Magnetocardiography (MCG) has been used to risk stratify patients in terms of sudden death or to detect ischemia. We evaluated the potential of this technique to assess myocardial viability in coronary artery disease.
Fifteen patients aged 36–75 (median, 59) years with stable single-vessel disease (≥70% diameter stenosis) and corresponding regional wall-motion abnormality underwent (1) echocardiography to evaluate wall motion, (2) 201Tl dipyridamole single-photon emission computed tomography to document perfusion and (3) quantitative 18F-fluorodeoxyglucose positron emission tomography to assess viability in 16 left-ventricular wall segments. MCG was performed in each patient using a shielded prototype 49-channel low-temperature superconducting quantum interference device (SQUID) system. Multiple time and area parameters were extracted automatically from each baseline-corrected data set.
Eleven patients had prior myocardial infarction. In each patient, four to 12 (median, seven) segments were lesion dependent, totalling up to 117 out of 240 segments. A total of 88 segments (75%) were viable and 29 segments (25%) represented scar. Patients were divided into three categories: (a) no scar segments (five patients), (b) scar in one to three segments (six patients) and (c) scar in ≥ four segments (four patients). The three MCG parameters with the best selectivity were identified using linear discriminant analysis with forward inclusion (P<0.10). The corresponding Fisher's discriminant functions classified all patients correctly (Wilks' λ=0.079).
Selected MCG parameters yielded accurate patient classification with regard to the extension of myocardial scar within the viable tissue in retrospect. These findings indicate that MCG may contribute to the assessment of myocardial viability. Further evaluation in a comprehensive multicenter study is warranted.