To the Editor:
The recent publication, showing the difference in antibody detection using single antigen beads (SAB) compared to complement-dependent cytotoxicity (CDC) or enzyme-linked immunosorbent assay (ELISA), is striking . We feel that the accompanying editorial  somewhat dismisses the implications to patients of using SAB in isolation. If a transplant program relies solely on the basis of antibodies detected by SAB then a significant proportion of patients could be denied the opportunity of transplant inappropriately. In this laboratory, we continually challenge SAB results, following results over time, querying test consistency according to operator or positive control values and considering removing unacceptable antigen mismatches only present on one occasion.
One of us (DL) has shown that death-censored 5-year graft survival of 80.4% in donor-specific alloantibody (DSA)+ve transplants is achievable. Those transplants that were crossmatch negative, SAB positive, experienced decreased graft survival only when pretransplant DSA levels exceeded 10 000 mean fluorescence intensity , suggesting that the threshold for clinically significant antibodies may often be set too low. The clinical urgency of the patient will obviously determine “safe” levels of antibody to transplant against. We firmly advocate the use of SAB in conjunction with luminex screening and cell-based techniques to enable more rigorous risk stratification.
In contrast to the views expressed in the editorial , we believe that unexplained reaction patterns in SAB assays can, by accumulation, have a negative impact on patients' access to compatible donors. Gombos et al also reported antibodies against common specificities in patients with no history of sensitization such as HLA-B*08:01, which is present at 29.9% in our North-West England population (www.allelefrequencies.net) compared to 12.5% quoted .
Determination of epitope sharing on different specificities will aid in reaching more meaningful conclusions on the validity of results. For example, a local patient was reactive solely with the A*34:01 bead on class I SAB. None of the four amino acid substitutions distinguishing A*34:01 from A*34:02 were exclusive to A*34:01, being present on other proteins in the SAB panel. Although not their primary aim, Gombos et al support this with the observation that SAB reactions are limited to fewer beads in patients positive by SAB only, compared to those also positive by CDC and ELISA . Furthermore, it is interesting that of their three reported specificities with the highest prevalence, two, HLA-B*37:01 and C*17:01, have no epitopes unique to that individual allele. It would be interesting to know if the samples containing these antibodies had other detected specificities. The very nature of epitope sharing suggests that true HLA-reactive patients would react with greater numbers of beads within the SAB assay.
In summary, we believe that Luminex technology has led to improvements in transplantation, giving the opportunity to transplant without crossmatching, and in allocation to patients with antibodies, especially in heart transplantation, but now it is a prerequisite for all to look again. The publications [1, 2] should stimulate that. What we and others suggest is that SAB results should be taken into consideration as a potential risk factor, not ruling out the transplant without forethought.