My Rituximab experience for ME

[Jonathan Edwards]

@Jonathan Edwards

What would be your view on idea that pathogens might trigger autoimmunity by infecting important immune cells or important immune organs such as the thymus, spleen or bone marrow? If a chronic infection occurs in these cells or organs, it may conceivably create an environment which increases the chances of autoimmunity occurring, even if its actual occurrence is due to random factors in antibody production.

If we take enterovirus, this appears to profoundly affect dendritic cells:
Wild-type coxsackievirus infection dramatically alters the abundance, heterogeneity, and immunostimulatory capacity of conventional dendritic cells in vivo

Echovirus infection causes rapid loss-of-function and cell death in human dendritic cells

And coxsackievirus B4 seems to infect the thymus:
How Does Thymus Infection by Coxsackievirus Contribute to the Pathogenesis of Type 1 Diabetes?

And of course we know that Epstein-Barr chronically infects B cells.

I guess it is possible that infections might create local conditions that allowed an autoimmune process to get started when it would not otherwise but I am not quite sure how. EBV infection of B cells causes a permanent shift in B cell behaviour and this might be a permissive factor in a lot of adult autoimmune disease but it is a bit hard to tell since there is no way of studying an EBV uninfected population of any size. I am not sure that dendritic cells have much to do with autoimmunity. Autoantibody production probably makes use of antigen presentation by B cells themselves rather than other cells.

I have not read the hepatitis C paper but the quote looks dubious to me. One thing we are pretty sure about now is that TH17 cells are not important in RA - and I rather doubt they are important in any of the autoimmune diseases mentioned. They seem to be important in seronegative spondarthropathies instead.

 

[nandixon]

I have not read the hepatitis C paper but the quote looks dubious to me.

Figure 1A from that paper provides an interesting graphic, showing significantly increased autoimmune disease in chronic hepatitis C (CH-C) infection:

The relation between CH-C and the phenotype of autoimmune-diseases.
The prevalence of these diseases in CH-C (n = 250) was significantly higher than in other chronic liver diseases (n = 98) (p = 0.0011) (A).

 

[Jonathan Edwards]

Figure 1A from that paper provides an interesting graphic, showing significantly increased autoimmune disease in chronic hepatitis C (CH-C) infection:

View attachment 12970

I would like to see some actual data. And you cannot call lymphoma an autoimmune disease. Moreover, the link with RA is tricky because Hep C on its own produces rheumatoid factors and also can produce arthritis. Cryoglobulins are also part of the Hep C response and not strictly autoimmunity. Thrombocytopenia is also tricky because chronic liver disease is associated with thronmocytopenia for other reasons. If they are going to try to claim a link they need to do it properly rather than chuck in mixed bag like this. There may be a weak link with thyroid disease and one or two other bona fide autoimmune conditions but when you see people over-egging things you tend to be sceptical about the whole thing. Moreover they do not seem to have any cases of autoimmune chronic liver disease (anti-smooth muscle associated), which would have bumped up the autoimmune association in the control group. Smells of cherry-picking to me.

 

[Hip]

EBV infection of B cells causes a permanent shift in B cell behaviour and this might be a permissive factor in a lot of adult autoimmune disease but it is a bit hard to tell since there is no way of studying an EBV uninfected population of any size.

That is certainly the problem with EBV, since of course around 95% of the adult population have it.

With coxsackievirus B / echovirus it's different though: I have seen studies which show the prevalence of IgG antibodies in the general population to be roughly around the 20% to 50% mark.

I am not sure that dendritic cells have much to do with autoimmunity. Autoantibody production probably makes use of antigen presentation by B cells themselves rather than other cells.

Mightn't the disruption that enterovirus causes to dendritic cells also affect the production of tolerogenic dendritic cells, which I believe help counter autoimmunity?

As an aside: what I found interesting is that estriol increases tolerogenic dendritic cell production, and blood levels of this hormone increase by 1000-fold during pregnancy. 1 Now some ME/CFS patients significantly improve during pregnancy, so possibly this estriol-dendritic cell connection may help explain this.

 

[Jonathan Edwards]

That is certainly the problem with EBV, since of course around 95% of the adult population have it.

With coxsackievirus B / echovirus it's different though: I have seen studies which show the prevalence of IgG antibodies in the general population to be roughly around the 20% to 50% mark.


Mightn't the disruption that enterovirus causes to dendritic cells also affect the production of tolerogenic dendritic cells, which I believe help counter autoimmunity?

As an aside: what I found interesting is that estriol increases tolerogenic dendritic cell production, and blood levels of this hormone increase by 1000-fold during pregnancy. Now some ME/CFS patients significantly improve during pregnancy, so possibly this estriol-dendritic cell connection may help explain this.

I have never heard of tolerogenic dendritic cells - it sounds like something based on transgenic mice. And I have certainly never heard of them being involved in autoimmunity - I cannot work out how the story would go. Trouble is that people who put in grants on dendritic cells are bound to write reviews saying they are important in autoimmunity. That is the way immunology works these days.

I am not sure the estriol connection would work. If autoimmunity is antibody mediated - as all the ones we know are so far, then the antibodies will still be there in pregnancy. The dendritic cells would presumably play a role in failing to tolerise or whatever when the disease started, beforehand. Upping the tolerising role would not seem to be much hel once tolerance was broken.

The problem for me with all these theories is that they fall aprat when one looks at the detailed dynamics - the way things would play out over time. Most immunologists never think about that. They think that if they can make a disease in a transgenic mouse that solves the problem!

 

[Hip]

@Jonathan Edwards
If you search PubMed, there are quite a few papers mentioning tolerogenic dendritic cells (TDCs). I saw one recent 2013 paper investigating the use of TDCs as a possible treatment for RA.

I am not sure of their mechanism of action; this paper looks at how TDCs induce regulatory T cells.

 

[Jonathan Edwards]

@Jonathan Edwards
If you search PubMed, there are quite a few papers mentioning tolerogenic dendritic cells (TDCs). I saw one recent 2013 paper investigating the use of TDCs as a possible treatment for RA.

I am not sure of their mechanism of action; this paper looks at how TDCs induce regulatory T cells.

There are always lots of papers on these trendy new cell types - from the people who want to ramp up the interest in grant applications. If there are cogent theories supported by evidence then OK, but all the stuff on regulatory T cells and even regulatory B cells in RA that I have come across was just ramping up interest without any cogent story.

As I have said before, just as there is psychobabble there is immunobabble - and rather more of it.

The question I would ask is why go chasing the possibility of some pre-preferred theory being relevant - especially when it is the sort of theory that has been around for decades and never added up - and why not look at the evidence first and pick a theory on its merits in terms of the evidence. The answer to that question for most people who write immunology papers as far as I can see is that they have no ideas of their own so latch on to old ideas that seem to make sense in naive terms and re-brew them around some new cell type. George W Bush was not the cause of the French Revolution for reasons that are not too hard to work out. Theories of autoimmunity tend to be at that sort of level of error.

A principle I have always found useful in science is that whatever is published in Nature this week is almost certainly NOT the answer to all known diseases including one you can put a grant in on!

 

[Jonathan Edwards]

I looked up that 2013 paper and saw the abstract. John Isaacs has been banging on about T cell mechanisms in RA for twenty years despite the fact that it was his work that very neatly showed that T cells do not matter much in RA (he reduced them to AIDS levels and the arthritis did not get better). I should perhaps make a small amendment to the post above - to put it all in bold type! He also submitted an abstract, together with the drug company, to the British Society for Rheumatology describing the trial that I had set up, without my name on it. I made sure the abstract was withdrawn. He was sidelined out of the B cell project (onto which I had personally invited him) by his old boss so went back to T cells and dendritic cells. He never seemed to understand how tolerance works.

 

[Eeyore]

Moreover they do not seem to have any cases of autoimmune chronic liver disease (anti-smooth muscle associated), which would have bumped up the autoimmune association in the control group. Smells of cherry-picking to me.

@nandixon - I think the problem here is that they are neither testing ALL autoimmune disease, nor 1 in particular - but rather choosing a rather hodgepodge group with no clear reason for it. This causes some statistical problems, such as correction for multiple testing. If you look at all possible subsets of autoimmune diseases (this is a combinations problem, sum of n choose k for k = 1 to n, a VERY large number for even modest sized n) the number is very large, so by chance alone you can find statistical significance. You can fix this with Bonferroni multiple test correction, but that would almost certainly eliminate statistical significance as combinations tend to ramp up very fast. Always be suspicious when you see researchers grouping things in seemingly random ways and cherry picking their way to statistical significance. You can almost always find extremely low p-values (measures of statistical significance) between a SNP and a disease if you look at the whole genome, because the odds of finding any (non predetermined) SNP enriched in a random population are high when you are looking at 5.8 billion base pairs. It's very different if your starting hypothesis was to prove SNP X important, and you do. It's rather like the lotto - your odds of winning it are very low, but there are so many people buying tickets that the odds of someone winning it are not low.

In this case, had they chosen only autoimmune disease of the liver, such as PBC, and seen statistical significance, that would be more meaningful because then the choice would not seem to be random. Similarly, if they had given a set of autoimmune diseases either meant to be comprehensive or based on theoretical prediction to be relevant in advance, and then checked it, and found statistical significance, that would be more relevant.

That is certainly the problem with EBV, since of course around 95% of the adult population have it.

I don't even have EBV. Negative antibodies and negative PCR, and no history of mono (in me or any family members, including extended to the extent I know). Didn't save me from ME. EBV has, however, long been implicated in MS (the risk may be 9-10 fold as high for EBV positives). I don't believe MS is infectious, or even that treating EBV would help (someone has probably tried this), but I do think it alters immune function in a way that predisposes to autoimmunity in some diseases. MS may be somewhat unique among autoimmune diseases though, and may not be subject to the same processes as the diseases treated by rheums more than neurologists (although I'm not sure where that distinction came in - maybe @Jonathan Edwards has an opinion on why MS is an autoimmune disease treated by neuros, and almost all others are treated by rheums). Is it just historical artifact?

As an aside: what I found interesting is that estriol increases tolerogenic dendritic cell production, and blood levels of this hormone increase by 1000-fold during pregnancy. Now some ME/CFS patients significantly improve during pregnancy, so possibly this estriol-dendritic cell connection may help explain this.

One big problem here is that you are talking about hormones found at much higher levels in women. Considering the ratio of women to men with ME is so high, it would argue strongly against this being very relevant.

This might be along the lines of Dr. Edwards analogy as to why GWB did not cause the French Revolution. I suppose one can posit time travel - you can always find away around an argument - but you have to examine if it's more likely that Dubya could time travel or that he simply wasn't the cause of the French Revolution.

 

[beaker]

Figure 1A from that paper provides an interesting graphic, showing significantly increased autoimmune disease in chronic hepatitis C (CH-C) infection:

View attachment 12970

I wonder if the thyroiditis listed is actually IIT - Interleukin Induced Thyroiditis. Interferon-alpha is often given as treatment for HepC and so it may be more of side effect of that.

 

[beaker]

As an aside: what I found interesting is that estriol increases tolerogenic dendritic cell production, and blood levels of this hormone increase by 1000-fold during pregnancy. Now some ME/CFS patients significantly improve during pregnancy, so possibly this estriol-dendritic cell connection may help explain this.

More likely that increased blood volume during pregnancy helps #MEcfs improve. Or at least a big chunck of the reason why.

 

[Valentijn]

More likely that increased blood volume during pregnancy helps #MEcfs improve. Or at least a big chunck of the reason why.

It sounds like there is also immune suppression during pregnancy, so that the fetus is tolerated. If there is an auto-immune disease, that immune suppression might be affecting it as well.

 

[Jonathan Edwards]

The effect of pregnancy is puzzling. Lupus actually gets worse in pregnancy - often much worse. So there is no general immune suppression (although RA gets better). The status of the fetus may not be quite what used to be suggested. Grafts are only rejected if the area of tissue containing the foreign cells is supplied by the host blood supply - because there can only be a host-based inflammatory reaction if the cells can traffic to and from host lymph nodes. In the placenta I think the fetal tissue is supplied by fetal vessels. So any maternal T cells getting into the fetal tissue that generated cytokines would be talking to the fetuses immune system, not the mother's immune system.

 

[Hip]

The question I would ask is why go chasing the possibility of some pre-preferred theory being relevant - especially when it is the sort of theory that has been around for decades and never added up - and why not look at the evidence first and pick a theory on its merits in terms of the evidence.

Unfortunately I do not have the expertise to make a judgement on whether a new cell type is a passing fad or something of significance.


Changing the subject slightly: it occurred to me that there must be some sort of feedback mechanism for antibody production, such that when a pathogen targeted by an antibody is completely wiped out (eg, when you get rid of a cold), then the antibody production is stopped, or reduced to a tick-over level.

I am guessing that a feedback mechanism must be in place, otherwise how would antibody production know to return to tick-over level once a pathogen is eliminated.

Obviously when the antibody is an autoantibody, that targets a component of the human body(autoantigen), that bodily component is never eliminated. So the feedback mechanism would never signal to B cells to stop antibody production in autoimmune cases, because this mechanism will always detect that the autoantigen is still present.

Thus my thought was, has anyone looked into trying too halt autoimmunity by modulating this feedback mechanism (whatever it is, I don't know what it is called)?

One big problem here is that you are talking about hormones found at much higher levels in women. Considering the ratio of women to men with ME is so high, it would argue strongly against this being very relevant.

This is not really a male/female issue, but rather why females with ME/CFS sometimes exhibit significant symptomatic improvements during pregnancy. Estriol in females becomes elevated by 1000 times during pregnancy (progesterone is also increased in pregnancy, but not to the same degree). I am guessing estriol may be part of the immune tolerance in pregnancy.

In pregnancy, the relapse rate of multiple sclerosis is reduced, and estriol is being investigated as a treatment for MS, as estriol treatment has also been shown to reduce MS relapse rate.

 

[Eeyore]

The effect of pregnancy is puzzling. Lupus actually gets worse in pregnancy - often much worse. So there is no general immune suppression (although RA gets better). The status of the fetus may not be quite what used to be suggested. Grafts are only rejected if the area of tissue containing the foreign cells is supplied by the host blood supply - because there can only be a host-based inflammatory reaction if the cells can traffic to and from host lymph nodes. In the placenta I think the fetal tissue is supplied by fetal vessels. So any maternal T cells getting into the fetal tissue that generated cytokines would be talking to the fetuses immune system, not the mother's immune system.

Patients with particular complement deficiencies (often involving impaired opsonization) VERY frequently develop SLE. Interestingly, one of the treatments for many complement deficiencies is androgen therapy. Hormones may have some role here. I am not sure of the mechanisms involved in using androgens to treat SLE resulting from complement deficiency.

 

[Biarritz13]

Hi everyone,

@funkyqueen
Hi! Where exactly do you follow the protocol please?

 

[Eeyore]

Changing the subject slightly: it occurred to me that there must be some sort of feedback mechanism for antibody production, such that when a pathogen targeted by an antibody is completely wiped out (eg, when you get rid of a cold), then the antibody production is stopped, or reduced to a tick-over level.

Not exactly - although some sort of feedback has been hypothesized in the role of IVIG as an immune modulator (along with a lot of other mechanisms). The body (unless something goes really wrong, like cancer) doesn't produce so many antibodies that the blood becomes hyperviscous, so there may be some general feedback inhibition tied to total antibody levels, although I'm not sure of this. I suspect it's more likely to be global rather than antigen specific though.

With antibody production, you need ongoing stimulus to continue to produce antibodies. Lack of that ongoing stimulus leads to a cessation of antibody production by short-lived plasma B-cells. You still have memory B-cells that allow for an anamnestic response when re-exposed to the same pathogen - which has obvious benefits - but these produce antibodies at a much lower level, and this too trails off over time w/o ongoing stimulus (this is why booster vaccines are needed). If the infection is lifelong (e.g. chicken pox) then there are always some viruses being produced, so there is always ongoing stimulation at some level.

 

[Jonathan Edwards]

Changing the subject slightly: it occurred to me that there must be some sort of feedback mechanism for antibody production, such that when a pathogen targeted by an antibody is completely wiped out (eg, when you get rid of a cold), then the antibody production is stopped, or reduced to a tick-over level.

I am guessing that a feedback mechanism must be in place, otherwise how would antibody production know to return to tick-over level once a pathogen is eliminated.

Obviously when the antibody is an autoantibody, that targets a component of the human body(autoantigen), that bodily component is never eliminated. So the feedback mechanism would never signal to B cells to stop antibody production in autoimmune cases, because this mechanism will always detect that the autoantigen is still present.

Thus my thought was, has anyone looked into trying too halt autoimmunity by modulating this feedback mechanism (whatever it is, I don't know what it is called)?

Antibody production does not return to tick over level. It continues at the level reached during the response for a considerable time. It then wanes as plasma cells die. Interestingly, there does not appear to be any feedback mechanism controlling antibody levels other than a fixed amount of space available to store plasma cells.

There are, however, mechanisms that will lead to saturation of the signal systems that expand B cells, which is presumably why the antibodies to antibodies in RA do not expand to infinity. The mechanisms are complicated and relate to steric interference between antibodies binding to different epitopes and a variety of other things like Fc receptor engagement.

We have indeed thought long and hard about trying to beef up the saturation mechanisms or other wise block further expansion of clones specifically making autoantibodies. However, there is no very obvious way to block signals specifically for the antibodies causing trouble. It may be that some of the traditional treatments for RA like gold in fact do this, by maybe interfering with complement or FcR binding but nobody has studied that in a convincing way.

 

[nandixon]

I wonder if the thyroiditis listed is actually IIT - Interleukin Induced Thyroiditis. Interferon-alpha is often given as treatment for HepC and so it may be more of side effect of that.

I don't think so, because the impression is given that the patients were not being treated prior to the study:

We collected the peripheral blood before the treatment (treatment naïve).

 

[nandixon]

I would like to see some actual data. And you cannot call lymphoma an autoimmune disease. Moreover, the link with RA is tricky because Hep C on its own produces rheumatoid factors and also can produce arthritis. Cryoglobulins are also part of the Hep C response and not strictly autoimmunity. Thrombocytopenia is also tricky because chronic liver disease is associated with thronmocytopenia for other reasons. If they are going to try to claim a link they need to do it properly rather than chuck in mixed bag like this. There may be a weak link with thyroid disease and one or two other bona fide autoimmune conditions but when you see people over-egging things you tend to be sceptical about the whole thing. Moreover they do not seem to have any cases of autoimmune chronic liver disease (anti-smooth muscle associated), which would have bumped up the autoimmune association in the control group. Smells of cherry-picking to me.

You might have missed it due to nomenclature differences between countries, but the Figure 1A graphic includes "AIH," i.e., autoimmune hepatitis, which I think may include what you describe as "autoimmune chronic liver disease (anti-smooth muscle associated)."(?)

With respect to the thrombocytopenia (labeled in the graphic as "ITP," i.e., idiopathic thrombocytopenic purpura) being associated with chronic liver disease for other reasons, note that the graphic is eliminating that bias/concern to some degree because in the pie charts ITP in chronic hepatitis C (CH-C) is being compared to ITP in "Other Chronic Liver Diseases" and it's shown to be increased in CH-C relative to those other chronic liver diseases.

I wouldn't think there's been cherry-picking here. The way the study is presented, my interpretation is that 250 patients with hepatitis C virus (HCV) were enrolled in the study and the percentage indicated by the pie chart were also found to have the comorbidities of the autoimmune diseases (and lymphoma) indicated. The pie charts indicate the remainder had "No Complication."

So it wouldn't seem that only certain diseases, autoimmune or otherwise, were selectively chosen and presented in order to give a good p-value. With the way they've presented things, to do otherwise would not only be unscientific but unethical - effectively lying, actually, given the "No Complication" labeling.

Lastly, I don't think we should overlook the possible relevance of the main finding of the study and its practical application (if any) to ME/CFS:

"In conclusion, we report the detailed mechanism of Th17 development and HCV infection, which might be involved in the pathogenesis of autoimmune-related disease in CH-C patients (Fig S2). Recently, a novel therapy targeting STAT-3 signaling was reported[23], [41], [42]. We should consider the clinical use of such treatments for autoimmune-related diseases in CH-C patients."

@nandixon - I think the problem here is that they are neither testing ALL autoimmune disease, nor 1 in particular - but rather choosing a rather hodgepodge group with no clear reason for it.
.............
Always be suspicious when you see researchers grouping things in seemingly random ways and cherry picking their way to statistical significance.

See my reply to Professor Edwards above.