hemorrhage in a Bombay

[Mabel Adams]

I was talking to a colleague that is studying for her SBB. We were discussing what one should do if a previously unknown Bombay patient should come in with a life-threatening hemorrhage and it was not possible to get Bombay units in until after the patient had to be transfused. Let's say that uncrossmatched blood was not needed so there was time to realize that there was a significant problem before units were chosen--maybe a severe upper GI bleed. I'm sure this will not happen to most of us in our lifetimes but it is an interesting mental exercise, nonetheless.

I think it would be good to start with thinking about how this would present in our own testing protocols. For instance, I think tube testing including IS would give more clues sooner than we would see in MTS gel testing.

[Malcolm Needs ☆]

One for the Pathologist-in-Charge of Blood Transfusion to work out, me thinks!

It would be a clinical decision for one thing, and, for another, that's why they get paid the big bucks.

One would like to say, give AB, as those units have the least H antigen expressed, except that Oh individuals also make anti-A and anti-B, but, in a situation like that, one would have to do something, so, to prevent the patient bleeding to death, I would give group O, with a HUGE warning to the docotor looking after the patient, and only under guidance from the Pathologist-in-Charge of Blood Transfusion.

Then I would cross every finger, every toe, both arms, both legs and hope like Hell!!!!!!!!

[PAWHITTECAR]

I'm with you Malcolm. Way above my pay grade. I think I would encourage them to transfer the patient to a facility better equiped to handle it. You know one with a pathologist in house to handle it.

[Mabel Adams]

Would it do any good to give AB (or a mix of AB & O) plasma in hopes of giving soluble ABH substances and absorbing out some of the antibody? Giving fluids to dilute out some of the antibody first???

My pathologists would need all of the input they can get. Hence my idea of creating a repository of such on this site in advance of need.

Transfer of the patient might not be fast enough here--or in Alaska.

[Malcolm Needs ☆]

Would it do any good to give AB (or a mix of AB & O) plasma in hopes of giving soluble ABH substances and absorbing out some of the antibody? Giving fluids to dilute out some of the antibody first???

No, to the first bit, but I suppose one could try giving FFP, in the hope that it would inhibit the anti-A, anti-B, anti-A,B and anti-H, together with diuretics - but I fear it would be a forlorn hope.

Even then, I would think you would need high dose IVIG and maybe methylprednisolone cover and sincere prayer.

Edited March 16, 2013 by Malcolm Needs

[Mabel Adams]

Why not give AB plasma? It should have soluble A & B antigens. Does it have less soluble H antigen or is there just less H on AB cells due to steric interference rather than a reduced production of the H antigen?

[Malcolm Needs ☆]

Steric hinderance - the L-fucose is still present.

As long as the AB donor is a secretor, I agree.

[Mabel Adams]

I figure if we gave several units of plasma the odds were good of getting mostly secretors. So no advantage to giving O plasma because the H substance is not more prevalent in O than in AB plasma? By sticking to AB we should be giving A, B & H substances in all units from secretors. After that, it sounds like we should just follow the advice on the California Blood Bank Society site for giving ABO incompatible blood as a last ditch effort as this would be very much like it. http://www.cbbsweb.org/enf/2003/urgenttxaboincompat.html Glad it will almost certainly never happen to me. Is there a role for IVIG or some other immunosuppressant?

[Malcolm Needs ☆]

Is there a role for IVIG or some other immunosuppressant?

Yes, probably, but not definately (see post 5).

I hope it never happens to me either, but it came close a couple of times.

A few years ago, one of our hospitals sent us a sample on a patient going for a THR (cold surgery) with an antibody. The Ortho surgeon decided to go ahead anyway (even though we were working on the sample at the time. The antibody turned out to be anti-H (strong too) and the patient Oh. Fortunately, the patient did not require a transfusion.

The other was a patient involved in an RTA (leg fractures), but with other underlying pathologies, who was an Oh, R2R2 of child-bearing age, and the only blood available in the UK was frozen Oh, R1R1, but again, fortunately, the patient did not require transfusion in the end.

[Whitney Poplin]

Hi Mabel-

The thought that I had was that the soluble A and B substance in AB plasma would create immune complexes with the patient's IgM Ab, which can bind autologous platelets and maybe even knock out the kidneys?? I'm not sure if this would even be a thought or a consideration in a massive transfusion situation.

[Malcolm Needs ☆]

Very good point Whitney, but then a massive acute transfusion reaction wouldn't do the renals the world of good either.

I honestly don't know the answer to Mabel's original post; I'm just making suggestions myself.

[Cliff]

Woohoo, the post right above this one is the 50,000th post!!!!!

[Malcolm Needs ☆]

Yee ha!!!!!!!!!!!!!!!!!!!!!!!!!

[Mabel Adams]

Surprise, surprise that it was Malcolm's, eh? Let's see, his 4000 posts out of 50,000--not quite 10%. But we are glad to have him sharing his copious wisdom so freely.

Back to the thread...

There are those that worry about the effects of immune complexes but we cause them every time we give O RBCs to non-O patients. Of course the residual plasma in an adsol unit doesn't have too much antibody, but, if the recipient is a secretor they are certainly making at least some immune complexes. Even more so when we have to give an O pheresis plt to a non-O patient. I will have to look again at the recent Transfusion article on the effects of mismatched plts and see what it says on the topic.

Malcolm, Whitney is the SBB student that started this thought.

[Malcolm Needs ☆]

Malcolm, Whitney is the SBB student that started this thought.

In that case Mabel, with a mind as enquiring as hers, she should go a very long way in her chosen profession, and I wish her all the very best for the future. I am very impressed.

[Dr. Pepper]

I think it would be good to start with thinking about how this would present in our own testing protocols. For instance, I think tube testing including IS would give more clues sooner than we would see in MTS gel testing.

Great thread so far, Malcolm & Mabel & Whitney. And congrats to the site and users for #50,000. While "our" famous Bombay patient was still alive and around, this scenario was our Sunday third shift nightmare. But getting to Mabel's original point, you do have to recognize the problem in the first place. Bumping into a Bombay is not an every day occurance for those of us who don't work in Malcolm's lab. Perhaps seeing tube rxn across the board would give more of a hint that it's anti-H than gel would, but how would you know to go that way? I think if you find an apparent group O with a strong pannaglutinin, WITH A NEGATIVE AUTO CONTROL, you can take the chance and type the cells for H. Whether it turns out to be anti-k, Lub, etc or anti-H, you will still end up in the same serological boat and have to give incompatible blood. I've never had to do the last ditch give-them-anything protocol myself, but an uncertain outcome has better odds than 100% bleeding to death.

[Malcolm Needs ☆]

Believe me, Phil, seeing an Oh patient with a genuine alloanti-H is not an everyday experience in my lab!.

One of the things that may point you in the right direction is the name (or, if you are lucky enough to actually be told, the ethnic origin - so rare in the UK these days, as it is not considered "politically correct" to give an ethnic origin, even though, under these circumstances, it is one of the most helpful clues you can have - in fact, I've even experienced one hospital that, point blank, refused to give me an ethnic origin of one of their patients). If the name is obviously of Asian origin, then the chances are that an antibody that is not an auto may be an anti-H or an anti-Inb. If the name is obviously of African origin (or obviously Black American/Black UK), then the chances are that an antibody that is not an auto is an anti-U, an anti-Fy3, an antibody directed against hrB, hrS, HrB, HrS or Js(. If the name is obviously of White origin, then start by looking for anti-k. If it is obviously of Latin American or Far Eastern origin, then start by looking for anti-Dib. If it is obviously of Scandinavian (particularly Swedish) start by looking for anti-Vel, but if Finnish or Latvian, think of anti-MAR.

Obviously, this is only a start, but it does help by starting with these. You may have to go further (I know of one Black patient - through Joyce Poole - who made an alloanti-Kpb - and if you read the books, this is impossible!), but it is a start.

To be honest, about the only difference between tube and gel that I've noticed with antibodies directed against high-frequency antigens is that it is easier to recognise Lutheran antibodies in gel, because of the apparent mixed-field reactions that you get with such antibodies.

[Mabel Adams]

Aren't Jk3's associated with Pacific Islanders? And let's not forget PP1Pk patients. Do they fit this non-auto panagglutinin picture too?

One difference with an Oh compared to all of the other nasties above would be that the transfusion reaction would be more likely to be intravascular hemolysis, like classic ABO incompatiblity, right?

[Malcolm Needs ☆]

Aren't Jk3's associated with Pacific Islanders? Yes, but not as much as is often quoted in the text books. I think it was in Issitt and Anstee where they say that the original authors of the paper about Jk(a-b-) frequency in this area's population were wrong, as they failed to take into account kindreds.

And let's not forget PP1Pk patients. Do they fit this non-auto panagglutinin picture too? Yes.

One difference with an Oh compared to all of the other nasties above would be that the transfusion reaction would be more likely to be intravascular hemolysis, like classic ABO incompatiblity, right?

.............and yes, although, one must not forget that quite a few people survive ABO mismatched (major side) transfusions. The most common, but by no means the only "killer" is A into O, whilst other ABO mismatches (major side) seem to be slightly less deadly (although still viscious and are quite capable of killing the patient - and sometimes do).

[Malcolm Needs ☆]

Actually, for those that have not had the chance to read it yet, there is a superb list of high prevalence antigens absent in certain ethnic populations in the third edition of The Blood Group Antigen FactsBook (on pages 699 to 701).

[galvania]

One other thing that might be possible if there's time is see whether the patient has any siblings who are within reach, and test them. OK, so you will have a lot of administrative problems to deal with, but it's still worth thinking about.

[Changezi]

A vary interesting discussion , Its happend with us long before a lady came with massive Bleeding and she was bomby .

I was on duty what i did

okay i called all the known blood bank with in the city for help

called the donor's we have in our own blood bank record.

with in one hour we find two adult donor (brother and sister ) who gave blood and immidiatly transfused to the patient with out any checking directly. and as i remember we got the secound unit again from the same donor after checking Hb after 24 hour and then latter uptill that time we arranged it from nighbouring city and country

All to say we have list of Donor with Bombay blood Group Record with the blood bank . i dont remember how many and all of them knows about this and are always ready at any time to donat .

when ever we find new patient or donor with bombay we call that person and let him know every thing about his blood group and let him be with us in contact so by the time and time we keep screen them for HbsAG HIV..... and all others donor test . and we have coordinate with other blood bank within the city as if any one find bombay we informed each other and share the donor detail for any kind of emergency.

[Mabel Adams]

In our area with fewer than 300,000 people (mostly of European extraction) over an area the size of the Netherlands we aren't likely to find even one Bombay donor--and then we would have to bypass the FDA rules and find someone to draw the donor(s) before the patient bled to death. (I guess that limits our likelihood of having any Bombay patients too.) Sounds like it can work in a big city with the right genetic background and good planning like you have, Riaz, but still a stressful situation, I'm sure.

Good point about siblings, Anna.

I hope this thread becomes the "go to" spot for information on emergency transfusion of Bombay patients so keep adding ideas. Now how do we get Google to find it effectively?

[PAWHITTECAR]

How do you get Google to find anything effectively??:pcproblem

[Cliff]

I hope this thread becomes the "go to" spot for information on emergency transfusion of Bombay patients so keep adding ideas. Now how do we get Google to find it effectively?

Surprisingly, Google likes this site. Their spiders are always crawling it and when I am searching for blood bank related items, I often see this site listed.

Here's an example, search for Bio-rad tango. The Bio-rad company comes up first, but we'll be listed in the top 5 or 10.