An integrative algorithm combining HLA epitope registry, PIRCHE-T2, and PIRCHE-B outcomes to improve immunological risk stratification in kidney transplantation

TL;DR

This study combines three algorithms to better predict kidney transplant rejection risks, enabling earlier interventions and improved outcomes.

Contribution

The novel integration of Epregistry, PIRCHE-T2, and PIRCHE-B scores improves immunological risk stratification in kidney transplantation.

Findings

High-risk patients had significantly lower dnDSA-free and ABMR-free survival compared to low-risk patients.

PIRCHE-T2 scores were significantly associated with T-cell mediated rejection.

Combining multiple algorithms improves predictive accuracy for kidney transplant outcomes.

Abstract

Kidney transplantation remains the most effective treatment for end-stage kidney disease. Still, the development of de novo donor-specific antibodies (dnDSA) increases the risk of rejection and allograft failure. While molecular matching algorithms assess B-cell and T-cell epitope mismatches, no single method fully captures rejection risk across immune pathways. This study combines the HLA Epitope Registry (Epregistry), PIRCHE-T2, and PIRCHE-B scores to enhance risk stratification, allowing for early intervention in high-risk recipients and improving long-term outcomes.

A retrospective study of 594 kidney transplant recipients in Saskatchewan (1981–2021), Canada, was conducted, tracking de novo donor-specific antibodies (dnDSA) development until January 2024. Epitope mismatch scores were calculated using Epregistry, PIRCHE-T2, and PIRCHE-B, and receiver operating characteristic (ROC) curve analysis determined the optimal cutoff values for predicting dnDSA formation. Patients were categorized into high-risk (all scores > cutoff), intermediate-risk (one algorithm > cutoff), and low-risk (all scores < cutoff) groups. Kaplan-Meier survival analysis evaluated dnDSA-free survival across risk categories.

Among 594 recipients, 104 individuals (17.5%) developed de novo DSA; of these, 29 patients developed more than one, resulting in a total of 146 dnDSA events. The most frequently targeted locus was HLA-DQ (72/146, 49.3%), followed by HLA-DR (25/146, 17.1%) and HLA-A (24/146, 16.4%). The optimal cutoff values for predicting dnDSA were 22.5 (Epregistry), 30.5 (PIRCHE-T2), and 5.5 (PIRCHE-B) for Class I, and 15.5 (Epregistry), 17.5 (PIRCHE-T2), and 5.5 (PIRCHE-B) for Class II (all p < 0.05). Across all molecular mismatch load metrics, Kaplan–Meier analysis demonstrated significantly lower dnDSA-free and antibody-mediated rejection (ABMR)-free survival among high-risk recipients compared with low-risk recipients (log-rank p < 0.001). In addition, both the PIRCHE-T2 score at HLA Class I loci and the overall PIRCHE-T2 score were significantly associated with T-cell mediated rejection (TCMR) (p < 0.01).

Integrating Epregistry, PIRCHE-T2, and PIRCHE-B enhances risk stratification for kidney transplant recipients. Epregistry and PIRCHE-B evaluate HLA antibody epitope mismatches, and PIRCHE-T2 focuses on T-cell mismatches. Applied in conjunction, the methods show improved predictive accuracy, making this multi-algorithm approach more effective in identifying high-risk patients. By enabling earlier interventions and personalized immunosuppressive strategies, this model has the potential to improve long-term transplant success.

Figures