62 male with pancreas CA,no history of transfusion,previous blood group(historic) was AB

anti-A 4+

anti-B 4+

anti-AB 4+

A1cell 4+, A2cell 2+, Bcell negative, Ocell negative,

Autocontrol negative,DAT negative, IAT negative,IAT(room)negative,slide anti-a4+,slide anti-B4+,

AB XM incompatible,B XM compatible, O XM compatible,

any suggesting wellcome

My first thought is that the patient's pathology gives a clue. Carcinoma often weakens ABO groups, and you say that this patient has a "historic" blood group of AB.

If the ABO antigens have weakened, and they can do so even to the extent of giving a frank mixed-field reaction, it could be that the patient is, indeed, producing either an anti-A or an anti-B(A).

A and B antigens on red cells is not weakened,

another possibility may be Tn

Well yes, but this would mean that both the A1 and A2 reverse grouping cells would have to be Tn activated, whilst the B and O cells were not, but if the patient's red cells were Tn activated, they would not react with monoclonal antibodies - only human polyclonal anti-A, anti-B and anti-A,B - in fact they would react with human "iert" AB plasma - so I don't think it is that.

I agree totally with Malcolm.

In this case ,I suspicious the anti-A reagent have something against low incidence antigen on the patient's red cells, try to use human anti-A and other lots or other brand of anti-A.

Anti-B(A) being similar to anti-A,B ? please elaborate. Thanks.

we used two diffent anti-A reagents,both of them was mouse monoclonal(Immucor,Tulip) with completely different types of clones.We havent human anti-A at the moment.

may be the answer is:

Incompatible A expression

Expression of `A-like antigen' in tumors of O individuals

is observed less frequently than deletion of

A or B determinant in A or B tumors (i.e. tumors

from blood group A or B individuals). This phenomenon,

termed `incompatible A expression,' is another

long-standing problem in human tumor immunology,

since tumor cells expressing A-like antigen in

O or B individuals could be a target of host immune

surveillance at the early stage of human cancer development

and are more likely to be immunologically

rejected. Incidence of tumor occurrence in blood

group A individuals compared to O individuals was

higher for gastric cancer (1.21) and ovarian cancer

(1.23), clearly higher for vulvar cancer (1.39) and

much higher for salivary gland cancer (1.55) [96].

A-like antigen was once considered to be an Across-

reacting antigen ..

S.-i. Hakomori / Biochimica et Biophysica Acta 1473 (1999) 247^266

Anti-B(A) being similar to anti-A,B ? please elaborate. Thanks.

Although I have used the term anti-B(A) myself in this thread, in actual fact the term, strictly speaking, should refer to the antigen, rather than the antibody. The A( and B(A) antigens result from overlapping specificities of A- and B- transferase enzymes.

B(A) antigen results from a Gly substitution from Ser at position 235 of the B enzyme, and A( antigen results from a, Ala substitution from a Pro at position234 of the A enzyme.

Monoclonal anti-A and monoclonal anti-B grouping sera are more prone to detecting these antigens than are human polyclonal antibodies - or, at least, that was what Douggie Voak told me, and he should know!

:bow::bow:

Edited December 4, 201014 yr by Malcolm Needs
I was talking rubbish!

Although I have used the term anti-B(A) myself in this thread, in actual fact the term, strictly speaking, should refer to the antigen, rather than the antibody. The A( and B(A) antigens result from overlapping specificities of A- and B- transferase enzymes.

B(A) antigen results from a Gly substitution from Ser at position 235 of the B enzyme, and A( antigen results from a, Ala substitution from a Pro at position234 of the A enzyme.

Monoclonal anti-A and monoclonal anti-B grouping sera are more prone to detecting these antigens than are human polyclonal antibodies - or, at least, that was what Douggie Voak told me, and he should know!

:bow::bow:

Thank you for the clarification.

I do not know who Douggie is, but you had me laughing out loud with your choice of emoticons!! Thank you for that too!

When Georges Kohler and Cesar Milstein described monoclonal antibodies in 1975, Doug Voak was working an the Chief Clinical Scientist at NHSBT-Cambridge Centre, and it was he that did the early work on the original ABO antibodies produced by Kohler and Milstein.

Infact in serum of the B(A) phenotype we can see anti-A1 (very rare) but not the anti-A2 ,is there any report contrary,so the only explanation is an acquired red cell antigen A in a patient with the cancer.

Infact in serum of the B(A) phenotype we can see anti-A1 (very rare) but not the anti-A2 ,is there any report contrary,so the only explanation is an acquired red cell antigen A in a patient with the cancer.

Sorry ziya, but I do't under stand this post.

Anti-A1 is far from rare, and can be detected in the plasma of approximately 2% of A2 individuals, but approximately 25% of A2B individuals, although clinically significant anti-A1 is VERY rare.

I have NEVER heard of anti-A2.

I have only ever heard of anti-A, anti-A1, anti-B and AntiA,B in terms of "pure" ABO antibodies.

Ziya, in my memory B(A) have anti-A reacted with A2 cells, this is the differ of B(A) from AsubgroupB.

And B(A) the A antigen is weaker than normal.

As to the human sourse anti-A, we will use patient serum to do it, although this kind of doing is not qualified.

Ziya, Malcolm has mentioned maybe the reverse cells you use is T antigen active, what about the result you get , you can do the test with another A patient (no abnormal antibody) react with the A1 and A2 cells you use before. To see if it react then maybe they are T active.

If it is not the T active, then I think you are right, it is the tumor change the group B patient to AB.

I don't know is the Incompatible A react the same strangth as acquired B, the 4+ reaction is so strong. If you have more test result about this type, please tell us, or you publish a paper about it and please tell us.

Edited December 5, 201014 yr by shily
ADD SOMETHING

sorry for confusing,of course there is no anti-A2 ,I would have to say that what the shily is, reaction with the A2 cells.There are number of questions to be answered.

1. I will find out the exact time of the historic blood group of the patient,if its an old history it could not be due to cancer

3.I think most of us use monoclonal reagents,but on literature reactions patterns usually belong to polyclonal antigens,so we get confused.

Edited December 5, 201014 yr by ziya

sorry for confusing,of course there is no anti-A2 ,I would have to say that what the shily is, reaction with the A2 cells.There are number of questions to be answered.

1. I will find out the exact time of the historic blood group of the patient,if its an old history it could not be due to cancer

2.I think most of us use monoclonal reagents,but on literature reactions patterns usually belong to polyclonal antigens,so we get confused.

Ziya, you are right, the historic blood group's time is very important, and maybe you can use saliva blood group substance and family member's blood group to get some information.

historic blood group is old(30 years with forward only), and prepared red cells of the patient with bromelin and find no change of reaction pattern of red cell antigens,so it is not aquired antigen or Tn activation.

historic blood group is old(30 years with forward only), and prepared red cells of the patient with bromelin and find no change of reaction pattern of red cell antigens,so it is not aquired antigen or Tn activation.

I don't know if enzyme can destroy the acquired antigen, maybe not. Waiting for certain answer.

From the crossmatch you give before, the patient has anti-A in his serum. So the A1 and A2 cells reaction pattern is no doubt.

But have looks normal A antigen and anti-A in his serum, but the anti-A not reacted with the autocells.

Partial A( just a borrow name from D antigen).What about the anti-A1 result in forword type?

Yanxia's idea of performing saliva testing is excellent; grouping reagents of 30 years ago were somewhat unreliable - to say the least!

If Iam understanding his correctly he is fronting as an AB could he be an A2B with anti-A1 and also have a low incidence antibody that is reacting with the corresponding antigen on the A2cells eg....anti-Vw or a anti-Lea that is only reacting at RT. Just a thought

Yes it would be nice to know if the patient is A1B or A2B. If they are A2B then an RT Saline panel would be useful ie use the same technique as you found the initial reaction with A2 cells with.