Acute Disseminated Encephalomyelitis

Acute disseminated encephalomyelitis (ADEM) is a disorder of inflammation and demyelination of the brain and/or spinal cord. It is an established disorder (first description dates back to 1724) with many components not well understood. ADEM is often difficult to distinguish from other disorders of demyelination such as multiple sclerosis (MS). ADEM is a clinical diagnosis and one of exclusion. There is no clear biological marker for the disease, and the diagnosis is made based on clinical presentation often with the assistance of neuroimaging. ADEM is classically described as an isolated episode of demyelination (i.e., monophasic), but reports of multiphasic forms of ADEM have led to significant controversy. Several guidelines exist to facilitate making the diagnosis of ADEM. Distinguishing ADEM from other disorders is important because treatments for the different disease processes vary.

ADEM is an uncommon disease with an incidence in children of 0.07-0.8 per 100,000 persons/year. The mean age of onset in children is approximately 6 years, but the incidence is highest in younger children (age <3 years). Boys may be slightly more affected than girls.

ADEM is an inflammatory demyelinating disorder thought to be secondary to an abnormal immune response to central nervous system antigens. An antecedent infection or vaccination is reported in as many as 70% of ADEM cases. Although ADEM typically follows an infection, the source of the infection is not definitively known in most cases.

Vaccination-associated ADEM is probably less common than infection-associated ADEM. Specific immunizations associated with ADEM include vaccinations for rabies, measles, Japanese encephalitis virus, poliovirus (oral preparation), tetanus, influenza, and recombinant hepatitis B, but others are reported as well.

ADEM is thought to be an autoimmune disorder. No genetic linkages are identified at this time.

The roles of infection and the immune system in ADEM are not entirely understood. Presently, the best animal model of ADEM is the experimental autoimmune encephalomyelitis (EAE). In this model, rodents receive injections of purified myelin components to yield clinical symptoms and a pathologic cellular response similar to demyelinating diseases in humans.

ADEM is thought to occur as a combination of autoimmunity and loss of regulation of the native immune response. It is believed that a form of “molecular mimicry” occurs, in which the preceding infection or vaccine has antigens that resemble self-derived epitopes, leading to inappropriate immune activation against both invading and native structures.

Preceding infection can also disrupt the normal process of prevention of self-immunity by removing regulatory cells that would otherwise have prevented inappropriate autoimmune responses. Finally, some viruses have been shown to damage, or even directly inhibit myelin, leading to identification of myelin as an appropriate immune target.

Children with ADEM present with a fairly rapid onset of symptoms (i.e., evolution of symptoms over days). Encephalopathy is a hallmark of the presentation though in small children encephalopathy may be difficult to appreciate. In mild forms of encephalopathy, a child may be fussy or want to sleep in excess. In more extreme cases, the child may be difficult to arouse or nonresponsive. Other ADEM symptoms include headache, nausea/vomiting, or neck pain (i.e., meningismus). The child may develop weakness presenting as inability to move an arm or leg. Problems with coordination or balance (i.e., ataxia) are also common. Seizures are often a part of the presentation, particularly in younger children. The typical presentation includes a combination of signs and symptoms that correlate with multiple areas of cerebral involvement referred to as multifocal or polyfocal. See Table I.

Table I.

ADEM is a disorder with overlapping features of encephalitis and other demyelinating disorders such as MS. Not uncommonly, children who are initially diagnosed with ADEM go on to have their diagnosis reassigned to an alternative diagnosis later when they experience progression of symptoms and/or recurrence of neurological deficits. See Table II.

Table II.

There is no specific laboratory or radiographic study to make a definitive diagnosis of ADEM. The diagnosis is made based on clinical characteristics, often with the support of neuroimaging. Laboratory studies can be helpful in excluding ADEM mimics such as encephalitis.

Children with ADEM typically present with a 2-5-day history of multifocal symptoms and encephalopathy. The presence of encephalopathy is essential to the diagnosis according to most diagnostic guidelines, though multiple publications include patients diagnosed with ADEM who do not have encephalopathy as part of their presentation. Given the retrospective design of the larger studies available, it is unclear if this variability is due to clinicians failing to consistently appreciate and/or document mild to moderate forms of encephalopathy or a truly broad range of presenting symptoms. A history of infection or vaccine less than 4 weeks prior to presentation is present for most, but not necessary to make the diagnosis.

Children considered for the diagnosis of monophasic ADEM should have no history of similar prior events or other diagnoses suggestive of a more chronic process (e.g., optic neuritis). Children with a single prior event may be considered for the diagnosis of multiphasic ADEM. Multiple prior events are no longer consistent with a diagnosis of ADEM.

Since the clinical appearance of ADEM overlaps considerably with other demyelinating diseases, comprehensive evaluations should include follow-up beyond the acute period. In fact, according to the most recent guidelines, the diagnosis of ADEM should only be applied retrospectively after a patient is seen in follow-up and found to have no new clinical events 3 months after the onset of initial symptoms.

Laboratory studies are useful to exclude other diagnoses that may explain a patient’s presenting symptoms. There is no specific laboratory study available at this time to make the diagnosis of ADEM.

Blood culture should be considered in the infectious work-up, particularly if the patient presents with fever and/or meningeal signs. Specific infectious studies should be considered based on the presentation, history of exposures and season. The Brighton Collaboration Encephalitis Working Group and others have published guidelines that include specific diagnostic labs to be included in the evaluation for encephalitis, myelitis, and ADEM.

Patients with clinical symptoms presenting over a longer period and/or evidence of multi-organ dysfunction may benefit from serological markers such as antinuclear antibody (ANA), antiphospholipid antibody and/or anti-dsDNA antibody to rule out a systemic rheumatologic disease with neurological manifestations. Patients with a history of neoplasm and/or psychiatric symptoms may benefit from paraneoplastic studies such as NMDAR antibody.

Aquaporin-4 antibody (AQP4) is an assay used to make the diagnosis of NMO or Devic’s disease. If a patient presents with optic neuritis and/or transverse myelitis, this serum study should be obtained in the acute period and repeated in follow-up if the suspicion remains high, as titers can be negative initially. AQP4 antibody seropositivity has been reported in cases of ADEM, and thus a positive result should be considered in the context of other features of the child’s course.

Children are at risk for accidental ingestion of prescription medications, illicit substances, alcohol, and a variety of household products which can produce acute encephalopathy associated with other neurological symptoms that mimic ADEM. A thorough history should be obtained to explore potential hazards in the environment which can guide laboratory studies obtained (e.g., lead, ethanol, Dilantin® level, etc.).

Standard studies including cell count, glucose, protein, and gram stain/culture should be obtained in most patients, but are not typically diagnostic except to exclude other disorders (e.g., positive gram stain and culture in bacterial meningitis). Cerebrospinal fluid (CSF) pleocytosis (WBC >5) and elevation of protein are common in ADEM, but not always abnormal. When elevated, the CSF WBC tends to be mildly elevated (<100 cells/uL).

Viral studies should be considered and tailored to presenting symptoms. HSV encephalitis is treatable with acyclovir, and thus an HSV polymerase chain reaction (PCR) should be strongly considered for patients with fever, seizure, CSF pleocytosis, and/or lesions in the temporal lobes. Other viral studies to consider include enterovirus PCR, EBV PCR, cytomegalovirus (CMV) PCR, and varicella-zoster virus (VZV) PCR.

The significance of CSF oligoclonal bands is variably reported in the literature with some studies reporting a greater likelihood of diagnosis with a competing demyelinating disorder, and others reporting presence of oligoclonal bands in the CSF of a percent of patients with ADEM.

Viral panels are now readily available to assess for a variety of upper respiratory infections such as influenza, parainfluenza, rhinovirus/enterovirus, or adenovirus. These panels can be helpful in diagnosing an acute infection as well as prior infections that may have precipitated a post infectious process.

Head computed tomography (CT) is a useful tool to exclude intracranial hemorrhage and acute space occupying lesions in children who present with severe encephalopathy and/or focal neurological symptoms, particularly if a lumbar puncture is anticipated. For children with more subtle symptoms who are neurologically stable, a CT may be deferred if a magnetic resonance image (MRI) is anticipated. A head CT is relatively insensitive in detecting small lesions and/or infratentorial lesions and thus can be normal, particularly in the first few days of symptom onset. Demyelinating lesions on head CT are hypodense and often nonspecific in appearance. If hypodense lesions are identified on CT, a MRI should be obtained to better characterize the lesions.

MRI is the neuroimaging modality of choice for making the diagnosis of ADEM. Lesions are best seen on T2 and fluid-attenuated tissue recovery (FLAIR) sequences. Lesions of increased T2 signal are typically bilateral, but asymmetric. White matter lesions are predominant, but grey matter involvement, particularly in the deep nuclei, is not uncommon. Gadolinium enhancement and diffuse restriction are often present in the acute period. ADEM lesions can range in size and distribution, from small homogeneous lesions, to large tumefactive lesions. Spinal cord involvement is seen in up to a third of cases. In children severely affected, it may be appropriate to empirically obtain imaging of both the brain and spine to define the extent of the disease process.

The radiologic appearance of ADEM can be difficult to distinguish from that of multiple sclerosis. Several groups have attempted to devise MRI criteria to discriminate between competing demyelinating disorders. For example, in 2009, Callen et al designed criteria with an 81% sensitivity and 95% specificity for distinguishing an initial manifestation of multiple sclerosis from ADEM in children. See Table III.

Table III.

There is suggestion that the Callen MRI criteria for MS can differentiate between ADEM and pediatric multiple sclerosis. The published sensitivity and specificity were 82% and 52%, respectively. This finding has not been verified in a large prospective study. The Callen criteria for diagnosing pediatric MS on MRI is as follows:

At least 2 of the following:

• 5 or more T2 lesions

• 2 or more periventricular lesions

• 1 or more brainstem lesions

MRI patterns are useful in distinguishing ADEM from other disorders as well. Leptomeningeal enhancement should steer the diagnosis away from ADEM and toward infectious meningoencephalitis. Bilateral thalamic lesions are typical of ADEM, but also may be suggestive of mitochondrial disease, thrombosis of the deep cerebral veins, and/or acute necrotizing encephalopathy of childhood (ANEC). Large tumefactive lesions are often the most difficult to distinguish by radiographic appearance alone. Neoplasm, malignant MS variants, and brain abscess should all be considered in addition to ADEM.

Electroencephalogram (EEG) should be considered in the evaluation of children who present with seizure and/or encephalopathy. EEG can assess for subclinical seizures as a potential cause of encephalopathy. Focal features and generalized slowing can be seen in ADEM, but are not specific to the disorder.

Several clinical algorithms for the diagnosis of ADEM are available.

The Brighton Collaboration Encephalitis Working Group (Brighton) published guidelines in 2007. The guidelines are written from the perspective of how to distinguish ADEM from encephalitis and myelitis. The guidelines provide levels of certainty for the diagnosis of each based on a combination of presenting signs and symptoms, histopathology, CSF studies, EEG, and neuroimaging.

The same year, the International Pediatric Multiple Sclerosis Study Group (IPMSSG) published diagnostic criteria for pediatric MS and immune-mediated CNS demyelinating disorders. The focus for the IPMSSG was to design criteria to accurately diagnose specific demyelinating disorders at their initial presentation. Both the Brighton and IPMSSG guidelines were definitive in the need for encephalopathy to be included in the clinical criteria for ADEM. The IPMSSG guidelines were revised in 2012. In the revision, the term recurrent ADEM was eliminated and “subsumed” under multiphasic ADEM. The definition of multiphasic ADEM was restricted to include a second event of ADEM associated with new, or re-emergence of prior, clinical and MRI findings occurring >3 months after the initial event. Children with more than 2 demyelinating events were no longer eligible for the diagnosis of multiphasic ADEM. They also removed timing in relation to steroids criteria for recurrent events. In the definition of monophasic ADEM, encephalopathy remained an essential feature, but stipulated if explained by fever, alternative diagnoses should remain in the differential. The group acknowledged that children with MS who are <12 years at the time of initial presentation may have a more nonspecific presentation of clinical and radiographic abnormalities. See Table IV.