ABC | Volume 113, Nº4, October 2019

Original Article Abolhasani et al. Serum levels of markers in CAD prediction Arq Bras Cardiol. 2019; 113(4):667-674 were found in all subgroups compared to controls (p<0.001 for all of them). Moreover, there were significant differences among subgroups that were not mentioned (previously reported in other references). 11,20,21 The critical values of VN, MDA, OX-LDL, PAI-1, hs‑CRP and SA levels were determined by ROC curve analysis. Both specificity and sensitivity of these six parameters in CAD were compared. A combined assay was then performed, using six indexes. In addition, using both healthy individuals and CAD patients as subjects, ROC analysis was performed, which showed the areas under the curve for OX-LDL, MDA, PAI, VN, hs‑CRP and SA (0.870, 0.804, 0.951, 0.799, 0.962 and 0.971, respectively). All risk factors had satisfactory diagnostic efficacy for CAD. The overall rank of efficacy was (from lower to higher): VN, MDA, OX-LDL, PAI-1, SA and hs‑CRP. PAI-1, SA and hs‑CRP had particularly high validity in the diagnosis of CAD (Figure 1 and Table 3). Notably, the criteria of variables were specifiedwith reference to appearance levels in both healthy individuals and CAD patients, Cutoff for OX-LDL, MDA, PAI1-, VN, hs‑CRP and SAwere 2.67 (ug/mL), 5.49 (mmol/mL), 67 (ng/mL), 254 (ng/mL), 3.4 (mg/dL), 7/89 (mg/dL), respectively. The sensitivity and specificity were 70% and 75%, 74% and 77 %, 92% and 90%, 70% and 83%, 94% and 93%, 94% and 96%, respectively. PAI1-, SA and hs‑CRP had the highest sensitivity and specificity in the test, compared to OX-LDL, MDA and VN (Table 4). The efficacy of the combined assay was then compared using six parameters (OX-LDL, MDA, VN, PAI1-, SA and hs‑CRP). Such combined assay increased the predictive value of sensitivity and specificity to 99% and 99%, respectively (Table 5). The area under ROC curve was 0.99 (95% CI: 0.975~1.005, Figure 2). Discussion Commonly, major risk factors causing atherosclerotic lesions in human coronary arteries comprise genetic factors, hyperlipidemia, diabetes, infections, hypertension or oxidative stress, with little correlation with patients’ age and environmental factors. 22 It is noteworthy that this pathological process includes macrophages and smooth muscle cells (SMCs), with addition and deposition of lipids and extracellular matrix proteins, especially glycoprotein. 22,23 Previous studies confirmed that serum levels of VN, PAI-1, OX-LDL, MDA, hs‑CRP and SA are significantly higher in CAD patients compared with healthy controls and positive correlation with severe diseases. A climb in these analyses in patients with CAD when compared to controls is already well distinguished in recent studies. 11,20,21,24 The present study considerably assessed the prognostic value of glycoprotein, fibrinolysis, oxidative stress and inflammatory biomarkers including VN, PAI-1, and OX-LDL, MDA, hs-CRP and SA in patients with CAD. Remarkably, VN is a multifunctional plasma glycoprotein with a multiple binding domain, which regulates processes such as platelet adhesion, aggregation and clotting. Besides, VN can be expressed and produced in the vessel wall, predominantly in atherosclerotic lesions. 25 Later studies showed that PAI-1 stimulates VN expression in SMCs by binding LDL receptor‑related protein-1 (LRP1), and controls vascular VN expression in vivo. Therefore, autocrine regulation of vascular VN expression by PAI-1 may play important roles in vascular homeostasis and pathologic vascular remodeling. 26 Table 1 – General characteristics of the study groups Characteristic Control No Stenosis 1VD 2VD 3VD p value Sample size 20 40 40 40 40 - Age(mean(SD)) 57.5(3.2) 58.80(7.5) 58.9(7.9) 61.0(11.8) 60.5(10.5) 0.37* Sex(male/female) 17/3 21/19 31/9 30/10 25/15 0.30** Smoking (N (%)) 0(0%) 13(32.5) 19(47.5) 16(40) 24(60) 0.004** Hypertension (N (%)) 0(0%) 25(62.5) 26(65) 12(30) 19(47.5) 0.01** Diabetes (N (%)) 0(0%) 7(17.5%) 11(27.5%) 12(30%) 15(37.5%) 0.02** * One-way analysis of variance; ** Chi square test. Table 2 – Comparison of the mean serum levels of cardiovascular risk factors among subgroups categorized based on number of occluded vessels of CAD patients p Value 3VD 2VD 1VD No Stenosis Control Variable 0.001* 3.13 ± 0.42 2.76 ± 0.38 2.62 ± 0.27 2.28 ± 0.32 1.41 ± 0.22 OX-LDL (ug/mL) 1 0.001* 7.12 ± 1.21 6.39 ± 0.66 5.25 ± 0.98 5.20 ± 0.44 4.32 ± 0.86 MDA (mmol/mL) 1 0.001* 86.8 ± 6.8 76.9 ± 4.7 75 ± 14.2 51.5 ± 10.8 41.7 ± 11.9 PAI-1 (ng/mL) 1 0.001** 361 (95.75) 264 (100.75) 304 (184.25) 208 (61.75) 200 (26) VT (ng/mL) 2 0.001** 5.23 (1.05) 7.53 (0.86) 5.21 (0.39) 1.52 (1.03) 2.54 (0.78) hs‑CRP (mg/dL) 2 0.001* 169.9 ± 15.3 138.3 ± 12.3 108.6 ± 9.2 60 ± 11.6 51.0 ± 5.0 SA (mg/dL) 1 OX-LDL: oxidation of low-density lipoprotein; MDA: Malondialdehyde; PAI-1: Plasminogen Activator Inhibitor; VT: Vitronectin; hs‑CRP: high-sensitivity C-reactive protein; SA: sialic acid; 1VD: Stenosis in one of vessels; 2VD: Stenosis in two of vessels; 3VD: Stenosis in three of vessel. 1 Mean ± standard deviation; 2 Median (inter quartile range). * Performed by ANOVA test. ** Performed by Kruskal-Wallis test. 669