Time to Eat Humble Pie is CAR-T Ready for MS?


Yesterday, I talked about thoughts that I have held about the poor level of efficacy of antibodies used to target events within the CNS.


Therefore, it is only appropriate that I report this new study as it may explode some of my old views and so I probably need to eat some humble pie. This study is not about MS, but it has some real implications for MS.

So first I have to explain this therapy and explain my concerns but also indicate the potential benefits.

As I have said antibodies penetrate into the central nervous system poorly, so antibodies targeting CD20 (rituximab, ocrelizumab, ofatumumab, ublituximab) are not going to get into the brain to any meaningful concentrations. Therefore they are unlikely to clear the B cells from the CNS.

But there is a way to do this that is currently approved for B cell cancers. This type of therapy is called CAR-T therapy. This means chimeric antigen receptor T-cell therapy.

T cells recognise their targets via their T cell antigen receptor. This sees a bit of the target but it has to be in the context of a transplanation/DNA fingerprint molecule called the major histocompatibility complex. The T cell receptor then activates the T cell via signalling by a molecule called CD3. If you had T cells that could recognise your B cells they could destroy them and importantly they can traffic into the inflammed CNS, just like any activated T cell. So they could clear the CNS,

However, you would have very few or in reality no T cells in your body that can recognise your own B cells. However you can make then using a chimeric antigen receptor (CAR). Here you generate an antibody against your target. You take the antibody binding element and generate a molecule that has CD3 elements so when the target is bound by the antibody it will deliver a T cell activating signal. You take T cells and then infect them with a gene construct that codes the CAR so it makes the T cell react to the target for the CAR plus it recognises targets through its own T cell receptor.

In this study they take over 100 million T cells from five different individuals and infect them with a B cell targeting CAR and so about 20-40% of them gain the potential to target B cells. There are then injected into the indivudals with the B cell autoimmunity Typically the T cells that are generated are CD8 killing T cells that recognise MHC molecules found on alll cells, but B-cells and macrophages express an MHC molecule that can select CD4 T cells. So in this study CD4 CAR-T cells are selected. You can make the CAR-T to target antibody forming cells using B cell maturation cell antigen-specific CARs. However the first type of CAR-T was to have a CAR-T receptor that targets CD19 and so this causes killing of essentially all B cells, except long lived antibody-forming cells called plasma cells. The T cells injected did not look Exhausted (low levels of CD57/PD-1). When put into people their numbers increased. The cyclophosphamide and fludarabine would help clear out the blood and bone marrow and would no doubt have beneficial effects as this approach is often used in HSCT with doses of chemotherapeutic agents that are not that disimilar.

This approach is already approved for use of B cell cancers that cause fatal disease. The capcity of these CAR-T cells to clear cancerous B cells is amazing and life-saving. This is great. However, the side effect is that healthly CD19 B cells will also get destroyed due to the CAR-T cells. Now there are people that are genetically deficient in B cells, meaning they are at risk of servere infection but this risk is treatable.

Cell numbers drop because they have chemotherapy but B cells dont recovery, CD8T cells do and CD4 T cells recover but remain low and significantly depleted largely below 200 cells/ul = CD4 lymphopenia

However, I thought that once these CAR-T targeting B cells are generated, they will wipe out all B cells everywhere including the blood, lymph glands, bone marrow and importantly brain, but this will be forever as the CAR-T cells once formed should last forever, just like immunity to infection.

So this is a risk as no B cells lead to reduced antibodies and risk severe infection. Whilst this may be acceptable to cure you of a disease that is rapidly going to kill you, MS is generally not such a rapidly fatal disease, so it is worth the risk? However, if you could destroy the CAR-T cells or turn the CAR-T cell off once they have done their job, it would be far safer. This is the next generation that I was hoping to see happen in cancer before MS trials. You could engineer the CAR-T to been a factor for expression or introduce something that allows you to kill the cells. This already achievable scientifically.

However, there are always people willing to take the risk so to my surprise this study was published in people with autoimmunity, using the bog-standard CAR-T CD19 B cell depleting therapy. This seemed to stop further disease activity, within 3 months in most people, which is interesting as the autoimmune disease was thought to be mediated by antibodies and CD19 is not expressed by long-lived antibody forming cells.

Mackensen, A., Müller, F., Mougiakakos, D. et al. Anti-CD19 CAR T cell therapy for refractory systemic lupus erythematosus. Nat Med (2022). https://doi.org/10.1038/s41591-022-02017-5

Autologous T cells from patients …..were transduced with a lentiviral anti-CD19 CAR vector, expanded and reinfused at a dose of 1 × million CAR T cells per kg body weight into the patients after lymphodepletion with fludarabine and cyclophosphamide (Chemotherapeutic agents that can deplete white blood cells) . CAR T cells expanded in vivo, led to deep depletion of B cells, improvement of clinical symptoms and normalization of laboratory parameters including inhibition of pathogenic antibodieso Remission was achieved in all five patients after 3 months. Drug-free remission was maintained during longer follow-up (median (range) of 8 months after CAR T cell administration) and even after the reappearance of B cells, which was observed after a mean (±s.d.) of 110 ± 32 d after CAR T cell treatment. Reappearing B cells were naïve and showed non-class-switched B cell receptors. T cell treatment was well tolerated with only mild cytokine-release syndrome. These data suggest that CD19 CAR T cell transfer is feasible, tolerable and highly effective

So why is this study interesting to MS?

MS is inhibted following both CD20 and CD19 B cell depletion so CD19 CAR-T should work in MS. Indeed about 3-4 years ago I heard someone said they were planning to do this in MS. So maybe my fears were unfounded.

Samples at BL= Baseline or in the RC=reconstitution period

B cells were rapidly depleted and stay depleted for some time, so the memory B cells are removed so that is a good sign for MS if my ideas are correct. But to my surprise in this study, B cells are not wiped out but recover. If I had done more reading it was evident that this has been seen before, but it has to be said some people remain B cell depleted for a long time and have to get supplementary immunoglobulin therapy. Likewise the CAR-T cells did not disappear in all people but did disappear in some and remember. That CAR-T is not 100% successful in people with cancer and the CAR-T cells are unnatural cells and may be removed by the immune system. However, this removal does not appear uniform. In the current study the recovering B cells were naive B cells so maybe they will respond to new vaccines and new targets and it shows a new B cell repertiore is appearing perhaps reset. Memory B Cells that have been previously activated were not seen.

How can they re-appear in the face of T cell destruction?…Maybe they stop making CD19 and so the CAR-T doesn’t work, maybe the CAR-T cells are lost or maybe the CAR-T cells become exhaused so they are no good at killing targets. This paper doesn’t really address these issues. However, the memory B cells are gone and so is disease in most individuals, remembering that the people in the study were refractory to many previous treatments.

Adapted from Mackensen et al. 2022

Next up they look at old infection and vaccination and can still find the vaccine antibody responses but the pathogenic antibodies disappear. It all seems a bit too good at this point. So how do you explain this?Is the pathogenic antibody coming from a different cell, but why maybe CD19 positive antibody forming cells that get depleted verses long-lived plasma cells not expressing CD19 that should be initially depleted although if the B cells are gone I suspect that may be the case.

Now it is appropriate to say that this approach also have side-effects notably you get a cytokine storm as the immune cells rapidly expand whilst others are dying. This is often seen with CD20-monoclonal antibody treatment when you get infusion reactionss, but this was not a major issue possibly because they had all been on B cell immunotherapy. There is also a potential for a neurotoxicity syndrome with CAR-T therapy but this was not the case here.

So the question is is the approach being done in MS?. I know suspect the answer is when not if. The logical starting place would be relapsing MS as you know if it works within a few weeks as it will stop new lesion formation, but it should work in active progression too. Will this get done at a current cost of about half a million dollars based on the cancer treatment. Maybe it would be tried in Neuromyelitis optica where CD19 depleting antibody is approved. and relapses are damaging.

Anyway it shows why sometimes you simply have to do the study rather than avoiding because of something you think..You never know you may not be thinking straight

CoI None

Disclaimer This is my views

About the author



  • Thanks MD.

    These naive B-cells, when they exposed to an antigen, they will still class change and hunted down by T-cells? Or they can class change to none CD-19 memory cells/plasma cells directly?

    • I would think if they expressed cd19 they would be hunted down. So I suspect they are cd19 negative unless the car t had died I will read again on a larger screen later it was too painful doing it on laptop last night.

  • (I’ve not read the paper, so maybe this is answered there)

    How long do CAR-T cells remain in your body after the transfusion? I thought one of the limiting factors for CAR-T against cancer was that the transgene cell count dropped and the cancer relapses. If the same thing happens here after the autoimmune memory cells are knocked down you might get a decent prolonged remission without prolonged immunosuppression.

    Also, do you not get T-cell exhaustion effects after a while? Presumably the patient’s bone marrow and whatnot is still producing CD19+ cells, which looks a lot like a chronic infection.

    • I don’t know I will do some reading and get back to you, if it does not last then that is good if the transgene gets in genome they should last forever like a memory t cell does. It was quite late when I did the post

  • I was reading about ATA188 just yesterday. How come this approach did not give wow results in the last update back in July? Is EBV the wrong target?

    • It has never given wow results, even when it was immunological sensible when they used autologous cells from the same recipient, going allogeneic where there is a mis match does not make alot of sense to me as it throws away alot of historical biology and allogenic cells can react to allogeneic MHC even when there is no specific antigen ….Sales of Car-T have not been great but as it costs half a million dollars it is going to have issues. It is not easy to do either

        • I would not confuse the T cell therapy of ATARA which is generation of a EBV specific T cell with proper CAR-T cell therapy (which does not need antigen presentation, however you could make an anti-EBV car-T
          Is there money flowing anywhere?….Not my way:-(

    • ATA188 has a hard road ahead of it if they want to appease investors looking for remyleination in such a short period of time with so few individuals. But back to the Car therapy-

      CD19 should be on all B cells after they recombine their Ig as a Pre-B cell until they terminally differentiate into a plasmablast. The signaling through the CD19 is necessary for survival after recombination of the receptor based on B cell development, so these B cells must have some expression on their surface. If these Car treated patients have CD19+ naive B cells detectable in circulation, likely the Car has petered out in them to a degree, possibly because of anti-Lentiviral activity by their own immune system destroying the Car T cells or exhaustion depending on the construction of the Car they used. I do know depending on the signaling domain and other transmembrane stitching for the Car it alters the possibility of CS and longevity of the Car itself. Maybe they picked a “de-tuned” version of a Car that worked well enough, but not so strong as to completely scrub the immune system as compared to one used against a cancer. Car construction is very involved at this point.

      Also, there was another paper that looked at EBV responses within normal and MS patients, compared against COVID infection in recent JEM.


      There is a response in MS patients against EBV- its much more effector memory as compared to central memory. They looked peripherally and in spinal fluid and found a broader anti EBV response, suggestive to them of recent activation. This may circle back to Atara’s question of whether increasing the frequency of CD8 response against EBV is going to work if MS patients already have that response (Hohendorf found anti lytic and latent response in MS patients). You could argue its an ineffective response, but then is more the answer? Would you focus on latent? Especially if its a fundamentally bad set of antigen presentation in the patient due to their unlucky set of MHC I/II molecules (DR15 rearing its
      head) which you can’t rework because that’s their genetics?

      To me, this paper suggests throwing more T cells at EBV may not be the way to go and you need to figure out an alternative way to eradicate B cells, which is why the Car therapy, mAb therapy, maybe lemtrada/tenofovir antiviral therapy specific for EBV is the way to go.

      • Personally it makes no sense…they have a T cell that is targeting EBV and they think it will repair….I ask what is the logic and what is the mechanism? Growth factors?…it seems to be clutching at straws

  • Time
    Only 3 months study?

    We need to wait

    Will send an email to the author of the study in a couple of months


By MouseDoctor



Recent Posts

Recent Comments