56 Retrospective Comparison Between HLA-Match and Nonbinary Platelet Crossmatch. (11th January 2018)

• Record Type: Journal Article
• Authors: Gniadek, Thomas; Gillen, Barbara; Novotny, Angela; Kinney, Stephanie; Van Buren, Nancy; Gorlin, Jed; Cohn, Claudia
• Is Part Of: American journal of clinical pathology. Volume 149(2018)Supplement 1
• Page Start: S193
• Page End: S193
• Publication Date: 2018-01-11
• Subjects: Diagnosis, Laboratory -- Periodicals; Pathology -- Periodicals; 616.07
• DOI: 10.1093/ajcp/aqx149.425
• Languages: English
• ISSNs: 0002-9173

Abstract

Abstract: This retrospective study examined the relationship between degree of platelet crossmatch incompatibility (XM) and patient anti-donor HLA antibody mean fluorescence intensity (MFI). A total of 10 patients were selected, each known to have had platelet crossmatch testing (Capture-P, Immucor Inc, Peachtree Corners, GA), recorded degree of crossmatch incompatibility (0 to 4+), and cotemporaneous HLA MFI testing (Luminex Corp, Austin, TX). Tests were performed to evaluate for antibody-mediated platelet refractoriness. MFI was normalized to self-allele beads. If multiple beads were present for a serotype, MFIs were averaged. In total, 155 units were tested by crossmatch; donor HLA type was known for 112, and these were analyzed further. For the four HLA-A/B alleles, there were zero 4-point, five 3-point, 15 2-point, 44 1-point, and 48 0-point matches. XM was 1+ for 26, 2+ for 1, 3+ for 3, 4+ for 3, and 0 for 79 units. Positive control results ranged from 2 to 4+. XM had a positive correlation with MFI against donor HLA (Pearson 0.48 for maximum MFI vs 0.49 for summed MFI). The correlation was not perfect: XM was positive for 9.5% of units with maximum MFI <1, 000 and negative for 24% of units with maximum MFI >20, 000. A multivariate linear mixed effect model (LM) considered the effect of patient, major ABO incompatibility, crossmatch positive control result, as well as MFI. MFI was included in the LM as either sum or maximum MFI of donor HLA-A/B or as sum or maximum MFI of donor HLA-A and HLA-B alleles separately. Maximum MFI against HLA-A showed the most significant correlation with XM ( P = 9.4e-5), compared with maximum MFI against HLA-B ( P = .48). Major ABO incompatibility was consistently significant ( P ≅ .0002). Maximum MFI against HLA-A correlated slightly better than either the summed MFI or maximum MFI against both HLA-A/B (t value = 4.08 vs 3.37 vs 2.99), but all were significant ( P < .03). Similarly, summed MFI against HLA-A was less significant than maximum MFI against HLA-A (t value = 2.79 vs 4.08). Of note, the average number of mismatched alleles was lower for HLA-A vs B (1.4 vs 1.8, P = 5e-6). Additionally, the average maximum MFI was higher for donor HLA-A vs B (5, 569 vs 2, 939, P = .01), despite a lower average Luminex bead MFI for HLA-A vs B (1, 775 vs 2, 660, P = .03). In conclusion, maximum MFI to HLA-A and major ABO incompatibility were patient/donor factors that correlated most with XM. The patient population appears preferentially alloimmunized against the HLA-A alleles in the donor population. Additional studies are needed to determine whether this was due to quantitative differences in levels of HLA-A vs HLA-B on donor platelets, allele-specific prior exposure, allele frequency, or differences in epitopes between the proteins used for MFI testing and the common HLA proteins in the platelet donor population.