Newly detected impaired glucose tolerance (IGT) or type 2 diabetes mellitus (T2DM) are common in patients with acute coronary syndrome (ACS; i.e. unstable angina/myocardial infarction) and related to disturbed beta-cell function. The aim of this study is to test the hypothesis that treatment with a dipeptidyl peptidase-4 inhibitor initiated soon after a coronary event improves beta-cell function.
Acute coronary syndromeACS patients with IGT or T2DM (n =71), screened by oral glucose tolerance test (OGTT) 4–23 days (median 6 days) after hospital admission, were randomly assigned to sitagliptin 100 mg (n =34) or placebo (n =37) and treated for a duration of 12 weeks. All patients received lifestyle advice but no glucose-lowering agents other than the study drug. The study end-point was beta-cell function assessed using the insulinogenic index (IGI = ΔInsulin30/ΔGlucose30), derived from an OGTT, and acute insulin response to glucose (AIRg) assessed by a frequently sampled intravenous glucose tolerance test.
The IGI and AIRg did not differ at baseline between the sitagliptin and placebo groups (69.9 vs. 66.4 pmol mmol−1 and 1394 vs. 1106 pmol L−1 min−1 respectively). After 12 weeks, the IGI was 85.0 in the sitagliptin and 58.1 pmol/mmol in the placebo group (P =0.013) and AIRg was 1909 and 1043 pmol L−1 min−1 (P <0.0001) in the sitagliptin and placebo groups respectively. Fasting glucose at baseline was 6.1 mmol L−1 in sitagliptin-treated patients and 6.0 mmol L−1 in those who received placebo compared with 5.8 and 5.9 mmol L−1 respectively, after 12 weeks of treatment. Post load glucose metabolism improved in significantly more sitagliptin-treated patients compared with the placebo group (P =0.003). Sitagliptin was well tolerated.
Sitagliptin improved beta-cell function and glucose perturbations in patients with ACS and newly diagnosed glucose disturbances.
BEta-cell function in Glucose Abnormalities and Acute Myocardial Infarction
Coronary care unit
Frequently sampled intravenous glucose tolerance test
Glucose Abnormalities and Acute Myocardial Infarction
Glucagon-like peptide 1
Homeostasis model assessment
High-sensitivity C-reactive protein
International Federation of Clinical Chemistry
Impaired glucose lolerance
N-terminal pro-B-type natriuretic peptide
Type 2 diabetes mellitus
Newly detected impaired glucose tolerance (IGT) and type 2 diabetes mellitus (T2DM) are common in patients with acute coronary syndrome (ACS) [1-3] and have an unfavourable prognostic impact [4, 5]. This may partly be due to a negative effect of hyperglycaemia per se, but factors such as endothelial dysfunction, inflammation, dyslipidaemia and still unknown reasons may also play a role .
Type 2 diabetes mellitus is characterized by insulin resistance in combination with progressive beta-cell failure. Deteriorating beta-cell function has been observed in longitudinal studies as patients progress from normal glucose metabolism to IGT and finally to T2DM . Initially plasma glucose remains normal as the beta-cells compensate for reduced insulin sensitivity by increasing insulin secretion, but progressive beta-cell dysfunction results in first postprandial and subsequently fasting hyperglycaemia [6, 8, 9]. Patients in the upper tertile of IGT have lost nearly 80% of their beta-cell function and it continues to decrease at a rate of approximately 4% per year irrespective of the initiation of glucose-lowering treatment [9, 10].
In the Glucose Tolerance in Patients with Acute Myocardial Infarction (GAMI) study, insulin resistance, expressed as homeostasis model assessment (HOMA-IR) index, did not differ between patients with acute myocardial infarction (AMI) plus newly discovered glucose disturbances and matched controls. By contrast, the proinsulin levels were significantly increased in patients compared to control subjects indicating impaired beta-cell function, also expressed as a compromised insulinogenic index (IGI) amongst the patients [11, 12]. An improvement in beta-cell function may delay or even reverse the progression to T2DM, which may be of prognostic importance in patients with AMI. The incretin glucagon-like peptide 1 (GLP-1), which is secreted in the gut in response to food ingestion, exerts a glucose-regulatory action via stimulation of insulin secretion and suppression of glucagon in a glucose-dependent manner. It is rapidly degraded by the enzyme dipeptidyl peptidase-4 (DPP-IV). Furthermore, GLP-1 has beneficial effects on the myocardium by increasing myocardial glucose uptake, improving endothelial function and potentially reducing infarct size; it also has anti-inflammatory and anti-atherogenic properties [13-18]. In addition to their glucose-lowering effect, GLP-1-based therapies seem to protect beta-cells by enhancing cell proliferation and differentiation, inhibiting apoptosis and stimulating insulin biosynthesis/secretion [19-22]. The DPP-IV inhibitor sitagliptin preserves the levels of endogenous GLP-1 by inhibiting degradation of the enzyme and thereby improves glycaemic control .
The aim of this study is to test the hypothesis that sitagliptin can improve beta-cell function in patients with recent ACS plus newly diagnosed IGT or T2DM, and that such treatment may be safely initiated soon after the coronary event.
Beta-cell function in Glucose abnormalities and Acute Myocardial Infarction (BEGAMI) was a multicentre, double-blind, randomized, parallel-group study (Clinicaltrials.gov: NCT00627744) compares the effects of once daily administration of 100 mg sitagliptin (Januvia™, Merck Sharp & Dohme AB, USA) and placebo on beta-cell function in patients with AMI or unstable angina pectoris and newly detected IGT or T2DM. Patients were recruited from the coronary care units (CCUs) at Karolinska University Hospital and Danderyds Hospital both in Stockholm, Sweden. Recruitment started in May 2008 and follow-up was completed in December 2010.
The primary end-point was improvement in beta-cell function after 12 weeks treatment measured using the IGI derived from an oral glucose tolerance test (OGTT). Secondary end-points included (i) improvement in glucose tolerance measured by an OGTT after 12 weeks; (ii) improvement in GLP-1-independent beta-cell function after 12 weeks measured as the acute insulin response to glucose (AIRg) derived from a frequently sampled intravenous glucose tolerance test (FSIGT).
Patients admitted to the CCU due to unstable angina pectoris or AMI according to the joint European Society of Cardiology and American College of Cardiology recommendations  were screened for study participation. Exclusion criteria were previously known type 1 or 2 diabetes, admission plasma glucose >12 mmol L−1, age <18 years, impaired renal function (Serum creatinine ≥130 μmol L−1), body mass index ≥34 kg m−2, heart failure (New York Heart Association classes III/IV) and inability to follow the study protocol or planned coronary revascularization.
A screening OGTT was performed under stable conditions on the morning of the day of discharge (at least 4 days after hospital admission) or post discharge during a scheduled visit within 3 weeks after admission; the median time of OGTT after admission was 7 days in the sitagliptin group, lower and upper quartile values (Q1;Q3; 4;12) and 6 (Q1;Q3 4;8) days in the placebo group. Patients with newly discovered IGT or T2DM returned to the clinic the following day for an FSIGT, and were subsequently randomized to study treatment.
Patients were assigned randomly to receive tablets of 100 mg sitagliptin or matching placebo once daily in a double-blind fashion with a ratio of 1:1 and block size of four via a computer-generated randomization sequence.
Examinations were performed at study entry and after 6 and 12 weeks. At baseline, all patients were given structured lifestyle advice in the form of oral and written information on diet and physical activity. Patients were given an Accu-Check Compact Plus glucometer (Roche Diagnostics Scandinavia, AB) and instructed to measure their glucose level regularly during the study period and record it in a diary (also provided). Patients underwent an OGTT and an FSIGT on two consecutive days both at baseline and during the final visit; biochemistry and demographic data were also collected as described below.
The visit at 6 weeks was a safety check-up including physical examination, evaluation of glucose control (self-monitored fasting and postprandial measurements recorded in the diary), reinforcement of lifestyle advice and completion of an adverse events record form.
The OGTT (oral administration of 75 g glucose in 200 mL water) was performed in the morning following an overnight fast for 12 h. A venous catheter for blood sampling was inserted into an antecubital vein and blood was drawn prior to and at 30, 60 and 120 min after glucose ingestion. The last dose of sitagliptin (or placebo) was administered on the day prior to the OGTT at the 12-week follow-up visit.
An FSIGT was performed on the day after the OGTT, also in the morning after fasting overnight. A catheter was inserted into an antecubital vein for blood sampling and a contralateral antecubital vein for glucose injection. A basal sample was drawn 5 min before injection of 300 mg kg−1 glucose over a period of 2 min. Blood samples were collected at 2, 5, 7, 10, 20, 30, 60 and 120 min after the start of injection.
During the OGTT and FSIGT blood was collected in EDTA-containing tubes for immediate analysis of plasma glucose by photometry (HemoCue® AB, Ängelholm, Sweden). Samples were kept on ice until centrifugation within 1 h at 2000 g for 20 min, and then plasma was stored at −70 °C until required for analyses. Plasma insulin was analysed after termination of the study and quantified with a commercially available specific immunoassay (DAKO Ltd, Cambridgeshire, UK) and plasma proinsulin was measured by enzyme-linked immunosorbent assay (Mercodia AB, Uppsala, Sweden).
At randomization and at the 12-week visit, fasting blood samples were collected for measurement of glycated haemoglobin A1c (HbA1c) and levels of glucose, cholesterol, HDL, LDL, triglycerides, creatinine, troponin I, high-sensitivity C-reactive protein (hsCRP) and N-terminal pro-b-type natriuretic peptide (NT-proBNP).
Data including blood pressure, heart rate, waist circumference, weight, ECG and adverse events were collected at baseline and at the 6- and 12-week visits. The level of physical activity was estimated before randomization and at the end of follow-up using a scale of 1–6 (1 = inactive, 6 = vigorous exercise ≥ 3 h per week).
Statistical analysis and calculations
Continuous variables are expressed as median and Q1;Q3 and the Wilcoxon rank-sum test was used to determine differences between groups. Categorical variables are expressed as numbers and percentages and analysed using Fisher's exact test.
Calculation of the required sample size was based on a previous study in which the IGI was measured in patients with AMI and newly discovered IGT or TDM2 . The mean ± standard deviation of the ΔIGI was 50 ± 35 pmol mmol−1. To detect a difference between the treatment groups of 25 pmol mmol−1 from baseline to the end of the observation period at 12 weeks, at a 5% significance level with 80% power using a two-tailed t-test, a sample size of 32 patients per group was considered to be sufficient. This number was increased by 10% to enable the use of nonparametric methods and to allow for potential loss of patients, resulting in a total sample size of 70 patients. P-values represent the comparison of delta values between groups after 12 weeks of treatment with study drug. Glucose tolerance categories were compared between groups at 12 weeks using the chi-square test for trend. All statistical analyses were performed using SAS software version 9.2 (SAS Institute).
The primary study end-point, beta-cell function after 12 weeks of treatment, was assessed as the IGI calculated as Δinsulin0–30/Δglucose0–30 obtained from the OGTT . The AIRg was calculated from the FSIGT as the incremental area under the curve from 0 to 10 min and the glucose disappearance constant (Kg) as the slope of the natural logarithm of the difference between the two glucose samples taken at 10 and 20 min (Kg = [Δ ln plasma glucose/Δ min] × 100). As a measure of insulin resistance the HOMA-IR index was calculated as fasting glucose (mmol L−1) × fasting insulin (mU L−1)/22.5 . Creatinine clearance (mL min−1) was calculated according to the Cockcroft–Gault formula: 140−age × weight (kg) × constant/Serum creatinine (μmol L−1), where the constant is 1.23 for men and 1.04 for women .
The BEGAMI study was conducted according to International Conference on Harmonisation and Good Clinical Practice guidelines and the Declaration of Helsinki. The trial was approved by the ethics committee at Karolinska Institutet. Informed consent, written and oral, was obtained from all patients prior to inclusion.
As shown in Fig. 1, overall 174 patients underwent the screening OGTT: 75 (43%) had a normal glucose tolerance, 63 (36%) had IGT and 36 (21%) had T2DM. Of the 99 eligible patients (i.e. those with IGT or T2DM), 79 were assigned randomly to receive sitagliptin (n =39) or placebo (n =40). Seven patients discontinued participation prior to the second OGTT and data were incomplete for one patient, therefore the total study cohort comprised 71 patients. Data from the FSIGT were available for 66 of these patients. Compliance to study drug, performed by pill count, was 100% in the sitagliptin group and 99% in the placebo group (P =0.4).
Biochemical and clinical characteristics
Of the 71 patients, 32 had ST segment elevation AMI (STEMI), 32 had non-STEMI and seven fulfilled the criteria for unstable angina. Medical history did not differ between the two treatment groups (Table 1). Biochemical and clinical characteristics at randomization and at follow-up after 12 weeks of treatment are listed in Table 2. There were no significant differences in the change in weight, waist circumference or physical activity from baseline to 12 weeks between the two groups. HbA1c improved from 40 to 38 mmol mol in the sitagliptin group, but remained unaltered in the placebo group (P =0.004).
Table 1. Patient characteristics at baseline. Continuous variables are presented as median (quartile 1; quartile 3) and categorical variables as numbers (%)
Table 2. Biochemical and clinical characteristics presented at randomization and after 12 weeks of treatment. Continuous data are presented as median (quartile 1; quartile 3) and categorical variables as numbers (%) unless otherwise stated. P-values represent comparison of values between groups at 12 weeks
At the time of randomization, 24 (71%) patients in the sitagliptin group had IGT and 10 (29%) had T2DM. The corresponding numbers in the placebo group were 23 (62%) and 14 (38%) (ns; Table 2). The OGTT performed after 12 weeks of treatment revealed an improvement in glucometabolic categorization in both groups, which was significantly more pronounced (P =0.003) in the sitagliptin group (Fig. 2). By 12 weeks, 26 (76%) of the sitagliptin-treated patients had a normal glucose tolerance whilst six (18%) had IGT and two had (6%) T2DM. The corresponding numbers in the placebo group were 15 (41%), 13 (35%) and nine (24%).
Beta-cell function expressed as IGI increased significantly from baseline to 12 weeks from 69.9 (50.0; 115.7) to 85.0 (59.4; 143.8) pmol mmol−1 in sitagliptin-treated patients compared to 66.4 (37.3; 88.6) to 58.1 (39.7; 96.8) pmol mmol−1 in the placebo group (P between groups = 0.013; Fig. 3a). At randomization, there was no significant difference in AIRg between the two groups: 1394 (908; 1929) vs. 1106 (628; 2172; P =0.45) pmol L−1 min−1. The AIRg increased significantly during the treatment period in the sitagliptin compared to that in the placebo group: 1909 (1235; 2651) vs. 1043 (662; 2044) pmol L−1 min−1 (P <0.0001; Fig. 3b). After 12 weeks of treatment, the fasting and 120-min post-load levels of proinsulin decreased in both groups, although to a greater extent in the sitagliptin group (Table 2). The glucose disappearance index (Kg) improved in the sitagliptin compared to that in the placebo group (Fig. 3c). Insulin resistance did not significantly differ between the two groups at the time of randomization or change in either of the groups during follow-up (Fig. 3d). The glucose and insulin response expressed as the area under the curve during a period of 120 min during the OGTT and FSIGT before and after 12 weeks of treatment with sitagliptin or placebo are shown in Fig. 4. Glucose levels were significantly lower in the sitagliptin compared with those in the placebo group during the OGTT (P <0.001) and the FSIGT (P =0.001; Fig. 4a, c). The total insulin response was not significantly different between the two groups (OGTT P =0.9; FSIGT P =0.5). However, as shown in Fig. 4d, the immediate insulin response was higher after sitagliptin treatment (see also Fig. 3b) but unchanged after placebo.
Adverse events were predominantly related to cardiovascular disease and equally distributed between the two treatment groups. None of these events was considered to be related to the study drug. Symptoms potentially related to study drug occurred in six sitagliptin-treated patients of whom three had nasopharyngitis and three had gastrointestinal symptoms. The latter side effect was also reported by two placebo recipients. The lowest measured plasma glucose concentration was 3.5 mmol L−1. None of the patients reported any symptoms of hypoglycaemia during the study. A detailed description of side effects is given in Table 3 (appendix).
Table 3. Number of patients reporting adverse events during the study period by type of event and treatment group allocation. Events labelled as serious are those which rendered a need for hospitalization
This study confirms previous observations that beta-cell function is disturbed in patients with ACS and newly detected glucose perturbations. We have also shown that the administration of the DPP-IV inhibitor sitagliptin improves beta-cell function. These findings suggest that sitagliptin could be safely instituted soon after a coronary event.
In general, the prognosis after an ACS has improved but to a lesser extent in patients with T2DM. Because newly detected glucose disturbance has negative prognostic implications [4, 5, 28], continued efforts to improve the impact of such conditions are required. The findings of this study are consistent with previous observations that unknown glucose abnormalities are common in ACS populations [1-3]. The majority of patients in this study had IGT and T2DM determined by elevated post load glucose rather than increased levels of fasting glucose. This underlines the importance of performing an OGTT as a part of total risk assessment in patients with ACS free from previously known glucose perturbations and with normal fasting glucose and HbA1c levels. Postprandial glycaemia is more strongly correlated with subsequent cardiovascular morbidity and mortality than an elevated fasting glucose [29, 30].
According to international guidelines , all patients received structured lifestyle counselling. Adherence to these recommendations is a likely explanation for some of the improvement in glucose tolerance seen in the placebo group, which was considerably less pronounced than that observed in sitagliptin-treated patients. Of note, no glucose-lowering drugs other than sitagliptin were prescribed in either group.
Previous investigations by our group have revealed high proinsulin levels reflecting beta-cell dysfunction in patients with ACS [11, 12]. These findings were replicated in the present patient population in which proinsulin levels were somewhat higher than those in our previous study, whilst the HOMA-IR index was of a similar magnitude. The likely explanation for the hyperinsulinaemia during the screening OGTT is the impaired insulin sensitivity, which improved over time in both groups (Fig. 4b) probably reflecting the impact of an improved lifestyle. A possible but less likely explanation for this improvement may be an increase in insulin sensitivity due to decreasing sympathetic activity after the acute event. Previous investigations by our group did, however, demonstrate that an OGTT performed 4–5 days and 3 months after an ACS gave similar results with regard to glucose categorization . The median time between the event and the screening OGTT was 6–7 days in this study.
The present findings support the notion that GLP-1 analogues and DPP-IV inhibitors may enhance beta-cell function by stimulating insulin secretion and synthesis, indicating that they may be a rational treatment option for ACS patients with abnormal glucose regulation and with pronounced beta-cell dysfunction. These agents act in a glucose-dependent manner eliminating the risk of hypoglycaemia, which could be of particular importance in patients with recent ACS [32, 33]. In contrast to GLP-1 analogues, DPP-IV inhibitors are orally available, which simplifies the initiation of treatment.
The primary end-point of this study, IGI, was derived from an OGTT because the evaluation of beta-cell function in terms of insulin secretion is more informative following a glucose challenge than in the fasting state. The IGI calculated from an oral glucose load reflects the insulinotrophic effects of incretins induced by increased levels of GLP-1 and gastric inhibitory polypeptide, and decreased glucagon secretion. An intravenous glucose tolerance test was also performed to evaluate a potential incretin-independent effect on beta-cell function measured as the AIRg. At the follow-up visit, the OGTT and FSIGT were performed in a fasting state on two consecutive days (the OGTT first) and the last dose of sitagliptin was administered on the morning prior to the OGTT. The half-life of sitagliptin is 12.4 h indicating a sustained effect at the time of the OGTT and probably, although attenuated, also at the time of the FSIGT. Thus, both investigations were to some extent influenced by sitagliptin.
An important consideration is the safety of the initiation of DPP-IV inhibitors close to an acute coronary event. Sitagliptin has primarily been investigated in patients with T2DM and in such studies has not been associated with cardiovascular events ; nevertheless it has not been previously studied in patients with ACS. Established coronary artery disease was indeed an exclusion criterion in many previous trials. The findings of this study suggest that early initiation of treatment after hospitalization for an ACS is safe. During the study period minor adverse events were equally distributed between the two treatment groups and the serious adverse events reported were those that could be expected in a population of ACS patients, and were considered unrelated to the study drug.
The practical implications of the present findings are that early intervention with sitagliptin is an interesting treatment modality that, by counteracting a pathophysiological deficit in patients with ACS, may offer the potential to prevent further progression of the dysglycaemic condition and of atherosclerotic disease manifestations. These assumptions need verification in a prospective trial designed not only to look at mechanistic features but also the long-term impact on glucose metabolism and outcome-related variables such as reinfarction, stroke and cardiovascular mortality. The results of ongoing trials, such as the Sitagliptin Cardiovascular Outcome Study (TECOS; ClinicalTrials.gov identifier: NCT00790205), are eagerly awaited.
Sitagliptin is able to restrain alpha-cell hyperactivity in T2DM thereby suppressing glucagon release that may have contributed to the reduced glucose levels in the sitagliptin group after 12 weeks of treatment . Glucagon was not measured during the OGTT and FSIGT in this study and therefore the specific effects on glucagon levels in this particular population are not known.
Sitagliptin improved beta-cell function and normalized glucose perturbations in patients with ACS and newly discovered glucose disturbances. Based on the present results, it is proposed that this DPP IV inhibitor, by targeting a core deficit in ACS patients with newly detected glucose perturbations, may prevent future development of T2DM as well as the accompanying cardiovascular complications. An important finding is that the drug could be safely introduced shortly after the acute coronary event.
Conflict of interest statement
L. Mellbin has received research grants from MSD and Sanofi Aventis, L. Rydén has received research grants from the Swedish Heart- and Lung Foundation, AFA Insurance, Karolinska Institutet, County Council of Stockholm, Astra Zeneca and Hoffmann-La Roche. He has received consultancy honoraria from Bristol Myers Squibb, Hoffman La Roche and Astra Zeneca.
The authors are grateful to E. Wallén Nielsen, J. Rasck and M. Asperen at Danderyd Hospital for blood sampling and patient care, to I-L. Wivall and E. Sandberg at Karolinska Institutet for laboratory analysis and to Drs A. Norhammar and M. Wallander at Karolinska Institutet for excellent advice.
This work was supported by grants from the Swedish Heart and Lung Foundation, AFA Insurance, Erling-Persson Family Foundation and the Medical Research Council (04224). The study drug and placebo, but no financial support, were provided by Merck Sharp & Dohme AB. Camilla Hage is the guarantor of this manuscript. Pia Lundman, Lars Rydén and Linda Mellbin contributed to data analysis and drafting of the manuscript. Kerstin Brismar and Suad Efendic contributed to discussions and reviewed and edited the manuscript.