Free-floating thrombus of the carotid artery detected on carotid ultrasound in patients with cerebral infarcts: a 10-year study.

ultrasound for diagnosis of thrombus ranged from 0.05% (1/2000)1 to 0.9% (14/1528).2 The close association with atherosclerosis was evident by the high prevalence of risk factors, male preponderance and carotid plaques. All patients (except 1) had one or more risk factors. Severe carotid stenosis (80% to 99%) was found in 2 patients (2 and 11). In contrast, Buchan et al4 found moderate carotid stenosis (50% to 90%) in 5 patients (17%), and severe stenosis (>90%) in 21 patients (70%). This refl ects a higher incidence of extracranial carotid disease in Caucasians compared to Asians. Two patients (5 and 20) had carotid dissection.6 Both did not give a history of trauma. Magnetic resonance imaging (MRI) brain for patient 5 showed an embolic distribution, suggesting the pathophysiology was artery-to-artery embolism. For patient 20, the watershed pattern of infarcts suggests that the mechanism was haemodynamic insuffi ciency. The association of malignancy with thrombus formation is wellknown.5,7 Patients 2 and 23 had a history of malignancy but no tumour recurrence at the time of stroke. Among patients without risk factors, there was a young man with low antithrombin III. This rare condition predisposes to thrombosis in patients with additional risk factors.8 Complete resolution of thrombus occurred in 58% of patients treated with anticoagulation. This compared with a thrombus resolution rate of 86% amongst anticoagulated patients in a systematic by Bhatti et al.3 This highlights the need for vigilant international normalised ratio (INR) monitoring.3-5,7,9 Surgical embolectomy1-5,7 and new intervention techniques10 should be considered when patients cannot be adequately anticoagulated or when they develop complications of anticoagulation. This study is limited by a lack of angiographic or histological correlation as all patients received medical therapy and none were surgically treated. In our study, 8% (n = 2) had recurrent stroke and 24% (n = 6) developed ischaemic heart disease. However, it is diffi cult to interpret the relationship between carotid artery occlusion and subsequent strokes as the time interval between strokes could be quite long and the patients also had risk factors which predisposed them to stroke. With regard to ischaemic heart disease, the time interval is shorter (4 months to 6 years), suggesting an increased risk of cardiovascular events. This is not surprising as both stroke and heart disease are different Free-fl oating Carotid Artery Thrombus—Hoe Chin Chua et al Letter to the Editor