A Reference Laboratory supervisor in a Rocky Mountain state reports that for the past ten or so years her laboratory has used an 'in-house' panel of "HTLA antigen negative" reagent red cells to help them determine (in many cases) the specificity of an HTLA antibody, rather then merely calling the antibody a "probable HTLA-like antibody". They 'make' this panel from red cells that they acquire from antibody identification panels, based on phenotype information provided by the manufacturers. Of late, they have noted that reagent red cell manufacturers are providing less 'extended phenotype information' with the red cell panels, so that her laboratory no longer has enough information to create the 'in-house' HTLA antigen negative panel. This has forced them to re-define their policy about trying to actually determine the specificity of an HTLA antibody. They wonder if they are no longer able to identify the specificity of an HTLA antibody, would colleagues think that it is sufficient to merely call the antibody "HTLA-like", and focus on ruling out the presence of additional (underlying) alloantibodies? In the absence of an HTLA antigen negative panel, how do others determine the specificity of the HTLA antibodies they detect, or does it even matter?

The following comments have been received.

ADDENDA Mar. 29, 2006

1. Professor W. John Judd at the University of Michigan Medical Center (attribution used with permission), reports that in his experience the term "HTLA" is a misnomer, usually applied to antibodies to KNOPS antigens, but some would consider Ch/Rg, Yta, Csa, JMH and LU antibodies to be part of that group. He reminds us that KNOPS antibodies are not always of high titer, and avidity has nothing to do with their weak reactivity; rather, it is a matter of low antigen site density on RBCs.

Professor Judd says, "We try and identify the aforementioned specificities to the greatest extent possible, based on the reactions of ficin-treated RBCs and DTT-treated RBCs, serum inhibition and titration studies, and the results of antiglobulin tests with a monoclonal anti-IgG that does not react with IgG4. Antibodies to Ch/Rg and JMH do not react with ficin treated RBCs; anti-Ch/Rg are inhibited by pooled plasma; anti-JMH is nonreactive with DTT-treated RBCs and most examples do not react in antiglobulin tests with the monoclonal anti-IgG. Antibodies to Kna and Yta react variably with ficin-treated RBCs and are usually nonreactive with DTT-treated RBCs; sometimes it is helpful titrate the antibody against ficin-treated, DTT-treated and untreated RBCs to see the effect of the treatments, and it is also helpful to put the tests up on anti-IgG columns in order to make meaningful comparisons. We have also noted that KNOPS antibodies react better in albumin-IATs than by LISS methods. We also use RBCs and sera from previously identified cases to help us in the identification process; KNOPS null cells are particularly useful and we are fortunate to have some frozen."

Professor Judd acknowledges that at his academic center they believe it is important to differentiate KNOPS antibodies from anti-Yta, since the latter can be clinically important.

ADDENDA Mar. 31, 2006

2. Two experienced SBB Technologists in Missouri indicate that they agree with Professor Judd's comments. They also agree with the Rocky Mountain Reference Lab Supervisor that there has been a steep decline in the "extended typing " information provided by reagent red cell manufacturers. They report that their laboratory does not refer to these antibodies as HTLA's in reports, although they admit to using that terminology in routine conversation. Like Professor Judd, they have also found that many of these antibodies do not have a high titer. They rarely try to determine an exact specificity for these antibodies, because many of their cells are not well characterized. They acknowledge that DNA genotype testing done by Joann Mould's group has demonstrated that the serological classification of several of such red cells was incorrect (not a new concept to red cell serologists). They do try to place these antibodies in a system/collection (i.e., Knops, Chido, Rogers, Cost...), and they use ficin, DTT, neutralization and their frozen cell collection to help classify these antibodies. Their method of choice for working with these antibodies is to use PEG enhancement and an indirect antiglobulin test. They have not had frequent success using a monoclonal anti-IgG that lacks reactivity with IgG4, as mentioned by Professor Judd, but do agree it can be a useful tool. In conclusion, like Professor Judd, their main concern is to determine if any clinically significant antibody might also be present in the patient's serum/plasma.

ADDENDA Sept. 17, 2007

3. A colleague in Virginia reports that there is little guidance or structure in the protocol at their facility for working up suspected HTLA antibodies, other than to rule out other significant alloantibodies if possible, and perform an antiglobulin crossmatch. They are capable of doing serum inhibition studies and titration studies, but more often than not, they do not do these studies. When the inquiring colleague suspects an HTLA antibody is present, six cord blood specimens are washed thoroughly and run against the patient's plasma in tube and/or gel micro column as cord red cells either weakly express HTLAs or fail to express them at all. The inquiring colleague's question is... does anyone else avail themselves of using cord red cells to discriminate between HTLA antibodies and other alloantibodies and is there a precedent in the literature for such an approach? The inquiring colleague suspects that such an approach might not have caught on because cord red cells are not standardized or phenotyped as one would find with commercially prepared red cells that are under stringent quality control. However, as was pointed out earlier in this discussion, HTLA "panels" and extended phenotypes have become more scarce.

ADDENDA Sept. 20, 2007

4. An experienced immunohematologist who is affiliated with an Immunohematology Reference Laboratory comments that the approach proposed on Sept. 17, 2007 by the colleague located in Virginia may be a fast way to assess a sample if the blood needs of a patient are urgent. However, the responding colleage cautions that it would be of concern if an Immunohematology Reference Laboratory relied solely on this practice as there are a series of antigens that are not or weakly expressed on umbilical cord cells. The cord cells could be phenotyped to ensure that common antigens are expressed in a double dose state thus ensuring antibodies to common antigens are detected and not missed. This would be an inherent risk of using non-typed umbilical cord cells solely. If this technique is used in parallel with other testing techniques, it may be useful. It would be valuable to provide parallel testing data on the use of this technique. As the author of the comment states, cord cells are not released from commercial companies, thus there is concern about the storage requirements being met and use of the cell within appropriate dating periods. There are also concerns about infectious disease testing or lack thereof that may expose the technologists to infectious agents not present in commercial cells.

Please submit comments to the e-Network Forum.

Ira A. Shulman, MD
CBBS e-Network Forum Editor & Moderator