Abstract
The aim of this study was to explore the possible association between markers of inflammation and oxidative stress (OS) and markers of cardiac function and necrosis in 100 NSTEMI (non-ST-elevation myocardial infarction) patients with various degrees of kidney dysfunction. At admission, ejection fraction (EF), brain natriuretic peptide (BNP), troponin (TnI), creatinine phosphokinase (CPK), alanine transaminase (ALT), aspartate transaminase (AST), high-sensitive C-reactive protein (hs-CRP), interleukins 6 and 10 (IL-6, IL10), myeloperoxidase (MPO), transforming growth factor beta (TGF-β1), glomerular filtration rate (GFR), and albuminuria were assessed. Study participants were divided into 2 subgroups based on the median level of EF. Compared to the high, patients in the low EF group had higher GFR, BNP, CPK, hs-CRP, IL-10, IL-6, and MPO values and lower albuminuria levels. The levels of EF decreased in parallel with the progression of CKD, whereas the levels of BNP, IL-6, and TGF-β were significantly higher in late stages of CKD. Spearman's rho correlation analysis showed that EF was inversely correlated with MPO (r = -0.20, p = 0.05) BNP (r = -0.30, p = 0.002), hs-CRP (r = -0.38, p < 0.0001), IL-10 (r = -0.30, p = 0.003), and IL-6 (r = -0.24, p = 0.02) and positively with GFR (r = 0.27, p = 0.008). TnI was correlated with CPK (r = 0.44, p < 0.0001), CPK-MB (r = 0.31, p = 0.002), ALT (r = 0.50, p < 0.0001), AST (r = 0.29, p = 0.004), IL-10 (r = 0.22, p = 0.03), and MPO (r = -0.28, p = 0.006). In multivariate regression analysis, only BNP (β = -0.011, p = 0.004), hs-CRP (β = -0.11, p = 0.001), and GFR (β = 0.12, p = 0.0029) were independent determinants of EF. Similarly, MPO (β = -1.69, p = 0.02), IL-10 (β = 0.15, p = 0.006), and AST (β = 0.04, p = 0.001) were the 3 major determinants of TnI. Based on these associations, we built a predictive model including markers of inflammation and OS (MPO, IL-10, and hs-CRP) to identify patients with the most severe cardiac injury (combined EF below median and troponin above median values). Receiver-operator characteristic (ROC) analysis showed that the area under the ROC curve of this model to detect patients with low EF and high TnI was 0.67 (p = 0.015, 95%confidence interval = 0.53-0.81).