A Short Corticosteroid Course Reduces Symptoms and Immunological Alterations Underlying Long-COVID, 2021, Utrero-Rico et al

[ola_cohn]

Abstract

Despite the growing number of patients with persistent symptoms after acute SARS-CoV-2 infection, the pathophysiology underlying long-COVID is not yet well characterized, and there is no established therapy. We performed a deep immune profiling in nine patients with persistent symptoms (PSP), before and after a 4-day prednisone course, and five post-COVID-19 patients without persistent symptoms (NSP). PSP showed a perturbed distribution of circulating mononuclear cell populations. Symptoms in PSP were accompanied by a pro-inflammatory phenotype characterized by increased conventional dendritic cells and augmented expression of antigen presentation, co-stimulation, migration, and activation markers in monocytes. The adaptive immunity compartment in PSP showed a Th1-predominance, decreased naïve and regulatory T cells, and augmentation of the PD-1 exhaustion marker. These immune alterations reverted after the corticosteroid treatment and were maintained during the 4-month follow-up, and their normalization correlated with clinical amelioration. The current work highlights an immunopathogenic basis together with a possible role for steroids in the treatment for long-COVID.

Keywords: long-COVID; immunological alterations; corticosteroids

Free full text:
https://www.mdpi.com/2227-9059/9/11/1540/htm

 

[Hutan]

Spanish study

Very small - 5 healthy controls who had had Covid-19
Nine with Long Covid after mild infections (3 to 13 months) - they contributed a baseline blood sample.
Only 8 of the 9 were treated - 4 days of prednisone (30mg/day), all providing a post-treatment sample.
4 months later, only 6 patients provided a followup sample.

One of the 8 treated recovered completely. But that is not really unexpected, given the short length of symptoms. It is said that that patient was discharged from care.


6/7 patients reported improvement in joint pain
6/8 people reported improvement in muscle pain
4/7 reported reported improvement in asthenia (weakness/lack of energy) - I think we might call that fatigue
There was no placebo treatment, and it was open label.

The differences in the immunology findings are interesting, but it's such a small sample. (The green are the healthy controls; the orange is the 9 Long Covid patients before treatment and the blue are 8 Long Covid patients after treatment.



There's a very nice listing of all the parameters that make up each dimension in that PCA analysis (the one in figure b.) and how each parameter changed for each individual. It's probably hard to read here, but I wanted you to see it because I think it's such a clear way to present a lot of detailed information. In c. there are the contributions of each parameter to the PCA dimension. In d., in the top row is the healthy controls, treated patients and untreated patients, and each column gives the results for a single patient.





I've run out of steam for now.

 

[Hutan]

See the CD11c DC in Figure d above? It's the first row of the red and blue analysis. The patients before treatment clearly had higher levels of CD11c (lots of red), and lower levels after treatment. They looked a lot like the healthy controls for that parameter after treatment.

Edit to add - this is from Figure 2. The green dots are % of CD11c dendritic cells in the healthy controls. The orange squares are for the Long Covid people, pre-treatment. And the blue triangles are for the Long Covid people, post-treatment. That is a pretty remarkable change.



About CD11c
Cd11c is part of complement receptor 4. It's part of the β2 integrin family of adhesion molecules. Complement receptor 4 can bind fibrinogen, and also heparin. It's involved in cell adhesion to fibrinogen of monocyte-derived macrophages and monocyte-derived dendritic cells. and is involved in cellular adherence, migration and phagocytosis.

This paper looks interesting but I've just started to read it:
CD11c regulates hematopoietic stem and progenitor cells under stress, 2020, Hou

Integrins, complement, fibrinogen - they keep coming up. For example, Snow Leopard said:

I'm focusing on integrins, which are viral (and bacterial) adhesion and entry proteins, particularly in endothelial cells. Integrins primarily mediate cell-extracellular matrix (focal adhesions), but also of interest for playing a role in regulating the activity of growth factor receptors (RTKs), cellular mechanotransduction, axon regeneration, aginogenisis and are responsive to oxidative stress among other things. They are also notable for their bi-directional signalling mechanism.

 

[Hutan]

The study used flow cytometry to find the different cell types. It would be good to know more about how to judge if that has been done well or not.

This paper (which is interesting besides its comment on flow cytometry)
Non-Classical monocytes display inflammatory features: Validation in Sepsis and Systemic Lupus Erythematous, 2015, Mukherjee et al said

However, lack of consistent gating strategies for quantification by flow cytometry has been a major limitation, leading to lack of reproducibility. Since CD16 is highly expressed on neutrophils and NK cells, these cell types need to be excluded from analysis for reproducible quantification and to avoid confounders. The other major limiting factor is that all functional studies so far have been performed with purified monocytes, contributing to highly variable results depending on method of isolation employed37.

Hopefully things have moved on since 2015 in the accuracy of flow cytometry.

 

[Hutan]

What they didn't find:

On the ongoing question about natural killer cells - for this small Long Covid sample, no differences:

We did not find any differences in natural killer (NK) cell subpopulations, except for decreased terminal NK cells in PSP compared with NSP (p < 0.05) that recovered after treatment (p < 0.01, Figure S1).

On autoantibodies - nothing consistent in what they tested:

Long-COVID has been proposed to be associated with the presence of autoimmune antibodies [28,29]. In order to elucidate if autoantibodies may associate with the persistence of symptoms in PSP, we analyzed a wide group of antibodies including antiphospholipid (aPL), antineutrophil cytoplasmic (ANCA), antinuclear (ANA), and myositis and neuronal-related autoantibodies. The great majority of autoantibodies were negative in all patients (Table S3), suggesting that autoimmunity did not play a role in the persistent symptoms in these patients.

 

[Hutan]

It looks as though the prednisone did change quite a number of blood/immunology characteristics. There's the question of whether that translated into any real change in symptoms though.

The results are presented in Supplementary Table 1. It's attached here - sorry about the size. Each of a number of symptoms is classified as either Partial Recovery, Complete Recovery or No (Recovery).



The decision to use those very broad categories for subjective recovery is probably a reasonable one, given the open label nature the study. It looks as though prednisone does improve symptom burden, but only one patient completely recovered after treatment.

For example, for brain fog, 4 patients reported brain fog at baseline, and none had even a partial recovery of it after treatment. For asthenia, which is essentially fatigue, 7 patients who were subsequently treated reported it at baseline. One had a complete recovery, three had no recovery and three had a partial recovery.

There's a statement that the improvements held at 4 month followup, but I couldn't find any detail on symptom level at the followup time.

So, sadly, that's a lot less dramatic change than some of the blood parameters. Still, these are interesting preliminary results. I can see why Avi Nath of NIH would be interested in doing a large trial of corticosteroids, if not to actually fix people, then to try to understand the pathology a bit more.

 

[Creekside]

I wonder whether long covid has the same 'treatments often work great the first few times, then stop working and never work again' syndrome.

 

[SNT Gatchaman]

PSP showed a perturbed distribution of circulating mononuclear cell populations.

So, sadly, that's a lot less dramatic change than some of the blood parameters. Still, these are interesting preliminary results. I can see why Avi Nath of NIH would be interested in doing a large trial of corticosteroids, if not to actually fix people, then to try to understand the pathology a bit more.

I wonder whether long covid has the same 'treatments often work great the first few times, then stop working and never work again' syndrome.

Thank you @Hutan for all the above summaries.

If I understand the diagram of the principal component analysis, there are two dimensions, with variables at minimum or maximum in each dimension. I understand this aims to be a non-biased data clustering analysis.

Panel B (the graph) shows these two dimensions as axes and treated patients return to the normal position on one axis/dimension but not the other. I.e. only some of the variables are returning to their normal range and some remain at the polar opposite level.

Presumably this reflects the changes they are observing being one or more steps removed from what is inciting them. We need to understand what causes the perturbed distribution of monocytes, which I assume are contributing to alterations in these variables - only some of which are correcting with corticosteroids.

This sort of deep immune profiling, with heat maps, are what I hope we might see with other trials, e.g. triple anti-platelet therapy (for platelet hyperactivation).

 

[Hutan]

Panel B (the graph) shows these two dimensions as axes and treated patients return to the normal position on one axis/dimension but not the other. I.e. only some of the variables are returning to their normal range and some remain at the polar opposite level.

Yes, I think that is exactly right. So the prednisone is not fixing at least some of the parameters listed as contributing to Dimension 2.

Those parameters are listed below, roughly in order of contribution. Levels are relative to those for the healthy controls (bearing in mind the small sample size)

CD86 classical - higher before and after prednisone
CD86 interim - higher before and after prednisone
CCR5 interim - higher before and after prednisone

CM CD8 T cells - lower (and getting quite a lot lower after prednisone)
EM CD8 T cells - lower than healthy controls (and getting quite a lot lower after prednisone)

CD4 T cells - a little bit lower than healthy controls (normalising a bit after prednisone)

CS B cells - highly variable before, but becoming uniformly and markedly lower than heathy controls after prednisone

Intermediate monocytes - variable before but mostly higher after prednisone