Watershed Infaction


12 pages

Please download to get full document.

View again

of 12
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Watershed Infaction
  The Pathophysiology of Watershed Infarction in InternalCarotid Artery Disease Review of Cerebral Perfusion Studies Isabelle Momjian-Mayor, MD; Jean-Claude Baron, MD, FRCP, FMedSci  Background and Purpose —In carotid disease, infarcts can occur in the cortical as well as internal watershed (WS), or both.Better understanding the pathophysiology of WS infarcts would guide treatment. Two distinct hypotheses, namelylow-flow and micro-embolism, are equally supported by neuropathological and physiological studies. Here we reviewthe evidence regarding the mechanisms for WS stroke in carotid disease and whether they differ between cortical andinternal WS infarcts. Summary of Review —After a brief account of the anatomy of the WS and the cerebrovascular physiology in circumstancesof low perfusion pressure, the literature concerning the mechanisms of WS infarction in carotid disease is reviewed anddiscussed with emphasis on imaging and ultrasound studies of the cerebral hemodynamics. Conclusion —The evidence strongly favors a hemodynamic mechanism for internal WS infarction, especially regarding theso-called rosary-like pattern in the centrum semiovale. However, the relationships between cortical WS infarction andhemodynamic compromise appear more complicated. Thus, although severe hemodynamic compromise appears to underliecombined cortical and internal WS infarction, artery-to-artery embolism may play an important role in isolated cortical WSinfarcts. Based on the high prevalence of microembolic signals documented by ultrasound in symptomatic carotid disease, arecent hypothesis postulates that embolism and hypoperfusion play a synergetic role, according to which small embolicmaterial prone to lodge in distal field arterioles would be more likely to result in cortical micro-infarcts when chronichypoperfusion prevails. Future studies combining imaging of brain perfusion, diffusion-weighted imaging, and ultrasounddetection of microembolic signals should help resolve these issues.  ( Stroke . 2005;36:567-577.)Key Words:  carotid artery occlusion    cerebral blood flow    stroke    tomography emission, computed. W atershed infarcts involve the junction of the distalfields of 2 nonanastomosing arterial systems. Classicneuropathologic studies 1 describe 2 distinct supratentorialWS areas: (1) between the cortical territories of the anteriorcerebral artery (ACA), middle cerebral artery (MCA), andposterior cerebral artery (PCA); and (2) in the white matteralong and slightly above the lateral ventricle, between thedeep and the superficial arterial systems of the MCA, orbetween the superficial systems of the MCA and ACA. Theformer, superficial areas have been commonly referred to asthe cortical watershed (CWS), and the latter have beenreferred to as the internal watershed (IWS).In autopsy studies, CWS and IWS infarcts—also termedexternal and internal border-zone infarcts, respectively—to-gether represent  10% of all brain infarcts. 2 However, becauseWS infarction is seldom fatal, this is probably an underestimate,and imaging studies in severe internal carotid artery (ICA)disease report an incidence ranging from 19% to 64%. 3–5 Althoughthepathological 6,7 andimagingcharacteristics 3,8,9 of WS infarcts are well-described, their pathogenesis remainsdebated. Based on the well-established notion that severe sys-temic hypotension can cause bilateral WS infarction, 7,10 hemo-dynamic failure is classically thought to cause WS infarcts inICA disease. 4,8,9,11 Susceptibility of the WS areas is thought toresult from their situation of “distal field,” where perfusionpressure is lowest, 12 and repeated episodes of hypotension in thepresence of severe ICA disease is regarded as facilitating WSinfarcts. However, the cortical distal field may not alwayscorrespond to the WS in situations in which a shift of the latter(as revealed angiographically by the pattern of leptomeningealanastomoses) has occurred because of additional hypoplasia orstenosis of the proximal ACA, PCA, or MCA. The occasionaloccurrence of syncope at onset of WS stroke, 13,14 and the typicalclinical presentation of episodic, fluctuating, or progressiveweakness of the hand, occasionally associated with upper limbshaking, are consistent with, and classically considered asmarkers of, hemodynamic failure. 3,15,16 This interpretation isfurther supported by radiological studies showing that WSinfarcts distal to ICA disease are more likely with a noncompe-tent circle of Willis. 17–19 In contrast, embolism from ICA disease Received November 12, 2004; accepted November 22, 2004.From the Department of Neurology and Stroke Unit, University of Cambridge, Cambridge, UK.Current affiliation for I.M.-M. is the Department of Neurology, University Hospital, Geneva, Switzerland.Correspondence to Dr J.-C. Baron, Department of Neurology, Addenbrooke’s Hospital Box 83, Cambridge CB2 2QQ, UK. E-mail jcb54@cam.ac.uk © 2005 American Heart Association, Inc. Stroke  is available at http://www.strokeaha.org DOI: 10.1161/01.STR.0000155727.82242.e1  567  Progress Review   b  y g u e  s  t   onD e  c  e m b  e r 1  3  ,2  0 1  7 h  t   t   p :  /   /   s  t  r  ok  e  . a h  a  j   o ur n a l   s  . or  g /  D o wnl   o a  d  e  d f  r  om   preferentially affects the stem and large branches of the MCA,producing cortical “wedge-shaped” and/or deep striato-capsularinfarcts. 17,19 In sharp contrast with this widely prevalent interpretation,several pathological reports emphasize the association of WSinfarction with microemboli arising from unstable carotidplaques or from the stump of an occluded ICA. Jorgensen andTorvik  2 and Torvik and Skullerud 6 were the first to report thatmost of the occlusions observed in the leptomeningeal arteriesover WS infarcts distal to ICA occlusion resulted from micro-emboli occluding the terminal vascular field, rather than beingsecondary to slowing of the cerebral blood flow (CBF). Beal etal 20 reported a patient in whom arm paresis developed aftermultiple transient ischemic attacks (TIAs) distal to ulcerativecarotid plaque and who was found at autopsy to have multiplepial arteries occluded by cholesterol emboli in the border zone.Pollanen and Deck  21 reported 3 cases in which embolization of thrombotic material (from the heart in 2 cases and from the ICAin 1) caused the CWS infarct. Masuda et al 22 found atheroma-tous embolism to cause CWS infarcts by occlusion of theterminal cortical branches with small emboli (50 to 300   m)mostly composed of cholesterol crystals, whereas territorialinfarcts were related to larger fibrin emboli. Importantly, there isexperimental evidence that small thrombi travel preferentially toWS areas because of their distinctly small size. 23 Interestingly,cerebralamyloidangiopathyhasrecentlybeenproposedasarisk factor for microinfarcts in the CWS areas. 24 Whereas these observations mainly applied to CWS infarc-tion, a recent pathological study 25 of 12 patients with IWSinfarcts suggested that ischemic lesions observed in the IWSarea may also involve an embolic mechanism, either cardiacor artery-to-artery. In the majority of lesions, histologyrevealed a significant component of incompletely infarctedbrain, which, according to the authors, would be consistentwith transient embolic occlusion. However, occluding mate-rial was not directly observed.Overall, therefore, there is considerable controversy re-garding the pathophysiology of WS infarcts in critical ICAdisease, with both the low-flow and the multi-embolic mech-anism being considered based on substantial evidence forboth. Interestingly, as is detailed, a synergetic association of these 2 mechanisms has been recently postulated. 26 The purpose of this article is to review the evidenceregarding the role of hypoperfusion versus emboli in thedevelopment of WS infarction in ICA disease, and to assesswhether the mechanisms may differ between cortical andinternal WS infarcts. Better understanding the underlyingmechanisms of WS infarction would help identify thosepatients at high risk for, and provide an evidence-basedrationale for preventing the occurrence and progression of,WS infarction. After a brief account of the anatomy of theWS and of the basic physiology of the cerebral circulation incircumstances of focally reduced perfusion pressure, theliterature concerning the mechanisms of WS infarction inICA disease is reviewed, with emphasis on the studies thatassessed brain perfusion and hemodynamics with ultrasoundtechniques or physiological imaging. The overall significanceof these findings is then briefly discussed. WS Infarcts: Anatomy, Structural Imaging,and Angiography The CWS regions are boundary zones where functional anasto-moses between the 2 arterial systems exist, ie, on the pial surfacebetweenthemajorcerebralarteries. 27 CWSinfarctsrepresentthemost familiar WS strokes. Anterior WS infarcts develop be-tween the ACA and MCA territories, either or both as a thinfronto-parasagittal wedge extending from the anterior horn of the lateral ventricle to the frontal cortex, or superiorly as a linearstrip on the superior convexity close to the interhemisphericfissure, whereas posterior WS infarcts develop between theACA, MCA, and PCA territories and affect a parieto-temporo-occipital wedge extending from the occipital horn of the lateralventricle to the parieto-occipital cortex. 1,28 IWS infarcts can affect the corona radiata (CR), betweenthe territories of supply of the deep and superficial (ormedullary) perforators of the MCA, or the centrum semiovale(CSO), between the superficial perforators of the ACA andMCA. 1 However, in carotid disease, ie, the focus of thisreview, it is unlikely that hemodynamic insufficiency willaffect equally the basal and the superficial MCA perforators,a situation that could, however, arise from eg, added MCAstem disease. Structural Imaging The impression that might be gained from the clinico-radiological literature is that the major WS areas occur atsymmetrical, predictable sites in the hemisphere. 29 However, forboth types of WS regions, there is substantial inter-individualand intra-individual variation. Using the minimum and maxi-mum areas of MCA supply (as defined by van der Zwan et al) 30 to assess whether infarcts distal to a hemodynamically signifi-cantICAdiseasewouldberegardedasterritorialorborder-zone,64% were considered as WS infarcts when using the minimumarea, but only 19% when using the maximum area of supply. 31 Identifying typical computed tomography (CT) or magneticresonance (MR) patterns associated with border-zone infarctionisthereforenotalwaysstraightforward.Inanyindividualpatient,it may be on occasions difficult to decide on the basis of brainimaging whether an infarct has arisen from occlusion of a smallcortical branch of the MCA or from hypoperfusion caused byestablished ICA disease.On the basis of their radiological appearance, IWS infarctshave been divided into confluent and partial infarcts. 15 Confluent infarcts correspond to large cigar-shaped infarctsalongside the lateral ventricle, whereas partial IWS infarctsmay appear either as a single lesion or in a chain-like (or“rosary-like”) pattern in the CSO. However, partial IWSinfarcts sometimes are difficult to distinguish from lacunar,medullary, or striatocapsular infarcts, as well as from leuko-ariosis. The latter, however, affects in a diffuse way theparaventricular WM bilaterally as it represents chronic dif-fuse white matter ischemia. Partial IWS infarct and leuko-ariosis may, however, coexist, particularly in the elderly.Regarding medullary infarcts, they correspond to small,immediately subcortical infarcts caused by occlusion of medullary arteries arising from the pial plexus. 32 They aregenerally smaller and more superficial than partial IWSinfarcts, 33 but IWS and white matter medullary infarcts have  568 Stroke  March 2005   b  y g u e  s  t   onD e  c  e m b  e r 1  3  ,2  0 1  7 h  t   t   p :  /   /   s  t  r  ok  e  . a h  a  j   o ur n a l   s  . or  g /  D o wnl   o a  d  e  d f  r  om   been sometimes lumped together as so-called subcorticalwhite matter infarcts because of the difficulty in distinguish-ing between them, further complicating classification.Illustrative examples of WS infarcts are shown in the Figure. Angiography: WS Infarct and the Circle of Willis Although a noncompetent circle of Willis should be regarded asan additional predisposing condition in WS infarcts from ICAdisease, 19 the role of supplency in the prevalence of WS infarctsis itself a matter of debate. The absence of collateral blood flowvia the anterior communicating (ACoA) and the posteriorcommunicating artery has been associated with WS infarcts(both CWS and IWS). 19,34 However, there are contradictoryopinions on the protecting role of collateral flow. Accordingly,collateralization through the posterior communicating artery hasbeen alternatively reported as protective 35,36 or without effect 37 ontheprevalenceofWSinfarction.Moreover,supplencyviatheanterior communicating artery was associated with a significantreduction in the prevalence and volume of IWS infarcts only,thus of no consequence for CWS. 37 Brain Perfusion and Hemodynamic Studies: BasicPhysiology, Methods, and Study Design The physiological response of the brain to reduced cerebralperfusionpressure(CPP)distaltoICAocclusionwasestablishedthanks to physiological imaging, initially positron emissiontomography (PET), 38–40 and subsequently SPECT. 41 The initialresponse to a decline in the CPP is an autoregulatory vasodila-tation of the resistive vessels (stage I of hemodynamic impair-ment). 42 This results in increased cerebral blood volume, longermean transit time, and impaired response to vasodilatory chal-lenge (hypercapnia or intravenously administered acetazol-amide). 38–40 With further reduction in the CPP, the autoregula-tory vasodilatation becomes inadequate and the CBF decreases.As neurons tend to maintain their oxidative metabolism, theoxygen tissue tension decreases and the oxygen extractionfraction (OEF) increases (“misery perfusion” 38 or stage II 42,43 ).Below the CBF penumbra threshold, neuronal function is im-paired and the affected tissue is at risk of infarction; 44 however,it is unknown if long-lasting reductions of CBF above thepenumbra threshold may also result in infarction—complete orincomplete. Materials and Methods Regardless of the technique used, all the studies reviewed here aimed todetect reduced CPP in the affected ICA territory, in the form of eithermisery perfusion or impaired perfusion reserve, ie, reduced vasodilatorycapacity, increased cerebral blood volume, or prolonged mean transittime. 43 In severe ICA disease, the CBF response may be abnormallyreduced or even absent because of maximal vasodilation; a focaldecrease in CBF may even occur (“steal phenomenon”). 45 Techniquesused have been either imaging-based, such as xenon CT, single-photonemission tomography (SPECT), PET, and MR-based perfusion(perfusion-weighted imaging [PWI]), or ultrasound-based, mainly trans-cranial Doppler sonography (TCD). One major difference betweenimaging techniques and TCD is that whereas the former allows one toassess perfusion directly in the brain region of interest (ROI), the latterassesses flow in the MCA trunk only, so it may lack sensitivity. Design Regarding design, studies reviewed here have either assessed therelationships between the presence of a WS infarct and the hemo-dynamic status of the carotid circulation, or directly measuredperfusion in or near the WS infarct per se, for instance by drawingROIs in the white matter of the affected CSO. Some studies,however, investigated patients without WS infarct and assessed thehemodynamic status in the WS areas. This diversity of designsoccasionally complicates the comparison among studies. Results Even though the study of cortical WS infarction has beenhistorically anterior and is still numerically superior to that of internal WS infarction, we review the latter first because, aswill be seen, its pathophysiology appears less controversialthan the former.Since 1981, 33 perfusion studies addressing the issue of WS pathophysiology in ICA disease have been publishedin the English language. Tables 1 to 4 list these studiesaccording to the investigative method used (PET in Table1; SPECT, Xenon 133, and xenon CT in Table 2; TCD in Table3; magnetic resonance imaging in Table 4). Note that 34 entriesappear in the Tables as 1 study used 2 different techniques.Although some studies lumped together CWS and IWS infarctsas “low-flow infarcts,” most did assess them separately or evendirectly compared them, which, as will be seen, turned out to becrucial.WithinIWSinfarcts,however,onlyaminorityofstudiesmade the important distinction between CR and CSO infarcts,another important issue. To facilitate reading, the findingsregarding the pathophysiology of IWS and CWS infarcts are Illustrative examples of watershed infarcts in patients with ICA disease. A, Right-hemisphere anterior watershed infarct on CT, affectinga superior strip between the ACA and MCA cortical territories. B, Right-hemisphere internal watershed infarct in the centrum semi-ovale(rosary-like pattern) on T2-weighted magnetic resonance imaging. C, Three DWI cuts from a single patient showing a right-hemisphereacute infarct involving both the cortical watershed (mainly the anterior but also slightly the posterior watershed) and the internal water-shed (rosary-like as well as confluent patterns). Momjian-Mayor and Baron Pathophysiology of Watershed Infarcts  569   b  y g u e  s  t   onD e  c  e m b  e r 1  3  ,2  0 1  7 h  t   t   p :  /   /   s  t  r  ok  e  . a h  a  j   o ur n a l   s  . or  g /  D o wnl   o a  d  e  d f  r  om   presented according to the following operational classification:(1) studies of “low-flow infarcts”; (2) studies that specificallyassessed the pathophysiology of IWS infarcts; (3) studies thatspecifically assessed the pathophysiology of CWS infarcts; and(4) studies that directly compared the pathophysiology of IWSand CWS infarcts. The studies that prospectively assessed theimpact of hemodynamic impairment on the risk of stroke inpatients with symptomatic ICA disease are also briefly reported. Studies of Low-Flow Infarcts Six studies, some early but also a few recent, did not separatelyassess IWS and CWS areas, 46–51 implicitly assuming that theirpathophysiology was similar. The number of patients enrolledranged from 11 to 102; all were symptomatic ICA disease(stenosis or occlusion). Two studies used TCD, 46,47 1 usedXenon-133, 48 and 3 used PWI. 49–51 All compared patients withlow-flow infarcts to patients with territorial infarcts, and allfound the former to be associated with a greater degree of hemodynamic compromise than the latter, in an area far exceed-ing the area of infarction. Studies That Specifically Assessed thePathophysiology of IWS Infarcts Nine studies focused on the pathophysiology of IWS infarcts(note that in none did the authors explicitly state thatassociated CWS infarcts were excluded). Three usedTCD, 52–54 1 used SPECT, 55 1 combined SPECT and TCD, 56 and 4 used PET. 57–60 They involved relatively small numbers TABLE 1. PET Studies  Authors and Year Reference Patient ICA Disease ROI Infarct Location (N) FindingsBaron et al 1981 38 1 Occlusion MCAT AWS, PWS None Misery perfusion in PWS, reversedafter EC–IC bypassSamson et al 1985 61 12 Occlusion (11) ACAT, MCAT, PCAT MCAT(4), ACAT(1),lacunar (1),Misery perfusion in AWS, PWSand MCAT, reversed in someMCAO (1) AWS, PWS CWS (5) pts after EC–IC bypassLeblanc et al 1987 63 7 Stenosis  80% MCAT IC (1) Impaired hemodynamics in AWS AWS, PWSLeblanc et al 1989 64 15 Occlusion (8) MCAT IC (3) Impaired hemodynamics in AWSUni/bil stenosis (7) AWS, PWSCarpenter et al 1990 65 32 Stenosis  50% orocclusionMCAT AWS, PWSNone No evidence of hemodynamicimpairment Yamauchi et al 1990 67 9 Occlusion MCAT, ACAT, PCAT CSO (8) Greater hemodynamic compromisein PWS AWS, PWS Yamauchi et al 1990 57 7 Occlusion PV WM, CSO CSO CSO lesions related to impairedhemodynamics Yamauchi et al 1991 58 16 Occlusion (11) ACAT, MCAT, PCAT CSO/CR (5) CSO lesions related to impairedStenosis (5) AWS, PWS Subcortical (6) hemodynamicsLevine et al 1991 68 18 Stenosis (8) MCAT None Impaired hemodynamics in AWSOcclusion (5) AWS, PWS with ICAS  50%Nonstenotic plaque (5)Levine et al 1992 69 16 Stenosis (8) MCAT None Impaired hemodynamics in AWSOcclusion (8) AWS, PWS Yamauchi et al 1999 59 7 Occlusion MCAT CSO Misery perfusion in the CSO AWS, PWSCSODerdeyn et al 2000 60 36 Occlusion MCAT MCAT/ No evidence of selectiveIWS BG; normal WM hemodynamic impairment in IWSDerdeyn et al 2001 76 110 Occlusion uni/bil MCAT MCAT Rosary-like pattern in IWS relatedCWS (9), IWS (17;rosary-like in 8)to hemodynamic impairment; nohemodynamic impairment in CWS Arakawa et al 2003 75 24 Occlusion (10)ICAS (6)IWS, AWS, PWSMCATLacunar or smallcorticalMisery perfusion more frequentlyobserved in IWSMCAS/MCAO (8) No hemodynamic compromiseobserved in CWS ACAT indicates anterior cerebral artery territory; AWS, anterior watershed; BG, basal ganglia; bil, bilateral; CR, corona radiata; CSO, centrum semiovale; CWS,cortical watershed; EC–IC, extracranial–intracranial; ICAO, internal carotid artery occlusion; ICAS, internal carotid artery stenosis; IWS, internal watershed; MCAO,middle cerebral artery occlusion; MCAS, middle cerebral artery stenosis; MCAT, middle cerebral artery territory; PCAT, posterior cerebral artery territory; pts, patients;PVWM, periventricular white matter; PWS, posterior watershed; ROI, region of interest; uni, unilateral; WM, white matter.  570 Stroke  March 2005   b  y g u e  s  t   onD e  c  e m b  e r 1  3  ,2  0 1  7 h  t   t   p :  /   /   s  t  r  ok  e  . a h  a  j   o ur n a l   s  . or  g /  D o wnl   o a  d  e  d f  r  om 
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!