Comparison of human leukocyte antigen immunologic risk stratification methods in lung transplantation

Outcomes after lung transplantation (LTx) remain poor, despite advances in sequencing technology and development of algorithms defining immunologic compatibility. Presently, there is no consensus regarding the best approach to define human leukocyte antigen (HLA) compatibility in LTx. In this study, we compared 5 different HLA compatibility tools in a high-resolution HLA-typed, clinically characterized cohort, to determine which approach predicts outcomes after LTx. In this retrospective single-center study, 277 donor-recipient transplant pairs were HLA-typed using next generation sequencing. HLA compatibility was defined using HLAMatchmaker, HLA epitope mismatch algorithm (HLA-EMMA), predicted indirectly recognizable HLA epitopes (PIRCHE), electrostatic mismatch score (EMS), and amino acid mismatches (AAMMs). Associations with HLA mismatching and survival, chronic lung allograft dysfunction (CLAD), and anti-HLA donor-specific antibody (DSA) were calculated using adjusted Cox proportional modeling. Lower HLA class II mismatching was associated with improved survival as defined by HLAMatchmaker (P < .01), HLA-EMMA (P < .05), PIRCHE (P < .05), EMS (P < .001), and AAMM (P < .01). All approaches demonstrated that HLA-DRB1345 matching was associated with freedom from restrictive allograft syndrome and HLA-DQ matching with reduced DSA development. Reducing the level of HLA mismatching, in T cell or B cell epitopes, electrostatic differences, or amino acid, can improve outcomes after LTx and potentially guide immunosuppression strategies.

The clinical utility and thresholds of Virtual and Halifaster Flow crossmatches in lung transplantation

INTRODUCTION

Immune sensitization, defined as the presence of alloreactive donor-specific antibodies (DSA), is associated with increased wait-times and inferior transplant outcomes. Identifying pre-transplant DSA with a physical cell-based assay is critical in defining immunological risk. However, improved solid phase antibody detection has provided the potential to forgo this physical assay. Here, we evaluated the association between DSA mean fluorescence intensity (MFI) and the recently introduced Halifaster Flow cytometry crossmatch (FXM) to determine if MFI could predict the outcome of FXM and whether a virtual crossmatch (VXM) would provide an accurate risk assessment.

METHODS

Sera from 134 waitlisted lung patients was retrospectively assessed by Halifaster FXM against lymphocytes preparations from 32 donors, resulting in 265 FXMs. HLA typing was performed to 2-field allelic level and Luminex single antigen beads (SAB) used to identify DSA. The association between FXM and Luminex MFI was calculated using ROC analysis. MFI threshold accuracy was confirmed using a separate validation cohort (174 recipient sera and 34 donors), whereby both virtual crossmatch (VXM) and FXMs were compared.

RESULTS

From the 265 FXM performed, 48 (18%) T-cell (TFXM) and 56 (21%) B-cell (BFXM) were positive. In the evaluation cohort, MFI thresholds of 2000 for HLA-A, B, DRB1 and >4000 for DQB1, were predictive of a positive FXM. The validation cohort of 233 paired FXM and VXM confirmed these MFI thresholds for both TFXM and BFXM with an accuracy of 91.4% and 89.3% respectively.

CONCLUSION

A positive VXM, defined with HLA-specific MFI thresholds predicts Halifaster FXM reactivity, and can potentially expedite organ allocation, by minimising the need for the more time-consuming flow cytometry crossmatch. This article is protected by copyright. All rights reserved.

HLA Matching at the Eplet Level Protects Against Chronic Lung Allograft Dysfunction

Donor selection in lung transplantation (LTx) is historically based upon clinical urgency, ABO compatibility, and donor size. HLA matching is not routinely considered; however, the presence or later development of anti-HLA antibodies is associated with poorer outcomes, particularly chronic lung allograft dysfunction (CLAD). Using eplet mismatches, we aimed to determine whether donor/recipient HLA incompatibility was a significant predictor of CLAD. One hundred seventy-five LTx undertaken at the Alfred Hospital between 2008 and 2012 met criteria. Post-LTx monitoring was continued for at least 12 months, or until patient death. HLA typing was performed by sequence-based typing and Luminex sequence-specific oligonucleotide. Using HLAMatchmaker, eplet mismatches between each donor/recipient pairing were analyzed and correlated against incidences of CLAD. HLA-DRB1/3/4/5+DQA/B eplet mismatch was a significant predictor of CLAD (hazard ratio [HR] 3.77, 95% confidence interval [CI]: 1.71-8.29 p < 0.001). When bronchiolitis obliterans syndrome (BOS) and restrictive allograft syndrome (RAS) were analyzed independently, HLA-DRB1/3/4/5 + DQA/B eplet mismatch was shown to significantly predict RAS (HR 8.3, 95% CI: 2.46-27.97 p < 0.001) but not BOS (HR 1.92, 95% CI: 0.64-5.72, p = 0.237). HLA-A/B eplet mismatch was shown not to be a significant predictor when analyzed independently but did provide additional stratification of results. This study illustrates the importance of epitope immunogenicity in defining donor-recipient immune compatibility in LTx.