ITP Types and Treatments by HAL

I've completed my retrospective study of PDSA user posts since 2010 and consolidated the results into a web page. Here is the link:
bottools.com/Hal/ItpTypes.html

Roughly 200 user stories were read and analyzed of which about 50 users had provided enough detail to support the conclusions given. There is no contradictory evidence that I know of. All 'unusual responses' could be accounted for by theorizing two ITP types were active at once or sequentially. Responses of those with other concurrent maladies may have been discounted in some cases.

This thread can be used to ask questions and report problems or suggestions. Cheers to your TPO agonist of choice

Wow Hal that must have taken some time to put together. Nice to see you included my link to the Nurses handbook.

How do you know which antibodies people have or did you base that on articles? Have you been able to confirm ITP antibodies with the responses that you quoted? What about people that have more than one antibody?

I was tested years ago and was told that I had a high presence of antibodies, but didn't know enough to ask which ones. I was newly diagnosed. Other antibodies have been discovered since then anyway so my information isn't complete even if I did try to find out now.

mrsb, yes it did take several months. But it was very intriguing reading the stories. Mesmerizing really. After a few weeks I thought about popping popcorn before I would start reading posts. Like a good book, it was hard to set some folks stories aside.

Looking back. What I could do, with just me alone, to produce some useful result took awhile to gel. Originally I was thinking of a prospective study where folks could enter their story data/drug responses into a database with a web page interface. This, as 'evidence based medicine', was running through my mind on this post , back in Dec last year.

Then Rob's post on the study which claimed there are exactly four ITP types. With such a small number, the ITP type problem seemed 'tractable'. And with that, a retrospective study could be started immediately without having to collect new data.

Anyway. I really like the nurse's handbook. Lots of good information in one place. I found myself referencing tables 5 and 6 often in trying to get up to speed about the different treatment regimens. Dosages, time to response, duration of response, toxicities, and such.

Sandi, those are very important questions.
It should be obvious I have no info on the antibodies a user may have. Folks rarely report that. So yes, the antibody associations are based on conclusions made in studies, reasonable deductions, and a process of elimination - based on the theory that only 4 antibodies account for all ITP cases as described here .

Let me do this as it came to me, and start with type 2. See this post and the study referenced. There you can see that the GPIb-IX antibody (type 2) is associated with poor steroid and IVIG responses. This is because of 'desialylation' process which occurs in the liver. IVIG doesn't affect the liver, it affects processes in the spleen, which explains the poor IVIG response. Also, I've read in an old study that good IVIG response predicts splenectomy possible success. This is corroborating evidence to the liver / IVIG connection.

Next was type 1. Unfortunately I have lost reference to the pertinent study - can't find it again. Blah. Basically it said something along the lines that ITP'ers with GPIIb-IIIa antibodies were typically associated with a good steroid response.

For TPO antibodies, type 3, no studies were used. I think it is reasonable to deduce it though. The logic is suggested in this post . The basics of the idea is that folks with TPO antibodies have a stronger Promacta response than other ITP types. This is because increased platelet production (due to Promacta) does not cause the immune system to ramp up antibody production - as it does in GPIIb-IIIa (type 1) response. In type 1, platelet increases cause a ramping up of the immune system, which causes a higher and higher Promacta dose, until eventually the immune system response is overcome and counts rise substantially. In type 3, immune system destruction of TPO is independent of platelet count and so an immune response doesn't have to be overcome. Finally, it is noted that ITP'ers with a strong Promacta response also have a strong IVIG response and a poor steroid response. So the type 3 steroid and IVIG response is backed into from actuarial Promacta response.

For type 4, TPO (cMpl) receptor antibodies, again, deductions were made. The idea was first suggested here . The idea is that TPO receptors do not circulate through the blood stream and never make it to the spleen - as glycoprotein and TPO antibodies do. The spleen is where IVIG retards antibody responses so there shouldn't be any sort of substantive response similar to the glycoprotein and TPO antibodies responses to IVIG. Because of this dynamic, and a process of elimination, the unique and odd 'weak' IVIG response characteristic of some ITP'ers is assigned to type 4.

Roomer has it there will be lab tests for the newly discovered antibodies soon - once it gets out of 'research' stage.

The usefulness of the study isn't really about antibodies. That column in the table can be ignored if desired. LOL, still I have answered your question, which makes for good documentation. The useful information is that steroid response X, and IVIG response Y, leads to a tailored response plan Z. Cheers to Probiotics !

It would be great if you could line up the antibodies with the treatment responses. I get it though, you don't have that information. The main thing that sticks out to me is that many people may have more than one antibody, which would really limit their responses and available treatments. We see very few refractory patients. I'm not saying that your information and deductions are not correct, but you'd think if you could work this out then the researchers should have too.

It will probably be many years before doctors perform routine antibody tests even after they become available. Everything with ITP takes forever to trickle down.

I just went through a folder with 19 years of lab results hoping I had a copy of my antibody test. No such luck.

As the four ITP type paper puts forth, those in remission still have antibodies present. Apparently the antibodies are held back (Treg?) by our immune system correctly recognizing the antibody is an attack on self. Besides the presence or absence of specific ITP antibodies, it is a matter of regulation and dis-regulation of those antibodies too. An antibody in regulation is asymptomatic.

In single ITP type/antibody dis-regulation knowing the antibody, or, knowing the first line drug responses is sufficient. In multiple ITP type/antibody dis-regulation, knowing just drug responses is less effective. In those situations knowing which antibodies are in play and knowing drug responses should make it vastly easier to determine which antibodies are in dis-regulation and which are in regulation. An analysis of drug 'partial response' and expected type responses becomes invaluable.

I can't speak for researchers. I can only guess doctors with a practice are too busy to notice and categorize nuances in first line responses - which may have further implication. Certainly for me, as a ITP novice, I would not have been able to go very far without knowing there are exactly 4 ITP types. For all I knew, there were 100 ITP types and the problem is completely intractable. It is the 4 ITP type paper, published in Dec 2016, that changed the landscape.

An aside. I have not published this information with the pretense that it is fact. I am publishing it so that it can be tested by others for possible validation or invalidation. If one doesn't publish, there are no advances. Same for testing - has to be done. The, now common, prevalence of 'software bugs' seems to have enlightened the general public as to this reality. A software bug means testing has been inadequate.

Hal, I do believe that there is something to this. However, I think it would be more accurate to state that there are four ITP antibodies rather than four ITP 'types'. There are probably people who have more than one antibody and in that case, you could start pairing them together and end up with more 'types'. We have to be careful making these characterizations here because we don't want people believing that there are four ITP types and telling their doctors this. We don't have the ability to categorize ITP nor do we want to start rumors. I think the research is headed in this direction and hopefully at some point we will have written articles and data to support it. When that happens, we can start sharing that information with doctors who might be unaware of matching treatment options. There are still only 40% to 60% of patients who have detectable antibodies (possibly due to undiscovered antibodies) and some of the tests are not available yet, so this may be a long way off. The info that you put together is amazing and I can see the time put into it. We just have to be careful how we phrase things.

I dont see this study and these charts reflected-
www.haematologica.org/content/100/5/623

www.haematologica.org/highwire/markup/65480/expansion?width=1000&height=500&iframe=true&postprocessors=highwire_figures%2Chighwire_math%2Chighwire_inline_linked_media%2Chighwire_embed

www.haematologica.org/content/haematol/100/5/623/F5.large.jpg?width=800&height=600&carousel=1

Jay,
I don't know the steroid, IVIG, or other drug responses of those in that study.

I know you are very concerned about TPO and how it relates to ITP. FWIW, I do find your case a bit odd in that department. I'm pretty sure you're the only person I have found on PDSA forum that has a great response to Nplate but never responded to Promacta. Those that don't respond to Promacta have needed a sizable Nplate dose.

HAL,
I've been taking a break from PDSA for a while now, and have just returned to find the results of your fascinating analysis. Well done! I gladly hand over to you the PDSA pit bull ceremonial tiara!
One question: under "other effective treatments" there is no mention of Rituxan for any type. What can I infer from this? Is Rituxan more effective for any particular group?

Rob, you should pop in more often. I haven't felt up to date in the latest studies with your absence.

For the 'other effective treatments' column I guess I was thinking 'other effective maintenance treatments'. Do I need a header change?

Ha, I published the table so that I can retire from the forum and not leave work incomplete. Or at least, take long breaks.

I wish I could get access to pre 2010 postings. It probably contains much more complete information on Imuran, MMF, Cyclosporin, Dapsone, and Danazol treatment responses. Once Promacta/Npate came out it looks like these other treatments have become very rare - and rightfully so. I also wonder if old postings have buried in it a remission drug for row/type 4. It bothers me I don't have this. Perhaps some folks in the past have achieved remission with Cyclosporin.

Something I'd like to call to your attention as you mentioned thoughts on it in the past and I didn't realize the importance then. Note the following sentence in the 4 ITP type study .
"
... 4/42 (10%) had anti-TPO autoantibodies only and 5/42 (12%) had anti-cMpl autoantibodies only and 15/42 (36%) had both.
"
Hell, there are more folks that have both type 3 and type 4 antibodies (36%) then those that have only type 3 and only type 4 (10+12). That is, type 3 along with type 4 is extraordinarily common. And with a cursory look at populations in this forum, it seems true. OUCH !
What does this mean? A poor steroid response and a IVIG response between the two types, or about 2-3 weeks, is common.

HAL, I was being dense. I understand now what you meant regarding Rituxan. Its remission is usually temporary and/or partial, so I was thinking of it as a treatment that occasionally gives long term remission, but a remission is a remission, regardless.

Perhaps you might add some combination rows to your table, especially Type 3 plus Type 4, since that combo is so common. You said that would be a "good" IVIG response (between the two)? I would have guessed the worse of the two, or "Weak". What is the Typical Promacta Response? Intermediate, or the highest dose?

Is there any way to aggregate your raw data and post it so that someone else might be able to see your work? For example, on what basis were people assigned to a specific type? How many people fell into each type, and how well did your data fit your model?

Have you summarized somewhere the theoretical or statistical basis for assigning each combination of steroid/IVIG response to a specific antibody type?

Hal
Over here in the UK IVIg is generally not used as a first line treatment unless count it dangerously low/bleeding present or high risk on initial diagnosis . I have never had IVIg so where am I going to fit into your classification?

Rob16 wrote: ...Perhaps you might add some combination rows to your table, especially Type 3 plus Type 4, since that combo is so common...

Yes, Sandi mentioned that too. I think the topic of combination antibodies and the resulting responses is a complex/advanced issue and would be better served by an entirely separate and new page. A page which would be linked to.

Rob16 wrote: ...You said that would be a "good" IVIG response (between the two)? I would have guessed the worse of the two, or "Weak". What is the Typical Promacta Response? Intermediate, or the highest dose? ...

This is an interesting question. A question which needs due diligence and a good answer. Can you convince me the response is the worse of the two instead of the sum of the two?

As I see it, if the combination response is the worst of the two responses:
- partial responses and partial remissions would not happen
- a 'poor steroid and 2-3 week IVIG' response would not happen (which it does), or, such a response could only be explained by yet another ITP type
- which antibody was dominant could change often and cause the response to look like one antibody was responsible for one treatment instance, and then the other antibody for another treatment instance. Gross response inconsistency, which does not happen.
- immune suppression drugs would likely change the dominant antibody with ease - then radically change one's counts with similar ease

Rob16 wrote: ...Is there any way to aggregate your raw data and post it so that someone else might be able to see your work? For example, on what basis were people assigned to a specific type? How many people fell into each type, and how well did your data fit your model? ...

Type assignments were made per the response definitions and the definitions were created from observation of the data. Obviously the data is only what someone reports, is not unbiased, and is often incomplete. This is the major problem innate to 'retrospective' studies.
Being the creator of the table, and own the reputation of it's accuracy, I would prefer to validate, invalidate, or fine tune the table going forward in time as folks arrive in the forum. LOL, I am not trying to speak for the PDSA or leverage its reputation. The forum is for educational purposes only. I am not a doctor of medicine, nor do I play one on TV.

Rob16 wrote: ...Have you summarized somewhere the theoretical or statistical basis for assigning each combination of steroid/IVIG response to a specific antibody type?

I've tried to explain the basis of the assignments already, earlier in this thread. If the basis looks weak, it is because it is weak.

Keep in mind, 'HAL9000' is the name of a fictional character - not the name of a real person.

I don't have a good answer for you mrsb. Best would be a process of elimination.
If you have a good IVIG response then row 1 would apply, and row 2b would apply with a poor IVIG response. So if you are on Promacta for a couple of years and dosage stays the same that implies row 1. Row 2b is implied if the dose decreases. Did your Nplate dose reduce over the months it was taken? AFAIK, losing Nplate response after a few months doesn't mean anything special.

The other possibility is an IVIG response somewhere between the two and would imply two ITP antibodies. From what I have seen, you would likely have a major change in steroid response sometime over the time you've had ITP for two antibodies to be the situation. I don't recall you saying anything like that so this doesn't seem to be the case.

Hal
My N plate dosing initial dose 1mcg/kg for first 2 weeks, then up to 2mcg/kg for 4 months then 3mcg/kg for 5 months followed by a loss of response.

Hal, I can guarantee you that pre-2010, there was very little information here regarding Imuran, MMF, Cyclosporin, Dapsone, and Danazol. Those treatments have always been rarely used. MMF and Cyclosporin have been used even less than the other three; MMF generally only for children and I've only seen it a few times over a span of 19 years. There is also CellCept, Vincristine and Cytoxin, also rarely used (thankfully). None of those treatments have seemed to work well and are usually last ditch effort drugs. You're not missing much.

There was a Rituxan explosion around 2002 and nearly everyone was using it. We actually had a scoreboard tracking patients and responses which was done by one of our members. His wife had ITP but she tragically died of sepsis. He left soon after and the scoreboard ended. The most popular treatments before the TPO's were always IVIG, Prednisone, Rituxan, Win-Rho, Decadron and splenectomy. Win-Rho got a black box warning and most doctors stopped using it, even though Rituxan and IVIG also have black box warnings.

Just a heads up.. MMF is the proper name for Cellcept, so they are the same drug.

I was thinking that as I wrote it, but was too half-asleep to look it up and know for sure. I should have remembered that; I took the darned stuff.

So did I. It didn't work and made me depressed.

I used it for Lupus, but it didn't work for that either. I wasn't able to think on that drug. I felt like my brain went black and I'd just sit and stare. Could not do my job while I was on it and quit the meds.

Oh wow mrsb. I wonder if increasing dose is a pattern which indicates upcoming loss of response.

Sandi, did not expect that. Must have been a lot of IVIG consumed back then. Gosh, and the folks that had no IVIG response? I haven't seen any immuno suppressive drug work for them. Brutal.
Speaking of WinRho. From the accounts I've read, that seems to be a real dud of a drug for ITP. I guess the only saving grace is that it is cheaper than IVIG. For Vincristine and Cytoxin, aren't those only tried when nothing seems to work to stop active bleeding? Too dangerous/risky to use?

Anon, leading up to your remission, how did your Nplate response go? Always decreasing, or, one day just stopped needing Nplate?

I don't like high counts on Nplate so as soon as my count was over 100 I cut the dose, then when I was on the lowest dose possible, cut the weeks, and then stopped as my count stayed up. I had no response to IVIG but do respond to immunosuppressants by the way.

Hal - Yes, there were people who had no response to anything prior to the TPO's. Those people usually managed with IVIG if they could. Some people just learned to live with low counts. Actually, Win-Rho worked quite well for a few people here and they were able to manage ITP with a Win-Rho treatment every two to three months which wasn't bad. Back then it was just a 5 minute IV push and they'd be on their way. I tried it; it didn't raise my counts at all. Vincristine and Cytoxin didn't stop active bleeding because they didn't work that fast. Those are a series of treatments that may or may not work to cause remission. Those are true chemos which had potential serious and permanent side effects, so they were not to be taken lightly. Personally, I would have never considered those and I'm glad they are not used very often.

Anon, you responded to immuno suppressants? I would guess that if you responded to steroids that other immuno suppressants, like Cellcept, could have worked. Was it a strong response - as strong as steroids?

Sandi, looking through my notes of the many people that tried WinRho, I found 9 people who had a good response to it - one unknown type, one type 3 and the rest type 1. Of those 9, two experienced some sort of remission with it.
Interestingly, for drug comparison purposes, everyone that tried Danazol and was type 3 (two people) had some sort of remission with it.
Right, of course Vincristine and Cytoxin treat ITP, not bleeding. Some folks who have ITP, don't respond to the more common drugs, and have a bleeding control problem because of the ITP, often receive those drugs. From what I've seen, these more extreme and dangerous drugs are used because the bleeding has to be controlled.

Hi Hal,

Is IVIG response weak or strong?
Am I Type 3 or Type 4.
Sept 13\14 <IVIG Given + Transfusion+Steroid>
September 21, Count 70 (80mg prednisone)
September 26, Count 45 (80mg prednisone)
Oct 7, = 31 (80mg prednisone)]

Nov 15\16 <IVIG Given + Transfusion+Steroid>
Nov 22, Count 113 (80mg prednisone)
Nov 29, Count 15 (80mg prednisone)
NPLATE STARTED

Thanks,
Jay

Jay, some data can be deceiving. For example, you were given the IVIG in a hospital, in the context of a emergency and not as treatment regimen from a hematologist, and they may have given you a small (single bottle) amount of IVIG. Plus the steroid and IVIG combo pollutes the data and makes it difficult/impossible to discern the IVIG response from the steroid response.

But, as you know, your Nplate response is phenomenal, which is typical of type 3. But also, there is a small number of type 1 folks that respond that well too, but their steroid response is way, way better than yours. So type 1 doesn't fit. A type 2 Nplate response would start out average and end up with a strong response.

Oh, by the way. I mentioned to you before that I thought it was odd that you didn't respond to Promacta but had a strong Nplate response. I was thumbing through my notes and found at least one and perhaps a second person besides you that has a similar response. That is, no Promacta response and yet a strong/good Nplate response. So it seems some folks just doesn't Promacta from an individual physiological reason.

Does this help? Still thinking of splenectomy?

My Wife is pushing me to get splenectomy done. My Father in Law who is physician is against stimulating bone marrow too much to produce platelets.

I was worried I was losing response to NPlate but recent tests show Iam not. Until now Iam holding steady with 3 week intervals but still there is still destruction.

I had annual followup.
I asked my Hematologist for Danazol but she is against it saying it is a nasty medication.
She is pushing me to do Rituximab, I ask her with such low steroid response what is point of having rituximab.