Chikungunya and the nervous system: what we do and do not know

Rev. Med. Virol. 2009; 19: 121–129.Published online 9 March 2009 in Wiley InterScience

(www.interscience.wiley.com)Reviews in Medical Virology DOI: 10.1002/rmv.606

Chikungunya and the nervous system:what we do and do not knowCarla Arpino1, Paolo Curatolo1 and Giovanni Rezza2*1Child Neurology Unit, Tor Vergata University, Roma, Italy2Department of Infectious Diseases, Istituto Superiore di Sanita, Roma, Italy

SUMMARY

Chikungunya virus (CHIKV), an alphavirus transmitted by mosquitoes of the Aedes genus, has recently re-emerged,causing epidemics on Indian Ocean Islands and the Indian subcontinent, and an unexpected outbreak in north–eastern Italy. CHIKV infection was first reported to affect the nervous system in the 1960s; in the early 1970s it wasfound to be associated with meningoencephalopathy, myelitis, and choroiditis, and animal studies appeared toconfirm that CHIKV was neurotropic. Nonetheless, CHIKV has never been considered as a ‘true’ neurotropic virus.The re-emergence of CHIKV infection in areas with efficient clinical facilities has allowed CHIKV-related neurologicaldisease to be better defined both in adults and children. Encephalopathy appears to represent the most commonclinical manifestation among newborns infected through mother-to-child transmission. Although data are still scarce,the ratio between cases with and without CNS involvement for CHIKV appears to be comparable with that for otherneurotropic viruses. Unfortunately, the neurotropism of CHIKV has not been completely defined, and differentanimal studies show inconsistencies with regard to the capacity of the virus to invade and replicate in the brainparenchyma. This merits further investigation in light of the emergence of the virus in previously unaffected areasand of the clinical evidence of CNS involvement in a considerable proportion of symptomatic cases. Copyright #2009 John Wiley & Sons, Ltd.

Received: 3 October, 2008; Revised: 24 December, 2008; Accepted: 30 December 2008

INTRODUCTIONArboviruses get their name from an abbreviationof ‘arthropod-borne viruses’, an ecologicaldefinition which includes any virus that affectsvertebrates and is biologically transmitted byarthropods. Arboviruses belong to several distinctfamilies of viruses and are, collectively, known tobe an important cause of acute disease of the ner-vous system worldwide. However, only some ofthe members of this group of viruses affect theCNS (a list of selected viruses is reported inTable 1). Currently, there is much concern overthe increased activity of some of these viruses,and the epidemic of infection with the West Nile

virus (WNV) in the United States may be consid-ered as a paradigm of the public-health threatrepresented by arboviruses [1].In the past 3 years, another arbovirus (i.e. the

Chikungunya virus, CHIKV) has re-emerged.CHIKV is an alphavirus of the Togaviridae familyand is transmitted by mosquitoes of the Aedesgenus. Three distinct clades have been identified:the West African genotype, the Central/East Afri-can genotype and the Asian genotype. CHIKV wasfirst isolated in the Newala district of Tanzania in1952–1953 [2]. Since then, sporadic cases and anumber of large epidemics have been reported inseveral African countries, on the Indian subconti-nent, and in southeast Asia [3,4]. In particular, arecent series of outbreaks began in Kenya in 2004and ravaged the Comoros Islands, the island of LaReunion, and other islands in the southwestIndian Ocean in early 2005. These outbreakswere followed by an epidemic on the Indian sub-continent in 2005–2006 [5,6] and by an unexpectedoutbreak in northeastern Italy [7]. According to themolecular analysis of the strains isolated on Indian

RR EE V II E W

Copyright # 2009 John Wiley & Sons, Ltd.

*Corresponding author: G. Rezza, Epidemiology Unit, Department ofInfectious Diseases, Istituto Superiore di Sanita, Roma, ItalyE-mail: [email protected]

Abbreviations usedCHIKV, Chikungunya virus; EEE, Eastern equine encephalitis; EEG,electroencephalography; GBS, Guillain-Barre syndrome; MRI, mag-netic resonance imaging; SF, Semliki Forest virus; SINV, Sindbisvirus; VEE, Venezuelan equine encephalitis; WEE, Western equineencephalitis; WNV, West Nile virus.

Ocean islands and in India, these epidemics werecaused by a variant of the Central/East Africangenotype [8,9].Although persons with CHIKV infection usually

have relatively mild disease, presenting with highfever, joint pain and skin rash, at times neurologi-cal complications occur, yet these complicationshave never been well documented. In the recentoutbreaks, a number of infected individuals hadatypical and/or severe manifestations of infection,including neurological consequences. Further-more, in a recent hospital-based study conductedon the island of La Reunion, encephalitis andmeningoencephalitis were two of the major causesof death among persons with severe atypicalCHIKV fever [10], stressing the need to focus onthe neurological consequences of CHIKV infectionand the potential neurotropism of this virus.

NEUROTROPIC AND NON-NEUROTROPICALPHAVIRUSES: A CLEAR CUT DEFINITION?The term ‘neurotropic’ indicates a particular affi-nity of the virus for neural tissue. It is thereforenecessary to demonstrate direct invasion of neu-roinvasiveness and a preferential affinity for neur-al tissue by virus in order to establishneurotropism. Aseptic meningitis, without neuraltissue involvement, or an associated post-infectiveinflammatory disease, such as Guillain–Barre syn-drome (GBS) or acute disseminated encephalo-myelitis (ADEM), therefore does not indicateneurotropism. The ability of a neurotropic virusto cause CNS disease is then defined as ‘neuro-virulence’.

Several members of the alphavirus genus, aswell as other arboviruses, are known to causeacute infection of the CNS, whose manifestationsrange from relatively mild disease, with feverand headache or aseptic meningitis, to severeencephalomyelitis [11]. The initial replication ofthe virus usually occurs at a site peripheral tothe CNS; the virus is then released into the blood-stream and enters the CNS by either retrogradeaxonal transport or blood-borne infection of thecerebral capillary endothelium or choroidalepithelial cells [12–14]. For all alphaviruses affect-ing the CNS, the primary target in the CNS is theneuron [11]. The severity of neurological manifes-tations depends on the neurovirulence of the virusand the maturity of the neurons: the risk of devel-oping encephalitis and adverse outcomes is age-dependent, with a higher risk among young chil-dren [15–18] and the elderly [19].CHIKV has never been considered as a ‘true’

neurotropic virus, like those of the flavivirusgenus, such as WNV, the Japanese encephalitisvirus (both of which belong to the mosquito-borneencephalitis complex), and the tick-borne encepha-litis virus. Even within its own genus (i.e. alpha-virus), CHIKV is classified among those virusesassociated with fever, rash and polyarthritis,which are found in the Old World (e.g., o’nyong-nyong and Ross River), and not among thoseviruses that cause encephalitis, which are primar-ily found in the New World (e.g. Eastern equineencephalitis [EEE], Western equine encephalitis[WEE] and Venezuelan equine encephalitis [VEE]viruses) (Table 2). Of these latter viruses, which

Table 1. Some examples of Arboviruses capable of causing encephalitis and/or other CNSinvolvement

Family Genus (Species) Vector

Flaviviridae FlavivirusWest Nile encephalitis, Japanese encephalitis,(St. Louis encephalitis, Murray Valley, Rocio) Mosquito(Tick-born encephalitis, Powassan, Kyasanur Forest) Tick

Togaviridae Alphaviruses (EEE, WEE, VEE) MosquitoBunyaviridae Orthobunyavirus (California, La Crosse) Mosquito

Phlebovirus (Toscana) Phlebotomus

WEE, Western equine encephalitis; EEE, Eastern equine encephalitis; VEE, Venezuelan equine encephalitis.

122122 C. ArpinoC. Arpino et al.et al.

Copyright # 2009 John Wiley & Sons, Ltd. Rev. Med. Virol. 2009; 19: 121–129.DOI: 10.1002/rmv

were first isolated from the brain of dead horses inthe 1930s, EEE is the most neurovirulent. It iswidespread in Central and South America and inparts of the USA, and it has an enzootic cyclebetween birds and ornithophilic mosquitoes.Humans and horses represent ‘dead-end’ hosts,in which the virus causes severe fulminant ence-phalitis, with case-fatality rates exceeding 30%,due to a combination of a direct cytopathic effect,inflammatory damage and vasculitis [20]. TheWEE virus causes more human infections thanthe EEE virus, yet the illness is less severe andthe case-fatality rate rarely exceeds 10%. Finally,VEE, the least neurovirulent of these viruses,usually causes a subclinical or self-limiting febrileillness with infrequent development of severeencephalitis and an overall case-fatality rate below1% [3,21].

Although CHIKV is not considered to be neuro-tropic, sporadic cases of neurological manifesta-tions during acute infection have been reportedfor a number of years. In particular, a few casesof CHIKV affecting the nervous system were firstreported in the 1960s during large outbreaks inThailand [22] and India [23]. In the early 1970s,CHIKV was associated with meningoencephalopa-thy, myelitis and choroiditis in Cambodia [24], andanimal studies appeared to confirm that CHIKVwas neurotropic [25].

The re-emergence of CHIKV infection in areaswith efficient clinical facilities has allowedCHIKV-related neurological disease to be better

defined. Recently, in La Reunion, 30children withCHIKV infection (25% of all hospitalised cases)had neurologic manifestations, such as encephali-tis, febrile seizures, meningeal syndrome andacute encephalopathy [26]. In a hospital-basedstudy in La Reunion, neurological manifestationswere detected in 61% of the children between 3and 18 years of age but in only 16% of thoseyounger than 3 years [27]. Neurologic abnormal-ities (e.g. altered mental function and focal neuro-logical deficits) were also reported during therecent epidemic in India [28]. Furthermore, in LaReunion, severe illness due to encephalopathyrepresented the most common clinical manifesta-tion among newborns infected through mother-to-child transmission [29]. Finally, a few cases ofGBS were diagnosed in adults with CHIKV infec-tion in La Reunion [30]. The main findings of thesestudies are discussed in detail below.

CLINICAL MANIFESTATIONS

Meningitis, encephalitis,encephalomyeloradiculitisNeurological manifestations of acute CHIKV dis-ease were first reported during an outbreak inMadras, India, in the summer of 1964. Signs ofmeningeal irritation with nuchal rigidity and Ker-nig’s sign, sluggish pupillary reaction, deliriumand reduced consciousness, with no significantCSF changes, were observed in six patients with

Table 2. Alphaviruses: antigenic complex, and clinical features

Genotype Antigenic complex Clinical features(virus species)

Sindbis Fever, rashBarmah Forest Barmah Forest Fever, rash, joint painMiddelburg Middelburg ?Nduma Nduma ?Semliki Forest Semliki Forest

(Semliki Forest) Fever, encephalitis(CHIK, Ross River) Fever, rash, joint pain

WEE/EEE WEE Fever, encephalitisEEE Fever, encephalitis

VEE VEE Fever, encephalitis

WEE, Western equine encephalitis; EEE, Eastern equine encephalitis; VEE, Venezuelan equine encephalitis.

Chikungunya and the nervous systemChikungunya and the nervous system 123123


suspected or confirmed CHIKV [23]. In Cambodia,two cases of acute encephalitis, one without focalneurologic signs and the other with corticospinaltract involvement and vestibular signs, possiblyassociated with CHIKV infection, were reportedin 1970 [24].A more detailed description of neurological

manifestations was recently provided by studiesconducted in La Reunion and India and is sum-marised in Table 3 [26,28]. In La Reunion, in a clin-ical series of 30 patients with neurologicalmanifestations associated with CHIKV infection,encephalitis was diagnosed in 12 children; 6 chil-dren had complex febrile seizures and four hadsimple febrile seizures; four children were diag-nosed with acute encephalopathy, defined asreduced consciousness in the absence of other neu-rological signs, and four children had meningealsyndrome [26]. In the Indian series, encephalitiswas diagnosed in 15 persons, encephalomyelitisin three persons, and optic neuritis in two persons[28].In these studies, neurological symptoms usually

began early in the course of the disease (i.e. lessthan 24 h after the onset of fever in 73% of the chil-

dren in La Reunion and on the 2nd or 3rd day inmost adults in India). In nearly all cases, the dis-ease started with abrupt high fever and jointpain; macular rash was reported in 67% of thecases in La Reunion but in only 25% in India. Inthese outbreaks, suspected CHIKV infection,based on epidemiological and clinical findings,was confirmed by standard laboratory procedures(i.e. PCR detection of the viral genome in the ser-um and/or CSF, and/or by the detection of serumanti-CHIKV IgM in the first 15 days after the onsetof symptoms).In most cases, the CSF findings were unremark-

able. The glucose level was normal except for twocases in La Reunion; increased protein level wasdetected in 85% of the Indian patients yet in onlytwo patients in La Reunion; mild pleocytosis (55white cells/mL) was found in 45% of the cases inIndia, whereas in La Reunion only one child hadpleocytosis (120 white cells/ml). Finally, theCHIKV genome was detected by RT-PCR in theCSF of 11 of the 18 patients tested in La Reunion;this test was not performed in the Indian study orin any of the other outbreak studies. Radiologicalfindings were largely non-specific. Head ultrasound

Table 3. Frequency distribution of neurological signs and symptoms of CHIKV disease in20 adults, 22 children (> 1 years old) and 8 infants (< 1 year old) recruited in India or on LaReunion (adapted from references 26 and 28)

Neurological sign* Adults (India) Children (Reunion)

> 1 year < 1 year

N (%) N (%) N(%)

Altered level of consciousness 20 (100) 15 (68) 2 (25)Cranial nerve deficit 20 (100) — —Seizures 15 (75) 13 (59) 5 (62)Decreased deep tendon reflexes 7 (35) — 1 (12)with faintnessPsychosis 6 (30) 3 (14) —Hemi/paraparesis, paraplegia 4 (20) — —Involuntary movements 4 (20) 1 (4) —Pyramidal syndrome — 2 (9) —Nuchal rigidity — 7 (32) —Hypotonia — — 2 (25)Tense fontanelle — — 2 (25)Status epilepticus — — 1 (12)

*not mutually exclusive



was performed in four newborns, all of whomshowed brain oedema and lenticulothalamostriatalvasculitis [26]. The brain CT performed on the chil-dren in La Reunion showed, for some, brain oede-ma and hyperdensities of periventricular whitematter and cerebellar hemisphere, suggestingthat there were areas of haemorrhage [26]; multi-ple small haemorrhages with diffuse cerebraloedema were also detected among adults in India[28]. Brain magnetic resonance imaging (MRI) wasperformed only in La Reunion: the findings werenormal in more than 60% of the cases; in the othercases, increased T2 signal in periventricular whitematter and/or other cerebral areas (i.e. bilateralcentrum semiovale, cingular and limbic areas, cer-ebellar hemispheres) was found. When diffusion-weighted imaging was performed, it showed areasof restricted diffusion in bilateral frontal, parietaland temporal white matter. In La Reunion, electro-encephalography (EEG) was performed for 18 chil-dren in the acute phase; 16 of them had abnormalfindings (i.e. diffuse slow waves predominant inanterior regions; status epilepticus with a burstsuppression pattern; polyspikes and waves on aslowed background activity and paroxysmal poly-spikes) [26]. In the adults in India, the EEG find-ings were normal [28].

Other CNS pathological manifestations havebeen described only in small series of patients.For example, two cases of encephalomyeloradiculi-tis were observed in eastern Maharashtra, India[31]. These patients, who presented with feverand joint pain, showed drowsiness and reducedor no limb movement. Electromyography showedsensorimotor peripheral neuropathy in one caseand motor neuropathy with no sensory componentin the other. Frontoparietal white matter lesions,which are common neuroimaging findings in theearly stage of viral encephalitis, and enhancementof the ventral cauda equina nerve roots, weredetected with brain and spinal MRI, respectively.

In these patient series, it cannot be excluded thatseizures and other non-focal neurological featurescould have resulted from a variety of indirectcauses, including fever, electrolyte and metabolicderangements, whereas focal features could havebeen due to a secondary vascular mechanism.However, the onset of neurological features earlyin the illness may favour a direct effect ratherthan ADEM, as does viral DNA in CSF, but neitheris conclusive.

GBS and peripheral neuritisGBS was reported in three children in La Reunion[30]. These children had quadriparesis (which wasmore pronounced distally) and facial palsy, whichoccurred between 3 days and 2 weeks after theonset of fever. The classical albuminocytologic dis-sociation (i.e. acellular rise of total protein in theCSF) was detected at lumbar puncture in twopatients. Elevated IgM CHIKV antibodies werefound in all patients, with positive PCR in the ser-um of one of them. GBS has been associated withseveral arboviruses, such as WNV and denguevirus. However, before the recent outbreak in LaReunion, only one case of polyneuropathy of theGuillain–Barre type had been described in a patientaffected by CHIKV fever, during the outbreak inMadras in 1964 [23]. Finally, four cases of acute flac-cid paralysis associated with CHIKV have beenreported in India [32]; the patients presented withweakness in their upper and lower limbs, whichwas probably due to peripheral neuritis. In generalterms, cases of GBS or peripheral neuritis (i.e., facialpalsy) coming on 1–2 weeks after the onset of feverare more in favour of a post-infective inflammatoryprocess than a direct viral damage.

Neurological manifestations in newborns as aconsequence of mother-to-child transmissionVertical transmission of CHIKV infection appearsto occur in nearly 50% of women with viremiaand does not seem to be affected by caesarean sec-tion. All neonates are asymptomatic at birth butthen develop neonatal disease at a median of 4days after birth, suggesting the occurrence of peri-natal transmission. Severe disease has beenobserved in more than 50% of patients; in about90% of whom, it consists of encephalopathy [29].In the nine neonates with encephalopathy studiedat the Maternity Department of the Groupe Hospi-talier Sud-Reunion, MRI findings showed, in mostcases, brain swelling, and, in a minority of thesevere cases, cerebral haemorrhage, with scatteredand sometimes hyperintense signal of the supraten-torial white matter in the early stage. About 40% ofthese neonates developed persistent disabilities. Inanother study conducted in La Reunion, hypotoniawas the most common neurologic feature in mater-nofetal forms, followed by coma, seizures and epi-leptic status. CSF was often positive for CHIKV,and MRI revealed scattered white matter lesionswith areas of intraparenchymal haemorrhage, but



no grey matter lesions [33]. A few cases of CNSinvolvement were also seen in Mayotte, Comorosarchipelago: of these, three had meningoencephali-tis and one severe hypotonia [34].

Outcome and sequelaeOf the 30 children described in the article of Robinet al. [26], all but one were hospitalised, and sevenof them (24%) were admitted to the paediatricintensive care unit: of these, four required mechan-ical ventilation. Two children died, one with mas-sive haemorrhage (epistaxis, emesis, haemoptysisand haematuria) and circulatory collapse, theother with circulatory collapse and massive brainswelling. Neurological sequelae were found infour patients at discharge.Of the 20 adults with neurological complications

observed in India, six died; three of them wereaffected by systemic disease, such as diabetes mel-litus and hypertension [28]. With the exception ofa 45-year-old patient, who developed multiplesmall haemorrhages and diffuse cerebral oedema,the patients were all over 65 years of age; one ofthese patients had developed total blindness dueto retro-bulbar neuritis. In another patient, whosurvived, temporal pallor of the optic disc wasrevealed by direct fundoscopy. Finally, encephali-tis and meningoencephalitis were the fourth lead-ing causes of death among 65 patients with severeatypical CHIKV fever in a hospital-based studyconducted in La Reunion [10].

NEUROPATHOLOGICAL FINDINGSASSOCIATED WITH INFECTIONWITH CHIKV AND OTHER OLDWORLD ALPHAVIRUSES

Animal modelsEncephalitis due to alphaviruses is associated withthe neurotropism of the virus. The disease is morelikely to be fatal in young children and the elderly,and this age effect may be particularly evident forthose viruses that are not considered highly neuro-tropic and/or neurovirulent. For example, moststrains of the Sindbis virus (SINV), another alpha-virus that, like CHIKV, is usually associated withfever and polyarthritis, have been shown to causefatal encephalomyelitis in young mice by inducingapoptosis of immature neurons. This is associatedwith efficient virus replication and rapid death of

the animal. As neurons mature, they become moreresistant to apoptosis, but cell death can still becaused by neurovirulent strains of SINV throughnecrotic rather than apoptotic processes [35].These findings are consistent with reports that, inmouse models, young age is a risk factor for severedisease involving the CNS in animals infectedwith CHIKV. In particular, viral disseminationand disease severity are strongly increased duringthe neonatal period. Another critical factor influen-cing viral replication and leading to severe diseaseis a defective type-I IFN signalling, thoughwhether this also applies to humans remains tobe determined [36].The Semliki Forest virus (SF), which is in the

same antigenic complex as CHIKV and whose dis-ease potential remains to be fully elucidated, hasalso been associated with fatal encephalitis. Inthe mouse, following systemic inoculation, SFV isefficiently neuroinvasive and CNS infection leadsto perivascular lesions of immune-mediateddemyelination, whereas intracerebral inoculationresults in selective and widespread infection ofthe major white matter tract of the brain, the cor-pus callosum [37]. Mature oligodendrocytes arethe predominantly infected cell type, and earlyvirus-induced necrotic death of infected cells is fol-lowed by apoptotic death of adjacent uninfectedcells, a strong inflammatory response and myelinloss.The neurotropism of CHIKV has also been stu-

died in animal models, but the findings are, tosome extent, inconsistent, especially with regardto the involvement of brain parenchyma. CHIKVappears to cause histological changes, such asnecrosis of the neurons in the cerebral cortex andspinal cord in suckling mice. Electron microscopystudies of suckling white mice, which were inocu-lated intracerebrally with CHIKV, revealed thepresence of two size groups of particles in boththe nucleus and cytoplasm of damaged brain cells[38]. Ultrastructural changes in the cerebral cortexand spinal cord were also found with electronmicroscopy in the CNS of newborn mice withthe African strain of the virus. Electron-dense par-ticles were seen in the extracellular spaces of theneuropil and within axon fibres, and alterationsof the endoplasmic reticulum, which appearsto be the main organelle involved in virusreplication, were observed in neurons and glialcells [39].



There is human clinical evidence that CHIKVmay spread to the CNS in the case of severe infec-tion, especially in neonates and young children orin older individuals with underlying conditions.CHIKV dissemination to the CNS does not corre-spond to non-specific spreading due to an over-whelming viral multiplication, and all strains ofthe virus exhibit the same ability to reach theCNS. However, studies using the mouse modelsuggest that CHIKV is not intrinsically encephalo-genic [36]. In fact, CHIKV has not been detected inthe brain microvessels or parenchyma but reachesthe CNS exclusively via the choroid plexus routeand undergoes a step of viral amplification at theependyma and leptomeningeal levels. Consis-tently, CHIKV has been detected in CSF in patientsaffected by severe CHIKV disease associated withCNS disease [36,40]. Thus, according to mousemodels, CHIKV appears to have a marked tropismfor the meninges (i.e., leptomeningeal tissues sharea common mesenchymal origin with peripheralfibroblasts, which are one of the main targets ofCHIKV), whereas it does not infect the brainmicrovessels or parenchyma. Overall, these find-ings suggest that CHIKV does not invade the brainparenchyma nor infect neurons [36]. By contrast,not only the American encephalitic alphaviruses,but also the more closely related SF virus, targetthe brain endothelium and also infect neurons[41,42].

Human autopsiesVery few brain autopsies are available frompatients who died of CHIKV-associated CNS dis-ease. In one case of encephalomyeloradiculitis,the brain was swollen and presented subarachnoidhaemorrhages upon gross examination. Micro-scopy revealed perivascular lymphocytic infil-trates, predominantly in the basal ganglia, similarto those seen in WNV encephalitis [31,43]. Smallfoci of demyelination in the subcortical white mat-ter were also seen. All of these findings are non-specific [31].

TO WHAT EXTENT IS THE NERVOUSSYSTEM INVOLVED IN CHIKV INFECTION?There are several similarities between CHIKV andother re-emerging arboviruses affecting the CNS,such as WNV. Firstly, both viruses have recentlyexpanded their activity range well beyond their

original geographic niche. Secondly, like CHIKV,WNV, first identified in Uganda in 1937, was initi-ally considered to cause only a mild febrile illness.Although WNV was classified for a long time as‘neurotropic’ because of its properties in mice[44], human cases of encephalitis were first docu-mented only in 1952 [45]. Whether the late recogni-tion of neurological consequences of WNV feverwas due to an increased neurovirulence of new cir-culating strains, to the higher probability of identi-fying CNS cases when a larger populationbecomes infected, or merely to improved diagnos-tics, is unclear. This is similar to the experiencewith CHIKV, in that the involvement of the ner-vous system has only recently been fully recog-nised.Another consideration regards the risk of devel-

oping neurological disease in CHIKV, comparedto other arboviral infectious that are traditionallyconsidered as neurotropic. Actually, the ratiobetween cases with apparent infection (i.e. caseswith acute CNS disease) and inapparent infectionappears to vary greatly for arboviral infections,from about 1:20 for EEE to 1:1000 for WEE [46].With regard to CHIKV, little information is avail-able from community-based studies, and datafrom hospital-based series of patients tend to over-estimate the rate of neurological manifestations. Inthe Italian outbreak, one case of encephalitis wasdetected among 205 confirmed cases of CHIKV[47]. The ratio of approximately 1:200 appears tobe lower than that reported for other viruses,such as WNV (i.e.< 1:100) [1] yet higher thanthat of several ‘neurotropic’ viruses, such asWEE and the Japanese encephalitis virus (appar-ent-to-inapparent case ratio of about 1:1000), forwhich CNS disease is the predominant clinicalmanifestation [46].

CONCLUSIONSThe recent outbreaks of CHIKV, which haveoccurred in countries with high level health facil-ities, have highlighted the frequency and severityof CNS involvement in this viral infection. In par-ticular, there is increasing evidence of the occur-rence of encephalitis and meningoencephalitisassociated with CHIKV infection. As with otherarboviral infections, neurological manifestationsare more likely to occur among children and olderindividuals, though the neurovirulence of CHIKVappears to be limited.



Several gaps in our knowledge of CHIKV-related CNS disorders still need to be addressed,concerning the limited availability of histology,brain imaging and electron microscopy studies inhumans affected by CHIKV-associated CNS dis-ease. Clinical descriptions of neurological symp-toms are also limited and sometimes confusing(i.e. lack of clear criteria to distinguish betweenencephalitis and encephalopathy, etc.). Findingsfrom neuropathology studies are not conclusive:for example, whether CHIKV may infect onlymeningeal and ependymal cells but not brain par-enchyma cells (i.e. neurons and glial elements), assuggested by some mouse models, remains contro-versial. In conclusion, there is clinical evidence ofthe involvement of the CNS in CHIKV infection,especially in infants (with particular regard tothe neonatal period) and older individuals withunderlying disease; however, further research isneeded to better define the clinical spectrum ofthe nervous system involvement in CHIKV infec-tion and the neurotropism, neuroinvasivenessand neurovirulence of the virus.

REFERENCES1. Johnson RT. Emerging viral infections of the nervous

system. J Neurovirol 2003; 9: 140–147.2. Ross RW. The Newala epidemic. III. The virus: isola-

tion, pathogenic properties and relationship to theepidemic. J Hyg (Lond). 1956; 54: 177–191.

3. Zuckerman AJ, Banatvala JE, Pattison JR, GriffithsPD, Schaub BD. Principle and practice of Clinical Virol-ogy, 5th edn. John Wiley & Sons, Ltd: West Sussex,England; 2005.

4. Pialoux G, Gauzere B-A, Jaureguiberry S, Strobel M.Chikungunya, an epidemic arbovirosis. Lancet 2007;7: 319–327.

5. World Health Organization.Chikungunya and den-gue, south-west Indian Ocean. Wkly Epidemiol Rec2006; 81: 105–116.

6. Charrel RN, de Lamballerie X, Raoult D. Chikungu-nya outbreaks––the globalization of vectorborne dis-eases. N Engl J Med 2007; 356: 769–771.

7. Rezza G, Nicoletti L, Angelini R, et al. Infectionwith Chikungunya virus in Italy: an outbreakin a temperate region. Lancet 2007; 370: 1840–1846.

8. Yergolkar PN, Tandale BV, Arankalle VA, et al.Chikungunya outbreaks caused by African geno-type, India. Emerg Infect Dis 2006; 12: 1580–1583.

9. Parola P, de Lambellerie X, Jourdan J, et al. NovelChikungunya virus variant in travelers returning

from Indian Ocean islands. Emerg Infect Dis 2006;12: 1493–1499.

10. Economopoulou A, Dominguez M, Helynck B, et al.Atypical Chikungunya virus infections: clinical man-ifestations, mortality and risk factors for severe dis-ease during the 2005–2006 outbreak in Reunion.Epidemiol Infect 2008; 11:1-–8 [Epub ahead of print].

11. Griffin DE. Arboviruses and the central nervous sys-tem. Springer Semin Immunopathol 1995; 17: 121–132.

12. Dropulic B, Masters CL. Entry of neurotropic arbo-viruses into the central nervous system: an in vitrostudy using mouse brain endothelium. J Infect Dis1990; 161: 685–691.

13. Johnson KP, Johnson RT. California encephalitis. II.Studies of experimental infection in the mouse.J Neuropathol Exp Neurol 1968; 27: 390–400.

14. Johnson RT. Virus invasion of the central nervoussystem: a study of Sindbis virus infection in themouse using fluorescent antibody. Am J Pathol1965; 46: 929–943.

15. Hurwitz ES, Schell W, Nelson D, Washburn J,LaVenture M. Surveillance for California encephali-tis group virus illness in Wisconsin and Minnesota,1978. Am J Trop Med Hyg 1983; 32: 595–601.

16. Longshore WA, Stevens IM, Hollister AC, GittelsohnA, Lennette EH. Epidemiological observations onacute infectious encephalitis in California with spe-cial reference to the 1952 outbreak. Am J Hyg 1956;63: 69–86.

17. Earnest MP, Goolishian HA, Calverley JR, Hayes RO,Hill HR. Neurologic, intellectual and psychologicsequelae following western encephalitis: a follow-up study of 35 cases. Neurology 1971; 21: 969–974.

18. Przelomski MM, O’Rourke E, Grady GF, Berardi VP,Markley HG. Eastern equine encephalitis in Massa-chusetts: a report of 16 cases, 1970–1984. Neurology1988; 38: 736–739.

19. Monath TP, Tsai TF. St. Louis encephalitis: lessonsfrom the last decade. Am Trop Med Hyg 1987: 37(Suppl): 40S–59S.

20. Peters CJ, Dalrymple JM. Alphaviruses. In Field’s Vir-ology, 2rd edn, Fields BN, Knipe DM (eds). RavenPress: New York, 1990.

21. Strauss JH, Strauss EG. Viruses and Human Disease,2nd ed. Elsevier: Oxford, UK; 2008.

22. Nimmannitya S, Halstead SB, Cohen SN, MargiottaMR. Dengue and Chikungunya virus infection inman in Thailand 1962–1964. Am J Trop Med Hyg1969; 18: 954–971.

23. Thiruvengadam KV, Kalyanasundarami V, RajgopalJR. Clinical and pathological studies on Chikungu-nya fever in Madras city. India J Med Res 1965; 53:729–744.

24. Mazaud R, Salaun JJ, Montabone H, Goube P,Bazillio R. Acute neurologic and sensorial disorders



in Dengue and Chikungunya fever. Bull Soc PatholExot Filiales 1971; 64: 22–30.

25. Precious SW, Webb HE, Bowen ETW. Isolation andpersistence of Chikungunya virus in cultures ofmouse brain cells. J gen Virol 1974; 233: 271–279.

26. Robin S, Ramful D, Le Seach F, Jaffar-Bandjee MC,Rigou G, Alessandri JL. J Child Neurol 2008; 23:1028–1035.

27. Ernould S, Walters H, Alessandri J-L, et al. Aspectspediatriques de l’epidemie de Chikungunya 2005–2006 a Saint-Denis, ile de la Reunion. Arch Pediatr2008; 15: 253–262.

28. Rampal SM, Meena H. Neurological complicationsin Chikungunya fever. J Assoc Physicians India 2007;55: 765–769.

29. Gerardin P, Barau G, Michault A, et al. Multidisci-plinary prospective study of mother-to-child Chi-kungunya virus infections on the island of LaReunion. PLoS Med 2008; 5: e60.

30. Wielanek AC, Monredon JD, Amrani ME, Roger JC,Serveaux JP. Guillain-barre syndrome complicatingChikungunya virus infection. Neurology 2007; 69:2105–2107.

31. Ganesan K, Diwan A, Shankar SK, Desai SB, SainaniGS, Katrak SM. Chikungunya encephalomyeloradi-culitis: report of 2 cases with neuroimaging and 1case with autopsy findings. AJNR Am J Neuroradihl2008; 29: 1626–1637.

32. Singhi SS, Manimunda SP, Suganan AP, Sahina,Vijayachari P. Four cases of acute flaccid paralysisassociated with Chikungunya virus infection. Epide-miol Infect 2008; 136: 1277–1280.

33. Ramful D, Carbonnier M, Pasquet M, et al. Mother-to-child transmission of Chikungunya virus infec-tion. Pediatr Infect Dis J 2007; 26: 811–815.

34. Sissoko D, Malvy D, Giry C, et al. Outbreak of Chi-kungunya fever in Mayotte, Comoros archipelago,2005–2006. Trans R Soc Trop Med Hyg 2008; 102:780–786.

35. Griffin DE. Neuronal cell death in Alphavirus ence-phalomyelitis. Curr Top Microbiol Immunol 2005; 289:57–77.

36. Couderc T, Chretien F, Schilte C, et al. A mousemodel for Chikungunya: young age and inefficienttype-I interferon signalling are risk factors for severedisease. PLoS Pathog 2008; 4 (2): e29.

37. Fazakerley JK, Cotterill CL, Lee G, Graham A. Virustropism, distribution, persistence and pathology inthe corpus callosum of the Semliki Forest virus-infected mouse brain: a novel system to studyvirus-oligodendrocyte interactions. Neuropathol ApplNeurobiol 2006; 32: 397–409.

38. Chatterjee SN, Sarkar JK. Electron miscroscopy stu-dies of suckling mouse brain cells infected with Chi-kungunya virus. Indian J Exp Biol 1965; 3: 227–234.

39. White JD. Electron microscopy of Chikungunya virusinfection in the nervous system of suckling mice. DefenseTechnical Information Centre: Fort Detrick, MD;1969.

40. Grivard P, Le Roux K, Laurent P, et al. Molecular andserological diagnostic of Chikungunya virus infec-tion. Pathol Biol 2007; 55: 490–494.

41. Soilu-Hanninen M, Eralinna JP, Hukkanen V, et al.Semliki Forest virus infects mouse brain endothelialcells and causes blood brain barrier damage. J Virol1994; 68: 6291–6298.

42. Pathak S, Webb HE. Possible mechanisms for thetransport of Semliki Forest virus into and withinmouse brain. An electron microscopic study. J NeurolSci 1974; 23: 175–184.

43. Ali M, Safriel Y, Sohi J, et al. West Nile virus infec-tion: MR imaging findings in the nervous system.AJNR Am J Neuroradiol 2005; 26: 289–297.

44. Smithburn KC, Hughes TP, Burke AW, Paul JH. Aneurotropic virus isolated from the blood of a nativeof Uganda. Am J Trop Med Hyg 1940; 20: 471–492.

45. Southam CM, Moore AE. Induced virus infections inman by the Egypt isolates of West Nile virus. Am JTrop Med Hyg 1954; 3: 19–50.

46. Johnson TR. Viral infections of the nervous system, 2ndedn. Lippincott-Raven: Philadelphia, PA; 1998.

47. Casolari S, Briganti E, Zanotti M, et al. A fatal case ofencephalitis associated with Chikungunya virusinfection. Scand J Infect Dis 2008; 40: 995–996.



Documents

Chikungunya and the nervous system: what we do and do not know