OR49 A pitfall of virtual crossmatching in pediatrics: Multiple solid phase reagents can show strong HLA antibodies that are not detectable in flow cytometry crossmatches

Abstract

The usefulness of virtual crossmatching (VXM) is limited by the accuracy of current assays for HLA antibodies (Ab). We describe two cases of a specific false-positive pattern of class II Ab in toddler-age heart transplant (HTx) candidates. Subject 1 was a 13 month old female awaiting HTx for dilated cardiomyopathy, with no history of sensitizing events other than routine immunizations. Ab binding to DR4, DR7, and DR9 molecules in single antigen reagents from One Lambda (lot 11, 5,000–8,000 MFI) and Immucor (11,000–15,000 MFI) was suspect because it included a self-antigen (DRB1∗07:01). These specificities were also detectable in sera pretreated with DTT or fetal calf serum as well as with phenotype reagents (One Lambda and Immucor). Surrogate flow cytometry crossmatches (SFCXM) with DR4+ or DR7+ B cells treated with pronase were negative (positive without pronase). DR4, DR7, and DR9 specificities persisted during the 3 mo transplant waiting period. The patient underwent successful HTx with a DR4+ donor. The FCXM with donor B cells was negative (positive without pronase). Soon after transplant the strong DR4, DR7, DR9-signature was not detected. Nearly 2 years after transplant, no clinical or pathologic Ab-mediated or cellular rejection has occurred. Subject 2 (DR1,13) was listed for HTx due to dilated cardiomyopathy as an 18 month old male, without history of sensitizing events other than routine immunizations. His HLA Ab profile with single antigen reagents (One Lambda lot 12) showed a similar DR4, DR7, DR9 pattern (9,000–14,000 MFI) which was supported by data from phenotype reagents. SFCXM with DR4+ and DR7+ B cells were negative. This patient is still awaiting heart transplantation, and donors with these HLA types will not be excluded by VXM. DR4, DR7 and DR9 avoidance would exclude 50% of the American population, which may tragically slow the HTx waiting process for this patient population. We are skeptical of this particular pattern of antibodies, particularly in toddlers without previous HLA presensitizing events. These observations suggest that exclusive use of VXM without cell-based crossmatches to support clinical significance of the antibodies may unnecessarily limit donors. Investigation is underway to determine the cause for this discrepancy between the solid phase assays and cell-based crossmatches.