Diabetes mellitus and coronary artery surgery : clinical and epidemiological studies

Abstract

Objectives: Assess early and late mortality and incidence of acute myocardial infarction (AMI) after coronary artery bypass grafting (CABG) in patients with and without diabetes mellitus (DM) in relation to type of treatment. Analyse mortality after CABG in patients with and without DM to examine if any difference was influenced by changes in prognosis related to time-period. Measure glycosylated haemoglobin 1 (HbA1c) before CABG to determine correlation to postoperative outcome. Determine if the use of thoracic epidural analgesia (TEA) during and after CABG reduced insulin requirements and hyperglycaemia in patients with and without DM. Seek novel markers for morbidity and hospital stay after CABG by using gene expression techniques.

Methods and Results: The risk of early mortality (≤30 days) was increased in patients with insulin-treatment (odds ratio [OR] 4.6, 95% Confidence Interval [CI], 2.5-8.4) and in those on oral antidiabetic drugs (OR 2.0, 95% CI 1.0-3.8), but not in diet treated patients compared with patients without DM among 6727 patients who had CABG during 1980-1995. At 10 years the relative risk of death or AMI was 1.8 (95% CI 1.5-2.2) in insulin-treated patients and 1.4 (95% CI 1.2-1.7) in patients on oral drugs but there was no increased risk in diet treated patients compared with patients without DM. Survival at 10 years without AMI was 40% in patients with insulin-treatment, 48% if on oral drugs, 59% if diet managed, compared with 66% in patients without DM.

Early mortality was 3.4% in patients with DM versus 1.8% in patients without DM (OR 2.0, 95% CI 1.4-2.7) among 12,557 patients who had CABG during 1970-2003. Early mortality was reduced in patients operated on 2000-03 compared with 1970-89 in patients with DM (OR 0.3, 95% CI 0.1-0.9) and in those without DM (OR 0.4, 95% CI 0.2-0.7). Five-year mortality was 14.6% in patients with DM versus 8.3% in those without DM (hazard ratio 1.8, 95% CI 1.5-2.0). Five-year mortality was reduced 40% in patients operated on 2000-03 compared to 1970-89 in patients with and without DM.

Superficial sternal wound infection after CABG occurred in 13,9% of patients with preoperative HbA1c level ≥6% versus in 5,5% if HbA1c <6% (p=0.007). Mortality at an average of 3.5 years after CABG was 18.9% in patients with HbA1c ≥6% compared with 4.1% if HbA1c <6% (p<0.001, hazard ratio 5.4, 95% CI 3.0-10.0).TEA was used during and three days after CABG in half of 44 patients with DM and 60 without DM. TEA reduced mean blood glucose (BG) and insulin requirements (p<0.02) during the initial 24 hours in patients without DM whereas in patients with DM mean BG level was reduced (p=0.017) with unchanged insulin requirements. TEA did not attenuate hyperglycaemia during the first three postoperative days or diminish the increased fasting BG on the third postoperative day in patients without DM. Metabolic gene expression profiles were analysed in biopsies obtained during CABG in 66 patients. Patients with DM and not diagnosed DM had prolonged hospitalization time. Levels of the anti-inflammatory gene dual-specificity phosphatase 1 (DUSP1) in skeletal muscle differed in patients with normal (≤8 days) versus long hospitalization (>8 days, p=0.003).

Conclusions: DM was associated with an increased risk of early and late mortality. Early and late mortality was reduced in patients with and without DM operated on more recently but the mortality disadvantage associated with DM was not eliminated. HbA1c level ≥6% was associated with increased risk ofwound infection and higher mortality at three years after CABG. TEA improved glucose homeostasis minimally during the initial 24 postoperative hours but did not attenuate hyperglycaemia during subsequent three postoperative days. Levels of DUSP1 expression predicted hospitalization time and may be of use to predict outcome after CABG.