Ipragliflozin Add-on Therapy to a GLP-1 Receptor Agonist in Japanese Patients with Type 2 Diabetes (AGATE): A 52-Week Open-Label Study

This multicenter (Electronic Supplementary Material 1) open-label study consisted of three periods: a 4-week washout period (only for patients using glucose-lowering agents other than GLP-1 receptor agonists and SUs), a 6-week observation period, and a 52-week open-label treatment period. During the observation period, we confirmed the stability of glycosylated hemoglobin (HbA1c). Following the observation period, patients were registered and began the open-label treatment period. The dose of the study drug was increased according to the dose-escalation criteria defined in this study. The 52-week open-label treatment period comprised an initial 20-week period of administration of ipragliflozin at a fixed dose of 50 mg/day. Patients with an HbA1c of > 7.0% at week 16 (considered to be poor responders to the 50-mg ipragliflozin dose) were eligible for a dose increase to 100 mg/day starting on week 20. In the following 32 weeks, patients received ipragliflozin at either 50 mg or 100 mg/day. The study design is shown in Fig. 1.

The study was conducted at 19 sites in Japan. All procedures performed in the study were in accordance with the ethical standards of the institutional review board of each site and with the 1964 Helsinki declaration and its later amendments. Additionally, the study complied with all relevant laws and regulations, including Good Clinical Practice and the International Conference on Harmonisation of Technical Requirements for Registration of Pharmaceuticals for Human Use Guidelines. Informed consent was obtained from all individual participants included in the study. This study was registered at ClinicalTrials.gov (identifier: NCT02291874).

The inclusion criteria were as follows: age ≥ 20 years; diagnosis of type 2 diabetes ≥ 12 weeks before enrollment; prescription of a stable dose/regimen of a GLP-1 receptor agonist or concomitant prescription of a stable dose/regimen of a GLP-1 receptor agonist and SU for ≥ 6 weeks; HbA1c ≥  7.5% and ≤ 10.5%, with a maximum change in HbA1c of ± 1% at visit 1 (week − 6) and visit 2 (week − 2); body mass index (BMI) of between 20.0 and 45.0 kg/m²; and fasting plasma glucose (FPG) level of ≥ 126 mg/dL at visit 2 (week − 2) for those receiving SU therapy.

The main exclusion criteria were as follows: diagnosis of type 1 diabetes mellitus; urinary symptoms (dysuria, anuria, oliguria, and urinary retention); history of recurrent urinary tract or genital infection; proliferative retinopathy; renal disease; history of pancreatitis or cerebrovascular or cardiovascular diseases; chronic disease(s) that required continuous use of non-topical corticosteroids or immunosuppressants; severe infections; malignant tumors; alcoholism or other addictions; unstable psychiatric disorder; women with child-bearing potential not using contraceptives appropriately; history of allergy to ipragliflozin and/or similar drugs; previous participation in a clinical study or postmarketing surveillance within 12 weeks of enrollment; and patients who were unable or unwilling to adhere to the study procedures.

Patients could be withdrawn at any time during the study for a number of reasons, including safety concerns and severe hypoglycemia (hypoglycemic coma, convulsions, or other conditions requiring infusion or injection of glucose or glucagon). Additionally, patients were withdrawn if their HbA1c was > 8.5% on two consecutive visits from week 24 onwards and if the FPG level exceeded 270 mg/dL on two consecutive measurements.

Patients underwent liraglutide monotherapy or treatment with liraglutide or other GLP-1 receptor agonist in combination with an SU at a constant dose/dose regimen for at least 6 weeks before visit 1. Visit 1 (week − 6) was then followed by the 6-week observation period, registration, and enrollment into the open-label treatment period. At the beginning of the 20-week open-label treatment period, all patients received 50 mg/day of ipragliflozin. At week 20, the ipragliflozin dose could be increased to 100 mg/day in patients whose HbA1c was ≥ 7.0% (poor responders) and for whom the investigators considered there were no safety concerns. Treatment was discontinued in patients whose HbA1c was > 8.5% at two consecutive visits from visit 10 (week 24) onwards. From the observation period to the end of the treatment period, the type or dose/dose regimen of GLP-1 receptor agonists and SU could not be changed. Once the dose of ipragliflozin was escalated, the patient continued to receive ipragliflozin at the escalated dose (100 mg/day), in principle, until the end of the treatment period. However, if safety concerns were raised after the dose increase, the dose could be reduced to 50 mg/day, but no further dose adjustment was allowed after the dose reduction.

Patients performed self-monitoring of fasting blood glucose every morning, or if they felt symptoms of hypoglycemia, and recorded the values in their diary. Treatment compliance was assessed by the investigator based on the patient diary, dispensed study drugs, number of collected study drugs, and number of lost study drugs. The use of glucose-lowering agents and insulin was prohibited, except for the use of ipragliflozin, GLP-1 receptor agonists, and SUs. In Japan, liraglutide is approved as monotherapy or as combination therapy with SUs, whereas other GLP-1 receptor agonists are approved for use only in combination with SUs. Because the Japanese clinical guidelines do not recommend any specific class of antidiabetic drug (including biguanides) as the first-line therapy, metformin was not included among the permitted concomitant oral antidiabetic drugs in this study. Temporary or topical use of corticosteroids or immunosuppressants was allowed.

The changes from baseline (week 0) in HbA1c, FPG, fasting insulin, leptin, adiponectin, glucagon, body weight, waist circumference, homeostasis model assessment (HOMA)-R, HOMA-beta, and self-monitored blood glucose (SMBG) values recorded in the patient diary were examined. HbA1c, FPG, fasting insulin, leptin, adiponectin, and glucagon levels were measured by LSI Medience Corporation (Tokyo, Japan). For HbA1c levels, National Glycohemoglobin Standardization Program values were used. SMBG was performed seven times a day (fasting after waking up, 1 h after the start of breakfast, before lunch, 1 h after the start of lunch, before dinner, 1 h after the start of dinner, and before bedtime) for 3 days per week before each scheduled visit, and the measured value, time of measurement, start time of each meal, and dinner end time were recorded in the diary.

Safety was assessed in terms of vital signs, treatment-emergent adverse events (TEAEs), and laboratory tests. TEAEs were defined as AEs observed after the first administration of ipragliflozin and were classified according to system organ class and preferred term (MedDRA version 1.61; https://​www.​meddra.​org/​how-to-use/​support-documentation); their relationship to the study drug, seriousness, and severity were evaluated.

An estimated sample size of 100 patients (approximately 70 patients to receive ipragliflozin in combination with a GLP-1 receptor agonist and approximately 30 patients to receive ipragliflozin in combination with a GLP-1 receptor agonist and SU) was planned based on the Guidelines for Clinical Evaluation of Oral Antihyperglycemic Drugs (Notification no. 0709-1 of the Evaluation and Licensing Division, Pharmaceutical and Food Safety Bureau, dated July 9, 2010) [16], in which 50–100 patients per group are recommended to ensure a sufficient number of patients to evaluate the safety of the test drug.

The study analysis sets were the full analysis set (all patients who received at least one dose of ipragliflozin, and in whom at least one efficacy variable was measured after administration of ipragliflozin) and the safety analysis set (all patients who received at least one dose of ipragliflozin). Overall data were analyzed, and data for both subgroups separately: 50/50 mg group (in which the dose of ipragliflozin was maintained at week 20 because of sufficient efficacy or safety concerns) and the 50/100 mg group.

Missing data were not imputed, except for the month of T2DM diagnosis and partial onset dates of AEs. Baseline characteristics were summarized descriptively using mean ± standard deviation (SD) or the number and percentage of patients for continuous and categorical variables, respectively. Efficacy data were analyzed descriptively in terms of the mean ± SD values at each time point together with the mean changes from baseline to each time point. The mean values and changes in efficacy variables from baseline to the end of treatment using the last observation carried forward were also calculated. TEAEs were presented as the number and percentage of patients in each treatment group. The statistical analyses were performed by Astellas Pharma Inc. using Version 9.2 or later of Windows SAS® (SAS Institute Inc., Cary, NC, USA).

The baseline characteristics of patients are shown in Table 1. Overall, the majority of patients were male (72/103, 69.9%), had a mean (± SD) age of 53.4 ± 10.5 years at the time of consent, BMI of 28.09 ± 4.15 kg/m², HbA1c of 8.81 ± 0.89%, FPG of 185.9 ± 44.2 mg/dL, and estimated glomerular filtration rate of 86.69 ± 18.39 mL/min/1.73 m². More than one-half of the patients (57 [55.3%]) had diabetes for < 120 months (10 years). The majority (90/103 [87.4%]) had been treated with liraglutide monotherapy, with only 13/103 (12.6%) treated with liraglutide or another GLP-1 receptor agonist in combination with an SU. The percentage of patients who underwent washout of glucose-lowering agents other than GLP-1 receptor agonists and SU agents was 18.4% (19/103). When both groups were compared, the 50/50 mg group had numerically lower baseline body weight, BMI, FPG, and HbA1c values, and shorter diabetes mellitus duration compared with the 50/100 mg group.

As shown in the patient flow diagram (Fig. 2), a total of 103 patients received ipragliflozin 50 mg/day, three of whom discontinued the treatment by week 20 because of an AE, and 100 of whom continued ipragliflozin 50 mg/day until week 20. Of these, 67 patients had not achieved adequate glycemic control, and their ipragliflozin dose was increased to 100 mg/day (50/100 mg group), while 33 patients maintained the initial ipragliflozin dose (50/50 mg group). Among patients in the 50/50 mg group, 32 completed the study, and one patient discontinued because of lack of efficacy. Among those in the 50/100 mg group, two patients underwent a dose decrease to 50 mg/day after week 20 because of AEs (one patient presented fatigue and the other, pollakiuria); both patients completed the study. Of the remaining 65 patients in the 50/100 mg group, 49 patients completed the study and 16 discontinued (14 for lack of efficacy, 1 for an adverse event, and 1 patient withdrew voluntarily).

The mean (± SD) change in HbA1c from baseline to end of treatment was − 0.92 ± 0.80% (95% confidence interval [CI] − 1.07, − 0.76]) in all ipragliflozin-treated patients (n = 103). The mean changes in HbA1c from baseline to week 20 and from baseline to end of treatment were − 1.18 ± 0.66% and − 0.84 ± 0.75%, respectively, in the 50/50 mg group and − 0.95 ± 0.71% and − 0.97 ± 0.83%, respectively, in the 50/100 mg group (Table 2). These values reflect an increase of 0.33% from week 20 to end of treatment in the 50/50 mg group, but a decrease of 0.02% in the 50/100 mg group.

The time courses of HbA1c and changes in HbA1c from baseline in each dose subgroup are shown in Fig. 3a, b. The 50-mg dose was administered until week 20, and HbA1c decreased in both subgroups. In the 50/50 mg group, the dose remained at 50 mg/day at week 20, and HbA1c slightly increased until week 52. In the 50/100 mg group, the dose was increased to 100 mg/day at week 20, and HbA1c further decreased until week 36, with a subsequent slight increase until week 52.

The mean (± SD) change in FPG from baseline to the end of treatment was − 38.9 ± 39.0 mg/dL (95% CI − 46.5, − 31.3) (Table 2). The time courses of FPG and changes in FPG from baseline in each dose subgroup are shown in Fig. 4a, b. In each dose subgroup, FPG decreased rapidly within 2 weeks of starting treatment with ipragliflozin. In the 50/50 mg group, the FPG level continued to decrease gradually up to week 20 and remained almost unchanged thereafter. In the 50/100 mg group, the FPG level continued to decrease gradually up to week 32 and remained almost unchanged thereafter. The decrease in FPG until week 52 was more marked in the 50/100 mg group.

The mean (± SD) change in SMBG from baseline to the end of treatment in all patients was − 38.9 ± 41.3 mg/dL in the fasting state, − 52.5 ± 61.2 mg/dL 1 h after breakfast, − 47.3 ± 53.5 mg/dL before lunch, − 51.3 ± 69.1 mg/dL 1 h after lunch, − 35.9 ± 46.8 mg/dL before dinner, − 40.9 ± 63.7 mg/dL 1 h after dinner, and − 54.9 ± 65.3 mg/dL before bedtime (Table 2). Figure 5 shows the SMBG profile in each dose subgroup. A sustained reduction from baseline in SMBG was observed over the 52 weeks of treatment in both ipragliflozin dose subgroups at all measurement time points. The same sustained reduction in SMBG was reflected by the changes in the seven-point SMBG profile from baseline to the end of treatment in both ipragliflozin dose subgroups, as shown in Fig. 6.

Table 2 shows the changes in body weight, waist circumference, and metabolic parameters from baseline to the end of treatment in all ipragliflozin-treated patients and according to the dose of ipragliflozin used. The mean (SD) change in body weight from baseline to the end of treatment in all patients was − 2.69 ± 2.39 kg (95% CI − 3.16, − 2.22). The mean change in body weight from baseline to the end of treatment in the 50/50 mg group was − 2.53 ± 1.48 kg and that in the 50/100 mg group was − 2.77 ± 2.76 kg. The time courses of body weight and changes in body weight from baseline in each dose subgroup are shown in Fig. 7a, b. During the course of the 52-week treatment, body weight decreased progressively in each ipragliflozin dose subgroup. In the 50/50 mg group, a marked decline in body weight was observed up to week 16, after which it showed a tendency to increase, and then remained approximately stable up to week 52. In the 50/100 mg group, body weight declined more gradually, with a further reduction after the dose was increased to 100 mg/day at week 20.

Glucagon increased from baseline to the end of treatment; the increase was 23.1 ± 27.6 pg/mL (95% CI 17.7, 28.5) in all patients. Leptin was unchanged in all patients (− 0.19 ± 4.29 ng/mL; 95% CI − 1.03, 0.65). Adiponectin increased in all patients (0.98 ± 1.63 μg/mL; 95% CI 0.66, 1.30) (Table 2).

Treatment-emergent adverse events occurring in all patients who received ipragliflozin are shown in Table 3. TEAEs occurred in 77.7% (80/103) of patients treated with ipragliflozin. Overall, 46.6% (48/103) of patients experienced drug-related TEAEs. Serious TEAEs occurred in 3.9% (4/103) of patients and included glaucoma, nasopharyngitis, cerebral infarction, and sleep apnea syndrome (in one patient each). Hospitalization was required in three cases (glaucoma, nasopharyngitis, and sleep apnea syndrome), and cerebral infarction was regarded as an event of medical significance. Of these, glaucoma and nasopharyngitis were considered by the investigator as not related to the study drug. TEAEs leading to permanent discontinuation occurred in 3.9% (4/103) of patients and included constipation, hypoglycemia, cerebral infarction, and rash (in one patient each). The latter patient developed a rash on day 29 after starting treatment, which disappeared completely (resolved) by day 50. The patient continued medical treatment for the rash up to day 47. The severity of the rash was mild, and it was considered as “probably related” to the study treatment. By preferred term, the most common drug-related TEAEs in all patients were pollakiuria (9.7% [10/103]), hypoglycemia (8.7% [9/103]), constipation (6.8% [7/103]), and thirst (5.8% [6/103]). Administration of ipragliflozin was not associated with any clinically significant changes in laboratory variables or vital signs, except for increased blood ketone bodies. No deaths were reported during this study.