Director-in-Trainee Brigham and Women's Hospital Boston, MA
Aim: vXM is purposed to reduce cold ischemia time and improve organ allocation, but false negatives could potentially risk graft rejection. This study evaluated vXM predictive accuracy for T-cell flow cytometry crossmatch (TFCXM) positivity using various DSA thresholds and assessed additional factors that could contribute to false-negative vXM outcomes.
Methods: A retrospective analysis of 3,803 TFCXM that were tested against sera screened by LABScreen single antigen beads (SA) or mixed bead (PSL) assays. TFCXM positivity was defined by delta fluorescence units (DFU) ≥150. DSA assessment for physical TFCXM was performed using two thresholds: standard (>3000) and low (>1000), per program’s decision. Retrospective vXM was conducted using low threshold and includes HLA-C antibodies, epitope sharing, and donor A/B locus homozygosity, which were not included in original DSA assessment.
Results: The low DSA threshold improved sensitivity (66.9% vs. 29.6%) and negative predictive value (NPV: 94.3% vs. 79.8%) for predicting TFCXM outcome, with moderate reduction in PPV (75.2% vs 83.1%). After excluding 42 BFCXM+ cases with Class II DSA only, Class I DSA at the lower threshold showed stronger correlation with TFCXM than with BFCXM (NPV 94.3% vs. 77.0%), while the true positive rate (sensitivity) became much lower for BFCXM (30.5% vs. 66.9%). ROC analysis of HLA A/B/C DSA MFI showed an AUC of 0.85, indicating good predictive value and an optimal cutoff of 1000 for TFCXM classification. Incorporating HLA-C antibodies, epitope analysis, and donor homozygosity further improved sensitivity (74.3%) and NPV (95.3%).
Conclusion: Lowering the DSA MFI threshold to 1000 significantly improves vXM sensitivity and NPV for predicting TFCXM. Adding HLA-C, epitope matching, and donor homozygosity enhances predictive accuracy.
