ADEM

Acute disseminated encephalomyelitis (ADEM) is a rare inflammatory demyelinating disease of the CNS affecting predominantly children and young adults. Typically, it is antedated by an infectious illness, most commonly measles, mumps, influenza A or B, Rocky Mountain spotted fever, or hepatitis A or B, or infection with herpes simplex, human herpesvirus 6, varicella, rubella, vaccinia, Epstein–Barr virus, cytomegalovirus, Mycoplasma pneumoniae, Chlamydia, Legionella, Campylobacter, or Streptococci. ADEM can also follow immunizations: rabies, diphtheria–tetanus–pertussis, smallpox, measles, Japanese B encephalitis, and hog.1,2⇓ Characteristic clinical features include sudden onset of multifocal neurologic disturbances such as bilateral optic neuritis, visual field defects, aphasia, motor and sensory deficits, ataxia, movement disorders, and signs of an acute meningoencephalopathy with meningismus, a depressed level of consciousness, focal or generalized seizures, and psychosis. Maximal deficits are reached within several days and remission may be similarly rapid. More often, however, resolution is partial and takes weeks or months. High mortality rates (up to 20%) were frequent in the past, particularly when ADEM followed measles. However, with the introduction of effective vaccination strategies and the decline of measles, death is rare.2 MRI reveals multifocal areas of increased signal intensity on T2-weighted sequences that commonly enhance synchronously with contrast medium.3,4⇓ Extensive perifocal edema may be seen. White matter involvement predominates but gray matter can also be affected, particularly basal ganglia, thalami, and brainstem. Lesions may also be restricted to the brainstem or cerebellum. Tumor-like lesions also occur. CSF usually shows pleocytosis but only rarely intrathecal oligoclonal immunoglobulin (Ig) G production, which almost invariably ceases as patients improve. Whereas high-dose steroids, plasma exchange, IVIg, immunomodulators, and hypothermia have been tried, no treatment is established.5-9⇓⇓⇓⇓ This textbook view of ADEM is based on a compilation of features extracted from single cases and small series in the literature. How faithfully these descriptions delineate the clinical entity is unclear because there have been no large series of ADEM. Three such series are now being published.

Series from Australia by Hynson et al.10 and the United Kingdom by Dale et al.11 look at children; the one from Germany, by Schwarz et al.,12 at adults. All are retrospective single-centered studies reporting clinical features, investigational laboratory findings, and MRI results on 37 children,10 31 children,11 and 40 adult patients.12 The two pediatric series10,11⇓ rely solely on chart and MRI review whereas nearly all adult patients identified in the third series were also examined at follow-up (8 to 137 months; mean 38).12 The initial diagnosis of ADEM was re-evaluated on that occasion. Only this adult study12 looks relatively systematically over a longer interval at the occurrence of new clinical episodes of multifocal demyelination. Of the initial 44 adult patients diagnosed with ADEM upon admission to the hospital, 40 entered the study, 14 of whom had a second clinical episode within the first year after initial presentation and were therefore re-classified as having MS.12 Obviously, all shortcomings of retrospective studies relating to nonuniform reporting, case ascertainment, nonuniform data acquisition, and center bias need to be taken into account. With MRI, utilization of different and development of more sophisticated scanners, lack of defined reading protocols, and the emergence of more advanced sensitive and specific measures additionally confound and weaken the statistical power of the analysis. Finally, in the absence of universally agreed-upon diagnostic criteria, the authors in these three studies apply somewhat different definitions, with the one used in the Australian series being partly tautologic10 because it stipulates MR to be consistent with ADEM. Nevertheless, given the rarity of the syndrome, all three series, the largest thus far published, provide useful clinical information. Overall, the data summarized in the table agree with previous descriptions of ADEM in children13 and underscore the strong link between an infective illness and the acute onset of an encephalopathic clinical phenotype with multifocal deficits and an often rapid recovery. Helpful is the collective information on MRI features. Asymmetric white matter lesions located around the ventricles and subcortically, the majority of which enhance with contrast medium, are characteristic. It is of note that the frequent MR involvement of basal ganglia is usually asymptomatic. Similar results were obtained across all three studies regarding resolution of lesions upon follow-up. Lesions resolved partially in two thirds and vanished in nearly a third of patients.

The MRI results confirm the notion of most neuroradiologists of just how weak conventional MRI is with respect to the specific diagnoses of ADEM, MS, and other white matter diseases. MRI criteria using lesion number and location are just not refined enough to render a definitive MS. More importantly, the impetus to treat patients with MS as early as possible using MRI criteria will result in misdiagnosis and improper administration of therapy. The authors are to be congratulated in alerting us to the shortcomings of MRI in the diagnosis of ADEM. What is needed now are more specific methods that potentially can separate these and other white matter lesions from MS. Currently, none exist and it is unclear that any of the newer methods (e.g., diffusion-weighted imaging, magnetization transfer, proton spectrocopy) will improve specificity. The bottom line is caution: in this case, “image isn’t everything.”

The study by Schwarz et al.12 is remarkable in two regards. First, it is the largest ever published series of ADEM in adults and, second, it shows that a significant proportion of patients presenting with a clinical phenotype and MRI findings consistent with ADEM go on to develop MS. Most adult patients presented clinically in a fashion similar to children with the notable exception of the relatively infrequent occurrence of headache, fever, and meningismus, and the higher frequency of sensory deficits. Optic neuritis was rare in adult ADEM.

In the cohort of 40 patients reported by Schwarz et al. that were initially diagnosed with ADEM, 35% developed clinically definite (Poser criteria) MS over a mean observation period of 38 months.12 This addresses a crucial issue. Can ADEM be diagnosed with certainty and differentiated from the initial manifestation of MS? Schwarz et al. look at possible discriminating clinical features but in accord with the previous literature fail to identify any exclusive to either condition, which is true for both children and adults. Similarly, CSF findings are not distinctive enough to allow discrimination between ADEM and MS in a single patient. Unfortunately, even MRI could not differentiate ADEM from MS. It is of particular note that 50% of the patients with ADEM had MRI features that, utilizing the Paty criteria, were suggestive of MS. Whereas the newly revised diagnostic criteria for MS allow the diagnosis to be made after one attack if stringent MRI criteria14 are met, they emphasize that in monophasic demyelinating disease such as ADEM a diagnosis should be withheld unless new symptoms and signs or imaging abnormalities appear more than 3 months after clinical onset.15 The Schwarz et al. study does not permit a calculation of “conversion rate” for ADEM to MS because the observation period was short and MRI examinations were not performed at predetermined intervals. In the longest follow-up (8 years) of 11 patients with ADEM, none experienced a new clinical attack during follow-up and new white matter lesions were detectable in only one patient.16 The issue of whether ADEM can be the presentation of MS is important because two recent trials support early initiation of therapy in MS.17,18⇓

No therapy has been established by controlled trials in ADEM but the two series published in this issue are consistent with earlier anecdotal reports that corticosteroids are useful.5 It seems reasonable to deliver pulsed IV steroids with an oral taper. Other treatments that have been found to benefit include plasma exchange and high-dose IVIg.6-8⇓⇓ If high-dose corticosteroids fail, it is reasonable to advise plasma exchange.19

What is the pathogenesis of ADEM? This remains elusive. Tempo of disease evolution and pathologic findings of multifocal inflammation and demyelination can be reproduced in the animal model of acute experimental autoimmune encephalomyelitis, which is mediated by autoreactive CNS-specific T cells.20 T cells directed to microbial epitopes may recognize amino acid sequences shared with myelin antigens and launch an autoaggressive attack on CNS structures (“molecular mimicry”), alone or in synergy with antibodies. Viral or bacterial superantigens could likewise trigger autoreactive T cells. Semple rabies vaccine, still used in India,21 contains neural antigens that could excite a cross-reactive T-cell response. Whether hepatitis B vaccination, associated with two instances of ADEM in the Australian series, can cause the disease remains unclear, particularly because the vaccine has recently been exculpated from precipitating attacks in MS.22,23⇓

With many similarities in clinical presentation, MR findings, and putative pathogenesis, it appears reasonable to consider ADEM as part of the spectrum of inflammatory demyelinating diseases that encompasses ADEM and its most severe variant, acute disseminating hemorrhagic encephalomyelitis; restricted disorders such as optic neuritis, transverse myelitis, and Devic’s syndrome; and MS, which in itself in all likelihood reflects heterogeneous pathology.