diabetes research and clinical practice 109 (2015) 199–205
Contents available at ScienceDirect
Diabetes Research and Clinical Practice journ al h ome pa ge : www .elsevier.co m/lo cate/diabres
Effect of sitagliptin on intrahepatic lipid content and body fat in patients with type 2 diabetes Hiroyuki Kato a, Yoshio Nagai a,*, Akio Ohta a, Ayumi Tenjin a, Yuta Nakamura a, Hidekazu Tsukiyama a, Yosuke Sasaki a, Hisashi Fukuda a, Toshihiko Ohshige a, Yuko Terashima a, Yukiyoshi Sada a, Akihiko Kondo a, Toshiyasu Sasaoka b, Yasushi Tanaka a a
Department of Internal Medicine, Division of Metabolism and Endocrinology, St. Marianna University School of Medicine, Kawasaki 216-8511, Japan b Department of Clinical Pharmacology, University School of Toyama, 930-0194 Sugitani, Toyama, Japan
article info
abstract
Article history:
Aims: To evaluate the effect of the DPP-4 inhibitor sitagliptin on intrahepatic lipid (IHL)
Received 26 July 2014
content and body fat in overweight Japanese patients with type 2 diabetes.
Received in revised form
Methods: A prospective, 24-week, single-center, open-label comparative study enrolled 20
7 February 2015
Japanese patients with type 2 diabetes (male: 11, female: 9) with a BMI 25 kg/m2 or fatty
Accepted 12 April 2015
liver. Subjects were randomly assigned to receive treatment with sitagliptin (25 mg titrated
Available online 20 April 2015
up to 50 mg: S) or glimepiride (0.5 mg titrated up to 1 mg: G). After starting each treatment, IHL and total fat mass were evaluated by 1H-magnetic resonance spectroscopy (1H-MRS) and
Keywords:
dual energy X-ray absorptiometry (DEXA), respectively at baseline and at 12 weeks and
Type 2 diabetes
24 weeks.
DPP-4 inhibitor
Results: After 24 weeks, HbA1c levels showed a similar significant decrease in both groups
Sitagliptin
from 7.2 (7.0, 7.5) to 6.6 (6.4, 6.8)%, (54 (53, 56) to 48(47, 49) mmol/mol) with S and 7.3(6.8, 7.4) to
Fatty liver
6.6 (6.3, 6.7)%, (55 (51, 56) to 48 (46, 49) mmol/mol) with G, median (interquartile range), p < 0.05 vs. baseline, with no significant differences between the two groups. The IHL and
Body fat
total body fat mass were decreased in S group from 24.5(18.9, 36.6) to 20.5 (14.6, 28.5)% ( p = 0.009) and 22.5 (20.6, 33.7) to 21.6 (19.7, 32.4)kg ( p = 0.028), respectively, but not in G group. Conclusions: Our findings indicate that sitagliptin and glimepiride achieved similar glycemic control, but only sitagliptin reduced IHL and total body fat (UMIN: 000013356). # 2015 Elsevier Ireland Ltd. All rights reserved.
1.
Introduction
The American Diabetes Association (ADA) and the European Association for the Study of Diabetes (EASD) have published a
position statement on type 2 diabetes mellitus (T2DM) that stresses the importance of a patient-centered approach [1,2]. In this statement, metformin (MET) is recommended as initial drug therapy to be provided simultaneously with or soon after commencing lifestyle modification because of its high efficacy
* Corresponding author. Tel.: +81 44 977 8111x3149; fax: +81 44 976 8516. E-mail address: [email protected] (Y. Nagai). http://dx.doi.org/10.1016/j.diabres.2015.04.008 0168-8227/# 2015 Elsevier Ireland Ltd. All rights reserved.
200
diabetes research and clinical practice 109 (2015) 199–205
for reducing HbA1c, low cost, low risk of hypoglycemia, and neutral effect on body weight (or weight loss). If the target HbA1c is not achieved, the second and third agents to be combined with MET are selected from among five drug classes: sulfonylureas (SU), thiazolidinediones (TZD), dipeptidyl peptidase-4 inhibitors (DPP-4I), glucagon-like peptide-1 (GLP-1) agonists, or basal insulin. The unique and important point of the statement is comparison of the characteristics of these five drug classes to allow selection of suitable therapy for each patient. In the statement, DPP-4I are characterized as showing intermediate efficacy for HbA1c reduction, with a low risk of hypoglycemia, neutral effect on body weight, and few major side effects. DPP-4I increase the plasma concentration of active GLP-1, which increases insulin secretion in a glucosedependent manner and simultaneously suppresses the secretion of glucagon. Thus, DPP-4I reduce fasting and postprandial plasma glucose levels with a relatively low risk of hypoglycemia [4,5]. A recent study showed that HbA1c reduction by DPP-4I is greater in Asian patients than Caucasians [3], so these agents are widely used for Japanese patients. The GLP-1 receptor (GLP-1R) has been detected on human hepatocytes and adipocytes [6], and a GLP-1 analog (exendin-4) has been demonstrated to attenuate triglyceride synthesis by primary cultured human hepatocyte [7]. Furthermore, another GLP-1 agonist (liraglutide) achieved 42% reduction of the intrahepatic lipid (IHL) content and 11% reduction of the visceral fat volume in obese patients with T2DM after 6 months of treatment [8]. However, the plasma concentration of active GLP-1 is much lower during DPP4-I treatment compared with the levels during treatment with these GLP1 agonists [9,10]. According to the ADA/EASD statement, the effect of DPP4-I on body weight is neutral, but not marked [1,2]. The effects of DPP-4I on body weight and hepatic fat accumulation have not been fully evaluated in patients with diabetes, although animal experiments have shown attenuation of hepatic steatosis by DPP-4I administration [11,12]. Therefore, the aim of the present study was to investigate the effects of sitagliptin, one of the DPP-4I, on IHL and body fat compared with low-dose SU therapy using glimepiride in Japanese patients with type 2 diabetes with a BMI 25 kg/m2 or fatty liver on abdominal ultrasonography.
2.
Materials and methods
2.1.
Subjects
The subjects were 20 Japanese patients with T2DM (11 men and 9 women aged 58.5 (40.0, 77.0) years, median (interquartile range (IQR)) with a BMI 25 kg/m2 or fatty liver detected by ultrasonography. The ultrasonography was carried out by using convex-array probe (3.5 MHz) for assessing the liver, and after a fasting period of 12 h. Diagnosis of fatty liver was attempted based on the difference between the echo intensities of the liver and kidney [13]. They were recruited from the outpatient clinic of St. Marianna University Hospital (Kawasaki, Japan) with following inclusion criteria: (1) stable, but inadequate,
glycemic control (9.4% (79 mmol/mol) > HbA1c > 7.4% (57 mmol/mol)) and (2) drug-naı¨ve or MET monotherapy. The exclusion criteria were pregnancy, severe illness, anemia, renal failure (serum creatinine 2.0 mg/dL) and/or overt proteinuria, chronic liver disease, thyroid disease, malignancy, severe hypoglycemia requiring assistance within the previous 6 months, and use of medications that could affect glycemic control (including systemic glucocorticoids).
2.2.
Study design
Diet and exercise therapy were given at diagnosis and the patient education staff encouraged their continuation throughout the study period. All the patients were instructed by a doctor to consume a balanced diet (25 kcal/kg of ideal body weight) with about 60% of energy intake as carbohydrate, about 25% as fat, and about 15% as protein. The patients were also advised to perform one or two sessions of walking exercise (20–30 min each) on 3–5 day per week. All patients visited the outpatient clinic of our hospital every month, and were encouraged by their doctors to continue calorie restriction and exercise. Alcohol consumption was restricted to less than 20 g/day. All patients gave written informed consent, and the study was approved by the Ethics Committee of St. Marianna University School of Medicine. Patients were randomly assigned to treatment with sitagliptin (S group, n = 10) or glimepiride (G group, n = 10). Administration of sitagliptin and glimepiride was respectively started at 25 mg/day and 0.5 mg/day after breakfast. These doses could be increased to 50 mg/day and 1.0 mg/day, respectively, in order to achieve a fasting plasma glucose (FPG) < 130 mg/dl and HbA1c < 7.0% (53 mmol/mol) up to 24 weeks. We started from half the dose of each in reference to effects of sitagliptin [14] and glimepiride [15] on blood glucose levels. Fasting blood samples were obtained at baseline, as well as at 12 weeks and 24 weeks after starting treatment.
2.3.
Data collection
The following parameters were investigated: glucose, lipids [HDL-cholesterol (HDL-C), LDL-cholesterol (LDL-C), triglycerides (TG), and free fatty acids (FFA)], HbA1c, glycated albumin (GA), high molecular weight adiponectin (HMW-Ad), leptin, and highly sensitive C-reactive protein (hsCRP). TG, HDL-C, LDL-C, FFA and glucose were measured by standard methods. HbA1c was measured by the latex cohesion method (Determiner HbA1c, Kyowa Medex, Tokyo, Japan). GA was determined by an enzymatic method with an albumin-specific protease, ketoamine oxidase, and an albumin assay reagent (Lucica GA-L, Asahi Kasei Pharma, Tokyo, Japan). Leptin was measured by a radioimmunoassay (LSI Medience Corporation, Tokyo, Japan). HMW-Ad was quantitated by a sandwich enzyme-linked immunosorbent assay that used two antihuman adiponectin monoclonal antibodies (Sekisui Medical Co., Ltd., Japan). Serum hsCRP was measured by latex agglutination turbidimetry (Siemens Healthcare Diagnostics, Japan). To evaluate the liver fat content and the total body fat mass, 1H-MRS and dual energy X-ray absorptiometry (DEXA) were performed at baseline, 12 weeks, and 24 weeks.
201
diabetes research and clinical practice 109 (2015) 199–205
2.4.
1
H-MRS analysis of intrahepatic lipid content
The method employed for 1H-MRS has been described previously [16,17]. Briefly, images of the liver were acquired with a 1.5-T whole-body system (Intera Achieva; Phillips Medical Systems, Netherlands). The volume of interest (20 20 20 mm3) was centered on segment 6 of the liver, and spectra were acquired using a point-resolved spectroscopy sequence (PRESS) with a repetition time (TR) of 2000 ms, echo time (TE) of 144 ms, and acquisition number of 128. Then intracellular triglycerides were quantified from the methylene proton peaks (–CH2) of fatty acids at 1.3 ppm. Spectrum fitting and analysis were done with LC-Model software (version 6.2; S. Provencher, PhD, Canada). IHL was calculated as a percentage as follows: area under the curve of the methylene proton peak (AUC-IHL)/[AUC-IHL + area under the water proton peak (AUCwater)] 100.
2.5.
Statistical analysis
Results are presented as median (interquartile range (IQR)). We used the Wilcoxon signed-rank test, Mann–Whitney U test. All analyses were performed with Stat-View software (Abacus
Concepts, Berkeley, CA) and statistical significance was accepted at p < 0.05.
3.
Results
The baseline clinical characteristics of the patients are shown in Table 1. There were no significant differences between the two groups with regard to the age, blood pressure, BMI, HbA1c, GA, FPG, LDL-C, HDL-C, TG, FFA, HMW-Ad, leptin, and hsCRP. A comparison of laboratory parameters between the two groups during study period is displayed in Table 2. At 24 weeks, HbA1c and GA showed a similar significant decrease from their baseline levels in both groups. FPG was significantly decreased from baseline at 12 and 24 weeks in G group and at 12 weeks in S group, while there was no significant difference of FPG between the two groups at any time. TG and hsCRP was significantly decreased from baseline at 12 weeks in S group. There were no differences from the baseline levels at 24 weeks for HOMA-R, insulin, plasma lipids, adipokines, and hsCRP, and these parameters did not differ between the two groups. We confirmed fatty liver in S group (n = 9) and G group (n = 9) by
Table 1 – Baseline characteristics of the patients. S group N (male/female) Age (yrs) Blood pressure Systolic (mmHg) Diastolic (mmHg) BMI (kg/m2) HbA1c (NGSP) (%) HbA1c (mmol/mol) GA (%) Fasting plasma glucose (mg/dL) IRI (mU/mL) HOMA-IR Plasma lipid LDL-cholesterol (mg/dL) HDL-cholesterol (mg/dL) Triglyceride (mg/dL) FFA (mmol/L) Adipo-cytokine HM-adiponectin (mmol/L) Leptin (mmol/L) HS-CRP (mmol/L) IHL (%) LBM (kg) Fat(kg) %-Fat (%) Other medications for diabetes, n (%) OHAs(a-GI/Met/TZD) Statins ARBs Calcium channel blocker Diuretics b-blocker
G group
p Value (S vs. G)
10 (6/4) 62 (56, 70)
10 (5/5) 55 (42, 62)
0.142
141 (134, 143) 85 (79, 90) 25.6 (24.7, 32.5) 7.2 (7.0, 7.5) 54 (53, 56) 18.3 (16.3, 19.7) 131 (126, 172) 8.6 (5.5, 9.9) 3.0 (1.7, 4.8)
131 (118, 140) 78 (74, 84) 26.6 (25.0, 32.4) 7.3 (6.8, 7.4) 55 (51, 56) 16.4 (16.0, 18.4) 138 (134, 143) 8.1 (5.6, 8.9) 2.6 (1.9, 3.0)
0.325 0.120 0.706 0.939 0.939 0.306 0.791 0.570 0.597
127 (115, 151) 48 (43, 53) 153 (119, 190) 0.6 (0.5, 0.8)
112 (108, 136) 50 (43, 56) 151 (111, 189) 0.6 (0.4, 0.7)
0.151 0.791 0.910 0.450
1.2 (0.7, 1.6) 12.6 (7.6, 14.6) 0.18 (0.05, 0.23) 24.5 (18.9, 36.6) 50.3 (46.5, 53.0) 22.5 (20.6, 33.7) 37.7 (29.6, 39.7)
0.8 (0.6, 3.1) 10.3 (6.0, 15.1) 0.19 (0.07, 0.28) 18.0 (10.8, 37.5) 53.2 (42.8, 60.2) 21.3 (18.9, 31.6) 35.3 (29.3, 36.6)
0.734 0.678 0.762 0.345 0.497 0.545 0.364
0(0)/6(60)/0(0) 3(30) 3(30) 1(10) 1(10) 1(10)
0(0)/5(50)/0(0) 4(40) 4(40) 3(30) 0(0) 1(10)
Data are the median (IQR); analyzed using Mann–Whitney U test. p < 0.05, S; sitagliptin, G; glimepiride BMI; body mass index, GA; glycoalbumin, IRI; immunoreactive insulin, HOMA-IR; homeostasis model assessment as an index of insulin resistance, LDL; low-density lipoprotein, HDL; high-density lipoprotein, FFA; free fatty acids, HMW; high molecular weight, HS; highly sensitive, IHL; intrahepatic lipid, LBM; lean body mass, OHAs; oral hypoglycemic agents, Met; metformin, TZD; thiazolidinedione.
202
diabetes research and clinical practice 109 (2015) 199–205
Table 2 – Changes of clinical data during treatment with sitagliptin or glimeplide. Group
baseline
12 weeks
p Value (vs. baseline)
p Value (D12 W S vs. D12 W G)
50.0 (25, 50) 0.5 0.5
24 weeks
p Value (vs. baseline)
p Value (D24 W S vs. D24 W G)
Drug dose (mg)
S G
50.0 (31,50) 0.5 0.5
BMI (kg/m2)
S G
25.6 (24.7, 32.5) 26.6 (25.0, 32.4)
25.9 (24.5, 31.8) 25.8 (24.9, 31.9)
0.799 0.959
0.821
25.7 (24.8, 32.1) 26.4 (24.9, 32.2)
0.386 0.241
0.496
HbA1c (NGSP)(%)
S G
7.2 (7.0, 7.5) 7.3 (6.8, 7.4)
6.7 (6.3, 6.9) 6.7 (6.4, 7.1)
0.005 0.028
0.096
6.6 (6.4, 6.8) 6.6 (6.3, 6.7)
0.005 0.011
0.151
HbA1c (NGSP)
S G
54 (53, 56) 55 (51, 56)
49 (46, 50) 49 47(, 52)
0.005 0.028
0.096
48 (47, 49) 48 (46, 49)
0.005 0.011
0.151
GA (%)
S G
18.3 (16.3, 19.7) 16.4 (16.0, 18.4)
15.7 (14.1, 17.1) 15.3 (14.5, 16.7)
0.007 0.022
0.450
14.4 (13.9, 15.9) 15.2 (13.5, 16.4)
0.011 0.028
0.597
FPG(mg/dL)
S G
131 (126, 172) 138 (134, 143)
128 (111, 142) 133 (125, 136)
0.037 0.037
0.450
124 (114, 145) 122 (113, 135)
0.075 0.044
0.880
IRI(mU/mL)
S G
8.6 (5.5, 9.9) 8.1 (5.6, 8.9)
7.5 (5.6, 9.9) 7.5 (5.6, 9.9)
0.221 0.541
0.257
9.1 (6.7, 11.5) 6.3 (4.0, 9.9)
0.839 0.759
0.496
HOMA-R
S G
3.0 (1.7, 4.8) 2.6 (1.9, 3.0)
2.3 (1.7, 3.2) 2.3 (1.8, 3.0)
0.139 0.114
0.290
2.8 (2.0, 3.3) 1.8 (1.0, 1.8)
0.333 0.374
0.31
LDL-C (mg/dL)
S G
132 (115, 151) 112 (108, 136)
127 (126, 104) 126 (111, 136)
0.139 0.678
0.227
133 (108, 159) 116 (103, 138)
0.647 0.759
0.130
HDL-C (mg/dL)
S G
48 (43, 53) 50 (43, 56)
47 (43, 50) 45 (41, 55)
0.185 0.636
0.369
46 (43, 51) 44 (41, 55)
0.285 0.919
0.650
TG (mg/dL)
S G
153 (119, 190) 151 (111, 189)
136 (93, 160) 160 (101, 208)
0.037 0.415
0.131
145 (114, 167) 126 (104, 215)
0.203 0.953
0.762
FFA (mmol/L)
S G
0.6 (0.5, 0.8) 0.6 (0.4, 0.7)
0.7 (0.5, 1.0) 0.6 (0.5, 0.7)
0.508 1.000
0.650
0.7 (0.4, 0.9) 0.7 (0.5, 0.8)
0.285 0.124
0.290
HM-adiponectin (mmol/L)
S G
1.2 (0.7, 1.6) 0.8 (0.6, 3.1)
1.2 (0.8, 2.0) 1.1 (0.8, 2.7)
0.684 0.333
0.762
1.3 (0.7, 1.9) 1.2 (0.7, 3.6)
0.139 0.124
0.821
Leptin (mmol/L)
S G
12.6 (7.6, 14.6) 10.3 (6.0, 15.1)
11.9 (8.8, 14.2) 12.2 (6.8, 15.1)
0.879 0.647
0.706
10.4 (8.3, 15.3) 11.0 (5.7, 16.9)
0.441 0.953
0.895
HS-CRP (mmol/L)
S G
0.18 (0.05, 0.23) 0.19 (0.07, 0.28)
0.05 (0.04, 0.11) 0.17 (0.03, 0.37)
0.009 0.859
0.041
0.09 (0.04, 0.15) 0.18 (0.03, 0.20)
0.059 0.859
0.191
Data are the median (IQR); analyzed using Wilcoxon signed-rank test, Mann–Whitney U test. p < 0.05 BMI; body mass index, GA; glycoalbumin, FPG; fasting plasma glucose, IRI; immunoreactive insulin, HOMA-IR; homeostasis model assessment as an index of insulin resistance, LDL; low-density lipoprotein, HDL; high-density lipoprotein, FFA; free fatty acids, HMW; high molecular weight, HS; highly sensitive, S; sitagliptin, G; glimepiride.
ultrasonography. Two patients did not have ultrasonography, but their BMI was >25 kg/m2. As shown in Fig. 1A, IHL was significantly decreased from baseline at 24 weeks in S group, but not in G group (S group: from 24.5 (18.9, 36.6) to 20.1 (14.6, 28.5)%, p = 0.009, G group: from 18.0 (10.8, 37.5) to 12.8 (10.2, 39.0)%, p = 0.959, median(IQR)). The percent change of IHL from baseline at 24 weeks was 15.1 ( 11.7, 18.8)% in S group and 1.6 ( 17.9, 20.1) % in G group (Fig. 1B). Fig. 2A displays the changes of total body fat mass at 24 weeks; there was a significant decrease in S group, but not in G group (S group: from 22.5 (20.6, 33.7) to 21.6 (19.7, 32.4) kg, p = 0.028, G group: from 21.2 (18.9, 31.6) to 22.6 (19.2, 31.1) kg, p = 0.241, media(IQR)). The percent change was greater in S group than in G group ( 4.7 ( 7.5, 0.0)% vs. 4.9 ( 2.3, 10.1)%, p = 0.007, median(IQR)). As shown in Fig. 2B, lean body mass did not differ from the baseline level at 24 weeks in both groups (S group: from 50.3 (46.5, 53.0) to 51.3 (46.8, 54.8) kg, p = 0.114, G group: from
53.2 (42.8, 60.2) to 50.9 (44.9, 57.2) kg, p = 0.799, median (IQR)). BMI showed no significant change in either group and there was no significant difference between the two groups at any point (S group: 25.6 (24.7, 32.5) at baseline, 25.9 (24.5, 31.8) at 12 weeks, and 25.7 (24.8, 32.1) at 24 weeks (kg/m2) median(IQR)), G group: 26.6 (25.0, 32.4) at baseline, 25.8 (24.9, 31.9) at 12 weeks, and 26.4 (24.9, 32.2) at 24 weeks (kg/ m2) median (IQR)). There were no significant differences between the two groups with regard to the number of subjects with concomitant metformin therapy and the doses (S group; n = 6, 750 (485, 875) mg, G group; n = 5, 1000 (750, 1500) mg, median(IQR)). The percent change of IHL from baseline at 24 weeks was not significantly different between patients with and without concomitant use of metformin (S group without metformin (n = 4) 12.8( 28.9, 3.2) % vs. S group with metformin (n = 6) 15.6( 18.1, 13.2)%, p = 0.999; G group without metformin (n = 5) 11.5( 29.8, 6.2)% vs. G group with metformin (n = 5) 17.3 (9.4, 21.0)%, p = 0.251,
diabetes research and clinical practice 109 (2015) 199–205
203
Fig. 1 – Absolute change (A) and percent change from baseline (B) of intrahepatic lipid (IHL) assessed by 1H-MRS during the 24-week study period.
median (IQR)). The percent change of total body fat mass from baseline at 24 weeks was also not significantly different between patients with and without concomitant use of metformin (S group without metformin (n = 4) 5.1 ( 6.8, 2.7)% vs. S group with metformin (n = 6) 3.9 ( 7.3, 0.02)%, p = 0.670; G group without metformin (n = 5) 1.7 ( 7.7, 2.6) % vs. G group with metformin (n = 5) 9.8 (7.1, 10.2)%, p = 0.175, median(IQR)). During the study period, there were no symptoms or signs of hypoglycemia and no FPG levels less than 70 mg/dl were recorded.
4.
Discussion
The present study demonstrated that the reduction of HbA1c and GA was similar after treatment with sitagliptin or lowdose glimepiride for 24 weeks, but only sitagliptin (not
glimepiride) significantly decreased IHL and total body fat mass. In the liver, TG are not only produced from free fatty acids (FFA) that are released from adipose tissue by lipolysis, but also by intrahepatic fatty acid synthesis from dietary glucose via the malonyl–CoA pathway. Intake of a carbohydrate (CH)restricted low energy diet (CH: 10% of total energy) for 48 h has been shown to markedly reduce IHL in obese subjects compared with an energy-matched high CH diet (CH: 65% of total energy) [18]. We also previously reported the effect of diet on lowering IHL in patients with type 2 diabetes or obese subjects with impaired glucose tolerance (IGT) [16,17,19]. However, patients did not change their dietary energy intake or nutritional balance throughout the present study and the improvement of glycemic control was similar between the two groups, suggesting that there was a direct effect of sitagliptin on IHL and body fat mass.
Fig. 2 – Changes of fat mass (A) and lean body mass (B) assessed by dual energy X-ray absorptiometry (DEXA) during the 24week study period.
204
diabetes research and clinical practice 109 (2015) 199–205
We previously reported that treatment with low-dose metformin (750 mg/day) was able to reduce IHL by 25% in obese Japanese subjects with T2DM [20]. Although the exact mechanism of IHL reduction was unclear, activation of hepatic AMP-activated protein kinase (AMPK) by metformin is thought to play a major role in suppressing intrahepatic TG production [21,22]. Hepatic AMPK is considered to be a key regulator of the intracellular energy balance by limiting anabolic pathways to prevent further ATP consumption and by facilitating catabolic pathways to increase ATP production. The activation of hepatic AMPK leads to increased b-oxidation of fatty acids along with simultaneous inhibition of lipogenesis and gluconeogenesis [23]. Recent studies have demonstrated that hepatic AMPK is also activated by GLP-1 in isolated rat hepatocytes [24,25], and DPP-4 deficient rats have a lower hepatic TG content compared with wild-type rats [24]. Cuthbertson et al. reported a marked reduction of IHL in human subjects treated with a GLP-1 analog, and they mentioned hepatic AMPK activation as a possible mechanism [8]. However, the peak plasma level of active GLP-1 after administration of sitagliptin to Japanese subjects was reported as 15–20 pmol/L [9], while injection of the GLP-1 analog liraglutide resulted in much higher levels (4000–8000 pmol/L) [10]. Therefore, the smaller reduction of IHL by sitagliptin in the present study compared with liraglutide in the above report (14% vs. 42%) may be partly explained by a difference of the plasma GLP-1 concentration. The incretin effect on fat cells in adipose tissue differs between GIP and GLP-1. While GIP enhances insulin-induced lipogenesis and inhibits the lipolytic action of glucagon [26], GLP-1 directly stimulates lipolysis through activation of both cAMP-dependent protein kinase (PKA) and AMPK in fat cells [27,28]. DPP-4I therapy simultaneously increases plasma GIP and GLP-1 levels, and Terauchi et al. have observed a similar increase of total GIP and active GLP-1 in Japanese subjects after sitagliptin administration [9]. While DPP-4I have a neutral effect on weight gain according to the ADA/EASD statement, a small but significant reduction of body fat was observed in the present study. Interestingly, recent studies of obese insulinresistant subjects have detected down-regulation of the GIP receptor and its signaling in subcutaneous fat cells (but not in visceral fat cells) compared with lean healthy controls, along with up-regulation of the GLP-1 receptor and lipolysis mediated by it in fat cells from both types of adipose tissue [28,29]. These reports may partly explain our results, but we did not perform analysis of fat cell receptors. A recent meta-analysis of glycemic control by DPP-4I revealed a relatively greater reduction of HbA1c in Asian than non-Asian subjects [3]. The authors pointed out that the baseline BMI was negatively correlated with the HbA1clowering effect of DPP-4I, and that the studies in which the average BMI was
Contents available at ScienceDirect
Diabetes Research and Clinical Practice journ al h ome pa ge : www .elsevier.co m/lo cate/diabres
Effect of sitagliptin on intrahepatic lipid content and body fat in patients with type 2 diabetes Hiroyuki Kato a, Yoshio Nagai a,*, Akio Ohta a, Ayumi Tenjin a, Yuta Nakamura a, Hidekazu Tsukiyama a, Yosuke Sasaki a, Hisashi Fukuda a, Toshihiko Ohshige a, Yuko Terashima a, Yukiyoshi Sada a, Akihiko Kondo a, Toshiyasu Sasaoka b, Yasushi Tanaka a a
Department of Internal Medicine, Division of Metabolism and Endocrinology, St. Marianna University School of Medicine, Kawasaki 216-8511, Japan b Department of Clinical Pharmacology, University School of Toyama, 930-0194 Sugitani, Toyama, Japan
article info
abstract
Article history:
Aims: To evaluate the effect of the DPP-4 inhibitor sitagliptin on intrahepatic lipid (IHL)
Received 26 July 2014
content and body fat in overweight Japanese patients with type 2 diabetes.
Received in revised form
Methods: A prospective, 24-week, single-center, open-label comparative study enrolled 20
7 February 2015
Japanese patients with type 2 diabetes (male: 11, female: 9) with a BMI 25 kg/m2 or fatty
Accepted 12 April 2015
liver. Subjects were randomly assigned to receive treatment with sitagliptin (25 mg titrated
Available online 20 April 2015
up to 50 mg: S) or glimepiride (0.5 mg titrated up to 1 mg: G). After starting each treatment, IHL and total fat mass were evaluated by 1H-magnetic resonance spectroscopy (1H-MRS) and
Keywords:
dual energy X-ray absorptiometry (DEXA), respectively at baseline and at 12 weeks and
Type 2 diabetes
24 weeks.
DPP-4 inhibitor
Results: After 24 weeks, HbA1c levels showed a similar significant decrease in both groups
Sitagliptin
from 7.2 (7.0, 7.5) to 6.6 (6.4, 6.8)%, (54 (53, 56) to 48(47, 49) mmol/mol) with S and 7.3(6.8, 7.4) to
Fatty liver
6.6 (6.3, 6.7)%, (55 (51, 56) to 48 (46, 49) mmol/mol) with G, median (interquartile range), p < 0.05 vs. baseline, with no significant differences between the two groups. The IHL and
Body fat
total body fat mass were decreased in S group from 24.5(18.9, 36.6) to 20.5 (14.6, 28.5)% ( p = 0.009) and 22.5 (20.6, 33.7) to 21.6 (19.7, 32.4)kg ( p = 0.028), respectively, but not in G group. Conclusions: Our findings indicate that sitagliptin and glimepiride achieved similar glycemic control, but only sitagliptin reduced IHL and total body fat (UMIN: 000013356). # 2015 Elsevier Ireland Ltd. All rights reserved.
1.
Introduction
The American Diabetes Association (ADA) and the European Association for the Study of Diabetes (EASD) have published a
position statement on type 2 diabetes mellitus (T2DM) that stresses the importance of a patient-centered approach [1,2]. In this statement, metformin (MET) is recommended as initial drug therapy to be provided simultaneously with or soon after commencing lifestyle modification because of its high efficacy
* Corresponding author. Tel.: +81 44 977 8111x3149; fax: +81 44 976 8516. E-mail address: [email protected] (Y. Nagai). http://dx.doi.org/10.1016/j.diabres.2015.04.008 0168-8227/# 2015 Elsevier Ireland Ltd. All rights reserved.
200
diabetes research and clinical practice 109 (2015) 199–205
for reducing HbA1c, low cost, low risk of hypoglycemia, and neutral effect on body weight (or weight loss). If the target HbA1c is not achieved, the second and third agents to be combined with MET are selected from among five drug classes: sulfonylureas (SU), thiazolidinediones (TZD), dipeptidyl peptidase-4 inhibitors (DPP-4I), glucagon-like peptide-1 (GLP-1) agonists, or basal insulin. The unique and important point of the statement is comparison of the characteristics of these five drug classes to allow selection of suitable therapy for each patient. In the statement, DPP-4I are characterized as showing intermediate efficacy for HbA1c reduction, with a low risk of hypoglycemia, neutral effect on body weight, and few major side effects. DPP-4I increase the plasma concentration of active GLP-1, which increases insulin secretion in a glucosedependent manner and simultaneously suppresses the secretion of glucagon. Thus, DPP-4I reduce fasting and postprandial plasma glucose levels with a relatively low risk of hypoglycemia [4,5]. A recent study showed that HbA1c reduction by DPP-4I is greater in Asian patients than Caucasians [3], so these agents are widely used for Japanese patients. The GLP-1 receptor (GLP-1R) has been detected on human hepatocytes and adipocytes [6], and a GLP-1 analog (exendin-4) has been demonstrated to attenuate triglyceride synthesis by primary cultured human hepatocyte [7]. Furthermore, another GLP-1 agonist (liraglutide) achieved 42% reduction of the intrahepatic lipid (IHL) content and 11% reduction of the visceral fat volume in obese patients with T2DM after 6 months of treatment [8]. However, the plasma concentration of active GLP-1 is much lower during DPP4-I treatment compared with the levels during treatment with these GLP1 agonists [9,10]. According to the ADA/EASD statement, the effect of DPP4-I on body weight is neutral, but not marked [1,2]. The effects of DPP-4I on body weight and hepatic fat accumulation have not been fully evaluated in patients with diabetes, although animal experiments have shown attenuation of hepatic steatosis by DPP-4I administration [11,12]. Therefore, the aim of the present study was to investigate the effects of sitagliptin, one of the DPP-4I, on IHL and body fat compared with low-dose SU therapy using glimepiride in Japanese patients with type 2 diabetes with a BMI 25 kg/m2 or fatty liver on abdominal ultrasonography.
2.
Materials and methods
2.1.
Subjects
The subjects were 20 Japanese patients with T2DM (11 men and 9 women aged 58.5 (40.0, 77.0) years, median (interquartile range (IQR)) with a BMI 25 kg/m2 or fatty liver detected by ultrasonography. The ultrasonography was carried out by using convex-array probe (3.5 MHz) for assessing the liver, and after a fasting period of 12 h. Diagnosis of fatty liver was attempted based on the difference between the echo intensities of the liver and kidney [13]. They were recruited from the outpatient clinic of St. Marianna University Hospital (Kawasaki, Japan) with following inclusion criteria: (1) stable, but inadequate,
glycemic control (9.4% (79 mmol/mol) > HbA1c > 7.4% (57 mmol/mol)) and (2) drug-naı¨ve or MET monotherapy. The exclusion criteria were pregnancy, severe illness, anemia, renal failure (serum creatinine 2.0 mg/dL) and/or overt proteinuria, chronic liver disease, thyroid disease, malignancy, severe hypoglycemia requiring assistance within the previous 6 months, and use of medications that could affect glycemic control (including systemic glucocorticoids).
2.2.
Study design
Diet and exercise therapy were given at diagnosis and the patient education staff encouraged their continuation throughout the study period. All the patients were instructed by a doctor to consume a balanced diet (25 kcal/kg of ideal body weight) with about 60% of energy intake as carbohydrate, about 25% as fat, and about 15% as protein. The patients were also advised to perform one or two sessions of walking exercise (20–30 min each) on 3–5 day per week. All patients visited the outpatient clinic of our hospital every month, and were encouraged by their doctors to continue calorie restriction and exercise. Alcohol consumption was restricted to less than 20 g/day. All patients gave written informed consent, and the study was approved by the Ethics Committee of St. Marianna University School of Medicine. Patients were randomly assigned to treatment with sitagliptin (S group, n = 10) or glimepiride (G group, n = 10). Administration of sitagliptin and glimepiride was respectively started at 25 mg/day and 0.5 mg/day after breakfast. These doses could be increased to 50 mg/day and 1.0 mg/day, respectively, in order to achieve a fasting plasma glucose (FPG) < 130 mg/dl and HbA1c < 7.0% (53 mmol/mol) up to 24 weeks. We started from half the dose of each in reference to effects of sitagliptin [14] and glimepiride [15] on blood glucose levels. Fasting blood samples were obtained at baseline, as well as at 12 weeks and 24 weeks after starting treatment.
2.3.
Data collection
The following parameters were investigated: glucose, lipids [HDL-cholesterol (HDL-C), LDL-cholesterol (LDL-C), triglycerides (TG), and free fatty acids (FFA)], HbA1c, glycated albumin (GA), high molecular weight adiponectin (HMW-Ad), leptin, and highly sensitive C-reactive protein (hsCRP). TG, HDL-C, LDL-C, FFA and glucose were measured by standard methods. HbA1c was measured by the latex cohesion method (Determiner HbA1c, Kyowa Medex, Tokyo, Japan). GA was determined by an enzymatic method with an albumin-specific protease, ketoamine oxidase, and an albumin assay reagent (Lucica GA-L, Asahi Kasei Pharma, Tokyo, Japan). Leptin was measured by a radioimmunoassay (LSI Medience Corporation, Tokyo, Japan). HMW-Ad was quantitated by a sandwich enzyme-linked immunosorbent assay that used two antihuman adiponectin monoclonal antibodies (Sekisui Medical Co., Ltd., Japan). Serum hsCRP was measured by latex agglutination turbidimetry (Siemens Healthcare Diagnostics, Japan). To evaluate the liver fat content and the total body fat mass, 1H-MRS and dual energy X-ray absorptiometry (DEXA) were performed at baseline, 12 weeks, and 24 weeks.
201
diabetes research and clinical practice 109 (2015) 199–205
2.4.
1
H-MRS analysis of intrahepatic lipid content
The method employed for 1H-MRS has been described previously [16,17]. Briefly, images of the liver were acquired with a 1.5-T whole-body system (Intera Achieva; Phillips Medical Systems, Netherlands). The volume of interest (20 20 20 mm3) was centered on segment 6 of the liver, and spectra were acquired using a point-resolved spectroscopy sequence (PRESS) with a repetition time (TR) of 2000 ms, echo time (TE) of 144 ms, and acquisition number of 128. Then intracellular triglycerides were quantified from the methylene proton peaks (–CH2) of fatty acids at 1.3 ppm. Spectrum fitting and analysis were done with LC-Model software (version 6.2; S. Provencher, PhD, Canada). IHL was calculated as a percentage as follows: area under the curve of the methylene proton peak (AUC-IHL)/[AUC-IHL + area under the water proton peak (AUCwater)] 100.
2.5.
Statistical analysis
Results are presented as median (interquartile range (IQR)). We used the Wilcoxon signed-rank test, Mann–Whitney U test. All analyses were performed with Stat-View software (Abacus
Concepts, Berkeley, CA) and statistical significance was accepted at p < 0.05.
3.
Results
The baseline clinical characteristics of the patients are shown in Table 1. There were no significant differences between the two groups with regard to the age, blood pressure, BMI, HbA1c, GA, FPG, LDL-C, HDL-C, TG, FFA, HMW-Ad, leptin, and hsCRP. A comparison of laboratory parameters between the two groups during study period is displayed in Table 2. At 24 weeks, HbA1c and GA showed a similar significant decrease from their baseline levels in both groups. FPG was significantly decreased from baseline at 12 and 24 weeks in G group and at 12 weeks in S group, while there was no significant difference of FPG between the two groups at any time. TG and hsCRP was significantly decreased from baseline at 12 weeks in S group. There were no differences from the baseline levels at 24 weeks for HOMA-R, insulin, plasma lipids, adipokines, and hsCRP, and these parameters did not differ between the two groups. We confirmed fatty liver in S group (n = 9) and G group (n = 9) by
Table 1 – Baseline characteristics of the patients. S group N (male/female) Age (yrs) Blood pressure Systolic (mmHg) Diastolic (mmHg) BMI (kg/m2) HbA1c (NGSP) (%) HbA1c (mmol/mol) GA (%) Fasting plasma glucose (mg/dL) IRI (mU/mL) HOMA-IR Plasma lipid LDL-cholesterol (mg/dL) HDL-cholesterol (mg/dL) Triglyceride (mg/dL) FFA (mmol/L) Adipo-cytokine HM-adiponectin (mmol/L) Leptin (mmol/L) HS-CRP (mmol/L) IHL (%) LBM (kg) Fat(kg) %-Fat (%) Other medications for diabetes, n (%) OHAs(a-GI/Met/TZD) Statins ARBs Calcium channel blocker Diuretics b-blocker
G group
p Value (S vs. G)
10 (6/4) 62 (56, 70)
10 (5/5) 55 (42, 62)
0.142
141 (134, 143) 85 (79, 90) 25.6 (24.7, 32.5) 7.2 (7.0, 7.5) 54 (53, 56) 18.3 (16.3, 19.7) 131 (126, 172) 8.6 (5.5, 9.9) 3.0 (1.7, 4.8)
131 (118, 140) 78 (74, 84) 26.6 (25.0, 32.4) 7.3 (6.8, 7.4) 55 (51, 56) 16.4 (16.0, 18.4) 138 (134, 143) 8.1 (5.6, 8.9) 2.6 (1.9, 3.0)
0.325 0.120 0.706 0.939 0.939 0.306 0.791 0.570 0.597
127 (115, 151) 48 (43, 53) 153 (119, 190) 0.6 (0.5, 0.8)
112 (108, 136) 50 (43, 56) 151 (111, 189) 0.6 (0.4, 0.7)
0.151 0.791 0.910 0.450
1.2 (0.7, 1.6) 12.6 (7.6, 14.6) 0.18 (0.05, 0.23) 24.5 (18.9, 36.6) 50.3 (46.5, 53.0) 22.5 (20.6, 33.7) 37.7 (29.6, 39.7)
0.8 (0.6, 3.1) 10.3 (6.0, 15.1) 0.19 (0.07, 0.28) 18.0 (10.8, 37.5) 53.2 (42.8, 60.2) 21.3 (18.9, 31.6) 35.3 (29.3, 36.6)
0.734 0.678 0.762 0.345 0.497 0.545 0.364
0(0)/6(60)/0(0) 3(30) 3(30) 1(10) 1(10) 1(10)
0(0)/5(50)/0(0) 4(40) 4(40) 3(30) 0(0) 1(10)
Data are the median (IQR); analyzed using Mann–Whitney U test. p < 0.05, S; sitagliptin, G; glimepiride BMI; body mass index, GA; glycoalbumin, IRI; immunoreactive insulin, HOMA-IR; homeostasis model assessment as an index of insulin resistance, LDL; low-density lipoprotein, HDL; high-density lipoprotein, FFA; free fatty acids, HMW; high molecular weight, HS; highly sensitive, IHL; intrahepatic lipid, LBM; lean body mass, OHAs; oral hypoglycemic agents, Met; metformin, TZD; thiazolidinedione.
202
diabetes research and clinical practice 109 (2015) 199–205
Table 2 – Changes of clinical data during treatment with sitagliptin or glimeplide. Group
baseline
12 weeks
p Value (vs. baseline)
p Value (D12 W S vs. D12 W G)
50.0 (25, 50) 0.5 0.5
24 weeks
p Value (vs. baseline)
p Value (D24 W S vs. D24 W G)
Drug dose (mg)
S G
50.0 (31,50) 0.5 0.5
BMI (kg/m2)
S G
25.6 (24.7, 32.5) 26.6 (25.0, 32.4)
25.9 (24.5, 31.8) 25.8 (24.9, 31.9)
0.799 0.959
0.821
25.7 (24.8, 32.1) 26.4 (24.9, 32.2)
0.386 0.241
0.496
HbA1c (NGSP)(%)
S G
7.2 (7.0, 7.5) 7.3 (6.8, 7.4)
6.7 (6.3, 6.9) 6.7 (6.4, 7.1)
0.005 0.028
0.096
6.6 (6.4, 6.8) 6.6 (6.3, 6.7)
0.005 0.011
0.151
HbA1c (NGSP)
S G
54 (53, 56) 55 (51, 56)
49 (46, 50) 49 47(, 52)
0.005 0.028
0.096
48 (47, 49) 48 (46, 49)
0.005 0.011
0.151
GA (%)
S G
18.3 (16.3, 19.7) 16.4 (16.0, 18.4)
15.7 (14.1, 17.1) 15.3 (14.5, 16.7)
0.007 0.022
0.450
14.4 (13.9, 15.9) 15.2 (13.5, 16.4)
0.011 0.028
0.597
FPG(mg/dL)
S G
131 (126, 172) 138 (134, 143)
128 (111, 142) 133 (125, 136)
0.037 0.037
0.450
124 (114, 145) 122 (113, 135)
0.075 0.044
0.880
IRI(mU/mL)
S G
8.6 (5.5, 9.9) 8.1 (5.6, 8.9)
7.5 (5.6, 9.9) 7.5 (5.6, 9.9)
0.221 0.541
0.257
9.1 (6.7, 11.5) 6.3 (4.0, 9.9)
0.839 0.759
0.496
HOMA-R
S G
3.0 (1.7, 4.8) 2.6 (1.9, 3.0)
2.3 (1.7, 3.2) 2.3 (1.8, 3.0)
0.139 0.114
0.290
2.8 (2.0, 3.3) 1.8 (1.0, 1.8)
0.333 0.374
0.31
LDL-C (mg/dL)
S G
132 (115, 151) 112 (108, 136)
127 (126, 104) 126 (111, 136)
0.139 0.678
0.227
133 (108, 159) 116 (103, 138)
0.647 0.759
0.130
HDL-C (mg/dL)
S G
48 (43, 53) 50 (43, 56)
47 (43, 50) 45 (41, 55)
0.185 0.636
0.369
46 (43, 51) 44 (41, 55)
0.285 0.919
0.650
TG (mg/dL)
S G
153 (119, 190) 151 (111, 189)
136 (93, 160) 160 (101, 208)
0.037 0.415
0.131
145 (114, 167) 126 (104, 215)
0.203 0.953
0.762
FFA (mmol/L)
S G
0.6 (0.5, 0.8) 0.6 (0.4, 0.7)
0.7 (0.5, 1.0) 0.6 (0.5, 0.7)
0.508 1.000
0.650
0.7 (0.4, 0.9) 0.7 (0.5, 0.8)
0.285 0.124
0.290
HM-adiponectin (mmol/L)
S G
1.2 (0.7, 1.6) 0.8 (0.6, 3.1)
1.2 (0.8, 2.0) 1.1 (0.8, 2.7)
0.684 0.333
0.762
1.3 (0.7, 1.9) 1.2 (0.7, 3.6)
0.139 0.124
0.821
Leptin (mmol/L)
S G
12.6 (7.6, 14.6) 10.3 (6.0, 15.1)
11.9 (8.8, 14.2) 12.2 (6.8, 15.1)
0.879 0.647
0.706
10.4 (8.3, 15.3) 11.0 (5.7, 16.9)
0.441 0.953
0.895
HS-CRP (mmol/L)
S G
0.18 (0.05, 0.23) 0.19 (0.07, 0.28)
0.05 (0.04, 0.11) 0.17 (0.03, 0.37)
0.009 0.859
0.041
0.09 (0.04, 0.15) 0.18 (0.03, 0.20)
0.059 0.859
0.191
Data are the median (IQR); analyzed using Wilcoxon signed-rank test, Mann–Whitney U test. p < 0.05 BMI; body mass index, GA; glycoalbumin, FPG; fasting plasma glucose, IRI; immunoreactive insulin, HOMA-IR; homeostasis model assessment as an index of insulin resistance, LDL; low-density lipoprotein, HDL; high-density lipoprotein, FFA; free fatty acids, HMW; high molecular weight, HS; highly sensitive, S; sitagliptin, G; glimepiride.
ultrasonography. Two patients did not have ultrasonography, but their BMI was >25 kg/m2. As shown in Fig. 1A, IHL was significantly decreased from baseline at 24 weeks in S group, but not in G group (S group: from 24.5 (18.9, 36.6) to 20.1 (14.6, 28.5)%, p = 0.009, G group: from 18.0 (10.8, 37.5) to 12.8 (10.2, 39.0)%, p = 0.959, median(IQR)). The percent change of IHL from baseline at 24 weeks was 15.1 ( 11.7, 18.8)% in S group and 1.6 ( 17.9, 20.1) % in G group (Fig. 1B). Fig. 2A displays the changes of total body fat mass at 24 weeks; there was a significant decrease in S group, but not in G group (S group: from 22.5 (20.6, 33.7) to 21.6 (19.7, 32.4) kg, p = 0.028, G group: from 21.2 (18.9, 31.6) to 22.6 (19.2, 31.1) kg, p = 0.241, media(IQR)). The percent change was greater in S group than in G group ( 4.7 ( 7.5, 0.0)% vs. 4.9 ( 2.3, 10.1)%, p = 0.007, median(IQR)). As shown in Fig. 2B, lean body mass did not differ from the baseline level at 24 weeks in both groups (S group: from 50.3 (46.5, 53.0) to 51.3 (46.8, 54.8) kg, p = 0.114, G group: from
53.2 (42.8, 60.2) to 50.9 (44.9, 57.2) kg, p = 0.799, median (IQR)). BMI showed no significant change in either group and there was no significant difference between the two groups at any point (S group: 25.6 (24.7, 32.5) at baseline, 25.9 (24.5, 31.8) at 12 weeks, and 25.7 (24.8, 32.1) at 24 weeks (kg/m2) median(IQR)), G group: 26.6 (25.0, 32.4) at baseline, 25.8 (24.9, 31.9) at 12 weeks, and 26.4 (24.9, 32.2) at 24 weeks (kg/ m2) median (IQR)). There were no significant differences between the two groups with regard to the number of subjects with concomitant metformin therapy and the doses (S group; n = 6, 750 (485, 875) mg, G group; n = 5, 1000 (750, 1500) mg, median(IQR)). The percent change of IHL from baseline at 24 weeks was not significantly different between patients with and without concomitant use of metformin (S group without metformin (n = 4) 12.8( 28.9, 3.2) % vs. S group with metformin (n = 6) 15.6( 18.1, 13.2)%, p = 0.999; G group without metformin (n = 5) 11.5( 29.8, 6.2)% vs. G group with metformin (n = 5) 17.3 (9.4, 21.0)%, p = 0.251,
diabetes research and clinical practice 109 (2015) 199–205
203
Fig. 1 – Absolute change (A) and percent change from baseline (B) of intrahepatic lipid (IHL) assessed by 1H-MRS during the 24-week study period.
median (IQR)). The percent change of total body fat mass from baseline at 24 weeks was also not significantly different between patients with and without concomitant use of metformin (S group without metformin (n = 4) 5.1 ( 6.8, 2.7)% vs. S group with metformin (n = 6) 3.9 ( 7.3, 0.02)%, p = 0.670; G group without metformin (n = 5) 1.7 ( 7.7, 2.6) % vs. G group with metformin (n = 5) 9.8 (7.1, 10.2)%, p = 0.175, median(IQR)). During the study period, there were no symptoms or signs of hypoglycemia and no FPG levels less than 70 mg/dl were recorded.
4.
Discussion
The present study demonstrated that the reduction of HbA1c and GA was similar after treatment with sitagliptin or lowdose glimepiride for 24 weeks, but only sitagliptin (not
glimepiride) significantly decreased IHL and total body fat mass. In the liver, TG are not only produced from free fatty acids (FFA) that are released from adipose tissue by lipolysis, but also by intrahepatic fatty acid synthesis from dietary glucose via the malonyl–CoA pathway. Intake of a carbohydrate (CH)restricted low energy diet (CH: 10% of total energy) for 48 h has been shown to markedly reduce IHL in obese subjects compared with an energy-matched high CH diet (CH: 65% of total energy) [18]. We also previously reported the effect of diet on lowering IHL in patients with type 2 diabetes or obese subjects with impaired glucose tolerance (IGT) [16,17,19]. However, patients did not change their dietary energy intake or nutritional balance throughout the present study and the improvement of glycemic control was similar between the two groups, suggesting that there was a direct effect of sitagliptin on IHL and body fat mass.
Fig. 2 – Changes of fat mass (A) and lean body mass (B) assessed by dual energy X-ray absorptiometry (DEXA) during the 24week study period.
204
diabetes research and clinical practice 109 (2015) 199–205
We previously reported that treatment with low-dose metformin (750 mg/day) was able to reduce IHL by 25% in obese Japanese subjects with T2DM [20]. Although the exact mechanism of IHL reduction was unclear, activation of hepatic AMP-activated protein kinase (AMPK) by metformin is thought to play a major role in suppressing intrahepatic TG production [21,22]. Hepatic AMPK is considered to be a key regulator of the intracellular energy balance by limiting anabolic pathways to prevent further ATP consumption and by facilitating catabolic pathways to increase ATP production. The activation of hepatic AMPK leads to increased b-oxidation of fatty acids along with simultaneous inhibition of lipogenesis and gluconeogenesis [23]. Recent studies have demonstrated that hepatic AMPK is also activated by GLP-1 in isolated rat hepatocytes [24,25], and DPP-4 deficient rats have a lower hepatic TG content compared with wild-type rats [24]. Cuthbertson et al. reported a marked reduction of IHL in human subjects treated with a GLP-1 analog, and they mentioned hepatic AMPK activation as a possible mechanism [8]. However, the peak plasma level of active GLP-1 after administration of sitagliptin to Japanese subjects was reported as 15–20 pmol/L [9], while injection of the GLP-1 analog liraglutide resulted in much higher levels (4000–8000 pmol/L) [10]. Therefore, the smaller reduction of IHL by sitagliptin in the present study compared with liraglutide in the above report (14% vs. 42%) may be partly explained by a difference of the plasma GLP-1 concentration. The incretin effect on fat cells in adipose tissue differs between GIP and GLP-1. While GIP enhances insulin-induced lipogenesis and inhibits the lipolytic action of glucagon [26], GLP-1 directly stimulates lipolysis through activation of both cAMP-dependent protein kinase (PKA) and AMPK in fat cells [27,28]. DPP-4I therapy simultaneously increases plasma GIP and GLP-1 levels, and Terauchi et al. have observed a similar increase of total GIP and active GLP-1 in Japanese subjects after sitagliptin administration [9]. While DPP-4I have a neutral effect on weight gain according to the ADA/EASD statement, a small but significant reduction of body fat was observed in the present study. Interestingly, recent studies of obese insulinresistant subjects have detected down-regulation of the GIP receptor and its signaling in subcutaneous fat cells (but not in visceral fat cells) compared with lean healthy controls, along with up-regulation of the GLP-1 receptor and lipolysis mediated by it in fat cells from both types of adipose tissue [28,29]. These reports may partly explain our results, but we did not perform analysis of fat cell receptors. A recent meta-analysis of glycemic control by DPP-4I revealed a relatively greater reduction of HbA1c in Asian than non-Asian subjects [3]. The authors pointed out that the baseline BMI was negatively correlated with the HbA1clowering effect of DPP-4I, and that the studies in which the average BMI was
