Introduction

Obesity is one of the biggest health challenges, affecting millions of individuals worldwide. It is associated with decreased quality of life and disabling complications, shortening lives, and costing billions in healthcare costs and loss of workability [1, 2]. Since 2006, glucagon-like peptide-1 receptor (GLP-1R) agonists have caused a dramatic change in the treatment of type 2 diabetes, with clinically relevant and sustained effects on glycaemic control and body weight combined with cardioprotective mechanisms [3, 4]. The body-weight lowering effect of GLP-1R agonists has subsequently been utilised for the treatment of obesity. Until recently, no therapeutic potential was associated with the other incretin hormone, glucose-dependent insulinotropic polypeptide (GIP), as preclinical studies have suggested potential obesogenic effects. However, the dual GIP receptor (GIPR)/GLP-1R agonist, tirzepatide, has demonstrated superior efficacy in reducing HbA1c and body weight in people with type 2 diabetes [5]. The emergence of GIPR/GLP-1R co-agonists has fostered a growing interest in the actions of GIP and glucagon-like peptide-1 (GLP-1) in metabolically relevant tissues, including liver, muscle and adipose tissue, regarding the control of glucose and lipid homeostasis. Interestingly, the discovery of poly-agonist drugs that activate multiple gut–brain pathways promises to further transform the management of metabolic diseases besides type 2 diabetes, including obesity and non-alcoholic fatty liver disease (NAFLD) [6,7,8]. With NAFLD affecting 25–30% of the global population, up to 60–70% of people with type 2 diabetes and almost all individuals with severe obesity [9, 10], an available pharmacotherapy for NAFLD is highly warranted.

This review highlights established and emerging concepts regarding the use of incretins for controlling body weight and treating obesity and, potentially, NAFLD.

Incretin hormones in the regulation of body weight

The knowledge of the complex mechanisms controlling body weight is evolving rapidly. However, the understanding of hunger, satiety and weight control is still mainly based on findings from preclinical studies with rather limited translational value in humans. In addition, limited data are available on the pathophysiological neuroendocrine regulation of satiety in the development of human obesity.

GIP was the first incretin hormone to be discovered; however, the understanding of GIP's biology was overshadowed by research focusing on GLP-1 and its impact in diabetes and obesity treatment [4, 11,12,13,14]. Today it is recognised that the hypothalamus is a key region of the brain that is implicated in homeostatic regulation and is suggested to be an integral centre for the control of feeding behaviour [15, 16]. It has also been shown that neuronal tissue has a dense expression of both the GIPR and the GLP-1R [17]. Interestingly, exogenously administered GIP and GLP-1 seem to access the brain predominately via leaks in the blood–brain barrier, where the underlying neuronal tissue has dense expression of their receptors [11,12,13, 17] (Fig. 1). However, early work indicated that GIP was obesogenic (Fig. 1), limiting interest in developing GIPR agonists to treat type 2 diabetes [18]. However, recent GIP research has reinvigorated interest in this peptide, and paradoxical observations with different approaches have been discussed for treating obesity, one promoting GIPR agonism and the other GIPR antagonism [19]. Different hypotheses are considered, including a compensatory relationship between incretin receptors by which GIPR enhances GLP-1R activity and/or that chronic GIPR agonism produces desensitisation and loss of GIPR activity, thus, mimicking antagonism. However, a more profound understanding of GIP biology in relation to satiety and body weight remains to be fully elucidated [19].

Fig. 1
figure 1

Simplified summary of the biological actions of GIP and GLP-1 as pharmacological targets in obesity and NAFLD. In the liver, GIP and GLP-1 have indirect effects that may be generated by changes in plasma insulin and glucagon concentrations (depicted by the image of the pancreas) and neural regulation signals (depicted by the image of the brain). Solid lines, direct effects; dashed lines, indirect effects. TG, triglycerides. This figure is available as part of a downloadable slideset

Full size image

In contrast to GIP, a substantial number of preclinical and clinical studies link GLP-1 as a central peptide that, through hormonal and neural pathways, not only regulates pancreas function but also hunger and satiety (Fig. 1). The importance of brain-derived GLP-1, the direct vs indirect actions of GLP-1 and GLP-1-induced control of neural activity is extensively discussed since many organs and cellular targets of GLP-1 action do not exhibit detectable levels of GLP-1R [20,21,22,23]. Studies using exogenous GLP-1 infusion in humans demonstrated reduced energy intake, reduced appetite and decreased brain responses to food anticipation and consumption without direct changes in energy expenditure, and data strongly suggest that body-weight loss ensuing from GLP-1R agonist administration in humans largely reflects reductions in food intake [24, 25]. Whereas the relative contribution of brain-derived GLP-1, which is synthesised in the nucleus of the solitary tract, on the reduction in food intake or body weight remains to be elucidated [16].

GLP-1R agonists were initially developed for the treatment of type 2 diabetes. However, due to their efficacy in inducing body-weight loss, extensive evidence with high doses of GLP-1R agonists have demonstrated clinically relevant and sustained effects on body weight in obesity [26]. Recently, through the engagement of their respective receptors, studies have shown that GIP and GLP-1 have a synergistic effect on appetite and body weight when the peptides are combined in pharmacology [6, 8], but the exact mechanisms underlying the enhanced weight loss exhibited by GIPR/GLP-1R co-agonism are still to be elucidated.

Incretin-based therapies for treatment of obesity

The concept of incretin-based treatment of overweight and obesity was driven by the effect of native GLP-1 and GLP-1R agonists on appetite and satiety in combination with an observed loss in body weight in trials investigating liraglutide for the treatment of type 2 diabetes [27]. The first incretin-based treatment regimen to be approved for the management of overweight and obesity was liraglutide 3.0 mg once-daily, which was followed by semaglutide 2.4 mg once weekly [16]. Recently, the dual GIPR/GLP-1R co-agonist, tirzepatide, has demonstrated a marked effect on body weight in individuals with overweight and obesity, and Phase III trials of this drug are expected to be completed in 2023 [28].

Body-weight reductions

The efficacy and safety of liraglutide for the treatment of overweight and obesity was investigated in the Satiety and Clinical Adiposity – Liraglutide Evidence (SCALE) programme [29,30,31,32,33]. The trials included individuals with overweight (BMI ≥27 kg/m2) and ≥1 weight-related complication, including dyslipidaemia, hypertension, sleep apnoea, impaired glucose tolerance and/or impaired fasting glucose and type 2 diabetes, or individuals with obesity (BMI ≥30 kg/m2) with or without complications. When combined with lifestyle changes and not preceded by body-weight loss due to low-energy diet, daily administration of 3.0 mg of liraglutide provided an additional body-weight loss of 4.0–5.4% compared with placebo [30,31,32] (Fig. 2), which was sustained after 3 years of treatment [33]. Although providing clinically meaningful body-weight loss, the efficacy of once-daily liraglutide has been markedly exceeded by 2.4 mg of semaglutide once weekly [34]. The efficacy and safety of weekly semaglutide (2.4 mg) for the treatment of overweight and obesity was investigated in the Semaglutide Treatment Effect in People with Obesity (STEP) programme, which included individuals with overweight (BMI ≥27 kg/m2) and complications, including dyslipidaemia, hypertension, sleep apnoea, cardiovascular disease and type 2 diabetes, or individuals with obesity (BMI ≥30 kg/m2) with or without complications [26, 34,35,36,37,38,39]. Compared with placebo, the smallest percentage body-weight loss was observed in individuals with type 2 diabetes, with an estimated treatment difference of 6.2% after 68 weeks of treatment [35], whereas an additional body-weight loss of 10.3–12.5% after 68 weeks of treatment (without prior run-in with semaglutide) was observed in individuals without diabetes in trials aiming to directly compare semaglutide with placebo [26, 36, 38, 39] (Fig. 3). In a head-to-head trial, mean body-weight loss was 15.8% with semaglutide compared with 6.4% in the liraglutide group [34] (Fig. 3). A continuous decline in body weight for the first 60 weeks of treatment with semaglutide has been demonstrated, followed by a stable body weight for up to 104 weeks with continued treatment [38]. Importantly, discontinuation of semaglutide after 20 weeks of treatment led to regain of body weight, whereas maintaining treatment led to continued body-weight loss over an additional 48 weeks of treatment, underlining that life-long treatment is necessary to maintain treatment effect [37]. Recently, semaglutide (2.4 mg once weekly) has also been investigated for the treatment of childhood obesity in children aged 12 years to <18 years with overweight (BMI ≥85th percentile) and ≥1 weight-related complication, or with obesity (BMI ≥95th percentile) [40]. In children receiving lifestyle intervention, semaglutide resulted in an additional 16.7% body-weight loss compared with placebo.

Fig. 2
figure 2

Overview of body-weight loss in the Satiety and Clinical Adiposity – Liraglutide Evidence (SCALE) trials. In all trials, participants were administered 3.0 mg of liraglutide daily or placebo. The SCALE trials included individuals with overweight (BMI ≥27 kg/m2) and ≥1 weight-related complication, including dyslipidaemia, hypertension, sleep apnoea, impaired glucose tolerance, impaired fasting glucose and type 2 diabetes, or individuals with obesity (BMI ≥30 kg/m2) with or without complications. The SCALE Obesity and Prediabetes trial included individuals with obesity/overweight without type 2 diabetes (n=3652), SCALE Diabetes participants had obesity/overweight with type 2 diabetes (n=846), SCALE Sleep Apnoea participants had obesity and sleep apnoea without type 2 diabetes (n=359), and SCALE Maintenance involved individuals who had obesity/overweight without type 2 diabetes and were randomised after achieving a ≥5% body-weight loss after following a low-energy diet for 12 weeks (n=382). Data are reported as observed means for change in body weight and proportions for individuals with body-weight loss >10%, using the last observation carried forward, except for SCALE Diabetes for which data are estimated means using multiple imputations. LED, low-energy diet; T2D, type 2 diabetes. This figure is available as part of a downloadable slideset

Full size image
Fig. 3
figure 3

Overview of body-weight loss in the STEP trials. The STEP programme included individuals with overweight (BMI ≥27 kg/m2) and complications, including dyslipidaemia, hypertension, sleep apnoea, cardiovascular disease and type 2 diabetes, or individuals with obesity (BMI ≥30 kg/m2) with or without complications. Participants were administered 2.4 mg of semaglutide once weekly or placebo and were observed for 68 weeks in all trials apart from the STEP 5 trial, in which they were observed for 104 weeks. The STEP 1 trial included individuals with obesity/overweight without type 2 diabetes (n=1961), STEP 2 participants had obesity/overweight with type 2 diabetes (n=1210), STEP 3 participants had obesity/overweight without type 2 diabetes and received semaglutide/placebo as an adjunct to intensive behavioural therapy (n=611), STEP 4 participants had obesity/overweight without type 2 diabetes and were observed after taking semaglutide for 20 weeks (n=803), STEP 5 participants had obesity/overweight without type 2 diabetes (n=304), STEP 6 participants had obesity/overweight with or without type 2 diabetes (n=401), and the STEP 8 trial included individuals who had obesity/overweight without type 2 diabetes who were either allocated (3:1:3:1) to receive semaglutide or matching placebo, or 3.0 mg of liraglutide daily or matching placebo (n=338). Data are reported as observed means for change in body weight and proportions for individuals with body-weight loss >10%, and represent the treatment policy estimands (intention-to-treat analysis). IBT, intensive behavioural therapy; T2D, type 2 diabetes. This figure is available as part of a downloadable slideset

Full size image

Safety

GLP1-R agonists have been utilised in clinical practice for decades and long-term safety is well described. Though adverse effects of the emerging co-agonists have been extensively studied in developmental programmes, potential long-term adverse effects cannot be excluded for obvious reasons. Adverse events in relation to GLP-1R agonists for the treatment of obesity are mainly gastrointestinal side effects, including nausea, diarrhoea, vomiting and constipation, which is consistent with observations in type 2 diabetes, and rates of discontinuation due to adverse events are generally higher with GLP-1R agonists treatment compared with placebo [26, 30,31,32,33,34,35,36,37,38,39]. Furthermore, there is an increased risk of gallbladder-related disorders, which is consistent with the known association between rapid weight loss and gallbladder-related disorders [41, 42]. The STEP 8 trial demonstrated similar rates of adverse events for semaglutide and liraglutide, but with less discontinuation due to adverse events with semaglutide [33]; however, it should be noted that this study had a relatively small number of participants (n=338) [34].

Future incretin-based therapies

Future research in incretin-based treatment of overweight and obesity aims to investigate the potential of combining GLP-1-R agonism with the targeting of other peptide hormones to achieve synergistic effects, with the most prominent targets of interest being the GIPR, the glucagon receptor (GCGR) and the amylin receptor [16]. GIPR/GLP-1R co-agonism has already proved efficient in clinical trials in obesity. Results from the first Phase III trial in obesity and overweight investigating the GIPR/GLP-1R co-agonist tirzepatide, which is currently approved for the treatment of type 2 diabetes and has resulted in marked body-weight reductions in type 2 diabetes [5, 43,44,45,46], were published in 2022 [28]. In this trial, three doses of tirzepatide (5 mg, 10 mg and 15 mg once weekly) were compared with placebo in individuals with overweight (BMI ≥27 kg/m2) and complications, or with obesity (BMI ≥30 kg/m2). Individuals with diabetes were excluded. Compared with placebo, tirzepatide provided an additive body-weight loss of 11.9%, 16.4% and 17.8% with an increasing dose of 5 mg, 10 mg and 15 mg, respectively, and 83.5% of participants in the 15 mg group had ≥10% body-weight loss [28]. Recently, after 72 weeks of treatment, adults with obesity or overweight and type 2 diabetes had a mean weight loss of 12.8% and 14.7% with 10 mg and 15 mg tirzepatide, respectively, vs 3.2% with placebo [47]. In the group treated with 15 mg of tirzepatide, 48% had ≥15% body-weight loss. Tirzepatide has not been directly compared with semaglutide 2.4 mg in individuals with overweight or obesity; however, when compared with semaglutide 1.0 mg in type 2 diabetes, in doses of 5 mg, 10 mg and 15 mg, tirzepatide resulted in significantly greater body-weight reductions [5]. As alluded to above, the rationale for combining GLP-1R agonism with either GIPR agonism or antagonism is, however, still debated, and how the GIPR agonistic properties of tirzepatide contribute to body-weight loss remains unclear [19]. It is noteworthy that the combined GIPR antagonist and GLP-1R agonist AMG 133 demonstrated a body-weight loss of 14.5% at the highest dose (420 mg every 4 weeks) after 12 weeks in a Phase I trial, and this compound will now enter Phase II trials [48]. Hence, research is needed to elucidate the underlying mechanisms of GIPR as a therapeutic target in obesity.

Currently, several co- and tri-agonists targeting the GLP-1R and the GIPR in combination with the GCGR are under investigation for metabolic conditions including obesity. Previous studies have demonstrated a synergistic effect of co-infusion of GLP-1 and glucagon on food intake, and GCGR agonism may provide an additive effect due to increased energy expenditure [49, 50]. However, balancing the body-weight lowering effect of GCGR agonism with its hyperglycaemic effect is a challenge. Hence, combining GCGR agonism with the glucose-lowering effect of GLP-1R and GIPR agonism is a necessity. The GIPR/GLP-1R/GCGR tri-agonist LY3437943 has been compared with dulaglutide 1.5 mg and placebo in a Phase I study in individuals with type 2 diabetes [51]. In this study, dose-escalation of LY3437943 to 12 mg once weekly over a period of 12 weeks resulted in an additional body-weight loss of 9 kg compared with both the placebo and dulaglutide, while demonstrating an acceptable safety profile.

Another approach has been to combine the long-acting acylated amylin analogue cagrilintide with semaglutide. Cagrilintide as a monotherapy has demonstrated an ability to induce significant body-weight reductions compared with placebo, with synergistic effects when added to semaglutide [52, 53]. In a study comparing different once-weekly doses of cagrilintide in combination with 2.4 mg of semaglutide with semaglutide (2.4 mg) monotherapy in individuals with overweight or obesity, the combination of cagrilintide 2.4 mg and semaglutide 2.4 mg resulted in a 17.1% body-weight loss compared with a 9.8% body-weight loss with semaglutide as a monotherapy [53].

Incretins: a potential therapy in NAFLD

Based on the impressive effect of GLP-1R agonists on body weight and also the ability of these drugs to improve liver enzymes [54], incretin-based therapy for the treatment of NAFLD is currently being explored. NAFLD is a group of metabolism-related liver conditions that are characterised by ectopic hepatic lipid accumulation not attributed to excessive alcohol consumption [55]. Increased fat accumulation can lead to inflammation (non-alcoholic steatohepatitis [NASH]) and liver fibrosis, and can further advance to NASH-related cirrhosis, which is the fastest-growing indication for liver transplantation in western countries [10]. The incidence of NAFLD is rapidly increasing worldwide, concurrently with the epidemics of obesity and type 2 diabetes [56]. The exact cause of NAFLD is unknown. The complex pathophysiology involves multiple features, such as metabolic disturbances, lipotoxicity, insulin resistance, chronic inflammation, fibrosis, intestinal function and the gut microbiome [57], and is closely associated with obesity, type 2 diabetes and the metabolic syndrome [58]. NASH has a strong genetic component, which may be amplified by co-existing obesity [59, 60]. Currently, no pharmacotherapies are available for the treatment of NAFLD/NASH and the disease is managed by targeting lifestyle changes and treating cardiometabolic risk factors [61]. Thus, novel treatment options for NAFLD/NASH are highly warranted. Incretin-based therapy may target the underlying metabolic and hormonal pathways that are thought to be involved in developing NAFLD. Incretin-based therapy may also have a role in the treatment of NASH, especially in those carrying a high polygenic risk for this condition.

The independent and combined hepatic effects of GIP and GLP-1 have been explored in preclinical studies and potential beneficial effects on NAFLD/NASH have been demonstrated [6, 62,63,64,65,66]. GIP and GLP-1 have indirect hepatic effects that are thought to be generated by changes in portal and peripheral plasma insulin and glucagon concentrations and may partly be related to neural regulation signals (Fig. 1). In preclinical studies, GLP-1R agonists have been demonstrated to improve hepatocyte mitochondrial function and hepatic insulin sensitivity [67], and reduce adipose tissue lipotoxicity both due to body-weight loss and mechanisms independent of body-weight reductions [68]. An overview of the randomised controlled trials conducted in humans to evaluate the potential of GLP-1R agonists for treating NAFLD/NASH is shown in Table 1. The largest and longest trial to date includes 320 individuals with biopsy-confirmed NASH and fibrosis randomised to treatment with once-daily semaglutide (0.1 mg, 0.2 mg or 0.4 mg) vs placebo [69]. Semaglutide dose-dependently reduced liver enzymes, and resulted in greater NASH resolution (reduced ballooning and lobular inflammation) without worsening of fibrosis, as compared with placebo after 72 weeks of treatment. Semaglutide 0.4 mg once-daily significantly improved NASH compared with placebo (NASH improved in 59% vs 17% of participants). In contrast, the difference in liver fibrosis resolution stage did not reach statistical significance (liver fibrosis staging improved in 43% of participants in the semaglutide arm vs 33% in the placebo arm), potentially owing to a lack of power within the study or the relatively short follow-up time of 72 weeks. Notably, there was a relatively high proportion of individuals with fibrosis resolution in the placebo arm (observed in 33% of participants). Liver fibrosis staging (NASH clinical research network [CRN] score) depends on a complex classification that is based on several subjective histological evaluations that have low sensitivity [70, 71]. It has therefore been considered suboptimal when evaluating treatment effects in clinical trials [71, 72]. Currently, the effects of 2.4 mg semaglutide once weekly are being explored in 1200 participants with biopsy-confirmed NASH in a Phase III clinical trial with a combined primary endpoint of resolution of NASH without worsening fibrosis, improvement in liver fibrosis without worsening of steatohepatitis, and time to first liver-related clinical event (NCT04822181) (see Table 2). A recent meta-analysis based on available data from placebo-controlled randomised trials confirms that GLP-1R agonists are associated with reduced liver fat content, improved liver enzyme levels, and greater histological resolution of NASH without worsening of liver fibrosis [73].

Table 1 An up-to-date overview of randomised controlled trials with GLP-1R agonists for the treatment of NAFLD/NASH
Full size table
Table 2 An overview of ongoing clinical trials with GLP-1R agonist and incretin combination therapies for the treatment of NAFLD/NASH
Full size table

While GIP is less-well studied, it is suggested to have an essential role in lipid metabolism, to enhance lipid deposition in the liver and to induce inflammatory changes within the liver [74, 75]. GIP appears to have both beneficial and potentially deleterious effects. The effects of GIP appear to be beneficial when combined with GLP-1 or when GIP is administrated in pharmacological doses [76]. Fifty-two weeks of treatment with the once-weekly dual GIPR/GLP-1R agonist tirzepatide has been demonstrated to reduce liver fat content and improve liver enzyme levels compared with insulin degludec; however, histological changes and liver fibrosis stage were not assessed in this study [77]. Once-weekly tirzepatide (5 mg, 10 mg and 15 mg) is currently being explored in 196 participants with overweight or obesity and biopsy-confirmed NASH in a Phase II clinical trial (NCT04166773; see Table 2).

GLP-1R agonists may be ideally suited for combination therapy in NAFLD/NASH, as most of their mechanisms do not directly target the liver. Thus, poly-agonist drugs may complement classical incretin actions on NASH pathophysiology, providing a promising therapeutic approach. Dual (GCGR/GLP-1R and GIPR/GLP-1R) and triple (GCGR/GIPR/GLP-1R) agonists have led to improvements in lipid metabolism and hepatic steatosis [6, 64,65,66]. Moreover, studies suggest that dual and triple receptor agonists may have additional effects on histological NASH features vs high-dose GLP-1R agonists alone, and that these may be independent of changes in body weight [78]. In addition, combining GLP-1R agonists with other peptide agonists or small-molecule therapeutics, including fibroblast growth factor 21 (FGF21), the farnesoid X receptor (FXR) agonist cilofexor with/without the acetyl coenzyme A carboxylase inhibitor firsocostat, a long-acting amylin agonist, cannabinoid receptor antagonists and sodium-glucose cotransporter 2 inhibitors (SGLT-2i), is currently being explored and may lead to additional improvements in NASH features vs GLP-1R agonists therapy alone [79,80,81,82,83]. An overview of ongoing clinical trials with GLP-1R agonist and incretin combination therapy is provided in Table 2. It is unclear how much of the hepatic improvements observed with use of GLP-1R agonists and GIPR/GLP-1R agonists are driven by a reduction in body weight. Larger Phase III trials with liver-associated histological endpoints are needed to establish the therapeutic role of incretins for NAFLD/NASH.

Conclusions and perspectives

Body-weight homeostasis relies on complex mechanisms and the development of obesity occurs on a background of genetic susceptibility and an environment promoting reduced physical activity and increased energy intake. The pathophysiology of common obesity links neuroendocrine and metabolic disturbances with behavioural changes, genetics, epigenetics and cultural habits. Diet and lifestyle modification are necessary but not sufficient for sustained body-weight loss in the majority of individuals with obesity [84]. As of today, no pharmacotherapies are available for the treatment of NASH and the disease is managed by lifestyle changes targeting weight loss [85]; however, most people with NASH cannot lose weight through lifestyle interventions or sustain the weight loss achieved [86]. The therapeutic potential of GLP-1R agonists alone or in combination with other peptide agonists or other small-molecule therapeutics for the treatment of overweight/obesity and NASH is under evaluation, with a specific focus on efficacy and safety. Ongoing trials in people with obesity will further clarify the safety of mono, dual and triple receptor agonists, and pivotal studies are underway in individuals with fatty liver disease [87] and cardiovascular disease [88] that will define whether these peptides represent effective and safe therapies for people suffering from diseases beyond diabetes. So far, the therapeutic potential of combination therapies seems to be huge with predictable side effects and no major safety concerns. Moreover, based on the reduction in major cardiovascular events observed with use of GLP-1R agonists in cardiovascular outcome trials in people with type 2 diabetes [89], large-scale outcome trials in overweight and obesity have been initiated (ClinicalTrials.gov registration no. NCT03574597 and NCT05556512). These trials will investigate the ability of semaglutide and tirzepatide to reduce cardiovascular disease and mortality, potentially expanding evidence of clinical benefit of these therapies and establishing a benchmark for future pharmacotherapies in overweight and obesity. As the development of additional gut-hormone poly-agonists progresses, pharmacotherapies used for the treatment of obesity may become as efficient as bariatric surgery for reducing weight. These developments have the potential to reshape the treatment landscape for individuals with metabolic disorders, offering new hope for effective management and improved quality of life.