guidline treatment up

8/19/2019 Guidline Treatment up

1/29

Management of suspected viral encephalitis inchildren e Association of British Neurologists andBritish Paediatric Allergy Immunology and InfectionGroup National Guidelines

R. Kneen a,b,s, *, B.D. Michael b,c,j,k,s , E. Menson d,l , B. Mehta a,m , A. Easton e,n ,C. Hemingway f,o , P.E. Klapper g,p , A. Vincent h,q , M. Lim i,l , E. Carrol a,r ,T. Solomon b,c,j,k , On behalf of the National Encephalitis GuidelinesDevelopment and Stakeholder Groups

a Alder Hey Children’s NHS Foundation Trust, Eaton Road, West Derby, Liverpool L12 2AP, UK b Institute of Infection and Global Health, University of Liverpool, 8thFloor Duncan Building, Daulby Street, Liverpool L69 3GA, UK c The Walton Centre Neurology NHS Foundation Trust, Lower Lane, Fazakerly, Liverpool L9 7JL, UK d Evelina Children’s Hospital London, Guys and St Thomas’, Westminster Bridge Road, London SE1 7EH, UK e Encephalitis Society, 32 Castlegate, Malton, North Yorkshire, Y017 7DT, UK f Great Ormond Street Hospital, 40 Bernard Street, London WC1N 1LE, UK g University of Manchester, 2nd Floor, Clinical Sciences Building 2, Manchester Royal Inrmary, Oxford Road,Manchester, M13 9WL, UK h Oxford Ion Channel and Disease Initiative, Department of Clinical and Experimental Neuroimmunology, WeatherallInstitute of Molecular Medicine, University of Oxford, Oxford OX3 9DS, UK i Guys and St. Thomas’ Hospital, London, Westminster Bridge Road, London SE1 7EH, UK j Department of Neurological Science, University of Liverpool, 8th Floor Duncan Building, Daulby Street, Liverpool L69 3GA, UK

Accepted 13 November 2011Available online 18 November 2011

* Corresponding author. Tel.: þ 44 151 228 4811; fax: þ 44 151 228 032.E-mail addresses: [email protected] (R. Kneen), [email protected] (B.D. Michael), [email protected]

(E. Menson), [email protected] (B. Mehta), [email protected] (A. Easton), [email protected] (C. Hemingway), [email protected] (P.E. Klapper), [email protected], [email protected] (A. Vincent),[email protected] (M. Lim), [email protected] (E. Carrol), [email protected] (T. Solomon).

k Tel.: þ 44 151 529 5460; fax: þ 44 151 529 5465.l Tel.: þ 44 20 7188 7188.m Tel.: þ 44 151 228 4811; fax: þ 44 151 228 032.n Tel.: þ 44 1653 692 583; fax: þ 44 1653 604 369.o Tel.: þ 44 207 405 9200x8308; fax: þ 44 20 7813 8279.p Tel.: þ 44 161 276 8853; fax: þ 44 161 276 5744.q Tel.: þ 44 1865 280528; fax: þ 44 1865 280535.r Tel.: þ 44 151 252 5160.s R. Kneen and B.D. Michael are joint rst authors.

0163-4453/$36 ª 2012 Published by Elsevier Ltd on behalf of The British Infection Association.doi:10.1016/j.jinf.2011.11.013

www.elsevierhealth.com/journals/jinf

Journal of Infection (2012) 64 , 449e 477


2/29

KEYWORDSEncephalitis;Viral encephalitis;Herpes simplex virus;Immunocompromised;Varicella zoster virus;Enterovirus;

Antibody-associatedencephalitis

Summary In the 1980s the outcome of patients with herpes simplex encephalitis was shownto be dramatically improved with aciclovir treatment. Delays in starting treatment, particu-larly beyond 48 h after hospital admission, are associated with a worse prognosis. Several com-prehensive reviews of the investigation and management of encephalitis have been published.However, their impact on day-to-day clinical practice appears to be limited. The emergencymanagement of meningitis in children and adults was revolutionised by the introduction ofa simple algorithm as part of management guidelines.

In February 2008 a group of clinicians met in Liverpool to begin the development process forclinical care guidelines based around a similar simple algorithm, supported by an evidencebase, whose implementation is hoped would improve the management of patients with sus-pected encephalitis.ª 2012 Published by Elsevier Ltd on behalf of The British Infection Association.

Introduction

Encephalitis is dened as a syndrome of neurologicaldysfunction caused by inammation of the brain paren-chyma. Encephalitis has many causes and some are specicto childhood, but fortunately it is relatively rare. Howeverdoctors whotreat acutely ill children shouldbe aware of howto managea child with suspected encephalitisas some of theindividual causes of encephalitis will respond to specictreatments and delays in the diagnosis in these children canbe devastating. Strictly speaking, inammation of the brainparenchyma is a pathological diagnosis, however due to thepractical limitations of this, surrogate clinical markers ofinammation are used (Table 1. Denitions).

Classication of encephalitis

Encephalitis can be caused by many individual diseaseprocesses but can broadly be divided into those associated

with infection (either directly or indirectly) and non-infectious causes. Direct infections of the central nervoussystem (CNS) can be caused by many viruses, bacteria(especially intracellular bacteria such as Mycoplasma pneu-moniae ), parasites and fungi (Table 2. Viral encephalitis;Table 3. Non-viralcauses of encephalitisor encephalopathy).Those indirectly associated with infection include an acutedemyelinating process, which is often temporally related toa prior infection outside of theCNS.This process mayalso fol-low immunisation and is known as acute disseminated en-cephalomyelitis (ADEM). Non-infectious causes includeantibody-mediated encephalitis, which may be paraneoplas-tic for example limbic encephalitis associated with ovarianteratomas or may be an isolated nding. Initially these disor-

ders were reported in adults, but they are being increasinglyrecognised in children. 1 Most viral encephalitides are acute,but sub-acute or chronic presentations are characteristic ofparticular pathogens, especially in the immunocompromised(Table 4. Sub-acute and chronic encephalitis).

Epidemiology

The incidence of encephalitis in children is difcult toestablish as reported studies have used different casedenitions, methodologies and different geographic

locations and study populations. However, in westernsettings reported incidences range from 6.3 to 7.4 per100,000 for all ages (adults and children) and approximately10.5e 13.8 per 100,000 children. 2 In the UK, this shouldequate to 1 e 2 children per year in a typical district generalhospital and 8 e 10 in a large tertiary children’s hospital. Inindustrialised nations, the most commonly diagnosed causeof encephalitis is herpes simplex virus (HSV) with an annualincidence of 1 in 250,000 to 500,000. 3 The age specic inci-dence is bimodal, with peaks in childhood and the elderly.Most HSV encephalitis is due to HSV type 1 but about 10% isdue to HSV type 2. The latter occurs typically in immuno-compromised adults and in neonates in whom it can alsocause a disseminated infection. Varicella zoster virus(VZV) is also a relatively common cause of viral encephali-tis, especially in the immunocompromised, whilst cytomeg-alovirus (CMV) occurs almost exclusively in this group.Enteroviruses most often cause aseptic meningitis but canalso be an important cause of encephalitis. Among theother non-infectious causes of encephalitis, immune medi-ated conditions are increasingly being recognised includingADEM and encephalitis associated with antibodies to thevoltage-gated potassium channel complex, or N-methyl-D-aspartate antibody (NMDA) receptors. 1,4

Table 1 Denitions.

Encephalopathy Clinical syndrome of altered mental status (manifestingas reduced consciousness or altered cognition,personality or behavior)

Has many causes including systemic infection, metabolicderangement, inherited metabolic encephalopathies,toxins, hypoxia, trauma, vasculitis, or central nervoussystem infection

Encephalitis Inammation of the brain Strictly a pathological diagnosis; but surrogate clinicalmarkers often used, including inammatory change inthe cerebrospinal uid or parenchyma inammation onimaging

Causes include viruses, small intracellular bacteria thatdirectly infect the brain parenchyma and some parasites

Can also occur without direct brain infection, forexample in acute disseminated encephalitis myelitis(ADEM), or antibody-associated encephalitis

450 R. Kneen et al.


3/29

Aims and scope of the guideline

In the 1980s the outcome of HSV encephalitis in adults wasshown to be dramatically improved by aciclovir treat-ment. 5,6 Delays in starting treatment, particularly beyond

48 h after hospital admission, are associated with a worseprognosis. 7,8 Several comprehensive reviews of the inves-tigation and management of encephalitis have been pub-lished, 9e 11 but their impact on day-to-day clinicalpractice appears to be limited. 12 e 14 The emergency

Table 2 Causes of acute viral encephalitis, with geographical clues modied from (Solomon and Whitley 2004; Solomon, Hartet al. 2007). 80,110 Note viral causes of chronic encephalitis such as JC viruses are not included here.

Groups Viruses Comments

Sporadic causes (not geographically restricted) listed by groupHerpes viruses

(family Herpesviridae)Herpes simplex virus type 1 Most commonly diagnosed sporadic encephalitisHerpes simplex virus type 2 Causes meningitis in adults (esp. recurrent);

Meningoencephalitis occurs typically in theimmunocompromised. Also causes a radiculitis.

Varicella zoster virus Post-infective cerebellitis, or acute infectiveencephalitis or vasculopathy

Epstein e Barr virus Encephalitis in the immunocompromisedCytomegalovirus Encephalitis in the immunocompromised; also retinitis

or radiculitis; often neutrophilic CSF with low glucoseHuman herpes virus 6 & 7 Febrile convulsions in children (after roseola);

encephalitis in immunocompromisedEnteroviruses (family

Picornaviridae)Enterovirus 70 Epidemic haemorrhagic conjunctivitis, with

CNS involvementEnterovirus 71 Epidemic hand foot and mouth disease, with aseptic

meningitis, brainstem encephalitis, myelitisPoliovirus MyelitisCoxsackieviruses, Echoviruses,

ParechovirusMostly aseptic meningitis

Paramyxoviruses(family Paramyxoviridae)

Measles virus Causes acute post-infectious encephalitis,sub-acute encephalitis and sub-acute sclerosingpanencephalitis

Mumps virus Parotitis, orchitis or pancreatitis may occur before,during or after meningoencephalitis

Others (rarer causes) Inuenza viruses, adenovirus,Erythrovirus B19, lymphocyticchoreomeningitis virus, rubella virus,

Arthropod-borne and zoonotic viruses a

Flaviviruses(family Flaviviridae)

West Nile virus North America, Southern Europe, Africa, Middle East,West and Central Asia associated with accid paralysisand Parkinsonian movement disorders

Japanese encephalitis virus Asia, associated with accid paralysis andParkinsonian movement disorders

Tick-borne encephalitis virus Travel in Eastern Europe, Former USSR; tick bite;upper limb accid paralysis

Dengue viruses (types 1 e 4) Causes fever, arthralgia, rash and haemorrhagicdisease, occasional CNS disease

Alphaviruses

(family Togaviridae)

Western, Eastern and Venezuelan

equine encephalitis viruses

Found in the Americas; encephalitis of horses and

humansChikungunya virus Asia Pacic, AfricaBunyaviruses Lacrosse virus Encephalitis in AmericaColtiviruses Colorado tick fever virus North AmericaRhabdoviruses Rabies, virus other lyssaviruses Non-arthropod-borne zoonitic viruses transmitted

by dogs, cats, bats, depending on locationChandipura virus Transmitted by sandies, causing outbreaks

in IndiaHenipah Viruses Nipah virus Transmitted in faeces of fruit bats in Malaysia,

Bangladesha Most are zoonotic e i.e. animals rather than humans are the main natural hosts, the exceptions being dengue and chikungunya

viruses.

Association of British Neurologists and British Paediatric Allergy Immunology and Infection Group National Guidelines 451


4/29

management of meningitis in children and adults was rev-olutionised by the introduction of a simple algorithm aspart of management guidelines. 15 e 17 In February 2008a group of clinicians met in Liverpool to begin the devel-opment process for clinical care guidelines based arounda similar simple algorithm (Fig. 1. Algorithm for the man-agement of patients with suspected viral encephalitis,

Table 3 Non-viral causes of encephalitis and its mimicsmodied from (Solomon and Whitley 2004; Solomon2009). 80,110

Encephalitis Mimics

CNS InfectionsBacteria

Small bacteria (mostly intracellular)Mycoplasma

pneumoniaeMycobacterium tuberculosis

Chlamydophila Streptococcus pneumoniaeRickettsiae (including

scrub typhus, RockyMountain spotted fever)

Haemophilus inuenza

Ehrlichiosis(anaplasmosis)

Neisseria meningitidis

Coxiella burnetti(Q fever)

Bartonella hensellae(cat scratch fever)

Tropheryma whipplei(Whipple’s disease)

Brucella sp.(brucellosis)

Listeria monocytogenesSpirochetes

Trepenoma pallidum(Syphilis)

Leptospirosis

Borrelia burgdorferi(Lyme neuroborreliosis)

Borrelia recurrentis(relapsing fever)Other bacteria

Nocardiosis Infective endocarditisActinomycosis Parameningeal infection

Abscess/empyemaParasites

Trypanosoma brucei gambiense andTrypanosoma bruceirhodesiense(African sleeping sickness)

Malaria

Naegleria fowleri,Balamuthia mandrillaris(Amoebic encephalitis)

Cysticercosis

Angiostrongyluscantonensis (rat lungworm)

Trichinosis

FungiCoccidioidomycosis CryptococcosisHistoplasmosisNorth American

blastomycosisPara/post-infectious causes

InammatoryAcute disseminated

encephalomyelitis (ADEM)Acute haemorrhagic

leukoencephalopathy(AHLE)

Table 3 ( continued )

Encephalitis Mimics

Acute necrotisingencephalitis (ANE) inchildren

Bickerstaff’sencephalitis

Toxic/MetabolicReye’s syndrome

Systemic infectionSeptic encephalopathyShigellosis

Non-infectious causesVascular

VasculitisSystemic lupuserythematosisBehçet’s diseaseSubarachnoid & subduralhaemorrhageIschaemic cerebrovascular

accidentsNeoplastic

Paraneoplasticencephalitis

Primary brain tumourMetastases

Metabolic encephalopathyHepatic encephalopathyRenal encephalopathyHypoglycaemiaToxins (alcohol, drugs)Hashimoto’s diseaseSeptic encephalopathyMitochondrial diseases

OtherAntibody-mediated

encephalitis: VGKCcomplex or NMDAreceptor

Drug reactions

Encephalitis lethargicaHaemophagocytic

Lymphohistiocytosis (HLH)syndrome (usually

children)

Epilepsy

Functional disorder

In this table some of the important aetiologies are classiedinto whether they cause an encephalitis, with inammatorychanges seen histopathologically in the brain parenchyma, orencephalopathy without inammatory changes in the paren-chyma, although for some aetiologies this is based on limited

evidence.Abbreviations: VGKC, voltage-gated potassium channel; NMDA,N-methyl-D-Aspartic acid.

452 R. Kneen et al.


5/29

supported by an evidence base, whose implementation, itis hoped, would improve the management of patients withsuspected encephalitis. The scope of the guideline is tocover the initial management of all patients with suspected encephalitis, up to the point of diagnosis and earlytreatment, in an acute care setting such as acute medicalunit or emergency room. They are thus intended as a readyreference for clinicians encountering the more commoncauses of encephalitis, rather than specialists managingrarer causes. The guidelines also cover the specic treat-ments and further management of patients for whom a di-agnosis of viral encephalitis is made, particularly that dueto HSV, VZV and enteroviruses. Encephalitis due to CMV isalmost exclusively seen in the immunocompromised and isnot covered in detail; its diagnosis and management iscovered in HIV guidelines. 18 At the end of the guidelinesthe special circumstances of returned travellers, immuno-compromised patients and encephalitis associated withantibodies are discussed. Many patients with suspected vi-ral encephalitis ultimately prove to have another infec-tious or non-infectious cause for their illness. Thefurther management and treatment of such patients is be-yond the scope of this guideline, but we have includeda section on follow-up and support for patients with en-cephalitis in both the healthcare and voluntary sectors af-ter discharge from hospital. Finally, we have included

some suggestions for audit standards to assess practicebefore and after implementation of the guidelines.

Methods

A literature search was performed on the Medline databasefor the years 1998 e 2008, to identify for all (English lan-

guage) publications using the key words (‘Encephalitis’ AND:‘Symptoms’; ‘Signs’; ‘Management’; ‘Diagnosis’; ‘Investiga-tion’; ‘Lumbar Puncture’; ‘Cerebrospinal Fluid’ (CSF);‘Computed Tomography (CT)’; ‘MagneticResonanceImaging(MRI)’; ‘Single Photon Emission Tomogrophy (SPECT)’; ‘Elec-troencephalography (EEG)’; ‘Treatment’; ‘Antiviral’; ‘Aci-clovir’; ‘Steroids/Dexamethasone’) separately and incombination with the following MESH terms: (‘HerpesSimplex Virus’; ‘Varicella Zoster Virus’; ‘Enterovirus’; ‘Hu-man Immunodeciency Virus (HIV)’; ‘Immune compromise’;‘Arbovirus’. This yielded a total of 6948 citations, includingmany case reports, which were grouped together in subjectareas including clinical presentation, diagnosis, imaging,treatment, outcome, immune compromise. These groups of

papers were each screened by at least 2 of the group andscored for relevance, level of evidence and need for in-clusion. Further sources were added from review of thebibliographies of these articles, textbooks, other reviewsand personal collections of the screening group.

Using these revised source reference lists each sub-section of the manuscript was composed by two authors ofthe Guidelines Writing Group, from the elds of neurology,infectious diseases, microbiology, virology, acute medicineand the patient-sector. This included members from pro-fessional bodies including the British Infection Society (nowBritish Infection Association), the British Paediatric AllergyImmunology and Infection Group, the British PaediatricNeurology Association, the Society for Acute Medicine and

the Encephalitis Society. Each subsection was internallypeer-reviewed. The contributions from the various sectionsof the guidelines that people wrote were assimilated intoa single document in accordance with the principles of theAGREE (appraisal of guideline research and evaluation)collaboration. 19 In rating the strength of evidence wehave used the GRADE approach, in which the strength ofrecommendations is rated from A to D, and the quality ofthe evidence supporting the recommendation is ratedfrom I to III (Table 5. GRADE).20

This document has again been internally peer-reviewedtwice by the Guidelines Development Group, and updated toinclude further comments from all contributing authors,incorporating references published in 2009 e 11. The guide-

line has also been peer-reviewed by the wider GuidelinesStakeholder Group. This included members from the RoyalCollege of Paediatrics and Child Health, the PaediatricIntensive Care Society, the Children’s HIV Association andthe Meningitis Research Foundation. The guidelines arestructured to answer common clinical questions posed duringthe work-up of a patient with possible encephalitis.

Denition of childhood for this document

This guideline is for the management of suspected viralencephalitis in children aged older that 28 days (outside the

Table 4 Sub-acute and chronic central nervous systempresentations e microbiological causes, modied from(Solomon, Hart et al., 2007; Solomon 2009). 110,211

VirusesIn immunocompromised patients

Measles virus (inclusion body encephalitis)Varicella zoster virus (causes a multi-focalleukoencephalopathy)CytomegalovirusHerpes simplex virus (especially HSV-2)Human herpes virus 6EnterovirusesJC/BK a virus (progressive multi-focal

leukoencephalopathy)HIV (dementia)

In immunocompetent patientsJC/BK a virus (progressive multi-focalleukoencephalopathy)Measles virus (sub-acute sclerosing panencephalitis)

BacteriaMycobacterium tuberculosisTreponema pallidum (syphilis)Borrelia burgdorferi (Lyme neuroborreliosis)Tropheryma whipplei (Whipple’s Disease)

FungiCryptococcus neoformans

ParasitesTrypanosoma brucei spp. (African trypanosomiasis)Toxoplasma gondii (toxoplasmosis)

PrionsCreutzfeldt-Jakob disease

a JC and BK viruses are named after the initials of the patientsfrom whom they were rst isolated.



6/29

Figure 1 Algorithm for the management of patients with suspected viral encephalitis.

454 R. Kneen et al.


7/29

neonatal period) and younger than 16 years. The manage-ment of neonatal encephalitis (including premature in-fants) is outside the scope of this document. Nationalguidelines for the management of suspected viral enceph-alitis in adults are also available as a separate document(Solomon, Michael, et al. 2012).

Diagnosing encephalitis

Which clinical features should lead to a suspicion of encephalitis in children, how do they differ fromother encephalopathies, and can they be used todiagnose the underlying cause?

Recommendation

The constellation of a current or recent febrile illnesswith altered behaviour, personality, cognition or con-sciousness or new onset seizures or new focal neurolog-ical signs should raise the possibility of encephalitis, oranother CNS infection, and should trigger appropriateinvestigations (A, II)

The differential diagnosis of encephalopathy (due tometabolic, toxic, autoimmune causes or sepsis outsidethe CNS) should be considered early (B, III), especiallyif there are features suggestive of a non-encephaliticprocess, such as a past history of similar episodes, sym-metrical neurological ndings, myoclonus, clinical signsof liver failure, a lack of fever, acidosis or alkalosis (B, III)

Patients presenting with a sub-acute (weeks to months)encephalitis should trigger a search for autoimmune,paraneoplastic, metabolic aetiologies (C, III)

The priority of the investigations shown in Table 9 is de-termined by the patient’s clinical history and clinicalpresentation (C, III)

EvidenceThe differential diagnosis of acute encephalitis in childhoodis broad encompassing infectious, para-infectious immune-mediated, autoimmune, metabolic, vascular, neoplastic,paraneoplastic, and toxic aetiologies as well as braindysfunction due to systemic sepsis (Tables 2and 3). 21,22

Nevertheless, dening the clinical features that shouldprompt the suspicion of acute encephalitis of childhood isessential in order to achieve prompt recognition,

investigation and management because delays have beenshown to impair outcome.

However, differentiating infection-associated encepha-litis from the other causes of encephalopathy on the basisof clinical ndings poses a signicant diagnostic challenge,especially in children in whom the clinical picture can bevague. For example, of Chaudhuri and Kennedy’s list‘useful clinical pointers to aid exclusion of non-infectivecauses of encephalopathy’, none are absolute. 23 In adults,fever and abnormal mental status, often with severe head-ache, nausea and vomiting, are the classical clinical fea-tures of infective encephalitis. Eighty-ve (91%) of 93adults with HSV-1 encephalitis in one study were febrileon admission 8; even those not febrile on admission will of-ten have a history of febrile illness (Table 6. History). Dis-orientation (76%), speech disturbances (59%) andbehavioural changes (41%) were the most common fea-tures, and one third of patients had seizures. 8 Howevera normal Glasgow coma score at presentation was seen insome patients in this, and other studies, reecting thefact that it is a crude tool for detecting subtle changes inbehaviour. 13 Alterations in higher mental function includelethargy, drowsiness, confusion, disorientation and coma(Table 7. Examination).

Denition of the spectrum of clinical ndings at pre-sentation, and the pattern of subsequent manifestations inchildren is more difcult as documentation of the clinicalpresentation of children with encephalitis is less welldescribed. Several studies include adults and childrentogether making it difcult to comment on whetherchildren may have a different presentation. 24 e 26 Further-more, given that children are susceptible to different aetio-logical agents than adults and also present differently fromadults with other causes of infection, the utility of the pro-les dened in characterising encephalitis presentations inchildhood is limited by the lack of breakdown by age. Themost relevant studies have only reported ndings in small

Table 5 GRADE rating system for the strength of theguidelines recommendations and the quality of the evi-dence (Atkins, Best et al., 2004). 20

Strength of therecommendation

Quality of the evidence

A Strongly recommended I Evidence fromrandomised controlled trials

B Recommended,but other alternativesmay be acceptable

II Evidence fromnon-randomised studies

C Weakly recommended:seek alternatives

III Expert opinion only

D Never recommended

Table 6 Questions to consider in the history when assess-ing a patient with suspected encephalitis, modied from(Solomon, Hart et al., 2007). 110

Current or recent febrile or inuenza-like illness? Altered behaviour or cognition, personality change oraltered consciousness?

New onset seizures? Focal neurological symptoms? Rash? (e.g. varicella zoster, roseola, enterovirus Others in the family, neighbourhood ill? (e.g. measles,

mumps, inuenza)Travel history? (e.g. prophylaxis and exposure for malaria,arboviral encephalitis, rabies, trypanosomiasis)

Recent vaccination? (e.g. ADEM) Contact with animals? (e.g. rabies) Contact with fresh water (e.g. leptospirosis)Exposure to mosquito or tick bites (e.g. arboviruses, Lymedisease, tick-borne encephalitis)

Known immunocompromise? HIV risk factors?

Abbreviations: ADEM Acute disseminated encephalomyelitis;HIV Human immunodeciency virus.



8/29

numbers of children and the entry criteria were not provenencephalitis, but suspected encephalitis and are primarilyhospital-based. 21,27 In the Toronto Acute Childhood En-cephalitis study, 50 children with suspected encephalitiswere reported with the most common presenting featuresbeing fever (80%), seizures (78%), focal neurological signs(78%) and decreased consciousness (47%). 27 In Wang’s studyfrom Taiwan, 101 children with a nal diagnosis of enceph-alitis were reported to have the following features; changein personality or reduction in consciousness (40%), seizures(33%), new neurological signs (36%) and meningism 22%, Thenumber presenting with fever was not reported. 28 In themore recent Liverpool study, 51 children were treated forsuspected encephalitis and their most common presentingfeatures included confusion, irritability or a behaviourchange (76%), fever (67%), seizures (61%), vomiting (57%)and focal neurological signs (37%). 14 However, 14 of thesechildren were ultimately not felt to have viral encephalitisand should not have received aciclovir so the list of pre-senting symptoms is likely to be inaccurate. Ill childrenare different to adults, young children cannot often ade-quately describe symptoms such as headache and infantsfrequently have non-specic symptoms and signs for manyacute illnesses including feeding and respiratory difcul-ties. In Wang’s study 54% of the children had concomitantsigns of a respiratory infection and 21% had gastrointestinalsymptoms. 28

With the advent of CSF PCR more subtle presentations ofHSV encephalitis have been recognised and described inadults and children. 30 These include low-grade pyrexiarather than a high fever, speech disturbances (dysphasiaand aphasia), and behavioural changes which can mistakenfor psychiatric illness, or the consequences of drugs or alco-hol, occasionally with tragic consequences. 31

Distinguishing HSV encephalitis from otherencephalopathies

Several studies have documented the potential mimics ofHSV encephalitis in adults and children. 12 e 14,32,33 Whitley

et al. demonstrated that of 432 (168 < 18 years old)patients undergoing brain biopsy for presumed HSV en-cephalitis: 195 (45%) had the diagnosis proven histologi-cally and in a further 95 patients (22%) an alternative,often treatable, diagnosis was established. 33 However,the clinical presenting features of these two groups werevery similar. Chataway et al. found that, of those patientsinitially considered to have HSV encephalitis, inamma-tory aetiologies such as ADEM or multiple sclerosis werethe most frequent mimics. 32 Some of the rarer paediatricdiagnoses included epileptic encephalopathies such asRasmussen’s encephalitis and Alper’s syndrome. Kneen etal. showed the broad range of nal diagnoses in childreninitially treated with aciclovir for possible HSV encephali-tis in children and Bell et al. and Michael et al. demon-strated that this was also similar in adult practice. 12 e 14

The clinical picture clearly varies with disease severitybut can also vary with aetiological agent; of the many vi-ruses that cause acute encephalitis in children, somehave a predilection for localised parts of the brain whichcan determine the initial clinical picture and the subse-quent clinical course. 34 Although, De Tiege et al. 35 en-dorse the view that the concept of a ‘‘classical’’ pictureof HSV encephalitis in children is now out-dated and re-mind clinicians that the most common reason for failureto diagnose HSV encephalitis is non-specic initial clinicalpresenting symptoms and signs. 7

Seizures are more frequently found in patients present-ing with encephalitic processes affecting the cortex, whichare more often infectious in aetiology, as opposed toencephalitic processes predominantly affecting subcorticalwhite matter that more frequently have an immune-mediated pathogenesis (e.g. ADEM). However, seizuresand movement disorders are also often seen in childrenwith encephalitis due to autoimmune antibody-mediateddisease (see ‘Special circumstances’ section). Seizures canalso be subtle and include subtle motor status: a syndromeof subtle continuous motor seizure activity. This oftenfollows overt convulsive seizures or status epilepticus ornon-convulsive status epilepticus (NCSE): a syndrome ofencephalopathy with no overt motor seizure activity but anelectrical seizure correlate on the EEG. A study of 144 (134children) patients with encephalitis due to Japaneseencephalitis virus found that 40 had witnessed seizures inhospital. Of these, 25 had one or more episodes of statusepilepticus including 15 who went onto develop subtlemotor status. Patients with witnessed convulsive or subtlemotor status epilepticus were more likely to die( p Z 0.0003). 36 However, it is very unusual for patientswith encephalitis or other CNS infections and encephalopa-thy to present with de novo NCSE. A study of 236 consecu-tive intensive care unit patients (11% < 16 years) during therst 3 days of an illness with coma (and no witnessed overtor subtle seizures) identied that 19 (8%) were in NCSE. Ofthese, 2 were children. Only one adult had a CNS infection(diagnosis unspecied). 37 In another study of 45 consecu-tive adults diagnosed with NCSE, 20 had no previous diagno-sis of epilepsy. Twenty-eight of the 45 patients hada remote risk factor for developing epilepsy includingprevious CNS infections in some (number not specied). 38

Despite its relative rarity, NCSE can only be diagnosedwith an EEG and as there are specic treatments available,

Table 7 Examination ndings of importance in assessinga patient with suspected encephalitis modied from (Solo-mon, Hart et al., 2007). 110

Airways, Breathing, Circulation Mini-mental state, cognitive function, behaviour (whenpossible)

Evidence of prior seizures (tongue biting, injury) Subtle motor seizures (mouth, digit, eyelid twitching) Meningism Focal neurological signs Papilloedema Flaccid paralysis (anterior horn cell involvement) Rash (purpuric e meningococcus; vesicular e hand footand mouth disease; varicella zoster; rickettsial disease)

Injection sites of drug abuse Bites from animals (rabies) or insects (arboviruses) Movement disorders, including Parkinsonism

456 R. Kneen et al.


9/29

an EEG should be considered in all patients with undiag-nosed encephalopathy. 39

In adult clinical practice the most frequently encoun-tered infection-associated encephalopathy is septic en-cephalopathy, being found in 50 e 70% of septic patients. 40

This syndrome usually occurs in elderly patients with anextracranial focus of sepsis where the encephalopathycannot be attributed to other organ dysfunction. Clinically,the diagnosis is one of exclusion. The syndrome is charac-terised neurologically by progression from a slowing ofmentation and impaired attention to delirium and coma.Neurological examination ndings are usually symmetrical.This syndrome is uncommon in paediatric practice but itcan occur and is most frequently seen in association withbacterial infections of the urinary tract. It can also beseen in association with other rarer infective encephalop-athies such as shigella or in association with typhoidfever. 41,42

Diagnostic features for specic aetiologiesThe history is important in dening the spectrum ofagents potentially responsible for encephalitis as this isinuenced by age, immunocompetence, geography andexposure. Geographical restrictions are laid out in theTable 2. These are particularly signicant for arthropod-borne infections.

As indicated above the features for HSV are non-specic:many patients with suspected HSV encephalitis ultimatelyprove to have a different diagnosis. In adults, the nding oflabial herpes (cold sores) has no diagnostic specicity forHSV encephalitis and is merely a marker of critical illness.However, in children who are more likely to developencephalitis with a primary HSV infection, labial herpesmay be noted. 43,44 Lahat reported 2 children with recent la-bial herpes in a series of 28 children aged from 3 months to16 years with proven HSV encephalitis due to a primary in-fection 43 and Elbers reported active or a recent history oflabial herpes in 4 out of 16 children with proven HSV en-cephalitis. 44 Elbers also reported that 3 further childrenwith positive CSF PCR for HSV-1 were excluded from his se-ries because of an atypical presentation. These children allhad a milder illness (all had fever, one had multiple sei-zures, one had a single seizure and ataxia and one had leth-argy and headache) and normal cranial imaging. Elbersconcluded that the CSF PCR results may be false positivesor due to reactivation of the virus but it is also conceivablethat HSV can cause a mild encephalitis and for this reason itshould be considered in the differential diagnosis of chil-dren with less severe symptoms. A mild HSV encephalitishas also been reported in 2 previous children aged 3.5 and15 years who recovered without treatment with aciclovir. 31

Children with HSV encephalitis may also present with anacute opercular syndrome (disturbance of voluntary controlof the facio-linguo-glosso-pharyngeal muscles leading tooro-facial palsy, dysarthria and dysphagia). 45,46 CNS diseasecaused by HSV-2 is rare outside the neonatal period. Themost common manifestation in adults is aseptic meningitiswhich may be recurrent. 47,48 This has also been reportedin children and the possibility of sexual abuse may needto be considered. 49

Varicella zoster virus (VZV) can cause central nervoussystem manifestations through a post-infective immune-

mediated cerebellitis, an acute infective viral encephalitisor a vasculopathy; the neurological presentation may bepreceded by the vesicular rash of by days or weeks, thoughit occasionally occurs before the rash or even in patientswith no rash. 50 e 52 Encephalitis is more common in adults,especially those with cranial dermatome involvement ora disseminated rash or the immunocompromised. The pre-sentation may be acute or sub-acute with fever, headache,altered consciousness, ataxia and seizures. A more commonneurological presentation associated with VZV infection inchildren is post-infectious cerebellitis particularly in youngchildren. This is usually a relatively mild and self limitingdisorder but children can become unwell due to hydroceph-alus secondary to swelling of the cerebellum in more severecases. 53,54 Children usually present with a short history ofunsteadiness or limb ataxia and nystagmus. The other rela-tively common association in childhood is between VZV in-fection and arterial ischaemic stroke, and is thought toaccount for up to one third of cases of arterial stokes inpaediatric practice. The majority present with acute butpermanent hemiparesis, acute chorea or facial weaknesswhich is commonly transient; 55 seizures and visual orspeech disturbances also occur. Patients usually present af-ter the rash has cleared and the time period can be very de-layed with a mean of 3 months (range 1 week to 48 months)reported in a recent London study. 55 However, early mani-festations can occur within days of exposure, 50 well beforethe vesicular eruption, which may be uncharacteristicallymild, 52 making diagnosis more challenging, especially as on-set of encephalitic features can be abrupt or gradual. 51 PCRfor VZV DNA in the CSF is positive in around a third of pa-tients. A more sensitive test (positive in over 90% patients)is measuring VZV specic IgG antibodies in CSF. The levelscan be compared to a concomitant serum sample as a re-duced serum/CSF ratio of VZV IgG conrms intrathecalsynthesis. 56

Epstein e Barr virus (EBV) encephalitis most commonlyaffects teenagers (median age 13 years; but generallypresents in the absence of signs of the typical mono-nucleosis clinical picture. 57 In Doja’s series of 21 patients,17 had a non-specic prodrome of fever and 14 had head-ache. Manifestations of EBV encephalitis and encephalomy-elitis may also include an altered level of consciousness,seizures and visual hallucinations. 57 e 59 However, the tem-poral relationship between symptoms is highly variable, in-cluding CNS disease as the presenting manifestation,making aetiological diagnosis difcult on clinical groundsand highlighting the need to consider EBV in all cases ofchildhood encephalitis irrespective of symptoms. 57

Encephalitis may be associated with respiratory illnessesin children: most common pathogens include the inuenzaviruses, paramyxoviruses and the bacterium M. pneumo-niae . There may be no preceding respiratory symptomsprior to the development of encephalitis in a signicantproportion of patients. 60,61 In a recent study of patientswith M. pneumonia encephalitis, the affected childrenwere an older cohort (median age 11 years old), presentingwith a short prodrome of fever (70%), lethargy (68%), andaltered consciousness (58%), while gastrointestinal (45%)and respiratory (44%) symptoms were less common. 60 Theirclinical course progressed rapidly (median 2 days from on-set to hospitalization), and commonly required intensive



10/29

care (55%). Seizures were less common in the clinical pic-ture. Symptoms of progressive symmetrical external op-thalmoplegia typify Bickerstaff brainstem encephalitis inassociation with M. pneumonia and can serve as a clue todiagnosis especially when accompanied by ataxia. 62 Inu-enza has been reported to be associated with a spectrumof neurological disorders in adults and children rangingthrough a mild encephalopathy with seizures, encephalitis,ADEM, encephalopathy with posterior reversible encepha-lopathy syndrome, malignant brain oedema syndrome andacute necrotising encephalopathy (ANE). 65,66 Patientswith inuenza (particularly inuenza B) can also have as-sociated severe myositis. 63,64 Patients with inuenza en-cephalopathy/encephalitis rarely have viral antigens orviral nucleic acid in CSF or neural tissue and the mecha-nisms for causing neurological illness are still unclear. In-uenza A in particular, has been reported in associationwith ANE, a severe encephalopathy often associated withfever and in which typical MRI abnormalities have been re-ported in the thalami, brainstem and cerebral whitematter. 65 e 67 ANE has most frequently been reported inyoung children in small outbreaks in Japan and otherSoutheast Asian countries. This disorder has been foundto have an autosomal dominant inheritance pattern insome families with genetic mutations identied. 68 Thereis some very recent evidence that the H1N1 strain of Inu-enza A that emerged in 2009 may cause more neurologicalmanifestations than seasonal u. Ekstrand reported 18children with H1N1and compared them to 16 with seasonalu. Children with the H1N1 strain were more likely to haveencephalopathy, focal neurological signs, aphasia and anabnormal EEG. 69

Encephalitis associated with gastrointestinal symptomsincludes infection with enteroviruses, rotavirus and humanparechoviuses. Enteroviral encephalitis can be associatedwith a brainstem syndrome. Large outbreaks of encephalitishave been reported with enterovirus 71 in Bulgaria 1975,Hungary 1997, Malaysia 1997 and Taiwan 1997. Childrenunder 5 are more commonly affected 70 and the highestmortality is in those aged 6 e 12 months. 71,72 Clues to infec-tion with this virus include the typical papular lesions onthe hands, feet and in the mouth but those with encepha-litis often develop neurogenic pulmonary oedema 73 onday 2e 3 of illness which can rapidly progress to fatal car-diorespiratory collapse despite intervention. 71 Rotavirusencephalopathy has been reported to cause convulsionsand cerebellar signs in some children. 74,75

Rashes may be seen in other encephalitides; for examplea maculopapular or vesicular rash is seen in Rickettsialinfections or the highly typical rash of measles virusinfection. Measles can cause three separate encephaliticillnesses and is of particular concern given the recent rise incases reported in children and young adults across Europe.The rst is either an acute encephalitis or acute dissemi-nated encephalomyelitis associated with the acute infec-tion, although patients may present without the typicalrash. 76 The second is a sub-acute encephalopathy aroundsix months after the primary infection in the immunocom-promised with measles inclusion bodies in the brain oftenwithout a rash. The third is sub-acute sclerosing panence-phalitis (SSPE) in the immunologically normal which can oc-cur several years after the primary infection. Patients with

the sub-acute forms usually present with a dementia, visualproblems and later with seizures. 77

HHV6 (and possibly HHV7) is a cause of encephalitiscausing severe disease and long-term sequelae far beyondself-resolving febrile convulsions. 78 Typical below age 2years 79 ataxia and prolonged convulsions are the majorneurological manifestations and gastrointestinal symptomscan accompany the high fever and rash systemically, thuscan be indistinguishable from the viral encephalitides typi-ed by gastroenteritis.

Sometimes the pattern of neurological decit can bea clue as to the possible aetiology. Thus autonomicdysfunction, myoclonus and cranial neuropathies can in-dicate brainstem encephalitis, which is seen in listeriosis,brucellosis, some viral infections or rarely tuberculosis(Table 8. Brainstem encephalitis); there may be tremorsand other movement disorders if the thalamus and otherbasal ganglia are involved, as seen in aviviruses, such asWest Nile virus and Japanese encephalitis, and alphavirusessuch as Eastern equine encephalitis virus. 80,81 An encepha-litis with an acute accid paralysis is characteristic of polio,and other enterovirsues, such as enterovirus 71, as well asaviviruses. 82

Which patients with suspected encephalitis shouldhave a lumbar puncture (LP), and in which shouldthis be preceded by a computed tomography (CT)scan?

Recommendation

All patients with suspected encephalitis should havea lumbar puncture as soon as possible after hospital

Table 8 Brainstem encephalitis (rhombencephalitis) -clues and causes, from (Solomon, Hart et al., 2007). 110

Suggestive clinical features Lower cranial nerve involvement Myoclonus Respiratory drive disturbance Autonomic dysfunction Locked-in syndrome MRI changes in the brainstem, with gadoliniumenhancement of basal meninges

Causes Enteroviruses (especially EV-71)Flaviviruses, e.g. West Nile virus, Japanese encephalitisvirus

Alphaviruses, e.g. Eastern equine encephalitis virus Rabies Listeriosis Brucellosis Lyme borreliosis Tuberculosis Toxoplasmosis Primary or secondary central nervous systemmalignancy

Paraneoplastic syndromes

458 R. Kneen et al.


11/29

admission, unless there is a clinical contraindication(Table 10. Contraindications to immediate lumbarpuncture) (A, II)

Clinical assessment and not cranial CTshould be used todetermine if it is safe to perform a LP (A, II)

If there is a clinical contraindication indicating possibleraised intracranial pressure due to or causing brainshift, a CT scan should be performed as soon as possi-ble, (A, II). An immediate LP following this should ide-ally be considered on a case by case basis, unless theimaging reveals signicant brain shift or tight basal cis-terns due to or causing raised ICP, or an alternative di-agnosis, or the child’s clinical condition changes (B, III).

If an immediate CT is not indicated, imaging (CT or,preferably, MRI) should be performed as soon as possi-ble after the LP (A, II)

In anticoagulated patients, adequate reversal (withprotamine for those on heparin and vitamin K,

prothrombin complex concentrate, or fresh frozenplasma for those on warfarin) is mandatory before lum-bar puncture (A, II). In patients with bleeding disorders,replacement therapy is indicated (B, II). If unclear howto proceed, advice should be sought from a haematolo-gist (B, III)

In situations where an LP is not possible at rst, the sit-uation should be reviewed every 24 hours, and an LPperformed when it is safe to do so (B, II)

If an initial LP is non-diagnostic, a second LP should beperformed 24-48 hours later (B, II)

Children and young adults should be stabilisedbefore performing a CT scan, and an anaesthetist,paediatrician or intensivist should be consulted.(A, III)

Lumbar punctures should be performed with needlesthat meet the standards set out by the National PatientSafety Agency (A, III)

Table 9 Additional investigations to consider in the differential diagnosis of encephalitis.

Differential diagnosis Investigations to consider

Para-infectious immune-mediatedencephalitis

MRI brain and spineAntiDNAse B and ASO titre, inuenza A and B PCR and/or

antibody in CSF and serumCSF examinationBrain and meningeal biopsy

Autoimmune/Inammatoryencephalitis

FBC, ESR, CRP, ANA, ENA, dsDNA, ANCA, C3, C4, lupusanticoagulant, cardiolipin, thyroglobulin, thyroperoxidaseantibodies, ferritin, brinogen, trigylcerides.Voltage-gated potassium channel complex and NMDA receptor

antibodiesSerum and CSF ACE, Serum 25OH Vitamin D, 24hr urinary calciumWhole body CTBiopsy: Brain, meninges, skin, lymph node, peripheral nerve/muscle

Metabolic Renal, liver, bone & thyroid prolesArterial blood gas analysisPlasma and CSF lactate, ammonia, pyruvate, amino acids, very

long-chain fatty acids, urinary organic acidsPorphyrins: blood/urine/faecesBiopsy: skin, lymph node, peripheral nerve/muscle

Vascular CT or MRI head with venogram and/or angiogramNeoplastic MRI brain and MR spectroscopy

CSF cytological analysisBrain and meningeal biopsyCT chest/abdomen/pelvisLDH, IgG/A/M, protein electrophoresis, urinary Bence-Jones protein

(in adults), bone marrow trephineParaneoplastic Anti-neuronal and onconeuronal antibodies

CT or PET chest, abdomen and pelvisBiopsy of non CNS visceraAlpha fetoprotein, beta human chorionic gonadatrophin

Toxic Blood lm; blood or urine levels of alcohol, paracetamol, salicylate,tricyclic, heavy metals

Urinary illicit drug screenSeptic Encephalopathy Serum microbiological cultures, serology and PCR

Abbreviations: MRI magnetic resonance imaging; ASO antistreptolysin; PCR polymerase chain reaction; CSF cerebrospinal uid; FBC fullblood count; ESR erythrocyte sedimentation rate; CRP C-reactive protein; ANA antinuclear antibodies; ENA extraneuclear antibodies;dsDNA double stranded deoxyribonucleic acid antibodies; C3/4 complement; ACE angiotensin converting enzyme; CT computed tomog-raphy; LDH lactate dehydrogenase; IgG/M/A immunoglobulin; PET positron emission tomography; CNS central nervous system.



12/29

EvidenceA lumbar puncture (LP) is an essential investigation inthe management of children with suspected encephalitisto conrm the diagnosis and rule out other causes.Therefore all children with suspected encephalitis shouldhave a LP unless a specic contraindication exists.Contraindications have been published in an evidencedbased guideline for the management of decreased levelconscious level in children (Table 10. Contraindicationsto immediate lumbar puncture). 83 The risk of presentingwith viral encephalitis following a febrile seizure is notknown although the risk for bacterial meningitis follow-ing febrile seizures is quoted to occur in between 0.4and 5%.84,85 This gure increases to 18% in children whopresent with febrile status epilepticus. 86 In addition,young children may present with meningitis without anysigns of meningeal irritation; 6 children aged less than18 months in a series of 95 children who presented witha simple ( n Z 87) or complex febrile seizure ( n Z 8)and without signs of meningeal irritation had underlyingbacterial meningitis. 87 Although others have reportedthat only 0.4 e 1.2% of children who present with a fever

and a seizure, in the absence of signs of meningeal irrita-tion will have bacterial meningitis. 85 Nevertheless, chil-dren who do not make a full recovery within an hourfollowing a typical, simple febrile convulsion shouldhave a LP. 88

There has been considerable controversy over the role ofcomputer tomography (CT) and LP in patients with sus-pected central nervous system infection, in particularwhether a CT is needed before an LP. 14,90,91 Although thereare few studies that specically address this issue in pa-tients with suspected encephalitis, much of the literatureabout suspected bacterial meningitis is pertinent becauseof overlap in the clinical presentations. As in patientswith meningitis, in encephalitis the CT scan is not a reliabletool for the diagnosis of raised intracranial pressure, andshould not be used to for this. 14,16

In patients with suspected encephalitis, an early CTscan has two roles: suggesting the diagnosis of viralencephalitis and indicating an alternative diagnosis. Aninitial CT scan soon after admission will show a suggestiveabnormality in about 80% of patients with herpes simplexvirus (HSV) encephalitis; 8 almost all those with HSV en-cephalitis, and a negative initial scan will have abnormal-ities on a second scan. 8 The sensitivity of the CT scan fordetecting abnormalities is increased with the use of intra-venous contrast media, however, it is not, on its own,diagnostic.

The second role of an early CT scan is suggesting analternative diagnoses, so that LP may no longer be neces-sary. For example in one study of 21 adults with suspectedencephalitis, 2 (10%) patients did not have a LP after the CTscan showeda stroke; 12 however, ina larger study only 2 (1%)of 153 patients with suspected CNS infections who had a CTrst did not subsequently need a LP. 13

If clinical contraindications to an immediate LP arepresent then an urgent CT should be performed. If thisidenties shift of brain compartments or tight basalcisterns, due to mass lesions and/or oedema a subsequentLP may be dangerous. In patients with brain shift, a LP, byreducing the cerebrospinal uid (CSF) pressure below thelesion, may precipitate herniation of the brainstem orcerebellar tonsils. 92,93 For example this may occur in pa-tients with brain abscess, subdural empyema, tumour, ora necrotic swollen lobe in HSV encephalitis. However unse-lected CT scanning all patients before a LP can cause un-necessary delays in many patients, in whom there wereno contraindications to an immediate LP. 12,13 For examplein one recent study of 21 adults with suspected encephali-tis, 17 had a LP, which was delayed for a CT scan in 15,though only one of them had any contraindications to animmediate LP. The median time to CT scan was 6 h, butthe median time to LP was 24 h. 12 Furthermore, in a largerstudy of 217 patients with suspected CNS infections the me-dian (range) time to LP was signicantly longer if the pa-tient had a CT scan rst (18.5 [2 e 384] versus 6 [1e 72]hours respectively, p < 0.0001). 13 The increasing availabil-ity of CT scans in emergency units since this work was pub-lished (due to implementation of national guidelines forstroke thrombolysis and acute head injury) will undoubt-edly result in easier access to imaging for children with sus-pected encephalitis who are managed in a hospital that alsotreats adult emergency patients.

Table 10 Contraindications to an immediate lumbarpuncture in patients with suspected CNS infections, modi-ed from (Kneen, Solomon, et al., 2002; Michael, Sidhu,et al ., 2010; Hasbun, Abrahams, et al., 2002;NICE).13,16,21,92,110

Imaging needed before lumbar puncture (to exclude brainshift, swelling, or space occupying lesion )

Moderate to severe impairment of consciousness(GCS < 13)a or fall in GCS of > 2

Focal neurological signs (including unequal, dilated orpoorly responsive pupils)

Abnormal posture or posturing Papilloedema After seizures until stabilised Relative bradycardia with hypertension Abnormal ‘doll’s eye’ movements Immunocompromise

Other contraindications Systemic shock Coagulation abnormalities:

B Coagulation results (if obtained) outside the normalrange

B Platelet count < 100 109/LB Anticoagulant therapy

Local infection at the lumbar puncture site Respiratory insufciency Suspected meningococcal septicaemia (extensive orspreading purpura)

a There is no agreement on the depth of coma that necessitatesimaging before lumbar puncture; some argue Glasgow comascore < 12, others Glasgow coma score < 9.

Patients on warfarin should be treated with heparin instead,and this stopped before lumbar puncture.

Consider imaging before lumbar puncture in patients withknown severe immunocompromise (e.g. advanced HIV).A lumbar puncture may still be possible if the platelet count is50 109/L; Seek haematological advice.

460 R. Kneen et al.


13/29

A series of studies have examined which clinical signscan be used to determine which patients with suspectedbacterial meningitis need a CT scan before LP. 90,92,94 In onestudy of 696 episodes of community acquired acute bacte-rial meningitis it was concluded that CT scan should pre-cede LP in patients with new onset seizures, focalneurological signs, excluding cranial neuropathies, or mod-erate to severe impairment of consciousness, as indicatedby a Glasgow coma score of 10 or less. 94 Papilloedema, a di-rect indicator of raised intracranial pressure, is also an in-dication for imaging before a LP. NICE guidelines for therole of CT and LP in children with suspected bacterial men-ingitis have also been produced recently. 16 Given the po-tential overlap of clinical features in patients withsuspected meningitis and suspected encephalitis, this ap-proach is also applied to patients with suspected encepha-litis. Although there is good agreement about most of theindications for a CT scan before a LP, there is disagreementabout the precise level of consciousness that should betaken as a contraindication to an immediate LP. Amongseven commentaries reviewed by Joffe 2007, 92 three sug-gested “increasing stupor progressing to coma”, three sug-gested a “deterioration in consciousness level”, twosuggested a GCS < 8, and one a GCS < 13. 14,95 e 100 The re-cent NICE guidelines for bacterial meningitis recommenda CT scan before a LP if the GCS is < 9 or uctuating > 2,so that an alternative diagnosis can be excluded. 16 Thereis also lack of clarity about whether in such patients, a LPshould then be performed at all, if the scan is normal, orwhether a low coma score is an absolute contraindication,whatever the scan shows. Occasionally deterioration afterLP has been reported in patients with bacterial meningitisand an apparently normal CT; but we are not aware of sim-ilar cases in patients with viral encephalitis; and mostwould argue that the information from the LP is essentialto make a diagnosis and guide treatment. In one retrospec-tive series of 222 adults with suspected encephalitis lessthan 5% of patients had imaging changes suggestive ofraised intracranial pressure. 32

Other contraindications to LPincludelocal skininfectionatthe site of puncture, a clinically unstable patient withcirculatory shock or respiratory insufciency, and any clinicalsuspicion of spinal cord compression. LP may also be harmfulin patients with coagulopathy, because of the chance ofneedle-induced subarachnoid haemorrhage or of the devel-opment of spinal subdural and epidural haematomas. Thestandard recommendation is to perform a lumbar punctureonly when the patient does not have a coagulopathy and hasa platelet count of 100 109/L or greater, although plateletcounts of 20 109/L or greater have also been recommen-ded. 16,101 A rapidly falling count is also a contraindication.Haemorrhage can occur in patients anticoagulated with hep-arin or warfarin, but inone large study, preoperative antipla-telet therapy with aspirin or nonsteroidal anti-inammatorymedications and subcutaneous heparin on the operative daywere not risk factors for spinal haematoma inpatients under-going spinal or epidural anaesthesia. 101

In summary, many children will need a CT before a LP,because of their clinical contraindications to an immediateLP; such patients should have a CT, and then ideally a LPshould be considered on a case by case basis (if stillindicated and no radiological contraindications are

identied) within 6 h and then decisions made on antiviraltreatment based on these results. In some children who donot have a clinical contraindication to immediate LP, and inwhom CT is not immediately available, a prompt LP may bethe most useful approach to get an early diagnosis.

Lumbar punctures should be performed with needlesthat meet the standards set out by the National PatientSafety Agency.

What information should be gathered from the LP?

Recommendations

CSF investigations should include:- Opening pressure when possible (A, II)- Total and differential white cell count, red cell count,

microscopy, culture and sensitivities for bacteria(A, II)

- If necessary, the white cell count and proteinshould be corrected for a bloody tap

- Protein, lactate and glucose, which should be com-pared with a plasma glucose taken just before theLP (A, II)

- A sample should be sent and stored for virological in-vestigations or other future investigation as indicatedin the next section (A, II)

- Culture for Mycobacterium tuberculosis when clini-cally indicated (A, II)

If there is a strong clinical diagnosis of encephalitis ina child, but the initial CSF results are normal, a secondLP should be undertaken and all CSF tests repeated, in-cluding consideration for antibody detection (A, II)

EvidencePatients with HSV encephalitis typically have moderateelevation of CSF opening pressure, a moderate CSFpleocytosis from tens to hundreds of cells 106 /L,a mildly elevated CSF protein and normal CSF to plasmaglucose ratio. 7,8,102 Whilst measuring the CSF openingpressure is part of a standard LP, it is often more difcultto achieve this in children and is frequently omitted.Whilst the evidence base would recommend undertakingopening pressure, it is recognised that it is often imprac-tical to do this in a child so the clinician responsible forundertaking the LP will have to make a balanced judge-ment as to the value of trying to achieve this. However,if the child is being anaesthetised for the procedure, theopening pressure should be measured with arterial bloodgas results stabilised and CO 2 noted. Occasionally poly-morphonucelar cells predominate, or the CSF may be nor-mal, especially early in the illness: in approximately 3 e 5%of adults with proven HSV encephalitis an initial CSF maybe normal with no pleocytosis and a negative HSV PCR. 8,10

The gure is even higher in patients with immunocompro-mise and in children, especially infants. However, if therst CSF is normal in HSV encephalitis, a second CSF exam-ination 24 e 48 h is likely to be abnormal with a positiveHSV PCR, although the viral load does reduce when pa-tients are receiving aciclovir. 8,10

A series of studies have shown the apparent difculty inmeasuring plasma glucose at the same time as CSF



14/29

glucose, 12 e 14 but without the former, interpretation ofthe CSF results is very difcult. HSV encephalitis can behaemorrhagic, and the CSF red cell count is elevated inapproximately 50% of cases. 103 An acellular CSF is also de-scribed for other viruses, VZV, EBV and CMV, and occursmore frequently in patients with immunocompromise. 25

Although a lymphocytic CSF pleocytosis is typical ofviral CNS infections, non-viral infection, particularly tu-berculosis and listeriosis, and partially treated acutebacterial meningitis can give a similar picture. Usuallythe clinical setting and other CSF parameters (low glucoseratio and higher protein) will suggest these possibilities.CSF lactate may be helpful in distinguishing bacterialmeningitis from viral CNS infections 104,105 ; in particulara CSF lactate of < 2 mmol/l is said to rule out bacterial dis-ease. 105 A high CSF lactate may also be an indicator ofa metabolic disorder, in particular a mitochondrial en-cephalopathy. If ADEM is in the differential diagnosis,the CSF (with a paired serum sample) should be sent foroligoclonal bands. 106

In a traumatic tap white blood cells and protein fromthe blood can contaminate the cerebrospinal uid. 107

There is little good evidence to support how to correctfor this in children. 108 However, the standard approxima-tion would be to subtract 1 white cell for every 700 redblood cells 106 /L in the CSF and 0.1 g/dl for every 100red blood cells. 109 This approximation will sufce in mostcircumstances, though more complicated formulae allow-ing for anaemia etc are available. 107 However, in patientswith HSV encephalitis a blood-stained CSF sample may re-ect the haemorrhagic pathophysiology of the condition,this is more likely if serial CSF specimens are blood-stained.

What virological investigations should be

performed ?

Recommendation

All patients with suspected encephalitis should havea CSF PCR test for HSV (1 and 2), VZV and enterovirusesas this will identify 90% of known viral cases and EBVconsidered (B, II)

Further testing should be directed towards specicpathogens as guided by the clinical features such astravel history and animal or insect contact (B, III)

EvidenceAlthough the list of viral causes of encephalitis is long, 110

HSV types1 & 2, VZV, and enteroviruses are the commonestcauses of viral encephalitis in immunocompetent individ-uals in Europe & the United States. 24,111 Our ability to diag-nose encephalitis caused by herpes viruses & enteroviruseshas been improved greatly by using polymerase chainreaction (PCR). 112,113 CSF PCR for HSV during day 2e 10of illness has overall sensitivity and specicity of > 95%for HSV encephalitis in immunocompetent adults. 10,39

Although HSV PCR may be negative in the rst few daysof the illness 113 a second CSF taken 2 e 7 days later willoften be HSV positive, even if aciclovir treatment hasbeen started. 11,114

Further microbiological investigations should be basedon specic epidemiological factors (age, animal and insectcontacts, immune status, recreational activities, geographyand recent travel history, season of the year and vaccina-tion history) and clinical ndings (hepatitis, lymphadenop-athy, rash, respiratory tract infection, retinitis, urinarysymptoms and neurological syndrome (Table 11.Microbiological investigation of encephalitis). 102,110,115,116

Table 11 Microbiological investigations in patients withencephalitis, modied from (Solomon, Hart et al., 2007). 110

CSF PCR1. All patients

HSV-1, HSV-2, VZVEnterovirus, parechovirus

2. If indicatedEBV/CMV (especially if immunocompromised)HHV6,7 (especially if immunocompromised, or children)Adenovirus, inuenza A &B, rotavirus (children)Measles, mumps

Erythrovirus B19Chlamydia3. Special circumstances

Rabies, West Nile virus, tick-borne encephalitis virus(if appropriate exposure)

Antibody testing (when indicated e see text )a

1. Viruses: IgM and IgG in CSF and serum (acute andconvalescent), for antibodies againstHSV-1 & 2, VZV, CMV, HHV6, HHV7, enteroviruses, RSV,Erythrovirus B19, adenovirus, inuenza A & B

2. If associated with atypical pneumonia, test serum forMycoplasma serologychlamydophila serology

Ancillary investigations (when indicated - Theseestablish carriage or systemic infection, but notnecessarily the cause of the CNS disease )

Throat swab, nasopharyngeal aspirate, rectal swab,faeces, urinePCR/culture of throat swab, rectal swab, faeces forenterovirusesPCR of throat swab for mycoplasma, chlamydophilaPCR/antigen detection of nose/throat swab ornasopharyngeal aspirate for respiratory viruses,adenovirus, inuenza virus (especially children)PCR/culture of parotid duct swab followingparotid massage or buccal swab for mumpsPCR/culture of urine for measles, mumps and rubella

Vesicle electron microscopy, PCR and cultureb

Patients with herpetic lesions (for HSV, VZV)Children with hand foot and mouth disease (forenteroviruses)

Brain Biopsy For culture, electron microscopy, PCR andimmunohistochemistry b

a Antibody detection in the serum identies infection (pastor recent depending on the type of antibodies) but does notnecessarily mean this virus has caused the CNS disease.

b Viral culture and electron microscopy less sensitive thanPCR.

462 R. Kneen et al.


15/29

What antibody testing should be done on serum &CSF?

Recommendation

Guidance from a specialist in microbiology, virology orinfectious disease specialists should be sought in decid-

ing on these investigations (B, III) In patients with suspected encephalitis where PCR ofthe CSF was not performed acutely, a later CSF sample(at approximately 10-14 days after illness onset) shouldbe sent for HSV specic IgG antibody testing (B, III)

In suspected avivirus encephalitis CSF should betested for IgM antibody (B, II)

Acute and convalescent blood samples should be takenas an adjunct to diagnostic investigation especiallywhen EBV, arboviruses, lyme disease, cat scratch dis-ease, rickettsiosis or ehrlichioses are suspected (B, II)

EvidenceWhilst all patients with suspected encephalitis should have

PCR requested for the common viruses, decisions aboutantibody testing of serum and CSF are best made inconjunction with the specialist microbiology/virology orinfectious diseases service.

Cerebrospinal uid Intrathecal synthesis of HSV-specic IgG antibodies isnormally detected after 10 e 14 days of illness, peaksafter one month and can persist for several years. 117 Thedetection of intrathecally synthesized HSV IgG antibodies,when available, may help to establish the diagnosis of HSVencephalitis in patients where the CSF is taken after day10e 12 of the illness. This is especially useful in patientsfor whom an earlier CSF was not taken, or was not testedfor HSV by PCR. A European consensus statement recom-mended the combined approach of testing CSF by PCRand antibody detection, such that a negative HSV-PCR re-sult early in the disease process coupled with a negativeHSV-specic CSF antibody study sampled 10 e 14 days aftersymptom onset effectively ruled out the disease 39 ; how-ever, intrathecal immune responses may be delayed orabsent when antiviral therapy is started early. 118 Addition-ally, many laboratories do not provide CSF antibodydetection services. The detection of oligoclonal bands inthe CSF is a non-specic indicator of an inammatory pro-cess in the CNS; immunoblotting of the bands against viralproteins from HSV can been used to detect anti-HSVantibody, although this is not routinely available. 32,39 Anti-body detection can also be particularly useful in VZVencephalitis. 119

The detection of virus specic IgM in CSF is usuallyindicative of an intrathecal antiviral immune response; thisis especially useful for aviviruses and other RNA viruses,which tend to be primary infections, rather than DNAviruses, which are often reactivations. 80

Blood Acute and convalescent blood samples should be taken forappropriate serological testing based on the likely organ-isms identied from specic epidemiological and clinical

features. 11 Examples of infectious causes of encephalitisthat can be diagnosed from serological investigations ofblood include: EBV, arthropod-borne viruses (arboviruses),Borrelia burgdorferi (lyme disease), Bartonella hensae(cat scratch disease), rickettsioses, ehrlichioses, and myco-plasma. Serological testing for antibodies in autoimmuneencephalitis is covered in the ‘Special circumstances’ sec-tion of the guideline.

What PCR/culture should be done on other samples(e.g. throat swab, stool, vesicle etc)?

Recommendation

Investigation should be undertaken through close col-laboration between a specialist in microbiology, virol-ogy, infectious diseases and the clinical team (B, III)

In all patients with suspected viral encephalitis throatand rectal swabs for enterovirus investigations shouldbe considered; and swabs should also be sent from ves-icles, if present (B, II)

When there is a recent or concomitant respiratory tractinfection, throat swab or sputum/lavage should be sentfor PCR for respiratory viruses (B, II)

When there is suspicion of mumps CSF PCR should beperformed for this and parotid gland duct or buccalswabs should be sent for viral culture or PCR (B, II)

EvidenceInvestigation of sitesoutside theCNS canbe usefulto providepointers as to possible aetiology (Table 11 Microbiologicalinvestigations); however it must be remembered that suchinfectionmight be coincidental rather than causal; this is es-pecially the case for non-sterile sites, or sites where longterm shedding of virus occurs (e.g. in the stool). In enterovi-rus encephalitis, the virus may be isolated by swabbing thethroat and rectum, or, if present, vesicles. 10 Of these vesic-ular swabs are most useful because they indicate acute andsystemic infection, whereas carriage in the faeces, and tosome extent the throat may be long term. 82 Although manypatients withenterovirus CNSinfectionsdo nothavevesicles.

If the clinical illness suggests a recent respiratoryinfection, samples taken from the respiratory tract (throatswab, nasal swab, nasopharyngeal aspirate, nasal washings,tracheal aspirate or brochoalveolar lavage) can be testedby PCR for respiratory viruses.

Mumps encephalitis is most accurately conrmed byPCR of the CSF; serum or salivary mumps antibodies arealso helpful. Viral culture or PCR performed on parotidgland duct swabs, taken after massaging the parotid glandfor 30 s, or buccal (saliva) swabs are useful for thediagnosis of recent mumps virus infection, within 9 daysof the onset of symptoms. A urine sample is less sensitivebut may be positive for at least 5 days after detection inthe mouth. Urine analysis if also useful if measles issuspected.

Viral infection may be demonstrated by culture, de-tection of viral genomes (by PCR), viral antibodies (byserology), viral antigens (by direct immunouoresence), or,if available, viral particles (by electron microscopy),although PCR is the most sensitive. 9,11



16/29

Which children with encephalitis should have anHIV test?

Recommendation

We recommend that an HIV test be performed on all pa-tients with encephalitis, or with suspected encephalitis

irrespective of apparent risk factors (A, II)EvidenceHIV is directly capable of causing an encephalopathy inyoung children 120 but infection with the virus also predis-poses children to CNS infections from other specic patho-gens. Antenatal screening for HIV infection is offered to allpregnant women in the United Kingdom and has a high up-take (90%). However it is not compulsory, and Mother toChild transmission of HIV can still occur in the UK. 121,122 Mi-grants and travellers to the UK, from areas dened by theWHO as areas with high endemicity of HIV infection 123 par-ticularly from sub-Saharan Africa, and Asia, including majorparts of the former Soviet Union, should be regarded as be-

ing at risk of Mother to Child Transmission of HIV. It must beremembered that while HIV infection is rapidly progressivein 20% of infants, 124,125 late presentation of perinatally ac-quired HIV infection can occur at 13 years of age orolder. 126,127 HIV should be considered in children with sus-pected encephalitis for three reasons. Children with undi-agnosed advanced HIV disease can present with CNSinfections from a number of the less common infectiouscauses, such as cytomegalovirus. 128,129 Secondly some ofthe more common CNS infections, such as Streptococcus pneumoniae or Mycoplasma tuberculosis have an increasedincidence in patients with HIV. Thirdly, although uncom-mon in children, primary HIV-1 infection can present withan acute meningoencephalitis as part of a seroconversion

illness.130

The current UK guidelines on HIV testing123,131

emphasise that “all patients presenting for healthcarewhere HIV, including primary HIV infection, enters the dif-ferential diagnosis” should be tested for HIV and offers de-tailed guidance on the testing of infants, children andyoung people.

What is the role of brain biopsy in children withsuspected viral encephalitis?

Recommendation

Brain biopsy has no place in the initial assessmentof suspected acute viral encephalitis in immunocompe-tent children. Stereotactic brain biopsy should be con-sidered in a child with suspected encephalitis in whomno diagnosis has been made after the rst week, espe-cially if there are focal abnormalities on imaging and ifthe ndings could change the child’s management(B, II)

If imaging shows nothing focal, an open biopsy, usuallyfrom the non-dominant frontal lobe, may be preferable(B, II)

The biopsy should be performed by an experienced pae-diatric neurosurgeon and the histology should be exam-ined by an experienced neuropathologist (B, III)

EvidenceFor many years brain biopsy was the preferred method fordiagnosing HSV encephalitis, because clinically many con-ditions mimic HSV encephalitis, 33 the chances of culturingthe virus from the CSF were low, and a biopsy was one ofthe few reliable means of making the diagnosis; althoughits sensitivity was low, specicity was high. 33 SubsequentlyCSF PCR for HSV DNA was developed, and proved a rapidand reliable diagnostic test, 112,113 largely replacing biopsyfor the diagnosis HSV encephalitis. However biopsy stillhas a role in the investigation of other patients. Althoughuntil recently it was considered highly invasive with a signif-icant mortality and morbidity (through intracranial haemor-rhage, or biopsy site oedema), with modern stereotacticapproaches the incidence of serious adverse events islow, and it is now considered a relatively safeinvestigation. 132,133

There is no role for a brain biopsy in the initialassessment of patients with suspected HSV encephalitis.However it may have a role in patients with suspected HSVencephalitis who are PCR negative and deteriorate despiteaciclovir, or to identify alternative causes, such as vascu-litis 134 ; biopsy is especially helpful if there is a focal lesionon imaging 132 or in patients with immune compromisewhere the differential diagnosis is often wide. Tissue needsto be sent for pathogen detection (electron microscopy,culture, PCR and immunoouresence) and for histopathol-ogy. Experienced neuropathologists are essential. In one se-ries one fth of patients with suspected HSV encephalitishad an alternative diagnosis made by biopsy, in half ofwhom it was a treatable condition. 33

What is the role of magnetic resonance imaging(MRI) and other advanced imaging techniques inchildren with suspected viral encephalitis?

Recommendation

MRI (including diffusion weighted imaging), should beperformed as soon as possible on all patients with sus-pected encephalitis in whom the diagnosis is uncertain;ideally this should be within 24 hours of hospital admis-sion, but certainly within 48 hours (B, II). Where the pa-tient’s clinical condition precludes an MRI, urgent CTscanning may reveal an alternative diagnosis (A, II)

MRI sequences obtained need to be chosen appropri-ately for a paediatric population and images shouldbe interpreted by an experienced paediatric neuroradi-ologist (B, II)

The role of MR spectroscopy is uncertain; SPECT andPET are not indicated in the assessment of suspectedacute viral encephalitis (B, II)

EvidenceMRI is signicantly more sensitive than CT in detecting theearly cerebral changes of viral encephalitis. In HSV enceph-alitis a CT obtained early may be normal, or have onlysubtle abnormalities; in one small series only a quarter ofpatients with HSV encephalitis had an abnormality on initialCT scanning. 135 In contrast, MRI obtained within 48 h ofhospital admission is abnormal in approximately 90% of

464 R. Kneen et al.


17/29

patients. 135 e 138 Early MRI changes occur in the cingulate gy-rus and medial temporal lobe, and include gyral oedema onT-1 weighted images, and high signal intensity on T2-weighted and T2 uid attenuated inversion recovery(FLAIR) images.139 Later there may be haemorrhage. Diffu-sion-weighted MRI may be especially sensitive to earlychanges. 140 e 142 Specic sequences, such as FLAIR andSTIR (short-tau inversion recovery) sequences may be ofparticular value in young children due to normal brain mat-uration processes. 143 The MRI should be interpreted by anexperienced neuroradiologist.

The changes seen on MRI are reported to be specic(87.5%) for PCR-conrmed HSV encephalitis but can alsoidentify alternative (often treatable) diagnoses in patientsthat are negative for HSV. 137 Thus it is important that anMRI is performed urgently. In small studies, the extentof MRI abnormality seen acutely in HSV encephalitis didnot correlate with the clinical evolution of the disease 137

nor with depression afterwards, 144 though a correlationbetween number of seizures in acute HSV encephalitis,and subsequent brain atrophy on MRI has been demon-strated. 145 In VZV CNS disease in immuncocompetent chil-dren, the most common pathogenesis is a large vesselvasculitis, which presents with an ischaemic or haemor-rhagic infarct often seen on MRI and angiography. 146 e 148

In immunocompromised children, VZV may cause a multi-focal leukoencephalopathy which may be seen to followa clear arterial distribution. 34 Other pathogens may havetypical abnormal ndings. M. pneumoniae may show focalcortical lesions, deep white matter lesions and large areasof demyelination. 61 Japanese B encephalitis typically in-volves the thalamus and basal ganglia with T2 hyperinten-sity. Enterovirus may result in generalised parenchymaldestruction, or may predominately affect the brainstem,occasionally spreading posteriorly to involve the cerebel-lar denate nuclei or superiorly to involve the thalamiand basal ganglia. 149 In young children, white matter oe-dema is not easily distinguished from unmyelinated whitematter, and without contrast may result in incorrectreporting. 150

Although MRI is the investigation of choice, in young,acutely ill, comatose or confused children a general anaes-thetic is usually required posing practical difculties formany hospitals. In one series 70% of children requiredsedation or general anaesthesia. 143 Some would recom-mend that sedation not be used in this patient group andonly anaesthetic support and formal anaesthesia used. 151

In these circumstances, CTscanning may be the only urgentimaging available. However, the CT scan may be normal inchildren with CNS infections including severe bacterialmeningitis and encephalitis so should not be relied uponto make or refute the diagnosis of these conditions. 152,153

A pragmatic approach is to perform a CT scan as the rstcranial imaging investigation and then an MRI can be under-taken as soon as it can be arranged, often after transfer tothe local tertiary centre.

Other modalitiesMR spectroscopy identies and quanties concentrations ofvarious brain metabolites, and so may help distinguishnormal from diseased brain tissue, and characterise thenature of the damage, particularly distinguishing

inammatory from neoplastic processes; however there areno prospective studies assessing its diagnostic role. Singlephoton emission computed tomography (SPECT) may showfocal hypoperfusion persisting after recovery from acuteviral encephalitis. 154 However, it has been used mainly asa research tool and appears to have little application in sus-pected acute encephalitis in practice. Fluorodeoxyglucosepositron emission tomography (PET) shows abnormalities inacuteviralencephalitis 154,155 with regions of FDG-PET hyper-metabolism seen most frequently in the medial temporallobes (sometimes reecting seizure activity). However PETscanning is not practical or sufciently informative to beused in children with suspected acute viral encephalitis.

Which children with suspected viral encephalitisshould have an electroencephalogram (EEG)?Recommendation

An EEG should not be performed routinely in all pa-tients with suspected encephalitis; however, in pa-tients with mildly altered behaviour, if it is uncertain

whether there is a psychiatric or organic cause an EEGshould be performed to establish whether there are en-cephalopathic changes (B, II)

EEG should also be performed if subtle motor, or sub-clinical seizures are suspected (B,II)

An EEG should be performed in children with suspectedchronic viral encephalitis for example SSPE (B, II)

EvidenceThe EEG is abnormal in most patients with encephalopathy,including more than 80% of those with acute viral enceph-alitis. 7,27 When patients have a more subtle presentation, itcan be helpful in determining whether abnormal behaviouris due to psychiatric causes or is an early feature of ence-phalopathies. EEG is also useful in determining whetheran individual has non-convulsive or subtle clinical seizures,which occur in both HSV encephalitis and otherencephalopathies. 156,157

In HSV encephalitis EEG abnormalities include non-spe-cic diffuse high amplitude slow waves, sometimes withtemporal lobe spike-and-wave activity and periodic lateral-ised epileptiform discharges (PLEDs). 158 Even though PLEDsoccur in many cases of HSV encephalitis 33 and were at onestage considered pathognomonic, they are now recognisedin otherviral encephalitides 36 and non-infectiousconditions,andit is accepted that there areno EEGchanges diagnostic ofHSV encephalitis. 159 For example when PLEDS are identiedin patients with a sub-acute or chronic encephalopathy this

would be suggestive of SSPE.77,160

Treatment of viral encephalitis

For which patients should aciclovir treatment bestarted empirically?

Recommendation

Children with suspected encephalitis should have intra-venous aciclovir started if the initial CSF and/or imag-ing ndings suggest viral encephalitis, and denitely



18/29

within 6 hours of admission if these results are awaited(A, II).

If the rst CSF microscopy or imaging is normal but theclinical suspicion of HSV or VZV encephalitis remains,aciclovir should still be started within 6 hours of admis-sion whilst further diagnostic investigations (as outlinedbelow) are awaited (A, II)

The dose of intravenous aciclovir should be:- 3 months-12 years 500mg/m 2 8 hourly- > 12 years 10mg/kg 8 hourly

The dose of aciclovir should be reduced in patients withpre-existing renal impairment (A, II)

Patients with suspected encephalitis due to infectionshould be notied to the appropriate Consultant inCommunicable Disease Control (In Scotland, only pa-tients with a proven aetiology or those occurring aspart of an unusual outbreak are notiable) (A, III)

If meningitis is also suspected, the child should also betreated in accordance with the NICE meningitis guide-line (A, II)

EvidenceAciclovir is a nucleoside analogue with strong antiviralactivity against HSV and related herpes viruses, includingVZV. Tworandomised trialshave shownthat aciclovir(10 mg/kg three times a day) improves the outcome in adults withHSV encephalitis from a mortality of about 70% to less than20e 30%.5,6 Even with aciclovir treatment the outcome is of-ten still poor, especially in patients with advanced age, a re-duced coma score, or delays of more than 48 h betweenhospital admission and starting treatment. 7,8 Because HSVencephalitis is the most commonly diagnosed viral encepha-litis in industrialised countries, treatment with aciclovir isusually started once the initial CSF and/or imaging ndingssuggest viral encephalitis, without waiting for conrmationof HSV by PCR. However, unlike meningococcal septicaemia,where children can die within a few hours and thus immedi-ate treatment with antibiotics is needed, in a patient withencephalopathy who has only mild confusion, investigationwith a lumbar puncture before considering treatment is sen-sible; especially given the very wide differential diagnosis,and relative rarity of HSV encephalitis. Moreover, empiricaluse of antimicrobial and antiviral agents can prematurelyhalt the diagnostic pathway because clinicians feel falselyreassured, and this delays the identication of otheraetiologies for which different treatments might beappropriate.

Experience from paediatrics has shown that the

guidline treatment up

Documents