J Neurosurg 76:752-758, 1992
Thromboendarterectomy of the symptomatic occluded internal carotid artery PATRICK W. McCoRIMICK, M.D., ROBERT F. SPETZLER, M.D., JULIAN E. BAILES, M.D., JOSEPH M. ZABRAMSKI, M.D., AND JAr~IES L. FREY, M.D.
Divisions of Neurological SurgeJ3' and Neurology. Barrow Neurological Institute, Phoenix, Arizona u- A retrospective review of 42 patients (mean age 61.4 years) with surgically managed symptomatic internal carotid artery, occlusion is reported. A standardized surgical protocol aimed at restoration of flow in the vessel was used. Presenting symptoms included hemispheric transient ischemic attacks in 68% of patients, new fixed neurological deficits in 28%, amaurosis fugax in 28%, and stroke-in-evolution in 9%. Twenty-four arteries were successfully reopened. A proximal remnant angioplasty (stumpectomy) was performed alone in nine patients or in combination with an external carotid endarterectomy in nine. In four patients with persisting symptoms who failed to achieve primary restoration of flow, a superficial temporal-to-middle cerebral artery bypass procedure was performed. The permanent surgical morbidity rate was 2% and the surgical mortality rate was 0%. Transient postoperative deficits were present in three patients (7%). Follow-up review at a mean of 40 months was obtained in 39 patients (93%). Following surgical intervention, five patients died of unrelated causes, two had neurological events consistent with a transient cerebral ischemic attack, and two had vertebrobasilar insufficiency. No patient suffered from stroke. Of the 24 successfully reopened vessels, follow-up ultrasound evaluations were obtained in 17 (73%) at a mean of 28 months after surgery. In 15 patients (88%) the vessels were widely patent, one (5.8%) had stenosis greater than 70%, and one (5.8%) showed asymptomatic reocclusion. Reopening occluded internal carotid arteries in selected patients is associated with low surgical morbidity and mortality rates. Further studies are necessary to determine the impact of this surgical therapy on the natural history' of this condition.
internal carotid artery endarterectomy extracranial-intracranial bypass thrombectomy KEY WORDS
HROMBOENDAR~ERECTOMYof the internal carotid artery (ICA) has been described as a surgical therapy to improve the natural history of ICA occlusion. In the 1960's, three series of patients with ICA reopening were reported. Thompson, et al.,~9 described 118 operations with a 6.2% operative mortality rate, Murphey and Maccubbin ~3 presented 50 patients with a surgical morbidity and mortality rate of 16%, and Hunter, et al.,8 reported on 21 patients with a surgical morbidity and mortality rate of about 15 %. Each of these studies demonstrated that the surgical morbidity and mortality rate was high if all patients with ICA occlusion underwent surgery, but that a subgroup fared better than the group as a whole. This subgroup included patients who presented early without "protbund" neurological deficits. Two groups of investigators have reported their series of surgical thromboendarterectomy for ICA occlusion
T
752
carotid artery occlusion
in the subgroup of patients presenting early without profound deficits. These patients were most likely to benefit from surgery. Hugenholtz and Elgie 7 excluded patients with drowsiness, major neurological deficits, and tandem lesions distal to the ICA occlusion. Based on these criteria, they reduced the surgical morbidity rate to 10% and the surgical mortality rate to 0% in a group of 35 patients. Hafner and Tew 6 used similar selection criteria in 47 patients and achieved surgical mortality and neurological morbidity rates of 0%. Despite the identification of a subgroup of patients expected to benefit from ICA restoration of flow and the publication of two clinical series showing favorable surgical morbidity and mortality rates in this group, a recent poll indicated that, according to expert opinion, ICA occlusion was a contraindication to ICA surgery? j In fact, after reviewing 1000 carotid endarterectomies, the responders categorized the 60 procedures (6%) per-
J. Neurosurg. / Volume 76~May, 1992
T h r o m b o e n d a r t e r e c t o m y for o c c l u d e d c a r o t i d a r t e r y formed for complete carotid artery occlusion as inappropriate. To help clarify this issue, we have reviewed our experience with ICA thromboendarterectomy using strict patient selection criteria, a uniformly applied surgical algorithm, standardized microsurgical technique, and aggressive intraoperative neuronal ischemic protection. Clinical Material and Methods
Patient Population From 1985 to 1990, 42 patients with angiographically documented ICA occlusion were evaluated, A uniform diagnostic protocol was used for all patients, with a computerized tomography (CT) brain scan obtained at presentation to rule out the presence of major infarct, hemorrhage, or other lesions. Four-vessel cerebral angiography was carried out immediately in all patients using selective catheterization, image subtraction, and image magnification. In every case, a complete occlusion of the ICA was documented; in no patient was there evidence of a "string sign" or delayed anterograde filling through an apparently occluded vessel segment. Twenty-one patients received heparin intravenously while being prepared for surgery. Surgery was performed immediately on every patient. Follow-up review was accomplished through office visits and telephone interviews. Operative Technique and Algorithm A consistent surgical protocol was applied to all patients (Fig. 1). The ICA thromboendarterectomy technique used is a modification of that described by others.6'7'~6A standard dissection was completed to expose the common, internal, and external carotid arteries, and care was exercised to avoid manipulating the occluded ICA. Without occlusion of any vessel, a small arteriotomy on the lateral ICA was made distal to the obvious atheromatous disease. At this point, one of three phenomena was noted: 1) the clot spontaneously expressed itself with good backflow; 2) the clot did not spontaneously exit and no backflow was observed; or 3) a small amount of clot exited with weak backflow. In the first instance, standard cross-clamping of the vessels was performed followed by an endarterectomy. In the latter two events, a No. 2 French Fogarty catheter with a 0.2-ml balloon was passed up the ICA 10 to 12 cm with very gentle pressure. The balloon was inflated with saline and gently withdrawn. This maneuver occasionally delivered loose clot and was followed by vigorous backflow; if it failed, it was repeated once or twice, after which further attempts to reopen the ICA were abandoned. If the vessel was atretic and of small caliber, indicating little chance for reopening, the operation was abandoned without passing the Fogarty catheter. When attempts to reopen the vessel failed, a clip was used to occlude the ICA from the circulation at the level of the bifurcation. This procedure, commonly referred to as "stumpectomy," ensured a smooth vascuJ. Neurosurg. / Volume 76/May, 1992
FIG. 1. Operative paradigm applied to patients with symptomatic internal carotid artery occlusion.
lar lumen from the common carotid to the external carotid artery without a blind pouch that can act as a source of embolic material. If angiographic evidence of significant external carotid artery atheromatous disease was present, an external carotid artery endarterectomy was performed. Throughout the surgical procedure, the patients were maintained with electroencephalographic (EEG) burstsuppression by means of titrated doses of pentobarbital? ~ Normotensive blood pressure was preserved, and in patients who received heparin preoperatively administration was continued throughout the operative procedure. Immediately after surgery, the patients were started on a course of aspirin. Vessel shunting was employed on one occasion when an asymmetric EEG burst-suppression pattern developed during the endarterectomy. A standardized microsurgical technique was used for endarterectomy to ensure a clean vessel wall and smoothly tapered intima proximal and distal to the atheroma dissection.~S Results
Of the 42 patients, 29 (70%) were males, and the average age was 61.4 years. Associated medical conditions included tobacco use in 67%, essential hypertension severe enough to warrant at least one instance of drug therapy in 72%, and previous myocardial infarction and/or stable angina in 26%. The clinical presentation included focal cortical transient ischemic attacks (TIA's) in 68%, amaurosis fugax in 28%, new fixed neurological deficits in 28%, and stroke-in-evolution in 9%. For patients who presented initially to our institution for treatment, the mean time from the neurological event to the first surgical procedure was 42 hours. Computerized tomography scans of the brain obtained within 4 hours of presentation demonstrated remote cerebral infarctions in 19%, established lacunar infarctions in 17%, and new changes consistent with 753
P. W. McCormick, et al.
FIG. 2. Upper:Computerized tomography scan showing infarction ipsilateral to a stenotic carotid artery lesion. Lower Left: Angiogram confirming the presence of the stenotic lesion. Lower Right: Six weeks later, the patient developed a new-onset hemiparesis due to a complete carotid occlusion. Emergency reoperation was successful, with resolution of the neurological deficit.
ischemia in 15%. Of those with new changes, four patients had watershed distribution infarcts. Over onehalf (56%) of the CT scans were normal. A total of 46 operations were performed on the 42 patients, including 24 (52%) successful vessel reopenings, nine (20%) stumpectomies with associated external carotid endarterectomy, and four (9%) stumpectomies with external carotid endarterectomies that eventually required extracranial-intracranial (EC-IC) bypass (Figs. 2 and 3). Transient surgical morbidity occurred in three patients (7 %) and included a postoperative seizure in one, vocal cord paralysis in one, and a mild contralateral 7fi4
FIG. 3. UpperLeft: Preoperative angiogram in a patient with complete internal carotid artery occlusion and a filling proximal remnant. Upper Right: Postoperative study showmg obliteration of the remnant. Transient ischemic attacks ceased after the stumpectomy and external carotid endarterectomy. Lower: Thrombotic material is seen in the proximal internal carotid artery remnant &the endarterectomized specimen (arrow).
hemiparesis in one. Permanent morbidity occurred in one patient (2%) whose left hemiparesis worsened following restoration of flow in the ICA. Postoperative TIA's occurred in four patients who initially underwent stumpectomy and external carotid endarterectomy; in accordance with the operative paradigm (Fig. 1), these patients subsequently underwent an EC-IC bypass procedure. There were no operative deaths in the series. Long-term arterial patency was assessed in 17 (73%) of the 24 patients, with successful restoration of flow. Carotid Doppler ultrasound studies were performed a mean of 28 months after surgery. In 15 of these patients (88 %), vessels were widely patent; of the other two, one had greater than 70% stenosis and one had reocclusion. The restenosed vessel was operated on again and a vein patch was placed. Three patients were lost to follow-up review; the follow-up period for the remaining 39 patients was an J. Neurosurg. / Volume 76/May, 1992
Thrornboendarterectomy for occluded carotid artery
FIG. 4. Left: Graph showing the long-term outcome for patients with successful restoration of flow in the internal carotid artery for the end-points of stroke, transient ischemic attacks, and stroke death (upper line) and the expected stroke event in patients left untreated (lower line). Right: Same analysis for patients who did not have successful reopening but who were surgically treated with stumpectomy, external carotid endarterectomy, or extracranial-intracranial bypass until their symptoms resolved (upper line) and for patients left untreated (lower line).
FIG. 6. Operative specimen of atherosclerotic plaque and organized thrombus with distal extension removed from a patient who presented with transient ischemic attacks and internal carotid artery occlusion. Measure is in millimeters.
FIG. 5. Carotid angiography in a patient with crescendo transient ischemic attacks from a left internal carotid artery occlusion. Left." Retrograde filling of the cavernous and petrous carotid artery is seen on delayed films predicting a greater likelihood of reopening of the vessel. Right." Retrograde filling at 56 seconds following injection. The patient's symptoms ceased following restoration of flow in the internal carotid artery. average of 40 months after their last operation. Five patients (13%) died, two of cancer and three of myocardial infarction. Four patients (10%) experienced neurological events: TIA's in two and vertebrobasilar insufficiency in two. Six patients (15%) have had new myocardial infarctions or new-onset angina. The outcome for patients with successful restoration of ICA flow was not significantly different from that of patients who were treated until asymptomatic with the alternative surgical strategies described. No patient has had a subsequent stroke, and the rates for neurological transient events (TIA's and vertebrobasilar insufficiency) for both groups are comparable with the expected natural history for completed stroke in nonsurgically treated patients (Fig. 4). The extent of reflux visualization in the ICA on
J. Neurosurg. / Volume 76 /May, 1992
preoperative angiograms was noted as an excellent predictor of successful vessel reopening. In 26 patients with delayed selective digital subtraction studies, retrograde reflux visualization of the occluded ICA was evaluated. Based on the Hugenholtz and Elgie grading system, v reflux to the supraclinoid segment only (Grade 2) occurred in four patients with successful reopening in one, for a success rate of 25%. Reflux to the cavernous portion of the ICA (Grades 3 and 4) occurred in eight patients with successful reopening in four, for a success rate of 50%. Reflux to the petrous ICA (Fig. 5) occurred in 14 patients with successful reopening in 10, for a success rate of 7 1%. Pathology specimens were obtained from 33 patients (nine patients had stumpectomy only), and 42% of these demonstrated acute intraplaque hemorrhage. A typical pathological specimen is shown in Fig. 6.
Discussion
Patient Selection Previous experience with surgery for symptomatic ICA occlusion has demonstrated the important effect of patient selection on the operative morbidity and mortality rates. Case series published in the 1960's 755
P. W. McCormick, et al. identified patients who did not have "profound" neurological deficits as responding much better to surgery. s~3'~9 Thompson, et at., ~9 showed a drop in the procedure mortality rate from 6.2% to 1.1% for carotid endarterectomy when patients with neurological deficits were excluded. Discussing ICA reopening, they stated: "Clinical considerations of the stroke itself are the most important factors determining operability. Patients with acute profound strokes of 12 hours' duration or more should not be operated upon because of the prohibitive mortality. Whether one should operate on such a patient if seen within six hours of onset is open to question. If the neurologic deficit is mild and is improving rapidly, or if a previously audible carotid bruit suddenly disappears, operation may be done promptly to ensure restoration of carotid flow." The influence of neurological deficit on the surgical morbidity and mortality rates associated with ICA reopening can be inferred from other recent case series. Ojemann, et al., 16 reviewed 16 cases of ICA occlusion and concluded that patients with profound deficits and decreased levels of consciousness should not be surgically treated. DeWeese 3 reported a series of 19 patients with severe progressive neurological deficits associated with carotid stenosis or occlusion; the surgical mortality rate in those patients was 37%. Meyer, et al., 12 reviewed their experience with restoration of ICA flow in patients with severe fixed deficits and decreased levels of consciousness. Although a subset of patients who benefited from the surgery could be identified, the overall surgical mortality rate in these very sick patients was 20.6%. It is not surprising, then, that case series excluding patients with severe neurological deficits show a favorable response to restoration of flow in the ICA. Two groups of authors have reported such series, both with no surgically related deaths (Table 1). 6.7 Kusunoki, et al., 9 divided their patients into two groups: those with neurological, medical, and angiographic risk factors and those without. The latter group of 14 patients had no operative mortality. Our data further strengthen the conclusion that a subgroup of patients, chosen primarily based on their presenting neurological examination, will have low surgical morbidity and mortality rates from restoration of ICA flow (Table 1). Our series excluded patients with profound neurological deficits and intracerebral
TABLE
hemorrhage visualized on CT and with documented chronic occlusion and no reflux visualization of the petrous ICA. The issue of timing of surgery remains unclear. Series where the precise timing between ICA occlusion and attempted restoration of flow is known are few. ~0.~4These include postangiographic or postsurgical occlusions that represent a pathophysiology other than spontaneous occlusion associated with atherosclerotic disease. The impression of those reporting such series is that the earlier surgery is performed, the better patients fare. Operative Technique
Certain methodological considerations concerning restoration of flow in the ICA deserve emphasis. Preoperatively, patients should be started on aspirin therapy, which can be given by suppository if necessary. Heparinization seems to be a logical medical treatment for symptomatic ICA occlusion before surgery, but no strong efficacy data are available. It is known that distal embolization can occur and is associated with a poor outcome. 12 In patients receiving heparin (partial thromboplastin time 1.5 x control), we continued heparin administration throughout the operation and withheld aspirin therapy until the heparin was stopped. During induction of anesthesia, great care must be taken to keep blood pressure normotensive. Reduction in blood pressure can decrease collateral flow to compromised vascular territories. Intravenous thiopental titrated to maintain the patient in EEG burst-suppression reduces the metabolic requirements of ischemic neurons. Careful use of this technique will protect the brain from fluctuations in collateral blood flow during surgical manipulation. 2~ It is important to monitor a parameter sensitive to cerebral oxygen delivery during the operation to assess the adequacy of collateral flow. Under barbiturate cerebral protection, this can be accomplished by observing the symmetry of the burst-suppression pattern between hemispheres or by direct assessment of cerebral oxygen saturation.~ ~,~8.20In our experience with ICA occlusion, cross-clamping the external carotid artery has caused monitoring asymmetries. When the arteriotomy is made, the clot should be given time to spontaneously decompress. Only if it fails to do so should an attempt be made to pass a Fogarty
1
Outcome of patients selected for surgical restoration of lCA flow in four series*
Authors & Year
NO.casesOf
Exclusion Criteria
Successful Morbidity Mortality Reopening
Kusunoki, et at., 1978 14 profounddeficit, major medical problem, severeangiographicdisease 1 (7) 0 NA Hugenholtz & Elgie, 1980 35 ICH,drowsy, major deficit 4 (11) 0 19 (54) Hafner & Tew, 1981 47 profounddeficit, ICH, symptoms> 45 days 0 0 32 (68) McCormick, et al., 1992 43 profounddeficit, ICH, chronic disease with poor reflux 1 (2) 0 24 (56) totals 139 6 (4.3) 0 75 (59) * All four seriesused similarexclusion criteria. Numbers in parenthesesare percentagesof the total in each study. ICA = internal carotid artery; ICH = intracerebralhemorrhage; NA = not available. 756
J. Neurosurg. / Volume 7 6 / M a y , 1992
Thromboendarterectomy for occluded carotid artery catheter. This maneuver requires experience and judgment. It should be noted that low morbidity and mortality rates are associated with a 50% to 60% successful restoration of flow (Table 1). Overvigorous attempts to reopen a vessel may well result in poor surgical outcome due to intimal injury, distal embolization, and distal dissection. Small atretic vessels should not be subjected to attempts to restore flow with a catheter. In the event that the ICA cannot be reopened and extracranial surgical procedures do not prevent subsequent neurological symptoms, surgical augmentation of cerebral collateral blood flow has been successful. Four patients in our series had anterior circulation TIA's following stumpectomy and external carotid endarterectomy and, therefore, a superficial temporal-tomiddle cerebral artery bypass was performed; one patient died 2 years later of a myocardial infarction and three remain asymptomatic an average of 37 months after surgery. This select group of patients may well have hemodynamic ischemic symptoms, which explains their response to bypass surgery. Preoperative Angiography An important observation in this study is the correlation between the extent of reflux visualization of the occluded ICA by angiegraphy and its successful reopening. Although other authors have mentioned this relationship].4.17 labeled "collateral filling," it probably does not represent true collateral filling for two reasons. There is minimal collateral supply to the petrous carotid artery and no such supply to the cervical carotid artery. The vessel fills retrograde on delayed-sequence films, and the vessel has no pressure differential across it, especially in the retrograde direction. In our opinion, the vessel fills with contrast medium in a delayed fashion as a result of refluxing of contrast medium into the distal portion, possibly aided by turbulent flow patterns in this region. Hugenholtz and Elgie7 described a five-tier classification system for grading this reflux pattern, but were unable to conclude if such a detailed classification was necessary. We have reduced the classification system to three tiers and have shown in this series that the further the refluxing of contrast medium into the vessel, the greater the likelihood of restoration of flow. Filling of the petrous vessel was associated with a 71% success rate for restoration of flow. This finding is important because it facilitates preoperative surgical planning and patient counseling. Patients with symptomatic chronic ICA occlusion and petrous vessel reflux are candidates for thromboendarterectomy. Furthermore, if a reflux pattern is not encouraging, an intraoperative decision to abort attempts at restoration of flow with a Fogarty catheter can be reached earlier. Morbidity and Mortality of Acute ICA Occlusion The natural history of spontaneous ICA occlusion associated with atherosclerotic disease has been estabJ. Neurosurg. / Voh~me 76/May, 1992
lished by prospective randomized clinical trials. Fields and Lemak 5 reported 359 patients with this diagnosis at a mean follow-up period of 44 months. New strokes occurred in 25 % of the patients; in those with complete records, 64% of the strokes were ipsilateral to the occluded ICA. Cote, et al., l followed 47 patients with ICA occlusion for 34.4 months; new strokes occurred in 23.5% of the patients and 66% were ipsilateral. Together, these studies demonstrate an annual stroke rate of 5% in the vascular territory ipsilateral to an occluded ICA. 3,5 At the time of ICA occlusion, our patients would be expected to have some associated morbidity and mortality, but the incidence is difficult to estimate. Meyer, et al.,~2 surveyed the literature and reported that acute ICA occlusion is associated with a 40% to 69% incidence of profound deficit and a 16% to 55% death rate. However, because patients with profound deficits were eliminated from our protocol, the morbidity and mortality rates associated with spontaneous occlusion alone would not be as severe as the high (42%) rate of ipsilateral events described by Nicholls, et al., 15 or that described by Meyer, et al. Surgical Outcome The permanent surgical morbidity rate for attempted ICA reopening in the 42 operations in this series was 2%, and the 30-day surgical mortality rate was 0%. These outcomes compare well with the anticipated morbidity and mortality rates of ICA occlusion even in this select group of patients. In an average followup period of 40 months, no patient with successful restoration of flow in this series has had a stroke; furthermore, no patients treated with alternative surgical approaches until their symptoms resolved have had subsequent strokes. The literature concerning thromboendarterectomy of spontaneously occluded symptomatic ICA's consistently demonstrates a subgroup of patients who have successful restoration of flow with little morbidity and mortality. This subgroup is best defined as those without severe neurological deficits, without a decreased level of consciousness, and without intracerebral hemorrhage. Clinical experience with such patients by our group and others shows a low surgical morbidity and mortality rate associated with carefully performed thromboendarterectomy. During long-term follow-up review, the vessels remained patent and the patients had fewer strokes than would be anticipated from natural history data. Our patients followed a prospectively designed diagnostic, anesthetic, and surgical protocol. Their evaluation and treatment were interdisciplinary. The data, however, were collected retrospectively and no simultaneous control group was available for comparison. This experience cannot definitively establish the efficacy of surgical management for ICA occlusion. However, based on the available data, we currently recommend surgery for patients with acute symptomatic ICA 757
P. W. McCormick, et al. occlusion who have not experienced a decreased level of consciousness, hemiplegia, or aphasia.
References 1. Cote R, Barnett HJM, Taylor DW: Internal carotid occlusion: a prospective study. Stroke 14:898-902. 1983 2. De Bakey ME, Crawford ES, Morris GC Jr, et al: Surgical considerations of occlusive disease of the innominate, carotid, subclavian, and vertebral arteries. Ann Sorg 154: 698-725, 1961 3. DeWeese JA: Management of acute strokes. Surg Clin North Am 62:467-472, 1982 4. Dyken ML, Klatte E, Kolar OJ, et al: Complete occlusion of common or internal carotid arteries. Clinical significance. Arch Neurol 30:343-346, 1974 5. Fields WS, Lemak NA: Joint study of extracranial arterial occlusion. X. Internal carotid artery occlusion. JAMA 235:2734-2738, 1976 6. Hafner CD, Tew JM: Surgical management of the totally occluded internal carotid artery: a ten-year study. Surgery 89:710-717, 1981 7. Hugenholtz H, Elgie RG: Carotid thromboendartemctomy: A reappraisal. Criteria for patient selection. J Nenrosurg 53:776-783, 1980 8. Hunter JA, Julian OC, Dye WS, et al: Emergency operation for acute cerebral ischemia due to carotid artery- obstruction. Review of 26 cases. Ann Surg 162:901-904, 1965 9. Kusunoki T, Rowed DW, Tator CH, et al: Thromboendarterectomy for total occlusion of the internal carotid artery: a reappraisal of risks, success rate and potential benefits. Stroke 9:34-38, 1978 10. Kwaan JHM, Connolly JE, Sharefkin JB: Successful management of early stroke after carotid endarterectomy. Ann Surg 190:676-678, 1979 11. McCormick PW, Stewart M, Goetting MG, et al: Regional cerebrovascular oxygen saturation measured by optical spectroscopy in humans. Stroke 22:596-602, 1991 12. Meyer FB, Sundt TM Jr, Piepgras DG, et al: Emergency carotid endarterectomy for patients with acute carotid oc-
758
clusion and profound neurological deficits. Ann Surg 203: 82-89, 1986 13. Murphey I:, Maccubbin DA: Carotid endarterectomy. A long-term follow-up study, d Neurosurg 23:156-168, 1965 14. Najafi H, Javid H, Dye WS, et al: Emergency carotid thromboendarterectomy. Surgical indications and results. Arch Surg 103:610-614, 1971 15. Nicholls SC, Bergelin R, Strandness DE: Neurologic sequelae of unilateral carotid artery occlusion: immediate and late. J Vasc Surg 10:542-548, 1989 16. Ojemann RG, Crowell RM, Roberson GH, et al: Surgical treatment of extracranial carotid occlusive disease. Clin Neurosurg 22:214-263, 1975 17. Shucart WA, Garrido E: Reopening some occluded carotid arteries. Report of four cases. J Neurosurg 45:442-446, 1976 18. Spetzler RF, Marlin N, Hadley MN, et al: Microsurgical endarterectomy under barbiturate protection: a prospective study. J Neurosurg65:63-73, 1986 19. Thompson JE, Austin DJ, Patman RD: Endartereclomy of the totally occluded carotid artery for stroke. Arch Surg 95:791-801, 1967 20. Wilkinson E, Spetzler RF, Carter LP, et al: Intraoperative barbiturate therapy during temporary vessel occlusion in man, in Spetzler RF, Carter LP. Selman WR, et al (eds): Cerebral Revascularizaliun for Stroke. New York: Thieme-Stratton, 1985, pp 397-403 21. Winslow CM, Solomon DH, Chassin MR, et al: The appropriateness of carotid endarterectomy. N Engl J Med 318:721-727, 1988 Manuscript received June 13, 1991. Accepted in final form October 25, 1991. Address for Dr. McCormick: St. Vincent's Hospital and Medical Center, Toledo, Ohio. Address for Dr. Bailes: Allegheny General Hospital, Pittsburgh, Pennsylvania. Address reprint requests to." Robert F. Spetzler, M.D., Editorial Office, Barrow Neurological Institute, 350 West Thomas Road, Phoenix, Arizona 85013-4496.
J. Neurosurg. / Volume 7 6 / M a y , 1992
Thromboendarterectomy of the symptomatic occluded internal carotid artery PATRICK W. McCoRIMICK, M.D., ROBERT F. SPETZLER, M.D., JULIAN E. BAILES, M.D., JOSEPH M. ZABRAMSKI, M.D., AND JAr~IES L. FREY, M.D.
Divisions of Neurological SurgeJ3' and Neurology. Barrow Neurological Institute, Phoenix, Arizona u- A retrospective review of 42 patients (mean age 61.4 years) with surgically managed symptomatic internal carotid artery, occlusion is reported. A standardized surgical protocol aimed at restoration of flow in the vessel was used. Presenting symptoms included hemispheric transient ischemic attacks in 68% of patients, new fixed neurological deficits in 28%, amaurosis fugax in 28%, and stroke-in-evolution in 9%. Twenty-four arteries were successfully reopened. A proximal remnant angioplasty (stumpectomy) was performed alone in nine patients or in combination with an external carotid endarterectomy in nine. In four patients with persisting symptoms who failed to achieve primary restoration of flow, a superficial temporal-to-middle cerebral artery bypass procedure was performed. The permanent surgical morbidity rate was 2% and the surgical mortality rate was 0%. Transient postoperative deficits were present in three patients (7%). Follow-up review at a mean of 40 months was obtained in 39 patients (93%). Following surgical intervention, five patients died of unrelated causes, two had neurological events consistent with a transient cerebral ischemic attack, and two had vertebrobasilar insufficiency. No patient suffered from stroke. Of the 24 successfully reopened vessels, follow-up ultrasound evaluations were obtained in 17 (73%) at a mean of 28 months after surgery. In 15 patients (88%) the vessels were widely patent, one (5.8%) had stenosis greater than 70%, and one (5.8%) showed asymptomatic reocclusion. Reopening occluded internal carotid arteries in selected patients is associated with low surgical morbidity and mortality rates. Further studies are necessary to determine the impact of this surgical therapy on the natural history' of this condition.
internal carotid artery endarterectomy extracranial-intracranial bypass thrombectomy KEY WORDS
HROMBOENDAR~ERECTOMYof the internal carotid artery (ICA) has been described as a surgical therapy to improve the natural history of ICA occlusion. In the 1960's, three series of patients with ICA reopening were reported. Thompson, et al.,~9 described 118 operations with a 6.2% operative mortality rate, Murphey and Maccubbin ~3 presented 50 patients with a surgical morbidity and mortality rate of 16%, and Hunter, et al.,8 reported on 21 patients with a surgical morbidity and mortality rate of about 15 %. Each of these studies demonstrated that the surgical morbidity and mortality rate was high if all patients with ICA occlusion underwent surgery, but that a subgroup fared better than the group as a whole. This subgroup included patients who presented early without "protbund" neurological deficits. Two groups of investigators have reported their series of surgical thromboendarterectomy for ICA occlusion
T
752
carotid artery occlusion
in the subgroup of patients presenting early without profound deficits. These patients were most likely to benefit from surgery. Hugenholtz and Elgie 7 excluded patients with drowsiness, major neurological deficits, and tandem lesions distal to the ICA occlusion. Based on these criteria, they reduced the surgical morbidity rate to 10% and the surgical mortality rate to 0% in a group of 35 patients. Hafner and Tew 6 used similar selection criteria in 47 patients and achieved surgical mortality and neurological morbidity rates of 0%. Despite the identification of a subgroup of patients expected to benefit from ICA restoration of flow and the publication of two clinical series showing favorable surgical morbidity and mortality rates in this group, a recent poll indicated that, according to expert opinion, ICA occlusion was a contraindication to ICA surgery? j In fact, after reviewing 1000 carotid endarterectomies, the responders categorized the 60 procedures (6%) per-
J. Neurosurg. / Volume 76~May, 1992
T h r o m b o e n d a r t e r e c t o m y for o c c l u d e d c a r o t i d a r t e r y formed for complete carotid artery occlusion as inappropriate. To help clarify this issue, we have reviewed our experience with ICA thromboendarterectomy using strict patient selection criteria, a uniformly applied surgical algorithm, standardized microsurgical technique, and aggressive intraoperative neuronal ischemic protection. Clinical Material and Methods
Patient Population From 1985 to 1990, 42 patients with angiographically documented ICA occlusion were evaluated, A uniform diagnostic protocol was used for all patients, with a computerized tomography (CT) brain scan obtained at presentation to rule out the presence of major infarct, hemorrhage, or other lesions. Four-vessel cerebral angiography was carried out immediately in all patients using selective catheterization, image subtraction, and image magnification. In every case, a complete occlusion of the ICA was documented; in no patient was there evidence of a "string sign" or delayed anterograde filling through an apparently occluded vessel segment. Twenty-one patients received heparin intravenously while being prepared for surgery. Surgery was performed immediately on every patient. Follow-up review was accomplished through office visits and telephone interviews. Operative Technique and Algorithm A consistent surgical protocol was applied to all patients (Fig. 1). The ICA thromboendarterectomy technique used is a modification of that described by others.6'7'~6A standard dissection was completed to expose the common, internal, and external carotid arteries, and care was exercised to avoid manipulating the occluded ICA. Without occlusion of any vessel, a small arteriotomy on the lateral ICA was made distal to the obvious atheromatous disease. At this point, one of three phenomena was noted: 1) the clot spontaneously expressed itself with good backflow; 2) the clot did not spontaneously exit and no backflow was observed; or 3) a small amount of clot exited with weak backflow. In the first instance, standard cross-clamping of the vessels was performed followed by an endarterectomy. In the latter two events, a No. 2 French Fogarty catheter with a 0.2-ml balloon was passed up the ICA 10 to 12 cm with very gentle pressure. The balloon was inflated with saline and gently withdrawn. This maneuver occasionally delivered loose clot and was followed by vigorous backflow; if it failed, it was repeated once or twice, after which further attempts to reopen the ICA were abandoned. If the vessel was atretic and of small caliber, indicating little chance for reopening, the operation was abandoned without passing the Fogarty catheter. When attempts to reopen the vessel failed, a clip was used to occlude the ICA from the circulation at the level of the bifurcation. This procedure, commonly referred to as "stumpectomy," ensured a smooth vascuJ. Neurosurg. / Volume 76/May, 1992
FIG. 1. Operative paradigm applied to patients with symptomatic internal carotid artery occlusion.
lar lumen from the common carotid to the external carotid artery without a blind pouch that can act as a source of embolic material. If angiographic evidence of significant external carotid artery atheromatous disease was present, an external carotid artery endarterectomy was performed. Throughout the surgical procedure, the patients were maintained with electroencephalographic (EEG) burstsuppression by means of titrated doses of pentobarbital? ~ Normotensive blood pressure was preserved, and in patients who received heparin preoperatively administration was continued throughout the operative procedure. Immediately after surgery, the patients were started on a course of aspirin. Vessel shunting was employed on one occasion when an asymmetric EEG burst-suppression pattern developed during the endarterectomy. A standardized microsurgical technique was used for endarterectomy to ensure a clean vessel wall and smoothly tapered intima proximal and distal to the atheroma dissection.~S Results
Of the 42 patients, 29 (70%) were males, and the average age was 61.4 years. Associated medical conditions included tobacco use in 67%, essential hypertension severe enough to warrant at least one instance of drug therapy in 72%, and previous myocardial infarction and/or stable angina in 26%. The clinical presentation included focal cortical transient ischemic attacks (TIA's) in 68%, amaurosis fugax in 28%, new fixed neurological deficits in 28%, and stroke-in-evolution in 9%. For patients who presented initially to our institution for treatment, the mean time from the neurological event to the first surgical procedure was 42 hours. Computerized tomography scans of the brain obtained within 4 hours of presentation demonstrated remote cerebral infarctions in 19%, established lacunar infarctions in 17%, and new changes consistent with 753
P. W. McCormick, et al.
FIG. 2. Upper:Computerized tomography scan showing infarction ipsilateral to a stenotic carotid artery lesion. Lower Left: Angiogram confirming the presence of the stenotic lesion. Lower Right: Six weeks later, the patient developed a new-onset hemiparesis due to a complete carotid occlusion. Emergency reoperation was successful, with resolution of the neurological deficit.
ischemia in 15%. Of those with new changes, four patients had watershed distribution infarcts. Over onehalf (56%) of the CT scans were normal. A total of 46 operations were performed on the 42 patients, including 24 (52%) successful vessel reopenings, nine (20%) stumpectomies with associated external carotid endarterectomy, and four (9%) stumpectomies with external carotid endarterectomies that eventually required extracranial-intracranial (EC-IC) bypass (Figs. 2 and 3). Transient surgical morbidity occurred in three patients (7 %) and included a postoperative seizure in one, vocal cord paralysis in one, and a mild contralateral 7fi4
FIG. 3. UpperLeft: Preoperative angiogram in a patient with complete internal carotid artery occlusion and a filling proximal remnant. Upper Right: Postoperative study showmg obliteration of the remnant. Transient ischemic attacks ceased after the stumpectomy and external carotid endarterectomy. Lower: Thrombotic material is seen in the proximal internal carotid artery remnant &the endarterectomized specimen (arrow).
hemiparesis in one. Permanent morbidity occurred in one patient (2%) whose left hemiparesis worsened following restoration of flow in the ICA. Postoperative TIA's occurred in four patients who initially underwent stumpectomy and external carotid endarterectomy; in accordance with the operative paradigm (Fig. 1), these patients subsequently underwent an EC-IC bypass procedure. There were no operative deaths in the series. Long-term arterial patency was assessed in 17 (73%) of the 24 patients, with successful restoration of flow. Carotid Doppler ultrasound studies were performed a mean of 28 months after surgery. In 15 of these patients (88 %), vessels were widely patent; of the other two, one had greater than 70% stenosis and one had reocclusion. The restenosed vessel was operated on again and a vein patch was placed. Three patients were lost to follow-up review; the follow-up period for the remaining 39 patients was an J. Neurosurg. / Volume 76/May, 1992
Thrornboendarterectomy for occluded carotid artery
FIG. 4. Left: Graph showing the long-term outcome for patients with successful restoration of flow in the internal carotid artery for the end-points of stroke, transient ischemic attacks, and stroke death (upper line) and the expected stroke event in patients left untreated (lower line). Right: Same analysis for patients who did not have successful reopening but who were surgically treated with stumpectomy, external carotid endarterectomy, or extracranial-intracranial bypass until their symptoms resolved (upper line) and for patients left untreated (lower line).
FIG. 6. Operative specimen of atherosclerotic plaque and organized thrombus with distal extension removed from a patient who presented with transient ischemic attacks and internal carotid artery occlusion. Measure is in millimeters.
FIG. 5. Carotid angiography in a patient with crescendo transient ischemic attacks from a left internal carotid artery occlusion. Left." Retrograde filling of the cavernous and petrous carotid artery is seen on delayed films predicting a greater likelihood of reopening of the vessel. Right." Retrograde filling at 56 seconds following injection. The patient's symptoms ceased following restoration of flow in the internal carotid artery. average of 40 months after their last operation. Five patients (13%) died, two of cancer and three of myocardial infarction. Four patients (10%) experienced neurological events: TIA's in two and vertebrobasilar insufficiency in two. Six patients (15%) have had new myocardial infarctions or new-onset angina. The outcome for patients with successful restoration of ICA flow was not significantly different from that of patients who were treated until asymptomatic with the alternative surgical strategies described. No patient has had a subsequent stroke, and the rates for neurological transient events (TIA's and vertebrobasilar insufficiency) for both groups are comparable with the expected natural history for completed stroke in nonsurgically treated patients (Fig. 4). The extent of reflux visualization in the ICA on
J. Neurosurg. / Volume 76 /May, 1992
preoperative angiograms was noted as an excellent predictor of successful vessel reopening. In 26 patients with delayed selective digital subtraction studies, retrograde reflux visualization of the occluded ICA was evaluated. Based on the Hugenholtz and Elgie grading system, v reflux to the supraclinoid segment only (Grade 2) occurred in four patients with successful reopening in one, for a success rate of 25%. Reflux to the cavernous portion of the ICA (Grades 3 and 4) occurred in eight patients with successful reopening in four, for a success rate of 50%. Reflux to the petrous ICA (Fig. 5) occurred in 14 patients with successful reopening in 10, for a success rate of 7 1%. Pathology specimens were obtained from 33 patients (nine patients had stumpectomy only), and 42% of these demonstrated acute intraplaque hemorrhage. A typical pathological specimen is shown in Fig. 6.
Discussion
Patient Selection Previous experience with surgery for symptomatic ICA occlusion has demonstrated the important effect of patient selection on the operative morbidity and mortality rates. Case series published in the 1960's 755
P. W. McCormick, et al. identified patients who did not have "profound" neurological deficits as responding much better to surgery. s~3'~9 Thompson, et at., ~9 showed a drop in the procedure mortality rate from 6.2% to 1.1% for carotid endarterectomy when patients with neurological deficits were excluded. Discussing ICA reopening, they stated: "Clinical considerations of the stroke itself are the most important factors determining operability. Patients with acute profound strokes of 12 hours' duration or more should not be operated upon because of the prohibitive mortality. Whether one should operate on such a patient if seen within six hours of onset is open to question. If the neurologic deficit is mild and is improving rapidly, or if a previously audible carotid bruit suddenly disappears, operation may be done promptly to ensure restoration of carotid flow." The influence of neurological deficit on the surgical morbidity and mortality rates associated with ICA reopening can be inferred from other recent case series. Ojemann, et al., 16 reviewed 16 cases of ICA occlusion and concluded that patients with profound deficits and decreased levels of consciousness should not be surgically treated. DeWeese 3 reported a series of 19 patients with severe progressive neurological deficits associated with carotid stenosis or occlusion; the surgical mortality rate in those patients was 37%. Meyer, et al., 12 reviewed their experience with restoration of ICA flow in patients with severe fixed deficits and decreased levels of consciousness. Although a subset of patients who benefited from the surgery could be identified, the overall surgical mortality rate in these very sick patients was 20.6%. It is not surprising, then, that case series excluding patients with severe neurological deficits show a favorable response to restoration of flow in the ICA. Two groups of authors have reported such series, both with no surgically related deaths (Table 1). 6.7 Kusunoki, et al., 9 divided their patients into two groups: those with neurological, medical, and angiographic risk factors and those without. The latter group of 14 patients had no operative mortality. Our data further strengthen the conclusion that a subgroup of patients, chosen primarily based on their presenting neurological examination, will have low surgical morbidity and mortality rates from restoration of ICA flow (Table 1). Our series excluded patients with profound neurological deficits and intracerebral
TABLE
hemorrhage visualized on CT and with documented chronic occlusion and no reflux visualization of the petrous ICA. The issue of timing of surgery remains unclear. Series where the precise timing between ICA occlusion and attempted restoration of flow is known are few. ~0.~4These include postangiographic or postsurgical occlusions that represent a pathophysiology other than spontaneous occlusion associated with atherosclerotic disease. The impression of those reporting such series is that the earlier surgery is performed, the better patients fare. Operative Technique
Certain methodological considerations concerning restoration of flow in the ICA deserve emphasis. Preoperatively, patients should be started on aspirin therapy, which can be given by suppository if necessary. Heparinization seems to be a logical medical treatment for symptomatic ICA occlusion before surgery, but no strong efficacy data are available. It is known that distal embolization can occur and is associated with a poor outcome. 12 In patients receiving heparin (partial thromboplastin time 1.5 x control), we continued heparin administration throughout the operation and withheld aspirin therapy until the heparin was stopped. During induction of anesthesia, great care must be taken to keep blood pressure normotensive. Reduction in blood pressure can decrease collateral flow to compromised vascular territories. Intravenous thiopental titrated to maintain the patient in EEG burst-suppression reduces the metabolic requirements of ischemic neurons. Careful use of this technique will protect the brain from fluctuations in collateral blood flow during surgical manipulation. 2~ It is important to monitor a parameter sensitive to cerebral oxygen delivery during the operation to assess the adequacy of collateral flow. Under barbiturate cerebral protection, this can be accomplished by observing the symmetry of the burst-suppression pattern between hemispheres or by direct assessment of cerebral oxygen saturation.~ ~,~8.20In our experience with ICA occlusion, cross-clamping the external carotid artery has caused monitoring asymmetries. When the arteriotomy is made, the clot should be given time to spontaneously decompress. Only if it fails to do so should an attempt be made to pass a Fogarty
1
Outcome of patients selected for surgical restoration of lCA flow in four series*
Authors & Year
NO.casesOf
Exclusion Criteria
Successful Morbidity Mortality Reopening
Kusunoki, et at., 1978 14 profounddeficit, major medical problem, severeangiographicdisease 1 (7) 0 NA Hugenholtz & Elgie, 1980 35 ICH,drowsy, major deficit 4 (11) 0 19 (54) Hafner & Tew, 1981 47 profounddeficit, ICH, symptoms> 45 days 0 0 32 (68) McCormick, et al., 1992 43 profounddeficit, ICH, chronic disease with poor reflux 1 (2) 0 24 (56) totals 139 6 (4.3) 0 75 (59) * All four seriesused similarexclusion criteria. Numbers in parenthesesare percentagesof the total in each study. ICA = internal carotid artery; ICH = intracerebralhemorrhage; NA = not available. 756
J. Neurosurg. / Volume 7 6 / M a y , 1992
Thromboendarterectomy for occluded carotid artery catheter. This maneuver requires experience and judgment. It should be noted that low morbidity and mortality rates are associated with a 50% to 60% successful restoration of flow (Table 1). Overvigorous attempts to reopen a vessel may well result in poor surgical outcome due to intimal injury, distal embolization, and distal dissection. Small atretic vessels should not be subjected to attempts to restore flow with a catheter. In the event that the ICA cannot be reopened and extracranial surgical procedures do not prevent subsequent neurological symptoms, surgical augmentation of cerebral collateral blood flow has been successful. Four patients in our series had anterior circulation TIA's following stumpectomy and external carotid endarterectomy and, therefore, a superficial temporal-tomiddle cerebral artery bypass was performed; one patient died 2 years later of a myocardial infarction and three remain asymptomatic an average of 37 months after surgery. This select group of patients may well have hemodynamic ischemic symptoms, which explains their response to bypass surgery. Preoperative Angiography An important observation in this study is the correlation between the extent of reflux visualization of the occluded ICA by angiegraphy and its successful reopening. Although other authors have mentioned this relationship].4.17 labeled "collateral filling," it probably does not represent true collateral filling for two reasons. There is minimal collateral supply to the petrous carotid artery and no such supply to the cervical carotid artery. The vessel fills retrograde on delayed-sequence films, and the vessel has no pressure differential across it, especially in the retrograde direction. In our opinion, the vessel fills with contrast medium in a delayed fashion as a result of refluxing of contrast medium into the distal portion, possibly aided by turbulent flow patterns in this region. Hugenholtz and Elgie7 described a five-tier classification system for grading this reflux pattern, but were unable to conclude if such a detailed classification was necessary. We have reduced the classification system to three tiers and have shown in this series that the further the refluxing of contrast medium into the vessel, the greater the likelihood of restoration of flow. Filling of the petrous vessel was associated with a 71% success rate for restoration of flow. This finding is important because it facilitates preoperative surgical planning and patient counseling. Patients with symptomatic chronic ICA occlusion and petrous vessel reflux are candidates for thromboendarterectomy. Furthermore, if a reflux pattern is not encouraging, an intraoperative decision to abort attempts at restoration of flow with a Fogarty catheter can be reached earlier. Morbidity and Mortality of Acute ICA Occlusion The natural history of spontaneous ICA occlusion associated with atherosclerotic disease has been estabJ. Neurosurg. / Voh~me 76/May, 1992
lished by prospective randomized clinical trials. Fields and Lemak 5 reported 359 patients with this diagnosis at a mean follow-up period of 44 months. New strokes occurred in 25 % of the patients; in those with complete records, 64% of the strokes were ipsilateral to the occluded ICA. Cote, et al., l followed 47 patients with ICA occlusion for 34.4 months; new strokes occurred in 23.5% of the patients and 66% were ipsilateral. Together, these studies demonstrate an annual stroke rate of 5% in the vascular territory ipsilateral to an occluded ICA. 3,5 At the time of ICA occlusion, our patients would be expected to have some associated morbidity and mortality, but the incidence is difficult to estimate. Meyer, et al.,~2 surveyed the literature and reported that acute ICA occlusion is associated with a 40% to 69% incidence of profound deficit and a 16% to 55% death rate. However, because patients with profound deficits were eliminated from our protocol, the morbidity and mortality rates associated with spontaneous occlusion alone would not be as severe as the high (42%) rate of ipsilateral events described by Nicholls, et al., 15 or that described by Meyer, et al. Surgical Outcome The permanent surgical morbidity rate for attempted ICA reopening in the 42 operations in this series was 2%, and the 30-day surgical mortality rate was 0%. These outcomes compare well with the anticipated morbidity and mortality rates of ICA occlusion even in this select group of patients. In an average followup period of 40 months, no patient with successful restoration of flow in this series has had a stroke; furthermore, no patients treated with alternative surgical approaches until their symptoms resolved have had subsequent strokes. The literature concerning thromboendarterectomy of spontaneously occluded symptomatic ICA's consistently demonstrates a subgroup of patients who have successful restoration of flow with little morbidity and mortality. This subgroup is best defined as those without severe neurological deficits, without a decreased level of consciousness, and without intracerebral hemorrhage. Clinical experience with such patients by our group and others shows a low surgical morbidity and mortality rate associated with carefully performed thromboendarterectomy. During long-term follow-up review, the vessels remained patent and the patients had fewer strokes than would be anticipated from natural history data. Our patients followed a prospectively designed diagnostic, anesthetic, and surgical protocol. Their evaluation and treatment were interdisciplinary. The data, however, were collected retrospectively and no simultaneous control group was available for comparison. This experience cannot definitively establish the efficacy of surgical management for ICA occlusion. However, based on the available data, we currently recommend surgery for patients with acute symptomatic ICA 757
P. W. McCormick, et al. occlusion who have not experienced a decreased level of consciousness, hemiplegia, or aphasia.
References 1. Cote R, Barnett HJM, Taylor DW: Internal carotid occlusion: a prospective study. Stroke 14:898-902. 1983 2. De Bakey ME, Crawford ES, Morris GC Jr, et al: Surgical considerations of occlusive disease of the innominate, carotid, subclavian, and vertebral arteries. Ann Sorg 154: 698-725, 1961 3. DeWeese JA: Management of acute strokes. Surg Clin North Am 62:467-472, 1982 4. Dyken ML, Klatte E, Kolar OJ, et al: Complete occlusion of common or internal carotid arteries. Clinical significance. Arch Neurol 30:343-346, 1974 5. Fields WS, Lemak NA: Joint study of extracranial arterial occlusion. X. Internal carotid artery occlusion. JAMA 235:2734-2738, 1976 6. Hafner CD, Tew JM: Surgical management of the totally occluded internal carotid artery: a ten-year study. Surgery 89:710-717, 1981 7. Hugenholtz H, Elgie RG: Carotid thromboendartemctomy: A reappraisal. Criteria for patient selection. J Nenrosurg 53:776-783, 1980 8. Hunter JA, Julian OC, Dye WS, et al: Emergency operation for acute cerebral ischemia due to carotid artery- obstruction. Review of 26 cases. Ann Surg 162:901-904, 1965 9. Kusunoki T, Rowed DW, Tator CH, et al: Thromboendarterectomy for total occlusion of the internal carotid artery: a reappraisal of risks, success rate and potential benefits. Stroke 9:34-38, 1978 10. Kwaan JHM, Connolly JE, Sharefkin JB: Successful management of early stroke after carotid endarterectomy. Ann Surg 190:676-678, 1979 11. McCormick PW, Stewart M, Goetting MG, et al: Regional cerebrovascular oxygen saturation measured by optical spectroscopy in humans. Stroke 22:596-602, 1991 12. Meyer FB, Sundt TM Jr, Piepgras DG, et al: Emergency carotid endarterectomy for patients with acute carotid oc-
758
clusion and profound neurological deficits. Ann Surg 203: 82-89, 1986 13. Murphey I:, Maccubbin DA: Carotid endarterectomy. A long-term follow-up study, d Neurosurg 23:156-168, 1965 14. Najafi H, Javid H, Dye WS, et al: Emergency carotid thromboendarterectomy. Surgical indications and results. Arch Surg 103:610-614, 1971 15. Nicholls SC, Bergelin R, Strandness DE: Neurologic sequelae of unilateral carotid artery occlusion: immediate and late. J Vasc Surg 10:542-548, 1989 16. Ojemann RG, Crowell RM, Roberson GH, et al: Surgical treatment of extracranial carotid occlusive disease. Clin Neurosurg 22:214-263, 1975 17. Shucart WA, Garrido E: Reopening some occluded carotid arteries. Report of four cases. J Neurosurg 45:442-446, 1976 18. Spetzler RF, Marlin N, Hadley MN, et al: Microsurgical endarterectomy under barbiturate protection: a prospective study. J Neurosurg65:63-73, 1986 19. Thompson JE, Austin DJ, Patman RD: Endartereclomy of the totally occluded carotid artery for stroke. Arch Surg 95:791-801, 1967 20. Wilkinson E, Spetzler RF, Carter LP, et al: Intraoperative barbiturate therapy during temporary vessel occlusion in man, in Spetzler RF, Carter LP. Selman WR, et al (eds): Cerebral Revascularizaliun for Stroke. New York: Thieme-Stratton, 1985, pp 397-403 21. Winslow CM, Solomon DH, Chassin MR, et al: The appropriateness of carotid endarterectomy. N Engl J Med 318:721-727, 1988 Manuscript received June 13, 1991. Accepted in final form October 25, 1991. Address for Dr. McCormick: St. Vincent's Hospital and Medical Center, Toledo, Ohio. Address for Dr. Bailes: Allegheny General Hospital, Pittsburgh, Pennsylvania. Address reprint requests to." Robert F. Spetzler, M.D., Editorial Office, Barrow Neurological Institute, 350 West Thomas Road, Phoenix, Arizona 85013-4496.
J. Neurosurg. / Volume 7 6 / M a y , 1992
