Interpretation of antibody panels

[CAROL BARNEY]

A question has come up at our facility concerning interpreting antibody panels. When testing is completed is the correct method to eliminate possible antibodies to use the negative cell results obtained in testing and eliminate any antibody which should be positive for that cell based on the profile provided by the manufacturer? Or is the correct method the opposite?

[Lcsmrz]

"Ruling Out" has a higher predictive value -- if the serum does not react with one or more cells containing an antigen, then the probability of having the corresponding antibody is very, very small.

One would hope that, after crossing off, that the remaining pattern fits the antigen that was not crossed off -- I've been fooled before!

[jhaig]

Ruling out using negative cells is the way to go. You want to get rid of any panel cell that is positive to effectively "rule-out" that antibody. Also, it is preferable to rule out using homozygous panel cells for Rh, Kidd, Duffy and MNS groups. Kell groups are tough to rule out based solely on a homozygous cell. With Kell, and certain other groups, several heterozygous rule-outs should be fine, especially if you have a clear pattern on the panel and a good idea of what the antibody ID is.

Other low frequency antigens, like Kpa, Jsa, Lua etc., usually can't be ruled out because of the lack of reactivity. But every patient is different and I've been fooled before. Rule-outs will usually give you a very good idea of what you're dealing with, but they're not fool-proof. Take it from a fool...:cool:

[ovrwkd]

Please don't say lack of reactivity with antibodies to low prevelence antigens.

Reactivity should be the same as any antibody as well as the possibility of patient transfusion reaction to a unit positive for that antigen. Once antibodies to common antigens have been ruled out you may find that the 2-3+ reaction you have is your Wra or Jsa!

What does occur is the lack of commercial antisera to test units for these antigens. Rule of thumb is if your patient has an antibody to a "low freq" and supplied units are negative for any antigens for other antibodies discovered, you can use your patient sample as your screening serum.

Back to the rule out technique. Cross out on homozygous (double dose) antigens which are non-reactive in your testing is ther better way to go. To prove the probability (P-values) we commonly use 3 pos for the antigen and 3 neg, 2 pos and 5 neg.

[kate murphy]

Agree with ovrwkd (love your name) - to properly rule out, you need to cross out cells with homozygous antigens only which are non-reactive. Otherwise, you may cross out anitbodies showing dosage - the ones that react stronger (or only react) with homozygous cells.

We also use 3 pos and 3 negs to get 95% probability.

[janet]

Do you use one or two homozygous to rule out (when possible?)

[Max001]

Do you use one or two homozygous to rule out (when possible?)

It's been a while but we used one negative cell to rule out for all but Kell, IIRC.

We'd routinely store about 20 panels, many of them well past their expiration date, in the cold room for rule outs. We used to have three different manufacturers, each in use too, Ortho, Gamma and Immucor.

[AMcCord]

Using only 1 rule out/rule in is begging for trouble. Coincidence can bite you if you are not careful. (Such as: By coincidence, that panel cell has a slightly weaker expression of the antigen and your patient failed to react with it for that reason or By coincidence, that K pos cell your patient reacts with is also Co( positive and they have anti-Co( and not anti-K after all.) Do I speak from experience, you ask??? Yes, I do...learned that lesson a long time ago and don't aim to repeat it.

[Max001]

Using only 1 rule out/rule in is begging for trouble.

I BEG your pardon folks. (smacks forehead) Told you it's been a long time. We used three not one to rule out.

It's been a long time, almost 7 years. Sorry.

[Joanne P. Scannell]

Interesting ... you all rule out with just ONE cell all day, every day as you call your Antibody Screens 'negative'. Why would a panel be treated any differently?

The 3-cell rule is too often misused and misquoted. Yes, you need 3 reactive positives to be on the statistical side of right (eg. 3 K1-pos cells to call it Anti-K1) OR you need 3 non-reactive negatives to confidently identify an antibody to a high incidence antigen (eg. 3 Kpb-neg to call it Anti-Kpb). The statistical 'rule' is not '3 pos PLUS 3 neg'. Nor is it 'you can't say it's not there without 3 cells ... otherwise, we'd have to run full panels instead of 'antibody screens'.

I, too, speak from experience ... and yes, we use only homozygous cells to rule out (with a few exceptions, such as Anti-K1 which is both almost impossible to find a homozygous cell and doesn't tend to show dosage anyway). (And two heterozygous does not equal a homozygous, as some people have expressed in these conversations, so we don't go that route.)

And yes, got to keep an open mind to those 'hints' that sometimes haunt us ... eg. reacts to some, not all, homozygous cells. But that's whole 'nother ball game ...

[Philip Thornton]

We rule-out exclusively with homozygous or a rarely the combo of 2 to 3 heterozygous cells. This practice eliminates much of the ambiguity in antibody diagnosis.

[annie467]

I agree with JPCroke. People really mis-interpret the so-called rule of three. It does not mean that you must cross of every antigen three times. We see that concept a lot in recently-taught techs. A good exercise to demonstrate this is to perform a "rule-out" on your negative antibody screens.

The object of crossing off is two-fold: to identify your antibody, and to rule out the coincidence of additional antibodies underlying the pattern of the first antibody.

The first object is to identify the antibody. Crossing off only homozygous antigens will help if the antibody is showing dosage. Using out-of-date panels can be helpful for multiple antibody resolution.

Once you have identified an antibody, you are only looking for underlying antibodies. So absolutely, you must use negative-reacting panel & screening cells to do that. At our facility, we recommend that each antigen except low frequencies be crossed off with one homozygous cell whenever possible. If it is not possible, then heterozygous cells may be used, recognizing the risk of missing an antibody showing dosage. If an antibody shows dosage, testing three heterozygous cells does not really help detect dosing antibodies. Using expired panels to identify can be helpful, but entails a risk when ruling out other antibodies: you may not see an antibody for which the antigen expression has been weakened by age; panels expire for a reason. Your AHG crossmatch will detect an incompatibility due to additional antibodies not detected.

[jhaig]

We run full IgG gel crossmatches on any patient with a positive antibody screen that is going to be transfused. How does this affect single vs. double dose rule-outs? Will a full crossmatch lessen the need to rule out on homozygous cells only?

[Joanne P. Scannell]

I don't see the crossmatches being related to the antibody identification.

Crossmatch: If there are clinically significant antibodies, a 'full' crossmatch is mandatory. Period. (Again, back to the actually statement of the 'law' ... 'clinically significant' being the key words people often leave out.)

Antibody ID: I send out a 'high caution' to anyone using heterozygous cells to rule out antibodies that tend to or can show dosage. If the antibody is not demonstrable with heterozygous cells, it doesn't matter how many you run ... you will get a negative result and erroneously eliminate that antibody. I've seen many antibodies that react ONLY with homozygous cells ... some quite strongly (eg. 2+). Don't put yourself into a false security situation.

Unless you like surprises during the crossmatching ...

[janet]

I don't see the crossmatches being related to the antibody identification.

.....I've seen many antibodies that react ONLY with homozygous cells ... some quite strongly (eg. 2+). Don't put yourself into a false security situation.

Unless you like surprises during the crossmatching ...

I'd like to add that uless you're testing dosage of your crossmatched unit, that antibody that doesn't react with the heterozygous exclusion cell won't react with a heterozygous unit either. Crossmatching antigen negative units may give a false sense of security if you don't know it's heterozygous unit.

[AVANHORN]

I don't see the crossmatches being related to the antibody identification.

Crossmatch: If there are clinically significant antibodies, a 'full' crossmatch is mandatory. Period. (Again, back to the actually statement of the 'law' ... 'clinically significant' being the key words people often leave out.)

Antibody ID: I send out a 'high caution' to anyone using heterozygous cells to rule out antibodies that tend to or can show dosage. If the antibody is not demonstrable with heterozygous cells, it doesn't matter how many you run ... you will get a negative result and erroneously eliminate that antibody. I've seen many antibodies that react ONLY with homozygous cells ... some quite strongly (eg. 2+). Don't put yourself into a false security situation.

Unless you like surprises during the crossmatching ...

I just read your response to abs reacting to homozygous cells only. We have a patient with anti-D. The C & E antigens on our panel are heterozygous and we will not find homozygous cells for these. Do you go to the trouble of antigen typing the patient to see if they are E-C-, which is most likely the case, and give them E-,C- AHG crossmached units only? I have never really worried about this in the past especially since most Rh negative units are E-,C-. What's your take on this one?