Increased risks of cancer and autoimmune disease among the first-degree relatives of patients with ME/CFS, 2022, Moslehi et al

Abstract

Background: Myalgic encephalomyelitis (ME)/chronic fatigue syndrome (CFS) is a disabling multi-system complex disorder with prevalence of 875 per 100,000 (up to 3.4 million people) in the United States. There are no known etiologic or risk factors and no approved treatments for ME/CFS. We conducted a molecular epidemiologic study to test the hypothesis that ME/CFS may be an autoimmune disease (AID) and explore the link between ME/CFS and cancer, specifically hematologic malignancies.

Methods: Our clinic-based study involved carefully selected cases with confirmed diagnosis of ME/CFS (n=59) and healthy controls (n=54) frequency matched to cases on age, gender and ethnicity. During structured interviews, detailed multi-generation pedigrees, epidemiologic and medical questionnaires, and biospecimen were obtained on all subjects. Statistical analysis of pedigree data involved comparison of cases and controls with respect to the prevalence and cumulative incidence of AID and cancer among their first-degree relatives. For unadjusted analysis, risk ratios, 95% confidence intervals (CI), and p-values were calculated. For age-adjusted analyses, cumulative incidence estimates were compared using the log-rank test.

Results: The prevalence of AID was significantly higher among the first-degree relatives of cases compared to those of controls (OR=5.30; 95%CI: 1.83-15.38; p=0.001). The prevalence of AID among mothers was 14% for cases and 1.9% for controls (p=0.03). 11.2% of the first-degree relatives of cases had an AID compared to 3.1% of the relatives of controls (prevalence ratio=3.71; 95% CI: 1.74-7.88; p=0.0007). The cumulative incidence of AID among the first-degree relatives of ME/CFS cases was 9.4% compared to 2.7% for those of the controls (p=0.0025). First-degree relatives of ME/CFS cases had a significantly higher prevalence of any cancer compared to the relatives of unrelated controls (OR=4.06, 95%CI: 1.84-8.96, p=0.0005). Age-adjusted analysis revealed significantly higher (p=0.03) cumulative incidence of any cancer among the first-degree relatives of cases (20%) compared to the relatives of controls (15.4%). The cumulative incidence of hematologic cancers was also significantly higher among the relatives of cases (p<0.05).

Conclusions: We found statistically significant increased risks of AID and cancer among the first-degree relatives of ME/CFS cases. Our findings implicate immune dysregulation as an underlying mechanism, providing etiologic clues and leads for prevention. Given symptomatic similarities between ‘long COVID’ and ME/CFS, it is predicted that there will be a significant increase in incidence of ME/CFS as the result of COVID-19 pandemic. Our findings may enable defining a subset of COVID-19 patients who could be at risk of developing ME/CFS, and who may benefit from treatments used for certain AIDs.

Abstract only

 

Interesting. My father had rheumatoid arthritis.

 

Leukemia killed my father.

 

My mother developed chronic ulcerative colitis (an autoimmune disease) in her 30's, which likely led to the development of liver cancer (actually bile duct cancer within the liver) in her late 50's - ultimately taking her life.


[ETA: clarified bile duct cancer]

 

I would hazard a guess if you ask a significant number of people about their close relatives health many would report the existence of autoimmune diseases . correlation is not causative .

 

Yes, this does not look to me to use reliable enough methodology to want to put much weight on it. A decent population based study with very good checking of data is needed.
There may be correlations and they may be important but we all have relatives with diseases.
Ulcerative colitis is probably not in fact an autoimmune disease but rather an auto inflammatory disease, although things are complicated.

The odd thing here is that although they found increased rates of both immune disease and cancer they say this suggests an immune cause for ME. But why not a neoplastic cause? The logic seems a bit woolly. There are links between autoimmunity and cancers but they are very very specific like the relation between PBC and liver tumours or RA and certain types of lymphoma. To show something that really looks to be relevant you need much more careful investigation I think.

 

I think they need a better acronym than AID. I keep reading AIDS, and I imagine many other people will, too. Minor quibble, but it seems really strange they didn't consider the high likelihood of confusion with a very common and familiar term.

 

I have a first degree and second degree relative with hematologic malignancy. I know others with ME who have a first degree relative with hematologic malignancy. I'm glad this is being looked into. I've always felt that it could provide a clue as to who is likely to get ME or which ME pts might respond to which treatments.

This is why I am hoping that DecodeME might include a question about relatives with hematologic malignancies in their questionairre.

 

A historical note of inconsequential interest: In the late 70s, a widely advertised diet suppressant taken in the form of a kind of candy or toffee was called AYDS. There were chocolate, chocolate mint, butterscotch, caramel and peanut flavors. Every ad was someone's story, with photos from the person's pre-AYDS days (fat) and post-AYDS (thinner). The product did not survive long into the '80s, once the epidemic gained public awareness and was renamed from GRID (gay-related immune disorder) to AIDS.
https://en.wikipedia.org/wiki/Ayds

 

A limitation of using first degree relatives is that without specific analysis it is difficult to separate social, behavioural and environmental features from purely genetic factors. Each family unit tending to have the same access to healthcare, same dietary habit, same addictive habits, same educational opportunities, same work status, same income level, same environmental exposures. Inherited genetic susceptibilities of course may respond to those exposures in similar ways but the task of teasing those susceptibilities out can't be achieved without detailed statistics.

The 3 fold excess of autoimmune disease compared to controls looks significant but assuming (an optimistic) 6 relatives per subject (2 x parents, 2 x siblings, 2 x children) in this study that produces just 26 individuals more with an AiD than would be expected in an average population, it's not inconceivable that these 26 fall into a very limited number of family groups.

The 7 fold excess amongst mothers looks very significant but only amounts to 8 individuals matched to 1 control. That would suggest some specificity in AiD type.

Maybe there's something in this but it will need bigger numbers to make sense of and I'd bet bigger numbers will mean diminishing effect. Which is a shame because the work itself looks to be thorough.

 

I'd be interested to compare the pwME who had a relative with autoimmd with those who had a relative with hematologic malignancies and those who had no affected relatives. Were they sudden/gradual onset? What is their severity? Have the improved/deteriorated since onset? Age/sex? There might be some subsets involved.

 

All 3 of my siblings have/had autoimmune disease - Coeliac, T1D, ITP

 

My sister (only sibling) has Hashimoto's thyroiditis and so do I. My mom has vitiligo, which is believed to be autoimmune.

I had sudden viral onset. I started at the lower end of moderate, initially (in the first 2-3 years) improved a lot to very mild, then started to fluctuate.

 

My mum has (remarkably) rheumatoid arthritis and my wife, her sister and I have all had two cancers each. Nobody has ME. These things are around so studies have to be very careful.

 

In my family there was 1 case of cancer in 4 generations and no autoimmune disease. Nothing major at least.

I think I may have some psoriasis but it's so mild that it's hard to tell.

My mother might have some autoimmune thing affecting her fingers but she has refused to get it diagnosed and it doesn't seem to be too bad most of the time.

 

AD variation runs in families so looking for an association for ME/CFS in first degree relatives risks picking up familial groupings with high AD prevalence but which are only randomly associated with ME/CFS.

How do autoimmune diseases cluster in families? A systematic review and meta-analysis
BMC Med. 2013; 11: 73.
Published online 2013 Mar 18. doi: 10.1186/1741-7015-11-73

Abstract
Background
A primary characteristic of complex genetic diseases is that affected individuals tend to cluster in families (that is, familial aggregation). Aggregation of the same autoimmune condition, also referred to as familial autoimmune disease, has been extensively evaluated. However, aggregation of diverse autoimmune diseases, also known as familial autoimmunity, has been overlooked. Therefore, a systematic review and meta-analysis were performed aimed at gathering evidence about this topic.

Methods
Familial autoimmunity was investigated in five major autoimmune diseases, namely, rheumatoid arthritis, systemic lupus erythematosus, autoimmune thyroid disease, multiple sclerosis and type 1 diabetes mellitus. Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) guidelines were followed. Articles were searched in Pubmed and Embase databases.

Results
Out of a total of 61 articles, 44 were selected for final analysis. Familial autoimmunity was found in all the autoimmune diseases investigated. Aggregation of autoimmune thyroid disease, followed by systemic lupus erythematosus and rheumatoid arthritis, was the most encountered.

Conclusions
Familial autoimmunity is a frequently seen condition.
Further study of familial autoimmunity will help to decipher the common mechanisms of autoimmunity.

 

I found the poster for this study, which gives a little more detail, including the list of auto-immune diseases:

https://twitter.com/user/status/1513233638569172994


Interesting to see they selected only ME patients with acute onset following infectious disease.

This definitely needs follow up, since it is such a small study. But the good thing is this that the follow up would be very easy to do. It looks like they had NIH funding for this work; maybe they can get more for a follow up.

 

It looks like the first result is looking at a binary variable: having one or more first degree relative with AID versus having none, so I don't think that would be an issue for that result (the first OR 5.30 one).

 

Looks like they plan genomic sequencing of this cohort and expanding to more patients in the future:

https://twitter.com/user/status/1538933286226915328

 

You learn something new every day! I'm not sure I'd want to eat benzocaine sweets anyway.