomized-controlled trials (RCTs) that included EFX in the intervention arm and placebo in the control arm in individuals with MASH. The primary outcome was the safety of EFX, while additional outcomes included its efficacy in altering hepatic and metabolic parameters. Meta-analyses were conducted using the RevMan web computer program with the random-effects model.

RESULTS

Four phase 2 RCTs (five reports), mostly with low risk of bias, involving 450 subjects, were analyzed. Compared to the placebo, EFX 50 mg was associated with higher risks of treatment-emergent adverse events (TEAEs) [risk ratio (RR) = 1.05], TEAEs leading to discontinuation (RR = 3.05), nausea (RR = 1.78), and diarrhea (RR = 1.9). EFX 28 mg increased risks of vomiting (RR = 2.17) and frequent bowel movements (RR = 8.98). Both doses of EFX were associated with higher risks of drug-related TEAEs (28 mg: RR = 1.45; 50 mg: RR = 1.67) and increased appetite (28 mg: RR = 3.16; 50 mg: RR = 5.66). EFX (28 and 50 mg) and placebo exhibited identical risks for severe TEAEs, serious AEs, abdominal pain, fatigue, headache, injection site erythema, and injection site reactions. EFX (28 and 50 mg) was associated with improvements in hepatic safety outcomes, including liver enzymes and urate levels. EFX outperformed the placebo in both relative and absolute reductions in hepatic fat fraction. Reductions in enhanced liver fibrosis score, Pro-C3, and liver stiffness were also more robust with EFX. EFX was superior in terms of MASH resolution and improvement in fibrosis stage, MASH resolution and no worsening of the fibrosis stage, and fibrosis regression by ≥ 1 stage and no worsening in steatohepatitis. Furthermore, EFX also improved metabolic parameters, including reductions in HbA1c and insulin resistance, as well as improvements in adiponectin and lipid parameters.

CONCLUSION

EFX demonstrates promising dual efficacy on liver histology and metabolic markers in MASH. However, gastrointestinal side effects and the need for parenteral administration require caution. Long-term data are still necessary to fully evaluate safety and long-term effectiveness.

Key Words: Efruxifermin; Fibroblast growth factor 21; Metabolic dysfunction-associated steatohepatitis; Non-alcoholic steatohepatitis; Safety; Hepatic fat fraction

Core Tip: This one is the first systematic reviews and meta-analyses specifically targeting efruxifermin (EFX)’s safety as well as liver histological and metabolic impacts in metabolic dysfunction-associated steatohepatitis (MASH). We analyzed four randomized-controlled trials (five reports) involving 450 subjects and found that EFX, imparts greater risks for drug-related treatment-emergent adverse events (TEAEs) and gastrointestinal AEs but does not increase the occurrence of severe TEAEs or serious AEs. EFX surpassed the placebo in lowering liver enzymes, urate levels, hepatic fat fraction, liver fibrosis score, Pro-C3, and liver stiffness, while also improving the fibrosis stage. Furthermore, EFX demonstrated metabolic benefits, including reductions in HbA1c and insulin resistance, along with improvements in adiponectin and lipid parameters. The outcomes of the current phase 3 trials are eagerly anticipated to verify its safety and effectiveness in MASH.

INTRODUCTION

The global rise in obesity and its metabolic comorbidities has led to an estimated 38.2% prevalence of nonalcoholic fatty liver disease (NAFLD), now referred to as metabolic dysfunction-associated fatty liver disease (MAFLD)[1]. MAFLD poses significant clinical and economic challenges due to its association with liver complications and cardiovascular risks. A more advanced form, metabolic dysfunction-associated steatohepatitis (MASH) (formerly called nonalcoholic steatohepatitis, NASH), is characterized by hepatic steatosis, lobular inflammation, and hepatocellular ballooning, which predispose to fibrosis, cirrhosis, end-stage liver disease, and hepatocellular carcinoma[2]. In the past decade, medications such as pioglitazone, sodium-glucose cotransporter-2 (SGLT-2) inhibitors, and glucagon-like peptide-1 (GLP-1) receptor agonists, which are effective in treating cardiometabolic diseases, have been investigated for MASH but have seen limited success. The March 2024 approval of resmetirom, a liver-selective thyroid hormone receptor beta agonist, represents a significant advancement in treating MASH with moderate to advanced liver fibrosis[3]. This approval is positive for tackling MASH and related cardiometabolic risks that may cause liver failure and cardiovascular diseas[4].

Despite ongoing clinical studies for approximately 40 compounds targeting MASH, an urgent need for effective treatments remains[5]. Fibroblast growth factor 21 (FGF21), a hepatokine, plays a vital role in regulating metabolism, enhancing mitochondrial function, and improving insulin sensitivity[6]. FGF21 analogs, including efruxifermin (EFX), pegbelfermin, and pegozafermin, are currently under investigation for their potential benefits in the treatment of MASH[5-7]. EFX, developed by Akero Therapeutics, is a long-acting FGF21 fusion protein that targets both the liver and adipose tissue, unlike the PEGylated FGF21 analogs pegbelfermin and pegozafermin, which primarily target adipose tissue. EFX has a longer half-life than both pegbelfermin and pegozafermin[7]. EFX has shown promise in improving liver fat and markers of liver injury, fibrosis, and key metabolic biomarkers in Phase 2 trials. The ongoing Phase 3 SYNCHRONY program is assessing EFX's efficacy in treating pre-cirrhotic MASH and compensated cirrhosis[8,9]. While some systematic reviews and meta-analyses (SR/MAs) have examined EFX, they have not provided a comprehensive focus on its specific effects or included all relevant studies[10-12]. This SR/MA aims to evaluate the safety and effectiveness of EFX, focusing on adverse events, liver histological improvements, and metabolic outcomes in individuals with MASH. The goal is to enhance understanding of EFX’s clinical implications and its potential role in modifying the progression of MASH.

MATERIALS AND METHODS

This SR/MA was conducted according to the procedures outlined in the Cochrane Handbook for Systematic Reviews of Interventions[13] and is reported according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses checklist[14]. It has been registered with PROSPERO (CRD420251054498), and the protocol summary can be accessed online.

Search strategy

Several databases and registers were systematically searched, including MEDLINE (via PubMed), Scopus, EMBASE, the Cochrane Central Register, and ClinicalTrials.gov; the search engine Google Scholar was also utilized to find additional articles. The search spanned these sources from their inception to May 15, 2025. Using the Boolean operators “AND” and “OR”, the following terms were searched: “Efruxifermin”, “AKR-001”, “AMG-876”, “EFX”, “metabolic dysfunction-associated steatotic liver disease”, “metabolic dysfunction-associated fatty liver disease”, “nonalcoholic fatty liver disease”, “fatty liver”, “metabolic dysfunction-associated steatohepatitis”, “nonalcoholic steatohepatitis”, “NAFLD”, “MAFLD”, “MASLD”, “NASH”, “MASH”, and “cirrhosis”. The search terms targeted titles, abstracts, and keywords to identify published studies in the English language. The details of the search strategies are provided as a supplementary material (Appendix 1 in Supplementary material). Furthermore, the process involved examining the references cited in the articles obtained for this research, as well as relevant journals.

Study selection

The Population, Intervention, Comparison, Outcomes, and Study design served as the framework for establishing eligibility criteria for the clinical trials in this SR/MA. The patient population (P) consisted of adults with MASH. The intervention (I) was once-weekly subcutaneous injections of EFX. The comparison or control (C) included adults receiving either a placebo or no interventions. The outcomes (O) included the safety and effectiveness of EFX compared to the control arm. Only randomized controlled trials (RCTs) with a duration of at least 12 weeks were considered as the study type (S) for inclusion. The trials required to have at least two treatment arms to be qualified for inclusion: One arm (EFX arm) receiving once-weekly subcutaneous injections of EFX at any dose, with or without lifestyle interventions or other drugs for treating MASH, and the other arm (control arm) receiving either a placebo or no interventions, with or without lifestyle interventions or other medications for treating MASH. Clinical trials involving animals or healthy humans, non-randomized trials, retrospective studies, pooled analyses of clinical trials, conference proceedings, preprints, letters to the editor, case reports, and articles that did not provide data on outcomes of interest were excluded from consideration.

Outcomes analyzed

The primary outcome of interest was the general and hepatic safety outcomes in the EFX vs placebo arms. The general safety outcomes included any treatment-emergent adverse events (TEAEs), drug-related TEAEs, severe TEAEs, TEAEs leading to treatment discontinuation, serious AEs, gastrointestinal (GI) AEs, fatigue, headache, injection site erythema, and injection site reactions. Hepatic safety outcomes included changes in liver enzymes and urate levels. The additional outcomes included hepatic and metabolic efficacy outcomes. Hepatic efficacy outcomes were changes from baseline in hepatic fat fraction (HFF) measured by Magnetic Resonance Imaging Proton Density Fat Fraction (MRI-PDFF), enhanced liver fibrosis (ELF) score, liver stiffness (E) during FibroScan^® measurement, N-terminal propeptide of type III collagen (Pro-C3), and proportions of study subjects achieving reductions in HFF by ≥ 30%, ≥ 50%, and normalization of liver fat (HFF ≤ 5%), proportions of study subjects achieving a decrease in NAFLD activity score (NAS) by ≥ 2 points, MASH resolution and improvement in fibrosis stage, MASH resolution and no worsening of fibrosis stage, fibrosis regression by ≥ 1 stage and no worsening in steatohepatitis, and fibrosis regression by ≥ 2 stage and no worsening in steatohepatitis. Metabolic outcomes included changes in body weight, glycated hemoglobin (HbA1c), homeostatic model assessment for insulin resistance (HOMA-IR), C-peptide, triglycerides (TG), low-density lipoprotein cholesterol (LDL-C), high-density lipoprotein cholesterol (HDL-C), non-HDL-C, adiponectin, apolipoprotein B (Apo-B), and apolipoprotein C3 (Apo-C3).

Data extraction and dealing with missing data

Four review authors independently performed data extraction, utilizing standardized forms. The results were consolidated upon identifying multiple publications from the same study group, and relevant data from each article were incorporated into the analyses. The following data were extracted for all eligible studies and included in the review: First author, year of publication, country where the study was conducted, study design, major inclusion criteria for study subjects, sample size, frequency of female participants and participants with type 2 diabetes (T2D), mean age, body mass index, concomitant medications, and trial duration. Additionally, as previously mentioned, data on primary and secondary outcomes were extracted. All disagreements were settled through consensus. The process of handling missing data has been described in detail in a previous publication[15].

Statistical analysis

The results of the outcomes were expressed as mean differences (MDs) for continuous variables and as risk ratios (RRs) for dichotomous variables with 95%CI. RRs were preferred for both safety and efficacy outcomes in prospective studies (e.g., RCTs) because they directly compare probabilities and are more intuitive. The Review Manager (RevMan) computer program, version 7.2.0, was used to generate forest plots, which illustrated the MD or RR for the observed outcomes[16]. The left side of the forest plot favored the EFX group, while the right side favored the placebo group. Random-effects analysis models were selected to address the expected heterogeneity resulting from variations in population characteristics and study duration. The inverse variance statistical method was utilized in all instances. CI calculated by the Wald-type method. Tau² was estimated by the DerSimonian and Laird method. The SR/MA included forest plots that combined data from at least two trials. A significance level of P < 0.05 was applied.

Risk of bias assessment

Three authors independently performed the risk of bias (RoB) assessment using version 2 of the Cochrane risk-of-bias tool for randomized trials (RoB2) in the RevMan web[16,17]. All disagreements were settled through consensus. The Risk-of-bias VISualization (robvis) web app was used to create RoB plots[18]. When applicable (i.e., a minimum of ten studies in a forest plot), publication bias was assessed using funnel plots generated by the RevMan web software[16,19].

Assessment of heterogeneity

The assessment of heterogeneity was initially performed by examining forest plots. Subsequently, χ² tests were conducted using N-1 degrees of freedom and a significance level of 0.05 to assess statistical significance. The I² test was also used in the subsequent analysis. Thresholds for I² values were 25% for low heterogeneity, 50% for moderate heterogeneity, and 75% for high heterogeneity[20].

RESULTS

Search results

Figure 1 illustrates the steps involved in selecting studies. The initial search identified 256 articles, which were narrowed to six after screening titles, abstracts, and subsequent full-text reviews. A detailed evaluation led to the inclusion of four RCTs with five published reports involving 450 subjects that met all the inclusion criteria[21-25]. Two studies were excluded: One was an animal study[9], and the other was a phase 1 pharmacodynamic study (Figure 1)[26].

Figure 1 PRISMA flowchart on study retrieval and inclusion in the meta-analysis.

Characteristics of included studies

All four RCTs in this SR/MA were multicenter phase 2 clinical trials; three were conducted in the United States, while the other took place in the USA, Puerto Rico, and Mexico. The BALANCED trial (NCT03976401) had two reports (Harrison 2021 and Harrison 2022) from two separate cohorts: One comprising individuals with biopsy-proven NASH and the other involving those with biopsy-confirmed compensated cirrhosis due to NASH[21,22]. The Harrison 2023 (HARMONY) and Harrison 2025 (Cohort D) trials involved individuals with biopsy-confirmed NASH[23] or MASH[24]. Adults with liver histologic features consistent with MASH and compensated cirrhosis made up the study subjects in the Noureddin 2025 (SYMMETRY) trial[25]. All four trials used a placebo as the control. In the Harrison 2025 (Cohort D) trial, all participants were maintained on a stable dose of GLP-1 receptor agonists at the time of randomization, ensuring a balanced effect of GLP-1 receptor agonists across both the EFX and placebo groups. In the intervention groups, Harrison 2021 (BALANCED) administered three doses of EFX (28, 50, and 70 mg)[21]. Harrison 2023 (HARMONY)[23] and Noureddin 2025 (SYMMETRY)[25] utilized two doses (28 and 50 mg), whereas Harrison 2022 (BALANCED)[22] and Harrison 2025 (Cohort D)[24] trials involved a single dose of 50 mg of EFX. As EFX 70 mg was used in only one trial, it was omitted. In every trial, EFX and the placebo were administered as subcutaneous injections once a week. Harrison 2021 (BALANCED)[21] and Harrison 2025 (Cohort D)[24] each had trial durations of 12 weeks. Meanwhile, Harrison 2023 (HARMONY)[23] extended over 16 weeks, while both Harrison 2023 (HARMONY)[23] and Noureddin 2025 (SYMMETRY)[25] spanned over 96 weeks. The baseline characteristics of the included RCTs were consistent across the trial arms. The details of the included and excluded studies are presented in Table 1 and Supplementary Table 1, respectively.

Table 1 Baseline characteristics of individual studies and study participants included in the meta-analysis, mean ± SD.

Ref.	Study design, name of the trial, trial registration number	Major inclusion criteria	Groups	Number of study subjects	Female (%)	T2D (%)	Age (year)	BMI (kg/m2)	Hepatic parameters	Concomitant medications (%)	Duration
Harrison et al[21], 2021, Multicenter in United States	Phase 2a RCT, BALANCED, NCT03976401	Adults with biopsy-proven NASH, fibrosis stage 1–3, NAS ≥ 4, HFF ≥ 10% and E > 7.0 kPa	EFX 28 mg	19	53	37	50.4 ± 12.4	38.8 ± 9.3	HFF: 21.4 ± 8.1; NAS: 5.6 ± 1.0; ELF: 9.5 ± 0.6	MFN (37), Statin (32)	12 weeks
			EFX 50 mg	20	50	50	52.6 ± 14.2	36.7 ± 6.8	HFF: 18.3 ± 6.3; NAS: 5.1 ± 1.2; ELF: 9.5 ± 0.9	MFN (42), Statin (37)
			EFX 70 mg	20	55	50	53 ± 13.2	37.2 ± 5.5	HFF: 19.4 ± 6.3; NAS: 5.6 ± 0.7; ELF: 9.5 ± 0.8	MFN (25), Statin (45)
			Placebo	21	71	67	52.4 ± 9.6	37.6 ± 4.8	HFF: 19.3 ± 6.9; NAS: 5.1 ± 1.0; ELF: 9.5 ± 1.0	MFN (48), Statin (33)
Harrison et al[22], 2023, Multicenter in United States	Phase 2a RCT, BALANCED (expansion cohort), NCT03976401	Adults with biopsy-confirmed compensated cirrhosis due to NASH, E > 13.1 kPa, ELF > 9.8, and no evidence of worsening levels of ALT and AST	EFX 50 mg	20	80	50	61.1 ± 10	36.0 ± 5.6	E: 22.1 ± 10.8 kPa; NAS: 4.1 ± 1.7; ELF: 10.4 ± 1.2	GLP-1RA (15), Statin (30)	16 weeks
			Placebo	10	30	50	57.1 ± 14.4	39.1 ± 8.3	E: 25.8 ± 13.2 kPa; NAS: 3.3 ± 2.1; ELF: 9.7 ± 0.8	GLP-1RA (30), Statin (30)
Harrison et al[23], 2023, Multicenter in United States	Phase 2b RCT, HARMONY, NCT04767529	Adults with biopsy-confirmed NASH, defined by NAS ≥ 4 and scores of 1 or higher in each of steatosis, ballooning, and lobular inflammation, with histological stage F2 or F3 fibrosis	EFX 28 mg	42	69	76.2	56.5 ± 9.3	38.3 ± 6.9	HFF: 18.5 ± 6.9; E: 13.8 ± 5.2 kPa; NAS: 5.1 ± 1.0; ELF: 9.7 ± 0.8	GLP-1RA (12), SGLT-2i (14), Statin (52.4)	96 weeks
			EFX 50 mg	43	54	69.8	52.4 ± 11.4	37.2 ± 6.6	HFF: 17.5 ± 6.4; E: 16.0 ± 7.1 kPa; NAS: 5.6 ± 1.1; ELF: 9.8 ± 0.8	GLP-1RA (9), SGLT-2i (12), Statin (32.6)
			Placebo	43	63	65.1	55.0 ± 10.1	38.7 ± 7.7	HFF: 17.1 ± 6.4; E: 14.5 ± 6.2 kPa; NAS: 5.4 ± 1.2; ELF: 9.8 ± 0.7	GLP-1RA (16), SGLT-2i (19), Statin (48.8)
Harrison et al[24], 2025, Multicenter in United States	Phase 2b RCT, Cohort D, NCT05039450	Adults with MASH, F1 to F3 (confirmed by biopsy), and T2D treated with a stable dose of GLP-1RA	EFX 50 mg	21	43	100	59	35.0	HFF: 10.4; E: 10 kPa; ELF: 9.2	GLP-1RA (100), MFN (76), SGLT-2i (33), Statin (81)	12 weeks
			Placebo	10	90	100	55	35.4	HFF: 15.2; E: 12 kPa; ELF: 9.6	GLP-1RA (100), MFN (70), SGLT-2i (20), Statin (50)
Noureddin et al[25], 2025, Multicenter in United States, Puerto, Rico, and Mexico	Phase 2b RCT, SYMMETRY, NCT05039450	Adults with liver histologic features consistent with MASH and compensated cirrhosis (defined as stage 4 fibrosis with a Child-Pugh score of 5 or 6, T2D or two components of MetS	EFX 28 mg	57	68	81	61.7 ± 8.3	36.1 ± 7.1	E: 24.1 ± 12.4 kPa; NAS: 3.9 ± 1.6; ELF: 10.6 ± 0.8	GLP-1RA (19)	96 weeks
			EFX 50 mg	63	70	78	59.4 ± 8.8	34.5 ± 5.9	E: 24.5 ± 13.4 kPa; NAS: 4.1 ± 1.5; ELF: 10.5 ± 0.8	GLP-1RA (33)
			Placebo	61	62	82	61 ± 7.5	36.7 ± 6.8	E: 24.7 ± 14.2 kPa; NAS: 3.7 ± 1.6; ELF: 10.4 ± 0.8	GLP-1RA (26)

ALT: Alanine aminotransferase; AST: Aspartate aminotransferase; BMI: Body mass index; E: Liver stiffness during FibroScan measurement; EFX: Efruxifermin; ELF: Enhanced liver fibrosis; GLP-1RA: Glucagon-like peptide-1 receptor agonists; HFF: Hepatic fat fraction (by Magnetic Resonance Imaging Proton Density Fat Fraction); kPa: Kilopascals; MASH: Metabolic dysfunction-associated steatohepatitis; MetS: Metabolic syndrome; NAS: Non-alcoholic fatty liver disease activity score; NASH: Nonalcoholic steatohepatitis; MFN: Metformin; RCT: Randomized controlled trial; SGLT-2i: Sodium-glucose cotransporter 2 inhibitor; T2D: Type 2 diabetes.

RoB in the included studies

Supplementary Figure 1 illustrates the specific and overall RoB in the included studies. Three studies (60%)[21,23,24] had a low overall RoB, and the other two (40%)[22,25] had some concerns. The trials with some concerns had biases due to missing outcome data (domain 3). Publication bias was not assessed because there were not enough RCTs (at least 10) in the forest plots[19].

Safety outcomes

Table 2 summarizes the safety outcomes of EFX 28 mg and 50 mg compared to placebo. Compared to the placebo, EFX at a dosage of 50 mg, but not at 28 mg, posed higher risks of any TEAE (RR = 1.05, 95%CI: 1.00-1.11, P = 0.04, I² = 0), TEAE leading to treatment discontinuation (RR = 3.05, 95%CI: 1.06 to 8.81, P = 0.04, I² = 0), nausea (RR = 1.78, 95%CI: 1.25-2.54, P = 0.001, I² = 0), and diarrhea (RR = 1.90, 95%CI: 1.04-3.49, P = 0.04, I² = 36). Higher risks of vomiting (RR = 2.17, 95%CI: 1.16-4.08, P = 0.02, I² = 0) and frequent bowel movements (RR = 8.98, 95%CI: 1.71-47.13, P = 0.009, I² = 0) were evident only with EFX 28 mg, but not with EFX 50 mg. Both doses of EFX imparted higher risks of drug-related TEAE (28 mg: RR = 1.45, 95%CI: 1.16-1.82, P = 0.001, I² = 0; and 50 mg: RR = 1.67, 95%CI: 1.36-2.05, P < 0.00001, I² = 0) and increased appetite (28 mg: RR = 3.16, 95%CI: 1.39-7.20, P = 0.0006, I² = 0; and 50 mg: RR = 5.66, 95%CI: 2.71-11.82, P < 0.00001, I² = 0). Other AEs, including severe TEAEs, serious AEs, abdominal pain, fatigue, headache, injection site erythema, and injection site reactions, were similar in both arms of EFX compared to placebo.

Table 2 Comparison of the general safety outcomes in the efruxifermin vs placebo arms.

Outcome variables	EFX dose (mg)	No. of study reports	No. of participants with outcome/participants analyzed (n/N)		Pooled effect size, RR (95%CI)	P value (EFX vs placebo)	I2 (%)	P value (EFX 28 vs 50)
			EFX arm	Placebo arm
Any TEAE	28	3	110/116	113/125	1.03 (0.96-1.10)	0.39	9	0.56
	50	4	141/145	121/135	1.05 (1.00-1.11)	0.04	0
Drug-related TEAE	28	3	79/116	58/125	1.45 (1.16-1.82)	0.001	0	0.38
	50	4	110/145	61/135	1.67 (1.36-2.05)	< 0.00001	0
Severe TEAE	28	2	1/59	1/64	1.11 (0.07-16.47)	0.94	NA	0.70
	50	3	2/82	1/74	2.21 (0.22-22.47)	0.50	NA
TEAE leading to treatment discontinuation	28	3	10/116	3/125	3.15 (0.96-10.35)	0.06	0	0.97
	50	5	16/166	3/145	3.05 (1.06-8.81)	0.04	0
Serious AE	28	3	16/116	11/125	1.51 (0.77-2.97)	0.23	0	0.89
	50	5	19/166	12/145	1.35 (0.30-6.11)	0.70	40
Abdominal pain	28	2	10/76	8/82	1.35 (0.56-3.21)	0.50	0	0.84
	50	2	8/82	8 /82	1.13 (0.27-4.69)	0.87	37
Nausea	28	3	36/116	29/125	1.38 (0.66-2.91)	0.39	53	0.55
	50	5	68/166	32/145	1.78 (1.25-2.54)	0.001	0
Vomiting	28	3	28/116	12/125	2.17 (1.16-4.08)	0.02	0	0.57
	50	4	27/145	12/135	1.80 (0.95-3.39)	0.07	0
Diarrhea	28	2	29/76	22/82	1.42 (0.90-2.24)	0.13	0	0.45
	50	4	59/123	26/102	1.90 (1.04-3.49)	0.04	36
Frequent bowel movements	28	2	12/59	1/64	8.98 (1.71-47.13)	0.009	0	0.26
	50	2	2/62	1/64	1.43 (0.09-22.30)	0.80	37
Increased appetite	28	3	21/116	7/125	3.16 (1.39-7.20)	0.0006	0	0.30
	50	4	45/146	7/135	5.66 (2.71-11.82)	< 0.00001	0
Fatigue	28	2	12/76	12/82	1.04 (0.51-2.14)	0.90	0	0.73
	50	2	11/82	12/82	0.87 (0.42-1.83)	0.72	0
Headache	28	2	14/97	13/104	1.15 (0.57-2.33)	0.69	0	0.68
	50	3	22/126	14/114	1.41 (0.75-2.65)	0.29	0
Injection site erythema	28	3	23/116	14/125	1.66 (0.88-3.14)	0.12	4	0.89
	50	4	30/145	14/135	1.77 (0.95-3.32)	0.07	5
Injection site reaction	28	3	16/116	14/125	1.17 (0.60-2.28)	0.65	0	0.51
	50	4	26/145	14/135	1.71 (0.67-4.39)	0.26	21

AE: Adverse events; EFX: Efruxifermin; NA: Not available; RR: Risk ratio; TEAE: Treatment-emergent adverse event.

Hepatic safety outcomes

Table 3 summarizes the liver-specific safety outcomes of EFX 28 mg and 50 mg in comparison to placebo. Both doses of EFX were more effective than placebo in lowering alanine aminotransferase (ALT) (28 mg: MD: -16.52 U/L, 95%CI: -29.54 to -3.51, P = 0.01, I² = 87; and 50 mg: MD: -16.93 U/L, 95%CI: -28.38 to -5.48, P = 0.004, I² = 92), aspartate aminotransferase (AST) (28 mg: MD: -14.62 U/L, 95%CI: -21.72 to -7.53, P < 0.0001, I² = 74; and 50 mg: MD: -13.35 U/L, 95%CI: -20.94 to -5.75, P = 0.0006, I² = 89), gamma-glutamyl transferase (GGT) (28 mg: MD: -16.52 U/L, 95%CI: -29.54 to -3.51, P = 0.01, I² = 87; and 50 mg: MD: -16.93 U/L, 95%CI: -28.38 to -5.48, P = 0.004, I² = 92), and urate (28 mg: MD: -0.73 mg/dL, 95%CI: -1.06 to -0.41, P < 0.0001, I² = 64; and 50 mg: MD: -0.59 mg/dL, 95%CI: -1.06 to -0.12, P = 0.01, I² = 85) levels from the baseline values. However, the reduction of alkaline phosphatase (ALP) was identical in the EFX and placebo (28 mg: MD: -2.13 U/L, 95%CI: -10.52-6.26, P = 0.62, I² = 86; and 50 mg: MD: -4.46 U/L, 95%CI: -9.61 to -0.70, P = 0.09, I² = 64) groups.

Table 3 Comparison of the hepatic safety outcomes in the efruxifermin vs placebo arms.

Outcome variables	EFX dose (mg)	No. of study reports	No. of participants with outcome analyzed		Pooled effect size, MD (95%CI)	P value (EFX vs placebo)	I2 (%)	P value (EFX 28 vs 50)
			EFX arm	Placebo arm
ALT (U/L)	28	3	118	125	-16.52 (-29.54 to -3.51)	0.01	87	0.96
	50	5	165	145	-16.93 (-28.38 to -5.48)	0.004	92
AST (U/L)	28	3	118	125	-14.62 (-21.72 to -7.53)	< 0.0001	74	0.81
	50	5	165	145	-13.35 (-20.94 to -5.75)	0.0006	89
ALP (U/L)	28	3	118	125	-2.13 (-10.52 to 6.26)	0.62	86	0.64
	50	5	165	145	-4.46 (-9.61 to 0.70)	0.09	64
GGT (U/L)	28	3	118	125	-16.52 (-29.54 to -3.51)	0.01	87	0.96
	50	5	165	145	-16.93 (-28.38 to -5.48)	0.004	92
Urate (mg/dL)	28	3	112	124	-0.73 (-1.06 to -0.41)	< 0.0001	64	0.63
	50	5	158	134	-0.59 (-1.06 to -0.12)	0.01	85

ALP: Alkaline phosphatase; ALT: Alanine aminotransferase; AST: Aspartate aminotransferase; EFX: Efruxifermin; GGT: Gamma-glutamyl transferase; MD: Mean difference.

Hepatic efficacy outcomes

Tables 4 and 5 summarize the hepatic efficacy outcomes of EFX compared to placebo. From baseline values, the relative reduction in HFF was more pronounced with EFX 28 mg (MD: -54.06%, 95%CI: -71.01 to -37.11, P < 0.00001, I² = 76) and EFX 50 mg (MD: -62.0%, 95%CI: -71.27 to -52.73, P < 0.00001, I² = 31) compared to placebo. EFX 50 mg also achieved a greater absolute reduction in HFF (MD: -9.3%, 95%CI: -16.36 to -2.23, P = 0.01, I² = 87). More participants taking EFX 28 mg and 50 mg experienced reductions in HFF of ≥ 30% (28 mg: RR = 4.73, 95%CI: 2.54-8.82, P < 0.00001, I² = 16; and 50 mg: RR = 4.79, 95%CI: 2.76-8.30, P < 0.00001, I² = 4) and ≥ 50% (28 mg: RR = 19.06, 95%CI: 4.82-75.40, P < 0.0001, I² = 0; and 50 mg: RR = 19.03, 95%CI: 5.68-63.78, P < 0.00001, I² = 0) by the end of the studies, compared to those on placebo. More participants experienced normalization of liver fat (HFF ≤ 5%) with EFX 28 mg (RR = 8.71, 95%CI: 2.06-36.75, P = 0.003, I² = 0) and EFX 50 mg (RR = 12.45, 95%CI: 4.09-37.93, P < 0.00001, I² = 0) compared to those receiving placebo. Both doses of EFX resulted in larger reductions in ELF score (28 mg: MD: -0.67, 95%CI: -0.83 to -0.51, P < 0.00001, I² = 0; and 50 mg: MD: -0.75, 95%CI: -0.89 to -0.61, P < 0.00001, I² = 0) and Pro-C3 Levels (28 mg: MD: -5.42 μg/L, 95%CI: -8.15 to -2.70, P < 0.0001, I² = 68; and 50 mg: MD: -6.11 μg/L, 95%CI: -8.06 to -4.16, P < 0.00001, I² = 53) compared to placebo. A significantly greater reduction in liver stiffness was observed with EFX 50 mg compared to placebo (MD: -2.0 kPa, 95%CI: -2.51 to -1.50, P < 0.00001, I² = 0), but not with EFX 28 mg (MD: -1.9 kPa, 95%CI: -3.88-0.08, P = 0.06, I² = 0). Comparable proportions of subjects receiving EFX 50 mg and placebo achieved a decrease in NAS ≥ 2 (RR = 1.96, 95%CI: 0.64-5.99, P = 0.24, I² = 0). Compared to the placebo, both doses of EFX were superior regarding MASH resolution and improvement in fibrosis stage (28 mg: RR = 4.59, 95%CI: 1.29-16.25, P = 0.02, I² = 0; and 50 mg: RR = 6.33, 95%CI: 1.83-21.90, P = 0.004, I² = 0), MASH resolution and no worsening of fibrosis stage (28 mg: RR = 3.19, 95%CI: 1.47-6.93, P = 0.003, I² = 0; and 50 mg: RR = 4.89, 95%CI: 2.41-9.91, P < 0.0001, I² = 0), and fibrosis regression by ≥ 1 stage with no worsening in steatohepatitis (28 mg: RR = 2.06, 95%CI: 1.28-3.30, P = 0.003, I² = 0; and 50 mg: RR = 2.44, 95%CI: 1.55-3.83, P = 0.0001, I² = 0). However, EFX was not better than the placebo in fibrosis regression by ≥ 2 stage with no worsening in steatohepatitis (28 mg: RR = 2.68, 95%CI: 0.70-10.18, P = 0.15, I² = 0; and 50 mg: RR = 2.68, 95%CI: 0.69-10.18, P = 0.15, I²= 0).

Table 4 Comparison of the hepatic efficacy (continuous) outcomes in the efruxifermin vs placebo arms.

Outcome variables (continuous)	EFX dose (mg)	No. of study reports	No. of participants with outcome analyzed		Pooled effect size, MD (95%CI)	P value (EFX vs placebo)	I2 (%)	P value (EFX 28 vs 50)
			EFX arm	Placebo arm
HFF (relative change) (%)	28	2	57	63	-54.06 (-71.01 to -37.11)	< 0.00001	76	0.42
	50	3	71	73	-62.00 (-71.27 to -52.73)	< 0.00001	31
HFF (absolute change) (%)	50	2	36	31	-9.30 (-16.36 to -2.23)	0.01	87	NA
ELF score	28	3	113	124	-0.67 (-0.83 to -0.51)	< 0.00001	0	0.46
	50	5	166	144	-0.75 (-0.89 to -0.61)	< 0.00001	0
Liver stiffness (kPa)	28	2	95	103	-1.90 (-3.88 to 0.08)	0.06	0	0.92
	50	4	138	123	-2.00 (-2.51 to -1.50)	< 0.00001	0
Pro-C3 (μg/L)	28	3	113	122	-5.42 (-8.15 to -2.70)	< 0.00001	68	0.69
	50	5	157	142	-6.11 (-8.06 to -4.16)	< 0.00001	53
Pro-C3 (%)	50	2	83	71	-20.85 (-36.07 to -5.64)	0.007	0	NA

EFX: Efruxifermin; ELF: Enhanced liver fibrosis; HFF: Hepatic fat fraction; kPa: Kilopascals; MD: Mean difference; NA: Not available; Pro-C3: N-terminal propeptide of type III collagen.

Table 5 Comparison of the hepatic efficacy (categorical) outcomes in the efruxifermin vs placebo arms.

Outcome variables (categorical)	EFX dose (mg)	No. of study reports	No. of participants with outcome/participants analyzed (n/N)		Pooled effect size, RR (95%CI)	P value (EFX vs placebo)	I2 (%)	P value (EFX 28 vs 50)
			EFX arm	Placebo arm
Reduction in HFF ≥ 30%	28	2	48/54	11/62	4.73 (2.54 to 8.82)	< 0.00001	16	0.98
	50	2	48/52	11/62	4.79 (2.76 to 8.30)	< 0.00001	4
Reduction in HFF ≥ 50%	28	2	35/54	2/62	19.06 (4.82 to 75.40)	< 0.0001	0	1.0
	50	2	44/52	2/62	19.03 (5.68 to 63.78)	< 0.00001	0
HFF ≤ 5%	28	2	17/54	2/62	8.71 (2.06 to 36.75)	0.003	0	0.70
	50	3	41/68	3/72	12.45 (4.09 to 37.93)	< 0.00001	0
Decrease in NAS ≥ 2	50	2	17/25	2/7	1.96 (0.64 to 5.99)	0.24	0	NA
MASH resolution and improvement in fibrosis stage	28	2	15/51	2/43	4.59 (1.29 to 16.25)	0.02	0	0.72
	50	2	19/47	2/43	6.33 (1.83 to 21.90)	0.004	0
MASH resolution and no worsening of fibrosis stage	28	2	24/51	6 43	3.19 (1.47 to 6.93)	0.003	0	0.43
	50	3	36/59	6/48	4.89 (2.41 to 9.91)	< 0.0001	0
Fibrosis regression by ≥ 1 stage and no worsening in steatohepatitis	28	3	42/108	19/104	2.06 (1.28 to 3.30)	0.003	0	0.61
	50	4	55/122	19/109	2.44 (1.55 to 3.83)	0.0001	0
Fibrosis regression by ≥ 2 stage and no worsening in steatohepatitis	28	2	9/51	2/43	2.68 (0.70 to 10.18)	0.15	0	1.00
	50	2	9/47	2/43	2.68 (0.69 to 10.38)	0.15	0

EFX: Efruxifermin; HFF: Hepatic fat fraction; MASH: Nonalcoholic steatohepatitis; NA: Not available; NAS: Non-alcoholic fatty liver disease activity score; RR: Risk ratio.

Metabolic outcomes

Table 6 summarizes the metabolic outcomes of EFX compared to placebo. EFX 50 mg was more effective than placebo for body weight reduction (MD: -0.149 kg, 95%CI: -2.74 to -0.25, P = 0.02, I² = 0), while EFX 28 mg did not show this effect (MD: 0.05 kg, 95%CI: -1.48-1.57, P = 0.95, I² = 0). Both doses of EFX outperformed placebo in reducing HbA1c (28 mg: MD: -0.24%, 95%CI: -0.46 to -0.02, P = 0.03, I² = 0; and 50 mg: MD: -0.4%, 95%CI: -0.55 to -0.25, P < 0.00001, I² = 0), and HOMA-IR (28 mg: MD: -5.7, 95%CI: -8.21 to -3.19, < 0.00001, I² = 0; and 50 mg: MD: -5.13, 95%CI: -6.94 to -3.32, P < 0.00001, I² = 0). A larger reduction in fasting C-peptide was achieved with EFX 50 mg compared to placebo (MD: -1.16 μg/L, 95%CI: -1.59 to -0.73, P < 0.00001, I² = 0). Increases in adiponectin levels were also more pronounced with EFX 28 mg (MD: 1.95 mg/dL, 95%CI: 0.13-3.76, P = 0.04, I² = 76) and EFX 50 mg (MD: 3.65 mg/dL, 95%CI: 2.06-5.24, P < 0.00001, I² = 85) compared to the placebo. EFX, at both doses, was more effective than placebo in reducing TG (28 mg: MD: -53.83 mg/dL, 95%CI: -86.82 to -20.84, P = 0.001, I² = 83; and 50 mg: MD: -63.21 mg/dL, 95%CI: -92.07 to -34.35, P < 0.0001, I² = 82), non-HDL-C (28 mg: MD: -16.1 mg/dL, 95%CI: -29.92 to -2.29, P = 0.02, I² = 68; and 50 mg: MD: -18.4 mg/dL, 95%CI: -26.99 to -9.81, P < 0.0001, I² = 7), and LDL-C (28 mg: MD: -11.1 mg/dL, 95%CI: -17.84 to -4.35, P = 0.001, I² = 14; and 50 mg: MD: -6.99 mg/dL, 95%CI: -12.9 to -1.08, P = 0.02, I² = 8), and in improving HDL-C (28 mg: MD: 10.77 mg/dL, 95%CI: 7.51-14.03, P < 0.00001, I² = 46; and 50 mg: MD: 12.77 mg/dL, 95%CI: 10.68-14.86, P < 0.00001, I² = 0). Both EFX doses reduced apo-B (28 mg: MD: -15.08 mg/dL, 95%CI: -22.02 to -8.14, P < 0.0001, I² = 0; and 50 mg: MD: -11.1 mg/dL, 95%CI: -17.01 to -5.19, P = 0.0002, I² = 0) more than placebo. However, significant apo-C3 reduction was achieved only with EFX 28 mg (MD: -2.24 mg/dL, 95%CI: -4.1 to -0.38, P = 0.02, I² = 69), not with EFX 50 (MD: -2.44 mg/dL, 95%CI: -4.89-0.01, P = 0.05, I² = 82).

Table 6 Comparison of the metabolic outcomes in the efruxifermin vs placebo arms.

Outcome variables	EFX dose (mg)	No. of study reports	No. of participants with outcome analyzed		Pooled effect size, MD (95%CI)	P value (EFX vs placebo)	I2 (%)	P value (EFX 28 vs 50)
			EFX arm	Placebo arm
Body weight (kg)	28	3	113	124	0.05 (-1.48 to 1.57)	0.95	0	0.13
	50	5	158	144	-1.49 (-2.74 to -0.25)	0.02	0
HbA1c (%)	28	3	113	124	-0.24 (-0.46 to -0.02)	0.03	0	0.24
	50	5	158	144	-0.40 (-0.55 to -0.25)	< 0.00001	0
HOMA-IR	28	3	113	124	-5.70 (-8.21 to -3.19)	< 0.00001	0	0.72
	50	4	135	130	-5.13 (-6.94 to -3.32)	< 0.00001	0
C-peptide (μg/L)	50	2	54	50	-1.16 (-1.59 to -0.73)	< 0.00001	0	NA
Adiponectin (mg/dL)	28	2	55	53	1.95 (0.13 to 3.76)	0.04	76	0.17
	50	3	72	72	3.65 (2.06 to 5.24)	< 0.00001	85
TG (mg/dL)	28	3	111	124	-53.83 (-86.82 to -20.84)	0.001	83	0.68
	50	4	137	134	-63.21 (-92.07 to -34.35)	< 0.0001	82
Non-HDL-C (mg/dL)	28	2	92	103	-16.10 (-29.92 to -2.29)	0.02	68	0.78
	50	3	117	113	-18.40 (-26.99 to -9.81)	< 0.00001	7
LDL-C (mg/dL)	28	3	111	124	-11.10 (-17.84 to -4.35)	0.001	14	0.37
	50	4	137	134	-6.99 (-12.90 to -1.08)	0.02	8
HDL-C (mg/dL)	28	3	111	124	10.77 (7.51 to 14.03)	< 0.00001	46	0.31
	50	4	137	134	12.77 (10.68 to 14.86)	< 0.00001	0
Apo-B (mg/dL)	28	2	56	53	-15.08 (-22.02 to -8.14)	< 0.00001	0	0.39
	50	3	75	73	-11.10 (-17.01 to -5.19)	0.0002	0
Apo-C3 (mg/dL)	28	2	55	63	-2.24 (-4.10 to -0.38)	0.02	69	0.90
	50	2	56	63	-2.44 (-4.89 to 0.01)	0.05	82

Apo-B: Apolipoprotein B; Apo-C3: Apolipoprotein C3; EFX: Efruxifermin; HbA1c: Glycated hemoglobin; HDL-C: High-density lipoprotein cholesterol; HOMA-IR: Homeostatic model assessment for insulin resistance; LDL-C: Low-density lipoprotein cholesterol; MD: Mean difference; NA: Not available; TG: Triglycerides.

Comparison of EFX 28 mg and 50 mg for safety and efficacy outcomes

EFX 28 mg and 50 mg demonstrated similar general and liver-specific safety profiles when compared to placebo (subgroup differences for EFX 28 mg and 50 mg for all outcomes were statistically insignificant, i.e., P ≥ 0.05; Tables 2 and 3). EFX 28 mg and 50 mg demonstrated similar efficacy for hepatic and metabolic outcomes (Tables 4, 5 and 6).

Management of heterogeneity

Since the number of included trials is small, subgroup analyses beyond EFX dose and meta-regression were not performed. Leave-one-out sensitivity analyses were conducted for primary and important secondary outcomes with high heterogeneity (when I² > 75%) to detect changes in statistical significance and significant heterogeneity (at least a 2-step change) (Supplementary Table 2). For ALT with EFX 28 mg, statistical significance was lost when either Harrison 2021 (BALANCED) or Harrison 2023 (HARMONY) studies were removed; the heterogeneity among the studies decreased after removing the study Noureddin 2025 (SYMMETRY). For ALT and AST with EFX 50 mg, removing either of the studies did not change the statistical significance or heterogeneity among the studies. For changes in ALP with EFX 28 mg, removing either of the studies did not alter the statistical significance or heterogeneity among the studies. For GGT changes with EFX 28 mg, statistical significance was lost when either Harrison 2021 (BALANCED) or Harrison 2023 (HARMONY) studies were removed; the heterogeneity among the studies decreased after removing the study Noureddin 2025 (SYMMETRY). For GGT with EFX 50 mg, removing either of the studies did not change the statistical significance or heterogeneity among the studies.

DISCUSSION

This meta-analysis examined the safety and efficacy of EFX, an analogue of FGF21, in treating MASH. The disease poses serious health risks to many worldwide, and few approved pharmacotherapeutic agents exist. Several novel agents, including FGF21 analogs, are being investigated to fill this treatment gap and prevent complications. FGF21 analogs, like EFX, target the multifaceted pathobiology of MASH. EFX modulates lipid metabolism, enhances insulin sensitivity, reduces inflammation, and decreases hepatic steatosis[27]. Our analysis pooled data from four RCTs involving 450 participants. The results indicated significant improvements in fibrosis-related, biochemical, and metabolic parameters with EFX, along with a relatively safe overall profile in patients with MASH. In addition to histological benefits, both doses of EFX significantly outperformed placebo, with notable reductions in liver fat measured by MRI-PDFF. Hepatic enzymes, such as ALT, AST, and GGT, also showed significant improvement, indicating that the drug can modify disease activity. Furthermore, EFX was linked to positive effects on metabolic parameters, including reductions in HOMA-IR, HbA1c, triglycerides, and Apo-B, as well as increases in HDL-C and adiponectin levels. These pleiotropic effects are particularly beneficial, as MASH commonly coexists with metabolic syndrome and T2D, which can worsen cardiovascular disease outcomes[28,29]. Despite its therapeutic benefits, EFX was linked to a higher occurrence of TEAEs and drug discontinuation compared to placebo. Drug-related TEAEs were higher with both doses of EFX. Interestingly, some GI adverse effects, such as vomiting and diarrhea, were observed with EFX 28 mg but not with EFX 50 mg. Other AEs, including severe TEAEs, were comparable to placebo, highlighting a reasonable safety profile for EFX.

Compared to other emerging therapies, EFX offers distinct advantages. Data indicate that SGLT-2 inhibitors may effectively treat MASLD, particularly by reducing liver fat content, likely due to their glucose-lowering and weight-reducing effects[3,30]. While GLP-1 receptor agonists and dual agonists effectively promote weight loss and improve glycemic control, they provide relatively modest histological benefits in MASH and pose higher risks of GI AEs[3,30,31]. However, the degree of weight loss and HbA1c reduction with EFX is not as strong as with injectable GLP-1 receptor agonists and dual agonists. So, EFX may not be as effective as these GLP-1-based therapies in the holistic management of individuals with MASH who have obesity and T2D, the most common comorbidities of MASH. FXR agonists, such as obeticholic acid, have demonstrated efficacy in improving fibrosis; however, factors including pruritus, LDL-C elevation, and poor tolerability limit their use[32]. Resmetirom is hepatoselective and effective in reducing hepatic fat, with only modest improvement in fibrosis. In contrast to EFX, which exhibits multisystem effects through both endocrine and metabolic pathways, resmetirom specifically targets intrahepatic lipid metabolism[33]. In a Phase 3 trial, resmetirom at daily doses of 80 mg and 100 mg achieved NASH resolution in 25.9% and 29.9% of patients, respectively, and an improvement in ≥ 1 stage of fibrosis in 24.2% and 25.9%, compared to 9.7% and 10.3% with placebo[34]. In a recent meta-analysis, resmetirom demonstrated a robust reduction in MRI-PDFF, with a mean relative hepatic fat reduction of 27.76% in the 80 mg group and 36.01% in the 100 mg group, respectively[35]. In contrast, EFX has a comparatively higher efficacy, with MASH resolution in up to 41% and fibrosis improvement in 48% of patients treated with EFX 50 mg. EFX improved body weight, insulin resistance, HbA1c, lipids, and pro-inflammatory markers; most of these benefits were not observed with resmetirom[35]. Thus, while resmetirom offers targeted hepatic lipid-lowering and some modest histological improvements, EFX seems to hold more promise for comprehensive MASH management, addressing hepatic, metabolic, and fibrotic aspects. While both agents were reasonably tolerated, the parenteral route of administration and GI side effects (e.g., nausea, increased appetite) may pose challenges to therapeutic compliance for EFX compared to oral resmetirom. Lanifibranor, a pan-peroxisome proliferator-activated receptor agonist targeting multiple metabolic and inflammatory pathways, is another oral drug under development for MASH and T2D. In the Phase IIb NATIVE trial, lanifibranor significantly improved key histological endpoints in patients with biopsy-proven MASH and fibrosis stages F2-F3[36]. These innovative agents are expected to transform the future treatment of MASH.

Strengths and limitations

This SR/MA provides the first comprehensive analysis of the safety and therapeutic benefits of EFX in MASH, based on the clinical trials published to date. Despite its potential, this meta-analysis has several limitations. The number of trials and study participants included is still small, and all are Phase 2 trials. All of the trials were conducted in North America (primarily in the United States), which limits their global representativeness. The inclusion criteria were not homogeneous throughout the trials, especially regarding liver stiffness and histology. Two trials had short durations (12-16 weeks), while the other two extended to 96 weeks. MASH being very prevalent and lifelong disorders, drug interventions should ideally involve larger cohorts with longer-term follow-up data to obtain more meaningful therapeutic outcome data. The calculated RRs can be sensitive to outcome frequency, especially in small studies, and may not accurately reflect alterations in the risk when dealing with rare outcomes. There is moderate to high heterogeneity for several outcomes, likely due to differences in patient characteristics and fibrosis stages. Some surrogate endpoints of the SR/MA, like MRI-PDFF and liver stiffness, do not always correlate with long-term clinical outcomes in these patients. Finally, long-term safety data are still awaited, especially concerning the risk of immunogenicity, oncogenic potential, and hepatic decompensation in patients with cirrhosis. Despite these limitations, the remarkable benefits and reasonable safety profile of EFX in comparison to other medications currently approved for MASH make our study promising for future large-scale research.

CONCLUSION

This meta-analysis supports EFX’s potential as a multifaceted and effective treatment for MASH. EFX distinguishes itself among new therapies for the disease, demonstrating notable histological, biochemical, and metabolic enhancements alongside a reassuring safety profile. Although more data from Phase 3 trials involving diverse groups of participants in terms of the MASH stage and ethnicity will be needed to validate these findings and evaluate long-term safety, existing evidence encourages ongoing development and clinical use.