Of Common "Strokes" F

Postgrad Med J: first published as 10.1136/pgmj.42.483.5 on 1 January 1966. Downloaded from POSTGRAD. MED. J. (1966), 42, 5 CURRENT VIEWS ON THE PATHOGENESIS OF COMMON "STROKES" F. J. FAWCETT, M.B., M.'R.C.P. W. THoMAS SMITH, M.D., M.C.Path. Lecturer in Pathology Reader in . Neuropathology University"I ofI Birmingham- I

THE TITLE of th(is review is deliberately pro- out the 19th and early 20th centuries it was vocative and only becomes meaningful when the widely held that;cerebral infarcts were simply term "stroke" is defined. According to "A explained by occlusion of the supplying artery, New English Dictionary" (Murray, 1919) a this belief being reinforced by Cohn'heim's "stroke" is "an apoplectic or .... . paralytic description in 1872 of clearly defined vascular seizure"; apoplexy is described as "a malady territories and by the belief in end-arteries...... usually caused 'by an effusion of blood The role of anastomotic channels was seldom or serum in the brain ... . ". The survey will considered. The clinical and pathological be arbitrarily restricted to current views on features of occlusion of all main arteries were pathological aspects of commoner seizures ca,refully documented and many tantalizing classifiable as apoplectic by the above definition. syndromes, beloved of examiners, were at- "Effusions of blood" undoubtedly means tributed to occlusion of specific braniches of cerebral haemorrhage, and "effusions of the circle of Willis. Occasional dissenters such serum" would seem to refer to forms of as Nothnagel '(1877) noted that relevant arteries cerebral infarction such as acute colliquaition were not always occluded in brains showing or more chronic encystation ("apoplectic cyst"). infarction. The possibility of resorption of "Paralytic seizures" due to other causes are thrombus and the concept of reactive vaso-

thus outside our selfimposed terms of refer- spasm to explain these discrepancies were by copyright. ence. The account will be further limited by neither generally accepted nor confirmed. confining it to the three varieties of apoplectic The importance of the carotid arteries has stroke 'that have attracted most attention in long been known and the word carotid is an recent years. These are: cerebral infarction adaptation of the Greek term for "to stupefy"; related 'to atheromatous disease of the intra- Galen showed that carotid compression could cranial and/or carotid and vertebral arteries; cause "carus" or insensilbilhity. Todd (1844), subarachnoid haemorrhage and other con- Gull,(1855) and Ramsay Hunt (1914) recognised sequences of ruptured berry ("congenital") the importance of carotid occlusion in the aneurysm of the circle of Willis; and massive pathogenesis of cerebral infarction but the http://pmj.bmj.com/ intracerebral bleeding 'associated with systemic knowledge that bilateral occlusion was con- arterial hypertension. Infarotion resulting from sistent with survival and that carotid ligation unusual forms of arterial disease or from was not necessarily associated with cerebral emrboli of cardiac origin, and bleeding due to lesions temporarily discredited the former haemorrhagic disorders, trauma, tumours etc. avant-garde observations. The advent of are'therefore not included. arteriography in the 1930's istimulated con- Cerebral siderable interest in internall carotid artery Infarction thrombosis, the clinical syndrome first being on September 30, 2021 by guest. Protected Historical reported by iMoniz, Lima and de Lacerda Though Morgagni (1761) distinguished bet- (1937). However, it was only recently that the ween cerebral 'haemorrhage ("sanguineous carotid and vertebral systems together with the stroke") and cerebral softening '("non-sanguine- circle of Willis were considered as an integral ous stroke") many years elapsed before the unit in a detailed clinical and pathological cause of infarction was suspected. Abercrombie investigation by 'Hutchinson and Yates (1957) (1828) defined its ischaemic nature and held and Yates and Hutchinson '(1961). that the effects of local arterial 'lesions and of extracranial factors were important. Through- The role of carotico-vertebral stenosis Based on a Joint communication presented at a Symposium on Cerebro-vascular Disease, organised Yates and Hutchinson fully investigated 100 by the Board iof Graduate Studies of the Faculty of cases in which the clinical picture suggested Medicine, University of Birmingham, on June 26th, that cerebral ischaemia had either caused or 1965. contributed to death; cases recognised as intra- Postgrad Med J: first published as 10.1136/pgmj.42.483.5 on 1 January 1966. Downloaded from 6 POSTGRADUATE MEDICAL JOURNAL January, 1966 cranial haemorrhage or cerebral embolism by arterial disease; both of these effects could complicating heart disease were excluded. Theey bring about temporary disturbances of cerebral examined the whole cerebrovascular system, function, especially when associated with sludg- including the carotid and vertebral arteries and ing of blood in the capillary bed. These conclu- circle of Willis, together with the brain. Because sion were in general supported by the findings of the technical difficulties involved in removing of McGee, McPhedran and Hoffman (1962); the intraosseous and intracranial parts of the Baker, Dahl and Sandler (1963); and Fisher, vertebral and internal carotid arteries at Gore, Okabe and White '(1965). necropsy thhis procedure had been neglected in The significance of non-occlusive stenosis previious work on cerebral infarction. The in reducing blood-flow has been questioned impressive results of 'Hutchinson and Yates by Mitchell and Schwartz (1965) on the grounds showed that atheromatous diseases of the extra- that the series of Hutchinson and Yates (1957), cranial arteries iis of paramount significance in selected because of cliical evidence of cerebral the pathogenesis of cerebral ischaemia and ischaemia, did not show a signilficantly different infarction. They found occlusion and/or serious prevalence of carotico-vertebral stenosis from stenosis of one or more neck arteries in 58 that found in their own entirely unselected of the 100 cases; in 18 only the carotids and necropsy series. Amongst 93 adults over the in 7 only the vertebrals were affected but in age of 35 only 27 'had clinical evidence of 33 both were affected. In 35 cases there were "strokes". They found stenoses at some point in all 74 separate infarcts; all but 3 of these along the vessels in 90%/, of men and 85% of 35 cases showed significant stenosis (defined women which were severe in 43%/O of men and as 50% or more reduction of lumen) or 35% of women. Furthermore, severe narrow- occlusion of the extracranial arteries; only ings were present in the carotid sinus in one 19 of these same cases (showing 22 of the man in 7 and in the vertebral arteries in one infarcts) were associated with significant changes man in 4. Mitchell and Schwartz commented in ithe intracranial arteries. Thus, in 16 (46%) that 90% of the peripheral resistance in the by copyright. of the cases with infarcts signfificant dfisease cerebral circulation is contributed by the of the intracranial arteries was excluded. The smaller "intracranial vessels and considerable site of the disease in the internal carotid reduction in the calibre of the major neck arteries was most commonly the carotid sinus. arteries can occur before their narrowing be- In the vertebral arteries atherosclerosis was comes a controlling factor in brain blood-flow. somewhat more widespread but also tended They argued that their own findings showed to involve the proximal 2 cm. In 19 cases the relation between carotico-vertebral stenosis showing thrombotic occlusion of one or more and "strokes" to be uncertain and suggested extracranial arteries all but 4 showed cerebral that the stenosed segments do not usually http://pmj.bmj.com/ infarction, and 2 of these almost certainly seriously impair blood-flow but are more im- died before gross changes had time to develop. portant 'as a source of emboli which block the Yates and Hutchinson noted that distortion smaller cerebral vessels. They also claimed that and obstruction of diseased vertebral arteries as the frequency of "strokes" and of carotico- could 'be produced or aggravated by osteo- vertebral stenosis are both correlated with age, arthritis of the cervical spine as subsequently it is inevitable that there will be a correlation appraised by Brain (1963). They also em- between them, even if causally unrelated. It phasised the role of general systemic factors is unfortunate that Mitchell and Schwartz on September 30, 2021 by guest. Protected such as anaemia and blood loss, anoxia, hypo- neither specified the variety of "stroke" to tension and traumatic shock, and concluded which they alluded, nor described whether that cerebral infarction rarely had a single cerebral infarction, massive haemorrhage or cause; it was usually the result of a combination other lesions were found post mortem in the of systemic disease and inadequacy of total small number of cases on which they based cerebral blood-flow due to stenosis or occlu- their conclusions; they also ignored the more sion of extracranial or intracranial cerebral detailed report of Ytates and Hutchinson (1961). arteries, or both, infarction being commoner The role of cervical artery stenosis in massive with extracranial ithan intracranial arteriopathy. haemorrhagic strokes could be of paradoxical They further suggested that atheroma of the significance, and even protect the bra'in against carotid sinus could cause intermittent hypo- the damaging effects of hypertension; the de- tension by affecting the carotid reflex mechan- monstration by Wilson and Byrom (1939) that ism, and that vasospasm of the major extra- hypertension induced experimentally by con- cranial cerebral arteries was sometimes induced trolled stenosis of a renal artery does not Postgrad Med J: first published as 10.1136/pgmj.42.483.5 on 1 January 1966. Downloaded from

January, 1966 FAWCETT and SMITH: Pathogenesis of Common 'Strokes' 7 result in homolateral renal vascular lesions, It was concluded that a stenosis will significantly is of interest in this respect. It is a common reduce blood-flow if its cross-sectional area is experience that any group of cases diagnosed less than 2 sq. mm. If the minimal area is on clinical grounds as "cerebral infarction" more than 5 sq. mm. blood-flow will not be will include a number of unexpected "haemor- reduced. Between 2 and 5 sq. mm. iblood-flow rhages", neoplasms etc. and in Yates and may possibly be reduced, depending on the Hutchinson's cases of "cerebral ischaemia", shape, length and number of stenoses. These from 'which clinically confirmed haemorrhages observations suggest that the lumen of the were excluded, neoropsy showed haemorrhages internal carotid artery can be narrowed to in 28% land neoplasms in 6%. Though data less than 2 mm. diameter before flow is reduced concerning the terminal neurological state and appreciably. Pathological assessments of arterial the appearances of the brains at necropsy were stenoses '(as recommended by the World Health not given, it is probable that some of Mitchell Organization) usually classify reduced lumen and Schwartz's patients dying from general diameters of 50% or more as "severe"', whereas medical diseases such as cardiac or respiratory the stenosis that Brice and others (1965) re- failure, cancer etc. did in fact succumb to gard as possibly significant (2-5 sq. mm.) cerebral disorder resulting from ithe combination corresponds to a reduction of the order of of carotico-vertebral stenosis and adverse 84-93%. If this is correct, many of the stenoses haemodynamtic effects brought albout by these classified by Yates and Hutchinson and by general diseases. This is a common mode of Mitchell and Schwartz as severe, probably death in hospital practice and often explains had negligible effect on resistance; this in- why other patients with similar diseases, ad- ference reinforces the suggestion that the mitted 'to hospital for simillar reasons, respond significance of 'stenotic segments in the cervical favourably to appropriate treatment. carotico-vertebral arteries may 'be related to Gunning, 'Pickering, Robb-Smith and Ross- embolus formation rather than reduction of Russell (1964) found 'that frialble emboli, com- blood-flow by coarctation. by copyright. posed of fused platelet masses detached from The result of carotico-vertebral occlusion mural thrombi in the internal carotid artery, Whilst there is therefore some doubt as to and possibly other sites, were the commonest the precise role of neck-artery stenosis in the cause for 'intermittent attacks of blindness and causation of cerebral infarction, the importance contralateral hemiparesis; the commonest of thrombotic occlusion is unquestionable. cause for intermittent attacks of blindness or Mitchell and Schwartz found occlusion in 19% hemiparesis; and a possible cause for a single of their 27 patients with clinical evidence of attack of monocular blindness or hemiparesis. "strokes" and Yates and Hutchinson in 19 of These authors also discussed four possiible their 100 cases of cerebral ischaemia, and 43%/, http://pmj.bmj.com/ effects of mural thrombi in the internal carotid of their 35 cases with macroscopical infarction arteries: emboli may become detached, imipact of the brain. Carotico-vertebral occlusion, how- at a branch of appropriate size, break up if ever, can occur in patients who have shown sufficiently friable in consistency, reimpact in neither clinical evidence of cerebral ischaemia smaller distal 'branches, and perhaps repeat nor necropsy evidence of infarction (Rob, 1961; this process several times till they ultimately Roberts, 1965). Agenesis of an internal carotid disappear; they may cause -permanent embolic artery may also be found without related on September 30, 2021 by guest. Protected blockage of cerebral or retinal arteries; they lesions in extreme old age (Turnbull, 1962). may remain in situ and become organised and In such cases the development of adequate incorporated into the artery wall; and they may anastomotic chiannels is the determining factor: propagate and eventually occlude the whole these are derived from the meningeal anasto- internal carotid artery and perhaps even its moses on the surface of the brain between all anterior 'and middle cerebral branches. of the major arteries ,(vander Eecken and Important observations on the haemodynamic Adams, 1953); the circle of Willis (Hutchinson, effects of carotid artery 'stenosis have recently 1962); and the anastomoses between the internal been made by Brice, Dowsett and Lowe (1965) and external carotid arteries (Weiss, 1938). on a series of patients undergoing carotid Collaterals develop much more effectively if ligation for the treatment of berry aneurysms. the occlusion 'is gradual rather than sudden. The dimensions of an artificial constriction The functional capacity of the circle of Willis which just produced a detectable change of may be 'impaired by disease or by anomalies pressure gradient and flow in the internal carotid such as aplasia or hypoplasia; the latter occur artery were deternmined under local anaesthesia. in at least 40% of all individuals (Alpers and Postgrad Med J: first published as 10.1136/pgmj.42.483.5 on 1 January 1966. Downloaded from 8 POSTGRADUATE MEDICAL JOURNAL January, 1966 Berry, 1963; Riggs and Rupp, 1963), though balance is prejudiced by episodic haemodynamic their true signlilficance in association with neck- crises as suggested *by Denny-Brown (1951). artery disease remains to be determined. It Di-sequilibrium is precipitated by a wide range is therefore probable that carotico-vertebral of systemic disorders such as heart disease, occlusion is most likely to cause cerebral in- haemorrhage, anaemia, anoxic anoxia, trau- farction when emboli are detached, when matic or surgical shock, medical or surgical thrombus is propagated into the anterior and hypotensive therapy, altered water/electrolyte middle cerebral or basilar arteries, or when the balance (with dehydration, diuretic therapy other cervical arteries or circle of Willis are etc.), as well as by local factors such as vertebral severely diseased or show a signlificant deve- artery compression 'by osteoarthritic spurs lopmental anomaly such as aplasia of an during movement of the head and neck important segment. (especially rotation and extension), or transient Mitchell and Schwartz found that neck vascular occlusion due to 'progressive fragmenta- artery lesions tend to be bilateral and occlu- tion 'and ultimate dissolu-tion of fragile micro- sion is frequently associated with stenosis on emboli, as previously noted. In the recently the contralateral side and Hutchinson and Yates described isubclavian-steal 'syndrome, reviewed emphasised that if one of the carotid or verteb- by Toole (1964) and Patel and Toole (1965) ral vessels is involved there is approximately stenosis or occlusion of the first part of the an 80% chance that one of the remain)ing subclavian or innominate artery causes reversal vessels will also be affected. Not only does this of blood-flow in the vertebral artery; haemo- prejudice the development of adequate collateral dynamic crises occur when blood is temporarily circulation but it also makes it difficult to shunted from the vertebro-basilar territory relate the site of vascular changes to the (brain-stem, cerebellum, hindbrain and inner distribution of 'brain lesions. Carotid artery ear), and the resultant ischaemia induces occlusion usually resuilts in infarction in the transient disturbances such as vertigo, weakness middle cerebral artery distribution while of the 'limbs, visual upsets, ophthalmoplegia, by copyright. vertebral occlusion usually results in brain- ataxia etc. Precipitating factors include head stem and hind-brain infarction (Yates and or homolateral upper limb movements and Hutchinson). When there is combined carotico- postural hypotension. vertebral stenosis the junctional territories will be The rising incidence of cerebral infarction 'most vulnerable to consequential ischaemia, Yates (1964) showed that during the pre- whatever the cause. Dickinson (1961) found vious five years cerebral infarction had become that the site of extracrantial stenosis or occlusion a commoner form of "stroke" than cerebral had little effect on ithe localisation of cerebral haemorrhage, though in 'the 1930's haemorrhage damage and that disproportionate arterial http://pmj.bmj.com/ involvement on one side was just as likely was three times as frequent. This change in to cause infarction in the contralateral as in pattern was apparently due to a gradual fall the homolateral hemisphere. A corollary to in deaths from haemorrhage and an abrupt rise these observations is that it is probable that in cerebral infarction between 1947 and 1955. cerebral ischaemia due to carotico-vertebral The evidence was adduced from the Registrar disease might be treated effectively by improv- General's loverall statistics for deaths, and con- ing the blood-carrying capacity of any of the firmed by comparison with the necropsy

findlings on 3,397 "strokes". TIhe incidence of on September 30, 2021 by guest. Protected four 'major neck-vessels, irrespective of their fatal cerebral haemorrhage in men and women apparent relation to the location of cerebral was the same at almost all ages. Cerebral changes. Evidence supporting this view is found infarction has always been commoner in men in recent papers by De Bakey, Crawford and than women of the same age and the incidence Fields (1961) and Humphries, Young, Beven, in women increases with age parallel to that Le Fevre and de Wolfe (1965). of men but with a constant lag of up to two The importance of haemodynamic crises years. Age and cohort analysis of the available Because of the variable distribution and statistics showed an apparent reduction of severity of carotico-vertebral and intracranial cerebral infarction, 'similar in all age groups atherosclerosis, and the anastomotic vagaties of at risk, during 1940-1945, the years of the war. a transformed cerebral circulation brought on It was suggested that this was due to suppres- by stenosis and/or occlusion, recurrent and sion of certain complications of atherosclerosis transient attacks of carotid and particularly -possi(bly thrombosis--by a rapidly acting un- vertebro-basillar ischaemia '(Marshall. 1964; known factor either introduced or removed in Williams, 1964) can occur when the delicate the war years; this factor was probably a dietetic Postgrad Med J: first published as 10.1136/pgmj.42.483.5 on 1 January 1966. Downloaded from January, 1966 FAWCETT and SMITH: Pathogenesis of Common 'Strokes' 9 restriction (such as fats or sugar) rather than Carmichael) which only occur in adults. Small increased physical activity, as the less active defects were numerous, increased in frequency elderly benefited as much as the more active with age, occurred throughout the circle of younger cohorts. Willis and were prevalent in the lateral angle and apex of birfurcations; large defects were Subarachnoid Haemorrhage, and Other Con- infrequent, uncommon in the posterior half of sequences of Ruptured Berry ("Congenital") the circle of Willis, only occurred at the apex Aneurysm of the'Circle of Willis. of bifurcations where aneurysms were found, The aetiology of cerebral aneurysms and were less abruptly demarcated from the Though the main recent interest of patho- normal 'adjacent media. The age distribution logists in this problem has been in the nature of aneurysms- mean age 50.2 years, o}nly 22.5% and mechanism of the cerebral lesions that occurring under the age of 40 years, very few commonly complicate isubarachnoid haemor- under the age of 20 years-favoured an rhage, the aetiology of the responsible berry acquired origin. Stehbens also reported that aneurysm is still in doubt. Medial defects at developmental anomalies of the cerebral arteries bifurcations of the cerebral arteries in man were no more frequent (in association with (Forbus, 1930; Glynn, 1940; Stehbens, 1959) aneurysms. and animals (Stehbens, 1963a) are very common, but their role in the development of human The consequences of rupture aneurysms has 'been questioned (Glynn, 1940; The main consequences of ruptured berry Stehbens, 1959), and comparable aneurysms aneurysms are subarachnoid and intrauerebral in animals are almost unknown (Stehbens, haemorrhages and cerebral infarction. The 1963a). Similar medial defects in the appearance of diffuse haemorrhage within the media of extracranial arteries of man subarachnoid space is well known, adequately (Forbus) and animals '(Stebbens, 1963a) are not described in standard texitbooks and presents associated with aneurysm formation. Glynn no problems. Little factual information has been by copyright. concluded that medial defects played no part available until recently however, on the intra- in the development of berry aneuryisms as their cerebral complications of ruptured aneurysms. occurrence in 80% of cerebral bifurcations does Intracerebral haematoma not furnish a 'satisfactory explanation of a Intracerebral haematoma may be the result lesion found in less than 10% of all sulbjects; of secondary rupture of a locaised subarachnoid he demonstrated experimentailly the importance haematoma situated in a large fissure (most of the internal elastic flatmella in maintaining commonly interfrontal with aneurysms of the the form of the cerebr-al arteries and anterior communicating or anterior cerebral ascriibed aneurysms to degeneration of elastica. arteries, or Sylvian with middle cerebral http://pmj.bmj.com/ Richardson and Hyland '(1941) thought aneurysms) or of direct rupture of an aneurysm that *the medial defect determined the site into the brain, particularly when the sac Is of some aneurysms but that an additional adherent to or buried in the bralin substance. acquired lesion played an essential part in their Tomlinson '(1959) and Crompton (1962) showed formation. Carmichael (1950) held that aneur- that: anterior communicating and anterior ysms always arose at the site of substantial cerebral aneurysms were most defects in the muscle and elastic coats and that commonly associated with 'initracerebral haematomas in the on September 30, 2021 by guest. Protected while developmental deficiency was mainly frontal lobes, the cavi,ty of the septum pelluci- responsible for the muscle lesions, degenerative dum or the white-matter of the corpus callosum; processes, mainly of atheromatous type, caused middle cerebral aneurysms usually resulted in changes in the elastica; the chance occurrence haematomas in the region of the external of elastic and muscle defects of adequate size capsule and less commonly in the frontal or probably explained the low incidence of temporal lobes; posterior communicating aneurysms. Crawford '(1959) reached gimilar aneurysms frequently ruptured into 'the temporal conclusions (but stressed the additional role of lobes; and aneurysms at the terminal bifurcation high blood pressure. Stehbens (1963 b and c) of the internal carotid were often embedded thought 'that berry aneurysms were probably in and ruptured directly into the frontal lobes. acquired. He concluded that there was a Intracerebral rupture of aneurysms can mimic distinction -between the small "congenital" massive hypertensive haemorrhage and the medial defects (of Forbus) which occur in the pathological diagnosis may necessitate careful foetus and infant as well as adults, and the study to establish a correct diagnosis. Robertson larger early aneurysmal changes (described by (1949) noted that recurrent haemorrhages from Postgrad Med J: first published as 10.1136/pgmj.42.483.5 on 1 January 1966. Downloaded from 10 POSTGRADUATE MEDICAL JOURNAL January, 1966 aneurysms were more ljikely to give rise to recent surgical intervention, systemic hypo- intracerebral rupture. Crompton '(19.62) found tension and atheroma or anomalies of the circle that recurrent haemorrhages were also more of Willis; these factors are rarely associated commonly associated with ventricular dilatation with comparable infarction in the absence of following intracerebral or intraventricular rup- aneurysm. ture, possibly as the result of secondary arach- Haematomas are important because the noidal adhesions obliterating the exit foramina vessels traversing them show changes such as of the 4th ventricle. Intracerebral haematoma kinking and compression, mural oedema, is more flikely to be associated with fatal coma thrombosiis and necrosis; blood may also track than is uncomplicated subarachnoid bleeding along the perivascular sheaths of basal perfora- and examination of the brains of patients who ting arteries and impair the blood supply to have survived previous haemorrhage(s) rarely basal ganglia. shows evidence of old 'haematomas. The existence and significance of cerebro- Cerebral infarction vascular spasm has been frequently discussed. The incidence of cerebral infarction following There seems no doubt from experimental evi- the rupture of cerebral aneurysms is un- dence (Raynor and McMurtry, 1963; Corday doubtedly high; recent figures are 55%10 overall and otheis, 1963), angiographic evidence (du (Schneck,1964) and 80% of women and 69% Boulay, 1963; Allcock and Drake, 1965) and the of men (Crompton, 1964a). In the past, infarc- testimony of neurosurgeons (Pool, 1958) that tion has often been overlooked because it is cerebral vasospasm occurs in laboratory animals often difficult to distinguish macroscopically and man under a wide variety of circumstances, and is often edlipsed by the more dramatic but its importance has been contended (Picker- appearance of sulbarachnoid or intracerebral ing, 1948). Small arteries of a diameter exceed- haematomas. The lesions are ing about 400-500u 'have limited but distinct patchy, palle lamellae of smooth muscle in their media and or rarely haemorrhagic in colour; they by copyright. may be abundant, and though usually most would seem to (be capable of constriction; numerous in the territory of the aneurysm- arterioles of a smaller calibre 'have little or no bearing artery, also occur throughout the brain, medial muscle and probably do not constrict. even in the opposite hemisphere. The grey- At necropsy on a case of subarachnoid matter of the cerebral cortex and basal ganglia haemorrhage due to a ruptured anterior cerebral and to a lesser extent the central white-matter aneurysm, Small and Smith'(1964) found fibri- are affected to a variable extent; the cerebellum noid necrosis of an aneurysm-bearing artery and brain-stem are rarely involved. Histologi- which had previously shown spastic narrowing cally the lesions show variable ischaemic on angiography; it therefore appeared probable changes, depending on the avatilability of anas- that the necrosis resulted from vasospasm. http://pmj.bmj.com/ tomotic circulation; the degree of phagocytic Comparable fibrinoid necrosis of cerebral arter- infiltration and reactive gliosis is correlated with ioles can follow-on the spasm which is the survival period and patients dying from characteristic of experimental hypertensive recurrent haemorrhage months or years after a encephalopathy in rats (Byrom, 1954). Recent previous episode may show marked gliotic investigations by one of us (W.T.S. unpublished scarring. "Oedematous" iswelling of the central data) indicates that vascular necrosis is a rare

white-matter is commonly associated with the sequel of sulbarachnoid 'bleeding. on September 30, 2021 by guest. Protected infarcts and is probably an important factor The vasospastic istimulus in subarachnoid in elevating the intracranial pressure and im- haemorrhage is conjectural. It could be the pairing blood-flow, so perpetuating a vicious reflex effect of aneurysmal rupture or the irri- circle of raised pressure; hippocampal, cingulate tative effect of 'blood or blood-clot, though and/or tonsillar herniations are frequently seen. homolateral vessels some distance fro-m signifi- The infarcts do not appear to be due to micro- cant haematomas, or even contralateral vessels thrombosis. can show spasm. Large haematomas and sys- temfic hypertension are commonly present The cause of infarction (Allcock and Drake, 1965). Operation certainly The mechanism of infarction has been dis- seems to bring on spasm in many cases, cussed by Tomlinson '(1959), Barbara Smith especially when there is a short period ;between (1963), Crompton '(1964b) and Schneck who bleeding and surgery, and when hypothermia, found a positive relationship to factors such as excessive hyperventilation and possibly fluo- the presence of large subarachnoid haematomas, thane anaesthesia are used; these findings led angiographic vascular spasm or narrowing, Allcock and Drake to suggest delaying operation Postgrad Med J: first published as 10.1136/pgmj.42.483.5 on 1 January 1966. Downloaded from January, 1966 FAWCETT and SMITH: Pathogenesis of Common 'Strokes' 11 for several weeks if possible. Surgery should be basal nuclei, where bleeding most commonly planned to avoid disturbing haematomas and occurs. the altered blood vessels Itraversing them. The adverse effects are Micro-aneurysms of angiography considered by In 1868 Charcot and Bouchard made the Schneck 1(1964). first significant investigation of cerebral ihaemor- Spasm could also be related to the liberation rhage and demonstrated saccular or fusiform of chemical suibstances, such as 5-hydroxytryp- miliary, or more correctly su(bmiliary tamine (5'HT; serotonin) from the blood clot. aneurysms on the small cerebral arteries of 5HT was shown experimentally to be a cerebral 250-400,u diameter. In the adventitia of these vasoconstrictor by Raynor, McMurtry and Pool vessels they descrifbed periarteritic changes, (1961). Honour and Mitchell (1964) found that responsilble for the aneurysms, rupture of both human and ra-bbit red blood cell extracts which resulted in the 'haemorrhage. The and 5HT gave rise to platelet clumps ("white aneurysms -were demonstrated iby immersing bodies") which built up and embolized when the brains in running water for many weeks, appllied to minimally injured rabbit cerebral until only the vessels were left; the unpleasant arteries; red cell extracts but not 5HT also smell associated with the procedure discouraged caused vasoconstriction; -these observations and further research along the same lines for the possible role of platelet clumping merit many years. Nevertheoless, later reports of sub- further study in relation to subarachnoid miliary aneurysms were occasionally made by haemorrhage. Buckell 1(1964) demonstrated sub- workers using other techniques (Green 1930; stances capable of contracting smooth muscle Matsuoka, 1952). Ellis (1909) however, con- in fluid taken from haematomas resulting from sidered that these lesions were in fact false rupture of cerebral aneurysms. If such sub- aneurysms, the resulit rather than the cause of stances do play a part in the development of arterial rupture; this contention 'has often been spasm, it is conceivable that antagonistic drugs in of their doubtful quoted support significance. by copyright. could be used therapeutically. Preliminary ex- Recently 'Ross-Russell (1963) carried out a periments along these lines using synthetic anti- careful examination of a series of hypertensive serotonin agents were not significantly effective brains, which were injected with a radio-opaque (Raynor and McMurtry, 1963). gelatin solution and examined by microradio- There ils little doulbt that cerebral infarction after is the cause of many of the residual clinical dis- graphy 'fixation. He found aneurysmal abilities, often profound, that Walton (1956) dilatation of 300-900,u on the small cerebral found in albout 70% of patients surviving arteries of 100-300ju diameter. These aneurysms subarachnoid occurred commonly but not exclusively at the haemorrhage due to ruptured site of branching and were present in 14 out aneurysms; larger, focal lesions due to haema- of 15 hypertensive brains and 10 out of 38 http://pmj.bmj.com/ tomas are much more likely to be immediately of a control normotensive group. The number lethal. of aneurysms in each brain varied fm 1 to 20 Massive Hypertensive Cerebral Haemorrhage and were most frequent in the putamen, pal- The association of cerebral haemorrhage and lidum and thalamus, but also occurred in the high blood pressure 'has been well-known since caudate nucleus, internal capsule, centrum Wepfer (1658) 'first associated strengthening of semiovale and cortical grey-matter. With one the pulse with apoplexy, and demonstrated in- exception, the 'brains showing more than 10 on September 30, 2021 by guest. Protected traventricular clot and destruction of the brain aneurysms 'belonged to the hypertensive group by haemorrhage and since Baglivi (1699), in the which were categorized by the presence of a necropsy report on 'Marcellus Malph'igi, the diastdlic blood pressure of 110 mm. Hg. or Pope's physician and renowned anatomist, higher. The heart of 'this exception weighed described both intraventricular blood in the 450 g. which in the absence of valvular or brain ischaemic heart disease could be held to indicate and 'hypertrophy of the left ventricle systemic hypertension. Histologically, the of the heart. Thoug,h the exact mechanism affected arteries showed marked medial filbrosis is still uncertain the massive nature and and loss of muscle fibres with fraying and sudden onset, wit-h rapid rise in intracranial reduplication of the single internail elastic pressure suggests that the haemorrhage is lamella. The aneurysms showed a fiibrous, non- arterial rather than venous or capillary. Various muscular wall into which remnants of elastica changes have been noticed in the cerebral extended for a short distance. The luminal arteries in such cases, but are not confined to surface was 'lined by endothelium or coated with the perforating arteries iin the region of the variable amounts of fibrin. Because none of the Postgrad Med J: first published as 10.1136/pgmj.42.483.5 on 1 January 1966. Downloaded from 12 POSTGRADUATE MEDICAL JOURNAL January, 1966 patients in this study were under 50 years of throughout the brain, particularly in the pons, age, final assessment will depend on the findings and resembled that seen in the arterioles of in the brains of young hypertensive patients, other organs in malignant hypertension, part- in whom the natural vascular degeneration of icularly in the kidney. She suggested that these ageing can be expected to be minimal. vascular changes explained the fact that primary pontine haemorrhage occurred more frequently The role of perivascular necrosis with malignant than benign hypertension, Globus and Strauss (1926) and Westphal and whereas wit-h 'haemorrhage elsewhere in the Bar (1926) favoured the view previously ex- brain the reverse was found. pressed by Rouchoux i(1844) that haemorrhage Okinaka and Kameyama (1960) also des- occurred from rupture of an artery at a site cribed fibrinoid necrosis Fin the cerebral arteries where the supporting tissue had undergone of cases of cerebral haemorrhage, most com- ischaemic softening. They considered that oc- monlly in patients over the age of 70 years. clu-sion of one or more cerobral arteries due They found that the incidence of necrosis was to either generalised or localised arterial de- directly proportional to the degree of hyper- generation (Globus and Strauss) or spasm tension, and occurred most frequently in the (Westphall and 'Bar) created a zone of necrosis putmen followed by the thalamus, pons and around the diseased artery, which was the most cerebellum. Arteries showing this change some- important factor 'in the development of hyper- times developed submiliary aneurysms. tensive cerebral haemorrhage. Hassler (1962) demonstrated defects in the Globus and Epstein (1953) provided ad- muscle coat in the small intracerebral arteries ditional evidence supporting this view in a study in hypertension, particularly at the site of based both on further data from human cases branching, and postulated that such defects and on experimental work in dogs and monkeys. could rupture without aneurysm formation. They occluded cerebral arteries and were able to demonstrate haemorrhagic areas, resembling Vascular necrosis induced by spasm massive cerebral haemorrhage in man, whlich Byrom (1954 and 1958-59) considered thatby copyright. sometimes developed spontaneously but could in hypertension a vasoconstrictor stimulus pro- be produced more frequently (by subsequent voked spasm of groups of cerebral arterioles injection of neosynephrine. and iterminal arteries leading to ischaemia of Bruetman, Fields, Crawford, and de Bakey the territory supplied. The effect depended on (1963) confirmed these findings with further the severity, duration and extent of the spasm. clinical material. In the course of evaluating An acute widespread crisis would cause anoxia corrective surgical treatment in 900 patients of the whole brain and commonly result in with internal carotid artery stenosis or occlu- acute 'hypertensive encephalopathy. Less wide- sion, they encountered six who developed spread but more lasting crises would give rise http://pmj.bmj.com/ postoperative cerebral haemorrhage in the to successive "crops" of lesions of three grades, affected hemisphere. Though all six patients viz. mild and reversible changes, consisting of felt that antecedent small zones of increased capillary permeability were hypertensive, they and focal oedema; small infarcts or haemor- infarction was of greater significance than the rhages; and focal arterial necrosis sometimes hypertension. resulting in major haemorrhages. Rosenblath (1918) 'had previously held a It is relevant at this point to consider the

similar view. He noticed that cerebral haemor- on September 30, 2021 by guest. Protected in with problem of hypertensive encephalopathy, which rhage frequently occurred patients Oppenheimer and Fishberg (1928) first asso- renal disease, and considered that the diseased ciated with cerebral vasospasm and "oedema". kidneys liberated an autolytic enzyme which Byrom (1954) induced hypertensive encephalo- caused cerebral softening into which 'haemor- pathy in rats and observed and filmed spasm rhage occurred. of the cerebral arteries during an attack Arterial degeneration secondary to hypertension through acrylic windows inserted in the skull. Dorothy Russell (1954) found two principal He also showed increased capillary permeability changes in the cerebral vessels in a series of distal to the site of spasm by noting escape of 232 cases of hypertensive cerebral haemorrhage. trypan blue after injecting the dye into the In benign hypertension the perforating cereibral circulation. He also found that the brains of arteries showed fibrous replacement of the rats killed during an attack of encephalopathy muscle in the media, which could predispose to showed a higher water content than a control rupture. In malignant hypertension 'fibrinoid group. Meyer, Waltz and Gotoh (1960) obtained necrosis of the arterioles was widely distributed similar changes in experiments on monkeys. Postgrad Med J: first published as 10.1136/pgmj.42.483.5 on 1 January 1966. Downloaded from January, 1966 FAWCETT and SMITH: Pathogenesis of Common 'Strokes' 13 Ziegler, Zosa and Zieli (1965) reviewed 43 tomas. Twenty of the cases associated with the patients discharged from hospital over a 10 examples); central cerebral (4 examples); and year period with the diagnosis of hypertensive cerebellar (6 examples). Bleeding occurred at encephalopathy. On retrospective analysis they a younger age than in classical hypertensive considered that only three patients had shown cerebral haemorrhage, all cases being under the classical clinical picture, formulated by 40 years, and 15 under 20 years of age. Clinic- Oppenheimer and Fishberg, of seizures with ally they showed evidence of catastrophic focal neurological signs and symptoms com- apoplexy or in less acute cases simulated plicating advanced hypertension. The remainder rapidly growing neoplasms. Krayenbuhl and were classified as uraemia (10 cases), cerebro- Siebenmann (1961) recently reported a further vascular insufficiency '(13 cases), cerebral in- 24 similar cases of spontaneous cerebral farction (10 cases), cerebral and subarachnoid haemorrhage and in 15 they were alble to haemorrhage (4 cases), and miscellaneous en- demonstrate vascular malformations on angio- tities (3 oases). These conclusions suggest that gra-phy; they assu'med -that the haemorrhage classical hypertensive encephalopathy is not to had destroyed the malformation in the other be regarded as a common diisorder, at least in 9 cases. recent years. Cerebral haemorrhage simulating massive Conclusions hypertensive haemorrhage Current views regarding the pathogenesis of Large intracerebral haematomas secondary to hypertensive cerebral haemorrhage are to some ruptured berry aneurysm on the circle of WiHis extent complementary and all implicate de- can simulate hypertensive haemorrhage and generation of small arteries prior to actual have already been considered. Massive haemor- rupture. Though the evidence that submiliary rhage due to blood diseases or true neoplasms aneurysms are frequently present is concluslive, etc. can also occur but will not be discussed 'n their significance in all cases has not been finally by copyright. this review. A comment on cerebral haemor- suibstantiated, as they can certainly develop rhage complicating hamartomatous vascular in normotensive brains without causing haemor- malformations is however appropriate as it has rhage. It could 'be that several factors, either recently -been the subject of several reports, and alone or in combination, operate in different is important from ithe diagnostic point of view. cases to 'produce the identical final catastrophe. Hawkins and Rewell (1946) described two It is not without interest that aetiological con- cases of apoplexy in early life which were siderations such as age, 'hypertension and subsequently shown to 'be caused by haemor- spasm are also relevan:t to the pathogenesis

rhage from cavernous haemangiomas. Though and complications of cerebral infarction and http://pmj.bmj.com/ haemorrhage into such malformations had berry aneurysm. Such inter-relationships should been descrilbed, previously it had not been not be overlooked when adjusting this whole associated with apoplexy without previous group of diisorders. symptoms '(Lindau, 1931; Elkington, 1935). Margolis, Odom, Woodhall, and Bloor (1952) reported four similar cases with spontaneous Summary intracranial 'haematomas due to small ruptured Current views on the pathogenesis of cerebral angiomatous malformations and suggested that infarction due to atherosclerosis, cerebral berry on September 30, 2021 by guest. Protected many examples of spontaneous cerebral haemor- aneurysms and their complications, and massive rhages of uncertain aetiology in young people hypertensive cerebral haemorrhage are dis- could be of this nature. Dorothy Russell (1954) cussed. analysed 461 cases of intracranial haemorrhages, Particular attention is directed to the role and attributed 21 to vascular hamartomas both of stenosis (and its relation to embolism) which she classified as telangiectases, cavern- and of occlusion of the neck-arteries in infarc- ous haemangiomas, and arterio-venous hamar- tion, to the cause of berry aneurysms and the latter type were subsequently described in complications which can result from their detail (Crawford and Dorothy Russell, 1956) rupture, and to the mechanism of massive and termed "cryptic hamartoma" on account of cerebral haemorrhage. their latent clinical characters and the difficulty Factors such as arterial degeneration, hyper- often encountered in locating and demonstrating tension, vasospasm, age and diet are to a them. These "cryptic hamartomas" were variable extent relevant to the pathogenesis divided into three groups on an anatomical and consequences of all three main types of basis: related to the cerebral convexities (10 "stroke". Postgrad Med J: first published as 10.1136/pgmj.42.483.5 on 1 January 1966. Downloaded from 14 POSTGRADUATE MEDICAL JOURNAL January, 1966 REFERENCES tracranial Arterial Occlusive Lesions, Neurology (Minneap.), 11, Supplement p. 145. ABERCROMBIE, J. (1828): Pathological and Practical DENNY-BROWN, D. (1951): The Treatment of Recur- Researches on Diseases of the Brain and the rent Cerebrovascular Symptoms and the Question Spinal Cord. Edinburgh: Waugh and Innes. of Vasospasm, Med. Clin. N. Amer., 35, 1457. ALLCOCK, J. M. and DRAKE, C. G. (1965): Ruptured DICKINSON, C. J. (1961): Functional of the Intracranial Aneurysms-the Role of Arterial Circle of Efficiency Spasm, J. Neurosurg., 22, 21. Willis, Brit. med. J., i, 858. ALPERS, B. J. and BERRY, R. G. (1963): The Circle VANDER EECKEN, H. M. and ADAMS, R. D. (1953): of in Cerebrovascular Arch. Anatomy and Functional Significance of Meningeal Willis Disorders, Arterial Anastomoses of the Human Brain, J. Neurol. (Chic.), 8, 398. Neuropath. exp. Neurol., 12, 132. BAGLIVI, G. (1699): De Praxi Medica ad Verum ELKINGTON, J. ST. C. (1935): Cerebral Vascular Observandi Rationem Revocanda, Libro Duo, Accidents Unassociated with Cardiovascular Dis- p. 403. Paris: Joannem Anisson. ease, Lancet, i, 6. BAKER, A. B., DAHL, E. and SANDLER, B. (1963): ELLIS, A. IG. (1909): The Pathogenesis of Spontane- Cerebrovascular Disease. Etiologic Factors in ous Cerebral Haemorrhage, Proc. path. Soc. Cerebral Infarction: Neurology (Minneap.), 13, 445. Philad., 12 (new series) 3, 197. DU BOULAY, G. (1963): Distribution of Spasm in the FORBUS, W. D. (1930): On the Origin of Miliary Intracranial Arteries after Subarachnoid Haemor- Aneurysms of the Superficial Cerebral Arteries, rhage, Acta radiol (Stockh.), 1, (new series), 257. Bull. Johns Hopk. Hosp., 47, 239. BRAIN, LORD (1963): Some Unsolved Problems of FISHER, C. M., GORE, I., OKABE, N. and WHITE, P. D. Cervical Spondylosis, Brit. med. J., i, 771. (1965): Atherosclerosis of the Carotid and Ver- BRICE, J. G., DOWSETT, D. J. and LOWE, R. D. (1965): tebral Arteries-Extracranial and Intracranial, J. Haemodynamic Effects of Carotid Artery Stenosis, Neuropath. exp. Neurol., 24, 455. Brit. med. J., ii, 1363. GLOBUS, J. H. and STRAUSS, I. S. Massive BRUETMAN, M. E., FIELDS, W. S., CRAWFORD, E. S., Cerebral (1926): DE BAKEY, M. E. (1963): Cerebral Haemorrhage Haemorrhage. Its Relation to Pre-existing in Carotid Artery Surgery, Arch. Neurol. (Chic.), Cerebral Softening. Arch. Neurol. Psychiat. (Chic.), 9, 458. 18, 215. BUCKELL, M. (1964): Demonstration of Substances GLOBUS, J. H. and EPSTEIN, J. A. (1953): Massive Capable of Contracting Smooth Muscle in the Cerebral Haemorrhage: Spontaneous and Experi- Haematoma Fluid from Certain Cases of Ruptured mentally Induced, J. Neuropath. exp. Neurol., 12, by copyright. Cerebral Aneurysm, J. Neurol. Neurosurg. Psy- 107. chiat, 27, 198. GLYNN, L. E. (1940): Medial Defects in the Circle BYROM, F. B. (1954): Pathogenesis of Hypertensive of Willis and Their Relation to Aneurysm Forma- Encephalopathy and Its Relation to ,the Malignant tion, J. Path., Bact., 51, 213. Phase of Hypertension, Lancet, ii, 201. GREEN, F. H. K. (1930): Miliary Aneurysms in the (-1958-59): The Significance of Hypertensive Brain, J. Path. Bact., 33, 71. Encephalopathy, in Lectures on the Scientific GULL, W. (1855): Thickening and Dilatation of the Basis of Medicine, vol. 8, p. 256, University of Arch of the Aorta, with Occlusion of the Innomi- London: The Athlone Press. nata and Left Carotid, Atrophic Softening of the CARMICHAEL, R. The of Non- Brain, Guy's Hosp. Rep., ser., 3, 1, 12. (1950): Pathogenesis GUNNING, A. PICKERING, G. Inflammatory Cerebral Aneurysms, J. Path. Bact., J., W., ROBB-SMITH, http://pmj.bmj.com/ 62, 1. A. H. T. and ROSS-RUSSELL, R. (1964): Mural CHARCOTT, J. M. and BOUCHARD, C. (1868): Novelles Thrombosis of Ithe Internal Carotid Artery and Recherches sur la Pathogenie de 1'Hemorrhagie Subsequent Embolism, Quart. J. Med., 33, 155. Cerebrale, Arch. Physiol. norm. et path., 1, 110, HASSLER, 0. (1962): Medial Defects in Intracerebral 634 and 275. Arteries, Acta neurol. scand., 38, 29. COHNHEIM, J. (1872): Untersuchungen uber die HAWKINS, C. F. and REWELL, R. E. (1946): Un- Embolischen Processe. Berlin: Hirschwald. heralded Fatal Haemorrhages in Haemangiomata of CORDAY, E., ROTHENBURG, S. F. and IRVING, D. W. the Brain, Guy's Hosp. Rep., 95, 88. (1963): Cerebral Angiospasm-A Cause of Cerebral HONOUR, A. J. and MITCHELL, J. R. A. (1964): Stroke, Amer. J. Cardiol., 11, 66. Platelet Clumping in Injured Vessels, Brit. J. exp. on September 30, 2021 by guest. Protected CRAWFORD, J. V. and RUSSELL, D. S. (1956): Path., 45, 75. Cryptic Arterio-venous and Venous Hamartomas HUMPHRIES, A. W., YOUNG, J. R., BEVEN, E. G., of the Brain, J. Neurol., Neurosurg. Psychiat., 19, LEFEVRE, FAY, A., and DE WOLFE, V. G. (1965): 1. Relief of Vertebrobasilar Symptoms by Carotid CRAWFORD, T. (1959): Some Observations on the Endarterectomy, Surgery, 57, 48. Pathogenesis and History of Intracranial HUNT, J. RAMSAY (1914): The Role of the Carotid Aneurysms, J. Neurol. Neurosurg. Psychiat., 22, Arteries in the Causation of Vascular Lesions of 259. the Brain, Amer. J. med. Sci., 147, 704. CROMPTON, M. R. (1962): Intracerebral Haematoma HUTCHINSON, E. C. and YATES, P. 0. (1957): Caro- Complicating Ruptured Cerebral Berry Aneurysm, tico-Vertebral Stenosis, Lancet, i, 2. J. Neurol. Neurosurg. Psychiat., 25, 378. HUTCHINSON, E. C. (1962): The Circle of Willis - (1964a): Cerebral Infarotion Following Rupture Today, in Modern Trends in Neurology, ed. by of Cerebral Berry Aneurysms, Brain, 87, 263. Denis Williams, p. 175, London: Butterworths. (1964b): The Pathogenesis of Cerebral Infarc- KRAYENBUHL, H. and SIEBENMANN, R. (1965): Small tion Following the Rupture of Cerebral Berry Vascular Malformations Causing Haemorrhage, Aneurysms, Brain, 87, 491. J. Neurosurg., 12, 7. DE BAKEY, M. E., CRAWFORD, E. S. and FIELDS, LINDAU, A. (1931): Discussion on Vascular Tumours W. S. (1961): Surgical Treatment of Patients with of the Brain and Spinal Cord, Proc. roy. Soc. Med., Cerebral Artery Insufficiency Associated with Ex- 24, 363. Postgrad Med J: first published as 10.1136/pgmj.42.483.5 on 1 January 1966. Downloaded from January, 1966 FAWCETT and SMITH: Pathogenesis of Common 'Strokes' 15 MCGEE, D. A., MCPHEDRAN, R. S. and HOFFMAN, ROBERTSON, E. G. (1949): Cerebral Lesions Due to H. J. (1962): Carotid and Vertebral Artery Disease, Intracranial Aneurysms, Brain, 72, 150. Neurology (Minneap.), 12, 848. ROSENBLATH, (1918): Ueber die Entstehung der MARGOLIS, G., ODOM, G. L., WOODHALL, B. and Hirnblutung bei dem Schlaganfall, Dtsch. Z. BLOOR, B. M. (1951): The Role of Small Angio- Nervenheilk, 61, 10. matous Malformations in the Production of Intra- ROUCHOUX, J. A. (1844): Du Ramollissement de cerebral Hematomas, J. Neurosurg., 8, 564. Cerveau et de sa Curabilit6, Arch. gen. Med., 6, MARSHALL, J. (1964): The Natural History of Tran- 265. sient Ischaemic Cerebrovascular Attacks, Quart. J. RUSSELL, D. S. (1954): The Pathology of Spontane- Med., 33, 309. ous Intracranial Haemorrhage, Proc. roy. Soc. Med., MATSUOKA, S. (1952): Histopathological Studies on 47, 689. the Blood Vessels in Apoplexia Cerebri, Proc. 1st. ROSS-RUSSELL, R. W. (1963): Observations on internat. Cong. Neuropath. (Rome), 3, 222. Intracerebral Aneurysms, Brain, 86, 425. MEYER, J. S., WALTZ, A. G. and GOTOH, F. (1960): SCHNECK, S. A. (1964): On the Relationship Between Pathogenesis of Cerebral Vasospasm in Hyperten- Ruptured Intracranial Aneurysm and Cerebral In- sive Encephalopathy, I and II, Neurology (Min- farction, Neurology (Minneap.), 14, 691. neap). 10, 735 and 859. SMALL, J. M. and SMITH, W. T. (1964): Unpublished MITCHELL, J. R. A. iand SCHWARTZ, C. J. (1965): Communications to the Society of British Neuro- Arterial Disease. Oxford: Blackwell Scientific Pub- surgeons and The British Neuropathological lications. Society. MONIZ, E., LIMA, A. and DE LACERDA, R. (1937): SMITH, BARBARA (1963): Cerebral Pathology in Sub- Hemiplegies par Thrombose de la Carotide Interne, arachnoid Haemorrhage, J. Neurol. Neurosurg. Presse med., 45, 977. Psychiat., 26, 535. MORGAGNI, J. B. (1761): De Sedibus et Causis Mor- STEHBENS, W. E. (1959): Medial defects of the borum. Translation by B. Alexander (1769). Cerebral Arteries of Man, J. Path. Bact., 78, 179. London: Millar and Cadell. (1963a): Cerebral Aneurysm of Animals Other MURRAY, J. (1919): ed. of A New English Dictionary, Than Man, J. Path. Bact., 86, 161. vols. I and IX. Oxford: Clarendon Press. - (1963b): Histopathology of Cerebral Aneurysms, NOTHNAGEL, C. W. H. (1877): Anaemia, Hyperaemia, Arch. Neurol. (Chic.), 8, 272. Haemorrhage, Thrombosis and Embolism of the (1963c): Aneurysms and Anatomical Variations Brain, in Cyclopaedia of the Practice of Medicine, of Cerebral Arteries, Arch. Path. (Chic.), 75, 45. Account of a Case of a Dis- ed. by H. von Ziemssen, 12, 3. London: Sams,on TODD, R. B. (1844): by copyright. Low. secting Aneurysm of the Aorta, Innominate, and OKINAKA, S. and KAMEYAMA, M. (1960): A Study of Right Carotid Arteries, Giving Rise to Suppression the Etiology of Hypertensive Cerebral Haemor- of Urine and White Softening of the Brain, Med. rhage from the Viewpoint of Vascular Pathology, chir. Trans., 27, 301. Trans. Amer. neur. Ass., 85, 114. TOMLINSON, B. E. (1959): Brain Changes in Ruptured OPPENHEIMER, B. S. and FISHBERG, A. M. (1928): Intracranial Aneurysm, J. clin. Path., 12, 391. Hypertensive Encephalopathy, Arch. intern. Med., TOOLE, J. F. (1964): Reversed Vertebral Artery Flow 41, 264. and Cerebral Vascular Insufficiency, Ann. intern. PATEL, A. and TOOLE, J. F. (1965): Subclavian Steal Med., 61, 159. Syndrome-Reversal of Cephalic Blood Flow, TURNBULL, I. (1962): Agenesis of the Internal Carotid Medicine (Baltimore), 44, 289. Artery, Neurology (Minneap.), 12, 588. PICKERING, G. W. (1948): Transient Cerebral Para- WALTON, J. N. (1956): Subarachnoid Haemorrhage, http://pmj.bmj.com/ lysis in Hypertension and in Cerebral Embolism, p. 175. Edinburgh: E. & S. Livingstone. With Special Reference to the Pathogenesis of WEISS, S. (1938): The Regulation and Disturbance Hypertensive Encephalopathy, J. Amer. med. Ass., of the Cerebral Circulation Through Extracerebral 137, 423. Mechanisms, Res. Publ. Ass. nerv. ment. Dis., 18, POOL, J. M. (1958): Cerebral Vasospasm, New Engl. 571. J. Med., 259, 1259. WEPFER, J. J. (1658): Historiae Apoplecticorum, RAYNOR, R. B., MCMURTRY, J. G. and POOL, J. L. Observationibus et Scholiis Anatomicis et Medicis (1961): Cerebrovascular Effects of Topically Ap- Quanplurimis Elaboratae et Illustratae. Amsterdam:

plied Serotonin in the Cat, Neurology (Minneap.), Janssonio-Waesbergios (Published posthumously, on September 30, 2021 by guest. Protected 11, 190. 1704). RAYNOR, R. B. and MCMURTRY, J. G. (1963): Pre- WESTPHAL, K., and BAR, R. (1926): Ueber die Ent- vention of Serotonin-Produced Cerebral Vaso- stehung des Schlaganfalles. I Pathologisch-anato- spasm, J. Neurosurg., 20, 1. mische Untersuchhungen zur Frage der Entstehung RICHARDSON, J. C. and HYLAND, H. H. (1941): des Schleganfalles, Dtsch. Arch. klin. Med., 151, 1. Intracranial Aneurysms. A Clinical and Patho- WILLIAMS, D. (1964): Vertebro-basilar Ischaemia, logical Study of Subarachnoid and Intracerebral Brit. med. J., i, 84. Haemorrhage Caused by Berry Aneurysms, WILSON, C. and BYROM, F. B. (1939): Renal Changes Medicine (Baltimore), 20, 1. in Malignant Hypertension, Lancet, i, 136. RIGGS, Helena E., and RUPP, C. (1963): Variations in YATES, P. 0. and HUTCHINSON, E. C. (1961): Cerebral Form of the Circle of Willis, Arch. Neurol. (Chic.), Infarction: The Role of Stenosis of the Extracranial 8, 8. Cerebral Arteries, Spec. Rep. Ser. med. Res. Coun. ROB, C. G. (1961): The Surgical Treatment of (Lond.), No. 300. Stenosis and Thrombosis of the Extracranial Por- YATES, P. 0. (1964): A Change in the Pattern of tions of the Carotid Arteries, J. cardiovasc. Surg., Cerebrovascular Disease, Lancet, i, 65. 2, 336. ZIEGLER, D. K., ZOSA, A., and ZIELI, T. (1965): ROBERTS, B. (1965): Carotid-Vertebral Artery Steno- Hypertensive Encephalopathy, Arch. Neurol. sis and Occlusion, Practitioner, 194, 214. (Chic.), 12, 472.