Archives ofDisease in Childhood 1991; 66: 702-705

Imaging in influenza A encephalitis

S M Protheroe, D H Mellor

AbstractTwo cases of influenza A encephalitis seenduring an outbreak of influenza types A/England/427/88 (H3N2) and A/Taiwan/1/86(HIN1) in December 1989 are described. Inboth children the encephalitis developedwithin three days of the respiratory symptomsand both became comatose within 48 hours.Virological studies showed that the patientshad had a recent influenza A infection. Sym-metrical localised hypodense lesions withinthe thalami and pons were demonstrated inboth cases on computed tomography of thebrain and striking findings in the pons in onecase on magnetic resonance imaging.Influenza A encephalitis is not easy to recog-nise clinically and serological confirmationcan only be made after 10 days. Imaging mayprovide evidence in the acute stage to supporta diagnosis of influenza encephalitis duringinfluenza outbreaks.

An outbreak of influenza occurred in Notting-hamshire in late 1989. Two different influenzaviruses A/England/427/88 (H3N2) A/Taiwan/l/86 (HINI) were implicated in the epidemic.'Two children admitted to our service duringthis period were found to have influenzalencephalitis. They showed unusual computedtomography of the head, which has been des-cribed previously in two cases of influenzalencephalitis.2 3 Magnetic resonance imagingwas carried out in our second case and showedstriking changes.

acids were also normal. Electroencephalographyshowed generalised slow wave activity.Computed tomography of the brain on the

second day showed low attenuation changes inthe thalami bilaterally (fig 1). There was somebrainstem swelling with low attenuation changein the pons (fig 2). Both putamina and globus

Figure I Axialcomputedtomogram ofthe brain incase Ishowinglowattenuation lesions in both thalami(arrowed).

Department ofPaediatrics,University Hospital,NottinghamS M ProtheroeD H MellorCorrespondence to:Dr S M Protheroe,Department of Paediatrics,Peterborough DistrictHospital,Thorpe Road,Peterborough PE3 6DA.Accepted 17 January 1991

Case reportsCASE 1A 4 year old boy was admitted after two shortgeneralised convulsions from which he had notgained full consciousness. He had been febrilethe night before admission with malaise andcough. He had a history of febrile convulsionswhen aged 2.On examination he was drowsy with a tem-

perature of 38-0°C. His optic fundi were nor-mal, but he had nystagmus and was hypertonicwith brisk deep tendon reflexes and extensorplantar responses. There was no meningeal irri-tation or sensory disturbance. He had a furtherfit and remained drowsy.

Cerebrospinal fluid showed a normal proteinconcentration (0-3 g/l) with no increase in cells.Full blood count, blood glucose, urea and elec-trolytes, hepatic enzymes, copper, caeruloplas-min, lactate, ammonia, and coagulation testswere all normal. Urine amino acids and organic

Figure2 AxialcomputedtomogramofthebrainincaseIshowingswellingandlowzattenuationchangeinthepons(arrowed).

702

on August 31, 2020 by guest. P

rotected by copyright.http://adc.bm

j.com/

Arch D

is Child: first published as 10.1136/adc.66.6.702 on 1 June 1991. D

ownloaded from

Imaging in influenza A encephalitis

pallidi showed small foci of high attenuationpossibly representing haemorrhage.On the third day his condition deteriorated

with decorticate posturing, early papilloedema,and a sixth nerve palsy on the right. Dexa-methasone was started, he was ventilated, andan intracranial pressure monitoring device wasinserted. He was treated with acyclovir andthiamine, although the features were not typicalor herpes simplex encephalitis or Leigh'sdisease.He was extubated after four days but

remained in a somnolent state. By day 10 he hadbegun to improve, although he showedchoreoathetoid movements of his trunk andlimbs on purposeful movement. Complementfixation titre for influenza A virus was less than16 on admission, but increased to +256 on the15th day, indicating recent influenzal infection.Serological tests failed to demonstrate any othervirus antibody response and we were not able tofind any other cause for coma or encephalo-pathy.He was discharged on day 28, walking with

an aid and speaking in sentences. Follow upthree months later showed he had made a fullneurological and intellectual recovery apartfrom a persisting right lateral rectus palsy.

CASE 2An 11 year old girl was admitted having beenfound semicomatose by her parents. For theprevious three days she had been febrile withheadache, cough, and malaise. Two familymembers also had influenza like symptoms. Shehad had an acute encephalitis at age 3, fromwhich she had recovered, but had been left withresidual behavioural problems, left divergentsquint, precocious puberty, and mild optic atro-phy. This episode had also followed an upperrespiratory tract infection and influenza A wasisolated from a throat swab. Computed tomog-raphy of the brain at that time had shown linearhypodensities in the external capsule region onboth sides.On this occasion she was drowsy but able to

respond to commands on admission. Pale opticdiscs were noted as before. She was pyrexial andhad brisk tendon reflexes with extensor plantarresponses. The cerebrospinal fluid showed araised protein concentration (1-2 g/l) with nowhite cells, 36 red cells/mm3, and normal glu-cose. Serum electrolytes, ammonia, hepaticfunction tests, copper, caeruloplasmin, lactate,and a full blood count were within normallimits, as were urine amino acids and organicacids. Trichrome staining of a muscle biopsyspecimen showed no 'ragged red muscle fibres'.

She became unrousable by the second dayand was unresponsive to pain with loss of thegag reflex. Central respiratory irregularities,namely hyperventilation and Cheyne-Stokesrespiration became apparent, and she was intu-bated and ventilated. She was treated with acyc-lovir and dexamethasone.Computed tomography of the brain showed

symmetrical well defined areas of low attenua-tion in both thalami extending to involve thecapsular fibres. Further diffuse low attenuation

changes were seen within the pons and mid-brain. Some swelling of the brainstem was seen.Low attenuation changes considered longstand-ing and a result of her previous encephalitiswere noted in the external capsules. Magneticresonance imagimg imversion recovery sequencesof the brain showed low signal changes in thethalami and pons (fig 3). Axial T2 weightedimages showed ring like areas of high signalintensity with central low signal in the pons (fig4). Complement fixation titres for influenza Avirus were less than 16 on admission butincreased to +256 by day 15.By day 17 of her illness it was possible to take

her off the ventilator and extubate her. Sheopened her eyes to sensory stimuli but did notrecognise her parents. Spasticity affected allfour limbs; she was unable to swallow andrequired tube feeding. She remained in this'sleeping beauty, akinetic mute' state for threemonths without much change. Progress wasvery slow and five months after the onset of herillness she was still immobile with expression-less facies and paucity of movement. She couldsit in a chair with support and recognise her

Figure 3 Saggital inversion recovery magnetic resonanceimaging brain scan in case 2 showing low signal changes inthe pons (arrowed).

Figure 4 Axial T2 weighted magnetic resonance imagingbrain scan in case 2 showing ring like areas ofhigh signal inthe pons (arrowed).

703

on August 31, 2020 by guest. P

rotected by copyright.http://adc.bm

j.com/

Arch D

is Child: first published as 10.1136/adc.66.6.702 on 1 June 1991. D

ownloaded from

704 Protheroe, Mellor

parents. Speech had not returned but she coulduse a simple communication board and under-stand simple sentences.

DiscussionThese two children were considered to haveinfluenza A encephalitis on the basis of theprodromal illness, clinical course, and virologystudies. Clinically they resemble the six casesdescribed by Delorme and Middleton, whodeveloped fever, reduced consciousness, andneurological signs after influenzal symptoms.4Analyses of the cerebrospinal fluid were normalin these six cases apart from a slight increase inprotein in one patient, but computed tomogra-phy was not done.True encephalitis is comparatively rare after

influenza A but benign confusional states andminor neurological signs (for example, ocularmuscle palsies) are observed more often.5 Para-infectious processes such as transverse myelitisand Reye's syndrome are also observed,F'0although virus isolation from the brain or cere-brospinal fluid is infrequent." 12 Neurologicalsymptoms and signs may appear several weeksafter the infection making it difficult to estab-lish an aetiological link in some cases.The finding of focal low attenuation lesions in

the frontotemporal regions on computed tomog-raphy can be helpful in the diagnosis of herpessimplex encephalitis.'3 14 Computed tomogra-phy is usually diagnostically unrewarding inother forms of encephalitis. However, our find-ings suggest that there may be a typical appear-ance on computed tomography in influenzaencephalitis, and this is in keeping with tworeports from Japan describing similar localisedhypodense lesions.2 3 Recognition of thesetomographic abnormalities in patients with pos-sible influenzal encephalitis may aid diagnosiswhile viral titres are awaited.The findings on magnetic resonance imaging

in brainstem encephalitis include pontineenlargement and increased signal lesions in thepons. 15 16 Magnetic resonance imaging is super-ior to computed tomography for detectingbrainstem encephalitis, as seen in our secondcase where obvious pontine abnormalities wereshown on magnetic resonance imaging but notwell seen on computed tomography. A similardiscrepancy between magnetic resonance imag-ing and tomography findings in brainstemencephalitis has been previously reported wherediscrete abnormal areas were seen in the pons,midbrain, and thalamus on magnetic resonanceimaging but the computed tomogram hadrevealed only a low density area in the pons.17

It has been observed that a normal computedtomogram on follow up correlates with fullclinical recovery in viral encephalitis ingeneral.'8 19 This is also seen in the previouslyreported case of influenza A encephalitis wherethe initial abnormal hypodense areas in the mid-brain, basal ganglia, and cerebral white matterhad disappeared on follow up computed tomog-raphy and the patient made a good recovery.3The cause of these lesions is unknown. It has

been postulated that hypodense areas on com-puted tomography in encephalitis reflect the

sequelae of ischaemia induced by compressionof end arteries by brain oedema or may be dueto localised areas of inflammation. It is knownthat the virus can have a direct cytopathic effecton vulnerable sites.2 7Our second case experienced two episodes of

acute encephalitis when infected with influenzaA, which could not be explained by any otheridentifiable cause. On both occasions there wasvirological evidence of recent influenza infec-tion. The original episode was associated withbilateral hypodense areas in the external capsuleregions on computed tomography, with lowattenuation lesions in both thalami, pons, andmidbrain on the second scan.

Localised hypodense lesions in the basalganglia have also been described on computedtomography in encephalitis after infections withECHO, mumps, measles, and Coxsackieviruses.18 20 21 Other causes of low density areasin the basal ganglia and thalamus have beendescribed in disorders such as Leigh's encepha-lopathy, Wilson's disease, and mitochondrialcytopathies as well as carbon monoxide poison-ing and anoxia.22 Wilson's disease, Leigh'sencephalopathy, and mitochondrial cytopathiesseem unlikely in view of the normal copperstudies and lactate concentrations in both casesand the normal muscle biopsy specimen incase 2.

Circumstantial evidence has implicated theinfluenza virus as the cause of the encephalitislethargica epidemic of the 1920s.23 Similar iso-lated cases have been described from year toyear often at the time ofinfluenza epidemics.2429von Economo first described the illness, post-ulated a viral aetiology, and reported foci ofinflammation in the grey matter of the midbrainand basal ganglia.30 The manifestations ofencephalitis lethargica were variable but a num-ber of clinical features have been proposed asmajor criteria to support the diagnosis.3' Ourfirst case showed signs of basal ganglia involve-ment with choreoathetoid movements, ophthal-moplegia, and somnolence. Our second caseexhibited akinetic mutism, central respiratoryirregularities, and somnolence. Combined withthe acute encephalitic illness, our cases fulfil thecriteria to support the diagnosis of encephalitislethargica. We are aware that unexpectedneurological sequelae such as parkinsonismwith oculogyric crises can appear after apparentrecovery as long as four years later.

1 PHLS Communicable Disease Surveillance Centre. Com-municable disease report. CDR 89/44. London: CDSC,1989; 3rd November: 3-4.

2 Hattori H, Kawamori J, Takao T, et al. Computed tomo-graphy in postinfluenzal encephalitis. Brain Dev 1983;5:564-7.

3 Okuno T, Takao T, Ito M, Mikawa H, Nakano Y. Contrastenhanced hypodense areas in a case of acute disseminatedencephalitis following influenza A virus. Comput Radiol1982;6:215-7.

4 Delorme L, Middleton PJ. Influenza A virus associated withacute encephalopathy. AmJ Dis Child 1979;13:822-4.

5 Sulkava R, Rissanen A, Pyhala R. Post-influenza encephalitisduring the influenza A outbreak in 1979/80. J NeurolNeurosurg Psychiatry 1981 ;44: 161-3.

6 Wells CEC. Neurological complications of so-called'influenza' winter study in S E Wales. BMJ 1971 ;i:369-73.

7 Booss J, Esiri M. Epidemic encephalitis I. Viral encephalitis.Pathology, diagnosis and treatment. Oxford: BlackwellScientific Publications, 1986:141 .

8 Flewett TH. Influenzal encephalopathy and post-influenzalencephalitis. Lancet 1958,ii: 11-5.

9 Dubowitz V. Influenzal encephalitis. Lancet 1958;ji:140.

on August 31, 2020 by guest. P

rotected by copyright.http://adc.bm

j.com/

Arch D

is Child: first published as 10.1136/adc.66.6.702 on 1 June 1991. D

ownloaded from

Imaging in influenza A encephalitis 705

10 Horner FA. Neurological disorders after Asian influenza.N EnglJ Med 1958;258:983-5.

11 Anonymous. News and Notes. Epidemiology. LaboratoryReports. BMJ 1970;i:311.

12 Frankova V, Jirasek A, Tumova B. Type A influenza virusisolations from different organs in human lethargica cases.Arch Virol 1977;53:265-8.

13 Thompson JLG. The computed axial tomography in acuteherpes simplex encephalitis. BrJ Radiol 1976;49:86-7.

14 Davis JM, Davis KR, Leinman GM, Kirchner HS, TaverasJM. Computed tomography of herpes simplex encephalitiswith clinicopathological correlation. Radiology 1978;129:409-17.

15 Davidson HD, Steiner RE. Magnetic resonance imaging ininfections of the central nervous system AJNR 1985;6:499-504.

16 Furman JM, Brownstone PH, Baloh RN. A typical brainstemencephalitis magnetic resonance imaging and oculographicfeatures. Neurology 1985;35:438.

17 Hosoda K, Tamaki N, Masumoa M, Matsumoto S. Magneticresonance images of brainstem encephalitis. J Neurosurg1987;66:283-5.

18 Charney R, Orecchio EJ, Simmerman RA, Berman RH.Computerised tomography in infantile encephalitis. AmJ Dis Child 1979;133:803-5.

19 Bittner T, Dorndorf W. Prognostic value of computerisedtomography and cerebrospinal fluid analysis in viralencephalitis. J Neuroimmunol 1988;20:16.

20 Goutieres F, Aicardi J. Acute neurological dysfunctionassociated with destructive lesion of the basal ganglia inchildren. Ann Neurol 1982;12:328-32.

21 Ochi J, Okuno T, Uenoyama Y, Narita H, Mikawa H.Symmetrical low density areas in bilateral thalami in aninfant with measles encephalitis. Comput Radiol 1986;10:137-9.

22 Aicardi J, Gordon N, Hagberg B. Holes in the brain. DevMed Child Neurol 1985;25:249-60.

23 Maurizi CP. Influenza caused epidemic encephalitis(encephalitis lethargica). The circumstantial evidence andchallenge to the naysayers. Med Hypotheses 1989;28:18.

24 Leigh AD. Infections of the nervous system occurring duringan epidemic of influenza B. BMJ 1946;ii:936.

25 Espir MLE, Spalding JMK. Three recent cases of encephali-tis lethargica. BMJ 1956;i: 1141-4.

26 Hunter R, Jones M. Acute lethargic type encephalitis. Lancet1966;ii: 1023-4.

27 Anonymous. Encephalitis of lethargica type in a mentalhospital [Editorial]. Lancet 1966;ii:1014-5.

28 Greenough A, Davis JA. Encephalitis lethargica. Mystery ofthe past or undiagnosed disease of the past? Lancet 1983;i:922-3.

29 Anonymous. Encephalitis lethargica [Editorial]. Lancet 1981;ii:1396.

30 von Economo C. Encephalitis lethargica: its sequelae and treat-ment. Translated by KO Newman. London: Oxford Uni-versity Press, 1931.

31 Howard RS, Lees AJ. Encephalitis lethargica. A report offour recent cases. Brain 1987;110:19-33.

AddendumSince this paper was accepted for publicationcase 1 was readmitted to hospital after a sei-zure at home. Vomiting and a further seizureoccurred after admission. He was drowsy andfebrile but there was no neck stiffness and nofocal neurological findings. Blood urea, electro-lytes, glucose, prothrombin time, liver functiontests, haemoglobin, and platelets were normal.Total peripheral white cell count was low at2 2 x 109/1.He was given an intravenous loading dose of

phenytoin. Computed tomography showed nochanges. The cerebrospinal fluid was sampledand contained one lymphocyte/mm3, a normalglucose, but a raised protein concentration of1-6 g/l, and a pressure of 35 cm H20. Furthershort seizures occurred and then a respiratoryarrest requiring resuscitation and mechanicalventilation. Despite passive hyperventilationand mannitol his pupils remained fixed anddilated. Brainstem function tests showed noactivity and ventilatory support was withdrawnafter a period of 12 hours. Permission for post-mortem examination was not given. Subse-quently the results of viral antibody titres senton admission became available. All were nega-tive including influenza A and B.

COMMENTThe previous encephalitis had occurred in as-sociation with a serologically proved influenzainfection. This fatal illness a year later raises thepossibility that the original illness could havebeen some kind of metabolic encephalopathyprecipitated by the influenza, although detailedmetabolic investigations were negative.

on August 31, 2020 by guest. P

rotected by copyright.http://adc.bm

j.com/

Arch D

is Child: first published as 10.1136/adc.66.6.702 on 1 June 1991. D

ownloaded from