CFI Spinal Fluid study from Lipkin and Hornig is out.

[Antares in NYC]

On Twitter.

"Interesting"? Funny anecdote, my artist friends tell me that one of the lamest things you can say when commenting on their work is "it's interesting". It's akin to saying "I don't get it" or "I have nothing to say about it".
I think it applies here.

 

[SOC]

"Interesting"? Funny anecdote, my artist friends tell me that one of the lamest things you can say when commenting on their work is to say it's "interesting". It's akin to saying "I don't get it" or "I have nothing to say about it".
I think it applies here.

Or, "I think it's crap, but I don't feel it's in my best interest to say so right now."

My guess is that W can't afford not to comment, being the UK's expert on CFS/ME and all , but knows he can't acknowledge it's sound research without destroying his life's work. He's in a bind -- he can't say it's good research, but he can't knee-jerk dismiss sound research from a researcher as major as Lipkin. He needs more time to dig down and look for ways to pick it apart because there's no obvious flaws. So, it's "interesting." It will be "interesting" to see where he decides to stand on this research after he's had time to scheme with his cronies.

We'll know we've got them running away with their pants around their ankles when we start hearing stuff like, "We were using the best information available to us at the time." We all know that's a crap excuse for their abuse and willful ignorance, but "I was doing the best I could at the time" is a classic excuse used by abusers to try to get out of responsibility for their earlier actions. Sadly, a lot of people in the mainstream will buy that excuse and W and Co will get off relatively easily.

 

[adreno]

We'll know we've got them running away with their pants around their ankles

Don't get all worked up, now! This isn't going to happen for a while

 

[Jonathan Edwards]

It surprises me that there are no standard comparison sets from other trials and that each trial has to test and match controls. Especially when the controls and samples seem to have been collected and stored in different ways.

I think the problem with cytokine assays is that they are so temperamental that you absolutely have to test your controls in the same batch, hopefully under a blind coding.

 

[jimells]

I think the problem with cytokine assays is that they are so temperamental that you absolutely have to test your controls in the same batch, hopefully under a blind coding.

Does this mean that cytokine tests regular patients might have access to are not reliable?

 

[Jonathan Edwards]

Does this mean that cytokine tests regular patients might have access to are not reliable?

I cannot be sure but I would pretty much assume so. The figures people quote for their IL-8 levels on PR bear no relation whatever to what Mady Hornig found in blood so something is badly wrong somewhere. Cytokine assays have been pretty dodgy for the last twenty years and last time I spoke to Mady she agreed that unless you have tight comparison in a research lab they are pretty difficult to interpret. All you need is for the blood to hang around a bit in the post or the reagents to be a bit old and you may get nonsense. When I worked on assaying for TNF in 1990, after a few months we concluded all our results were nonsense and we moved on.

 

[Kati]

I cannot be sure but I would pretty much assume so. The figures people quote for their IL-8 levels on PR bear no relation whatever to what Mady Hornig found in blood so something is badly wrong somewhere. Cytokine assays have been pretty dodgy for the last twenty years and last time I spoke to Mady she agreed that unless you have tight comparison in a research lab they are pretty difficult to interpret. All you need is for the blood to hang around a bit in the post or the reagents to be a bit old and you may get nonsense. When I worked on assaying for TNF in 1990, after a few months we concluded all our results were nonsense and we moved on.

Indeed, I have a cytokine pannel from early on in my illness, with fresh blood, by Dr Fletcher's lab. In Miami and my Interferon Gamma, which is supposed to be elevated, is in the lowest of the lows. How do you make sense of all that? It's very hard to interpret and likely you need to take a look at the whole picture for each individual.

 

[Tuha]

Will be there a small explication for non-scientific brains? Are these findings important/interesting? Do they confirm previous immune studies?

 

[Kyla]

Will be there a small explication for non-scientific brains? Are these findings important/interesting? Do they confirm previous immune studies?

http://simmaronresearch.com/2015/04...ramatic-differences-chronic-fatigue-syndrome/

 

[user9876]

I think the problem with cytokine assays is that they are so temperamental that you absolutely have to test your controls in the same batch, hopefully under a blind coding.

That makes it all sound very dodgy. Does this mean that absolute measures of cytokines become largely meaningless but relative measures with in the batch are valid?

 

[jimells]

I've always found it interesting that medical lab reports never include error bars. Do the labs assume doctors can remember the margins of error for each test, or that the margins of error are so low they don't matter to clinicians?

Lab report footnotes like "Add x% for African Americans" doesn't give me much confidence in the precision of the test. What if the patient has one European parent and one African parent - add 1/2 of x% to the lab result?

 

[Jonathan Edwards]

That makes it all sound very dodgy. Does this mean that absolute measures of cytokines become largely meaningless but relative measures with in the batch are valid?

It is a long time since I actually tried to measure cytokines myself but I suspect yes - absolute measures are probably pretty useless. The variation in the complex chemistry that occurs while doing the test is likely to be bigger than the variation between true normal and abnormal samples. I think the basic problem is that the amount of cytokine in chemical terms is so tiny that the test relies on a massive amplification system.

The idea of 'error bars' mentioned by jimells is interesting. Certainly medical biochemistry labs tend not to give any indication of the consistency of repeated assays - which I think is the issue. Error bars given in scientific papers do not, of course, have anything to do with this. They are a technical measure that is only concerned with the likelihood that a statistical comparison can be made with a certain degree of precision. They actually tell us nothing at all about variability of the assay because the figure is divided by the number of data points. In simple terms, in my book, error bars are a complete con because people think they give some guide to variability and they do not. They always look conveniently tiny. Scatter plots reveal the real truth - dots all over the place!

 

[halcyon]

The variation in the complex chemistry that occurs while doing the test is likely to be bigger than the variation between true normal and abnormal samples.

If that were the case wouldn't the results be stochastic? Instead they are finding patterns, and the patterns for patients match, and differ from controls. Sure, maybe if the patient samples were all tested one day, and then the controls a different day, or by a different lab, you might expect some oddities, but I doubt this is the case. Dr. Lipkin isn't some amateur scientist. I highly doubt he would put his name on a paper with dodgy science, especially in such a controversial area of research.

 

[Daisymay]

If that were the case wouldn't the results be stochastic? Instead they are finding patterns, and the patterns for patients match, and differ from controls. Sure, maybe if the patient samples were all tested one day, and then the controls a different day, or by a different lab, you might expect some oddities, but I doubt this is the case. Dr. Lipkin isn't some amateur scientist. I highly doubt he would put his name on a paper with dodgy science, especially in such a controversial area of research.

Quite so. And I would have thought that Molecular Psychiatry would be extremely careful of what they publish on ME/CFS.

 

[Jonathan Edwards]

If that were the case wouldn't the results be stochastic? Instead they are finding patterns, and the patterns for patients match, and differ from controls. Sure, maybe if the patient samples were all tested one day, and then the controls a different day, or by a different lab, you might expect some oddities, but I doubt this is the case. Dr. Lipkin isn't some amateur scientist. I highly doubt he would put his name on a paper with dodgy science, especially in such a controversial area of research.

I think you have missed the line of argument. I was talking about cytokine assays done one off in diagnostic labs - that is what was asked about. I am not criticising this paper or the blood paper from Hornig. They run samples alongside controls in the same batch.

The effects of the unpredictable shifts in the assay chemistry from run to run will not actually be random. (Whether in a routine lab or on different days in a research lab.) They will be unpredictable but not Gaussian. Some days all the results will be low because a reagent is old. Some days they will be high because the weather was warm - and the effect will be systematic, on all samples. This has nothing to do with being an amateur scientist. It has to do with the fragility of the chemical process when looking at tiny levels of signal. Any good scientist will assume that control and test samples have to be run alongside each other on the same plate.

 

[cigana]

I cannot be sure but I would pretty much assume so. The figures people quote for their IL-8 levels on PR bear no relation whatever to what Mady Hornig found in blood so something is badly wrong somewhere. Cytokine assays have been pretty dodgy for the last twenty years and last time I spoke to Mady she agreed that unless you have tight comparison in a research lab they are pretty difficult to interpret. All you need is for the blood to hang around a bit in the post or the reagents to be a bit old and you may get nonsense. When I worked on assaying for TNF in 1990, after a few months we concluded all our results were nonsense and we moved on.

The levels quoted here on PR are virtually all from RED Labs in Belgium. I can't vouch for their accuracy, but what I do know is that the blood is taken from KDM's clinic and gets to the lab usually the same day or next day, and they are well aware there of the need for proper storage (the lab and clinic are in the same city, so it is convenient, and so many samples are taken on a daily basis the process is consistent from one week to the next). That might explain some of the differences, as it seems like a better set up than any of the "stored" results that appear in the literature (I don't know, just FYI).
I also know KDM tends to alternate between testing serum proteins and mRNA expression of cytokines in the same patient in case of any differences (I have not seen any in the small ~8 number of reports I've seen).

Something that I always wonder about: perhaps it doesn't matter what the absolute concentration of cytokines is reported to be. In a way, in terms of a marker, all that matters is that a certain lab can consistently differentiate between a PWMC and a control?

 

[Jonathan Edwards]

The levels quoted here on PR are virtually all from RED Labs in Belgium. I can't vouch for their accuracy, but what I do know is that the blood is taken from KDM's clinic and gets to the lab usually the same day or next day, and they are well aware there of the need for proper storage (the lab and clinic are in the same city, so it is convenient, and so many samples are taken on a daily basis the process is consistent from one week to the next). That might explain some of the differences, as it seems like a better set up than any of the "stored" results that appear in the literature (I don't know, just FYI).
I also know KDM tends to alternate between testing serum proteins and mRNA expression of cytokines in the same patient in case of any differences (I have not seen any in the small ~8 number of reports I've seen).

Something that I always wonder about: perhaps it doesn't matter what the absolute concentration of cytokines is reported to be. In a way, in terms of a marker, all that matters is that a certain lab can consistently differentiate between a PWMC and a control?

As far as I am aware we are a very long way away from a lab being able to distinguish someone with ME from a healthy control. Even in the reports in the literature we are talking about statistical differences between populations. I have not yet seen anything like a situation where samples from PWME have been reported as consistently outside a 'normal range'. The diagnostic labs may give this impression but if their assays are that good at differentiating why have they not got papers in Nature saying so!!? There is a rather large credibility gap here.

 

[duncan]

That could be explained easily by the fact that most samples of PWME historically have been diluted with non-PWME serum or whatever.

There are probably diagnostics and metrics that work and are applicable, e.g, NK cell function, but because of polluted sampling, few sound studies have been able to demonstrate those.

 

[Sidereal]

NK cell number and/or function tend to be low in major depression. This has been known since 1980s. There are many old papers on this topic.

 

[duncan]

Perhaps. But that was offered only as an example.

My point stands: The greater history of testing ME/CFSers has been challenging because of lack of adequate control over who is included in the samples. That includes inadvertent sampling issues due simply to poor definitions and selection criteria.

So, it is very difficult to state concretely whether or not there are already tests out there that can discriminate between us and healthy controls.