Radiologic assessment of tumor and the carotid artery: correlation of magnetic resonance imaging, ultrasound, and computed tomography with surgical findings.

Carotid artery fixation is an extremely poor prognostic sign in squamous cell carcinoma of the head and neck. Methods more precise than physical exam are necessary to determine whether a neck dissection is sufficient, to exclude patients from surgery, or to plan a carotid artery resection. Thirty-three patients with a suspicion of carotid wall attachment by physical examination or computed tomography (CT) were further evaluated by magnetic resonance imaging (MRI) and ultrasound (US). The criterion for possible carotid involvement on CT was effacement of greater than 25% of the circumference of the artery. On MRI, the criterion that we proposed was any loss of the fascial plane around the internal or common carotid artery. This was considered effacement of the carotid artery wall. For US, the criterion we used was the loss of the echogeneity along the carotid wall. Histopathological correlation was available in 23 patients (24 necks) who either underwent surgery or necropsy (1 patient); 11 patients were felt not to be surgical candidates or declined surgery, though 1 of these 11 patients did undergo surgical resection after radiotherapy. MRI accurately predicted an uninvolved carotid artery in 13 of 15 cases; US accurately predicted an uninvolved carotid artery in 15 of 20 cases. Demonstration by either MRI or US that the carotid artery is free of tumor had a high degree of reliability. Both US and MRI appear superior to CT and physical examination in assessing the carotid wall in suspicious cases. Individually and in combination, MRI and US offer advantages in selecting patients for possible surgery when the issue of carotid artery fixation is raised.