The efficacy of selective shunting during carotid endarterectomy (CEA) using intraoperative monitoring (IOM) for detection of cerebral ischemia is well established. There is mounting evidence that monitoring of both electroencephalography (EEG) and somatosensory evoked potentials (SSEPs) increases the sensitivity of cerebral ischemia detection. Predictors of cerebral ischemia requiring selective shunt placement using IOM of both EEG and SSEPs have not been previously identified.Methods
Consecutive patients who underwent CEA between January 1, 2000, and December 31, 2010, were retrospectively analyzed. Primary end points were IOM changes at any time during the operation or IOM changes with carotid cross-clamping. Risk factors assessed included demographics; baseline comorbidities; severity of ipsilateral and contralateral disease; symptomatic status; and use of statin, antiplatelet, and beta-blocker medications. Univariate and multivariate logistic regression was used for analysis.Results
During the 11-year study period, a total of 758 patients underwent 804 CEAs (mean age, 70.6 ± 9.5 years; 59.8% male; 39.2% symptomatic) using IOM of both SSEPs and EEG for selective shunting guidance. Postoperative stroke rate was 1.37%; 27.1% of patients had significant SSEP or EEG changes, and 49.1% of these were clamp induced (within 5 minutes of cross-clamping). Of these patients, 83.2% received a shunt (11.4% overall). The most common reason that a shunt was not placed after cross-clamp-induced changes was that the changes resolved with further blood pressure elevation (8 of 17 patients). Clamp-induced IOM changes were predictive of postoperative stroke (odds ratio [OR], 5.5; P = .005). Risk factors for clamp-induced IOM changes were contralateral carotid occlusion (OR, 2.5; P = .01), symptomatic stenosis (OR, 1.8; P = .006), and diabetes (OR, 1.6; P = .03), whereas there was a trend toward increased risk with female sex (OR, 1.5; P = .08). Risk factors for any IOM change (clamp and nonclamp induced) were symptomatic carotid stenosis (OR, 1.8; P < .001), use of beta blockers (OR, 1.5; P = .03), and female sex (OR, 1.5; P = .02).Conclusions
Whereas some patients can be expected to experience IOM changes by monitoring of SSEPs and EEG, a much smaller percentage will receive a shunt. Contralateral carotid occlusion, symptomatic stenosis, diabetes, and female sex increase the risk of clamp-induced IOM changes and should be anticipated to need a shunt. Patients receiving beta blockers are likely to experience IOM changes during the operation that are not associated with clamping.