Abstract
Objectives: To describe the clinical, CSF, and radiologic findings and long-term follow-up in a cohort of patients with acute disseminated encephalomyelitis (ADEM), and to determine possible prognostic factors for progression to MS.
Methods: Forty adults (28 women, mean age 33.5 years) diagnosed with ADEM were analyzed. Clinical symptoms, cranial MRI and CSF findings, and the response to a standardized treatment during the acute phase of the disease were analyzed by chart review. The final diagnosis of ADEM or clinically definite MS was established upon follow-up examination after 8 to 137 months. The patients with ADEM and MS were compared to detect differences between the two groups.
Results: Fifteen patients had a preceding infection (n = 14) or immunization (n = 1). The most frequent clinical signs were motor deficit (80%), followed by sensory deficits, brainstem signs, and ataxia. CSF findings were highly variable; normal results were present in 20% of patients. Oligoclonal bands were positive in 65% of patients. Ninety-five percent of all patients improved during the acute phase of the disease. Upon follow-up, 14 patients had developed clinically definite MS. Of the 26 patients with the final diagnosis of ADEM, two patients had died, nine had minor deficits, three had moderate deficits, and 12 patients had no remaining symptoms. Patients with the final diagnosis of ADEM were older, and more often had a preceding infection, clinical signs of brainstem involvement, a higher CSF albumin fraction, and infratentorial lesions.
Conclusions: Many patients initially diagnosed with ADEM develop clinically definite MS upon long-term follow-up. The authors found no useful diagnostic criteria for the differentiation of a first episode of MS from monophasic ADEM. The term ADEM may still be employed as a description of a clinical syndrome, but should not be used as a distinct entity until reliable diagnostic criteria have been developed.
Among the demyelinating diseases, MS occurs most frequently and is the best defined. Acute disseminated encephalomyelitis (ADEM) has been described as an uncommon, acute demyelinating disorder of the CNS, typically occurring after infections or vaccinations.1 Although the pathophysiology of ADEM is not known, an autoimmune response to myelin basic protein triggered by infection or immunization is considered to be a possible etiologic factor. Thus, ADEM may be the clinical counterpart to experimental allergic encephalomyelitis (EAE).1-3⇓⇓ In contrast to MS, ADEM is usually defined as a monophasic disease with a favorable long-term prognosis. The differentiation of ADEM from MS in a patient after a single clinical episode has prognostic and therapeutic implications because early treatment of MS may be advantageous. There are no generally accepted diagnostic criteria for ADEM; thus, distinguishing ADEM from the first episode of MS is often difficult. Moreover, cases of recurrent or multiphasic ADEM have been described.
Until now, only single case reports or small case series of patients with ADEM have been published, and most of these patients were children. Although the monophasic course of ADEM has been repeatedly emphasized, larger follow-up studies have not yet been undertaken.
We describe the presenting clinical, radiologic, and CSF findings and the long-term follow-up of adult patients initially diagnosed with ADEM to develop diagnostic criteria for the differentiation of ADEM from MS.
Patients and methods.
Patients.
Between 1988 and 1999, 42 adult patients diagnosed with ADEM were treated at the University of Heidelberg Neurologic Hospital. For follow-up examination, two patients could not be located; neither was residing in Germany. Of the remaining 40 patients, follow-up of two patients was based on chart review alone: one of these patients had died of colon cancer in the interim, and the second patient had moved after several outpatient clinic visits and could not be located.
Criteria for the initial diagnosis of ADEM.
Generally accepted diagnostic criteria for ADEM have not yet been established. Although various “typical” features have been proposed, until now, reliable clinical or radiologic signs that allow a clear distinction of ADEM from MS or other demyelinating diseases have not been identified.1,4,5⇓⇓ We have, however, defined the following operational criteria for the initial diagnosis of ADEM:
-
Clinical examination—Acute neurologic symptoms without a history of previous, unexplained neurologic symptoms. We excluded patients with transverse myelitis or unilateral optic nerve neuritis6 as the only neurologic deficit.
-
Cranial MRI—One or multiple supra- or infratentorial demyelinating lesions. Absence of “black holes” on T1-weighted MRI as a sign of a previous destructive inflammatory–demyelinating process. Because MRI abnormalities in ADEM are highly variable,4,5,7⇓⇓ more specific radiologic criteria are not available.
-
Lumbar puncture—Exclusion of CNS infection, vasculitis, or other autoimmune disease with CSF analysis, including additional serologic and microbiologic tests.
We were not able to define more specific criteria from the previously published evidence. Therefore, we re-evaluated the initial diagnosis of ADEM at the follow-up examination. The final diagnosis of ADEM was only established if there was no evidence of a second clinical episode of CNS demyelination.
Clinical evaluation at initial presentation.
The findings at initial presentation were collected by chart review. All patients with the initial diagnosis of ADEM were hospitalized. We collected the following information from the clinical evaluation upon first presentation: age, sex, history of prior neurologic disease, duration of symptoms, and history of preceding disease—particularly, infection or immunization within the last 4 weeks. From the neurologic examination, the presence of motor or sensory deficits, ataxia, brainstem symptoms, loss of consciousness, fever, seizures, aphasia, and meningism was noted. Neurologic disability was assessed using the Kurtzke Expanded Disability Status Scale (EDSS). In patients who continued to deteriorate after admission, we used the most severe symptoms to describe the deficits at onset of disease. The neurologic examination was repeated upon discharge and the length of hospital stay was documented.
Neuroradiologic examination.
From the study population of 40 patients, 39 were examined with cranial MRI before initiating any treatment. We also included one patient with a pathologically confirmed diagnosis of ADEM in whom a contrast-enhanced CT scan only was available. This patient presented with a fulminating course and died within a few days of admission.
Our study population was collected over a period of 11 years, and many patients referred to our institution had external MRI studies. Therefore, the MRI scans analyzed for this study were performed with various scanners, using different scanning protocols. Standard T1- and T2-weighted images were performed in all patients. In 31 patients, the initial radiologic workup also included images after IV injection of gadolinium–diethylenetriaminepentaacetic acid (DTPA).
We scored the following items from the T2-weighted images: total number of lesions; the presence of supratentorial, periventricular, subcortical, basal ganglia (including internal capsule), thalamus, brainstem, cerebellar, and peduncular lesions; and lesions within the corpus callosum. We analyzed the gadolinium-DTPA–enhanced T1-weighted images for the number of all enhancing lesions and the pattern of enhancement. We also tested whether Paty’s MRI criteria for MS were useful to discriminate between ADEM and MS.8 MRI results were classified as strongly suggestive of MS if 1) four lesions were present (Paty A), or 2) three lesions were present, one periventricular (Paty B).
A short-term follow-up examination during the initial treatment period was performed if a patient progressively deteriorated following admission. Because early follow-up MRI studies were not done systematically, only the first diagnostic MR images were employed for comparative analysis.
CSF examination.
Lumbar puncture was performed immediately after admission in all patients. CSF analysis included leukocyte count and cytology, and determination of total protein, albumin fraction, and levels of immunoglobulin (Ig) G, M, and A. The intrathecal synthesis of these immunoglobulins was determined according to an empirical differentiation scheme.9 Oligoclonal bands of IgG were evaluated by isoelectric focusing. Oligoclonal bands were simultaneously analyzed in serum and CSF and were considered positive if they were absent in the serum. CSF culture and an extensive microbiologic workup for bacterial and viral CNS infection were undertaken in all patients.
Treatment.
We treated all patients according to a standardized institutional protocol. The initial, standard treatment consisted of methylprednisolone 500 mg daily IV for 5 days. In patients who responded well to methylprednisolone clinically, no further therapy was employed. In patients who responded positively to methylprednisolone but in whom moderate to severe deficits persisted after the treatment period of 5 days, methylprednisolone was continued orally and slowly tapered over 4 to 6 weeks. Patients with poor response or no response or who progressively deteriorated during therapy with methylprednisolone were treated with cyclophosphamide (1 g IV, in some patients repeatedly, depending on the clinical course), or immunoglobulin (0.4 g/kg IV daily for 5 days).
Follow-up examination.
At follow-up examination, any neurologic symptom that had developed in the meantime was documented and a clinical examination was performed. Based on history and clinical evaluation, each patient was classified for the final diagnosis of either ADEM or MS. To establish the diagnosis of MS, we employed the criteria of Poser et al.10 for clinically definite MS. In 34 patients cranial MRI examination was repeated. We could not perform a follow-up MRI in six patients owing to death (n = 2), pregnancy (n = 1), and claustrophobia (n = 3).
Statistical analysis.
Statistical analysis was performed on a personal computer, using statistical software (Statview 5.0, SAS Institute, Cary, NC). Chi-square and Fisher’s exact tests were used to determine differences between the groups of patients with MS and monophasic ADEM. Continuous variables were tested with the nonparametric Mann–Whitney test to detect differences between the groups. The level of significance (α) chosen was p < 0.05 for all statistical analyses.
The study protocol was approved by the local ethics committee. Informed consent for performing additional clinical and MRI examinations and data analysis was obtained from each patient.
Results.
A total of 40 patients, 28 women and 12 men, with a mean age of 33.5 years (15 to 68 years), entered the study. The mean interval between initial presentation and follow-up examination was 38 months (8 to 137 months). At initial presentation, all patients fulfilled our criteria for ADEM. In four patients, a brain biopsy was done to exclude CNS lymphoma. In one patient, the diagnosis was established postmortem. At follow-up examination, 14 patients had experienced a second clinical episode and thus, according to the criteria of Poser et al.,10 were finally diagnosed with MS. In all patients who progressed to MS, the second episode occurred within the first year of initial presentation.
Clinical signs.
The basic patient characteristics, separated for the final diagnosis ADEM or MS, are depicted in table 1. Patients with ADEM had a higher rate of infection before initial presentation (p = 0.04). Of the 14 patients with previous infections, 10 patients had nonspecific upper respiratory tract infections, one patient had pleural empyema after lung surgery, and three patients had systemic signs suggestive of viral infection. The duration of symptoms before admission was shorter in patients with ADEM (p < 0.01), indicating a more abrupt onset of symptoms. One patient, who later progressed to MS, experienced the first symptoms a few days after active immunization against diphtheria and tetanus.
Basic characteristics
Table 2 shows the most frequent clinical findings at onset in both patient groups. The incidence of fever, loss of consciousness, and meningism was low; however, these symptoms were uniformly restricted to patients with ADEM. Many patients (62%) with ADEM presented with brainstem signs, whereas in patients with MS brainstem signs were only infrequently present (p = 0.02). In the patients with ADEM, the most frequent brainstem sign was ocular motor deficit, which present in 7 patients (27%), followed by dysarthria (19%) and various other brainstem signs.
Clinical symptoms at onset
CSF analysis.
The CSF characteristics are shown in table 3. There were no significant differences between the CSF findings in patients with a final diagnosis of ADEM or MS except for the CSF albumin fraction, which was higher in patients with ADEM (p = 0.006), though there was a wide zone of overlapping values. In particular, the presence of oligoclonal bands of IgG did not discriminate ADEM from MS.
CSF characteristics
Radiologic data.
Table 4 presents the results of the radiologic evaluation. Of all patients, 13 had solitary lesions, eight patients had two or three lesions, and four or more lesions were present in 19 patients. Infratentorial lesions were associated with ADEM (p = 0.045), but there were also four patients with infratentorial lesions who progressed to MS. Patients with MS had a higher rate of periventricular lesions, although this difference did not reach significance.
Radiologic data from the initial examination
In 31 patients, imaging was performed after IV administration of gadolinium-DTPA. Of the 21 patients with ADEM in whom gadolinium-DTPA was administered, enhancement of all lesions was present in 15 patients (71%); in six patients, (24%) enhancement was present in at least one, but not in all lesions; and one patient had no enhancing lesion. Ring enhancement was present in seven patients (35%). There were no differences in the frequency or pattern of contrast enhancement after gadolinium-DTPA between the patients with the final diagnosis of ADEM or MS. Of the 10 patients with MS in whom gadolinium-DTPA was administered, six patients showed enhancement of all lesions; in two patients, enhancement was present in some but not all lesions; and two patients had no enhancement at all. Ring enhancement was seen in four patients (44%) with MS.
A short-term follow-up MRI during the acute phase was performed in 13 patients who progressively deteriorated following admission. New lesions were seen in six of these patients, and in nine patients, an extension of pre-existing lesions was present.
At follow-up examination, a cranial MRI was performed in 34 patients. The MRI showed new lesions in 13 of 14 patients with definite MS. One patient with MS had developed transient symptoms typical for MS in the interim without radiologic evidence of new lesions. Follow-up MRI was performed in 20 of 26 patients with the final diagnosis of ADEM. In six of these patients, the MRI findings were completely normal. In 11 patients, partially resolved lesions were still present. Three patients with ADEM had new lesions but no new clinical symptoms, therefore not fulfilling the criteria for clinically definite MS. Because we did not systematically perform short-term follow-up MRI examinations during the acute phase of the disease, we could not determine whether the new lesions on follow-up MRI in these three patients still originated from the initial phase or from a subsequent, subclinical episode.
Treatment and clinical course.
Most patients had a favorable outcome. All patients except for one woman who had minor symptoms resolving spontaneously were treated with IV methylprednisolone. In 22 patients who responded well to methylprednisolone but in whom moderate to severe deficits were still present after 5 days, methylprednisolone was continued orally and slowly tapered over the course of several weeks. In seven patients with a poor response to methylprednisolone or progressive deterioration, cyclophosphamide pulse therapy was initiated, and one patient received immunoglobulin. With this treatment regimen, comparing the worst clinical state with the findings at discharge, 31 patients had significantly improved, seven were without any remaining symptoms, one remained unchanged, and one patient died. Patients with a final diagnosis of ADEM tended to have more severe initial symptoms. However, patients with the final diagnosis of ADEM seemed to respond better to the treatment; in patients with ADEM, the median reduction of the EDSS from the worst clinical state to discharge was 3 points (range, 0 to 9.5 points), whereas the mean reduction of the EDSS in patients with MS was 1.5 points (range, 1.5 to 3.5 points). At follow-up examination, 12 (46%) patients with ADEM did not have any symptoms; 9 (35%) had residual, minor, and nondisabling symptoms; and 3 (12%) had moderate deficits. Among patients with MS, only 2 (14%) had no clinical symptoms; 6 (43%) had minor, nondisabling symptoms; and 6 (43%) had disabling, moderate to severe deficits (median EDSS score 2; range, 0 to 6.5 points).
Discussion.
In our population of 40 adults, we found that 35% of all patients initially diagnosed with ADEM developed clinically definite MS over a mean observation period of 38 months. There are several reasons for our findings. Until now, there have been only case reports or small case series of adult patients with ADEM, and follow-up examinations were not performed in most studies. However, 17 of 121 children initially diagnosed with ADEM later developed various categories of MS.11 In our study, preceding infection or immunization was not a prerequisite for the diagnosis of ADEM. This also could have affected the high rate of patients progressing to MS. In patients with ADEM, a preceding infection was present significantly more often than in patients with MS. In two small case series, no patient with postinfectious ADEM developed MS over the observation period.4,12⇓ Nevertheless, most authors agree that infection or immunization is not a prerequisite for the diagnosis of ADEM.7,13,14⇓⇓ In elderly patients in particular, the rate of prior infection could possibly be lower than in children.15 In our patients, the restriction to patients with postinfectious or postvaccinal ADEM would not have changed our main results: two of 14 patients with the final diagnosis of MS had an infection before the first presentation, and the only patient with postvaccinal ADEM eventually progressed to MS. Fever and infection are well-known trigger factors for MS attacks,16,17⇓ and therefore, the presence of prior infection cannot be expected to reliably distinguish a first episode of MS from monophasic ADEM.
ADEM seems to occur more frequently in children, and most previous studies were conducted in the pediatric population. The development of MS in small children is rare, and therefore, the greater rate of progression to MS in our adult patient population could be partly explained by this age difference. In a large cohort study of patients with bilateral optic neuritis, only the adult patients presented a certain risk of developing MS.18
To further complicate the diagnosis of ADEM, many authors have described patients with epi-sodes of “recurrent ADEM,” usually triggered by infections.13,19-23⇓⇓⇓⇓⇓ However, criteria for the differentiation between recurrent ADEM and multiple episodes of MS have not yet been developed. In our study, patients with recurrent ADEM would have fulfilled the criteria of Poser et al.10 for clinically definite MS.
Our study has its limitations owing to the different length of follow-up and, in some patients, the relatively short observation period. Although all patients who progressed to MS experienced the second clinical episode within 1 year, an even higher rate of conversion to MS might be possible with a longer observation period.
There were some significant differences between the patients with ADEM and those with MS. Patients with ADEM were younger, had a more acute onset of symptoms, frequently had a preceding infection, more often had a disturbed blood–brain barrier, and presented more often with clinical and radiologic signs of infratentorial lesions. However, there was a wide zone of overlap for all of these factors, and we found no radiologic, clinical, or CSF symptoms that clearly indicated the final diagnosis. In general, patients with ADEM tended to present with a more acute, widespread CNS disturbance, causing loss of consciousness and multifocal neurologic signs. Fever, loss of consciousness, and meningism were infrequently observed, but were seen only in patients with the final diagnosis of ADEM. These symptoms are rare in patients with MS. In our patient population, motor deficit was the most frequent symptom upon presentation. Our results confirm the findings of a study in pediatric patients with ADEM in whom motor deficits were present in 89% of 79 children.13 Bilateral optic neuritis, which has been described as a common finding in ADEM,12,14⇓ was only found in one of our patients.
No CSF result satisfactorily predicted the final diagnosis. It has been proposed that the presence of CNS-specific oligoclonal bands predicts MS.7 Our findings, as well as several case reports of patients with ADEM in whom oligoclonal bands were positive,24-27⇓⇓⇓ do not confirm this point of view.
MRI findings in ADEM have been the subject of several studies.4,7,12,14,28-31⇓⇓⇓⇓⇓⇓⇓ Various “typical” signs for ADEM have been advocated, including the involvement of the basal ganglia and thalamus,14,29,31⇓⇓ presence of cortical lesions,5,7,14,31⇓⇓⇓ and predominant brainstem involvement.32 In our study, none of the MRI findings had a high specificity for either disease, and we agree with the results from previous studies, which indicate that in the individual patient, the MRI pattern may be indistinguishable and cannot predict the further course.4,5⇓ Cortical involvement or lesions in the basal ganglia were exclusively present in patients with the final diagnosis of ADEM; however, these findings were uncommon.
Demyelinating lesions in ADEM evolve over a short period of time. It has been speculated that after administration of gadolinium, “all lesions should enhance since all lesions should be active.”32,p 801 However, this hypothesis has been contradicted by newer studies, which demonstrated that lesions in ADEM may evolve over several weeks. There may be enhancement of some of the lesions without enhancement of others, or there may also be no enhancement at all.5,14,29⇓⇓ Our results confirm these findings: In 29% of our patients with confirmed ADEM, some of the lesions showed no enhancement. With conventional MRI techniques, the age of nonenhancing lesions cannot be determined with certainty except for “black holes” on T1-weighted images, indicating a previous destructive inflammatory–demyelinating process. New MRI techniques may allow determination of the age of demyelinating lesions with greater reliability; thus, previously clinically silent episodes of CNS demyelination may be detected in patients presenting with a first clinical episode.
Treatment of ADEM has not been systematically studied, and spontaneous improvement has been repeatedly noted.11,28⇓ With our therapeutic regimen of methylprednisolone as the first-line treatment and cyclophosphamide or Ig in patients who did not respond to methylprednisolone, nearly all patients improved during the acute phase. Response to steroid therapy has been reported by most authors.11,15,33,34⇓⇓⇓ Ig, plasmapheresis, and cytostatic drugs are alternative treatment options in patients who do not respond to steroid therapy.27,35-39⇓⇓⇓⇓⇓ Except for its fulminant form, the acute hemorrhagic necrotizing encephalomyelitis (Hurst syndrome), the prognosis of ADEM has been uniformly reported as favorable. However, nearly half of our patients with ADEM had remaining symptoms at follow-up.
- Received August 10, 2000.
- Accepted December 2, 2000.
References
Letters: Rapid online correspondence
REQUIREMENTS
You must ensure that your Disclosures have been updated within the previous six months. Please go to our Submission Site to add or update your Disclosure information.
Your co-authors must send a completed Publishing Agreement Form to Neurology Staff (not necessary for the lead/corresponding author as the form below will suffice) before you upload your comment.
If you are responding to a comment that was written about an article you originally authored:
You (and co-authors) do not need to fill out forms or check disclosures as author forms are still valid
and apply to letter.
Submission specifications:
- Submissions must be < 200 words with < 5 references. Reference 1 must be the article on which you are commenting.
- Submissions should not have more than 5 authors. (Exception: original author replies can include all original authors of the article)
- Submit only on articles published within 6 months of issue date.
- Do not be redundant. Read any comments already posted on the article prior to submission.
- Submitted comments are subject to editing and editor review prior to posting.
