%0 Conference Paper %F Oral %T Activated natural killer cell transcripts predict graft outcome in kidney transplantation: The BIOMARGIN study %+ University Hospitals Leuven [Leuven] %+ Laboratoire d'analyse des données et d'intelligence des systèmes (LADIS) %+ Département de Néphrologie [CHU Necker] %+ Hannover Medical School [Hannover] (MHH) %+ Pharmacologie des Immunosuppresseurs et de la Transplantation (PIST) %A Yazdani, Saleh %A Gazut, Stéphane %A Anglicheau, Dany %A Gwinner, Wilfried %A Essig, Marie %A Marquet, Pierre %A Naesens, Maarten %< avec comité de lecture %B 18th Congress of the European Society for Organ Transplantation %C Barcelona, Spain %S Transplant international %V 30 %N S2 %P p110 - OS308 %8 2017-09-24 %D 2017 %R 10.1111/tri.13051 %K kidney transplantation %K immune cell subtypes %K graft failure %K ABMR diagnosis %K NK cell transcripts %K NK cell genes %K deconvolution algorithm %K RNA transcripts %K biopsy sample %K transcriptomics %K Cox %K biomarkers %Z Life Sciences [q-bio]/Biochemistry, Molecular Biology/Genomics [q-bio.GN] %Z Life Sciences [q-bio]/Quantitative Methods [q-bio.QM] %Z Life Sciences [q-bio]/Human health and pathology/Urology and NephrologyConference papers %X Background: Immune cells have been implicated in the graft failure after kidney transplantation. However, the best therapeutic target(s) among immune cells subtypes have not been identified yet. We aimed therefore to identify which immune cell subtypes possess the highest prognostic value for graft failure.Methods: We recently described the role of activated NK cell transcripts in ABMR diagnosis and activity in BIOMARGIN. Here, we applied an innovative deconvolution algorithm to estimate the relative fraction of immune cells subsets, using RNA transcripts expression of 547 genes in a dataset of 282 indication biopsy samples. These estimates were corrected for the globalinflammation in each biopsy.Results: Analysis of estimated immune cells subsets infiltration, based on deconvolution of transcriptomic data, showed that activated NK cells have the highest discriminative power in predicting graft failure at both 1 and 2 years post-biopsy in all biopsies; irrespective of rejection (AUC = 0.74; p < 0.0001 and AUC = 0.67; p = 0.0006, respectively), and within biopsies with rejection (AUC = 0.79 (p = 0.002) and AUC = 0.68 (p = 0.03), respectively). Cox proportional hazard analyses confirmed that activated NK cells were the cell type most robustly associated with graft failure (HR 3.60; p < 0.0001), especially in biopsies with acute rejection (HR 3.89; p = 0.002). We then evaluated individual NK cell transcripts expression in relation to graft failure. Using both the differentially expressed genes from the deconvolution algorithm and a literature-based NK-cell related transcript set, the prognostic performance of individual activated NK cell genes for post-biopsy kidney allograft failure was confirmed.Conclusion: These findings, in addition the emerging role of NK cells infiltration in ABMR diagnosis and activity, offer promising new avenues for development of NK-cell targeted prognostic biomarkers and therapeutic strategies to prevent or attenuate graft failure. %G English %L cea-01883442 %U https://cea.hal.science/cea-01883442 %~ CEA %~ UNILIM %~ CNRS %~ APHP %~ OPENAIRE %~ UMR-S850 %~ GEIST %~ DRT %~ CEA-UPSAY %~ UNIV-PARIS-SACLAY %~ CEA-UPSAY-SACLAY %~ LIST %~ IPPRITT %~ CHU-UNIV-PARIS5 %~ DM2I %~ TEST-HALCNRS %~ GS-ENGINEERING %~ GS-COMPUTER-SCIENCE %~ OMEGAHEALTH %~ DIN