Segmented Middle Cerebral Artery

Occlusion in Primates:

An Experimental Method Requiring

Minimal Surgery and AnesthesiaBY G. F. MOLINARI, M.D.,* J. I. MOSELEY, M.D., AND J. P. LAURENT, M.D.

Abstract:Segmentat MiddleCerebral ArteryOcclusion inPrimates:An ExperimentalMethodRequiringMinimal Surgeryand Anesthesia

• Due to certain anthropomorphic features of the cerebral circulation, monkeys are generallypreferable to lower species in experimental models of stroke. By injecting specially moldedsilicone cylinders through internal carotid artery cannulas, segmental occlusions of the middlecerebral artery were produced in macaques. This embolic method produced cerebral infarctionin all instances while the integrity of skull and intracranial collateral circulation were preserved.Because only neck surgery was required, local anesthesia could be used, permitting observationof acute infarctions in sedated, conscious animals.

Additional Key Wordsembolic method

primate stroke model local anesthesia

Introduction• Both direct ligation of blood vessels1 and intra-arterial injection of foreign substances2 have providedthe means for studying experimental cerebrovascularlesions for well over a century.

Over the years, direct extrinsic ligation ofcerebral vessels using spring clips has been used exten-sively and effectively in dogs,3 cats,4 and primates,6

paralleling the development of improved opticalequipment and surgical techniques. While applicationof a metal clip offers the desirable features of directvisual control and reversibility after known time inter-vals, a prolonged period of anesthesia is required atthe outset because of the prerequisite craniotomy ororbital surgery.6 Moreover, single clips occlude vesselsat points no more than 1 or 2 mm in width and allowperfusion of distal branches proportionate to thequality and quantity of collateral circulation.

Embolism with plastic materials7"8 has been usedsuccessfully in dogs to occlude longer cerebrovascularsegments, but segmental occlusion in primates usuallyhas been accomplished by application of multipleclips.10- u

This report describes an embolic methodoriginally developed in dogs,8'B adapted to theanatomical variables encountered in the primate

From the Sections on Head Injury and Stroke, Applied NeurologicResearch Branch, Collaborative and Field Research, NationalInstitute of Neurological Diseases and Stroke, and the ArmedForces Radiobiology Research Institute, Defense Nuclear Agency,Bethesda, Maryland.

•National Institutes of Health, Building 36, Room 4A03,Bethesda, Maryland 20014.

species, Macaca mulatto. The technique permitssegmental middle cerebral artery occlusion afterminor extracranial surgery and under minimalanesthesia.

MethodsSmall amounts of the silicone rubber compound, Microfil®,were mixed in sterile containers using the manufacturer'srecommended formula. Avoiding air bubbles and bacterialcontamination, the liquid mixture was gently drawn byvacuum suction into polyethylene tubes of selected internaldiameter (Intramedic®, PE 200, PE 205). Filled tubes weresealed with surgical clamps and set aside for 24 hours.Yellow Microfil® (MU 122)* was used most often, while theorange-colored (MU 117)* material was preferred whenrepeat experiments were performed upon a single animal.The lead contained in both pigments allowed visualization ofthe polymer on x-ray films.

Emboli molded in Intramedic tubing, PE 200, had adiameter of 1.41 mm, while those cast in PE 205 measured1.58 (1.6) mm. Segments were cut from the tubing con-taining continuous filaments of the polymerized siliconerubber compound, and cylindrical emboli were extrudedfrom the cut segments by simple digital compression. Freesegments of the elastic polymer were inserted into the con-necting tips of disposable syringes containing 3 to 5 cc ofsterile physiological saline and cut to selected lengths rang-ing from 4 to 8 mm.

Healthy adolescent monkeys (Macaca mulatto) ofeither sex, weighing 2.8 to 4.7 kg, were premedicated withphencyclidine (0.5 mg per kilogram) and atropine (0.1 mgper kilogram). Initially animals were given a repository doseof pencillin in oil, 300,000 U, I.M. after surgery, but, later,

'Manufacturer's reference number.

334 Stroke, Vol. 5, May-June 1974

by guest on May 6, 2018

http://stroke.ahajournals.org/D

ownloaded from

SEGMENTAL MIDDLE CEREBRAL ARTERY OCCLUSION I N PRIMATES

another 300,000 U of aqueous penicillin was added to thepremedication schedule.

Twenty-two common carotid artery cutdowns were per-formed in 18 animals. Twelve of these operations were doneunder general anesthesia using intravenous pentobarbital, 4mg per kilogram. Ten animals received only local skin in-filtration with 2 ml of 1% procaine at the surgical site.

Sham operations were performed on two monkeys inthe group receiving general anesthesia; the common and in-ternal carotid arteries were tied in the neck, but emboli werenot injected.

Twenty embolisms were produced in the remaining 16animals using the following injection techniques: Medicut®cannulas (No. 16 or No. 18 gauge) were inserted into the in-ternal carotid artery through longitudinal common carotidarteriotomies; the arterial incisions were made between twotemporary ligatures just proximal to the carotid bulb.Bleeding was controlled by a stopper in the cannula and ten-sion on the two untied ligatures. Single cylinders of elastic,pigmented, radiopaque, silicone rubber of preselecteddiameter and length were then injected through the internalcarotid artery cannula and flushed into the arterial system.

Since diameters of 1.4 to 1.6 mm exceeded the bores ofNo. 18 and No. 16 gauge cannulas, respectively, embolipassed through the cannulas by stretching to greater lengthbut smaller diameter. Due to the inherent plasticity andelasticity of the material, emboli returned toward preinjec-tion diameters and lengths as permitted by the diameter ofthe surrounding conduit. Therefore, gentle forward pressurewas necessary to propel the emboli through the cannulas.After minimal experience, a palpable pressure drop signaledclearance of the embolus from the cannula into the internalcarotid artery. Backfiow did not appear, however, until theembolus was advanced to a point distal to the sources ofcollateral circulation to the internal carotid artery. Thus,after injection, emboli were advanced by intermittentsyringe pressure to allow both palpation of the pressure dropand observation of the first appearance of backfiow. Thelatter event marked the intracranial arrival of the embolus,since retrograde carotid flow in the monkey could originateonly from the ophthalmic artery or the circle of Willis.

Arteriotomy sites were closed using 9-0 microsurgicalsuture in seven operations; proximal and distal commoncarotid ligatures were tied to isolate the incised segment inall other cases.

Animals were killed after one hour to 30 days of clinicalobservation. Brains were usually perfused, in situ, with 10%formalin solution and then immersed in formalin for one totwo weeks. Fixed brains were studied for correlation of siteand extent of vascular occlusion with distribution andquality of cerebral lesions. Representative sections weremade from blocks of infarcted tissue and selected controlareas.

ResultsCarotid injection of elastic cylinders producedsegmental occlusions in the middle cerebral artery dis-tribution in all cases. Seventeen occlusions (80%) in-volved the horizontal segment of the middle cerebralartery proximal to the candelabra formation, the M-lsegment as classified by Fischer.12 In three ex-periments, multifocal segmental occlusion of the

pericallosal and middle cerebral branches were causedby fractionation of the embolus during or after injec-tion.

The two sham neck operations produced noclinical or pathological evidence of cerebral infarctionor ischemia.

The diameter of the proximal segment of the mid-dle cerebral artery (MCA) in fixed brain specimensranged from 0.7 mm to 1.1 mm in this series ofmacaques. In a previous report8 using this method,emboli of 1.6 mm diameter were found to occlude the1 mm diameter segment of MCA in dogs. Therefore,in order to selectively occlude the main stem of theMCA near its origin from the anastomotic circle ofWillis, a ratio of approximately 1.5:1 was applied inselecting Intramedic® tubing sizes PE 200 (internaldiameter 1.4 mm) and PE 205 (internal diameter 1.6mm) for molding emboli for use in monkeys.

Slight variations in diameter of the MCA wereindependent of body weight in animals of this size andage. The internal carotid arteries in the neck, however,were more variable and were weight-dependent. Theinternal carotid easily admitted a No. 18 gauge can-nula in animals under 3.5 kg, and a No. 16 gaugecould be used in larger monkeys. Emboli molded inPE 200 tubing (1.4 mm diameter) passed through thelumina of the No. 18 gauge cannulas with minimalresistance. Similarly, emboli of 1.6 mm diameter (PE205) easily passed through No. 16 gauge cannulas.Either size was satisfactory for occlusion of the prox-imal MCA segment. The choice between sizes wasdetermined by the gauge of the largest cannula accom-modated by the surgically exposed internal carotidartery.

The final placement of emboli within the cerebralarterial tree was dependent upon the size and qualityof the embolic particle injected and, to a lesser extent,upon anatomical variables in the branching of themiddle cerebral artery.

Figure 1 is an x-ray film showing an embolus inthe proximal middle cerebral artery segment justlateral to the optic foramen. The film was madewithout special preparation, using ordinaryroentgenographical equipment. Projection wasanterior to posterior in the long axis of the right orbit.The embolus was 7 mm X 1.5 mm at injection.

The distribution of infarctions varied with thelength of the embolus injected. Emboli measuring 7mm or more before injection occluded the entirehorizontal segment from the anastomotic circle to thebranching at the candelabra formation in the Sylvianfissure. Figure 2 shows the same embolusdemonstrated radiographically in figure 1 in the brainspecimen. The embolism extended from a point prox-imal to the origin of the posterior communicatingartery through the MCA to the origin of the large or-bitofrontal branch. Longer emboli of this typeobstructed the origins of all the penetrating branches

Stroke, Vol. 5, May-June 1974 335

by guest on May 6, 2018

http://stroke.ahajournals.org/D

ownloaded from

MOLINARI, MOSILIY, LAURENT

Plain A -P roentgenogram made In the long axis of the right orbit 24hours after embolism. The arrow denotes the curvilinear positiveshadow of an experimental embolus in the proximal MCA. Note therelationship of the embolus in MCA to the negative shadow of theoptic foramen medially.

of MCA including the anterior choroidal artery, prox-imally.

Distally, the orbitofrontal artery was usually thefirst branch of the MCA causing an abrupt reductionin the caliber of the parent vessel in the Sylvian fissure.However, in some hemispheres, that branch arosemedial to the fissure as a small vessel from theanterior wall of the M-l segment. In such instancesthe first major branch contributing to thecandelabrum was the anterior temporal. The distal ex-tent of the occlusion was slightly more lateral inhemispheres with the latter vascular pattern; thesmaller orbitofrontal branch was then occluded at itsorigin in the same manner as the penetrating vessels.

Figure 3 is a coronal section of the brain shown infigure 2. Twenty-four hours after embolism, the brainshows pallor, loss of vascular markings and slightswelling throughout the territory supplied by the sur-face and penetrating branches of MCA. The caudate,pallidum, putamen, claustrum, internal and externalcapsules are all involved. Shorter emboli of ap-proximately 4 to 5 mm occluded the lateral portion ofthe horizontal segment, but the first few millimeters ofthe MCA remained patent. Therefore, deep structuressupplied by the anterior choroidal artery, the firstpenetrating branch of MCA, were spared.

Outdated Microfil (shelf-life four months) andmaterial molded more than two days before use

produced friable emboli of generally poor quality. Inthree of our experiments (20%), segmental occlusionsinvolved distal MCA branches in the Sylvian fissureand the pericallosal artery. Patchy, bland areas of cor-tical infarction were produced distal to each occludedsegment. Deep hemispheral structures were spared,however, unless occlusions also involved the horizon-tal MCA segment proximal to the candelabra forma-tion. Multifocal occlusions did not occur when freshlyprepared emboli were used.

In general the severity of focal residual signsvaried directly with the completeness of horizontalsegment occlusion. Occlusions of the entire segmentfrom the origin at the carotid trifurcation to thebranching at the candelabra caused severe con-tralateral sensory, motor and visual deficits; shorterocclusions, sparing the proximal MCA, caused lesssevere impairments. Occlusions of distal corticalbranches of MCA and the pericallosal artery occlu-sion produced only transient signs and negligibleresidual impairment.

Focal neurological signs such as hemiplegia andtonic eye deviation were observed immediately afterembolism in animals given only phencyclidine andlocal procaine, while six hours or more were requiredto evaluate the effects of embolism when barbiturateanesthesia was used.

Some lesions obtained from animals used inchronic experiments involving more than four days'survival showed histological evidence of bacterial con-tamination at the periphery. Lesions from animalsmeticulously treated with antibiotics before and afterembolism were free of biological artifacts.

Detailed clinicopathological correlation will bethe subject of a subsequent report.

DiscussionIn applying this previously reported model toprimates, a number of species-dependent, anatomicalvariables were encountered. In the dog the major ex-tracranial anastomosis between the maxillary and in-ternal carotid arteries and the large, well-developedposterior communicating arteries" differs markedlyfrom primate vascular patterns. Intracranially, thebranching of the M-l segment of MCA into the so-called "candelabra" occurs on the hemispheral sur-face of canine brains, while in the anatomically morecomplex simian brain, this arborization occurs deep inthe Sylvian fissure on the insular cortex. The relativeproportions of the middle cerebral to posterior com-municating arteries of the macaque may be seen infigure 2. These anatomical features make the primatepattern the more credible model of the human cerebralcirculation.

Despite the relative constancy of patterns of thecerebral vasculature among primate species includingman,14 the anthropomorphism of the circle of Willis isfar from perfect even in the macaque, particularly

336 Strok; Vol. 5, May-Jun 1974

by guest on May 6, 2018

http://stroke.ahajournals.org/D

ownloaded from

SEGMENTAL MIDDLE CEREBRAL ARnRY OCCLUSION IN PRIMATIS

• ' - . a

A7-U0

PKMJM2

Experimental embolus in the right MCA of specimen taken from the same animal shown in figure 1. The embolus measured 7 mm X 1.5 mmwhen injected. The occlusion extends from the internal carotid a., proximal to the posterior communicating a., and is impacted at the origin ofthe orbitofrontal a., distally. Emboli of this length occlude the origins of all of the major penetrating branches of MCA, including the anteriorchoroidal.

Gross cross section of specimen shown in figure 2 showingdistribution of ischemia. The affected side is pale with a paucity ofvascular markings in both the superficial and deep territoriessupplied by the MCA. Note the degree of brain swelling indicatedby the ventricular asymmetry. This specimen was obtained 25 hoursafter experimental embolism.

anteriorly. In our specimens the anterior cerebralarteries anastomosed in a gently sweeping arc givingrise to a large single pericallosal artery in the midline.This configuration could allow crossover of smalleremboli or embolic fragments from the side injected tothe contralateral middle cerebral or pericallosalartery, if flow gradients were created by vigorous in-jection pressures. The use of small volumes of salineand intermittent pressure to advance the emboliminimized the risk of crossover through the anteriorportion of the anastomotic circle.

Because the configuration of the candelabra wassomewhat variable in this species, the distribution ofinfarctions varied slightly with the vascular anatomy.This was considered acceptable since similar variablesoccur in human pathology and because preciseclinicopathological correlations can be made using ourmodel. Furthermore, horizontal segment occlusionsalways caused deep infarctions in the distribution of

Strokt, Vol. 5. May-Jun* 1974 337

by guest on May 6, 2018

http://stroke.ahajournals.org/D

ownloaded from

MOLINARI, MOSELEY, LAURENT

the lateral lenticulostriate arteries despite variations inthe angioanatomy and length of embolus. The use oflonger emboli, approximately 7 mm in length, com-pensated for variations in the distal patterns of ar-borization and assured obstruction of the more medialpenetrating branches as well.

The diameters of emboli optimal for occlusion ofthe horizontal segment of the MCA were ap-proximately one and one-half times the diameters ofthe formalin-fixed arterial segments. Although thisratio was determined empirically, further precisionbased upon measurements of fixed arterial specimenswould not be fruitful because of the inconstantshrinkage artifact. Moreover, these segments are reac-tive in vivo, and the diameters vary withpathophysiological changes. Using this ratio, the in-herent elasticity of the silicone cylinders maintainscentrifugal pressure upon the intimal surface ofoccluded segments despite a considerable range ofreactive vasodilatation.

This method produced segmental occlusion of theMCA while leaving the meningocerebral collateralcirculation anatomically and physiologically intact.Meyer10- n has discussed the need to apply multipleligatures to isolate a vascular segment from collateralflow. In studies comparing the effects of temporaryocclusion to permanent ligation of cerebral arteries,the model used for permanent ligation has usuallybeen surgical division of the artery between twoclips.15-16 While such occlusions are permanent in-deed, they are also segmental. The severity and dis-tribution of the cerebral lesions might well be differentfrom those caused by nonsegmental occlusion withsingle removable clips. Moreover, surgical divisionmay cause autonomic neurotomy, with denervation ofthe pial vessels distal to the occlusion. Vasospasmcaused by extrinsic compression of cerebral vesselsalso may introduce physiological if not morphologicalartifacts in experimental models of infarction andischemia.

Because our only surgical defect is in the neck,the brain specimen is free from the pathological ar-tifacts produced by traumatic manipulation and elec-trocautery.17 The entire cranium remains intact andsuitable for acute or chronic electrophysiologicalmonitoring. Physiological data recorded in vivo maybe collated with precise localization of the vascularocclusion by x-ray or postmortem examination. Theabbreviated preparation time and limitation of sur-gery to soft tissues in the neck eliminate the need forprolonged anesthesia and therefore the specific effectsof barbiturate drugs are removed from morphologicaland pathophysiological studies.

Using microsurgical techniques, the internalcarotid arteriotomy may be closed, leaving the lumenpatent for craniad blood flow after embolism.

This embolic technique is subject to certaintechnical and biological artifacts which can be avoided

or controlled. These include the accidental fragmenta-tion of embolic cylinders during injection and secon-dary infection of cerebral lesions in chronic ex-periments.18 Both may be controlled by using freshmaterials, meticulous sterile technique and generousantibiotic treatment before and after surgery.

Because of minor variations in the branching ofindividual middle cerebral arteries, the distal limit ofthe occluded segment may vary by a few millimeters.These anatomical variables may be compensated forby using longer emboli (7 mm) of the largest diameteradmissible through an internal carotid artery cannula.

ConclusionIn this series infarctions of the putamen and internalcapsule were constant findings in horizontal segmentocclusion regardless of length of embolus used or con-figuration of the candelabra. The overall rate ofhorizontal segment occlusion (80%) improved with ex-perience and approached 100% as technical artifactswere identified and corrected. The method describedoffers a reasonable alternative to craniectomy andgeneral anesthesia in the laboratory investigation ofcerebral infarction. Intrinsic segmental occlusion ofthe middle cerebral artery may be preferable to extrin-sic ligation in certain pathophysiological experiments.

AcknowledgmentsThe authors wish to thank Dr. Julio Garcia and Dr. Y. Kamijyo forprofessional assistance and technical support in this work.

References1. Cooper A: Some experiments and observations on tying the carotid

and vertebral arteries. Guys Hosp Rep 1:458-475, 18362. Flourens JPM: Note touchant I'action de diverses substances injectees

dans les arteres. Acad Sci 24:905-908, 18473. Petersen JN, Evans JP: The anatomical end results of cerebral arterial

occlusion; experimental and clinical correlation. Trans Amer NeurolAssoc 63:88, 1937

4. Sundt TM Jr, Waltz AG: Experimental cerebral infarction. Retro-orbital, extradural approach for occluding the middle cerebralartery. Mayo Clin Proc 41:1959-1968, 1966

5. Hudgins WR, Garcia JM: Transorbital approach to the middlecerebral artery of the squirrel monkey. Stroke 1:107-111, 1970

6. O'Brien MD, Waltz AG: Transorbital approach for occluding themiddle cerebral artery without craniectomy. Stroke 4:201-206, 1973

7. Whisnant JP, Millikan CH, Wakim KG, et ah Experimental productionof cerebral infarction in animals. Mayo Clin Proc 29:613-617, 1954

8. Molinari GF: Experimental cerebral infarction. I. Selective segmentalocclusion of intracranial arteries in the dog. Stroke 1:224-231, 1970

9. Molinari GF: Experimental cerebral infarction. II. Clinicopathologicalmodel of deep cerebral infarction. Stroke 1:232-244, 1970

10. Meyer JS: Circulatory changes following occlusion of the middlecerebral artery and their relation to function. J Neurosurg 15:653-673, 1958

11. Meyer JS: Importance of ischemic damage to small vessels in ex-perimental cerebral infarction. J Neuropath Exp Neurol 17:571-585, 1958

12. Fischer E: Die Lageabweichungen der vorderen hirnarterie imgefassbild. Zentralbl Neurochir 3:300-313, 1938

13. de la Torre E, Netsky MG, Mesenhan L: Intracranial and extracranial

338 Stroke, Vol. 5, May-June 1974

by guest on May 6, 2018

http://stroke.ahajournals.org/D

ownloaded from

SEGMENTAL MIDDLE CEREBRAL ARTERY OCCLUSION IN PRIMATES

circulation in the dog: Anatomic and angiographic studies. Amer J 16. Crowell RM, Olsson Y, Klotzo I, etal: Temporary occlusion of themid-Anat 105:343-362, 1959 die cerebral artery in the monkey: Clinical and pathological obser-

14. Gillilan LA: The arterial and venous blood supplies to the forebrain vations. Stroke 1:439-448, 1970(including the internal capsule) of primates. Neurology (Minneap) 17. HudginsWR, Garcia JH: The effect of electrocautery, atmospheric ex-18:653-670, 1968 posure, and surgical retraction on the permeability of the blood-

15. Harvey J, Rasmussen T: Occlusion of the middle cerebral artery. Arch brain-barrier. Stroke 1:375-380, 1970Neurol Psychiat 66:20-29, 1951 18. Molinari GF: Septic cerebral embolism. Stroke 3:117-122, 1972

Sfroie, Vol. 5, May-June 1974

by guest on May 6, 2018

http://stroke.ahajournals.org/D

ownloaded from

G. F. MOLINARI, J. I. MOSELEY and J. P. LAURENTRequiring Minimal Surgery and Anesthesia

Segmented Middle Cerebral Artery Occlusion in Primates: An Experimental Method

Print ISSN: 0039-2499. Online ISSN: 1524-4628 Copyright © 1974 American Heart Association, Inc. All rights reserved.

is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231Stroke doi: 10.1161/01.STR.5.3.334

1974;5:334-339Stroke. 

http://stroke.ahajournals.org/content/5/3/334World Wide Web at:

The online version of this article, along with updated information and services, is located on the

  http://stroke.ahajournals.org//subscriptions/

is online at: Stroke Information about subscribing to Subscriptions: 

http://www.lww.com/reprints Information about reprints can be found online at: Reprints:

  document. Permissions and Rights Question and Answer available in the

Permissions in the middle column of the Web page under Services. Further information about this process isOnce the online version of the published article for which permission is being requested is located, click Request

can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial Office.Stroke Requests for permissions to reproduce figures, tables, or portions of articles originally published inPermissions:

by guest on May 6, 2018

http://stroke.ahajournals.org/D

ownloaded from