Encephalitis is considered a severe manifestation of dengue virus infection, and found in the three classical disease groups. Diagnosis of dengue encephalitis is based on criteria proposed by Soares and Marzia (2014).

However, normal CSF cellularity cannot exclude dengue encephalitis. According to Soares et al., dengue is the first reason for encephalitis with normal CSF cellularity in 75% of patients with viral meningitis and encephalitis in a dengue endemic region, followed by Herpes Simplex Virus 1 (HSV1) and mild encephalitis (Soares et al., 2011). Neuroimaging of dengue encephalitis yields divergent data, with normal findings in most cases (Baldaçara et al., 2013; Madi et al., 2014). In case of abnormal neuroimaging findings, magnetic resonance imaging (MRI) has advantages over cranial computed tomography (CT) in revealing cerebral lesions in dengue encephalitis. However, changes are usually non-specific (Garg et al., 2017). Decisive characterizations of MRI properties in dengue encephalitis remain undefined (Mathew et al., 2015). Treatment is nonspecific, with mostly symptomatic treatment provided. Most patients have good recovery.

Encephalopathy caused by dengue fever can be reflected by reduced sensitivity, cognitive impairment, convulsions, and personality and behavior disorders, including acute mania, depression, emotional lability, anxiety, psychosis, and agoraphobia (Rittmannsberger et al., 2010; Baldaçara et al., 2013; Srivastava et al., 2013). In a review reporting cases of dengue fever associated with neurological disorders in 2012, encephalopathy was considered by far the most encountered complication (Oehler et al., 2012). Most encephalopathy cases occur in children of developing countries, and do not show CSF abnormalities (Angibaud et al., 2001). Dengue associated encephalopathy is generally very serious, with around 50% of the affected patients succumbing (Angibaud et al., 2001).

In the past, encephalopathy was considered to be exclusively associated with Dengue hemorrhagic fever/Dengue shock syndrome (DHF/DSS). Brain edema, anoxia, hemorrhage, intense hyponatraemia, liver or kidney failure, release of toxic substances, metabolic acidosis, and direct organ invasion are commonly reported precursors of encephalopathy in patients with serious DHF/DSS (Sumarmo et al., 1983; Nimmanitya et al., 1987; Srivastava et al., 1990; Hendarto and Hadinegoro, 1992; Thisyakorn and Thisyakorn, 1994; Sirivichayakul et al., 2000; Solbrig and Perng, 2015).

A substitution of alanine by a valine residue at position 173 of the envelope glycoprotein was reported in encephalopathy-associated dengue type 2 virus in 1996 (Sistayanarain et al., 1996). However, whether such mutation is directly involved in envelope-receptor interactions has not been clearly elucidated. Burst suppression, seizures, focal patterns, or epilepsia partialis continua can be observed on EEGs (electroencephalographs) of patients with encephalopathy (Kalita and Misra, 2006; Misra et al., 2006; Liou et al., 2008a,b).

Dengue infection associated meningitis is rarely encountered (Soares et al., 2006). In the related literature, three reports about dengue meningitis are found (Soares et al., 2010, 2011; Mamdouh et al., 2013). Soares et al. described a 24-year-old woman with the symptoms of fever, headache, and nuchal rigidity, in whom diagnosis was confirmed by polymerase chain reaction (PCR) and positive results in CSF evaluation (Soares et al., 2010). Two additional cases of dengue meningitis were reported by Mamdouh in 2013. Both cases presented with fever, headache, neck rigidity, and low platelet count in a dengue endemic area. Diagnosis was confirmed by positive IgM in the CSF in the absence of the typical clinical spectrum of infection. CSF cellularity was normal in one case. Brain CT or MRI scan was normal. Dengue meningitis might cause migraine such as headaches with poor response to anti-migraine therapy and common analgesics. The two patients achieved full recovery after several months without any residual neurological deficit (Mamdouh et al., 2013).

Ischemic and hemorrhagic strokes have been reported in a few cases. Liou et al. reported a dengue fever patient who showed thrombocytopenia and ischemic stroke in 2008 (Liou et al., 2008a). In 2013, Rajesh Verma et al. reported a 68-year-old male patient presenting with moderate grade, continuous fever for 15 days; the patient also had sudden onset of weakness of the left half of his body as well as facial asymmetry, 10 days prior to hospital admission. Dengue infection was confirmed and brain MRI revealed acute infarction in the right parietal lobe. After 2 months of medication and physiotherapy, a partial improvement was observed in limb weakness (Verma et al., 2013). In cases with ischemic stroke, meningovasculitis, or a transient hypercoagulable state during dengue infection was postulated as the pathogenetic mechanism (Liou et al., 2008a; Verma et al., 2013).

A summary of intracerebral hemorrhage cases associated with dengue virus infection in various countries from 2001 to 2014 was reported by Vargas-Sánchez et al. (2014). In the above study, patient age ranged from 9 to 68 years, and there were 5 females and 7 males. Outcome was unspecified in one patient, while 7 and 5 patients recovered and died, respectively. Meanwhile, DENV2 was detected in 5 patients; one patient was infected with DENV3 and precision was not provided for the remaining seven. The involved regions included the pontine, basal gangliar, cerebellar, parietal, temporal, and frontal lobes of the brain (Vargas-Sánchez et al., 2014).

Current studies suggest that in dengue virus infection, cytokine overproduction results in immune-mediated endothelial cell damage (Seet and Lim, 2007; Basu and Chaturvedi, 2008). Hemorrhage may be caused by elevated vascular permeability, plasma leakage, and vasculitis due to dengue (Seet and Lim, 2007; Basu and Chaturvedi, 2008).

Cerebellar syndrome was mentioned by Weeratunga et al. (2014b). They described three patients diagnosed with dengue infection based on combined clinical and laboratory findings, fulfilling the WHO criteria in an endemic area. Within 2 weeks of diagnosis, they developed cerebellar symptoms. The manifestations of cerebellar syndrome include bilateral vertical and horizontal nystagmus, dysarthria, bilateral limb, and gait ataxia. Other causes of cerebellar dysfunction were excluded, and all cases were self-limiting. MR brain scans showed cerebellar T2 hyperintense lesions in one case, while the remaining two showed normal signals. The CSF in all cases had normal protein levels and cell count. All cases showed antibodies for dengue in blood and CSF samples. A low-grade inflammatory process was the proposed mechanism (Weeratunga et al., 2014b).

The diagnosis of transversemyelitis (TM) is established by a somewhat sudden onset of sensorimotor and sphincter disturbances resulting from inflammatory demyelinating lesions (Scott et al., 2011). In longitudinally extensive transverse myelitis (LETM), a spinal lesion covers three or more vertebral segments (Wolf et al., 2012). According to Jacob and Weinshenker's opinions, TM can be broadly classified into four categories: (1) demyelination (monofocal clinical isolated syndrome) or multifocal demyelination [acute disseminated encephalomyelitis (ADEM) or multiple sclerosis]; (2) combined systemic connective tissue disease; (3) infection; (4) idiopathic illness (Jacob and Weinshenker, 2008).

Positive findings by spinal MRI are crucial in reaching the diagnosis of TM/LETM. Hyperintensity in T2-weighted signals found in spinal MRI scans support transverse myelitis diagnosis. In addition, inflammatory reactions in the CSF are found in most patients. TM/LETM in dengue is rarely encountered, and no more than 12 case reports are available (Renganathan et al., 1996; Leao et al., 2002; Kunishige et al., 2004; Seet et al., 2006; Chanthamat and Sathirapanya, 2010; Verma et al., 2011a; Larik et al., 2012; Singh et al., 2013; Weeratunga et al., 2014a; Fong et al., 2016; Mo et al., 2016; Mota et al., 2017).

However, the spectrum of TM could be broad, and more attention should be paid to atypical cases. For example, Mota reported a patient confirmed with acute TM without paraparesis following a dengue virus infection last year (Mota et al., 2017).

Treatment with intravenous steroids is useful, and individuals with parainfectious-dengue TM show satisfactory recovery (Fong et al., 2016). Fong et al. reported that a previously healthy 12-year-old girl with parainfectious-dengue TM and concomitant spinal epidural haematoma had good clinical recovery without surgical intervention after 6 months (Fong et al., 2016). It was shown that post-infection autoimmune reactions and direct infection are associated with transverse myelitis (Solomon et al., 2000b; Sindic et al., 2001; de Sousa et al., 2014; Mo et al., 2016).

ADEM is characterized by an acute inflammatory demyelinating ailment affecting the CNS, a monophasic course, and multifocal white matter involvement which occurs during or after dengue virus infection (Puccioni-Sohler et al., 2013). PubMed and Scopus combinely reported only 7 cases prior to May 2017 (Yamamoto et al., 2002; Brito et al., 2007; Gera and George, 2010; Sundaram et al., 2010; Gupta et al., 2013; Viswanathan et al., 2016). ADEM onset occurs within a limited time (averaging 5.6 days) after initial dengue signs (Gupta et al., 2013). An abnormal CSF contributes to ADEM diagnosis, while a normal CSF cannotrule out its possibility (Viswanathan et al., 2016).

Few studies reporting radiologically confirmed dengue associated ADEM are available. MRIs show extensive involvement of the white matter of the frontal, parietal, or temporal lobe, basal ganglia, brainstem, cerebellum, corpus callosal, and periventricular lesions (Puccioni-Sohler et al., 2012, 2013; Gupta et al., 2013; Domingues and Kuster, 2014; Mudin, 2015; Viswanathan et al., 2016). Perivenous demyelination, macrophage influx, and perivascular infiltration of lymphocytes with hemorrhagic foci were reported after histological examination of such lesions (Sundaram et al., 2010).

Dengue related ADEM results from immune reactions (Gupta et al., 2013). In Murthy's study, its pathophysiology was considered a transient autoimmune reaction to myelin or unknown self-antigens (Murthy, 2010). There is no established treatment for ADEM, but the use of steroids is effective during its active phase (Carod-Artal et al., 2013; Gupta et al., 2013; Domingues and Kuster, 2014).

Guillain-Barré syndrome (GBS) features a quickly rising paralysis, reflecting an inflammatory demyelinating or axonal polyneuropathy (Soares et al., 2008). In most cases, a protein-cytological dissociation in the CSF is indicative of GBS (Allos, 1998). Presentations of reduced conduction velocity, conduction blockage in motor nerves, extended distal latency, and prolongation or absence of F responses are often observed by electromyography in GBS patients (Soares et al., 2008).

GBS following dengue virus infection is uncommon (Meena et al., 2011; Qureshi et al., 2012; Chaudhary et al., 2014). In the literature, about 20 case reports of GBS associated with dengue infection are available (Paul et al., 1990; Sainte Foie et al., 1993; Esack et al., 1999; Santos et al., 2004; Sulekha et al., 2004; Kumar and Subhashini, 2005; Soares et al., 2008; Oehler et al., 2012; Carod-Artal et al., 2013; Dupont-Rouzeyrol et al., 2014; Simon et al., 2016). Most cases were pediatric patients, and only few adults were involved (Chew et al., 1998; Gupta P. et al., 2009; Qureshi et al., 2012).

Neurological signs develop at 1–19 days after the onset of dengue. Simon et al. found GBS occurs early in DF with an unusual progression; in their report, average time elapsed from fever to neurological signs was 2 days, ranging between 1 and 3 days (Simon et al., 2016). A patient with Miller Fisher syndrome caused by dengue has been described (Gaultier et al., 2000). Umapat et al. reported that asymptomatic dengue infection may also trigger Guillain-Barré syndrome, and found four GBS cases after dengue diagnosis, including 1 and 3 with acute motor axonal neuropathy and acute inflammatory demyelinating polyneuropathy, respectively (Umapathi et al., 2016).

The exact mechanism of dengue associated GBS remains unclear. It is considered a neurological disease modulated by immunocytes (Chew et al., 1998; Gupta P. et al., 2009). Pro-inflammatory substances such as tumor necrosis factor, interleukins, and the complement may play important roles in GBS pathogenesis (Shah, 2007). Immune reactions induced by dengue virus infection might involve peripheral nerve constituents; for example, myelin or axon may be targeted by the immune response (Carod-Artal et al., 2013).

Plasma exchange seems to be more effective than supportive treatments in the treatment of GBS in randomized clinical trials (Comi et al., 1985; Qureshi et al., 2012). Immunoglobulins administered intravenously show similar effectiveness to plasma exchange, and might be more advantageous (Chew et al., 1998; Gupta P. et al., 2009; Qureshi et al., 2012). Corticosteroids alone do not make a difference, and there is not enough data supporting their usefulness (Comi et al., 1985; Qureshi et al., 2012). Other treatment options, e.g., CSF filtration, are still being assessed (Qureshi et al., 2012).

Dengue myositis diagnosis is based on clinical manifestations of dengue virus infection, positive serum IgM for the dengue virus, high creatine phosphokinase levels, normal CSF, and exclusion of other causes (Paliwal et al., 2011). The clinical spectrum is broad, from mild asymmetrical weakness of lower extremities to sudden progressive severe limb and trunk weakness, and even lung failure (Paliwal et al., 2011). However, muscle weakness is distinctly uncommon in dengue virus infection (Malheiros et al., 1993; Sangle et al., 2010). Doctors should pay more attention to patients suffering from early pulmonary impairment, elevated creatine phosphokinase amounts, and serious myalgia (Paliwal et al., 2011).

The pathogenesis of myositis remains unclear. The proposed mechanisms comprise direct muscular invasion by the virus and immunity associated destruction of muscle fibers, particularly by tumor necrosis factor (Malheiros et al., 1993; Paliwal et al., 2011). Dengue myositis is reflected by perivascular infiltration of mononuclear cells, mitochondrial proliferation, fat accumulation, nuclear centralization, fiber-type grouping, and/or myonecrosis foci on muscle biopsy (Malheiros et al., 1993; Misra et al., 2012).

Normal motor unit potentials are of reduced duration and amplitude, and polyphasic on electromyographic examination. Fibrillations, sharp waves, and complex repetitive discharges do not occur (Kalita et al., 2005; Misra et al., 2012).

However, dengue myositis is considered a relatively benign and self-limiting disease in pediatric reports (Rajajee et al., 2005; Misra et al., 2006; Ahmad et al., 2007; Pimentel et al., 2011). In adult patients, dengue myositis is often more severe, even leading to severe rhabdomyolysis (Kalita et al., 2005; Finsterer and Kongchan, 2006; Siriyakorn and Insiripong, 2015). Sangle et al. reported a 16-year-old girl with dengue shock syndrome showing myositis and myocarditis. Symptomatic treatment and rehabilitation were provided, and she had recovered well at discharge after 1 month of hospitalization (Sangle et al., 2010). Finsterer et al. found dengue can cause persistent and serious myositis, which is resolved after administration of corticosteroids (Finsterer and Kongchan, 2006).

Dengue-associated hypokalemic paralysis was reported by several authors, but most studies were based on isolated case reports and short series (Esack et al., 1999; Santos et al., 2004; Kalita et al., 2005; Jha and Ansari, 2010; Roy et al., 2011; Jain et al., 2014; Maurya et al., 2016). Maurya and colleagues assessed 12 dengue patients with hypokalemic paralysis, who had complete and rapid recovery after potassium supplementation (Maurya et al., 2016).

It has been suggested that elevated creatine phosphokinase amounts result from vasoconstriction and muscular ischemia caused by hypokalemia (Comi et al., 1985). In a study of 12 patients with acute neuromuscular weakness in a dengue epidemic, 10 individuals showed hypokalemia, while Guillain-Barré syndrome and myositis were found in one patient each. Of the latter 2 patients, the one with myositis showed slow improvement and normal serum potassium levels, and the other with Guillain-Barré syndrome recovered within 6 weeks (Hira et al., 2012). In a retrospective study of 7 patients with dengue myositis who had generalized weakness, a mild serum creatin phosphokinase (CPK) elevation was noted in 3 individuals with hypokalemia, whereas three cases with fulminant presentation and respiratory muscle involvement had markedly elevated CPK levels (16,590-117,200 U/L) with normal potassium amounts. All patients completely recovered within 4 weeks (Paliwal et al., 2011).

The severity of weakness may not correlate with potassium levels in Maurya's report (Maurya et al., 2016). Hypokalemia does not always lead to paralysis. In 1,342 patients with dengue fever assessed in China, hypokalemia occurred in up to 28% individuals but weakness was not mentioned (Ying et al., 2007).

Hypokalemic paralysis differs from dengue myositis and idiopathic hypokalemic paralysis according to clinical, biochemical features, and consequences (Maurya et al., 2016). It represents a systemic complication of dengue fever (Murthy, 2010; Jain et al., 2014). The pathogenesis of hypokalemic paralysis in dengue remains obscure. A probable explanation for hypokalemia in dengue is serum potassium redistribution in cells and transient kidney tubular dysfunction increasing urinary potassium elimination (Jha and Ansari, 2010; Jain et al., 2015; Maurya et al., 2016). In addition, infection-related stress may lead to catecholamine or insulin release, which results in an intracellular shift of potassium (Jha and Ansari, 2010; Hira et al., 2012). Supplementation of potassium can achieve satisfactory recovery in patients with hypokalemia paralysis (Jha and Ansari, 2010; Kayal et al., 2013). Jain et al. reported a 30-year-old man with dengue and hypokalemic paralysis alongside hypomagnesemia. Weakness persisted after potassium supplementation, while low serum magnesium levels were detected. However, he completely recovered within 48–72 h, with a normalization of serum potassium and magnesium levels. The possible mechanism may be that muscle Na+ and K+-ATPase activity is inhibited by hypomagnesemia, leading to a reduced ion influx into muscle fibers and secondary kaliuresis (Jain et al., 2015).

Neuritis after dengue infection is believed to rarely occur. However, recent studies have reported more and more dengue fever patients showing uncommon neurological signs. Dengue-associated neuritis, such as brachial neuritis, long thoracic nerve palsy, phrenic nerve palsy, abducens nervepalsy, and peripheral facial palsy have been reported in different areas of the world (Patey et al., 1993; Chappuis et al., 2004; Chien et al., 2008; Verma et al., 2011b; Shivanthan et al., 2012; Mishra et al., 2013; Peter et al., 2013).

Its diagnosis is established, after other reasons for neuritis, such as tumor and demyelinating diseases, infections, trauma, and stroke, are accordingly excluded. Dengue-associated nerve palsy is mostly managed by supportive treatment. Some cases improve even without specific treatment, e.g., with steroids or intravenous immunoglobulins (Peter et al., 2013; Biswas and Pal, 2014), while others have a good response to steroids (Verma et al., 2011b).

The pathogenesis of dengue-associated neuritis is likely related to immune reactions, although the mechanisms remain largely unclear (Carod-Artal et al., 2013). Most patients achieve a remarkable recovery.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.