Autoimmune encephalopathies in children

Abstract

Immune-mediated disorders affecting children's brains have been regarded as a rather rare, if intriguing, group of diseases. Demyelinating conditions, Sydenham chorea, opsomyoclonus, and their associations have been well recognized for decades. In the meantime, in adult neurology multiple syndromes such as such as limbic encephalitis, Morvan syndrome, or Ophelia syndrome have been delineated, mostly as paraneopastic phenomena, with a host of associated antibodies. In addition, individual antibodies are being associated with a wider range of phenotypes, for example aquaporin 4 with brainstem encephalitis as well as neuromyelitis optica. However, over the last few years it has also become apparent that a much wider range of previously unexplained problems in children may have an autoimmune basis and might be treatable.1 One major difference from adults is that in children few are paraneoplastic. Opsomyoclonus and ROHHAD (rapid-onset obesity with hypothalamic dysfunction, hypoventilation and autonomic dysregulation) are associated with neuroblastoma, while limbic encephalitis can be linked with ovarian teratoma or other tumours. However, most limbic encephalitis, especially in younger children, is isolated. Recognized triggers include infection or immunization. Pre-existing autoimmune disorders such as diabetes mellitus, or immunodeficiency states, are also risk factors. Many autoimmune syndromes include very striking behavioural changes. This has been well recognized in, for example, Sydenham chorea or opsomyoclonus, but not necessarily as strongly emphasized as it should be. There is usually one or a combination of symptoms such as hallucinations, withdrawal, anxiety, disinhibition, and/or psychosis, which may be associated with mutism, eating problems such as hyperphagia, and sleep disorders, typically insomnia.1 As a result affected children may initially present to psychiatrists. If they go on to develop more obvious neurological features such as movement disorders, seizures, an impaired conscious level, or cognitive regression the possibility of an organic cause will be recognized. Changes in endocrine function, hypoventilation, autonomic features such as tachycardia or hypertension, and incontinence may not be so immediately obvious. Behaviour changes of this nature in a child or young person with a pre-existing psychiatric diagnosis may be particularly difficult to discern as organic. Similarly now that some children are also recognized to only have isolated organic symptoms (see the case report by Hacohen et al in this issue) it is clear that autoimmune disorders enter the differential of a wide range of acute or subacute neurological and psychiatric presentations. Some conditions are now more clearly proved to be autoimmune, due mainly to advances in antibody assay methodology. Some proven viral encephalitides, especially herpes simplex, can follow a biphasic course with a short-lived improvement after the initial illness, then recurrence of encephalopathic features and often a prominent choreo-athetoid movement disorder followed by severe impairment. This was traditionally considered to be due to reactivation of the viral infection and led to protocols with prolonged courses of antiviral therapy. It now appears more likely to be immune mediated with the rather counter-intuitive implication that immune suppression is required for treatment.2 Cases of suspected encephalitis lethargica are now recognized to be forms of anti-NMDA antibody, and more recently anti-dopamine 2 receptor antibody-mediated encephalitis.3 Some acute or subacute epileptic syndromes are strongly suspected to have an autoimmune basis.4 That narcolepsy might be immune mediated is supported by the association with HLA DQ B1*0602 and more recently with a specific influenza immunisation.5 A wide variety of other syndromes such as Rasmussen encephalitis, acute epileptic encephalopathy with multiple acronyms (FIRES, DESC etc.), Tourette syndrome, PANDAS, Klein-Levin syndrome, or childhood disintegrative disorder, some of which have long been suspected to have a possible immune mediated pathology, are being reviewed. There are even hints that features suggesting autistic regression may occasionally be immune mediated. The importance of promptly recognizing autoimmune causes is not just that of diagnostic accuracy, and avoiding unnecessary investigations or treatment, but that many respond to a variety of immunosuppressive regimens such as steroids, drugs like azathioprine, intravenous immunoglobulin, plasma exchange, and some monoclonal antibodies. Not all children respond and those that do may not recover fully, while others may remit spontaneously, but there is also emerging data that early treatment may lead to a better outcome. Actively considering these treatable conditions is now essential in children with many different presentations.