In-hospital peak glycemia in non-diabetic patients with heart failure complicating acute coronary syndrome

Abstract

Acute heart failure syndromes (AHF) remain a major cause of morbidity andmortality, in part because of the lack of definitive evidence for AHFmanagement and risk stratification parameters [1], and there is a clinical need to identify clinical and/or laboratory tools useful for risk stratification in the AHF population [2]. The relation between glucose values and insulin resistance and mortality, both at short and long term [3–5], is still poorly understood and previous studies were based only on admission glucose level [6–10], while data on in-hospital glucose values are still lacking. The present investigation was aimed at assessing the prognostic role of admission and peak in-hospital glycemia in 273 consecutive patients with AHF complicating acute coronary syndrome (ACS) and without previously known diabetes, admitted to our Intensive Cardiac Care Unit (ICCU) from June 30, 2009 to December 31, 2012. Even if in all cases datawere collected prospectively, the current study is a retrospective analysis of those data. The study population includes 141 patients with ST-elevation myocardial infarction (STEMI) (51.6%) and 132 (48.4%) patients with Non ST Elevation myocardial infarction (NSTEMI). All patients were submitted to coronary angiography. STEMI patients were admitted to our ICCU after mechanical revascularization. Among NSTEMI patients, 24 (18%) patients underwent urgent coronary artery bypass graft, while percutaneous coronary interventionwas performed in 65 NSTEMI patients (49%) and medical therapy was administered to the remaining 43 NSTEMI patients (33%). The presence of comorbidities was determined by taking the patients' history directly [11]. On ICCU admission, in a fasting blood sample, the following parameters were measured: glucose (g/l), insulin (UI/L), troponin I (Tn I, ng/ml), uric acid (mg/dl), glycated hemoglobin (%), NT-pro Brain Natriuretic Peptide (NT-proBNP) (pg/ml), leukocytes count (*10/μl), fibrinogen (mg/dl) and lactate (mmol/l). Creatinine (mg/dl) was also measured in order to calculate glomerular filtration rate (ml/min/1.73 m) [12]. In all patients, glucose levels and Tn I were measured three times a day, and peak glycemia [13,14] and peak Tn I were considered, respectively. Acute insulin resistancewas investigated bymeans of thehomeostaticmodel assessment index (HOMA index), as previously described [15–17]. Intensive insulin therapy was administered in patients with significant hyperglycemia (that is plasma glucose N180 g/l) [18]. The study population was divided according to tertiles of peak glycemia. Outcome was in-ICCU mortality. The study protocol was in accordance with the Declaration of Helsinki and approved by the local Ethics Committee. Informed consent was obtained in all patients before enrolment. Categorical values have been reported as frequency (percentage); continuous variables have been reported as mean± SD or median (interquartile range, IR) as appropriate. Between-group comparisons were made by means of chi-square and ANOVA or Kruskal–Wallis test as needed. Logistic regression analysis, both uniand multivariable, were carried out considering as outcome intra-ICCU mortality. Variables were chosen as those clinically related to in-ICCU death. A receiver operating characteristic (ROC) curve was plotted to identify a cutoff of peak glycemia with respect to in-ICCU death. A p value less than 5% has been taken as statistically significant (IBM-SPSS 20.0 for Windows® statistical software (IBM-SPSS Inc., USA)). Clinical characteristics of the study population are depicted in Table 1. Patients in the third tertile showed the highest percentage of STEMI, thehighest incidenceof COPDand the lowest valuesof admission and discharge LVEF. The highest in-ICCUmortality rate was observed in patients in the third tertile. Patients with the highest values of peak glycemia showed the highest admission glycemia levels and the highest incidence of HOMA-index and HOMA positivity. In the third subgroup, it was observed a progressive increase in leukocytes count, in NT-pro BNP levels and in peak Tn I. Devices and inotropes were more frequently used in patients in the third tertiles (Table 2). A cutoff of 1.91 g/l peak glycemia was identified with C-statistic as significantly related to an in-ICCU higher mortality rate: AUC, 0.81 (95% CI = 0.74 to 0.87), p b 0.001; sensitivity, 75% (95% CI = 69% to 81%); specificity, 76% (95% CI = 59% to 87%); predictive value of positive test,