Nature: ‘It’s all gone’: CAR-T therapy forces autoimmune diseases into remission

Dual is the best but perhaps too aggressive & only BCMA has good results too but its all up to 2 years follow up so might be too optimistic to say so
BAFF is against all 3 receptors; BCMA, TACI and BAFF-R.

 

Yes, 'seronegative RA' is probably a mixture of people with small immune complex arthritis who don't happen to have their complexes involve rheumatoid factor or anti-CCP and people with psoriatic type T cell arthropathy without psoriasis and a whole lot of other smaller groups we don't understand. People with abnormal Ig profiles tended to respond to rituximab so presumably have complexes for some other type of B cell dysregulation. And of course a small proportion of myeloma and lymphoma patients with monoclonal gammopathies have a rheumatoid type arthritis - which makes sense. Some of them responded to rituximab but most myeloma patients probably need plasma cell ablative therapies.

 

I discussed this this evening with Jo Cambridge, who did the first set of antibody profiles on this. Rituximab depletes plasmablasts very effectively simply because it kills any B cells that might be turning into plasmablasts and any plasmablasts already present will either die or transform into plasma cells. So autoantibodies generated by palsmablasts disappear with ritux.

We discussed our Anti-SM data. From what we could remember it is very patient dependent. Some Sm goes down and she does not. There is no clear pattern, unlike anti-DNA that nearly always goes down. RNP seemed pretty resistant. If an antibody goes down it is probably produced by short lived splenic plasma cells. If it doesn't it is probably bone marrow long lived PC. The plasmablasts are really just a transient population that people can measure easily but almost certainly don't mean much. There was a lot of fuss about the importance of plasma blasts in RA but our impression was that it was based on some fairly wooly thinking.

 

Thank you Dr. Edwards. I attached a figure from Dr. Schett's review. However, after reading your comments on ritux depleting plasmablasts, I realized the review figure could be misleading. According to this "static and fixed" view of marker expression, ritux will not touch plasmablasts, CD19 CAR T will not deplete LLPC. But the reality is B cells do not have such rigid clear-cut lineage commitment. Many "early committed" LLPC will still express CD19; and plasmablasts survival still responds to ritux. Although we cannot name/identify the distinct population separated by CD19 vs CD20 marker, it is possible that CD19 CAR T still deplete a significant amount of B cells, eg CD19+ LLPC, that express CD19 but not CD20. And/or, CD19 CAR T can penetrate better to deplete those deep-in-tissue cells beyond ritux's reach. The net effect is more cells that account for the SLE immune complexes are depleted, leading to longer remission, better responsiveness and efficacy.

If that was true, then we have an explanation for the Anti-SM, dsDNA, and RNP mystery: these autoreactive Ab-producing B cells are at the spectrum of CD19/CD20 expression. Those B cells, regardless what we call them, are increasing autoantibodies production amid their lineage commitment to LLPC with downregulating CD20, and eventually CD19.

 

Ah thank you. That's good to know. Is the Luminary BAFF including all three?

Question: I was checking Novartis ELIANA trial and found out that FDA issued approval of Kymriah based on ELIANA, which is only a phase II with 88 patients. So does it mean we do not require Phase III for CAR T for SLE if the efficacy/safety profile looks good? Is it possible to have CAR T approved after the ongoing trials (a long list with a couple of Phase I/II trials) or we must wait until a Phase III?

 

Do you know about anti-ku?
A person I know is doing CAR-T with anti-ku. He has myositis & SLE. low anti-dsdna. But he goes for CAR-T mostly for his myositis as he is in a wheelchair. We were wondering is CD19 enough for him...

 

I really think plasmablasts are irrelevant to any of this because they are a brief transient state. LLPC are not depleted much if at all by any of these regimens targeting CD19 or CD20. Short lived plasma cells in spleen deplete, but not because they express CD19 or CD20, simply because they are short lived. They die over a period of about six months. The decay in autoantibody levels seen for rheumatoid factor or anti-DNA occurs over months, it is not due to immediate killing of cells but attrition of cells over a long period. Ig half life, and particularly autoantibody half life, is expected to be much shorter.

 

That is since my time in the lab. I haven't seen any kinetic data after depletion.

 

https://www.nejm.org/doi/full/10.1056/NEJMc2403705

A new study showing scl70 can continue go down if cd19 car T cells sustained longer. Also chemo drug was given longer in this study. Just a Case study of course with limitations, maybe because spleen plasma cells are accountable for those antibody? Or raises hope for systemic sclerosis patients?

 

Hopefully, yes.

The main problem over the last 25 years has been that the rheumatology community has not even twigged to the fact that it is the antibodies causing the endothelial damage and the whole disease. Nobody much has even been trying to target B cells in scleroderma. One day people might realise that it all made sense in 1999.

 

Random thought - if scleroderma is accurately diagnosed (i.e. not a mixed bag of diseases) then would GWAS [or rare variant genetic studies] provide strong evidence that it is a B cell autoimmune disease i.e. that could be treated using rituximab (or similar B-cell depleting drugs)?

 

Scleroderma is known to be associated with an MHC Class II type, exactly as expected from a B cell disease. It is also associated with a range of signalling genes, some specifically B cells related. There are different genetic risk factors for scleroderma mediated by the three main antibody routes. (Scleroderma is really three almost identical and easily identifiable specific antibody diseases.) There is a nice review on the more recent findings.https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3803145/
In fact the genetics would almost fill a book.

 

Thank you.
Obviously I'm way out of my depth here [& I haven't looked at the link yet] but given the evidence, it would appear strange if there haven't been attempts to adopt prove treatments e.g. the approach you piloted i.e. rituximab.

 

Rituximab haș been tried and may have some benefit but we knew early on that antibodies of the sort seen in scleroderma tended not to go away. It is also a tough disease to treat because it is likely that a major part of the damage to blood vessels goes on silently even before symptoms appear and once fibrosis has occurred it is hard to see it changing. Perhaps more than anything the worst complication of rituximab is pneumonitis and the lung is already badly affected in scleroderma. I think at the time I considered trying it I was too concerned that we would kill people with respiratory failure.

But the scleroderma docs haven't even caught up with the fact that they are studying a B cell disease. You would be surprised how dumb some doctors can be - or maybe not so surprised knowing what we do.

 

@Jonathan Edwards so do you think that the persistence of CAR-T cells for 11 months (vs. average 3) is responsible for the complete disappearance of anti-scl70 in this case, or it was more just a natural death of plasmablasts, not happening all at once?

From the case report:

interesting that when they stopped mycophenolate (7 months) there was a rise and then again fall of antibodies.

 

I don't see plasmablasts as having much role - they are just a transient state in blood. I presume the antibodies continue to go down because splenic plasma cells are dying off and not being replaced while there are no B cells.

 

Not keeping up/late to the party (as usual!) - CAR-T seems a remarkable tool* - would it help re T cells i.e. are testing whether T cells are relevant ME/CFS (Mark Davis Stanford)/not?
While the sort of understanding Jonathan & Jo et al developed, i.e. re rituximab/successful application RA, is preferable, I guess there are other potential ways to discover whether CAR-T may be useful-

• serendipity - someone being treated for cancer - actually how the rituximab trial came about;
• genetic clues e.g. via DecodeME, or a rare variant (whole genoe sequence) study, i.e. indicating that ME/CFS could potentially be treated via CAR-T.

Other routes in e.g. are people who opt for assisted dying able to request a treatment they think might help (hope no-one finds that offensive)?


*https://www.cancer.gov/about-cancer/treatment/research/car-t-cells

 

Yes, so in any case it is not likely because as the authors stated because of CAR-T cells persistence or prolonged nycophenolate?!

 

Well, the continued existence the CAR-T cells would be responsible for the continued B cell depletion to some degree. It is not clear quite what determines B cell return after deep depletion. They remain absent quite long after therapeutic levels of rituximab are present, for instance.

Mycophenolate is very interesting because in the context of B cell depletion adding mycophenolate can lead to very profound hypogammaglobulinemia even though it does not do so alone.

It is time that some careful pharmacodynamic dose ranging studies were done with combinations - starting with just drugs rather than CAR-T I would say because nobody has investigated this well. When IgG levels fell with the mycophenolate combination the investigators got frightened and stopped the study - which may have been very silly since at some point it may be necessary to induce very low Ig levels to break a feedback loop. But it needs to be done with a very clear understanding of what you are trying to achieve and that has been lacking in the autoimmune field. It is easier to justify for cancer because aggressive curative treatment is seen as the best way to go and people are courageous.

 

Could you expand on this a bit? What type of combinations do you mean? Mycophenolate also suppresses T cells, correct? Do you think it is a potential drug candidate for MECFS, given that you have also suggested investigating Campath, which also depletes T cells?