Genetic basis for acute necrotizing encephalopathy of childhood

Acute necrotizing encephalopathy of childhood (ANE) is a distinct form of acute encephalopathy triggered by respiratory infections, particularly viral infections such as Influenza A and B.1 It was initially described as a novel disease entity by Mizuguchi et al.2 in 1995. The condition affects previously healthy children aged 5 months to 10 years, who present with a febrile respiratory illness followed within 3 days by features of an acute encephalopathy including recurrent vomiting, seizures, altered consciousness rapidly progressing to coma, hepatomegaly with abnormal liver function, elevated cerebrospinal fluid protein levels, and diffuse slow waves on electroencephalogram in the acute stage. The hallmark of ANE is the neuroradiological findings which take the form of multiple, symmetric lesions on T2-weighted magnetic resonance imaging (MRI) or low density areas on computed tomography (CT) affecting the thalami, posterior limb of the internal capsule, periventricular white matter, the tegmentum of the midbrain and pons, and around the dentate nucleus of the cerebellum. These changes can be seen within 12 hours after the onset of coma and are associated with cerebral oedema. ANE is known to have a high mortality and survivors are often left with neurological and cognitive sequelae. Although ANE has mainly been reported from Japan, it is being increasingly recognized in other countries.3 In 2003, Neilson et al. 4 reported a large multigenerational family with autosomal dominant ANE. Half of the affected individuals in this family presented with recurrent episodes of ANE that produced more severe neurological sequelae. The condition demonstrated incomplete penetrance (estimated to be about 40%). The authors subsequently mapped this disorder to a 6.5 Mb region on chromosome 2 (2q12.1–2q13).5 In January 2009, the authors identified a missense mutation (c.1880C>T; p.Thr585Met) in the RANBP2 gene at 2q13 as the cause of recurrent ANE in this family.6 The authors also identified the same RANBP2 mutation in nine of 15 additional families with suspected familial or recurrent ANE and different missense mutations (c.2085C>T; p.Thr653Ile and c.2094A>G; p.Ile656Val) in two of these families. No mutation could be identified in the RANBP2 gene in the remaining four families. In one of these families linkage to RANBP2 was excluded by linkage analysis, and in another family affected siblings with parental consanguinity and the absence of affected individuals in previous generations raised the possibility of autosomal recessive inheritance. The authors also looked for mutations in the RANBP2 gene in 19 patients with isolated ANE but did not find any mutations. The authors proposed the term ANE1 to refer to patients with ANE who had an associated RANBP2 mutation. Gika et al.7 report a 9-year-old female with recurrent ANE who was heterozygous for the recurrent RANBP2 mutation, c.1880C>T (p.Thr585Met). This was one of the families included in the seminal paper by Neilson et al.6 that were found to have the recurrent RANBP2 mutation. Publication of the detailed phenotypes in this family is both interesting and instructive. The affected child presented with ANE at the age of 9 months following a febrile illness from which she made a slow but complete recovery. At 2 years of age she presented with acute onset of right-sided abducens nerve palsy following a viral illness, from which she made a slow recovery. CT during this episode was normal. At the age of 9 years the patient again developed ANE following an episode of Influenza A. This time she remained in coma for several months and was left with severe cognitive and neurological sequelae. The patient’s mother had an episode of encephalitis/polyneuritis following a viral illness at the age of 19 and was left with persistent foot drop. She was also found to be heterozygous for the RANBP2 mutation that was identified in her daughter. What does this report tell us? It makes us aware that ANE may, in some instances, represent a familial disorder or may be the result of an inherited predisposition due to a mutation in the RANBP2 gene. It also broadens the phenotype of neurological problems that can be seen in individuals that harbour a RANBP2 mutation. In addition to viral illnesses resulting in the phenotype of ANE in these individuals, they may result in a less severe phenotype of mononeuritis or polyneuritis affecting not only the peripheral nerves but also the cranial nerves. An important question that emerges is which patients with ANE should be tested for RANBP2 mutations? Clearly, any patients with ANE that have a family history of this condition (parent, sibling, or grandparent) or an ‘encephalitic’ illness should be screened for RANBP2 mutations. All patients with recurrent ANE should also be tested. The report by Gika et al. suggests that patients with ANE that have other acute-onset neurological problems following a viral illness such as a cranial nerve palsy or polyneuritis may also be candidates for RANBP2 mutation analysis. Are there any differences in the phenotype of ANE1 and isolated ANE? Neilson et al.6 show that patients with ANE1 do not have elevated serum transaminase levels, which can be seen in the majority of patients with isolated ANE, and they also show that patients with ANE1 can present with disease onset in adolescence or adult life whereas isolated ANE mainly affects young children. Neilson et al.6 also suggest that the MRI findings may help to differentiate ANE1 from isolated ANE. Patients with ANE and RANBP2 mutations have involvement of the external capsule, extreme capsule, claustrum, medial temporal lobes, amygdale, hippocampi, and spinal cord in addition to the typically affected areas of the brain. In a patient with ANE without evidence of family history or recurrent disease, these findings may help the astute clinician to consider RANBP2 mutation analysis. There are undoubtedly some key questions that still require answers. These include explanations for the incomplete penetrance of RANBP2 mutations, and how mutations in RANBP2, which encodes a nuclear pore protein with many different roles, predispose to ANE. The work of Neilson et al.6 provides evidence that mutations in genes other than RANBP2 can also cause familial ANE and recurrent ANE. The discovery of these genes may help to address these question.