ischaemia in cerebral - adc.bmj.com · health organisation's criteria for cerebral malaria2...

Archives of Disease in Childhood 1994; 70: 281-287

Brain swelling and ischaemia in Kenyans withcerebral malaria

C R J C Newton, N Peshu, B Kendall, F J Kirkham, A Sowunmi, C Waruiru, I Mwangi,S A Murphy, K Marsh

Kilifi Research Unit,Kenya MedicalResearch Institute,Kilifi, KenyaC R J C NewtonN PeshuA SowunmiC WaruiruI MwangiS A MurphyK Marsh

Department ofPaediatrics, OxfordUniversityC R J C Newton

Department ofNeuroradiology,Hospital for SickChildren, LondonB Kendall

Neurosciences Unit,Institute of ChildHealth, LondonF J Kirkham

Department ofPharmacology andTherapeutics,University of Ibadan,NigeriaA Sowunmi

Nuffield Departmentof Clinical Medicine,Oxford UniversityS A MurphyK Marsh

Correspondence to:Dr C R J C Newton, JohnsHopkins, Anesthesiology andCritical Care Medicine,Blalock 1404, 600 NorthWolfe Street, Baltimore, MD21287-4961, USA.

Accepted 20 November 1993

AbstractComputed tomography was performed on14 unconscious Kenyan children recover-ing from cerebral malaria (seven ofwhomhad another scan 12-120 days later) toelucidate the cause of intracranial hyper-tension and neurological sequelae. Brainswelling, defined as a loss ofcerebrospinalfluid spaces, was documented in sixchildren, while a further two had con-

spicuously small ventricles only. Therewas severe intracranial hypertension inthe two children with definite brainswelling in whom intracranial pressurewas monitored. There was no evidenceof acute hydrocephalus or vasogenicoedema. Four children with brain swellingalso had widespread low density areas

suggestive of ischaemic damage. Thepatterns of damage were not uniform butwere consistent with a critical reduction incerebral perfusion pressure (which wasdocumented in the two in whom thiswas monitored), hypoglycaemia, or statusepilepticus. All four had serious neuro-logical sequelae. These data suggest thatbrain injury in cerebral malaria may bedue in part to secondary systemic andintracranial factors as well as to the directeffect ofintravascular sequestration.(Arch Dis Child 1994; 70: 281-287)

Cerebral malaria is the most importantpaediatric encephalopathy in Africa, account-ing for a large proportion of the estimated0 5-2 million childhood deaths each year fromfalciparum malaria in this region.' Themortality of children admitted to Africanhospitals with malaria is 1 0o40%2 and a

further 9-11% are discharged from hospitalwith neurological sequelae including motordisorders (hemiparesis and quadriparesis),blindness, epilepsy, and aphasia.3 4 The patho-logical basis of these handicaps is not clear.4Widespread blockage of small vessels byparasitised erythrocytes has been proposed as amechanism for ischaemia,5 but hypoglycaemia,seizures, and anaemia, which are all clinicalfeatures of cerebral malaria, might also con-tribute to brain damage.6 Occlusion of basalcerebral arteries has been shown in somechildren who developed hemiparesis.7 In otherencephalopathies, intracranial hypertension isassociated with poor neurological outcome,either by precipitating transtentorial herniationor by reducing cerebral perfusion pressure(CPP=mean arterial pressure (MAP)-intra-

cranial pressure (ICP)) below the threshold forischaemia.8 Increased ICP has been docu-mented in African children with cerebralmalaria9 10; possible mechanisms include anincrease in blood volume, cerebral oedema,and acute hydrocephalus. Cerebral oedemahas been described at necropsy in children"and detected by computed tomography in afew adults with cerebral malaria.'2 13Tomographic findings in adults may not reflectthe situation in African children, however, asintracranial hypertension is not thought to playan important part in the pathophysiology ofcerebral malaria in adults.6 The role of cerebraloedema in the pathogenesis of coma andneurological sequelae remains controversial.Fourteen unconscious Kenyan children withcerebral malaria were investigated usingcomputed tomography to investigate the causeof increased ICP and neurological sequelae.

Patients and methodsFourteen children who fulfilled the WorldHealth Organisation's criteria for cerebralmalaria2 were studied. Parasite counts wereperformed on admission and blood glucosewas measured at least every six hours and atany, clinical deterioration using Dextrostix(Bayer Diagnostics), confirmed with an AnaloxGM6 microstat analyser (London). Thechildren were treated with parenteral anti-malarial drugs, either quinine dihydrochloride(20 mg/kg as a loading dose followed by10 mg/kg every eight hours) or artemether (3.2mg/kg intramuscularly and 1-6 mg/kg daily)until they could drink, when they were given asingle dose of sulphadoxine/pyrimethamine(Fansidar). Antimicrobial drugs were givenfrom the time of admission until a lumbarpuncture was performed to exclude meningitis.Initial hypoglycaemia and dehydration werecorrected and thereafter the children weregiven O-18% normal saline and 4% dextroseat a rate of 3 ml/kg/hour while unconscious.The level of consciousness was assessed atleast every six hours by one of the clinicalinvestigators using the Adelaide coma scale.'4The ICP was monitored with a fibreoptic

system (Camino Laboratories, USA)15 ifthe children had signs of brain stem compro-mise (decorticate or decerebrate posturing,impaired oculocephalic reflexes, or dilated andsluggish pupils) or if they had no response topainful stimulation other than non-specificextension. The ICP and MAP were recordedevery 15 minutes during the monitoringperiod. Mannitol (0-25 g/kg or 0-5 g/kg

281

on June 3, 2020 by guest. Protected by copyright.

http://adc.bmj.com

/A

rch Dis C

hild: first published as 10.1136/adc.70.4.281 on 1 April 1994. D

ownloaded from

http://adc.bmj.com/

Newton, Peshu, Kendall, Kirkham, Sowunmi, Waruiru, Mwangi, Murphy, Marsh

Table 1 Grading of diffuise brain swelling on computedtomography

1N

2

3

Normal scanNarrowing of sulci and fissures. Small ventricles compared

with early recovery scansLoss of sulci and fissures. Small ventricles and narrow

basal cistems compared with early recovery scans

Complete loss of sulci, fissures, and supratentorial basalcisterns

intravenously over 20 minutes) was adminis-tered if the ICP was greater than 20 mm Hg for20 minutes or increased above 40 mm Hg.Severe intracranial hypertension was definedas a maximum ICP >60 mm Hg and a

minimum CPP <40 mm Hg, intermediateintracranial hypertension as maximum ICP20-60 mm Hg and minimum CPP <50 mmHg, and mild intracranial hypertension as

maximum ICP 10-20 mm Hg. The childrenwere scanned while they were still unconscious(not yet able to localise pain) but neurologi-cally stable. Scans were performed at least 36hours after admission as the children had to betransferred to other hospitals for computedtomography. Most children were scanned witha Siemens Somatogram DR, but three scans

were performed with a Shimadzu SCT-3000TE (both scans of patient 11 and thefollow up scan ofpatient 14). Contrast iopami-dol (300 mg/kg) was given in four patients,three during the scan in the acute phase of ill-ness (acute scan) and one on a follow up scan.

The scans were examined by an experiencedneuroradiologist (BK). Brain swelling was

assessed by examining the size of the followingcerebrospinal fluid spaces: (a) the cerebralsulci; (b) the perimesencephalic and chias-matic cisterns; and (c) the ventricular system.Table 1 shows the grading scale used.16 TheEvans ratio (width of the frontal horns dividedby the internal skull diameter) was used to

compare the ventricular size on the scans

obtained during the acute phase with thoseobtained on recovery.16 The Evans ratio was

not calculated in children who did not have

follow up scans as small ventricles are seen inscans which are interpretated as normal.

Comparisons of proportions was performedusing Fisher's exact test.

ResultsTable 2 presents the clinical features. Theadmission parasitaemia ranged from 1200 to

1 108800 parasites/mm3. On admission thesummated Adelaide coma score varied from 6to 9. There was a deterioration in coma score

in 12 children after admission. Two childrenshowed signs of brain stem compromiseon admission and a further six childrendeteriorated with abnormal brain stem signs(compatible with transtentorial herniation)6-60 hours later. All except one had seizureseither on presentation or during admission.The children were scanned three to nine

(median five) days after the onset of the illness.Six children had clear evidence of diffuse brainswelling while still unconscious, as defined by a

loss of sulci (n=6), loss of spaces around thecisterns (n=2), and resolution of the swellingwithout atrophy on the follow up scans (n=2)(fig 1). In addition, of the eight children whosescans were reported as normal, two had smallventricles. The Evans ratio increased in thethree children (patients 2, 9, and 10) in whomtwo measurements were made from 0 27, 0-25,and 0-26 on the initial scans to 0 30, 0-28, and0'27 respectively on the follow up scans.

Tissue density was not increased in any of thechildren with diffuse brain swelling. None ofthe children had features of vasogenic oedemaor midline shift on computed tomography.Four children (all with diffuse brain

swelling) had low tissue density and a loss ofgrey/white matter differentiation affecting thecerebral hemispheres (fig 2), compatible with a

diffuse ischaemic or hypoglycaemic insult. Allwere unconscious for more than 120 hours. Intwo children there was also low tissue densityin the basal ganglia. The pattern of brain

Table 2 Clinicalfeatures of 14 children with cerebral malaia

Deteriorationin coma score Duration

Patient Age AMS on (hours after ofcoma ICP HypoglycaemicNo (years) Sex admission* Seizurest Statu4 admission) (hours) pattern§ episodesj Outcome

1 2 M 3 GT and R UC 1 None 38 NP NA+ 1 Normal2 7 M 3 GTC 4 22 87 MIH 0 Mild left dystonic hemiparesis and learning

difficulties3 2-5 F 3 R and L UC 1 None 42 NP A Normal4 1-5 F 3 R and L UC 1 4 48 MIH 0 Normal5 3 M 3 LUC 0 10 78 NP A+1 Normal6 3 F 3 GTC 0 18 120 MIH A Transient mild right hemiparesis7 3-8 F 3 GT 1 27 108 NP 0 Normal8 3 M 3 GTC and UC 0 36 and 48 106 NP A+1 Normal9 4-5 M 4 RUC 0 12 42 MIH 0 Normal10 4 M 3 GTC and UC 0 12 and 39 114 IIH A Normal11 2-5 F 3 GTC and L UC 0 20 >192 SIH Asymmetrical dystonic/spastic motor disorder

with recovery of vision and speech over fourmonths

12 4-5 M 1 GTC and UC 2 6 >192 NP 0 Blind, no language, normal gait13 2 M 3 GTC and L UC 0 6 >192 SIH 0 Blind, no language, spastic quadriparesis,and

epilepsy14 2 F 2 L UC 1 12 >192 NP A Vegetative state, with mixed

pyramidal/extrapyramidal motor disorder

*AMS=Adelaide motor score: 4, localising pain; 3, flexing to pain; 2, extending to pain; and 1, no response to pain.tSeizures: G=generalised; U=unilateral; T=tonic; C=clonic; L=left; R=right.*Status: time (hours) with one seizure lasting >30 minutes or more than three consecutive seizures.§lntracranial pressure (ICP) monitoring: mild intracranial hypertension=maximum ICP 10-20 mm Hg; intermediate intracranial hypertension=maximum ICP20-60 man Hg and minimum cerebral perfusion pressure (CPP) <50 mm Hg; severe intracranial hypertension=maximum ICP >60 mm Hg and minimum CPP<40 mm Hg. NP=not performed; MIH=mild intracranial hypertension; IIH=intermediate intracranial hypertension; SIH=severe intracranial hypertension.lHypoglycaemia <2-2 mmol/l. A=admission; NA=not on admission.

282


http://adc.bmj.com

/A

rch Dis C


ownloaded from

http://adc.bmj.com/

Brain swelling and ischaemia in Kenyans with cerebral malaria

Figure I Patient 10. (A) Acute scan: diffuse brain swelling with loss ofsulci and compression of ventricles (arrow).(B) Convalescent scan: resolution of brain swelling, with increase in the sulci and ventricles but without atrophy.

damage was different in each patient. Onechild (patient 11) had a typical superficialwatershed distribution on the initial scan withlow density in the basal ganglia, and subse-quently developed an intracerebral haemor-rhagic lesion in the watershed area between theleft middle and posterior cerebral arteries,which enhanced with intravenous contrast(fig 3). Another child (patient 12) had amaximum watershed distribution with sparingof small areas in the anterior cerebral arteryand posterior cerebral artery territories only.The third child (patient 13) had diffuse lowdensity tissue in the cerebral hemispheres, butthe basal ganglia and posterior fossa werespared; he went on to develop cerebral atrophy

and evidence of infarction in the right frontaland parietal regions (fig 2). The other child(patient 14) had scattered irregular areas ofhypodensity throughout the cerebral hemi-spheres and basal ganglia compatible withmicrovascular obstruction (fig 4). None of thechildren had any evidence of thalamic or

posterior fossa abnormality. Generalised hypo-density on the acute scans reliably predictedprolonged coma, followed by significantneurological sequelae associated with cerebralatrophy on the convalescent scans.There was no association between the find-

ing of low density on computed tomographyand hypoglycaemia (blood glucose <2-2mmol/1) either on admission or during the stay

Table 3 Computed tomography findings in 14 children with cerebral malaria

Acute scan Convalescent scan

Time since Time since Grade* Time sincePatient onset of admission of diffuse initial scanNo coma (days) (hours) Appearances on computed tomography brain swelling (days) Appearances on computed tomography

1 4 48 Normal N NPt2 4 84 Normal N 120 Mild cerebral atrophy3 1-5 48 Normal N NP4 4 48 Normal N NP -

5 6 70 Normal N NP -

6 6 44 Normal N NP -

7 3 36 Normal with small ventricles N NP8 4-2 71 Normal with small ventricles N NP -

9 3-4 40 Swollen brain 1 12 Normal10 4 72 Swollen brain 1 24 Normal11 5 120 Generalised hypodensity of hemispheres 3 11 Generalised infarction of hemispheres and

and basal ganglia but sparing posterior basal ganglia with enhancement offossa. Watershed distribution watershed area between left posterior and

middle cerebral arteries12 4 72 Ischaemia of hemispheres not affecting 1 49 Cerebral atrophy with infarction of posterior

basal ganglia or posterior fossa. temporal and parietal regionsMaximum watershed distribution

13 6 148 Ischaemia of hemispheres not affecting 1 70 Cerebral atrophy with infarction of rightbasal ganglia or posterior fossa. Not anterior and posterior regions (withwatershed distribution occipital sparing) and left parietal regions

14 5 76 Ischaemia of hemispheres and basal ganglia. 2 31 Cerebral atrophy without any focal infarctsScattered hypodense areas. Notwatershed distribution

*Criteria for grading in table 1.tNP=Follow up computed tomography not performed.

283


http://adc.bmj.com

/A

rch Dis C


ownloaded from

http://adc.bmj.com/

Newton, Peshu, Kendall, Kirkham, Sowunmi, Waruinu, Mwangi, Murphy, Marsh

_ , ",___' _- t) ...._r .24fe ;§ i | '_ jl* S . , __ s. ?s R _

t: , ._> z _s z _

_p x_|_ _E .qEs ia__F t_w ::F_ t .r 'Bi *t 'S. wX rX ww _ F [ wc,5Z, *s3SF,-ti_.}X-. -.j_ _ .Sty. w.. - ^.

Figure 2 Patient 13. (A) Acute scan: brain swelling with diffuse hypodensity sparing the basal ganglia (arrows).(B) Convalescent scan: cerebral atrophy with infarction (arrows) of rightfrontal and parietal regions.

!a-

S ';Swit' l_ W.fl,_......

'3: ..> Ss t> trFigure 3 Patient 11. (A) Acute scan: diffuse brain swelling with extensive hypodensity of the superficial watershed areas(arrows) and the basal ganglia. (B) Convalescent scan: contrast scan showing enhancement of the border zone (arrows)between the left middle and postenor cerebral artery areas.

in hospital (Fisher's exact test one tail p=0 58)or status epilepticus (three or more seizureseach hour, or a seizure lasting more than 30minutes) (Fisher's exact test one tail p=0 59).There was no relation between the parasit-aemia on admission and the presence ofbrain swelling or hypodensity on computedtomography.The ICP was monitored in seven children

(three others fulfilled the criteria but were notmonitored, either for technical reasons (twochildren) or because the platelet count was<40X 109/1 (one child)). Four had mildintracranial hypertension of whom one haddiffuse brain swelling without evidence ofbraindamage and three had normal scans, one hadintermediate intracranial hypertension anddiffuse brain swelling without brain damage,

284


http://adc.bmj.com

/A

rch Dis C


ownloaded from

http://adc.bmj.com/


Figure 4 Patient 14. (A) Acute scan: diffuse brain swelling with scattered areas of hypodensity (arrows) throughoutincluding the basal ganglia. (B) Convalescent scan: cerebral atrophy.

and two had severe intracranial hypertensionwith minimum CPP <40 mm Hg andhad diffuse brain swelling with widespreadhypodensity. Detailed analysis of the ICP datais in preparation.

DiscussionThis is the first report describing computedtomography findings in African children withcerebral malaria. Computed tomographyperformed during the recovery phase in 14children showed a wide range of appearances,some of which have not been documentedpreviously in adults. Eight scans were normal,though two of these had ventricles which werenoticeably small, but within the normal range.Six children had diffuse brain swelling; onlytwo of these had no evidence of brain damage.Their follow up scans showed complete resolu-tion of the swelling and they recovered withoutsequelae. In the other four children diffusebrain swelling was associated with widespreadareas of low density on the initial scan, and allthese children had severe cerebral atrophy andinfarction on their follow up scans associatedwith significant neurological sequelae.

Diffuse brain swelling is characterised incomputed tomography by small ventricles, theabsence of sulci, and compression of theperimesencephalic and chiasmatic cisternswith resolution on follow up scan.17 18 In headinjury, diffuse brain swelling occurs morecommonly in children than adults19 and hasbeen associated with a greater tomographicdensity in the white matter, reflecting anincrease in blood volume.'7 The swelling mayresolve quickly,'7 and thus we cannot excludethe possibility that we, did not detect this insome of the normal scans, as these were

performed while the child was recoveringconsciousness. The degree of brain swellingcorrelated with ICP in a study of childrenwith non-traumatic coma; however, theappearances on computed tomography couldnot be used to predict the ICP as somechildren with normal scans developedincreased ICP and children with focal abnor-malities in the basal ganglia or cerebral hemi-spheres had a loss of cerebrospinal fluid spacewithout a generalised increase in ICP.'6 In thechildren with cerebral malaria the most severeswelling occurred in those with the highestICP, whether or not there were also lowdensity areas. In the four children with themost severe swelling the tomographic densitieswere decreased, probably from an accumula-tion of intracellular fluid secondary to impairedcellular metabolism (cytotoxic oedema) andinterstitial fluid after the breakdown of theblood-brain barrier (vasogenic oedema).20Acute hydrocephalus was not seen in any ofthechildren and thus is unlikely to be the cause ofincreased ICP in children with cerebralmalaria.

In the two children with diffuse brainswelling without low density areas, the tomo-graphic densities in the white matter werenormal and there was no enhancement withthe administration of contrast medium. Thusthere is no evidence for an increase in bloodvolume or for vasogenic oedema, but thesepossibilities cannot be excluded as computedtomography may not detect mild increases incerebral blood volume or oedema. Further-more, if these disorders coexist, the changes intomographic density will tend to cancel the twoout.

In this series all the children with severeneurological sequelae had evidence of brain

285


http://adc.bmj.com

/A

rch Dis C


ownloaded from

http://adc.bmj.com/

Newton, Peshu, Kendall, Kirkham, Sowunmi, Waruiru, Mwangi, Murphy, Marsh

damage on computed tomography which waseither diffuse or affected the boundary zonesbetween the anterior and middle cerebralarteries, or the middle and posterior cerebralarteries. There was no evidence of posteriorfossa abnormality, though the basal gangliawere affected in two children. These patternsof damage have been seen on computedtomography after low CPP,21 hypogly-caemia,22 and status epilepticus.23 Boundaryzone ischaemia is thought to arise from asudden precipitate decrease in CPP,24 but hasalso been seen by computed tomography afterhypoglycaemia.22 The more diffuse changesaffecting in order of preference the cortex, thebasal ganglia, and the cerebellum are morelikely to be seen in patients in whom there hasbeen a moderate but prolonged decrease inCPP24 or status epilepticus.25 In cerebralmalaria such patterns could result from severalpossibly interacting mechanisms. Firstly, aglobal reduction in CPP may be caused byhypotension or intracranial hypertension, orboth; two children who had ICP monitoringand who developed neurological sequelae hadsevere intracranial hypertension and in bothchildren the CPP was reduced into the rangeassociated with poor outcome in otherencephalopathies.8 Because of the timing ofICP monitoring and computed tomography,however, it is not clear whether intracranialhypertension has a primary role or is itself asecondary response to cellular damage causedby other mechanisms. Status epilepticus andhypoglycaemia are common features ofcerebral malaria and although the mechanismsof damage may be different from low Cpp,2526both may produce additive neuropathologicalchanges when the CPP is reduced. It is alsopossible that there is a synergistic interactionbetween decreased CPP and the sequestrationof parasite infected cells. Sequestration maylead to reduced peripheral perfusion and asinfected cell cytoadherence is enhanced by lowshear stress,27 a decrease in CPP from anycause may promote sequestration in areas oflow flow. A further potential mechanism forbrain damage in cerebral malaria would betranstentorial herniation secondary to intra-cranial hypertension. Although children whodie of cerebral malaria often show progressiveneurological deterioration consistent with her-niation,9 the distribution of damage observedin this study of survivors is different, withoutthe expected prominent involvement of struc-tures in the territory of the posterior cerebralartery. However, this possibility could not beexcluded in those with very diffuse hypodensityon computed tomography. It is likely that incerebral malaria herniation is a cause of deathrather than of sequelae.The features seen on computed tomography

in these childrencontrastwith those in adultswith cerebral malaria and provide furtherevidence of the differences in the pathophysi-ology between these two groups. In non-immune adults brain swelling is associatedwith cerebral oedema and in a study of10patients cerebral oedema was only detectedduring the agonal phases in two adults.'2

Cerebral oedema has since been reported in anadult who survived.13 Focal areas of alteredtomographic density have been seen inadults,12 13 but this feature was not associatedwith clinical signs or neurological sequelae.

Brain swelling is a feature of children withsevere cerebral malaria and it is likely to con-tribute directly to the intracranial hypertensionseen in these children. Although the mostsevere brain swelling was associated with cyto-toxic oedema and brain damage, two childrenwith definite swelling recovered without anyneurological sequelae. These features seen oncomputed tomography suggest that secondaryevents such as brain swelling, hypotension,hypoglycaemia, and status epilepticus may beimportant in the pathogenesis of cerebralmalaria and its sequelae. Further understand-ing of brain damage in children with cerebralmalaria and of the importance of intracranialhypertension requires clarification fromdetailed neuropathological studies and neuro-radiological studies undertaken earlier in thecourse of the illness.We thank the director of KEMRI, Dr D Koech, for permissionto publish these results. We thank DrJ Crawley, Dr J Davies,Dr P A Winstanley, and the nurses on the KEMRI unit, Kilififor help in looking after the patients; Drs P Emurwon and Z AKaka at the Aga Khan Hospital, Mombasa and Dr F 0 Nondiat the Coast Imaging Clinic, Mombasa for performing thecomputed tomography; and Professor B G Neville andProfessor A D Edwards for comments on the manuscript. Thework was funded by the Wellcome Trust, UK and formed partof a collaborative study of the pathophysiology of malaria inchildren. Dr C Newton is a Wellcome Trust AdvancedTraining Fellow and Dr K Marsh is a Wellcome Trust SeniorResearch Fellow in clinical science.

1 Snow RW, Armstrong-Schellenberg JRM, Peshu N, et al.Periodicity and time-space clustering of severe childhoodmalaria on the coast of Kenya. Trans R Soc Trop Med Hyg1993; 87: 386-90.

2 Warrell DA, Molyneux ME, Beales PF. Severe and compli-cated malaria. Trans R Soc Trop Med Hyg 1990; 84 (suppl2):1-65.

3 Molyneux ME, Taylor TE, Wirimani, Borgstein A. Clinicalfeatures and prognostic indicators in paediatric cerebralmalaria: a study of 131 comatose Malawian children. Q JMed 1989; 71: 441-59.

4 Brewster DR, Kwiatkowski D, White NJ. Neurologicalsequelae of cerebral malaria in children. Lancet 1990; 336:1039-43.

5 Phillips RE, Warrell DA. The pathophysiology of severefalciparum malaria. Parasitology Today 1986; 2: 271-82.

6White NJ, Ho M. The pathophysiology of malaria. AdvParasitol 1992; 31: 84-173.

7 Omanga U, Ntihinyurwa M, Shako D, Mashako M. Leshemiplegies au cours del'access pernicieux a plasmodiumfalciparum del'enfant. Ann Pediatr 1983; 30: 294-6.

8 KirkhamIJ. Intracranial pressure and cerebral bloodflow innon-traumatic coma in childhood. In: Minns RA, ed.Problems of intracranial pressure in childhood. London:MacKeith Press, 1991: 283-348.

9 Newton CRJC, Kirkham Fl, Winstanley PA, et al.Intracranial pressure in African children with cerebralmalaria. Lancet 1991; 337: 573-6.

10 Waller D, CrawleyJ, Nosten F, et al. Intracranial pressure inchildhood cerebral malaria. Trans R Soc Trop Med Hyg1991; 85: 362-4.

11 Thomas JD. Clinical and histopathological correlation ofcerebral malaria. Trop GeogrMed 1971; 21: 232-8.

12 LooareesuwanS, Warrell DA, White NJ, et al. Do patientswith cerebral malaria have cerebral oedema? A computedtomography study. Lancet 1983; i: 434-7.

13 Pham-Hung G, Truffert A, Delvallee G, Michel G, LaporteJP, Duval G. Infarctus cerebral au cours d'un accespernicieux palustre. Interet diagnostique de la tomo-densitometrie. Ann FrAnesth Reanim 1990; 9: 185-7.

14 Simpson D, Reilly P. Pediatric coma scale. Lancet 1982; ii:450.

15 Tasker RC, Matthew DJ. Cerebral intraparenchymalpressure monitoring in non-traumatic coma: clinicalevaluation of a new fiberoptic device. Neuropediatrics1991; 22: 47-9.

16 Tasker RC, Matthew DJ, Kendall B. Computed tomo-graphy in the assessment of raised intracranial pressure innon-traumatic coma. Neuropediatrics 1990; 21: 91-4.

17 Bruce DA, Alavi A, BilaniukL, Dolinskas C, Obrist W.Uzzeli B. Diffuse cerebral swelling following head injuriesin children: the syndrome of 'malignant brain edema'.JNeurosurg 1981; 54: 170-8.

286


http://adc.bmj.com

/A

rch Dis C


ownloaded from

http://adc.bmj.com/


18 Teasdale E, Cardoso E, Galbraith S, Teasdale G. CT scan insevere diffuse head injury: physiological and clinical corre-lations. JNeurolNeurosurgPsychiatry 1984; 47: 600-3.

19 Aldrich EF, Eisenberg HM, Saydjari C, et al. Diffuse brainswelling in severely head-injured children. J Neurosurg1992; 76: 450-4.

20 Go KG. The cerebral blood supply. Cerebral pathophysiol-ogy. Amsterdam: Elsevier, 1991: 208-77.

21 Kjos BO, Zawadzki MB, Young RG. Early CT findings ofglobal central nervous system hypoperfusion. AJ3R 1983;141: 1227-32.

22 Iwai A, Sakamoto T, Kinoshita Y, Yokotai J, Yoshioka T.Sugimoto T. Computed tomographic imaging of the brainin after hypoglycemia coma. Neuroradiology 1987; 29:398-400.

23 Aicardi J, Cheverie n. Consequences of status epilepticus ininfants and children. Adv Neurol 1983; 34: 115-25.

24 Adams JH, Brierley JB, Connor RCR, Treip CS. The effectsof systemic hypotension upon the human brain. Clinicaland neuropathological observations in 11 cases. Brain1966; 89: 235-68.

25 Zimmerman HM. The histopathology of convulsivedisorders in children. Jf Pediatr 1938; 13: 859-90.

26 Meldrum BS, Horton RW, Brierley JB. Insulin-inducedhypoglycaemia in the primate: relationship betweenphysiological changes and neuropathology. In: BrierleyJB, Meldrum BS, eds. Brain hypoxia. Clinics in develop-mental medicine 39/40. London: Spastics InternationalMedical Publishers/Heinemann Medical, 1971:207-24.

27 Nash GB, Cooke BM, Marsh K, Berendt A, Newbold C,Stuart J. Rheological analysis of the adhesive interactionsof red blood cells parasitised by Plasmodium falciparum.Blood 1992; 79: 798-807.

Evening primrose oil

It would be right and proper if evening primrose oil were 'a goodthing' simply because the name has such implications of beauty.As you know, there has been considerable debate and conflictingdata on its use in eczema. Recently published data fromLeicester (J Berth-Jones and R A C Graham-Brown, Lancet1993; 341: 1557-60) add to the 'con' side of the argument.

It is postulated that the effect of evening primrose oil isdependent on its content ofn6 series essential fatty acids (EFAs)which might effect the metabolism of the mediators of inflam-mation such as prostaglandins and leukotrienes. Fish oil, whichprovides n3 series EFAs, has also been suggested as a treatmentfor eczema. The Leicester workers performed a double blindtrial in which patients received one of three possibilities: eveningprimrose oil alone, evening primrose oil with fish oil, or placebo.There were 41 patients in each group and half were children of12 years and under.The treatment was given for 16 weeks and the results assessed

using clinical severity scores, patient diary scores, and use oftopical steroid. No improvement was demonstrated with eitherof the active treatments. The patients' diary scores tended to bebetter on placebo but there were no significant differencesbetween the three groups at 16 weeks. There was nodemonstrable difference in response between children andadults.These authors point to methodological problems with

previQus studies which have shown a beneficial effect of eveningprimrose oil, though they do not claim to be able to explain awayall such results. No doubt the debate will continue.

ARCHIVIST

287


http://adc.bmj.com

/A

rch Dis C


ownloaded from

http://adc.bmj.com/

ischaemia in cerebral - adc.bmj.com · health organisation's criteria for cerebral malaria2...

Documents