Whole genome sequencing uncovers a misclassified case of glycogen storage disease type 13 previously diagnosed as ME/CFS, 2021, Brown, Younger et al

I very much believe that by doing a proper (deep) work up of each and every patient who fails to recover after a couple of years into ‘ME’, we’d have patients who would have a different diagnosis, or at least further understanding on what is the potential problem for them. We could also stratify groups by most prevalent problematic system (energy, muscle, GI, immune, CNS) and dig deeper. For this we’d need a patient registry, a biobank, and big data kind of people.

 

Happily, the NIH in house study is already making a good start on this. They have probably the most sophisticated diagnostic setup in the world and I'm fairly sure it includes whole genome sequencing. But I think their misdiagnosis rate is around 25%. So most remain with an ME/CFS diagnosis.

The great strength of GWAS studies is that you don't need to go looking for stuff in particular — effectively, it scans the whole of human biology and if something is significant, it will show up.

There is a caveat: to avoid false positives (they will be checking around a million different DNA positions) there is fierce statistical correction. And for decodeme, initially there will only be 20,000 people, which isn't huge. There will almost certainly be DNA differences that stand out from the crowd but don't reach the threshold for statistical significance. (Though there will hopefully be further GWAS in future and the results can be pooled to give greater statistical power to confirm/refute these potential finds.)

 

So 1 out of 4 people from the NIH study gets a different diagnosis than ME. It is a high percentage to me, considering that they are pre-screening the patients really well (i am sure they are). But then i would be surprised if they processed 40 patients yet meaning that the 25% is likely to be different if 1000 patients received the same comprehensive testing to rule out other diseases.

ETA: the 25% is still big, and considering the little screening that is being done at the family practice level, even at the specialist level, it makes me wonder who among us did not receive the proper diagnosis.

 

It would be good to sequence the genome of families with multiple ME/CFS cases (especially if other unexplained chronic illness is also present).

We could learn what genetic diseases are misdiagnosed as ME/CFS which will be very helpful for those patients where this is the case. It could be another way to find rare mutations in genes that increase the risk of ME/CFS (to complement the GWAS which will only be able to find the more common mutations associated with ME/CFS).

For example I was found to have an unusual and potentially devastating mutation in the muscle cadherin gene. Cadherins were the protein family most strongly associated with ME/CFS in the recent Hanson paper. There could be a connection here.

 

I'd be curious to know what the alternative diagnoses turned out to be.

 

25% misdiagnosis rate is unacceptably high IMO.

 

ME is not like rheumatoid arthritis or similar diseases. Even something like migraine which is diagnosed symptomatically can't be compared because of the variability of problems in ME. Ramsay actually felt that variability was a key diagnostic feature of ME. He said it could vary by the hour, day, week, month and year.

It is not helped by the way things like PEM are used by doctors to mean a worsening of symptoms after exercise that is more post exertional fatigue with none of the immune aspects it has in ME. Even the emphasis on fatigue when many people with ME do not experience it causes confusion.

Then there is a belief that people with CSF should not get extensive testing for other diseases because it will confirm they actually have a disease (!) so it would be no surprise that rare diseases are being misdiagnosed as ME.

When there is a diagnostic test many of us will turn out to have something else.

 

Maybe drop Maureen an email - she may be interested.

 

This work and this specific finding was discussed in a Solve ME webinar in 2019.
Thread : Solve M.E. webinar on 28th May re Whole Genome Sequencing and Analysis of ME/CFS
https://www.s4me.info/threads/solve...genome-sequencing-and-analysis-of-me-cfs.9324

Link to slide image showing ENO3 variant (Time stamp 40min in video in thread I just linked)
https://www.s4me.info/attachments/screen-shot-2019-05-28-at-21-00-05-png.7394/

The presenter in the video states that a muscle biopsy would be needed to confirm.

It does seem the researchers submitted this to Clinvar
https://www.ncbi.nlm.nih.gov/clinvar/RCV001172283/

Here is the gnomAD database info which shows a predicted loss of function and high impact CADD score of 33.
https://gnomad.broadinstitute.org/variant/17-4956111-C-CCAGAT?dataset=gnomad_r3

And here is dbSNP
https://www.ncbi.nlm.nih.gov/snp/rs1298530904

[I hope I got those two right]. The variant does seem very rare.

The author and the video presentation doesn't say if this is a heterozygous or homozygous mutation. This is what wikipedia states

Source : https://en.wikipedia.org/wiki/ENO3



In regards to the intramural NIH study that folks are talking about, my understanding was that some people could not continue in the study due to the presence of other found issues, but those issues did not necessarily mean they did not have ME. I think it was Brian Vastag that explained that. I believe he had an abnormal muscle biopsy so could not continue to the 2nd visit. They did not want comorbidities to affect their analysis.

 

With regard to comorbidities what % of people are eliminated, from the intramural NIH, on that basis?

Just to show my ignorance here but would mild [heterozygous] disease not have been picked up by Doctors? Need to look at he above links etc. Thanks.

 

As far as i know they are still recruiting so it would be hard to tell unless you are part of that research team.

 

I found this link to the Webinar https://solvecfs.org/you-can-power-the-next-wave-of-m-e-research/

 

Bear in mind aptamer-based technology Maureen Hanson's published paper*. If this turned out to be a common problem then it might be possible to find something in the blood i.e. to base diagnosis on - rather than biopsy's.

*
https://www.mdpi.com/2227-7382/9/1/6/htm

 

25% misdiagnosis rate in NIH intramural study
@Milo it's worth noting that the NIH only studied people with a post-infectious onset. I wouldn't be surprised if the misdiagnosis rate was very much higher for people with a gradual onset. But I'm not aware of any data on that.

I'm pretty sure that @Chris Ponting has commented on this before and I think a fairly large cohort would be needed to have any chance of success. Building such a large cohort would be very difficult, but 10 but potentially very valuable. Perhaps the Solve You and ME registry could be part of the answer (the aim is that decodeme will funnel people from the study to the registry).

@Hutan

Yes, that could certainly be done once the data is in. In fact, one value of GWAS is that if you have a hypothesis, you can search a GWAS for genetic evidence to support (or otherwise the hypothesis).

But I don't see how this would ever provide evidence for routine genetic testing. It's important to note that going from genome sequencing and detecting rare variants to understanding why any patient is ill requires a big leap. As someone pointed out, we don't even know if the EN 03 variant was homozygous. And even if it was, whether or not that would explain the patient's illness. Or if there were treatment if it did.

And it's notable that this is the one case they focused on in the pilot. I saw the Solve presentation, and while I think in every case there was a possible story, but whether or not it explained patient's illness is unknown. I gather you need about a thousand genomes to be able to tell if you're getting robust results.

But Then it is only meant to be a pilot. And I do think the whole approach is very interesting

it might, or it might not. You're talking about rare variants. The common variants that GWAS normally study have only tiny impact on the risk of disease and are unlikely to explain the disease of any individual. Rare variants (which are more likely to affect the protein produced by genes, as opposed to common variants which tend to affect the amount of protein produced) do have bigger effects. But they're not like the genes in genetic diseases, such as cystic fibrosis and sickle cell anaemia where if you have a homozygous condition, you going to get the disease.

And even if it does explain a patient's illness, there may be no treatment available.

I don't mean to sound negative. I'm just saying that there is a gulf between interesting findings from whole genome sequencing and a method that could guide treatment for a good chunk of patients.

 

I agree with your impression.

i do hope we find out one day.