Complement dysregulation is a predictive and therapeutically amenable feature of long COVID, 2023, Morgan et al

E

EndME

Senior Member (Voting Rights)
Published abstract and link here



Preprint
Complement dysregulation is a predictive and therapeutically amenable feature of long COVID


Background: Long COVID encompasses a heterogeneous set of ongoing symptoms that affect many individuals after recovery from infection with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). The underlying biological mechanisms nonetheless remain obscure, precluding accurate diagnosis and effective intervention. Complement dysregulation is a hallmark of acute COVID-19 but has not been investigated as a potential determinant of long COVID.

Methods: We quantified a series of complement proteins, including markers of activation and regulation, in plasma samples from healthy convalescent individuals with a confirmed history of infection with SARS-CoV-2 and age/ethnicity/gender/infection/vaccine-matched patients with long COVID.

Findings: Markers of classical (C1s-C1INH complex), alternative (Ba, iC3b), and terminal pathway (C5a, TCC) activation were significantly elevated in patients with long COVID. These markers in combination had a receiver operating characteristic predictive power of 0.794. Other complement proteins and regulators were also quantitatively different between healthy convalescent individuals and patients with long COVID. Generalized linear modeling further revealed that a clinically tractable combination of just four of these markers, namely the activation fragments iC3b, TCC, Ba, and C5a, had a predictive power of 0.785.

Conclusions: These findings suggest that complement biomarkers could facilitate the diagnosis of long COVID and further suggest that currently available inhibitors of complement activation could be used to treat long COVID.

https://www.medrxiv.org/content/10.1101/2023.10.26.23297597v1.full.pdf
 
Last edited by a moderator:
This sounds interesting. It's too late in the day for me to read it today, but I look forward to hearing more e.g. sample size, how the individual complement proteins compared, whether any of the findings have also been found in ME/CFS.
 
we first quantified six markers of complement activation, including classical, lectin, alternative, and terminal pathway products, in plasma samples obtained from healthy convalescent individuals (controls, n = 79) and patients with long COVID (cases, n = 166).
Click to expand...

All participants had a clearly defined episode of acute COVID-19 confirmed via molecular evidence of infection with SARS-CoV-2. Groups were matched for age (cases, median = 47 years; controls, median = 45 years), ethnicity (cases, white = 88.0%; controls, white = 84.8%), gender (cases, female = 76.5%; controls, female = 78.5%), infection wave (cases, 54.8% infected more than 2 years before sample acquisition; controls, 46.8% infected more than 2 years before sample acquisition), and vaccination status (cases, median number of vaccinations before infection = 2; controls, median number of vaccinations before infection = 3), a parameter known to mitigate the risk of long COVID.18 Of note, obesity (BMI > 30) was significantly more common in cases versus controls (48.8% versus 34.6%, respectively; p = 0.042), and although employment status was comparable between groups pre-acute COVID-19, only 43.6% of cases remained in full-time work post-acute COVID-19 compared with 89.1% pre-acute COVID-19 (p < 0.00001).
Click to expand...
Well age matched and not too focused on older age groups. However almost 1/2 of LC patients in this study are overweight so possibly a lot of noise here.

In 46.8% of controls and 54.8% of cases, the index infection occurred >2 years prior to sample acquisition, which was limited to a time window between February and October 2022.
Click to expand...

The largest chunk of data (including basic data such as Cohort demographics, symptomatology, and other key features) is in the supplementary tables, which however aren't part of the preprint (or I can't find them) so it currently isn't possible for me to assess this work (at least they do state that the long-Covid patients weren't hospitalised).

Edit: As pointed out below the supplementary material is available, I was just too stupid to find it.
 
Last edited:
EndME said:
The largest chunk of data (including basic data such as Cohort demographics, symptomatology, and other key features) is in the supplementary tables, which however aren't part of the preprint (or I can't find them) so it currently isn't possible for me to assess this work (at least they do state that the long-Covid patients weren't hospitalised).
Click to expand...

Supplementary material here.
 
Hutan said:
whether any of the findings have also been found in ME/CFS
Click to expand...
Yea, wonder if anyone has looked at this in ME/CFS, Lyme --- anything else that looks like ME/CFS?
My usual reply - I'm guessing that a GWAS study, which had a large cohort with this problem*, would have a strong (related) signal? Even if ME/CFS doesn't have such a high proportion of people with this pathology* it might be interesting to examine the DecodeME data i.e. to see it there's a related signal?

*"Generalized linear modeling further revealed that a clinically tractable combination of just four of these markers, namely the activation fragments iC3b, TCC, Ba, and C5a, had a predictive power of 0.785."
 
Yes, good sample size, healthy controls had also had Covid-19.

Figure 1 - Dot plots for the 6 complement activation products, plasma concentrations. Overlaps, but some great P values. Long covid on the left.

Screen Shot 2023-10-30 at 7.27.21 am.png
Screen Shot 2023-10-30 at 7.27.35 am.png
The C1s-C1INH complex, a product of classical pathway activation, was significantly elevated in cases versus controls (1.25 versus 1.09 μg/ml, respectively, p = 0.0089), whereas no such differences were observed for the mannose-associated serine protease 1 (MASP1)-C1INH complex, generated during lectin pathway activation (65.0 versus 56.8 ng/ml, respectively, p = 0.15) (Figure 1A, B). The fragments iC3b and Ba, which indicate alternative pathway activation, were also significantly elevated in cases versus controls (iC3b, 20.7 versus 14.4 μg/ml, respectively, p < 0.0001; Ba, 0.39 versus 0.22 μg/ml, respectively, p < 0.001) (Figure 1 C, D). In addition, C5a and the terminal complement complex (TCC), which demarcate terminal pathway activation, were both significantly elevated in cases versus controls, with the latter demonstrating a substantial increase (C5a, 7.45 versus 5.09 ng/ml, respectively, p < 0.02; TCC, 5.56 versus 3.55 μg/ml, respectively, p < 0.0001) (Figure 1E, F).
Click to expand...

They investigated levels of complement components and regulators too, and again there was evidence for increased complement activity in cases. Some very good p-values. Two complement regulators, factor D and properdin look particularly interesting.

To extend these findings, we quantified a series of complement components and regulators in the same samples, again comparing healthy convalescent individuals (controls, n = 79) and patients with long COVID (cases, n = 166). C1q, the trigger for classical pathway activation, was significantly lower in cases versus controls (109.2 versus 130 μg/ml, respectively, p < 0.05), likely reflecting consumption (Figure 2A). In contrast, C3, C5, and C9 were all significantly elevated in cases versus controls (C3, 0.89 versus 0.83 mg/ml, respectively, p < 0.01; C5, 198.4 versus 174.5 μg/ml, respectively, p < 0.005; C9, 92.5 versus 83.2 μg/ml, respectively, p < 0.01) (Figure 2 D–F). All three of these proteins are positive acute phase reactants, likely explaining the increased concentrations in patients with long COVID. Levels of C4 and factor B (FB) were also higher in cases versus controls, albeit not significantly (Figure 2 B, C).
Click to expand...

Most of the complement regulators selected for measurement were also significantly elevated in cases versus controls, including C1INH, the key regulator of classical and lectin pathway activation (113.7 versus 100.9 μg/ml, respectively, p < 0.001) (Figure 3A), factor D (FD), factor H (FH), and properdin, which are involved in regulation of alternative pathway activation (FD, 0.92 versus 0.70 μg/ml, respectively, p < 0.0001; FH, 263.4 versus 232.3 μg/ml, respectively, p < 0.01; properdin, 3.88 versus 2.99 μg/ml, respectively, p < 0.0001) (Figure 3 B–D), and clusterin, a regulator of the terminal pathway (145.6 versus 130.6 μg/ml, respectively, p < 0.05) (Figure 3E). Plasma levels of the key alternative pathway regulator factor I (FI), the soluble form of complement receptor 1 (sCR1), and the FH-related (FHR) proteins (FHR4 and FHR125) were not significantly different between healthy convalescent individuals and patients with long COVID (Figure 3F, Supplementary Figure 1 A–C). Of note, there were also no differences in plasma haemolytic activity or anti-SARS-CoV-2 spike protein receptor-binding domain (RBD) antibody titers between healthy convalescent individuals and patients with long COVID
Click to expand...
 
Figure 5 is a correlogram. 5a is for all of the substances tested, for the controls, and 5b is for the cases. (I don't understand why these two correlograms weren't combined - what is the point of duplicating the same data above and below the diagonal line? I mean, you could put the case data above the diagonal line and the control data below the diagonal line.) It is quite clear that relationships between the various substances are very different between cases and controls. I think it's reasonable to conclude the complement system is dysregulated.
Screen Shot 2023-10-30 at 8.04.56 am.png

I was hoping that they would do a PCA, and they did.
We then conducted a principal component analysis (PCA) (Supplementary Figure 3A). Visualization of individual sample contributions to the first two principal components (PCs) revealed considerable overlap between healthy convalescent individuals and patients with long COVID (Supplementary Figure 3B). The first two principal components (PCs) explained a total of 35% of the total variance (20.2% for PC1, 14.8% for PC2) (Supplementary Figure 3C). C3 made the greatest contribution to PC1 (17.27%), followed by C9 (12.7%) and C5 (11.21%) (Supplementary Figure 3D), whereas Ba made the greatest contribution to PC2 (9.74%), followed by C1INH (9.05%) and FHR-125 (9.03%) (Supplementary Figure 3E).
Click to expand...

Screen Shot 2023-10-30 at 8.16.27 am.png
I'm not surprised that the PCA was put in the Supplementary Data, because it is very underwhelming. There is very little separation of the groups - the orange circle almost completely overlaps the green circle. It does make me feel more worried about the point that EndME mentioned, is the obesity affecting the results? I'm surprised there was no investigation of the effect of BMI - it would have been relatively easy to do. For example, the sample was big enough that they could have had a look at a sample that excluded obese individuals.
 
The mismatch on BMI could definitely be an issue:
The complement system is dysfunctional in metabolic disease: Evidences in plasma and adipose tissue from obese and insulin resistant subjects, 2019, Moreno-Navarette et al
Moreno-Navarette et al said:
Alternative pathway occurs on microbial surfaces by spontaneous hydrolysis of C3 in combination with factor B (FB), adipsin and properdin. Adipsin is identical to factor D (FD), is mainly produced by adipose tissue
Click to expand...
For example the Factor D finding could have been affected by the different prevalence of obesity in the samples.
 
Although the p values are strong, it looks from most of the dot plots like the majority of the long covid patients are in the same range of values for each test as the controls. So not a test that can diagnose Long Covid. It looks like there is a subgroup with significantly different results. I think the really need to see whether that correlates with obesity or some other factor before they conclude that it's a key factor in Long Covid.
 
https://meassociation.org.uk/2023/12/medscape-new-tests-may-finally-diagnose-long-covid/


By Sara Novak

Extracts
Researchers at Cardiff University School of Medicine in Cardiff, Wales, United Kingdom, tracked 166 patients, 79 of whom had been diagnosed with long COVID and 87 who had not. All participants had recovered from a severe bout of acute COVID-19.

In an analysis of the blood plasma of the study participants, researchers found elevated levels of certain components. Four proteins in particular — Ba, iC3b, C5a, and TCC — predicted the presence of long COVID with 78.5% accuracy.

The study revealed that long COVID was associated with inflammation of the immune system causing these complement proteins to remain dysregulated. Proteins like C3, C4, and C5 are important parts of the immune system because they recruit phagocytes, cells that attack and engulf bacteria and viruses at the site of infection to destroy pathogens like SARS-coV-2.

The more doctors understand about the mechanism causing immune dysregulation in long COVID patients, the more they can treat it with existing medications. Zelek’s lab has been studying certain medications like pegcetacoplan (C3 blocker), danicopan (anti-factor D), and iptacopan (anti-factor B) that can be used to break the body’s cycle of inflammation and reduce symptoms experienced in those with long COVID
 
Moreover, we found that plasma concentrations of several complement regulators, namely C1INH, FD, properdin, FH, and clusterin, were relatively elevated in patients with long COVID.
Click to expand...

Clusterin is a multifunctional plasma lipoprotein that inhibits assembly of the MAC. It is notable here that plasma levels of clusterin were previously found to be reduced in severe acute COVID-19.
Click to expand...
 
Now published in Med as —

Complement dysregulation is a prevalent and therapeutically amenable feature of long COVID
Kirsten Baillie; Helen E. Davies; Samuel B.K. Keat; Kristin Ladell; Kelly L. Miners; Samantha A. Jones; Ermioni Mellou; Erik J.M. Toonen; David A. Price; B. Paul Morgan; Wioleta M. Zelek

BACKGROUND
Long COVID encompasses a heterogeneous set of ongoing symptoms that affect many individuals after recovery from infection with SARS-CoV-2. The underlying biological mechanisms nonetheless remain obscure, precluding accurate diagnosis and effective intervention. Complement dysregulation is a hallmark of acute COVID-19 but has not been investigated as a potential determinant of long COVID.

METHODS
We quantified a series of complement proteins, including markers of activation and regulation, in plasma samples from healthy convalescent individuals with a confirmed history of infection with SARS-CoV-2 and age/ethnicity/sex/infection/vaccine-matched patients with long COVID.

FINDINGS
Markers of classical (C1s-C1INH complex), alternative (Ba, iC3b), and terminal pathway (C5a, TCC) activation were significantly elevated in patients with long COVID. These markers in combination had a receiver operating characteristic predictive power of 0.794. Other complement proteins and regulators were also quantitatively different between healthy convalescent individuals and patients with long COVID. Generalized linear modeling further revealed that a clinically tractable combination of just four of these markers, namely the activation fragments iC3b, TCC, Ba, and C5a, had a predictive power of 0.785.

CONCLUSIONS
These findings suggest that complement biomarkers could facilitate the diagnosis of long COVID and further suggest that currently available inhibitors of complement activation could be used to treat long COVID.

FUNDING
This work was funded by the National Institute for Health Research (COV-LT2-0041), the PolyBio Research Foundation, and the UK Dementia Research Institute.


Link | PDF (Med)
 
Levels of C4 and factor B (FB) were also higher in cases versus controls, albeit not significantly (Figures 2B and 2C).
Click to expand...
It says that in the text, but Table 2 shows higher values for controls for both these proteins. Table 2 also doesn't match the scatter plots in Figure 2, at least for C4. The table says mean of 511.8 for controls but the mean line in the plot is below 500. Table S2 in the supplementary file uses medians instead, but this agrees with the text that these proteins are elevated in long COVID. I think the data for these two might be wrong in Table 2.

Most of the complement regulators selected for measurement were also significantly elevated in cases versus controls, including C1INH, the key regulator of classical and lectin pathway activation (113.7 versus 100.9 μg/ml, respectively, p < 0.001) (Figure 3A), factor D (FD), factor H (FH)
Click to expand...
Also, the text says FH was significantly elevated, but Table 2 shows it is not significant. It isn't significant in Table S2 either, which used a different statistical test.
 
You must log in or register to reply here.
Back
Top Bottom