Lack of platelet autoantibodies and non-responsiveness to Rituxan

There is not even an abstract for this study, but maybe the title says it all:

Lack of detectable platelet autoantibodies is correlated with non-responsiveness to rituximab treatment in ITP patients
www.bloodjournal.org/content/early/2017/05/03/blood-2016-11-751719
Leendert Porcelijn, Elly Huiskes, Martin Schipperus, Bronno van der Holt, Masja de Haas and Jaap Jan Zwaginga
Blood 2017 :blood-2016-11-751719; doi: doi.org/10.1182/blood-2016-11-751719

I did find an abstract, where the lead author presented this paper in January 2017 at a hematology conference. There was a strong correlation
between antibodies and response to Rituxan:

www.hematologiecongres.nl/en/program-registration/program-11dhc/program-11dhc/clinical-abstract-session-1/platelet-autoantibodies-and-response-to-rituximab-in-itp-patients
Introduction
We tested if response to rituximab in ITP patients, enrolled in a multi-center randomized open label phase II trial (HOVON 64), was associated with the presence of platelet-associated autoantibodies.
Methods
The HOVON 64 trial included adult patients with an ITP relapse or refractoriness (at least 2 platelet counts less than 30x109/L) and at least three weeks after high-dose corticosteroids (≥ 1mg/kg) before start of rituximab. Starting before the first rituximab dose, platelet-associated autoantibodies were weekly, for ten weeks, assessed by the direct platelet immunofluorescence test (PIFT). The direct monoclonal antibody immobilization of platelet antigens assay (MAIPA) was performed if sufficient platelets were isolated. Autoantibody results were compared with response to rituximab treatment.
Results
In 99 of 112 patients samples (88%), the PIFT could be performed. Of these 99 patients, 47 (47%) responded to rituximab while 52 (53%) were non-responders. Antibodies were present in 79 patients of which 43 (54%) responded with 16 (21%) complete responders. The absence of antibodies in 20 patients in contrast was associated with 4 (20%) responses (p=0.006) of which only 1 (5%) was complete. In reverse, 16 (94%) of patients with a complete response to rituximab showed positive to very strong positive direct PIFT reactions while considerably less (69%) non-responsive patients showed positive direct PIFT results. In addition to direct PIFT, direct MAIPA could be carried out for 30 patients (7 non-responders, 23 responders). Both PIFT and MAIPA results appeared in full concordance.
The negative predictive value (NPV) of undetectable platelet autoantibodies in the direct PIFT for less than complete response to rituximab was 95.1% (95%CI 75.1-99.9%). Response to rituximab treatment and increasing platelet counts coincided inversely with the presence and score of reactivity of autoantibodies (p<0.0001, r =-0.6251).
Conclusion
Rituximab-treatment related increase in platelet counts is strongly associated with a decrease in platelet autoantibody levels. Second, lack of detectable platelet autoantibodies at treatment initiation correlated with non-responsiveness to rituximab. These results support the presence of an antibody-independent platelet clearance in specific ITP patients and platelet autoantibody detection might enable a more individualized therapeutic approach in this group of ITP patients.

This is what I am excited about most for the near future in ITP treatment: being able to determine in advance what treatments are likely to be effective for a specific patient's particular form of ITP, saving time, money, and unnecessary treatment risk.

Gosh Rob, you and me both. I've been working on answers to those questions by sifting through PDSA forum member reports. But it is very, very time consuming. I think I've got the four ITP type signatures identified. Steroid and IVIG response seems to be enough. This is all a 'retrospective study' so it's difficult to make conclusions on others without being able to query the person. Therefore the ITP type of many, perhaps most, folks cannot be identified because of lack of information. Once the ITP type is identified one can then create lists of what drugs do and do not contribute to higher counts for that type. Perhaps a concerted effort would be more productive.

About this study. Do you suppose the author(s) don't know the theory that there are four types of ITP? That he only knows about 'GPIIb/IIIa' and 'GPIb/IX' antibodies only?

Thank you thank you thank you for doing the legwork to match up the paper with an abstract you found elsewhere! I'm still in my infancy of reading through these technical articles and every bit of help matters.

I am really interested in this in particular because of my experiences in my pregnancies. All three of my boys were born with my antibodies in their systems (temporarily depressed platelet counts which returned to normal within the first weeks after birth), which would seem to indicate that I am in the group that would be more likely to achieve some kind of a response with Rituxan. When I go for my consultation with the new ITP doctor this understanding may affect how I might want to proceed. I'm having a few mouth blood blisters (tiny ones, but still) which have me a bit down on my Promacta response at the moment.

Extremely exciting research coming out though! If we can do better than running through a list of treatments by rote and can instead group ITP patients by specific manifestations of antibodies/destruction mechanisms/etc., that will be huge for treatment options!

Here's my problem with all of it. I've been saying for years that there could be antibodies that have not yet been discovered. I actually thought that was a bit of an assumption until I saw Drew Provan say it in an article or video. I always thought it was a good explanation for ITP in people with no detectable antibodies.

I realized it wasn't too crazy of an idea when they recently discovered the antibodies that affect production (within the last 10 years or so). There is still so much that is not known about ITP, so I don't think we can presume to make sense of it with the available info just yet. When I was diagnosed, ITP was strictly a disorder of destruction. Production was not even an issue ever discussed as a possible reason for low counts. I will never forget the day I saw the first article about production being a factor. It was huge news! There is so much more to learn though.

I'm with you guys as far as hoping that one day soon we can have treatments better tailored to fit patients. Having to go through so many until one is found to work is ridiculous. My concern is always the toxicity, especially if it's unnecessary.

Sandi, There might be undiscovered antibodies in addition to the ones already known, but it appears that all (or virtually all) cases of ITP can be accounted for by the known autoantibodies. I previously posted this:

www.bloodjournal.org/content/128/22/2548?sso-checked=true&utm_source=TrendMD&utm_medium=cpc&utm_campaign=Blood_TrendMD_0
Autoantibodies to Thrombopoietin and the Thrombopoietin Receptor in Patients with Immune Thrombocytopenia
Conclusions: Testing the entire panel of autoantibodies that included anti-TPO, anti-cMpl and anti-GP, we were able to identify all patients with active ITP; however, we could not distinguish between patients with ITP and other thrombocytopenic syndromes.

Here is a study showing that lack of responsiveness to prednisone can largely be predicted by high levels of C-reactive proteins:

www.thrombosisresearch.com/article/S0049-3848(17)30034-8/fulltext
Increased C-reactive protein levels at diagnosis negatively predict platelet count recovery after steroid-treatment in newly diagnosed adult immune thrombocytopenia patients
Y. Rama Kishore, B. Prashantha'Correspondence information about the author B. PrashanthaEmail the author B. Prashantha, M. Girish, B. Manaswitha
DOI: dx.doi.org/10.1016/j.thromres.2017.02.012 |
Highlights
•An elevated CRP level amplifies antibody mediated platelet destruction thus aggravating thrombocytopenia.
•The diagnostic potential of CRP in predicting the severity and treatment response in ITP patients should be considered.
•Therapeutic opportunities targeting CRP levels could lead to better treatment outcomes in ITP patients in the future

Wouldn't it be nice to know ahead of time whether that awful prednisone even has a decent chance of working?

Another study on using antibodies to predict response to Rituxan:

onlinelibrary.wiley.com/doi/10.1111/bjh.14782/full
GPIIb/IIIa autoantibody predicts better rituximab response in ITP
Rui Feng, Xinguang Liu, Yajing Zhao, Yuanyuan Zhu, Jun Peng, Ming Hou, Chunyan Chen
First published: 23 May 2017 DOI: 10.1111/bjh.14782
In summary, our study showed that ITP patients with anti-GPIIb/IIIa autoantibodies were more responsive to rituximab treatment; therefore, autoantibodies might be useful predictors for rituximab response in ITP treatment.

Recently, I replied to Sandi:

Sandi, There might be undiscovered antibodies in addition to the ones already known, but it appears that all (or virtually all) cases of ITP can be accounted for by the known autoantibodies.

What I did not take into account is the possibility of sub-types of autoantibodies, which are probably not all known.
For example, in 2014, Sandi posted this study:

www.ncbi.nlm.nih.gov/pubmed/25231551
Glycoprotein Ibα clustering induces macrophage-mediated platelet clearance in the liver.
Thromb Haemost. 2015 Jan;113(1):107-17. doi: 10.1160/TH14-03-0217. Epub 2014 Sep 18.
Abstract
Many immune thrombocytopenia (ITP) patients, particularly patients with anti-glycoprotein (GP) Ib-IX autoantibodies, do not respond to the conventional treatments such as splenectomy. However, the underlying mechanism remains unclear. Here we found that anti-GPIbα N-terminus antibody AN51, but not other anti-GPIbα antibodies (AK2, HIP1, VM16d, or WM23), induced GPIbα clustering that led to integrin αIIbβ3-dependent platelet aggregation. After intravenous injection, AN51 dose-dependently induced thrombocytopenia in guinea pigs, and the platelets were mainly removed by macrophages in the liver. N-acetyl-D-glucosamine, previously shown to inhibit integrin αMβ2-mediated phagocytosis of refrigerated platelets, dose-dependently inhibited AN51-induced platelet clearance. Furthermore, AN51 but not VM16d, induced rapid platelet clearance in the liver of cynomolgus macaques. Five of 22 chronic ITP patients had anti-GPIbα autoantibodies, and the autoantibodies from four of the five patients competed with AN51 for binding to platelets. These data indicate that GPIbα clustering induced by anti-GPIbα N-terminus antibody causes integrin αIIbβ3-dependent platelet aggregation, phagocytosis, and rapid platelet clearance in the liver. Our findings reveal a novel Fc-independent mechanism underlying the pathogenesis of ITP, and suggest new therapeutic strategies for ITP patients with anti-GPIbα autoantibodies.

In this article there are five different sub-types of anti-GPIbα antibodies identified: AN51, HIP1, AK2, VM16d, WM23.
There is no reason to think this list is complete, and it seems reasonable to assume that there exist sub-types for the other types of antibodies as well.

Ha, getting too confusing for me! I'm glad they keep identifying auto-antibodies. When I was diagnosed, there were only two and the fact that I had them didn't mean anything other than a slight confirmation that I had ITP. There were many people who tested negative.

My CRP is always pretty normal and I responded to Prednisone. Wonder what that means? There are so many variables.

In my retrospective analysis of PDSA forum stories, the only hard and fast rule I have found in regards to Rituxan are those with a poor response to steroids and strong response to IVIG. That is, those that seem to have antibodies to TPO. In every case, they fail a full response to Rituxan. A partial response should be possible when a Rituxan treatable antibody is present concurrent with the TPO antibody.

From a treatment cost perspective, I think that is saying something.

So you're saying who responds to Rituxan? Those who fail steroids but respond to IVIG?

Rob, thanks for your research! Very interesting and helpful. Regarding the C-reactive protein study: I see how prednisone responsiveness is the focus of their study. But am I correctly understanding them to say that high levels of C-reactive protein correlated with higher platelet destruction? And so reducing C-reactive protein numbers might improve platelet counts? The line I'm reading is this-

"An elevated CRP level amplifies antibody mediated platelet destruction thus aggravating thrombocytopenia."

I have high C-reactive protein when measured 2 years ago. Don't know what it was on ITP diagnosis. Will be getting it checked again in a couple weeks. This is very interesting- if I'm reading it correctly. I have inflammatory arthritis, joint pain. So I'm very interested in getting my C-reactive protein numbers down and reducing inflammation. And after reading this study-- C-reactive protein being a marker of inflammation, if a person reduces inflammation through diet, exercise, whatever- perhaps they could improve platelet counts? Not sure if thats what its saying, but I like the idea.

here is a link if anyone is interested to general info on c-reactive protein testing and inflammation- heart disease, autoimmune arthritis etc.
www.drweil.com/health-wellness/body-mind-spirit/heart/elevated-c-reactive-protein-crp/

Oops, wasn't clear. Just the opposite Sandi. Those with a poor response to steroids and a strong IVIG response, consistently fail to respond to Rituxan. LOL, I am an example of this.

This notion of 'sub types' is very interesting to me Rob. Have been in a pickle about how to categorize the ITP type of folks with good steroid response but poor IVIG response. With sub types as an extension of GPIb/IX antibodies (type 2), makes this troublesome response reasonably part of that (type 2) population. Previously, I had flat line response to steroids and IVIG as the only response characteristic indicative of type 2 population. The reasoning for the flat line responses is because platelet destruction occurring in the liver, via desialylation, instead of the spleen.

Hello Hal,

Does the TPO-RAs work for you?
My opinion is Rituxin has a poor response rate overall about 25%. For the cost and side effects not worth the effort.

Jay, yes Promacta is working fine. Very sensitive to it just as you are to NPlate. I was re-reading your first PDSA post and drug responses. The super sensitivity I think is because our immune system is destroying TPO and not platelets. I suspect Rituxan would not work for you as it did not for me.

I was tested for anti bodies and it was negative. I did not tespond to IVIg or rituxan but great response to steriods, promacta and Nplate. Just trying to add to the discussion!

Emilyk, I've got you and 'ann' with the same ITP response/type in my records. Good steroid response, poor IVIG response, and no Rituxan response. Both of you have gone into remission with a TPO receptor agonist. Yea ! ! Best I can tell, if you want to go into remission with Promacta/NPlate, this response seems to provide the best chances of it occurring.

I don't have it in my notes. Was your IVIG response more like a flat line, or, did it just go up a bit, say from 13 to 30?

I do not have my notebook at home to give you numbers but I know they went up for 24 hours and then came right back down.

Poseymint - Sorry for the slow response, but I've been traveling.

Your comments on c-reactive protein (CRP) seem right on the mark. It is reasonable to conclude from this article that lowering an elevated CRP would improve platelet counts just as it improves cardiovascular risk and improves inflammatory arthritis. There is evidence from cardiology that lowering CRP even below normal levels is beneficial, so this may apply to ITP as well.

The best ways to reduce CRP are diet and exercise. Exercise lowers CRP even if there is no weight loss, but together they benefit CRP even more.
www.ncbi.nlm.nih.gov/pmc/articles/PMC1336715/
C-Reactive Protein, Inflammation, and Cardiovascular Disease

drhoffman.com/article/12-natural-ways-to-protect-your-heart-and-lower-your-crp/
12 natural ways to protect your heart and lower your CRP