HSCT in MS, more a remission than a cure


In a prose to seduce the weary palate, the British neurologist, I recollect saying: “What should be the aims of HSCT treatment? At worst, remission. At best, a cure“.

I may just have to eat my words…

In a small study from Uppsala University (n=45), Larsson et al. report the effects of HSCT on neuronal damage (using neurofilament levels) and inflammation (IgG index and oligoclonal bands (OCBs)). The latter is very important as the presence of antibodies in the CNS space is a good refection of the underlying disease process. In fact >95% of those with MS are OCB positive; the remaining probably don’t have the disease or the lab technique/interpretation was poor.

I remember during my PhD re-evaluating the oligoclonal band status of a group of individuals with SPMS ~15-20y after diagnosis (many of whom have never received treatment), and to my surprise all were OCB positive.

It therefore goes without saying that OCB status in MS is very important, and furthermore ridding the body of OCBs, even more so.

So, what is the effect of HSCT on OCB status?

At >1500 day follow-up (~49 months or 4 years), in the 45 participants studied, 98% were OCB positive prior to HSCT. However, 74% remained positive afterwards (see Figure 1 below). Whilst, the proportion of those with normal IgG index (a quantitative measure of antibody production in the CSF) increased from 30% to 54%.

In my opinion, this rate of elimination of the underlying autoimmunity is insufficient to declare HSCT a success, in so far as a cure is concerned.

In a recent publication by Rejdak et al. sub-cutaneous cladribine achieved OCB negativity in 55% of PwMS at 10 year follow-up.

On the other hand, with regards to the proportion of those with elevated neurofilament levels in their CSF before HSCT (=72%), at the latest follow-up only 24% had levels exceeding the upper-limit of normal (see Figure 1 below). This indicates that HSCT has a greater chance of achieving long-term remission in selected individuals, than a cure.

Figure 1: Normalization of IgG Index, IgG OCB, IgM OCB and NFL in CSF. Proportion of patients with normal values of CSF IgG Index (a), NFL (b), IgG OCB (c) and IgM OCB (d). Baseline values were compared with values at latest follow-up. ns, not significant.

My earlier enthusiasm towards HSCT somewhat deflated, I am in other ways strangely buoyed by the knowledge that clinicians are now starting to ask the right questions about treatment efficacy in MS. I eagerly await further data on HSCT and its comparative treatments.

I would also like to take this opportunity to thank all those who participated in this study, your contribution has been invaluable.


Mult Scler. 2019 Jul 26:1352458519863983. doi: 10.1177/1352458519863983. [Epub ahead of print]

Intrathecal immunoglobulins and neurofilament light after autologous haematopoietic stem cell transplantation for multiple sclerosis.

Larsson D, Åkerfeldt T, Carlson K, Burman J.


Oligoclonal bands (OCB) are widely believed to be stable over time and rarely affected by disease-modifying treatment in MS. It is presently unknown how intrathecal immunoglobulin production and other cerebrospinal fluid (CSF) biomarkers are impacted by a highly efficacious procedure such as autologous haematopoietic stem cell transplantation (aHSCT).


To describe the evolution of intrathecal immunoglobulin and neurofilament light (NFL) over time in MS patients treated with aHSCT.


In this retrospective study, available data from previously made CSF investigations in 46 patients treated with aHSCT were analysed.


After a median follow-up time of 745 days, immunoglobulin G (IgG) OCB remained detectable in 74% of patients, the proportion of patients with a pathological IgG index went down from 70% to 46%, and the proportion of patients with a pathological NFL went down from 72% to 24%. In patients with follow-up time >1500 days, IgG OCB were detectable in 50% of patients, 14% had a pathological IgG index and none a pathological NFL.


Intrathecal immunoglobulin production and NFL were lower after treatment with aHSCT, decreased over time and were normalised in a significant portion of patients. This challenges the notion that OCB are unaffected by therapeutic intervention in MS.

About the author

Neuro Doc Gnanapavan


  • Maybe I’m missing something, but according to my reading of the abstract after 745 days 74% of patients remained OCB positive, but after 1500 days this number dropped to 50%.

    To my relatively untrained eye, this would seem to indicate that the treatment effect gains in potency as time goes on, so perhaps looking at HSCT as a potential cure isn’t completely off the mark. Furthermore, the abstract states that after 1500 days only 14% of patients had pathological antibody indexes, and none had elevated NFL levels. Again, seems pretty impressive.

    Please, correct me if I am wrong on this…

    • So, the OCB status does improve over time. This tells us something about how HSCT may possibly be working. The effect is reliant on probably interaction between the new cells produced and any left over cells residing in the brain. I would expect this to take time. 24% still had elevated NFL so the disease has by no means disappeared in some. The NFL decrease like the IgG index reduce over time suggesting a link between the two pathological processes! Very interesting findings

      • Tissue immunology is in its infancy as is, and the human CNS tissue immunology is of the most challenging to study for obvious reasons(excluding using CSF as a proxy in the same we use blood as a proxy in other organs).

        I feel like we are potentially jumping the gun on the significance of OCB loss. Especially when the disease seems “cured” clinically by other measure and the patients have improved! But I might be wrong, of course it is best to remain agnostic.

        I don’t know, sometimes unconscious bias can work as a gravitational pull in many ways.

    • So the number of participants does drop between 500 and 1500 days but not by much, and the LPs are mostly on the repeats performed at 2y rather than the at least 4y (>1500 days) noted in the results. Hopefully the team will continue to do more and we’ll know further in the near future.

  • In a recent publication by Rejdak et al. sub-cutaneous cladribine achieved OCB negativity in 55% of PwMS at 10 year follow-up.

    Mixed results
    CSF examination
    The number of oligoclonal bands did not change in any

    Cladribine in treatment of chronic progressive multiple sclerosis


    Ocb in ms remains an hot topic this “Maven” thinks they are not specific

    Background: It is uncertain whether there are autoantibodies detectable by indirect immunofluorescence
    in the serum of patients with multiple sclerosis (MS).
    Objective: To determine whether there are anti-central nervous system (CNS) autoantibodies detectable
    by indirect immunofluorescence in the serum of MS patients.
    Methods: Sera and in some cases cerebrospinal fluid from 106 patients with multiple sclerosis, 156
    patients with other neurological diseases, and 70 healthy control subjects were examined by indirect
    immunofluorescence using cryostat sections of rat cerebrum fixed by perfusion with paraformaldehyde.
    Results: Autoantibodies were detected that recognized more than 30 neuronal, glial, and mesodermal
    structures in 28 of 106 MS cases. Most were also detected in patients with other related and unrelated
    neurological diseases and several were also found in healthy controls. Novel anti-CNS autoantibodies
    recognizing particular sets of interneurons were detected in both normal controls and in subjects with
    CNS diseases.
    Interpretation: Serum anti-CNS autoantibodies of diverse specificities are common in MS patients.
    The same anti-CNS autoantibodies are not uncommon in patients with other neurological diseases. The
    findings provide no support for the proposition that myelin breakdown in MS is caused by exposure of
    intact myelin sheaths or oligodendrocytes to a pathogenic serum anti-myelin or anti-oligodendrocyte

    John W Prineas and John D.E Parratt
    Multiple sclerosis: Serum anti-CNS

    DOI: 10.1177/

  • Considering a significant proportion went to be OCB- and most normalised their NFL, would this not be remission at worst and ”cure” at best.

    Seems about right. I guess it adds more weight to the inclusion of Cladribine as a follow-up treatment or added to the HSCT cocktail?

    • So follow on treatments after HSCT is not something that’s discussed and in those who relapse it’s often onto the low efficacy drugs to avoid PML risk mainly.

  • Non-ablative regimen? Nevertheless, the data is disappointing and brings the risk / benefit equation back into focus (or at least it should).

    • Why is it disappointing?

      Patients have very little-disease progression, improving OCB indicators and normal NFL. It is still somewhat nebular as to what OCBs actually represent, treat the patient/clinical picture not the lab result in isolation!

      If anything these are great results, especially considering 31 of them had CyATG.

      Of course, we want to get everything as “right” as possible but considering that patients seem to respond very well to HSCT everything is pointing north at the moment. Maybe the HSCT process needs improvement but these results are still very positive.

      Alemtuzumab is more convenient but seems to be more risky!

      • OCBs have been linked to conversion from CIS to MS, and have a greater risk of developing SPMS (particularly the IgM pattern). Relying solely on clinical findings has left as open for not really understanding what the disease is and more importantly what treatments are doing. It’s time to really get a proper grasp on this condition

        • Sorry I should say, the importance of the “persistence of OCBs” is nebular. But even so, it is more an association than anything, and yes provides some ontological clues potentially, but the presence of OCBs in the CIS to MS conversion is a different biological state in many ways.

          I think that the expectation that HSCTs will provide a “cure” in every patient is optimistic at best and unrealistic at worst.

          For all we know, these OCBs persist due to plasma cell clones that eventually die off without appropriate T cell help following HSCT?

          But it is putting a lot on OCBs when we still do not even know their significance. We might end up focussing on something that ultimately, in post-HSCT patients, is not important. But maybe it is, I guess we do not know yet.

          Just saying.

      • 15 in fact had the myeloablative BEAM chemotherapy,

        This is incorrect BEAM protocol is not myeloablative

        BEAM is (RIC) Reduced intensity regime

        Conditioning regimens for autologous HSCT

        Conditioning regimens were divided retrospectively into (a) ‘high intensity’, including TBI or high-dose busulphan-containing regimens, (b) ‘low intensity’, referring to CY alone, melphalan alone and fludarabine-based regimens, or (c) ‘intermediate intensity’, including other combinations, such as BEAM, and, in most patients, the combined use of ATG with high-dose CY or other chemotherapy. There was a significant relationship between efficacy and intensity, balanced by the inverse relationship with toxicity as reflected by TRM. Overall, ‘intermediate intensity’ conditioning regimens were associated with significantly improved outcomes compared with ‘low-‘ and ‘high-‘ intensity regimens.17, 18

        Although more profound responses are possible with high-intensity regimens, caution should be exercised given the substantial short- and long-term toxicity. A further analysis of the MS data supports the significantly higher TRM with oral high-dose busulphan.52 The use of irradiation-containing protocols outside of clinical trials has been questioned owing to the long-term adverse effects, including malignancies, even after low-dose irradiation.14

        Haematopoietic SCT in severe autoimmune diseases: updated guidelines of the European Group for Blood and Marrow Transplantation


        • So there are in the literature a lot of attempts to standardise these definitions among haemoncologists. You’ve picked up on one. In simple terms a myeloablative regimen is one that uses total body irradiation and/or alkylating agent at sufficient doses that don’t allow the self haematological cells to recover. Now melphlan in BEAM like Busulfan that you’ve mentioned are alkylating agents but the later from what I’ve seen is more toxic on marrow cells. But combining different regimens greatly skews where you’d land on the sliding scale. So probably a more simpler way is to talk about high-dose chemotherapy regimen versus reduced intensity and non myeloablative chemotherapy regimens. A combination of cyclophosphamide +/-ATG would be easily considered in the non myeloablative end.

          • Even that its not so simple

            It depends on the Cy dose

            For some conditioning regimens, classification
            may be not straightforward. One example is Cy 200
            mg/kg, with or without thymic radiation [47,48].
            This is a truly nonablative regimen, because it does
            not kill SCs, but it does cause profound cytopenia, especially
            when given over 4 days (50 mg/kg/day 4) in
            patients with SAA, and is followed by allogeneic hemopoietic
            SC. Some of these patients may recover an autologous
            hematopoies, and Cy 200 mg/kg has also
            been given in aplastic anemia, without SC support
            [49], although the rate of lethal infections, because of
            prolonged cytopenia, was very high [50]. Therefore,
            Cy 200 mg/kg does not fit our working definition of
            a MA conditioning nor of a NMA conditioning, and
            falls in the category of RIC conditioning

            Defining the Intensity of Conditioning Regimens:
            Working Definitions

            Biol Blood Marrow Transplant 15: 1628-1633 (2009) 2009 American Society for Blood and Marrow Transplantation

            This is familiar because i “drink” this particular one almost 2 years ago


  • “It is not obvious how to interpret these findings in individual patients. Many patients, who to all appearances were doing well, without clinical relapses or MRI disease activity, still retained IgG OCB and pathological values of IgG index and NFL in the first years after aHSCT. Over time, IgG OCB tend to diminish and the values of IgG index and NFL decrease in stable patients. In most patients, this is a very slow process and abnormal CSF findings should not be interpreted as treatment failure. In our experience, patients who experience relapses follow a slightly different pattern, with sudden increases in the IgG index and NFL in conjunction with the relapse”….

    Would love to see some further data related to which patients showed which patterns, was there a correlation to BEAM or CyATG? What association did age or length of disease or EDSS have? What were the prior treatments?

    Not sure about Uppsala, but Karolinska leans heavily on rituximab as a first-line. Why did these patients have HSCT vs other treatments, or what other treatments had they been on?

    • Likewise I’d be interested to know how BEAM did versus Cyclophosphamide, also how the other treatments do by comparison!

  • I copy my comment from a different post
    (the important aspect of OCBs as an epiphenomenon)

    Maybe we should not count much on OCBs to verify a cure for MS. OCBs can be absent in a lot of MS patients, can be present in non MS patients, can vanish with DMTs that are inferior to HSCT, can persist in otherwise normal MRI (including brain atrophy) HSCT MS patients. We dont know what they do and why they are present, so they should be investigated but not be a part of the cure criteria for now.
    NFl levels, absolutely.

    From Joachim Burman’s PhD thesis out of Uppsala …”intrathecal antibody production can be demonstrated in a vast majority of MS patients. It has been claimed that
    oligoclonal IgG bands persist in the CSF of almost all patients, but in another study with a high intensity regimen 75% retained oligoclonal bands.143 In Swedish patients, only 69 % retained oligoclonal bands after a
    low-intermediate conditioning. This discrepancy may be due to that Swedish
    patients were on average treated earlier, most of them still in the relapsing remitting phase. The biological significance of these antibodies is not clear and for all we know they could be an epiphenomenon. It is well known that
    oligoclonal bands are present to variable degrees in other inflammatory CNS
    disease, not least in infection with B burgdorferi, when antibody
    production can persist for years after the infection has been cleared.

    The latter illustrates that is not necessary to lose the oligoclonal bands to
    allow us to speak of cure”…


    • The latter illustrates that is not necessary to lose the oligoclonal bands to
      allow us to speak of cure”…

      I think that is deeply mistaken. Levels of antibody in the CNS interacting with populations of activated microglia is not what the doctor ordered.

      • Recovery from and consequences of severe iatrogenic lymphopenia (induced to treat autoimmune diseases)

        Surprisingly, antibody levels did not drop substantially, in spite of the severe CD4 T and B lymphopenia during the first several months posttransplant. Consistent with that, oligoclonal immunoglobulins in the cerebrospinal fluid of most multiple sclerosis patients treated with autologous transplantation did not disappear [3,37,38]. As the half life of IgM and IgA is only ~5 days and that of IgG only ~23 days [39], the antibodies detected in the sera of our patients were continuously produced, presumably by plasma cells generated pretransplant. Plasma cells are radiation-resistant and long-lived [40–42]

        From the infectious disease point of view, the persistent production of antibodies by plasma cells generated before the autologous transplantation may be beneficial, as most frequent pathogens are likely encountered pretransplant. Unfortunately, from the point of view of autoimmune diseases caused by autoantibodies like pemphigus (anti-desmoglein) or Lambert–Eaton myasthenic syndrome (anti-voltage gated calcium channel), the persistent production of autoantibodies may be deleterious. This may not apply to systemic sclerosis or multiple sclerosis in which the role of autoantibodies is uncertain

        This patients had Hsct Tbi (8Gy) Cy and atg,
        still ocb did´nt disappear


        • Exactly, furthermore, not all Fc Receptors are activating. In the same way, not all “inflammation” is destructive.

          • But why would you want that in an organ that ordinarily goes out of its way to maintain the most rigorous homeostasis?

        • This review is based on serum antibody levels. They do reference an assessment on OCB post 1yr HSCT by Saiz Et al. In Neurology 2001, that only had 4 participants, and CSF was only collected up to year 1 post HSCT. I don’t think you can conclude anything from this…

          • @MouseDoctor2. Because homeostasis is a process to balance inhibition and activation. From an evolutionary perspective, it one must have breaks to stop or slow acceleration, otherwise we would also have encephalitis/meningitis all the time?

          • *Because “homeostasis” is a term to describe the process of balancing inhibition and activation

            *encephalitis through either excess activation on one hand or uncleared infection on the other

  • One would like to believe that geting rid of Ocb would be a potential “cure ”

    So in this study patients had no plasma cells no b cell no t cells no lymphoid structures in the meninges and

    Guess wath?

    They progress 🙂

    Autologous haematopoietic stem cell transplantation
    fails to stop demyelination and neurodegeneration in
    multiple sclerosis
    Surprisingly, we found no B cells, plasma cells or
    lymphoid follicle-like structures in our cases, although it is
    suggested that pretransplant plasma cells and, accordingly,
    possible pathogenetic antibodies persist in the CNS after
    extremely lymphoablative conditioning (Storek et al., 2004).
    Thus, the pathogenic role of plasma cells or antibodymediated
    damage remains uncertain.


    This one had no Ocb after hsct and

    Still Prgression

    Progression and CSF Inflammation after Eradication of Oligoclonal Bands in an MS Patient Treated with Allogeneic Hematopoietic Cell Transplantation for Follicular Lymphoma

    DOI: 10.1159/000339738

    • I wish you would all comment so that a lay person understands what you are actually saying! Perhaps a closing easy précis would be useful

    • The follicular cell lymphoma case with MS would probably have a very confused immune system. Fiddling with that one may have unexpected side effects. I had a patient who received glivec for chronic myeloid leukaemia and developed MS – the question is was that individual always genetically predisposed to develop MS or was it the drug that caused it. Fortunately the cases of MS after glivec are only a handful, so probably the former!



Recent Posts

Recent Comments