Clinical manifestations

Presentation and course

Because of their size, location, and potential for associated mass effect and edema, tumefactive demyelinating lesions cause neurologic symptoms atypical for multiple sclerosis. Among the largest cohort of biopsy-proven tumefactive demyelinating lesions analyzed to date (n=168), common presenting symptoms, in decreasing order of frequency included motor, cognitive, cerebellar, and brainstem dysfunction (47). Additional symptoms included headache, seizures, aphasia, cortical sensory loss, or psychosis. Unlike typical CNS demyelinating lesions, tumefactive lesions are more likely to cause cortical signs, such as altered mental status and visual field deficits. The tumefactive episode represented the initial event in 61% of cases; however, 29% of patients had a prior history of relapsing neurologic symptoms, and 5% carried an established diagnosis of multiple sclerosis.

Research suggests that two thirds of patients with tumefactive demyelinating lesions subsequently develop a relapsing-remitting disease course with more typical multiple sclerosis lesions, although a small subset will have tumefactive lesions at relapse (47; 02). Their median time to the second relapse was 4.8 years, an interval longer than reported for prototypic multiple sclerosis. The remaining one third of patients had a monophasic illness at last follow-up. These findings contrast with an earlier shorter duration study, with less than 5 years of follow-up for most patients, suggesting that patients with biopsy-proven tumefactive demyelinating lesions uniformly have a monophasic disease course (37). Another study, also with less than 5 years of follow-up for most patients, found that overall, two thirds of patients had a monophasic course within this time frame, with 50% of those with a ring-enhancing lesion on initial presentation relapsing, though Lucchinetti did not find a similar association. The presence of concomitant typical multiple sclerosis-appearing lesions also increased the probability of relapse (41.3% vs 9.6%, p< 0.005) independent of the imaging characteristics of the initial tumefactive lesions (88). Based on Lucchinetti’s finding that the average time to relapse after a tumefactive demyelinating lesion is 4.8 years, it is likely that an overall higher relapse rate would be detected with follow-up beyond 5 years. In infiltrative lesions, ring enhancement, multiple lesions, and infratentorial involvement were independent risk factors for relapse (43). Of this cohort, 48.3% had a definite etiology, most commonly multiple sclerosis, followed by MOG-associated disease, Balo concentric sclerosis, neuromyelitis optica spectrum disorders, and acute disseminated encephalomyelitis.

The prognosis of a cohort of 39 patients in India with tumefactive demyelinating lesions differed markedly from Lucchinetti’s, demonstrating a median time to relapse of 4 months, with the majority of patients having tumefactive lesions at relapse (57). Importantly, in an Indian cohort the incidence of multiple sclerosis is lower than in the United States and the incidence of ADEM is higher, and the cohort included children, who have a higher incidence of ADEM. The discrepancies between findings in these various studies highlight the challenges of studying tumefactive demyelinating lesions, which are rare, may require long-term follow-up to detect relapses, and are caused by a heterogeneous group of diseases.

Prognosis and complications

The prognosis for patients with tumefactive demyelinating lesions is impacted primarily by the underlying CNS idiopathic inflammatory demyelinating disease, the location and size of the tumefactive lesions, the extent of mass effect, and the response to initial treatment. In general, long-term prognosis for tumefactive demyelinating lesions is favorable if recognized early and treated appropriately. Prognosis depends on whether the patient has a prior diagnosis of multiple sclerosis or is presenting with tumefactive demyelinating lesions as their first clinical event. Reportedly, 6% to 70% of patients presenting with a tumefactive demyelinating lesion as the first clinical event will develop a clinically definite multiple sclerosis (27). Despite this broad range, it is generally accepted that about two thirds of patients will follow a relapsing-remitting course and about one third will have no further clinical attacks (01). A small subset of patients presenting with tumefactive demyelinating lesions may relapse in a strictly tumefactive fashion. As noted above, the presence of ring enhancement on MRI and other typical multiple sclerosis-appearing lesions at presentation are associated with a higher incidence of relapse (88). Most studies agree that tumefactive demyelinating lesions do not typically evolve into a more active subtype of multiple sclerosis and recovery from tumefactive lesions is, in general, favorable (47; 02; 88; 11). Long-term disability associated with tumefactive demyelinating lesions was similar to a population-based multiple sclerosis cohort matched for disease duration (47). Index lesion size, location, and presence of mass effect or edema are not predictive of conversion to multiple sclerosis (47). However, very large initial lesion size (greater than 35 mm), mass effect, perilesional edema on MRI, and age greater than 30 years at initial presentation predicts a more aggressive disease course (64). Isolated intramedullary tumefactive lesions carry a worse prognosis, with only one of 14 patients in a mixed pediatric and adult cohort achieving full recovery (62).

Clinical vignette

A 23-year-old woman initially presented at 12 years of age with new-onset progressive left hemiparesis. The symptoms started in the left hand, then progressed to involve the left face and left lower extremity over a period of 1 week. Brain MRI revealed a large open-ring, gadolinium-enhancing lesion involving the right frontal lobe.

She underwent a craniotomy, and brain biopsy confirmed inflammatory demyelinating disease. She was treated with intravenous corticosteroids and completely recovered over a period of 5 weeks. At the age of 18 years, she had new-onset difficulty in language comprehension that progressed over 3 days. Brain MRI showed a new, large ring-enhancing lesion involving the left frontal lobe.

She was treated with intravenous corticosteroids and again had complete recovery over 4 to 5 weeks. At the age of 22 years, she developed severe early morning headaches. Brain MRI showed a new large, ring-enhancing lesion involving the right frontal lobe.

She was treated with intravenous corticosteroids and had a rapid clinical improvement. At the age of 23 years, she had ascending sensory loss to the umbilicus associated with urinary urgency. She was treated with intravenous steroids and completely recovered 11 days later. She was started on multiple sclerosis disease-modifying therapy with interferon beta1a weekly intramuscular injections for relapsing-remitting multiple sclerosis. Her EDSS score was 1.0 at last follow-up, 13 years after disease onset.

Biological basis

Etiology and pathogenesis

Pathology and pathogenesis differ depending on the type of CNS idiopathic inflammatory demyelinating disease.

Tumefactive demyelinating lesions.

Pathology. Tumefactive demyelination typically features classic characteristics of active inflammatory demyelinating disease including focal demyelination, variable inflammation, gliosis, and relative axonal sparing.

Up to 31% of diagnostic biopsies from patients with demyelinating lesions are misinterpreted. The most common misdiagnoses include glioma and infarction. Pathological features that cause difficulty in interpretation include hypercellularity, presence of necrosis and cystic changes, and the presence of reactive astrocytes with bizarre appearance, multiple nuclei, and mitotic figures, referred to as Creutzfeldt cells (91; 03; 19; 01). Zagzag and colleagues noted that the presence of foamy macrophages is an important feature suggestive of a demyelinating etiology. Once this is suspected, staining for myelin and axons can confirm the diagnosis (91). Sampling may also contribute to misinterpretation of demyelinating biopsies. Samples from the lesion edge may be misinterpreted as glioma; whereas samples from the lesion center may resemble infarct due to overabundance of macrophages (03; 01).

Pathogenesis. The immunopathogenesis of tumefactive demyelinating lesions remains uncertain. Early case reports suggested an association between tumefactive demyelinating lesions and vaccinations, leading to the hypothesis that these lesions represent an intermediate phenotype between multiple sclerosis and acute disseminated encephalomyelitis. However, this association was not found in subsequent studies (37; 47; 02; 27). It has been argued that as these lesions can develop in patients who typically go on to develop classic multiple sclerosis, this suggests that the lesions are part of the heterogenous spectrum of multiple sclerosis and not a transitional phenotype between acute disseminated encephalomyelitis and multiple sclerosis (37), recurrent acute disseminated encephalomyelitis (10), or a tumefactive multiple sclerosis variant (67).

The association of various immunosuppressants and occurrence of cases of tumefactive demyelinating lesions in association with HIV and acquired lymphopenia suggests a possible relationship between an adaptive cell-mediated immune defect and tumefactive demyelinating lesion pathogenesis rather than limiting the etiology to multiple sclerosis alone (41).

Marburg acute multiple sclerosis.

Pathology. Lesions may be disseminated throughout the brain and spinal cord, ranging in size from small to large confluent plaques. In some cases, there is diffuse demyelination throughout the brain and spinal cord. Microscopically, lesions are more destructive than typical multiple sclerosis lesions and characterized by significant macrophage infiltration, acute axonal injury, necrosis, and cavitation. Despite the degree of tissue destruction, variable remyelination is also evident.

Pathogenesis. The etiology of Marburg variant of multiple sclerosis is unclear. Myelin basic protein in patients with Marburg acute multiple sclerosis was found to have 18 citrullinyl residues compared with six in patients with chronic multiple sclerosis. The number of citrullinyl residues is thought to result in structural instability of myelin, causing severe demyelination as is seen in these patients (07). In a case report of a serial biopsy from a patient with Marburg acute multiple sclerosis, the initial biopsy (done on day 33) showed marked inflammation in the absence of demyelination, which became evident by the subsequent biopsy (day 109 of disease) (09). These findings are consistent with inflammation preceding demyelination.

Balo concentric sclerosis.

Pathology. Patients with Balo concentric sclerosis may have one or more lesions in the cerebral white matter, often sparing the cortical U fibers. Less common sites of demyelination associated with Balo concentric sclerosis include the basal ganglia, pons, cerebellum, and, rarely, the spinal cord and optic nerves. The pathological hallmark consists of concentric demyelination and oligodendrocyte loss characterized by alternating bands of demyelination with normal myelination or partial demyelination giving the appearance of onion bulbs. Lesions exhibit pattern III demyelination, characterized by distal oligodendrocyte dystrophy, composed of T lymphocytes and microglia and an absence of immunoglobulin and completement (35).

Pathogenesis. This pathological pattern of demyelination and tissue injury resembles a pattern seen in tissue ischemia. Similar to acute ischemia, a preferential loss of myelin-associated glycoprotein and oligodendrocyte apoptosis is seen in a subset of multiple sclerosis lesions. Because myelin-associated glycoprotein is localized at the distal inner glial loop, its early loss is interpreted as evidence of a dying-back gliopathy where the oligodendrocyte is unable to support the distal axon.

Different cellular responses to injury are found in the different rings of Balo concentric sclerosis lesions and display similarities to cellular mechanisms observed in hypoxia. In active areas of demyelination, macrophages, and microglia express nitric oxide synthase, indicating an attempt, although unsuccessful, to mitigate damage. In areas with rings of preserved myelin, oligodendrocytes express protective proteins involved in tissue preconditioning (eg, hypoxia-inducible factor 1alpha and heat shock protein 70). Astrocytes and macrophages in the same location also express these proteins, though to a lesser extent. This cellular rim is more resistant to further damage and forms the layer of preserved myelin in the lesion. This overexpression of tissue preconditioning molecules near the layer of active demyelination supports a role for histotoxic hypoxia in mediating the concentric pathology of demyelination (75). Studies have also found decreased levels of aquaporin 4 and connexins, highlighting the role of astrocyte dysfunction in Balo concentric sclerosis (45).

Epidemiology

The incidence of tumefactive demyelinating lesions is low, and precise estimates vary. A 2021 review of patients with tumefactive demyelination in Olmsted County, Minnesota found that out of 792 multiple sclerosis cases, 15 (1.9%) had tumefactive multiple sclerosis (22). It was the first clinical multiple sclerosis attack in half of the patients. The adjusted overall annual incidence rate was 0.57 out of 100,000.

A number of authors have tried to estimate the prevalence of tumefactive demyelinating lesions based on reviews of brain biopsies. Sugita and colleagues reviewed brain biopsies from 1231 cases coded as brain tumor or “unclassified” and found demyelination in three patients (0.24%) (76). Hunter and colleagues reported five of 1220 brain biopsies done over a 5-year period were consistent with inflammatory demyelinating disease (30). Two of the five biopsies were from the same patient. Annesley-Williams reported 0.09% of 15,394 retrospectively reviewed brain mass biopsy and autopsy specimens over a 22-year period demonstrated evidence of acute demyelination. Two of their patients had a prior history of multiple sclerosis (03).

Tumefactive demyelinating lesions typically present in the third decade (47; 02; 88) although Schilder disease and ADEM typically occur in children. There is some variability in age of presentation. Case reports describe a 13-month-old child (48) and an 87-year-old woman with pathologically confirmed tumefactive CNS demyelination (80). In the biopsy cohort of Lucchinetti and colleagues, 4% were younger than 18 years and 4% older than 65 years. A predilection for women may be present, although findings from studies differ. The ratio of women to men in Lucchinetti’s cohort was 1.2:1, suggesting an almost equal ratio. However, selection bias may have played a role in this biopsy-based cohort if clinicians are less likely to suspect inflammatory demyelination in men and, therefore, are more likely to biopsy. Two studies with patients enrolled based on radiographic findings of tumefactive demyelinating lesions found 62% to 68% of patients were women (02; 88).

Differential diagnosis

Making the diagnosis of tumefactive CNS idiopathic inflammatory demyelinating disease can be challenging. In addition to mimicking other diseases, such as brain tumors, tumefactive demyelinating lesions can be seen in other conditions (see Table 1). A detailed history and examination are essential to determine if there is evidence of previous demyelinating relapses. If CNS idiopathic inflammatory demyelinating disease is suspected, empiric therapy with intravenous steroids can be initiated. Tumefactive demyelinating lesions can look radiographically identical to high-grade gliomas on initial imaging; however, tumefactive demyelinating lesions will regress with serial imaging and generally respond clinically to high-dose steroids whereas high-grade gliomas will grow over time and not respond substantially to steroids (44). However, if CNS idiopathic inflammatory demyelinating disease is considered unlikely, steroids should be delayed to avoid obscuring the diagnosis, particularly of CNS lymphoma (56). Both tumefactive demyelinating disease and CNS lymphoma often demonstrate a rapid radiographic response to steroids, making diagnosis in the absence of biopsy challenging. Additional important considerations are that tumefactive demyelination may be the only finding on biopsy in patients with CNS lymphoma not previously treated with steroids (90) and that patients with known multiple sclerosis can develop brain tumors (12). Onset age of patients with tumefactive demyelinating lesions is younger than in CNS lymphoma, with an age of 37 +/- 14 years in tumefactive demyelinating lesions compared to 58 +/- years in primary CNS lymphoma (77). Creutzfeldt-Peters cells and reactive astrocytes with fragmented nuclear inclusions on biopsy correlated with tumefactive demyelinating lesions. Pathological immune features, including type of demyelination, infiltrating cell type distribution, specific astrocyte pathology, and complement deposition, can be helpful in differentiating inflammatory etiology between multiple sclerosis, MOG-associated disease, or neuromyelitis optica spectrum disorder secondary to AQP4 (86).

Clinical and radiographic follow up are key, and repeat biopsy should be considered in cases where the diagnosis remains unclear. The differential diagnosis of tumefactive demyelinating lesions is wide and should be tailored to the clinical and radiological features. Table 1 summarizes the differential diagnosis of CNS idiopathic inflammatory demyelinating disease.

It is important to recognize that patients with multiple sclerosis may also develop neoplasms. A review of gliomas in patients with multiple sclerosis found that 30% of gliomas associated with multiple sclerosis were multicentric or diffusely infiltrative as compared to 2.5% to 5% of gliomas not associated with multiple sclerosis, suggesting a causal relationship to the multifocal disease of multiple sclerosis (70). Case reports vary as to whether the neoplastic process was adjacent to or infiltrated the multiple sclerosis lesions. Shuangshoti and colleagues hypothesized that neoplastic cells may originate from reactive astrocytes in inflammatory lesions (74). In patients with multiple sclerosis and oligodendrogliomas, a causal relationship can be hypothesized, with damage to oligodendrocytes causing oligodendrocyte proliferation in an effort to remyelinate axons, making the oligodendrocytes susceptible to malignant transformation (70). Based on the existing data, it is difficult to draw conclusions about whether there is a true association between multiple sclerosis and CNS neoplasia. However, the potential for concurrence should raise suspicion of an alternative diagnosis in a multiple sclerosis patient developing a new atypical mass lesion. Therefore, there should be careful clinical and radiographic follow-up of patients with multiple sclerosis presenting with new mass lesions.

Table 1. Differential Diagnosis of CNS Idiopathic Inflammatory Demyelinating Disease

Diagnostic workup

Diagnostic work-up should be focused on the specific clinical scenario, with careful attention to the symptoms, signs, and setting. In a patient known to have multiple sclerosis, the diagnosis is more obvious. The workup of tumefactive demyelination is associated with higher cost and more adverse events than evaluation of conventional multiple sclerosis. One review of 32 patients with tumefactive demyelination and 32 with conventional multiple sclerosis noted that most of this financial burden and morbidity was related to biopsy, which is often necessary to rule out malignancy (73). Table 2 shows a list of possible diagnostic studies that can be considered in a patient suspected of having tumefactive demyelination.

Table 2. Diagnostic Work-up For a Patient with Possible Tumefactive Demyelination

Magnetic resonance imaging (MRI). Tumefactive demyelinating lesions tend to be well-circumscribed and typically involve white matter; however, gray matter lesions may also occur. Most patients have multifocal lesions; however, typically there is a large dominant lesion (17; 47). The large lesion is often supratentorial and usually involves the frontal or parietal lobes; however, periventricular, juxtacortical, subcortical, or corpus callosal lesions may also occur. Lesions involving the corpus callosum may have a butterfly configuration. This was observed in 12% of lesions in Lucchinetti’s cohort (47). In addition, 45% of lesions were associated with mass effect and two thirds were associated with edema, though the degree of mass effect and edema tend to be less than is observed with tumors. Almost all lesions enhance with gadolinium with a variety of enhancement patterns. The most common patterns of enhancement are irregular rim and ring enhancement (either open- or closed-ring); however, other patterns are also observed (40; 47).

Ring-like lesions are associated with a higher rate of relapse (88). A pattern of open-ring enhancement with the ring open to the gray matter side can be an important clue to the presence of tumefactive demyelination (52). A radiographic-pathologic correlation study found that MRI patterns of open-ring and irregular rim enhancement were associated with infiltration of macrophages and angiogenesis, whereas a pattern of inhomogeneous enhancement was associated with perivascular lymphocytic cuffing (40). Lesions with an inflammatory core can show increased intensity on diffusion-weighted sequences with correlated low signal of apparent diffusion coefficient sequences. Diffusion restriction may also be seen peripherally in the lesion. These correlations are similar to prior studies in multiple sclerosis without tumefactive demyelination. T2-weighted MRI allows calculation of cerebral blood volume to determine vascularity of lesions. Tumefactive demyelinating lesions exhibit normal vasculature, whereas tumors provoke neovascularization and angiogenesis. A ratio of relative cerebral blood volume to contralateral normal white matter yielded a statistically significant difference in intracranial neoplasms versus demyelinating lesions (15). Follow-up MRI can be helpful in distinguishing etiology for tumefactive demyelinating lesions, with resolution more common in MOG-associated disease than multiple sclerosis or neuromyelitis optica spectrum disorders associated with AQP-4 (13).

Magnetic resonance spectroscopy. The role of magnetic resonance spectroscopy in the diagnosis of tumefactive demyelination is not clear and studies have had conflicting results. There are differences between pseudotumoral demyelinating lesions and solid brain masses in several metabolites, including myo-inositol, NAA, glutamine, and choline (49). A serial study of a demyelinating lesion demonstrated an initial peak in lipid and lactate that normalized after 4 weeks. Reduction in NAA (neuronal marker) correlates with the degree of axonal damage in the lesion (20). Gliomas and acute demyelinating plaques show a similar spectral pattern of increased Cho/Cr and reduction in NAA/Cr (42). A larger study was able to differentiate tumefactive demyelinating lesions from high-grade gliomas by means of the mean Cho/NAA ratio, with encouraging sensitivity and specificity. However, this study, like the others, was small, with only four patients with tumefactive demyelinating lesions. Although this technique shows promise for minimizing costs and morbidity, it is not yet as effective as tissue biopsy in securing an accurate diagnosis. However, Cho/NAA as an adjunct to conventional MRI does appear to improve diagnostic ability (46).

Computed tomography. CT may serve as a helpful adjunct to MRI in distinguishing tumefactive demyelinating lesions from tumors. Areas of MRI enhancement are hypodense on unenhanced CT scan significantly more often with tumefactive demyelination (93%) than with tumors (4%), including CNS lymphoma (38). Patients with primary CNS lymphoma are more likely to have hyperdense lesions (77).

Positron emission tomography. A small study of five patients using fluoro-deoxyglucose PET showed a marginal increase in metabolism in tumefactive lesions compared to the typical marked increase in metabolism seen in tumors, making PET potentially informative in distinguishing these lesions from neoplasms (79).

Another PET modality, using the translocator protein (TSPO) radiotracer, was used to compare tumefactive and Balos concentric sclerosis lesions to relapsing-remitting multiple sclerosis lesions. TSPO detects activated microglia and CNS macrophages and is a promising approach for assessing inflammation. Tumefactive multiple sclerosis shows diffuse TSPO tracer uptake, even beyond the region of contrast enhancement on MRI. In Balos concentric sclerosis, PET shows concentric circular areas of inflammatory cell activation that correlate with enhancement on MRI (87).

CSF studies. CSF studies are often avoided in patients with large tumefactive lesions associated with mass effect due to concerns of raised intracranial pressure. As such, there are limited large retrospective reviews on this topic. In the study of Lucchinetti and colleagues, 62 of the patients had CSF studies prior to brain biopsy. In 33%, oligoclonal bands were present, and 35% had an elevated IgG synthesis rate (47). This study was retrospective, and CSF analyses were done in different laboratories under different conditions; therefore, firm conclusions are difficult to draw. However, it is possible that patients with tumefactive demyelinating lesions have a lower frequency of oligoclonal bands and elevated IgG synthesis rate compared to prototypic multiple sclerosis. This may also be a result of selection bias in a biopsied cohort in that patients with positive oligoclonal bands were perhaps less likely to undergo brain biopsy.

A retrospective review of the CSF findings in patients diagnosed with Balos concentric sclerosis found that oligoclonal bands were only present in 31% to 37% and suggests that the CSF profiles of these patients were more similar to neuromyelitis optica or myelin oligodendrocyte glycoprotein-associated encephalomyelitis than to multiple sclerosis (33). A similar retrospective analysis by the same group targeting those diagnosed with Schilder disease found similar findings. Most of these cases were negative for oligoclonal bands (74%). Additionally, Epstein-Barr virus antibodies were absent in 5 cases of Schilder disease, but are present in nearly all patients with multiple sclerosis (34). In addition to a decrease in oligoclonal bands, there is a higher median white cell count in MOG-associated disease and neuromyelitis optica spectrum disorders compared to multiple sclerosis (13).

Brain biopsy. Brain biopsy has an important role in diagnosis of tumefactive demyelination. It is indicated in situations when the diagnostic work-up is inconclusive and the patient is either severely ill or rapidly declining despite therapy and there is urgent need for a definitive diagnosis.

Management

There are no controlled trials of acute management of patients with tumefactive demyelinating lesions as it is a rare condition. High-dose intravenous corticosteroids (methylprednisolone 1 gram intravenous for 5 days) are typically the first-line management for tumefactive relapses. A subgroup of patients with fulminant attacks of CNS idiopathic inflammatory demyelinating disease refractory to steroid therapy may respond to plasma exchange (31; 89; 50). For patients who continue to worsen or have relapses of disabling tumefactive lesions, escalation to rituximab or cyclophosphamide may be appropriate (21; 71; 72). A retrospective review found neurologic improvement with cyclophosphamide treatment in 10 out of 12 patients with tumefactive multiple sclerosis who had been refractory to corticosteroids and plasma exchange treated with cyclophosphamide (23). Aggressive, supportive management in the acute phase, including measures to reduce raised intracranial pressure, is crucial because the anticipated long-term outcome of many of these patients is good. Decompressive surgery may be considered in patients with large demyelinating lesions causing impending herniation (68). For those patients diagnosed to have definite multiple sclerosis, treatment using currently approved multiple sclerosis immunomodulatory agents should be initiated.

Outcomes

It is generally accepted that about two thirds of patients that present with tumefactive demyelinating lesions will follow a relapsing-remitting course and about one third will have no further clinical attacks (01). In those who do have further attacks, the median time for a second clinical attack is close to 5 years (47), which is longer than the 1.9 to 3 years for a typical multiple sclerosis demyelinating event (18). A potential explanation is the more aggressive immunomodulatory therapy provided for many patients with tumefactive demyelinating lesions. In patients with a prior diagnosis of multiple sclerosis, tumefactive demyelinating lesions are associated with a higher risk of clinical relapse and development of new T2 lesions but are not predictive of faster disability progression (83).

Special considerations

Pregnancy

Although tumefactive demyelinating lesions are not thought to be particularly associated with pregnancy, there is one case report of initial presentation of tumefactive multiple sclerosis during pregnancy (60). After poor response to high-dose intravenous steroids and plasma exchange, this patient had an elective termination and had clinical improvement with initiation of multiple sclerosis disease-modifying therapy and rehabilitation.