1 INTRODUCTION

The prevalence of non-alcoholic fatty liver disease (NAFLD) is approximately 25% in both Asian and western countries, with an incidence that continues to increase.1, 2 NAFLD is not only one of the most common liver diseases, but also is regarded as a cluster and interacts with other metabolic diseases. In patients with type 2 diabetes (T2D), the prevalence of NAFLD is up to 70%.3-5 Moreover, the presence of T2D accelerates the progression of NAFLD and is a predictor of poor prognosis of NAFLD.5 Conversely, the presence of NAFLD promotes incident T2D and accelerates hyperglycaemia.6 In this context, both control of hyperglycaemia in T2D and fatty burden reduction in NAFLD have prognostic implications.

Dipeptidyl pepdiase-4 (DPP-4) is an enzyme that breaks down glucose-dependent incretin hormones, glucose-dependent insulinotropic polypeptide, and glucagon-like peptide-1. Inhibition of DPP-4 leads to increased insulin secretion and suppressed glucagon release from the pancreas. DPP-4 is distributed highly in the liver, and increased expression of DPP-4 in liver is associated with NAFLD.7 Evogliptin is a novel, selective DPP-4 inhibitor, and its potential for treatment of NAFLD has been shown in recent studies.8-10 In an experimental study, evogliptin improved insulin resistance and reduced hepatic lipid accumulation by inhibiting transcription factors for lipogenesis.8 In terms of glycaemic control, evogliptin has proven clinical efficacy and safety in randomized trials for T2D, both as monotherapy and as dual therapy in combination with metformin.11, 12 Accordingly, evogliptin is expected to show efficacy for treatment of T2D complicated by NAFLD. In this study, we performed a head-to-head comparison of the efficacy and safety of evogliptin versus pioglitazone for treating patients with T2D complicated by NAFLD in decreasing fatty liver burden.

2 MATERIALS AND METHODS 2.1 Study design

This study was a 24-week, double-blind, active-controlled, randomized, parallel, phase IV clinical trial in South Korea (NCT03910361). The trial was conducted at seven sites. The first participant was enrolled in May 2019, and the last patient visit was in July 2020. The trial subjects were Korean patients with T2D, aged 19-70 years, with an HbA1c level of 6.5% to 9.0%, fasting plasma glucose (FPG) less than 270 mg/dl, and a body mass index (BMI) of 20.0 to 40.0 kg/m2. Patients were eligible for the study if they had been receiving metformin monotherapy for at least 8 weeks and were naive to treatment with any antidiabetic medication other than metformin. In addition, the alanine aminotransferase (ALT) level was required to be above normal but lower than three times the upper normal limit. All patients exhibited NAFLD, as diagnosed by abdominal ultrasound and confirmed by radiology specialists within 6 months. This study complied with the Declaration of Helsinki and good clinical practice guidelines. All the study subjects provided informed consent prior to inclusion. The exclusion criteria for the current study are described in Appendix S1.

Eligible patients were randomly assigned in a 1:1 ratio using a computer-generated randomization sequence to either receive 5 mg/day of evogliptin or 15 mg/day of pioglitazone. No dosage adjustments were made to the study medications in either study arm during the study period. All patients received diet and exercise counselling at the beginning of the study and were reminded at each study visit to follow their recommended plan. Specifically, the dietary plan recommended that women, or those who were inactive, had 1400 to 1600 kcal/day (otherwise 1800 to 2000 kcal/day), and all the patients were asked to reduce high amounts of fat, carbohydrates, or sodium in their diets. In addition, all subjects were asked to exercise for at least 35 min/day on at least 5 days each week.

2.2 Laboratory and imaging studies

All individuals underwent physical examination and clinical laboratory tests after an overnight (≥8 h) fasting period at baseline. The liver fat content of study subjects was measured using magnetic resonance imaging-derived proton density fat fraction (MRI-PDFF) (Philips, Ingenia 3.0 T [CX] R5.3.1) at baseline and at 24 weeks. Details regarding MRI-PDFF are provided in Appendix S1.

2.3 Endpoints and assessments

The primary endpoint of this study was the change from baseline in liver fat content at week 24, as quantified by MRI-PDFF. Secondary endpoints were the change from baseline in body weight and the changes from baseline in liver enzymes of aspartate transaminase (AST), ALT, and gamma glutamyl transferase (γ-GT). We also investigated the changes in hepatic inflammatory markers such as interleukin 6 (IL-6), tumour necrosis factor-α (TNF-α), fibroblast growth factor 21 (FGF21), cytokeratin 18 (CK-18) fragments, and ferritin levels from week 0 (baseline) to week 24 as an exploratory outcome. Other secondary endpoints were the median change from baseline at week 24 in HbA1c, FPG, and lipid profile. Changes in insulin resistance and secretory function of beta cells were quantified using the homeostasis model assessment of insulin resistance (HOMA-IR) and the homeostasis model assessment of beta cell function (HOMA-β), respectively.13 Variables required for these calculations were measured at baseline and the final visit. Safety variables were adverse events, hypoglycaemic episodes, and the findings of standard laboratory analyses, physical examination, vital signs, and electrocardiography. The statistical analysis plan is described in Appendix S1.

3 RESULTS 3.1 Demographic and baseline characteristics

The mean age of patients was 52.33 ± 11.31 years, and 33 participants were male (Table 1). The mean duration of T2D was 4.59 ± 3.51 years, and the mean BMI was 28.61 ± 3.82 kg/m2. Approximately 80% of study participants had moderate to severe fatty liver disease, as assessed by ultrasonography at baseline. In the overall study group, 74.5% patients were treated with metformin at baseline (92.0% in the evogliptin group and 57.7% in the pioglitazone group).

TABLE 1. Baseline characteristics Evogliptin Pioglitazone P value Age (y) 54.24 ± 10.91 50.50 ± 11.60 .2296 Male 20 (80.00) 13 (50.00) .0250 Diabetes duration (y) 5.08 ± 3.12 4.12 ± 3.84 .1175 Body weight (kg) 80.32 ± 14.63 76.24 ± 12.74 .2926 BMI (kg/m2) 28.76 ± 3.18 28.46 ± 4.41 .7973 HbA1c (%) 7.16 ± 0.64 7.19 ± 0.63 .8956 FPG (mg/dl) 134.40 ± 36.37 134.12 ± 26.37 .5245 HOMA-IR 5.38 ± 4.22 5.20 ± 3.72 .8808 HOMA-β 88.32 ± 55.60 85.46 ± 49.32 .9179 eGFR (mL/min/1.73m2) 93.98 ± 15.86 96.14 ± 18.09 .6526 AST (IU/L) 44.00 ± 21.53 41.88 ± 12.86 .8073 ALT (IU/L) 58.52 ± 28.72 66.00 ± 27.73 .1453 γ-GT (IU/L) 49.24 ± 27.87 63.42 ± 41.61 .2193 Total cholesterol (mg/dl) 152.52 ± 38.51 159.19 ± 42.51 .5601 Triglyceride (mg/dl) 212.16 ± 131.75 163.54 ± 72.79 .2921 HDL-C (mg/dl) 41.68 ± 10.76 44.12 ± 10.69 .4105 LDL-C (mg/dl) 88.44 ± 32.47 98.04 ± 43.53 .3901 Free fatty acids (uEq/L) 642.88 ± 389.15 685.35 ± 323.06 .4602 Degree of fatty livera Mild 3 (12.00) 7 (26.92) .3774 Moderate 15 (60.00) 14 (53.85) Severe 7 (28.00) 5 (19.23) Concomitant OADs Metformin 23 (92.00) 15 (57.69) .0049 Concurrent medication Antihypertensive agents 17 (68.00) 10 (38.46) .0346 Lipid-lowering agents 14 (56.00) 15 (57.69) .9029 Note: Values are expressed as mean ± SD or n (%). Abbreviations: ALT, alanine aminotransferase; AST, aspartate transaminase; BMI, body mass index; eGFR, estimated glomerular filtration rate; FPG, fasting plasma glucose; γ-GT, gamma-glutamyl transferase; HDL-C, high-density lipoprotein cholesterol; HOMA-β, homeostatic model assessment for β-cell function; HOMA-IR, homeostatic model assessment for insulin resistance; LDL-C, low-density lipoprotein cholesterol; OADs, oral antidiabetic agents. 3.2 Effect on hepatic fat content assessed using MRI-PDFF

Hepatic fat content quantified by MRI-PDFF was significantly reduced in the pioglitazone group compared with the evogliptin group. Pioglitazone-treated participants showed a greater reduction in liver fat content compared with evogliptin-treated participants at week 24 (–6.02% ± 1.04% vs. –1.69% ± 0.98%, respectively; P = .0047). Similar changes were seen in absolute liver fat content after treatment (Table 2). A decreased trend without significance in end-of-treatment liver fat content in the evogliptin group (from 16.15% to 14.38%, P = .0834) and a significant decrease in the pioglitazone group (from 15.62% to 9.69%, P = .0003) were found after 24 weeks compared with baseline.

TABLE 2. Changes from baseline in liver fat content as assessed by MRI-PDFF from baseline following 24 weeks of treatment Evogliptin Pioglitazone P valueb Liver fat content (%) Baseline 16.15 ± 6.50 15.62 ± 7.55 Week 24 14.38 ± 7.47 9.69 ± 4.68 Change from baselinea −1.69 ± 0.98 −6.02 ± 1.04 .0047 P valuec 0.0834 0.0003 Note: Values are expressed as mean ± SD (except in the case of a, expressed as LS mean ± SE). Abbreviation: MRI-PDFF, magnetic resonance imaging-derived proton density fat fraction. 3.3 Effects on fatty liver-related variables

Changes in fatty liver-related variables following 24 weeks of treatment compared with baseline are presented in Table S1. After 24 weeks of treatment, the median (IQR) change in the subcutaneous area was 0.25 (−16.35, 10.70) cm2 in the evogliptin group and 14.05 (2.15, 23.80) cm2 in the pioglitazone group. Furthermore, the median (IQR) change in the visceral fat area was 4.75 (−10.60, 10.50) cm2 in the evogliptin group and 1.45 (−7.75, 6.95) cm2 in the pioglitazone group. Compared with baseline, there was a decrease in end-of-treatment AST and ALT levels in both groups. The end-of-treatment γ-GT and FGF21 levels decreased in the pioglitazone group and showed a tendency to increase in the evogliptin group. The end-of-treatment IL-6, TNF-α, CK-18 fragments, and ferritin levels in both groups also decreased compared with levels at baseline.

3.4 Other clinical endpoints

Both the evogliptin- and pioglitazone-treated groups showed a reduction in HbA1c level after 24 weeks, with a median (IQR) change of −0.31 (−0.73, 0.18) in the evogliptin group and of −0.48 (−0.80, −0.10) in the pioglitazone group. Both groups also presented reductions in FPG, triglyceride, LDL-C, and free fatty acid levels after 24 weeks of treatment. Insulin resistance as assessed by HOMA-IR showed a tendency to decrease in both groups, whereas the secretory function of beta cells quantified by HOMA-β showed a tendency to increase. While the pioglitazone group showed a remarkable increase in body weight with a median (IQR) change of 2.50 (0.00, 4.40) kg, the evogliptin group showed a reduction in body weight (–0.30 [–1.40, 1.20] kg) after 24 weeks.

3.5 Adverse events

During the study period, the proportion of patients with one or more adverse events was similar between the two groups (Table S2). The most common adverse event in the evogliptin group was gastrointestinal disorder (n = 2 [8.33%]), while in the pioglitazone group it was infection (n = 3 [11.54%]). One serious adverse event (varicose veins) occurred in one subject in the evogliptin group. There were no drug-related adverse events or hypoglycaemic events in either group during the study period.

4 DISCUSSION

The current study showed that the hepatic fatty burden was decreased with both evogliptin and pioglitazone treatment over 12 weeks in subjects with T2D and NAFLD, and that more statistically significant attenuation was observed in the pioglitazone-treated group. Although several clinical studies have focused on fatty liver,14 none of the new medical treatments for NAFLD have been introduced in the clinic. T2D is one of the risk factors for NAFLD, and vice versa.3, 5, 6 This suggests that hypoglycaemic agents that reduce metabolic burden would be of benefit for NAFLD. Therefore, we investigated the benefit of evogliptin, which showed proven efficacy in NAFLD in preclinical studies,8-10 on NAFLD compared with pioglitazone.

Our study has several clinical implications. First, the hepatic fatty burden was decreased after pioglitazone use. Pioglitazone is a unique drug among antihyperglycaemic agents recommended for non-alcoholic steatohepatitis because of its improvement in liver histology.15 Additionally, unlike previous studies, we showed that a lower dose of pioglitazone (15 mg/day) compared with the proven dosages (30~45 mg/day)15 was effective in attenuating fatty content in liver.

Second, although the beneficial effect on hepatic fatty burden was lower in the evogliptin group compared with the pioglitazone group, the side effects and body weight gain were also lower in evogliptin-treated subjects. Pioglitazone-treated subjects gained weight (median 2.50 kg), whereas the evogliptin-treated group maintained their body weight (median 0.30 kg) over the 24 weeks. The body weight changes mostly resulted from gain of subcutaneous fat. Analysis of body fat composition by MRI-PDFF showed that the subcutaneous fat was increased (median 14.05 cm2) in the pioglitazone treatment group, whereas it was slightly increased (median 0.25 cm2) in evogliptin-treated subjects. Although treatment with evogliptin did not show statistical significance in decreasing liver fat, we observed a tendency for attenuating the hepatic fat burden while maintaining body weight in the evogliptin-treated group. The current study was a pilot and explorative trial, and we believe that evogliptin has the potential to attenuate the hepatic fat burden in NAFLD. Additionally, regarding the side effects of pioglitazone, including heart failure, oedema, and osteoporosis, evogliptin might be an option for ameliorating hepatic fatty burden.

Third, we showed attenuated hepatic fatty burden as well as glycaemic lowering efficacy in evogliptin. HbA1c and FPG were decreased in both the evogliptin-and pioglitazone-treated groups, and those changes were comparable. Despite this trial being a pilot study, we believe that our results suggest a potential benefit of evogliptin for subjects with T2D and NAFLD. Compared with the previous sitagliptin study, which reported a reduction in hepatic fat content of 1.2% over 24 weeks,16 the current trial showed a slightly better outcome for evogliptin of reducing hepatic fat content by 1.69% from baseline.

Finally, although the mean T2D duration of evogliptin- and pioglitazone-treated subjects was comparable at baseline, the proportion of patients with metformin combination therapy was higher in the evogliptin-treated group. This suggests that the antihyperglycaemic drug-naïve subjects may have tended to be assigned to the pioglitazone group. Additionally, the HOMA-IR value in the study population was much higher compared with that of the general T2D population in Korea.17, 18 Moreover, half of the study participants had moderate to severe NAFLD, and the proportion of mild NAFLD was higher in the pioglitazone group than in the evogliptin group (26.9% vs. 12.0%). Considering the action mechanism of pioglitazone, which is mainly attenuating insulin resistance,19, 20 an insulin resistance state in drug-naïve subjects with mild hepatic fatty burden was more assigned to pioglitazone, and would therefore produce a better outcome in the pioglitazone treatment group.

Despite the clinical relevance of the current study, it has several limitations. First, the sample size of this study was not calculated for the statistical test, and it was not large enough to have sufficient power to evaluate the efficacy in the liver fat content. Additionally, as the recruited hospitals were tertiary university hospitals, the generalization and representativeness of the study are not guaranteed. We believe that a large, and long-term follow-up study, would show a definitive outcome for the impact of evogliptin in attenuating hepatic contents in NAFLD. Because of the absence of a placebo-treated group, we could not assess the natural course of NAFLD in a study population. Moreover, we did not include a control group with diet and exercise, which could influence the severity of NAFLD. Finally, it is still possible that metformin could influence attenuating NAFLD.

In summary, this 24-week trial showed that evogliptin was safe and showed a trend of reducing liver fat in patients with T2D and NAFLD. Our results also showed a clinically comparable change in glycaemic control with evogliptin therapy compared with pioglitazone treatment.

ACKNOWLEDGEMENTS

The authors wish to acknowledge Professor Jin Young Choi, Department of Radiology, Yonsei University College of Medicine, for help with analysing MRI-PDFF images, and Professor Seung Up Kim, Division of Hepatology, Department of Internal Medicine, Yonsei University College of Medicine, for help with interpreting the significance of the results of this study. This study was supported by Dong-A ST Co. Ltd.

CONFLICT OF INTEREST

The authors have no conflicts of interest to declare.

AUTHOR CONTRIBUTIONS

Conception and design: EH, JHH, and B-WL; development of methodology: EH, JHH, and B-WL; analysis and interpretation of data: EH, JHH, and B-WL; writing, review, and/or revision of the manuscript: EH, JHH, and B-WL; administrative, technical, or material support: EYL, JCB, SWC, SHY, SHK, and KSP; and study supervision: B-WL.