Clinical Electrophysiology -> SCA Risk Assessment: -> Clinical Risk Assessment Techniques D-PO04 - Poster Session IV (ID 15) Poster

D-PO04-209 - Electrocardiogram Measures And Cardiac Troponin T Predict Replacement Fibrosis On Cardiac Magnetic Resonance Imaging In Middle-aged And Older Adults: The Multi-ethnic Study Of Atherosclerosis (ID 1236)


Background: Abnormal myocardial substrate detected on cardiac MRI portends adverse prognosis
Objective: Whether 12-lead ECG and biomarkers together predict abnormal substrate detected by CMR in subjects free of known CVD
Methods: We performed a cross-sectional analysis of subjects in Multi-Ethnic Study of Atherosclerosis (MESA) without CVD and with measurements of cMRI, biomarker, and ECG. We used multiple linear regression to estimate associations of each ECG metric in addition to biomarker (highest tertile), with abnormal substrate(highest quartile ECV via T1 mapping[diffuse fibrosis] and/or detectable LGE[focal fibrosis]), adjusting for demographic and clinical risk factors (RFs). We also analyzed whether a biomarker (highest tertile) when added to each ECG metric improved prediction (using c-statistic) of diffuse or focal fibrosis
Results: There were 1170 participants (50% women;51% white;mean age 67.1±8.6 yrs) eligible for analysis. Among ECG measures, QRSd [OR 1.7(1.4, 2.1)] and abnormal frontal QRS-Ta [OR 7.4(3.8, 14.1)] were strongly associated with focal fibrosis and abnormal QRS-Ta [OR 3.27(1.9, 5.6)] with diffuse fibrosis, even after adjustment for clinical RFs. When QRSd and abnormal QRS-Ta were added to a model containing RFs and hscTnT, the ECG measures improved prediction of focal (c-statistic 0.827) but not diffuse fibrosis
Conclusion: QRSd and abnormal frontal QRS-Ta, when added to a prediction model considering RFs and evidence of myocardial injury, improve prediction of focal fibrosis in older middle-aged subjects without CVD. These findings may help refine selection of adults who harbor myocardial scars in the general population and merit a larger study.