Clinical manifestations

Presentation and course

Initial symptoms of acute disseminated encephalomyelitis may occur spontaneously without antecedent illness or, more typically, begin between 2 days and 4 weeks after a febrile illness of presumed viral etiology (01). The actual microbial illness is rarely determined. Clinical presentation includes encephalopathy (which may manifest as behavioral change, profound irritability, or altered consciousness) that cannot be explained by fever and polyfocal neurologic deficits (30; 53). Systemic symptoms such as fever, headache, and fatigue often precede the development of neurologic deficits. Once neurologic symptoms begin, the course is rapidly progressive, and patients typically develop maximal symptoms within 3 to 5 days (53).

Neurologic signs and symptoms vary in acute disseminated encephalomyelitis and are determined by the location of lesions in the CNS. In a prospective series of 84 patients, the most common presenting features of acute disseminated encephalomyelitis included long tract signs (85%), acute hemiparesis (76%), and cerebellar ataxia (50%), in addition to encephalopathy (51). Less common presenting features included cranial nerve palsies, spinal cord involvement, aphasia, seizures, and visual dysfunction.

The severity of symptoms varies in acute disseminated encephalomyelitis as some children may present with mild somnolence and subtle neurologic dysfunction, whereas others rapidly develop obtundation and respiratory depression necessitating intensive care. Respiratory failure secondary to brainstem involvement or severely impaired consciousness has been reported in 11% to 16% of cases (52). In fulminant cases, cerebral edema and increased intracranial pressure can occur and result in death or severe neurologic dysfunction if not intervened on quickly.

Prognosis and complications

Although the majority of patients with acute disseminated encephalomyelitis completely recover, the acute phase can be severe and life-threatening, and minor residual deficits have been reported in 20% to 30% of children (47; 53). Of these residual deficits, the most frequently reported include mild motor deficits, visual problems, and seizures (51). Children with a history of acute disseminated encephalomyelitis have also been shown to report increased fatigue and have decreased exercise capacity when compared to healthy peers (55).

The average time to full recovery ranges from 1 to 6 months, though patients often experience immediate improvement of symptoms after beginning treatment with corticosteroids. In a cohort of 24 children with acute disseminated encephalomyelitis, time until initial improvement was 6.0+4.5 days, and time until full recovery was 26+34 days (45). Sixty-five percent of this cohort improved within 3 days of high dose methylprednisolone administration (45). The mortality of acute disseminated encephalomyelitis had previously been reported to be as high as 20% (42). However, in the era of modern treatment agents for inflammation and elevated intracranial pressure, this figure has been markedly reduced. One group aimed to identify clinical and radiologic risk factors that might predict poor functional outcomes by retrospectively reviewing patients with ADEM with moderate to severe disease severity, and found that a higher percentage of those who presented with dystonia and myoclonus had poor recovery as did those with more cerebellar lesions (13). Of note, only a few of the patients in this cohort were tested for MOG antibodies.

Cognitive and learning deficits have also been reported as long-term consequences of acute disseminated encephalomyelitis. Subtle deficits in executive function, attention, and behavior have been reported in children who have otherwise completely recovered from acute disseminated encephalomyelitis (31; 49). These deficits have been noted to be more prominent in children who were younger than 5 years of age at the time of their diagnosis (24). Attention may improve over time as patients tested further from the diagnosis tend to have better neurocognitive test scores (49).

It has been shown that although the clinical characteristics and presenting features of adult-onset acute disseminated encephalomyelitis can be similar to pediatric-onset disease, the outcome and disease course can be more severe. In fact, in a retrospective review looking at prognosis and disease course in adults and children with acute disseminated encephalomyelitis, it was found that adults were more likely to require longer hospitalization and intensive care admission. Outcome in adult-onset disease was also worse as fewer adults had complete motor recovery (28).

Acute disseminated encephalomyelitis is typically a monophasic disorder. However, some children with a first demyelinating attack meeting criteria for acute disseminated encephalomyelitis will ultimately be diagnosed with multiple sclerosis. The reported percentage of patients with an ultimate diagnosis of multiple sclerosis varies in different cohorts based on the inclusion criteria and definition of acute disseminated encephalomyelitis used. For instance, although one prospective study reported that 18% of their acute disseminated encephalomyelitis cohort eventually had a confirmed diagnosis of multiple sclerosis, a prospective study of 302 children with acute demyelinating syndromes and strict definitions for acute disseminated encephalomyelitis reported that only 4 of 77 patients (5%) initially diagnosed with acute disseminated encephalomyelitis were subsequently diagnosed with multiple sclerosis (41; 06).

Though multiple sclerosis and neuromyelitis optica spectrum disorder must be considered in a patient with relapsing demyelination following an initial episode of acute disseminated encephalomyelitis, other disorders such as myelin oligodendrocyte glycoprotein antibody-associated disease (MOGAD) should also be considered (18; 48). Proposed diagnostic criteria for MOGAD have been published in an effort to improve identification of individuals with this disorder (Table 2) (07).

Table 2. Diagnostic Criteria for MOGAD (A, B, and C are required)

Table 3. Supporting Clinical or MRI Features for MOGAD Diagnosis

Although patients with acute disseminated encephalomyelitis often make a full clinical recovery, one study has shown that these children (along with children with multiple sclerosis and clinically isolated syndromes) have impaired cerebellar growth and cerebellar volume reduction (59). This has not yet been correlated with disability or persistent clinical symptomatology, but further studies are needed to better understand the clinical implication of this finding (59).

It has also been shown that children with acquired monophasic demyelination, especially those with acute disseminated encephalomyelitis, demonstrated reduced age-expected brain growth, which is thought to be primarily driven by reduced white matter growth (05). The clinical implications of this are still unknown. Additionally, when looking at serial measurements of normal appearing white matter using diffusion tensor imaging in children with acquired demyelinating syndromes, there was a decline from age-expected values not only in those with multiple sclerosis, but also in those with acute disseminated encephalomyelitis (33). A study has also corroborated that patients with acute disseminated encephalomyelitis showed reduced brain volume at baseline compared with matched controls, in addition to a reduction in age-expected brain growth over time (10). Volume loss was more pronounced in patients with acute disseminated encephalomyelitis who were seronegative for MOG antibodies compared to those who were seropositive (10).

Clinical vignette

A developmentally normal 4-year-old boy was admitted with depressed consciousness and inability to walk or sit up independently. His lethargy and truncal instability had progressed over the preceding 4 days. Of note, he had a diarrheal illness 1 week prior to his presentation and had developed low-grade fever 4 days prior to his admission. His vital signs and general medical exam on admission were normal. His neurologic exam was notable for a somnolent mental status, dysarthria, and severe truncal and appendicular ataxia. Initial labs, including a complete blood count and electrolyte panel, were normal. Cerebrospinal fluid (CSF) studies showed 48 white blood cells per mm³ (predominantly lymphocytes), few red blood cells, normal glucose, and a mildly elevated protein. Oligoclonal bands were negative. Viral PCR studies and a bacterial culture from the CSF were negative. MRI of the brain revealed multifocal T2-hyperintense lesions involving both cerebral hemispheres, the pons, the middle cerebellar peduncles, and the cerebellum, with some lesions showing patchy enhancement with gadolinium. Serum MOG antibody was positive.

Intravenous methylprednisolone was started after the MRI was completed. Within 24 hours of starting corticosteroids, the patient become more alert and was able to sit independently. Over the following 2 days, the patient’s clinical condition improved rapidly; his dysarthria resolved, and he was able to walk with support. The patient was discharged home with an oral prednisone taper. Two weeks later, the patient was asymptomatic and had returned to baseline. A brain MRI was repeated 6 months after discharge and showed complete resolution of his initial lesions and no new lesions.

Biological basis

Etiology and pathogenesis

Nonspecific upper respiratory tract infections have been reported to be the most common illnesses preceding acute disseminated encephalomyelitis (26; 43; 51). Less common precedent infections include varicella, herpes simplex virus encephalitis, mycoplasma, enterovirus, Zika virus (44), Chikungunya (50), and SARS-CoV-2 (36; 61).

The level of evidence for a direct association between vaccines and acute disseminated encephalomyelitis remains anecdotal. In the past, acute disseminated encephalomyelitis was associated with specific vaccinations that were produced in neural tissue culture, such as the Semple form of the rabies vaccine and the Japanese B encephalitis vaccine. These cases have become less frequent as recombinant protein vaccines have replaced those cultured from neural tissue (53).

Though the pathogenesis of acute disseminated encephalomyelitis has not been fully elucidated, existing evidence suggests that acute disseminated encephalomyelitis results from an autoimmune response toward myelin-derived antigens. Much of the current thinking regarding the pathogenesis of acute disseminated encephalomyelitis comes from study of the animal model of experimental autoimmune encephalomyelitis, as many features of the induced murine model are notable in human acute disseminated encephalomyelitis. Experimental autoimmune encephalomyelitis is an acute encephalomyelitis that has been induced in a variety of species with introduction of myelin proteins or myelin-derived peptides. Although specific mouse genetic strains can develop a relapsing form of experimental autoimmune encephalomyelitis, most models are characterized by a monophasic illness associated with diffuse CNS demyelination.

Using this animal model, it has been hypothesized that both primary and secondary autoimmune responses contribute to CNS inflammation and subsequent demyelination in acute disseminated encephalomyelitis. The two most widely accepted possible pathogenic mechanisms for acute disseminated encephalomyelitis are molecular mimicry and self-sensitization.

The hypothesis of molecular mimicry is based on the idea that antigenic epitopes are shared between pathogens or vaccines and host myelin antigens. These antigens from pathogens or vaccines subsequently activate myelin-reactive lymphocytes, which can then migrate into the CNS and attack myelin. Many myelin peptides such as myelin basic protein and proteolipid protein have shown similarity to viral sequences and can elicit cross-reactive T-cell responses.

The hypothesis of self-sensitization suggests that direct infection with a neurotropic pathogen results in the release of myelin peptides and a secondary inflammatory response. In the setting of this diffuse inflammation, there is breakdown of the blood-brain barrier, and myelin-derived autoantigens are released into peripheral circulation where they are exposed to naive lymphocytes. These sensitized T-cells may then migrate into the CNS and attack host myelin, causing demyelination in the CNS.

The role of autoantibodies to specific myelin proteins in the pathogenesis of acute disseminated encephalomyelitis and other demyelinating diseases is an area of active research. Multiple studies have shown that myelin oligodendrocyte glycoprotein (MOG) antibodies are found in at least one third of children with acquired demyelinating syndromes. In one study, MOG antibodies were present in 64% of patients with acute disseminated encephalomyelitis, and greater than 95% of those with relapsing acute disseminated encephalomyelitis (ie, multiphasic acute disseminated encephalomyelitis, acute disseminated encephalomyelitis followed by optic neuritis) (18). In this study, about half of the children with MOG antibodies relapsed (18). Another study has shown MOG antibodies to be present in about 57% of children with acute disseminated encephalomyelitis (22). In this study, a high MOG antibody titer was associated with a relapsing non-multiple sclerosis course.

Given the strong association of MOG antibodies with both monophasic and relapsing acute disseminated encephalomyelitis, we recommend that all children with a clinical suspicion for acute disseminated encephalomyelitis undergo serum MOG antibody testing with a cell-based assay, and that those who are positive undergo serial evaluation to determine whether antibodies are persistent over time. CSF MOG antibody testing is typically not required as MOG-IgG is produced mostly extrathecally, resulting in lower CSF than serum titers (25).

Although the presence of MOG antibodies at presentation are predictive against a subsequent multiple sclerosis disease course (19; 27), a single positive MOG result does not appear sufficient to predict relapsing non-multiple sclerosis clinical phenotypes. One study has suggested, however, that relapse may be more likely in patients with persistent MOG antibodies after acute disseminated encephalomyelitis, defined as those who remained seropositive at a time point at least 3 months after initial presentation (34). Specifically, 15 of 17 patients with persistent MOG seropositivity after acute disseminated encephalomyelitis experienced a relapse, compared to one of eight patients with transient seropositivity and 8 of 24 that were seronegative for MOG (34).

Importantly, demographics and initial clinical features of children with acute disseminated encephalomyelitis who are found to be seropositive for MOG antibodies are indistinguishable from those who are seronegative (12). Multiphasic disseminated encephalomyelitis, a relatively rare clinical disorder, is strongly associated with MOG antibodies (11). By definition, these patients experience at least two discrete episodes of acute disseminated encephalomyelitis.

It is important to emphasize that the spectrum of pediatric MOG antibody-associated syndromes is wider than previously reported and includes encephalitis different from acute disseminated encephalomyelitis. In a prospective study of 239 children with demyelinating syndromes and 296 with encephalitis other than acute disseminated encephalomyelitis, 116 had MOG antibodies. The presenting syndromes in these 116 patients included acute disseminated encephalomyelitis (46 [68%]), encephalitis other than acute disseminated encephalomyelitis (22 [19%]), optic neuritis (20 [17%]), myelitis (13 [11%]), neuromyelitis optica spectrum disorders (6 [5%]), and other disorders (9 [8%]). Children with anti-MOG encephalitis different from acute disseminated encephalomyelitis had a worse clinical outcome. In children with MOG antibodies and acute disseminated encephalomyelitis 13% had a modified Rankin score of at least 2 compared to 36% of children with anti-MOG encephalitis who did not fulfill criteria of acute disseminated encephalomyelitis (04).

In a study analyzing the serum of 15 patients with acute disseminated encephalomyelitis and 26 patients with multiple sclerosis, distinct profiles of autoantibody reactivity to myelin peptides were found in patients with each disease. Specifically, the serum of patients with acute disseminated encephalomyelitis was characterized by IgG autoantibodies targeting myelin basic protein and myelin-associated oligodendrocyte basic protein, whereas the serum of patients with multiple sclerosis was characterized by autoantibodies of the IgM subtype targeting myelin peptides, including proteolipid protein, myelin-associated oligodendrocyte basic protein, and oligodendrocyte-specific protein (57). The finding that IgG autoantibodies, which are class-switched, are predominant in acute disseminated encephalomyelitis supports the notion that acute disseminated encephalomyelitis is a postinfectious process where the immune response is primed prior to the onset of demyelination.

Other autoantibodies, including N-methyl-D-aspartate receptor, glycine receptor, and voltage-gated potassium channel complex antibodies, have been found in some patients with acute disseminated encephalomyelitis although not necessarily at the time of clinical manifestations of acute disseminated encephalomyelitis (20). These results have to be taken with caution. The antibodies when detected only in serum may represent a false positive result. In addition, voltage-gated potassium channel complex antibodies without specificity for LGI1 or CASPR2 are not considered biomarkers of autoimmunity (32).

Perivenous inflammatory infiltrates of T-cells and macrophages associated with perivenular demyelination is the pathologic hallmark of acute disseminated encephalomyelitis, which differs from the more confluent demyelination that is pathologically seen in patients with multiple sclerosis (60). Lesions in acute disseminated encephalomyelitis predominantly involve white matter, though deep gray structures and cortex can also be involved, and are typically of similar histologic age (53).

Epidemiology

Acute disseminated encephalomyelitis is primarily a pediatric disease, with the mean age of clinical presentation ranging from 5 to 8 years (23; 03; 53). Although acute disseminated encephalomyelitis can occur at any time of the year, it has been reported to be more common in the winter and spring in temperate climates (37). Incidence rates vary in different cohorts and range from 0.07 to 0.64 per 100,000, though different inclusion criteria and definitions of acute disseminated encephalomyelitis were used in each study (37; 46; 08; 54). In terms of gender predilection, a male predominance has been described in few pediatric cohorts (43; 51; 08).

Prevention

There is currently no evidence or accepted strategy for primary prevention of acute disseminated encephalomyelitis.

Differential diagnosis

Children and adults with acute disseminated encephalomyelitis are acutely ill, sometimes with fever, and, by definition, have altered consciousness. As such, the first priority is to exclude acute CNS infection. CSF analysis and prompt administration of antimicrobial therapy, including acyclovir, is essential until CNS infection is excluded. As the diagnosis of acute disseminated encephalomyelitis is based solely on clinical history and radiographic studies, a number of other inflammatory and infectious diseases should be considered in the differential before a definitive diagnosis is made (Table 1). Although the thalamus is often involved in acute disseminated encephalomyelitis, it can also be involved with deep cerebral vein thrombosis and metabolic disease, which are two entities important to consider in the differential in the proper clinical context. Fulminant cases of acute hemorrhagic leukoencephalitis may mimic acute herpes encephalitis, acute bacterial meningitis, or venous sinus thrombosis. Large tumefactive lesions on MRI may suggest alternate diagnoses such as Marburg variant of multiple sclerosis, tumor, abscess, or CNS vasculitis.

Table 4. Differential Diagnosis of Acute Disseminated Encephalomyelitis

Though differentiating an initial attack of multiple sclerosis from acute disseminated encephalomyelitis may be difficult at the time of the first incident attack, some epidemiologic, clinical, and radiographic features may help to distinguish a first attack of multiple sclerosis from monophasic acute disseminated encephalomyelitis (Table 2). Additionally, as neuromyelitis optica spectrum disorder and MOG antibody-associated disease can initially present with an acute disseminated encephalomyelitis-like event, they should also be considered in the differential for acute disseminated encephalomyelitis.

Table 5. Characteristic Features of Acute Disseminated Encephalomyelitis and Multiple Sclerosis

Diagnostic workup

Investigations when evaluating cases of possible acute disseminated encephalomyelitis are prioritized to first exclude active CNS infection and to evaluate recent viral exposures. CSF analysis (cell count, protein, glucose) should be completed. A mild to moderate CSF lymphocytic pleocytosis is often present (40% to 85% of cases). In severe cases and those with a hemorrhagic component, a pleocytosis of greater than 1000 per mm³, a polymorphonuclear predominance, and red blood cells in the CSF can be seen. CSF protein is elevated in about one half of cases, but levels above 100 mg/dL are rare. Oligoclonal bands are typically negative. This is in contrast to multiple sclerosis, where intrathecal oligoclonal bands are frequently present and persist over time.

Four radiographic patterns of cerebral involvement have been described in acute disseminated encephalomyelitis. These patterns include acute disseminated encephalomyelitis with (1) small lesions (less than 5 mm), (2) large confluent lesions with associated edema and mass effect, (3) additional symmetric bithalamic involvement, and (4) acute hemorrhagic encephalomyelitis (51). Meningeal enhancement is uncommon in acute disseminated encephalomyelitis (though can be seen in some cases of MOGAD), as is the presence of complete ring-enhanced lesions and lesions confined to the corpus callosum (38; 52). Lesions are commonly visible in the thalamus and basal ganglia in acute disseminated encephalomyelitis.

Of note, cases where imaging reveals only small lesions, well-demarcated lesions, or involvement of just one anatomical area are less likely to be MOG-Ab positive (12).

Cerebral MRI abnormalities in children with neuromyelitis optica spectrum disorder can sometimes mimic those seen in cases of acute disseminated encephalomyelitis (39). For example, large confluent lesions affecting the brainstem, hypothalamus, juxtacortical, and periventricular areas have been described in pediatric patients presenting with an acute disseminated encephalomyelitis—like attack who ultimately meet criteria for neuromyelitis optica spectrum disorder (09; 39).

The spine can be affected in up to 28% of patients with acute disseminated encephalomyelitis, and lesions can be longitudinally extensive (43; 51).

In these cases, it is important to also consider the diagnosis of neuromyelitis optica spectrum disorder and MOG antibody-associated disease. Spinal cord imaging features that may be more suggestive of MOG antibody-associated disease include grey matter involvement and conus lesions (17). Spinal cord lesions in multiple sclerosis are typically more focal (40; 15).

Management

Drug treatment. To date, there have been no formal treatment trials for acute disseminated encephalomyelitis, and standard of care is based on observational data and expert consensus (35). Initial treatment of acute disseminated encephalomyelitis involves supportive care. Depending on the clinical situation, antibiotics and antivirals should be started empirically until an infectious etiology has been excluded. When the diagnosis of acute disseminated encephalomyelitis has been confirmed, treatment with corticosteroids should be initiated (methylprednisolone at a dose of 30 mg/kg/day up to 1 gram/day) intravenously for 3 to 5 days. If symptoms resolve, subsequent treatment with oral prednisone is not required. For children with improvement, but still ongoing deficits, a prednisone taper can be used (dose starting at 2 mg/kg/day, with a max of 60 mg/day, tapering over 10 to 14 days) (23; 47).

For patients with acute disseminated encephalomyelitis who fail to demonstrate clinical improvement by the third to fifth day of corticosteroid treatment, or children with life-threatening demyelination at onset, plasma exchange should be considered (14; 47).

Intravenous immunoglobulin has been reported to be effective in some children with acute disseminated encephalomyelitis who do not respond fully to corticosteroids, or in some patients who seem to experience recurrence of neurologic deficits on corticosteroid withdrawal (52; 35). Maintenance monthly intravenous immunoglobulin is typically given to those children who are ultimately diagnosed with MOG antibody-associated disease if they should relapse after their initial attack of acute disseminated encephalomyelitis (21).

Some children with acute disseminated encephalomyelitis experience fulminant, life-threatening cerebral edema refractory to conventional medical management. In these cases, decompressive craniectomy may lead to rapid improvement. The most severe form of acute disseminated encephalomyelitis is acute hemorrhagic leukoencephalitis and is thought to be rapidly fatal if not treated promptly. Survival has been reported with combination treatment using corticosteroids, plasmapheresis, and decompressive craniectomy if needed.

Treatments on the horizon. There are no additional acute or preventative treatments available for acute disseminated encephalomyelitis on the horizon at the current time.

Outcomes

The outcome of monophasic acute disseminated encephalomyelitis is generally favorable as most patients recover without sequelae. Treatment with methylprednisolone is generally well tolerated though providers should be aware of potential side effects associated with high-dose corticosteroids, including hyperglycemia, hypertension, and mood lability. Prolonged use of steroids is not indicated for acute disseminated encephalomyelitis.

Special considerations

Pregnancy

Although there are case reports of acute disseminated encephalomyelitis in pregnant women, there is no known increase in ADEM in the context of pregnancy. Risks and benefits of treatment during pregnancy must be considered on an individual basis. Although immunomodulatory treatment may be indicated to prevent additional morbidity, there may be associated risks with therapies during pregnancy.

Anesthesia

As patients with acute disseminated encephalomyelitis are encephalopathic, use of anesthetic medications should be managed by trained anesthesiologists or intensive care physicians when required. Patients should be monitored closely while receiving anesthesia and in the period after the medication has been administered.