The possibility of autoimmunity or auto-reactivity in ME/CFS

Jonathan Edwards said:
not being an autoantibody but a 'loose' regulation of B cell clonality allowing a wider range of antibody specificities.
Click to expand...
Forgive the probably elementary question. Is this anything like the Lipkin finding in the 1999 Borna study of "non-specific polyclonal B-cell activation"? (I did not find a thread for that paper, but @wigglethemouse has a nice summary post and video of Lipkin discussing this here. )

Lipkin's phrase has always stuck in my mind, because it was written verbatim on my bone marrow biopsy report, and I never understood what it meant. But I feel like I might have just had a small lightbulb moment reading the way @Jonathan Edwards put it.

That the issue might not be a single pathogenic antibody, but that there are lots of antibodies being made against "stuff" in general. The problem might not be a single ingredient, but the whole soup. Maybe even a "the whole is greater than the sum of its parts" situation?
 
I'm here mostly because you'd be shocked how hard it is to change 'autoantibodies.' And by that I mean reproducible signatures of pan-IgG autoantibodies on a protein microarray or PhIP-Seq assay - no titer, too expensive to do for titers - but - what antigens one individual has reactions to before and after a year or decades is quite stable. They often don't move much at all! Stable for years in the way you could identify someone at a crime scene and tell them apart from their identical twin. A true 'immune fingerprint' - and Rituxumab doesn't really change them observed in this way: https://pubmed.ncbi.nlm.nih.gov/38753445/

Seeing 'a change' induced by daratumumab 'yes' might be expected - but is inherently interesting as so many things don't affect these profiles. For example checkpoint blockade immunotherapy, cancer vaccines - the autoreactome is essentially rock stable or a few new abs - you never see things that go away. Something really profound changed in these patients post dara. Magnitude of this correlated with response. But tiny tiny trial and that of course does not prove mechanism. The 'responders' are somewhat confounded by the fact they have 'more' autoantibodies at baseline too (and higher NK tendency) which could represent some more general dysregulation prior to treatment.

I don't believe in 'autoantibodies' as this vague generic thing - I believe in discoverable antigen-specific mechanisms. In particular - I am looking for 'true' mechanistic events. Ie I want to find "a stark difference in some specific autoantibody level that could plausibly produce symptoms" and I want to see that go away in responders and remain in non-responders. I have looked for this for the study in-discussion and there is a real mechanistic direction pointed to from a few different angles with 'responder autoabs that go away' as to their functional ontology enrichment (not public yet) but the study is so small and there are no healthy controls - even when this is published will not be conclusive.
 
Mortal Machinery said:
Forgive the probably elementary question. Is this anything like the Lipkin finding in the 1999 Borna study of "non-specific polyclonal B-cell activation"? (I did not find a thread for that paper, but @wigglethemouse has a nice summary post and video of Lipkin discussing this here. )
Click to expand...

Not quite. Polyclonall B cell activation means that whatever B cells you have are activated, maybe expanding, generating plasma cells and antibody. What we proposed in the Qeios paper is that selection of B cell clones might follow looser rules - leading perhaps to a greater diversity of antibody species and possibly more autoreactive ones. There are a whole range of shifts in regulation of B cell populations that could come under either or both terms though.
 
Tyler Hulett said:
I'm here mostly because you'd be shocked how hard it is to change 'autoantibodies.' And by that I mean reproducible signatures of pan-IgG autoantibodies on a protein microarray or PhIP-Seq assay - no titer, too expensive to do for titers - but - what antigens one individual has reactions to before and after a year or decades is quite stable.
Click to expand...

I agree that autoantibody specificities tend to stay much the same in qualitative terms but you can change levels dramatically using rituximab (as we showed in our original studies on rheumatoid and lupus in the early 2000s). Also some antibodies, like anti-DNA, will change by four or five titre levels spontaneously over months in lupus. There have also been reports of new autoantibodies appearing on B cell return after rituximab - associated with things like new myasthenia gravis.
Tyler Hulett said:
I don't believe in 'autoantibodies' as this vague generic thing - I believe in discoverable antigen-specific mechanisms.
Click to expand...

I agree that specific pathogenic autoantibodies are what we have a handle on in most diseases. Nothing of that sort is seen in ME/CFS as far as I am aware. The idea of immunoglobulin populations mediating disease through a general mechanism that did not depend on such a classic specificity was raised in our Qeios paper (Edwards Cambridge and Cliff) as a speculation. I don't have a strong attachment to the idea but I think for ME/CFS we may need to think outside the box. The effector mechanism would not be through antigen binding but through activation of immune cells via interaction with complexes (ironically, similar to what happens in rheumatoid, but RA is maybe the exception that probes the rule amongst the known autoimmune diseases).
 
My mother had meningoencephalitis as child, for a month, after erroneously receiving a second dose of a vaccine. She has since suffered from poorly definable health issues. My brother and I also suffer from poorly definable illness (gradual slow onset). I suspect that this meningoencephalitis might indicate a genetic predisposition which could also play a role in my and my brother's illness. I'm curious to know if anyone here has some ideas about this.

I've also wondered if non-genetic inheritance of health issues plays a role.
 
Last edited:
Jonathan Edwards said:
I agree that autoantibody specificities tend to stay much the same in qualitative terms but you can change levels dramatically using rituximab (as we showed in our original studies on rheumatoid and lupus in the early 2000s). Also some antibodies, like anti-DNA, will change by four or five titre levels spontaneously over months in lupus. There have also been reports of new autoantibodies appearing on B cell return after rituximab - associated with things like new myasthenia gravis.


I agree that specific pathogenic autoantibodies are what we have a handle on in most diseases. Nothing of that sort is seen in ME/CFS as far as I am aware. The idea of immunoglobulin populations mediating disease through a general mechanism that did not depend on such a classic specificity was raised in our Qeios paper (Edwards Cambridge and Cliff) as a speculation. I don't have a strong attachment to the idea but I think for ME/CFS we may need to think outside the box. The effector mechanism would not be through antigen binding but through activation of immune cells via interaction with complexes (ironically, similar to what happens in rheumatoid, but RA is maybe the exception that probes the rule amongst the known autoimmune diseases).
Click to expand...
Right on multiplex platforms 'all the old antigen-specific pan-IgG signatures come back' post rituxumab according to Bodansky et al 2024 - I haven't actually seen ritux data in my own stuff but dara really 'does something' that I'm not used to seeing. There are also 'new autoabs' that appear post dara to fill that space in some patients. My guess is you reset a repertoire and you're getting autoabs back again after most always due to chronic viral infections, etc.

There haven't really been any proper autoab surveys yet for ME/CFS or IMHO most any disease so pointing to that negative data doesn't eliminate autoabs as a possible cause - think the data I'm looking at is among the first larger scale ones and its a really small study. 'Thousands of antigens' or 'this company's GPCR assay' isn't really that big of a survey. There is no such thing as a pan-human antibody discovery platform - nobody has all the GPCRs properly isolated and folded mostly just messy cell overexpression assays; we have 80% of canonical human SwissProt on HuProt but you can't always trust soluble domains to be folded as an autoab binds outside a membrane. Could find if we get lucky and mechanism is vs a soluble signal pathway. Will also miss it entirely if it depends on post-translational modification. If we citrullinate the array we get an entirely different autoab repertoire to those transformed targets.

To 'really check' and eliminate autoantibodies as a possible mechanistic cause for X given chronic disease - you need to survey for autoabs to all protein and non-protein structures in a human body - and multiply that by all likely posttranslationally-modified variants thereof that are occurring naturally or uniquely in the patient population. Nobody on the planet is yet capable of doing that at-scale. Sometimes you can get lucky with native proteins via epitope spreading (the anti-glialCAM / HEPACAM on this that dominates is a phosphorylated epitope but was found on a 'native' HuProt): https://pubmed.ncbi.nlm.nih.gov/35073561/

In fact - I expect a drug-like autoab to be a likely path to direct activation or pathway skew "effector mechanism... through activation of immune cells via interaction with complexes" - ie a lever thrown in the wrong direction permanently by an autoab like a checkpoint drug you're continuously living with after acquisition - expect something like this is more liklely than a 'direct killing cytotoxic autoab.' Could also be entirely T cell driven - but have a hunch that permanent weird chronic immune behavior is downstream of weird 'anti-immune autoantibodies' more often than people expect...!
 
Jonathan Edwards said:
Not quite. Polyclonall B cell activation means that whatever B cells you have are activated, maybe expanding, generating plasma cells and antibody. What we proposed in the Qeios paper is that selection of B cell clones might follow looser rules - leading perhaps to a greater diversity of antibody species and possibly more autoreactive ones. There are a whole range of shifts in regulation of B cell populations that could come under either or both terms though.
Click to expand...
Yes the autoabs I am discussing are not necessarily 'high affinity' - more 'there and very reproducibly part of that person's serum background longitudinally.' Everyone seems to have these to some degree and the impacts of this are less well established.
 
I am glad you have such well developed opinion @Tyler Hulett. As the person who did the proof of concept study for B cell depletion in autoimmunity 25 years ago I have come to a number of slightly different views!! I don't think viral infections have anything to do with autoantibody production for instance.

I think we have pretty exhaustive data indicating that specific autoantibodies are not driving ME/CFS, although the largest screens are not published yet. It remains possible but what I see mostly is commercial vested interest keeping the idea alive.
 
Tyler Hulett said:
Yes the autoabs I am discussing are not necessarily 'high affinity' - more 'there and very reproducibly part of that person's serum background longitudinally.' Everyone seems to have these to some degree and the impacts of this are less well established.
Click to expand...
If everyone have them, how can they explain disease? You’d need something other than what’s normal to get disease, because normal = healthy.

Tyler Hulett said:
Right on multiplex platforms 'all the old antigen-specific pan-IgG signatures come back' post rituxumab according to Bodansky et al 2024 - I haven't actually seen ritux data in my own stuff
Click to expand...
Then how can Ritux produces many years up to decades of remission in certain cases of autoimmune disease?
 
Utsikt said:
Then how can Ritux produces many years up to decades of remission in certain cases of autoimmune disease?
Click to expand...
Yes, this part makes little sense to me as well. If as @Tyler Hulett the assay "sees nothing" for Rituximab then presumably you're not seeing what is of interest. Because if you can get long-term remission and also long-term fall of specific antibodies linked to certain autoimmune diseases at least in a substantial subset of individuals and this isn't visible then it doesn't necessarily look like a good proxy.
 
EndME said:
Yes, this part makes little sense to me as well. If as @Tyler Hulett the assay "sees nothing" for Rituximab then presumably you're not seeing what is of interest. Because if you can get long-term remission and also long-term fall of specific antibodies linked to certain autoimmune diseases at least in a substantial subset of individuals and this isn't visible then it doesn't necessarily look like a
Click to expand...
 
@Tyler Hulett In case you are interested, here is the thread on JE et al's hypotheses paper from last year that was mentioned earlier in this discussion
Nightsong said:
Abstract:
Evidence bearing on possible mechanisms for the clinical syndrome of ME/CFS is reviewed. The evidence is used to argue for a hypothesis that centres on a form of persistent, inappropriate, ‘neuroimmune hypervigilance’ mediated primarily by T lymphocyte-macrophage interaction but influenced by IgG antibody binding to the gamma interferon-inducible high affinity immunoglobulin receptor FcγRI. This proposed mechanism could explain why the illness resembles post-infective T cell-mediated autoinflammatory syndromes in age of onset and time course but has a female preponderance similar to autoantibody-mediated disease.

Link | PDF (Qeios preprint, May 2025, open access)
Click to expand...
 
Tyler Hulett said:
It’s more the long lived autoabs there before ritux are still there after treatment - there are probably cases where full repertoire you need to remove is circulating only and thus removing those with CD20 only can be curative. It’s not clear who the culprit B cells are in which people
Click to expand...

Which is what Jo Cambridge, Vikki Abrahams and I published in 1999 before we even had the results of treatment. But none of this has anything to do with profiles of low affinity noise autoantibodies that don't cause disease anyway, as far as i can see. I am not convinced that any of the recent interventions have actually told us more than we knew in 2005.
 
You must log in or register to reply here.
Back
Top Bottom