Large scale phenotyping of long COVID inflammation reveals mechanistic subtypes of disease, 2023, Liew et al.

[SNT Gatchaman]

Large scale phenotyping of long COVID inflammation reveals mechanistic subtypes of disease

Spoiler: Multiple Authors, esp National Heart and Lung Institute, Imperial College

Felicity Liew; Claudia Efstathiou; Sara Fontanella; Matthew Richardson; Ruth Saunders; Dawid Swieboda; Jasmin K. Sidhu; Stephanie Ascough; Shona C. Moore; Noura Mohamed; Jose Nunag; Clara King; Olivia C. Leavy; Omer Elneima; Hamish J.C. McAuley; Aarti Shikotra; Amisha Singapuri; Marco Sereno; Victoria C Harris; Linzy Houchen-Wolloff; Neil J Greening; Nazir I Lone; Matthew Thorpe; A. A. Roger Thompson; Sarah L. Rowland-Jones; Annemarie B. Docherty; James D. Chalmers; Ling-Pei Ho; Alexander Horsley; Betty Raman; Krisnah Poinasamy; Michael Marks; Onn Min Kon; Luke Howard; Daniel G. Wootton; Jennifer K. Quint; Thushan I. de Silva; Antonia Ho; Christopher Chiu; Ewen M Harrison; William Greenhalf; J. Kenneth Baillie; Malcolm G. Semple; Rachael A. Evans; Louise V. Wain; Christopher Brightling; Lance Turtle; Ryan S. Thwaites; Peter J.M. Openshaw; ISARIC4C Investigators and the PHOSP-COVID collaborative group

One in ten SARS-CoV-2 infections result in prolonged symptoms termed ‘long COVID’, yet disease phenotypes and mechanisms are poorly understood. We studied the blood proteome of 719 adults, grouped by long COVID symptoms.

Elevated markers of monocytic inflammation and complement activation were associated with increased likelihood of all symptoms. Elevated IL1R2, MATN2 and COLEC12 associated with cardiorespiratory symptoms, fatigue, and anxiety/depression, while elevated MATN2 and DPP10 associated with gastrointestinal (GI) symptoms, and elevated C1QA was associated with cognitive impairment (the proteome of those with cognitive impairment and GI symptoms being most distinct).

Markers of neuroinflammation distinguished cognitive impairment whilst elevated SCG3, indicative of brain-gut axis disturbance, distinguished those with GI symptoms. Women had a higher incidence of long COVID and higher inflammatory markers. Symptoms did not associate with respiratory inflammation or persistent virus in sputum. Thus, persistent inflammation is evident in long COVID, distinct profiles being associated with specific symptoms.

Link | PDF (Preprint: MedRxiv)

Link to published article here

 

[SNT Gatchaman]

A few quotes —

In this prospective multicentre study, we measured 360 plasma proteins in 719 adults six months after hospitalisation for COVID-19, confirmed either clinically or by PCR (n=621). Within our cohort, including the 626 patients previously analysed, 250/719 (35%) reported full recovery (“Recovered”), the remaining 469 (65%) reported persistent symptoms consistent with long Covid.

Mediators suggestive of persistent monocytic inflammation were associated with all symptoms. Specifically, elevated IL1R2 and/or Matrilin-2 (MATN2) were consistently associated with the highest odds of all symptoms except cognitive impairment (cardiorespiratory IL1R2 OR=1.16; fatigue IL1R2 OR= 1.53; anxiety/depression IL1R2 OR=1.13; GI MATN2 OR=1.08).

IL1R2 is expressed by monocytes and macrophages and modulates IL1-driven inflammation. MATN2 is an extracellular matrix (ECM) protein which promotes inflammation by activating toll-like receptors and enhancing monocyte infiltration into tissues. Although the effect was smaller, CSF3 (G-CSF, which promotes inflammation by stimulating neutrophil production), was associated with GI symptoms (OR=1.05), fatigue (OR=1.12) and anxiety/depression (OR=1.05). IL-6 was also associated with increased odds of cardiorespiratory symptoms (OR=1.06) and fatigue (OR=1.09).

Although these findings are indicative of systemic inflammation, C-reactive protein levels, a commonly used clinical marker of inflammation, were not significantly different between long COVID and recovered groups.

Elevated sCD58 (lymphocyte-function antigen, an immunosuppressive factor), was associated with reduced odds of all long COVID symptoms and this was most pronounced for fatigue (OR=0.71) and cardiorespiratory symptoms (OR=0.79).

C1QA is a component of the complement system, initiating inflammation via the classical pathway. Notably, C1QA was associated with the third highest odds of cognitive impairment

Consistent with previous reports, we found women were more likely to experience long COVID symptoms. Since oestrogen is known to influence immunological responses, we compared protein levels between men and women younger and older than 50 years. Oestrogen-dependent differences would be expected to be most pronounced in pre-menopausal women, but we found no evidence of this.

Several inflammatory markers were significantly elevated in post-COVID adults; in addition to TNFα, IL8 and IL-1α, several epithelial growth factors were elevated including VEGF, TGFα, GDF and HGF.

 

[DokaGirl]

It would be amazing if this depth of research was applied to ME.

But, wait a minute, in-depth research has been done by the NIH, but we're still waiting to see results. It's been almost 7 years so far - the study started in late 2016. The first publication is still many months to one year away, according to journalist Brian Vastag.

See this thread:
USA: National Institutes of Health (NIH) intramural ME/CFS study

 

[Hutan]

In this prospective multicentre study, we measured 360 plasma proteins in 719 adults six months after hospitalisation for COVID-19, confirmed either clinically or by PCR (n=621).

Note that all of the participants had been hospitalised with COVID-19. Measurement 6 months after the infection.

Pre-existing conditions that might predispose to adverse outcomes (e.g., chronic lung disease and prior cardiorespiratory symptoms) were a risk factor for all symptoms (except GI); age and acute disease severity were not associated with any symptom.

So, those findings and others in the paper need to be interpreted in the light of the whole sample having had a severe enough infection to have been hospitalised.

 

[Hutan]

fatigue IL1R2 OR= 1.53

That's a pretty solid odds ratio - so elevated IL1R2 at six months was a predictor of fatigue at six months.

Note that IL1R2 isn't the same as IL-1. It's a receptor of IL-1, so it binds IL-1. It apparently inactivates IL-1, unlike a competing receptor IL1R1 which generates a signal.

From Wikipedia:

Interleukin 1 receptor, type II (IL-1R2) also known as CD121b (Cluster of Differentiation 121b) is an interleukin receptor. IL1R2 also denotes its human gene.

The protein encoded by this gene is a decoy receptor for certain cytokines that belongs to the interleukin-1 receptor family. This protein binds interleukin-1α (IL1A), interleukin-1β (IL1B), and interleukin 1 receptor antagonist (IL1Ra), preventing them from binding to their regular receptorsand thereby inhibiting the transduction of their signaling. IL-1R2 protein also interacts non-productively with the second component of the signaling IL-1 receptor, namely IL-1RAcP,[5] and a complex of the IL-1R2 and IL-1RAcP extracellular domains with interleukin-1 beta has been solved by X-ray crystallography.[6] Interleukin 4 (IL4) is reported to antagonize the activity of interleukin 1 by inducing the expression and release of this cytokine. This gene and three other genes form a cytokine receptor gene cluster on chromosome 2q12. Two alternatively spliced transcript variants encoding the same protein have been reported.[7]

From IL-1 receptor 2 (IL-1R2) and its role in immune regulation (2013):

In contrast, much less is known about the type 2 IL-1 receptor (IL-1R2), which acts as a decoy receptor for IL-1. While IL-1R2 is structurally similar to the type 1 IL-1 receptor (IL-1R1) responsible for IL-1 signal transduction, its truncated cytoplasmic domain and lack of Toll-IL-1 receptor (TIR) region renders IL-1R2 incapable of transmembrane signaling. IL-1R2 competes with IL-1R1 for ligands and for the IL-1R1 co-receptor, IL-1 receptor accessory protein (IL-1RAP). Additionally, IL-1R2 exists in both a membrane bound and soluble form (sIL-1R2) that has biological properties similar to both a decoy receptor and a binding protein. Thus far, IL-1R2 has been implicated in arthritis, endometriosis, organ transplantation, sepsis/sickness behavior, diabetes, atherosclerosis, autoimmune inner ear disease (AIED), Alzheimer’s disease and ulcerative colitis.

That paper has a good diagram that helps to understand how IL-1 and various receptors are thought to work.

 

[Jonathan Edwards]

I think we need to keep in mind that these are people in whom no actual inflammation was found (as far as I can see) and the CRP, a good marker of macrophage-driven inflammation (monocyte is not the right term here), was normal. Also I don't see any big difference reported from controls? Just correlations of some things with some symptoms. The correlation of IL1R2 with fatigue at 1.5 is reasonably substantial but the others look minimal.

I don't see this revealing 'mechanistic subtypes' to be honest. I can think of a lot of confounders in the analysis.

 

[Jonathan Edwards]

It is good to see people at leat trying to measure these things. I guess I wish they would be more realistic about interpretation.

I have not had time to read the whole paper. I am wondering if we can be sure that the people with the symptoms and raised markers have symptoms because of Covid or maybe they are ill for other reasons and these markers tend to go with those symptoms. It would be easy to discount that if the picture was black and white but statistical relations can so often turn up for other reasons.

 

[Hutan]

That 2013 paper is given as a reference in this 2023 Long Covid paper, for this statement in the 2023 paper:

IL1R2 is expressed by monocytes and macrophages and modulates IL1-driven inflammation.17

Blocking the IL-1/IL-1R1 interaction in neutrophils and monocytes does prevent IL-1-induced production of IL-6, IL-8 and TNF-α

The 2013 paper tells us that IL-1R2 is found on cell membranes and free in the plasma (where it's called sIL-1R2). It's the plasma sIL-1R2 that is reported in this study.

sIL-1R2, on the other hand, is ubiquitously present in the plasma of healthy individuals (Giri et al., 1994) at a relatively high concentration when compared to serum IL-1β and IL-1RA. In healthy women, serum concentrations of sIL-1R2 are nearly two-fold higher than IL-1RA (570.5 ± 79.1 pg/mL vs 11,328.9 ± 384.9 pg/mL) (Chun et al., 2012) while serum IL-1β is essentially undetectable in healthy human subjects (0.3 ± 0.5 pg/mL)

Even though the 1.5 of 'elevated IL-1R2' is a decent odds ratio and the difference is statistically significant, the scatter chart isn't terribly convincing. There's an awful lot of people with fatigue with perfectly normal sIL-1R2. I guess it's possible that the people with ME/CFS-like Long Covid are the ones with high IL-1R2, but it's also possible that it is those with a co-morbidity of some sort, such as heart disease.

 

[Jonathan Edwards]

That 2013 paper is given as a reference in this 2023 Long Covid paper, for this statement in the 2023 paper:

Yes. My grouse is that you need to distinguish between monocytes in blood that might by drawn out into tissues full of cytokines from all sorts of sources, like neutrophils, and inflammation full of macrophages making cytokines that call in more cells and therefore drive the inflammation.

Monocytic inflammation is an oxymoron because monocytes are what we call cells that have not yet reached a site of inflammation. It is a technical distinction but people do not make the mistake if they understand the compartmental aspects of inflammation - the same problem as calling it inflammation when there are cytokines in blood but no tissue changes.

 

[Trish]

Some posts discussing possible publication timing for the NIH ME/CFS study have been moved to: USA: National Institutes of Health (NIH) intramural ME/CFS study

 

[John Mac]

Now published
https://www.nature.com/articles/s41590-024-01778-0


https://www.bbc.co.uk/news/health-68762171

 

[John Mac]

Science Media Centre

expert reaction to study looking at long Covid subtypes and inflammatory signs in people with long Covid after Covid hospitalisation

A study published in Nature Immunology looks at long Covid subtypes and inflammatory signs.

Prof Eleanor Riley, Professor Emerita, Immunology and Infectious Disease, University of Edinburgh.
Dr Emily Fraser, Consultant in Respiratory Medicine, Oxford University Hospitals NHS Foundation Trust, and clinical lead of the Post COVID clinic in Oxford.
Prof Kevin McConway, Emeritus Professor of Applied Statistics, Open University.

https://www.sciencemediacentre.org/...-with-long-covid-after-covid-hospitalisation/

 

[Lindberg]

Is the article comparable with this one in any way? Or are they unrelated regarding their findings?

https://www.medrxiv.org/content/medrxiv/early/2024/02/13/2024.02.11.24302636.full.pdf

 

[Simon M]

Science Media Centre
expert reaction to study looking at long Covid subtypes and inflammatory signs in people with long Covid after Covid hospitalisation
A study published in Nature Immunology looks at long Covid subtypes and inflammatory signs.

Prof Eleanor Riley, Professor Emerita, Immunology and Infectious Disease, University of Edinburgh.
Dr Emily Fraser, Consultant in Respiratory Medicine, Oxford University Hospitals NHS Foundation Trust, and clinical lead of the Post COVID clinic in Oxford.
Prof Kevin McConway, Emeritus Professor of Applied Statistics, Open University.SMC

I'm not up pon LC research, but I wonder about the relevance of this to ME, or even most LC.
It only looks at people hospitalised with Covid, and effectively operationalises LC as anyone with ANY LC symptoms at 3 months, which is pretty broad. The vast majority of people with LC were not hosptitalised.

Fig 1b maps the cohort by symptoms:

E.g. the largest group have fatigue and cardio-respiratory symtpoms but not e.g. cognitive. Could this be linked to lung damage from a severe respiratory infection?

Almost all ME folk have cognitive symptoms yet these make up the smallest group in this study (<100), and about 20% of these have no fatigue. I'm sure it's a very useful study looking at why people don't recover after hospitalised infections, but there is quite a lot of broader speculation in the media.

[QUOTE Fig 1 legend ]b, An UpSet plot describing pooled LC groups. The horizontal colored bars represent the number of patients in each symptom group: cardiorespiratory (Cardio_Resp), fatigue, cognitive, GI and anxiety/depression (Anx_Dep). Vertical black bars represent the number of patients in each symptom combination group. To prevent patient identification, where less than five patients belong to a combination group, this has been represented as ‘<5’. The recovered group (n = 233) were used as controls. .[/QUOTE]

That said, the results for the 'cognitive' symptom group showed the biggest effect, for neurofascin (shown below, Fig 1h, the abstract mentions C1QA):

Would be interesting to see results for this/these molecules in LC and ME.

 

[SNT Gatchaman]

Commentary: Immune dysregulation in long COVID (2024, Nature Immunology, paywall). Concludes —

Collectively, this study supports the hypothesis of a dysregulated immune response that causes long COVID with its different manifestations. Persistent dysregulation and activation of the complement system leading to thrombo-inflammation and endothelial cell damage in long COVID has been reported. In addition to dysregulated levels of soluble complement factors in the blood, a previous study found an altered monocyte phenotype characterized by platelet monocyte aggregation 4. CD41, a platelet integrin usually absent from monocytes, was detected on the surface of monocytes in individuals with active long COVID 12 months after acute infection 4, which suggests an altered coagulation cascade, which in turn can activate the complement system.

This link between a prothrombotic environment, infiltration of myeloid cells and complement dysregulation is also supported by the findings of Liew et al. 3, who identified additional factors related to monocyte activation (MATN2 and IL-1R2), endothelial cell damage (COLEC12), and complement activation (C1QA). Moreover, the authors 3 showed these findings were shared across individuals with long COVID with different symptoms, suggesting a common underlying disease mechanism that could provide an opportunity for therapeutic intervention.

Notably, several drugs that target myeloid and vascular inflammation, including IL-1 antagonists (NCT05926505), are currently being explored in clinical trials for long COVID. Moreover, these findings 3,4 support therapeutic investigation of the complement system and its regulation.

(3) This paper
(4) Persistent complement dysregulation with signs of thromboinflammation in active Long Covid (2024, Science)

 

[SNT Gatchaman]

a previous study found an altered monocyte phenotype characterized by platelet monocyte aggregation 4. CD41, a platelet integrin usually absent from monocytes, was detected on the surface of monocytes in individuals with active long COVID 12 months after acute infection 4, which suggests an altered coagulation cascade, which in turn can activate the complement system.

CD41a was also noted on extracellular vesicles in Comparative Analysis of Extracellular Vesicles in Patients with Severe and Mild Myalgic Encephalomyelitis/Chronic Fatigue Syndrome (2022, Frontiers in Immunology) —

EVs bearing markers from a variety of cellular lineages were examined [...] Of the platelet markers CD41a and CD62P, CD41a was also higher in the severe ME/CFS participants compared to the healthy control participants (p=0.015).

Significant higher levels of the B cell marker CD19 and platelet marker CD41a were seen on EVs from the severe compared to mild ME/CFS cases (P < 0.05). Platelet derived EVs are the most abundant EV population found circulating in humans (18). Interestingly, these EVs have been shown by some groups to possess anti-inflammatory properties when taken up by myeloid or lymphoid cells, including induction of TGF-b secretion by myeloid cells, and are associated with low NK cell function (17, 18).

 

[Liie]

Video by James Gills about this article: