Marburg acute multiple sclerosis: the rarest and most aggressive form of the disease
Marburg MS moves in weeks rather than years, and it is so rare that it still has no official diagnostic criteria. Here is what it looks like, why it is mistaken for a brain tumour, what systematic reviews of every published case now show, and why the outcome is no longer what it was twenty years ago.
When people first read the words "Marburg multiple sclerosis" they usually get frightened. I understand why: every article about it carries the words fulminant, malignant and fatal. So let me say two things straight away. First, this is an extraordinarily rare form — so rare that every case ever published in the world literature fitted into a review of twenty-three patients [2](#ref-2). Second, the outcomes printed in older textbooks no longer describe what happens in hospitals today.
Marburg acute multiple sclerosis, also called acute fulminant MS or Marburg disease, is named after the Viennese neurologist Otto Marburg, who described it in 1906. It sits among the so-called borderline forms of MS, alongside neuromyelitis optica, Baló concentric sclerosis and Schilder's disease — a group of conditions that some authors treat as MS variants and others as separate diseases [1](#ref-1).
> The difference is not which symptoms appear. It is how fast they appear. In ordinary MS we talk in years. In Marburg MS we talk in weeks.
How it differs from "ordinary" multiple sclerosis
Typical relapsing-remitting MS comes in waves: relapse, recovery, quiet period. Marburg MS usually has a single monophasic onslaught that does not stop. Within weeks to a few months the person reaches severe disability, and in the worst cases a state requiring intensive care [2](#ref-2), [5](#ref-5).
In a 2025 systematic review pooling twenty-three patients, the median age was thirty-two, ranging from fourteen to sixty-three, and 78 per cent were women [2](#ref-2). A second review published the same year reached similar numbers: a mean age of 34.9 years and 88.9 per cent women [3](#ref-3). So: young people, usually women, with no earlier disease history.
Another difference is that many of these patients never meet the full criteria for classic MS. In a meta-analysis of thirty-five cases, 81 per cent had lesions disseminated in space on imaging, yet only 41 per cent met the complete McDonald criteria, and oligoclonal bands were found in just a third [4](#ref-4). The disease simply never gets the chance to "disseminate in time" the way the classic definition requires.
What it looks like at onset
In the meta-analysis of published cases, the commonest presenting features were weakness of one side of the body (77 per cent), cranial nerve deficits (57 per cent), acute encephalopathy and clouding of consciousness (51 per cent), sensory changes (49 per cent) and unsteady walking, or ataxia (37 per cent) [4](#ref-4).
What raises suspicion is not any single symptom but the tempo. When a young person accumulates new neurological deficits over a few weeks and none of them recede, that picture fits no ordinary relapse. I wrote separately about what a normal relapse looks like and how to recognise it in the article on [recognising an MS relapse](/en/blog/ms-relapse-recognize).
Why it is mistaken for a brain tumour
When Marburg disease presents as a single large lesion, it can be radiologically indistinguishable from a brain tumour or an abscess [1](#ref-1), [8](#ref-8). That is why some patients are first referred to neurosurgery rather than neurology, and why a biopsy is sometimes performed.
On MRI the lesions are hyperintense on T2 and FLAIR sequences in every described patient; in one review 55 per cent were periventricular, 22 per cent subcortical, and ring enhancement after contrast was seen in 28 per cent [3](#ref-3). In the most severe cases the number of contrast-enhancing lesions exceeds a hundred [5](#ref-5). What is generally read on a scan I broke down in the article on [MRI in multiple sclerosis](/en/blog/what-doctors-look-mri-scans-multiple-sclerosis-detailed-guide).
A lumbar puncture helps but does not decide. Of the fourteen patients who had one, 71 per cent showed lymphocytic pleocytosis, 64 per cent raised protein, and only half had positive oligoclonal bands [3](#ref-3). The crucial point: all were negative for anti-MOG and anti-AQP4 antibodies [3](#ref-3). Those two tests are now mandatory, because they separate Marburg MS from conditions that look similar but are treated differently — anti-MOG associated encephalomyelitis and neuromyelitis optica [11](#ref-11).
There are still no official criteria — but they are being proposed
This deserves saying plainly: there are no officially accepted diagnostic criteria and no standard of care for Marburg MS, precisely because it is so rare [4](#ref-4). The diagnosis is made by excluding other causes, alongside the characteristic clinical course and imaging.
That has begun to change. Two groups have independently proposed criteria: one based on a meta-analysis of thirty-five cases [4](#ref-4), the other through a 2025 systematic review proposing Definite, Probable and Possible categories, based on a fulminant course, characteristic MRI patterns, exclusion of mimics in CSF and serum and, where available, histopathology [2](#ref-2).
What is used in treatment
Nearly everyone receives the same first steps: high-dose corticosteroids (96 per cent in the 2025 review) and, very often, plasma exchange, given to 61 per cent [2](#ref-2). When that fails to halt the disease, cytotoxic drugs follow.
Cyclophosphamide is the treatment the literature agrees on most. In the meta-analysis of thirty-five cases, only high-dose induction cyclophosphamide was associated with significant EDSS improvement, while steroids alone, plasma exchange alone, low-dose cyclophosphamide and mitoxantrone were not [4](#ref-4). The most striking published example is a twenty-six-year-old woman with more than a hundred gadolinium-enhancing lesions who recovered almost completely after four days of high-dose cyclophosphamide and was then stabilised on ocrelizumab [5](#ref-5), [13](#ref-13).
After the acute phase, high-efficacy maintenance therapy is used — ocrelizumab, rituximab or alemtuzumab [3](#ref-3), [9](#ref-9). A favourable long-term outcome after autologous stem cell transplantation has also been reported in individual cases [10](#ref-10); I covered that procedure and its risks in the [guide to stem cell transplantation](/en/blog/stem-cell-transplantation-hsct-ms).
One caveat: mitoxantrone has been used historically and has successful case reports [9](#ref-9), but it showed no advantage in pooled analyses and carries serious cardiac and blood-count risks. None of these drugs is formally licensed for Marburg MS — they are used on the basis of case experience.
What the outcome looks like today
This is where the change is most visible. Historically, acute MS was a fatal disease, with death usually within a year of onset, most often from extensive brainstem demyelination [1](#ref-1), [12](#ref-12). In the meta-analysis of older published cases, mortality was 37 per cent [4](#ref-4).
The 2025 systematic review paints a different picture: of twenty-three patients, six died (26 per cent), four were left with a stable deficit (17 per cent), nine improved (39 per cent) and four recovered almost completely (17 per cent) [2](#ref-2). So more than half of patients today improve or recover nearly fully.
It is only fair to say why these numbers should be read carefully. They all come from case reports, not clinical trials. Cases with good outcomes are published more readily than bad ones, so the true picture is probably somewhat bleaker. But the direction is not in doubt: fast diagnosis and aggressive treatment change the outcome.
What this means for someone living with MS
If you already have a diagnosis of multiple sclerosis, the chance that you have the Marburg form is negligible. It does not develop out of relapsing-remitting MS after years of disease — it is different from the start, lightning-fast and monophasic [2](#ref-2). A worsening that lasts days and then recedes is not Marburg disease; it is a relapse or a pseudo-relapse.
What does warrant urgent medical attention, without waiting for a scheduled appointment: sudden weakness on one side of the body, double vision or visual loss worsening day by day, confusion or reduced consciousness, difficulty swallowing and speaking, or a run of new neurological deficits within a few weeks. Those situations mean emergency care, not an outpatient booking.
And if you are reading this because someone close to you has just been given this diagnosis — what the literature shows most clearly is that time is everything, and that this belongs in a centre experienced in neuroimmunology with access to plasma exchange and intensive care. The first practical steps are collected in the article on [recognising MS and what to do first](/en/blog/how-recognize-ms-diagnosis-first-steps).
This article is informational and does not replace examination and advice from a doctor. Marburg multiple sclerosis is an emergency and requires hospital assessment and treatment.