Descriptive study of acute disseminated encephalomyelitis and evaluation of functional outcome predictorsJN Panicker1, D Nagaraja1, JME Kovoor2, DK Subbakrishna3
1 Department of Neurology, National Institute of Mental Health and Neurosciences, Bangalore-560 029, India
2 Department of Neuroimaging and Interventional Radiology, National Institute of Mental Health and Neurosciences, Bangalore-560 029, India
3 Department of Biostatistics, National Institute of Mental Health and Neurosciences, Bangalore-560 029, India
Correspondence Address: Source of Support: None, Conflict of Interest: None DOI: 10.4103/0022-3859.62425
Source of Support: None, Conflict of Interest: None
Background: Outcome following Acute Disseminated Encephalomyelitis (ADEM) is variable and there are only limited studies from India. Aim: The study aims to evaluate the predictors of functional outcome in a cohort of patients with ADEM. Setting: Tertiary-care teaching hospital. Materials and Methods: Patients admitted with the diagnosis of ADEM from 1999 to 2004 have been included. Clinical features and radiological findings were evaluated. Functional outcome at discharge was scored using modified Rankin Scale and patients were followed up regularly. Statistical Analysis: Chi-Square test or Fisher's exact test, and Student's t test for comparison of categorical and continuous variables, respectively, and logistic regression for multivariate analysis. Results: Sixty-one patients were evaluated (mean age 22±15.9 years, 1-65). Fifty-two patients had preceding febrile illness or vaccination with mean 9.1±12.5 days interval to first neurological symptom. Non-specific febrile illnesses were the commonest trigger. Commonest findings were motor signs (n=41), impaired consciousness (n=33), bladder symptoms (n=21), ataxia (n=15), and seizures (n=14). Between adult (mean age 30.1±13.1 years, 13-65, n=38), and pediatric (mean age 6.2±2.8 years, 1-12, n=23) patients, language disturbances were more common in the latter (P=0.047). MR imaging (n=35) demonstrated lesions mostly in frontoparietal white matter (n=23) and thalamus (n=15). Nine patients expired. Patients with poor functional outcome at discharge more often had impaired consciousness (P=0.038) and seizures (P=0.06). At follow-up (n=25), deficits included motor signs (n=15) and bladder symptoms (n=5). Conclusions: ADEM has a wide range of neurological presentations and language disturbances are more common in pediatric patients. The presence of impaired consciousness, and possibly seizures, predict poor functional outcome at hospital discharge.
Keywords: Acute disseminated encephalomyelitis, impaired consciousness, modified rankin scale, outcome, prognosis, seizures
Acute Disseminated Encephalomyelitis (ADEM) is a monophasic, inflammatory demyelinating disorder of the central nervous system (CNS). , Although the exact pathogenesis is unknown, there is evidence to suggest that it represents an autoimmune phenomenon, usually triggered by infections or vaccinations. ,, The clinical presentation is characterized by polysymptomatic neurological involvement and imaging shows multifocal demyelinating lesions in the brain and spinal cord. Clinical spectrum of disease, cerebrospinal fluid (CSF) findings, radiological features and distinguishing features from the first presentation of multiple sclerosis (MS) have been studied extensively. ,,,,,
Earlier reports have documented considerable morbidity amongst survivors of ADEM.  Although the prognosis has significantly improved with the advent of various immunomodulatory treatments, , studies that specifically evaluate predictors of functional outcome are few.  Therefore, we undertook this study to evaluate the clinical features and radiological findings that could predict functional outcome in a cohort of patients with ADEM from south India.
This study was carried out from 1999 to 2004 at a tertiary-care teaching hospital in south India. The study had Institutional Review Board approval; consent was obtained from all patients. Patients were diagnosed to have ADEM if they fulfilled the following criteria: 
Clinical features: Presentation with an acute polysymptomatic neurological disease, without a previous history of unexplained neurological symptoms.
Imaging: Cranial/spinal imaging (CT or MRI) showing one or multiple lesions suggestive of demyelination.
Serology and lumbar puncture: Exclusion of CNS infections, vasculitis and other autoimmune disorders, as described below.
Patients presenting with a clinically isolated demyelinating syndrome (eg. optic neuritis or transverse myelitis) or with clinical features to suggest additional peripheral nervous system (PNS) involvement were excluded. Any patients with clinical or radiological evidence for dissemination in time and space suggesting MS were excluded as well. Age, gender, history of antecedent infections or immunizations and symptoms at presentation and during the course of illness were evaluated. Patients were assessed by a standard neurological examination. Level of consciousness was assessed using the Glasgow Coma Scale (GCS).
Relevant investigations were performed to exclude infective or inflammatory etiologies: complete blood count, erythrocyte sedimentation rate, rheumatoid factor, LE cell, antinuclear antibody vasculitis screen and serological testing for human immunodeficiency virus. Lumbar puncture was done in all patients and CSF analyzed for cells, protein and sugar levels, viral serology (ELISA for Japanese Encephalitis and Herpes Simplex Virus antibodies), bacterial and fungal culture and cytology.
Computed tomography (CT) and/or MRI (superconducting 1.5 Tesla scanner) were performed during the acute stage of illness. CT was done using a third generation scanner and contrast enhanced images were obtained. MRI T1-weighted, T2-weighted and fluid-attenuated inversion recovery (FLAIR) sequences were acquired in the axial, sagittal and coronal planes. When feasible, contrast enhanced images were obtained using gadopentetate dimeglumine (0.1 mmol/kg). CT and MR images were read independently by a radiologist (JMEK) and number of lesions, distribution, hemorrhage, midline shift and pattern of contrast enhancement were noted. Standard 30-minute interictal surface electroencephalogram (EEG) was recorded in patients with impaired consciousness or seizures.
Patients were treated with high-dose intravenous corticosteroids, either methyl prednisolone (10-20 mg/kg) or dexamethasone (1 mg/kg), daily for three to five days. Subsequently, prednisolone (1 mg/kg orally) was started and continued for four weeks. Plasma exchange, intravenous immunoglobulin (2 gm/kg divided over five days) or repeat high-dose intravenous methyl prednisolone were given for patients who continued to deteriorate. Acutely ill patients were monitored in the intensive care unit and were ventilated when indicated. Seizures were managed with intravenous phenytoin. Functional outcome at discharge was graded using the modified Rankin Scale (mRS). This is a seven point scale from 0 (no symptoms) to six (death) and patients' outcome was classified as good (mRS≤3) or poor (mRS>3)  according to the evaluation at the time of discharge.
Patients stabilized on medical management were transferred to a comprehensive rehabilitation program. They were regularly reviewed at the out-patient clinic at six weeks and thereafter every three to six months, depending upon their clinical condition for 12 months. At each visit, the neurological condition was reassessed through history and a formal neurological examination. Repeat brain imaging was arranged if there was suggestion of progression or recurrence of neurological deficits.
Statistical analysis was performed using SPSS for Windows version 16.0 (SPSS Inc. Chicago, IL, USA). Comparison of categorical and continuous variables was performed using chi-square test or Fisher's exact test, and Student's t test, respectively. Multivariate analysis was performed using logistic regression analysis. P value of <0.05 was considered significant.
A total of 61 patients, 34 men and 27 women, were included in the study. Age ranged from 1 to 65 years (mean age 22±15.9 years). Clinical features are listed in [Table 1]. Neurological deficits of adult (mean age 30.1±13.1 years, range 13-65 years, n=38), and pediatric (mean age 6.2±2.8 years, range 1-12, n=23) patients were compared [Table 1] and language disturbances were more common in the latter (P=0.047). In 37 patients, progression of neurological deficits occurred over several days, ranging from 1 to 30 days (mean 9.1±6.4 days). A clear infectious event or vaccination preceded the illness in 52 patients (85%) [Table 2]. The interval between the preceding event and the first onset of neurological symptoms ranged from 1 to 60 days (mean 9.1±12.5 days).
Cerebrospinal fluid was abnormal in 41 (67%) patients. CSF protein was elevated in 26 patients and pleocytosis was seen in 32 (lymphocyte predominant in 29 and neutrophil predominant in 3) [Table 1]. Laboratory investigations for inflammatory or infective etiologies were normal in all patients.
CT imaging was performed in 44 patients and demonstrated hypodense lesions in the frontoparietal white matter in 31 (71%), with contrast enhancement in 3 (10%). Hyperdense lesions suggesting hemorrhage were seen in two patients. MR imaging was performed in 35 patients and the distribution of lesions in the CNS are given in [Table 3]. Contrast was administered in nine patients and lesions demonstrated uniform (n=2) or open ring (n=1) enhancement.
EEG recording was performed in 21 patients and the commonest abnormality was diffuse theta range background slowing (n=10). Other abnormal recordings included focal spike and wave discharge (n=1), diffuse beta range background slowing (n=1) and diffuse nonresponsive alpha background activity (alpha coma) (n=1). EEG was normal in eight patients. Following intravenous high-dose corticosteroids [methyl prednisolone (n=53), dexamethasone (n=8)], 46 patients showed improvement in neurological deficits. Seven continued to worsen and were treated with intravenous immunoglobulin (n=2), second course of methyl prednisolone (n=2) or plasma exchange (n=1). Four patients required mechanical ventilation. Nine patients expired, all adults (P=0.01), and causes for death included sepsis (n=7), bronchopneumonia (n=3) and cardiovascular autonomic dysfunction (n=2).
Hospital stay ranged from 5 to 16 days (mean 11.4 days). At the time of discharge, 45 patients had residual neurological deficits. Commonest deficits were motor weakness (n=36), ataxia (n=15) and visual impairment (n=3). Eleven patients had bladder symptoms. Functional outcome at the time of discharge was assessed using the modified Rankin scale (mean scores included in [Table 1]) and 34 patients had good outcome (scores≤3) while 18 had poor outcome (scores 4 or 5). Patients with poor functional outcome at discharge more often had impaired consciousness (GCS<10) at admission (P=0.038) and seizures, though the latter association did not attain statistical significance (P=0.06) [Table 4]. No clinically significant results were derived by multivariate analysis.
Twenty-five patients (48%) came for review to the outpatient clinic. Follow-up ranged from one to 12 months (mean 3.4±2.9 months). None of the patients developed recurrence of neurological symptoms during this period. Persisting complaints included motor signs (n=15) and bladder symptoms (n=5). Sixteen patients (64%) had improved score on the modified Rankin scale.
In our series, 61 patients presented over six years with clinical and radiological features suggesting ADEM; 85% of patients had a preceding infection or vaccination, most commonly a nonspecific febrile illness which was presumably viral in etiology. The nature of the preceding event did not influence the clinical features or radiological findings.  Prevalence of ADEM following immunization was low in our study compared to previous studies from India  and exclusion of patients with PNS involvement is likely to have contributed to this under-representation. 
Most patients presented with impaired consciousness, multifocal motor deficits and cranial neuropathies. As seizures, meningism and headache were common, viral or bacterial meningoencephalitis was the immediate differential diagnosis and was excluded by CSF analysis. The pattern of neurological deficits did not differ significantly from those previously reported in the literature. ,,,, Inclusion of adult and pediatric patients allowed for comparison between the two groups and language disturbances occurred more often in the latter. Though language disturbances have been described previously, ,, they are considered rare in ADEM. Diaschisis has been suggested as a possible mechanism, whereby lesions affecting the white matter tracts anatomically connected to the cortical language centers may result in aphasia.  About 50-70% of patients experience full clinical and radiological recovery over weeks , and follow-up in our patients suggested a similar pattern.
It was observed that mortality was higher amongst adults. This is similar to the findings of studies done previously that report mortality in adult patients to be 8-25% while in children to be less that 5%. ,, In spite of advances in the treatment of ADEM and care of seriously ill patients, morbidity continues to be high amongst survivors. It would therefore be pertinent to identify the subset of patients who are predicted to fare poorly. Our study demonstrates that patients with impaired consciousness are more likely to have a poor outcome at hospital discharge. The reasons for this may be multifactorial and could reflect more severe disease or increased risk for complications. ,, A similar observation has been made in a cohort of seriously ill patients being treated in an intensive care unit setting. 
The occurrence of seizures has also been show to be associated with poor outcome,  and this is due to increased risk for neurological and systemic complications. Previous series report an incidence of seizures ranging from 4% to 35%, and there appears to be no correlation with age group or imaging findings, ,,, [8, 12, 15] although it has been shown to be higher in seriously ill patients.  Though 23% of our patients developed seizures, the association with poor outcome showed only a trend for statistical significance. The semiology was most often generalized tonic clonic seizures, which is in contrast to partial motor status epilepticus that has most often been reported in the literature.  Three patients developed status epilepticus, which was controlled by intravenous phenytoin. Similar to previous observations,  diffuse background slowing was the commonest abnormality recorded in EEG. Though seizure recurrence has been reported, , this was not seen in any of our patients.
Other parameters found to predict poor outcome include extent of lesions in the white matter  and brainstem.  Patients with acute hemorrhagic leukoencephalitis are expected to have poor outcome.  However, limited number of patients with this condition restricted the analysis of this variable.
Neuroradiology is paramount in diagnosing ADEM. In contrast to the high sensitivity of MRI in detecting lesions,  CT imaging may miss small lesions especially when performed early in the illness  and findings are often nonspecific. , MRI was abnormal while CT was normal in 7 patients who underwent both CT and MR imaging. Frontal and parietal white matter, thalamus and cerebellar white matter/peduncles were most commonly involved. Deep grey matter involvement is common in ADEM and is often bilateral and symmetric. ,,,, In our series, 48% demonstrated lesions in the thalamus and 23% in the basal ganglia. In spite of extensively distributed lesions, mass effect may be relatively small  and was seen in only six patients. None of the radiological parameters were shown to predict functional outcome.
The heterogeneous nature of the study cohort was a limitation to our study and restricted the evaluation of some variables. A comprehensive search for potential ADEM-triggering pathogens was not performed and hence the pattern of antecedent infections in our geographical region was not accurately reported. Imaging was performed during the acute stage of illness and hence MRI could not be done in all patients, either because of poor general condition or due to financial constraints. The possibility that some patients had the first presentation of MS rather than ADEM could not be excluded as oligoclonal bands were not analyzed and patient review was limited to only a few months. Long-term follow-up would allow for the evaluation of persisting subtle deficits, such as neuropsychological impairment, and their impact on school or work performance, and this is an obvious area for further research.
In conclusion, ADEM has a wide range of neurological presentations and language disturbances are more common in pediatric patients. In this cohort of patients, the presence of impaired consciousness and seizures were the only clinical features shown to predict poor functional outcome at hospital discharge.
[Table 1], [Table 2], [Table 3], [Table 4]