What are the other mechanisms by which RTX might be working?

[Jonathan Edwards]

@Marco
I read the paper (most of it). They gave rituximab on day zero - so before the immune response got started. You would expect a T cell mediated disease to be suppressed in that situation because B cells are very potent specific antigen presenters and without them the T cell response it likely to be greatly reduced. So the effect of rituximab in this model just tells us that the textbooks of immunology are roughly right.

They make the mistake of citing the immediate blockade of new lesions in patients with rituximab as being against antibody mediation. However, as indicated above this is non sequitur. In RA the treatment is designed to damp down existing inflammation. In MS it is designed to block new inflammation. I am not aware of any suggestion that rituximab speeds up recovery from existing lesions. So the half life of pre-existing antibodies is not relevant. They talk of plasma half life of antibody but plasma antibody is not the problem in MS - if it was the whole brain would be inflamed for months on end - or permanently. The lesions in MS are focal, so any role for antibody must be local. You probably only need one plasma cell in a piece of brain to make enough antibody locally to cause microglial activation and demyelination.

The pity is that most clinical immunologists will buy this as further evidence of MS being T cell mediated after all.

I do not think there is any suggestion in this paper that rituximab is doing anything other than killing B cells. The effect on microglial activation will be downstream of a reduction in T cell activation. Microglial activation will occur distant from where the T cells are because when a nerve cell is damaged in one place microglia get activated in all the places where that nerve cell has connections that have been knocked out. So after almost any focal brain damage you get brain stem microglial activation because more or less all neurons connect to the brain stem. (You see that after that lecture I am now an expert on this as well!!)

Thanks for bringing this to my attention. Angela Vincent made a reference to this sort of work at the workshop. Angela is brilliant but I do suspect a blind spot on the old T cell issue here. (She has reciprocal suspicions judging by her eyebrows at times.) Next time I will tell her why she is wrong.

 

[Snow Leopard]

And note that the human disease is more or less defined by oligoclonal bands of antibody in the CSF.

Hmm, why haven't there been any serious studies investigating whether there are oligoclonal bands in CSF or serum in ME patients?

Byron Hyde claimed the following in the "Handbook of Chronic Fatigue Syndrome" (2003, editor- Lenny Jason)

The second point to remember is that many patients with acute onset ME/CFS may demonstrate IgG oligoclonal bands in their spinal fluid. These do not usually go on to develop multiple sclerosis (MS).

 

[Marco]

@Marco
I read the paper (most of it). They gave rituximab on day zero - so before the immune response got started. You would expect a T cell mediated disease to be suppressed in that situation because B cells are very potent specific antigen presenters and without them the T cell response it likely to be greatly reduced. So the effect of rituximab in this model just tells us that the textbooks of immunology are roughly right.

They make the mistake of citing the immediate blockade of new lesions in patients with rituximab as being against antibody mediation. However, as indicated above this is non sequitur. In RA the treatment is designed to damp down existing inflammation. In MS it is designed to block new inflammation. I am not aware of any suggestion that rituximab speeds up recovery from existing lesions. So the half life of pre-existing antibodies is not relevant. They talk of plasma half life of antibody but plasma antibody is not the problem in MS - if it was the whole brain would be inflamed for months on end - or permanently. The lesions in MS are focal, so any role for antibody must be local. You probably only need one plasma cell in a piece of brain to make enough antibody locally to cause microglial activation and demyelination.

The pity is that most clinical immunologists will buy this as further evidence of MS being T cell mediated after all.

I do not think there is any suggestion in this paper that rituximab is doing anything other than killing B cells. The effect on microglial activation will be downstream of a reduction in T cell activation. Microglial activation will occur distant from where the T cells are because when a nerve cell is damaged in one place microglia get activated in all the places where that nerve cell has connections that have been knocked out. So after almost any focal brain damage you get brain stem microglial activation because more or less all neurons connect to the brain stem. (You see that after that lecture I am now an expert on this as well!!

Thanks Jonathan - that all makes sense.

Thanks for bringing this to my attention. Angela Vincent made a reference to this sort of work at the workshop. Angela is brilliant but I do suspect a blind spot on the old T cell issue here. (She has reciprocal suspicions judging by her eyebrows at times.) Next time I will tell her why she is wrong.

Gotta love academia. It reminds me of my undergrad thesis which was poised to support or refute either my thesis supervisor's thesis or the PhD thesis of a rival lecturer from the same department. That was fun. I ended up being explicitly banned from talking to the rival in case I was 'contaminated'.

 

[Jonathan Edwards]

Hmm, why haven't there been any serious studies investigating whether there are oligoclonal bands in CSF or serum in ME patients?

Byron Hyde claimed the following in the "Handbook of Chronic Fatigue Syndrome" (2003, editor- Lenny Jason)

I think if there was any rise in immunoglobulin level, whether oligoclonal or not, in ME CSF then it would have stood out like a sore thumb in the proteomics study. All CSF samples get tested for protein and I think oligoclonal bands will show up on that at least as an increase in protein and if an increase in protein is found it should be pretty routine to look for bands. My guess is that they are rarely found.

Dr Hyde's comment seems a bit back to front to me. If someone has neurological symptoms and oligoclonal bands I would personally classify them in scientific terms as having the disease we call MS even if they never progressed clinically. There must be a lot of subclinical MS because there are a lot of people who only ever have one mild episode. That might mean that MS is one of the many causes of CFS or ME or SEID.

 

[Snow Leopard]

I think if there was any rise in immunoglobulin level, whether oligoclonal or not, in ME CSF then it would have stood out like a sore thumb in the proteomics study. All CSF samples get tested for protein and I think oligoclonal bands will show up on that at least as an increase in protein and if an increase in protein is found it should be pretty routine to look for bands. My guess is that they are rarely found.

How common are they in autoimmune disorders other than MS (I mean in terms of specificity/sensitivity of the finding)? The Wikipedia article claims:

Oligoclonal bands are found in:

Multiple sclerosis
Lyme Disease
Neuromyelitis optica (Devic's disease)
Systemic lupus erythematosus
Neurosarcoidosis
Subacute sclerosing panencephalitis
Subarachnoid hemorrhage
Syphilis
Primary central nervous system lymphoma
Sjögren's Syndrome
Guillain-Barre Syndrome
Meningeal carcinomatosis
Multiple Myeloma
Parry–Romberg syndrome

 

[Jonathan Edwards]

How common are they in autoimmune disorders other than MS (I mean in terms of specificity/sensitivity of the finding)? The Wikipedia article claims:

I cannot give any informed figures for that but the list you give includes several diseases in which there may be plasma monoclonal bands with the same in CSF which is sort of cheating (myeloma, lymphoma, Sjogren's). There are also some disorders where chronic immune reactions set up in brain and if so one would expect the antibody produced locally to be relatively oligoclonal anyway. (And I think this is probably a list that includes any single case reports found in the literature. Oligoclonal bands with subarachnoid haemorrhage seems pretty unlikely.) I think maybe the main point is that most of these have obvious structural changes in brain tissue that makes them easily separable from MS. I guess maybe the diagnostic feature of MS is oligoclonal bands in the absence of some fairly obvious structural immune boot of some other type being put in.

 

[Snow Leopard]

I cannot give any informed figures for that but the list you give includes several diseases in which there may be plasma monoclonal bands with the same in CSF which is sort of cheating (myeloma, lymphoma, Sjogren's)..

What I'm trying to say is, if we don't observe these bands in either plasma or CSF, should we avoid going fishing for specific autoantibodies in ME patients?

 

[Jonathan Edwards]

What I'm trying to say is, if we don't observe these bands in either plasma or CSF, should we avoid going fishing for specific autoantibodies in ME patients?

Most autoantibodies are polyclonal so they do not show up as bands. Although all the antibodies may be anti-thyroglobulin there will be 100+ different clones contributing to that, each with a different way of binding thyroglobulin and so a different electrophoretic mobility. Autoantibodies are invisible on standard electrophoresis.

 

[Snow Leopard]

Most autoantibodies are polyclonal so they do not show up as bands. Although all the antibodies may be anti-thyroglobulin there will be 100+ different clones contributing to that, each with a different way of binding thyroglobulin and so a different electrophoretic mobility. Autoantibodies are invisible on standard electrophoresis.

Ah, I misunderstood.

Still, now I'm curious as to what conditions would limit the response to a oligoclonal pattern vs polyclonal, but I guess that is a question for another time.

 

[Jonathan Edwards]

Ah, I misunderstood.

Still, now I'm curious as to what conditions would limit the response to a oligoclonal pattern vs polyclonal, but I guess that is a question for another time.

The common reason for one or sometimes two or three bands is a malignant clone producing antibody - as in lymphoma or myeloma. In conditions like Sjogren's and RA certain clones seem to break through the normal mechanism that 'evens out' how much antibody any one clone can make and you can get a handful of bands. The reasons for that are not very clear but if feedback loops are deranged maybe sometimes this rule can be broken too.

The reason for oligoclonal bands in CSF in MS or other brain disorders with lymphocytes in CSF is that all the antibody is being produced by a few clones sitting in the wrong place (meninges). Not much antibody gets in from plasma so all the background of millions of plasma cell clones producing millions of antibodies does not blot out the few local clones.

 

[Bob]

@Jonathan Edwards, excuse me if we've discussed this before (perhaps someone could direct me to the discussions if we have?), but might the apparent preliminary success that Drs Fluge & Mella are having with Cyclophosphamide, have any influence on your thoughts about ME and the mechanisms of rituximab in ME? I don't know anything about Cyclophosphamide except that Wikipedia suggests it works by acting on t-cells rather than b-cells. Also, Fluge and Mella say that Cyclophosphamide works quickly as a treatment for their ME patients (but I don't know how quickly they mean) suggesting that it doesn't rely on waiting for autoantibodies to be naturally depleted over time.

 

[leokitten]

If we postulate that ME in the chronic phase is due to a hypersensitivity of brain stem/autonomic circuits to very low level immune danger signals it might be that what generates symptoms are the normally trivial levels of cytokine or other signal that occur as part of normal daily clearance of toxic material and low grade microbes.

Is this an alternate theory as to why Fluge/Mella/Newton see vascular dysfunction? That it might not be an autoantibody to vascular epithelium but actually coming from the brain stem?

 

[Jonathan Edwards]

@Jonathan Edwards, excuse me if we've discussed this before (perhaps someone could direct me to the discussions if we have?), but might the apparent preliminary success that Drs Fluge & Mella are having with Cyclophosphamide, have any influence on your thoughts about ME and the mechanisms of rituximab in ME? I don't know anything about Cyclophosphamide except that Wikipedia suggests it works by acting on t-cells rather than b-cells. Also, Fluge and Mella say that Cyclophosphamide works quickly as a treatment for their ME patients (but I don't know how quickly they mean) suggesting that it doesn't rely on waiting for autoantibodies to be naturally depleted over time.

I don't think we have discussed it.

The trouble with cyclophosphamide is that it induces death of a proportion of a whole range of cells - T cells, Tregs, b cells and plasma cells. A proportion of B cells are very sensitive to cyclo. It may be that short lived plasma cells are sensitive to cyclo - which would explain an earlier response through earlier antibody decline. But cyclo also knocks down neutrophils and monocytes and so has a direct rapid anti-inflammatory effect. For instance in RA it produces a rapid effect that has gone by 3 months. So the bottom line is that it is not a drug that one can easily use to infer mechanisms from.

 

[Jonathan Edwards]

Is this an alternate theory as to why Fluge/Mella/Newton see vascular dysfunction? That it might not be an autoantibody to vascular epithelium but actually coming from the brain stem?

I guess that the change in flow mediated dilatation could be mediated by autonomic signals coming down from the brain stem but I am not sure. If this vascular dysfunction stands up on repeating it is rather strong evidence for a continuing peripheral immune drive maybe.

 

[Kati]

I don't think we have discussed it.

The trouble with cyclophosphamide is that it induces death of a proportion of a whole range of cells - T cells, Tregs, b cells and plasma cells. A proportion of B cells are very sensitive to cyclo. It may be that short lived plasma cells are sensitive to cyclo - which would explain an earlier response through earlier antibody decline. But cyclo also knocks down neutrophils and monocytes and so has a direct rapid anti-inflammatory effect. For instance in RA it produces a rapid effect that has gone by 3 months. So the bottom line is that it is not a drug that one can easily use to infer mechanisms from.

@Jonathan Edwards would there be greater benefit (albeit more risky) in cyclophosphamide vs Rituximab in resetting the immune system by apoptosis of a greater range of cells? did Fluge and Mella elaborated on that?

i am interested in whether non-responders to Rituximab would benefit from Cyclo because a greater range of immune-related cells are killed.

thank you.

 

[Jonathan Edwards]

@Jonathan Edwards would there be greater benefit (albeit more risky) in cyclophosphamide vs Rituximab in resetting the immune system by apoptosis of a greater range of cells? did Fluge and Mella elaborated on that?

i am interested in whether non-responders to Rituximab would benefit from Cyclo because more immune-related cells are killed.

thank you.

I think the chances of resetting the immune system with cyclo are not very great unless you using bone marrow conditioning doses (like 10 grams) and barrier nursing. So that is a good point but I suspect they will not get long remissions with cyclo maybe (you never know though).

 

[Kati]

I think the chances of resetting the immune system with cyclo are not very great unless you using bone marrow conditioning doses (like 10 grams) and barrier nursing. So that is a good point but I suspect they will not get long remissions with cyclo maybe (you never know though).

But wouldn't a responde at low dose provide further clues as of patho-physiology of the disease?

 

[Jonathan Edwards]

But wouldn't a responde at low dose provide further clues as of patho-physiology of the disease?

I am not sure. Cyclo is so non-specific in its targets.

 

[nandixon]

I mentioned above that:

[It was just recently (2014) found that Tregs, unlike other T cells, lack the ABCB1 transporter/efflux pump and thus are especially vulnerable to cyclo due to a resulting increased intracellular concentration of that agent. (Ref 2)]
...
2. Human regulatory T cells lack the cyclophosphamide-extruding transporter ABCB1 and are more susceptible to cyclophosphamide-induced apoptosis (Note: Abstract only.)

So, in theory - at low levels of cyclo - one would think that Tregs should be killed more efficaciously and more selectively than other T cells. Quoting from the above reference:

In vitro, the ABCB1-substrate CPA [cyclophosphamide] was cytotoxic for Treg cells at a 100-fold lower dose than for nonregulatory counterparts... Thus, Treg cells lack expression of ABCB1, rendering them selectively susceptible to CPA.

I'm not sure what other types of immune cells might also not express ABCB1, though (if there are any others that don't).

Note: ABCB1 is the name of the gene that makes a protein known as the p-glycoprotein transporter which is responsible for clearing out many toxic substances (e.g., drugs), but also some endogenous compounds, from within cells.

 

[greeneagledown]

I suspect they will not get long remissions with cyclo

Given its risk profile, it seems like cyclophosphamide isn't going to be worth using at all in CFS if it's not going to produce long remissions. Even if it leads to remissions lasting for a year or two, it seems like it would be exceedingly dangerous for someone to do a 6-month, 6-infusion cycle every one or two years -- so dangerous that it's not really an option.