Hypothermic circulatory arrest in neurovascular surgery: evolving indications and predictors of patient outcome.

OBJECTIVE: One of the largest contemporary neurosurgical experiences with hypothermic circulatory arrest was analyzed for trends in patient selection and clinical variables affecting outcome. METHODS: During a 12-year period, 60 patients underwent 62 circulatory arrest procedures: 54 had posterior circulation aneurysms, 4 had anterior circulation aneurysms, and 2 had other lesions (hemangioblastoma and vertebral artery atherosclerosis). RESULTS: The surgical morbidity and mortality rates associated with hypothermic circulatory arrest were 13.3 and 8.3%, respectively. At late follow-up, 76% of the patients had good outcomes (Glasgow Outcome Scale scores of 1 and 2), 5% had poor outcomes (Glasgow Outcome Scale scores of 3 and 4), and 18% had died. After 1992, circulatory arrest was limited to posterior circulation aneurysms and included increasing numbers of basilar trunk aneurysms. Patient outcome correlated with preoperative neurological condition (admission Glasgow Coma Scale score, P < 0.001; Hunt and Hess grade, P = 0.037; and age, P = 0.007). Preservation of perforating arteries was paramount to achieving a good outcome (P = 0.005); duration of circulatory arrest was not. CONCLUSION: Current indications for hypothermic circulatory arrest include only giant and complex posterior circulation aneurysms that cannot be treated using conventional techniques or that recur after endovascular coiling. Surgical morbidity and mortality rates reflect the increasing complexity of the aneurysms treated but are still more favorable than the natural history of these lesions. This experience demonstrates that management in specialized neurovascular centers can minimize the morbidity associated with circulatory arrest so that it remains a viable treatment option for complex posterior circulation aneurysms.

Aneurysms of the basilar artery treated with circulatory arrest, hypothermia, and barbiturate cerebral protection.

Complete circulatory arrest, deep hypothermia, and barbiturate cerebral protection are efficacious adjuncts in the surgical treatment of selected giant intracranial aneurysms. These techniques were utilized in seven patients, one with a large and six with giant basilar artery aneurysms; four had excellent results, one had a good result, one had a fair outcome, and one died. The rationale for the use of complete cardiac arrest with extracorporeal circulation, hypothermia, and barbiturate cerebral protection is outlined. The surgical and anesthetic considerations are reviewed. The perioperative morbidity and long-term results support the use of these techniques in selected patients with complex intracranial vascular lesions.

Surgical management of large AVM's by staged embolization and operative excision.

A series of 20 patients with giant arteriovenous malformations (AVM's) managed with staged embolization and surgical resection is presented. Complete excision was accomplished in 18 of these patients. There were no deaths and only three complications, of which one was disabling. Further evidence for the presence of low perfusion surrounding the AVM, emphasizing the risk of normal perfusion pressure breakthrough, is provided by cortical perfusion pressure, cortical cerebral blood flow (CBF), and stable xenon computerized tomography CBF measurements. The staged approach to giant AVM management is a proposed method to render AVM's that were previously considered inoperable or marginally operable into totally excisable lesions, while maintaining an acceptable level of morbidity and mortality.

Microsurgical endarterectomy under barbiturate protection: a prospective study.

Carotid endarterectomy has the potential to improve on the natural history of untreated carotid artery disease with respect to subsequent infarction in symptomatic patients with causative angiographic lesions. This benefit of a reduced risk of stroke can be realized only if the perioperative morbidity and mortality rates are kept low. An approach to symptomatic carotid artery bifurcation disease is outlined, with a defined protocol of microsurgical endarterectomy utilizing barbiturate protection during the period of potential focal temporary cerebral ischemia. This protocol includes preoperative antiplatelet therapy, barbiturate anesthesia, the avoidance of an internal shunt, the use of the operating microscope, and strict control of postoperative hypertension. A series of 200 consecutive endarterectomies performed within this protocol in 180 patients and the resultant combined permanent morbidity and mortality rate of 1.5% are reported.

Postoperative paraplegia with preserved intraoperative somatosensory evoked potentials. Case report.

Intraoperative somatosensory evoked potentials (SSEP's) are being used with increasing frequency to monitor neurological function during spinal surgery. The authors report a case of postoperative paraplegia that occurred despite preserved intraoperative SSEP's in an achondroplastic dwarf who underwent correction of a congenital kyphoscoliosis. Surgeons and anesthesiologists involved with SSEP monitoring should be aware that false-negative results may occur with this technique.

Somatosensory evoked potentials and cortical blood flow during craniotomy for vascular disease.

It has been demonstrated that central conduction time (CCT) is slowed and that attenuation of cortical potentials occurs with reduced cerebral perfusion. During 11 craniotomies for aneurysm, arteriovenous malformation (AVM), or extracranial-intracranial (EC-IC) bypass, we continuously monitored somatosensory evoked potentials (SSEPs) and regional cortical blood flow (rCBF) as determined by a thermal flow probe. The CCT was calculated and correlated with the rCBF. In 8 of the 11 cases, the rCBF values varied within 1 SD of normal values derived from 25 measurements of nonischemic cerebral cortex. All initial CCT values were within 1 SD of normal, but 4 of 11 patients had a prolonged CCT intraoperatively. Three of these were associated with a low rCBF (14 to 31 ml/100 g/minute). One patient had postoperative confusion at the time of discharge. One patient who underwent AVM embolization had a permanent loss of SSEPs postoperatively, and his preexisting hemiparesis was more profound after operation. Finally, 1 patient's CCT improved after EC-IC bypass. This is a preliminary study that demonstrates the feasibility of monitoring CCT and rCBF during craniotomy. When rCBF values fall and the CCT slows, neurological deficit will probably occur.

Surgical treatment of an abscess in a Rathke's cleft cyst.

An abscess in a Rathke's cleft cyst was surgically treated in a 39-year-old man. The patient presented with headaches, fever, and visual deficits. Transcranial decompression of the optic chiasm was carried out first. The abscess recurred, however, and drainage of the abscess and removal of its wall via the transsphenoidal route was carried out 4 weeks later. Visual evoked responses were useful in the successful management of this lesion. The patient regained normal pituitary function and visual function after the operation.

Intraoperative monitoring of brain-stem auditory evoked potentials.

Intraoperative brain-stem auditory evoked potentials (BAEP's) were monitored in 46 patients undergoing intracranial surgery for a variety of pathological conditions to determine whether this technique was capable of providing useful information to the operating surgeon. Intraoperative BAEP's were unchanged throughout surgery in 34 patients (74%), and these individuals had no postoperative hearing deficits. Four patients (9%) developed an abrupt ipsilateral loss of all waveform components beyond Wave I and had postoperative evidence of a pronounced hearing loss in the affected ear. An additional patient demonstrated BAEP loss contralateral to the side of surgery, and this was associated with subsequent signs of severe brain-stem dysfunction. Seven patients (15%) developed intraoperative delays of BAEP waveform latency values, but maintained recognizable waveforms beyond Wave I. Postoperatively, their hearing was either normal or mildly impaired, and there were no indications of other brain-stem abnormalities. This group represents the individuals who may have been benefited by evoked potential monitoring, since corrective surgical measures were taken when latency delays were observed. Intraoperative BAEP's can be reliably and routinely recorded in an operating room environment. They provide a good predictor of postoperative auditory status, and may have prevented permanent neurological deficits in a small segment of patients by alerting the surgeon to potentially reversible abnormalities.

Monitoring of intraoperative auditory brain stem responses.

We believe that our initial experience establishes the fact that ABRs can be routinely and reliably performed in an operating room environment. There was no added risk to the patient, and operative delays were minimal. We did note transient fluctuation in latency values up to 1.5 msec. These changes would revert to baseline levels within five to ten minutes. Changes noted during drilling were probably related to the random noise produced. Case 3 was worrisome in that hearing was lost after maintenance of the ABRs during the surgical procedure. Evidently the vestibule was damaged, and a labyrinthitis caused the hearing loss. We feel that this procedure will be most useful in those situations in which the cochlear nerve and blood vessels are at risk. Such procedures as acoustic tumor removal with attempts to preserve hearing, vestibular nerve sections, and facial nerve problems in the IAC should be routinely monitored. Further experience will, of course, be most helpful in explaining and recognizing these changes.

Intraoperative monitoring of evoked potentials.

Sensory EPs were recorded intraoperatively in 173 neurosurgical procedures (71 VEPs, 66 BAEPs, and 31 SSEPs) to evaluate the utility of this technique. EPs could be safely recorded in all cases, but the yield of useful results varied with each sensory modality. BAEPs were recorded reliably in 100% of the cases and intraoperative latency changes accurately predicted postoperative hearing deficits in 10%. Potential hearing deficits were detected in another 15%. BAEP changes were associated with brainstem dysfunction in only one case. VEP changes were difficult to interpret intraoperatively because of contamination by a high degree of variability and both false negative and false positive results. Changes in VEP amplitudes related to surgical manipulation of the optic chiasm were only suggested. SSEP changes were recorded reliably in only 75% of the cases and no correlations between SSEP changes and postoperative sensory function were established. Again, intraoperative amplitude attenuation of the SSEP waveform with surgical manipulation only suggested a potential sensory deficit. Intraoperative EP monitoring is a valuable technique that provides a functional analysis of the sensory nervous system during surgical procedures. Specific sensory stimuli and improved data analysis will increase the utility of this CNS monitor.