Neurological Aspects of Cardiovascular Surgery

Advances in Vertebral Artery Surgery at the Skull Base

Andrew L. Carney, M.D.,* Evelyn M. Anderson, M.D.,*and Daniel M. Martinez, M.D.t

Between July 1975 and May 1985, 219 procedures were performedon the vertebral system; two were intracranial, 118 were at the skullbase and 99 were at the neck base. Of the 108 patients operated on atthe skull base, procedures were multiple in five and bilateral in two.

Bypass to the vertebral artery between the transverse process ofCland C2 was performed 91 times. The blood supply for the bypass wasthe common carotid (70), the external carotid (9), the internal carotid(9), the subclavian (3), and the occipital artery (3). Decompression wasperformed in three patients, segmental resection in seven, and ligationin two.The primary objective was to increase the flow capacity of the

vertebral artery. This flow to the Circle of Willis supplies the fore-brain and the hindbrain if the internal carotid artery is obstructedand the posterior communicating artery is patent, or it may supplyonly the hindbrain when no communication exists. The pathologicprocesses include atherosclerosis, thrombosis, dissection, compres-sion, and vasospasm.

The male:female ratio is 2:3. Long-term mortality is 19%for the maleand 4.6% for the female. Hindbrain symptoms in the male predatescardiac symptoms by 2 to 4 years. Operative mortality and stroke rateis less than 3% combined. Long-term graft patency is 87%.

Vertebral artery surgery at the skull base produces results superiorto thosefollowing proximal segment reconstruction when measured bydynamic computerized tomographic scanning, neurofunctional testing,and symptom relief. Success following reconstruction depends oncareful patient selection and surgical expertise. The techniques ofreconstruction are well established and results have been durable forover a decade.

DYNAMIC COMPUTERIZED tomo- This type of scanning permits a hemodynamicgraphic scanning and vertebral artery approach to brain ischemia because its diag-

surgery at the skull base are major develop- nostic capability is not limited to the cerebralments in the field of neurovascular surgery. hemispheres, but extends into the brainstem

From *Mercy Hospital and the University ofIllinois, Chicago; and tPresbyterian Hospital,Southwest Medical School, Dallas, Texas.

Address for reprints: Andrew L. Carney, M.D., Box 2009, 222 Forest Avenue, Oak Park,Illinois 60302.

lexas Heart Institute Journal 83

and the cerebellum. The hindbrain is sensory,homeostatic and works in conjunction with theorgans of special sense. The ability to distin-quish central ischemia from peripheral dys-function is critical. Cortical blindness andcentral deafness can be distinguished fromend-organ dysfunction. 1

Vertebral artery surgery at the skull baseaddresses the complexity of anatomy, path-ology, and motion at the skull base and recog-nizes the importance of the vertebral artery tobrain hemodynamics.2'3 Vertebral arterysurgery at the skull base has produced resultssuperior to that performed on more proximalsegments when measured by the criteria ofdynamic computerized tomography, neuro-functional testing, and symptom relief.Success following reconstruction depends oncareful case selection and surgical expertise.The techniques of vertebral artery reconstruc-tion at the skull base are established and theresults are durable.4

PATIENTS AND METHODS

Between July 1975 and May 1985, 219procedures were performed on the vertebralsystem: two were intracranial, 118 were at theskull base, and 99 were at the neck base. Ofthe 108 patients operated on at the skull base,procedures were multiple in five and bilateralin two.

Bypass to the vertebral artery between thetransverse process of Cl and C2 was per-formed 91 times. The blood supply for thebypass was the common carotid artery in 70cases, the external carotid artery in nine cases,the internal carotid artery in nine cases, thesubclavian artery in three cases, and theoccipital artery in three cases. Decompressionwas performed as the sole procedure in threecases, and segmental resection with end-to-end anastomosis in seven cases. The vertebralartery was ligated without reconstruction intwo cases because of dissection. Sympatheticdenervation was accomplished in two cases:(1) by superior cervical ganglionectomy, and(2) by perivascular denervation. 'The male to female ratio was 2:3. Long-

term mortality was 19% in the males and 4.6%in the females. Hindbrain symptoms in themale predated cardiac symptoms by 2 to 4

years. Operative mortality and stroke ratewere less then 3% when combined. Long-termgraft patency was 87%.

CASE REPORTS

Case 1Common Carotid to Distal VertebralArtery Bypass5On August 29, 1977, a 77-year-old man

presented with transient right hemiparesis,confusion, slurred speech and the inability toswallow. A computerized tomographic scanrevealed a watershed infarction in the leftparietal-occipital region. Angiography dis-closed total occlusion of the left internalcarotid artery. Injection of the dominant leftvertebral artery showed sharply-angleddynamic obstructions at the origin of theproximal segment within the cervical spine,and at the level of C2. The right internalcarotid artery did not supply the forebrain onthe left side.On October 5, 1977, a vein bypass from the

common carotid artery to the distal vertebralartery between Cl and C2 was performed.This bypass resulted in increased flow andpressure to the forebrain and hindbrain byincreasing the flow capacity of the vertebralartery and the posterior communicating artery.Comparison of his preoperative and postopera-tive angiograms revealed marked increase inthe number and size of the visualized second-ary vessels. Despite dramatic improvement,the patient died within a year of coronaryartery disease.

CommentCarotid obstruction proximal to the circle of

Willis will result in the diversion ofblood flowfrom the hindbrain into the forebrain if theposterior communicating artery is large.6 Ifthe vascular bed exceeds the capacity of thevertebral artery, though its flow be great, theischemia will be maximal in the parietal-occipital watershed. If the carotid is stenotic,endarterectomy will relieve symptoms byrestoring normal flow to the carotid artery andreducing the demand on the vertebral system.

However, if the internal carotid artery isoccluded, then the vertebral artery must

Vol. 13, No. 1, March, 198684

supply an overextended vascular bed, theforebrain and hindbrain. Only by increasingthe flow capacity of the vertebral artery can theblood flow be increased and improve brainperfusion. Dynamic arterial obstructionsdecrease flow capacity. Bypass to the verte-bral artery at the skull base increases flowcapacity (Fig. 1).

Fig. 1 Brain hemodynamics. The circle of Willis,including the vertebral arteries, is the anatomical basisfor competitive flow within the system. McDonaldand Potter3 define the "deadpoint" as the point of noflow where pressures are balanced. Occlusion of asingle vessel results in the rapid redistribution ofbloodflow within the circle of Willis and may affect avascular bed2 remote from the occluded vessel.(Reprinted from Carney AL: Advances in Neurology,by permission of Raven Press, New York, N.Y.).

Case 2Subclavian to Distal VertebralArtery Bypass7

A 65-year-old male, who was a retiredcement worker, presented with dramatic lossof strength precipitated by head rotation.Noninvasive evaluation revealed a markedreduction in right carotid blood flow. Angi-ography disclosed total occlusion of the nghtexternal carotid artery, but no apparentobstruction of the common or internal carotid

artery. Left carotid compression was nottolerated. There were dynamic obstructions ofthe dominant left vertebral artery.On May 5, 1978, a vein bypass was per-

formed between the subclavian artery and thevertebral artery between Cl and C2 on the leftside. The subclavian artery was selected bynecessity. Left carotid compression was nottolerated. Subsequently, the patient wasdramatically relieved of his symptoms.

Postoperatively, dynamic computerizedtomographic scanning was done with thebypass occluded and open. This revealed thatthe contralateral right frontal lobe wasischemic when the bypass was closed, andperfusion was normal when the subclavian:distal vertebral artery bypass was open. Highflow through the left vertebral artery wasnecessary to restore perfusion of the rightfrontal lobe to normal.The patient died 4 years later in 1982 from

progressive pulmonary fibrosis.

CommentObstruction of a single pre-circle artery,

internal carotid artery or vertebral artery mayresult in ischemia in a portion of the brain thatis well tolerated. When flow is further reducedby obstruction of a second pre-circle vessel,the impact can be catastrophic.

In this patient, the right carotid angiogramrevealed no apparent obstruction of the inter-nal carotid artery, but the blood flow wasreduced by noninvasive testing, and the rightfrontal lobe was ischemic. Occlusion ofneither the left common carotid artery nor theleft vertebral artery could be tolerated. Post-operative angiograms disclosed filling of thedistal external carotid artery segment on theright side. Thus the left vertebral artery nowcontributed to the perfusion of the right frontallobe and increased the flow to the right exter-nal carotid artery (Fig. 2).

Case 3Segmental Resection at C1-C2

A 39-year-old woman presented with recur-ring syncope precipitated by head hyperexten-sion. The carotid arteries were normal. Bothvertebral arteries were extremely redundantand acutely angled at the level ofC2, but muchworse on the right.

ftxas Heart Institute Journal 8s

A B C

Flg. 2 Bypass to the distal vertebral artery. (A) Common carotid takeoff requires limited dissection and pennitsunencumbered access to the carotid artery if required at a later date. (B) Internal carotid takeoff can utilize apreviously occluded vessel, and the bypass is short and direct. (C) Subclavian artery takeoff avoids clamping ofthe carotid artery if not tolerated.(Reprinted from Camey AL: Advances in Neurology, by permission of Raven Press, New York, N.Y.)

On July 31, 1975, a segmental resection ofthe vertebral artery between Cl and C2 wasperformed with primary end-to-end repair.Her symptoms have been relieved for the 10years of follow-up without requiring contra-lateral surgery.

CommentHead motion at the skull base has a tremen-

dous range. Stressed head positions normallyresult in obstruction of one or two vessels. Ifcardiac output is adequate, if the circle ofWillis is complete, and if the primary vesselsto the circle of Willis are patent, there are nosymptoms. If the flow capacity of patent

vessels is decreased, the dynamic occlusion ofthe vertebral artery can be catastrophic. Headrotation primarily compromises vertebralartery flow between Cl and C2, but it mayoccur within the cervical spine. Head exten-sion can result in obstruction in the proximalsegment within the cervical spine, and pos-teriorly between the arch of the atlas and theskull base.

In this patient, head hyperextension resultedin the obstruction of both vertebral arteries atthe level of C2. Reconstruction of the worstvessel resulted in the relief of symptoms.

Vol. 13, No. 1, March, 198686

Case 4CC-DVA Bypass to a HypoplasticVertebral Artery

(CC = common carotid; DVA = the verte-bral artery between Cl and C2)

In 1977, a 59-year-old mechanic suffered abrainstem stroke that resulted in painfuldysaesthesias and difficulty in walking. Thecarotid arteries were normal but the carotidinjection filled the basilar artery to its mid-portion. The vertebral system was hypoplastic.On April 11, 1979, a vein bypass was per-

formed from the left common carotid artery tothe 2.0 mm vertebral artery between Cl andC2. His symptoms were markedly improved,and his recovery was uneventful. Two yearslater, in 1981, he underwent coronary bypassfor progressive left ventricular dysfunctionand did not experience any neurologic compli-cations. Angiograms performed in April 1985,6 years after surgery, revealed continuedpatency of the vein bypass.

CommentWhen the carotid vascular bed is large, the

vertebral bed is correspondingly smaller.Ordinarily the internal carotid artery suppliesthe anterior cerebral artery and middle cerebralartery. When it also supplies both posteriorcerebral arteries, the basilar artery ends withthe superior cerebellar arteries, and the verte-bral artery is smaller in diameter, resulting inan increase of high resistance. A reduction ofcardiac output can cause hindbrain ischemia orbrain infarction in this vulnerable territory.

Coronary insufficiency may first be man-ifested by symptoms of hindbrain ischemia.These symptoms may precede manifest car-diac disability by 2 to 4 years, and carry aserious prognosis for males.

Hypoplastic vertebral systems have highresistance, and any reduction ofcardiac outputor perfusion pressure can result in severeischemic damage to this vulnerable territory. Ifa hypoplastic vertebral artery does notcommunicate with the circle of Willis andterminates in the end posterior inferior cere-bellar artery, extracranial vertebral arteryocclusion can result in brainstem and cere-bellar infarction.

Case 5Carotid Distal Vertebral Artery Bypass4

In 1978, a 63-year-old woman came into ourinstitution with dysequilibrium, ataxia andsyncope. The carotid arteries were normal.Mild sinus bradycardia was present. On March7, 1978, a vein bypass was performed from thecommon carotid to the vertebral arterybetween Cl and C2. A year later, in 1979, aphysiologic pacemaker was implanted forworsening bradycardia. Angiography per-formed 7 years after surgery demonstratedcontinued satisfactory function.

CommentHypertension and bradycardia are common

findings in patients with hindbrain ischemia.The control of hypertension is often successfulin controlling symptoms, but vertebral arterysurgery and cardiac pacing may also berequired. Ventricular pacing should not beused if sinus rhythm is present, becausesymptoms may be severely aggravated.

Over the course of 5 years, one-third of thepatients undergoing vertebral artery surgerywill require pacemakers. Physiologicpacemakers are preferable. If the bradycardiais profound, initial pacemaker implantation isemployed. However, if the patient fails torespond, then vertebral artery reconstructionmay be necessary. Vertebral artery reconstruc-tion does not prevent the necessity of cardiacpacing at a later date.

Case 6ICA Ligation; ICA:DVA Anastomosis,End-to-Side

(ICA = internal carotid artery; DVA = thevertebral artery between Cl and C2)A 63-year-old man, blinded by ocular

enucleation, presented with severe dizziness,ataxia, and confusion. He had functioned as askilled cabinet maker using power tools andwas dependent upon his seeing-eye dog. Thesymptoms and the death of his dog destroyedhis well-being. He was not considered eligiblefor a dog because of his symptoms. Angi-ography revealed severe internal carotid arterysiphon stenosis, which was bilaterally worseon the right, stenosis of the right internalcarotid artery at the bifurcation, and bilateralvertebral artery stenosis at the origin from thesubclavian artery.

Ibxas Heart Institute Journal 87

On November 11, 1984, the right carotidbifurcation was endarterectomized and thedistal internal carotid artery ligated anddivided at the level of C1. The divided end ofthe internal carotid artery was then anas-tomosed end-to-side to the vertebral arterybetween Cl and C2. The postoperative angio-gram revealed good filling of the carotidsiphon through the posterior communicat-ing artery.The patency of the internal carotid artery is

assured and no other procedure appears war-ranted at this time. The patient obtained aseeing-eye dog and is doing well.

CommentEighty-five percent of superficial temporal

artery-middle cerebral artery bypasses aredone for internal carotid occlusion or inacces-sible stenosis. In the 5-year cooperative studyby Barnett and Peerless, this procedure wasdeemed of no benefit or detriment.8 Theanastomosis of the superficial temporal arteryto the middle cerebral artery is associated withthe postoperative occlusion of the proximalmiddle cerebral artery. The point where thereversed flow in the middle cerebral arteryencounters the circle ofWillis becomes a dead-point, a watershed where pressures arebalanced, stasis is produced, and thrombosisenhanced. A similar phenomenon is observedwith coronary bypass.The same competition for flow occurs

between the CC-DVA bypass and the internalcarotid artery. If the internal carotid arterysiphon stenosis is severe, then the internalcarotid artery will occlude - and hasoccluded following a CC-DVA bypass. Anartery is a better conduit than a vein. Utilizingthe internal carotid artery as a conduit to theDVA requires the bold stroke of dividing thedistal internal carotid artery at the skull baseand anastomosing it to the DVA. Because ofthe adequate posterior communicating arteryin this patient, the procedure was well toler-ated and produced a satisfactory result.

Case 7Suboccipital Denervationof the Vertebral ArteryA 44-year-old motorcyclist exhibited a

history of severe migraine headaches, whichincapacitated him from work. The headache

was localized on the side of a hypoplasticvertebral artery. Intense medical therapy for 6years proved to be both expensive and fruit-less. Selective angiography of the right ver-tebral artery precipitated a typical migraineheadache. Magnification angiography- of theright skull base revealed prominence ofoccipital artery and spasm of the posteriormeningeal artery.On April 5, 1985, the vertebral artery

between the occiput, dura and foramen trans-versarium was denervated. Postoperatively,he was able to work without limitation. Theheadaches that occurred were more easilycontrolled and appeared related to musclespasm in the flap. The migraine headacheshave been controlled since surgery.

CommentNeurovascular access to the posterior skull

base requires access to a blood supply. Only bycombining the lateral with the posteriorapproach is this possible. In 1954, Hauge9showed that contrast injection into the verte-bral system results in sensory disturbances. Itcan, in fact, reproduce the clinical manifesta-tions of hindbrain ischemia. Followingsurgery, the relief of the severe migraineheadache in this patient was astounding.Long-term follow-up is required to satisfac-torily evaluate this approach.

DISCUSSION

Clinical symptoms of hindbrain ischemiacannot localize the arterial lesions, identify thevascular system involved, nor distinguish end-organ from central hindbrain dysfunction,Preoccupation with the end organ may resultin missing the central pathology. In respiratoryarrest, for example, attention is usuallydirected to the lung, but not to the respiratorycenter in the brainstem or the vertebral arterywhich perfuses it. Distinguishing dysfunctionof the brain stem from that of the vestibularapparatus is commonly done, but identifyingischemia as the pathologic mechanism isuncommon. Dynamic computerized tomo-graphic scanning permits the qualitativemeasurement of focal tissue perfusion withhigh resolution in the cerebral hemispheres aswell as the brainstem.3

Vol. 13, No. 1, March, 198688

The structure of the brain and regionalperfusion must be considered separately fromvascular anatomy. Brain structure is constant;arterial supply is variable. Hindbrain perfu-sion studies cannot identify the site, the side,nor the obstructed artery, which, whenrepaired, will improve flow. Dynamic com-puterized tomographic studies can only doc-ument the presence and location of brainischemia preoperatively and the changes inperfusion that result following arterialreconstruction.The selection of surgical therapy requires a

careful evaluation of vascular anatomy, brainhemodynamics, and risk. The objective ofsurgery is to maximize the flow capacity of thevertebral artery. The integrity of the circle ofWillis, especially the posterior communicat-ing artery, influences the risk of stroke.'0 Theprocedures are multiple and varied, dependingupon the anatomy and lesions.

In the presence of adequate posterior com-municating arteries, the internal carotid arteryobstruction will markedly increase the size ofthe intracranial vascular bed supplied by thevertebral artery. Flow demand is increasedbecause the vertebral artery now supplies theforebrain, the frontal, and the parietal lobes. Alarge caliber vein bypass to the dominantvertebral artery at the skull base can enhanceflow capacity by dilating the system, enhanc-ing arterial flow, and normalizing the perfu-sion of the forebrain, the hindbrain, or both.

In the absence of the posterior communicat-ing arteries, the vascular bed of the hindbrainis limited to the occipital and temporal lobes,the brainstem and cerebellum. In the hypoplas-tic vertebral system, two vertebral arteriessupply the brainstem and cerebellum. Theasymmetrical hypoplastic vertebral artery mayend in the posterior inferior cerebellar artery,supplying one side of the brainstem andcerebellum. When a single vertebral arterysupplies a vascular bed without obviouscollateral, the risk of brain infarction duringsurgical occlusion is increased. The caseprocess requires careful clinical judgment.

Vertebral artery surgery at the skull base hasled to a better understanding of the dynamicanatomy and pathology in this region, and toan appreciation of the sensory nature of thehindbrain. The primary manifestation of

hindbrain ischemia is disability due todysfunction of the sensory, the sensory inte-grating, or the autonomic homeostatic sys-tems. Relief of disability is the primaryindication for surgery.The vertebral system is very sensitive to

reductions of cardiac output. Symptoms ofhindbrain ischemia are particularly ominousfor males and may antedate cardiac crisis by 2to 4 years in our cases. Vertebral artery surgerywas associated with a 19% mortality in malesover the period of follow-up (a four-foldincrease over females). Brain ischemia mayalso antedate symptoms by as much as 2years." Progressive brain ischemia appearsassociated with a falling cardiac output.

Cardiac surgeons have aggressively cor-rected carotid stenosis without signficantlyreducing the incidence of neurologic compli-cations. We propose several questions: Howdo patients with hindbrain ischemia toleratecardiopulmonary bypass? Should the hind-brain and the vertebral artery be more carefullyconsidered and aggressively pursued prior toheart surgery? We find that cardiopulmonarybypass has been well tolerated by the patientswho have had vertebral artery reconstructionat the skull base as the initial procedure.

Ischemia of the hindbrain is a centralhemodynamic problem of the central nervoussystem, with peripheral neurologic manifesta-tions. Only a hemodynamic approach encom-passing brain hemodynamics and vertebralartery surgery will resolve the issue. Thediagnostic instrumentation, the surgicaltechnique, and the knowledge are available.The cost of disability should be factored intoclinical decisions because stroke and deathare not reliable indicators of sensory or cog-nitive defects.

REFERENCES

1. Foley J, Gordon N. Recovery from corticalblindness. Develop Med Child Neurol 1985;27:383.

2. Nornes H, Grip A, Wikeby P. Intraoperativeevaluation of cerebral hemodynamics usingdirectional Doppler technique: Part 2, Saccularaneurysms. J Neurosurg 1979; 50:570.

3. McDonald DA, Potter JM. The distribution ofblood flow to the brain. J Physiol 1951; 114:356.

Texas Heart Institute Journal 89

4. Carney AL. Vertebral artery surgery: Historicaldevelopment, surgical technique and clinicalexperience of 102 cases. Adv Neurol 1981;30:249.

5. Carney AL, Anderson EM. Carotid distalvertebral bypass for carotid occlusion. Clin EEG1978; 9:105.

6. Bogousslavsky J. Vertebrobasilar transientischemic attacks in internal carotid arteryocclusion or right stenosis. Arch Neurol 1985;42:64.

7. Dobben GD, Anderson EM, Bums EM, CarneyAL, Mafee MF, Valvassori G. Measurement ofbrain blood circulation by computed tomog-raphy. Adv Neurol 1981; 30:307.

8. Bamett HJM, Peerless SJ. Five-year cooperativestudy of the EC-IC bypass. N Engl J Med 1985(in press).

9. Hauge T. Catheter vertebral angiography. ActaRadiol (Suppl-Stockholm) 1954; 109:1.

10. Battacharji ER, Hutchinson EC, McCall AJ. Thecircle of Willis: The incidence of developmentalabnormalities in normal and infarcted brains.Brain 1967B; 90:747.

11. Meyer JS, Roger M. Progressive cerebralischemia antedates cerebrovascular systems bytwo years. Ann Neurol; 16:314.

Vol. 13, No. 1, March, 198690