Annals of Vascular Surgery
Volume 21, Issue 1 , Pages 30-33, January 2007

Cerebral Monitoring of Somatosensory Evoked Potentials during Carotid Surgery: A Review of 100 Cases

  • Olivier Baton, MD
    • ,
  • Pascal Szym, MD
    • ,
  • Jean-Jacques Hoffmann, MD
    • ,
  • Marc Borne, MD
    • ,
  • Yves Diraison, MD
    • ,
  • Bernard Baranger, MD

      Affiliations

    • Department of Visceral and Vascular Surgery, Val de Grace Military Teaching Hospital, Paris, France; Department of Clinical Neurophysiology, Val de Grace Military Teaching Hospital, Paris, France; Department of Anesthesia and Intensive Care, Val de Grace Military Teaching Hospital, Paris, France
    • Corresponding Author InformationCorrespondence to: Bernard Baranger, MD, Service de Chirurgie Viscéral et Vasculaire, HIA Val de Grâce, 74 bd de Port-Royal, 75005 Paris, France

Paris, France

Article Outline

The purpose of this study was to evaluate immediate and middle-term results of surgical carotid artery revascularization (CAR) with cerebral monitoring of intraoperative somatosensory evoked potentials (SEPs). Between 1998 and 2004, a total of 100 CARs in 86 patients were performed under general anesthesia with SEP monitoring. A shunt was inserted if SEP amplitude decreased by 50% or latency time increased by 10%. Immediate and middle-term results were analyzed retrospectively. The shunt insertion rate was 5%. Two transient ischemic attacks were observed, and one patient died postoperatively due to myocardial infarction. The cumulative stroke and death rate was 1% at 30 days. Intraoperative SEP monitoring with selective shunt placement can be used safely for carotid surgery. Randomized studies will be necessary to determine the respective indications for various cerebral monitoring techniques.

 

Back to Article Outline

Introduction 

Clamp application for occlusion of the internal carotid artery during carotid surgery induces cerebral ischemia in 8-15% of patients.1, 2 Cerebral ischemia is one of the main causes of intraoperative neurologic morbidity and mortality during cross-clamping. Two other causes are thromboembolism, which is by far the most common, and hyperperfusion, which can result in stroke. Several techniques are available to monitor cerebral function. The most frequently used are clinical examination under locoregional anesthesia (LRA), electroencephalography (EEG), and somatosensory evoked potentials (SEPs) under general anesthesia (GA). There have been no controlled studies to document the relative efficacy of these techniques. The purpose of this report is to describe the results of our 5-year experience using SEP monitoring as the technique of choice for cerebral monitoring during carotid surgery.

Back to Article Outline

Patients and Methods 

The results of surgery on 100 carotid artery revascularization procedures performed in 86 patients between January 1998 and March 2004 were retrospectively analyzed. Narrowing was >70% according to the criteria of the North American Symptomatic Carotid Endarterectomy Trial. Both symptomatic and asymptomatic cases were included. In case of recent ischemic attack, carotid surgery was postponed for the 45 days after the event. Routine preliminary work-up included Doppler ultrasound of the supra-aortic trunks (SATs), magnetic resonance angiography (MRA) with assessment of the circle of Willis, and cerebral magnetic resonance imaging (MRI) to detect stroke damage. Preoperative EEG and SEP measurement were performed. All patients were operated on under GA with arterial blood pressure monitoring, SEP monitoring, and heparin therapy starting as soon as the cross-clamp was applied. A neurophysiologist continuously monitored SEP tracings. A shunt was inserted if SEP amplitude decreased by 50% or latency time increased by 10%. Intraoperative completion arteriography was performed in all cases. Patients underwent follow-up examinations with Doppler ultrasound at 1 month, 6 months, and 1 year. Analysis of follow-up data was purposefully limited to 1 year.

Back to Article Outline

Results 

The study population included 66 men and 20 women with a median age of 71 years (range 49-86). Stenosis was symptomatic in 30% of cases (recent stroke in 14 cases and transient ischemic attack [TIA] in 16 cases) and asymptomatic in 70% of cases. The etiology of all lesions was atherosclerosis. Four patients had no history of vascular disease. Concurrent factors included hypertension in 66 cases, diabetes in 20, and coronary artery disease in 46. Previous treatment included percutaneous coronary artery revascularization in 20 and surgical repair of the aorta or a peripheral artery in 29 cases. Assessment of the SAT demonstrated unilateral carotid artery lesions in 72 cases, significant contralateral carotid artery lesions in 11 cases including thrombosis in four and stenosis in seven, vertebral artery lesions in six cases, and subclavian artery lesions in three cases. The circle of Willis was devoid of communicating arteries in three cases (3%). Evidence of stroke damage was observed in 10 patients (14%) undergoing treatment for asymptomatic stenosis. Carotid artery revascularization consisted of endarterectomy in 90 cases including 44 procedures with patch angioplasty, venous bypass using an autologous internal saphenous vein graft in seven cases, and prosthetic bypass in three cases. Vertebral artery reimplantation was performed in two cases. The mean duration of cross-clamping was 30 min (range 13-54). Monitoring of SEP demonstrated ischemic changes after clamping application in eight patients. In five of these cases (5%), internal shunts were inserted. In the remaining three, the procedure was well advanced when the ischemic changes were detected and it was deemed quicker to finish the anastomosis than to insert a stent.

Two TIAs (transient amaurosis) were observed during postoperative recovery including one involving an intimal flap that required revision of the anastomosis on day 1. One patient died on day 1 due to myocardial infarction (MI) that occurred during reoperation for cervical hematoma. Reoperation for cervical hematoma was necessary in seven patients. One patient presented acute lower extremity ischemia on day 1. Paralysis of the hypoglossal nerve and seizure was observed in one patient with documented epilepsy. At the end of the 1-year follow-up period, four patients presented restenosis including one who was symptomatic. Three patients required redo, including venous bypass in two cases and patch angioplasty in one.

The cumulative stroke and death rate was 1%. The rate of freedom from neurologic deficits was 96.5% for patients treated for asymptomatic stenosis and 98.5% for patients treated for symptomatic stenosis.

Back to Article Outline

Discussion 

For carotid artery stenosis >70%, surgical treatment can be more beneficial than medical treatment when performed on symptomatic patients by a surgical team whose cumulative stroke and death rate is 5% or on asymptomatic patients whose cumulative stroke and death rate is 3%.1, 2 The patient's response to cross-clamp application is a pivotal factor due to the risk of cerebral ischemia. Cerebral ischemia is irreversible when the flow is <10-12 mL/100 g/min and reversible below 20 mL/100 g/min depending on the quality of collateral circulation and on circulatory and metabolic adaptation. Recognized protective factors against cerebral ischemia include heparin therapy, careful surgical technique, and intraoperative detection of morphologic abnormalities. However, intraoperative arteriography does not rule out the possibility of complications, as shown by the patient who required surgical revision on day 1 to correct an intimal flap causing transient amaurosis.

The value of shunting is controversial. Some groups recommend systematic use of shunts, while others consider them unnecessary. Most groups recommend selective shunting. Imparato et al.3 listed the following as absolute indications for shunting: detection of ischemic changes after cross-clamp application, recent stroke, and occlusion of the contralateral internal carotid artery. Pulli et al.4 showed that the shunt insertion rate was significantly higher in patients with than in those without contralateral carotid artery occlusion (25.6% vs. 6.9%). Conversely, in a prospective randomized study including 86 patients undergoing surgery within 30 days after a regressive stroke, Ballotta et al.5 found no significant difference with regard to shunt insertion or morbidity/mortality rate. The possibility of cerebral ischemia after clamp application is a compelling argument for intraoperative monitoring of cerebral function. The most frequently used techniques for cerebral monitoring are LRA, transcranial Doppler ultrasound, EEG, and SEP. Few studies have been performed to compare the efficacy of these monitoring techniques.

LRA with deep and superficial cervical plexus blockade is the most widely used technique. It allows clinical assessment of cerebral function and provides excellent hemodynamic stability. A randomized controlled study was carried out to compare the impact of LRA and GA on cerebral flow.6 Findings showed that LRA achieved better preservation of ipsilateral cerebral flow with no correlation with variations in systemic blood pressure. The cumulative stroke and death rate was the same in the LRA and GA groups, but the need for shunting was 67% in the GA group vs. 15% in the LRA group. Regarding the risk of postoperative MI, a randomized controlled study by Sbarigia et al.7 showed that the MI rate was lower after LRA. Despite the poor statistical power of that study, LRA appears to be indicated for patients at high risk for cardiac complications. However, the success of LRA depends on the patient's ability to remain cooperative despite the uncomfortable operative position. Conversion to general anesthesia with emergency induction of anesthesia and intubation is necessary in 3% of cases.

In addition to enhancing the comfort of the patient and surgeon, GA decreases cerebral metabolic requirements and oxygen requirements, thus improving tolerance to hypoxia. However, use of GA raises the need for monitoring of cerebral function to detect ischemic changes after cross-clamp application. EEG is frequently used despite a non-negligible 5% false-negative rate8 and the ability of anesthetic agents to affect signals. Transcranial Doppler ultrasound study (TDUS) allows immediate reactivity in case of significant changes in blood flow in the middle cerebral artery. The diagnostic value of TDUS has been documented for detection of postoperative hypoperfusion, but the procedure cannot be used in 10-20% of patients due to the thickness of the temporal bone window.

Determination of sensory function using SEP depends on stimulation of the median nerve. An evoked potential amplitude reduction >50% or a decrease in latency time >10% is considered pathologic. Measurements are made every minute. In a series including 994 endarterectomies, Haupt and Horsch9 reported a false-negative rate of 0.1% and a positive predictive value for complications of 99.9%. A prospective series comparing the SEP under GA to LRA10 supported the protective effect of anesthetic agents by showing that a 30% decrease in amplitude could be considered pathologic when SEP monitoring was performed under LRA. However, unlike EEG and LRA, monitoring focused on the cortical sensory complex can overlook early-stage ischemia in the motor zone. This problem might explain the three cases of clamp-related ischemia that occurred during endarterectomy. In patients with a history of stroke, SEP monitoring demonstrates asymmetry between the two hemispheres that must be used as the new baseline amplitude. In this way, SEP monitoring based on the usual criteria can be used for reliable detection of cerebral ischemia after clamp application with no significant difference.11

Back to Article Outline

Conclusion 

In our experience, general anesthesia with monitoring of evoked potentials appeared to provide good results in terms of patient and surgeon comfort during the procedure and freedom from neurologic deficit. We recommend selective shunting on the basis of intraoperative electrophysiologic findings. The best cerebral function monitoring technique remains unclear due to the lack of randomized controlled studies with sufficient statistical power. The choice of monitoring technique depends on the internal organization of each surgical facility and on a concerted choice by the surgeon and anesthesiologist.

Back to Article Outline

References 

  1. North American Symptomatic Endarterectomy Trial Collaborators . Beneficial effect of carotid endarterectomy in symptomatic patients with high-grade carotid stenosis. N Engl J Med. 1991;325:445–453
  2. European Carotid Surgery Trialists Collaborative Group . Randomised trial of endarterectomy for recently symptomatic carotid stenosis: final results of the MRC European Carotid Surgery Trial. Lancet. 1998;351:1379–1387
  3. Imparato AM, Ramirez A, Riles T, et al. Cerebral protection in carotid surgery. Arch Surg. 1982;117:1073–1078
  4. Pulli R, Dorigo W, Barbanti E, et al. Carotid endarterectomy with contralateral carotid artery occlusion: is this a higher risk group?. Eur J Vasc Endovasc Surg. 2002;24:63–68
  5. Ballotta E, Da Giau G, Baracchini C, et al. Early versus delayed carotid endarterectomy after a non disabling ischemic stroke: a prospective randomized study. Surgery. 2002;131:287–293
  6. McCarthy RJ, Nasr MK, McAteer P, Horrocks M. Physiological advantages of cerebral blood flow during carotid endarterctomy under local anesthesia: a randomised clinical trial. Eur J Vasc Endovasc Surg. 2002;24:215–221
  7. Sbarigia E, Darrio Vizza C, Antonini M, et al. Locoregional versus general anesthesia in carotid surgery: is there an impact on perioperative myocardial ischemia? Results of a prospective monocentric randomized trial. J Vasc Surg. 1999;30:131–138
  8. Stoughton J, Nath RL, Abbott W. Comparison of simultaneous electroencephalographic and mental status monitoring during carotid endarterectomy with regional anesthesia. J Vasc Surg. 1998;28:1014–1023
  9. Haupt WF, Horsch S. Evoked potential monitoring in carotid surgery: a review of 994 cases. Neurology. 1992;42:835–838
  10. Sbarigia E, Schioppa A, Misuraca M, et al. Somatosensory evoked potentials versus locoregional anesthesia in the monitoring of cerebral function during carotid artery surgery. Eur J Vasc Endovasc Surg. 2002;21:413–416
  11. Manninen PH, Tan TK, Sarjeant RM. Somatosensory evoked potential monitoring during carotid endarterectomy in patients with a stroke. Anesth Analg. 2001;1:39–44

PII: S0890-5096(06)00012-4

doi:10.1016/j.avsg.2006.10.006

Annals of Vascular Surgery
Volume 21, Issue 1 , Pages 30-33, January 2007

Article Tools