The following are comments and questions from members of the public, posted on Prof Ponting’s blog when it originally appeared on a previous MEGA website, and his replies.
Thank you, Professor Ponting. This is very interesting, especially, as you may know, that there’s a lot of interest in the mitochondria in ME/CFS, following the recently published work of Professor Robert Naviaux in the US. To what extent could MEGA shed light on these kind of hypotheses which are a level up from metabolites/DNA etc. (i.e. at the cellular/organelle level, rather than the molecular level)? Which do you think is the most relevant level to look at in terms of determining the underlying mechanism(s) of a disease such as ME/CFS, and in finding effective treatments?
What sort of timeframe could MEGA expect to make those sorts of discoveries? Given that it’s a 10,000-patient study, would we have to wait years until all those patients were recruited before we’d have useful results? What’s the reason for having so many patients when Prof. Naviaux and Prof. Ron Davis at Stanford are looking at tens of patients and finding interesting things?
I’m sorry that that’s rather a lot of questions but this kind of study is new to me (and I’m sure to a lot of us) and as patients, our focus is on speed! Thank you for your interest in our disease – it’s very much appreciated that someone of your calibre wants to be involved.
Reply from Prof Ponting:
Thanks for your great questions. The problem that I have with this ME/CFS study is that it doesn’t tell us whether these “metabolic features” are acquired because of illness, or whether they reflect its underlying cause. In Tweets I’ve read that some people with ME/CFS are relieved that these features might be diagnostic of your disease. Nevertheless, my own thoughts are that we need to find out the (likely several) root-causes of ME/CFS so that therapy not just for symptoms, but also for causes, can be developed.
Unfortunately, studies that do this require very large numbers of patients. This is because they need to look across all parts of their DNA which vary in different people to test whether DNA differences cluster in people with ME/CFS. And lots of people’s DNA is needed so that this clustering is not found just because of random chance.
Yes, the timeframe is long-term, certainly many years and I cannot forecast more precisely when, because the study (and its funding) are not in place. Why should we embark on this long-term project when interesting studies, such as Prof Naviaux’s, can be done faster? My answer to this is that we – as scientists – need to be humble about what could be causing ME/CFS: we should expect the unexpected, and not only test every potential cause of the disease – one by one – as they come to mind, but test *all* possible DNA causes, including those whose biology is entirely unexpected. (Although we know more about fundamental biology than we did a decade or two back, there is so very much more that we still have to learn.)
So I think the value of the MEGA study is that it seeks underlying causes, by being objective (in scientific jargon: “hypothesis free”) and being as comprehensive as possible, whilst knowing that there are likely to be several underlying causes. It is likely that many of these causes will be environmental – not from DNA changes. Yet, when we find DNA-based causes these will – I strongly believe – lead us quickly to what these environmental causes are. This is probably the best way of finding environmental explanations of disease, because scientists are unable to easily compare environmental exposures between people with ME/CFS and those without.
I hope this helps. One last thing from me, if I may: Scientists are people, just like anyone else. The reason I started doing biology research was the hope that I could make a difference to people and, in large part, those people (friends and family) around me whose lives have been blighted by disease. There would be no better – and more thrilling – outcome of this study than its definition of the causes of this debilitating disease.
“Hypothesis free?”I understand that some people argue gathering data – number crunching – is hypothesis free (and this has been debated), but what one does with the data and how it is interpreted afterwards is inevitably underpinned by hypothesis, which may be implicit rather than explicit. Steven Wiley has written well about this.
I have huge concerns about this project especially bearing in mind the mega data research used to inform the Occupational Guidelines in the UK on CFS which had and still has a psychological bias (it excluded all research into ME but still applies to people with CFS / ME). A classic example of what you look for is what you get and misuse of mega data research. We need more robust biological research especially at time when research is showing up clear results using clear patient criteria (so we know the same disease is being studied). This study would take up valuable resources which could be better used elsewhere.
I’m confused because the psychological research (I’m assuming you mean PACE and similar) was not a big or mega-data study by any means. They were clinical trials. This is using big data in the sense of omics studies generating vast amounts of data eg gene SNPs on each individual in the study. No relation to the kind of study you are referring to.
The guidelines are for the Occupational Medicine Physicians drawn up in 2004. Full title is “Occupational Aspects of the Management of Chronic Fatigue Syndrome: a National Guideline.” It is applied to employees with ME / CFS – the summary leaflets for patients and employees use the term CFS / ME. It used a mega literature review similar to the one being cited for this research.
It highlights the importance of what you are looking for, what you exclude and the difficulty of challenging something once it is out there as “robust” i.e. large numbers. No research is neutral. It has a starting point and an end point abd ask questions along the way. These guidelines- totally biased towards CFS ( ME) being a psychological illness – are still being used.
Elaine, a literature review is a very different thing to what is being proposed here. It looks at research already published and if the research it looks at is already biased, the conclusions it comes to will also be biased.
This, however, is a completely new study of thousands of new patients, which aims to look at the genetic variants present in people with ME/CFS and hence uncover, as Chris says above, ‘clues about what physiological processes have become dysfunctional and in what cells.’ He also says, ‘I would be keen to apply my group’s long-standing computational and experimental experience in these areas to defining the *molecular pathology* of CFS/ME.’
Research that aims to uncover physiological processes and molecular pathology *is* biological research. Assuming that the data that goes in is good (the right patients are chosen and there are enough of them) the data coming out will also be good – and incredibly useful to biological researchers all over the world. Also, rather than ‘taking up valuable resources’, research like this could help attract new researchers (and hence funding) into the field. That’s what happens when big studies like this produce results. Other researchers become interested.
As you say Rosalind, assuming the data that goes in is good and the right patients are chosen. But that is the issue here isn’t it. Unless severe and very severe ME patients are properly represented then the data going in will not be ‘good’. It will as usual be skewed towards the merely fatigue side of things, not CFS and certainly not ME. And, for a number of reasons, with the current structure of MEGA the severe and very severe will not be properly represented. Which means that from the off the data cannot possibly be good.
‘It will as usual be skewed towards the merely fatigue side of things, not CFS and certainly not ME.’ I don’t see how we can possibly make that judgement. Firstly, we’re not the ones diagnosing the patients or meeting them at the clinic. Secondly, we don’t know what ME is anyway. We only know that people with it have certain symptoms in common. Thirdly, just because someone is capable of getting to a clinic doesn’t mean that they don’t have ME. I do understand your concerns about the more severely affected not being included. I, too, would like to see that happen. But I don’t think it’s fair to dismiss anyone who can and does attend a clinic as having an illness/problem that is ‘certainly not ME’.
Margaret, why will the severe and very severe not be represented? The N.H.S. clinics M.E.G.A. is proposing to draw it’s population from can and do deal with the severe and very severe element of C.F.S./M.E. You don’t have to physically attend a clinic in order to be under that clinic.
Most NHS clinics do not deal with severe, very severe or long term ME patients in anything like a quantity that truly represents their proportionality. If recruitment is only from NHS clinics then they will not pick up the vast majority of severe, very severe or long term ME patients. That is a fact known by the vast majority of severe, very severe and long term ME patients, you only have to look on social media to ascertain that. I know it from personal experience as do all the other ME patients that I personally know, none of whom live in the same NHS area as I do or each other.
This fact will inevitably leads to the study being skewed further towards the merely fatigue side of things that past studies, most notably PACE, always tend to be. I don’t see how it can be otherwise unless the 25% severe and very severe and the long term ME patients are proportionally included as 25% of the study, and the vast majority of us don’t attend NHS clinics and those that do are usually very quickly signed off straight back to their GPs with zero follow up. This is something I and the ME patients I know have experienced and you here it very frequently voiced on social media. I don’t know which particular clinics you are referring to Neil, but that is not the experience of the vast majority of severe, very severe and long term ME patients.
Rosalind, you have misunderstood my point. I didn’t actually dismiss individual patients at clinics as not having ME. What I said was if selection is only from NHS clinics then the study participants will certainly not be a true representation of ME because most more severely affected patients will not be found there. The reason I say certainly not ME is because the vast majority of ME clinics only offer CBT and GET. We know as ME patients that this can only help depression or fatigue. Yes CBT can help you come to terms with living with any chronic illness. It does nothing for true ME and can all too often make it worse, particularly when it is based on blaming the patient for having false illness beliefs, which it all too often is. Therefore NHS clinics do not represent a true picture of ME, they are skewed more towards the fatigue and depression side of a very broad range of fatigue based illnesses and conditions.
Elaine, you are referring to a review of literature – completely different from this new study which will generate a warehouse full of biological data on metabolites, gene variants, gene expression and proteins.
Simon, thanks for replying. Yes, it was a mega data review of literature. But the concerns are still appropriate – the basic (unstated) hypothesis, what is omitted from the research being an important component: the broad definition basis etc. I’ll leave it at that. I don’t support this study. I hope that all involved will consider carefully the mistakes made from previous research.
Thank you for this information. As a long term ME sufferer I am delighted that scientists of your calibre are expressing an interest in studying ME. I am however concerned that such a large scale project will take many years to come to fruition and be prohibitively expensive.
I am also concerned that the method for locating patients to include in the study is to use fatigue clinics. While this would speed the process, it would exclude the moderate and severe sufferers and long term sufferers who do not use these clinics, giving a skewed sample. I am sure this is anathema to a population…See more
Reply from Prof Ponting:
Thanks for your comments. Yes, this project would unfortunately take several years to come to fruition. I’m not sure that it would be prohibitively expensive, however, because funders such as the Wellcome Trust, Medical Research Council and the National Institutes of Health (in the USA) have previously funded similarly sized projects. It should also be said that any funds given to this project are very unlikely to redirect (the limited) funds currently available to other studies.
I’m not an expert in patient recruitment for such studies butI think you’re right that the most severe sufferers would be excluded, but as I understand it there are many moderate patients in NHS clinics. This is not a show-stopper because what we learn from the more moderate sufferers is likely to be highly relevant to the more severe sufferers. Or at least this is what is seen in other studies.
Unfortunately the 300+ ME biobank data is just not enough with which to explain the genetics of ME/CFS. If it were, then we would immediately do the study. I entirely agree: “Patients have waited too long already. We need answers soon.” I do hope that we can get this study funded and running ASAP.
Finally, I have always collaborated and shared data with scientists from across the world, and am a strong advocate of open consented data. In particular, it is clear that the MEGA study would have to work alongside the USA based studies.
Thank you Professor Ponting for such a rapid and thoughtful reply. I am a supporter of fundamental research, having done a genetics degree myself (nearly 50 years ago) and do understand the need for a large sample. I wonder whether this could be speeded up by involving clinics in other countries.
I have to admit I’m not reassured on the lack of severe and very severe sufferers. It seems likely to me that if there are genetic subgroups within ME, those at the severe end of the spectrum may well be genetically different. The only way to find this out is to include them. A call could be put out via the patient groups to find sufficient severely effected and long term sufferers to participate. I for one would be happy to do so.
Like others, Professor Ponting, I am extremely grateful to you and any other biomedical specialists who want to study this extremely debilitating condition. However, like others I am also very concerned about the lack of representation of severe and very severe ME. As an ME patient of over 20 years whose condition is steadily worsening, my feeling is that the inclusion of severe ME is paramount to this study.
Without it you are not really studying ME, just CFS or even just fatigue depending on just how diluted the participant criteria becomes. Not that I am saying fatigue does not need studying, but please don’t conflate ME with fatigue. They are very very different and fatigue has been studied far far more than ME. It is the 25% that are desperate for such biomedical research as they have never been properly represented to date in any research.
This is partly because severe and very severe ME patients are very costly and time consuming to include in studies due to their being housebound (like myself) or all too often bedbound. However, as they represent around 25% of the ME population, without them being 25% represented then you are not really studying ME, just fatigue.
Please, I would ask you and your fellow researchers to take this concern, that I am sure many ME patients share, on board. I have to say that whilst I am really keen to support a biomedical study into ME, I cannot support one that does not properly and proportionately represent the 25% severe and very severe ME. As I have said anything else is simply yet more study of fatigue.
I look forward to hearing if the MEGA project has any proposals to represent the severe and very severe ME patients that make sense to the ME community as I can assure you this is far more likely to gain our support.
Again Prof Ponitng, I would like to express my gratitude in your interest and also your throughtful replies. My main concern is still patient recruitment. The clinics do not see many moderate patients either who are mainly hosuebound too or may only be at the clinics for a short time. Would it be inconceivable to also put out a call to patients to proffer their own inclusion if they can give a Drs / consultants diagnosis? Just as the ME biobank did? I can assure the word will spread quickly around the UK ME community and would get to the moderate and severe sufferers.
You suggest that we can learn a lot from moderate patients which can apply to severe patients yet this study seems to suggest that there were clear differences in findings between moderate and severe patients. Is this therefore relevant to your work?
I fully agree there are no severe and very severe patients in NHS clinics – but there are plenty of moderate patients. If you look at PACE, for instance, which recruited from NHS clinics, the average SF36 physical function score (a widely used measure of disability) it’s about 37/100 which is at least moderate. Over 90% of the healthy working age population score at least 85/100, most score 95 or 100/100.
Simon, I understand although I’m not sure how effective a tool it is for working out ME severity. Someone with Mod-Severe ME will have the same score more or less as someone with very severe ME. It is very useful to workout how stark the differences are between ME and othe conditions though. I class myself asthe severe end of moderate, I am housebound most of the time and have lost a lot of function and ability. However, in comparison to my friends who are at the very severe end I am doing well and they would love even the little I am able to do.
Simon O commented:
Thanks for your blog and answers to comments, Prof Ponting, and thank you for your interest in helping us. I’m just wondering how many patients would be needed to make the study worthwhile. You say 300 is not enough, but do we really need as many as 12000?
Reply from Prof Ponting:
Unfortunately the patients’ sample size really does need to be much larger than 300. The first reliable studies used data from 14,000 individuals. It is possible that a strong signal would be found with, say, 3000. But funders (and ourselves) need larger numbers to be convinced that there will be value for money: that something will be found in this study.
Simon O commented:
Many thanks for the info which is much appreciated. So it seems that even 12,000 would be a bit on the low side? Bearing in mind that this sample will be *very* heterogeneous, embracing numerous subgroups which are believed to constitute neurological ME plus patients with other fatigue conditions (including depression, burnout etc) which will be included in the extremely broad NICE criteria, I just wonder if this is really a viable study even at 12,000 patients? Sorry to be negative (and I really appreciate you and the other researchers wishing to be involved in this) but patients are concerned. Oh, yes, and the 12,000 will apparently include children as well, which again seems a lot to add to the mix. (Would it be usual to include adults and children together like this ?)
Reply from Prof Ponting:
The selection of individuals for this study is critical and all that I have heard on this makes me believe that post-exertional malaise criteria should lie at the heart of this. The patient advisory groups that will be set up need to have a strong voice on this issue.
Is this study viable at 12k patients and given the heterogeneity? We do not know – which is why the results of this research are needed. We could go round-and-round on this issue, but until we have data/evidence no-one will know. CFE/ME may present clinically in a heterogeneous manner, but this might belie a less heterogeneous set of genetic contributions.
For example, are adult and adolescent CFS/ME the same or different? My guess is somewhere in between and, if so, understanding one will help to understand the other. Even if a 12k study does not produce clear results, when combined with other studies’ results (e.g. from the USA), it might yet. I think it worthwhile to attempt this study: scientifically it is the right thing to do next, although as with any study there is no guarantee of success
Thanks for the information, Dr. Ponting. This is all very interesting. A couple of questions for you: One is, how likely is it that there is a genetic signal at all? And if there is, will it necessarily point us to the root cause? I don’t intend those questions skeptically — I just don’t know the answer. Looking at other diseases, do they have genetic signals? MS, diabetes, heart disease, cancer, ALS (particularly the type that isn’t strongly hereditary)?
Also, I would think that the need for a large sample would also mean that it would be especially important to have that sample be well defined. Wouldn’t including someone who is very tired but doesn’t have ME/CFS at all be worse than useless? Both a waste of money and a dilution of the signal? It seems like the argument is being made that it’s actually good to have some of those folks, and I don’t understand that at all.
Also, I understand that the ME biobank does not currently have enough samples for this type of study, but would it not make sense to fund them to collect more samples rather than starting again from scratch? The ME biobank team seem to have the complete confidence of patients and have a system in place for collecting samples from severely affected patients. My perception is that an alliance between the ME biobank team and the MEGA scientists would have far more support than the current group, which includes some people in whom patients have no confidence for reasons which are well documented.
Reply from Prof Ponting:
Julie, Thank you for these questions. It is highly likely that genetics will contribute, but clearly environment (e.g. infection) contributes too. Here is an analogous situation: of all who are infected by the flu virus, you are far more likely to be hospitalised if you have a rare change in your DNA (IFITM3 rs12252 polymorphism).
Yes, adding a person who clearly does not have ME/CFS will reduce the study’s power. This goes back to the dialogue that is needed on the inclusion criteria which should include input from people with CFS/ME (see previous comments). Looking in from the outside on this debate, however, I think that there is a risk here: specifically, because (even using criteria e.g. PEM/Canadian) there is no consensus / strong evidence-base about who should be included or excluded then this issue will delay any evidence being acquired. And I do understand the urgency of gathering evidence sooner rather than later to better understand CFS/ME. With the genetic data to hand, it is plausible that the genetics signals will help separate patients according to their different underlying causes.
The science is fascinating. Could you give an example of where this kind of approach has led to insight into disease mechanisms? Thanks very much
Reply from Prof Ponting:
Thanks Simon. One study used 1,924 type 2 diabetes patients, and then a further 3,757 patients for replication, to find that if someone has two copies of a DNA change then they weigh – on average – 3 kg more.
A detailed -omics study showed that this is likely due to altered regulation of heat generation in fat cells. This is a large effect; for a smaller effect to be detected more individuals in the cohort would have been needed
Thanks very much, Chris.I’d encourage people to look at that paper on heat generation, because it involves mitochondria, which may play a role in mecfs, plus it was in the top medical journal, the NEJM – and used some elegant science including gene editing with CRISPR. I realise that things might now work out so neatly with this illness but it looks to me like a pretty impressive example of using genetics and other technologies to likely get at causal mechanisms.
Thank you, Prof Ponting for your careful explanations and for taking the time to answer so many questions. Regarding patient selection, you said: ‘The selection of individuals for this study is critical and all that I have heard on this makes me believe that post-exertional malaise criteria should lie at the heart of this.’
I would agree with this wholeheartedly and I think the vast majority of the patient community would, too. I would add that I think a major concern of many of those who have posted here is that ‘post-exertional malaise’ is defined as something more than ‘feeling unusually tired afterwards’. In other words, that it includes symptoms such as nausea, dizziness, weakness etc. However, that, in itself, raises all sorts of other questions.
My first question would be: If post-exertional malaise is foundational to ME/CFS, it implies that something significant happens in the body after exertion. For example, I’m fairly sure I’ve seen a study that looked at inflammatory markers before and after exertion in ME/CFS and found differences. So do you plan to collect samples before exertion, after exertion or just randomly? Does it matter?
One reason I ask is because, obviously, there are ethical issues involved in researching post-exertional malaise, particularly where the severely ill are concerned. Many patients seems to be expressing concerns that the severely ill are not being included in the study. But, there will be some for whom even the taking of a blood test counts as ‘exertion’. If we can get equally meaningful results from patients who are less ill, then, from an ethical standpoint, that looks to me like the better way to go?
My second question would be: What of other illnesses that manifest in a similar way? It’s frequently touted that the post-exertional malaise of ME/CFS is unique to the condition. But is it? Do we know this for sure? Is it likely that some patients will have other undiagnosed or subclinical conditions? If so, are you likely to find them?
As an example: 15 yrs ago I went down with a text-book case of ICC ME/CFS and was in bed/housebound for many months. Over the following years, my condition slowly improved. Then, about 4 years ago, it began to deteriorate again. As far as I could tell, the symptoms were largely the same. I just found I could do less before they impacted me. About a year later, a chance blood test revealed that I had Pernicious Anaemia. By then, my B12 levels were rock bottom and I was living in constant fog. Since then, my condition has improved again, though not to the level that it was a year prior to the new diagnosis.
This is interesting to me because I now know that subclincal forms of PA exist – where the B12 deficiency doesn’t show up on the standard blood test, but other tests can find it. Also, that B12 seems to have helped a proportion of patients with ME/CFS, but not others. Finally, I know that B12 and folate are essential to both energy production and the neuro-immune system. In other words, you’d expect PA patients to look a bit like ‘ME/CFS’ patients and vice versa.
So, why, oh why, do the NICE guidelines for ME/CFS state: ‘Tests for vitamin B12 deficiency and folate levels should not be carried out unless a full blood count and mean cell volume show a macrocytosis.’ ?!!!!!! In the absence of thorough testing, how is either patient or clinician supposed to tell the difference? Further, in a case like mine, are we talking about one illness or two? If two, how are they related? Did one trigger the other? Or is there some other causal factor that is common to both? Similarly, a friend of mine has PA and FM. So are they linked? What about FM and ME/CFS?
As far as I am aware, there isn’t anyone who can answer these questions. As you say, we are talking about complex systems and relationships and categorising people by symptoms alone can only tell us so much. But that’s precisely why I am excited by a study such as this. If the patients are well phenotyped, there is so much that it could tell us, particularly if the results are (eventually) compared with other patient groups (including the 300 ME/CFS patients for whom we already have samples).
If it was important that people be NOT in a state of PEM for the blood testing this could be specified. The numbers for who taking blood itself would be exertion are quite small portion of the severely ill (which means largely housebound ) the weakest are at deaths door function but aren’t a reason why the severe and many very severe can’t participate IMO.
There has been the research quoted above showing differences in the moderate and severe and indeed the question of why some push and get incapacitated and others don’t might be in the genes too?. As someone severely ill years and generally neglected in ME research I wouldn’t be satisfied with low representation here when the nature of the research means that is something many severe could participate in long as attempts to go the extra miLe and accomodate were made
As a side note, this might be relevant. I remember hearing Prof Ron Davis saying at the UK invest in ME conference that his severely ill sons tests, which I think were metabolomic, were way off scale. If this study is going to involve similar testing then the severe might yield staggering findings as they are so ill.
‘…The severe might yield staggering findings as they are so ill.’
I think that would almost certainly be true and I would like to see the severe included for precisely that reason. However, I have a friend for whom the taking of blood would be an issue, so I think that’s also worth bearing in mind.
Meanwhile, in moderate cases, I suspect that the results would be a lot more telling if the tests were done after ‘exertion’, since this seems to be what distinguishes PWME from ‘controls’. Certainly, that’s when I feel most ill. But it does depend a bit on what one is looking for, which is why I asked about it.
I am still stuck on the severe patients issue. If we are not investigated then we cannot be helped. As Prof Holgate suggests there may be as many as 15 different conditions.
What if those of us with severe ME have a different condition to those who don’t deteriorate and never become severe?
It is vital that we are studied. It is those with severe ME who are deteriorating and dying after all and are left with the most awful symptoms with very little help.
3,000 severe patients need to be included to represent 25% . If that costs too much then why not start with us and include milder patients as finances permit. Is anyone from MEGA addressing this issue?
Interesting stuff is coming out using these tools in the USA when they use well characterised patient cohorts. It would be wonderful if the Uk did this as well but this proposal for 12,000 people with any type of fatigue is NOT a study of the neurological disease they the USA is looking at.