4. Discussion
In this systematic review and meta-analysis of 8 RCTs, including 967 patients, the safety and efficacy of FGF21 analogs were compared to placebo in patients with NASH. The main findings from the combined analysis were: (1) NASH resolution was not different between the two groups; (2) FGF21 analogs reduced glycated hemoglobin; (3) triglyceride levels were lower in patients using FGF21 analogs; (4) the intervention group had a higher incidence of adverse events.
The mechanism of NASH development, progression, and complications is likely to be related to metabolic syndrome, diabetes, and dyslipidemia.21-23 Ideally, treatment for this condition would simultaneously address both local hepatic and metabolic factors.24-26 In this regard, FGF21 analogs demonstrate efficacy in inhibiting adipose tissue lipolysis and liver lipogenesis, contributing to a significant decrease in intrahepatic triglyceride concentrations.27,28 However, in the FALCON 1 trial, 197 patients with stage of III or IV fibrosis were randomized to receive once-weekly subcutaneous injections of pegbelfermin (PGBF) or placebo for 48 weeks. In the 148 patients who underwent PGBF, a higher NASH resolution was not found.10 The study also did not meet its primary outcomes, such as 1-stage fibrosis improvement without NASH worsening or NASH improvement without fibrosis worsening probably due to a lack of dose-related changes in PGBF response rates. The effect of FGF21 analogs on NASH resolution in the remained trials, however, found a benefit in the same endpoint for patients who were randomized to receive Pegbelfermin,11 Pegozafermin,18 and Efruxifermin 15,16 as compared with placebo.
There also exists uncertainty as to the optimal dose of FGF21 analogs for NASH disease. The dosage of FGF21 analogs varied substantially between the included studies, ranging from 3 mg to 70 mg. Accordingly, the lack of clinical benefit with FGF21 analogs for NASH in previous studies could be related to the relatively reduced number of patients exposed to FGF21 analogs within groups by dose. Therefore, we performed a dedicated analysis of studies with low, medium, and high doses for NASH resolution, finding no difference between groups. However, a leave-one-out analysis excluding the FALCON-1 trial10showed a higher incidence of NASH resolution in patients receiving FGF21 analogs, regardless of the dose. In that trial, the study population had significant NASH disease activity and fibrosis at baseline in non-invasive assessments. In addition, the FALCON 1 study cirrhosis progression rate at week 24 was higher than those seen in previous studies at later time points 29,30 and in a natural history study.31 Thus, in the absence of definitive improvement in clinical outcomes, our findings do not appear to support FGF21 analogs in all NASH patients, considering the administration of these therapeutic agents in non-severe fibrosis and NASH disease.
Furthermore, NAFLD is intrinsically linked to insulin resistance, with a prevalence about 2 to 5 times higher in individuals with type 2 diabetes compared to non-diabetics.32-35 This association, in turn, amplifies the risks of hepatic fibrosis and mortality.36,37 Therefore, controlling HbA1c is essential for a better prognosis among patients 38,39Recent studies demonstrate benefits in insulin resistance, suggesting a potential positive impact on the development of type 2 diabetes (DM2).15-17As observed in the HARMONY study, by suppressing lipolysis, FGF21 analogs increase the action of insulin in adipose tissue.17 These data corroborate the findings of our meta-analysis, in which the drug is associated with a reduction in glycated hemoglobin.
It is well established that the function of fibroblast growth factor 21 is related to the regulation of glucose and lipid metabolism, as well as the development of adaptive responses to stress and nutrient deprivation.40 In addition to lipid metabolism, FGF-21 stimulates lipolysis in brown adipose tissue during fasting, hepatic ketogenesis, and free fatty acid oxidation, acting in an autocrine and paracrine manner in the liver.41 The resulting free fatty acids from the lipolysis process provide substrates for these metabolic processes and stimulate PPARα activity in the liver, leading to increased protein and FGF21 gene transcription.42,43 This is consistent with our analysis, which suggests that FGF21 analogs are associated with a reduction in triglyceride levels, probably due to free fatty acid oxidation.
Although the data is promising, addressing potential side effects is essential before considering these drugs as standard treatment for NASH. Overall, FGF21 analogs have been considered safe, with frequent but manageable adverse events such as gastrointestinal disorders and injection site reactions.44 In a recently published RCT, adverse events were reported in 79% of patients receiving weekly pegofazemin. However, only one severe event, acute pancreatitis, was deemed by the investigator to be drug-related.18 In the HARMONY study, among the 63 adverse events reported in patients in the intervention group, only one case of ulcerative esophagitis in a patient with a history of gastroesophageal reflux disease was considered severe and was attributed to the use of efruxifermin.17
Our study has some important limitations. First, the follow-up time differs significantly between studies. Second, the drug dosage used was also heterogeneous among studies. To address this, we performed a subgroup analysis by dosage (low, median, and high) for the main outcomes. Third, patients with different fibrosis stages were included. Unfortunately, the intended subgroup analysis by fibrosis stage was not possible due to lack of data, since the studies had a variety of fibrosis worsening and they did not report outcomes according to this baseline characteristic. These factors contributed to high heterogeneity in the pooled analysis. In light of this, leave-one-out sensitivity analyses were performed.