Wednesday, 27 August 2014

Future outcome measures in MS

How will the definition of NEDA evolve? #MSBlog #MSResearch

"As promised the following is my presentation from a meeting I spoke at in Copenhagen yesterday. I covered NEDA and how I see its definition evolving in the future. This presentation explains why we replaced DAF (disease activity free) with NEDA (no evident disease activity); DAF implies we know everything about MS disease activity, whereas NEDA is a more humble term and allows us to adapt the definition as new technologies come online."

"My talk was very upbeat, particularly for people with progressive MS. The audience were very responsive to my proposed asynchronous model of progressive MS and the therapeutic lag hypothesis. As a result of these hypotheses I have given the odds of the fingolimod PPMS, siponimod SPMS, natalizumab SPMS and ocrelizumab PPMS trials being positive as being greater than 50% for each study; please note that according the the therapeutic lag hypothesis the odds of the studies being positive go up the longer the blinded phase of the trials run."

CoI: multiple; please note this meeting was hosted by Professor Per Soelberg Sorensen and his centre and sponsored by Biogen-Idec.


  1. Question related to therapeutic lag: A non-responder injected beta-interferon for a few years before moving to more effective treatment. The lesion load kept increasing during that time.

    Was all that DMT a waste or could it possibly have some long-term benefit?

  2. It's really great to see use of stats gathered from this blog. As an MSer, I find it empowering!

  3. The chart on slide 5 seems to indicate that SC IFN-1a is significantly - almost 2x - as effective as dimethyl fumarate. Do you believe this is really the case? What activity is DMF not controlling (given it seems to perform equal to or better than the injectables on relapse rate, disability, and MRI lesions)?

    I always find the slides from your public talks to be very informative, so please continue to post them!

    1. Looking through the paper, it appears that DMF compares closely on clinical measures to cladribine and fingolimod. However, it is significantly less effective than these two on radiographic measures. Natalizumab is clearly superior on all dimensions, but the IFN-1a story is a little trickier..

      The paper only shows glatiramer acetate + IFN-1a combination therapy vs treatment with either one, so I can't determine where Prof G's IFN-1a number comes from (I assume he had access to the data). However, adding IFN-1a on top of GA therapy provided 1.57x the patients DAF while adding GA on top of IFN-1a therapy provided 1.74x the patients DAF. If you assume these are entirely independent, you would estimate about 3.05x the patients would be DAF with IFN alone (1.57/(1-1.74)) and 2.12x the patients would be DAF with GA alone. However, this independence seems too strong considering the trial conclusion was that combination treatment is not effective.

      This raises a distribution question on how to define NEDA. You may have a drug, like DMF, that reduces the average lesions by a huge amount but leaves some lingering activity for most people. Another drug, like IFN-1a, may not reduce the average lesion load as much, but it has a larger subset that "super respond" and show zero activity.

      If we can't identify why patients "super respond" quickly (or ahead of treatment), the question is whether we have patients cycle through treatments until they find the magic drug? Or do we go with the one that works for almost everyone but leaves some lingering damage? I guess this requires knowing more about how the early years influence the disease course, so as always, "more research is needed".


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