New era for ME/CFS research as top cytokine study attracts media headlines

Simon submitted a new blog post:


Simon McGrath describes ME/CFS research presently in the media spotlight ...

The immune systems of patients who have recently developed ME/CFS look markedly different from those who have been ill for much longer, according to a major new study from Drs. Ian Lipkin and Mady Hornig at Columbia University. This shift in immune function hadn’t been seen before.


Study leader Dr. Mady Hornig in her lab

“Perhaps the most significant evidence yet that chronic fatigue syndrome has a biological basis”, said the Wall Street Journal. The immune signature discovered might eventually be the “basis of the first diagnostic test for the illness”, said The New York Times. The prestigious New Yorker magazine mentions the study too.

This feels like a new era of ME/CFS research: big, rigorous studies by top researchers and clinicians, with media interest to match. Many more such studies are in the pipeline, including Dr. Montoya’s related immune profiling work, the large Open Medicine Institute immune gene sequencing project, and the huge CDC multi-clinic studies.

If other researchers can confirm the finding that there is an important difference between short and long duration patients — an important qualification — this work could change how we understand, study and look for treatments in ME/CFS.


Doing it right

One of the most exciting aspects of this study was simply that it was big: nearly 300 patients and 350 controls. Most ME/CFS studies to date have had just a few dozen patients.

And it was designed meticulously. Patients were carefully diagnosed (mainly using Canadian Consensus criteria) by the clinics of leading U.S. ME/CFS physician-researchers including Drs. Montoya, Klimas, Bateman, Peterson and Komaroff. The study matched healthy controls and patients by age and sex, as both affect the immune system.

"The study is an important step forward in trying to track the biological basis of ME/CFS."
- Professor Jonathan Edwards

And they went further in trying to minimise anything else that might skew immune profiles. The immune system has daily cycles and is sensitive to stress, so all blood samples were drawn within two hours of midday and after a common mild stressor (the necessary patient paperwork!).

Controls were also matched by season of the year, as infections (even ones we aren’t aware of but our immune system are busy fighting) are more common in winter, as well as also by region, as infections vary by location.

Not so much difference between all patients and controls

The researchers, led by
Dr. Hornig, analysed cytokines — small messenger molecules that play a central role in regulating the immune system. Because cytokines are a window to the immune system, ME/CFS researchers have focused on them before, and numerous smaller studies did find unusual cytokine levels in patients.

But cytokines come in many different flavours that perform distinct roles in the immune system, and there was little consistency among these previous studies about which types of cytokines had unusual levels. Generally, cytokines linked with inflammation were higher in patients, but some studies found no differences at all.

The current researchers shared the view that cytokines were critical immune markers and believed that variation in diagnostic criteria and laboratory methods may have contributed to the variation. So they looked at more different kinds of cytokines than most previous studies had — 51 — and they hoped a large rigorous study would give a clear picture.

It didn’t — at least not at first. Several inflammatory cytokines were a little lower in patients than controls, contradicting both previous studies and intuition. But overall, there didn’t seem to be a huge amount going on, which surprised the researchers.


However, when they looked at the raw data more closely, Hornig and her colleagues could see that something was up.

Most research cohorts are made up of patients who have been ill for a long time, but Hornig’s group had carefully included newly-ill patients as well: 52 who had been ill for under three years (an average of 20 months) alongside 246 patients who had been ill for longer (an average of 13 years).

When they split the patients into those two groups, the data suddenly formed a striking pattern: cytokine levels were mostly up in the short duration patients and down in the long duration.

The graphic below makes this clear.


]

Changes in cytokine levels mirror each other in short- and long-duration patients. Modified with permission from study Table S6 (click image to access original)

Red bars indicate lower cytokine levels than controls while green bars indicate higher. White bars indicate there was no significant differences between the two groups.

Because the cytokine levels mirror each other (e.g., up in short and down in long duration) they tend to cancel each other out in the combined group, which explains why there was not much to see when researchers compared all patients with controls.

More than half of the 51 cytokines studied were significantly different between the short and long duration patients.

A few of these differences may be false positives, but their high number makes for a compelling pattern. Also, for 11 of the cytokines, the statistics showed that there was less than 1 in 100,000 chance of a false positive (i.e., the p value was .00001 or less, when the accepted standard for significance is .05 or less).

“It appears that ME/CFS patients are flush with cytokines until around the three-year mark, at which point the immune system shows evidence of exhaustion and cytokine levels drop ... This shows there are distinct stages to the disease.” — Dr. Mady Hornig

What’s going on?

So do these findings explain ME/CFS? Not yet. At this stage they are more like clues. The average cytokine differences between short- and long-duration patients are modest, and furthermore, the patients whose cytokine levels were more abnormal didn’t necessarily have worse symptoms. Abnormal cytokine levels, then, cannot be the direct, sole cause of symptoms.

Professor Jonathan Edwards explained on Phoenix Rising that immunological diseases such as rheumatoid arthritis set off very complex networks of cytokine signals, but it’s very hard to tell which of these signals actually causes the trouble. In fact, what you can measure may not be the signals that cause the trouble, but chatter about it.

“What this study is saying is that if you listen at the door of the immune system you can definitely hear some plotting of crime inside, but you don’t know which of the people speaking is going to go out and commit it”, he said.


Trouble elsewhere?

Dr. Hornig agrees that the cytokine pattern may be a sign of a problem in another part of the immune system, perhaps as a consequence of problems with particular immune cells. Each immune cell secretes many cytokines, so a problem with one type of immune cell is likely to affect the levels of many cytokines.

Intriguingly, one clue points at B cells, the very immune cell targeted by the drug rituximab (which has shown promise in treating CFS in early trials). Short-duration patients have less of a cytokine called CD40L that is critical to the development of properly functioning B cells.

In addition, CD40L levels in short-duration patients break free from the levels of other cytokines, varying up or down out of step with the rest. In healthy controls and long-duration patients, on the other hand, they correlate closely.

Another possibility is that what matters most of all are the levels of cytokines in the brain, and that the abnormal cytokine levels seen in the blood in this study are an echo of a bigger problem there. Because patients can have such striking cognitive problems, the brain has long been a suspect in ME/CFS. Hornig and colleagues have a new paper on cytokines in cerebrospinal fluid (which bathes the brain) coming out soon.

Viral hit and run?

One of the most eye-catching findings of the study was that the level of the inflammatory cytokine interferon gamma was crucial in distinguishing short-duration from long-duration patients. Short-duration patients consistently had higher interferon gamma levels than long-duration patients (albeit not much higher).

The strength of this association can be quantified through a number called an “odds ratio,” which in this case was a staggering 100 — normally anything over 3 is considered pretty exciting.


Virus-driver-5wee.png


The increased cytokine levels could be the results of a 'hit-and-run' virus

The authors say the increase in interferon gamma is consistent with a viral trigger (or disrupted immune regulatory networks), and of course many patients report a viral-like illness triggering their illness. However, to date, searches in blood from the same patients have yet to find signs of an infection.

Dr. Hornig believes a hit-and-run scenario is possible, where a virus infects someone and is defeated, but for some reason the immune system remains 'stuck in a high gear'. It’s possible that around the three-year mark, the immune system becomes exhausted, leading to the lowered cytokine levels in long-ill patients.

The finding here of elevated cytokines in short-duration patients differs from the famous Dubbo post-infectious studies, which found that cytokine levels were normal in ME/CFS after the initial acute illness, a disparity that’s currently hard to explain.

But other studies have found a link between cytokines and ongoing fatigue — for example, researchers found elevated levels of interferon gamma in those left with fatigue after West Nile virus infection.

What next?

These fascinating findings suggest a whole range of possible research avenues. The first thing needed is confirmation, as Hornig herself has stressed. One way to check them is to see if the cytokine levels for individual patients are changing over time the way this study suggests, and Columbia already has such a study in progress looking at some of these patients one year on.

But the best way to really understand what’s happening with cytokines in this illness is to get in there at the beginning and see what happens in the immune system as the illness first kicks off and then progresses. Hornig is exploring ways to make such a study happen.


“This is an exciting step forward in a historically challenging field."
- Professor Julia Newton


A biomarker – and a unique opportunity for treatment?
A biomarker has long been a ‘Holy Grail’ of ME/CFS research. Is the cytokine signature found in this study for early patients finally ‘it’, a test that will say “yes, you have ME/CFS”? And can it help researchers be sure they are studying people with the same illness?

Not yet. First, the findings need confirmation in other studies. In addition, the pattern of cytokine changes needs to be compared with those from other fatiguing illnesses, like multiple sclerosis or even unknown causes of fatigue.

Doctors can usually tell ME/CFS patients from the healthy. It’s separating these patients from those with other illnesses that’s so important. Hornig is eager to do the biomarker research that would take the study from the lab to the clinic. All they need is the funding.

The findings also raise the possibility that targeting the early immune dysregulation can prevent the illness settling into a long-term pattern. The first and crucial step is to identify what’s driving the cytokine abnormalities. Then, Hornig points out, therapies might be able to target those abnormalities in short-duration patients to stop the illness before it develops into its long-duration form.

A few years of ME/CFS are bad enough, but for many of us that would have been a dream outcome. And if these results are right, then it’s likely that newly ill patients will need different treatments from long-ill patients.

Clearly there is a long way to go before patients can expect a diagnostic test or, better still, effective treatment. But this new study, and others under way, gives me hope that we have entered a new era of bigger and better research that is finally capable of cracking ME/CFS.


You can donate to Dr. Hornig's and Dr. Lipkin’s work on ME/CFS
at The Center for Infection and Immunity, Columbia University, here.

Simon McGrath tweets on ME/CFS research: @sjmnotes


Thanks to Dr. Mady Hornig for talking to me and for supplying the photo; to Anne Griffith for 'hit-and-run' graphic created from Open Clip art images; to Sasha and oceiv (oceiv on UK media) for their summaries of media coverage.

Phoenix Rising is a registered 501 c.(3) non profit. We support ME/CFS and NEID patients through rigorous reporting, reliable information, effective advocacy and the provision of online services which empower patients and help them to cope with their isolation.

There are many ways you can help Phoenix Rising to continue its work. If you feel able to offer your time and talent, we could really use some more authors, proof-readers, fundraisers, technicians etc. We’d also love to expand our Board of Directors. So, if you think you can help in any way then please contact Mark through the Forums.

And don’t forget: you can always support our efforts at no cost to yourself as you shop online! To find out more, visit Phoenix Rising’s Donate page by clicking the button below.

Continue reading the Original Blog Post
 
Last edited:
Not that I know much about this, but I wonder...

Could an unusually strong immune reaction to a hit-and-run virus alter the balance of organisms in the microbiome, somewhat as an antibiotic can alter the balance? Might then the altered microbiome look like a chronic infection to the immune system?
 
Not that I know much about this, but I wonder...

Could an unusually strong immune reaction to a hit-and-run virus alter the balance of organisms in the microbiome, somewhat as an antibiotic can alter the balance? Might then the altered microbiome look like a chronic infection to the immune system?

You may be right, you may not be.

The possibilities are endless you see.

That's why Lipkin has approached this with the "it could be anything" attitude and testing cytokines, bacteria, viruses, metabolites in the blood, spinal fluid, intestine, pharynx. Etc.
 
Not that I know much about this, but I wonder...

Could an unusually strong immune reaction to a hit-and-run virus alter the balance of organisms in the microbiome, somewhat as an antibiotic can alter the balance? Might then the altered microbiome look like a chronic infection to the immune system?
Mady Hornig and Ian Lipkin are actively looking to see if changes in the microbiome are what's driving the changes they see in the immune system - that was one of the driving ideas behind the crowdfund project
www.microbediscovery.org
 
This article really helped me to understand this study in a complex and in-depth way. Great science to English translations are always appreciated. The thorough explanations of the study, the contextual information of where the field is and of what the future holds because of this study's findings are particularly interesting to see laid-out in one place.

Table 6 provides both an overall visual snapshot of the illness and a detailed richness. It's interesting to note that leptin wasn't singled out in this study, but did correlate with symptom severity in the much smaller Jared Younger study (if I'm reading Table 6 correctly - all the scientific terms in the table are new to me). I wonder what the in-progress Columbia study will show about how cytokine levels for individual patients are changing over time. I also hope that we see replication studies soon for both this Hornig/Lipkin study and the Jared Younger leptin study. As the saying goes "reading is fundamental." In the ME/CFS world, our new saying could be "funding is fundamental."

I loved this metaphor from Professor Jonathan Edwards (@Jonathan Edwards). It humorously explains and sums up this study well:

“What this study is saying is that if you listen at the door of the immune system you can definitely hear some plotting of crime inside, but you don’t know which of the people speaking is going to go out and commit it”,
 
Last edited:
Excellent article, @Simon, very clear - even I understood it (which is saying something)!

Yes, thank you for the excellent article, Simon!

In reading the Lipkin/Hornig paper, I don't recall if they mentioned the length of the shortest duration since onset for the samples taken from patients who had had the disease <3 years. I'm just wondering if the cytokine profiles might show yet more differences the closer you get to onset, or if they just get stuck in the "on" position following some infection, as has been mentioned. One problem with looking for that kind of thing is, obviously, that you can't technically get a diagnosis prior to onset+6 months. I had blood drawn within 3 days after onset, but, of course, only routine tests were run on it. That was a long time ago, but, even today, no one is going to preserve your blood sample if they think you just have a prolonged case of the flu.
 
This article really helped me to understand this study in a a complex and in-depth way. Great science to English translations are always appreciated. The thorough explanations of the study, the contextual information of where the field is and of what the future holds because of this study's findings are particularly interesting to see laid-out in one place.
You certainly know how to make a blogger happy - thank you.

able 6 provides both an overall visual snapshot of the illness and a detailed richness. It's interesting to note that leptin wasn't singled out in this study, but did correlate with symptom severity in the much smaller Jared Younger study (if I'm reading Table 6 correctly - all the scientific terms in the table are new to me). I wonder what the in-progress Columbia study will show about how cytokine levels for individual patients are changing over time. I also hope that we see replication studies soon for both this Hornig/Lipkin study and the Jared Younger leptin study.
The paper also highlights this about Leptin
Among short-duration subjects, the levels of the adipose-related cytokine leptin were only correlated with IL-12p40, CSF2 (GM-CSF), and serpin E1 [plasminogen activator inhibitor-1 (PAI1)], whereas leptin was tightly related to most of the other 51 cytokines in both the long-duration ME/CFS and the control groups. Leptin also had an inverse relationship to PDGFBB among controls.
So Leptin seems oddly out of whack in short duration patients, suggesting it is still a cytokine to investigate.

And yes, great line from @Jonathan Edwards.

I loved this metaphor from Professor Jonathan Edwards. It humorously explains and sums up this study well:
Yes, thank you for the excellent article, Simon!

In reading the Lipkin/Hornig paper, I don't recall if they mentioned the length of the shortest duration since onset for the samples taken from patients who had had the disease <3 years. I'm just wondering if the cytokine profiles might show yet more differences the closer you get to onset, or if they just get stuck in the "on" position following some infection, as has been mentioned.
You're welcome!

When I spoke to Mady Hornig she said that it seemd fairly stable up to three years, but they weren't sure how fast it dropped afterwards as they didn't have too many patients in the 3-8 year bracket. Their ongoing longitudinal study (tracking cytokine levels in the same patient one year on - or and and two years on for a few) should help throw light on this. I they plan more analysis and research on this, still early days.
 
This study also highlights a major failing in just about all psychology research.
Complete failure to take into account all confounding variables. :bang-head:

Thank goodness these researchers had the basic common scientific sense to "eyeball the data" - the phrase used in labs to describe the very first thing you do with it.:):thumbsup::)

Thanks for this, Simon. A beautiful, clear explanation. :thumbsup:
 
Simon, this is a nice piece of writing.

Why the emphasis on fatigue? What do you mean by "fatiguing illnesses"? Don't most diseases cause varying degrees of fatigue? Are there any febrile infections, for instance, that don't generate fatigue as a by-product?
 
One problem with looking for that kind of thing is, obviously, that you can't technically get a diagnosis prior to onset+6 months.
That is why you need large prospective cohort studies. Recruit a whole lot of people, do baseline tests. Some get sick. Follows those more closely, do all the tests. Some of those get ME or SEID or whatever. You now have a timeline of tests up until they got ME, and can continue into the future.
 
The hit and run idea, is a good one, But certain things for me seem to stick out like a sore thumb.
Not so much hit and run. More like Hit run, and come back again.

1 I had numerous Viral Flu like attacks that occurred over this three to five year period. At first, classic flu symptoms, With measurable very high fevers. And respiratory chest symptoms. Cold shivering. Burning hot. Classic Flu. Often at first, with long with gaps in time, between the attacks. EG twelve weeks, very early on.

2 No one in my family seemed to catch these Flu type illnesses. Not once ?

3 The illness, at first, was being controlled to some extent by the immune system.
I have the distinct impression. The immune system was putting the virus to bed, so to speak. But not ridding it. As it reactivated numerous times. Which eventually led to ME/CFS.

4 A virus that becomes part of the persons DNA like a retro virus, or chicken pox, type virus. or other ? One that can not be rid from the body, But can re activate itself after a period of time. seems apparent to me.

5 The longer this went on for, the more the symptoms altered. Meaning the long gaps that had initially happened.
Became shorter and shorter. But almost as a consequence, the symptoms started to change in parallel with this change.

They became less sever, the temperatures stopped. It seemed to become a balancing act between the immune system coping with such a virus, The symptoms becoming less sever, But also more constant in time. Now with no long gaps. Gaps now measured in days. Not weeks. With many of the Flu symptoms that started the ball rolling. But now to a less sever extent.

6 Over many years, this working out of the immune system, how to keep such a virus in check, continued. As the years rolled by. So the attacks strength and time duration weakened, and shortened more and more. The more the years passed. It has now been twenty years for me, and I am not as ill as I was. Infact much less so. I consider myself partly recovered.

I can only speculate for me, what all this means. But for me, this was NOT just a simple hit and run Scenario.

But more numerous HIT AND HIDES and combacks.

likely reactivation of the same virus, clearly the body was NOT able to rid itself of. But had to learn to co exist with, and learn to fight the best way it could. over the rest of my life.

Yes it become Chronic, But not just fatigue. More like chronic virally ill.

What viruses behave like this. What could it be ? why hasn't it been discovered ?

Either I did not have ME. Or clearly there is indeed subsets. As not everyone has the same historical picture as mine ?
Not sure what to make of all this. And please forgive my simplistic, story, and explanations.
But I do know I was more ill the first five years. which seemed to suggest cytokines would have been more super elevated early on than later on.
Hope this means something to those that understand these things more than I do. I apologize if it does not.
Lets hope this research is really the first key to unlock the door.
Thank you to all those involved in this study. Its heartening to see such brilliant minds trying to crack a very confusing, misleading, and often contradictory nut.
Thank you Doctors Lipkin and Hornig
Thank you Simon for your hard work.
 
A very good summation of the paper, but I think the viral hit-and-run theory was a bit too prominent. I haven't heard Dr. Hornig's comments on this, but just going by the paper it would seem that this is not the only thing the results might suggest. I assume the reference to viruses is because the same pattern of cytokines appears in HCV and HIV infection (as stated in the paper), but these are chronic infections, not hit-and-runs. So a chronic infection seems equally (if not more) likely, based on the paper's findings.
 
A very good summation of the paper, but I think the viral hit-and-run theory was a bit too prominent. I haven't heard Dr. Hornig's comments on this, but just going by the paper it would seem that this is not the only thing the results might suggest.
I suspect the graphic (can't control the size it appears in the forum version) may have given too much prominence to this :). I've also edited the blog to say IFN-g levels as consistent with a viral trigger trigger 'or disrupted immune regulatory networks', as per paper. You'll have seen the suggestion about CD40L too, a cytokine critical in B cell development, and the possibility that cytokine levels in the brain are driving this:
Intriguingly, one clue points at B cells, the very immune cell targeted by the drug rituximab (which has shown promise in treating CFS in early trials). Short-duration patients have less of a cytokine called CD40L that is critical to the development of properly functioning B cells.

In addition, CD40L levels in short-duration patients break free from the levels of other cytokines, varying up or down out of step with the rest. In healthy controls and long-duration patients, on the other hand, they correlate closely.

Another possibility is that what matters most of all are the levels of cytokines in the brain, and that the abnormal cytokine levels seen in the blood in this study are an echo of a bigger problem there. Because patients can have such striking cognitive problems, the brain has long been a suspect in ME/CFS. Hornig and colleagues have a new paper on cytokines in cerebrospinal fluid (which bathes the brain) coming out soon.

I assume the reference to viruses is because the same pattern of cytokines appears in HCV and HIV infection (as stated in the paper), but these are chronic infections, not hit-and-runs. So a chronic infection seems equally (if not more) likely, based on the paper's findings.

IFN-g is a general antiviral cytokine, not specific to HCV or HIV. Certainly a chronic infection remains a possibility, but to date the pathogen hunt using blood samples from the same cohort hasn't found any sign of chronic infections. Dr Ian Lipkin reported somewhere last year that intial testing on WBCs for HCV drew a blank.

edit: interesting re persistent WNV in kidney tissue.
 
Yeah, I probably wouldn't have said anything without the graphic...I don't know enough about cytokines to be sure on this, but I think IFN-g is also raised in bacterial infection, at least I saw one paper where it was raised in acute Lyme. I asked on another thread, and no one knew, but you seem very informed, so I will ask you: do you know when the pathogen paper is coming out? I would like to know exactly which pathogens they tested for, and whether some of these were bacteria.
 
Oh, and when I said the same pattern of cytokines I was remembering it slightly wrongly, the 3-year separation was used because this they had observed a similar divide in HCV, HIV and West Nile Virus.
 
Back