autoantibody cross match advice

[Lisa J]

Hi there, I would be really interested to hear other blood bankers opinions on the following scenario:

We have a patient with an autoantibody who is transfused 2 units red cells approx once every 2-3 weeks. I have always understood that patients with autoantibodies should have the autoantibodies adbsorbed from the sample (either in house or by a reference lab) in order to detect any underlying red cell antibodies that may be produced, and that this should be perfromed on every occasion that the patient requires blood to ensure that the lab is detecteding any recently produced red celll antibodies from recent transfusions.

Some colleagues of mine have stated that this does not need to be done and that you can cross match red cell units against the patients plasma and as long as they are less reactive than an auto then it is OK to issue units to the patient. They do not do panels on this patient as the autoantibody results in psoitive reactions in all panel cells so 'it is of little use'.

What do other labs do in this scenario? I am not happy to issue units in this way and would prefer an adsorbtion to be done first and then run a panel to determine the presence of any other underlying masked red cell antibodies and then do a cross match. If the patient has a very weak underlying red cell antibody that is masked by an autoantibody and the unit that is selected for xm has a heterozygous expression of the corresponding antigen, surely this weak reaction in the xm could be masked by the presence of the autoantibody and the reaction may well be as weak as the auto leading to the issdue of the unit and a possable delyed transfusion reaction?

Would be very interested to hear others opinions

I

[Malcolm Needs ☆]

Lisa, you are absolutely 100% correct that the plasma of the patient should be alloadsorbed each time and an antibody panel performed.

I'm not sure if you are the Lisa I know (in the Tooting area?), but if you are from the UK, you will know that your colleagues are NOT following BCSH Guidelines, and are taking risks with other peoples' lives. They would find themselves in very hot water with the CPA and MHRA, and if they had to defend themselves in a court of law,....well good luck.

All that having been said, individuals with autoantibodies tend either to make lots of antibodies, or very few indeed. It is probably to do with the condition of their own immune system (because of the causative pathology, or the drugs they are on).

[Lisa J]

Many thanks Malcolm, just needed to hear what I already thought....and yes I am Lisa from the tooting area! But no longer in Tooting....now in Kent! But still working in London !

[Malcolm Needs ☆]

With Marion I believe.

[irshadaad]

The practice should be we first look for antbodies present both allo and auto and try our best to prvide antigen negative blood so that aftermoth will be little problematic especialy those pts which are regularly on transfusion its mostly because of a delayed transfusion reaction which cause heamolysis in a slow phase and same thing happens when you give same antigen as the auto anti body so you are creating an unending delayed reaction in the pt.......

[Malcolm Needs ☆]

The practice should be we first look for antbodies present both allo and auto and try our best to prvide antigen negative blood so that aftermoth will be little problematic especialy those pts which are regularly on transfusion its mostly because of a delayed transfusion reaction which cause heamolysis in a slow phase and same thing happens when you give same antigen as the auto anti body so you are creating an unending delayed reaction in the pt.......

I am sorry, but this advice is completely against any that I have ever read.

If you are giving blood that is antigen negative for an auto-antibody (which is pretty difficult in the case of a warm auto-antibody, as most of these are mimicking specificities, but actually antibodies directed against a high incidence antigen), you will usually be giving blood that is positive for an antigen that the patient does not express. In doing so, not only will you be exposing the patient to sensitisation, resulting (possibly) in the production of an alloantibody, but, if such an antibody is produced, you will be making it far, far more difficult to determine antibody specificity in future.

In the case of, for example, an auto-anti-e (or, more probably, an auto-anti-e-like antibody), you will be exposing the patient to the E antigen (homozygous expression) and if they produce an alloanti-E, then what are you going to do?

I know that if the auto-anti-e is virulent enough, you may have to switch to e negative blood, but you do not want to do this before you absolutely have to.

[David Saikin]

I like Malcolm's advice a lot!

[irshadaad]

chapter 20 AABB TECHNICAL MANUAL 15th ED. PAGE463....PLS READ SELECTING BLOOD IN CHRONIC WAIHA.

[Malcolm Needs ☆]

chapter 20 AABB TECHNICAL MANUAL 15th ED. PAGE463....PLS READ SELECTING BLOOD IN CHRONIC WAIHA.

Please read AABB Technical Manual 16th edition, page 510, Chapter 17 (Positive DAT and Immune-Mediated Hemolysis), Section, Selection of Blood for Transfusion.

Pay particular attention to the sentences,

"If the autoantibody has apparent and relatively clear-cut specificity for a single antigen (e.g. anti-e) and there is active ongoing hemolysis, (my bold font throughout this quote) blood lacking that antigen may be selected. There is evidence that, in some patients, such red cells survive better than the patient's own red cells. In the absence of hemolysis, autoantibody specificity is not important, although donor units negative for the antigen may be chosen because this is a simple way to circumvent the autoantibody and detect alloantibodies." In other words, this use of antigen negative blood is not because it will last longer in the patient, but because it will show up serological incompatibility due to alloantibodies.

The section goes on to say,

"It may be undesirableto expose the patient to Rh antigens absent from the autologous cells, especially D and especially in females who may bear children later, merely to improve serologic compatibility testing with the autoantibody (e.g. when a D- patient has autoanti-e, available e- units will be D+; D-e- units are extremely rare)" They are at a frequency of approximately 1 in 7, 000.

[ckcheng]

I agree with Malcolm's advice. In addition, in order to avoid further alloimmunize the patient, you may consider XM the patient with extended phenotype-matched blood for Rhese, Duffy, Kidd, Kk, and Ss. Other additional antigens depend on race.

By the way, how often do you repeat the adsorption procedures? A week, Ten days, or per admission?

CK Cheng, MSc, SBB(ASCP), CQA(ASQ)

Hong Kong

July 22, 2009

Edited July 21, 2009 by ckcheng

[Malcolm Needs ☆]

I agree with Malcolm's advice. In addition, in order to avoid further alloimmunize the patient, you may consider XM the patient with extended phenotype-matched blood for Rhese, Duffy, Kidd, Kk, and Ss. Other additional antigens depend on race.

By the way, how often do you repeat the adsorption procedures? A week, Ten days, or per admission?

CK Cheng, MSc, SBB(ASCP), CQA(ASQ)

Hong Kong

July 22, 2009

We do the adsorption about twice a week when the patient is being actively transfused, but left often when they are not being transfused.

If it were a "normal" patient, we would have to have a sample taken within 24 hours of the next transfusion they are to have, if they have been recently transfused, but patients with WAIHA, like sickle or thalassaemic patients tend either to make no alloantibodies, or every antibody under the sun!

The fact that some (much more than the 20% one might expect) do not make alloantibodies is probably because their immune system is affected by their underlying condition.

If, however, they do make an alloantibody, we may well do alloadsorption studies more frequently than twice a week.

[eric1980]

Am I right to think that it's actually a question of balancing?

Case: Patient has non-specific autoantibodies mimicking anti-e

A school of thought is to preach safety and issue e- blood. The reason is because there are indeed antibodies in the patient that will attack any e antigen on the donor cells. This may trigger haemolytic transfusion reactions.

Another school of thought is to issue e+ blood. The reason is save precious e- blood to other anti-e patients instead of this patient who have autoantibodies mimicking anti-e, since all the RBCs will be sensitised with IgG anyway. It's also to look into the future by prevent the production of anti-E in the patient, which if it happens, mean that the patient will be at very high risk of transfusion for the rest of the patient's life.

Am I right to say the above?

I think I am more inclined to subscibe to the second school of thought... But I'm sure I won't gain any support from my lab in this...

[Malcolm Needs ☆]

Am I right to think that it's actually a question of balancing?

Case: Patient has non-specific autoantibodies mimicking anti-e

A school of thought is to preach safety and issue e- blood. The reason is because there are indeed antibodies in the patient that will attack any e antigen on the donor cells. This may trigger haemolytic transfusion reactions.

Another school of thought is to issue e+ blood. The reason is save precious e- blood to other anti-e patients instead of this patient who have autoantibodies mimicking anti-e, since all the RBCs will be sensitised with IgG anyway. It's also to look into the future by prevent the production of anti-E in the patient, which if it happens, mean that the patient will be at very high risk of transfusion for the rest of the patient's life.

Am I right to say the above?

I think I am more inclined to subscibe to the second school of thought... But I'm sure I won't gain any support from my lab in this...

The first school of thought is correct in as much as, if the patient is haemolysing their own red cells to such an extent that it is becoming a life or death situation, you would transfuse e- red cells. BUT, the auto-antibody is much more likely to be a high frequency Rh antibody that happens to react more strongly with e+ red cells than e- red cells. e-, E+ red cells will still adsorb this antibody and will be sensitised in vivo by the antibody and will be destroyed by the RES, but at a slower rate than e- red cells. In addition, however, transfusion of any red cells can exacerbate the auto-antibody and strengthen the auto-antibody (at the same time making the specificity broaden) and, of course, the patient may also make a genuine allo-anti-E.

This, to a certain extent, also answers the second school of thought (but also r"r" blood, if the patient happens to be, for example, a rr female of child-bearing potential, is as rare as hen's teeth).

I hope all my blather helps!

[sshel55]

Once an initial workup has been done and the patient is being chronically transfused, we perform an adsorption one every two weeks. We also utilize phenotypically and genotypically similar antigens matching of transfused units, particularly in patients who have formed antibodies.

[Malcolm Needs ☆]

Once an initial workup has been done and the patient is being chronically transfused, we perform an adsorption one every two weeks. We also utilize phenotypically and genotypically similar antigens matching of transfused units, particularly in patients who have formed antibodies.

Phenotypically-matched units I can understand, but it must cost the Earth to use genotyped donors each time.

[David Saikin]

I try to match the Rh phenotype . . . anytime I have not, the next time I see the patient they have been sensitized to the Rh ag I exposed them to. Now I allow no exceptions.

[sshel55]

Phenotypically-matched units I can understand, but it must cost the Earth to use genotyped donors each time.

Not each time, Malcolm. But for patients who have complicated serology that includes Warm Autos along with allo-antibodies or for African American patients with D-CE hybrid genes, it is well worth the cost. Many patients we see have been multi-transfused before we get them, so sometimes phenotyping is not an option and genotyping fits the need perfectly. We are in the process of genotyping all of our repeat donors.

[eric1980]

The first school of thought is correct in as much as, if the patient is haemolysing their own red cells to such an extent that it is becoming a life or death situation, you would transfuse e- red cells. BUT, the auto-antibody is much more likely to be a high frequency Rh antibody that happens to react more strongly with e+ red cells than e- red cells. e-, E+ red cells will still adsorb this antibody and will be sensitised in vivo by the antibody and will be destroyed by the RES, but at a slower rate than e- red cells. In addition, however, transfusion of any red cells can exacerbate the auto-antibody and strengthen the auto-antibody (at the same time making the specificity broaden) and, of course, the patient may also make a genuine allo-anti-E.

This, to a certain extent, also answers the second school of thought (but also r"r" blood, if the patient happens to be, for example, a rr female of child-bearing potential, is as rare as hen's teeth).

I hope all my blather helps!

OMG... For all that has happened that I missed this reply all these years!

But what you said deepens the topic. I was pondering the possibility of the second option being selected in such a context.

I hope regardless school of thought BBers choose when they are faced with this situation, their decision will be a more informed one.

[Malcolm Needs ☆]

Not each time, Malcolm. But for patients who have complicated serology that includes Warm Autos along with allo-antibodies or for African American patients with D-CE hybrid genes, it is well worth the cost. Many patients we see have been multi-transfused before we get them, so sometimes phenotyping is not an option and genotyping fits the need perfectly. We are in the process of genotyping all of our repeat donors.

Thanks. Now I see from where you are coming.

[kirkaw]

I am glad to find this thread. I have a patient who was determined to have WAIHA in 2009. We did not see the patient again until June of this year, when it was determined that the patient had likely developed allo-antibodies to little c and big E. At that point we phenotyped the patient and decided to give phenotypically matched red cells. On the current admission, phenotypically matched red cells are still slightly incompatible. Without doing an adsorption, we transfused 2 units (phenotypically matched) and the patient seems to be tolerating them well with appropriate increase in hemoglobin. The question is, should adsorptions have been done this time before transfusion? We do not have the resources to do them and and have to send the specimen to a reference lab, who typically turns the work around in 24 to 48 hours. Although this patient was hemodynamically stable at the time of the crossmatch request, I felt that the cost of the delay was too great since we were giving phenotypically matched red cells.

[Malcolm Needs ☆]

I am glad to find this thread. I have a patient who was determined to have WAIHA in 2009. We did not see the patient again until June of this year, when it was determined that the patient had likely developed allo-antibodies to little c and big E. At that point we phenotyped the patient and decided to give phenotypically matched red cells. On the current admission, phenotypically matched red cells are still slightly incompatible. Without doing an adsorption, we transfused 2 units (phenotypically matched) and the patient seems to be tolerating them well with appropriate increase in hemoglobin. The question is, should adsorptions have been done this time before transfusion? We do not have the resources to do them and and have to send the specimen to a reference lab, who typically turns the work around in 24 to 48 hours. Although this patient was hemodynamically stable at the time of the crossmatch request, I felt that the cost of the delay was too great since we were giving phenotypically matched red cells.

I'm sorry kirkaw, but to my mind I think that you took a tremendous risk here if, as you say, the patient was hemodynamically stable.

The problem is that you have not seen this patient for three years, and you don't know (presumably) if he or she has received a transfusion elsewhere during this time, or if the anti-c+E were the result of a transfusion at your hospital, or at another hospital. Even if the antibodies had been produced as a result of a transfusion at your own hospital, you do not know if there were any other atypical alloantibodies underlying the auto-antibody that could have caused a delayed haemolytic transfusion reaction due to an anamnestic responce. Of course, it could be that he or she had made, for example, an anti-Jka, which may not have been detected even after alloadsorption studies, as such antibodies are notoriously labile in vivo, and a delayed haemolytic transfusion reaction could have resulted anyway, but, under such circumstances, I think that you would have been safe in court, but, if another weak antibody had been identifiable after alloadsorption studies, and the patient had undergone a delayed haemolytic transfusion reaction (I don't think that an acute haemolytic transfusion reaction would have been "on the cards"), you would find it hard to justify not doing a full investigation prior to the transfusion.

I would be the first to say I could be wrong, but I would be very worried indeed if it were me, if you ever did this again.

Just my two penny worth.