Signs of Intracranial Hypertension, Hypermobility and Craniocervical Obstructions in patients with ME/CFS (Pre-print 2019/published 2020) Bragée et al

There are lot of changes and add-ons in the peer-review process, for example we have digged in to fibromyalgia concomittant diagnosis, and brushed up comparisons with expected, normal values of prevalance. The preprint hade some 2300 downloads, but as always, it is the peer-reviewed article that counts.

 

Thanks. Earlier in this thread based on your pre print there’s been a few pages of discussion. I was wondering whether you plan to do a study which is blinded and which will have control subjects?
Although you refer to what values in healthy people may be, there aren’t actually matched controls to see if these values actually do correlate to the healthy population.
The fact it is unblinded & doesn’t have controls is pointed out in this comment earlier:

https://www.s4me.info/threads/signs...h-me-cfs-2019-bragée-et-al.12553/#post-221315

 

As long as we study MRI findings and CSF samples and so forth, there is no need for blinding, it is not interventional or randomized. Of course there is a point to have a matched control cohort, we plan such a study, yes.

 

I think the point of the blinding is that bias could have an effect on how the scans are read by a radiologist, and the measurements that are taken.

 

In that way this study was blinded, no examiner/radilogist knew more than the scan he scrutinized. I thought you meant blinded to patient, normally used in clinical pharmacological trials.

 

I was more referring to the fact that the radiologist could be blinded as to who the person was and also, whether or not they had ME/CFS or were healthy. But this is only possible in a control-matched study.

In this study, the radiologist and those checking the scans presumably knew that the study was being carried out on ME/CFS patients (even if they didn’t know who exactly they were) - and knowing that the study is about structural abnormalities in ME/CFS patients, this is where the bias can come in.

 

I think it would be of the utmost importance as well to have more research where the researchers interpreting the findings are blind to whether those belong to patients or controls.

 

In next study we will use two radiologists, blinded whether it is patient or control. This time blinding was only on other findings.
The setting was sort of proof of concept, I was stunned by all hypermobile patients with clear neurological findings, earlier not described in their medical records and absolutely not in literature - very surprising. And all neck trauma patients.... I have always looked upon MRI scans my self, when I started study MRI in ME patients I was immediatly hit by all pathological findings too, often not mentioned. I have been in the pain field for three decades and had never seen anything like it. On the other hand, most of the patients had widespread pain, that was more normal to me.
That was the start of this retrospective study. We now have some 20 projects started or running or planned on different biomedical findings and treatments.

 

And the MRI findings are objective, I mean, its a question of millimeters and angles, not subjective impressions.

 

Yes, but the interpretations might differ from researcher, to researcher /clinician etc?

Check this video out by dr Bolognese. He talks about this at 20.45

https://www.youtube.com/watch?v=zUdwvBDnWpE

.

 

As we state - if you read the article - we made such a Inter Rater Reliability test with good results.

 

Thank you! I just have many questions. Great to have you here explaining about the study.

For me this is a bit of a difficult subject as I'm a former professional (male) ballet dancer. I'm saying male because I was rather on the stiff side compared to my friends. With my capabilities I would've never been a female professional ballet dancer. I was more a high jumper, pirouette turner and power dancer than flexibility.

Many of my friends/colleagues however were extremely flexible. They could easily fit the criteria of hEDS just like Dr. Bolognese suggests.

I now of course have certain (neurological) symptoms that would fit the clinical picture. But where do you draw the line with hypermobility and that they actually explain the symptoms of ME? They might be completely unrelated?

I'm looking forward to see if the healthy controls in the next study will have less pathological findings such as IH, cervical spine obstructions/tonsil position etc. that were according to this study more prominent in ME patients.

Maybe it would be even better to ask some dancers of the Royal Swedish ballet or Cullberg ballet to join the study.

I do think the IH is very interesting.

 

“MRI of the Brain
MRI scans of the brain were offered to all participants. MRIs were conducted in different labs and included T1- or T2-weighted scans with or without fluid-attenuated inversion recovery (FLAIR) with a section thickness of 3 mm. Scans were assessed by an experienced radiologist through the Regions MRI database, using Sectra® software, which offers built-in measurement tools. An interrater reliability (IRR) study of the MRI assessments was performed on 100 randomized participants for whom the radiologist assessments were compared to those made by a resident physician and a medical student.”

Again, it doesn’t look like any of the other people looking at the scans would be blinded either, since there was no control group, and so everyone involved knew they were ME/CFS patients and the study is related to structural abnormalities.

I am interested in the study you do in the future where the radiologists will be completely blinded as to whether they are in the ME/CFS group or the control group. And so compare, without possible bias, whether the measurements in ME/CFS patients are different to those in healthy controls.

 

Dear @Bjorn Bragee,


Thanks for flagging up your paper.


I appreciate your enthusiasm but I have to be frank and say that publications like this are included in what I regard as unwarranted and irresponsible. Retrospective studies with this type of methodology are guaranteed to be subject to major referral and ascertainment bias. The patient recruitment description in the paper is sketchy but, however it was done, I think we can guarantee bias if human investigators were involved. The resulting data, like the papers from Rowe and from Knoop’s group with similar problems, are certain to perpetuate the story of a relation between hypermobility/spinal issues and chronic illness. The consequences for patients may be very serious. As a peer reviewer (and as an editor) I would not want to see this published.

That may sound harsh but I draw on my own experience with the hypermobility story. I started the first hypermobility clinic in the UK at Guys Hospital in the 1970s with Rodney Grahame. I was asked to take over his NHS clinic when he retired in 2005. I have watched the story unfold. I never saw any evidence for the existence of a ‘hypermobility syndrome’ or ‘hEDS’ of clinical relevance. I have seen population-based studies from the UK and Scandinavia that indicate that there is no link between hypermobility and ME.

The rate of hypermobility in Rodney’s clinics was 100%. The rate of chronic fatigue in 1978 was 0% but by 2005 probably 60%. The reason is simple. His clinic took referrals from people who knew he was interested in hypermobility. Later people knew he was interested in chronic fatigue and hypermobility. You cannot publish data on tertiary referral clinic populations and imply that they are representative. This is the most basic mistake of any epidemiological or demographic study. You may argue that you think bias was slight but science is about providing data that is secure from doubts about such things. And the experience with CBT and GET in ME illustrates the fact that when bias can come in, it does. The ME community desperately needs more rigorous science. The situation is still a complete mess, with people investigating pseudoscientific treatments like the Lightning Process using invalid methodology. I worked in the field of rheumatoid arthritis where methodology of this sort was considered unacceptable.

If you are as concerned as I am about unwarranted surgery, I would make a plea to withdraw this paper. It may do untold harm and lead to deaths. Jennifer Brea is almost certain to capitalise on it.

 

I see from the thread on structural spinal problems that I am out of date.
This paper may cause serious harm.

 

This is a public thread and I've Tweeted a link to this post on my own Twitter timeline and RTd a copy to Julie Rehmeyer - I hope that's OK with you, Prof Jo?

 

There have been anecdotal reports and surveys that suggest that lots of ME/CFS patients have Craniocervical instability (CCI). This study does not really support this hypothesis.

The mean clivo-axial angle (CXA) was 148◦ ± 10◦ which doesn't seem that abnormal. It's fairly similar to what has been reported in the control group of some studies (for example this one). I'm not sure why a CXA of 150 should be considered abnormal, given that the range in healthy controls is usually somewhere from 135 to 175 degrees.

Instead of the CXA axis, the paper focuses on the finding that 80% of tested patients had craniocervical obstructions. The paper explains what this is about:

These obstructions were more frequent that have been reported in the general population but selection bias may be at play here. Not only are people with symptomatic obstructions more likely to reach out to the centre that is interested in studying this, but in only half of the participants (55%) was an MRI scan of the cervical spine performed. So the figure of 80 should be taken with a grain of salt.

 

Is there a link to these?

I’m interested because I think Ron Davis has also said that in the severe patient population from the study they did, 50% were hyper mobile (or had EDS?), but that was from a small sample.

 

Bit confusing this. The results section says:

But the discussion says:

 

The UK ME Biobank team found no evidence of increased hypermobility or EDS in their population sample but this is not published.

Dr Kate Armin (paediatrician) has a presentation that mentions A study by Seckin from Denmark which I looked up a while back. Essentially not much link between hypermobility and problems. There is also a Raine study from Western Australia that suggests that about 50% of females qualifying for hypermobility (basically what is loosely called hEDS now) is normal.

The Beighton score is a nonsense as far as I can see. A lot of people with genuinely problematic hyper mobile individual joints do not score particularly high on Beighton and each of Beighton's criteria fall well within normal.