The early predictors of ventricular remodeling after myocardial infarction: the role of tumor necrosis factor-alpha

• Published in: Anadolu kardiyoloji dergisi : AKD Vol. 9; no. 2; pp. 84 - 90
• Main Authors: Kaya, Ergün Bariş, Ozer, Necla, Deveci, Onur Sinan, Kepez, Alper, Tülümen, Erol, Aksöyek, Serdar, Atalar, Enver, Ovünç, Kenan, Ozmen, Ferhan, Ozkutlu, Hilmi
• Format: Journal Article
• Language: Turkish
• Published: Turkey 01-04-2009
• Subjects: Adult; Aged; Aged, 80 and over; Analysis of Variance; Case-Control Studies; Cohort Studies; Coronary Angiography; Echocardiography; Female; Humans; Logistic Models; Male; Middle Aged; Myocardial Infarction - pathology; Predictive Value of Tests; Prospective Studies; ROC Curve; Statistics, Nonparametric; Tumor Necrosis Factor-alpha - blood; Ventricular Remodeling
• ISSN: 1308-0032

Ventricular remodeling (VR) is a pathologic process characterized by progressive ventricular dilatation occurring after acute myocardial infarction (MI) leading to left ventricular systolic dysfunction. The purpose of the study was to evaluate the efficacy of plasma tumor necrosis factor alpha (TNF-alpha) levels to predict the left VR.
This prospective observational cohort study included 72 consecutive patients with newly diagnosed MI with age ranging between 38-87 years (mean 59 +/- 12 years). Control group was consisted of 30 patients with no additional systemic disease and normal coronary arteriograms. Transthoracic echocardiography was performed to all patients and controls both in the beginning of the study and in the 6th follow-up. A coronary arteriography was also performed to all patients. Patients with an increment in the diastolic volume index more than 20% in the follow-up compared with basal values included in the VR subgroup. The patient subgroup with VR consisted of 19 patients. Statistical analyses were performed using ANOVA and Kruskal Wallis tests for comparison of variables between groups. Logistic regression and ROC analyses were used for evaluation of accuracy of TNF-alpha in prediction of VR.
There were no significant differences between groups according to demographic characters. The basal plasma levels of TNF-alpha were higher in the patient subgroup with VR as compared with patients without VR and controls (14.59 +/- 4.28 pg/ml vs 7.30 +/- 4.48 pg/ml, and 1.64 +/- 1.49 pg/ml, p< 0.001). In logistic regression analysis only TNF-alpha predicted the VR (OR-1.356, 95% CI 1.117-1.647). Plasma TNF-alpha levels with a cut-off > or = 10.33 pg/ml were found to have 89.5% sensitivity and 79.3% specificity to predict the development of VR.
These results demonstrate the increment of plasma TNF-alpha levels in the acute phase of MI and the close relationship between the TNF-alpha levels and VR in the patients with first MI.