Preprint Effect of Immunoadsorption on clinical presentation and immune alterations in COVID-19-associated ME/CFS, 2024, Anft

[Dolphin]

Effect of Immunoadsorption on clinical presentation and immune alterations in COVID-19-associated ME/CFS

Moritz Anft, Lea Wiemers, Kamil Rosiewicz, Adrian Doevelaar, Sarah Skrzypczyk, Julia Kurek, Sviatlana Kaliszczyk, Maximilian Seidel, Ulrik Stervbo, Felix S. Seibert, Timm H. Westhoff, Nina Babel

doi: https://doi.org/10.1101/2024.09.25.24314345

https://www.medrxiv.org/content/10.1101/2024.09.25.24314345v1

Abstract

Autoreactive antibodies (AAB) are currently being investigated as causative or aggravating factors during post-COVID.

In this study we analyze the effect of immunoadsorption therapy on symptom improvement and the relationship with immunological parameters in post-COVID patients exhibiting symptoms of Myalgic Encephalomyelitis/Chronic Fatigue Syndrome (ME/CFS).

This observational study includes 12 post-COVID patients exhibiting a predominance of ME/CFS symptoms alongside increased concentrations of autonomic nervous system receptors (ANSR) autoantibodies and neurological impairments.

We found that following immunoadsorption therapy, the ANSR autoantibodies were nearly eliminated from the patients' blood.

The removal of IgG antibodies was accompanied by a decrease of pro-inflammatory cytokines including IL4, IL2, IL1β, TNF and IL17A serum levels, and a significant reduction of soluble spike protein.

Notably, a strong positive correlation between pro-inflammatory cytokines and ASNR-AABs β1, β2, M3, and M4 was observed in spike protein-positive patients, whereas no such correlation was evident in spike protein-negative patients. 30 days post-immunoadsorption therapy, patients exhibited notable improvement in neuropsychological function and a substantial amelioration of hand grip strength was observed.

However, neither self-reported symptoms nor scores on ME/CSF questionnaires showed a significant improvement and a rebound of the removed proteins occurring within a month.

https://twitter.com/user/status/1840362020664709611

 

[Nightsong]

Now published in Molecular Therapy:

Link | PDF (open access)

 

[Jonathan Edwards]

I think members can see what things are wrong with this paper but just in case new members are unfamiliar with all the pitfalls and might be encouraged by it I will chip in with the usual comment that it tells us nothing.

The title should not include the word 'effect'. Some improvements are reported - although the subjects apparently didn't feel better (!!) - but there is no reason to call these effects without a controlled observation. Improvements would have been expected without the treatment.

 

[forestglip]

but there is no reason to call these effects without a controlled observation. Improvements would have been expected without the treatment.

Even without a control group, we do see that these may be real effects of the intervention, since they rebounded afterwards:

In addition to ANSR-AABs, it effectively removes the soluble spike protein, decreases pro-inflammatory cytokines and increase inflammatory cytokine TGFβ. However, these improvements are highly transient, with a resurgence of the eliminated proteins happening within a month.

Before the first IA, soluble spike protein was detected in 6 out of 12 patients, with a median concentration of 207.7 pg/mL. Following the IA-therapy, the concentration of soluble spike protein significantly dropped to 22.8 pg/mL, but rebounded to a median of 207.5 pg/mL by the 30-day follow-up examination (Figure 4A). Except for one patient, the IA-therapy also significantly reduced the concentration of SARS-CoV-2 spike protein specific antibodies, but these levels also returned to their initial values within the 30 days following IA (Figure 4B). However, we found no differences in the frequency of spike protein specific CD4+ or CD8+ T cells after the IA-therapy (Figure 4C).

I don't know if any of that is actually relevant, though. Why would IgG removal decrease spike protein in the blood, and only temporarily?

And though self-reported symptoms didn't improve, these did:

Patients exhibited significantly increased hand strength post-therapy (Figure 1C). Moreover, the overall score of the CERAD test battery was significantly higher at the one-month follow-up measurement, primarily attributed to enhancements in word learning, fluent speaking, and the Trail Making Test B (Figure 1D).

Like you said, it doesn't mean much without a control group, but I'd like to see it followed up with a controlled study.

 

[Jonathan Edwards]

Even without a control group, we do see that these may be real effects of the intervention, since they rebounded afterwards:

Of course there were real effects on antibodies because that is what the treatment can be expected to do, but that is not an effect on clinical presentation (OK it is an effect on antibodies but you can't put the two together).

It is a bit like anti-IL-6 receptor, which blocks CRP production. CRP levels are usually a good guide to how bad rheumatoid is clinically but if you give anti-IL-6R and CRP goes down you have proven nothing in terms of an effect on clinical rheumatoid. The CRP is not itself the inflammation but an after effect of inflammation. So the inflammation could be just as bad but the CRP zero.

 

[forestglip]

there were real effects on antibodies because that is what the treatment can be expected to do

Well, spike protein too. Do you know why that would happen?

Edit: From discussion about spike protein findings:

Interestingly, spike protein was also largely removed from the patients’ blood with IA-therapy, even though they were not expected to bind to the Ig-column. However, since high titers of SARS-CoV-2 spike protein specific antibodies were detected in all patients, it is likely that these antibodies bound to soluble spike protein, leading to their removal along with the antibodies during IA-therapy.

Notable, there was no improvement in self-reported symptoms after IA-therapy associated with the removal of spike proteins in patients with soluble spike protein, which is consistent with the observation that spike positive patients do not show significant correlations between spike level and ME/CSF test scores.

Whether and how soluble spike protein itself has a direct pathogenic effect or if the active virus reservoirs lead to the manifestation and exacerbation of post-COVID symptoms is not yet clear. It is discussed that soluble spike protein directly leads to the formation of fibrin and amolytic microclots, which should be prevented by removing spike protein from the body.55 Additionally, it has been shown in mice that spike protein injected in the brain can lead to neuropsychological impairments that are similar to post-COVID symptoms,56 although it remains highly unclear, whether the spike protein crosses the blood-brain barrier in humans and can be responsible for similar mechanisms.

However, since IA-therapy does not address the source of the soluble spike protein, the ongoing production of spike protein due to persistent infection may prevent a beneficial effect. This would be supported by the fact that the level of spike protein almost returned to its original level shortly after the last IA session.

It is also noteworthy that a strong correlation between the amount of all ANSR autoantibodies tested and most pro-inflammatory cytokines was only seen in the spike positive patients, which is only the case for β1 AAB in the spike negative patients. It could be assumed that a persistent infection leads to increased inflammation and generation of auto-reactive antibodies. Moreover, previous studies demonstrated a clear association between inflammation and autoimmunity.57

So likely just: "However, since high titers of SARS-CoV-2 spike protein specific antibodies were detected in all patients, it is likely that these antibodies bound to soluble spike protein, leading to their removal along with the antibodies during IA-therapy."

 

[Jonathan Edwards]

Well, spike protein too. Do you know why that would happen?

If the patient had any anti-spike protein antibody, as is likely, then it would presumably come out with the antibody. I don't think it tells us anything useful.

In other words maybe they should have 'expected'.