Risk of Ischemic Stroke in Patients With Symptomatic Vertebrobasilar Stenosis Undergoing Surgical Procedures
Background and Purpose— There is little information to provide an estimate for stroke risk in patients with established stenosis or occlusion in the basilar or intracranial vertebral arteries undergoing surgical procedures. The objective of this study was to determine the ischemic stroke risk in this specific patient population.
Methods— A medical records linkage system retrospectively identified patients with a diagnosis of symptomatic vertebrobasilar stenosis or occlusion matched with surgical procedures. Patients were selected if they had stenosis or occlusion of the basilar or intracranial vertebral arteries identified on vascular imaging before undergoing surgical procedures under general anesthesia. Clinical and radiographic features were reviewed, along with the nature of the surgeries and details of the perioperative management. Records were reviewed for the diagnosis of stroke occurring within 1 month of surgery.
Results— Thirty-eight patients with a history of symptomatic vertebrobasilar ischemia underwent 50 operations under general anesthesia, and 3 had ischemic strokes in the vertebrobasilar territory immediately after surgery, a per-procedure rate of 6.0% (95% confidence interval, 1.2 to 16.6) All 3 had episodes of prolonged hypotension (systolic blood pressure <100 mm Hg for >10 minutes) during surgery.
Conclusions— The risk of perioperative stroke in patients with vertebrobasilar stenosis undergoing surgery under general anesthesia is 6.0%, which is notably higher than the risk for patients with other patterns of cerebrovascular disease.
The assessment of patients with cerebrovascular disease before surgical procedures is a well-recognized neurological consultation. Patients and their surgeons may wish to know an estimate of the stroke risk associated with various procedures and if specific preventive measures might be undertaken to reduce this risk.
There are robust data on the stroke risk associated with general,1–3 cardiovascular,4–7 and carotid surgery.8–10 Additionally, there are data regarding perioperative stroke risk in the specific scenarios of having asymptomatic carotid bruits,11 carotid occlusion, and stenosis of varying degrees.5,12–15 The overall risk of stroke after general surgical procedures is exceedingly low (0.08%1 and 0.2%2), but in patients with a prior history of stroke, the risk is substantially higher (2.9%16), and this has been consistently shown to be a predictor of postoperative stroke.1–3,13 Other predictors include peripheral vascular3,5,12 and chronic obstructive lung disease,3 hypertension,1 smoking,1 and abnormal ECG rhythm.1
Patients with asymptomatic carotid bruits do not seem to carry a higher risk for stroke11; however, those with bruits or a history of stroke or transient ischemic attack and a documented stenosis of >50% may have a risk as high as 3.6%.13 The risk of stroke after carotid endarterectomy ranged from 2.5%8 to 5.8%9 and 7.0%10 in 3 large trials of this procedure. Contralateral internal carotid occlusion increases the perioperative endarterectomy stroke risk.17
The stroke risk associated with coronary bypass grafting (CABG) ranges from 2.2%5 to 5.2%6 and is highest when concurrent valve surgery or other intracardiac procedures are performed.7 Ultrasound4,5,12,14 and angiographic15 studies of the carotid circulation in patients undergoing cardiac procedures have shown variable risk rates for perioperative stroke. Some early studies18 have not shown severe carotid stenosis to be a risk for stroke after cardiac surgery, whereas others have confirmed an association with increasing degrees of stenosis.4,5,14 Similar to the carotid endarterectomy data, carotid occlusion appears to have a particularly high risk for stroke after CABG, up to 27.3% in 1 series.5
There is less information pertaining to perioperative stroke in the posterior circulation, although some studies have suggested that neck positioning is a contributing factor.19,20 We are not aware of any studies that specifically provide a stroke risk estimate for patients with established vertebrobasilar occlusive disease who are to have surgery. As noninvasive imaging techniques such as MR angiography (MRA) improve, the threshold to perform these studies is lowered, and patients of increasing age undergo more surgical procedures, it is likely that cerebrovascular clinicians will be increasingly asked to evaluate patients in this situation.
The objective of this study was to determine the risk of posterior circulation ischemic stroke in patients with known disease in the basilar or intracranial vertebral arteries who undergo surgical procedures under general anesthesia.
The Mayo Clinic medical records linkage system was used to identify a population of patients coded with the diagnosis of vertebrobasilar stenosis or occlusion matched with any surgical or invasive diagnostic procedure between 1995 and 2002. Medical records were then reviewed to identify patients with symptomatic stenosis or occlusion within the intracranial vertebral or basilar artery documented by vascular imaging (cerebral angiogram, MRA, or CT angiogram) who underwent any form of surgical or interventional diagnostic procedure after the vascular imaging study. Only patients >18 years of age undergoing general anesthesia were selected to form the study population. This population included patients referred from other centers, but all surgeries were performed at 1 of the 2 Mayo-affiliated hospitals in Rochester (Minn).
From eligible patients, clinical notes, radiology reports, operative notes, and neurological consultations were reviewed in detail. Demographic data abstracted included patient sex, age, and cerebrovascular risk factors. Clinical data were reviewed for prior vertebrobasilar ischemic symptoms. Records were reviewed for the diagnosis of ischemic stroke occurring within 1 month of the procedure. If present, the nature of symptoms of vertebrobasilar ischemia was recorded. From the operative and anesthesia records, the date and nature of the surgery or procedure, baseline blood pressure recorded on the anesthetic chart, lowest blood pressure on the anesthetic chart, and presence of sustained hypotension (systolic blood pressure [SBP] <100 mm Hg for >10 minutes) were noted. Records were reviewed for follow-up to the last entry or through the end of 2002 (whichever was latest), specifically looking for the diagnosis of recurrent stroke or TIA or surgery performed outside the study center.
Radiographic imaging included conventional arteriography, MRA, and CT angiogram. All studies were reviewed by consultant neuroradiologists, and these reports were reviewed for details of the site and degree of the stenosis, as well as the presence of any associated vertebral or basilar lesion. The sites were divided into intracranial vertebral artery (ICVA), vertebrobasilar junction, or proximal, mid, or distal basilar artery. The degree of stenosis was graded as severe (>60%) or moderate (40% to 60%).
From the Mayo Clinic database, there were 655 matches for the diagnosis of vertebrobasilar stenosis and any surgical or invasive diagnostic procedure. Eight patients were excluded for miscellaneous reasons (5 patients had missing charts, 2 patients had “basilic vein thrombosis,” and 1 patient was <18 years of age). The term “vertebrobasilar stenosis” was frequently interchanged by clinicians for the clinical diagnosis of “vertebrobasilar insufficiency” or “posterior circulation” TIA. It was clear that in many patients this diagnosis was tentative and was listed as part of a differential diagnosis or as a condition that required exclusion. Two hundred forty-four patients were given this diagnosis and did not have vascular imaging. Thus, 403 patients underwent appropriate imaging.
Imaging failed to show arterial stenosis or occlusion in 267 patients. Among 136 patients with vertebrobasilar stenosis confirmed by imaging, 48 did not have a procedure after the diagnosis. Forty-five patients underwent 75 minor surgical or diagnostic procedures with local or spinal anesthesia or with sedation; none developed a periprocedural stroke. Five patients had asymptomatic vertebrobasilar stenosis. This left 38 symptomatic patients who had 50 procedures under general anesthesia to form the study population.
The clinical and radiographic characteristics of these patients are outlined in Table 1, and the details of the operative procedures are shown in Table 2. Of the 38 patients who underwent 50 procedures, 28 had single operations, 9 had 2 operations, and 1 patient had 4 operations. Eighteen patients also had 29 minor surgical or diagnostic procedures that did not require general anesthesia during the study period. All patients were taking antiplatelet, anticoagulants, or both at baseline, although only a minority (18 of 50) received perioperative heparin. BP recordings were available during most of the operations (43 of 50), with transient hypotension (SBP <100 mm Hg) common (22 of 43) but sustained hypotension (SBP <100 mm Hg for >10 minutes) uncommon (6 of 43).
Long-term follow-up (mean time, 19 months) was available for all of 35 patients without stroke but only 1 of the 3 stroke patients. Twenty-three were seen by neurologists or neurosurgeons after the index surgical procedure. Three patients were found to have strokes within 1 month of surgery. Seven patients had cerebrovascular events occurring >1 month after surgery: 5 transient ischemic attacks (2 vertebrobasilar, 1 carotid, 2 undefined vascular territory), 1 middle cerebral artery territory stroke, and 1 intraventricular hemorrhage. None of these events occurred within a month of another surgical procedure.
Patients With Postoperative Stroke
The 3 stroke patients were diagnosed with stroke in the immediate postoperative period (on awakening). A consultant neurologist was involved in the diagnosis of all cases, and 2 had confirmatory imaging findings. One patient had a brainstem ischemic syndrome diagnosed by one of the authors (K.D.F.) with no obvious infarct on CT scan but did not have an MRI performed. One further patient had hemorrhagic basal ganglia stroke 2 days after a carotid endarterectomy that was thought to be a reperfusion syndrome. This patient was not considered to have had a stroke related to vertebrobasilar stenosis or to the surgery and anesthetic. Thus, the perioperative stroke risk on a per-procedure basis was 3 of 50 or 6.0% (95% confidence interval, 1.2 to 16.6).
This 71-year-old man with hypertension and hypercholesterolemia had a cerebellar stroke in 1997. MRA at that time demonstrated moderate stenosis of the proximal basilar artery, 2 short segments of ≈70% stenosis in the right ICVA, and poor flow in the left ICVA (Figure 1). There were no residual neurological deficit and no recurrent cerebral ischemic symptoms. At a preoperative evaluation 5 years later, a carotid ultrasound was negative, and his aspirin was discontinued 8 days before elective revision of a right hip prosthesis under general anesthesia. During the 520-minute procedure, there was protracted hypotension with an SBP <80 mm Hg for ≈40 minutes and <100 mm Hg for 150 minutes. Postoperatively, he was somnolent and poorly responsive and was transferred to an intensive care unit. The following day, he was able to follow commands but was agitated and complained of blurred vision. Neurological examination revealed dysarthric speech, bilateral gazed evoked nystagmus, and bilateral appendicular ataxia.
MRI (Figure 2) showed acute infarcts in the cerebellar hemispheres and occipital cortices bilaterally. MRA showed 50% stenosis of the proximal basilar artery (unchanged compared with 1997), along with stenosis of the intracranial vertebral arteries measuring 60% on the left and 90% on the right, which appeared to have progressed since 1997. A transesophageal echocardiogram (TEE) showed no cardioembolic source.
This 70-year-old man with hypertension and hypercholesterolemia presented with vertebrobasilar TIAs in 1992, and a cerebral angiogram showed 95% stenosis of the proximal basilar artery. There was mild stenosis of the intracranial left vertebral artery and a large right posterior communicating artery and no significant stenosis within the anterior circulation. MRI showed a small right cerebellar infarct and a pontine lacune. Anticoagulation was begun. He had no cerebrovascular symptoms until July 1998, when anticoagulation had been temporarily withheld for a coronary angiogram, and he had a vertebrobasilar TIA. In late 2000 (at 79 years of age), he developed exertional angina that became unstable in January 2001. At this time, examination by an author (K.D.F.) was normal. MRA showed findings similar to the digital subtraction angiography of 1992 with no signs of progressive stenosis. Heparin was substituted for warfarin, and the patient underwent general anesthesia for a 4-vessel CABG and aortic valve replacement (Carpentier Edwards bioprosthesis). The aorta was cross clamped for 114 minutes, and cardiopulmonary bypass was maintained for 142 minutes. There was a 10-minute period with SBP <100 mm Hg, and 2 other separate recordings showed 80/55 mm Hg. Intraoperative TEE showed no significant aortic atheroma.
Postoperatively, he was slow to wake, and a right hemiparesis was noted. Repeat neurological examination (K.D.F.) showed a moderate (4 to 5) hemiparesis and a left internuclear ophthalmoplegia suggestive of a left pontine lesion. CT showed the old cerebellar infarction but no new lesion.
In September 1994, a 67-year-old male smoker had transient left hemiparesis and dysarthria that was thought to represent a posterior circulation TIA. MRI/MRA performed elsewhere (but reviewed at this institution) showed no discrete infarct but possible occlusion or very slow flow in the basilar artery, patent posterior communicating arteries, and no stenosis within the intracranial anterior circulation. Anticoagulant and antiplatelet agents were administered, and no further TIAs occurred.
Extracranial Doppler ultrasound in January 1996 also showed no significant stenosis. In that month, warfarin was changed to heparin, and he underwent general anesthesia for a 2-vessel bypass and aortic and mitral valve replacements (St Jude’s). Intraoperative TEE described the valve lesions but made no mention of any aortic arch atheroma. Total bypass time was 404 minutes. An intra-aortic balloon pump was also required. SBP was <100 mm Hg for 137 minutes before bypass.
The balloon pump was removed after 4 days. Postoperatively, there was a dense left hemiparesis, left homonymous hemianopia, and a new left extensor plantar. CT scan showed infarction of the right posterior cerebral artery territory involving the right temporal and occipital lobes, as well as some involvement of the posterior limb of the internal capsule.
Our study demonstrates a 6.0% perioperative stroke risk in patients with basilar or ICVA stenosis undergoing a variety of surgical procedures under general anesthesia. All 3 strokes were diagnosed in the immediate postoperative period, which increases the likelihood that they were related to the surgery. In other series,1,2 strokes as long as 36 days1 after the index surgery were deemed postoperative. Although our study is too small to have the statistical power to make meaningful comparisons between stroke and nonstroke patients, we observed 2 interesting clinical features. Most notably, all 3 patients had prolonged intraoperative hypotension. This was noted in only 3 of 47 procedures without stroke. In addition, all 3 stroke patients had evidence of combined ICVA and basilar disease compared with 9 of 35 patients without stroke. This may suggest that hypoperfusion is an important and potentially preventable factor.
The relationship of stroke to vertebrobasilar occlusive lesions is of substantial interest. Although all 3 patients had documented intraoperative hypotension, none had a typical border-zone infarction on imaging. Patient 1 was unlikely to have had a cardiac source of embolism, given that he had noncardiac surgery and a negative transesophageal echocardiogram. In patients 2 and 3, a cardiac source of embolism was clearly possible, although patient 2 had a syndrome typically described in association with basilar artery stenosis or occlusion.21 Because intracardiac procedures have been associated with stroke rates as high as 5.9%,7 this factor may have skewed our results. Other cardiac series have suggested that embolization rather than hypoperfusion is the predominant stroke mechanism,22 further minimizing the role of large-artery occlusive disease in this setting. Limiting our study to patients undergoing nonvascular procedures may have made the link between stroke and vertebrobasilar disease more certain but less useful clinically because patients undergoing carotid endarterectomy17 and cardiac procedures6 formed the majority of patients. This is not surprising, given the association between craniocervical large-artery intracranial and extracranial disease and coronary artery disease.23
Another proposed mechanism of perioperative stroke in patients with vertebrobasilar disease is hemodynamic stress related to intubation and neck positioning. An MRA study19 of the vertebrobasilar circulation during neck extension that simulates the position for endotracheal intubation demonstrated a reduction in flow. A clinical series20 describing brainstem and cerebellar infarcts after noncardiac and nonvascular surgeries proposed that neck positioning may have an important role by leading to thrombus formation in compressed extracranial vertebral arteries, which later moves intracranially when neck motion become free. Such factors might further exacerbate tenuous posterior circulation flow resulting from established ICVA or basilar stenosis.
Our study is the first to provide an estimate of the perioperative risk of ischemic stroke (6.0%) in patients with documented vertebrobasilar stenosis or occlusion undergoing surgical procedures under general anesthesia. It suggests a risk twice as high as in patients with significant occlusive carotid artery disease.13 Marked hypotension was evident in all patients who awoke with stroke. This could imply that aggressive maintenance of intraoperative BP is warranted in patients with symptomatic vertebrobasilar occlusive disease.
- Received March 25, 2003.
- Revision received June 18, 2003.
- Accepted July 2, 2003.
Limburg M, Wijdicks E, Li H. Ischemic stroke after surgical procedures: clinical features, neuroimaging, and risk factors. Neurology. 1998; 50: 895–901.
Dashe J, Pessin M, Murphy R, Payne D. Carotid occlusive disease and stroke risk in coronary artery bypass surgery. Neurology. 1997; 49: 678–686.
Breuer A, Furlan A, Hanson M, Lederman R, Loop F, Cosgrove D, Greenstreet R, Estafanous G. Central nervous system complications of coronary artery bypass graft surgery: prospective analysis of 421 patients. Stroke. 1983; 14: 682–687.
Wolman R, Kanchuger M, Newman M, Roach G, Nussmeier N. Adverse neurological outcome following cardiac surgery. Anesth Analg. 1994; 78: S483.Abstract.
North American Symptomatic Carotid Endarterectomy Trial Collaborators. Benefit of carotid endarterectomy in patients with symptomatic moderate or severe stenosis. N Engl J Med. 1998; 339: 1414–1425.
Evans B, Wijdicks E. High-grade carotid stenosis detected before general surgery: is endarterectomy indicated? Neurology. 2001; 57: 1328–1330.
Furlan A, Craciun A. Risk of stroke during coronary artery bypass graft surgery in patients with internal carotid artery disease documented by angiography. Stroke. 1985; 16: 797–799.
Gasecki A, Eliasziw M, Ferguson G, Hachinski V, Barnett H, for the North American Symptomatic Carotid Endarterectomy Trial (NASCET). Long term prognosis and effect of endarterectomy in patients with symptomatic severe carotid stenosis and contralateral stenosis or occlusion: results from NASCET. J Neurosurg. 1995; 83: 778–782.
Weintraub M, Khoury A. Cerebral hemodynamic changes induced by simulates tracheal intubation: a possible role in perioperative stroke? Stroke. 1998; 29: 1644–1649.
Tattenborn B, Caplan L, Sloan M, Estol C, Pessin M, DeWitt L, Haley C, Price T. Postoperative brainstem and cerebellar infarcts. Neurology. 1993; 43: 471–477.
Amarenco P, Caplan L, Pessin M. Vertebrobasilar occlusive disease. In: Barnett H, Mohr J, Stein B, Yatsu F. Stroke: Pathophysiology, Diagnosis and Management. 3rd ed. Philadelphia, Pa: Churchill Livingstone; 1998:chap 2.
Hise J, Nipper M, Schnitker J. Stroke associated with coronary artery bypass surgery. Am J Neuroradiol. 1991; 12: 811–814.
Chimowitz M, Poole R, Starling MR, Scwaiger M, Gross MD. Frequency and severity of asymptomatic coronary artery disease in patients with different causes of stroke. Stroke. 1997; 38: 941–945.