Outcomes of urgent carotid endarterectomy for stable and unstable acute neurologic deficits.

OBJECTIVE: The aim of the study was to assess the outcomes of carotid endarterectomy (CEA) performed in an urgent setting on acutely symptomatic patients selected through a very simple protocol. METHODS: From January 2002 to January 2012, 193 symptomatic patients underwent CEA. Of these, 90 presented with acute symptoms, and after a congruous carotid stenosis was identified, underwent urgent operations (group 1): 27 patients had transient ischemic attack (group 1A), 52 patients had mild to moderate stroke (group 1B), and 11 patients had stroke in evolution (group 1C). The remaining 103 patients with a nonrecent neurologic deficit were treated by elective surgery in the same period (group 2). End points were 30-day neurologic morbidity and mortality. RESULTS: The median delay of urgent CEA (U-CEA) from deficit onset was 48 hours (interquartile range, 13-117 hours). Groups 1 and 2 were comparable in demographics. Acute patients showed a higher rate of stroke at presentation (70% vs 37%; P = .001) and of history of coronary artery disease (30% vs 13.5%; P = .007). Acute patients sustained six postoperative strokes (6.6%). Neurologic outcomes were correlated to clinical presentation: no strokes occurred in group 1A patients, and 5.8% group 1B patients and 27.3% group 1C patients had postoperative stroke (P < .01). Postoperative mortality was 4.4% for U-CEA: one fatal myocardial infarction, one intracranial hemorrhage, and two thromboembolic strokes. Elective patients sustained four postoperative strokes (3.9%), with one death (0.9%) as a consequence of hyperperfusion cerebral edema. U-CEAs performed ≤48 hours from symptom onset had a lower postoperative stroke rate than those performed >48 hours (4.4% vs 8.8%; P = .3). Among patients presenting with a stroke (group 1B), the National Institutes of Health Stroke Scale (NIHSS) assessment at discharge showed improvement in 79% (although only 25% had ≥4 points in reduction), stability in 17%, and deterioration in 4%. Patients with moderate stroke were slightly better in NIHSS improvement than those with mild stroke (median NIHSS variation at discharge, -3 vs -1; P = .001). CONCLUSIONS: Our results with U-CEA confirm that this population has a higher risk profile compared with elective surgery. The type of acute presentation is correlated with perioperative risk. U-CEA was safe when performed on patients presenting with transient ischemic attack. An acceptable complication rate was achieved for patients with minor to moderate strokes. The poorest outcomes occurred in patients presenting with stroke in evolution: U-CEA in these patients should be offered with extreme caution, although we are aware that a conservative treatment may not grant a better prognosis.