original articles intracerebral haemorrhage after the ... · intracerebral haemorrhage in childhood...

Archives of Disease in Childhood, 1986, 61, 538-544

Original articles

Intracerebral haemorrhage after the neonatal periodJ H LIVINGSTON AND J K BROWN

Department of Paediatric Neurology, Royal Hospital for Sick Children, Edinburgh

SUMMARY Intracerebral haemorrhage is rare in childhood. We have reviewed the last 10 years'experience, in our referral area, of parenchymatous intracerebral haemorrhage in children from 1month to 16 years of age. There were 27 cases, five of which were intracerebellar and twopredominantly intraventricular. The commonest aetiology was vascular malformation (10),followed by haemorrhage into tumour (four), and coagulopathies (five).

Clinical features were non-specific, but altered consciousness, headache, vomiting, and focalsigns were the most common. Focal signs were, however, rare in the patients with intracerebellarhaemorrhage. There was an overall mortality of 54% (14 out of 27). Nine patients were

handicapped on follow up, but none severely so. For the diagnosis of intracerebral haemorrhagea high level of clinical suspicion is needed with early use of computed tomography. Maintenanceof homeostasis, relief of raised intracranial pressure, and evacuation of haematoma are the aimsof management.

Intracerebral haemorrhage in childhood is a rareand often devastating event. There are a largenumber of published works on this type of haemor-rhage in the adult, hypertension with or with-out atherosclerosis being the major aetiologicalfactor.' 2There have been few studies of intracerebral

haemorrhage in childhood and these have beenmainly concerned with one aetiological group or,have been isolated case reports. In an attempt toobtain an overview of the aetiology, presentation,management, and prognosis of intracerebral haemor-rhage in childhood we have reviewed the last 10years' experience of this type of haemorrhage inEdinburgh.

Patients and methods

Records were reviewed from the Royal Hospital forSick Children, Edinburgh, and the Regional Neuro-surgical Centre. All cases of intracranial haemor-rhage in children from 1 month to 16 years of agewere sought, including extradural, subdural, sub-arachnoid, and intracerebral haemorrhages. Casescoded in the records as intracranial aneurysm andarteriovenous malformation were also sought, andin addition the autopsy book kept by the RegionalNeuropathology Service was examined to detectthose cases who presented and died at other

hospitals in the region who were referred forautopsy. The 10 year period from June 1974 to May1984 was considered.

Results

Over the 10 year period there were a total of 114cases of intracranial haemorrhage. Of these, 50 weresubdural, 34 were extradural, and three were puresubarachnoid.The remaining 27 cases were parenchymatous

intracerebral haemorrhages. There were 20 cases ofsupratentorial intracerebral haemorrhage, five ofintracerebellar haemorrhage, and two of predomi-nantly intraventricular haemorrhage. All subse-quent discussion refers to this group of parenchyma-tous haemorrhages, and the non-parenchymatoushaemorrhages will not be discussed further.

Table 1 Age and sex distribution of the 27 cases ofparenchymatous intracerebral haemorrhage

No %

MF 16:11Age (years):

(-2 3 112-10 9 331(-16 15 56

Total 27 100

538

on 21 May 2019 by guest. P

rotected by copyright.http://adc.bm

j.com/

Arch D

is Child: first published as 10.1136/adc.61.6.538 on 1 June 1986. D

ownloaded from

http://adc.bmj.com/

Intracerebral haemorrhage after the neonatal period 539

Age and sex. The age and sex distribution is shownin Table 1. There were three children under the ageof 2 years. The aetiology in these cases was; latehaemorrhagic disease of the newborn, haemophiliaand associated trauma, and ruptured middle cer-ebral artery aneurysm. There were nine casesbetween 2 and 10 years and 15 cases between 10 and16 years. All the cases of intracerebellar haemor-rhage occurred in the oldest age group. There was amale preponderence of 16 to 11, a male/female ratioof 1 45/1.

Aetiology. The aetiological groups are shown inTable 2.

Vascular malformationThe largest group were those with haemorrhage dueto vascular malformation, which comprised 10 out of27 (37%) of the total. One of these had Marfan'ssyndrome, which is a recognised association. Thesites of the vascular malformations were as follows:middle cerebral (two), posterior cerebral (two),anterior cerebral (two), around the splenium drain-ing into the vein of Galen (one), microvascularanomaly in the parietal region (one), posterior fossa(one), temporal region of uncertain origin (one),and frontal lobe of uncertain origin (one).

CoagulopathiesAs a group, coagulopathies comprised four of the 27cases (15%). Two of these were thrombocytopenicand one had haemophilia. The other child was a 9week old baby who had a massive right sidedintracerebral haemorrhage associated with pro-longed prothrombin and partial thromboplastintime. This was thought to be due to late haemor-rhagic disease of the newborn. Autopsy revealedpancreatic changes suggestive of cystic fibrosis, butthe relation of this to the haemorrhages is unclear.

Table 2 Aetiology of the 27 cases of parenchymatousintracerebral haemorrhages

No %

Vascular malformation 1() 37Aneurysm 3 11Acute lymphoblastic leukaemia 1 4Aplastic anaemia 1 4Haemophilia 1 4Other coagulopathy 1 4Meningioma 1 4Lymphoma 1 4Astrocytoma 1 4Cerebellar glioma 1 4Pontine glioma 1 4Hypertension 1 4Unknown 1 4

27 100

*One patient with Marfan's syndrome.

AneurysmHaemorrhage was due to ruptured aneurysm inthree cases, two from the middle cerebral and onefrom the posterior inferior cerebellar arteries.

Haemorrhage into neoplasmThere were five cases of haemorrhage into tumour;one each of meningioma, pontine glioma, lym-phoma, cerebellar glioma, and cerebellar astro-cytoma. In four of these cases haemorrhage was theinitial presenting feature of the tumour.

OthersThere were only four cases in whom no cause couldbe found. One of these was a 4 year old boy whopresented with multifocal seizures and drowsiness.His computed tomogram showed small bilateralfrontal haematomas. On reviewing the history itbecame apparent that he had been taking largeamounts of aspirin for three days before presenta-tion. A platelet function defect was suspected butnever subsequently proven. There was also one caseof intracerebral haemorrhage in the region of theright basal ganglia due to hypertension in a teenagerwith advanced glomerulo nephritis.

Clinical features. The symptoms and signs onpresentation are shown in Table 3. Onset ofsymptoms wassudden in 16 (59%) cases and gradualin 11 (41%). Onset was defined as sudden whensymptoms developed over a period of one hour orless. The commonest symptoms were headaches andvomiting, occurring in 19 or 20 (70-75%) cases.Altered consciousness occurred in 25 (92%) cases.Thirteen of these were merely drowsy on presenta-

Table 3 Symptoms and signs at presentation

No %

Onset:SuddenGradual

HeadacheVomitingAltered consciousness:

DrowsinessStuporComa

Focal signs:HemiparesisFocal seizureDysphasiaPupillary changesHemisensoryOther cranial nerve

SeizuresVisual symptomsApnoeaHistory of trauma

16111920

13

71713

3

2392

59417074

6348191177

1133774

*Occurred in only one patient after intracerebellar haemorrhage.



j.com/

Arch D


ownloaded from

http://adc.bmj.com/

540 Livingston and Brown

tion, five were stuporose, and seven were comatose.Focal signs were common and occurred in 17 (63%)cases. Hemiplegia occurred in 13 (48%), focalseizures in five (19%), dysphasia in three (11%),cranial nerve signs in three (11%), and pupillarychanges in two (7%), and two of the older childrencomplained of hemisensory symptoms. Seizuresoccured in nine (33%) cases and, with the exceptionof one child who presented in grand mal status,these were always focal. Two children had visualsymptoms on presentation, one had visual failure,and one an homonymous hemianopia. Only onechild gave an unequivocal history of trauma, thisbeing an 18 month old boy with haemophilia whowalked against a cupboard and subsequently diedfollowing a massive parieto-occipital haemorrhage.

It is of note that focal signs occurred in only onecase of intracerebellar haemorrhage, and this con-sisted of tonic deviation of the eyes to one side.None of those with intracerebellar haemorrhage hada hemiplegia on presentation.

Investigations. Excluding the patients with coagulo-pathies, the results of initial investigations on arrivalin hospital were not particularly helpful. Neutrophilleucocytosis was present in nine cases, hypergly-caemia in eight, and fever > 38°C in seven. Onlyone patient had a low haemoglobin concentration onpresentation, and no patients were thrombocy-topenic.

Method of diagnosis. Computed tomography wasused for diagnosis in 18 (67%) cases. In two casesdiagnosis was made by angiography, both of thesepresenting in the era before computed tomography.Diagnosis was made at autopsy in five (19%) casesand by head ultrasound in two (7%).Lumbar puncture was performed in 13 (48%)

cases, in all but one of whom the cerebrospinal fluidwas blood stained or xanthrochromic. It is of notethat four of the five patients with intracerebellarhaemorrhage had a lumbar puncture performed andthat all these patients subsequently died. Computedtomography was performed in only three of thepatients with intracerebellar haemorrhage, theothers having died before scanning was possible.

Site of bleed and angiographic findings. There werefive posterior fossa and two predominantly in-traventricular haemorrhages. The sites of the re-maining haemorrhages are shown in Table 4. Therewere four midline haemorrhages, two midbrain, onepons, and one callosal. Of the remaining 16, 10 wereright sided and six left sided. There was only onedeep or so called ganglionic haemorrhage, the restbeing lobar haemorrhages. There was one case of

Table 4 Site of haemorrhages in the 27 cases withparenchymatous intracerebral haemorrhage

No %

Midline: 4 15Midbrain 2Pons 1Callosal 1

Right hemisphere 10 37Left hemisphere 6 22Frontal 2 7Bilateral frontal 1 4Frontoparietal 2 7Parietal 3 11Temporal 4 15Basal ganglia 1 4Parieto-occipital 1 4Multiple 1 4Massive right sided 1 4Posterior fossa 5 19Intraventricular 2 7

bilateral frontal haemorrhage and one of multiplehaemorrhages in a child with leukaemia.Angiograms were carried out in 14 patients. In

four cases the angiograms yielded normal results.There was only one false negative; a small vascularmalformation was found at surgery in a child whoseangiograms had shown normal results. Vascularmalformations were found in eight cases andaneurysms in two. The site of these lesions has beendiscussed above.

Management. Surgery was carried out in 19 (70%)cases. This usually comprised evacuation of haema-toma and resection of a malformation if present. Insix of the patients surgery comprised emergencyinsertion of intraventricular drains in an attempt torelieve massively raised intracranial pressure. All ofthese patients subsequently died. Three of these sixpatients had posterior fossa haemorrhages, two ofthem intraventricular haemorrhage, and one amidline intracerebral haemorrhage. Of the patientswho had posterior fossa haemorrhages, only onesurvived long enough for exploration of the pos-terior fossa to be attempted.

Outcome. There was an overall mortality of 52% (14out of 27) (Table 5). The mortality after eitherintracerebellar haemorrhage or pure intraventricu-lar haemorrhage was, however, 100% (seven cases).The aetiologies of haemorrhage in the remaining

seven fatal cases were pontine glioma, angioblasticmeningioma, acute lymphoblastic leukaemia, aplas-tic anaemia, haemophilia, haemorrhagic disease ofthe newborn and hypertension. Only two of thesepatients had had surgery. One died shortly afterinsertion of intraventricular drains with massivelyraised intracranial pressure after haemorrhage into



j.com/

Arch D


ownloaded from

http://adc.bmj.com/


cases with parenchymatous

No % Surgical Conservativemanagement management

Normal 3 11 2 1Dead 14 52 7 7Handicap: 9 33 9 -

Hemiparesis 5 19Seizures 4 15Dysphasia 4 15Visual field 4 15Clumsiness 1 4Memory difficulties 1 4

Unknown 1 4 1

Mortality after intracerebellar or intraventricular haemorrhage=100%.

an angioblastic meningioma. The other patient hadhaemophilia and underwent successful evacuation ofa parietal-occipital haematoma. He returned threeweeks later with a massive recurrence and diedshortly after admission. Four out of five of thepatients who did not have surgical management inthis group were at the end stages of malignant orprogressive diseases and it is uncertain whethersurgical intervention would have altered the out-come.Nine patients were handicapped on follow up and

three were normal. Five (19%) have hemiparesis,but in all cases it was mild, all having independentlocomotion. Seizures have occurred in four (15%),but in none of these have they been a majorproblem. Visual field defects and dysphasia haveeach occurred in four (15%). One patient hasmemory difficulties and another has slight clumsi-ness. There are no severely handicapped survivors.

Discussion

There are many causes of intracerebral haemorrhagein childhood. The fairly high proportion of vascularmalformations in our study is in keeping with mostother reports.3 4

In childhood, haemorrhage is the commonestpresenting feature of an arteriovenous malfunction.This may be a subarachnoid or intracerebralhaemorrhage. These malformations extend deepinto the brain parenchyma,6 and it is not surprising,therefore, that if haemorrhage occurs it is moreoften intracerebral than purely subarachnoid.Vascular malformations are found in all parts of thecentral nervous system, including the posterior fossaand spinal cord.7 Margolis et al first drew attentionto the problem of microvascular malformations as acause of massive intracerebral haemorrhage, andsince that time many reports have confirmed thatwhat had previously been called a spontaneous

intracerebral haemorrhage was often a haemorrhagefrom a microangioma.9Y" These microangiomasmay obliterate themselves after haemorrhage, mak-ing subsequent angiographic or histological identi-fication impossible.'2 In our series there were fourhaemorrhages for which no cause was found. Onepatient had bilateral haemorrhages and it seemsunlikely that these were due to microangiomata.The other three, two of which were cases ofintracerebellar haemorrhage, could have been dueto microangiomata. In only one case was a microan-gioma shown, this being in a 7 year old girl with alarge left parietal haemorrhage after chickenpox.No malformation was seen at surgery, and histologyshowed encephalitic tissue. However, angiographysubsequently revealed a microangioma in the leftparietal region. This raises the interesting possibilitythat the angioma may have developed secondary tothe viral infection.Aneurysms accounted for only a small proportion

(11%) of this group. This is in keeping with the factthat aneurysms rarely bleed in childhood, althoughthey can do at any age from neonate onwards.4 13 14

The comprehensive study by Sedzimir et al revealeda fairly high incidence of aneurysms, 40% of 124subarachnoid haemorrhages in the 0-20 age range.'5This was a study of subarachnoid haemorrhage andit is not stated how many of the children also hadintracerebral haemorrhage, so it is not directlycomparable with this series.Why aneurysms bleed in childhood is uncertain. It

is not associated with atherosclerosis, trauma, exer-cise, or stress.4 13 In one study 12% of rupturedaneurysms were associated with coarctation of theaorta and 3-5% with polycystic kidneys. In ourseries we had one ruptured aneurysm associatedwith renal artery stenosis and hypertension present-ing as a pure subarachnoid haemorrhage.

Histologically, most aneurysms are the congenitaltype and arise most commonly from the terminalcarotid, anterior communicating, and middle cere-bral arteries, but they may also arise from the basilarsystem.

Coagulopathies comprised a major aetiologicalgroup in this series (15% of the total). There aremany different causes of coagulopathy associatedwith intracerebral haemorrhage. Intracerebralhaemorrhage complicating idiopathic thrombocy-topenic purpura occurs in up to 1% of cases. 16Intracerebral haemorrhage complicating acute lym-phoblastic leukaemia occurs in 1% of cases in theacute stage and can be due to thrombocytopenia,disseminated intravascular coagulation, haemor-rhage into infiltrating tumour, or treatment withdrugs-for example, with L-asparaginase. 17-18 Therewas one case of haemophilia in our series and as

Table 5 Outcome of the 27intracerebral haemorrhage



j.com/

Arch D


ownloaded from

http://adc.bmj.com/


with other reported cases haemorrhage followedminor trauma.'9

Apart from the child with haemophilia none ofthe haemorrhages followed trauma. In a previousreview by Ingraham and Matson of 1330 cases ofhead injury in children they found no intracerebral,30 extradural, and 319 subdural haemorrhages.5

Intracerebellar haemorrhage accounts for a 10thof spontaneous intracerebral haemorrhage in adultsand is usually due to hypertension.211 In this seriesthere were five cases (19%) of intracerebellarhaemorrhage. Two were due to haemorrhage intotumour and one to a vascular malformation, and twowere of unknown cause. All of these occurred inteenage children. Again microangiomas have beenemphasised as a cause of intracerebellar haemor-rhage.The two cases of intraventricular haemorrhage in

this series were due to vascular malformation inone and aneurysm in the other. Intraventricularhaemorrhage had a uniformally fatal outcome. Bothpatients were over 7 years of age, and this out-come is in keeping with the high mortality ofintraventricular haemorrhage in adults.22 23 In-traventricular haemorrhage often occurs fromventricular extension of a ganglionic/thalamichaemorrhage. It is thought that it is the proximity tovital centres and the greater density of nerve fibresin the region of the initial haemorrhage with anadditional aggravating effect of the intraventricularhaemorrhage that accounts for the poor outcome.24

Cardinal presenting features of intracerebralhaemorrhage (Table 3) were altered consciousness,headache, vomiting, and focal signs. The consciouslevel ranged from coma to mild drowsiness. None ofthese features are specific for haemorrhage and allare similar to those occurring in other acute en-cephalopathies of childhood, such as encephalitis,status epilepticus, and toxic encephalopathy. Like-wise, the speed of onset of symptoms was not areliable diagnostic pointer, onset being sudden in59% of cases and gradual in 41%. Sudden onset ofsymptoms is much less likely, however, in the otheracute encephalopathies of childhood. The groupwith intracerebellar haemorrhage is different in thatnone of the patients had hemiplegia on presentationand only one had focal signs, comprising tonicdeviation of the eye to one side. Four out of fivepatients with intracerebellar haemorrhage had asudden onset of symptoms.The presence of neutrophil leukocytosis, hyper-

glycaemia, or fever was of little help in the diagnosisand as with the clinical features is just as likely to bedue to other causes of acute encephalopathy. Themethod of diagnosis in all recent cases was com-puted tomography. Before its availability diagnosis

was implied by angiography or shown at surgery orautopsy. Computed tomography has a diagnosticaccuracy for intracerebral haemorrhage of 94% withvirtually no false positives.25 Taking this fact withthe fairly non-specific presenting features, it will beseen that for early diagnosis to be made a high levelof suspicion is needed in the clinical situationoutlined above with recourse to early computedtomography. In this situation lumbar punctureshould be deferred until after the scan is available.In the presence of acutely raised intracranial press-ure lumbar puncture is clearly hazardous and all itcan show is subarachnoid blood, which tells onenothing about whether intracerebral haematoma ispresent or not. This point needs to be emphasised tojunior doctors receiving emergencies as the diagno-sis of meningitis can, in most cases, wait a few hoursuntil an emergency scan has been obtained. Ourfigures add weight to this argument, particularly inrelation to posterior fossa haemorrhage. Four out offive of the cases of intracerebellar haemorrhage hadlumbar punctures and all of these patients subse-quently died with massively raised intracranialpressure and brain stem compression.There is little uniformity in the management of

intracerebral haemorrhage,26 and there have beenno studies specifically addressing this issue in child-hood. Adult studies, which are made up predomi-nantly of older patients with hypertension andatherosclerosis, have found that in general earlysurgical evacuation of haematoma and resection of avascular malformation (if present) is superior to latesurgical or conservative management.2 15 26 In thesestudies the best results of early intervention havebeen in young, non-hypertensive patients.2 Thereare theoretical grounds for early clot evacuation inthat after formation of haematoma secondary cere-bral insult occurs due to perivascular haemorrhage,oedema, and perifocal cortical necrosis.27 The mainindication for early clot evacuation, however, is torelieve the raised intracranial pressure. Untilsurgery can be arranged, before starting monitoringof intracranial pressure, routine brain orientatedintensive care should be begun. Thus the aims ofmanagement are control of homeostasis with moni-toring of central venous pressure, arterial pressure,and urine output and measurement of electrolytes,glucose, and calcium concentrations. Seizuresshould be controlled aggressively as these are amajor cause of secondary damage. Elective hyper-ventilation with infusion of mannitol and possiblysteroids should be used if there is clinical indicationof raised intracranial pressure. Monitoring of in-tracranial pressure will in general not be possibleuntil neurosurgical intervention is arranged. So, if adelay is likely, full supportive measures, as de-



j.com/

Arch D


ownloaded from

http://adc.bmj.com/


scribed above, should be begun on clinical groundsalone. Monitoring of cerebral blood flow in thissituation using ultrasonographic techniques wouldclearly provide a rapid non-invasive assessment ofcerebral haemodynamics and allow more rationalmanagement decisions.

In this series all of the patients with intraventricu-lar or intracerebellar haemorrhage died from mas-sively raised intracranial pressure and brain stemfailure. Five out of seven of these patients hadintraventricular drains inserted with failure to re-lieve the raised intracranial pressure.Only one of the patients with intracerebellar

haemorrhage survived long enough for explorationof posterior fossa to be attempted but he died duringsurgery. In all cases the surgical intervention waswithin a few hours of presentation. Two of thepatients with intracerebellar haemorrhage had to betransferred from another hospital in the region. It isunlikely that earlier surgical intervention wouldhave altered the outcome in any of these cases and itseems that in these cases it is the site of haemor-rhage rather than the aetiology that relates to thepoor outcome. Only one of the survivors of in-tracerebral haemorrhage was not surgically man-aged. This was the child with small bilateral frontalhaematomas. The aetiologies in those who diedafter supratentorial haemorrhage (excluding in-traventricular haemorrhage) were tumour (two),coagulopathy (four), and hypertension (one). Onlytwo of these had surgery. Whether the outcomewould have been different in the others aftersurgical intervention is uncertain. Although surgicalevacuation of haematoma in tumour and due tocoagulopathy has been described,'9 28 prognosis inthese groups tends to reflect the prognosis of theunderlying disease.

In contrast to this group all patients with vascularmalformation or aneurysm underwent surgical eva-cuation of haematoma, with resection of the under-lying abnormality. Only one patient in this grouphad had a respiratory arrest before surgery, and ingeneral these patients had milder increases inintracranial pressure than the previous group. Thisis reflected in the outcome in that all of this groupsurvived and morbidity was not severe. Althoughmild handicap was present in nine out of 13survivors, all are independent in the activities ofdaily living. Whether this outcome relates to surgicalintervention or not is uncertain. As there is a realchance, however, of re-bleeding from arteriovenousmalformation or aneurysm15 as well as the theoreticalsecondary damage that can occur after haematomaformation early surgical intervention is preferable.

Prognosis in adult series has been related to thedegree of depression of the conscious level on

presentation.' 2 24 In this series, however, no clearrelation was shown. Six of those who died weremerely drowsy on presentation, and three of thosewho survived with mild handicap were comatose onpresentation. Perhaps it is of importance that noneof those who are normal on follow up werecomatose on presentation.

For early diagnosis of intracerebral haemorrhageto be made a high level of clinical suspicion isneeded. Whether early diagnosis and interventionappreciably alter the ultimate prognosis is not clearfrom this retrospective study. Survival after haema-toma evacuation in sites such as the brain stem andposterior fossa and after haemorrhage into tumouror due to coagulopathy has been described,2 19-21 28and further prospective studies addressing the issueof intervention in these situations are needed.

References

Garde A, Bohmer G, Selden B, et al. 100 cases of spontaneousintracerebral haemorrhage. Eur Neurol 1983;22:161-72.

2 Cahill DW, Ducker TB. Spontaneous intracerebral haemor-rhage. Cliti Neurosurg 1982;29:722-79.

3 Lagos J, Siekert R. Intracranial haemorrhage in infancy andchildhood. Cliti Pediatr (Phila) 1969;8:90-7.

4 Matson D. Intracranial arterial aneurysms in childhood.J Neurosurg 1965;23:578-83.

5 Ingraham FD, Matson D. 7he neurosurgersy of inifancy andchildhood. Springfield: CT Thomas. 1954.McCormick WF. The pathology of vascular 'arteriovenous"malformations. J Neurosurg 1966;24:807-16.

7 McCormick WF, Norfzinger JD. Cryptic vascular malforma-tions of the central nervous system. J Neurosurg 1966:24:865-75.Moyes PD. Intracranial and intraspinal vascular anomalies inchildren. J Neurosurg 1969;31:271-8.Margolis G. Odom GL, Woodhall B, et al. The role of smallangiomatous malformations in the production of intracerebralhaematomas. J Neurosurg 1951;8:564-75.Margolis G, Odom GL, Woodhall B. Further experiences withsmall vascular malformations as a cause of massive intracerebralbleeding. J Neuropathol Exp Neurol 1961;20:161-7.Crawford J, Russell DS. Cryptic arteriovenous and venoushamartomas of the brain. J Neurol Neurosurg Psychiatry1956;19:1-1 1.

12 Conforti P. Spontaneous disappearance of cerebral arterio-venous malformation. J Neurosurg 1971;34:432-4.

13 Patel A, Richardson A. Ruptured intracranial aneurysms in thefirst two decades of life. J Neurosurg 1971;35:571-6.

14 Keren G, Barzilay Z. Cohen BE. Ruptured intracranial arterialaneurysm in the first year of life. Arch Neurol 1980;37:392-3.

5 Sedzimir CB, Robinson J. Intracranial haemorrhage in childrenand adolescents. J Neurosurg 1973;38:269-81.

16 Humphreys R, Hocksley A, Freedman M, et al. Management ofintracerebral haemorrhage in idiopathic thrombocytopaenicpurpura. J Neurosurg 1976;45:700-4.

7 Campbell RHA, Marshall WC. Chessells JM. Neurologicalcomplications of childhood leukaemia. Arch Dis Child1977;52:850-8.Urban CH, Sager WD. Intracranial bleeding during therapywith L-asparaginase in childhood acute lymphoblastic leukaemia.Eur J Pedaitr 1981;137:323-7.

'9 Seeler RA. Imano RB. Intracranial haemorrhage in paticntswith haemophilia. J Neurosurg 1973;39:181-5.



j.com/

Arch D


ownloaded from

http://adc.bmj.com/


2"1 Kneeland WF. Spontaneous cerebellar haemorrhage in childrenand adolescents. Am J Dis Child 1981;135:167-70.

21 Kazimiroff P, Weichsel M, Grinnell V, et al. Acute cerebellarhaemorrhage in children: aetiology, diagnosis and treatment.Neurosurgery 1980;6:524-8.

22 Pai HW. The prognosis of intraventricular haemorrhage. ProgBrain Res 1968;30:463-70.

23 Taneda M, Kaneda H, Minami T, et al. The prognosis ofintraventricular haemorrhage: an analysis of 103 cases ofputaminal and/or thalamic haemorrhage by Ct scan. No toshinkei 1978;30:1265-70.

24 Hungerbuhler JP, Regli F, VanMelle G, et al. Spontaneousintracerebral haemorrhage. Clinical and CT features: immediateevaluation of prognosis. Schweiz Arch Neurol NeurochirPsychiatr 1983;132:13-27.

25 Tans JT. Computed tomography of intracerebral haematoma.Clin Neurol Neurosurg 1977;79:285-95.

26 Sano K, Yoshida S. Cerebellar haematomas: indications andprognosis. In: Pia HW, Langmaid C, Zierski J, eds. Spon-taneous intracerebral haematomas: advances in diagnosis andtherapy. Berlin: Springer-Verlag, 1980:348-60.

27 Suzuki J, Ebina T. Sequential change in tissue surrounding ICH.In: Pia HW, Langmaid C, Zierski J, eds. Spontaneous in-tracerebral haematomas: advances in diagnosis and therapy.Berlin: Springer-Verlag, 1980:121-8.

28 Vincent PM, Bartone JR, Jones MZ. Cerebellar astrocytomapresenting as a cerebellar haemorrhage in a child. Neurology(Minneapolis) 1980;30:91-3.

Correspondence to Dr J H Livingston, The Royal Hospital for SickChildren, Sciennes Road, Edinburgh, Scotland.

Received 28 March 1986



j.com/

Arch D


ownloaded from

http://adc.bmj.com/

original articles intracerebral haemorrhage after the ... · intracerebral haemorrhage in childhood...

Documents