GEL and Jka; something to think about

[Brenda K Hutson]

SORRY FOR THE LONG E-MAILS; I like to explain myself....

Being new to this website, it is possible this topic has been previously discussed; but if not, I think it is worth mentioning. It is something I discovered in the first Hospital I worked in that used GEL, then saw in work-ups coming from our client Hospitals when I was a Red Cross Reference Lab Supervisor, and have since seen at 2 other Facilities I have worked at which use GEL (including the current). Problem is, none of these Facilities knew about this, so they all called it whatever term "their" Institution used for an unidentified antibody. And this is NOT an Antibody you want to mis-identify!

A number of years ago, I was in charge of a Blood Bank in a 400 bed Hospital. One day, one of the Generalist Techs. brought me his panels for review (as much as possible, I wanted to review work-ups before blood went out). He said he had ruled everything out (we used 1 homozygous or 3 heterozygous to rule out), and that the positive reactions did not "fit" anything. He was going to call it a non-specific antibody.

My approach to work-ups is always the same, making every effort to avoid "biased" blood banking. After performing rule-outs with the Negative reacting cells, I was in fact able to cross everything off. However, I then look at the cells with Positive reactions, to see what they all "have in common." In this case, what they all had in common was being homozygous for Jka. However, there were other homozygous Jka cells on the Panel with negative reactions. Nevertheless, knowing the Kidds can be tricky as well as nasty, I was not going to leave it just yet. I told the Tech. to type the patient for Jka; the patient was Jka Negative. Hmmm...that was suspicious, given it only fit 25% of the population. I instructed the Tech. to transfuse with Jka Negative for now. Being a small community Hospital with a number of Generalists rotating through, we did limited Antibody ID; complicated work-ups were sent to a Reference Lab (too difficult to keep competency among staff otherwise). We used GEL and we kept LISS on hand, however, we did not stock PEG. I sent the specimen to our Reference Lab who used PEG; they pulled up a clear-cut Anti-Jka.

From then on, I instructed staff to approach work-ups in this way, and to pursue any suspicious Jka, just as I had. There were several more in my time there.

At the Red Cross, a number of our clients sent work-ups, getting the same type of reactivity with GEL. We would work it up with PEG (and type the patient if possible). As I came across more of these Jka reactions, I would then educate the Hospitals. Again, in my current position, we have run across the same phenomenon (though the staff had been calling them a "weak, unidentified antibody;" our Institution's term for "I DON'T KNOW!"

NOTE: This is not meant to put GEL down; I love GEL! Any of the methods have their problems. But given what I have seen now in my 26 years, and how many in the field approach work-ups, I just thought you might want to know about this. Something to think about when your next GEL work-up doesn't "fit" anything. Also, I have not noticed this with Anti-Jkb, but then I certainly see that much less anyway; so not sure about that one.

Brenda Hutson, CLS(ASCP)SBB

[Malcolm Needs ☆]

A good point, well made Brenda. We see quite a few of these too.

It is probably because, although it is always quoted that there are about 14, 000 Kidd antigen sites per red cell, this is, of course, an average. Some cell samples will have more (the ones that you see reacting) and some less (the ones that you don't see reacting).

I totally agree with you that anti-Jka is not an antibody that you want to miss. You only need to look at the antibodies that have caused most clinically significant delayed haemolytic transfusion reactions in any haemovigilance schemes around the world, and anti-Jka always seems to come out on top.

We now always run a panel of enzyme-treated red cells by IAT, as well as our normal untreated panel by IAT, and we find that this is much more sensitive than running just the normal IAT panel with an enzyme panel (i.e. we didn't used to take the enzyme panel on to IAT). We get some cracking reactions like that.

The problem comes, however, in two ways. Firstly, if the anti-Jka is not detected in the screen (which, in the UK, you only need to do by IAT). Secondly, if there is an "enzyme-auto" pan-reacting antibody present too, as, of course, this masks the anti-Jka (mind you, you soon pick it up by IAT after the transfusion - especially in the eluate!).

I think this is the one drawback with using plasma, rather than serum. You do not have the complement system around to help you detect complement binding antibodies. We had a fatal case of anti-Vel in the UK only about 5 years ago, where the anti-Vel could only be detected in serum; but too late I'm afraid.

That having been said, I wouldn't want to go back to clotted samples (except for certain specific tests) and I wouldn't want to go away from gel; I think it is a fantastic system too.

Like you, I can't recall having seen a case of anti-Jkb reacting in a similar way, but, also as you say, they are considerably more rare in any case.

:)

[Brenda K Hutson]

A good point, well made Brenda. We see quite a few of these too.

It is probably because, although it is always quoted that there are about 14, 000 Kidd antigen sites per red cell, this is, of course, an average. Some cell samples will have more (the ones that you see reacting) and some less (the ones that you don't see reacting).

I totally agree with you that anti-Jka is not an antibody that you want to miss. You only need to look at the antibodies that have caused most clinically significant delayed haemolytic transfusion reactions in any haemovigilance schemes around the world, and anti-Jka always seems to come out on top.

We now always run a panel of enzyme-treated red cells by IAT, as well as our normal untreated panel by IAT, and we find that this is much more sensitive than running just the normal IAT panel with an enzyme panel (i.e. we didn't used to take the enzyme panel on to IAT). We get some cracking reactions like that.

The problem comes, however, in two ways. Firstly, if the anti-Jka is not detected in the screen (which, in the UK, you only need to do by IAT). Secondly, if there is an "enzyme-auto" pan-reacting antibody present too, as, of course, this masks the anti-Jka (mind you, you soon pick it up by IAT after the transfusion - especially in the eluate!).

I think this is the one drawback with using plasma, rather than serum. You do not have the complement system around to help you detect complement binding antibodies. We had a fatal case of anti-Vel in the UK only about 5 years ago, where the anti-Vel could only be detected in serum; but too late I'm afraid.

That having been said, I wouldn't want to go back to clotted samples (except for certain specific tests) and I wouldn't want to go away from gel; I think it is a fantastic system too.

Like you, I can't recall having seen a case of anti-Jkb reacting in a similar way, but, also as you say, they are considerably more rare in any case.

:)

Wow, sorry to hear about the Vel! And speaking of Anti-Vel; when I was the Red Cross Supervisor, I put on 2 Seminars a year which Techs. from all area Hospitals attended for CEUs (usually around 200 people). I would bring in great speakers like Dr. Garratty and Phyllis Walker. I recall 1 speaker (could have been Phyllis) discussing the issue of Prewarm (a HUGE pet peeve of mine) and literature of Anti-Vel's being prewarmed away (among other antibodies). Within a week of that Seminar, 1 of our client Hospitals almost succeeded in doing that! Unvelievable! I think they were at the Seminar too. Luckily for this patient, they were not successful in prewarming it away. I have seen way too many instances of Techs. who just want the antibody to "go away!" If I were to look at 2 issues right now that I would give a talk about, they would be: Prewarming, and the Reality of Biased Blood Banking.

We had one client who had the following Policy regarding Antibody ID: They would run a Panel, then "automatically" try to Prewarm it; if it Prewarmed away, it was "nothing;" if it didn't, they needed to pursue it! After 2 Hemolytic Transfusion Reactions (1 patient with an Anti-E and 1 with an Anti-E and Anti-c), I pursuaded their Medical Director to change their Procedure immediately! But what was even more appalling, was that in looking at their Panels (we asked our clients to send xerox copies of their work-ups), these Antibodies were clear-cut, > 2+; they obviously did not even take the time to perform rule-outs, or even look at the pattern! I have to say, after having that position, it made me a little more concerned about being a patient in a Hospital; I had NO idea these kinds of things occurred out there. Scary!

Anyway, enough said (too much said)!! Sorry...

Brenda Hutson, CLS(ASCP)SBB

[Malcolm Needs ☆]

Wow, sorry to hear about the Vel! And speaking of Anti-Vel; when I was the Red Cross Supervisor, I put on 2 Seminars a year which Techs. from all area Hospitals attended for CEUs (usually around 200 people). I would bring in great speakers like Dr. Garratty and Phyllis Walker. I recall 1 speaker (could have been Phyllis) discussing the issue of Prewarm (a HUGE pet peeve of mine) and literature of Anti-Vel's being prewarmed away (among other antibodies). Within a week of that Seminar, 1 of our client Hospitals almost succeeded in doing that! Unvelievable! I think they were at the Seminar too. Luckily for this patient, they were not successful in prewarming it away. I have seen way too many instances of Techs. who just want the antibody to "go away!" If I were to look at 2 issues right now that I would give a talk about, they would be: Prewarming, and the Reality of Biased Blood Banking.

We had one client who had the following Policy regarding Antibody ID: They would run a Panel, then "automatically" try to Prewarm it; if it Prewarmed away, it was "nothing;" if it didn't, they needed to pursue it! After 2 Hemolytic Transfusion Reactions (1 patient with an Anti-E and 1 with an Anti-E and Anti-c), I pursuaded their Medical Director to change their Procedure immediately! But what was even more appalling, was that in looking at their Panels (we asked our clients to send xerox copies of their work-ups), these Antibodies were clear-cut, > 2+; they obviously did not even take the time to perform rule-outs, or even look at the pattern! I have to say, after having that position, it made me a little more concerned about being a patient in a Hospital; I had NO idea these kinds of things occurred out there. Scary!

Anyway, enough said (too much said)!! Sorry...

Brenda Hutson, CLS(ASCP)SBB

"Unvelievable". I love it!

As it intimated, I am a great fan of gel, but, in the right hands, I am a great fan of pre-warmed, warm-washed, tube IAT, using either a broad-spectrum HAG or a monospecific anti-IgG. I had the huge honour (and luck) of being trained in red cell serology by both Dr. Carolyn Giles, when she was in charge of the Red Cell Reference Department of the WHO International Blood Group Reference Laboratory (when Joyce Poole was a Senior Medical Laboratory Technician in her Department) and, of course course, by Joyce Poole herself (who is now Head of this Department). Both taught me how to perform these tests, and, in Carolyn's case, not that she ever did swear, she swore by the technique. In Joyce's case, I know that she still uses the test in her Department on a regular basis.

I honestly don't think that this technique is intrinsically insensitive; I think that it is the way people are taught how to perform the technique, and in particular, how they are taught to read the results that is wrong (that and trying to perform a test that is, essentially, unfamiliar to them through lack of practice.

I have seen people bash the living daylights out of the tubes are gently spinning them to make the red cell button after addition of the AHG; so much so that they could make a strong anti-D negative! THey are not taught the gentle "tip and roll" required for this technique, with the emphasis onnthe word "gentle". I further think that they are so used to the robustness of the CAT that they do not realise just how gentle the resuspension has to be, and just how weak a positive reaction can be, and still be clinically significant.

Without doubt though, it is a technique that needs practice, practice, practice, and performing it once in a blue moon is NOT sufficient to maintain competence.

We use it in the Reference Laboratory that I manage quite a lot, but I make sure that my staff/colleagues are well and truely competent in the technique before I let them loose anywhere near a patient's sample.

I do agree wholeheartedly with you though about the fact that, what goes on out there is scary! AS the saying goes: "Be afraid, be very afraid!".

:eek::eek:

[TimOz]

Funny this should come up. I was messing around with serum samples in Polyspecific Gel cards in North Asia a week before last and found a range of reactive samples that "went away" when DiaMed ID-Dil was used but were very apparent when other Low Ionic solutions were used. The were seen as pure complement in a monospecific DAT card. I could reproduce the neutralising effect by adding edta to the serum samples. I did not get the time to fully ID them but it was interesting. Seems that ID-Dil 2 has edta or certainly behaves in the same way. So these reactions, (whatever they were) would not be seen in edta plasma no matter what card type was used.

Also, have a look at this reference Vox Sang. 1998;74(1):53-5. Complement-binding anti-Jka not detectable by DiaMed gels.O'Brien P, Hopkins L, McCarthy D, Murphy S.

They draw the same conclusion.

I guess we all know that we can miss a certain proportion of Jk antibodies if we use plasma and maybe gel is also not so good at detecting weak Jk antibodies that can be seen in a tube. When I am asked questions about this and especially if serum is better than plasma I usually say that it is a balance. Serum is able to improve you chances at Jk detection BUT is a lot slower (clot retraction before spinning etc), may give more confusing false positives and may be prone to give fibrin and top line effects. Maybe on balance it is better for the patient to have a faster, cleaner sample that have the ability to detect a small number of (admitedly nasty) Jk antibodies. And the clincher is automation makes edta plasma almost a must.

Am I right?

By the way, the DiaMed monospecific DAT cards are great and very informative.

[Malcolm Needs ☆]

Hi Tim.

I have no doubt whatsoever that you are right.

The kind of antibody I was talking about in my post above is disappearingly rare, and I would not want to go back to routine tube IAT for all the tea in China. I am a huge fan of gel.

On the other hand, for certain specialised tests performed in my laboratory, LISS tube IAT can be an essential tool.

Edited July 28, 2009 by Malcolm Needs
Spelling was okay (for once); the syntax was terrible!

[L106]

I agree with Tim also. This post made me stop and think about all those times before we switched to plasma a few years ago when we would have problems with patient samples that wouldn't clot (or kept making fibrin clots.) It caused significant delays in testing and providing blood products fairly frequently, which can be detrimental to patient care (as Tim pointed out.)

Our experience during the last several year since switching to plasma has brought me to the conclusion that those strictly "complement-dependent Anti-Kidd antibodies" are far, far less common than I was led to believe during my original training.

Edited July 28, 2009 by L106
Oops! Spelling.

[Brenda K Hutson]

Well, your points are probably all valid, however, I have to say that in the instances where the Jka was a hit and miss on homozygous cells in GEL, the Jka still came up perfectly with PEG, even using an EDTA. It is also interesting to note that in these instances, the GEL "is" hitting some of the homozygous cells; so why not all of them? Also, in talking with an Ortho Technical Support person some time ago, she did say that even with EDTA, it is possible for C' activation to still occur in certain scenarios (and don't I wish right now I could remember what her technical explanation was ).

No doubt you could explain all of this; you seem to understand it at a deeper, more technical level than myself.

Thanks for the feedback......

Brenda Hutson, CLS(ASCP)SBB

[Malcolm Needs ☆]

Our experience during the last several year since switching to plasma has brought me to the conclusion that those strictly "complement-dependent Anti-Kidd antibodies" are far, far less common than I was led to believe during my original training.

Edited July 28, 2009 by Malcolm Needs

[L106]

I understand your point and agree with you, Malcolm. And I agree that we may not be detecting these complement-dependent antibodies if we are using plasma and/or a technique that is not sensitive to them. But what makes me feel comfortable is that the majority of our clientele population are frequently-transfused patients who consistently use our facility. (Oncology, GI Bleeding, repeat-joint replacement patients, etc., who we see repeatedly.)

If we are missing some of these antibodies, surely we are transfusing antigen-positive donor units (albeit "compatible" by our technique) to these patients. Typically, the patient would be stimulated to increase their antibody production and I would expect that with subsequent testing the antibody would rise to a level that would be detected by our technique, or I would observe some other signs such as unexplained drop in hemoglobin, icteric plasma, positive Direct Antiglobulin Test, etc. (And we are not seeing this.) Am I being a little too dangerous in my reasoning?

[Malcolm Needs ☆]

No, you are being perfectly logical.

I just thought that I should make the point, otherwise the "science" is not proven.