Ipragliflozin and sodium glucose transporter 2 inhibitors to reduce liver fat: will the prize we sought be won?

KEYWORDS:

• Ipragliflozin
• NAFLD
• Sodium glucose co-transporter 2 inhibitors
• type 2 diabetes

1. Introduction

Non-alcoholic fatty liver disease (NAFLD) is a general term encompassing conditions from increased intrahepatic lipid content (IHLC) (steatosis with triglyceride deposition >5% of hepatocytes) to non-alcoholic steatohepatitis (NASH) with inflammation and/or liver fibrosis [1]. This medical condition has very recently acquired special attention, given the following associations of NAFLD with (1) type 2 diabetes mellitus (T2DM); (2) cardiovascular disease (CVD); and (3) hepatocellular carcinoma (HCC) [2,3]. Indeed, it is speculated that the prevalence of NAFLD approaches 70% in T2DM patients over 50 years old [3]. T2DM patients carry an even higher risk of NASH, malignancy, and CVD, whereas the allegedly benign isolated steatosis is nowadays considered to predict disease progression [1].

In the light of these observations, implementation of specific screening and treatment algorithms for T2DM patients with NAFLD seems imperative [1]. Deplorably, we still lack licensed pharmacological agents for NAFLD treatment [4]. Therefore, drugs available for T2DM therapy have been tested in NAFLD. Among them, recent randomized controlled trials with metformin have yielded negative results [1]. Conversely, 51% of prediabetes (n = 49) or T2DM patients (n = 52) with biopsy-proven NASH who were administered pioglitazone (45 mg/day) additionally to hypocaloric diet for 36 months achieved NASH resolution with two-thirds of them accomplishing reduction in the NAFLD activity score (both p < 0.001) [5], whereas, in a very recent meta-analysis, pioglitazone was associated with improved advanced fibrosis, fibrosis of any stage, and NASH resolution among patients with NASH, with or without diabetes [6]. Although we currently cannot be sure whether the aforementioned effects change the natural history, experts now suggest that pioglitazone alone or in combination with statins or ezetimibe merits special consideration in NAFLD/NASH patients at high CVD or HCC risk [4].

Additional agents have also been tested in NAFLD [7–18]. In a double-blinded, randomized, phase 2 trial [7], liraglutide (1.8 mg/day) or placebo was administered for 48 weeks to 52 obese insulin-resistant NASH patients (1/3 had T2DM). Definite NASH histological resolution was observed in 39% of liraglutide-treated patients versus 9% with placebo. Liraglutide achieved significant body weight reduction but failed to improve fibrosis [7].

Finally, a number of other therapies have biopsy-proven effects on NASH [1,8,9]. Among them, vitamin E (in a daily dose of 800 IU) has been associated among non-diabetic adults with amelioration of hepatic cell injury (NASH improvement vs. placebo was: 43% vs. 19%, p = 0.001; number needed to treat: 4.2) but only borderline significance was attained in NASH resolution (36% vs. 21%, p  =  0.05 vs. placebo) [1,9]. Moreover, a number of studies have suggested increased risk of prostate cancer with long-term exposure [1]. Likewise, bariatric surgery seems promising in the improvement of biochemical and histological features of NAFLD [8], but long-term trials are needed to clarify its role [1]. Of note, exogenous administration of ursodeoxycholic acid, a non-toxic bile acid led to significant reduction of aminotransferases, but this did not always correlate with histological improvement [9].

2. The role of sodium glucose co-transporter 2 inhibitors

Sodium glucose co-transporter 2 inhibitors (SGLT2is) are modern antidiabetic agents which induce glycosuria by blockade of proximal renal tubule glucose reabsorption [10,11]. These agents manifest pleiotropic beneficial actions: attenuation of hyperglycemia, hypertension, and albuminuria and reduction of body weight and serum uric acid along with nephroprotection attributed among others to reduction of hyperfiltration and intraglomerular pressure [19,20]. With respect to NAFLD, tofogliflozin administered (for 3–5 weeks) to obese diabetic mice was associated with decrease of liver weight and IHLC and with improvement of hepatic steatosis along with attenuation of body weight gain (body weight of control group increased by 8.6 ± 0.5 g over 5 weeks) [12]. Four-week treatment with remogliflozin etabonate in a mouse NAFLD model was also reported to reduce plasma aminotransferases, liver weight, and IHLC, but without change in body weight gain [13]. Similarly, in another study, luseogliflozin decreased liver weight and serum alanine aminotransferase levels in NASH mice, but again body weight was not affected [13]. Unfortunately, this reduction in alanine aminotransferase can only be considered as a surrogate marker of NASH and, hence, far too indirect to state that this is an effective therapy.

Among agents approved by the US Food and Drug administration (FDA), empagliflozin alone or combined with linagliptin (for three weeks) was associated with anti-inflammatory, anti-steatotic and anti-fibrotic effects in a NASH T2DM mouse model, but none of the treatments achieved significant body weight reduction [14]. Importantly, reduced expression of genes related to de novo fatty acid synthesis (lipogenesis) was reported [14]. In a randomized, placebo-controlled trial, dapagliflozin (10 mg/day) was administered for 24 weeks to T2DM patients (mean body mass index 31.9 kg/m² and body weight 91.5 kg) inadequately controlled on metformin [15]. Results from 67 participants (30 on dapagliflozin and 37 on placebo) suggested weight loss (−2.96 kg vs. −0.88 kg, p < 0.0001), decreased visceral fat volume (VFV) (−9.2% vs. −1.4%, p  =  0.0084), and subcutaneous fat volume (SFV) (−6.7% vs. −2.6%, p  =  0.0385), but there was no change in IHLC (−2.4% vs. −1.5%, p  =  0.4499). Finally, a very recent study in Japanese T2DM patients having high levels of alanine aminotransferase (>30 U/l) [16] reported that canagliflozin (100 mg once daily) was linked with significant reduction of alanine aminotransferase levels at 12 weeks of treatment in comparison with placebo (−10.3 ± 11.7 vs. −3.2 ± 17.6 U/L, p = 0.0206). This effect persisted through 52 weeks of treatment with a comparable incidence of adverse events between these two treatment groups [16]. However, in multiple regression analysis, the contribution of body weight for the change in alanine aminotransferase was modest.

Ipragliflozin is another SGLT2i which gained first global approval in January 2014 for marketing in Japan for T2DM treatment in doses of 25 or 50 mg once daily [10]. Ipragliflozin administered (for 4 weeks) to T2DM NAFLD mice was associated with reduction of plasma aminotransferases and body weight [10]. Administration for 8 weeks was associated with reduction of serum free fatty acids (FFA), IHLC, and areas of fibrosis, but not with body weight reduction [11].

More recently, Ohta et al. [17] have investigated the effect of 24-week therapy with this agent (50 mg per day) in 20 T2DM Japanese patients (with mean body mass index: 29.7 ± 3.2kg/m², body weight: 82.2 ± 11.3 kg, HbA_1c: 8.2 ± 1.3%, VFV: 6303.9 ± 1907.3 cm³, SFV 7904.2 ± 2097 cm³). Primary outcome measures included change in VFV, abdominal SFV (assessed by whole abdominal computed tomography scanning), and IHLC (measured by proton magnetic resonance spectroscopy) after 12 and 24 weeks of treatment. Secondary end points included change of body fat and lean mass (assessed by dual X-ray absorptiometry) [17]. The authors reported that 12-week ipragliflozin administration resulted in reduction of body weight, body mass index, HbA_1c, aminotransferases, fasting plasma glucose (FPG), and homeostasis model assessment of insulin resistance (HOMA-IR). However, only the first four aforementioned parameters decreased further at 24 weeks of therapy, whilst at that time point, a marginal increase of 3-hydroxybutyric acid was observed [17]. Further reductions were reported for fat mass, limbs lean mass, SFV, VFV, and IHLC at 12 and 24 weeks (p < 0.05 for all parameters at weeks 12 and 24 in comparison with baseline), but only VFV was further significantly reduced at week 24 as compared with reduction described at week 12 [17].

3. Conclusion

Ipragliflozin has very recently demonstrated efficacy in reduction of visceral and hepatic fat [17]. This is very important, granted the need of NAFLD treatment in the population of T2DM patients. Although a number of important issues remain to be clarified, the results of this new trial add to the potential of SGLT2i in NAFLD resolution.

4. Expert opinion

Normally, there is a precise balance between de novo lipogenesis and FFA uptake from plasma versus triglyceride disposal by fatty acid oxidation or export of triglycerides in very-low-density lipoproteins. This results in low and steady triglyceride content in the liver. Misbalance of the aforementioned mechanisms [14,16], enhanced expression of pro-inflammatory adipokines, and diminished adiponectin levels [4,14,16] trigger NAFLD. Insulin resistance plays a pivotal role by increasing FFA uptake [4,14].

Pioglitazone is already currently recommended for adult patients (aged ≥21 years) with biopsy-proven NASH with or without diabetes [21], and it is expected to become the first-line agent for NAFLD (even for those patients without biopsy-proven advanced fibrosis), especially in T2DM [1]. Indeed, this agent specifically ameliorates insulin resistance, diverts FFA from the liver to adipose tissue, and enhances secretion of adiponectin by adipocytes. However, adverse events (notably, weight gain, fluid retention, fracture risk) have so far prevented its widespread acceptance [1,4].

On the contrary, ipragliflozin increases 3-hydroxybutyrate levels, which suggests a metabolic shift from carbohydrate to FFA oxidation in the liver (as a compensatory mechanism for glucose loss) [13] and demonstrates efficacy in attenuation of gene expression related to de novo lipogenesis in the liver [13]. This concurs with the marked reduction in IHLC (by 42.5 ± 24.3%) observed by Ohta et al. after 24 weeks of treatment [17]. Another study with NASH mice also suggested reduction of insulin resistance [11].

Ohta et al. [17] have attributed the similar percent reduction of VFV and SFV reported in their study mainly to energy loss. However, in their study, appendicular skeletal muscle showed a slight (albeit significant) reduction. Importantly, FPG and HOMA-IR, which reflect hepatic insulin resistance, did not further reduce at 24 weeks of treatment [17]. This may pertain to the fact that hepatic insulin resistance is more severe in diabetic patients with NASH than in matched obese with only isolated steatosis [1]. Notably, the general applicability of their results is further limited by the fact that some studies report greater efficacy of ipragliflozin in Japanese than Western populations [10].

Although the exact mechanism by which ipragliflozin mediates its effects on NAFLD is rather controversial [10,11,13,17], it seems reasonable that SGLT2i merit special consideration as add-on or even alternative to pioglitazone treatment for NAFLD, by virtue of their combined mechanisms of action and the potential of SGLT2i to attenuate any fluid retention pertaining to piogltazone therapy (by promoting salt/water loss) [10]. However, applicability of results in animal models [10–14] and reliability of non-invasive assessment of NAFLD [15–18] is associated with controversy and so, in the coming years, longitudinal human trials with biopsy proven effects are expected to provide additional evidence in this field. In July 2017, a randomized, active-controlled trial including 66 T2DM NAFLD patients [18] has reported that ipragliflozin (50 mg per day) achieved equally beneficial effects on NAFLD and glycemic control in comparison with pioglitazone (15–30 mg per day). The former also exerted a significant effect on body weight and abdominal fat area [18]. In the light of these encouraging results with ipragliflozin [17,18], it appears that the long-sought progress in NAFLD treatment might be won.

Declaration of interest

N Papanas has been an advisory board member of TrigoCare International, Astra-Zeneca, Boehringer Ingelheim, MSD, Novo Nordisk and Pfizer; has participated in sponsored studies by Astra-Zeneca, GSK, Novo Nordisk, Novartis and Sanofi-Aventis; has received honoraria as a speaker for Astra-Zeneca, Boehringer Ingelheim, Eli-Lilly, ELPEN, MSD, Mylan, Novo Nordisk, Pfizer, Sanofi-Aventis and Vianex; and attended conferences sponsored by TrigoCare International, Eli-Lilly, Galenica, Novo Nordisk, Pfizer and Sanofi-Aventis. E Maltezos has participated in sponsored studies by Astra-Zeneca, GSK, Novo Nordisk, Novartis and Sanofi-Aventis; and attended conferences sponsored by Wyeth, Pfizer, and Bayer.

Additional information

Funding

This manuscript has not been funded.