Current surveys

The differential diagnosis of 'little strokes'

JOHN MARSHALLM.D., F.R.C.P.(Ed.), F.R.C.P.(Lond.), D.P.M.

Reader in Clinical Neurology, University of LondonNational Hospital for Nervous Diseases, Queen Square, London, W. C. 1

SummaryTransient focal disturbances of cerebral func-

tion are often indicative of atherosclerosis. Theymay occur, however, in association with anumber of other conditions which are illustratedby fourteen brief case histories.

IntroductionThe more widespread appreciation of the fact

that transient ischaemic cerebrovascular attacks(TIAs) often give warning of an impendingcompleted stroke has been a significant advancein the management of cerebrovascular disease.An increasing number of patients are nowreferred for investigation after they have exper-ienced only one or two TIAs while there maystill be time to prevent the development of acompleted stroke.

Whilst the pathological basis for the greatmajority of TIAs is atheroma, it should not beforgotten that they may be associated with othertypes of pathology. Similarly, focal neurologicaldisturbances which are too prolonged to justifythe label TIA, and yet, by remitting, do not earnthe title of a completed stroke, may also be asso-ciated with a pathology other than atheroma.The purpose of the present paper is to callattention to some of these by recounting briefcase histories of fourteen patients who wereencountered in a series of 900 personal cases ofcerebrovascular disease.

TIA and cardiac dysrhythmiaThe occurrence of TIAs in association with

episodic cardiac dysrhythmia is illustrated by thetwo following cases.

Case No. 778A man of 41 years gave a 5-year history of

sudden attacks of unsteadiness of gait or stance,each lasting about 1 min and recurring at monthlyintervals. They were not associated with vertigo,visual symptoms or feelings of faintness, but hewas aware on each occasion of a fast, regularbeating of his heart.

Neurological examination was negative, theheart was normal, and the blood pressure 150/

100 mmHg. There were no cervical bruits. Bloodcount, cerebrospinal fluid, radiographs of skulland chest, audiograms and caloric responseswere normal. The ECG showed a sinus rhythmat 60/min, a P-R interval of 0-16 sec, QRS of008 sec and normal complexes.A diagnosis of transient vertebro-basilar

ischaemia associated with attacks of paroxysmaltachycardia was made.

Case No. 609A right-handed woman of 59 years suddenly

lost vision to her left for a period of 15 min inJuly 1962. In March 1963 she had an episode ofdysphasia lasting about 2 hr. Finally, in April1963 she suddenly developed pins and needlesin the left upper limb and face and was againunable to see to her left; these symptoms clearedafter some minutes. She had, therefore, exper-ienced first a right, then a left, and finallyanother right cerebral hemisphere disturbance.Examination revealed no abnormal neurologicalor cardiovascular signs. Blood pressure was 130/80 mmHg; pulse 76/min and regular. ECG wasnormal with a P-R interval of 0 14 sec and aQRS of 0-08 sec.

Whilst under observation she had two episodesof atrial fibrillation both of which subsided spon-taneously and over the succeeding 4 years shecontinued to have occasional attacks of atrialfibrillation the cause of which was never found.

TIAs and mitral stenosisTIAs may occur in association with rheumatic

heart disease without atrial fibrillation (Hutchin-son & Stock, 1963). The following case is anexample of this.

Case No. 787A woman of 45 years developed at 18.30 hr

a right hemiparesis and dysphasia which clearedafter 1 hr. She gave no history of cardiacirregularity at the time of the attack. When seen5 days later there were no abnormal neurologicalsigns, but she had a previously undiagnosedmitral stenosis and incompetence with regularrhythm. She was referred to a cardiologist who

543copyright.

on July 28, 2020 by guest. Protected by

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.44.513.543 on 1 July 1968. D

ownloaded from

Current surveys

did not advise surgery. Anticoagulant therapywas commenced and she has remained wellover 4 years.

It is often impossible to decide whether anepisode of this kind is embolic or due to transientchange in cardiac output. Emboli are unusualin the absence of atrial fibrillation, but an attackof fibrillation could have passed unnoticed bythe patient. Equally, an attack of fibrillationcould have been responsible for a transient fallin cardiac output, the focal character of thecerebral manifestations being determined by thesite of pre-existing lesions in the cerebral arteries.Because of her age, the embolic hypothesis waspreferred in the present instance and treatmentinstituted accordingly.

TIAs associated with heart blockHeart block commonly produces Stokes-

Adams attacks, but when there is associatedcerebrovascular disease, transient focalischaemic attacks may also occur as in thefollowing case.

Case No. 757A man of 71 years began to experience in

Spring 1963 sudden attacks of loss of conscious-ness without convulsion, lasting 20 min, andoccurring when he was sitting at a table half-way through a meal. He had ten attacks in8 weeks, always during a meal, then remainedfree for 6 months; four more attacks, which werenot associated with the taking of food, thenfollowed. He next noticed that if he turned hishead to left or right, his vision became 'grey'and if he persisted in the movement, he lostconsciousness. He also had occasional, briefepisodes of ataxia, seemingly unassociated withneck movement.His pulse was found to be 36/min and regular

and did not increase on gentle exercise. Therewas a systolic bruit over the left carotid arteryand neck movement was restricted in all direc-tions. On turning his head to the left he com-plained of greying of vision so the neck wasreturned to the neutral position. There were noabnormal neurological signs. An electrocardio-gram confirmed that he had a complete heartblock. He had further attacks of loss of cons-ciousness whilst under observation in which thepulse was counted at 25/min. He was trans-ferred to the National Heart Hospital, where apace-maker was successfully inserted. He had nofurther attacks over the next 4 years.

TIAs and anaemiaSevere anaemia is a well recognized cause of

TIAs (Siekert, Whisnant & Millikan, 1960) whichis illustrated by the following case.

Case No. 699A man of 63 years suddenly developed vertigo,

ataxia and diplopia with vertical separation ofthe images. This recurred intermittently through-out 1 day, being associated for part of the timewith tingling across his neck and down botharms. One week later he had an episode ofslurred speech; his tongue felt stiff and he haddifficulty swallowing. His wife noted the left sideof his face was drooping. The symptoms clearedafter 6 hr.Examination 6 weeks after the onset revealed

only an extensor plantar response on the left.The blood pressure was 160/105 mmHg in theright arm and 140/95 mmHg in the left arm.The heart was normal and there were no vascularbruits. His haemoglobin was 5-4g/100ml, MCV68 ju3 and MCHC 22%. There were 1-1% reticu-locytes. Tests for occult blood in the stools werepositive on two occasions, but a barium mealand follow-through and a barium enema werenormal. He received a transfusion of 4 pints ofblood which brought his haemoglobin up to9.3g/100ml. He had no further neurologicalepisodes. He was re-admitted 8 weeks later witha history of increasing exertional dyspnoea of afew weeks duration and evidence of cardiacfailure. His haemoglobin had fallen to 8-2 g/100 ml and after further transfusion rose to10-2g. Serum iron was 105,ug/100ml; the occultblood was still positive.He was referred to a general medical unit

where a further barium meal showed a smallhiatus hernia, but neither barium swallow noroesophagoscopy revealed any varices. Prolongedtreatment with oral iron brought his haemo-globin up to 12-3 g/lOOml with a MCHC of32% and occult blood gradually disappeared fromhis stools. He had no further TIAs.

This patient had experienced two TIAs in thevertebro-basilar territory in association with asevere anaemia. The latter was presumably due togastro-intestinal bleeding, the source of whichwas never found.

TIAs and polycythaemiaPolycythaemia may also give rise to TIAs

(Millikan, Siekert & Whisnant, 1960) whichrespond to the treatment of the polycythaemia.Case No. 512A woman of 70 years began in November

1960 to have attacks of loss of vision in theleft eye each lasting 5 min; the attacks recurredevery 2-3 weeks. In Febraury 1961 she also

544copyright.

on July 28, 2020 by guest. Protected by

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.44.513.543 on 1 July 1968. D

ownloaded from

Current surveys

developed episodes of numbness and weaknessof the right upper limb of 10 min duration.Investigation at that time showed her to have ablood pressure which varied from 195/100 to220/130 mmHg, a haemoglobin of 16.1 g/100 mland a left carotid stenosis. She was treated withhypotensive and anticoagulant drugs without anyclear-cut reduction in the frequency of herischaemic episodes. In November 1963 shedeveloped a left hemiparesis and re-investigationshowed her haemoglobin to be 18 g with 6-5million erythrocytes/mm3, 18,000 leucocytes, anda haematocrit of 63%. Further investigation (DrC. S. Pitcher) revealed a blood volume of 5 9litres (100 ml/kg), plasma volume 43 ml/kg, redcell mass 56.6 ml/kg. She was treated with radio-active phosphorus (Dr P. Wade) with consider-able reduction in the frequency of her attacks.She required further treatment in September 1964but since then has kept well with only occasionalTIAs.The patient had a carotid stenosis, hyperten-

sion, hypotensive therapy and polycythaemia, allof which could have contributed to her TIAs.The response to the treatment showed that thepolycythaemia was an important factor.

TIAs and leukaemiaTIAs may be the presenting symptom in

leukaemia as the next case shows.Case No. 755A man of 67 suddenly developed vertigo,

ataxia and vomiting for I hr without tinnitus,deafness or other symptoms. Three weeks laterhe had a second attack lasting several hours.Over the preceding 7 months he had lost1 stone in weight and had several night sweats.Examination showed venous engorgement of

the retinae with extensive haemorrhages contain-ing white centres along the veins. Some of thehaemorrhages were old, some new; there was nopapilloedema. The liver and spleen wereenlarged, but there was no lymphadenopathy. Inthe nervous system there was slight Rombergismand ataxia of gait.Blood examination showed a haemoglobin of

5-8g/100ml with 700,000 white cells/mm3, 27%of which were lymphocytes. A diagnosis oflymphatic leukaemia was made.

This patient was also severely anaemic, butTIAs may occur in patients with leukaemia whoare not anaemic.

'Little strokes' and thrombotic thrombocytopeniapurpuraA rare blood disorder which may present with

cerebrovascular episodes is thrombotic thrombo-

cytopenia purpura of which the next case is anexample.

Case No. 611A woman of 57 years was found speaking

incoherently and unable to use her right arm.These symptoms were improving when shedeveloped a raised, red rash all over the body.This faded, but 2 weeks later she suddenlydeveloped a left hemiplegia, hemianopia andneglect of the left side of her body. She wasadmitted to hospital on 21 June 1963 under thecare of Dr Q. J. G. Hobson and after 24 hr shedeveloped extensive bruising. Her haemoglobinwas 10-8 g/ 100 ml, white cells 5800/mm3, plate-lets 92,000/mm3, reticulocytes 8-4%. Bone-marrow examination revealed hyperactivenormoblastic erythropoiesis and a normal granu-locyte series. The megakaryocytes numbered0-5% of the differential count, and there was littleevidence of platelet formation. The Quick one-stage prothrombin time was 20 sec (control19 sec) coagulation time (Dale and Laidlaw)2 min. Direct Coombs' test was negative withserum dilutions of 1: 4 to 1: 1024.Serum bilirubin 1-4 mg/ 100 ml. Serum protein60 g/100ml with normal electrophoretic strip.Over the course of 10 days her platelets fell

to 27,000/mm3, thrombotic thrombocytopeniapurpura was diagnosed and heparin 10,000 unitsintravenously 4-hourly was given. She improvedon this regime, the platelet count rising to209,000/mm3, though her hemiplegia remainedunchanged. She was able to return home on14 September but was re-admitted on8 November, having developed a brain-stemlesion with dysarthria, dysphagia and ataxia.Her haemoglobin was 131 g and her plateletcount 121,000/mm3. Her condition slowly deter-iorated and she died on 28 February 1964.Necropsy (Dr A Knudsen) showed areas of

haemorrhage into the muscle of the anterior wallof both cardiac ventricles, and on the surface ofthe kidneys. There were some renal infarctsmeasuring up to 2cm diameter. The spleen wasfirm, but the follicles were ill-defined. The brainshowed multiple areas of cortical softening,measuring up to 3 cm in diameter. The whitematter was not involved. The spinal cord wasnormal. Sections showed plugging of capillaryvessels by hyaline and granular eosinophilicmaterial, best seen in the kidney, heart, pancreasand bladder.

'Little strokes' associated with hypertensionHypotension is a well-recognized cause of

TIAs (Denny-Brown & Meyer, 1957), but their

545copyright.

on July 28, 2020 by guest. Protected by

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.44.513.543 on 1 July 1968. D

ownloaded from

Current surveys

occurrence in association with hypertension isless well recognized. Indeed, when hypertensionis being treated with hypotensive drugs, the mis-take of attributing TIAs to hypotension ratherthan hypertension may be made. This dilemmais illustrated by the next case.

Case No. 810A woman of 64 years awoke one morning

in December 1964 with headache, vomiting, weak-ness of the right leg and unsteadiness of gaitwhich was greater than could be accounted forby her weakness. The vomiting and weaknessceased after a few hours, but she continued tobe unsteady for about 2 weeks. She was seen ata hospital where her blood pressure was foundto be raised and treatment with hypotensive drugsbegun.Over the next 3 months she had repeated

attacks of non-rotatory 'giddiness'. As there wasdoubt as to whether this symptom was due tohypotension produced by therapy or hyperten-sion she was referred for a further opinion. Whilein the supine position on the examination couchshe suddenly complained of feeling faint andgiddy. A second-degree nystagmus to the leftappeared, accompanied by a left facial weaknessof nuclear type, ataxia in the left upper limband increase of reflexes down the right side. Shevomited several times.

Unfortunately the examination had not reachedthe stage at which the blood pressure was re-

corded before the onset of these symptoms, buta few minutes after the onset it was 260/150 mmHg. Over the next few hours it fell to220/130 mmHg.Treatment of her blood pressure with methyl-

dopa was intensified and a level of 160/110 mmHgmaintained. The nystagmus and ataxiadisappeared over the course of 2 weeks, but therewas persistent increase in the reflexes on theright.

Five weeks later she had a further attack ofgiddiness with a re-appearance of nystagmus; herblood pressure had been rising slowly over theprevious 48 hr, reaching a level of 220/140 mmHg at the time of the attack; controlwas again established at a level of 170-150/110-100 mmHg and she had no further symptoms.

'Little strokes' with multiple causesThe difficulty in determining which of several

possible causes is responsible for cerebrovascularepisodes was mentioned in the report of thepatient with polycythaemia (Case No. 512) andis further illustrated by the next case in whichhypertension and blood loss were both present.

Case No. 808A man of 58 years developed melaena in

October 1964. His haemoglobin was reduced to49 g and a transfusion of 5 pints of blood wasgiven. He had no neurological symptoms at thattime. A barium meal showed scarring of theduodenal cap.He remained well until March 1965 when both

lower limbs suddenly became weak as he steppedout of his car. He was helped into the housewhere he observed that the left arm was weakand that there was a sensation of pins andneedles in the left lower limb. The symptomsbegan to improve within 3 days of the onset but7 days later he had a severe melaena causingsyncope and necessitating a further transfusionof 6 pints of blood. This was not accompaniedby an exacerbation of his neurological symptoms.Examination 5 weeks later showed only slight

weakness in the left upper limb with more severeweakness in the left lower limb; the tendonreflexes were brisker on the left and the plantarwas extensor. Blood pressure was 180/110 mmHg.There was a late systolic murmur at the apex,the retinae showed only grade I changes, therewere no vascular bruits but the ECG showedevidence of left ventricular hypertrophy. Anarch aortogram showed no abnormality of thecarotid and vertebral arteries.The problem as to whether his neurological

symptoms were assocated with his blood loss orwith his hypertension was discussed. The formerwas thought to be unlikely because there wasno exacerbation of his neurological deficit at thetime of the second melaena which was severeenough to produce syncope. Accordingly hisblood pressure was reduced with methyldopa250mg t.d.s. and over a period of 21 years hehas remained well with a blood pressure of150/100 mmHg.Acute blood loss poses a special problem

because it may be associated with hypotension,so that it may be difficult to determine whetherthe anaemia or the fall in blood pressure is thecause of the cerebrovascular symptoms. Thetime-relationships in the present case indicate thatprobably neither was responsible, but the diffi-culty in apportioning responsibility amongseveral causes and so deciding treatment, shouldnot be underestimated.

'Little strokes' associated with arterial constrictionNot every arterial stenosis is due to an

atheromatous narrowing; a healthy vessel maybe constricted from without and give rise tosymptoms.

546copyright.

on July 28, 2020 by guest. Protected by

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.44.513.543 on 1 July 1968. D

ownloaded from

Current surveys

Case No. 928A man of 64 years woke on the morning of the

25th January 1967 with weakness of the left armand leg. At noon the following day the rightside began to tingle and became weak. He lostconsciousness for a few moments and on recoverywas found to be dysarthric. Examination at thisstage revealed no other abnormality in the cranialnerves. All four limbs were weak with increasedreflexes and extensor plantar responses. Positionsense was impaired in the left lower limb. Bythe 29 January power had returned except in theleft leg which dragged as he walked. Bloodpressure was 150/95 mmHg in both arms andneck movements were full. Arch aortographyshowed severe constriction of the left vertebralartery at the level of C6-7 space. The rightvertebral artery was small in calibre, the carotidswere normal.At operation (Professor G. W. Taylor) the left

vertebral artery was found to be constricted byfibro-muscular bands at the C6-7 level just beforeit entered the bony canal in the transverse processof C6. The vessel itself appeared healthy. Thebands were divided to free the artery. Post-operative angiography showed the constrictionto have been relieved; the patient has had nofurther TIAs.

'Little strokes' and meningo-vascular syphilisA time-honoured cause of stroke which now-

adays tends to be overlooked is neurosyphilis.The following case underlines the need to ex-amine the cerebrospinal fluid in every case ofstroke.

Case No. 720A man of 56 years woke from an afternoon

nap to find his left arm and leg to be paralysed.On admission to hospital he had a left hemi-plegia involving the face, arm and leg and ablood pressure of 200/90 mmHg. The pupillaryreflexes were normal. The cerebrospinal fluidcontained ninety-five lymphocytes and 64mg ofprotein/ 100 ml. The WR was positive in theblood and cerebrospinal fluid. He made analmost complete recovery following treatmentwith 1-2 mega-units of procaine penicillin for21 days.

'Little strokes' and cerebral tumoursIt is not only vascular lesions which may

present as TIAs; cerebral tumours may first bemanifest in this way.

Case No. 933A man of 54 years awoke in September 1965

with numbness of the left upper limb and clum-

siness in the performance of tasks such as thefastening of buttons. This improved over 4 days.When seen in October, neurological examinationwas negative. Radiographs of the skull showedthe calcified pineal to be central. Theelectroencephalogram showed diffuse irregularslow activity in both hemispheres with oneepisode of focal sharp waves, which could havebeen artefact, in the right frontal region. Asecond examination 1 month later showed diffuseslow activity as before, but no right frontal dis-turbance.

In July 1966, a second episode occurred, butagain improved. He was admitted to hospital inSeptember 1966 when examination revealed noth-ing abnormal except a threshold of 1 cm to two-point discrimination in the left hand. Arch-aortography showed no lesion of the carotid andvertebral arteries in their cervical course.Electroencephalography showed diffuse activityat 4-5 cycles/sec with sporadic delta activity inthe right mid-temporal region on overbreathing.Following discharge he returned to work, butbegan to have morning headache, vomited occas-ionally and had difficulty using his left arm andleg. He was re-admitted on 14 October 1966when a right carotid angiogram revealed a rightparietal neoplasm which, at operation (Mr Lind-say Symon), proved to be a meningioma. Thepatient made an excellent recovery.

Case No. 789A man of 61 years gave a 3-year history of

episodes of headaches and vomiting lasting 24 hron three or four occasions each year. Threemonths prior to being seen he began to haveoccasional, brief attacks of numbness of theright hand. Sixteen days before being seen hesuddenly developed numbness of the right indexand middle finger; this disturbance spread withinseconds to include the whole of the right upperlimb and right side of face and trunk, but notthe lower limb. The symptom persisted for 10 minduring which time the right arm hung limplyby his side and he was unable to speak properly.These attacks recurred frequently but betweenthem he had no disability. He was started onheparin, whereupon the attacks ceased for 48 hr.Warfarin had been started with the heparin,hence the latter was stopped after 48 hr, where-upon the patient had first a short-lived attackand then a more prolonged attack lasting somehours. A continuous intravenous infusion ofheparin was begun and he had no further attacksover 48 hr. The heparin was again stopped priorto angiography and he had three slight attacks.The angiogram showed a left parietal neoplasm

547copyright.

on July 28, 2020 by guest. Protected by

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.44.513.543 on 1 July 1968. D

ownloaded from

548 Current surveys

which at biopsy proved to be a grade IL astro-cytoma.Not only was the description of the attacks

typical of TIAs but they also showed a strikingarrest following the administration of heparin ontwo occasions with recurrence when the heparinwas stopped.

DiscussionTransient disturbances of neurological func-

tion may be associated with a number of condi-tions. They may for instance occur in epilepsyas in Todd's paralysis, though usually the historyof a preceding convulsion makes clear the natureof the transient weakness. Similarly, the reasonfor the temporary paresis which may follow focalmotor seizures is revealed by the account ofclonic movements of the affected part. Focalsensory attacks, however, present more of aproblem; it is often extremely difficult to deter-mine whether transient sensory symptoms are apositive phenomenon due to an epileptic dis-charge from the sensory cortex, or a negativephenomenon due to ischaemic paralysis of thatarea. The duration of the disturbance may helpas it is unusual for an epileptic discharge to lastmore than 1 or 2 min; a numbness lasting 10 minor so is, therefore, more likely to be ischaemic.The motor, sensory and visual symptoms which

may occur in association with migraine may alsopose problems. The symptoms which may usherin an attack may be very like TIAs especiallyin those cases in which the succeeding headacheis mild. Indeed, it may be argued that theseare in effect TIAs, as it is commonly held thatthe symptoms at the onset of an attack ofmigraine are due to constriction of intracranialarteries. Further difficulty is created by thefact that TIAs associated with atheroma may beaccompanied, or followed, by localized headachedue to dilatation of collateral vessels in anattempt to ward off the threatened ischaemia.Even when it seems clear that a transient focal

cerebral disturbance is due to ischaemia, thepossibility that it is the result of a pathologyother than atherosclerosis must be borne inmind, as has been abundantly illustrated by thecase histories presented here. A careful historyand examination should reveal most of these.In addition, simple investigations such as bloodcount, ECG and radiographs of the cervical spineshould bring to light most of the remainder.

It is the cerebral tumour presenting with TIAswhich offers the most difficulty. Echoencephal-ography and electroencephalography are of greathelp in this situation. The finding of a focal, slowwave disturbance in the EEG in a patient

currently experiencing TIAs should alwaysarouse suspicion of a tumour as transient focalischaemia, unlike an infarct, does not produce adisturbance of this kind. If the focal disturbancepersists over 3 weeks, it should certainly beregarded as due to a tumour until proved other-wise. Cerebral angiography or a radioactive brainscan must be undertaken.Even when it is decided to carry out angio-

graphy, the difficulties are not at an end. Thestudy of patients experiencing TIAs has placedincreasing emphasis upon the great vessels in theneck and their origins from the aortic arch, asthe site of the relevant lesions; these are bestdemonstrated by aortography via a catheterintroduced into the arch of the aorta by oneor other route. This procedure does not, how-ever, permit adequate visualization of the intra-cranial vessels, for which selective puncture orcatheterization of single vessels with radiographscentred on the skull is preferable. There is aconflict of interests here which cannot be resolvedby any simple rule-of-thumb; the most appro-priate procedure must be determined for eachcase in the light of the clinical features and theresults of other investigations. When there is sus-picion of a cerebral neoplasm, selective angio-graphy of a single artery is the better procedure.But when cervical bruits or other evidence makesatheroma the more likely cause of TIAs,aortography is more informative, as it may assessthe status of all four vessels supplying the brain.

In all these circumstances the most importantfactor remains the clinician's awareness of thepossibilities. Because atherosclerotic vasculardisease is so common, there is always the dangerof a false sense of diagnostic certainty arising.These cases illustrated here indicate how mis-placed this may be.

AcknowledgmentI wish to thank the British Heart Foundation for financial

support of my studies in cerebrovascular disease, of whichthis forms a part.

ReferencesDENNY-BROWN, D. & MEYER, J.S. (1957) The cerebral

collateral circulation. II. Production of cerebral infarctionby ischemic anoxia and its reversibility in early stages.Neurology (Minneap.), 7, 567.

HUTCHINSON, E.C. & STOCK, J.P.P. (1963) Paroxysmalcerebral ischaemia in rheumatic heart-disease. Lancet, i,653.

MILLIKAN, C.H., SIEKERT, R.G. & WHISNANT, J.P. (1960)Intermittent carotid and vertebral basilar insufficiencyassociated with polycythaemia. Neurology (Minneap.), 10,188.

SIEKERT, R.G., WHISNANT, J.P. & MILLIKAN, C.H. (1960)Anaemia and intermittent focal cerebral arterial insuffi-ciency. Arch. Neurol. (Chic.), 3, 386.

copyright. on July 28, 2020 by guest. P

rotected byhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.44.513.543 on 1 July 1968. Dow

nloaded from