Biomarkers for ME/CFS - discussion thread on the next steps for testing biomarkers, and why we need them

[AknaMontes]

Moved post

A simple biomedical test that could identify ME/CFS might help those Very Severe who are trapped in terrible situations in the UK being denied assisted nutrition on the basis they must be psychologically refusing to eat - resulting in malnutrition to the point where they can no longer be saved. (ie inquest ongoing re case in Exeter and Dr Weir and Dr Speight's article on malnutrition in VSME). There are also children being taken from parents with false FII accusations and forced to exercise on the assumption the disease doesn't exist. I appreciate that there are psychs out there who will never be convinced by any amount of scientific evidence that the disease is biomedical - but a clear diagnostic test has to be of some benefit perhaps preventing these most vulnerable patients from ever ending up in front of substandard doctors in the first place.

 

[Barry]

What different characteristics would people with ME/CFS who had paid for a test have over those who did not? I could understand that possibly being an issue with a large population study but most ME/CFS studies use convenience samples and it's unclear to me what the different characteristics would be in a scenario like this.

I thought the whole point of such things was that you may not know up front. What if there are different disease variants that might be influenced by poverty issues, pre-ME/CFS private healthcare for other health conditions, etc. I thought the nearer to random selection you can achieve the better?

 

[Barry]

biology is complicated that otherwise you just get coincidences flagged up as meaning something - or worse.

I guess this is what the BPS brigade exploit so much of the time. Coincidences abound within complexity, and if you cherry pick the ones you like, and studiously ignore the ones you don't, presenting your favoured ones as causal relationships ... then making silk purses out of sows' ears becomes routine.

 

[Barry]

Another thought, while my mind is focused on this.

A study may find that there is a statistically significant difference between the mean value of substance X in the blood of pwME and the blood of healthy controls. This doesn't necessarily mean substance X is a biomarker for ME/CFS.

And it could simply be a biomarker for something else, that has some correlation with having ME/CFS, though that could presumably still be very useful to know.

 

[Dolphin]

What different characteristics would people with ME/CFS who had paid for a test have over those who did not? I could understand that possibly being an issue with a large population study but most ME/CFS studies use convenience samples and it's unclear to me what the different characteristics would be in a scenario like this.

I thought the whole point of such things was that you may not know up front. What if there are different disease variants that might be influenced by poverty issues, pre-ME/CFS private healthcare for other health conditions, etc. I thought the nearer to random selection you can achieve the better?

The perfect can be the enemy of the good.

My guess is many ME/CFS studies do not have a truly representative sample of the population.
Certainly I imagine that's the case in many US ME/CFS studies.

I think publishing data on the results from tests that had been paid for (esp. if you had other information on the sample population age, gender, etc. and restricted it to those with a confirmed diagnosis) would be better to not having any data.

 

[Dolphin]

The perfect can be the enemy of the good.

My guess is many ME/CFS studies do not have a truly representative sample of the population.
Certainly I imagine that's the case in many US ME/CFS studies.

I think publishing data on the results from tests that had been paid for (esp. if you had other information on the sample population such as their diagnosis, age, gender, etc.) would be better to not having any data.

From listening to the accents of callers to the helpline for the association I help run over the last 25 years, I can tell that people from lower socio economic groups/more deprived backgrounds are less likely to call. Similar things can be seen from their addresses from people who enquired.

My guess this is similar in many ME/CFS groups around the world and indeed I recall seeing some studies which suggest that also i.e. diagnostic rates are not even throughout society even in countries such as the UK with fairly universal healthcare.

Edited to add: I imagine such bias extends to partipation in much research in the field.

 

[Barry]

Edited to add: I imagine such bias extends to partipation in much research in the field.

I'm sure you are right, but is that necessarily a justification for consciously introducing a potential source of bias, unless there is a way to correct for it?

 

[Dolphin]

I'm sure you are right, but is that necessarily a justification for consciously introducing bias, unless there is a way to correct for it?

As I understand it, the proposal was to get research done on the cheap by using existing data.

Future studies could use a different methodology.

 

[Barry]

The perfect can be the enemy of the good.

Yes, (nothing is perfect of course), but that argument is immediately then prone to what someone's definition of perfection is, and what their motivation might be to label something as such. I'm speaking generally here, I emphasize.

I worry that line of argument can easily morph into the BPS' line of argument - that GET must be OK for pwME "because it is the best we currently have"; could almost paraphrase as "we can't have the perfection you seem to insist on, so what we already have must be OK".

Not for a moment suggesting that is what is going on in this case, but just a general observation, and how others could misuse.

 

[Dolphin]

Yes, (nothing is perfect of course), but that argument is immediately then prone to what someone's definition of perfection is, and what their motivation might be to label something as such. I'm speaking generally here, I emphasize.

I worry that line of argument can easily morph into the BPS' line of argument - that GET must be OK for pwME "because it is the best we currently have"; could almost paraphrase as "we can't have the perfection you seem to insist on, so what we already have must be OK".

Not for a moment suggesting that is what is going on in this case, but just a general observation, and how others could misuse.

Well in this case it relates to the level of bias and how important it is.

In some scenarios, not having a truly socio-economically balanced cohort could mean there was a large level of distrust about the findings. However I'm not sure where possible biomarkers and the like were being investigated in ME/CFS, that it would be a huge reason to doubt the findings in most cases?

And as I say the same question mark could be put over lots of existing research in the field. Biomedical studies in the field generally don't report the socio-economic status of participants and I imagine if they did, many would not be socio-economically balanced.

 

[Sean]

then making silk purses out of sows' ears becomes routine.

Indeed, it becomes your 'methodology'.

 

[Sasha]

Posts have been moved from Replicated blood-based biomarkers for Myalgic Encephalomyelitis not explicable by inactivity, 2024, Beentjes, Ponting et al

Ponting's study does not claim to provide any sort of diagnostic test. Moreover, diagnostic tests are not what we need. The comment shows a lack of understanding of the basic process of clinical decision-making.

I agree that this study doesn't provide a diagnostic test and that it could give us useful info on mechanism, but you can add me to the list of people who don't understand the basic process of clinical decision-making! I thought we did need diagnostic tests, and that this has been a major research goal of ME/CFS for decades. What am I missing?

 

[EndME]

I agree that this study doesn't provide a diagnostic test and that it could give us useful info on mechanism, but you can add me to the list of people who don't understand the basic process of clinical decision-making! I thought we did need diagnostic tests, and that this has been a major research goal of ME/CFS for decades. What am I missing?

The point that is being made I believe, is that we don't need a diagnostic test because we already have one: The diagnosis itself (for example via the CCC). If we were to have a diagnostic test that is perfect, that is to say 100% accurate, i.e. 100% sensitive and 100% specific, we wouldn't have anything new, one would just be diagnosing those people we are already diagnosing and no one else. It would accomplish absolutely nothing. If it is for example less accurate by being less specific than it is no different to just using a different definition (for example Fukuda).

What is needed is a finding that delivers a pathological clue, something that tells us where things could be going wrong in a subset of people. In that case sensitivity and specificity become far less important and the crucial part is that you can actually explain pathology in a subset of people.

Now it might seem a bit contradictory at first that one could have a "diagnostic test" that doesn't deliver any biological clues, but that is precisely what has occured over and over again for years. There's hundreds of papers claiming to have discovered a "diagnostic test" for Long-Covid, stating a certain specificity and sensitivity, yet as anyobody can witness, none of those have brought any understanding, nor will they ever help any patient. The reasons why one could obtain a "diagnostic test" without it somehow having identified anything pathological are plentiful and range from it being a test for deconditioning, it capturing a different behavioural consequence of being ill, to it capturing the notion of a patient behaving like a patient rather a healthy control etc. That is to say it's extremely easy to obtain a meaningless test, which is precisely why it occurs over and over again, year after year.

Sometimes people argue that a "diagnostic test" would end the psychological debate. But that is likely to not be true. It would only do so if it actually gives a vital clue to biomedical pathology, otherwise one could just as well argue that the test simply identifies people that are lazy or whatever notion is suitable to the scenario.

 

[Jonathan Edwards]

I thought we did need diagnostic tests, and that this has been a major research goal of ME/CFS for decades. What am I missing?

This is a pervasive myth. I worked on rheumatoid arthritis for 35 years and we made huge progress - enough to keep almost all patients well all the time. But there is no diagnostic test for rheumatoid. GPs think there is and that is a huge problem because people with rheumatoid with negative tests (a significant proportion early on) can get sent away without a diagnosis.

There is no need for a diagnostic test for a disease that is recognised by symptoms and signs. Once those symptoms and signs are explained by lab tests the old 'disease' can be forgotten and people can be treated for what we can see is wrong with them.

Does someone with the symptoms of ME/CFS not have ME/CFS if a test is negative?
Does someone without symptoms of ME/CFS have ME/CFS if the test is positive?

Ah, you might say, but doctors cannot be sure without a test whether people have ME/CFS.
In which case it is impossible for any test to come up with 100% of patients with ME/CFS and nobody without - because the samples tested on are going to be partly 'wrong' - whatever that means - if doctors are unsure who should be included. It is a completely circular exercise.

If patients are not being diagnosed in good time it is because doctors are not trained properly to listen to the symptoms. A test doesn't help make them better doctors.

 

[duncan]

There is no need for a diagnostic test for a disease that is recognised by symptoms and signs. Once those symptoms and signs are explained by lab tests the old 'disease' can be forgotten and people can be treated for what we can see is wrong with them.

Does someone with the symptoms of ME/CFS not have ME/CFS if a test is negative?
Does someone without symptoms of ME/CFS have ME/CFS if the test is positive?

As long as a disease like ME/CFS remains contested, a diagnostic test can benefit patients. With disability and health insurance claims challenged far too frequently, and a skeptical public that can include family and friends, a diagnostic that can rise above those challenges would be welcome. It has been MIA for too long.

 

[Mij]

The pathophysiology of delayed PEM biomarker would be very useful. Early signs of molecular changes that occur that could possibly prevent harm and worsening before they occur would be very helpful.

 

[V.R.T.]

This is a pervasive myth. I worked on rheumatoid arthritis for 35 years and we made huge progress - enough to keep almost all patients well all the time. But there is no diagnostic test for rheumatoid. GPs think there is and that is a huge problem because people with rheumatoid with negative tests (a significant proportion early on) can get sent away without a diagnosis.

There is no need for a diagnostic test for a disease that is recognised by symptoms and signs. Once those symptoms and signs are explained by lab tests the old 'disease' can be forgotten and people can be treated for what we can see is wrong with them.

Does someone with the symptoms of ME/CFS not have ME/CFS if a test is negative?
Does someone without symptoms of ME/CFS have ME/CFS if the test is positive?

Ah, you might say, but doctors cannot be sure without a test whether people have ME/CFS.
In which case it is impossible for any test to come up with 100% of patients with ME/CFS and nobody without - because the samples tested on are going to be partly 'wrong' - whatever that means - if doctors are unsure who should be included. It is a completely circular exercise.

If patients are not being diagnosed in good time it is because doctors are not trained properly to listen to the symptoms. A test doesn't help make them better doctors.

Agree with what you say about the goal being understanding enough of what's going on to develop treatments, not a diagnostic test.

But having one would have prevented my decline from mild, because a doctor couldn't say that they couldn't find anything wrong with me and withhold diagnosis and the knowledge that comes from it.

A lot of people push themselves into a condition where they cannot work or care for themselves because of no diagnosis. And unfortunately retraining doctors not to gaslight and psychosomatise ME will probably take a generation. So it would be really handy to have a test before then.

 

[Jonathan Edwards]

But having one would have prevented my decline from mild, because a doctor couldn't say that they couldn't find anything wrong with me and withhold diagnosis and the knowledge that comes from it.

But that's the point. We need a test that shows that 'something's wrong', not a test for ME/CFS, which is merely a placeholder term we use at the moment because we have no real idea which patients have the same thing wrong and which have some other thing wrong.

Any test that tells us that something is wrong, say S metabolism, immediately changes the concept of ME/CFS to a disorder of S metabolism - a diagnostic test for S disease.

It is conceivable that nearly everyone with a diagnosis of ME/CFS has something wrong with the same pathway and that a test could pick that out but it is just as likely that it is more complicated. If you start out assuming that you want to find a test that fits as precisely as possible with ME/CFS symptoms you are very likely to end up with a lemon, as has been said above. SO the idea that you are looking for a 'diagnostic test for this' is a logical error that almost everyone falls into until they see the absurdity of it. Yes, peo0le with ME/CFS would be much better off if there were tests that showed something was wrong but their value lies in showing what is wrong - that is what gets believed.

 

[V.R.T.]

But that's the point. We need a test that shows that 'something's wrong', not a test for ME/CFS, which is merely a placeholder term we use at the moment because we have no real idea which patients have the same thing wrong and which have some other thing wrong.

Any test that tells us that something is wrong, say S metabolism, immediately changes the concept of ME/CFS to a disorder of S metabolism - a diagnostic test for S disease.

It is conceivable that nearly everyone with a diagnosis of ME/CFS has something wrong with the same pathway and that a test could pick that out but it is just as likely that it is more complicated. If you start out assuming that you want to find a test that fits as precisely as possible with ME/CFS symptoms you are very likely to end up with a lemon, as has been said above. SO the idea that you are looking for a 'diagnostic test for this' is a logical error that almost everyone falls into until they see the absurdity of it. Yes, peo0le with ME/CFS would be much better off if there were tests that showed something was wrong but their value lies in showing what is wrong - that is what gets believed.

Agreed.

 

[Mij]

Jarod Younger was saying on youtube recently (2024 Stanford ME/CFS meeting) that different working groups are seeing subgroups with tests showing different pathologies.