Diagnosis of left ventricular thrombi by magnetic resonance imaging and comparison with angiocardiography, computed tomography and echocardiography

Abstract

Left ventricular (LV) thrombi are common in patients with acute and healed anterior myocardial infarction. Although the reported incidence of clinically diagnosed embolism varies widely, 1–3 diagnosis of thrombus seems to carry an increased risk for this potentially life-threatening complication. It is not entirely clear whether long-term anticoagulation can reduce the embolism rate in patients with chronic aneurysms with an acceptable risk of bleeding. However, it is widely assumed that this is so; thus, the diagnosis of LV thrombus has far-reaching consequences. It is difficult to diagnose LV thrombus using currently available imaging techniques. LV angiography carries the risk of precipitating systemic emboli and has a low sensitivity in most reported studies. 4 Echocardiography has been reported to be fraught with false-positive results. 5,6 Indium-111 platelet scintigraphy reflects thrombus activity but has suboptimal sensitivity, especially for small thrombi. 7 More recently, computed tomography (CT) 8 and magnetic resonance imaging (MRI) 9,10 have shown encouraging results in the detection of LV thrombi. However, the diagnostic accuracy of MRI has not been evaluated with independent confirmation by surgery or autopsy. This study describes the sensitivity and specificity of MRI in the diagnosis of LV thrombi in patients with LV aneurysms undergoing surgical aneurysmectomy, and compares the results with those of angiography, 2-dimensional echocardiography and CT performed in the same patients.