Efficacy and Safety of GLP-1 Receptor Agonists in Patients With Type 2 Diabetes Mellitus and Non-Alcoholic Fatty Liver Disease: A Systematic Review and Meta-Analysis

The rising incidence of type 2 diabetes mellitus (T2DM) is a major concern in health care worldwide. According to global epidemiological data, approximately 462 million individuals were confirmed to have T2DM in 2017 (1). Non-alcoholic fatty liver (NAFLD) and T2DM commonly coexist in clinical practice. Epidemiological evidence suggests a strong bidirectional relationship between T2DM and NAFLD (2). Approximately 50%-70% of person with diabetes have NAFLD (3). NAFLD is a strong clinical signal for insulin resistance and metabolic syndrome and is considered to be a confirmative risk factor for T2DM (4). According to a meta-analysis of 80 trials from twenty countries (5), the global prevalence rates of NAFLD, non-alcoholic steatohepatitis (NASH), and advanced fibrosis in patients with T2DM were 55.5%, 37.3% and 17.0%, respectively. The existence of T2DM is closely associated not only with advanced fibrosis in cross-sectional data (6, 7), but also with rapid progression of hepatic fibrosis (8, 9). In Japan, liver-related disease is the third leading cause of mortality (9.3%) in patients with T2DM (10).

NAFLD is a major cause of liver disease worldwide. In US, NASH has become the leading cause of end-stage liver disease and the main indication for liver transplantation. The major risk of NAFLD include alcohol consumption, obesity, T2DM, and metabolic syndrome. It has become increasingly clear that NAFLD is the hepatic manifestation of metabolic syndrome and is highly prevalent in obese and diabetic subjects (11). NAFLD encompasses a broad clinical spectrum ranging from non-alcoholic fatty liver to NASH, advanced fibrosis, cirrhosis, and hepatocellular carcinoma (12). NASH can be called “diabetic hepatopathy”. A meta-analysis of epidemiology studies suggested that NAFLD has a global prevalence of 25.24%, with the highest prevalence rates in the Middle East and South America and the lowest rate in Africa (13). NASH has been calculated to be present in 2%-5% of the general population (8), while NAFLD accounts for approximately 20% of cases of NASH (6). Until now, liver biopsy has been the gold standard for identifying simple steatosis, NASH, or fibrosis. From a histological point of view, only more than 5% of hepatocytes undergo degeneration called simple steatosis. Diagnosis of NASH necessary requires steatosis (more than 5%), and both lobular inflammation and ballooning degeneration of hepatocytes with a mainly zone 3 distributions (14). NASH gradually progresses to fibrosis, cirrhosis, and hepatocellular carcinoma. Hepatic fibrosis shows intrahepatic connective hyperplasia and deposition. There are 4 stages according to severity, stage1 (perivenular, perisinusoidal, or periportal fibrosis), stage 2 (both zone 3 and periportal fibrosis), stage 3 (bridging fibrosis), and stage 4 (cirrhosis) (15).

The pathogenesis of NAFLD involves a complex interaction between environmental factors, obesity, changes in the microbiota, and predisposing genetic variants and results in a disturbed lipid homeostasis and an excessive accumulation of triglycerides and other lipid species in hepatocytes (16, 17). Lipotoxicity and inflammation are the main pathogenic factors associated with NASH (18, 19). Moreover, there is growing evidence that overnutrition negatively interfere with immune system. In the progression of NAFLD, there is a specific link between fat accumulation and inflammation. Adipokines (the most abundant adipokines included leptin and adiponectin) produced by white fat are key factors between metabolism and immunity. At the same time, the two adipokines have different roles in inducing inflammatory response. Leptin upregulates TNF-a and IL-6, and is associated with insulin resistance and T2DM. In contrast, adiponectin has anti-inflammatory properties and downregulates the expression and release of proinflammatory immune mediators, for example, κB activation and TNF-α expression (20, 21). Besides, adiponectin anti-aliphatic activity in hepatocytes through increasing oxidation of free fatty acids and decreasing gluconeogenesis, free fatty acid flow and de novo lipogenesis (22). It has evidence support that serum leptin levels increased in NAFLD patients, while serum adiponectin levels decreased (23), which is mainly related to the characteristics of two kinds of adipokines. In addition, systemic insulin resistance to progression of non-alcoholic fatty is the main driver of nonalcoholic fatty liver hepatitis, cirrhosis, and liver cancer (17, 18). NAFLD and NASH impose a substantial economic burden and are responsible for poor health-related quality of life (11, 24).

As we all know, liver biopsy is the gold standard for diagnosing NAFLD. However, it has well-known limitations, including invasiveness, poor acceptability, sampling variability, and cost. As a result, noninvasive strategies are gradually replacing liver biopsy diagnosis NAFLD. These strategies including a “biological” approach based on the quantification of biomarkers in serum samples or a “physical” approach based on the measurement of liver stiffness, using either ultrasound or magnetic resonance-based elastography and so on techniques. However, in previous researches suggested that there was no significant correlation between serum biomarkers level, for example liver enzymes, and NAFLD (25, 26). In addition, patients with advanced liver disease show decreased alanine aminotransferase (ALT) levels. Thus, it is very unreasonable for us to diagnosis NAFLD only rely on the levels of serum biomarkers. It is important that we should combine serum biomarkers and physical measurements. We know that Fibroscan allows a rapid assessment and is reasonably accurate for diagnosing the presence of steatosis (27). However, the controlled attenuation parameter (CAP), measured by Fibroscan, is limited by high failure rates in obesity, lack of exact anatomic localization, and low accuracy for quantifying the amount of steatosis. Compared with MRI, studies found that the detection sensitivity is poor for patients with low fat concentrations using computerized tomography (CT) (28, 29). Magnetic Resonance Imaging (MRI) can detect small amounts of liver fat and is considered the most accurate non-invasive method to quantify liver fat. MRI proton density fat fraction (MRI-PDFF) corrects for imaging confounders that can affect liver fat measurement and has emerged as an accurate, reproducible biomarker of liver fat (30, 31). In a longitudinal assessment, MRI-PDFF was more accurate at detecting changes in liver fat than liver biopsy (32, 33). MRI, as noninvasive approach, diagnosis of NAFLD is very promising in future.

GLP-1 receptor agonists (GLP-1RAs) are gut-derived incretin hormone that induces weight loss and insulin sensitivity. These agents act directly on human hepatocytes in vitro, reducing steatosis by decreasing de-novo lipogenesis and increasing fatty acid oxidation (34, 35). In the LEAN trial, which investigated the effects of 48 weeks of treatment with liraglutide 1.8 mg on liver histology, resolution of NASH without worsening of fibrosis was found in nine of 23 patients randomized to liraglutide but in only two of 22 patients randomized to placebo (36).

Furthermore, several randomized controlled trials (RCTs) have compared the efficacy and safety of GLP-1RAs with that of other active agents or placebo in patients with T2DM and NAFLD. However, the sample sizes were small in all of these studies. Therefore, comprehensive evaluation of the various studies requires a meta-analysis. We hypothesized that more significant improvement of hepatic fibrosis and steatosis may be achieved by GLP-1RAs than by other antidiabetic agents or placebo in patients with T2DM and T2DM. This systematic review and meta-analysis were performed to summarize current evidence for the efficacy and safety of GLP-1RAs in these patients.

We have performed a comprehensive meta-analysis of the most commonly prescribed GLP-1 RAs based on the largest pool of GLP-1 RA trials for patients with T2DM and NAFLD to date. Our results show that, compared with placebo or other active comparators, GLP-1 RAs can significantly improve IHA, SAT, VAT, ALT, AST, body weight, BMI, waist circumference, FBG, percent HbA_1C, HoMA- IR, TC and TG. In contrast, use of a GLP-1 RA did not lead to statistically significant changes in FIB-4, NFS, PBG, LDL, HDL, SBP or DBP compared with controls. The major adverse events reported for GLP-1RAs were mild-to-moderate gastrointestinal discomfort and asymptomatic hypoglycemia, which resolved within a few weeks.

Improved insulin resistance and weight loss are the cornerstones of NAFLD treatment. Insulin resistance in NAFLD patients is associated with reduced adiponectin secretion by adipocytes, which is mainly manifested by adiponectin mediated signaling down-regulation of fatty acid β -oxidation (FAO), inhibition of glucose utilization and fatty acid synthesis. What matters is that in patients with NAFLD, mechanistic studies demonstrated that GLP-1RAs is associated with improvements in de novo lipogenesis, β-oxidation, and IR (systemic, adipose, and hepatic) (45, 46). In addition, GLP-1 RAs can influence IR through promotion of weight loss (delayed gastric emptying, appetite suppression). HoMA-IR is an indicator to evaluate insulin resistance, which runs through the whole process of occurrence and development of NAFLD. In our study we found that GLP1-RAs can improve HoMA- IR, which further prove that the treatment of NAFLD is benefit from GLP-1RAs. We know that the most serious risk of death from NAFLD is cardiovascular disease. We know that GLP-1 RAs have additional roles in reducing the risk of major adverse cardiovascular events and cardiovascular deaths in high-risk patients with T2DM, especially in the Asian population (47–49). T2DM and NAFLD often coexist, GLP-1RAs has the potential to reduce cardiovascular risk in NAFLD patients (48, 50). However, this benefit was only shown with liraglutide, injectable semaglutide and dulaglutide long-term use.

Previous studies have shown that GLP-1RAs act directly on human hepatocytes to decrease steatosis by preventing regeneration of fat and increasing oxidation of fatty acids and have shown that GLP-1RAs reduce IHA without insulin existed (34, 51, 52). These findings indicate that GLP-1RAs decrease IHA and have no relationship with weight loss. However, one study found that weight reduction alleviated intrahepatic fat content and aminotransferase levels (53). Moreover, a systematic review that included 23 RCTs of the effect of lifestyle interventions, such as diet, physical activity, and/or exercise, on the hepatic indicators of steatosis showed strong associations of reduction in IHA and liver aminotransferase levels with weight loss (54). Moreover, although weight loss in the range of approximately 5% -7% can decrease steatosis, a weight reduction of 8%-10% is needed to reverse steatohepatitis (55). In addition, reduction in NAFLD, resolution of steatohepatitis, and regression of fibrosis occurred in patients with a weight loss of ≥10% (56). Exercise and diet control appeared to have limited effects on NAFLD, given that 3-6% of patients achieved weight loss by long-term lifestyle interventions and less than 50% attained a weight reduction of 7%.

In recently years, the clinical role of GLP-1RAs has been explored in the treatment of patients with T2DM and NAFLD. A large number of clinical studies has shown that GLP-1RAs are better than placebo or other active comparators in their ability to achieve a significant decrease in IHA in these patients (37–43, 57–61). Surprisingly, in one study (32), IHA was not significantly reduced by administration of liraglutide for 12weeks, but was significantly decreased by insulin. However, Yan et al. (26, 30) demonstrated that liraglutide could significantly reduce intrahepatic fat content when compared with the effect produced by insulin 24/26 weeks of follow-up. The inconsistent findings of these studies may reflect differences in follow-up duration, the baseline characteristics of the study population, and other factors. Further clinical studies are needed to clarify the effects of liraglutide and insulin in patients with T2DM and NAFLD.

Several studies suggested that GLP-1RAs produce a significant change in IHA that correlates with changes in weight and HbA_1C in patients with T2DM and NAFLD (37, 58, 62). These studies divided patients into two subgroups according to whether weight loss was <5% or >5%. Eventually, they concluded that the reduction in IHA in patients who achieved marked weight loss was significantly greater than that in those with less weight loss. Furthermore, Feng et al. (58) reported patients with T2DM and NAFLD whose HbA_1c decreased by≥2.5% showed more significant reductions in IHA than those whose HbA_1c decreased by<2.5%. Patients with an HbA_{1c <}6.5% had a markedly lower IHA content than those with HbA_1c≥6.5%. These results are similar to our present findings of a significant decrease in the content of hepatic adipose, BMI, and HbA1C in patients who received a GLP-1RA in comparison with controls. Therefore, these findings suggest that GLP-1RAs are effective in improving fatty liver and are closely related to weight loss, but have limited effects. Further studies are needed to understand the role of a GLP-1 RAs in combination with weight loss in the treatment of T2DM and NAFLD.

Our study showed that GLP-1RAs significantly decreased SAT and VAT content compared with the control, which is consistent with the results of previous studies. Likewise, this research suggests that GLP-1RAs are able to achieve significant decreases in ALT, AST, weight, waist circumference, fasting blood glucose, HbA_1c, HoMA-IR, triglycerides and total cholesterol, which is consistent with previous RCTs (63–67). We further demonstrated the efficacy of GLP-1 RAs in improving liver enzymes, anthropometric variables, and some metabolic indices. However, in this study we did not find statistically significant difference in GGT or PBG levels when compared with controls, which differs from previous studies (63, 66, 67). However, these inconsistent findings may reflect differences in the baseline characteristics of the study populations and follow-up duration between studies. Further studies are needed to clarify the effects of GLP1 RAs on GGT and postprandial blood sugar levels.

We know that NASH is the most severe phase of NAFLD, and the primary objective of treatment for NASH is to prevent the development of cirrhosis. It is surprising that GLP-1RAs had a resolution effect on NASH, besides reducing IHA, liver enzymes and improving metabolic index. In a small phase II clinical trial of 52 patients with biopsy-proven NASH, patients receiving subcutaneous injections of liraglutide (1.8 mg per day for 48 weeks) achieved resolution of NASH than those receiving placebo (36). Although this finding was impressive, liraglutide may not be convincing enough to study histological improvements in small samples. Subsequently, in a phase II trial, 320 patients with biopsy-proven NASH were randomized to receive daily subcutaneous semaglutide 0.4/2.4mg or placebo for 72 /48weeks, found that subcutaneous injection of 0.4/2.4mg of semaglutide had a resolution effect on biopsy-confirmed NASH, but no effect was found on fibrosis reversal (68) (NCT02970942↗ /NCT03987451↗). Based on those results, semaglutide is expected to prove its benefit as a treatment for NASH through additional phase 3 clinical trials. Recently study suggested that semaglutide versus placebo reduced liver steatosis but not liver stiffness in subjects with non-alcoholic fatty liver disease assessed by MRI (69), which results of this study suggest that Semaglutide has no benefit in the improvement of fibrosis in NAFLD patients, which is consistent with previous studies. Collectively, significant improvement in biopsy‐confirmed liver histology with GLP‐1RAs treatment provides the most substantial evidence for the efficacy of GLP‐1RAs in the management of NASH, although the role of GLP‐1 RAs is still needed to be validated in large sample controlled trials with long‐term follow‐up.

This research has two important strengths. First, it is the first to evaluate the effect of GLP-1 RAs on IHA in patients with T2DM and NAFLD. Second, it provides further confirmation that GLP-1 RAs have a positive effect in these patients and improve IHA, liver enzymes, anthropometric variables, and metabolic indices.

However, the study also has some limitations that should be acknowledged. The first and foremost is the limited number of eligible studies and the small sample size of each individual study. Second, we integrated different RCTs that showed wide clinical heterogeneity in terms of duration, type of medication, drug dosages, and choice of the control group, we used a random-effects model and sensitivity analysis where possible to reduce the influence of these factors. Third, due to the invasive nature of liver biopsy, none of the included studies reported on liver biopsy as a diagnostic tool but relied on MRI to diagnose NAFLD, which is less accurate in demonstrating improvements in NASH.

In conclusion, treatment with an GLP-1RA improves IHA, SAT, VAT, inflammation markers, body composition, some glycemic indices and serum lipids to some extent. GLP-1RAs could be considered as a potential treatment strategy for patients with T2DM and NAFLD, if there are no contraindications. At present, there are few large prospective studies on GLP-1RAs in these patients. Further studies of this type are needed to understand the direct and indirect effects of GLP-1 RAs on the pathogenesis and prognosis of T2DM and NAFLD.

The original contributions presented in the study are included in the article/. Further inquiries can be directed to the corresponding author. 1

Conception and design of the study: SL. Screening of databases, extraction and analysis of the study data: YZ, JX, and DZ. Drafting of the initial manuscript: YZ and JX. Verification of the study methodology: YS, DZ, SC, and ZW. Revision of the manuscript: SL, YZ, and JX. All authors contributed to the manuscript and approved the submitted version.

This study was supported by Sichuan Science and Technology Program (No.2017SYZF0002 and No.2021YFH0168).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fendo.2021.769069/full#supplementary-material↗