Dengue is the second most common mosquito-borne disease affecting human beings. In 2009, WHO endorsed new guidelines that, for the first time, consider neurological manifestations in the clinical case classification for severe dengue. Dengue can manifest with a wide range of neurological features, which have been noted—depending on the clinical setting—in 0·5—21% of patients with dengue admitted to hospital. Furthermore, dengue was identified in 4—47% of admissions with encephalitis-like illness in endemic areas. Neurological complications can be categorised into dengue encephalopathy (eg, caused by hepatic failure or metabolic disorders), encephalitis (caused by direct virus invasion), neuromuscular complications (eg, Guillain-Barré syndrome or transient muscle dysfunctions), and neuro-ophthalmic involvement. However, overlap of these categories is possible. In endemic countries and after travel to these regions, dengue should be considered in patients presenting with fever and acute neurological manifestations.
Clinical suspicion is essential for diagnosis of dengue because many symptoms are non-specific. Various methods are available for laboratory confirmation. During the first days of infection, dengue virus is present in blood; thus, at that time, detection of NS1 antigen or RNA by RT-PCR and viral culture are appropriate diagnostic methods.1 Dengue virus-specific IgM antibodies are present in serum samples 3—10 days after disease onset.1 IgM capture (MAC)-ELISA is the most widely used serological test. Antibodies against other flaviviruses (eg, Japanese encephalitis, West Nile virus, yellow fever) might cross-react with dengue virus, leading to false-positive reactions.136
In endemic countries, or among travellers who recently (<14 days) returned from such regions, dengue should be ruled out in patients with fever and neurological features (panel 2). If possible, lumbar puncture should be done and CSF analysed for abnormalities and for dengue virus-specific antibodies, NS1 antigen, or dengue virus RNA, depending on available laboratory facilities. Differential diagnosis in patients with febrile encephalopathy includes malaria, tuberculosis, leptospirosis, rickettsial infection, and other bacterial or viral diseases (caused by, for example, Japanese encephalitis, West Nile virus, or herpes simplex virus [HSV]), depending on the local epidemiology. In a prospective hospital-based study in Vietnam, most children with acute encephalitis of presumed viral origin were infected with Japanese encephalitis (26%), followed by enteroviruses (9%) and dengue virus (5%).30 In adults and adolescents in Brazil, dengue was the leading cause of viral encephalitis (47%), followed by infections with HSV-1.31
To differentiate dengue encephalitis from encephalopathy, detection of dengue virus, NS1 antigen, or dengue virus-specific IgM antibodies in CSF is helpful. Nevertheless, sensitivity of serological techniques can be low. Dengue virus-specific IgM antibodies have been recorded in CSF of 22—33% of patients diagnosed with dengue encephalitis (Table 1, Table 2).90Detection of dengue virus in CSF could be hampered by low sensitivity of RT-PCR in CSF, compared with findings in serum, because of a lower viral load.137 Moreover, measurement of IgM antibodies in CSF might not be a reliable diagnostic marker of dengue CNS involvement, owing to low titres in CSF.138 Abnormalities in CSF—such as lymphocytic pleocytosis—support the diagnosis of dengue encephalitis, but they are not always present (Table 1, Table 2, Table 3). A mild increase in CSF protein has been recorded.28 In a series of patients with neurological complications of dengue, four of seven with encephalitis had no alterations in CSF.90 Therefore, normal CSF cellularity should not exclude dengue encephalitis.
The case definitions in panel 2 are designed to be used epidemiologically and clinically and to guide diagnosis and prognosis. Although we propose criteria for a classification scheme, a topic as challenging and as controversial as dengue encephalitis needs to be addressed in a standard way. Prospective studies are needed to assess the specificity and sensitivity of the proposed case definitions and to generate supporting evidence. Cases fulfilling neither the definition for encephalitis nor that for encephalopathy—eg, without CSF testing or when categories are overlapping—can be categorised as other or non-specified dengue CNS involvement.
Neuroimaging might provide additional clues in the diagnosis of neurological complications of dengue. In dengue encephalitis, brain MRI can be normal or show focal parenchymal abnormalities.22, 41 Nevertheless, no specific MRI findings suggestive of dengue encephalitis have been reported. Neuroimaging features of patients with dengue are diverse, with cerebral oedema the most commonly reported finding.77 Meningeal enhancement on post-contrast MRI has been reported occasionally as well.77
Finally, EEG abnormalities can be seen in dengue patients with neurological complications. In a study of 23 patients with dengue virus infection and neurological symptoms, EEG abnormalities were recorded in 12 people.139 Slowing on EEG can be seen, but this finding is unspecific and could be attributable to seizures, intracranial haemorrhage, and viral infection per se, besides encephalopathy.77
Currently, no effective antiviral agents are available to treat symptomatic dengue virus infection.140 Therefore, management remains supportive. In mild cases, antipyretic drugs and oral fluids could be useful. Acetyl-salicylic derivatives and other non-steroidal anti-inflammatory drugs should be avoided. Management of haemorrhagic complications should be initially conservative. Precise management of intravenous fluids is needed, and blood or platelet transfusion is only necessary when severe bleeding takes place.1
In patients with severe dengue and signs of plasma leakage, prompt fluid resuscitation is imperative, with close monitoring of packed-cell volume to avoid fluid overload. Isotonic crystalloid solutions should be used, with isotonic colloid solutions reserved for patients presenting with profound shock or those who do not have a response to initial crystalloid treatment.140, 141 In a randomised controlled trial from Vietnam,142 use of oral prednisolone during the early acute phase of dengue infection was not associated with a reduction in the development of shock or other recognised complications of dengue virus infection.
For supportive management of patients with neurological manifestations, possible underlying causes such as intracranial bleeding, liver failure, hyponatraemia, hypokalaemia, or metabolic acidosis should be ruled out and—if possible—corrected. Management of dengue encephalitis remains supportive and should include adequate hydration, nutrition, monitoring of consciousness, and maintenance of airways.143 Symptomatic seizures should be treated with non-hepatotoxic anticonvulsants. Decompressive craniotomy and cerebral haematoma evacuation were done in two patients with dengue after correction of prothrombin time and platelet count.92 Nevertheless, prognosis is not good and, in one case series, two of five patients died.92 At this moment, haematoma surgery cannot be proposed as a routine treatment for dengue virus intracranial bleeding.
Some clinicians recommend treatment of immune-mediated dengue CNS involvement with pulses of intravenous methylprednisolone for several days.50, 69, 72 However, up to now, no randomised controlled trial has been undertaken to show the efficacy of this approach in patients with dengue myelitis or acute disseminated encephalomyelitis. High doses of intravenous immunoglobulin might be useful to treat post-dengue Guillain-Barré syndrome. Supportive treatment—including hydration and analgesic drugs—is used for myalgia and transitory muscle dysfunction. The effectiveness of corticosteroids in dengue myositis remains to be proven.
No treatment has been approved for neuro-ophthalmic manifestations of dengue. Steroids have been administered previously because of possible underlying immune mechanisms, although up to now no randomised trials have been done. Topical steroids have been used to treat anterior uveitis, whereas pulsed intravenous methylprednisolone or systemic oral steroids might be indicated for extensive retinal vasculitis.120
Currently, no vaccine is available for protection against dengue. However, several vaccine candidates are in development.
Conclusions and future research
Dengue should be included in the differential diagnosis of acute febrile disease with neurological manifestations in dengue-endemic countries and in patients with a recent travel history to an endemic region. Many neurological manifestations of dengue have been recorded, ranging—with substantial overlap—from encephalitis and encephalopathy to immune-mediated syndromes and muscle involvement. Recent evidence suggests that dengue virus has neuroinvasive capacity. In several studies in endemic areas, a large proportion of viral encephalitis was caused by dengue virus.26—32 However, even though CNS involvement is included now as a criterion for severe dengue in the 2009 WHO case classification,1 no standardised case definitions or diagnostic criteria for dengue encephalitis or encephalopathy have been agreed, which leads to inconsistent use of these terms in published work.
An updated WHO dengue guideline should include a case definition for dengue encephalitis and encephalopathy, to guide clinicians and clinical epidemiological researchers into this topic. A case classification—such as the one proposed in panel 2—could serve as a starting point, which could be reviewed by WHO, agreed by consensus and best available current evidence, and refined as additional data become available from prospective studies. For this reason, assessment of CSF in patients with suspected neurological manifestations of dengue should be standardised. Very few published reports present findings of CSF testing for dengue virus, dengue virus-specific IgM antibodies, or NS1 antigen combined with CSF cellularity and confirmation of dengue in serum samples in a consistent way. Further epidemiological and neuropathological studies are needed to ascertain the true incidence and burden of neurological complications of dengue, to elucidate the underlying pathophysiology, and to assess the sensitivity and specificity of diagnostic markers for dengue encephalitis.