Abstract 200: The Utility of Cardiac Magnetic Resonance Imaging in Detecting Myocardial Involvement in Connective Tissue Disease
Background: Myocardial involvement can be present in all types of connective tissue disease. Immune complex deposition and mononuclear infiltration of the microvasculature lead to capillary leakage, edema and later fibrosis of the myocardium. These alterations are reflected by changes in T1 global Relative Enhancement (gRE), T2-Edema-Ratio (ER) and Late Gadolinium Enhancement (LGE) on cardiac MRI (CMR). This study examines the different CMR techniques for the diagnosis of myocardial involvement.
Methods: 28 patients with known connective tissue disease and symptoms suggestive of myocardial involvement (dyspnea, chest pain or arrhythmia) underwent CMR. Myocardial involvement was assumed if at least two of the following criteria were present: troponin-, BNP- or CK- elevations, ST-T-segment alterations, conduction disturbances or extrasystole. T2 measurements were made using turbo spin echo (TSE) sequences in three short axis orientations. T1 measurements were performed in the same orientations pre and post contrast injection of 0.2 mmol gadodiamide per kg body weight. ER was computed as ratio of mean myocardial signal and skeletal reference signal. gRE was computed as ratio of absolute enhancement of myocardium and skeletal muscle. LGE was deemed either present or absent 10 to 15 minutes post contrast. ER was assumed positive, if > = 1.9, gRE was assumed positive if > = 4.0. Sensitivity and specificty of were computed from SPSS cross tabs procedure. Continuous variables were compared using ANOVA and are presented as mean±SD.
Results: High likelihood of myocardial involvement was assumed in 20 of the 28 patients. The sensitivity, specificity and diagnostic accuracy of the CMR parameters as compared to the clinical likelihood were respectively 83%, 80% and 83% for gRE; 63%, 54%, 50% for ER and 82%, 94% and 90% for LGE. Mean T1 values were 4.9±2.9 for patients with high clinical likelihood vs. 2.9±0.9 for patients with low clinical likelihood, p = 0.18. Mean T2 values were 2.0±0.5 for patients with high clinical likelihood vs. 1.8±0.67 for patients with low clinical likelihood, p = 0.62.
Conclusion: CMR is well suited for detection of myocardial involvement in connective tissue disease. LGE and gRE are most accurate measures of myocardial involvement.