Sitagliptin Attenuates the Progression of Carotid Intima-Media Thickening in Insulin-Treated Patients With Type 2 Diabetes: The Sitagliptin Preventive Study of Intima-Media Thickness Evaluation (SPIKE): A Randomized Controlled Trial

Type 2 diabetes mellitus (T2DM) increases the likelihood of developing cardiovascular disease (CVD), which is one of the major causes of mortality and morbidity in these patients (1). Recent studies have cast doubt on the benefits of strict glycemic control, especially using insulin, on CVD in patients with established atherosclerosis or long-standing T2DM (2–4) for two main reasons: 1) It is assumed that frequent episodes of severe hypoglycemia might reduce their beneficial effects (5), and 2) weight gain may adversely affect the prognosis. Therefore, reducing insulin dose by stimulating endogenous insulin secretion and increasing insulin sensitivity using oral hypoglycemic agents (OHAs) (6,7) may reduce these adverse effects of insulin therapy. However, the effects of treatment with insulin plus other traditional OHAs on the progression of atherosclerosis remain largely unknown. New OHAs with the least risk of undeliverable effects and multiple beneficial effects on cardiovascular (CV) profiles when used with insulin are essential for the treatment of T2DM.

Sitagliptin, a dipeptidyl peptidase 4 (DPP-4) inhibitor, increases insulin secretion and suppresses glucagon release by protecting the degradation of incretins without increased risk of hypoglycemia and clinically relevant weight gain (8,9). Furthermore, the addition of DPP-4 inhibitors to ongoing insulin therapy was reported to have specific advantages in reducing the frequency of hypoglycemia and weight gain, in addition to the expected benefits associated with glycemic control and limiting insulin dose (10–12). In addition, evidence from preclinical studies suggests that DPP-4 inhibitors could have beneficial effects on atherosclerosis in both GLP-1-dependent and -independent manners (13–15). Thus the addition of DPP-4 inhibitors to insulin therapy is expected to have beneficial effects on CVD in patients with T2DM.

This study is a multicenter, prospective, randomized, open-label, blinded end point (PROBE) study designed to determine the effect of adding sitagliptin to insulin treatment (sitagliptin group) compared with conventional treatment (non-sitagliptin group) on carotid artery intima-media thickness (IMT) in CVD-free patients with T2DM. IMT, a marker of progression of atherosclerosis, was evaluated by a simple and noninvasive procedure.

Japanese patients with T2DM who regularly attended the outpatient diabetes clinics at 12 institutions in Japan were asked to participate in this study, as previously described in detail (16). Patients with T2DM in whom the target of blood glucose control specified in the Treatment Guide for Diabetes in 2010 (edited by the Japan Diabetes Society) (17) was not achieved despite insulin therapy, in addition to dietary/exercise therapy or concomitant therapeutic drugs for T2DM (other than DPP-4 inhibitors) over a period of ≥3 months, were included in the study. Patients who withdrew from previous treatment with DPP-4 inhibitor for more than 12 weeks were also included; additional inclusion criteria included ≥30 but <80 years of age (regardless of sex) and signing the consent form for participation in the study. The following exclusion criteria also were applied: 1) type 1 or secondary diabetes; 2) presence of severe infectious disease, before or after surgery, or severe trauma; 3) history of myocardial infarction, angina pectoris, cerebral stroke, or cerebral infarction; 4) retinopathy requiring laser photocoagulation and/or vitrectomy or history of these treatments within 1 year; 5) moderate or severe renal dysfunction (serum creatinine: males, >1.4 mg/dL; females, >1.2 mg/dL); 6) severe liver dysfunction (aspartate aminotransferase ≥100 IU/L); 7) moderate or severe heart failure (New York Heart Association stage III severity or higher); 8) treatment with an incretin preparation, such as other DPP-4 inhibitors, at the start of the study; 9) treatment with drugs that are not concomitantly administrable with incretin preparations with regard to the national health insurance, such as DPP-4 inhibitors, at the start of the study; 10) pregnant, lactating, or possibly pregnant women or those planning to become pregnant during the study period; 11) medical history of hypersensitivity to investigational drugs; and 12) those patients judged as ineligible by the clinical investigators.

The subjects were screened consecutively, and patients who met the above eligibility criteria were asked to participate. All patients who agreed to participate were entered into the study. The protocol was approved by the institutional review board of each participating institution, in compliance with the Declaration of Helsinki and current legal regulations in Japan. Written informed consent was obtained from all participants after a full explanation of the study. This study is registered in the University Hospital Medical Information Network Clinical Trials Registry (UMIN000007396), which is a nonprofit organization in Japan, and meets the requirements of the International Committee of Medical Journal Editors.

Patients were registered at the administration office of this trial via the Internet; once enrolled, they were randomly assigned to either the sitagliptin group or the control group receiving conventional treatment consisting of drugs other than DPP-4 inhibitors. Randomization was performed using a dynamic allocation method based on the number of insulin injections, with/without concomitant pioglitazone, age, and sex.

Patients in the sitagliptin group were started on sitagliptin (25 mg once daily) in addition to ongoing insulin therapy. The dose of sulfonylurea was tapered when considered clinically appropriate to avoid hypoglycemia when starting sitagliptin. Initiation of treatment with sitagliptin 50 mg once daily was permitted in patients who were not treated with a sulfonylurea. In patients treated with sitagliptin 25 or 50 mg once daily for 12 weeks, the dose of sitagliptin was increased to a maximum of 100 mg once daily when HbA_1c was ≥7.0% (8.6 mmol/mol) (17). The participating physicians were allowed to reduce sitagliptin to 25 or 50 mg/day if treatment with 50 or 100 mg/day was considered poorly tolerated. The insulin dose could also be adjusted, with priority given to achieve fasting blood glucose or 2-h postprandial blood glucose as recommended in the Treatment Guide for Diabetes (17). In the control group, either increasing the dose of current therapy (e.g., insulin) or adding a sulfonylurea, glinide, or α-glucosidase inhibitor was allowed, with the goal of achieving the target value specified in the Treatment Guide for Diabetes (usually HbA_1c <6.9% [8.44 mmol/mol]) (17). The addition of other DPP-4 inhibitors or GLP-1 analogs was banned in the control group. Dose adjustment and addition of metformin and pioglitazone during the study were banned in both groups. In cases of hypoglycemia, the dose of insulin and/or OHA was titrated. The use of antihyperlipidemic and antihypertensive drugs was allowed during the study.

The study period was 104 weeks after the patients were registered. All randomized participants were followed for 104 weeks until the end of the study, regardless of adherence to or discontinuation of study medication for any reason. Clinical outcome, adherence, and adverse events (AEs) were confirmed, and clinical and biochemical data were collected at 0, 26, 52, 78, and 104 weeks after randomization.

The primary study outcomes were changes in mean IMT of the common carotid artery (mean-IMT-CCA) and maximum IMT of the common carotid artery (max-IMT-CCA) after the 104-week treatment period. Investigations were carried out at the start of the study, after 52 weeks, and after 104 weeks.

AEs were recorded during study as described in the Supplementary Data online.

Ultrasonographic scans of the carotid artery were performed by expert sonographers who were specifically trained to perform the prescribed study examination, as reported previously (16). To avoid intersonographer variability, each participant was examined by the same sonographer with the same equipment (high-resolution B-mode ultrasound scanner equipped with a high-frequency [>7.5-MHz] linear transducer, with a limit of detection of <0.1 mm) throughout all the visits. The mean intrainvestigator coefficient of variation for measurement of mean-IMT-CCA and max-IMT-CCA (± SD) in each institution was 1.6 ± 1.2% and 2.4 ± 2.3%, respectively. The extracranial common carotid artery, the carotid bulb, and the internal carotid artery in the neck were scanned bilaterally in at least three different longitudinal projections (anterior, lateral, and posterior, which approximately corresponded to 60, 90, 150 degrees for the right carotid artery, and 210, 270, and 300 degrees for the left carotid artery marked on the Meijer arc) as well as transverse projections, and the site of greatest thickness, including plaque lesions, was identified along the arterial walls. IMT represents the distance between two parallel echogenic lines corresponding to the vascular lumen and the adventitia. To avoid interreader variability, all scans were electronically stored and e-mailed to the central office (IMT Evaluation Committee, Osaka, Japan) to be read, in a random order, by a single experienced reader blinded to the clinical characteristics of the patients using automated, digital edge-detection software (IntimaScope; Media Cross, Tokyo, Japan) (16). The software system averages 60 points of IMT values in the segment 2 cm proximal to the dilation of the carotid bulb (mean-IMT-CCA). In addition, the largest IMTs, including plaque lesions in the common carotid arteries (max-IMT-CCA), were also measured separately. Reproducibility analysis of replicate measurements in 20 subjects yielded absolute mean differences of 0.02 ± 0.01 and 0.01 ± 0.01 for mean-IMT-CCA and max-IMT-CCA, respectively. The intraobserver coefficient of variation for the measurement was 1.1% and 0.7%, respectively.

Yokoyama et al. (18) reported previously that the mean ± SD rate of increase in carotid IMT in Japanese patients with diabetes was 0.034 ± 0.054 mm/year and that 1% improvement in HbA_1c was associated with a 0.02-mm/year improvement in IMT. Given these results, it was assumed that during a 2-year observation period, registration of at least 232 patients was required to obtain 80% power to detect a difference of 0.04 mm in IMT between the two treatment groups, assuming an SD of 0.108, 15% dropout, and a 0.05 level of significance. On the basis of this calculation, the target number of enrolled patients was set at 274 for the 2-year registration period.

Efficacy was analyzed using mainly the full analysis set based on the intent-to-treat principle and secondarily using the protocol set. Primary analysis was performed using a mixed-effects model for repeated measures, with treatment group, time (weeks), interactions between treatment group and time (weeks), and baseline IMT as fixed effects; an unstructured covariate was used to model the covariance of within-subject variability. For the occurrence of CV events, time to onset (one of the secondary end points) was analyzed using a log-rank test and Cox proportional hazards model.

Baseline and follow-up group comparisons were assessed with the Student t test or Wilcoxon rank sum test for continuous variables, and the Fisher exact test for categorical variables. Changes from baseline to treatment visits were assessed with the one-sample t test and Wilcoxon signed rank test within the group. The number and percentage of patients reporting AEs were presented by treatment group and compared between the two treatment groups using Fisher exact test. All statistical tests were two-sided with a 5% significance level. All analyses were performed using SAS software version 9.4 (SAS Institute, Cary, NC).

A total of 282 participants were randomly allocated to either the sitagliptin group (n = 142) or the conventional treatment group (n = 140). After excluding 8 patients from analyses (they withdrew from the study and/or objected to the inclusion of their data in the analysis), 137 in the sitagliptin group and 137 in the conventional treatment group were included in the full analysis set (Supplementary Fig. 1). As shown in Tables 1, 2, and 3, baseline clinical characteristics, including potential risk factors for carotid atherosclerosis, were comparable between the groups.

In the mixed-effects model for repeated measures in the full-set analysis, sitagliptin significantly slowed the worsening of the mean-IMT-CCA and left max-IMT-CCA, but not right max-IMT-CCA, at 104 weeks (i.e., primary end points of the study) compared with conventional treatment (Table 2). Similar findings were noted even in the adjusted mixed-effect models, including model 1 (age and sex); model 2 (model 1 plus BMI, HbA_1c, total cholesterol, HDL cholesterol, triglycerides, and systolic blood pressure); model 3 (model 2 plus smoking, ACE/angiotensin II receptor blocker, statin, and antiplatelets); and model 4 (model 3 plus OHA) (Supplementary Table 1). Also, ANCOVA models that included treatment group, age, sex, baseline IMT, systolic blood pressure, and administration of statins produced findings similar to those of the mixed-effects models (Supplementary Table 2). Over 104 weeks, sitagliptin treatment, but not conventional treatment, significantly reduced the mean-IMT-CCA and left max-IMT-CCA, but not right max-IMT-CCA, relative to baseline values (Table 2).

Although both types of treatment significantly reduced HbA_1c levels from baseline, the improvement in HbA_1c (value at end of study − value at baseline) was significantly greater in the sitagliptin group (−0.5 ± 1.0%) than the conventional group (−0.2 ± 0.9%; P = 0.004) (Table 3). Similarly, there was a significant difference in the change in serum C-peptide concentrations (value at end of study − value at baseline); serum C-peptide concentration was significantly increased in the sitagliptin group but not in the conventional group (Table 3). There were no differences in the value and changes in fasting blood glucose concentrations at each visit during the observation period. Notably, the use of glinides was significantly higher during the study, and the use of α-glucosidase inhibitors tended to be higher among the conventional group than among the sitagliptin group at the end of the study (Supplementary Table 3).

Similarly, there were no differences between the two groups in other risk factors for atherosclerosis, such as BMI, blood pressure, and lipid parameters, at baseline or in their changes during the observation period (Table 3). Blood pressure and lipid metabolism were well controlled in both groups throughout the study. There were also no significant differences in the use of antihypertensive drugs and lipid-lowering agents (Supplementary Table 4). Also, no significant changes were observed in the levels of various markers of inflammation and endothelial damage (Table 4).

During the study, 123 patients (65 in the sitagliptin group and 58 in the conventional group) developed AEs, and 17 patients (8 and 9, respectively) developed serious AEs. There were no significant differences in the incidence of any AEs and serious AEs between the two groups. We recorded 104 hypoglycemic events (in 52 patients in the sitagliptin group and 52 patients in the conventional treatment group) (Supplementary Table 5). There was no significant difference in the average number of hypoglycemic events between the two groups (0.34 ± 0.85 times/month/person in the sitagliptin group vs. 0.36 ± 0.80 times/month/person in the conventional treatment group). One severe hypoglycemic event occurred in each group. Five patients developed CV events, and 6 patients were diagnosed with cancer. There was no significant difference in the incidence of CVD and cancer between the two groups.

This is the first PROBE trial to demonstrate that the addition of sitagliptin to insulin treatment significantly diminished the progression of mean-IMT-CCA and left max-IMT-CCA, but not right max-IMT-CCA, compared with conventional treatment. Because of these discrepant results, it might be viewed as an exaggeration if we were to conclude that sitagliptin attenuates the progression of carotid IMT. However, the mean-IMT-CCA is the most reliable parameter among these three prespecified primary outcomes of this study: It is generally believed that the mean-IMT-CCA is more reproducible compared with max-IMT-CCA and that max-IMT-CCA is more accurately measured on the left side than the right side since the left wall of the common carotid artery is thicker and more elastic than the right (19). Therefore, the positive results for the mean-IMT-CCA and left max-IMT-CCA are more reliable than the negative result on the right max-IMT-CCA, which could be a result of the underpowered sample. Accordingly, it is feasible to conclude with confidence that sitagliptin has a beneficial effect on carotid IMT in this population. It should be noted that mean-IMT-CCA and left max-IMT-CCA were significantly regressed in the sitagliptin group during the 104-week observation period.

Previous studies demonstrated that thiazolidinedione drugs had only partial effects on the prevention of carotid IMT progression in insulin-treated patients with T2DM (20,21), whereas they reduced carotid IMT in non-insulin-treated patients with T2DM (22,23). Thus it may be more difficult to slow carotid IMT progression in insulin-treated patients with T2DM than in others. In this regard, the higher rate of increase in carotid IMT among insulin-treated compared with non-insulin-treated patients with T2DM is probably related to the longer duration of T2DM, poor glycemic control, higher frequency of hypoglycemic episodes, increased chance of weight gain, and/or more advanced atherosclerosis. The higher rate of worsening of carotid IMT seems to correlate with the less-than-ideal effect of sitagliptin on carotid IMT progression at 52 weeks. In turn, treatment with sitagliptin resulted in regression of the mean-IMT-CCA and left max-IMT-CCA at the end of the sitagliptin treatment protocol. Thus long-term treatment may be required to lessen the worsening of carotid IMT in insulin-treated patients with T2DM.

In this study, the addition of sitagliptin to insulin therapy had a sustained glucose-lowering effect amounting to about 0.5% over 2 years from baseline, which is almost similar to previous reports (10–12,24). On the other hand, conventional treatment reduced HbA_1c level by only 0.2%, despite our effort to achieve an optimal glycemic target by adjusting the OHA dosage. In this regard, the dose of insulin was not increased during the study because of concerns about the potential AEs of insulin, such as hypoglycemia and weight gain. Accordingly, sitagliptin treatment was superior to conventional treatment in terms of HbA_1c reduction without increasing the incidence of hypoglycemic episodes and weight gain in insulin-treated patients, consistent with previous studies (10–12). However, the difference in HbA_1c between the two treatment groups does not seem to correlate with the reduction in carotid IMT, since changes in HbA_1c did not correlate with changes in mean-IMT-CCA and right and left max-IMT-CCA in both groups (data not shown). In addition, sitagliptin treatment still attenuated the progression of mean-IMT-CCA and left max-IMT-CCA compared with conventional treatment, even after adjusting for changes in HbA_1c from baseline (Supplementary Table 2). Thus the differences in carotid IMT progression could not be explained by the difference in HbA_1c, suggesting that these changes in the carotid artery could be independent of the glucose-lowering effects. In addition, it is possible that the frequent use of glinides and α-glucosidase inhibitors in the conventional treatment group over time affected the results. In this regard, the effect of sitagliptin on the reduction in carotid IMT may not be overestimated; previous reports demonstrated beneficial effects of those medications on carotid IMT (25,26). Indeed, almost similar findings were observed in ANCOVA models after adjusting for the use of glinides and α-glucosidase inhibitors (data not shown).

Previous reports demonstrated that short-term treatment with DPP-4 inhibitors decreased the levels of various markers of inflammation and endothelial injury in patients with T2DM (27,28). However, in this study of longer duration, no such beneficial effects of sitagliptin on these markers were found. Although the exact reason for these inconsistent findings is not clear, several factors could have affected the proinflammatory conditions, such as both transient and chronic nonatherosclerotic diseases and additional treatment within a relatively short time. In addition, a subanalysis of the Diabetes Control and Complications Trial/Epidemiology of Diabetes Interventions and Complications (DCCT/EDIC) demonstrated that the relationships between serum biomarkers of chronic inflammation and endothelial injury at multiple time points and the progression of carotid IMT remain unclear in patients with type 1 diabetes (29). Interestingly, a recent report suggested that evaluation of local inflammation and oxidative stress in the vascular wall, not in serum, was helpful in the full assessment of “non-glycaemic-dependent” antiatherosclerotic effects of DPP-4 inhibitors (30).

Although the exact mechanism by which DPP-4 inhibitors induce regression of carotid IMT remains uncertain at present, previous studies postulated a few scenarios. Inhibition of macrophage accumulation and inflammation by enhancing GLP-1 signaling with sitagliptin or GLP-1 receptor agonists reduced atherosclerotic lesion formation in a mouse model of atherosclerosis (13,31). Our group also demonstrated that another DPP-4 inhibitor suppressed macrophage-related inflammation and DPP-4-induced smooth muscle proliferation in vitro (14). Furthermore, Nagashima et al. (32) demonstrated that GLP-1 suppressed foam cell formation from murine macrophages. These mechanisms could potentially contribute to reduced atherosclerosis. In the future, measurement of local inflammation and the ability to form foam cells in human monocytes isolated from peripheral blood may be helpful in investigating the mechanism behind the beneficial effects of sitagliptin on carotid IMT.

Three recent randomized clinical studies showed that the use of DPP-4 inhibitors did not increase nor decrease the CVD event rate among patients with T2DM compared with placebo (33–35). Since these safety outcome studies were originally designed as noninferiority clinical trials, they probably lacked sufficient power to evaluate the beneficial effects of DPP-4 inhibitors on CVD. In general, the number of CV events required in a superiority trial is larger than that in a noninferiority trial (36). Therefore a larger-scale prospective clinical trial with a longer observation period is required to assess the CV efficacy of DPP-4 inhibitors. On the other hand, we focused on the effects of sitagliptin on the progression of carotid atherosclerosis, not CVD, as a first step because of practical constraints, including trial costs and concern about feasibility in relation to long-term intervention; progression or regression of carotid atherosclerosis remains a controversial surrogate for CV effects (37,38). In addition, the subjects of the above-mentioned three trials (29–31) were patients with T2DM with a history of CVD or at high risk for CVD. In comparison, our study enrolled insulin-treated subjects with T2DM who were free from apparent CVD. Considering the results of this study, we believe that early and effective intervention with DPP-4 inhibitors before the development of advanced atherosclerosis in patients without a history of apparent CVD is likely to be beneficial in the prevention of carotid IMT progression.

In addition to the limitations described above, we used a PROBE design, which may bias the assessment of outcomes. The reason for using an open-label study was practical constraints, including trial costs in an investigator-initiated trial. In an effort to overcome possible bias, a single experienced reader, who was blinded to the clinical characteristics of the subjects and type of treatment, measured carotid IMT using automated, digital edge-detection technology. Further, all statistical methods and data handling were prespecified under the blinded condition before the database was locked. Second, there may have been measurement errors in IMT as a result of intersonographer differences among institutions, which were not evaluated in this study. However, high reproducibility of the measurement of IMT by each sonographer may reduce these errors. Third, the chance of false-positive findings may increase because the multiplicity of testing for the primary end points was not specifically adjusted in our analysis. Finally, it is likely that other yet unknown factors and known factors such as change in lifestyle, duration and/or severity of CV risk factors, and family history of CVD that we did not evaluate in this study can explain the differences in changes in carotid IMT between the two treatment groups. Patients lost to follow-up are unlikely to have a major influence on the results because there were no significant differences in baseline characteristics and AEs between the two groups.

Compared with the conventional treatment, sitagliptin treatment attenuated the progression of carotid IMT in insulin-treated patients with T2DM who were free of a history of apparent CVD. Although our data suggest a promising effect of sitagliptin on carotid IMT progression, a large-scale prospective trial is required to establish the usefulness of DPP-4 inhibitors in the primary prevention of CVD in patients with T2DM.

Acknowledgments. The authors thank all the clinical staff and members of the committee (Yoshimitsu Yamasaki, Nishi-Umeda Clinic for Asian Medical Collaboration; Kazunori Shimada and Hirotoshi Ohmura, Department of Cardiovascular Medicine, Juntendo University Graduate School of Medicine; and Ryota Tanaka, Department of Neurology, Juntendo University Graduate School of Medicine) for their assistance with the execution and completion of the clinical trial.

Funding. Financial support for this study was provided by the Japan Society for Patient Reported Outcome research fund from Mitsubishi Tanabe, Ono, and Novo Nordisk.

Duality of Interest. T.M. received research funds from Merck Sharp and Dohme K.K. and Takeda Pharmaceutical Co. and has received lecture fees from AstraZeneca, Boehringer Ingelheim, Eli Lilly & Co., Kowa Pharmaceutical Co. Ltd., Mitsubishi Tanabe Pharma Co., Merck Sharp and Dohme K.K., Ono Pharmaceutical Co., and Takeda Pharmaceutical Co. N.Ka. is a staff member of the endowed chair (Department of Atherosclerosis and Metabolism) established by funds from Kowa Pharmaceutical Co. Ltd. and has received research funds from Merck Sharp and Dohme K.K. and lecture fees from Astellas Pharma, Inc.; AstraZeneca; Boehringer Ingelheim; Daiichi Sankyo Co.; Sumitomo Dainippon Pharma Co.; Eisai Co.; Eli Lilly & Co.; Kowa Pharmaceutical Co. Ltd.; Mitsubishi Tanabe Pharma Co.; Novartis Pharmaceuticals Corp.; Novo Nordisk Pharma Ltd.; Ono Pharmaceutical Co.; Otsuka Pharmaceutical Co.; Shionogi & Co. Ltd.; Takeda Pharmaceutical Co.; and Sanofi. T.S. has received lecture fees from Boehringer Ingelheim; Sanofi; Novo Nordisk Pharma Ltd.; Novartis Pharmaceuticals Corp.; Eli Lilly & Co.; Abbott Japan Co.; Takeda Pharmaceutical Co.; Sanwa Kagaku Kenkyusho Co.; Mitsubishi Tanabe Pharma Co.; Daiichi Sankyo Co.; Astellas Pharma, Inc.; Ono Pharmaceutical Co.; Merck Sharp and Dohme K.K.; Shionogi & Co. Ltd.; and Taisho Toyama Pharmaceutical Co. T.On. has received lecture fees from Ono Pharmaceutical Co. N.Ku. has received lecture fees from Sanofi and Novartis Pharmaceuticals Corp. T.Os. has received lecture fees from Novo Nordisk, Inc.; Astellas Pharma, Inc.; Mitsubishi Tanabe Pharma Co.; Sanwa Kagaku Kenkyusho Co.; Takeda Pharmaceutical Co.; and Kowa Pharmaceutical Co. Ltd. and has received research funds from Novo Nordisk, Inc.; Astellas Pharma, Inc.; Mitsubishi Tanabe Pharma Co.; Sanwa Kagaku Kenkyusho Co.; Kowa Pharmaceutical Co. Ltd.; Novo Nordisk Pharma Ltd.; Sumitomo Dainippon Pharma Co.; Eli Lilly & Co.; Taisho Pharmaceutical Co.; GlaxoSmithKline; Sumitomo Dainippon Pharma Co.; Astellas Pharma US, Inc.; Bayer HealthCare; and AbbVie GK. H.K. has received lecture fees from Boehringer Ingelheim; Sanofi; Ono Pharmaceutical Co.; Merck Sharp and Dohme K.K.; Novo Nordisk Pharma Ltd.; Novartis Pharmaceuticals Corp.; Daiichi Sankyo Co.; Takeda Pharmaceutical Co.; Kissei Pharmaceutical Co.; Sumitomo Dainippon Pharma Co.; Mitsubishi Tanabe Pharma Co.; Kyowa Hakko Kirin Co. Ltd.; Eli Lilly & Co.; Pfizer; AstraZeneca; and Astellas Pharma, Inc. and has received research funds from Takeda Pharmaceutical Co.; Merck Sharp and Dohme K.K.; Mochida Pharmaceutical Co.; Sanofi, Novartis Pharmaceuticals Corp.; Novo Nordisk Pharma Ltd.; Eli Lilly & Co.; Daiichi Sankyo Co.; Shionogi & Co. Ltd.; Teijin Pharma Ltd.; Sumitomo Dainippon Pharma Co.; Otsuka Pharmaceutical Co.; Kissei Pharmaceutical Co.; Mitsubishi Tanabe Pharma Co.; Ono Pharmaceutical Co.; AstraZeneca; Astellas Pharma, Inc.; and Kyowa Hakko Kirin Co. Ltd. K.K. has received lecture fees from Boehringer Ingelheim, Sanofi, Novo Nordisk Pharma Ltd., Novartis Pharmaceuticals Corp., Eli Lilly & Co., Takeda Pharmaceutical Co., Merck Sharp and Dohme K.K., Kowa Pharmaceutical Co. Ltd., Mitsubishi Tanabe Pharma Co. and research funds from Sysmex Co. H.Yok. has received lecture fees from Boehringer Ingelheim; Sanofi; Ono Pharmaceutical Co.; Novo Nordisk Pharma Ltd.; Novartis Pharmaceuticals Corp.; Sanwa Kagaku Kenkyusho Co.; Daiichi Sankyo Co.; Takeda Pharmaceutical Co.; Merck Sharp and Dohme K.K.; Sumitomo Dainippon Pharma Co.; Astellas Pharma, Inc.; AstraZeneca; Kyowa Hakko Kirin Co. Ltd.; and Kowa Pharmaceutical Co. Ltd. and has received research funds from Sanwa Kagaku Kenkyusho Co. M.G. has received lecture fees from Novartis Pharmaceuticals Corp. and Taiho Pharmaceutical Co. Ltd. and has received travel fees from Takeda Pharmaceutical Co. I.S. has received lecture fees from Astellas Pharma, Inc.; AstraZeneca; Merck Sharp and Dohme K.K.; Ono Pharmaceutical Co.; Kyowa Hakko Kirin Co. Ltd.; Kowa Pharmaceutical Co. Ltd.; Sanofi; Sanwa Kagaku Kenkyusho Co.; Daiichi Sankyo Co.; Takeda Pharmaceutical Co.; Mitsubishi Tanabe Pharma Co.; Teijin Pharma Ltd.; Eli Lilly Japan; Nippon Boehringer Ingelheim Co.; Novartis Pharmaceuticals Corp.; Novo Nordisk Pharma Ltd.; Bayer Yakuhin Ltd.; Pfizer Japan, Inc.; Bristol-Myers Squibb; Mochida Pharmaceutical Co.; Shionogi & Co. Ltd.; Taisho Toyama Pharmaceutical Co.; and Shionogi & Co. Ltd. and has received research funds from Astellas Pharma, Inc.; AstraZeneca; Eisai Co.; Merck Sharp and Dohme K.K.; Otsuka Pharmaceutical Co.; Ono Pharmaceutical Co.; Kaken Pharmaceutical Co.; Kissei Pharmaceutical Co.; Kyowa Hakko Kirin Co. Ltd.; Sanofi; Shionogi & Co. Ltd.; Daiichi Sankyo Co.; Sumitomo Dainippon Pharma Co.; Takeda Pharmaceutical Co.; Mitsubishi Tanabe Pharma Co.; Teijin Pharma Ltd.; Nippon Boehringer Ingelheim Co.; Novartis Pharmaceuticals Corp.; Novo Nordisk Pharma Ltd.; Pfizer Japan, Inc.; Bristol-Myers Squibb; Mochida Pharmaceutical Co.; Eli Lilly Japan; Kowa Pharmaceutical Co. Ltd.; and Taisho Toyama Pharmaceutical Co. H.W. has received lecture fees from Novo Nordisk, Inc.; Eli Lilly & Co.; Sanofi; Sumitomo Dainippon Pharma Co.; Fujifilm; Bayer HealthCare; Kissei Pharmaceutical Co.; Mochida Pharmaceutical Co.; Merck Sharp and Dohme K.K.; Takeda Pharmaceutical Co.; Boehringer Ingelheim; Daiichi Sankyo Co.; Ono Pharmaceutical Co.; Novartis Pharmaceuticals Corp.; Mitsubishi Tanabe Pharma Co.; AstraZeneca; Kyowa Hakko Kirin Co. Ltd.; Sanwa Kagaku Kenkyusyo Co.; Kowa Pharmaceutical Co. Ltd.; and Astellas Pharma, Inc.; has received advisory fees from Novo Nordisk Pharma Ltd., Inc.; Mochida Pharmacuetical Co.; AstraZeneca; Kowa Pharmaceutical Co. Ltd.; Astellas Pharma, Inc.; Sanofi; Boehringer Ingelheim; Merck Sharp and Dohme K.K.; Mitsubishi Tanabe Pharma Co.; Novartis Pharmaceuticals Corp.; Sumitomo Dainippon Pharma Co.; Takeda Pharmaceutical Co.; Ono Pharmaceutical Co.; Pfizer; and Kowa Pharmaceutical Co. Ltd.; and has received research funds from Boehringer Ingelheim; Pfizer; Mochida Pharmaceutical Co.; Sanofi; Novo Nordisk Pharma Ltd.; Novartis Pharmaceuticals Corp.; Sanwa Kagaku Kenkyusho Co.; Terumo Medical Corp.; Eli Lilly & Co.; Mitsubishi Tanabe Pharma Co.; Daiichi Sankyo Co.; Takeda Pharmaceutical Co.; Merck Sharp and Dohme K.K.; Shionogi & Co. Ltd.; Sumitomo Dainippon Pharma Co.; Kissei Pharmaceutical Co.; and AstraZeneca. No other potential conflicts of interest relevant to this article were reported.

Author Contributions. T.M. and N.Ka. drafted the manuscript. M.G. analyzed research data. T.M., N.Ka., T.S., H.Yos., T.On., N.Ku., T.Os., H.K., K.K., Y.U., T.Y., K.M., H.Yok., M.T., M.G., I.S. and H.W. were involved in analysis and interpretation of data, reviewed/edited the manuscript, and approved the final manuscript. The authors meet the criteria for authorship recommended by the International Committee of Medical Journal Editors and take full responsibility for all content in the manuscript and all editorial decisions. I.S. and H.W. are the guarantors of this work and, as such, had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.