Abstract: PO2174
Sparse Intragraft Molecular Classifiers for Antibody-Mediated and T Cell-Mediated Kidney Transplant Rejection: Development, Validation and Clinical Value
Session Information
- Transplantation: Clinical - Noninvasive Biomarkers, Immune Regulation, and Fascinomas
November 04, 2021 | Location: On-Demand, Virtual Only
Abstract Time: 10:00 AM - 12:00 PM
Category: Transplantation
- 1902 Transplantation: Clinical
Authors
- Callemeyn, Jasper, Department of Microbiology, Immunology and Transplantation, KU Leuven, Leuven, Belgium
- Nava-Sedeño, Josue Manik, Technische Universitat Dresden, Dresden, Sachsen, Germany
- Gwinner, Wilfried, Medizinische Hochschule Hannover, Hannover, Niedersachsen, Germany
- Anglicheau, Dany, Hopital Universitaire Necker-Enfants Malades Bibliotheque Jean Hamburger Pierre Royer, Paris, Île-de-France, France
- Marquet, Pierre, Universite de Limoges, Limoges, Nouvelle-Aquitaine, France
- Deutsch, Andreas, Technische Universitat Dresden, Dresden, Sachsen, Germany
- Hatzikirou, Haralampos, Technische Universitat Dresden, Dresden, Sachsen, Germany
- Naesens, Maarten, Department of Microbiology, Immunology and Transplantation, KU Leuven, Leuven, Belgium
Background
Although the transcriptional landscapes of antibody-mediated rejection (ABMR) and T-cell mediated rejection (TCMR) have been largely elucidated, applying these gene signatures in transplant clinics is hampered by the large number of features and difficult integration with histological findings. We aimed to develop and validate a sparse molecular classifier for ABMR and TCMR.
Methods
In a discovery cohort of 224 kidney transplant biopsies, microarray gene expression was applied to build two separate prediction models for presence of ABMR or TCMR. Variable selection for logistic regression was performed by lasso regularization. The diagnostic accuracy and prognostic value of the obtained ABMR and TCMR classifiers were assessed in two external validation cohorts.
Results
From the discovery cohort, a 2-gene ABMR classifier (PLA1A, GNLY) and 2-gene TCMR classifier (IL12RB1, ARPC1B) were derived. In the first validation cohort (N=403 biopsies), diagnostic accuracy was retained for both ABMR (ROC-AUC 0.80, 95% CI 0.75-0.85) and TCMR (ROC-AUC 0.83, 95% CI 0.77-0.89), also allowing discrimination between pure and mixed phenotypes. In the second validation cohort (N=282 biopsies), molecular ABMR and TCMR scores predicted graft failure (respective time-integrated AUC of 0.82 and 0.83) and identified kidneys at risk for graft failure which were not picked up by routine histology.
Conclusion
We identified and validated an intragraft 2-gene ABMR classifier and 2-gene TCMR classifier that can be used as diagnostic and prognostic tools. Robust variable selection models can yield parsimonious molecular classifiers for kidney transplant rejection, facilitating their interpretation and clinical implementation.