ECTRIMS2018. Relapses with B cells present

Disease activity with ocrelizumab in the absence of B cells
Should we start panicing about the B cell hypothesis?

P1216 – Unexpected highly inflammatory ms activity uncharacteristic for prior disease course in patients treated with ocrelizumab.G. von Geldern, L. Jayagopal, M. Persenaire, A. Wundes Neurology, University of Washington, Seattle, WA, United States

Introduction: Ocrelizumab (OCR), a B-cell depleting agent, demonstrated significant reduction in clinical and MRI disease activity in relapsing remitting MS (RRMS) and a moderate effect on disability progression in primary progressive MS (PPMS) in clinical trials. Since its approval in spring 2017, real world experience contributes to our knowledge of efficacy and safety of this new medication.
Objective: To report unexpected highly inflammatory disease activity uncharacteristic for prior clinical course in 3 MS patients treated with OCR.
Results: 126 MS patients received OCR at our site since its approval. Among those, 3 patients experienced highly inflammatory clinical and MRI disease activity uncharacteristic for their disease. All had non-detectable B cells at that time
(1) A 50 y/o man with SPMS with no clinical relapses in many years experienced his life-time worst relapse with more than 20 enhancing lesions on brain and spine MRI 5 months after starting OCR (previously rituximab x1 year). Non-responsive to two 5-day courses of iv methylprednisolone, he was admitted for expedited infectious work up (non-yielding) and improved clinically after plasmapheresis while short-term repeat MRI remained highly inflammatory. Anti-OCR antibodies were negative. 
(2) A 48 y/o woman with PPMS experienced her first ever relapse 3 months after OCR initiation. She developed right sided weakness and ataxia; a large new enhancing cerebellar peduncle lesion was found on MRI. After iv methylprednisolone, her symptoms improved. A repeat MRI showed ongoing enhancement 3 months later. 
(3) A 55 y/o man with PPMS with no prior enhancing lesions on MRI underwent routine surveillance MRI 6 months after starting OCR. A new large enhancing lesion in the right cerebellar hemisphere was found. No new symptoms or exam findings were seen but enhancement continued for several months despite treatment with iv methylprednisolone.
Conclusions: While OCR is a potent treatment for RRMS and has demonstrated benefit in PPMS, we here report unusual and unexpected inflammatory disease activity in 3 patients, who had a progressive disease course prior to OCR. Due t o the degree of immunosuppression associated with OCR and the amount of inflammation being uncharacteristic for these patients, infection or other complication of treatment mimicking acute MS disease activity had to be considered.Clinical experience with this new medication may help optimize patient selection and safety surveillance.

P933 – Can ocrelizumab prevent MS reactivation after discontinuation of natalizumab? P. Repovic, K. Smoot 

Introduction: Natalizumab (NTZ) and ocrelizumab (OCR) are both highly effective therapies for relapsing multiple sclerosis (MS). Current evidence suggests that there is a small but appreciable risk of increased inflammatory activity (“rebound”) after NTZ discontinuation in some patients. It is unknown whether OCR is capable to preventing this phenomenon.
Objective: To describe 2 patients who experienced rebound MS activity after transitioning from NTZ to OCR.
Summary: Two female MS patients, aged 56 and 63, had been treated with NTZ for 105 and 29 months, respectively. While on NTZ, neither patient had MS relapses or additional MRI lesions; however, patient 1 had a history of severe relapses after self-discontinuing NTZ in 2009 and 2014. Due to increasing PML risk, patients transitioned from NTZ to OCR. The interval between last NTZ infusion and first OCR infusion was 7 weeks for patient 1, and 4 weeks for patient 2. Their MRI scans at the time of last NTZ infusion showed no new lesions. 
Patient 1 experienced a relapse 11 weeks after starting OCR (18 weeks after last NTZ infusion); her MRI showed 2 contrast-enhancing lesions, including a PML-like lesion in left temporal lobe. PML was ruled out, and she was treated with steroids and plasma exchange, with good functional recovery. 
Patient 2 reported ambulatory difficulty with MRI evidence of 3 contrast-enhancing juxtacortical lesions 12 weeks after starting OCR treatment (16 weeks after last NTZ infusion). 
Serum CD19 count was undetectable at the time of relapse for patient 1.
Conclusion: During transition from OCR to NTZ, even with a short (< 8 weeks) washout period and effective CD19 suppression, there remains a risk of MS reactivation that, in some cases, may mimic PML. We postulate that the factor(s) driving MS rebound are intrinsic to the central nervous system.

These papers say something about MS, we just have to figure out what they are telling us! 

Do you have any ideas?

These studies show us what we have known from rheumatoid arthritis treated with rituximab, in that you can have flares of disease in the apparent absence of peripheral B cells. 

ProfG says that there were a few cases of early disease activity in the OPERA trials and we have heard from other neurologists that they have seen flares shortly after treating with ocrelizumab and rituximab. How common was this in the ORATORIO study.

In some cases it will be because they did not deplete, but here they looked and they could not see B cells in the blood.

Does it say that B cells are not important?

Maybe. Is this the evidence that the T celll biologists will use to say that B cells are not important. However, on balance there is more evidence to support B cell involvement than against it. 

The blood only harbours about 2% of the lymphocytes, the bone marrow about 10-15%, the spleen about 10-15% and the lymph glands about 55%. So the main action of the treatments will have to be in the lymphoid tissues, if the drugs are to work.

We know that drugs like rituximab and alemtuzumab do not purge these lymphoid tissues, including the bone maarow, completely. Therefore they still harbour lymphocytes that will repopulate and these may be the site of where new lesions formed from. That they cannot detect B cells in the blood does not mean that they have not come from the blood and into the CNS, because you would only need very few cells to trigger an immune response and very few cells will be deemed as the cells being adsent. 

In this case they did not report on CD27+ memory B cells, but other people seeing this phenomena have, apparently. So it does say that monitoring the blood for recurrence of memory B cell may not work for everyone. 

Will it work for anyone? We simply have to do the study…..properly. Why haven’t we done it already?…Tells you something about us and the ethical review process:-(. 

Genzyme wouldn’t give us the data. Maybe they will publish what they have. However, seems they are too busy trying to disprove that autoimmunity after alemtuzumab has anything to do with absence of effective T cell regulation and B cell repoplation.

These studies may suggest that dosing every 6 months is not enough as maybe the pathogenic cells can recover quickly. Natalizumab given every month kept them in check. In NMO there is a need to redose and it is not always successful in stoping relapses. In this study these relapses occurred 3-5months after the infusion of ocrelizumab. So it tells us that ocrelizumab is not invincible and people can develop a relapse, whilst on ocrelizumab. But we know this based on the trial results.

This is the same for many other agents including HSCT, because they are not 100% effective. It could tell us that it takes time for ocrelizumab to work. People seem to have really bought this story with copaxone and indeed it is probable that lesions that were occuring at the start of treatment, develop into a relapse. This is the reason why the participants were rebasedline three-six months after the start of treatment in the phase III trial to exclude this type of relapse from the analysis to better shown the influence of ocrelizumab.

It also suggests that in some people with progressive MS it can reboot to form relapsing MS, as it seems to have precipitated a relapse, possibly meaning that for some people dosing every 6 months in too long a time interval. Perhaps the low grade inflammation that does occur in progressive MS was halted and kepted in check and when the effects of ocrelizumab wore off the inflammation came back with a vengence. We have seen this will people with progressive MS taking fingolimod. In people with relapsing MS it goes unnoticed as it will be seen as a relapse.

Maybe it tells us that the driving force for lesions eminates from within the CNS as it is likely that ocrelizumab does not clear the brain of immune cells. So for the EAEers out there we are back to the “inside-out” theory. This would not surprise me that activity in CNS can trigger activity. The myth created by the MRI imagers is that gadolinium-enhancing lesions show infiltrating cells. It doesn’t. You can look at animals to see this. There you see gadolinium lleakage before the main wave of infiltrating cells arrive and the area enhancing is far bigger than the lesion and it is simiply showing fluid movement from the blood into the brain. However, to address this we need to get B cell depleting agents into the brain. Fingolimod does do this but its action is to stop migration so the CNS penetration is not relevant in this context. Protein molecules are too big to effectively do this. Cladribine is the only current drug to penetrate the CNS to any significant level. 

Maybe it tells us that B cells are a a variety of different cell types that exist in a balance and whilst in most people the pathogenic cells really outway the influence of regulation, but in some people regulatory B cells are very important and removing them, which ocrelizumab does do may be one of the unwanted consequences of this type of  all-encompassing therapy. This suggests that we need to be more surgical in which type of B cells are removed. This what acaticept would have done, but they got it wrong and removed the wrong sort of B cells leading to disease activation, Is this what happened in these people?

As you can tell..I don’t know the answer…yet

However, it is understanding these unusual cases that give us clues of what really is going on.

About the author



  • Thankyou for your post – very informative although rather scary to read as an MS patient myself taking one of the more potent DMTs. One thing that is noticeable is the older age of the participants who developed increased desease activity. Do you reckon this could be the key contributory factor? Perhaps it really is wise to stop all treatments for folk over a certain age. Are the risks simply too great with these supposedly highly effective therapies and the benefits not outweighing the potential risks?

  • When u blanket bomb b cells you take out b reg cells as well. Besides b cells to travel to cms. It needs to travel via blood supply and if the count is 0 then where r they coming from?. This also means ocrelizumab is not a ind8ction therapy. And if your theory of inside out is correct. Caldribine should be the most effective treatment and it's not! Have you considered once b cells have activated t cells. T cells no longer need b cells to drive the inflammation?

  • what a post!
    Thank you MD. This is the reason why I have avidly been reading this blog for 7 years now.
    Keep them coming.

  • I think those cases are very interesting. Solution to the conundrum can be diverse. First, MS is not a single disease. It does not matter how many times we try to convince ourselves. Second a good explanation can be found from other diseases that respond well to B cell depletion. B cells are gone quickly. Antibodies are not. No matter how many times we say that the B cell effect depends on antibody independent mechanisms, nature created B cells to sense antigens with their B cell receptor and switch to antibody secreting cells right after. Finally, as pointed out, availability of OCR in the tissues is likely to be highly variable across individuals depending, for example, on the FcG receptor polymorphisms… But, who knows. Maybe those are pure T cell MS patients…

  • Hi,

    Simplest explanation I think because all the relapses are some months after initiation of B cell depletion treatment:
    Autoantigens are in B cells (including EBV?) and are presented by B cells to T cells inducing relapses. When B cells die by ocrelizumab, B cells are phagocytosed by macrophages or B cell proteins enter the serum-> these B cell proteins are taken up by macrophages which now (instead of B cells) present them to T cells to enhance autoreactive T cells and relapses within a few months after initiation of treatment.
    Question is: are there any data about what happens immunologically in the immune system during the first weeks when B cells die?

  • Dr Mouse calm down 🙂

    You are "all over the place" 🙂

    3/126=0.023=2,3% failure (A very good rate)

    You want a drug that works 100%?

    Show me one?

    "These papers say something about MS"

    The 3 patients had progressive ms all the other 123 patients where rrms?

    Maybe is telling something about progressive ms

    I would lay more on the last bit of your comments (regulatory B 10


    Since you know that if you inject mice with rituxan before inducing

    EAE disease gets far worst, than when rituxan is given after disease


    EAE is a mouse T cell–mediated autoimmune disease of the CNS used to model the human condition MS. The
    contributions of B cells to EAE initiation and progression are unclear. In this study, we have shown that EAE
    disease initiation and progression are differentially influenced by the depletion of B cells from mice with otherwise
    intact immune systems. CD20 antibody–mediated B cell depletion before EAE induction substantially
    exacerbated disease symptoms and increased encephalitogenic T cell influx into the CNS. Increased symptom
    severity resulted from the depletion of a rare IL-10–producing CD1dhiCD5+ regulatory B cell subset (B10 cells),
    since the adoptive transfer of splenic B10 cells before EAE induction normalized EAE in B cell–depleted mice.
    While transfer of regulatory B10 cells was maximally effective during early EAE initiation, they had no obvious
    role during disease progression. Rather, B cell depletion during EAE disease progression dramatically
    suppressed symptoms. Specifically, B cells were required for the generation of CD4+ T cells specific for CNS
    autoantigen and the entry of encephalitogenic T cells into the CNS during disease progression. These results
    demonstrate reciprocal regulatory roles for B cells during EAE immunopathogenesis. The therapeutic effect
    of B cell depletion for the treatment of autoimmunity may therefore depend on the relative contributions and
    the timing of these opposing B cell activities during the course of disease initiation and pathogenesis.

    Ps: Also think that you should think about T cells for time to time 😉


    • DrK presented or should I say Stephania for the lab at ECTRIMS presented our real world data currently on 200 people with generic cladribine. The real world data of the first 50 is in the pipeline. The dose adjustment works to limit the occurrence of significant lymphopenia.

  • "..Disease activity with ocrelizumab in the absence of B cells
    Should we start panicking about the B cell hypothesis..?"

    Yes. Totally. B cell depletion is useless can't even stop relapses.

  • For those on this blog that are positing this as proof that B cell depletion is ineffective as a treatment for MS, why is it that there are plenty of case reports of early alemtuzumab failures (i.e. early breakthrough disease after initiation of this supposed 'immune system reboot therapy) despite profound/persistent drops in T CD4+ T cells?

    And guess what those patients respond to? B cell depleting therapy.

  • If the lesion was expanding due to brain activity and this crossed into a tract that was clinically imortant, a relapse (new symptoms) would occur and you may not need any new lesions or gadolinium enhancement.

By MouseDoctor



Recent Posts

Recent Comments