Intraoperative multimodal evoked potential monitoring during carotid endarterectomy: a retrospective study of 264 patients.

Background: Methods for detecting intraoperative cerebral ischemia arising from internal carotid artery (ICA) cross-clamping during carotid endarterectomy (CEA) should be sensitive, specific, and rapid to prevent intraoperative stroke. We had 3 objectives pertaining to this: (1) investigation of the rates of success of multimodal evoked potential (mEP) monitoring using a combination of median nerve (m) somatosensory (SS) EPs, tibial nerve SSEPs (tSSEPs), and transcranial electrical stimulated motor EPs (tcMEPs); (2) evaluation of the rates of false-negative mEP results; and (3) analysis of the relationship between different time periods associated with ICA cross-clamping and the postoperative outcome of motor function in patients with significant changes in mEP monitoring.
Methods: Two hundred sixty-four patients undergoing CEA using general anesthesia with monitoring of bilateral mSSEPs, tSSEPs, and tcMEPs were retrospectively reviewed between 2009 and 2012. The rates of successful assessment of mEPs were investigated, and the rate of false-negative mEP results was analyzed. Different time periods (T1--time of clamping, T2--clamping to significant mEP changes, T3--significant mEP change to intervention, and T4--intervention to recovery of EP) were tested using Welch t test for significant association with postoperative motor deficit.
Results: (1) Multimodal EP monitoring was achieved in 241 patients (91.3%, point estimate [PE] 0.91, confidence interval [CI] 0.87 to 0.94), whereas none of the modalities were recordable in one case (PE 0.0038, CI 0.0002 to 0.019). Additionally, tSSEP was not recordable in 21 patients (PE 0.08, CI 0.05 to 0.12), and we found one case of isolated failure of tcMEP recording (PE 0.0038, CI 0.0002 to 0.019). (2) False-negative mEP results were found in 1 patient (0.4%; PE 0.0038, CI 0.0002 to 0.019). Significant mEP changes occurred in 32 patients (12.1%), and thus, arterioarterial shunt was performed in 17 (6.4%) patients. Eleven patients (4.2%) showed transient and 1 showed permanent postoperative motor deficit. (3) There was no significant difference regarding any of the time periods associated with ICA cross-clamping and postoperative alteration of motor function (T1: P = 0.19, CI -30.1 to 6.8 minutes; T2: P = 0.38, CI -23 to 9.5 minutes; T3: P = 0.25, -9.7 to 2.8 minutes; T4: P = 0.42, CI to -15.5 to 7.0 minutes).
Conclusions: Multimodal EP monitoring is applicable during CEA. The 0.4% false-negative rate suggests an advantage of mEP monitoring when compared with isolated mSSEP monitoring. Our data suggest that periods of time during cross-clamping were not significantly associated with postoperative motor deficit. However, the small number of patients limits the conclusiveness of these findings. mEP monitoring could not prevent a postoperative motor deficit in all patients, but our results suggest that it is a useful adjunct to mSSEP monitoring.