Long-term Effects of Ipragliflozin on Adipose Tissue in Japanese Patients with Obese Type 2 Diabetes

Objective and Methods : A long-term effect of ipragliflozin on adipose tissue mass reduction in Japanese patients with obese type 2 diabetes (T2D) was investigated. Ipragliflozin was administered (50 mg/day) once daily for 12 months. At 0, 3, 6 and 12 months, visceral and subcutaneous adipose tissue area was determined by two different bioelectrical impedance methods, and blood samples for HbA1c, renal function, lipids and liver function obtained, and body weight and blood pressure recorded. The primary endpoint was decrease in body fat mass. Secondary endpoints included changes in body weight and the laboratory data. Results : Seventeen of 20 participants (mean body mass index (BMI) 35.1±1.1 kg/m 2 ) completed this prospective observational study. Visceral fat area (cm 2 , mean±SD) at 0, 3, 6 and 12 months was 166.0±49.7, 149.7±46.1, 149.7 ±42.4 and 148.5±40.2, respectively : the value at 3 months was significantly lower than baseline (P＝0.027). Sub­ cutaneous fat at the corresponding time points was 359.3±110.5, 316.6±87.1, 326.8±87.2 and 325.9±90.4, respec­ tively : the values at each post treatment period was significantly less than the baseline (P＝0.003, 0.018 and 0.036 for the three points, respectively). Body weight was significantly reduced by 12 months (P＝0.045). Serum alanine aminotransferase (ALT), aspartate aminotransferase (AST), and γ-glutamyl transpeptidase (γ-GTP) lev­ els decreased significantly. There were no significant correlations between serum hepatobiliary enzyme levels and Δbody weight or Δvisceral fat, but Δγ-GTP was correlated with Δsubcutaneous fat (Spearman’s P＝ 0.004). Conclusion : During the 1 year interval, ipragliflozin significantly reduced subcutaneous adipose tissue and serum AST, ALT, and γ-GTP levels. Shinshu Med J 66 : 29―37, 2018

pose tissue lipolysis and enhanced lipid metabolism 6) .
Ipragliflozin reduced body fat in rats 7) , and clinical reports indicated visceral fat reduction by ipragli flozin in Asian people. Ipragliflozin significantly decreased visceral adipose tissue in 4-week observa tion in 25 Japanese T2D patients 8) . Visceral fat area was significantly reduced in 6-month observation of 64 diabetic patients 9) . However, obesity is difficult to treat as many subjects regain weight after tempo rary weight loss 10) , so it is important to observe re search on obesity for a long term. Body fat reduction by sodium-glucose cotransporter 2 (SGLT2) inhibi tors has been observed for up to 1 year in Caucasian patients 11) . Nonetheless, there have been no long-term (12-month) surveys of SGLT2 inhibitors in obese T2D patients in Japan.
Therefore, we examined long-term effects of ip ragliflozin in obese T2D patients and evaluated the influence of ipragliflozin on adipose tissue and liver function by monitoring changes in serum AST, ALT, and γ-GTP levels.

A Ethics
This study conformed to the Declaration of Helsin ki, received approval from our university ethics com mittee (Study no. 3049), and subjects provided writ ten informed consent before participation.

E Clinical parameter measurement
The following variables were monitored before and at 3, 6, and 12 months after commencement of ipragliflozin treatment : HbA1c, body weight, BMI, estimated visceral fat area, estimated subcutaneous fat area, systolic blood pressure, diastolic blood pres sure, serum ALT, AST, γ-GTP, blood urea nitrogen, creatinine, uric acid, glomerular filtration rate (eGFR), and serum levels of low-density lipoprotein choles terol (LDL-C), and high-density lipoprotein choles terol (HDL-C). The data were collected before or≥2 hours after a meal. There were variations in time after meals, so we did not evaluate blood glucose level and TG value. All P-values for comparison before and after admin istration were subjected to Bonferroni adjustment.
Relationships between changes in adipose tissue and ALT, AST, and γ-GTP were assessed using Spear man's rank correlation coefficients. In all analyses, P<0.05 was taken to indicate statistical significance. Ⅲ

Results
Twenty T2D patients were enrolled in this study ( Table 1). Three subjects dropped out : one failed to be evaluated for urinary tract infections, and the other two discontinued ipragliflozin due to develop ment of eruptions within 1 week after administra tion. Seventeen subjects completed the full protocol and were included in statistical analyses. Antidiabet ic drugs other than SGLT2 inhibitors are shown in Table 2. Table 3 shows changes in test items every 3 months during ipragliflozin treatment.
Mean body weight decreased over the observation period ( Table 3,  Systolic and diastolic blood pressures tended to decrease during the treatment period (not signifi cant) ( Table 3).
Changes in various clinical parameters with ad ministration of ipragliflozin are shown in Table 3.
Mean HbA1c level improved, and decreased signifi cantly from 7.7 % at baseline to 7.2 % at 6 months (P＝0.032) ; there was no significant difference at 12 months. Blood urea nitrogen and creatinine showed no significant changes. Mean uric acid level de crea sed significantly from 6.4 mg/dL at baseline to 5.1 mg/dL at 6 months (P＝0.030) ; however, there was no significant difference at 12 months. eGFR re mained unchanged during the 12 months of treat ment ; it tended to increase during the treatment period, but the differences were not statistically significant. AST, ALT, and γ-GTP levels at the end of treatment were significantly decreased compared to baseline (Table 3, Fig. 4). Serum LDL-C tended to decrease, while serum HDL-C tended to increase, but the differences were not significant ( Table 3).
There were no correlations between changes in ALT or AST levels and those in visceral adipose tissue (Fig. 5A, B) or subcutaneous adipose tissue ( Fig. 5D, E). Change in γ-GTP levels was not related to those in visceral adipose tissue (Fig. 5C), but correlated with change in subcutaneous adipose tis sue volume (Spearman's ρ＝0.664, P＝0.004) (Fig.   5F).        5 Correlations between changes in serum hepatobiliary enzyme levels and visceral adipose tissue and subcutaneous adipose tissue after 12 months of ipragliflozin treatment The relationships between changes in adipose tissue and hepatobiliary enzymes were assessed using Spearman's rank correlation coefficients. In all analyses, P<0.05 was taken to indicate statistical significance. A : Correlation between changes in serum ALT levels and visceral adipose tissue (n＝17). B : Correlation between changes in serum AST levels and visceral adipose tissue (n＝17). C : Correlation between changes in serum γ-GTP levels and visceral adipose tissue (n＝17). D : Correlation between changes in serum ALT levels and subcutaneous adipose tissue (n＝17). E : Correlation between changes in serum AST levels and subcutaneous adipose tissue (n＝17). F : Correlation between changes in serum γ-GTP levels and subcutaneous adipose tissue (n＝17). The change in subcutaneous adipose tissue was significantly correlated with that in serum γ-GTP level. Oral SGLT2 inhibitor administration reduces body weight. Weight loss was reported with oral dapa gliflozin administration 15) , and with short-term (10 days) ipragliflozin treatment 2) . Japanese subjects giv en oral ipragliflozin showed body weight re duction of 3.3 % in 16 weeks 8) . In our study, 12-month ipra gliflozin administration resulted in weight loss of 3.6 %. The mean BMI in Yamamoto et al.'s cohort 8) was 28.9 kg/m 2 , while that in our study was 35.1 kg/ m 2 . Moreover, 65 % of subjects in our study were se verely obese. Nevertheless, there was no difference in the weight reduction effect, suggesting that ip ragliflozin is also effective in severely obese patients.
SGLT2 inhibitors were reported to reduce visceral fat and weight. Here, visceral fat was reduced by 7.8 % in 3 months, which was similar to the previous report of 8.2 % reduction by 16 weeks of 50 mg ip ragliflozin 8) . A visceral fat loss trend was recognized after 6 months but was not significant. Visceral fat reduction by 8.1 % with 300 mg of canagliflozin for 52 weeks was reported 11) . The lack of significant dif ference in our study was probably due to the small sample size. Subcutaneous fat decreased at all time points. Body weight decrease may be attributed to visceral fat tissue lipolysis due to SGLT2 inhibitor induced enhancement in lipid metabolism 6) . Long-term empagliflozin treatment significantly reduced weight of subcutaneous but not visceral fat in rats 16) .
The authors concluded that the decrease in body weight of rats treated with SGLT2 inhibitor was due to a decrease in subcutaneous rather than visceral fat. Our results were consistent with this previous study. They also showed SGLT2 inhibitor signifi cantly reduced the size of visceral adipocytes and increased the number of smaller size adipocytes, which was associated with the attenuation of oxida tive stress. Detailed analyses of body fat content contributing to weight loss in humans are necessary.
Canagliflozin improved liver dysfunction in pa tients with T2D, assessed by monitoring serum AST, ALT, and γ-GTP levels 16) . Ogawa and colleagues reported that ipragliflozin improved liver function in clinical and basic research 17) . They showed that This study had several limitations. First, the sam ple size was small. We think that the mean HbA1c level at 12 months did not differ from baseline be cause the sample size was small. Second, patients using drugs that affect lipid metabolism (insulin, pi oglitazone, and GLP1 analogs) were included. Third, this was a single-arm study. However, the number of participants met the minimum requirement for a prospective observational study, and concurrent use of antidiabetic drugs is inevitable in clinical practice.

Ⅴ Conclusion
Where proper dietary and nutritional guidance are provided, administration of the SGLT 2 inhibitor, ipragliflozin, induced weight loss and subcutaneous fat reduction over 12 months. Furthermore, ipragli flozin also improved liver function in obese patients, which was correlated with decrease of subcutaneous fat.