Monday, 5 December 2016

#ClinicSpeak & #NeuroSpeak: when to rebaseline?

Cherry-picking data; responders need to be analysed differently to delayed-responders or non-responders. #ClinicSpeak #NeuroSpeak

When we discuss NEDA (no evident disease activity) I always make the point of rebaselining. The reason is if you don't rebaseline then you will be taking into account disease activity that occurs before the drug works and blaming the drug for being ineffective when it may not be. With maintenance therapies the rebaselining needs to be done quite early, typically 3-6 months after starting the treatment, with the exception of glatiramer acetate that takes longer to have a full impact on MRI activity. In reality, an for pragmatic, reasons rebaseling at 6 or 12 months is fine.

When it comes to PIRTs, rebaselining needs to be done much later. Why? The mode of action of pulsed immune reconstitution therapies (PIRTs) is due to both depletion and reconstitution. The depletion may happen quickly, for example in the case of alemtuzumab and HSCT, or be more delayed in the case of cladribine. However, most consider the mode of action to be what happens post-reconstitution, i.e. when the immune system resets itself with new regulatory mechanisms. The reconstitution takes many months, if not years, to occur. Therefore it only makes sense to rebaseline at a point in time when you can do something about ongoing disease activity, i.e. in the case of alemtuzumab and cladribine offer another course of treatment. This is why we recommend 24 months in our Barts-MS algorithm.

Not all people agree with using the 24 months for rebaseling with alemtuzumab. Some have suggested using 12 months, i.e. before the second course has had time to work. The advantage of using the second year is that it at least allows you to compare alemtuzumab with other maintenance therapies. In the NEDA study below we simply looked at what happened to the 50% (175/349) of subjects who were rendered NEDA in year 2 (prior to alemtuzumab having had a chance to have its full effect). The other 50% of subjects had activity and hence should be looked at in year-3 and beyond; this analysis is being done. What this study tells you that if you are in the 50% of subjects that responds quickly to alemtuzumab by being NEDA in year 2 (prior to full activity) then you have 60% chance of being NEDA over the whole time period to year 5. That is all this poster says; it is not trying to make any other claim but that. Clearly, what happens to the other 50% is important and this information will be presented in a future publication. But more importantly is what happens to all these subjects over the next 10-15 years. This information will tell us if the promise of a 'potential cure' is real or not. 

People need to realise that PIRTs are not necessarily miracle drugs; all they offer is a different apporach to treating MS. What is clear is that the way we use them and monitor their activity is very different. 

CoI: multiple


  1. Replies
    1. Re: "Is rituxan considere a PIRT?"

      The million, or billion, dollar question. I suspect that when ocrelizumab is launched the community will run trials comparing ocrelizumab as a maintenance therapy compared to it as a PIRT. Based on how I think the drug is working there will be substantial proportion of people who will go into long-term remission with only 2-years of treatment.

  2. Thanks for posting this. It clears up some uncertainty I had about why/how the study re-baselined each year. I'd still like to see longer term data -- especially for those in the "lower" 50% -- but this goes a long way toward at least setting a frame of reference for understanding the data.

    1. I agree. I think in the future we are going to have to present NEDA data in a similar way oncologists present remission-free survival, i.e. using a survival curve. We set the survival curve ticking after the DMT has had time to work and then see what happens.

    2. Can I ask why (without picking on anyone) it has taken so long for these concepts (which are alive in other areas of medicine) to reach MS?

      We are now a decade into DMTs, since the first ones became available?

    3. RE: "Can I ask why (without picking on anyone) it has taken so long for these concepts (which are alive in other areas of medicine) to reach MS?"

      They have been around in MS for quite awhile; say 10 years or more. The person to push most was Rick Rudick at the Cleveland Clinic. I covered all the issues in an 2012/3 Editorial and we have been practising T2T for several years. There many places that practice it as well, but are less vocal than us as they don't have a blog.

      But you are right; why hasn't the MS community adopted these principles quicker? May be it is because of slow adoption or because MS is more difficult to monitor.

      Please note the issues raised by T2T-NEDA are not trivial.

    4. I am grateful for the answer and read your 2013 article with interest. But the 2013 study- Characteristics influencing therapy switch behavior after suboptimal response to first-line treatment in patients with multiple sclerosis - found that of 606 suboptimal responders, 214 (35.3%) switched therapy.

      That hardly paints a picture of treating to target?

      It really saddens me, I still hear of neurologists who use phrases like "just a small lesion" or "it's ok, i've seen worse, you've only got 3 new lesions since last year'

  3. For those you who are having doubts about alemtuzumab, please focus on the end-organ damage marker, i.e. the annualised rate of brain atrophy. In this cohort of pwMS it is well within the normal range. The only other data that comes close to this is the myelo-ablative HSCT data from Canada. I say this, but the HSCT patients take a massive hit in year one with brain atrophy rates way above what you expect for an anti-inflammatory. Presumably this is due to the neurotoxic effects of the chemotherapy they receive.

  4. Thank you for this post, it is a lot clearer on your rebaslining logic. However it does throw up another question for me. You always state the need for early aggressive treatment 'time is brain' which I personally agree with. However in your PIRT therapy you have no definitive way to know if new MRI/relapse activity is due to the patient being a non responder, or a patient just needs to give the drug 'more time.' This is why rebaslining is flawed, as you have no definitive marker. I'm sure if after 2 alemtuzumab courses if a patient relapses you retreat, and unless you definitively know if they are non responded or need to give the drugs more time to work you can't classify them differently.

    As soon as a patient relapses, their NEDA should be reset to zero and they would be considered NEDA from last treatment in my opinion, same way chemo drugs are tracked for cancer. If you have cancer then chemo and your cancer then reappears after 3 years, they retreat and your cancer goes again from year 4 onwards. They don't say you were in remission/cured for years 1,2,3,5,6 and ignore year 4.

    1. Re: "They don't say you were in remission/cured for years 1,2,3,5,6 and ignore year 4."

      The argument above varies depending on whether, or not, you are on a maintenance vs. a PIRT. If you are on a PIRT then breakthrough, provided you are taking the drug, implies a sub-optimal response. The latter may be okay if your are on top of the treatment pyramid with nothing else to escalate or switch to. With PIRTs we know that sometime it takes more than the one (HSCT), or two (Alemtuzumab and cladribine) to put you into remission. You may need additional courses. With HSCT this may be tricky, but with the others we have data to show that some people need more courses. How many courses would you give before you call someone a non-responder? With alemtuzumab you would check to make sure they are depleting and if not try and test the patient for NABs. The latter is a cause of failure in some patients. At present we are thinking about calling it a day after 4 courses; for two reasons - (1) funding and (2) all the cases that I have been involved in with 5 courses have not done that well. With cladribine we only have data on four courses; going beyond that we would have to consider safety and whether or not the immune system can take endless cycles of depletion followed by reconstitution. At some stage the stem cells start to be depleted and we will run into premature immunosenescence problems.

    2. Re NM HSCT: Dr Burt recommends top up chemo for those who relapse/show activity within a certain time from having HSCT. Does it matter if the top is chemo or a monoclonal antibody? I don't know but we'll find out in future. I suspect top up monoclonal could have a similar effect to top up chemo... but clearly I'm no doc.

      You could conceivably choose to have HSCT again in future (how many times is dictated by your bone marrow, not decisions regarding efficacy).

      I think re-treatment decisions post HSCT are easier than post Lemtrada: with HSCT, after bloods recover (12-24 months later), world is one's oyster in terms of future treatment.

      Now I'm just far too optimistic.

  5. "HSCT patients take a massive hit in year one with brain atrophy rates way above what you expect for an anti-inflammatory. Presumably this is due to the neurotoxic effects of the chemotherapy they receive."

    Looks like you and Joan agree on one thing..she links to one of your oldies.

  6. I have been hearing snippets here and there which suggest that there is some thought/theory/discussion/decision regarding a link between efficacy of Lemtrada and lymphocyte counts either recovering too quickly or not dropping low enough after first dosing.

    I don't know where what I am hearing is coming from. I am away that some US neuros seem to be basing rituximab redosing decisions on lymphocyte recovery after the last rituximab dose. What is this based on? Where are these ideas coming from? Is there data coming out that will explain it?

    Am I missing something?

    Thank you in advance.


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