Septic Cerebral EmbolismBY GAETANO F. MOLINARI, M.D.*

Abttract:SepticCerebralEmbolism

• In order to study the effects of septic embolism on neurovascular structures,cerebral infarction was produced in a series of dogs by injecting silicone rubbercylinders through an internal carotid artery cannula. Each embolic particle wasfirst incubated with one of four known bacterial pathogens. Death occurred inthree animals on the second postoperative day, at which time ten othermoribund dogs were sacrificed. Eleven animals made apparently uneventfulrecoveries from early hemiplegia and appeared well when sacrificed electivelyone to five weeks after embolism. Autopsies revealed subarachnoid and acutesubdural hemorrhages in the early group with gross and microscopic evidenceof mycotic aneurysm in each instance. Chronically surviving animals showedhistological lesions in the putamen or temporal pole consistent with brainabscess. In this series, mycotic aneurysm with hemorrhage was an extremelydramatic complication of emboli infected with the virulent pathogens,Staphylococcus aureus and Escherichia coli, while brain abscess developedinsidiously in ischemic areas, after embolism with the opportunistic pathogens,Streptococcus viridans and Enterococcus.

ADDITIONAL KEY WORDS brain abscess subarachnoid hemorrhageEscherichia coli acute subdural hemorrhage Staphylococcus aureusStreptococcus viridans Enterococcus

Introduction• Although mycotic aneurysm and brainabscess are well-recognized complications ofbacterial endocarditis,1 a the precise mecha-nisms responsible for their development areobscure. Retrospective studies suggestseptic embolism as a primary event inpathogenesis,4'r> but this can rarely be provedat autopsy. The purpose of this paper is toreport the clinical and pathological effects ofseptic cerebral embolism in experimentalanimals.

Associate Professor of Neurology, Department ofNeurology, University of Rochester School of Medi-cine; Director, Neurological Unit, Monroe Com-munity Hospital, 435 East Henrietta Road, Rochester,New York, 14620.

This study was supported by the Waasdorp StrokeResearch Laboratory, University of Rochester, Mon-roe Community Hospital, and Grant No. NS 05084,NINDS.

'Received the Irving S. Wright award for younginvestigators in the field of stroke (1971) for thispaper.

Stroke, Vol. 3, March-April 1972

MethodsIn a previous report, we have described a methodfor producing segmental cerebral arterial occlu-sions in dogs, using elastic silicone rubberemboli.0 Cylindrical emboli of preselected diame-ter, introduced through an internal carotid arterycannula, lodge in the proximal middle cerebralsegment, obstructing lenticulostriate arteries attheir origins. This technique produces unilateralbland infarction in the basal ganglia and internalcapsule on the side injected.7

In order to study septic embolism, this basicmethod was slightly modified. Embolic cylindersmeasuring 1.6x6 mm were placed on blood agarplates containing pure growths of known patho-genic bacteria. Using sterile technique, thematerial was gently rotated to obtain a uniformdistribution of organisms on the surface of theembolus.

Twenty-four dogs were given septic emboliprepared in this way.

The organisms used were Staphylococcusaureus (9) , Escherichia coli (3) , Streptococcusfecalis (2 ) , and Streptococcus viridans (10) .

Animals were observed clinically for signs ofcomplications for up to five weeks after embolism.Euthanasia was performed promptly on acutely ill

117

by guest on July 9, 2018http://stroke.ahajournals.org/

Dow

nloaded from

MOLINARI

STREPTq EECAllS

§ SIfiEPI^ VtRIDANS

5 STAPH§ AUREUS

$ ECOLI

MYCOTICANEURYSM

•

••

BRAIN ABSCESS

•D

[~i D Q •

n n • c

1 7 3 5 10 15 20 25 30 35TIME ( I VSI

Q-EIEUIVE MCRIflCE• - SPONTANEOUS HEMORRHAGE

FIGURE 1

Scattergram summarizing results. Three animals inthe Staph. aureus group were sacrificed electi vely onthe first day after embolism, to obtain specimens ofunruptured aneurysms. Closed squares indicate ani-mals which died spontaneously or were sacrificed ina moribund state after subarachnoid hemorrhage liadoccurred.

animals by intravenous barbiturate injection. Theothers were sacrificed electively at varying timeintervals after septic embolism.

At postmortem examination, swab cultureswere obtained from cerebral fluid, subdural andsubarachnoid blood, and brain surface at the siteof the arterial lesion. Cultures were also made ofneedle aspirates taken from areas of encephalo-malacia, deep to surface emboli.

After fixation, brain specimens were studiedand blocked for representative sections of neuraland vascular pathology. Histological sections wereprepared with hematoxylin and eosin, van Giesonelastic and trichrome stains.

ResultsFigure 1 is a scattergram which summarizes theclinical and pathological results. The closedsquares represent spontaneous deaths or ani-mals killed in a moribund state. Open squaressignify animals with hemiplegia or minorresidual signs of infarction when sacrificedelectively.

On the first day after embolism, alltwenty-four animals in this series were alertand responsive but hemiplegic on the sideopposite the injection. These were the charac-teristic features of middle cerebral arteryocclusion in previous studies of bland infarc-tion.7

Three hemiplegic animals with staphylo-coccal emboli were sacrificed 24 hours afterembolism. All three showed unruptured aneu-rysms at the site of vascular occlusion.

118

Another six animals given emboli infectedwith Staph. aureus and the three given E. colidied abruptly or showed catastrophic deteriora-tion on the second or third day after embolism.All nine had subarachnoid hemorrhages atautopsy; many were associated with acutesubdural hematomas on the side of theembolism.

In contrast to this group, 11 of 12 dogsgiven streptococcal emboli survived the acutephase and showed continuous improvementand adaptation to their hemiplegia, as theyprogressed into the chronic postoperationcourse. Many had no apparent deficit at allwhen sacrificed several weeks after embolism.Only one dog in the streptococcal groupdeveloped mycotic aneurysm and died in theacute phase from subarachnoid hemorrhage.Despite the clinical improvement, all of theothers had brain abscess at autopsy.

Therefore, both clinical course and brainpathology were closely related to the type oforganism used.

FIGURE 2

Subarachnoid hemorrhage in the dog. Brain specimen,in situ, after removal of the calvarium and reflectionof the dura. Blood fills the sulci of the subaraclinoidspace on the side of the aneurysm. The arrow marksthe extravascular embolic material in the subduralclot adjacent to the reflected dura mater.

Stroke, Vol. 3, March-April 7972

by guest on July 9, 2018http://stroke.ahajournals.org/

Dow

nloaded from

SEPTIC CEREBRAL EMBOLISM

FIGURE 3

Mycotic aneurysm. Upper, Ventral surface of dog brain with a mycotic aneurysm on theright middle cerebral artery. Lower. Magnification of the aneurysm outlined above. Theright optic nerve is at the top of the picture. The saccular aneurysm contains small perforationfilled by clotted blood. The proximal end of the embolus is seen in the upper right quadrantof the aneurysm, while the remainder of the embolic material is contained within the sac.

Strokt, Vol. 3, March-April 1972 119

by guest on July 9, 2018http://stroke.ahajournals.org/

Dow

nloaded from

MOLINARI

Bacteriological CorrelationsPure cultures of Staph. aureus or E. coli wererecovered from intracranial blood, cerebralfluid, and brain surface in all cases ofaneurysm with hemorrhage. Staphylococciwere also grown from the arterial lesion on thebrain surface in the three animals killed beforehemorrhage occurred.

In addition, the organism was recoveredfrom the one case of streptococcal aneurysm,but streptococci were grown from brain abscessspecimens only if obtained during the firstweek after embolism. Even in abscesses lessthan a week old, swabs taken from the brainsurface, occluded arterial segment, or cerebralfluid grew the injected streptococcal organism,but aspirates of the deep lesion yielded mixedflora or no growth. Recovery of the injectedorganism from any source was rare in matureabscesses more than a week old.

PathologyOn gross inspection, two types of hemorrhagewere seen in specimens of ruptured mycoticaneurysm. Through the intact dura, bright redblood was seen throughout the subarachnoidspace bilaterally. There was also a large, darksubdural clot on the side of the aneurysm.Upon reflecting the dura, the embolus wasalways found in the subdural clot. Unclottedsubarachnoid blood could be seen in the sulciof both hemispheres, but was more abundanton the side of the aneurysm. Figure 2 showsthe typical appearance of subarachnoid hemor-rhage as found at autopsy.

The brilliant yellow embolus was easilyidentified and recovered. In specimens ofruptured aneurysms, the embolus was extra-vascular, having escaped the artery through theruptured aneurysm.

Figure 3A shows the ventral surface of adog's brain. Note the component arteries of thecircle of Willis and the course and position ofthe unaffected middle cerebral artery, sweepinganterior to the left temporal pole. The outlinemarks the right middle cerebral artery contain-ing an aneurysm overlying an intravascularembolus.

Figure 3B is a tenfold magnification of theoutlined area showing a portion of the embolusin the artery just proximal to a saccularaneurysm, containing a minute perforation.

On microscopic study, aneurysmal seg-ments showed fusiform or saccular dilatation

120

compared to control contralateral segments.Aneurysmal segments, in both longitudinal andcross section, showed an intense inflammatorycell response on the adventitial surface andirregular penetration of the muscularis layerwith polymorphonuclear infiltrate. Only theintima and internal elastic membrane remainedintact in unruptured aneurysms. Within themedia, at the periphery of zones of inflamma-tion and frank necrosis, the vasa vasorum werecongested with polymorphonuclear inflamma-tory cells. Sections through ruptured segmentsdemonstrated coiling and retraction of theintima and elastica on either side of the rent atthe site of rupture.

Gross brain specimens from the animalswith chronic lesions each showed an emboluswithin a surface artery. Small accumulations ofpus surrounded the occluded segment, butaneurysmal dilatation was not observed. Local-ly, flattening of gyri and sulci indicated focalbrain swelling. Larger lesions produced easilyrecognized asymmetries, occasionally involvingmost of the affected hemisphere.

The most common site for abscessformation was in the distribution of thepenetrating arteries of the lateral striate group,involving the deep brain substance.

In section, older abscesses were consis-tently larger than those found in specimenstaken within the first week after embolism,despite similarity in size and site of segmentalartery occlusions.

Brain sections from the animals givenstreptococcal embolism revealed typical brainabscesses in gross and microscopic prepara-tions. Lesions were usually lateral to theinternal capsule, involving the basal gangliaand temporal pole. White matter was rarelyinvolved directly but in the older lesions, whichwere larger, some displacement and distortiondid occur. This factor plus the slow growth ofthese lesions probably accounts for the ap-parent benignity of the clinical course in thebrain abscess group.

Abscess lesions had sharply definedborders, were greenish in color, and werealways found in relation to an intravascularembolus in a surface artery. Cavities werefound only as artifacts produced by needleaspiration; otherwise, the necrotic centerscontained homogeneous gelatinous pus.

Figure 4 is a hematoxylin stain through abrain abscess. The capsule is clearly seen,

Stroke, Vol. 3, March-April 1972

by guest on July 9, 2018http://stroke.ahajournals.org/

Dow

nloaded from

SEPTIC CEREBRAL EMBOLISM

FIGURE 4

Brain abscess. Magnification of a horizontal section through right anteriorquadrant of canine brain. Note the sharp line of demarcation and intensecellular response of the capsule. The cavity and tract to the surface wereproduced by needle aspiration at autopsy. Hematoxylin and eosin stain.

composed of densely packed, deeply stainingpolymorphonuclear leukocytes. Cellular reac-tion of this type at five weeks, the age of thislesion, is distinctive of abscess and is inmarked contrast to bland infarction.

Microscopically, abscess walls showedthree layers. Each had an inner necrotic layer,a thick middle band of acute inflammatorycells, and an outer zone of gliosis, fibrosis, andprolific vascularization. Within the outer zone,more central blood vessels were frequently

Slrok; Vol. 3, March-April 1972

surrounded by inflammatory cells, while in theperipheral, transitional regions, there wasbland proliferation of endothelial cells aroundpatent lumina.

DiscussionThis series of experiments demonstrates thepathogenesis of mycotic aneurysm and brainabscess from a single mechanism. The con-trolled variable was the organism used toproduce each type of lesion.

121

by guest on July 9, 2018http://stroke.ahajournals.org/

Dow

nloaded from

MOLINARI

The experimental models presented paral-lel most of the clinical and pathological featuresof the naturally occurring prototypes. Thismethod offers the additional advantage of anembolism that is easily traced and recovered atautopsy.

This study shows that in canine brain theinterval from septic embolism to aneurysmformation is remarkably short. The entireprocess including embolism, local arteritis,aneurysm formation and hemorrhage usuallyoccurred within two days.

Despite delivery of bacteria to the intimalsurface of occluded arteries by septic emboli,the inflammatory response in mycotic aneu-rysm proceeded from the adventitial surfacetoward the media. The internal elastic mem-brane and intima were last affected. Thissupports the hypothesis of other investigatorsthat infection of the arterial wall is mediatedthrough stasis and sepsis in the vasavasorum.4'8

While there have been several previousmodels of brain abscess,0"11 all have incommon a primary brain injury, usuallysurgical introduction of a foreign body,allowing invasion by bacteria. Our methoddemonstrates that both ischemic injury andinfection may be caused simultaneously by asingle event, septic embolism.

Virulence of the organism was thedeterminant of morbidity and mortality in thisseries. The more virulent pathogens, Staph.aureus and E. coli, caused immediate anddramatic infections of the arterial wall, withaneurysm and hemorrhage. In contrast, the so-called "opportunistic" streptococci usuallycaused progressive lesions, beginning in areasof infarction or ischemia.

Summary and ConclusionsSeptic cerebral embolism may cause eithermycotic aneurysm or brain abscess. In thisseries, the virulence of the organism deter-mined the pathology and clinical course of the

cerebral lesions. The experimental modelspresented are representative of the clinical,pathological and histological features of natu-rally occurring disease states, and thereforeoffer opportunity for further research inmedical and surgical treatment.

References1. Osier W: The Gulstonian lectures on malig-

nant endocarditis. Brit Med J 1: 467-470 (Mar7) 1885

2. Kernohan JW, Wolrman HW, Baines AR:Involvement of the nervous system associatedwith endocarditis. Arch Neurol Psychiat 42:789-809 (Nov) 1939

3. Winkleman NW, Eckel JL: The brain inbacterial endocarditis. Arch Neurol Psychiat23: 1161-1182 (June) 1930

4. Bell WE, Butler C: Cerebral mycotic aneurysmin childrerv—fwo case reports. Neurology(Minneap) 18:81-86 (Jan) 1968

5. Rabinowitz MA, Marcus IH, Weinstein J:Subacute bacterial endocarditis with largebrain abscess. JAMA 98: 806-807 (Mar 5)1932

6. Molinari GF: Experimental cerebral infarction

I. Selective segmental occlusion of intracranialarteries in the dog. Stroke 1: 224-231 (July-Aug) 1970

7. Molinari GF: Experimental cerebral Infarction

II. Clinicopathological model of deep cerebral

infarction. Stroke I : 232-244 (July-Aug)

1970

8. Nakata Y, Shionaya S, Kamiya K: Pathogeneslsof mycotic aneurysm. Angiology 19: 593-601(Nov) 1968

9. Falconer MA, McFarlan AM, Russel DS:Experimental brain abscess in the rabbit. BritJ Surg 30: 245-260 (Jan) 1943

10. Hassler O, Forsgren A: Experimental abscessesin brain and subcutis. Acta Path MicrobiolScand62: 59-67, 1964

11. Long WD, Meacham WF: Experimental meth-od for producing brain abscess In dogs withevaluation of the effect of dexamethasone andantibiotic therapy on pathogenesis of intrace-rebral abscess. Surg Forum 19: 437-438,1968

122 Strok; Vol. 3, March-April 1972

by guest on July 9, 2018http://stroke.ahajournals.org/

Dow

nloaded from

MOLINARI

The experimental models presented paral-lel most of the clinical and pathological featuresof the naturally occurring prototypes. Thismethod offers the additional advantage of anembolism that is easily traced and recovered atautopsy.

This study shows that in canine brain theinterval from septic embolism to aneurysmformation is remarkably short. The entireprocess including embolism, local arteritis,aneurysm formation and hemorrhage usuallyoccurred within two days.

Despite delivery of bacteria to the intimalsurface of occluded arteries by septic emboli,the inflammatory response in mycotic aneu-rysm proceeded from the adventitial surfacetoward the media. The internal elastic mem-brane and intima were last affected. Thissupports the hypothesis of other investigatorsthat infection of the arterial wall is mediatedthrough stasis and sepsis in the vasavasorum.4'8

While there have been several previousmodels of brain abscess,0"11 all have incommon a primary brain injury, usuallysurgical introduction of a foreign body,allowing invasion by bacteria. Our methoddemonstrates that both ischemic injury andinfection may be caused simultaneously by asingle event, septic embolism.

Virulence of the organism was thedeterminant of morbidity and mortality in thisseries. The more virulent pathogens, Staph.aureus and E. coli, caused immediate anddramatic infections of the arterial wall, withaneurysm and hemorrhage. In contrast, the so-called "opportunistic" streptococci usuallycaused progressive lesions, beginning in areasof infarction or ischemia.

Summary and ConclusionsSeptic cerebral embolism may cause eithermycotic aneurysm or brain abscess. In thisseries, the virulence of the organism deter-mined the pathology and clinical course of the

cerebral lesions. The experimental modelspresented are representative of the clinical,pathological and histological features of natu-rally occurring disease states, and thereforeoffer opportunity for further research inmedical and surgical treatment.

References1. Osier W: The Gulstonian lectures on malig-

nant endocarditis. Brit Med J 1: 467-470 (Mar7) 1885

2. Kernohan JW, Wolrman HW, Baines AR:Involvement of the nervous system associatedwith endocarditis. Arch Neurol Psychiat 42:789-809 (Nov) 1939

3. Winkleman NW, Eckel JL: The brain inbacterial endocarditis. Arch Neurol Psychiat23: 1161-1182 (June) 1930

4. Bell WE, Butler C: Cerebral mycotic aneurysmin childrerv—fwo case reports. Neurology(Minneap) 18:81-86 (Jan) 1968

5. Rabinowitz MA, Marcus IH, Weinstein J:Subacute bacterial endocarditis with largebrain abscess. JAMA 98: 806-807 (Mar 5)1932

6. Molinari GF: Experimental cerebral infarction

I. Selective segmental occlusion of intracranialarteries in the dog. Stroke 1: 224-231 (July-Aug) 1970

7. Molinari GF: Experimental cerebral Infarction

II. Clinicopathological model of deep cerebral

infarction. Stroke I : 232-244 (July-Aug)

1970

8. Nakata Y, Shionaya S, Kamiya K: Pathogeneslsof mycotic aneurysm. Angiology 19: 593-601(Nov) 1968

9. Falconer MA, McFarlan AM, Russel DS:Experimental brain abscess in the rabbit. BritJ Surg 30: 245-260 (Jan) 1943

10. Hassler O, Forsgren A: Experimental abscessesin brain and subcutis. Acta Path MicrobiolScand62: 59-67, 1964

11. Long WD, Meacham WF: Experimental meth-od for producing brain abscess In dogs withevaluation of the effect of dexamethasone andantibiotic therapy on pathogenesis of intrace-rebral abscess. Surg Forum 19: 437-438,1968

122 Strok; Vol. 3, March-April 1972

by guest on July 9, 2018http://stroke.ahajournals.org/

Dow

nloaded from

GAETANO F. MOLINARISeptic Cerebral Embolism

Print ISSN: 0039-2499. Online ISSN: 1524-4628 Copyright © 1972 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.3.2.117

1972;3:117-122Stroke. 

http://stroke.ahajournals.org/content/3/2/117World 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 July 9, 2018http://stroke.ahajournals.org/

Dow

nloaded from