Pilot assessment of low NK cell-mediated ADCC and FCGR3A genetics in(ME/CFS):.. low ADCC unsuitable as diagnostic biomarker: Sung, Bateman et al 2019

[Sly Saint]

Pilot assessment of low NK cell-mediated ADCC and FCGR3A genetics in myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS): Based on inclusion of family members without ME/CFS as controls, low ADCC is unsuitable as a diagnostic biomarker

Alexander P. Sung, Jennifer J-J Tang, Michael J Guglielmo, Julie Smith-Gagen, Lucinda Bateman, Doug D. Redelman, Dorothy Hudig

Preprint, not peer reviewed.

Abstract
ADCC (antibody-dependent cell-mediated cytotoxicity) is dependent on the varying capacity of NK cells to kill, the affinities of FCGR3A-encoded CD16A receptors for antibody, and the presence of antigen-specific antibodies. In vivo ADCC depends on the number of CD16A receptor-positive NK cells in blood. We hypothesized that low ADCC cell function or low effector cell numbers could be biomarkers or risk factors for myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS).

We measured NK cell ADCC lytic capacity and antibody recognition, CD16Apositive NK cells/ul blood, and FCGR3A homozygosity for the F allele that encodes low affinity CD16A antibody receptors. ME/CFS patients met the Fukuda 1994 diagnostic criteria. In this pilot report, we examined 5 families, each with 2 to 5 ME/CFS patients, and compared 11 patients, 22 family members without ME/CFS, and 16 unrelated healthy controls. ADCC was measured as CX1:1 cytotoxic capacity (the percentage of 51Cr-Daudi tumors with obinutuzumab anti-CD20 antibody that were killed at a 1:1 ratio of CD16Apos NKs to Daudis) and CX-slope. Individual CX1:1 capacities varied from 16.2% to 81.8% and were comparable between patients and unaffected family members, while the ADCC of both family groups was lower than the unrelated healthy controls.

The lack of difference between patients and their unaffected family members indicates that low ADCC is unsuitable as a diagnostic biomarker for ME/CFS. Familial CD16Apos NK blood cell counts were lower than unrelated healthy controls. The potential for synergistic effects of combined low CX1:1 and low effector cell counts occurring in the same individual was 24-fold greater for CFS family members than for unrelated controls. FCGR3A of the families was predominantly F/F homozygous, correlating with the observed low EC50 for NK recognition of target cell-bound antibody. In summary, low ADCC is unsuitable as a biomarker, but could be a familial risk factor, for ME/CFS.

https://www.medrxiv.org/content/10.1101/2019.12.20.19015438v1

 

[Ravn]

Disappointing of course that their hunch for a biomarker didn't work out but refreshing to see them state their null result so clearly. Too many others would have tried to dress it up as something it isn't. Ruling things out is valuable; we can move our scarce resources on to the next line of enquiry.

Much of the paper is too technical for me but I found the inclusion of healthy family members as controls interesting. I wonder why it's not being done more often? It seems a worthwhile thing to do.

The inclusion of family members without ME/CFS as controls in the present report resulted in the rapid identification of unsuitable diagnostic ADCC biomarkers (which would have “passed” evaluation if comparisons were made only of patients vs. the unrelated healthy controls).

Basically they found significant differences in ADCC between patients and healthy (unrelated) controls which would normally create a great deal of excitement. But when they looked at the healthy family relatives, their ADCC levels were much the same as the patients', so ADCC couldn't distinguish between healthy and ill family members which makes it useless as a biomarker (though it still could still mark a predisposition for ME).

On the basis of these observations, we suggest that, while low cellular ADCC activity alone is definitely not a biomarker of CFS, low ADCC may be a risk factor for CFS.

 

[Jonathan Edwards]

I have not looked at the stud details thoroughly but I wonder if there may be some selection bias in the finding of a difference between ME families and controls. These are families with multiple ME cases - which are not a representative sample of ME per se. The Fc receptor polymorphisms may have been detected earlier and the associated shift in NK function assay results may not be entirely prospective.

I agree, however, that the comparison with family members is useful in showing that it is not the ME that causes the poor NK function results.