Efficacy of Adding Once- and Thrice-Daily Voglibose in Japanese Type 2 Diabetic Patients Treated with Alogliptin

Endocrine Journal 2014, 61 (5), 447-456

original Efficacy of adding once- and thrice-daily voglibose in Japanese type 2 diabetic patients treated with alogliptin

Mitsuyoshi Takahara1), Toshihiko Shiraiwa2), Naoto Katakami1), Hideaki Kaneto1), Taka-aki Matsuoka1) and Iichiro Shimomura1)

1) Department of Metabolic Medicine, Osaka University Graduate School of Medicine, Osaka, Japan 2) Shiraiwa Medical Clinic, Osaka, Japan

Abstract. We investigated the efficacy of once- and thrice-daily voglibose, an alpha-glucosidase inhibitor, as an add-on therapy to alogliptin, a dipeptidyl peptidase-4 inhibitor, on glycemic control in Japanese type 2 diabetic patients. In this 12-week, parallel-group, randomized, open-label, three-arm trial, 151 participants treated with alogliptin were randomly allocated to the following three arms; one was the group to initiate once-daily voglibose, another was to initiate thrice daily voglibose, and the other was the control group. The primary endpoint was the change of hemoglobin A1c levels at the end of the study, which was revealed to be significantly different among groups (p < 0.001). The once- and thrice-daily voglibose groups had a significantly greater reduction than the control group; the difference was -0.27% and -0.33% in the once- and thrice-daily voglibose group, respectively (both p < 0.001). No significant difference was observed between the two voglibose groups (p = 0.615). On the other hand, the increase of 1,5-anhydroglucitol levels were 3.3 and 5.5 μg/mL greater in the once- and thrice-daily voglibose groups than the control group (both p < 0.001). The thrice-daily voglibose group had a greater increase of 1,5- anhydroglucitol levels compared to the once-daily voglibose group (p = 0.005). In conclusion, once- and thrice-daily voglibose as an add-on to alogliptin significantly improved glycemic control in Japanese type 2 diabetic patients.

Key words: Alpha-glucosidase inhibitor, Dipeptidyl peptidase-4 (DPP-4) inhibitor, Once- and thrice-daily administration

Numerous studies have reported the efficacy of Recently, a favorable effect of alpha-glucosidase dipeptidyl peptidase-4 (DPP-4) inhibitors on glycemic inhibitors as an add-on to DPP-4 inhibitors has been control in type 2 diabetes mellitus [1-15]. However, suggested by some clinical studies [30-32]. In addi- in the clinical setting, some patients receiving DPP-4 tion, it is also suggested that although the agents are inhibitors fail to achieve strict glycemic control [16-19], usually administered thrice a day, only a once-daily and they often require an add-on therapy of another oral administration might have some efficacy on glycemic hypoglycemic agent. Although previous studies demon- control [33]. However, all of these studies were per- strated the efficacy of the combined use of DPP-4 inhib- formed for only a several days. Their longer-term glu- itors with other oral hypoglycemic agents, most studies cose-lowering effect, e.g., the change of hemoglobin focused on the efficacy of DPP-4 inhibitors as an add-on A1c levels, remains unknown. to other agents [20-29]. To date, few data are available The current study therefore investigated the impact about the glucose-lowering effect of other agents as an of 12-week administration of voglibose, an alpha-glu- add-on to DPP-4 inhibitors. It remains unknown to cosidase inhibitor, as an add-on to alogliptin, a DPP-4 what extent hemoglobin A1c levels will be decreased inhibitor, on glycemic control in Japanese type 2 dia- after adding other agents to DPP-4 inhibitors. betic patients.

Submitted Nov. 5, 2013; Accepted Jan. 27, 2014 as EJ13-0466 materials and methods Released online in J-STAGE as advance publication Feb. 22, 2014 Correspondence to: Mitsuyoshi Takahara, M.D., Ph.D., Department of Metabolic Medicine, Osaka University Graduate School of Study design Medicine, 2-2 Yamadaoka, Suita, Osaka 565-0871, Japan. This 12-week, parallel-group, randomized, open-la- E-mail: [email protected] bel, three-arm trial assessed the efficacy and safety of

©The Japan Endocrine Society 448 Takahara et al. once- and thrice-daily voglibose as an add-on to alo- Statistical analysis gliptin in Japanese type 2 diabetic patients. The cur- Data are given as means and standard deviations rent study was conducted at Shiraiwa Medical Clinic, (SD) for continuous variables, or as numbers and per- Osaka, Japan, between March 2013 and September centages for discrete variables, if not otherwise men- 2013. We performed the current study in accordance tioned. A p value less than 0.05 was considered to be with the Declaration of Helsinki, and it was approved significant. The difference among three groups was by the local ethics committee. Written informed con- assessed by the one-way analysis of variance (ANOVA) sent was obtained from every participant in the current for continuous variables, and the chi-square test for dis- study. The trial was registered as UMIN000009516. crete variables. Post hoc intergroup comparison was performed using the Turkey’s honestly significant dif- Study population and procedures ference (HSD) test for continuous variables, and the The study population comprised male and female Fisher’s exact test with the Bonferroni’s correction for Japanese adults (≥ 20 years old) with type 2 diabetes dichotomous variables. The adherence to the adminis- mellitus, who were treated with 25 mg per day of alo- tration of voglibose was evaluated by the generalized gliptin. The exclusion criteria were: difficulty of regu- linear mixed model in which the time of medication lar visit, severe hepatic, and/or renal disease, and preg- (i.e., before breakfast, lunch, or dinner) as well as the nant or breast feeding females. intervention arms (i.e., the thrice- or once-daily voglib- Eligible participants were randomly allocated to the ose group) were included as the fixed effects. following three arms. One arm was to initiate 0.2-mg The sample size of at least 45 per group was enough voglibose three times a day before meals (thrice-daily to detect the difference of 0.2% in hemoglobin A1c lev- voglibose group), whereas another was to initiate els, with SD of 0.25%, a two-sided significance level of 0.2-mg voglibose once a day, before the very meal at 5%, and a power of 80%. The sample size was also which alogliptin was administered (once-daily boglib- enough to detect the difference of 0.8% in glycoalbu- ose group). The other arm was the control group with min levels with SD of 1.0%, and 3.0 μg/mL in 1,5-AG no medication initiated. Alogliptin was continued, levels with SD of 4.0 μg/mL. without change in dose, in all the three arms. All partic- In the current study, we additionally investigated ipants in the current study were asked to visit the clinic whether there were any clinical baseline parameters 4, 8, and 12 weeks after the allocation. Hemoglobin associated with the response to add-on of voglibose. A1c, glycoalbumin, and 1,5-anhydroglucitol (1,5-AG) The impact of clinical parameters on the response to levels were measured at each visit. Data on side effects voglibose was assessed by the interaction effect with were collected to assess safety. the intervention arms on the primary efficacy endpoint (i.e., the change of hemoglobin A1c levels) in the lin- Study outcomes ear regression model. We also performed the logis- The primary efficacy endpoint was the change of tic regression analysis to investigate the association hemoglobin A1c levels at the end of the study, using of clinical baseline parameters with the prevalence of the last observation carried forward (LOCF) data anal- observed side effects. ysis. Secondary efficacy endpoints were the change of glycoalbumin and 1,5-AG levels at the end of the results study, using the LOCF data analysis. Safety was evaluated with the prevalence of observed side effects. The A total of 156 patients were randomly allocated. change of body weight and serum transaminase lev- After the allocation, one patient withdrew the partic- els, i.e., aspartate aminotransferase (AST) and alanine ipation, and another was lost to follow up before the aminotransferase (ALT) levels, were also assessed. In first visit. Finally, the full analysis set comprised the the current study we additionally evaluated the adher- remaining 154 patients (49 patients in the thrice-daily ence to the administration of voglibose in the thrice- voglibose group, 50 patients in the once-daily voglib- and once-daily voglibose groups. The adherence was ose group, and 55 patients in the control group). They judged when the extent to which patients took vogli- were 64 ± 12 years old and 92 patients (60%) were bose as prescribed was 90% or more. The data were male. The hemoglobin A1c levels were 6.8 ± 0.5% at based on their self-reports. the initiation of the study and 6.8 ± 0.5% one month Addition of volibose to alogliptin 449

Table 1 Baseline characteristics of the study population Thrice-daily Once-daily Control group voglibose group voglibose group p value (n = 55) (n = 49) (n = 50) Male 29 (59%) 29 (58%) 34 (62%) 0.919 Age (years) 62 ± 11 65 ± 12 65 ± 11 0.377 Body mass index (kg/m2) 24.5 ± 4.1 24.8 ± 3.6 24.2 ± 4.0 0.758 Body weight (kg) 62.3 ± 17.2 63.8 ± 10.5 61.9 ± 15.0 0.783 Hemoglobin A1c (%) at the study initiation 6.7 ± 0.4 6.8 ± 0.6 6.7 ± 0.5 0.844 one month before 6.7 ± 0.5 6.8 ± 0.6 6.7 ± 0.4 0.738 change for the one month 0.0 ± 0.1 0.0 ± 0.2 0.0 ± 0.1 0.592 Glycoalbumin (%) 17.6 ± 2.8 17.9 ± 2.7 18.2 ± 2.6 0.502 1,5-anhydroglucitol (μg/mL) 13.9 ± 7.1 13.5 ± 8.3 12.5 ± 5.9 0.594 Sulfonylurea use 1 (2%) 5 (10%) 4 (7%) 0.263 Metformin use 16 (33%) 17 (34%) 15 (27%) 0.731 Pioglitazone use 1 (2%) 1 (2%) 3 (5%) 0.515 Hypertension 12 (24%) 21 (42%) 18 (33%) 0.180 Dyslipidemia 24 (49%) 20 (40%) 25 (45%) 0.663 Administration of alogliptin 0.341 at breakfast 42 (86%) 43 (86%) 52 (95%) at lunch 0 (0%) 1 (2%) 0 (0%) at dinner 7 (14%) 6 (12%) 3 (5%) Data are mean ± SD or n (percentage).

before; the difference was 0.0 ± 0.1%, with no statistical significance (p = 0.555). There were no significant differences in baseline characteristics among the three arms (Table 1).

Efficacy assessment Fig. 1 shows the change of hemoglobin A1c levels during the study period. At the end of the study, the change was significantly different among groups (p < 0.001 by the one-way ANOVA). The post-hoc Tukey’s HSD test revealed that the once- and thrice-daily voglibose groups had a significantly greater reduction of hemo- Fig. 1 Change of hemoglobin A1c levels during the study period globin A1c levels than the control group; the difference Data are the change of hemoglobin A1c levels from from the control group was -0.27% (95% confidence baseline as well as their SD. Asterisks indicate p < 0.05 tested by the Tukey’s HSD method, while n.s. represents interval (CI): -0.40% to -0.15%) in the once-daily vogli- “not significant”. bose group, and -0.33% (95% CI: -0.45% to -0.20%) in the thrice-daily voglibose group (both p < 0.001). The change of hemoglobin A1c levels in the thrice-daily had a significantly greater reduction of glycoalbumin voglibose group was not statistically different from that levels than the control group, with the difference of in the once-daily voglibose group; the difference was -1.0% (95% CI: -1.6% to -0.4%, p = 0.001) and -1.2% -0.05% (95% CI: -0.18% to 0.08%, p = 0.615). (95% CI: -1.8% to -0.7%, p < 0.001), respectively. The change of glycoalbumin and 1,5-AG levels is There was no significant difference between the once- shown in Fig. 2. At the end of the study, the change of and thrice daily voglibose group -0.3% (95% CI: -0.9% both measurements was significantly different among to 0.3%, p = 0.496). On the other hand, as shown in groups (p < 0.001 by the one-way ANOVA). As shown Fig. 2B, the once- and thrice-daily voglibose group had in Fig. 2A, the once- and thrice-daily voglibose group a significantly greater increase of 1,5-AG levels than 450 Takahara et al.

Fig. 2 Change of glycoalbumin and 1,5-AG levels during the study period Data are the change of glycoalbumin (A) and 1,5-AG levels (B) from baseline as well as their SD. Asterisks indicate p < 0.05 by the Tukey’s HSD method, while n.s. represents “not significant”.

Table 2 Observed side effects in the study period Thrice-daily Once-daily Control voglibose group voglibose group group p value Side effects 16 (33%)* 10 (22%)* 0 (0%) < 0.001 Flatulence 5 (10%)* 3 (6%) 0 (0%) 0.062 Constipation 5 (10%)* 2 (4%) 0 (0%) 0.044 Abdominal distension 4 (8%)* 0 (0%) 0 (0%) 0.012 Loose stools 2 (4%) 1 (2%) 0 (0%) 0.323 Borborygmus 1 (2%) 1 (2%) 0 (0%) 0.569 Dizziness 2 (4%) 1 (2%) 0 (0%) 0.323 Nausea with stomachache 0 (0%) 1 (2%) 0 (0%) 0.351 Decreased appetite 1 (2%) 0 (0%) 0 (0%) 0.340 Others 2 (4%) 3 (6%) 0 (0%) 0.148 Data are numbers (percentages) of the participants with side effects. Asterisks indicate that the post-hoc analysis revealed a significant intergroup difference compared to the control group. Note that thepost-hoc analysis failed to demonstrate a significant difference between the thrice- and once-daily voglibose groups in any of the side effects. Others included increased appetite, thirst, vertigo, and daze (all n = 1).

the control group, with the difference of 3.3 μg/mL vertigo (n = 1), although the prevalence was not statis- (95% CI: 1.7 to 4.9 μg/mL) and 5.5 μg/mL (95% CI: tically different among groups (p = 0.122). 3.9 to 7.1 μg/mL) (both p < 0.001). Furthermore, the The change of body weight and the fold-change of the change of 1,5-AG levels in the thrice-daily group transaminase levels were not significantly different was significantly higher than that in the once-daily among groups (Table 3). group, with the difference of 2.2 μg/mL (95% CI: 0.6 to 3.9 μg/mL, p = 0.005). Adherence to administration of voglibose Adherence to administration of voglibose was 96% Safety and body weight change (47/49) before breakfast, 73% (36/49) before lunch, and Table 2 summarizes side effects observed during 94% (46/49) before dinner in the thrice-daily voglib- the study period. Most of the side effects were gas- ose group, and 91% (39/43) before breakfast, 0% (0/1) trointestinal symptoms. Side effects were more likely before lunch, and 83% (5/6) before dinner in the once- observed in the voglibose groups, with statistical sig- daily voglibose group. Compared to the administration nificance. Two patients in the once-daily voglibose before breakfast, that before lunch, but not before din- group, but none in the other groups, voluntarily discon- ner, was significantly associated with a lower - adher tinued voglibose because of constipation (n = 1) and ence (p = 0.022 and 0.518, respectively). No signifi- Addition of volibose to alogliptin 451

Table 3 Change of body weight and transaminase levels during the study period Thrice-daily Once-daily Control voglibose group voglibose group group p value (n = 49) (n = 50) (n = 55) Body weight change (kg) at 4 weeks -0.2 ± 1.2 -0.7 ± 4.3 -0.1 ± 0.8 0.503 at 8 weeks -0.8 ± 1.4 -1.5 ± 6.1 -0.4 ± 1.2 0.332 at 12 weeks -0.8 ± 2.1 -0.5 ± 1.3 -0.7 ± 1.4 0.725 LOCF -0.9 ± 2.0 -1.1 ± 4.4 -0.6 ± 1.3 0.686 Fold change of AST levels at 4 weeks 1.01 ± 0.05 1.01 ± 0.05 1.01 ± 0.05 0.904 at 8 weeks 1.01 ± 0.06 1.00 ± 0.05 1.00 ± 0.05 0.465 at 12 weeks 1.02 ± 0.07 1.01 ± 0.05 1.00 ± 0.52 0.467 LOCF 0.99 ± 0.06 1.01 ± 0.07 1.01 ± 0.05 0.210 Fold change of ALT levels at 4 weeks 1.10 ± 0.15 1.13 ± 0.16 1.12 ± 0.17 0.797 at 8 weeks 1.11 ± 0.18 1.13 ± 0.15 1.10 ± 0.16 0.785 at 12 weeks 1.09 ± 0.16 1.12 ± 0.17 1.09 ± 0.16 0.611 LOCF 1.08 ± 0.14 1.12 ± 0.17 1.11 ± 0.17 0.459 Data are mean ± SD.

Table 4 Clinical parameters influencing the association of voglibose administration with the change of hemoglobin A1c levels Thrice-daily voglibose Once-daily voglibose Male -0.058 (p = 0.607) -0.116 (p = 0.276) Age (years) -0.001 (p = 0.812) -0.002 (p = 0.619) Body mass index (kg/m2) -0.002 (p = 0.863) 0.005 (p = 0.740) Hemoglobin A1c (%) -0.209 (p = 0.065) -0.105 (p = 0.218) Glycoalbumin (%) -0.051 (p = 0.012) -0.044 (p = 0.024) 1,5-anhydroglucitol (μg/ml) 0.011 (p = 0.222) 0.017 (p = 0.025) Sulfonylurea use 0.109 (p = 0.737) -0.004 (p = 0.983) Metformin use 0.011 (p = 0.929) -0.066 (p = 0.562) Pioglitazone use -0.507 (p = 0.129) -0.048 (p = 0.877) Hypertension -0.173 (p = 0.152) 0.045 (p = 0.675) Dyslipidemia 0.101 (p = 0.367) -0.021 (p = 0.839) Alogliptin administered at breakfast -0.217 (p = 0.250) -0.174 (p = 0.324) Data are partial regression coefficients of the interaction term between the clinical parameters and the intervention arms (the thrice- or once-daily voglibose group vs. the control group), as well as their p values.

cant association was observed between the intervention globin A1c levels, although the impact was not statisti- arm and the adherence (p = 0.340). cally significant (p for interaction = 0.065). No other baseline parameters had a significant interaction effect Association of clinical parameters with the response on the efficacy of thrice-daily voglibose (Table 4). to voglibose On the other hand, once-daily voglibose administration Table 4 shows the impact of clinical baseline param- brought a greater reduction of hemoglobin A1c levels eters on the association of voglibose administration in patients with higher glycoalbumin levels and with with the change of hemoglobin A1c levels. Thrice- lower 1,5-anhydroglucitol levels (p for interaction = daily voglibose administration brought a greater reduc- 0.024 and 0.025, respectively). No baseline param- tion of hemoglobin A1c levels in patients with higher eters were associated with the presence of any side glycoalbmin levels (p for interaction = 0.012). A simi- effects (all p > 0.05) (data not shown). lar tendency was observed in those with higher hemo- 452 Takahara et al.

Discussion adding once-daily rapid-acting insulin to ongoing long- acting insulin did lower hemoglobin A1c levels in type The current parallel-group, randomized, open-la- 2 diabetic patients [35, 36]. In addition, it was demon- bel, three-arm study demonstrated that the introduction strated that once-daily rapid acting insulin as an add-on of once- and thrice-daily 0.2 mg of voglibose brought to long-acting insulin brought similar hemoglobin A1c a significant reduction of hemoglobin A1c levels in levels compared to twice- and thrice-daily rapid-act- Japanese type 2 diabetic patients treated with 25 mg ing insulin [37]. These reports indicate that correcting of alogliptin. No significant difference was observed even one glucose excursion would lead to a consider- in the change of hemoglobin A1c levels between the able reduction in hemoglobin A1c levels. Some clini- once- and thrice-daily voglibose group, whereas the cal studies using premixed insulin might also support thrice-daily voglibose group had a greater increase of the efficacy of a minimized correction of postprandial 1,5-AG levels compared to the once-daily voglibose hyperglycemia [38-41]. Voglibose, an alpha-glucosi- group. Most of the side effects observed in the voglib- dase inhibitor, suppress postprandial glucose excur- ose groups were gastrointestinal symptoms. sions right after the administration, by delaying carbo- DPP-4 inhibitors are now widely used in clinical hydrate absorption from the intestine [42]. The effect practice. The monotherapy of the agents has a consid- of once-daily voglibose on lowering hemoglobin A1c erable glucose-lowering effect without hypoglycemia levels observed in the current study would come from or weight gain, and have become one of the first-line the correction of one postprandial glucose excursions. medications in the management of type 2 diabetes [1-8]. Second, the combined use of alogliptin might favor- However, some patients receiving DPP-4 inhibitors fail ably modulate the glucose-lowering effect of voglib- to achieve sufficient glycemic control. Although sev- ose, via enhancing its effect on incretin, especially eral studies assessed the efficacy of DPP-4 inhibitors as glucagon-like peptide 1 (GLP-1) [30, 31, 43, 44]. an add-on therapy to alpha-glucosidase inhibitors [28, Alpha-glucosidase inhibitors prolong and enhance the 29, 34], few data are so far available about the efficacy secretion of GLP-1 from the intestine [45-50], whereas of alpha-glucosidase inhibitors as an add-on to DPP-4 DPP-4 inhibitors suppress the inactivation of the inhibitors. To the best of our knowledge, this is the first secreted incretin. The co-use of alogliptin might max- report investigating the 12-week efficacy of alpha-glu- imize the effect of once-daily voglibose on incretin. cosidase inhibitors as an add-on to DPP-4 inhibitors in Indeed, one recent clinical study indicates that once- type 2 diabetic patients. daily alpha-glucosidase inhibitors before breakfast In the current study, the once-daily voglibose group would increase GLP-1 levels and possibly decrease had a significantly greater reduction of hemoglobin postprandial glucose levels, not only after breakfast A1c levels than the control group. In addition, inter- but also after lunch in type 2 diabetic patients treated estingly, the reduction was almost similar and lacked with DPP-4 inhibitors [33], although their intervention a significant difference between the once- and thrice- period was only 2 days. daily voglibose groups (Fig. 1). The similar results Third, a similar reduction in hemoglobin A1c lev- were observed in glycoalbumin levels (Fig. 2A). els between once- and thrice-daily voglibose might Thrice-daily, i.e., frequent administration schedule is come from the limitation of hemoglobin A1c as a one major disadvantage of alpha-glucose inhibitors. It marker of glycemic control. Hemoglobin A1c lev- would be attractive in clinical practice that a less fre- els in general represent the average of daily glycemic quent administration could bring a similar effect in profiles, and postprandial glucose excursions are not lowering hemoglobin A1c levels. always accurately reflected by hemoglobin A1c levels It remains unclear why once-daily voglibose dem- [51]. Indeed, 1,5-AG levels, more sensitive to glucose onstrated a considerable impact on the reduction of excursions [51], were higher in the thrice-daily voglib- hemoglobin A1c levels similarly to thrice-daily voglibose group than in the once-daily voglibose group (Fig. ose. However, several possible explanations would be 2B). This finding indicates that thrice-daily voglibose given. First, as some clinical studies on insulin therapy would more strictly correct postprandial glucose excur- indicated, the correction of at least one glucose excursions compared to once-daily voglibose. The measure- sion might be actually effective in lowering hemoglo- ments of hemoglobin A1c might be unable to detect bin A1c levels. Some clinical trials demonstrated that this difference. Addition of volibose to alogliptin 453

In the current study, most of the side effects observed There were some limitations in the current study. in the voglibose groups were gastrointestinal symp- First, the study period was limited to 12 weeks and toms, e.g., flatulence, constipation, and abdominal dis- therefore a longer-term efficacy of once- and thrice- tension, as was previously reported [52, 53]. The sup- daily voglibose remains unknown. Second, daily plementary analysis showed that baseline parameters blood glucose profiles were not assessed in the current were not associated with side effects, indicating the dif- study. Future studies examining the profiles will be ficulty to predict the presence of side effects by base- need to validate the glucose-lowering effect of once- line parameters. The once-daily voglibose group had a and thrice-daily voglibose, which was evaluated using relatively low frequency of gastrointestinal symptoms hemoglobin A1c, glycoalbumin, and 1,5-AG levels in compared to the thrice-daily voglibose group, although the current study. Third, we did not assessed insulin or two patients in the once-daily voglibose group, but incretin levels. Detailed data on glucose metabolism none in the thrice-daily group, voluntarily discontin- were therefore unrevealed. Fourth, the patients’ satis- ued voglibose because of the sides effects. Future faction with the treatment was not surveyed. Previous studies with a larger population size will be needed to studies revealed that the treatment satisfaction was pos- reveal if the prevalence of side effects would be differ- itively associated with the glucose-lowering effect of ent between the once- and thrice-daily administration the medications, whereas it was inversely associated of voglibose. with the presence of side effects [56-58]. The admin- In the current study, the adherence analysis sug- istration of voglibose was associated with both glu- gested that as a whole, patients were less likely to cose-lowering effect and the presence of side effects. be adherent to the administration of voglibose before It remained to be investigated whether the treatment lunch. For the avoidance of non-adherence, it might be would satisfy the patients. better to prescribe voglibose before breakfast or dinner, In conclusion, once- and thrice-daily voglibose as but not before lunch, if it is to be administered once an add-on to alogliptin significantly lowered hemoglo- daily, although the current study sample size was too bin A1c levels in type 2 diabetic patients. There was small to conclude that. Future prospective studies will no significant difference in the change of hemoglobin be needed to validate this hypothesis. A1c levels between the once- and thrice-daily voglib- The additional analysis showed that poorer glycemic ose groups, although the change of 1,5-AG levels were control at baseline was associated with a greater reduc- significantly higher in the thrice-daily voglibose group tion of hemoglobin A1c levels after voglibose admin- than in the once-daily voglibose group. istration (Table 4). These findings would be consistent with previous studies on hypoglycemic agents, which Acknowledgements demonstrated a greater improvement of glycemic control in those with poorer glycemic control at baseline Mitsuyoshi Takahara is a Research Fellow of the [54, 55]. No other baseline characteristic was associ- Japan Society for the Promotion of Science. There ated with the effect of voglibose on glycemic control in is no conflict of interest concerning this manuscript. the current study, indicating a similar efficacy of vogli- The authors thank Akane Seo, Osaka Branch, Takeda bose across background characteristics except baseline Pharmaceutical Company Limited, Osaka, Japan, for glycemic control. her help in retrieving drug information and literature.

references

1. Pratley RE, Jauffret-Kamel S, Galbreath E, Holmes tidase-4 inhibitor sitagliptin as monotherapy in patients D (2006) Twelve-week monotherapy with the DPP-4 with type 2 diabetes mellitus. Diabetologia 49: 2564- inhibitor vildagliptin improves glycemic control in sub- 2571. jects with type 2 diabetes. Horm Metab Res 38: 423- 3. Aschner P, Kipnes MS, Lunceford JK, Sanchez M, 428. Mickel C, et al. (2006) Effect of the dipeptidyl pepti- 2. Raz I, Hanefeld M, Xu L, Caria C, Williams-Herman D, dase-4 inhibitor sitagliptin as monotherapy on glycemic et al. (2006) Efficacy and safety of the dipeptidyl pep- control in patients with type 2 diabetes. Diabetes Care 454 Takahara et al.

29: 2632-2637. T, et al. (2013) Addition of sitagliptin to ongoing met- 4. Scott R, Wu M, Sanchez M, Stein P (2007) Efficacy and formin monotherapy improves glycemic control in tolerability of the dipeptidyl peptidase-4 inhibitor sita- Japanese patients with type 2 diabetes over 52 weeks. J gliptin as monotherapy over 12 weeks in patients with Diabetes Invest. doi:10.1111/jdi.12116. type 2 diabetes. Int J Clin Pract 61: 171-180. 16. Nakamura A, Terauchi Y (2013) Factors associated 5. Pi-Sunyer FX, Schweizer A, Mills D, Dejager S (2007) with the glucose-lowering effect of vildagliptin identi- Efficacy and tolerability of vildagliptin monotherapy in fied from the results of the oral glucose tolerance test drug-naive patients with type 2 diabetes. Diabetes Res in Japanese patients with type 2 diabetes. Endocr J 60: Clin Pract 76: 132-138. 45-49. 6. Dejager S, Razac S, Foley JE, Schweizer A (2007) 17. Iwasaki M, Hoshian F, Tsuji T, Hirose N, Matsumoto Vildagliptin in drug-naive patients with type 2 diabe- T, et al. (2012) Predicting efficacy of dipeptidyl pep- tes: a 24-week, double-blind, randomized, placebo-con- tidase-4 inhibitors in patients with type 2 diabetes: trolled, multiple-dose study. Horm Metab Res 39: 218- Association of glycated hemoglobin reduction with 223. serum eicosapentaenoic acid and docosahexaenoic acid 7. Nonaka K, Kakikawa T, Sato A, Okuyama K, Fujimoto levels. J Diabetes Invest 3: 464-467. G, et al. (2008) Efficacy and safety of sitagliptin mono- 18. Chung SH, Suh S, Kim MY, Kim SK, Kim HK, et al. therapy in Japanese patients with type 2 diabetes. (2013) Predictive factors of durability to sitagliptin: Diabetes Res Clin Pract 79: 291-298. Slower reduction of glycated hemoglobin, older age and 8. DeFronzo RA, Fleck PR, Wilson CA, Mekki Q (2008) higher baseline glycated hemoglobin. J Diabetes Invest Efficacy and safety of the dipeptidyl peptidase-4 inhibi- DOI: 10.1111/jdi.12127. tor alogliptin in patients with type 2 diabetes and inad- 19. Kozawa J, Kitamura T, Nishizawa H, Yasuda T, Maeda equate glycemic control: a randomized, double-blind, N, et al. (2013) Dipeptidyl peptidase-4 inhibitors are placebo-controlled study. Diabetes Care 31: 2315- effective in Japanese type 2 diabetic patients with sus- 2317. tained endogenous insulin-secreting capacity, a higher 9. Shimoda S, Iwashita S, Ichimori S, Matsuo Y, Goto R, body mass index and insulin resistance. J Diabetes et al. (2013) Efficacy and safety of sitagliptin as add-on Invest 4: 190-194. therapy on glycemic control and blood glucose fluctua- 20. Charbonnel B, Karasik A, Liu J, Wu M, Meininger G tion in Japanese type 2 diabetes subjects ongoing with (2006) Efficacy and safety of the dipeptidyl peptidase-4 multiple daily insulin injections therapy. Endocr J 60: inhibitor sitagliptin added to ongoing metformin ther- 1207-1214. apy in patients with type 2 diabetes inadequately con- 10. Katsuno T, Ikeda H, Ida K, Miyagawa J, Namba M trolled with metformin alone. Diabetes Care 29: 2638- (2013) Add-on therapy with the DPP-4 inhibitor sita- 2643. gliptin improves glycemic control in insulin-treated 21. Rosenstock J, Brazg R, Andryuk PJ, Lu K, Stein P Japanese patients with type 2 diabetes mellitus. Endocr (2006) Efficacy and safety of the dipeptidyl peptidase-4 J 60: 733-742. inhibitor sitagliptin added to ongoing pioglitazone ther- 11. Takahara M, Shiraiwa T, Kaneto H, Katakami N, apy in patients with type 2 diabetes: a 24-week, mul- Matsuoka TA, et al. (2012) Efficacy of sitagliptin on ticenter, randomized, double-blind, placebo-controlled, blood glucose fluctuation in Japanese type 2 diabetic parallel-group study. Clin Ther 28: 1556-1568. patients with basal-supported oral therapy. Endocr J 59: 22. Bosi E, Camisasca RP, Collober C, Rochotte E, Garber 1131-1136. AJ (2007) Effects of vildagliptin on glucose control 12. Tajiri Y, Tsuruta M, Ohki T, Kato T, Sasaki Y, et al. over 24 weeks in patients with type 2 diabetes inade- (2012) Long-term efficacy of sitagliptin for the treat- quately controlled with metformin. Diabetes Care 30: ment of type 2 diabetic patients in Japan. Endocr J 59: 890-895. 197-204. 23. Garber AJ, Schweizer A, Baron MA, Rochotte E, 13. Seino Y, Hiroi S, Hirayama M, Kaku K (2012) Efficacy Dejager S (2007) Vildagliptin in combination with pio- and safety of alogliptin added to sulfonylurea in glitazone improves glycaemic control in patients with Japanese patients with type 2 diabetes: A randomized, type 2 diabetes failing thiazolidinedione monotherapy: double-blind, placebo-controlled trial with an open-la- a randomized, placebo-controlled study. Diabetes Obes bel, long-term extension study. J Diabetes Invest 3: 517- Metab 9: 166-174. 525. 24. Garber AJ, Foley JE, Banerji MA, Ebeling P, 14. Tanaka T, Goto H, Araki R, Yamamoto M, Tanaka T, et Gudbjornsdottir S, et al. (2008) Effects of vildaglip- al. (2013) Efficacy and safety of sitagliptin in Japanese tin on glucose control in patients with type 2 diabetes patients with type 2 diabetes switched from glinides. J inadequately controlled with a sulphonylurea. Diabetes Diabetes Invest. doi:10.1111/jdi.12140. Obes Metab 10: 1047-1056. 15. Kadowaki T, Tajima N, Odawara M, Nishii M, Taniguchi 25. Pratley RE, Reusch JE, Fleck PR, Wilson CA, Mekki Q Addition of volibose to alogliptin 455

(2009) Efficacy and safety of the dipeptidyl peptidase-4 pre-prandial dose of insulin glulisine while continuing inhibitor alogliptin added to pioglitazone in patients titrated insulin glargine in type 2 diabetes: a 6-month with type 2 diabetes: a randomized, double-blind, pla- ‘proof-of-concept’ study. Diabetes Obes Metab 13: cebo-controlled study. Curr Med Res Opin 25: 2361- 1020-1027. 2371. 36. Lankisch MR, Ferlinz KC, Leahy JL, Scherbaum WA 26. Nauck MA, Ellis GC, Fleck PR, Wilson CA, Mekki (2008) Introducing a simplified approach to insulin Q (2009) Efficacy and safety of adding the dipepti- therapy in type 2 diabetes: a comparison of two single- dyl peptidase-4 inhibitor alogliptin to metformin ther- dose regimens of insulin glulisine plus insulin glargine apy in patients with type 2 diabetes inadequately con- and oral antidiabetic drugs. Diabetes Obes Metab 10: trolled with metformin monotherapy: a multicentre, 1178-1185. randomised, double-blind, placebo-controlled study. Int 37. Davidson MB, Raskin P, Tanenberg RJ, Vlajnic A, J Clin Pract 63: 46-55. Hollander P (2011) A stepwise approach to insulin ther- 27. Pratley RE, Kipnes MS, Fleck PR, Wilson C, Mekki Q apy in patients with type 2 diabetes mellitus and basal (2009) Efficacy and safety of the dipeptidyl peptidase-4 insulin treatment failure. Endocr Pract 17: 395-403. inhibitor alogliptin in patients with type 2 diabetes inad- 38. Ushakova O, Sokolovskaya V, Morozova A, Valeeva F, equately controlled by glyburide monotherapy. Diabetes Zanozina O, et al. (2007) Comparison of biphasic insulin Obes Metab 11: 167-176. aspart 30 given three times daily or twice daily in com- 28. Seino Y, Fujita T, Hiroi S, Hirayama M, Kaku K (2011) bination with metformin versus oral antidiabetic drugs Alogliptin plus voglibose in Japanese patients with type alone in patients with poorly controlled type 2 diabetes: 2 diabetes: a randomized, double-blind, placebo-con- a 16-week, randomized, open-label, parallel-group trial trolled trial with an open-label, long-term extension. conducted in russia. Clin Ther 29: 2374-2384. Curr Med Res Opin 27 Suppl 3: 21-29. 39. Bebakar WM, Chow CC, Kadir KA, Suwanwalaikorn 29. Tajima N, Kadowaki T, Okamoto T, Sato A, Okuyama S, Vaz JA, et al. (2007) Adding biphasic insulin aspart K, et al. (2013) Sitagliptin added to voglibose mono- 30 once or twice daily is more efficacious than optimiz- therapy improves glycemic control in patients with type ing oral antidiabetic treatment in patients with type 2 2 diabetes. J Diabetes Invest. doi:10.1111/jdi.12116. diabetes. Diabetes Obes Metab 9: 724-732. 30. Aoki K, Masuda K, Miyazaki T, Togashi Y, Terauchi Y 40. Strojek K, Bebakar WM, Khutsoane DT, Pesic M, (2010) Effects of miglitol, sitagliptin or their combina- Smahelova A, et al. (2009) Once-daily initiation with tion on plasma glucose, insulin and incretin levels in biphasic insulin aspart 30 versus insulin glargine in non-diabetic men. Endocr J 57: 667-672. patients with type 2 diabetes inadequately controlled 31. Kusunoki Y, Katsuno T, Myojin M, Miyakoshi K, Ikawa with oral drugs: an open-label, multinational RCT. Curr T, et al. (2013) Effect of additional administration of Med Res Opin 25: 2887-2894. acarbose on blood glucose fluctuations and postprandial 41. Kalra S, Plata-Que T, Kumar D, Mumtaz M, hyperglycemia in patients with type 2 diabetes mellitus Sondergaard F, et al. (2010) Initiation with once-daily under treatment with alogliptin. Endocr J 60: 431-439. BIAsp 30 results in superior outcome compared to insu- 32. Kurozumi A, Okada Y, Mori H, Arao T, Tanaka Y (2013) lin glargine in Asians with type 2 diabetes inadequately Efficacy of α-glucosidase inhibitors combined with controlled by oral anti-diabetic drugs. Diabetes Res Clin dipeptidyl-peptidase-4 inhibitor (alogliptin) for glu- Pract 88: 282-288. cose fluctuation in patients with type 2 diabetes mellitus 42. Hara T, Nakamura J, Koh N, Sakakibara F, Takeuchi by continuous glucose monitoring. J Diabetes Invest. N, et al. (1996) An importance of carbohydrate inges- doi:10.1111/jdi.12059. tion for the expression of the effect of alpha-glucosidase 33. Aoki K, Kamiyama H, Yoshimura K, Shibuya M, inhibitor in NIDDM. Diabetes Care 19: 642-647. Masuda K, et al. (2012) Miglitol administered before 43. Yamazaki K, Inoue T, Yasuda N, Sato Y, Nagakura T, breakfast increased plasma active glucagon-like pep- et al. (2007) Comparison of efficacies of a dipeptidyl tide-1 (GLP-1) levels after lunch in patients with type 2 peptidase IV inhibitor and alpha-glucosidase inhibitors diabetes treated with sitagliptin. Acta Diabetol 49: 225- in oral carbohydrate and meal tolerance tests and the 230. effects of their combination in mice. J Pharmacol Sci 34. Masuda K, Aoki K, Kamiko K, Takihata M, Ito Y, et al. 104: 29-38. (2013) Glycemic control after addition of the dipepti- 44. Moritoh Y, Takeuchi K, Hazama M (2010) Combination dyl peptidase-4 inhibitor alogliptin in patients with type treatment with alogliptin and voglibose increases active 2 diabetes showing inadequate response to thrice-a- GLP-1 circulation, prevents the development of diabe- day treatment with alpha-glucosidase inhibitors. Expert tes and preserves pancreatic beta-cells in prediabetic db/ Opin Pharmacother 14: 1111-1118. db mice. Diabetes Obes Metab 12: 224-233. 35. Owens DR, Luzio SD, Sert-Langeron C, Riddle MC 45. Goke B, Fuder H, Wieckhorst G, Theiss U, Stridde E, et (2011) Effects of initiation and titration of a single al. (1995) Voglibose (AO-128) is an efficient alpha-glu- 456 Takahara et al.

cosidase inhibitor and mobilizes the endogenous GLP-1 52. Inzucchi SE, Bergenstal RM, Buse JB, Diamant M, reserve. Digestion 56: 493-501. Ferrannini E, et al. (2012) Management of hypergly- 46. Qualmann C, Nauck MA, Holst JJ, Orskov C, Creutzfeldt cemia in type 2 diabetes: a patient-centered approach: W (1995) Glucagon-like peptide 1 (7-36 amide) secre- position statement of the American Diabetes Association tion in response to luminal sucrose from the upper and (ADA) and the European Association for the Study of lower gut. A study using alpha-glucosidase inhibition Diabetes (EASD). Diabetes Care 35: 1364-1379. (acarbose). Scand J Gastroenterol 30: 892-896. 53. Kawamori R, Tajima N, Iwamoto Y, Kashiwagi A, 47. Ranganath L, Norris F, Morgan L, Wright J, Marks V Shimamoto K, et al. (2009) Voglibose for prevention (1998) Delayed gastric emptying occurs following acar- of type 2 diabetes mellitus: a randomised, double-blind bose administration and is a further mechanism for its trial in Japanese individuals with impaired glucose tol- anti-hyperglycaemic effect. Diabet Med 15: 120-124. erance. Lancet 373: 1607-1614. 48. Seifarth C, Bergmann J, Holst JJ, Ritzel R, Schmiegel 54. Goodman M, Thurston H and Penman J (2009) Efficacy W, et al. (1998) Prolonged and enhanced secretion of and tolerability of vildagliptin in patients with type 2 glucagon-like peptide 1 (7-36 amide) after oral sucrose diabetes inadequately controlled with metformin mono- due to alpha-glucosidase inhibition (acarbose) in Type 2 therapy. Horm Metab Res 41: 368-373. diabetic patients. Diabet Med 15: 485-491. 55. Iwamoto Y, Tajima N, Kadowaki T, Nonaka K, Taniguchi 49. Lee A, Patrick P, Wishart J, Horowitz M, Morley JE T, et al. (2010) Efficacy and safety of sitagliptin mono- (2002) The effects of miglitol on glucagon-like pep- therapy compared with voglibose in Japanese patients tide-1 secretion and appetite sensations in obese type 2 with type 2 diabetes: a randomized, double-blind trial. diabetics. Diabetes Obes Metab 4: 329-335. Diabetes Obes Metab 12: 613-622. 50. Narita T, Katsuura Y, Sato T, Hosoba M, Fujita H, et al. 56. Davies M, Speight J (2012) Patient-reported outcomes (2009) Miglitol induces prolonged and enhanced gluca- in trials of incretin-based therapies in patients with type gon-like peptide-1 and reduced gastric inhibitory poly- 2 diabetes mellitus. Diabetes Obes Metab 14: 882-892. peptide responses after ingestion of a mixed meal in 57. Ali MK, Feeney P, Hire D, Simmons DL, O’Connor Japanese Type 2 diabetic patients. Diabet Med 26: 187- PJ, et al. (2012) Glycaemia and correlates of patient-re- 188. ported outcomes in ACCORD trial participants. Diabet 51. Dungan KM, Buse JB, Largay J, Kelly MM, Button Med 29: e67-74. EA, et al. (2006) 1,5-anhydroglucitol and postpran- 58. Hauber AB, Mohamed AF, Johnson FR, Falvey H dial hyperglycemia as measured by continuous glucose (2009) Treatment preferences and medication adher- monitoring system in moderately controlled patients ence of people with Type 2 diabetes using oral glucose- with diabetes. Diabetes Care 29: 1214-1219. lowering agents. Diabet Med 26: 416-424.