Carotid Doppler high-intensity transient signals in dilated cardiomyopathy

BACKGROUND: Thromboembolism is an infrequent but serious complication in dilated cardiomyopathy (DCM), and the role of primary preventive antithrombotic or anticoagulation therapy is undetermined. High-intensity transient signals (HITS) by Doppler ultrasound representing microemboli have been described in various clinical settings associated with increased risk of stroke. This study assessed the feasibility, reproducibility, and prevalence of HITS in patients with DCM. METHODS: Thirty patients with severely reduced left ventricular ejection fraction (< or = 35%, mean 25%) and New York Heart Association class II to III who were not receiving antithrombotic or anticoagulant therapy and 20 age-matched normal subjects were prospectively examined. Patients with atrial fibrillation, significant cardiac valvular heart disease, a history of cerebrovascular disease, and those who otherwise required antithrombotic or anticoagulation therapy were excluded. One-hour pulsed-wave Doppler recordings over the common carotid artery (CCA) were performed on 3 separate days in each subject by a single, experienced, blinded sonographer with a 4-MHz probe (TC-2000, Nicolet/EME) with a specially designed probe holder. Studies were read in a blinded, random fashion by 2 independent, experienced HITS Doppler recording readers. RESULTS: HITS in the CCA were detected in 6 (20%) of 30 patients with DCM and in 3 (15%) of 20 volunteers. This difference was not statistically significant. Intrareader and interreader reproducibility were high (kappa = 0.91 and 0.84, respectively; P <.001), whereas intrasubject reproducibility over the 3 visits was moderate to low (kappa = 0.22). There was no significant difference between HITS characteristics, that is, intensity and duration, in patients versus controls. CONCLUSIONS: The prevalence of CCA HITS in patients with clinically stable heart failure who are not receiving anticoagulation/antithrombotic therapy and are not in atrial fibrillation is low and not significantly different from normal patients. These data suggest that HITS monitoring is not a viable surrogate marker for increased thromboembolic risk in such patients with DCM.