Preoperative magnetic resonance angiography as a predictive test for cerebral ischemia during carotid endarterectomy.

Background: To evaluate whether the preoperative magnetic resonance angiography (MRA) can predict the risk of cerebral ischemia associated with the carotid endarterectomy (CEA).

Methods: Between January 2004 and December 2010, 382 consecutive patients (mean age: 56.6 years; range: 45-78 years) were identified to have undergone preoperative MRA and the CEA under regional anesthesia. It was determined that the patient needs shunting during the CEA by intraoperative monitoring of patient's neurology. All patients were divided into two subgroups: shunt group or no-shunt group. Imaging findings on preoperative MRA were correlated to shunting using univariate and multivariate logistic regression analyses combined with patient's demographic and clinical features to identify predictors of cerebral ischemia during the CEA.

Results: In 37 of 382 CEA cases (9.7%), shunting had been performed intraoperatively because the patient had a neurologic deficit. At multivariate analysis, preoperative MRA findings such as the absence of patent communicating arteries (odds ratio [OR], 5.56; 95% confidence interval [CI], 3.05-9.69; p = 0.013) and the increase of intracranial arteries which were not patent in the contralateral hemisphere (OR, 4.277; 95% CI, 2.575 to 7.104; p < 0.0001) were significantly associated with shunting.

Conclusions: Preoperative MRA is valuable when predicting cerebral ischemia leading to an inevitable shunting during CEA. Therefore, if there are preoperative MRA findings such as multiple occlusive intracranial arteries in the contralateral hemisphere or the absence of patent communicating arteries, it is recommended that CEA be performed under general anesthesia with routine shunting to avoid a serious shunt-related complication.