Increased expression of activation antigens on CD8+ T lymphocytes in Myalgic Encephalomyelitis/chronic fatigue syndrome, Maes et al, 2015

OrganicChilli

OrganicChilli

Senior Member (Voting Rights)
Abstract

Background: There is now evidence that specific subgroups of patients with Myalgic Encephalomyelitis / chronic fatigue syndrome (ME/CFS) suffer from a neuro-psychiatric-immune disorder. This study was carried out to delineate the expression of the activation markers CD38 and human leukocyte antigen (HLA) DR on CD4+ and CD8+ peripheral blood lymphocytes in ME/CFS.

Methods: Proportions and absolute numbers of peripheral lymphocytes expressing CD3+, CD19+, CD4+, CD8+, CD38+ and HLA-DR+ were measured in ME/CFS (n=139), chronic fatigue (CF, n=65) and normal controls (n=40).

Results: The proportions of CD3+, CD8+, CD8+CD38+ and CD8+HLA-DR+ were significantly higher in ME/CFS patients than controls, while CD38+, CD8+CD38+, CD8+HLA-DR+ and CD38+HLA-DR+ were significantly higher in ME/CFS than CF. The percentage of CD19+ cells and the CD4+/CD8+ ratio were significantly lower in ME/CFS and CF than in controls. There were highly significant inverse correlations between the increased expression of CD38+, especially that of CD8+CD38+, and the lowered CD4+/CD8+ ratio and CD19+ expression. There were no significant associations between the flow cytometric results and severity or duration of illness and peripheral blood biomarkers of oxidative and nitrosative stress (O&NS, i.e. IgM responses to O&N modified epitopes), leaky gut (IgM or IgA responses to LPS of gut commensal bacteria), cytokines (interleukin-1, tumor necrosis factor-α), neopterin, lysozyme and autoimmune responses to serotonin.

Conclusions: The results support that a) increased CD38 and HLA-DR expression on CD8+ T cells are biomarkers of ME/CFS; b) increased CD38 antigen expression may contribute to suppression of the CD4+/CD8+ ratio and CD19+ expression; c) there are different immune subgroups of ME/CFS patients, e.g. increased CD8+ activation marker expression versus inflammation or O&NS processes; and d) viral infections or reactivation may play a role in a some ME/CFS patients.

Link

CD38 mentioned
 
What is causing more CD38 expression?

More importantly why did the authors know out of all the CDs which ones to test for? It doesn’t say any rationale for why they chose these specific CDs.

Wouldn’t it make more sense to measure all the CDs rather than choose some arbitrary subset and they didn’t explain why they chose these markers?
 
ryanc97 said:
More importantly why did the authors know out of all the CDs which ones to test for? It doesn’t say any rationale for why they chose these specific CDs.

Wouldn’t it make more sense to measure all the CDs rather than choose some arbitrary subset and they didn’t explain why they chose these markers?
Click to expand...
it’s a common marker for T cells in an “activated” state. All the markers they chose here are known functional markers for a certain T cell phenotype, they weren’t random.

ryanc97 said:
What is causing more CD38 expression?
Click to expand...
That’s the golden question, isn’t it. On a hunch I checked an interferon database since we were discussing CD38’s role in calcium signaling and downstream interferon response in another thread. It looks like CD38 is stimulated in response to interferon beta, which is nice since it sets up another potential feedback loop.

I can’t seem to link to the specific result but if anyone wants to confirm, you can search “Cd38” and check off all the T cell subsets under “cell” on this website https://interferome.org/interferome/search/searchGene.jspx

It’s not the only thing that stimulates CD38 though, just one possible candidate I was already interested in.
 
I'll note this seems to be just an abstract as far as I can tell? I can't find a full manuscript anywhere. Until the methodological details are known, probably best to take these results with a hefty grain of salt.
 
Utsikt said:
I think I found it
Click to expand...
Thanks!

Things that stood out to me from a quick skim:
Fukuda criteria
A potential founder is starting with very different magnitudes of cell counts between groups, but it looks like they got pretty similar leukocyte counts and proportions of lymphocytes across groups.
Huge difference in sex ratios between groups.

I'll be honest I'm having a hard time following the logic for the way they structured this statistical analysis. The tests that I would be interested in are the differences in proportions of these subsets corrected for age and sex. It looks like they separately tested association with age and sex, and they were strongly significant. Then they mention using a GLM for further analysis but I don't see mention of those being included as covariates, which would be the logical thing to do if you identified those as significant confounders. It also looks like they didn't include those as variables in the stepwise analysis.

Maybe I'm missing something? If I am interpreting it correctly, these results just show slight variations in cell subset frequency driven by sex ratios.
 
Last paragraph of the discussion (breaks added by me):
Maes et al. said:
It is interesting to note that age and gender significantly modified our flow cytometry results. Thus, females showed higher proportions of CD3+, CD4+, CD38+ and CD4+CD38+ T cells than men. Age was significantly and positively related to the CD4+ proportion and the CD4+/CD8+ T cell ratio, and negatively to CD8+, CD38+ and CD8+CD38+ T cell proportions.

These findings extent the results of Jiao et al. (2009) who found that CD8+CD38+ T cells were inversely related to age. However, in contrast to these authors we were unable to find significant inverse relationships between age and the CD3+, CD19+ and CD8+HLD-DR+ levels.

These results show that all results of flow cytometry should always be controlled for possible effects of age and sex.
Click to expand...

If someone has the energy and feels up to reading some statistics jargon, I'd greatly appreciate if they could verify that sex and age were not corrected for in any of the ME/CFS findings that were reported. This paragraph just baffles me if they included it after not correcting for age and sex in their own analysis.
 
jnmaciuch said:
Then they mention using a GLM for further analysis but I don't see mention of those being included as covariates
Click to expand...

The RESULTS page says gender and age (and interaction diagnosis * gender) were GLM explanatory:
Table 2 shows the results of a multivariate GLM
analysis with the 12 flow cytometric measurements asR
dependent variables and diagnosis (i.e. entered as three
groups that is controls, CF and ME/CFS), gender, the
interaction diagnosis X gender, and age as explana-
tory variables.
Click to expand...
 
arnoble said:
The RESULTS page says gender and age (and interaction diagnosis * gender) were GLM explanatory:
Click to expand...
That's what's confusing me because I initially interpreted that part to mean they were included as covariates (since that's the next logical step and I was reading in good faith), but looking closer at table 2 made me realize that what they actually did was a series of separate multivariate GLMs:

proportions of all measurements ~ age
proportions of all measurements ~ sex
proportions of all measurements ~ diagnosis
proportions of all measurements ~ diagnosis*sex

And then based on which ones were significant (all except diagnosis*sex) they did associations with individual measurements like:

feature1 ~ age
feature1 ~ sex
feature1 ~ diagnosis

feature2 ~ age
feature2 ~ sex
feature2 ~ diagnosis

etc.

So they never actually did the associations with ME/CFS accounting for sex, which is featureX ~ diagnosis + sex (+ age bracket, if the sample size allows). Age and sex were "explanatory" variables insofar as they ran separate associations with age and sex, I don't think they were ever included as covariates. They might not have done it because they just didn't have enough males in each diagnosis group to get appropriate power with the extra covariate--but if that was the case, they should have at least downsampled the ME/CFS cohort.
 
Last edited:
You must log in or register to reply here.
Back
Top Bottom