D-D RNLTOA SYMPOSIUM TRANSLATIONAL IDF-ADA Diabetes Care Volume 39, January 2016 139

Tomoya Mita,1,2 Naoto Katakami,3,4 Alogliptin, a Dipeptidyl Peptidase 4 Hidenori Yoshii,5 Tomio Onuma,5 Hideaki Kaneto,3 Takeshi Osonoi,6 Inhibitor, Prevents the Progression Toshihiko Shiraiwa,7 Keisuke Kosugi,8 Yutaka Umayahara,9 of Carotid Atherosclerosis in Tsunehiko Yamamoto,10 Hiroki Yokoyama,11 Patients With Type 2 Diabetes: Nobuichi Kuribayashi,12 Hideaki Jinnouchi,13 Masahiko Gosho,14 The Study of Preventive Effects of Iichiro Shimomura,3 and Hirotaka Watada,1,2,15 on behalf of the AlogliptinonDiabetic Collaborators on the Study of Preventive Effects of Alogliptin on Diabetic Atherosclerosis (SPEAD-A) Atherosclerosis (SPEAD-A) Trial Diabetes Care 2016;39:139–148 | DOI: 10.2337/dc15-0781

1Department of Metabolism & Endocrinology, Juntendo University Graduate School of Medi- cine, Tokyo, Japan 2Center for Molecular Diabetology, Juntendo University Graduate School of Medicine, Tokyo, Japan 3Department of Metabolic Medicine, Osaka Uni- versity Graduate School of Medicine, Osaka, Japan 4Department of Metabolism and Atherosclero- OBJECTIVE sis, Osaka University Graduate School of Medi- cine, Osaka, Japan Recent experimental studies have shown that dipeptidyl peptidase 4 (DPP-4) 5Department of Medicine, Diabetology & Endo- inhibitors have antiatherosclerotic benefits in glucagon-like peptide 1–dependent crinology, Juntendo Tokyo Koto Geriatric Medi- and –independent manners. The current study investigated the effects of alogliptin, a cal Center, Tokyo, Japan 6 DPP-4 inhibitor, on the progression of carotid atherosclerosis in patients with Naka Memorial Clinic, Ibaraki, Japan 7Shiraiwa Medical Clinic, Osaka, Japan type 2 diabetes mellitus (T2DM). 8Osaka Police Hospital, Osaka, Japan 9Osaka General Medical Center, Osaka, Japan RESEARCH DESIGN AND METHODS 10Kansai Rosai Hospital, Hyogo, Japan This prospective, randomized, open-label, blinded-end point, multicenter, parallel- 11Internal Medicine, Jiyugaoka Medical Clinic, group, comparative study included 341 patients with T2DM free of a history of Hokkaido, Japan 12Misaki Naika Clinic, Chiba, Japan apparent cardiovascular diseases recruited at 11 clinical units and randomly allo- 13Jinnouchi Hospital, Kumamoto, Japan cated to treatment with alogliptin (n = 172) or conventional treatment (n =169). 14Department of Clinical Trial and Clinical Epi- Primary outcomes were changes in mean common and maximum intima-media demiology, Faculty of Medicine, University of thickness (IMT) of the carotid artery measured by carotid arterial echography Tsukuba, Ibaraki, Japan 15Center for Therapeutic Innovations in Diabetes, during a 24-month treatment period. Juntendo University Graduate School of Medi- cine, Tokyo, Japan RESULTS Corresponding author: Tomoya Mita, tom-m@ Alogliptin treatment had a more potent glucose-lowering effect than the conven- juntendo.ac.jp. tional treatment (20.3 6 0.7% vs. 20.1 6 0.8%, P = 0.004) without an increase Received 14 April 2015 and accepted 5 October of hypoglycemia. Changes in the mean common and the right and left maximum 2015. IMT of the carotid arteries were significantly greater after alogliptin treatment than Clinical trial reg. no. UMIN000005311, http:// after conventional treatment (20.026 mm [SE 0.009] vs. 0.005 mm [SE 0.009], www.umin.ac.jp/ctr/. P =0.022;20.045 mm [SE 0.018] vs. 0.011 mm [SE 0.017], P = 0.025, and This article contains Supplementary Data online 2 2 P at http://care.diabetesjournals.org/lookup/ 0.079 mm [SE 0.018] vs. 0.015 mm [SE 0.018], = 0.013, respectively). suppl/doi:10.2337/dc15-0781/-/DC1. CONCLUSIONS T.M. and N.Ka. contributed equally to this study. Alogliptin treatment attenuated the progression of carotid IMT in patients with © 2016 by the American Diabetes Association. Readers may use this article as long as the work is T2DM free of apparent cardiovascular disease compared with the conventional properly cited, the use is educational and not for treatment. profit, and the work is not altered. 140 Effects of Alogliptin on Atherosclerosis Diabetes Care Volume 39, January 2016

Patients with type 2 diabetes mellitus interventions on the progression of 9) under treatment with therapeutic (T2DM) are at high risk for cardiovascular atherosclerosis. To our knowledge, no drugs not concomitantly administrable (CV) diseases (CVD), which is one of the published studies have reported the with incretin preparations with regard major causes of morbidity and mortality long-term antiatherosclerotic effects to the National Health Insurance pro- in these patients (1). Thus, one of the of DPP-4 inhibitors in patients with gram, such as DPP-4 inhibitors, at the main goals of T2DM management is T2DM free of CVD. The current study start of the study; 10)pregnancy,lacta- to reduce the incidence of CVD. In this investigated the effects of alogliptin tion, possible or planned pregnancy; 11) regard, although clinical studies have on the intima-media thickness (IMT) in medical history of hypersensitivity to in- shown that HbA1c is a risk factor for apparent CVD-free patients with T2DM. vestigational drugs; or 12) judged as in- CVD in patients with T2DM (2), there eligible by clinical investigators. is little evidence that glycemic control RESEARCH DESIGN AND METHODS The study subjects were screened and/or treatment with any particular Study Design consecutively, and those who met the oral hypoglycemic agents actually reduces The Study of Preventive Effects of Alogliptin above criteria were invited to partici- the incidence of CVD (3,4). on Diabetic Atherosclerosis (SPEAD-A) trial pate in the current study. All patients Dipeptidyl peptidase 4 (DPP-4) was a multicenter prospective, randomized, who agreed to participate were regis- inhibitors, a new class of oral hypoglyce- open-label, blinded-end point (PROBE) tered. The protocol was approved by mic agents, inhibit the degradation of study, as described previously (20). This the institutional review board at each active glucagon-like peptide 1 (GLP-1) study was registered on the University Hos- participating institution, and the study and glucose-dependent insulinotropic pital Medical Information Network Clinical was conducted in compliance with the polypeptide and increase their biological Trials Registry (UMIN000005311), a non- Declaration of Helsinki and current legal effects (e.g., augmentation of glucose- profit organization in Japan that meets regulations in Japan. dependent insulin secretion and suppres- the requirements of the International Com- sion of glucagon release) (5). In addition, mittee of Medical Journal Editors (ICMJE). Randomization and Study these agents have potential antiathero- Intervention sclerotic properties. In a rodent model Study Population Patients were registered at the adminis- of atherosclerosis, GLP-1 and GLP-1 re- Japanese patients with T2DM who period- tration office of the SPEAD-A trial via ceptor agonists were reported to inhibit ically attended the diabetes outpatient the Internet, and once enrolled, were atherosclerosis and inflammation (6–9), clinics at 11 institutions in Japan (listed randomly assigned in equal numbers and DPP-4 inhibitors, including alogliptin, in the Supplementary Data) were ap- into the alogliptin treatment group or inhibit these pathological processes in proached to participate in this study. the conventional treatment group (on GLP-1–dependent and –independent The inclusion criteria were as follows: 1) drugs other than DPP-4 inhibitors). manners (10–12). patients with T2DM in whom the target Randomization was performed using a These studies are promising and blood glucose control specified in the dynamic allocation method based on suggestthatDPP-4inhibitorscould Treatment Guide for Diabetes (edited by the with/without administration of also reduce CV risk in T2DM. However, the Japan Diabetes Society) (21) could pioglitazone, age, and sex. three recent randomized clinical trials not be achieved despite dietary/exercise Treatment was continued until the reported that DPP-4 inhibitors did not therapy or concomitant treatment for target value of HbA1c specified in the reduce CV risk, but they also did not T2DM other than DPP-4 inhibitors admin- Treatment Guide for Diabetes (21) was increase the risk compared with placebo istered for 3 months or longer and whose achieved (usually HbA1c level ,7.0%) in in patients with T2DM with history of HbA1c was below 9.4% (patients were also all patients. In the conventional treatment CVDorathigh-riskforCVD(13–15). included after a 12-week or longer with- group, the dosage of current therapy was Subjects in these studies had already drawal of previous treatment with DPP-4 increased or a concomitant oral glucose- received a multitude of therapies for inhibitors); 2)age$30 years, irrespective lowering drug (excluding other DPP-4 in- other pathologies. Thus, these factors of sex; and 3) signed consent for participa- hibitors, GLP-1 analogs, and insulin) was may possibly have influenced the re- tion in the study after a full explanation of added. In the alogliptin treatment group, sults of the above three trials. On the the study. The exclusion criteria were as alogliptin was administered orally at other hand, early and effective interven- follows: 1) type 1 or secondary diabetes; 25 mg, once daily. However, the addition tion before the development of advanced 2) severe infections before or after surgery of an alternative glucose-lowering atherosclerosis in patients without his- or severe trauma; 3) myocardial infarction, agent (excluding other DPP-4 inhibitors, tory of CVD may increase the chance of angina pectoris, cerebral stroke, or cere- GLP-1 analogs, and insulin) was permitted. significant reduction not only of micro- bral infarction; 4) moderate or severe re- In the case of hypoglycemia, the dose of vascular disease but also CVD (16). How- nal dysfunction (serum creatinine: male, any concomitantly used oral glucose- ever, assessing the long-term effects of a .1.4 mg/dL; female, .1.2 mg/dL); 5)se- lowering drug was titrated. The use of single drug on primary CVD in a clinical vere liver dysfunction (aspartate amino- antihyperlipidemic and antihypertensive setting is no doubt difficult. transferase $100 IU/L); 6)moderateor drugs was permitted during the study. Progressive thickening of the carotid severe heart failure (New York Heart As- artery intima-media is considered a sur- sociation stage III or higher); 7) under Observation Items and Schedule rogate marker for CVD in patients with treatment with an incretin preparation, The study period was 2 years after reg- T2DM (17–19) and has been used in such as other DPP-4 inhibitors, at the start istration (registration period: March theevaluationoftheeffectsofvarious of the study; 8) under insulin treatment; 2011 to June 2013). All patients were care.diabetesjournals.org Mita and Associates 141

monitored for 2 years, regardless of or even resulted in persistent or significant segment 2 cm proximal to the dilation adherence to or discontinuation of disability/incapacity and needed interven- of the carotid bulb (mean-IMT-CCA). study medications for any reason. tion to prevent permanent impairment In addition, the greatest thicknesses of Clinical outcomes, adherence, and ad- or damage. If participants suffered any IMT, including plaque lesions in the CCA verse events (AEs) were ascertained AEs/serious AEs, all details were docu- (max-IMT-CCA), the carotid bulb (max- and adjudicated by each investigator in mented and reported. Serious AEs were IMT-Bulb), and the ICA (max-IMT-ICA), an open fashion. Clinical and biochemi- reported to the principal investigator were also measured separately. Reproduc- cal data were collected at 0, 26, 52, 78, and the ethics committee. Both of them ibility analysis of replicate measurements and 104 weeks after randomization. judged whether the diagnosis was appro- in 20 subjects yielded absolute mean 6 SD priate or made a decision on whether the differences of 0.02 6 0.01, 0.01 6 0.01, Study Outcomes patient should be withdrawn from the 0.02 6 0.01, and 0.01 6 0.01 mm for The primary study outcomes were trial. CV events were diagnosed and mean-IMT-CCA, max-IMT-CCA, max-IMT- changes in right and left maximum fully assessed by members of the cardio- Bulb, and max-IMT-ICA, respectively. The IMT of the common carotid artery vascular end point committee (including intrareader coefficient of variation for (max-IMT-CCA) and mean IMT of the CCA two cardiologists and a neurologist). measurement was 1.1%, 0.7%, 0.7%, and (mean-IMT-CCA) during the 104-week 0.8%, respectively. treatment period measured by carotid Measurement of Carotid IMT arterial echography. These measure- Ultrasonographic scanning of the carotid Biochemical Tests ments were performed at the start of arteries was performed by expert sonog- Blood samples were obtained after over- the study and repeated after 52 and raphers who were specifically trained night fast. Serum lipids (total cholesterol, 104 weeks and at the time of any dis- to perform the prescribed study exam- HDL cholesterol, LDL cholesterol, triglyc- ination. The mean intrainvestigator co- continuations or changes in medication erides), HbA1c (NGSP), glucose, insulin, and/or dose. efficient of variation for measurement and creatinine were measured with stan- The secondary outcomes were 1) of mean-IMT-CCA and max-IMT-CCA dard techniques. Measurements of hs- changes in parameters related to gly- (6 SD) in each institution was 1.9 6 CRP, interleukin-6, intercellular adhesion 6 cemic control (HbA1c, fasting plasma 1.1% and 2.8 2.2%, respectively. To molecule 1, and VCAM-1 were out- glucose, and immunoreactive insulin); avoid intersonographer variability, each sourced to a private laboratory (SRL Lab- 2) changes in parameters related to participant was examined by the same oratory, Tokyo, Japan). Urinary albumin diabetic nephropathy, including uri- sonographer using the same equipment excretion was measured by the improved nary albumin excretion and estimated (high-resolution B-mode ultrasound bromocresol purple method using a spot glomerular filtration rate (eGFR); 3) scanner equipped with a high-frequency urine sample. The eGFR was calculated by changes in lipid parameters (total cho- [.7.5-MHz] linear transducer, with a the formula: eGFR (mL/min per 1.73 m2)= lesterol, HDL cholesterol, triglyceride, limit of detection of ,0.1 mm) through- 194 3 age20.287 3 serum creatinine20.1094 and LDL cholesterol); 4) changes in bio- out all the visits. Scanning of extracranial (30.739 for females) (23). chemical parameters, including serum CCAs, the carotid bulb, and the internal intercellular adhesion molecule 1, vascu- carotid arteries (ICA) in the neck was per- Sample Size lar cell adhesion molecule 1 (VCAM-1), formed in at least three different longi- The progression of carotid IMT in interleukin-6, and hs-CRP; 5) occurrence tudinal projections (anterior, lateral, and patients with T2DM is considered to be of CV events, including sudden death, cor- posterior, which approximately corre- 0.034 6 0.054 mm/year, and a 1% 8 8 8 onary heart disease, and stroke; and 6) spondedto60,90,and150 for the improvement in the HbA value is asso- 8 8 1c appearance of any AE. right carotid artery and 210 , 270 , and ciated with 0.02 mm/year improvement 8 300 for the left carotid artery marked on in IMT (24). Therefore, in the 2-year Safety and CV Events Evaluation the Meijer arc) as well as transverse pro- observation period, registration of at For the sake of patient safety, all AEs jections. The site of greatest thickness, least 324 patients was required to obtain were recorded during the treatment including plaque lesions, was sought 90% power to detect a difference of fi and follow-up. AEs were de ned as any along the arterial walls. The IMT was mea- 0.04 mm in IMT between the two treat- untoward medical occurrence in a clinical sured as the distance between two paral- ment groups, assuming a SD of 0.108, trial subject administered a medicinal lel echogenic lines corresponding to the 5% dropout, and 0.05 level of signifi- product and that did not necessarily vascular lumen and the adventitial layer. cance. According to this calculation, have a causal relationship with this treat- To avoid interreader variability, all the target number of enrolled patients ment. The association between AEs and scanswereelectronicallystoredand was set at 324 for the 2-year registration fi fi the study medication was classi ed as sent to the central of ce (IMT Evaluation period. related or not related to the study drug Committee, Osaka, Japan) and read by a by one of the investigators. All related single experienced reader blinded to Statistical Analysis AEs that resulted in a withdrawal of the the clinical characteristics of the subjects Efficacy, regardless of adherence, was subject from the study were monitored and type of treatment, in a random order, analyzed using an intent-to-treat until resolution. Serious AEs were defined using automated digital edge-detection approach. Results are presented as as death or life-threatening events that software (Intimascope; Media Cross, mean 6 SD or median (interquartile required inpatient hospitalization, caused Tokyo, Japan) (22). The software system range) for continuous variables or prolongation of existing hospitalization, averaged ;200 points of IMT values in the number (proportion) of patients for 142 Effects of Alogliptin on Atherosclerosis Diabetes Care Volume 39, January 2016

categorical variables. The primary end Table 1—Clinical characteristics of the patients in the two groups point was a change in IMT from baseline Alogliptin Conventional to week 104. Primary analysis was per- treatment group treatment group formed using the mixed-effects model n =172 n =169 P value for repeated measures with treatment Male sex 101 (63) 98 (61) 0.72 group, time (week), interactions between Age (years) 64.4 6 9.8 64.8 6 9.1 0.82 treatment group and time (week), and Current smoking 43 (27) 33 (21) 0.24 baseline IMT as fixed effects; an un- Hypertension 90 (56) 91 (57) 1.00 structured covariate was used to Dyslipidemia 86 (53) 94 (58) 0.43 model the covariance of within-subject variability. For the occurrence of CV Duration of T2DM (years) 9 (5.0, 15.0) 8.2 (4.0, 15.0) 0.94 events, one of the secondary end Use of oral glucose-lowering agents d d Metformin 84 (52) 75 (47) 0.37 points time to onset was analyzed Sulfonylurea 80 (50) 90 (56) 0.32 using a log-rank test and Cox propor- Glinides 9 (6) 16 (10) 0.21 tional hazards model. Thiazolidinediones 35 (22) 38 (24) 0.79 Baseline and follow-up group compar- a-Glucosidase inhibitor 56 (35) 51 (32) 0.64 isons were assessed with the Student Use of antihypertensive drugs t test or Wilcoxon rank sum test for con- ACE inhibitors 7 (4) 4 (2) 0.54 tinuous variables and the Fisher exact test Angiotensin II receptor blockers 72 (45) 68 (42) 0.74 for categorical variables. Changes from Direct renin inhibitor 0 (0) 1 (1) 1.00 Calcium channel blocker 47 (29) 58 (36) 0.23 baseline to treatment visits were as- Diuretic drugs 7 (4) 8 (5) 1.00 sessed with one-sample t test and the a-Adrenergic receptor antagonist 3 (2) 0 (0) 0.25 Wilcoxon signed rank test within the b-Adrenergic receptor antagonist 3 (2) 8 (5) 0.22 group. The number and percentage of Others 1 (0) 4 (2) 0.37 patients reporting AEs was presented by Use of lipid-lowering agents treatment group and compared between Statins 61 (38) 74 (46) 0.18 the two treatment groups using the Ezetimibe 5 (3) 2 (1) 0.45 Fisher exact test. All statistical tests Resins 0 (0) 1 (1) 1.00 Fibrates 6 (4) 9 (6) 0.60 were two-sided with 5% significance Use of antithrombotic agents level. Analyses were performed using Antiplatelet agents 19 (12) 23 (14) 0.62 SAS 9.4 software (SAS Institute Inc., Anticoagulants 1 (1) 3 (2) 0.62 Cary, NC). The administrative office of Others 0 (0) 1 (1) 1.00 the SPEAD-A trial analyzed the data Data are n (%), mean 6 SD, or median (interquartile range). based on instructions from an indepen- dent biostatistician.

RESULTS population. Over 104 weeks, alogliptin treatment group at 104 weeks A total of 341 participants were ran- treatment, but not conventional (P =0.003,Table3).TheDchange in domly allocated into the alogliptin group treatment, significantly reduced the HbA1c (value at end of study 2 value at (n = 172) or the conventional treatment mean-IMT-CCA and the right and left baseline) improved significantly in the group (n = 169). After excluding from max-IMT-CCA relative to the baseline alogliptin group (20.3 6 0.7%) but not the analyses 19 patients who withdrew (Table 2). In a mixed-effects model for in the conventional treatment group from the study and/or objected to the repeated measures, alogliptin signifi- (20.1 6 0.8%, P =0.004;Table3).The inclusion of their data in any analysis, cantly prevented the progression effect of alogliptin on the reduction in 161 in the alogliptin treatment group in mean-IMT-CCA and right and left HbA1c may be underestimated because and 161 in the conventional treatment max-IMT-CCA (i.e., primary end points the use of a-glucosidase inhibitors and group were included in the full analysis of the study) compared with conven- glinides was significantly higher in the set (Supplementary Fig. 1). The baseline tional treatment (Table 2). Similar conventional treatment group than in demographic and clinical characteris- findings were noted even in the the alogliptin group at the end of study tics of the 322 study participants are mixed-effects models adjusted for age (Supplementary Table 2). On the one reported in Tables 1–4. Most subjects and sex (data not shown). ANCOVA mod- hand, ANCOVA models that included had previously attended educational els that included treatment group, changes in HbA1c from baseline in addi- programs about diet and exercise ther- age, sex, baseline IMT, systolic BP, and tion to the factors in model 1 showed apy and received appropriate medical administration of statins also produced that alogliptin significantly prevented treatments. Blood glucose, lipids, and findings similar to those in the mixed- the progression in carotid IMT compared blood pressure (BP) levels were well con- effects models (model 1, Supplementary with conventional treatment (model 2, trolled in all subjects. Table 1). Supplementary Table 1). On the other All patients of both groups met the cri- The mean change in BMI was 0.3 6 hand, fasting blood glucose and plasma teria for inclusion in IMT analysis based 1.9 kg/m2 in the alogliptin group versus insulin levels were not different between on the criteria of the intention-to-treat 20.3 6 1.7 kg/m2 in the conventional the two groups. BP and lipid metabolism care.diabetesjournals.org Mita and Associates 143

Table 2—Effects of alogliptin on IMT Treatment effect (alogliptin- conventional treatment) Alogliptin treatment Conventional treatment P value between n group n group Mean change (95% CI) P value groups Common mean IMT Baseline (mm) 161 0.83 6 0.15 161 0.83 6 0.17 0.87 52 weeks (mm) 152 0.79 6 0.14 157 0.82 6 0.16 0.18 104 weeks (mm) 151 0.80 6 0.16 153 0.84 6 0.18 0.052 Mean change (SEM) 52 weeks 20.029 (0.009)# 20.012 (0.009) 20.017 (20.042, 0.007) 0.17 104 weeks 20.026 (0.009)# 0.005 (0.009) 20.030 (20.057, 20.004) 0.022 Right maximum IMT Baseline (mm) 161 1.04 6 0.32 161 1.03 6 0.26 0.77 52 weeks (mm) 152 0.97 6 0.23 157 1.02 6 0.30 0.099 104 weeks (mm) 151 0.99 6 0.27 153 1.04 6 0.30 0.10 Mean change (SEM) 52 weeks 20.062 (0.019)# 20.008 (0.018) 20.053 (20.105, 20.002) 0.041 104 weeks 20.045 (0.018)* 0.011 (0.017) 20.056 (20.105, 20.007) 0.025 Left maximum IMT Baseline (mm) 161 1.09 6 0.33 161 1.10 6 0.42 0.72 52 weeks (mm) 153 1.05 6 0.37 157 1.08 6 0.36 0.50 104 weeks (mm) 151 1.01 6 0.28 153 1.10 6 0.40 0.032 Mean change (SEM) 52 weeks 20.040 (0.019)* 20.031 (0.019) 20.009 (20.062, 0.043) 0.72 104 weeks 20.079 (0.018)§ 20.015 (0.018) 20.064 (20.114, 20.014) 0.013 Data are mean 6 SD unless otherwise stated. Comparisons of IMTs during treatment with those at baseline were performed by one-sample t test based on a mixed-effects model for repeated measures. Differences in IMT between groups at each point were analyzed by the Student t test. Differences in Dchange in IMT from baseline at 52 and 104 weeks between groups at each point (treatment effect) were analyzed with mixed-effects model for repeated measures. Treatment group, week, interactions between treatment group and week, and baseline IMT were included as fixed effects. *P , 0.05; #P , 0.01; §P , 0.001. were well controlled in both groups CONCLUSIONS the reduction was achieved without in- throughout the study, but there were In this study, the rate of mean-IMT-CCA creasingtheriskofhypoglycemia. no significant differences in other risk and right and left max-IMT-CCA diminished However, the difference in HbA1c factors for atherosclerosis at the end of significantly in the alogliptin treatment between the two groups is probably the study (Table 3). A modest difference group compared with the conventional- not closely related to the reduction in in VCAM-1 was noted, but no changes treatment group. Interestingly, there carotid IMT, because changes in HbA1c were observed in other markers of was substantial regression of the were only very modestly associated with inflammation and endothelial damage mean-IMT-CCA and of the right and changes in mean-IMT-CCA and right (Table 4). left max-IMT-CCA at the end of the alog- max-IMT-CCA and were not associated During the study, 37 patients devel- liptin treatment protocol. These results with changes in left max-IMT-CCA oped any AEs and 19 developed serious suggest that alogliptin treatment pre- (mean-IMT-CCA: r = 20.18, P , 0.05; AEs. There were no significant differ- vents the progression of atherosclerosis right max-IMT-CCA: r = 20.18, P , 0.05; ences in the incidences of any AEs and in patients with T2DM free of past history left max-IMT-CCA: r = 20.12, P =NS)in serious AEs between the alogliptin of apparent CVD. the alogliptin treatment group. In addi- group and the conventional treatment In the current study, we confirmed that tion, alogliptin treatment still attenuated group. The most frequent AE was hypo- alogliptin had a sustainable glucose- the progression of carotid IMT compared glycemia, followed by gastrointestinal lowering effect by ;0.3% over 2 years, a with the conventional treatment even disorders (Supplementary Table 3). finding almost identical to that of a re- after adjustment for changes in HbA1c Overall, AEs in 8 patients resulted in dis- cent randomized clinical study with alog- from baseline (Supplementary Table 1). continuation of alogliptin. Hypoglycemic liptin (13) and other DPP-4 inhibitors Furthermore, significant differences in events were recorded in 11 patients (14,15). The glucose-lowering effect of mean-IMT-CCA and right max-IMT-CCA (5 patients of the alogliptin group and alogliptin is not subtle, considering that were still observed in the post hoc 6 patients of the control group). How- most patients had already achieved rela- matched-pair set (n = 103, each group) ever, none of the patients experienced tively good glycemic control at baseline. for changes in HbA1c from baseline severe hypoglycemia. Only a few This glucose-lowering effect was probably (data not shown). These data suggest patients developed CV events (n =5)or achieved at least by lessening fluctuations that the glucose-lowering effect of alog- were diagnosed with cancer (n =5); in blood glucose levels (25), which was not liptin alone could not explain carotid therefore, there was no significant differ- evaluated in this study because alogliptin IMT regression. ence in the incidence of CVD between produced only a modest reduction in the In human studies, short-term treat- the two groups. fasting blood glucose level. Importantly, ment with DPP-4 inhibitors reduced 144 Effects of Alogliptin on Atherosclerosis Diabetes Care Volume 39, January 2016

Table 3—Effects of alogliptin on BMI, glucose metabolism, lipid metabolism, and blood pressure n Alogliptin treatment group n Conventional treatment group P value BMI at baseline (kg/m2) 161 24.6 6 4.3 161 24.9 6 3.7 0.49 Change from baseline 26 weeks 153 0.2 6 1.3 158 0.0 6 1.8 0.18 52 weeks 154 0.1 6 1.3 156 20.2 6 1.6 0.037 78 weeks 150 0.2 6 1.5 153 20.2 6 1.6 0.045 104 weeks 147 0.3 6 1.9 150 20.3 6 1.7 0.003

HbA1c at baseline (%) 158 7.3 6 0.8 160 7.2 6 0.8 0.54

HbA1c at baseline (mmol/mol) 158 56.3 6 8.5 160 55.7 6 9.3 0.54 Change from baseline (HbA1c %) 26 weeks 158 20.4 6 0.7§ 160 0.0 6 0.9 ,0.001 52 weeks 153 20.4 6 0.6§ 157 20.1 6 0.8 ,0.001 78 weeks 152 20.4 6 0.8§ 154 0.0 6 1.1 ,0.001 104 weeks 150 20.3 6 0.7§ 153 20.1 6 0.8 0.004 Fasting blood glucose at baseline (mmol/L) 160 7.81 6 1.5 161 7.85 6 1.93 0.85 Change from baseline 26 weeks 158 20.55 6 1.64§ 159 20.06 6 2.19 0.026 52 weeks 150 20.59 6 1.48§ 157 20.32 6 1.81 0.15 78 weeks 147 20.54 6 1.84§ 152 20.13 6 2.52 0.11 104 weeks 149 20.45 6 1.56§ 153 20.28 6 2.03 0.41 Insulin at baseline (pmol/L) 158 57.2 6 68.0 161 58.3 6 43.9 0.86 Change from baseline 52 weeks 145 22.0 6 47.3 154 25.0 6 37.0 0.54 104 weeks 147 2.7 6 72.8 152 0.0 6 43.5 0.70 Total cholesterol at baseline (mmol/L) 160 5.00 6 0.77 159 5.01 6 0.75 0.86 Change from baseline (%) 26 weeks 122 0.0 6 11.7 138 21.7 6 11.7 0.25 52 weeks 153 22.1 6 11.4* 154 23.0 6 11.6# 0.50 78 weeks 152 23.8 6 12.4§ 150 23.0 6 13.5# 0.59 104 weeks 150 22.2 6 13.9 151 23.5 6 14.4# 0.43 LDL cholesterol at baseline (mmol/L) 158 2.89 6 0.68 160 2.93 6 0.64 0.62 Change from baseline (%) 52 weeks 145 0.0 6 17.1 153 20.9 6 17.3 0.25 104 weeks 146 20.4 6 22.0 150 22.6 6 20.2 0.37 HDL cholesterol at baseline (mmol/L) 160 1.47 6 0.38 161 1.41 6 0.36 0.16 Change from baseline (%) 26 weeks 156 21.6 6 12.9 160 21.5 6 15.8 0.95 52 weeks 153 22.2 6 19.3 157 21.4 6 14.7 0.70 78 weeks 151 23.9 6 12.3§ 153 0.2 6 16.5 0.015 104 weeks 149 22.0 6 12.5 153 1.1 6 15.7 0.065 Triglyceride at baseline (mmol/L) 160 1.19 (0.82, 1.76) 161 1.25 (090, 1.68) 0.31 Change from baseline (%) 26 weeks 151 0.0 (214.7, 28.6) 159 20.5 (217.2, 23.8) 0.55 52 weeks 150 27.3 (224.4, 27.9) 157 23.7 (222.7, 23.1) 0.57 78 weeks 147 26.3 (223.5, 16.9) 153 25.5 (224.7, 19.8) 0.62 104 weeks 149 21.3 (224.3, 24.3) 152 25.5 (224.2, 19.8) 0.35 Systolic BP (mmHg) 161 130 6 16 161 132 6 15 0.34 Change from baseline 26 weeks 156 21 6 16 159 24 6 15§ 0.10 52 weeks 154 0 6 16 156 21 6 16 0.70 78 weeks 147 21 6 19 152 24 6 14# 0.19 104 weeks 150 2 6 19 152 0 6 15 0.35 Diastolic BP (mmHg) 161 75 6 12 161 75 6 11 1.00 Change from baseline 26 weeks 156 22 6 10 159 22 6 11* 0.62 52 weeks 154 21 6 11 156 21 6 16* 0.48 78 weeks 147 23 6 13# 152 24 6 14§ 0.75 104 weeks 150 0 6 12 152 0 6 15 0.35 Data are mean 6 SD or median (interquartile range). Differences in parameters between groups at baseline were analyzed by the Student t test or Wilcoxon rank sum test. Differences in parameters from baseline to 52 and 104 weeks within group were analyzed by one-sample t test or the Wilcoxon signed rank test. Differences in parameters from baseline to 52 and 104 weeks between groups were analyzed by the Student t test or the Wilcoxon rank sum test. *P , 0.05; #P , 0.01; §P , 0.001. care.diabetesjournals.org Mita and Associates 145

Table 4—Effects of alogliptin on markers of renal function, inflammation, and endothelial injury n Alogliptin treatment group n Conventional treatment group P value eGFR (mL/min/1.73 m2)161786 20 161 77 6 18 0.59 Change from baseline 26 weeks 156 21 6 10 160 1 6 90.025 52 weeks 154 21 6 915716 11 0.18 78 weeks 152 21 6 10 157 0 6 11 0.44 104 weeks 150 21 6 10 153 0 6 10 0.27 UAE at baseline (mg/g creatinine) 158 14.0 (7.8, 54.2) 160 15.8 (7.4, 46.5) 0.96 Change from baseline 26 weeks 103 0.6 (25.0, 8.6) 111 0.8 (23.5, 18.5) 0.14 52 weeks 144 20.3 (29.2, 5.2) 141 0.7 (24.4, 7.5) 0.067 78 weeks 128 21.0 (29.2, 5.1) 119 1.0 (24.1, 13.8) 0.014 104 weeks 145 0.3 (25.3, 8.8) 144 0.4 (25.1, 14.0) 0.48 hs-CRP (ng/dL) 158 443 (209, 924) 161 545 (244, 868) 0.42 Change from baseline 52 weeks 145 69 (265, 245) 153 23 (2189, 280) 0.28 104 weeks 146 56 (2103, 250) 153 13 (2205, 200) 0.18 Interleukin-6 at baseline (ng/dL) 158 2.1 (1.4, 2.7) 160 2.2 (1.5, 2.9) 0.26 Change from baseline 52 weeks 145 0.0 (20.7, 0.6) 153 20.3 (20.9, 0.4)§ 0.041 104 weeks 145 0.1 (20.3, 0.7)* 147 0.0 (20.6, 0.6) 0.12 ICAM-1 at baseline (ng/mL) 158 230 (187, 286) 160 218 (185, 297) 0.35 Change from baseline 52 weeks 145 22(228, 31) 153 26(226, 20) 0.62 104 weeks 144 26(237, 22)* 147 29(239, 10) 0.42 VCAM-1 at baseline (ng/mL) 148 664 (551, 821) 160 718 (580, 879) 0.065 Change from baseline 52 weeks 145 32 (278, 101) 153 9 (268, 89) 0.45 104 weeks 145 20 (257, 90)* 147 28(2102, 81) 0.030 Data are mean 6 SD or median (interquartile range). Differences in parameters between the groups at baseline were analyzed by the Student t test or the Wilcoxon rank sum test. Differences in parameters from baseline to 52 and 104 weeks within the group were analyzed by the one-sample t test or the Wilcoxon signed rank test. Differences in parameters from baseline to 52 and 104 weeks between the groups were analyzed by the Student t test or the Wilcoxon rank sum test. ICAM-1, intercellular adhesion molecule 1; UAE, urinary albumin excretion. *P , 0.05; §P , 0.001.

the levels of various markers of chronic patients with T2DM (28). Furthermore, patients never treated by such therapy, inflammation and endothelial injury short-term statin treatment is reported although the same treatments did not (26,27). In contrast, in the present to improve inflammation (29). In addition, have any effect on serum inflammatory longer study, no such beneficial effects our results showed a significant reduction cytokine levels in both groups (30). for alogliptin were noted as judged by of BMI in the conventional treatment Thus, evaluations of local inflammation these markers. Although the exact rea- group. Thus, these factors may have had and oxidative stress in the vascular wall son for these inconsistent findings is still beneficial effects on the markers in the are probably helpful in the assessment of unclear, we propose the following pos- conventional treatment group and may “non–glycemic-dependent” antiathero- sible scenarios: First, both transient and have canceled the difference between sclerotic effects of DPP-4 inhibitors. chronic nonatherosclerotic diseases the two groups. Furthermore, in this study, In this study, alogliptin seemed to could have coincidentally affected the we evaluated only a few inflammatory have already resulted in carotid IMT proinflammatory conditions. cytokines, including hs-CRP, which is regression at 52 weeks. A similar rapid Second, the levels of these markers mainly produced by hepatocytes. Thus, reduction (within 24 weeks) in mean ca- could have been altered by the additional measurement of other cytokines, includ- rotid IMT was reported in patients with treatment within a relatively short time. ing tumor necrosis factor-a, is probably T2DM treated with pioglitazone (31,32); In this study, the use of a-glucosidase needed in future studies. In addition, however, other studies could not con- inhibitors and glinides was significantly Balestrieri et al. (30) used immunohis- firm this effect (33). Although the direct higher (Supplementary Table 2), and the tochemistry to determine the expres- action of pioglitazone on atherosclerosis use of statins and calcium channel block- sion levels of various inflammatory has been reported (34), several reports ers tended to be higher in the control cytokinesandmarkersofoxidative showed the direct effects of DPP-4 inhib- group than in the alogliptin group at stress in tissue samples obtained by itors on atherosclerosis. We recently the end of study (Supplementary Table 4). carotid endarterectomy. They showed demonstrated that another DPP-4 inhibi- Indeed, one previous study showed that that the expression levels of such tor suppressed DPP-4–induced smooth 3-week treatment with a-glucosidase markers were lower in the atheroscle- muscle proliferation and macrophage inhibitors reduced the levels of inflam- rotic lesions of patients who received inflammation in vitro (11). In addition, matory and cell adhesion markers in incretin-based therapy compared with Nagashima et al. (35) demonstrated 146 Effects of Alogliptin on Atherosclerosis Diabetes Care Volume 39, January 2016

that GLP-1 suppressed foam cell forma- high-risk for CVD. In comparison, our progression or regression of carotid tion from murine macrophages. These study enrolled subjects with T2DM who atherosclerosis remains a controversial mechanisms may potentially contribute were free from apparent CVD and had surrogate for CV effects (19,40). to the reduced atherosclerosis. However, not received insulin treatment. Thus, Third, there may have been mea- the exact mechanism by which DPP-4 not surprisingly, given the earlier stage surement errors in IMT due to interso- inhibitors reverse carotid IMT remains of disease, our subjects had lower HbA1c nographer differences, which were not unknown at present. DPP-4 inhibitors levels and a lower prevalence of hyper- evaluated in this study. However, this block the cleavage and inactivation of tension and dyslipidemia, reflecting parameter was measured by the same stromal cell-derived factor-1, which is lower uses of therapies for other pathol- expert sonographer in each institution known to modulate the mobilization of ogies, than did subjects in the above-cited throughout all of the visits based on endothelial progenitor cells from the studies. In the present clinical trial, we the study protocol. In addition, we bone marrow. A study of patients with used surrogate end points due to practi- did not find significant heterogeneity T2DM demonstrated that DPP-4 inhibi- cal constraints, including trial costs and in changes in IMT among institutions tors increased the number of circulating concern about feasibility in relation to (data not shown). endothelial progenitor cells with con- long-term intervention. We demon- Fourth, multiple testing in primary comitant upregulation of stromal cell- strated the efficacy and benefits of alog- and secondary end points increases derived factor-1 (36). These effects of liptin used at the early stage of disease in the chance of false-positive findings, DPP-4 inhibitors may contribute to preventing the progression of IMT. and thus, our results should be inter- slowing the progression of atheroscle- Consistent with our results, a few preted with caution. rosis. Further studies are needed to in- recent studies reported possible benefi- Finally, it is likely that other yet vestigate these issues. cial effects for DPP-4 inhibitors on the unknown factors can explain the differ- Our results showed a low rate of pro- progression of carotid IMT (38,39). ence in Dchange in IMT between the gression of IMT, even in the conven- Short-term treatment with both sita- two treatment groups. tional group (0.005 mm). Generally, gliptin and vildagliptin reduced the pro- In conclusion, alogliptin treatment therateofcarotidIMTprogressionis gression of IMT in a subanalysis of a attenuated the progression of carotid high in untreated patients with T2DM small number of patients with T2DM, IMT in patients with T2DM free of a his- (24). However, few studies reported although with no control group (38). tory of apparent CVD compared with the that progression of carotid IMT could Another study involving a small group conventional treatment. A large-scale almost be prevented in patients with of subjects showed that sitagliptin treat- prospective trial is required to establish T2DM, even in the control/placebo ment attenuated the progression of IMT the usefulness of DPP-4 inhibitors for group of intervention trials (24), although compared with the control groups in pa- primary prevention of CVD in patients the reports were not always consistent tients with coronary artery disease and with T2DM. (24,37). The exact reason for differences impaired glucose tolerance or mild among these studies was not investi- T2DM (39). However, the current study gated, but we believe that several factors, was different from previous studies with Acknowledgments. The authors thank the such as differences in T2DM status (e.g., respect to study design, sample size, clinical staff who participated in this trial. The authors thank the members of the committee HbA1c, type of medications), control of length of the observation period, and several CV risk factors (e.g., stage of the clinical characteristics of the partic- (Yoshimitsu Yamasaki, Nishi-Umeda Clinic for fi Asian Medical Collaboration; Kazunori Shimada atherosclerosis), prevalence of compli- ipants. This is the rst PROBE trial that and Hirotoshi Ohmura, Department of Cardio- cations (e.g., nephropathy), duration investigated the long-term effects of vascular Medicine, Juntendo University Gradu- of the follow-up period, and race, can alogliptin on the progression of carotid ate School of Medicine; and Ryota Tanaka, affect carotid IMT progression. In this IMT in patients with T2DM free of CVD Department of Neurology, Juntendo University study, BP and total cholesterol levels were that included a relatively large sample Graduate School of Medicine) for their assis- tance with the execution and completion of the significantly lower at 52 weeks relative size. clinical trial. tothebaselineintheconventional The current study has certain limita- Funding and Duality of Interest. Financial treatment group. These changes were tions. First, we used the PROBE design, support for this study was provided by Astellas probably due to the use of statins and which may cause bias in the assessment Pharma Inc., AstraZeneca K.K., Bayer Holding, Daiichi Sankyo Co., Sumitomo Dainippon Pharma calcium channel blockers and probably of outcomes. The reason for the open Co., Eli Lilly Japan K.K., MSD K.K., Nippon Boeh- at least partly explain the low rate of ca- label was practical constraints, including ringer Ingelheim, Novartis Pharma K.K., Novo rotid IMT progression at 52 weeks. It is trialcosts,inaninvestigator-initiated Nordisk Pharma Ltd., Pfizer Japan Inc., Sanofi possible that this effect on carotid IMT trial. In an effort to overcome possible K.K., Sanwa Kagaku Kenkyusho Co., Shionogi & progression was attenuated in the second bias, a single experienced reader, who Co., Ltd., and Takeda Pharmaceutical Co. half of the study by a weaker BP control. was blinded to the clinical characteristics T.M. received research funds from MSD and Takeda Pharma K.K. and has received lecture Three recent randomized clinical of the subjects and type of treatment, fees from AstraZeneca K.K., Boehringer Ingel- studies showed that DPP-4 inhibitors measured carotid IMT using automated heim, Eli Lilly and Co., Kowa Pharmaceutical neither reduce nor increase the risk for digital edge-detection technology. Co., Mitsubishi Tanabe Pharma Co., Mochida CVD in patients with T2DM compared Second, we used surrogate markers Pharmaceutical Co., MSD, Ono Pharmaceutical – Co., and Takeda Pharmaceutical Co. N.Ka. is a staff with placebo (13 15). The subjects of as primary end point and our study member of the endowed chair (the Depart- these three trials were patients with lacked sufficient power to detect dif- ment of Metabolism and Atherosclerosis) estab- T2DM with a history of CVD or at ferences in onset of CVD. In addition, lishedbyfundsfromKowaPharmaceuticalCo.,has care.diabetesjournals.org Mita and Associates 147

received research funds from MSD, and has Pharmaceutical Co., Kissei Pharmaceutical Co., conventional treatment and risk of complica- received lecture fees from Astellas Pharma Kyowa Hakko Kirin Co., Sanofi K.K., Shionogi & tions in patients with type 2 diabetes (UKPDS Inc., AstraZeneca K.K., Boehringer Ingelheim, Co., Ltd., Daiichi Sankyo Co., Sumitomo Dainip- 33). Lancet 1998;352:837–853 Daiichi Sankyo Inc., Sumitomo Dainippon Pharma pon Pharma Co., Takeda Pharma K.K., Mitsubishi 4. Bolen S, Feldman L, Vassy J, et al. Systematic Co., Eisai Co., Eli Lilly and Co., Mitsubishi Tanabe Tanabe Pharma Co., Teijin Pharma, Nippon review: comparative effectiveness and safety of Pharma Co., Mochida Pharmaceutical Co., MSD, Boehringer Ingelheim, Novartis Pharma K.K., oral medications for type 2 diabetes mellitus. Novartis Pharmaceuticals, Novo Nordisk Pharma, Novo Nordisk Pharma, Pfizer Japan Inc., Bristol- Ann Intern Med 2007;147:386–399 Ono Pharmaceutical Co., Otsuka Pharmaceutical, Myers K.K., Mochida Pharmaceutical Co., Eli Lilly 5. Lovshin JA, Drucker DJ. Incretin-based Shionogi & Co., Ltd., Takeda Pharmaceutical Co., Japan K.K., Kowa Pharmaceutical Co., and Taisho therapies for type 2 diabetes mellitus. Nat Rev Teijin Pharma, and Sanofi. T.On. has received Toyama Pharmaceutical Co. H.W. has received Endocrinol 2009;5:262–269 lecture fees from Ono Pharmaceutical Co. H.K. lecture fees from Novo Nordisk, Inc., Eli Lilly 6. Arakawa M, Mita T, Azuma K, et al. Inhibition has received lecture fees from Boehringer and Co., Sanofi,SumitomoDainipponPharma of monocyte adhesion to endothelial cells Ingelheim, Sanofi, Ono Pharmaceutical Co., Co., Fujifilm, Bayer HealthCare, Kissei Pharmaceu- and attenuation of atherosclerotic lesion by a MSD, Novo Nordisk Pharma, Novartis Pharma- tical Co., Mochida Pharmaceutical Co., MSD, glucagon-like peptide-1 receptor agonist, exen- ceuticals, Daiichi Sankyo Inc., Takeda Pharma- Takeda Pharmaceutical Co., Boehringer Ingelheim, din-4. Diabetes 2010;59:1030–1037 ceutical Co., Kissei Pharmaceutical Co., Daiichi Sankyo, Ono Pharmaceutical Co., Novartis 7. Goto H, Nomiyama T, Mita T, et al. Exendin-4, a Sumitomo Dainippon Pharma Co., Mitsubishi Pharmaceuticals, Mitsubishi Tanabe Pharma Corp., glucagon-like peptide-1 receptor agonist, reduces Tanabe Pharma Co., Kyowa Hakko Kirin Co., Eli AstraZeneca LP, Kyowa Hakko Kirin Co., Sanwa Ka- intimal thickening after vascular injury. Biochem Lilly and Co., Pfizer, AstraZeneca, and Astellas gaku Kenkyusyo, Kowa Co., and Astellas Pharma Biophys Res Commun 2011;405:79–84 Pharma Inc. and research funds from Takeda Inc.; advisory fees from Novo Nordisk, Inc., Mo- 8. Gaspari T, Welungoda I, Widdop RE, Simpson Pharmaceutical Co., MSD, Mochida Pharma- chida Pharmaceutical Co., AstraZeneca LP, Kowa RW, Dear AE. The GLP-1 receptor agonist liraglu- ceutical Co. Sanofi, Novartis Pharmaceuticals, Co., Astellas Pharma Inc., Sanofi, Boehringer Ingel- tide inhibits progression of vascular disease via Novo Nordisk Pharma, Eli Lilly and Co., Daiichi heim, MSD, Mitsubishi Tanabe Pharma Corp., No- effects on atherogenesis, plaque stability and Sankyo Inc., Shionogi Pharma, Teijin Pharma, vartis Pharmaceuticals, Sumitomo Dainippon endothelial function in an ApoE(-/-) mouse Sumitomo Dainippon Pharma Co., Otsuka Phar- Pharma Co., Takeda Pharmaceutical Co., Ono model. Diab Vasc Dis Res 2013;10:353–360 fi maceutical, Kissei Pharmaceutical Co., Mitsubishi Pharmaceutical Co., P zer, and Kowa Co.; and re- 9. Wang Y, Parlevliet ET, Geerling JJ, et al. fi Tanabe Pharma Co., Ono Pharmaceutical Co., search funds from Boehringer Ingelheim, P zer, Exendin-4 decreases liver inflammation and fi AstraZeneca, Astellas Pharma Inc., and Kyowa Mochida Pharmaceutical Co., Sano ,NovoNor- atherosclerosis development simultaneously Hakko Kirin Co. T.Os. has received lecture fees disk Pharma, Novartis Pharmaceuticals, Sanwa by reducing macrophage infiltration. Br J Phar- from Novo Nordisk, Inc., Astellas Pharma Inc., Kagaku Kenkyusho, Terumo Corp., Eli Lilly, Mitsu- macol 2014;171:723–734 Mitsubishi Tanabe Pharma, Sanwa Kagaku bishi Tanabe Pharma, Daiichi Sankyo, Takeda 10. Matsubara J, Sugiyama S, Sugamura K, et al. Kenkyusho, Takeda Pharmaceutical Co., and Pharmaceutical Co., MSD, Shionogi Pharma, Su- A dipeptidyl peptidase-4 inhibitor, des-fluoro- Kowa Co. and research funds from Novo Nordisk, mitomo Dainippon Pharma Co., Kissei Pharma- sitagliptin, improves endothelial function and Inc., Astellas Pharma Inc., Mitsubishi Tanabe ceutical Co., and AstraZeneca. No other reduces atherosclerotic lesion formation in apo- fl Pharma, Sanwa Kagaku Kenkyusho, Kowa Co., potential con icts of interest relevant to this ar- lipoprotein E-deficient mice. J Am Coll Cardiol Novo Nordisk Pharma, Sumitomo Dainippon ticle were reported. 2012;59:265–276 Pharma Co., Eli Lilly and Co., Taisho Pharmaceu- Author Contributions. All authors contributed 11. Ervinna N, Mita T, Yasunari E, et al. tical Co., Ltd., GlaxoSmithKline, Astellas Pharma to the study design, were involved in analysis Anagliptin, a DPP-4 inhibitor, suppresses prolif- US, Inc., Bayer HealthCare, and AbbVie GK. T.S. and interpretation of data, and were involved eration of vascular smooth muscles and mono- at all stages of manuscript development, has received lecture fees from Sanofi. K.K. has cyte inflammatory reaction and attenuates reviewed and edited the manuscript, and ap- received lecture fees from Boehringer atherosclerosis in male apo E-deficient mice. proved the final manuscript. T.M. and N.Ka. Ingelheim, Sanofi, Novo Nordisk Pharma, Endocrinology 2013;154:1260–1270 drafted the manuscript. M.G. contributed to Novartis Pharmaceuticals, Eli Lilly and Co., Takeda 12. Shah Z, Kampfrath T, Deiuliis JA, et al. Long- analysis of research data. I.S. and H.W. are the Pharmaceutical Co., MSD, Kowa Co., and term dipeptidyl-peptidase 4 inhibition reduces guarantorsofthisworkand,assuch,hadfullaccess Mitsubishi Tanabe Pharma and research funds atherosclerosis and inflammation via effects on to all the data in the study and take responsibility from Sysmex Co. H.Yok. has received lecture monocyte recruitment and chemotaxis. Circula- for the integrity of the data and the accuracy of fees from Boehringer Ingelheim, Sanofi,Ono tion 2011;124:2338–2349 the data analysis. Pharmaceutical Co., Novo Nordisk Pharma, Novartis 13. White WB, Cannon CP, Heller SR, et al.; Prior Presentation. Parts of this study were Pharmaceuticals, Sanwa Kagaku Kenkyusho, Daiichi EXAMINE Investigators. Alogliptin after acute presented in abstract form at the 75th Scientific Sankyo Inc., Takeda Pharmaceutical Co., MSD, coronary syndrome in patients with type 2 di- Sessions of the American Diabetes Association, Sumitomo Dainippon Pharma Co., and Kowa abetes. N Engl J Med 2013;369:1327–1335 Boston, MA, 5–9 June 2015, and at the Interna- Co. and research funds from Sanofi,NovoNor- 14. Scirica BM, Bhatt DL, Braunwald E, et al.; tional Diabetes Federation’s 2015 World Diabetes disk Pharma, Novartis Pharmaceuticals, Sanwa SAVOR-TIMI 53 Steering Committee and Investi- Congress, Vancouver, Canada, 30 November–4 Kagaku Kenkyusho, Takeda Pharmaceutical Co., gators. Saxagliptin and cardiovascular outcomes December 2015. and MSD. N.Ku. has received lecture fees from in patients with type 2 diabetes mellitus. N Engl J Novo Nordisk Pharma, Novartis Pharmaceuticals, Med 2013;369:1317–1326 Takeda Pharmaceutical Co., and Eli Lilly and Co. 15. Green JB, Bethel MA, Armstrong PW, et al.; M.G. has received lecture fees from Novartis. I.S. References TECOS Study Group. Effect of sitagliptin on has received lecture fees from Astellas Pharma 1. Haffner SM, Lehto S, Ronnemaa¨ T, Pyor¨ al¨ aK,¨ cardiovascular outcomes in type 2 diabetes. Inc., AstraZeneca K.K., MSD K.K., Ono Pharmaceu- Laakso M. Mortality from coronary heart N Engl J Med 2015;373:232–242 tical Co., Kyowa Hakko Kirin Co., Kowa Pharma- disease in subjects with type 2 diabetes and in 16. Holman RR, Paul SK, Bethel MA, Matthews ceutical Co., Sanofi K.K., Sanwa Kagaku Kenkyusho nondiabetic subjects with and without prior DR, Neil HA. 10-year follow-up of intensive Co.,DaiichiSankyoCo.,TakedaPharmaK.K., myocardial infarction. N Engl J Med 1998;339: glucose control in type 2 diabetes. N Engl J Mitsubishi Tanabe Pharma Co., Teijin Pharma, 229–234 Med 2008;359:1577–1589 Eli Lilly Japan K.K., Nippon Boehringer Ingelheim, 2. Stratton IM, Adler AI, Neil HA, et al. Association 17. Yamasaki Y, Kodama M, Nishizawa H, et al. Novartis Pharma K.K., Novo Nordisk Pharma, of glycaemia with macrovascular and micro- Carotid intima-media thickness in Japanese Bayer Yakuhin, Pfizer Japan Inc., Bristol-Myers vascular complications of type 2 diabetes type 2 diabetic subjects: predictors of progres- K.K., Mochida Pharmaceutical Co., Shionogi & Co., (UKPDS 35): prospective observational study. sion and relationship with incident coronary Ltd., and Taisho Toyama Pharmaceutical Co. BMJ 2000;321:405–412 heart disease. Diabetes Care 2000;23:1310–1315 and research funds from Astellas Pharma Inc., 3. UK Prospective Diabetes Study (UKPDS) 18. Yoshida M, Mita T, Yamamoto R, et al. AstraZeneca K.K., Eisai Co., MSD K.K., Otsuka Group. Intensive blood-glucose control with Combination of the Framingham risk score and Pharmaceutical, Ono Pharmaceutical Co., Kaken sulphonylureas or insulin compared with carotid intima-media thickness improves the 148 Effects of Alogliptin on Atherosclerosis Diabetes Care Volume 39, January 2016

prediction of cardiovascular events in patients peptidase-IV inhibition. Diabetes Care 2012;35: on carotid intima-media thickness in type 2 di- with type 2 diabetes. Diabetes Care 2012;35: 2076–2082 abetes: a randomized trial. JAMA 2006;296: 178–180 26. Tremblay AJ, Lamarche B, Deacon CF, 2572–2581 19. Okayama KI, Mita T, Gosho M, et al. Carotid Weisnagel SJ, Couture P. Effects of sitagliptin 34. Wang N, Yin R, Liu Y, Mao G, Xi F. Role of intima-media thickness progression predicts therapy on markers of low-grade inflammation peroxisome proliferator-activated receptor- cardiovascular events in Japanese patients and cell adhesion molecules in patients with gamma in atherosclerosis: an update. Circ J with type 2 diabetes. Diabetes Res Clin Pract type 2 diabetes. Metabolism 2014;63:1141–1148 2011;75:528–535 2013;101:286–292 27. Satoh-Asahara N, Sasaki Y, Wada H, et al. 35. Nagashima M, Watanabe T, Terasaki M, et al. 20. Katakami N, Mita T, Yoshii H, et al.; A dipeptidyl peptidase-4 inhibitor, sitagliptin, Native incretins prevent the development of ath- Collaborators on the Study of Preventive Effects exerts anti-inflammatory effects in type 2 dia- erosclerotic lesions in apolipoprotein E knockout of Alogliptin on Diabetic Atherosclerosis Trial. betic patients. Metabolism 2013;62:347–351 mice. Diabetologia 2011;54:2649–2659 Rationale, design, and baseline characteristics 28. Satoh N, Shimatsu A, Yamada K, et al. An 36. Fadini GP, Boscaro E, Albiero M, et al. The of a trial for the prevention of diabetic athero- alpha-glucosidase inhibitor, voglibose, reduces oral dipeptidyl peptidase-4 inhibitor sitagliptin sclerosis using a DPP-4 inhibitor: the Study of oxidative stress markers and soluble intercellu- increases circulating endothelial progenitor Preventive Effects of Alogliptin on Diabetic Ath- lar adhesion molecule 1 in obese type 2 diabetic cells in patients with type 2 diabetes: possible erosclerosis (SPEAD-A). J Atheroscler Thromb patients. Metabolism 2006;55:786–793 role of stromal-derived factor-1alpha. Diabetes 2013;20:893–902 29. Gensini GF, Gori AM, Dilaghi B, et al.; Care 2010;33:1607–1609 21. The Japan Diabetes Society. Treatment Guide Achieve Cholesterol Targets Fast with Atorvas- 37. Fleg JL, Mete M, Howard BV, et al. Effect of for Diabetes [in Japanese]. Editorial Committee tatin Stratified Titration Investigators. Effect of statins alone versus statins plus ezetimibe on Members for the Treatment Guide for Diabetes, atorvastatin on circulating hsCRP concentrations: carotid atherosclerosis in type 2 diabetes: the Ed. Tokyo, Japan, Bunkodo Co., Ltd., 2014 a sub-study of the Achieve Cholesterol Targets Fast SANDS (Stop Atherosclerosis in Native Diabetics 22. Yanase T, Nasu S, Mukuta Y, et al. Evaluation with Atorvastatin Stratified Titration (ACTFAST) Study) trial. J Am Coll Cardiol 2008;52:2198– of a new carotid intima-media thickness mea- study. Int J Cardiol 2010;142:257–264 2205 surement by B-mode ultrasonography using an 30. Balestrieri ML, Rizzo MR, Barbieri M, et al. 38. Barbieri M, Rizzo MR, Marfella R, et al. De- innovative measurement software, intimascope. Sirtuin 6 expression and inflammatory activity creased carotid atherosclerotic process by con- Am J Hypertens 2006;19:1206–1212 in diabetic atherosclerotic plaques: effects of in- trol of daily acute glucose fluctuations in 23. Matsuo S, Imai E, Horio M, et al.; cretin treatment. Diabetes 2015;64:1395–1406 diabetic patients treated by DPP-IV inhibitors. Collaborators developing the Japanese equa- 31. Langenfeld MR, Forst T, Hohberg C, et al. Atherosclerosis 2013;227:349–354 tion for estimated GFR. Revised equations for Pioglitazone decreases carotid intima-media 39. Ishikawa S, Shimano M, Watarai M, et al. estimated GFR from serum creatinine in Japan. thickness independently of glycemic control in Impact of sitagliptin on carotid intima-media Am J Kidney Dis 2009;53:982–992 patients with type 2 diabetes mellitus: results thickness in patients with coronary artery dis- 24. Yokoyama H, Katakami N, Yamasaki Y. from a controlled randomized study. Circulation ease and impaired glucose tolerance or mild di- Recent advances of intervention to inhibit pro- 2005;111:2525–2531 abetes mellitus. Am J Cardiol 2014;114:384–388 gression of carotid intima-media thickness in 32. Koshiyama H, Shimono D, Kuwamura N, 40. Lorenz MW, Polak JF, Kavousi M, et al.; patients with type 2 diabetes mellitus. Stroke Minamikawa J, Nakamura Y. Rapid communica- PROG-IMT Study Group. Carotid intima-media 2006;37:2420–2427 tion: inhibitory effect of pioglitazone on carotid thickness progression to predict cardiovascular 25. Rizzo MR, Barbieri M, Marfella R, Paolisso G. arterial wall thickness in type 2 diabetes. J Clin events in the general population (the PROG-IMT Reduction of oxidative stress and inflammation Endocrinol Metab 2001;86:3452–3456 collaborative project): a meta-analysis of indi- by blunting daily acute glucose fluctuations in 33. Mazzone T, Meyer PM, Feinstein SB, et al. vidual participant data. Lancet 2012;379:2053– patients with type 2 diabetes: role of dipeptidyl Effect of pioglitazone compared with glimepiride 2062