Monday, 6 April 2015

BREMSO SCORE: predicting the course of MS

What is your risk of developing SPMS? #MSBlog #MSResearch

"The study below uses early clinical predictors to classify MSers into different risk categories for developing SPMS. The investigators use a special type of statistics called Bayesian statistics to generate their score, which they refer to as the BREMSO score. The BREMSO is calculated using only onset variables according to the following algorithm: 
  1. 0:05 x age (in decades) 
  2. - 1:07 (if female gender)
  3. + 0:93 (if sphincter onset) 
  4. + 0:62 (if pure motor onset) 
  5. + 0:81 (if motor and sensory onset)
  6. + 0:32 x number of neurological functional systems involved at onset 
  7. +0:52 (if incomplete recovery after onset)
As the BREMSO score does not incorporate information about relapses within the first year it allows easier and earlier score calculation, i.e. at the onset of the disease. What they find that MSers with a high BREMSO score (above the median or in the 3rd and 4th quartiles) do a lot worse than those with a low BREMSO score (in the first quartile or less that the 25th-percentile). This is an odd comparison as it excludes subjects who fall in the second quartile (between the 25th and 50th-percentile). It is a pity that the paper does not actually give you the data to calculate the cut-offs, which makes it impossible to know if your own BREMSO score falls into a good (1st quartile) or poor (3rd & 4th quartiles) prognostic groups, or for that matter 'no man's land (2nd quartile)'.

Despite this I am still impressed by the weightings of scores as they support my length-dependent axonopathy hypothesis. Sphincter involvement at onset is the worst prognostic variable; the fibres to the sphincters are the longest followed by the motor fibres to lower limbs. If both motor and sensory systems are involved is the next most significant predictor; this would indicate that the lesion causing the first attack is large or multifocal, i.e. involving at least 2 sites. 

Is this study useful for you as an individual with MS? Possibly; for example if you do have a high BREMSO score you may decide to go onto highly-effective DMTs first-line (flipping the pyramid) in comparison if you have a low score you may select less risky low to intermediate efficacy DMTs (maintenance-escalation). This prescribing pattern is kind of what we do already. 

What is missing from the BREMSO score are the baseline MRI metrics (T2 lesion load, number of Gd-enhancing T1 lesions, number and volume of T1 hypodensities and baseline brain volume), the presence or absence of oligoclonal bands and the gap between the 1st and 2nd relapse; all of these factors are linked to clinical course and prognosis. Despite this the BREMSO score allows us to profile MSers at baseline to make treatment decisions more rational and scientific."
"This paper illustrates how useful real-life data can be in generating new hypotheses and outcomes. Well done to MSBase for putting together this study, I can see us building on this model going forward."

An example of a risk score predicting the development of SPMS

Epub: Bergamaschi et al. BREMSO: a simple score to predict early the natural course of multiple sclerosis. Eur J Neurol. 2015 Mar 25. doi: 10.1111/ene.12696.

BACKGROUND AND PURPOSE: Early prediction of long-term disease evolution is a major challenge in the management of multiple sclerosis (MS). Our aim was to predict the natural course of MS using the Bayesian Risk Estimate for MS at Onset (BREMSO), which gives an individual risk score calculated from demographic and clinical variables collected at disease onset.

METHODS: An observational study was carried out collecting data from MS patients included in MSBase, an international registry. Disease impact was studied using the Multiple Sclerosis Severity Score (MSSS) and time to secondary progression (SP). To evaluate the natural history of the disease, patients were analysed only if they did not receive immune therapies or only up to the time of starting these therapies.

RESULTS: Data from 14 211 patients were analysed. The median BREMSO score was significantly higher in the subgroups of patients whose disease had a major clinical impact (MSSS≥ third quartile vs. ≤ first quartile, P < 0.00001) and who reached SP (P < 0.00001). The BREMSO showed good specificity (79%) as a tool for predicting the clinical impact of MS.

CONCLUSIONS: BREMSO is a simple tool which can be used in the early stages of MS to predict its evolution, supporting therapeutic decisions in an observational setting.


  1. Question:

    what is the definition of "onset"?
    The first relapse or a timeframe e.g. 3 months.

    Because my "onset" was 4 relapses within 5 months and every time a different system was affected and since the 3rd relapse I had incomplete recovery.

    So if I calculate my first relapse I get a low score but if I calculate all relapses within 4 months I get a high score (YAY highscore!).

  2. "As the assessment of a cut-off value for BREMS was impossible
    we divided patients into quartiles, and compared
    patients with the highest BREMS (fourth quartile, BREMS
    value ≥ 0.97) with patients with the lowest BREMS (first
    quartile, BREMS value ≤ -0.56). Among patients with
    higher BREMS, 17% reached SP within 10 years of disease
    onset, while only 4.4% of patients with lower BREMS
    reached SP."

    Well, found in an online resource, they at least defined what is the highest quartile and the lowest quartile - but not the rest. ( Found at : on Page 4)

  3. Can you define what you would consider to be motor onset? Would leg tightness be considered that?

  4. what is the scoring system as I can only get it up to a 2-3 not anywhere above currently

  5. On the subject of the guys at MsBase
    Have you seen the Curve and data associated with it?
    Is the average time to edss 3.5 from onset really 20years? That seems a long time?! It would be great if correct


Please note that all comments are moderated and any personal or marketing-related submissions will not be shown.