Historical Type A now types as O

[amhamilt]

We had a cancer outpatient typed for platelets the other day. She has historically typed as an A pos, but now types as a very obvious, very clean O pos. We had the patient redrawn with the same results. Furthermore, after calling other hospitals, we found that we we are not the first hospital to see the same phenomenon in this patient. Previously, she has typed as an A at another hospital, then typed as an O some months later. Four days ago she was typed in our lab as an A. What could cause this? Is it possible that some of the chemotherapeutic drugs could cause an A to appear as an O?

Other info: The patient is being treated for a myelodysplastic cancer. She has NOT had a bone marrow or stem cell transplant. She claims to have not received any blood products in the past four days. (4 days ago she typed as an A) Her blood smear showed junky looking red cells; lots of hypochromia, a lot of size varience, pappenheimer bodies, etc.

This mystery is driving me crazy!

[adiescast]

Lots of reticulocytes? Any possibility of a chimera? We found an interesting phenomenon that involved where in the sample you pull your cells from (top versus bottom). You got a different answer depending on where the cells came from. I suspect that it has to do with the weight of younger cells versus older cells and layering out on centrifugation. Most of the time we don't pay any attention to where we sample from.

[Yanxia]

What about A subgroup with alloantibody?

[Malcolm Needs ☆]

We had a cancer outpatient typed for platelets the other day. She has historically typed as an A pos, but now types as a very obvious, very clean O pos. We had the patient redrawn with the same results. Furthermore, after calling other hospitals, we found that we we are not the first hospital to see the same phenomenon in this patient. Previously, she has typed as an A at another hospital, then typed as an O some months later. Four days ago she was typed in our lab as an A. What could cause this? Is it possible that some of the chemotherapeutic drugs could cause an A to appear as an O?

Other info: The patient is being treated for a myelodysplastic cancer. She has NOT had a bone marrow or stem cell transplant. She claims to have not received any blood products in the past four days. (4 days ago she typed as an A) Her blood smear showed junky looking red cells; lots of hypochromia, a lot of size varience, pappenheimer bodies, etc.

This mystery is driving me crazy!

It is not unknown for patients with MDS to have weakened ABO antigens; so much so that they appear to lose their ABO antigens entirely (see Geoff Daniels' book Human Blood Groups and/or MArion Reid and Christine Lomas-Francis' book THe Blood Group Antigen Facts Book).

Can I assume that this was the forward group, and not both the forward group and the reverse group, with an anti-A appearing? The latter situation is extremely unusual, although, once again, not unknown.

Julie Mallory is, of course, also correct in saying that it can depend also upon from where the red cells in your sample were taken (top or bottom). A colleague of mine, Bob Stamps, has also shown this to be true.

In this case, if the MDS is in relapse, and you take the red cells from the top of the sample, they may well appear to be group O.

If the patient is in remission, you may fnd that the ABO antigens strengthen, and that the cells at the top (the younger cells) are group A, and those at the bottom are group O (or, at least, mixed-field with anti-A).

Whatever the condition of the patient, if you take off all the plasma, and then take a sample of the red cells from the top, and one from the bottom, mixed the two together in saline, and then group them, you may find a mixed-field reaction with the anti-A.

Edited November 19, 2011 by Malcolm Needs

[Auntie-D]

I've found this with analyser v/s manual technique too - the Diamed Gelstation samples from the bottom whereas people tend to sample nearer the top or middle.

[Liz]

I've found this with analyser v/s manual technique too - the Diamed Gelstation samples from the bottom whereas people tend to sample nearer the top or middle.

Auntie-D, What were the differences that you experienced?

[Auntie-D]

Auntie-D, What were the differences that you experienced?

There were two 'phenomena' seen. on was dual populations the the groups of chemo patients, the other was a persistent missed weak antibody - never clinically significant ones, just anti-Cw.

[Mabel Adams]

I disagree that anti-Cw is not clinically significant.

There was a poster at the AABB meeting on different blood types depending on where the specimen was sampled as those above have mentioned.

I've always thought that principle of density separation might be important in fetalscreen samples. If someone spins them out of habit and then samples close to the top will they get more fetal cells (large and early cells more likely in babies) than if the sample had been well-mixed? Worse, what happens if they sample from the middle and miss the fetal cells? Sorry, shouldn't side-track this thread.

[Malcolm Needs ☆]

I disagree that anti-Cw is not clinically significant.

As far as I am aware (and I am happy to be proved wrong) there has only ever been one case in the literature that blamed hydrops fetalis on anti-Cw (and samples were never submitted to the Reference Laboratory to prove this - I know this, because the paper came from one of "my" hospitals):

Gopalam Kollamparambil T, Rameshchandra Jani B, Aldouri M, Soe A, Ducker DA. Anti-Cw alloimmunization presenting as hydrops fetalis. Acta Paediatrica 2005; 94: 499-501.

Other than that, anti-Cw has only ever been implicated in very mild cases of HDFN, and, although anti-Cw has been reported to have caused severe transfusion reactions, it has not been reported now for many, many years (despite, undoubtedly, Cw+ blood being given to patients with anti-Cw not detected in a screen with Cw- red cells, by electronic issue). This lack of recent cases brings doubt as to the authenticity of the early papers.

Certainly, in their paper, Daniels G, Poole J, De Silva M, Callaghan T, MacLennan S, Smith N. The clinical significance of blood group antibodies. Transfusion medicine 2002; 12: 287-296, anti-Cw is not considered clinically significant.

"Anti-Cw is a relatively common antibody. There is no report of anti-Cw causing a transfusion reaction, and IAT-compatible blood may be selected. HDN as a result of anti-Cw has been reported. About 97% of {British} donors are Cw-."

[Mabel Adams]

I checked the listing under antigens on this site and it says anti-Cw can cause reactions and some degree of HDFN so I would want to consider it significant (i.e. do AHG xms etc.) for now anyway. I saw a patient once that I know had anti-Cw (among several antibodies) and had a transfusion reaction because someone missed an incompatible crossmatch. What I can't remember is whether the Cw was the new one found with the reaction or whether it was a Co b. I might have to contact my old lab and have them look her up!

[DOGLOVER]

Years ago, I had a patient who had had several fetal demises. Antibody screen was negative. We were sent a cord blood this time and it had a strong pos DAT. Mom and Dad were both B pos so I crossmatched them and 4+ incompatible. I was able to identify the antibody as Cw, so I am definitely a beliver in its clinical significance.

[Malcolm Needs ☆]

I cannot (and never would) dispute your findings, but why oh why did you not publish? This would mean that there would have been at least two cases in the literature.

I still maintain though, that anti-Cw is not clinically significant as far as transfusion reactions are concerned, as long as the cross-match is compatible. There are plenty of antibodies that cause trans fusion reactions, but not HDFN, and vice versa.

[DOGLOVER]

Hi Malcolm, I guess it was so early in my career that it never occurred to me that it was something that would have merited publishing.

[Malcolm Needs ☆]

Thanks for that DOGLOVER. It's a shame that your then bosses didn't think of it though!!!!!!!!!!1