John Eggington

What do you do/advise if the patient has a 'cold' antibody, and do you treat specific and non-specific antibodies differently?

You could argue that a screen is just that, a screen. It's positive, so identify the antibody by panelling the sample. Cold comfort, maybe, but a positive screen is all you're after, isn't it (preferably not a false pos!)?

Jk antibodies are often not detected by 'standard' enzyme techniques. If you use an enzyme IAT technique, they're almost always enhanced. Gel cards make performing an enzyme IAT very easy; just use enzyme treated cells in combination with an IAT (Coombs) card.

Also, I think that using a 'post amnio' sample to look for antibodies that may have been stimulated by the 'amnio' may be a bit unreliable. It will have been taken very shortly after the event (to get a reliable FMH estimation), so a later sample is more likely to detect additional/new antibodies, as well as giving a better idea of whether existing antibodies have been 'boosted'.

I was referring specifically to the administration of prophylactic anti-D, rather than considering other antibodies

In the UK, one of the most relevant documents is the BCSH guideline ‘Guidelines on the estimation of fetomaternal haemorrhage (September 2009 update)’. This identifies amniocentesis (amongst other things) as a potential sensitising event, i.e, may give rise to a FMH. After 20 weeks gestation it is possible that the size of a FMH may exceed the capacity of a ‘standard’ dose of prophylactic anti-D (RhoGam) to clear all fetal cells before they cause ‘sensitisation’ of the mother. To ensure a sufficient dose of prophylactic anti-D is given, a FMH estimation must be performed after any sensitising event (after 20 weeks, anyway) to determine if a significant, e.g. over 4ml, FMH has occurred and what dose of prophylactic anti-D is required to ‘cover’ this bleed. So, I guess that the ABO/Rh will go some way to confirming that you have the right patient, by matching the current results to any existing results. It will also serve to identify the D type of a ‘new’ patient. The antibody screen will give you an idea as to whether anti-D is already present (no doubt there is a policy in place for dealing with determining if this is ‘immune’, or due to a previous administration of prophylactic anti-D. Finally, your fetal screen will give you an idea as to whether there has been a (significant) FMH and what dose of prophylactic anti-D you’ll need to give (although you do this from 17 weeks, rather than 20 weeks, I guess this is adding a safety factor).

What tests do you do on the 'post amnio' sample?

DADE used to make a reagent called 'NeutrAB', which neutralised ABO antibodies. If you can find out just who, if anyone, makes it anymore, this could be what you're looking for (if you find it, let me know, I used to like it!).

We do, sometimes, encounter anti-M in patients who type as M+. I'm not convinced they are all truely auto antibodies, but that's what we call them. The most recent patient set they appeared amongst were very young 'cardiac' patients, so I wonder if it's something to do with treatment/pre-conditioning (a bit like the 'dialysis' anti-N that we used to see).

Are there any results for monospecific DAT tests, for example, with anti-IgA? If the patient had only an IgA class allo, or auto, antibody a 'regular' IAT, or 'poly' DAT, would not detect it (rare, but not impossible. MDS patients can throw up some 'unusual' stuff!). Was an elution perormed on any of the patient samples? The DAT may be neagtive, but an elution may still contain detectable antibody. Although, in this instance, the DAT may be negative because all (or most) of the cells that had been coated with any antibody had been destroyed prior to testing (the samples are lysed, after all). I've also found that 'gel' DATs are good when all red cells are coated with antibody, but may not detect a 'minor' population of (even heavily coated) red cells, that may be found in a transfusion-reaction-type investigation (particularly if a lot of the coated cells have already been destroyed). A tube DAT, checked by lens/microscope, might pick up a positive 'minor' population better than a 'gel' card.

Try 'How I treat autoimmune hemolytic anemia in adults', by Lechner and Jager (Blood, 2010. 116. 1831-1838). It's a good start and has plenty of references.

Those taking the 'BBTS certificate' undertake a single exam, which examins candidates on their specialist knowledge of transfusion science. The topic areas examined match, very closely, those listed at http://www.ascp.org/FunctionalNavigation/certification/GetCertified/SpecialistCertification.aspx, but no equivalence is claimed. It is expected that those taking the exam are HPC (the UK Healthcare Professions Council) registered, with a minimum of 2 years experience in a transfusion laboratory (the experience component is a recommendation, rather than mandatory).

Are your typing reagents 'direct spin' or IAT?

Maybe my choice of the term ‘mimicking’ wasn’t so good. I wasn’t thinking of the classic mimic, which could be adsorbed ‘out’ by cells lacking the relevant antigen. More something, or things, that behaved as ‘anti-Jka’, when the conditions were right (e.g., the unnatural ones found in the various techniques we use to detect antibodies). That’s not to say they wouldn’t be adsorbed ‘out’ by Jk(a-) cells, just that a failure to adsorb wouldn’t rule out the possibility of it not being a ‘real’ anti-Jka. Hmmm...I think I know what I mean, anyway!?

There is something (or a range of ‘somethings’) out there that is/are mimicking anti-Jka. I saw it a lot, 15-20 years ago (in gel, mostly, by enzyme technique), but it seemed to have stopped happening. The past year or so seems to see it happening again, but now we’re finding it with a, gel, enzyme IAT technique (although our workload has risen by a great deal, so it could be a ‘numbers’ issue). So, if the patient happens to be Jk(a+), you’ll call it ‘auto’, if your patient happens to be Jk(a-) you’ll call it ‘allo’, but actually it isn’t either. This thread shows the phenomenon happens with other techniques. Maybe one of these techniques is better at detecting these ‘non-anti-Jka’, and so looks as if it’s more sensitive for this specificity? Whatever it is, it’s a nuisance!

Took my time, but here now. Thanks for the welcome.

A and B cells seem to be an underused resource, when it comes to antibody identification.

Did you test the cells with monspecific AHG reagents and/or perform an elution?

What do you use to control antibody screens (e.g. weak anti-D with each batch of tests, etc.) and how did these controls perfrom on the day of the original investigation?