НЕЗАВИСИМЫЕ ПРЕДИКТОРЫ ГЕМОДИНАМИИЧЕСКИ ОСЛОЖНЕННОГО ТЕЧЕНИЯ РАННЕГО ПОСЛЕОПЕРАЦИОННОГО ПЕРИОДА У ПАЦИЕНТОВ С ИБС: РОЛЬ СОВРЕМЕННЫХ БИОМАРКЕРОВ

Abstract

Cardiac biomarkers are recommended to stratify the risk of surgery in general surgical practice, but their adoption in cardiac surgery patients is limited. The aim of our study is assessing the prognostic possibility and predictive significance of modern cardiac-specific biomarkers (NT-proBNP, sST-2, hsTn-I, Galectin-3, hsCRP and IL-6) in the determining of patient-oriented treatment strategies. In the open prospective cohort study 352 patients with moderate or severe functional IMR were included. The inclusion criteria in the study were the reduced myocardial contractility (LV EF < 40 %) in chronic CAD patients. CABG combined with mitral valve repair was performed in 239 patients (67.9 %), mitral valve replacement in 35 patients (9.9 %), and isolated in 78 patients (22.2 %). The primary end-point of the study was defined as a complicated postoperative period with worsening of heart failure. Complicated postoperative period has been registered in 80 patients (22.7% of cases). The complicated early postoperative period did not relate to surgical strategies (χ2 = 0.398, p = 0.528). We didn’t find any difference in the degree of LV remodeling between patients with complicated or not postoperative course (mean LV EDD - 68.2 mm, LV ESD - 56.3 mm, iEDV - 118.9 ml/m2 and iESV - 81.2 ml/m2, p > 0.05). By contrast, all of the used pre-operative biomarker tests differed in the examined groups of patients (p < 0,001). The complicated postoperative course was observed in 24.6 % of cases with isolated elevation of NT-proBNP > 136 pg/ml, in 44.4 % of cases with isolated elevation of sST2 > 35.8 ng/ml, and 59.3 % of cases in combined increase in thresholds and sST2 and NT-proBNP (χ2 = 71.67, p = 0.001). The highest quality of the model and the estimate forecast of the postoperative low-output syndrome was measured for sST2 and NT-proBNP biomarkers (AUC = 0,792, 95 % CI 0,709–0,846). The inclusion of both thresholds of sST2 and NT-proBNP retained their reliable influence on the forecast, with the model showed a high level of consent (79.7 % of correctly recognized cases).