Chronic inflammation, neutrophil activity, and autoreactivity splits long COVID, 2023, Woodruff et al

[EndME]

Chronic inflammation, neutrophil activity, and autoreactivity splits long COVID

Abstract

While immunologic correlates of COVID-19 have been widely reported, their associations with post-acute sequelae of COVID-19 (PASC) remain less clear. Due to the wide array of PASC presentations, understanding if specific disease features associate with discrete immune processes and therapeutic opportunities is important.

Here we profile patients in the recovery phase of COVID-19 via proteomics screening and machine learning to find signatures of ongoing antiviral B cell development, immune-mediated fibrosis, and markers of cell death in PASC patients but not in controls with uncomplicated recovery. Plasma and immune cell profiling further allow the stratification of PASC into inflammatory and non-inflammatory types. Inflammatory PASC, identifiable through a refined set of 12 blood markers, displays evidence of ongoing neutrophil activity, B cell memory alterations, and building autoreactivity more than a year post COVID-19.

Our work thus helps refine PASC categorization to aid in both therapeutic targeting and epidemiological investigation of PASC.

https://www.nature.com/articles/s41467-023-40012-7#article-info

 

[Jonathan Edwards]

I am afraid that any abstract that contains an instruction to the reader as to the value of the work described is a clear signal to stop bothering to read anything further.
Conclusions should be factual. Whether the work is useful is up to the reader.
And there are no data - so presumably there isn't anything important found.

 

[EndME]

As far as I can see there's a ton of data and it seems every bit of data is easily accessible open source.

 

[SNT Gatchaman]

A matched cohort of 26 donors with uncomplicated recoveries from COVID-19 at similar intervals post symptom onset were included as COVID-recovery (CR) controls (Table 1).

Matching was good in terms of time points, a little less well matched for initial disease severity. Doesn't look that well matched in terms of basic demographics. I also couldn't see inclusion dates for SARS2 variant discrimination, or mention of vaccination status. Perhaps that doesn't matter in terms of the analysis they are reporting on.

 

[SNT Gatchaman]

On initial read-through their findings seem to match some from Persistent serum protein signatures define an inflammatory subcategory of long COVID (2023).

That paper's abstract said —

Our analysis indicates heterogeneity in PASC and identified subsets with distinct signatures of persistent inflammation. Type II interferon signaling and canonical NF-κB signaling (particularly associated with TNF), appear to be the most differentially enriched signaling pathways, distinguishing a group of patients characterized also by a persistent neutrophil activation signature.

This paper said —

Testing for established biomarkers of neutrophil degranulation (calprotectin) and NETosis (citrullinated histone H3) revealed high levels of neutrophil activity observed exclusively in the majority of inflPASC patients tested. Altogether, these findings suggest that while available clinical tests fail to independently discriminate the inflPASC subgroup, they directly correlate with markers of neutrophil activity with strong established implications in neutrophil-based immunopathology

While elevated protein signatures were diverse in function, many of the most significant hits were inflammatory in nature and have been repeatedly identified as associates of the acute phase of severe COVID-19 including IL-6, IL-8, and NF-kB

Strong increased abundances of markers associated with cell death, including caspase 8 and the TNF death receptor, DR4, provide another potential explanation, suggesting that increased cellular debris from active cell death may be generally more abundant in these patients.

(IFNG looks to be just below threshold on the inflPASC subset volcano plot.)

 

[Hutan]

Doesn't look that well matched in terms of basic demographics.

Indeed, that's really disappointing. To not be anywhere matched on sex or age surely is a problem in proteomics?

The PCA (Figure 1a) describes relatively little of the variance (PC1 explains 21% of the variance; PC2 explains 6% of the variance), and there's no clear separation of people with Long covid and controls. I so wish Long Covid researchers would select their samples more. Only 64% reported fatigue. Why mix cohorts that were hospitalised with severe Covid-19 with people who had it only mildly? Only 59% of the patient cohort were not hospitalised.

If they can't select more tightly, it would be good to see more Principle Components Analyses done on homogeneous subsets - e.g. everyone who had mild Covid-19 and with just the people with Long Covid reporting fatigue.

 

[Hutan]

Further on the heterogeneity of the sample, they found that the Long Covid sample divided into two clusters, one of which looked like the healthy controls (on the left in the Fig 3a chart below) and one that was almost completely different (on the right). Pink is PASC, grey is healthy controls.

Fig 3b is unsupervised sorting of the PASC sample, with the division into the two groups (blue and red).


The researchers have termed these inflammatory and non-inflammatory PASC, and, if I'm assuming right, the non-inflammatory PASC looks most similar to the controls on inflammatory markers.

Of note, IL-8 and IL-1B signatures in the inflPASC cohort well-exceeded levels seen in severe/critical COVID-19 patients sampled in the acute phase of disease suggesting an inflammatory process unique to the recovery phase in this cohort.

Fig 3e and 3f show that the differences between the two subgroups aren't just related to the severity of the acute infection. The pattern shows in non-hospitalised patients.

(At this point, I note that the authors clearly have tried to dig into the data to find subgroups in the PASC sample.)

 

[Hutan]

I'm not up to much this afternoon, and have run out of steam to copy charts. But there do look to be some interesting findings. Have a look at the Figure 3 charts.

On symptom differences:

Although broad epidemiologic studies are necessary, clinical dif- ferences in disease presentation could also be observed between the niPASC and inflPASC cohorts (Supplementary Data 2). While many of the most prominent symptoms such as dyspnea and fatigue were present roughly equivalently between the groups, niPASC designation was associated with more than two-fold increased reporting of joint pain (23% vs. 6%), heart palpitations (14% vs. 6%), and anxiety (25% vs. 13%) compared to inflPASC. Muscle weakness was reported with increased frequency in inflPASC (13% vs. 33%). These differences in presentation became more prominent with time, with new dis- crepancies emerging between the groups with myalgia (17% vs. 29%) and numbness (7% vs. 13%) increased in inflPASC patients, alongside general weakness, at 90+ DPSO, thereby further confirming the per- sistence of this subset of patients well beyond the onset of disease. At more than 3 months following acute infection, almost 75% of niPASC patients reported brain fog in contrast to only 29% of inflPASC patients. Altogether, these data confirm a clinically distinct subset of PASC patients, independent of recovery time point, with differential inflammatory signaling, neutrophil activity, and clinical manifestations of disease.

 

[Hutan]

They note that the non-inflammatory PASC group still had proteomic differences compared to the controls

It is important to note that while these clinical markers such as neutrophil counts and fibrinogen are elevated within the inflPASC group, they do not necessarily reflect ‘abnormal’ test results in all cases. That is, the testing of any marker independently may not, by itself, indicate clear disorder. Instead, the elevation of multiple mar- kers, even when within “normal” ranges, seem to best reflect the broad inflammatory signals identified in the proteomics screen. This finding only emphasizes the need to develop tools capable of providing nuanced assessment across a variety of clinical parameters in patient classification. Similarly, it is important to acknowledge that niPASC is defined only as the absence of the robust inflammatory signature identified in inflPASC by comparison, and not as the absence of disease. As others have now shown20, and we show here (Fig. 2b, h),
biological associations such as EREG upregulation and cortisol levels8,20 can be readily identified across a wide spectrum of PASC even
independent of clear inflammatory signaling. It will be critical to understand how all of these signatures predict, and potentially con- tribute to, long-term patient morbidity.

Despite what they noted about some symptom differences earlier, in the discussion they comment about how symptoms aren't good at discriminating between people in the two groups they identified. They suggest that it will be important to move beyond symptomatic presentation when selecting people for therapy trials.

A surprising finding from these data, in combination with the published literature25, is the difficulty in discriminating disease subtype through symptom presentation alone. While clear trends do emerge based on subclassification, symptom presentation alone is a poor discriminator of the inflPASC and niPASC groups despite their discordant underlying biology. It is notable, however, that in chronic autoimmune disorders, differences in underlying biology can heavily impact treatment success independently of overall disease presentation37. In the case of PASC, and based on the data presented
here, two patients with highly similar symptomatic presentations might respond differently to immunomodulatory therapy. As a result, it is important to move beyond symptomatic presentation as a primary method for the classification of patients in therapeutic trial designs.

They go on to suggest that the inflammatory group might actually be an expression of a phase in the illness, an illness flare, for example, as might be the case if the cause is reactivation of a viral reservoir.