How many rule-out cells does your lab require for antibody ID?

[tupton]

Just wondering what the prevailing practice is for antibody ID rule-out cells.

At our facility, current SOP states 2 cells are needed if 1 is homozygous, or 3 heterozygous cells.

I'm looking to change it to 1 homozygous cell would be sufficient for rule-out.

Opinions?

[Malcolm Needs ☆]

Hi Tom,

 

I would be extremely cautious about using one cell showing apparent homozygous expression of an antigen to rule out the presence of an antibody; and I chose my words carefully, as antigens cannot be homozygous, heterozygous or, in the case of Xg(a), hemizygous.  Only the genes encoding the antigens (or, in some cases, such as ABO, the enzymes that confer antigenicity to a red cell) can be so described.

 

The reason for my caution is that, unless the actual genotype of the donor has been performed for each gene encoding as above, it is almost impossible (serologically) to tell the difference between a cell expressing apparent homozygosity and one expressing heterozygosity for a "normal" gene and a mutant gene.  For example, within the Black population, one only has to look at the situation within the Duffy Blood Group System.  Approximately 68% of that population type as Fy(a-b-).  Most of these individuals have the FYB gene in a homozygous state, but the antigen cannot be expressed on the red cell, as they are also homozygous for the GATA-1 mutation.

 

On the other hand, some will carry the genuine, silent FY gene in a heterozygous state, but will type (serologically) as Fy(a+b-) or Fy(a-b+) - depending upon the other allele they inherit at the locus, and so appear to express homozygosity, but will actually be expressing heterozygosity.

 

As Duffy antibodies are known to show "dosage", this means that an antibody within this Blood Group System could be missed.

 

The situation of the Duffy Blood Group System is probably the most striking, but this situation can apply to all Blood Group Systems, albeit the situation would be much less common.

 

I would always go for two cells showing apparent homozygous expression of a particular antigen for a rule out (where possible - even as a Reference Laboratory, with access to a huge number of cells, we cannot do this every single time - but we say so in our reports, if this is the case).

[jayinsat]

When I first look at my antibody screen and ID panel, the first thing I do is go to my negative reactions and cross off cells that are single dose "homozygous" for the antigen group.  Next, I look at the double-dose "heterozygous" cells and, if I have 3 or more, I will cross those out as possiblities.  Normally, I have a clear cut answer at this point, which is then evaluated to make sure I have 3 positive cells that reacted and 3 negative that did not.  If there is more than one possibility, I do "rule-out" panels to either prove the existence of the addtional antibody or rule it out.  I believe Mabel Adams had an excellent power-point that perfectly illustrated this approach.  It has been very helpful.  Of course, there are often "stray" reactions or cells you would expect to react that don't.  You just have to use technical skills and techniques to investigate these reaction to a conclusion that your pathologist will accept.

[mollyredone]

Malcolm,

 

Anytime you do an antibody screen of three cells and it is negative, you are automatically "ruling out" antibodies using just one homozygous cell in many cases (and with our screening cells one heterozygous K).  Doesn't the "more than one" rule come into play when identifying antibodies instead of ruling them out?

[Malcolm Needs ☆]

That's true mollyredone, but the samples we deal with already have an antibody (or antibodies) and so we know that the patient is a responder.  I suppose that is the difference in "point-of-view" between someone who works in a Reference Laboratory and someone who works "at the sharp end" in a hospital - but I do see what you mean.

[mollyredone]

Malcolm,

 

Although your job sounds really interesting, I think I'm glad I don't just have problem children to deal with on a daily basis!

[Malcolm Needs ☆]

You'd love it mollyedone; you would really love it!!!!!!!!!!!!!!

 

Mind you, my colleagues, my friends and my family (even my wife) all think I'm a bit strange!!!!!!!!!!!!

Edited May 30, 2013 by Malcolm Needs

[ChrisH]

I thought most people who love Blood Banking were strange. Who else would stay in this field for a long time?

 

Mind you, my colleagues, my friends and my family (even my wife) all think I'm a bit strange!!!!!!!!!!!!

[aafrin]

Couldn't agree more Chris.

[SMILLER]

We try to do 3 rule outs/rule ins for each significant antibody when following up a positive screen.  For antigens that show dosage, we try to have at least one of the rule outs on a homozygous cell. 

 

This may indeed be overkill for many cases. As has been pointed out -- most patients have a negative screen and in these instances we are routinely "ruling out" willy-nilly most antigens with only one or two cells.

 

Scott

[R1R2]

For Transfusion Services - I think if you are getting extraneous or nonspecific reactions you should increase your rule outs to at least 2 homozygous (with exceptions, of course), but for the majority of workups I think one homozygous is perfect, especially with the sensitive methods most of us are using today.  Also, IMO, ruling out on heterzygous cells is a waste of time and dangerous for some antibodies that show dosage. 

 

R1R2

Edited June 1, 2013 by R1R2