Preprint A Proposed Mechanism for ME/CFS Invoking Macrophage Fc-gamma-RI and Interferon Gamma, 2025, Edwards, Cambridge and Cliff

[Kitty]

Really good to see a positive response from MEA, but honestly the original paper is a lot more understandable to me than that summary, as someone who knows nothing about the subject!

 

[Snow Leopard]

I wonder about it acting on a neuron, like the DRG, with the message passed to the brain?

Interferons (of several stripes) like other cytokines don't cause sensory signals directly, but there are lots of studies showing they can sensitise neurons including in the DRG. (and I will be including discussion of this in the manuscript I am writing that may or may not ever be finished)

 

[rvallee]

I’m not sure I follow here. Who «sees» us, and why would we be trying to please them?

Yes, but it’s also being used very wrong, on everyone, and as the basis for supposedly curative interventions with great potential for harm.

I’m not suggesting that we don’t raise them, just that it is done in a way that is also less open for misinterpretation. Although I guess it could be argued that pseudoscience doesn’t follow logic, so they are going to end up at their favoured conclusion regardless of which starting point we give them..

It's definitely fair to say that, generally speaking, women take better care of their health than men. This has been taken to absurd levels onto misogynistic models, but it's still true. It means little more than women are more likely to report such problems. It doesn't even mean they are more likely to notice them, just report them. Because its counterpart is also known: men generally downplay symptoms and functional decline. Sometimes all the way to an early death.

It shouldn't mean much in this context, and ideologues will always say their thing no matter what, but it's fair to mention it, including as one potential reason for a female predominance in this illness. Which could just as well be the same reason why autoimmune diseases are more common with women. Most likely a combination of factors in which this is just one variable.

Saying the truth isn't about pleasing anyone, it's about being honest to what the facts are. How people choose to interpret those facts is a different issue. If the ideologues choose to latch on to that, it doesn't really matter because they are dishonorable, are not interested in the truth. They'll say all sorts of false things regardless. Just as if the answer happens to be in the brain, they'll go around screaming "Ah AAHHH!", even if it has zero overlap with anything they said before, and obviously can't be treated with happy thoughts or recreational activities.

Although I really can't see any potential role for neuroplasticity here. It just obviously doesn't fit with the facts. Maybe down several levels, in a similar way where stress, in a general sense, and cellular stress (or even mechanical stress) have completely different meanings and mechanisms, but even then it's just not supported by anything but speculation.

 

[rvallee]

Successful de-escalations often involve giving one party a way to withdraw from their position without losing face. Maybe, as and when new evidence arrives, the BPS advocates will feel readier to cede some influence if they can say "oh yes, your 'neural hypervigilance' is what we were driving at all along," instead of having to take it as a humiliating defeat.

At least in the transition phase. After that all bets are off

But, definitely, a good transition usually involves some escape valve for the defeated to scamper through.

 

[Karen Kirke]

Thank you so much, @Jonathan Edwards , Jo Cambridge and Jackie Cliff for all your thinking and work on this paper. Please pass on our thanks. I hope you gave them the good biscuits when they came to your house to work on it.

I really appreciate all the time and energy you have put into trying to figure out how salient features of the clinical picture and literature could fit together. What you’ve come up with is fascinating.

I think it will take multiple reads and extensive googling for me to fully understand. But it’s very readable, even when you don’t understand all the concepts.

Would your proposed mechanism predict higher prevalence of those conditions in relatives of people with ME/CFS @Jonathan Edwards ? My family is heaving with these conditions:

The conditions that might be most closely analogous are the seronegative autoinflammatory diseases, including psoriasis, Reiter’s syndrome, ankylosing spondylitis, other MHC Class-I associated seronegative arthropathies[21] and Crohn’s disease.

Could this bit explain how my cognitive function unexpectedly improved with a low-FODMAP diet, which would alter gut flora by feeding them less? And why probiotics always made me feel worse, specifically, more fatigued?

These antibodies may have high affinity for previously encountered microbes and may have undergone affinity maturation in some past immune response but are relevant because they bind with relatively low affinity to what might be called ‘everyday junk’ antigens that may include antigens derived from gut flora, low-level persistent viruses such as Herpes, partially degraded self proteins with post-translational modifications, or possibly normal host signal pathway proteins, such as heat shock proteins and proteasome components, upregulated during immune responses.

This sentence made me chuckle:

Until recently, treatment for ME/CFS has tended to focus on rehabilitative programs that encourage patients to undertake physical exertion of the sort that triggers symptoms.

A few suggestions:

The phrase “pain and ‘fatigue’” appears a couple of times. I’d prefer if it was “‘fatigue’ and pain”. By putting pain first, it can sound like fibromyalgia, and since people with fibro do much better than us with exercise (as evidenced by FM studies of exercise that are mostly of exercise classes of 45+ minutes vs ME/CFS studies), I think there may be important differences in mechanism.

I share @Hutan ’s misgivings about the first mention of resolution:

ME/CFS has a variable course. Spontaneous improvement or even resolution may occur, chiefly with onset before the age of 20[20]. Otherwise, it is a long-term fluctuating illness, in some cases with progressive deterioration. Adult-onset illness mostly never fully resolves.

First, there’s the issue of resolution in the early post-viral/post-infectious years. ME/CFS can now be diagnosed at 3 months post-onset. We’ve got the Dubbo study https://pubmed.ncbi.nlm.nih.gov/16950834/ that shows that of those who are sick at 3 months, 57% have recovered by 6 months (38/67) and 67% have recovered by 12 months (45/67).

The case rate for provisional post-infective fatigue syndrome was
35% (87/250) at six weeks,
27% (67/250) at three months,
12% (29/250) at six months, and
9% (22/250) at 12 months.

And things look similar with long COVID, with resolution continuing as long as patients are followed https://s4me.info/threads/looking-f...vid-in-year-one-and-beyond.44229/#post-610364.

By not acknowledging that most with an ME/CFS-like presentation at 3 months post-infection recover or at least remit in this sentence, the door is open for perfectly valid arguments that recovery is being underestimated. The sentence on p.4 comes too late:

Relatively short-term post-viral fatigue, lasting a matter of months, is also common for a number of infections, notably EBV, and diagnostic boundaries are uncertain.

Second, I had a look at the datasets for Collin & Crawley’s 2017 study of NHS specialist centres, which were published in 2019 by Crawley and Gaunt. There’s a sample of 385 patients for whom we have data at median 36 months post-onset and 2-5 years after that, i.e. 5-8 years post-onset. They attended 7 different centres 2010-2013, median age 43, 76% female.

12.7% of 385 were in the SF36 physical function range of 90-100 at 5-8 years post-onset, compared to 2.6% of 385 who were in that range at baseline i.e. 3 years post-onset. Wilshire et al. 2016 identified that range as being where 90% of the healthy population are. (There’s no real suggestion that anything they did at the specialist centres contributed to this. The mean change in SF36PF is an increase of 7.7 points in 5-8 years and I would say that upward drift is down to regression to the mean, spontaneous change and changes in how people fill in questionnaires more than anything else.)

Now, not all of the 49/385 in the SF36PF range 90-100 at 5-8 years post-onset will be fully recovered. 13/49 still have Chalder Fatigue scores ≥15/33. 12/49 are at an SF36PF score of 90, which is likely to mean they are still not able to do vigorous exercise at all. (They could also plausibly be limited a little in vigorous exercise and be limited a lot in something else like carrying groceries due to arthritis, for example.) 15 of the 49 in the range 90-100 at 5-8 years post-onset were already in that range at median 4 years post-onset (aka 1 year after first attending the specialist centre).

These numbers are not impressive at all, especially when divided by 7 centres and 3 years, but I don’t think “mostly never” summarises them. I think something like “Complete resolution of adult-onset ME/CFS is rare after the first few years” would be more accurate.

 

[hotblack]

While reading up on MAIT cells I saw that they recognise vitamin B compounds. Could one reason that some of us respond well to doses of B 12 or B complexes be nothing to do with “energy” but that we’re giving MAIT cells something else to be interested in so distract some from their roles in this feedback loop for a bit?

 

[voner]

@Jonathan Edwards ,

I was trying to understand some of your statements with only having a hotdog bun education (at best) in this area….

in the “myalgia not arthralgia" section you say,

"Reports of systemic cytokine levels in ME/CFS have been inconsistent, but one relatively recent study found raised serum levels of gamma interferon and transforming growth factor beta but not TNF, IL-6, or IL-1".

and you reference the 2017 paper by Mark Davis and others…

https://www.pnas.org/doi/epdf/10.1073/pnas.1710519114

I took a look at this 2017 paper and I don’t see where it shows raised serum levels in gamma interferon. The charts I looked at seemed to show a rising gamma interferon that correlated with increased severity. In mild patients it was below the healthy controls and in moderate patients it was about the same as healthy controls and in severe patients it was elevated over healthy controls.

am I interpreting this information incorrectly?

 

[Jonathan Edwards]

While reading up on MAIT cells I saw that they recognise vitamin B compounds. Could one reason that some of us respond well to doses of B 12 or B complexes be nothing to do with “energy” but that we’re giving MAIT cells something else to be interested in so distract some from their roles in this feedback loop for a bit?

I don't think it is to do with B12 is it? I think it is another B.

 

[Theresa]

Unfortunately I am not able to read this paper indepth but I was wondering whether the mechanism suggested provides an explanation for delayed PEM?

 

[Jonathan Edwards]

Thanks Karen,

Would your proposed mechanism predict higher prevalence of those conditions in relatives of people with ME/CFS @Jonathan Edwards ?

If you are referring to the seronegative spondarthropathies mentioned below, then no, because the HLA risk gene is different for them (B27 or C6)

First, there’s the issue of resolution in the early post-viral/post-infectious years. ME/CFS can now be diagnosed at 3 months post-onset.

This is a misunderstanding. We have to be clear that clinical diagnostic criteria are quite separate from research criteria, not to mention the fact that neither are that useful if you are doing things properly.

Suggestions that ME/CFS can be diagnosed at 3-4 months is to stop idiot doctors from saying that they cannot diagnose and advise patients about ME/CFS until 6 month have passed. For schoolchildren in particular this is a breach of human rights. But what an intelligent doctor should be doing is ignoring the criteria and saying that if someone is disabled by fatigue-type symptoms for three months then there is a significant possibility that in a year's time they will be understood to have ME/CFS. But if they get better at 5 months then I personally would say that they did not have ME/CFS in any useful sense, especially if the fatigue followed a viral infection because we already have another useful category for such people - post viral fatigue.

So no, fatigue for there months is not part of a useful concept of ME/CFS. A useful concept of ME/CFS is of an illness with a dynamic that is not simply a monophasic period of fatigue after infection. The key thing we want to focus on is an illness with a more complicated dynamic with unexplained worsenings over months and years. Four months of PVF isn't the issue. There may be common features involved but the problem of interest is the long term illness.

I am not sure why my paragraph is being interpreted as if I am saying that for this long term illness recovery is rare. I have simple said the mostly it does not fully recover, which is generally agreed. I have pointed out that sometimes it does. And the paragraph isn't about how often anyway. The argument being made is about the dynamic complexity.

I am aware that a small proportion of people during the early months following an infection, including Covid 19, have what appears to qualify as PEM. No doubt some of them will continue to have problems and turn out to have ME/CFS. But the criteria for PEM are hard to pin down and I suspect that the majority have had what I had with Long Covid. I would probably have answered yes to PEM questions because I did get fluctuations. But having listened to 500 people with ME/CFS for ten years I don't for a minute think I had your PEM. The definition of PEM is inherently problematic because it involves causal ascription to exertion and we know that that sort of causal ascription is hugely unreliable. For people with ME/CFS who have monitored their symptoms over months and years it may be very reliable but for anyone in the sort of situation I experienced I wouldn't put any weight on it.

I may review my wording a bit but so far I disagree with people's gripes on this one, firstly because that is not what the paragraph is about and secondly because I don't think what I am saying is actually wrong or misleading. Fatigue that turns out to last no more than 6 months is not usefully regarded as ME/CFS. Which why that time frame was originally suggested. It remains just as valid now.

 

[Jonathan Edwards]

Unfortunately I am not able to read this paper indepth but I was wondering whether the mechanism suggested provides an explanation for delayed PEM?

In as much as it allows for PEM to be dependent on re-activating or further activating immune cell interactions that may take many hours to evolve, yes that is the hope. Time needed for those immune cells to re-circulate to tissues like muscle and nerve might also contribute.

 

[Simon M]

@Jonathan Edwards, thanks to you, Jo Cambridge and Jackie Cliff for putting together such a thoughtful model that, at the very least, shows how a serious explanation of ME/CFS should look. Apart from the detail, it seems to go out of its way to acknowlege various factors that don't obviously fit, and adapt to them.

Hopefully, this model will take us a lot further. Not least because you have previously said that one reason for your optimism is what you are seeing of not-yet-published research.

Some questions and comments:

This proposed mechanism could explain why the illness resembles post-infective T cell-mediated autoinflammatory syndromes in age of onset and time course but has a female preponderance similar to autoantibody-mediated disease.

Is this building up from three repeated, notable findings that are unique (?) in combination and as yet unexplained?

all quotes are from the paper

There is no clear time relation between infective illness and onset of ME/CFS, which may be immediate or delayed for weeks or months in patient reports

The only hard data I have seen on this is from prospective studies of mono (and Q-fever and Ross River Virus for the Dubbo study) and that appears to show a steady decay curve rather than symptoms starting erratically after a delay. Though other infective and non-infective triggers may have a different pattern.

Antibody responses to a range of microbial antigens would not be expected to manifest as the same clinical syndrome; in autoantibody-mediated disease, different antibody specificities are reflected in different clinical syndromes.

That's very interesting. Are you suggesting for this model that ME/CFS is basically a single mechanism with numerous ways in?

Nevertheless, all these findings tend to point to a broad shift in B cell function rather than a single antigen-specific adaptive response.

How confident are you that the various quite subtle findings are reliable rather than unreliable (e.g. due to publication bias)? I don't have an opinion, but I think the question is worth asking when there are no strong results.

(Viral) myalgia
I had thought the mechanism was sickness response/behaviour resulting from cytokines acting directly and on the brain. @Hutan gave the examples of IFN alpha and gamma when administered. Is this just speculative.?

An ME/CFS-like illness has been reported after exposure to acetylcholinesterase inhibitors.

There was no reference for this - can you say more?

This might also be relevant to the finding that cytokine and lymphocyte population shifts in ME/CFS appear to be more pronounced in the first 3-5 years after onset compared to the longer-term situation[56][5].

I had the impression that the plasma cytokine signature paper finding was post hoc, as well as not being a large effect.

Think I should stop there.

 

[Jonathan Edwards]

Is this building up from three repeated, notable findings that are unique (?) in combination and as yet unexplained?

One of the subtexts of the piece is to point out to the thousands of immunologists who have muddled ideas about these things that autoantibody-mediated disease does not follow infection and has nothing to do with microbial triggers and 'cross-reactivity'. So we have a problem if we have a disease that is triggered by infection (like some known T cell mediated things) but is more common in women. Just getting this message across to immunologists who might be researching ME/CFS seems to me would justify the piece. It is based on something Jo C and I have worked on for thirty years. Non-clinical immunologists who don't actually see patients very easily get completely the wrong context about what they are studying. Even for Jackie, I think some of this sort to of stuff was, in her words 'very thought-provoking'.

The argument actually gets complex, because lupus has an age of onset rather like Reiter's. But it is unusual amongst autoimmune diseases in that respect. The proposal is that it may all make sense if we insist on making all the details fit. That is how we came to an answer for RA.

The only hard data I have seen on this is from prospective studies of mono (and Q-fever and Ross River Virus for the Dubbo study) and that appears to show a steady decay curve rather than symptoms starting erratically after a delay.

The problem here is that Dubbo is mostly about people who will turn out not to have ME/CFS but PVF. I don't know whether they assessed time delay to onset of fatigue. My information on this comes partly from asking individuals but mostly from the poll I did here. You do not need a very big poll to see that time delays before fatigue onset are common and very variable - some people say six months. Which of course begs the question as to whether it is relevant, but there was no doubt that there was no set pattern.

That's very interesting. Are you suggesting for this model that ME/CFS is basically a single mechanism with numerous ways in?

Yes. Things are again a bit more complicated because the disease pattern for RA is in a sense independent of the antibody specificity in that it relates to the pattern of distribution of FcRIII. But for most other autoimmune diseases the clinical picture is determined by the antibody specificity - anti-topoisomerase, anti-thyroid peroxidase etc. And also, the proposal is that in ME/CFS there isn't even any particular rogue antibody to anything. So in a sense the way in is always the same - a broad range of antibodies to junk, that have a particular Fc receptor binding capability connive with expanded T cell subsets to screw things up.

How confident are you that the various quite subtle findings are reliable rather than unreliable

Not very. They are interesting but they are not critical to the model.

I had thought the mechanism was sickness response/behaviour resulting from cytokines acting directly and on the brain. @Hutan gave the examples of IFN alpha and gamma when administered. Is this just speculative.?

That was my understanding. But just because you can mimic an effect with large amounts of soluble cytokine does not mean that is how it works in infection. In infection the action might be more local. There are also a lot of unanswered questions it seems to me. This is another example of a rag-bag concept (sickness behaviour) distracting people from specific mediators and specific responses. In some cases you get fever, in others not. Sometimes you get photophobia, sometimes nausea, and so on.

And in ME/CFS you get this hard to describe mix of symptoms, without fever or raised CRP. Maybe that means it has nothing to do with cytokines. But if so, what the heck is it due to? It must be something.

I just think that maybe malaise needs to be looked at more carefully. It reminds me of the days when people talked of RA being due to 'fixation of immune complexes' when nobody actually knew what they meant by that. I remember asking several eminent immunologists what they thought was the difference between immune complexes in RA and lupus and they just shrugged. By the time we had reached our model, Jo and I had worked out for ourselves what was going on and how it made couples sense. RA was just small complexes because complement worked fine and cleared any big ones. Lupus was both because complement wasn't working fine and nothing was cleared properly. The big complexes got stuck in glomerular basement membranes so lupus has nephritis. The little ones leak straight through to the tissues with FcRIII - joints and pericardium. Easy. But because of rag-bag thinking nobody had sorted it out.

There was no reference for this - can you say more?

You would have to ask Margaret Mar who got her ME/CFS from a sheep dip. She is a peer of the realm so her evidence does not require a formal citation!

I had the impression that the plasma cytokine signature paper finding was post hoc, as well as not being a large effect.

I agree. Others have also found some differences with time but we don't have anything very hard.

 

[hotblack]

I don't think it is to do with B12 is it? I think it is another B.

You’re right. Looks like it’s vitamin B2 (riboflavin) and B9 (folic acid) metabolites. May be relevant for them but was just a passing thought really.

 

[hotblack]

The more I listen to and think about this paper it does have the whiff of a Grand Unified Theory of chronic auto-inflammatory/immunity-like diseases.

It would be good if it can start getting others outside the ME/CFS world interested. So how much this should be seen as a call to arms for the willing and interested as opposed to trying to convince those set in their beliefs is something to probably think about.

Some may need to think differently to look at the combinations of pathways, signalling and subtle shifts rather than for a nice big ‘this one thing is wrong’ signpost. But that seems to be the direction of biology and medicine in general though so perhaps convincing the old guard isn’t the path. Embracing the new is.

I think that’s my wider view, I’ve really enjoyed going through the paper, particularly the synthesis section, multiple times. I have notes with more specific questions and thoughts but will go through what’s already been covered here before waffling further.

 

[forestglip]

You would have to ask Margaret Mar who got her ME/CFS from a sheep dip. She is a peer of the realm so her evidence does not require a formal citation!

There's this study, but it was studying CF, not CFS: Chronic fatigue and organophosphate pesticides in sheep farming: a retrospective study amongst people reporting to a UK pharmacovigilance scheme (2003, Ann Occup Hyg)

There's the link between gulf war illness, which can manifest as ME/CFS, and AChE inhibitors: Recent research on Gulf War illness and other health problems in veterans of the 1991 Gulf War: Effects of toxicant exposures during deployment (2015, Cortex)

 

[Simon M]

The proposal is that it may all make sense if we insist on making all the details fit. That is how we came to an answer for RA.

I'm assuming the RA/Lupus/omplement example also ties in with your point about making all the details fit.

The problem here is that Dubbo is mostly about people who will turn out not to have ME/CFS but PVF. I don't know whether they assessed time delay to onset of fatigue. My information on this comes partly from asking individuals but mostly from the poll I did here.

I don't think they looked at onset time to fatigue, but then I think there is a lot more to be gained from prospective studies if done better. In particular, can we distinguish those that go on to have ME/CFS from those that don't at an early stage?
Do you (or does anyone) have a link to that poll of fatigue delay?

Yes.

[re "are you proposing a single mechanism]
Thanks for confirming. Many researchers have argued that the wide variation of presentation is due to it being a collection of different illnesses. What would be this model's explanation for the variation?

I just think that maybe malaise needs to be looked at more carefully.

Given its role in ME/CFS, that would be helpful.

You would have to ask Margaret Mar who got her ME/CFS from a sheep dip. She is a peer of the realm so her evidence does not require a formal citation!

Ha! I do remember her mentioning it publicly now.

 

[forestglip]

Do you (or does anyone) have a link to that poll of fatigue delay?

https://s4me.info/threads/was-there-a-gap-between-trigger-and-onset-of-your-me-cfs.40505/

 

[Jonathan Edwards]

Many researchers have argued that the wide variation of presentation is due to it being a collection of different illnesses. What would be this model's explanation for the variation?

To me this variation in presentation is perhaps similar to that seen with tuberculous infection (a huge variation) where you have a common type of problem that can manifest in all sorts of ways dependent on maybe genetic or contextual factors. With autoantibodies, each antibody is consistently associated with a problem of a different nature. There are caveats to this, yes, but it is the rule.

If there are a range of genes associated with synapse maintenance the cover risk for ME/CFS each one might skew the clinical pattern in a slightly different way. And if there are several genes linked to T cell function, the same. There might also be a major bimodality between cases with mostly brain risk genes and those with immune risk genes but that remains to be seen.

 

[Hutan]

The problem here is that Dubbo is mostly about people who will turn out not to have ME/CFS but PVF. I don't know whether they assessed time delay to onset of fatigue.

Here's the Dubbo Study Results:

Results
Prolonged illness characterised by disabling fatigue, musculoskeletal pain, neurocognitive difficulties, and mood disturbance was evident in 29 (12%) of 253 participants at six months, of whom 28 (11%) met the diagnostic criteria for chronic fatigue syndrome. This post-infective fatigue syndrome phenotype was stereotyped and occurred at a similar incidence after each infection. The syndrome was predicted largely by the severity of the acute illness rather than by demographic, psychological, or microbiological factors.

People were assessed against the Fukuda CFS criteria at 6 months by clinicians familiar with (ME)/CFS - 28/29 of the people with the 'post-infective fatigue syndrome phenotype' were assessed as having CFS. We have seen that PEM is hard to diagnose. I think it's reasonable to assume that the Dubbo study tells us a lot about ME/CFS.

On the question of time delay to the onset of fatigue:

The 94 family practitioners and all four diagnostic pathology laboratories that serve the region cooperated to provide us with coded reports of all IgM positive serological results indicating acute Epstein-Barr virus, Q fever, or Ross River virus infections. Patients aged 16 years or over, who provided written informed consent, were enrolled through their family doctor. We excluded patients who had symptoms present for more than six weeks or reported pre-existing medical disorders or drug use likely to be associated with prolonged fatigue. After the baseline assessment, we followed up participants at three weeks, six weeks, and three months, after which we further evaluated a matched case-control series (see below). We reviewed all enrolled participants again at 12 months.

We classified participants as provisional cases of post-infective fatigue syndrome if their SOMA scores at all time points up to and including three months exceeded the established threshold score.16 We invited these cases, and control participants matched by age and sex who had recovered promptly from the same infection, at six months for a medical interview, examination by a physician (AL), and laboratory investigation to exclude alternative medical explanations for ongoing symptoms...Where appropriate, AL and IH diagnosed the chronic fatigue syndrome (termed here confirmed post-infective fatigue syndrome) by consensus at six months after the onset of symptoms, according to the international diagnostic criteria.1

It's a bit hard to tell. There was a gap between the test for the acute disease and the baseline assessment, with the serological results coming through, the person being enrolled in the study and arrangements being made for the baseline assessment in between. There's an indication that the researchers tried to get the baseline assessment done within six weeks of acute illness onset. The requirement that people classified as having PIFS had physical symptoms at all times from the baseline assessment to 3 months, suggests that the researchers didn't really see participants having a delay in the onset of symptoms after the baseline assessment).