Multiple Early Factors Anticipate Post-Acute COVID-19 Sequelae, 2022, Su et al

[Wyva]

Highlights

• Longitudinal multiomics associate PASC with autoantibodies, viremia and comorbidities
• Reactivation of latent viruses during initial infection may contribute to PASC
• Subclinical autoantibodies negatively correlate with anti-SARS-CoV-2 antibodies
• Gastrointestinal PASC uniquely present with post-acute expansion of cytotoxic T cells

SUMMARY

Post-acute sequelae of COVID-19 (PASC) represent an emerging global crisis. However, quantifiable risk-factors for PASC and their biological associations are poorly resolved. We executed a deep multi-omic, longitudinal investigation of 309 COVID-19 patients from initial diagnosis to convalescence (2-3 months later), integrated with clinical data, and patient-reported symptoms.

We resolved four PASC-anticipating risk factors at the time of initial COVID-19 diagnosis: type 2 diabetes, SARS-CoV-2 RNAemia, Epstein-Barr virus viremia, and specific autoantibodies. In patients with gastrointestinal PASC, SARS-CoV-2-specific and CMV-specific CD8+ T cells exhibited unique dynamics during recovery from COVID-19. Analysis of symptom-associated immunological signatures revealed coordinated immunity polarization into four endotypes exhibiting divergent acute severity and PASC.

We find that immunological associations between PASC factors diminish over time leading to distinct convalescent immune states. Detectability of most PASC factors at COVID-19 diagnosis emphasizes the importance of early disease measurements for understanding emergent chronic conditions and suggests PASC treatment strategies.

Open access (in PDF): https://www.cell.com/cell/fulltext/S0092-8674(22)00072-1

 

[Wyva]

This study didn't only involve people with mild covid infections but many were hospitalized. There is a New York Times article where some experts comment on it, including Avindra Nath: https://www.nytimes.com/2022/01/25/health/long-covid-risk-factors.html

But the thing is that it mentions Epstein-Barr reactivation as one of the risk factors:

That some patients had reactivated Epstein-Barr virus also made sense, Dr. Nath said, because other diseases have reawakened that virus, and its reactivation has been linked to conditions like chronic fatigue syndrome, which some cases of long Covid resemble, and multiple sclerosis. Dr. Deeks said it might be possible to give antivirals or immunotherapy to patients with reactivated Epstein-Barr virus.​

 

[Mithriel]

I would be more impressed by the work being done on EBV if they hadn't been looking at it for the last 30 years yet getting nowhere. I'd love to be wrong.

 

[rvallee]

This paper seems to be the cover issue of the March edition of Cell journal:

 

[John Mac]

Merged thread

Study Looks for Long COVID Risk Factors Su et al 2022

https://covid19.nih.gov/news-and-stories/study-looks-risk-factors-long-covid

The researchers enrolled 209 people ages 18 to 89 who had laboratory-confirmed SARS-CoV-2 infections. The participants’ COVID-19 experiences ranged from having mild symptoms and never having to be hospitalized to needing mechanical ventilation in the intensive care unit.

With the participants’ consent, researchers studied their electronic health records, interviewed them about their symptoms, and took blood samples. All participants were asked to come back 60 and 90 days after their initial COVID-19 symptoms started.

The researchers compared the 209 patients with people who had not had COVID-19 and checked their findings against a separate group of 100 people who had COVID-19 and were 60 to 90 days beyond developing their initial symptoms.

Three months after being diagnosed with COVID-19, half of the 209 participants reported fatigue. A quarter of them had a cough, and 18% had loss of smell or taste.

The researchers found that people were more likely to have symptoms 2 to 3 months after diagnosis if they had any of several risk factors at the time they were diagnosed, including:

• Type 2 diabetes
  
  
• Reactivated Epstein-Barr virus in their blood. Many people are infected with this virus in childhood. After infection, the virus persists in the body in an inactive form but may reactivate.
  
  
• Autoantibodies. While antibodies should bind only to materials from outside the body, some people make antibodies against their own tissues. The researchers checked for a few different autoantibodies.

The researchers also made other observations. For example, people who had cold-like symptoms at 3 months also had low levels of the hormone cortisol.

These risk factors are only part of the picture, the researchers noted — a person's genetics likely also play a factor in their Long COVID risk.

 

[LarsSG]

It looks like a significant part of the study was only done on the cohort that was 70% hospitalized and 30% ICU. It's not clear which parts included which cohorts (the EBV part, for instance, was just in the first cohort). The second cohort was only 10% hospitalized and reported significantly fewer LC symptoms.

 

[Hutan]

This paper has been cited as providing evidence of low cortisol in people with Long Covid. I had a closer look. Summary: just go to the last chart. The data don't support the idea of low cortisol in people with Long Covid.


This paper had several samples, but the primary one and the one that the cortisol finding seems to relate to was of 209 people - 71% had been admitted to hospital, 30% had been admitted to ICU, 18% had intubation and mechanical ventilation. There is T3 metabolomics data for only 114 people out of the 209 people with Long Covid.

This study reported a low level of cortisol in a subset of people who showed 'respiratory viral' symptoms. The paper says

Symptoms were also grouped as follows: respiratory viral (cough, fatigue, shortness of breath, fever or chills, muscle/body aches, nausea), gastrointestinal (diarrhea, abdominal pain), neurologic (anxiety, blurred vision, depression, memory problems, difficulty concentrating, difficulty sleeping, dizziness, headache), and anosmia/dysgeusia (loss of taste, loss of smell).

so 'respiratory viral' may include fatigue - 2 or more of the listed symptoms had to be present for someone to be in the 'respiratory viral' group.
People with 'respiratory viral' symptoms are reported to account for 42% of the sample. However, if you go to S1.3 PASC data and do a count, there are 42 participants out of 126 participants, so 33%. If you count how many people reported fatigue, there are 56, 44%.

Here is the chart, (middle chart, z scores, relative to healthy) at Time 3 (2-3 months). Note the cortisol finding only relates to people with 'respiratory viral' symptoms ( shown in orange in that middle chart). People who didn't have the 'respiratory viral' symptoms (in blue in the middle chart), looked just like the healthy controls. Even for the people with respiratory viral symptoms, there was quite a lot of overlap with the healthy controls.

Figure 1E



The authors note:

Suppression of endogenous cortisol production could be caused by steroid treatments, as certain steroids are structurally similar to cortisol and may cause feed-back inhibition of cortisol production

This is an interesting possible reason for some people with the catch-all Long Covid label to be presenting with lower cortisol. The study separated out the results according to whether people had been treated with steroids or not:
Figure S1D

The first three charts are for cortisol, at Times 1, 2 and 3.(Clinical diagnosis; about 15 days after onset; 2-3 months after onset). These are z scores relative to the healthy population. Some people given steroid treatment had lower cortisol during the acute illness but by 2-3 months after onset, levels had normalised. So, that doesn't seem to be the answer. But, note here that no group seems to have low cortisol results at T3 - not healthy people, not people given steroid treatment and not people not given steroid treatment.

The authors then looked at cortisol levels by the number of symptom groups that people reported. Look at the middle chart.

Figure S1E

I don't know why all of the means of the samples are above 0 on the z scores - I don't know what population could be providing the baseline. For other charts the baseline population is reported, but not for this one. Those with 4 or more symptom groups had lower cortisol levels than the other people with no persistent symptoms and the healthy controls. ( the text says 3 or more symptom groups, but that isn't correct). The numbers of people in that "4 or more symptom groups" may be quite small.

It's a bit hard to make sense of, so I went to Table S2.2 which has the metabolomics data for each participant and plotted the cortisol data for T3, for the healthy sample and the Long Covid sample. This is the primary sample, although just 114 people had T3 data.

Plotting the S2.2. data

It looks to me as though there is no evidence that cortisol scores in the Long Covid sample are significantly lower than those of the healthy sample. Sure there are a few outliers, but no more than that. Both distributions look pretty symmetrical around zero. Quite a few of the Long Covid sample were intubated and ventilated, and may be in quite a bad way even at the 3 month mark. It's also possible that there has been some data error to produce those low outliers.


On other things:
People with fatigue at T3 accounted for 44% of the Long Covid sample.
People with anxiety at T3 accounted for 3% of the Long Covid sample
People with depression at T3 accounted for 2% of the Long Covid sample

On the latter two, clearly is depends how you ask the questions, although the paper doesn't say how they determined them.