Title: Posterior Circulation Ischemia: Then, Now, and Tomorrow
Abstract: HomeStrokeVol. 31, No. 8Posterior Circulation Ischemia: Then, Now, and Tomorrow Free AccessOtherPDF/EPUBAboutView PDFView EPUBSections ToolsAdd to favoritesDownload citationsTrack citationsPermissions ShareShare onFacebookTwitterLinked InMendeleyReddit Jump toFree AccessOtherPDF/EPUBPosterior Circulation Ischemia: Then, Now, and Tomorrow The Thomas Willis Lecture—2000 Louis Caplan Louis CaplanLouis Caplan From the Department of Neurology, Beth Israel Deaconess Medical Center, Boston, Mass. Originally published1 Aug 2000https://doi.org/10.1161/01.STR.31.8.2011Stroke. 2000;31:2011–2023To know and appreciate where we are now and where we are going in the future, it is essential to know where we have been. We cannot afford to relive and repeat the history of stroke every several decades. Posterior circulation stroke represents a microcosm of stroke in general. In this presentation I first review the development of ideas regarding brain and posterior circulation ischemia and its recognition and treatment. I then share some recent data from a large prospective registry of patients with posterior circulation ischemia. Finally, I look ahead to reflect on what I believe should be the future directions for research and for the care of patients with posterior circulation disease.Patients who present to physicians and hospitals with symptoms that suggest posterior circulation ischemia are handled differently from patients who have symptoms that suggest anterior circulation disease in the great majority of medical facilities in the United States and in the world. A patient who has an attack of dizziness with diplopia and ataxic gait usually has a brain image but seldom has vascular or cardiac investigations. A diagnosis of "vertebrobasilar insufficiency" (VBI) is often made, and physicians then debate whether or not to treat with warfarin-type anticoagulants, and, if so, for how long and at what intensity. In contrast, a patient who has right-hand weakness and aphasia is usually evaluated and treated quite differently at the very same facilities. Brain imaging, cardiac investigations, noninvasive vascular tests of the carotid and intracranial anterior circulation with the use of extracranial and transcranial ultrasound and/or MR angiography (MRA) and CT angiography, and catheter angiography are often pursued, depending on the local technological capabilities and experience of the treating physicians. An effort is made to identify the etiology and mechanism of the ischemia. Treatment is then chosen among a variety of possibilities (including carotid artery surgery, angioplasty, anticoagulants, and antiplatelet aggregants) depending on the nature, location, and severity of the occlusive disease and the mechanism of ischemia.Why should anterior and posterior circulation ischemia be handled so differently? Does this schizophrenic approach make sense? After all, the internal carotid artery and its branches and the vertebral (VA) and basilar arteries (BA) and their branches are just a few inches apart; they are made of the same coats and look the same under the microscope except for size. These vessels carry the same blood under the same blood pressure. The diseases that affect the blood vessels in the 2 circulations are the same. Do stroke mechanisms really differ between the 2 circulations? How did this differing approach originate, and does it continue to make sense today? These are some questions that I will attempt to answer as I review the development of ideas about posterior circulation ischemia and as I report recent data.Development of IdeasHerein I review how knowledge about the posterior circulation evolved. To be as concise as possible, I have eclectically selected key individuals and their contributions. I was fortunate to have been mentored by some of the individuals who have made key contributions during the second half of the 20th century.Clinicoanatomic CorrelationsThe first important question that physicians asked concerned the anatomy of the brain. What did the brain look like? How did it work? Which areas were responsible for which functions? One of the very first important observers was Sir Thomas Willis. Willis (1621–1675) was born soon after the deaths of Shakespeare and Queen Elizabeth. Great Britain was still basking in the artistic and cultural bloom of Elizabethan England. Willis was a very successful practicing physician and an accomplished organizer, teacher, and researcher. He performed necropsies on his patients and did extensive anatomic dissections, especially on the brain. His coworkers included the physicists Robert Hooke and Robert Boyle; Richard Lower, an anatomist, physiologist, and clinician who administered the first blood transfusion1 ; and Sir Christopher Wrenn, the renowned architect and artist. Wrenn is responsible for the engraved plates from which the illustrations in Willis' The Anatomy of the Brain and Nerves23 are derived.Willis became the Sedleian Professor of Natural Philosophy at Oxford University. His anatomy text contains detailed description of the brain stem, the cerebellum, and the ventricles, with extensive hypotheses about the functions of these brain parts. He was the first person to use the term neurology. Willis knew and collaborated with other 17th century giants: Sir Isaac Newton; John Locke, physician and philosopher; and William Harvey.After Willis, there was a relative lull in activity concerning brain anatomy and function until the latter years of the 19th century, when physicians, mostly in France, Germany, and the United Kingdom, reported case studies of patients that helped to elucidate the anatomy and functioning of the brain stem. The so-called classic brain stem syndromes, all eponymic and named after the original describers of the syndromes, were stimulated by a fascination of the authors with the anatomy and functions of the brain stem.4 We still recognize today these various constellations of findings as the midbrain syndromes of the following: Weber45 (ipsilateral third nerve paresis and contralateral hemiparesis); Benedikt4678 (ipsilateral third nerve paresis and contralateral hemiparesis, tremor, and involuntary movements); Claude8910 (ipsilateral third nerve paresis and contralateral limb ataxia with gait ataxia); the pontine syndromes of Millard-Gubler1112 (ipsilateral facial palsy and contralateral hemiparesis) and Foville13 (ipsilateral facial palsy and conjugate gaze paresis with contralateral hemiplegia); the medullary syndromes of Wallenberg1415 (lateral medullary syndrome) and Babinski-Nageotte16 (lateral medullary syndrome with a contralateral hemiparesis); and the thalamic syndrome of Dejerine-Roussy17 (contralateral hemisensory loss with contralateral ataxia and clumsiness and delayed onset of pain). Many of the lesions described in these reports were not vascular in etiology; some were tuberculomas, tumors, and focal infections. Although most reports were single necropsy-based case reports, some had no necropsy confirmation. Wallenberg's reports were particularly exemplary. He reported detailed clinical findings, predicted the location of the medullary lesion, and then later described the necropsy findings.1415The next important contributor was Joseph Jules Dejerine (1849–1917). Dejerine was a master clinician and anatomist.1819 He was a large man who created an imposing image on ward rounds (Figure 1). Dejerine was associated with the Salpetriere and Bicetre hospitals in Paris, and in 1910 he assumed the Charcot chair. His wife, Augusta Klumpke, was an accomplished clinician and artist.20 She is responsible for the elegant illustrations in Dejerine's 2 major contributions: his anatomy21 and semiology22 texts. Dejerine and Dejerine-Klumpke drew illustrative cartoons that depicted the symptoms and signs in patients with various brain stem and cerebral lesions. Figure 2 shows one of the cartoons depicting the anatomy and findings in a patient with a hemimedullary infarct. Dejerine described the findings in patients with different varieties of reading abnormalities and first described the syndrome of alexia without agraphia.23Charles Foix (1882–1927) was probably the first modern stroke neurologist. Foix was born in Salies-de-Bearn, a small village in southern France.2425 He spent his entire medical career within the hospital systems of Paris (Hotel Dieu, Necker, Bicetre, Salpetriere). He was a clinician, anatomist, revered teacher, writer, and poet. Within a 3-year period (1924–1927), he and his coworkers published an astonishing array of reports concerning the clinicoanatomic correlation of symptoms and signs with softenings at various sites in the cerebral hemispheres and the brain stem.26 Especially important in relation to posterior circulation disease were his studies of the thalamic syndromes,27 syndromes related to occlusions of the posterior cerebral arteries,28 and the lateral medullary syndrome.29Later clinicians clarified the clinical findings in patients with pontine infarction related to basilar artery occlusion30 ; patients with cerebellar infarction at various loci in the cerebellum3132333435 ; midbrain, thalamic, and occipital and temporal lobe infarction in patients with embolism to the "top-of-the-basilar" artery3536 ; and patients with small localized infarcts in the pons, medulla, and thalamus caused by disease of the penetrating artery supply.3537383940414243444546Vascular AnatomyConcurrent with the interest in how the brain looked and how it worked was an interest in how the different parts of the brain were supplied with blood. Thomas Willis was probably the first to study the circulatory supply of the brain in detail. He wrote the following about the anatomy of the vertebral circulation:as the Carotides carry the tribute of the blood to the brain; so the Vertebrals serve chiefly for watering the cerebellum and the hinder part of the oblong marrow. . . . The Vertebral Artery passes through little holes cut in the extuberances of the Vertebrae till it comes near the base of the skull and is admitted through the last hole. . . . Beneith the Cerebellum the Vertebral branches are united.2Willis is usually remembered as the describer of the vascular composition of the large arteries at the base of the brain, the so-called circle of Willis. He emphasized the capability for collateral circulation if an artery became blocked and the interconnection of blood vessels (Figure 3). A section in his anatomy text is devoted to "for what use the wonderful net is made, and the reason for it."2Charles Foix and his colleagues dissected and described in detail the arteries of both the anterior and the posterior circulation. They described the arterial supply of the thalamus,2747 the posterior cerebral artery and its branches,28 and the blood supply of the pons4849 and the medulla oblongata.29 Especially important was the description of the pattern of blood supply of the pons (Figure 4).4849The pattern of large median arteries, smaller paramedian arteries, and circumferential arteries is a model for the circulatory supply of the brain stem and also the cerebral hemispheres. Foix also analyzed the clinical findings expected in case of occlusion of the various pontine penetrating and circumferential supply arteries.49Duret5051 and Duvernoy52 in France, Stopford53 in England, and Gillilan54 and Stephens and Stilwell55 in the United States were also important contributors to knowledge of the arterial and venous anatomy of the posterior circulation.Vascular Pathology and the Mechanism of Brain InfarctionDuring the first half of the 19th century, the terms encephalomalacia, softenings, and ramollissements were in general use. These were all descriptive terms and did not indicate etiology. Not until the observations of Rudolf Virchow (1821–1902) was it established that arterial occlusions and diminished blood flow to brain regions were the cause of softenings and that these lesions were infarctions. Laennec had already used the term infarction for pulmonary apoplectic lesions.56 In 1846, Virchow performed 76 necropsies and found blood clots in 18 peripheral veins and 11 pulmonary arteries.5657 He concluded that the blood stream allowed transport of venous coagula for distances from their origins. He then described necropsy material in which thrombi originating in the left atria or cardiac valves blocked cerebral, splenic, and renal arteries. In animals, Virchow showed that foreign materials placed into the jugular vein traveled to the lungs and foreign materials placed in arteries also traveled to distant arterial sites.565758 Virchow showed that thrombi that formed within arteries were often caused by lesions of the arterial wall. Before his work, blockage of arteries was usually attributed to inflammation. Virchow introduced the terms thrombus, thrombosis, embolus, and embolism and deduced the general principles of thrombosis and embolism.56 Virchow's triad explained localized thrombus formation and consisted of the following: (1) an abnormality of the intima and vascular wall, (2) an abnormality of blood flow, and (3) an abnormality of blood coagulability. Virchow's pathological studies revolutionized thinking about brain infarction, thrombosis, and embolism.The early studies of Charles Foix related strictly to the localization of ramollissements (brain softenings) and their vascular supply and accompanying clinical findings. He and his predecessors had shown little interest in the nature and mechanisms of the vascular occlusive process. Several weeks before his death (Foix died at the age of 45 years, likely of a ruptured appendix), Foix and his colleagues Hillemand and Ley delivered a paper at a meeting of the Medical Society of the Hospitals of Paris concerning a study that they performed on the arteries that led to brain infarcts. Although an abstract of this report was published,59 a full article never appeared.Among 56 brains with infarcts, the artery supplying the infarcts was totally occluded in only 12 and subtotally in 14. In 30 patients the arteries were open. Foix and his colleagues speculated on possible explanations of the arterial patency: (1) arterial occlusion might follow softenings, (2) embolism with distal passage before necropsy, (3) insufficiency (l'insuffisance arterielle), that is, more proximally located circulatory failure, and (4) vasospasm (spasme arterielle).The next important contributor was Raymond Adams, a neuropathologist and clinical neurologist. With Charles Kubik, then director of the neuropathology laboratory at the Massachusetts General Hospital (MGH), Adams, who at the time was director of the neuropathology laboratory at the Mallory Institute of the Boston City Hospital, described the clinical and necropsy findings in 18 patients who at necropsy had occlusion of the basilar artery.30 Eleven occlusions were thought to arise in situ, while 7 were considered embolic. Adams and Kubik described the clinical findings and diagrammed in each case the location of the arterial occlusion and the resulting brain stem and cerebellar infarcts (Figure 5). They noted morphological distinctions between thrombosis and embolism, as follows:Thrombosis of the basilar artery could usually be recognized at a glance. The thrombosed portion of the vessel was distended, firm, and rigid and the thrombus could not be displaced by pressure. . . . In embolism, the embolus was usually lodged in the distal portion of the basilar artery.30Thrombosis was engrafted on arteriosclerotic lesions, while a displaceable embolus often blocked a normal-appearing artery. Thrombosis was often superimposed on emboli distally and/or proximally.30 Adams later became chairman of the Neurology Department at MGH, where he and his protege Charles Miller Fisher performed many important clinical and pathology studies of various stroke conditions.C. Miller Fisher is the individual probably most responsible for furthering information about stroke and stroke mechanisms during the 20th century. Fisher, a Canadian by birth, came to the Boston City Hospital and later to MGH to study neuropathology with Raymond Adams. He created the Stroke Service at MGH, the first of its kind in the United States. I was a Stroke Fellow with Dr Fisher in 1969–1970, at which time I also came under the tutelage of Raymond Adams.Fisher's 1951 report on occlusion of the internal carotid artery was a benchmark in the history of stroke.60 This article emphasized that occlusions commonly developed in the neck engrafted on atherosclerosis and that transient ischemic attacks (TIAs) often preceded and warned of the ensuing stroke. The carotid artery stenosis was possibly approachable surgically. Before this report, although in 1905 Chiari had described a patient with embolism arising in an occlusion of the internal carotid artery in the neck,61 anterior circulation infarcts were invariably attributed to middle cerebral artery disease. In 1954, Fisher reported subsequent observations on internal carotid artery disease.62 Later Fisher and his colleagues described the distribution of atherosclerotic lesions found at necropsy within the extracranial and intracranial anterior and posterior circulations.63 Within the posterior circulation, Fisher described occlusions of the vertebral artery in the neck64 ; with Kubik and Karnes he described the vascular pathology found at necropsy in patients with lateral medullary infarcts65 and emphasized that intra-arterial embolism ("local embolism") was an important mechanism of stroke in the posterior circulation as well as in the anterior circulation.66 In a series of meticulous analyses of serial sections from patients with small deep infarcts, many located in the brain stem, Fisher described the pathology in the penetrating arteries, lipohyalinosis and atheromatous branch occlusions, that caused penetrating artery territory infarcts.46676869Physiology of the Cerebral CirculationDerek Denny-Brown (1901–1980), a neurophysiologist, introduced and popularized the term cerebrovascular insufficiency to explain TIAs and the fluctuating nature of brain ischemia. Denny-Brown was born in New Zealand and trained there at the University of Otaga in Duneeden. A Beit Memorial Fellowship allowed him to work in the neurophysiology laboratory of Sir Charles Sherrington during 1925–1928. In 1928 he became house physician at the National Hospital Queens Square and St Bartholomew's Hospital in London. In 1941 he was appointed to the James Jackson Putnam Chair at the Harvard Neurological Unit at the Boston City Hospital. Raymond Adams was the neuropathologist in Denny-Brown's Neurology Department at the Boston City Hospital, and Miller Fisher also spent some time there working with Adams. Denny-Brown was my department chairman and mentor during 1966–1969 during my neurology training on the Harvard Neurological Unit. I was in the last group of neurology residents that Denny-Brown trained. Denny-Brown was a physiologist by training. Although he had assisted in necropsies in the neuropathology laboratory of Godwin Greenfield at the National Hospital and performed his own staining and microphotography,70 he considered that physiology was a dynamic, living discipline that explained many clinical neurological phenomena better than morphological analyses performed at necropsy. During the1950s and early 1960s he studied and wrote about hemodynamic considerations in patients with brain ischemia in the anterior and posterior circulations.717273 At that time, "vasospasm" was the popular explanation for TIAs.We therefore postulated an explanation alternative to that of vasospasm, namely, a state of carotid insufficiency determined by either stenosis or occlusion of the internal carotid artery, with its vascular territory left supplied by collateral branches. . . . A similar situation in relation to the basilar artery accounted for insufficiency of supply of the brain stem and posterior cerebral artery territory of that artery. On this basis, carotid or basilar insufficiency was a physiological, potential hemodynamic state, in which reversible hemodynamic crises could be elicited by any factor that impaired the collateral circulation.73Denny-Brown and John Sterling Meyer, his associate, attempted experimentally, using tilt-tables and blood pressure manipulation, to demonstrate the sensitivity of the circulation to hemodynamic perturbations, but, in general, these experiments failed.At about the same time, clinicians at the Mayo Clinic, Bob Siekert and Clark Millikan, reported a series of patients who had fluctuating symptoms affecting brain structures supplied by the posterior circulation arteries that they termed VBI.74 Siekert later served during 1976–1981 as the first program chair of the International Stroke meeting of the American Heart Association.75 Other clinicians, including Fang and Palmer in California76 and Denis Williams in the United Kingdom,7778 wrote about symptoms and signs in patients with VBI, and the term became popular on both sides of the Atlantic Ocean.TreatmentDuring the first half of the 20th century, there was little interest among neurologists in active treatment, other than rehabilitation, of patients with strokes. During the 1950s and 1960s, anticoagulants of the heparin and warfarin types were being used to treat patients with myocardial infarction and pulmonary embolism. Although Craven7980 had written about anticoagulant effects of aspirin, it was not yet used extensively. During this time, reports from the Mayo Clinic stroke service (Millikan, Siekert, and Whisnant) enthusiastically endorsed anticoagulant treatment of posterior circulation vascular disease (VBI).818283 Millikan et al81 first described 21 patients with progressing vertebrobasilar territory strokes who were treated with anticoagulants. Only 3 (14%) died compared with 10 of 23 deaths (43%) in a retrospective search for similar patients not anticoagulated. Attacks stopped in all 5 patients who had VBI spells. Later Whisnant described the results among 140 patients with progressive posterior circulation ischemia.83 Twelve (8.5%) died compared with 23 of 39 deaths (59%) in similar patients not anticoagulated. VBI attacks often stopped after anticoagulation.Although these studies were retrospective and uncontrolled and the causative vascular lesions were most often not studied, anticoagulants were widely accepted as the proven treatment for patients with vertebrobasilar territory ischemia. After all, the prevailing view at that time, from the report of Kubik and Adams30 and others, was that VBI was nearly always fatal or disabling. The fact that most of the anticoagulant-treated Mayo Clinic patients survived, many without major neurological deficits, and attacks of brain stem and cerebellar ischemia stopped was considered persuasive by clinicians.1960 to Mid 1980sAfter Fisher's reports on carotid artery disease,6062 angiography was widely used to detect extracranial vascular disease, and surgical carotid endarterectomy became popular. With the introduction of CT in the 1970s and more widespread use of extracranial ultrasound, patients with anterior circulation ischemia were usually investigated for the causative vascular lesions and stroke mechanisms. Various etiologies and mechanisms were defined, including the following: internal carotid artery disease in the neck (plaques, stenosis, or occlusion), intracranial internal carotid artery stenosis and occlusion, cardiogenic embolism, middle cerebral artery and anterior cerebral and anterior choroidal artery occlusive lesions, and penetrating artery disease with lacunar infarctions. The term carotid insufficiency was abandoned. Various treatments were used, depending on the nature and severity of the vascular lesions found. Popular treatments included aspirin, other platelet antiaggregants, anticoagulants, neck surgery, and extracranial-to-intracranial bypass surgery.Although angiography was widely used in patients with anterior circulation disease, it was considered risky in patients with posterior circulation disease. Early-generation CT scans were ineffective in showing brain stem and cerebellar infarcts, and ultrasound was rarely used to study the vertebral arteries. Cardiogenic embolism was considered a rare cause of posterior circulation ischemia. Patients with posterior circulation ischemia were classified as having VBI and seldom had cardiac and vascular testing. Therapeutic debate centered solely around anticoagulation.Posterior Circulation Disease: 1985 to the PresentThe past 15 years have seen a technological revolution. The advent of MRI, with its ability to better image posterior fossa structures, made investigation of vertebrobasilar territory infarcts more feasible. Transcranial Doppler ultrasound (TCD) and MRA provided methods of studying the vertebral and basilar arteries safely and quickly. Extracranial ultrasound was more often used, especially in Europe, to define lesions within the extracranial subclavian arteries and the vertebral arteries. Cardiac investigations also improved with better echocardiography, which showed lesions in the aorta as well as the heart. It became possible to investigate the brain and cardiovascular lesions and stroke mechanisms quickly and noninvasively in patients with posterior circulation ischemia. During this time a number of other treatments—surgery on the extracranial vertebral arteries (ECVA), extracranial and intracranial angioplasty, newer antiplatelet aggregants, and thrombolysis—were introduced and applied to patients with posterior circulation ischemia.I have always advocated thorough investigation of patients with both anterior and posterior circulation ischemia.3584858687 Beginning in 1988, when the technology became available, my colleagues Michael Pessin and Dana DeWitt and I, together with our stroke fellows, began to prospectively collect data in a computerized registry on all of our personally examined patients with posterior circulation strokes and TIAs. The New England Medical Center (NEMC) Posterior Circulation Registry was continued until 1996 and accumulated 407 patients. The clinical data and imaging studies on each patient were reviewed on multiple occasions, and a consensus of the findings and stroke mechanisms was made. The diagnostic criteria and a review of the results among the first 300 patients have been published.35NEMC Posterior Circulation Registry ResultsInvestigationsAll patients were thoroughly investigated. All had brain imaging (CT or MRI); >80% had MRI. All had vascular imaging (MRA, angiography, or TCD). Most patients had catheter angiography, nearly all in the early years but fewer as the diagnostic quality of MRA improved. Most patients had extracranial ultrasound and TCD of the posterior circulation arteries, and many had echocardiography and cardiac rhythm monitoring.Distribution of Brain InfarctsTo describe the location of infarcts, we subdivided the posterior circulation into proximal, middle, and distal intracranial territories. The proximal intracranial posterior circulation territory included regions supplied by the intracranial vertebral arteries (ICVA): the medulla oblongata and the posterior inferior cerebellar artery (PICA)–supplied cerebellum. The middle intracranial posterior circulation territory included the brain supplied by the basilar artery up to its superior cerebellar artery (SCA) branches: the pons and the anterior inferior cerebellar artery (AICA)–supplied cerebellum. The distal intracranial posterior circulation territory included all the territory supplied by the rostral basilar artery, SCAs, and the the posterior cerebral arteries (PCAs), and the penetrating branches of these arteries to the midbrain and thalamus. These subdivisions are shown diagrammatically in Figure 6, modeled after a figure in the Duvernoy atlas.52 The location of infarcts within the cerebellum is particularly useful in localizing the rostrocaudal location of infarction. The cerebellar blood supply is shown in Figure 7.We used both clinical and imaging data to localize infarcts. For example, if a patient had a lateral medullary syndrome and a hemianopia on examination, but MRI showed only a lateral medullary infarct, the patient was classified as having both proximal and distal territory infarcts. Infarcts were localizable in 347 of 407 patients (85.3%). The others had either repeated TIAs or persistent deficits that could not be definitively localized clinically or by brain imaging. The frequency of the brain locations of these 347 infarcts is shown in Table 1. The table notes those patients in whom proximal, middle, and distal territory segments were included as well as the specific locations, eg, proximal only, middle and distal, and proximal and distal. The most common location of infarcts was in the distal segment. Distal+ infarcts (ie, infarcts in the distal territory and also in other territories) were especially common.Stroke MechanismsThe distribution of stroke mechanisms among the registry of 407 patients is shown in Table 2. This table shows the single most likely diagnoses and also the range of diagnoses thought plausible. For example, in the first column embolism of cardiac origin was listed only when it was considered the most likely mechanism, but in the second column all patients with a potential cardiac embolic source were included. The most common stroke mechanism was embolism. The most common donor sources were the heart and vertebral arteries. Table 3 shows the localization of infarcts in patients in whom the single most likely stroke mechanism was embolism. Embolism caused mostly distal, proximal, and proximal and distal territory infarcts. The most common recipient arteries in patients with embolism were the ICVA and its PICA branches and the distal basilar artery and its SCA a