Distinguishing features of Long COVID identified through immune profiling, 2022, Klein, Iwasaki et al

Abstract

SARS-CoV-2 infection can result in the development of a constellation of persistent sequelae following acute disease called post-acute sequelae of COVID-19 (PASC) or Long COVID. Individuals diagnosed with Long COVID frequently report unremitting fatigue, post-exertional malaise, and a variety of cognitive and autonomic dysfunctions; however, the basic biological mechanisms responsible for these debilitating symptoms are unclear.

Here, 215 individuals were included in an exploratory, cross-sectional study to perform multi-dimensional immune phenotyping in conjunction with machine learning methods to identify key immunological features distinguishing Long COVID. Marked differences were noted in specific circulating myeloid and lymphocyte populations relative to matched control groups, as well as evidence of elevated humoral responses directed against SARS-CoV-2 among participants with Long COVID. Further, unexpected increases were observed in antibody responses directed against non-SARS-CoV-2 viral pathogens, particularly Epstein-Barr virus.

Analysis of circulating immune mediators and various hormones also revealed pronounced differences, with levels of cortisol being uniformly lower among participants with Long COVID relative to matched control groups. Integration of immune phenotyping data into unbiased machine learning models identified significant distinguishing features critical in accurate classification of Long COVID, with decreased levels of cortisol being the most significant individual predictor.

These findings will help guide additional studies into the pathobiology of Long COVID and may aid in the future development of objective biomarkers for Long COVID.

Preprint

Edit:
Preview published Sept 2023, see this post:
https://www.s4me.info/threads/disti...-klein-iwasaki-et-al.29009/page-3#post-495609

 

Twitter thread from Prof Iwasaki:

https://twitter.com/user/status/1557391752889307138

 

Decreased cortisol is very non-specific and can simply reflect different behavioural patterns of people who are ill, namely a lack of intense activity in the morning.

The increased EBV and VZV antibodies are the most interesting - perhaps reactivated herpesviruses is part of the pathogenisis of one of the LongCOVID syndromes.

 

Looking at the plots, statistically significant differences do not seem to be very important, apart from antibodies to SARS-CoV-2 S and RBD (5 to approx. 8 fold difference) and possibly to EBV gp42 when compared to recovered controls (CC).

The long Covid patients seem to be quite heterogeneous with regards to some immune parameters such as the quantities of non-conventional monocytes and IL-{4, 6}+ CD4 T cells — wouldn’t this rule out their involvement in the pathophysiology of long Covid?

I concur with @Snow Leopard that low cortisol may be an unspecific finding, possibly due to sleep patterns in long Covid patients.

 

The cortisol difference is pretty large though and they looked for an association with time of sample collection and found none. I think you'd expect more overlap if this were due to sleep patterns or morning activity levels. Could be nothing, but I wouldn't be so quick to dismiss cortisol, especially with 101 patients.

 

I've posted a couple of papers under the galectin tag. There may be more papers to post. Galectin-1 links in to autophagy, kynurenine, mTOR, cell adhesion, fibroblast growth factor receptor. Other galectins, eg -3 negatively regulate T cell function.

 

short article

https://www.straitstimes.com/world/...mone-cortisol-predicts-long-covid-in-us-study

 

Bloomberg Striking Drop in Stress Hormone Predicts Long Covid in Study

quote:

Low cortisol has been reported in people with myalgic encephalomyelitis, or chronic fatigue syndrome, and boosting it with hydrocortisone treatment has provided a modest improvement in symptoms, researchers Akiko Iwasaki, David Putrino and their co-authors wrote in the study, released ahead of peer-review and publication on Aug. 10.

 

Thought I'd create a thread for the referenced paper on hydrocortisone treatment in CFS, Low-Dose Hydrocortisone for Treatment of Chronic Fatigue Syndrome - A Randomized Controlled Trial, 1998, McKenzie et al

 

The problem is people are told they must collect it at a certain time (morning) - and they are likely to comply since this is a scientific study, but they get less sleep than usual to do so. This is not taken into account in their analysis - they didn't record sleep-wake times during the week before the test as well as on the day.

 

They collected the samples between 10am and noon, roughly, so it's not a waking cortisol test. Assuming LC patients have a latter sleep schedule on average than the controls, wouldn't taking the sample at this time actually bias results in the opposite direction (because LC patients would be used to waking up closer to the test time than controls, so would have higher cortisol at this time)?

 

That would be a not-long-after waking test for some people.

(Relatively) Low cortisol levels in the morning are associated with sleepiness due to sleep-wake cycle disruption (and snoozing in bed).

Aside from focusing on sleep patterns and not the real reason for the cortisol cycle - daily activity patterns. Acclimatisation to low intensity of activity in the hours after awakening also leads to a lower and flatter cortisol curve throughout the day.

In the bigger picture, we've seen 30 years of both low-quality (cortisol test and questionnaires) and in-depth neuroendocrine investigations in ME/CFS patients and it's lead to no progress whatsoever.

 

It should be noted that recent studies of hair cortisol levels (i.e. long term levels, as opposed to transient upon-awakening or daily levels) in CFS have found no differences compared to healthy controls, patients with so-called somatic disorders or major depression.

Hair and salivary cortisol in a cohort of women with chronic fatigue syndrome, Roerink, …, van der Meer, Knoop, Nater, 2018

Cortisol levels in chronic fatigue syndrome and atypical depression measured using hair and saliva specimens, Herane-Vives, … Chalder, Young, Cleare, 2020

Hair cortisol levels in women with medically unexplained symptoms, Fischer et al, 2022

Further, Wessely and Cleare had conflicting results with regards to urinary (24h) free cortisol:

- In a 2001 study, 61 CFS (Fukuda) patients with no psychiatric history had a mean urinary free cortisol level of 70.4 ± 38.9 nmol/24h compared to 97.0 ± 52.9 nmol/24h for controls (p < 0.0005).

62 CFS patients with no comorbid psychiatric disorders and who had been drug free for at least 2 months had mean urinary free cortisol levels of 73.4 ± 28.3 (most disabled quartile) and 74.8 ± 56.2 nmol/24h (least disabled quartile), but these subgroups were not compared to the control group.

- In a 2006 study, urinary free cortisol levels were measured for 27 CFS (Fukuda) patients without comorbid psychiatric disorders and who had been free of drugs affecting the HPA axis for at least 2 months and 26 healthy controls. The median urinary free cortisol levels were 57.5 (interquartile range 37.8 –97.3) nmol/24h in CFS and 76 (55–107) nmol/24h in controls. This was not statistically significant (p = .17) and, importantly, between the full groups of participants (40 CFS, 40 controls), neither total cortisol metabolite levels nor cortisol metabolite ratios were different.

Nonetheless, Moss-Morris et al (2012) conducted a meta-analysis of salivary cortisol levels in CFS and other fatigued populations. For the cortisol awakening response in CFS, they included two studies, Roberts et al (2004) and Nater et al (2008), with a combined total of 131 CFS patients and 145 healthy control, where participants sampled their saliva at home up to an hour after awakening. CFS patients had a somewhat blunted increase in cortisol over baseline (i.e. immediately at awakening) — a statistically significant but small effect (Cohen’s d = -0.34; no heterogeneity, I2 = 0%). No differences were found for other markers of salivary cortisol.

Of note, Roerink 2018 also found a lower cortisol awakening response in 107 CFS patients compared to 59 healthy controls (4.2 nmol/L ± 5.4 vs 6.1 nmol/L ± 6.3, p = 0.036), but total salivary cortisol output during the day did not differ.

Further, similarly to the finding that low cortisol predicted long Covid severity, Hall et al (2014) found that greater cortisol awakening response was associated with less severe post-exertional malaise, as measured by the CDC CFS inventory. If we suppose that less severe PEM is associated with shorter sleep duration, the latter study suggests that differences in cortisol awakening response may solely owe to sleep duration. Nijhof et al (2014) provide more evidence in favour of this hypothesis as they found that in adolescents with CFS, lower cortisol levels were significantly correlated with longer sleep duration (about 1 more hour on average), but not with physical activity.

Finally, Papadopoulos and Cleare (2011) argue that changes in the HPA axis in CFS are likely not the cause but rather consequences of the illness, because 1) low salivary cortisol levels were not associated with persistent fatigue at 6 months in two prospective studies of of fatigue after EBV infection and surgery and 2) CBT can purportedly increase salivary cortisol levels per Roberts et al (2009) (see the graph below), presumably because patients would sleep less and/or increase their activity. However, this study was not controlled.

Overall, there seems to be little support for causal HPA axis dysfunction in CFS. The cortisol awakening response seems to be blunted to a small extent, but it appears, on the one hand, that total daily cortisol output, whether salivary or urinary, cortisol metabolite and hair cortisol (over the longer term) levels remain unaffected, and on the other hand, that this only reflects differences in sleep duration. Given the similar findings in the long Covid study, this probably applies to it too.

 

Sort of agree but given that the range/line for the LC box comes down to 0, with some people at 0 (which would have a pretty heavy impact on a mean) I'd have to have a closer look to see whether how much this is one or a few with a very significant issue impacting the mean. I haven't gone searching for the descriptor of precisely how the box-plot or graph shape represents the data ie is the shape representative of the distribution in the middle (ie would be similar shape to a scatter) or something else if going by the dots and size of square (in which case it looks like there is more likelihood of skewing at either end too).

 

Am I right to assume the refs that don't have Fukada in brackets don't use Fukada? If so was it tighter or looser criteria for these studies?

 

All of the studies above used Fukuda, with the exception of Nater 2008 which used the Reeves 2005 (“modified” Fukuda) criteria. However, cortisol levels across these studies were consistent — Hall 2014 mention that their values were within one standard deviation of Nater 2008 and Nijhof 2014 (and also Roerink 2018).

 

Thread from David Putrino, one of the authors, mostly discussing the cortisol findings:

https://twitter.com/user/status/1558838404107976707


One interesting point he makes is that 40% of the LC cohort in this study have POTS.

 

How quickly does that happen - if someone was highly active and then gets long covid?

 

As great as this team is for doing an in-depth study, the focus on cortisol will only hurt us as docs will make one measurement at the time of day of your appointment and it will likely be in the normal range for most people with ME or LC. The authors put far too much emphasis on the cortisol levels, measured once, anytime between 8:30am and 3pm. They plotted "median" values, I suppose to avoid having the outliers affect the "mean" value. It would be better if they plotted all values and use a violin-type plot. Note the LC folks did generally have blood drawn later in the day, even tho the stats don't say so.
Too bad, of all the interesting immunological data to emphasize, this cortisol stuff is getting the headlines, and so it will be easier for docs to get rid of sick people they don't know how to treat.