Unusual cases may hold key to understanding

Case reports can off insight into MS, 

B cells at the forefront but does this end of year case report burst the B memory bubble?

Rinaldi F, Federle L, Puthenparampil M, Perini P, Grassivaro F, Gallo P.Evidence of B-cell dysregulation in severe CNS inflammation after alemtuzumab therapy.

This is another case report about disease activation within a few months of alemtuzumab. This person was doing very well on natalizumab but stopped to have a baby. 

Post-partum the person had a relapse and decided to switch to alemtuzumab. 

They had a substantial relapse 4 months after dosing and there were only 0.8 lymphocytes x 10*9 /L in the blood.  

There were 0.18 x 10*9 CD19+ B/L and 0.14 x 10*9 CD3 T cells in the blood. 

In the CSF, CD19+ were 12% of the lymphocytes of which 40% where CD20- so they were most likely plasmablasts (plasma cells would be CD20-, and largely CD19-), there were extra oligioclonal bands in the CSF. 

The inference was that the problem was B cell dysfunction. 

Does this tell us the problem is in the plasmablasts and memory are not the problem? 


It is cases like this that can teach us stuff. It is open access so you can all read and we need to try explain them whether it is a problem of T or B or both.

Antibody in the blood was not the problem as removal of antibody made no effect. 

However, this would not remove antibodies being produced in the CSF, so maybe those in the brain were the problem.

There was no EBV detected.

Is this an issue. Only if you think they are the cause of the problem?

Maybe the relapse was destined before the alemtuzumab was started. 

However, the questions are what were the other 0.4 x 10*9 lymphocytes that were not CD19+ T cells and CD3+ T cells?

Also in the CSF what were the 60% CD19 B cells that were CD20+. 

Were they plasmablasts but these would be CD38+ as would mature and immature B cells so there is about 60%% CD38-ve, CD20+, CD19- . Probably would be memory B cells. But we dont know. 

Remember they can become plasma blasts.

So the idea is not quite dead yet.

This person was destined for HSCT. and probably myoablative (replace the immunsystem) at that.

If you don’t do myoablative HSCT it may not be that good.

Nothing quite like HSCT to get the comments flooding in. 

However, I will be tucking into my Christmas Pud rather than answering the comments I’m afraid.

The study below, suggests that relapses after 7 months can occur with non-ablative HSCT. So this appears no better than using DMT.

This also shows that HSCT is not immune to the risk of PML. 

The high intensity treatment failed after ten years.

There is no question that HSCT can be very effective.

Frau J, Carai M, Coghe G, Fenu G, Lorefice L, La Nasa G, Mamusa E, Vacca A, Marrosu MG, Cocco E.Long-term follow-up more than 10 years after HSCT: a monocentric experience. J Neurol. 2017. doi: 10.1007/s00415-017-8718-2. [Epub ahead of print]

BACKGROUND:Autologous hematopoietic stem cell transplantation (aHSCT) is used in aggressive relapsing and progressive multiple sclerosis (MS). The multicentre studies and case series reported have relatively short follow-up.
AIM:To evaluate long-term effect and safety of HSCT in MS.
MATERIALS AND METHODS: Patients referred to the MS centre of Cagliari and undergoing HSCT were included. Variations in relapses and EDSS before and after HSCT were evaluated by Wilcoxon test. A descriptive analysis was made for other clinical data.

RESULTS:Nine patients (female 6, males 3; 5 relapsing-remitting, 2 secondary progressive, 1 primary progressive, and 1 progressive relapsing) performed HSCT (1999-2006). The median follow-up was 11 years (11-18). Eight patients underwent aHSCT, seven using a low intensity conditioning regimen, and one an intermediate intensity. The primary progressive underwent allogeneic HSCT, due to onco haematological disease. The relapses number decreased in the 2 years following the procedure compared to the two preceding years (p = 0.041). New relapses or disease progressions were observed after a range of 7 (low intensity regimen)-118 (intermediate intensity) months. At last follow-up, the EDSS was stable in two patients, improved in two, and worse in five (maximum 2 EDSS in one patient). Six patients showed new lesions, and seven gadolinium-enhancing on brain MRI after a mean of 23.3 and 19.8 months, respectively. Two serious adverse events were reported: melanoma, and progressive multifocal leukoencephalopathy.
CONCLUSIONS AND DISCUSSION: Our results confirm in a long follow-up the efficacy of HSCT in reducing relapses and disability progression. The risk/benefit profile is better for intermediate intensity regimens.

About the author



  • This doesn't offer much to the discussion on plasma cells, but I have to say that the data of the abstract are not very clear. It is already known that the low intensity lympho-ablative regimen is not effective:

    Very aggressive RRMS patients are even suggested to undergo a second round of HSCT to have the desireable results, so it is not the magic potion to work without regimen.

    • This gets tricky with terminology. The study you've linked to, referencing "low intensity/lympho-ablative HSCT) uses 120mb/kg Cyclophosphamide, which is roughly half of the dose typically applied in what is commonly referred to (now) as low-intensity HSCT – and crucially also lacks the high dose secondary agent used in modern protocols (usually a Mab) to further ablate the remaining population of T/B cells.

      Interesting though. I agree – not a magic bullet, but definitely a high pretty calibre one!

    • The "low intensity regimen" is referring to the intensity of the chemotherapy, yet the regimen can vary widely inside this term. The interpretation that only myeloablative chemotherapy works seems little arbitrary when the rest of the data we have are contradicting to that…

  • Does this not sound like a problem with age related b cells as article couple of months ago suggested. All autoimmunity is driven by b memory cells which become corrupt over time due to antigens being clustered together on b cells. A bit like a hard disk that has bad sectors and need re-formatting. Not sure how the first article is related to 2nd article on HSCT! Other than the person was destined to have HSCT.

    • Merry Christmas MD1 (and MD2, Prof G, and the wider Barts' team). A big thanks for all of your hard work and truly awesome achievements this year, and for your time in keeping this blog alive and such an invaluable source of information, debate and good-hearted banter! Legends – hope you all enjoy some well-earned time off with your families over the festive period.

      I can't resist a quick bite on your HSCT post though – it just wouldn't be Christmas without it! 🙂

      I can't see from the abstract – what protocol did they use for low & intermediate HSCT? Was low intensity high dose Cyc + ATG/Rituximab, and intermediate used BEAM?

      Also, would be good to know EDSS and disease duration of the cohort going into the study, plus how the outcomes/NEDA periods mapped to disease type (RRMS, SPMS, PPMS, RPMS). Can you share? I'm guessing we're talking significantly advanced disease at the point of HSCT, rather than hitting hard and early…

      Difficult to put too much value on the outcome for intermediate versus low-intensity when n=1 for intermediate (and without context of which patients had progressive vs relapsing disease, or what their respective EDSS was)…

      "If you don't do myoablative HSCT it may not be that good" feels like an over-reach based on this data (at least in isolation).

      When looking at Nash/HALT-MS (BEAM) vs Burt/MIST (CYC & ATG) there actually wasn't a huge difference in efficacy demonstrated between low/medium at 3 years – albeit that this may change as the results play out over a longer time period. We shall see.

      On a side note (not specific to HSCT), do you think they need to look at trial design in long term studies, or at least the way they represent results? Presumably, even if disease activity is stopped, if you're a patient with long disease duration and significant damage/EDSS, presumably a small degree of progression over such a long period (though not MRI or relapse) is to be expected from natural degeneration/accelerated ageing based on existing damage already sustained, independent of new disease activity? Should that be reflected, rather than writing off as a failure?

      I guess, to my mind, a better way of representing and measuring success may be plotting average the pace of progression over time in treated patients versus control (e.g. if someone said to me take X drug, and you'll only accumulate 1 EDSS point of damage over 10-15 years and then slow/plateau at that, versus an expected 3-4 EDSS progression over the same period otherwise and accelerated worsening after that, I'd consider the former to be a massive success rather than a failure).

      Anyway – enjoy your Crimbo dinner and thanks again all. Look forward to 2018 being the year of the cure! Have a good one guys! 🙂


    • A small degree of progression over such a long period ) is to be expected from natural degeneration/accelerated ageing based on existing damage already sustained, independent of new disease activity? Should that be reflected, rather than writing off as a failure?

      I agree with you

      Chrimbo dinner thanks….but will probably stuck on floor for another day…. unable to walk at moment.

  • "Does this tell us the problem is in the plasmablasts"

    I think you are not alone

    "Given the stability of oligoclonal
    bands in multiple sclerosis patients who had undergone
    bone marrow transplantation, it is tempting to speculate
    that persistent memory B cells in the CNS are the source for the
    continuous local development of short-lived PB in the brain
    (Saiz et al., 2001). However, little overlap of B cell clonotypes in
    the CD138+ and CD19+ subset was observed in a recent study
    (Ritchie et al., 2001). Although this study did not differentiate
    between CD19+CD138+ and CD19+CD138– cells, the
    results would rather argue against continuous maturation of
    memory B cells to PB in the brain. Alternatively, PB primed in
    the lymph nodes by antigens released from the CNS may
    continuously be recruited to the brain throughout the disease

    Short-lived plasma blasts are the main B cell
    effector subset during the course of
    multiple sclerosis



By MouseDoctor



Recent Posts

Recent Comments