Assessment of Myocardial Scar; Comparison Between 18F-FDG PET, CMR and 99Tc-Sestamibi

OBJECTIVE Patients with heart failure and ischaemic heart disease may obtain benefit from revascularisation if viable dysfunctional myocardium is present. Such patients have an increased operative risk, so it is important to ensure that viability is correctly identified. In this study, we have compared the utility of 3 imaging modalities to detect myocardial scar. DESIGN Prospective, descriptive study. SETTING Tertiary cardiac centre. PATIENTS 35 patients (29 male, average age 70 years) with coronary artery disease and symptoms of heart failure (>NYHA class II). INTERVENTION Assessment of myocardial scar by (99)Tc-Sestamibi (MIBI), (18)F-flurodeoxyglucose (FDG) and cardiac magnetic resonance (CMR). OUTCOME MEASURE The presence or absence of scar using a 20-segment model. RESULTS More segments were identified as nonviable scar using MIBI than with FDG or CMR. FDG identified the least number of scar segments per patient (7.4 +/- 4.8 with MIBI vs. 4.9 +/- 4.2 with FDG vs. 5.8 +/- 5.0 with CMR, p = 0.0001 by ANOVA). The strongest agreement between modalities was in the anterior wall with the weakest agreement in the inferior wall. Overall, the agreement between modalities was moderate to good. CONCLUSION There is considerable variation amongst these 3 techniques in identifying scarred myocardium in patients with coronary disease and heart failure. MIBI and CMR identify more scar than FDG. We recommend that MIBI is not used as the sole imaging modality in patients undergoing assessment of myocardial viability.