Combined Assessments of Multi-panel Biomarkers for Diagnostic Performance in Coronary Artery Disease: Case-Control Analysis

Abstract

Amir Al-Mumin, Hayder Abdul-Amir Makki Al-Hindy, Mazin Jaafar Mousa

Background: Understanding the etiopathology of coronary artery diseases (CAD) has directed studies towards the evaluation of novel serum biomarkers as potential diagnostic tools for the clinical setting. In addition to the current gold standard cardiac troponin-I (TnI), several other biomarkers had been studied for diagnosis, prediction, and prognosis of CAD. Of these biomarkers, N-terminal-proatrial natriuretic peptide (NT-ProANP), Cystatin-C (CyC), and highly-sensitive C-reactive protein (HSCRP).
Aim of Study: Evaluation of the diagnostic performance of combined multiple biomarkers in patients with CAD
Subjects and methods: the study included 136 patients diagnosed as CAD by cardiologists and 44 healthy group. All participants underwent echocardiographic examinations and LVEF % < 40 selected as a cutoff value for expressing LV-systolic dysfunction. Hence, CAD subjects were classified into two subgroups: LVEF ≤ 40% versus > 40%. Moreover, hematological tests for creatinine, urea, total lipid profile, TnI, NT-ProANP, CyC, and HSCRP within the first 24 h of admission were done. An angiographic study was completed by expert interventional cardiologists. Biostatistical scrutiny was finalized with SPSS Version-25, and GraphPad Prism 8.0.2. ROC study performed for all four markers to evaluate their, diagnostic accuracy for CAD. Pairwise assessment of ROC for the analytic ability of the multi-biomarkers was also performed using MedCalc 19.3.1 software.
Results: There were no significant differences regarding serum creatinine, urea nitrogen, and BMI between two study groups. The incidence of diabetes, hypertension, lipid profile components, and smoking were significantly higher in CAD patients. There was a significant difference between patients and control groups concerning mean LVEF%, however, only 1/4th of the CAD patients had an echocardiographic finding of LVSD. There were significantly higher concentrations of all four markers in CAD than in controls. ROC curves revealed that AUCs of TnI, NT-ProANP, CyC and HSCRP were, respectively, 0.93 (p-0.001), 0.69 (p-0.001), 0.82 (p-0.4) and 0.69 (p-0.001); the highest was for TnI. There was no statistically significant difference between the diagnostic performance power (differences in AUC) of NT-ProANP and HSCRP. Nevertheless, TnI has the highest AUC followed by CyC. Pairwise comparison of the combined biomarkers of TnI and HSCRP appeared to perform best in predicting CAD (P-0.0001). The next best performing couple is the addition of TnI and NT-ProANP (P-0.001). Using HSCRP with TnI and NT-ProANP in a three-markers panel, or a four-markers panel with CyC yielded no further diagnostic value (results not shown). Combining either CyC or HSCRP and NT-ProANP, provided no extra information in the diagnosis of CAD than it would if either were studied uncoupled