In support of the findings of Christopher F. Bryan et al.

Abstract

We have read with interest the article by Dr. Bryan and co-workers (1) regarding the detection of HLA IgG class I alloantibodies (aAb) using enzyme-linked immunoassays (EIA) (2). We have similarly performed an evaluation of the PRA-STAT EIA kit (manufactured by SangStat Medical Corp. and marketed by Nextran, an affiliate of Baxter Healthcare) and compared it with the standard “NIH” C-dependent cytotoxicity (CDC) and antiglobulin (AHG) CDC PRA screening methods (3). We find that in our laboratories as well, this commercial EIA kit cannot reliably detect HLA class I aAb. In our study, 30 HLA-specific class I aAb (3 IgM and 27 IgG aAb) of reagent typing quality and 150 sera, including blind duplicates and dilutions obtained from renal transplant candidates, were screened using all three procedures. The former sera had previously been characterized for public and private HLA aAb by differential adsorption with type-specific platelets (4) or purified HLA antigen affinity columns (5), including the recovery of adsorbed aAb by acid elution. These sera, and others, had also been screened (by both CDC methods using local cell panels) in eight major U.S. histocompatibility typing laboratories by a Public HLA Antigen Workshop sponsored by the American Society for Histocompatibility and Immunogenetics (ASHI) (6). Since we intend to publish a full account of our comparative studies, we will focus only on the results of the 27 IgG HLA typing sera screened in parallel by standard CDC and PRA-STAT. Like Bryan et al. (1), we found that PRA-STAT is not always concordant to even the less sensitive standard CDC method (2) for detection of complement-fixing HLA IgG aAb: 19 of 27 HLA-specific sera (70.3%) did not have the same HLA specificity when screened using these two methods. Most troublesome, 11 of these 19 sera were nonreactive in PRA-STAT assays even though there was clear HLA specificity by NIH CDC (one HLA-A1-36 aAb, one A11 aAb, three B7, two B8, one B14, one B51, one B35-51, and one Cw4 aAb). (The lack of reactivity of these sera in EIA was subsequently confirmed at the SangStat laboratory facilities as well.) In addition to these 19 sera, four sera had EIA PRA >40% points greater than standard CDC, two sera were nonreactive by CDC and EIA but reactive by AHG-CDC, one serum had ambiguous reactivity, and one serum was concordant between EIA and CDC. Similar poor correlations in percentage PRA and specificity assignment were observed when we compared PRA-STAT with AHG-CDC. Notable was a pronounced inability to detect underlying public HLA aAb by EIA; these public specificities were fully documented by AHG-CDC in the ASHI workshop and in topographic analysis of HLA epitopes using flow cytometry (4, 5). In summary, we fully support the conclusions of Bryan et al. (1) that the PRA-STAT EIA procedure in its present commercial form does not represent a satisfactory screening procedure for detection of HLA-specific IgG aAb, even though interlaboratory reproducibility of this assay was excellent. We believe that, possibly with further standardization and modification in test sensitivity, these deficiencies can be overcome. We would like to convey one final note of caution to other laboratories that are considering use of PRA-STAT for clinical HLA aAb screening. Even though the U.S. Food and Drug Administration has approved this EIA kit for diagnostic use in screening for the presence of HLA aAb and thus for determination of the corresponding HLA class I specificity, laboratories should perform extensive in-house comparative studies with their conventional CDC HLA aAb screening procedures before embarking on the use of PRA-STAT in a clinical setting. Furthermore, the current ASHI laboratory standards for clinical histocompatibility testing (section H1.100) do not recognize alternative PRA screening procedures such as PRA-STAT EIA as acceptable methods for HLA antibody detection. Thomas C. Fuller1,2; Anne Fuller1; Jeffrey M. McCormack3,4; Glenn E. Rodey3 Department of Pathology; University of Utah School of Medicine; Salt Lake City, Utah 84132 Department of Pathology and Laboratory Medicine; Emory University School of Medicine; Atlanta, Georgia 30322 Department of Pathology; Baylor University Medical Center; Dallas, Texas 75246