Intensification of Therapeutic Approaches Reduces Mortality in Diabetic Patients With Acute Myocardial Infarction: The Munich registry

The data of the Munich MI registry of 1999 was analyzed with regard to mortality and therapeutic approaches. It was then the aim to intensify treatment in diabetic patients. In 2001, the goal was to apply strategies at comparable levels between diabetic and nondiabetic patients. Furthermore, in view of the results of the DIGAMI study (10), glucose-insulin infusion was administered in diabetic patients. The results of hospital outcome were compared.

The diagnosis of diabetes in the registry was based on the following criteria. The presence of diabetes was defined if the patient had been informed of this diagnosis or was on prescribed antidiabetic treatment (diet, tablets, or insulin). Patients without the diagnosis but with a blood glucose ≥200 mg/dl (hexokinase reaction) were also classified as having diabetes.

Acute MI was diagnosed in the presence of two or more of the following criteria (9): persistent angina pectoris for ≥15 min, ST segment elevation in at least two contiguous standard leads of ≥1 mm and ≥2 mm in precordial leads, respectively, and elevation of serum creatinine kinase above the normal range of 70 units/l. Patients were classified according to the recommendations of the European Society of Cardiology and the American College of Cardiology (11).

Patients with acute MI were treated according to the guidelines of the American College of Cardiology and the American Heart Association (12). Fibrinolytic therapy with alteplase or acute percutaneous intervention were applied as reperfusion strategies. In Germany, the use of thrombolysis in diabetic patients is not limited by local restrictions. Early percutaneous transluminal coronary angioplasty (PTCA) was defined as performance of PTCA within 6 h after admission.

In 2001, diabetic patients with acute MI and who were aged <75 years obtained a glucose-insulin infusion for a minimum of 24 h. The scheme of insulin infusion was a modification of the protocol of the DIGAMI study (10): 10% glucose was infused at 30 ml/h. Regular insulin (50 IU) was dissolved in 50 ml of 0.9% NaCl. Insulin was infused according to the blood glucose level: >250 mg/dl (13.9 mmol/l), 3 IU/h; 200−249 mg/dl (11.1–13.8 mmol/l), 2.5 IU/h; 150–199 mg/dl (8.3–11.0 mmol/l), 1.5 IU/h; 100–149 mg/dl (5.6–8.2 mmol/l), 1.0 IU/h; and <100 mg/dl (5.6 mmol/l), insulin infusion ceased. Insulin was applied seperately from glucose solution and was administered with an external pump (Perfusor FM; B. Braun Melsungen, Melsungen, Germany). Blood glucose was tested every 2 h (Glucometer Elite; Bayer Vital, Leverkusen, Germany). Administration of glucose-insulin infusion was succeeded by an intensive insulin therapy that comprised four to six injections of regular insulin and two injections of basal insulin.

Discharge from the intensive care unit was determined by the decline of serum creatinine kinase to the near-normal range and was succeeded by admittance to the general ward of the Department of Cardiology.

Plasma glucose, HBA_1c (normal range <6.2%), creatinine kinase, electrolytes, blood count, lipids, and C-reactive protein (CRP) were analyzed. Albuminuria was defined as >300 mg/l. Hypertension was defined according to World Health Organization criteria, which include multiple measurements of blood pressure of ≥140/90 mmHg (13). The history of peripheral artery disease was assessed in all patients. Resuscitation was defined as a cardiopulmonary resuscitation, which included cardiac massage, use of a heart defibrillator, and/or artificial respiration.

If not otherwise specified, the results are reported as means ± SD. Group comparisons were performed by the Mann-Whitney test for continuous variables and χ² test for categorical variables. Logistic regression analyses were used to identify the influence of clinical variables (age, sex, BMI, heart rate, blood glucose on admission, total cholesterol, triglycerides, creatinine, CRP, maximal value of creatinine kinase, hypertension, and previous coronary artery disease) and the number of treatment approaches performed in the patients (coronary angiography, PTCA, stenting, GbIIb/IIIa, fibrinolysis, coronary artery bypass grafting, and glucose-insulin infusion). Crude odds ratios (ORs) and CIs were calculated. They were adjusted for age, sex, and clinical variables (adjusted OR). The statistical analyses were performed with SAS version 8.2.

Clinical characteristics of diabetic and nondiabetic patients with acute MI who were admitted in 1999 and 2001 are presented in Table 1. In diabetic patients, age, sex, and smoking distribution, BMI, heart rate on admission, duration of diabetes, glycemic control, lipid profile, kidney function, CRP, and maximal value of creatinine kinase were comparable between 1999 and 2001. All characteristics, except sex distribution and age, were comparable in nondiabetic patients.

Therapeutic approaches changed between 1999 and 2001 (Table 2). In 1999, coronary angiography (P < 0.01), PTCA (P < 0.001), and stenting (P < 0.001) were performed less frequently in diabetic patients than in nondiabetic patients. In 2001, performance of these therapeutic options was substantially increased (P < 0.001 vs. 1999) to levels comparable with those of nondiabetic patients. Also, GbIIb/IIIa was administered more frequently in 2001 than in 1999 (P < 0.001).

In nondiabetic patients, early PTCA (P < 0.01), stenting (P < 0.05), and GbIIb/IIIa (P < 0.001) were applied more frequently in 2001 than in 1999. In both diabetic and nondiabetic patients, the application of fibrinolysis decreased (P < 0.05). Performance of coronary artery bypass grafting did not change significantly in the groups.

The mortality rates were consistently higher in diabetic patients than in nondiabetic patients. In 1999, total hospital mortality of diabetic patients was nearly twofold increased compared with nondiabetic patients (29 vs. 16%, P < 0.01). In 2001, total hospital mortality was 17% in diabetic patients and 14% in nondiabetic patients (P = NS). In 1999, hospital mortality was nearly threefold higher within the first 24 h after admission (P = 0.01) (Fig. 1), whereas subsequent in-hospital mortality was not significantly different between the two groups (Fig. 1).

In 2001, mortality within 24 h after admission and subsequent in-hospital mortality did not differ significantly between diabetic and nondiabetic patients (Fig. 1). In diabetic patients, total hospital mortality was reduced by 44% (P = 0.028, 2001 vs. 1999), and 24-h mortality decreased by 67% (P = 0.027, 2001 vs. 1999). Hospital mortality among nondiabetic subjects, however, did not change significantly between 1999 and 2001 (Fig. 1).

The outcome of the logistic regression analysis in diabetic patients (1999 and 2001) is presented in Table 3. It reveals that an increase in the number of therapeutic approaches in diabetic patients with acute MI (also when adjusted for age, sex, and other clinical variables) is associated with a reduction in mortality.

Blood glucose on admission of diabetic and nondiabetic patients was comparable between 1999 and 2001 (Table 1). In 2001, nearly one-half of diabetic patients received a glucose-insulin infusion at 1.6 ± 2.2 IU/h for 33 ± 13 h. This substantiated in a decrease of blood glucose levels from 234 ± 88 mg/dl (13.0 ± 4.9 mmol/l) to 152 ± 52 mg/dl (8.4 ± 2.9 mmol/l) (P < 0,001) within 12 h. Severe hypoglycemia did not occur in the course.

In 1999, lengths of hospitalization at the intensive care unit was 5.8 ± 5.5 days in diabetic patients and 5.5 ± 5.1 day in nondiabetic patients. Also, lengths of entire hospitalization were comparable between diabetic and nondiabetic patients: 14.1 ± 9.5 vs. 14.5 ± 9.0 days. In 2001, hospitalization at the intensive care unit was longer in diabetic patients (4.1 ± 2.4 days) than in nondiabetic patients (3.6 ± 2.5 days, P < 0.05), whereas the entire hospitalization was not significantly different (13.2 ± 9.0 vs. 12.1 ± 7.6 days). Differences between 1999 and 2001 were not statistically significant.

Total mortality in patients with and wihout previous coronary artery disease was as follows: 1999: diabetic patients 32 vs. 26%, nondiabetic patients 19 vs. 13%; 2001: diabetic patients 21 vs. 14%, nondiabetic patients 15 vs. 13%. Differences between diabetic and nondiabetic patients were not significanly different.

In the diabetic patients, previous antidiabetic medication (insulin and/or oral antidiabetic agents) did not influence the hospital outcome.

In both 1999 and 2001, heart rate on admission was higher in diabetic patients than in nondiabetic patients. Also, creatinine levels were increased and albuminuria was more frequently detected in diabetic patients (Table 1).

In 1999 and 2001, blood glucose on admission of the entire group of patients was higher in those who died during hospialization than in those who survived: 1999, 191 ± 95 mg/dl (10.6 ± 5.2 mmol/l) vs. 158 ± 85 mg/dl (8.8 ± 4.7 mmol/l) (P < 0.001); 2001, 209 ± 91 mg/dl (11.1 ± 5.0 mmol/l) vs. 163 ± 72 mg/dl (9.0 ± 4.0 mmol/l) (P < 0.001).

In 1999, recuscitation was performed more frequently in diabetic patients than in nondiabetic patients (24 vs. 11%, P < 0.01). In 2001, it decreased significantly to 9% (P < 0.01) in diabetic patients. The frequency of recuscitation in diabetic patients was comparable with that of nondiabetic patients. In 1999, catecholamines were infused more frequently in diabetic than in nondiabetic patients (29 vs. 12%, P < 0.01). In 2001, differences between the groups were not significantly different (12 vs. 14%).

In 1999, arterial hypertension was reported in 61% of diabetic and 44% nondiabetic patients (P < 0.01) compared with 57 and 46% in 2001 (P < 0.05). Frequencies of known coronary artery disease did not change significantly between 1999 and 2001 (diabetic patients 49% vs. 37 and 45% and 33% in nondiabetic patients).

Initially, aspirin and clopidogrel were recommended less frequently as discharge medication in diabetic than in nondiabetic patients. ACE inhibitors and diuretic agents were administered more frequently (Table 4). In 2001, the recommendation of aspirin, clopidogrel, and ACE inhibitors was enforced and levels were comparable to those of nondiabetic patients (Table 4). In nondiabetic patients, only the increase in recommendation of ACE inhibitors was significant.

This study demonstrates that the reenforcement of early therapeutic approaches (revascularization procedures and glucose-insulin infusion) in diabetic patients with acute MI is accompanied by a substantial reduction in hospital mortality. In 2001, the decrease resulted in a mortality rate of diabetic patients that was comparable with that of nondiabetic patients. It is interesting to note that frequencies of both diagnostic and therapeutic interventions were comparable between diabetic and nondiabetic patients. Furthermore, glucose-insulin infusion was successfully introduced as a new treatment strategy for diabetic patients.

The analysis of both years confirms the high prevalence of diabetes among patients with acute MI. Recently, 20.3% of patients with acute MI who were <80 years of age were reported to be diabetic patients (7). The Augsburg MONICA registry observed a prevalence of diabetes in 21% of patients with acute MI who were admitted to the hospital (4).

In 1999, early hospital mortality of diabetic patients was nearly threefold higher in diabetic patients than in nondiabetic patients. This substantial increase was reduced by 67% in 2001. An early hospital mortality rate of 4%, which was comparable to the rate of nondiabetic patients, was achieved. Since patients who died in the first 24 h were excluded from the Augsburg MONICA registry, data of early mortality are not available in the publication (4).

The study emphasizes the necessity to reduce hospital mortality in diabetic patients and highlights that the goal is achievable in the clinical setting. It demonstrates that early hospital mortality, which is pivotal for the future prognosis, can be successfully targeted. A more extensive application of established treatment has a potential to ameliorate the poor prognosis in diabetic patients with acute MI and is therefore supported.

The data of the Munich registry of 1999 reveals a reduced utilization of several diagnostic and therapeutic approaches in diabetic patients. This has recently been reported for the Swedish registry (RIKS-HIA [Register of Information and Knowledge about Swedish Heart Intensive Care Admission]), which included patients with MI who were admitted between 1995 and 1998 (7). The Augsburg MONICA registry, which investigated patients from 1985 to 1992, also observed an underutilization of treatment strategies (4).

As in RIKS-HIA, reperfusion strategies were initially performed less frequently in diabetic patients (7). In our registry, the gap of diagnostic procedures and therapeutic interventions between diabetic and nondiabetic patients was closed in 2001.

Admission plasma glucose levels have been previously shown to be a predictor of mortality of MI (9,14,15). Patients who died during hospitalization demonstrated higher blood glucose levels on admission than those who survived. Therefore, the importance of admission plasma glucose levels on hospital outcome in patients with MI is reemphasized.

It has been demonstrated that glucose metabolism is altered in a large proportion of patients with acute MI and no previous diagnosis of diabetes (16). Of patients, 35 and 40% presented with impaired glucose tolerance at discharge and after 3 months, respectively, and 31 and 25% had undiagnosed diabetes (16).

A limitation of the present analysis is that patients who qualified as nondiabetic patients might have had unknown diabetes. The study emphasizes the need for standardized oral glucose tolerance tests in the postinfarction period to detect previously unknown abnormalities of glucose metabolism. Also, larger studies are required to separately analyze the effects of therapeutic approaches in patients.

Glucose-insulin infusion as an adjunct to reperfusion after acute myocardial ischemia has become an effective strategy in the salvage of ischemic myocardium (17–19). Insulin has not only been suggested to act via modulation of cardiac and circulating metabolites but also to promote tolerance against ischemic cell death via activation of cell-survival pathways of the heart (18). There is evidence that early hospital mortality can be reduced by optimization of blood glucose control. The DIGAMI study, which assessed the effect of glucose-insulin infusion during 24 h after admission followed by an intensified therapy of diabetes for 3 months, reported a reduction of both early and long-term mortality in diabetic patients (9,12). Glucose-insulin infusion was applied to a large proportion of diabetic patients who were admitted with an acute MI in 2001. Since a reduction of mortality was mainly seen within the first 24 h after admission, a contribution of glucose-insulin infusion is suggested.

Autonomic incompetence plays a key role in the disturbed performance of the diabetic heart (1,20–22). Increased heart rate has also been suggested to be an indicator of cardiac incompetence (1). This observation also supports the use of β-blockers, which have been reported to be underused in diabetic patients with MI (23).

The present study demonstrates that a combination of multiple advanced treatment strategies in diabetic patients with acute MI enables a substantial reduction of hospital mortality in the patients. The analysis emphasizes the need for early intervention that targets both arterial occlusion and metabolic disturbances in an appropriate hospital setting. Resistance to early diagnostic and therapeutic interventions in diabetic patients should be abandoned.

We highly acknowledge the expertise and outstanding support of Gabriele Hölscher (Department of Medical Informatics, Biometry, and Epidemiology, Ludwig-Maximilians University, Munich) with regard to the statistical analysis of the data.