A randomized, double-blind, placebo-controlled, phase II clinical trial to investigate the efficacy and safety of oral DA-1229 in patients with type 2 diabetes mellitus who have inadequate glycaemic control with diet and exercise.

BACKGROUND: DA-1229 is a novel, potent and selective dipeptidyl peptidase-4 (DPP-IV) inhibitor that is orally bioavailable. We aimed to evaluate the optimal dose, efficacy and safety of DA-1229, in Korean subjects with type 2 diabetes mellitus suboptimally controlled with diet and exercise.
METHODS: We enrolled 158 patients (mean age, 53 years and a mean BMI, 25.6 kg/m(2) ). The mean baseline fasting plasma glucose level, HbA1c and duration of diabetes were 8.28 mmol/L, 7.6% (60 mmol/mol) and 3.9 years, respectively. After 2 or 6 weeks of an exercise and diet program followed by 2 weeks of a placebo period, the subjects were randomized into one of four groups for a 12-week active treatment period: placebo, 2.5, 5 or 10 mg of DA-1229.
RESULTS: All three doses of DA-1229 significantly reduced HbA1c from baseline compared to the placebo group (-0.09 in the placebo group vs. -0.56, -0.66 and -0.61% in 2.5, 5 and 10-mg groups, respectively) but without any significant differences between the doses. Insulin secretory function, as assessed by homeostasis model assessment β-cell, the insulinogenic index, 2-h oral glucose tolerance test (OGTT) C-peptide and post-OGTT C-peptide area under the curve (AUC)0-2h, significantly improved with DA-1229 treatment. The incidence of adverse events was similar between the treatment groups and DA-1229 did not affect body weight or induce hypoglycaemic events.
CONCLUSIONS: DA-1229 monotherapy (5 mg for 12 weeks) improved HbA1c, fasting plasma glucose level, OGTT results and β-cell function. This drug was well tolerated in Korean subjects with type 2 diabetes mellitus.

RCT Entities:   Population Interventions Outcomes

Introduction

Type 2 diabetes mellitus is a complex and progressive disease with a prevalence that is increasing worldwide 1. Although there is general agreement on the first-line use of metformin in most patients with type 2 diabetes mellitus, the ideal drug sequence after metformin failure is an area of increasing uncertainty 2.

Dipeptidyl peptidase-4 (DPP-IV) inhibitors were introduced as antidiabetic agents in 2006, with sitagliptin being the first available, followed by vildagliptin, saxagliptin, linagliptin, alogliptin and gemigliptin 3–8. DPP-IV inhibitors reduce plasma DPP4-IV activity by 70–90% in a sustained manner for 24 hours with an increase in glucagon-like peptide-1 (GLP-1) and glucose-dependent insulinotrophic polypeptide (GIP), which are hormones released in response to food intake that have glucose-lowering effects by stimulating insulin secretion by pancreatic β-cells and inhibiting glucagon secretion by pancreatic α-cells 9.

Although the above DPP-IV inhibitors differ in their potency, target selectivity, oral bioavailability, half-life, metabolism and excretion rate, they both possess considerable glucose-lowering effects and favorable safety profiles 10.

DA-1229 is a novel, potent and selective DPP-IV inhibitor that is orally bioavailable 11. In a single-and multiple-dose phase I clinical trial in healthy Korean subjects. DA-1229 was shown to be well tolerated without serious adverse events (SAEs) 12. The half-life of DA-1229 was approximately 30–40 h (1.25–60 mg), the pharmacokinetics of which were not influenced by food 12. In a pharmacodynamic study, more than 80% of DPP-IV was inhibited by a single dose of 5 mg or higher of DA-1229, and this level of inhibition was maintained for at least 24 h after a single dose of DA-1229 of 10 mg or higher 12. Therefore, DA-1229 is likely to be efficacious when administered as a once-daily antidiabetic agent 12.

This phase II clinical trial was designed to evaluate the efficacy and safety of oral DA-1229 and to determine the optimal dose to use for a phase III clinical study in Korean subjects with type 2 diabetes mellitus.

Materials and methods

Study design

This study was a phase II, multicenter, randomized, double-blind, placebo-controlled study. Study participants were recruited from 26 university hospitals throughout Korea. Participants who gave written informed consent were screened using the inclusion and exclusion criteria. After 2 or 6 weeks of an exercise and diet program designed based on general diabetes treatment guidelines, all participants received a placebo for 2 weeks. When the 2-week placebo run-in period was completed, the participants were randomly assigned to one of the following groups: placebo, 2.5, 5 or 10 mg of DA-1229. The block randomization method was applied at each study site, and the randomization schedule was generated using SAS System 9.2 (SAS Institute, Cary, NC, USA). Participants were instructed to take one tablet of the investigational drug once daily at the same time in the morning for 12 weeks. Participants with persistent hyperglycaemia, defined as having a fasting plasma glucose (FPG) level exceeding 14.99 mmol/L, were made to stop taking the study medications (placebo or DA-1229) and initiate additional rescue therapy, and were dropped from this study. Figure1 summarizes the trial profile and study scheme.

Figure 1

Trial profile (A) and study scheme (B)

Subjects

Men and women between the ages of 20 and 75 years with type 2 diabetes mellitus who satisfied either of the following conditions at screening (visit 1) were eligible for this study: patients with HbA1c in the range of 7.0% (53 mmol/mol) to 10.0% (86 mmol/mol) who were diagnosed with T2DM for the first time in the 4 weeks prior to screening or patients with HbA1c in the range of 7.0% (53 mmol/mol) to 10.0% (86 mmol/mol) who had not been treated with oral glucose-lowering agents in the 6 weeks prior to screening. Newly diagnosed subjects underwent 6 weeks of an exercise and diet program, while oral glucose-lowering agent-naïve subjects underwent a program of 2 weeks of exercise and diet, which allowed them to have periods of lifestyle modification lasting at least 8 and 10 weeks in drug-naïve subjects and newly diagnosed subjects, respectively. Patients with either HbA1c in the range of 7.0% (53 mmol/mol) to 10.0% (86 mmol/mol) or FPG in the range of 6.99 mmol/L to 14.99 mmol/L at visit 2 were administered a single-blinded placebo once daily for 2 weeks, while continuing the exercise and diet program, from the date of the second visit to the day before the third visit. These periods consisting of lifestyle modification ranging from 8 to 10 weeks and a placebo run-in of 2 weeks were used to identify stable HbA1 levels not affected by external conditions.

Exclusion criteria were as follows: body mass index (BMI) <20 or >40 kg/m2; a history of type 1 diabetes, secondary diabetes, or gestational diabetes; a history of myocardial infarction, transient ischemic attack or coronary artery bypass surgery in the 6 months prior to screening; a history of a major skin allergy; use of corticosteroids or sodium channel blockers in the 3 months prior to screening; use of insulin or thiazolidinediones in the 6 months prior to screening; current use of warfarin, docoumarin or digoxin; thyroid dysfunction (except cases that had been treated stably with thyroid hormonal replacement therapy for at least 3 months prior to entry); serum creatinine level >132.6 µmol/L (1.5 mg/dl); creatinine phosphokinase level exceeding 3× the upper normal limit with symptoms; alanine aminotransferase or aspartate aminotransferase level exceeding 2.5× the upper normal limit; history of alcohol or drug abuse in the 2 months prior to screening; participation in another clinical trial in the 2 months prior to screening; pregnancy or lactation and/or current use of drugs or food that could affect CYP3A4 metabolism in the liver.

The study protocol was approved by the independent ethics committee/institutional review board at each study site, and written informed consent was obtained from all participants. This study was performed according to the Good Clinical Practice guidelines and the Declaration of Helsinki.

Study endpoints

The primary endpoint was the mean change in HbA1c from baseline to week 12. The secondary endpoints included the HbA1c response rate [<7.0% (53 mmol/mol) or <6.5% (48 mmol/mol); changes in FPG, glycated albumin (GA), fasting insulin, fasting proinsulin, the proinsulin/insulin ratio, fasting lipid parameters, including total cholesterol, low-density lipoprotein-cholesterol (LDL-C), high-density lipoprotein-cholesterol (HDL-C) and triglycerides (TG); and changes in the homeostasis model assessment of insulin resistance (HOMA-IR), the homeostasis model assessment of β-cell function (HOMA-β) and the quantitative insulin sensitivity check index (QUICKI) 13.

An oral glucose tolerance test (OGTT) was performed by administering 75 g of glucose and obtaining blood samples for glucose, insulin, proinsulin and C-peptide at 0, 30, 60 and 120 min. We analysed changes from baseline in the 2 h postload levels of glucose, insulin, proinsulin and C-peptide at 12 weeks. The area under the curve (AUC)0–2h was calculated for each parameter. The following parameters were also set as secondary endpoints: 2 h OGTT for glucose, insulin, proinsulin and C-peptide; AUC0–2h for glucose, insulin, proinsulin and C-peptide; the insulinogenic index 13 and body weight.

Safety endpoints included adverse events (AEs), vital signs and laboratory test results. AEs and vital signs were evaluated, and laboratory tests were performed at every visit.

Statistical analysis

The planned sample size for this study was approximately 188 patients. This number was calculated to detect with 85% power a treatment mean difference of 0.56% (standard deviation, 0.8) in HbA1c from baseline with an overall significance level of 0.15 and assuming a 20% dropout rate. As the dropout rate was lower than previously estimated, this study was terminated when tests from 158 patients were completed.

Demographic data and baseline characteristics were compared among the treatment groups. Continuous variables were analysed by using analysis of variance (ANOVA) or the Kruskal–Wallis test, and categorical variables were analysed using the chi-square test or Fisher’s exact test. The efficacy variables were described using descriptive statistics for each group and analysed to identify all differences between treatment groups. For inter-group comparisons in efficacy variables, either ANOVA or the Kruskal–Wallis test was performed, with the post-hoc analysis using multiple comparisons of Dunnett’s test or multiple comparisons of the step-down Bonferroni method. [The Wilcoxon rank sum test was performed for comparisons between each DA-1229 dose and the placebo, followed by correction of p-values using the step-down Bonferroni method]. For intra-group comparisons, the results at baseline and at week 12 were analysed by a paired t-test or the Wilcoxon signed rank test. The primary end point was assessed using an analysis of covariance (ANCOVA) model, including terms for treatment, baseline HbA1c and the diagnosis group (i.e. newly diagnosed or drug-naïve group) as covariates. The least square (LS) mean differences and standard errors (SE) were estimated for the comparisons of placebo versus DA-1229. The chi-square test (or Fisher’s exact test) was used for categorical data. The full analysis set was used to evaluate efficacy and included all patients randomly assigned to different treatment groups who received at least one dose of the double-blind study drug and who had at least one post-baseline measurement. The last observation carried forward method was used in cases of missing data.

Safety and tolerability were assessed in patients who received at least one dose of the study drug, and safety data were collected. For each AE, the number of subjects who experienced one or more events was recorded, and the percentage was reported according to the treatment group. Events were described with regard to gravity, severity, relationship to the investigational drug, treatment and outcome.

The statistical software SAS System 9.2 was used for all statistical analyses, and two-sided tests were performed to determine statistical significance with a significance level of ≤5%.

Results

Two-hundred and eighty-nine people with type 2 diabetes mellitus gave consent to participate in the study and were screened for eligibility (Figure1A). Among the 187 eligible participants who were placed on an exercise and diet program, 158 completed the placebo run-in period. Of the 158 randomized subjects, five without efficacy data were excluded from the efficacy analysis and; thus, a total of 153 subjects were included in the full analysis set. Thirty-eight patients were randomized to the placebo group, and 40, 43 and 37 patients to the 2.5, 5 and 10 mg DA-1229 treatment groups, respectively.

Baseline characteristics of the subjects

Demographic information and baseline characteristics of subjects enrolled in this clinical trial are summarized in Table1. There were no significant differences among the subjects’ baseline characteristics. The mean age of the study subjects was 53 years, and 93 subjects (59%) were men and 65 (41%) were women. The mean body mass index (BMI) was 25.6 kg/m2, and the mean duration of diabetes was 3.9 years. The mean HbA1c, GA and FPG levels at baseline were 7.6% (60 mmol/mol), 19.1% and 8.28 mmol/L, respectively. Of the 153 subjects, the baseline HbA1c was >8.5% (69 mmol/mol) in 21 subjects and ≤8.5% (69 mmol/mol) in 132 subjects.

Table 1

Baseline characteristics of study participants according to treatment groups

	Placebo	DA-1229
Variables	(n = 38)	2.5 mg (n = 40)	5 mg (n = 43)	10 mg (n = 37)	P-valuea
Male (n, %)	26 (68.42)	22 (55.00)	25 (58.14)	20 (54.05)	0.5636
Age (years)	54.42 (9.85)	52.10 (11.48)	54.21 (9.74)	53.16 (9.60)	0.7268
Height (cm)	165 (10.00)	164 (8.00)	164 (9.00)	162 (10.00)	0.7629
Weight (kg)	70.66 (10.95)	69.04 (11.54)	68.40 (11.67)	67.21 (10.85)	0.6641
BMI (kg/m2)	25.98 (2.35)	25.67 (3.37)	25.34 (3.37)	25.39 (3.02)	0.3457
Duration of diabetes mellitus (years)	3.55 (2.85)	4.00 (3.57)	3.98 (3.88)	3.95 (3.88)	0.9900
Proportion of newly diagnosed diabetes mellitus (n, %)	7 (18.4)	9 (22.5)	9 (20.9)	6 (16.2)	0.9047
HbA1c (%) at baseline	7.57 (0.87)	7.67 (0.83)	7.58 (0.71)	7.71 (0.88)	0.8509
GA (%) at baseline	19.41 (3.78)	19.41 (4.30)	18.68 (3.85)	19.14 (4.14)	0.9405
FPG (mmol/L)	8.15 (1.99)	8.51 (2.27)	8.08 (1.81)	8.40 (1.88)	0.7695
Insulin (pmol/L)	62.57 (27.57)	60.21 (23.06)	57.37 (21.11)	60.28 (26.95)	0.6671
Proinsulin (pM)	19.04 (10.25)	18.15 (11.98)	15.87 (10.20)	15.48 (12.01)	0.2377
HOMA-ß (%)	45.14 (28.69)	40.76 (23.13)	41.48 (24.57)	39.80 (22.47)	0.8406
HOMA-IR	3.20 (1.39)	3.26 (1.38)	2.97 (1.25)	3.19 (1.42)	0.7561
Insulinogenic index	10.87 (13.51)	10.57 (18.70)	12.97 (16.90)	6.41 (17.41)	0.3267
QUICKI	0.57 (0.11)	0.56 (0.08)	0.57 (0.07)	0.56 (0.05)	0.7502
Proinsulin/insulin (%)	35.76 (21.54)	35.40 (36.99)	30.78 (25.08)	25.09 (15.40)	0.1393
Total cholesterol (mmol/L)	4.68 (0.97)	4.85 (1.06)	4.99 (1.16)	4.82 (1.06)	0.5032
TG (mmol/L)	1.72 (0.84)	1.68 (1.04)	1.88 (1.75)	2.08 (2.26)	0.9568
LDL-C (mmol/L)	2.80 (0.78))	2.98 (0.97))	3.06 (1.04)	2.79 (0.86)	0.4836
HDL-C (mmol/L)	1.27 (0.26)	1.25 (0.31)	1.26 (0.27)	1.29 (0.35)	0.9244

Data are presented as mean (standard deviation, SD) unless indicated otherwise.

P-values were derived from the analysis of variance (ANOVA) or the Kruskal–Wallis test for continuous variables and the chi-square test for categorical variables.

Efficacy

After 12 weeks of treatment, all doses of DA-1229 resulted in clinically and statistically significant reductions in HbA1c compared to the placebo group (Figure2 The mean placebo-subtracted reductions [95% confidence interval, (CI)] of HbA1c were −0.47% (−0.80, −0.15; p = 0.0014) in the 2.5-mg group, −0.57% (−0.86, −0.29; p < 0.0003) in the 5-mg group and −0.53% (−0.83, −0.23; p < 0.0002) in the 10-mg group.

Figure 2

Changes in HbA1c from baseline to week 12. Data are expressed as mean (SD). aP-values were derived from paired t-test or the Wilcoxon signed rank test. bP-values were derived from the Dunnett’s test or multiple comparison of the step-down Bonferroni’s method

The adjusted LS means (SE) of HbA1c reduction from baseline were −0.13 (0.11), −0.59 (0.10), −0.70 (0.10) and −0.64 (0.11) in the placebo, 2.5 mg, 5 mg and 10 mg DA-1229 groups, respectively (Table2). No differences were seen between the diagnosis groups (i.e. newly diagnosed or drug-naïve group, p = 0.3490, data not shown), while the baseline HbA1c was a significant covariate (p = 0.0096, data not shown).

Table 2

Adjusted changes in HbA1c from baseline to week 12

		DA-1229
	Placebo (N = 34)	2.5 mg (N = 39)	5 mg (N = 43)	10 mg (N = 37)	P-valuea
Change from baseline	−0.13 (0.11)	−0.59 (0.10)	−0.70 (0.10)	−0.64 (0.11)	0.0004
Difference from placebo		0.46	0.57	0.51
P-valueb		0.0054	0.0002	0.0018

Data are presented as LS mean (SE).

P-value was derived from the analysis of covariance (ANCOVA).

P-values were derived from Dunnett’s test.

Subgroup analysis was carried out for high and low HbA1c levels at baseline [>8.5% (69 mmol/mol) or ≤8.5% (69 mmol/mol)]. All three doses of DA-1229 (2.5, 5 and 10 mg) resulted in significant decreases in HbA1c from baseline in each stratum, except for the 10 mg subgroup with HbA1c >8.5% (69 mmol/mol) at baseline (p = 0.0985). The treatment effect was greatest in the group with HbA1c >8.5% (69 mmol/mol) (Online Supplementary Table S1). When the subjects were divided into two groups by age, subjects >60 years of age had a greater reduction in HbA1c from baseline in the 5-mg group compared to subjects ≤60 years of age, although this comparison was not statistically significant (−0.72 vs. −0.64%).

The proportion of patients with HbA1c <7.0% (53 mmol/mol) at week 12 was 50.0% (17/34) in the placebo group, 51.3% (20/39) in the 2.5-mg group, 62.8% (27/43) in the 5-mg group and 62.2% (23/37) in the 10-mg group. Among the four groups, the 5-mg group had the highest proportion of subjects with HbA1c <7.0% (53 mmol/mol) at week 12, although these results were not statistically significant (p = 0.5344). The proportion of patients with HbA1c <6.5% (48 mmol/mol) at week 12 was 8.8% (3/31) in the placebo group, 23.1% (9/39) in the 2.5-mg group, 27.9% (12/43) in the 5-mg group and 10.8% (4/37) in the 10-mg group.

All three doses of DA-1229 decreased the levels of FPG and GA from baseline through to week 12, while both levels increased in the placebo group (Table3). All three doses of DA-1229 significantly improved glycaemic control based on the levels of FPG and GA compared with the placebo (Table3). Among the DA-1229 groups, the 5-mg group showed the greatest change in FPG from baseline, while the 10-mg group showed the greatest change in GA from baseline. There were no statistically significant differences in fasting serum insulin or proinsulin compared to the placebo group, while the proinsulin/insulin ratio at week 12 showed a significant change from baseline in the 2.5 and 5-mg groups, although these changes were not significantly different from the placebo group (Table3).

Table 3

Baseline, week 12 and changes in results from baseline for FPG, GA, and parameters related to insulin

				Changes from baseline
Treatment	n	Baseline mean (s.d.)	Week 12 mean (s.d.)	Mean difference from baseline at week 12	Mean difference from placebo	95% CI	P-value (vs. baseline)a	P-valueb (vs. placebo)c
FPG (mmol/L)								0.0003
Placebo	34	8.15 (1.99)	8.50 (2.69)	0.35 (1.67)			0.3242
DA-1229 2.5 mg	39	8.51 (2.27)	7.72 (1.38)	−0.78 (1.59)	−1.14 (1.62)	(−1.90, −0.38)	0.0006	(0.0023)
DA-1229 5 mg	43	8.08 (1.81)	7.24 (1.28)	−0.84 (1.69)	−1.20 (1.68)	(−1.96, −0.43)	0.0005	(0.0026)
DA-1229 10 mg	37	8.40 (1.88)	7.75 (1.77)	−0.66 (0.74)	−1.01 (1.27)	(−1.62, −0.41)	<0.0001	(<0.0003)
GA (%)								<0.0001
Placebo	26	19.41 (3.78)	20.69 (6.70)	1.28 (3.76)			0.1266
DA-1229 2.5 mg	32	19.41 (4.30)	17.16 (3.22)	−2.25 (2.48)	−3.53 (3.12)	(−5.18, −1.88)	<0.0001	(<0.0001)
DA-1229 5 mg	37	18.68 (3.85)	16.23 (2.59)	−2.45 (2.20)	−3.73 (2.94)	(−5.23, −2.23)	<0.0001	(<0.0001)
DA-1229 10 mg	28	19.14 (4.14)	16.34 (2.81)	−2.69 (2.13)	−3.97 (3.02)	(−5.62, −2.32)	<0.0001	(<0.0001)
Insulin (pmol/L)								0.7873
Placebo	34	62.57 (27.57)	59.66 (20.70)	−2.99 (25.00)			0.4942
DA-1229 2.5 mg	38	60.21 (23.06)	63.48 (23.75)	4.10 (23.27)	7.01 (24.10)	(−4.31, 18.40)	0.2871
DA-1229 5 mg	43	57.37 (21.11)	66.46 (72.30)	9.10 (60.49)	12.01 (48.20)	(−10.00, 34.10)	0.8586
DA-1229 10 mg	37	60.28 (26.95)	60.91 (30.28)	0.63 (24.72)	3.54 (24.86)	(−8.20, 15.35)	0.7730
Proinsulin (pM)								0.7365
Placebo	26	19.04 (10.25)	17.97 (8.51)	−0.57 (9.60)			0.7656
DA-1229 2.5 mg	33	18.15 (11.98)	16.69 (13.94)	−1.17 (10.31)	−0.60 (10.01)	(−5.86, 4.65)	0.4350
DA-1229 5 mg	37	15.87 (10.20)	12.56 (7.80)	−3.16 (9.52)	−2.60 (9.55)	(−7.48, 2.29)	0.1056
DA-1229 10 mg	29	15.48 (12.01)	16.37 (17.61)	1.16 (8.03)	1.73 (8.81)	(−3.04, 6.50)	0.7599
Proinsulin/insulin (%)								0.4156
Placebo	26	35.76 (21.54)	32.00 (14.60)	−4.01 (20.42)			0.4957
DA-1229 2.5 mg	32	35.40 (36.99)	28.72 (29.18)	−7.37 (43.52)	−3.37 (35.14)	(−21.95, 15.22)	0.0435
DA-1229 5 mg	37	30.78 (25.08)	22.85 (15.08)	−8.12 (16.91)	−4.11 (18.43)	(−13.55, 5.32)	0.0046
DA-1229 10 mg	29	25.09 (15.40)	26.47 (26.08)	1.46 (18.36)	5.47 (19.36)	(−5.02, 15.95)	0.1474
HOMA-β (%)								0.0651
Placebo	34	45.14 (28.69)	42.87 (24.65)	−2.28 (20.18)			0.5153
DA-1229 2.5 mg	38	40.76 (23.13)	46.91 (22.50)	6.67 (22.80)	8.94 (21.61)	(−1.23, 19.12)	0.0797
DA-1229 5 mg	43	41.48 (24.57)	54.20 (50.73)	12.72 (43.11)	15.00 (34.93)	(−0.97, 30.96)	0.0050
DA-1229 10 mg	37	39.80 (22.47)	46.38 (25.65)	6.58 (14.94)	8.86 (17.64)	(0.50, 17.22)	0.0004
HOMA-IR								0.8170
Placebo	34	3.20 (1.39)	3.13 (1.10)	−0.07 (1.38)			0.7631
DA-1229 2.5 mg	38	3.26 (1.38)	3.16 (1.31)	−0.06 (1.14)	0.01 (1.26)	(−0.58, 0.60)	0.7352
DA-1229 5 mg	43	2.97 (1.25)	3.17 (3.68)	0.19 (3.22)	0.27 (2.58)	(−0.91, 1.44)	0.2663
DA-1229 10 mg	37	3.19 (1.42)	3.02 (1.72)	−0.17 (1.49)	−0.10 (1.44)	(−0.78, 0.59)	0.3654
QUICKI								0.7092
Placebo	34	0.57 (0.11)	0.55 (0.05)	−0.01 (0.11)			0.6884
DA-1229 2.5 mg	38	0.56 (0.08)	0.57 (0.12)	0.01 (0.10)	0.02 (0.11)	(−0.03, 0.07)	0.6322
DA-1229 5 mg	43	0.57 (0.07)	0.58 (0.08)	0.01 (0.08)	0.03 (0.09)	(−0.01, 0.07)	0.2210
DA-1229 10 mg	37	0.56 (0.05)	0.57 (0.06)	0.01 (0.04)	0.03 (0.08)	(−0.01, 0.06)	0.1214
Insulinogenic index								0.0217
Placebo	26	10.87 (13.51)	10.51 (11.40)	0.53 (11.32)			0.8123
DA-1229 2.5 mg	33	10.57 (18.70)	21.90 (33.44)	10.83 (27.04)	10.29 (21.61)	(−1.05, 21.64)	0.0104	(0.1088)
DA-1229 5 mg	37	12.97 (16.90)	22.30 (31.57)	9.03 (29.45)	8.50 (23.76)	(−3.66, 20.66)	0.0376	(0.1084)
DA-1229 10 mg	28	6.41 (17.41)	23.55 (30.04)	16.59 (22.68)	16.06 (18.13)	(6.15, 25.96)	<0.0001	(0.0036)

P-values were derived from the paired t-test or the Wilcoxon signed rank test.

P-values were derived from ANOVA or the Kruskal–Wallis test.

P-values were derived from the Dunnett’s test or multiple comparison of the step-down Bonferroni’s method.

HOMA-β significantly improved in the 5 and 10-mg groups at week 12 compared to baseline, although the mean differences from the placebo group were not statistically significant (Table3). Only the 10-mg group had a significant improvement in the insulinogenic index at week 12 compared to the placebo group (p = 0.0036, Table3). However, no significant changes in HOMA-IR and QUICKI were observed. When we further analysed the change from baseline of HOMA-β according to the diagnosis group (i.e. newly diagnosed or drug-naïve group), there was no difference in the change from baseline of HOMA-ß between the newly diagnosed and drug-naïve groups (Online Supplementary Table S2).

Compared with patients who received placebo, those who received DA-1229 experienced a significant reduction in mean 2 h OGTT C-peptide, post-OGTT glucose AUC0–2h, and post-OGTT C-peptide AUC0–2h (p = 0.0048 and p = 0.0018 for 2 h C-peptide in DA-1229 2.5 mg and 5 mg, respectively; p = 0.0288 and p = 0.0302 for post-OGTT glucose AUC0–2h in DA-1229 5 mg and 10 mg, respectively; p = 0.0058, p = 0.0048 and p = 0.0082 for post-OGTT C-peptide AUC0–2h in DA-1229 2.5, 5 and 10 mg, respectively, Table4). The post-OGTT 2 h glucose concentration decreased from baseline at week 12, while the post-OGTT 2-h insulin concentration and insulin AUC0–2h increased from baseline at 12 weeks; however, these changes were not significantly different from the placebo group (Table4).

Table 4

Baseline, week 12 and change in results from baseline for parameters related to OGTT

				Changes from baseline
Treatment	n	Baseline mean (s.d.)	Week 12 mean (s.d.)	Mean difference from baseline at week 12	Mean difference from placebo	95% CI	P-value (vs. baseline)a	P-valueb (vs. placebo)c
2 h PPG (mmol/L)								0.3511
Placebo	26	16.07 (4.19)	15.46 (4.67)	−0.56 (4.53)			0.5360
DA-1229 2.5 mg	33	15.77 (4.65)	13.56 (3.89)	−1.88 (3.71)	−1.32 (4.09)	(−3.47, 0.83)	0.0066
DA-1229 5 mg	37	15.38 (4.20)	13.03 (3.81)	−2.30 (3.90)	−1.74 (4.17)	(−3.88, 0.39)	0.0004
DA-1229 10 mg	29	15.20 (3.78)	13.08 (3.31)	−1.73 (3.10)	−1.17 (3.84)	(−3.26, 0.91)	0.0065
2 h insulin (pmol/L)								0.2899
Placebo	26	150.08 (117.30)	136.19 (96.81)	0.76 (91.60)			0.9659
DA-1229 2.5 mg	33	218.63 (179.88)	204.04 (111.68)	−12.92 (184.95)	−13.68 (151.26)	(−93.13, 65.77)	0.4723
DA-1229 5 mg	37	165.01 (125.29)	239.95 (231.96)	72.30 (214.53)	71.53 (174.94)	(−17.99, 161.05)	0.0249
DA-1229 10 mg	29	179.32 (126.40)	233.70 (181.96)	64.52 (171.89)	63.76 (139.87)	(−12.01, 139.53)	0.0616
2 h C-peptide (ng/ml)								0.0031
Placebo	26	7.11 (2.47)	6.59 (3.18)	−0.58 (2.11)			0.0409
DA-1229 2.5 mg	33	7.27 (3.06)	8.15 (2.36)	0.87 (2.91)	1.45 (2.59)	(0.09, 2.81)	0.0942	(0.0048)
DA-1229 5 mg	37	6.99 (2.67)	8.13 (3.41)	1.22 (3.43)	1.80 (2.96)	(0.28, 3.32)	0.0053	(0.0018)
DA-1229 10 mg	29	7.39 (2.40)	7.79 (2.30)	0.36 (2.04)	0.94 (2.07)	(−0.18, 2.07)	0.3454	(0.0687)
2 h proinsulin (pM)								0.3571
Placebo	26	48.96 (20.93)	45.62 (22.54)	−2.13 (19.79)			0.5883
DA-1229 2.5 mg	33	54.80 (47.16)	63.29 (65.49)	8.24 (29.57)	10.37 (25.74)	(−3.15, 23.88)	0.2062
DA-1229 5 mg	37	47.08 (25.32)	50.51 (28.34)	3.77 (24.06)	5.90 (22.41)	(−5.57, 17.36)	0.3471
DA-1229 10 mg	29	47.56 (31.15)	49.21 (31.09)	1.34 (13.98)	3.46 (16.97)	(−5.73, 12.66)	0.6107
AUC0–2h (glucose)								0.0356
Placebo	25	32 350 (6237)	31 639 (7305)	−423 (5616)			0.7098
DA-1229 2.5 mg	33	32 405 (7988)	28 095 (6434)	−3705 (5457)	−3282 (5526)	(−6217, −347)	0.0005	(0.0544)
DA-1229 5 mg	37	31 535 (7169)	26 474 (6342)	−5109 (6825)	−4686 (6369)	(−7984, −1387)	<0.0001	(0.0288)
DA-1229 10 mg	29	31 302 (5920)	27 002 (5250)	−3531 (3499)	−3108 (4599)	(−5627, −590)	<0.0001	(0.0302)
AUC0–2h (insulin)								0.0756
Placebo	25	2305 (1382)	2101 (1104)	−41 (901)			0.8199
DA-1229 2.5 mg	32	2775 (1425)	3549 (2723)	811 (2198)	852 (1754)	(−86, 1791)	0.0323
DA-1229 5 mg	37	2508 (1546)	3183 (2123)	705 (1777)	747 (1490)	(−24.61, 1518)	0.0334
DA-1229 10 mg	28	2451 (1542)	3290 (2261)	879 (1592)	921 (1313)	(195, 1646)	0.0032
AUC0–2h (C-peptide)								0.0053
Placebo	25	639 (226)	598 (266)	−41.18 (121.88)			0.1041
DA-1229 2.5 mg	33	639 (203)	751 (307)	107.87 (232.42)	149.10 (192.96)	(46.56, 251.60)	0.0040	(0.0058)
DA-1229 5 mg	37	642 (249)	692 (241)	58.15 (172.22)	99.33 (154.07)	(19.54, 179.10)	0.0168	(0.0048)
DA-1229 10 mg	29	613 (190)	684 (186)	67.54 (153.59)	108.70 (139.85)	(32.13, 185.30)	0.0250	(0.0082)
AUC0–2h (proinsulin)								0.5827
Placebo	25	4365 (1839)	4055 (1816)	−253 (1552)			0.4223
DA-1229 2.5 mg	33	4607 (3599)	5395 (7073)	765 (4090)	1018 (3254)	(−710, 2747)	0.3378
DA-1229 5 mg	37	3995 (2199)	3849 (1986)	−127 (1884)	127 (1759)	(−784, 1038)	0.6841
DA-1229 10 mg	29	3877 (2559)	4036 (3045)	140 (1204)	394 (1375)	(−359, 1147)	0.5757

P-values were derived from the paired t-test or the Wilcoxon signed rank test.

P-values were derived from ANOVA or the Kruskal–Wallis test.

P-values were derived from the Dunnett’s test or multiple comparison of the step-down Bonferroni’s method

At week 12, none of the treatment groups had significant changes from baseline in fasting lipid parameters (total cholesterol, LDL-C, HDL-C and TG) or body weight (data not shown).

Safety results

The safety set included 157 randomized subjects out of 158 randomized subjects; 121 were exposed to DA-1229 and 36 were given the placebo. AEs occurred in 52 (33.12%, 84 cases) out of 157 subjects, and SAEs occurred in one subject (2.56%) in the DA-1229 2.5-mg group and two subjects (5.26%) in the DA-1229 10-mg group (Table5). The reported SAEs were rotator cuff syndrome, which occurred in one subject in the 2.5-mg group and one subject in the 10-mg group, and hemorrhoids, which occurred in one subject in the 10-mg group, but none of the SAEs were related to the study drug (Table5). Hyperglycaemic events, defined as FPG exceeding 14.99 mmol/L, occurred in one subject in the placebo group and one subject in the DA-1229 10-mg group.

Table 5

Incidence of treatment-emergent adverse events and adverse drug reactions

		DA-1229
		2.5 mg (N = 39)	5 mg (N = 44)	10 mg (N = 38)
	Placebo (N = 36)	Number of Subjects (%) [Case]			P-valuea
Adverse events (AE)	13 (36.1) 18	12 (30.8)17	12 (27.3) 25	15 (39.5) 24	0.6561
Mild	16	14	20	20
Moderate	2	2	5	2
Serious (SAE)	[0]	1c	[0]	2d
Adverse drug reactions (ADR)b	4 (11.1) 4	2 (5.13) 4	4 (9.1) 7	4 (10.5) 4	0.7992
Mild	3	4	5	3
Moderate	1	[0]	2	1
Severe	[0]	[0]	[0]	[0]

P-values were derived from the chi-square test or the Fisher’s exact test for inter-group comparison in the incidence of adverse events.

Adverse drug reactions were defined as adverse events where the causal relationship with the investigational product could not be ruled out (when the causal relationship with the investigational product was evaluated as ‘definitely,’ ‘probably’ or ‘possibly’ related).

One case of rotator cuff syndrome.

One case of rotator cuff syndrome and one case of hemorrhoids.

Adverse drug reactions (ADRs) in which a causal relationship with the investigational product could not be ruled out occurred in 14 subjects (8.92%, 19 cases), including four subjects (11.11%, four cases) in the placebo group, two subjects (5.13%, four cases) in the 2.5-mg group, four subjects (9.09%, seven cases) in the 5-mg group and four subjects (10.53%, four cases) in the 10-mg group (Table5). Among the 15 ADRs in ten patients in all DA-1229 groups, most were nervous system disorders, such as dizziness and headache (two events in the 2.5-mg group, four events in the 5-mg group, and one event in the 10-mg group). The remaining eight ADRs were abdominal discomfort and chest discomfort in the 2.5-mg group; insomnia, sleep disorder and urticaria in the 5-mg group; and hyperglycaemia, pruritus and urticaria in the 10-mg group (one case, each). Of the 15 ADRs in the drug-treated groups, 12 were mild in severity, and the remaining three were moderate and included hyperglycaemia, insomnia and urticaria. There were no statistically significant differences among treatment groups in the incidences of overall AEs and ADRs (p = 0.6561 for AEs and p = 0.7992 for ADRs, Table5). There were no hypoglycaemic events. and there were no clinically significant findings with respect to vital signs, ECG or laboratory tests.

Discussion

This study on Korean subjects with type 2 diabetes mellitus demonstrated that patients treated with DA-1229 experienced significantly reduced HbA1c levels after 12 weeks compared to those who received a placebo. Although the reduction in HbA1c from baseline was greatest in the 5-mg group, there was no statistical difference between the dosage groups. These results indicate that a dose-dependent reduction in HbA1c may not occur at doses greater than 5 mg. The degree of HbA1c reduction was more profound in the subgroup with a higher baseline HbA1c [>8.5% (69 mmol/mol)]. Insulin secretory function as assessed by the insulinogenic index improved significantly in the 10-mg group compared to the placebo group. Other markers of insulin secretory function, including 2 h OGTT C-peptide and post-OGTT C-peptide AUC0–2h, also showed significant increases in the 2.5 and 5-mg groups for 2 h OGTT C-peptide and in all three groups for post-OGTT C-peptide AUC0–2h compared to the placebo group. Post-OGTT glucose AUC0–2h significantly decreased in the 5 and 10-mg groups compared to the placebo group. In summary, efficacy was demonstrated in all DA-1229 groups compared to the placebo group in the primary efficacy evaluation, and the greatest efficacy was shown in the 5-mg group. In the secondary efficacy evaluation, significant changes were demonstrated in all DA-1229 groups, with the greatest efficacy in the 5-mg group. However, there were no significant changes in lipid parameters and body weight in any group. Furthermore, the incidence of AEs was similar in all groups, including the placebo group. Based on the results of this phase II trial, the optimal dose for this drug in the treatment of type 2 diabetes mellitus is 5 mg once daily.

In our study, the administration of DA-1229 (5 mg) once daily significantly reduced the mean HbA1c by 0.57% compared with the placebo in Korean subjects with type 2 diabetes mellitus after 12 weeks of treatment (Figure2). This degree of HbA1c reduction by DA-1229 was comparable to other DPP-IV inhibitors 14–16. In a phase II trial involving sitagliptin (100 mg daily), the mean HbA1c was significantly reduced by approximately 0.55% compared with the placebo after 12 weeks of treatment 14. In a 12-week phase II trial of vildagliptin (100 mg daily), a 0.53% reduction in the mean HbA1c after 12 weeks of treatment was observed 15. The significant HbA1c-lowering effect of DA-1229 was supported by a large proportion of patients achieving an HbA1c of less than 7.0% (53 mmol/mol) and 6.5% (48 mmol/mol), lower FPG and GA levels (Table3) and decreased post-OGTT glucose AUC0–2h results compared with the placebo group (Table4). These effects were the greatest in subjects who received 5 mg of DA-1229; therefore, the optimal dose for this drug in type 2 diabetes mellitus patients should be 5 mg.

A high baseline HbA1c level was significantly associated with a greater reduction in HbA1c after DA-1229 treatment. When subgroup analysis was performed for the baseline HbA1c subgroup [>8.5% (69 mmol/mol) or ≤8.5% (69 mmol/mol)], there was a tendency towards a greater treatment effect in the subgroup with the higher baseline HbA1c (Online supplementary Table S1). These results are in line with those of trials of other DPP-IV inhibitors 17, and with results from studies using hypoglycaemia agents other than DPP-IV inhibitors 18.

Recent in vitro and in vivo experiments demonstrated that DPP-IV inhibitors have an islet-preserving effect through the proliferation and prevention of apoptosis of pancreatic β cells 19,20. This beneficial effect of DPP-IV inhibitors on pancreatic β cells has largely been attributed to an increase in the GLP-1 level mediated by the inhibition of the DPP-IV enzyme 19,20. In our study, insulin secretory function, as measured by the insulinogenic index (Table3) and post-OGTT C-peptide AUC0–2h (Table4), was significantly improved in the 10-mg group and in all DA-1229 groups, respectively, compared with the placebo group. Although we did not measure changes in GLP-1 levels before and after treatment with DA-1229, these findings are in agreement with results from other DPP-IV inhibitors 16,21. The long-term effects of this drug on human pancreatic β cell function need to be investigated further.

The primary physiological stimuli for the secretion of GLP-1 are fat- and carbohydrate-rich meals, but mixed meals or individual nutrients, including glucose and other sugars, sweeteners, fatty acids, amino acids and dietary fiber, can also stimulate GLP-1 secretion 22. Therefore, a mixed meal tolerance test (MMTT) appears to be more appropriate and physiological than OGTT. However, we used an OGTT to measure ‘glucose-dependent’ insulin release and increased insulin synthesis in response to our study drug similarly to previous studies 3,23. Furthermore, the protocol for MMTT for the measurement of the incretin effect has not been standardized yet.

Growing evidence demonstrated that GA, an intermediate-term glycaemic index, in conjunction with the GA/HbA1c ratio might be more accurate than HbA1c alone for assessing insulin secretory dysfunction, which resulted in glycaemic fluctuation and variability 24. In our study, all three doses of DA-1229 significantly reduced the GA/HbA1c ratio compared with the placebo group (Online supplementary Table S3). These results might indicate a beneficial effect of DA-1229 on glycaemic fluctuations, which is considered to be the third component of dysglycaemia along with hyperglycaemia at fasting and hyperglycaemia during postprandial periods 25. Further studies aimed at comparing the effects of DA-1229 with other DPP-IV inhibitors on glycaemic fluctuations are warranted.

Treatment with DA-1229 was well tolerated in this clinical trial. The mean treatment compliance ranged from 93.35% to 95.41% across all subjects. Of those treated with DA-1229 (n = 121), 39 patients (32.23%) experienced at least one AE, although most AEs were mild in severity (Table5). Although three SAEs occurred in three patients in different DA-1229 groups (two cases of rotator cuff syndrome and one case of hemorrhoids), they were unrelated to the study drug. The incidence of ADR was generally similar across DA-1229 and placebo treatment groups (Table5). ADRs occurred in ten patients in drug treatment groups, most of which were nervous system symptoms. There were no clinically significant hypoglycaemic events. The low incidence of hypoglycaemia observed with DA-1229 treatment, despite effective glucose lowering and stimulation of insulin release, is consistent with evidence that GLP-1 stimulates insulin release in a glucose-dependent manner 26. Treatment with DA-1229 had no effect on body weight relative to the placebo, similar to other DPP-IV inhibitors 9.

In summary, once-daily administration of DA-1229, a novel DPP-IV inhibitor, resulted in significant glucose-lowering effects for 12 weeks in Korean subjects with type 2 diabetes mellitus. Of the DA-1229 doses examined in this study, treatment with 5 mg of DA-1229 showed the greatest reduction in HbA1c, and this reduction is comparable to those achieved with other currently prescribed DPP-IV inhibitors 14–16. DA-1229 was also generally well tolerated, with no hypoglycaemic events and no changes in body weight. Based on this favorable clinical profile, it was determined that once-daily administration of 5 mg of DA-1229 is the optimal clinical dose, and it is expected that this drug has a high potential to be developed as another treatment modality for type 2 diabetes mellitus. The long-term efficacy of DA-1229 and its glucose-lowering effect combined with other hypoglycaemic agents need to be determined through additional studies.