Journal Article Epublish
PubMed: 42621113 DOI: 10.1016/j.ekir.2026.106702
Urinary CXCL9 Across the Spectrum of Kidney Allograft Inflammation.
Kidney Int Rep 11(10):106702
Authors: Esteban Cortes Garcia, Karolien Wellekens, Claire Tinel, Stéphanie Chhun, Charlotte Debiais-Deschamps, Juliette Leon, Antoine Troger, Camille Roger, Carole Burger, Marie-Bénédicte Le Stang, Emilie Lebraud, Marie-Noelle Peraldi, Christophe Legendre, Frank Martinez, Rebecca Sberro-Soussan, Julien Zuber, Marion Rabant, Fabiola Terzi, Marie Essig, Wilfried Gwinner, Pierre Marquet, Elisabet Van Loon, Maarten Naesens, Dany Anglicheau, Olivier Aubert
Affiliations
- Université Paris Cité, Inserm UMR1151, Necker Enfants Malades Institute (INEM), Paris, France.
- Department of Microbiology, Immunology and Transplantation, KU Leuven, Leuven, Belgium.
- Department of Nephrology and Kidney Transplantation, Dijon University Hospital, Dijon, France.
- Faculty of Medicine, Paris Cité University, Paris, France.
- Institut Necker-Enfants Malades (INEM), INSERM U1151-CNRS UMR 8253, Paris, France.
- AP-HP, Laboratory of Immunology, Necker-Enfants Malades Hospital, Paris, France.
- Department of Kidney and Metabolic Diseases, Transplantation and Clinical Immunology, Necker Hospital, Assistance Publique - Hôpitaux de Paris, Université Paris Cité, Paris, France.
- Department of Pathology, Necker-Enfants Malades Hospital, APHP, Université Paris Cité, Paris, France.
- Department of Nephrology, Ambroise-Paré hospital, Assistance Publique - Hôpitaux de Paris, Paris-Saclay University, Villejuif, France.
- Department of Nephrology and Hypertension, Hannover Medical School, Hannover, Germany.
- INSERM U1248 Pharmacology & Transplantation, University of Limoges, Limoges University Hospital, Limoges, France.
- Imelda Ziekenhuis, Bonheiden, Belgium.
Abstract
BACKGROUND
Kidney allograft inflammation spans a heterogeneous spectrum of phenotypes defined by the Banff 2022 classification. We assessed whether urinary C-X-C motif chemokine ligand 9 (uCXCL9) normalized to creatinine (uCXCL9:Cr) captures inflammatory burden across this spectrum and provides incremental diagnostic utility beyond standard variables.
METHODS
We analyzed 3181 biopsy-urine pairs from a development cohort (1341 biopsies from 921 patients) and an independent external validation cohort (1840 biopsies from 917 patients). Associations with Banff 2022 diagnostic categories were evaluated using multinomial logistic regression with stability-based selection. Incremental performance was assessed using the Hand & Till multiclass area under the curve (AUC) and decision curve analysis.
RESULTS
uCXCL9:Cr demonstrated a graded increase across the inflammatory spectrum, with lowest levels in biopsies without specific lesions (median 4.55 ng/mmol), intermediate levels in antibody-mediated phenotypes (15.14), higher levels in acute T cell-mediated rejection (TCMR; 46.11), and the highest levels in mixed inflammatory patterns (53.18-134.05; P < 0.001). Levels increased proportionally with microvascular and tubulointerstitial activity scores (all P < 0.001). In multivariable multinomial analyses, uCXCL9:Cr was independently associated with each inflammatory category across the Banff 2022 spectrum (relative risk ratio [RRR] range: 1.68-3.44; all P ≤ 0.004). Addition of uCXCL9:Cr increased multiclass AUC from 0.797 to 0.831 in the development cohort and from 0.801 to 0.821 in the validation cohort (both P = 0.001). Incorporation of uCXCL9 was associated with a potential reduction in unnecessary biopsies in decision-curve analyses.
CONCLUSION
uCXCL9 reflects active kidney-graft inflammation across modern Banff phenotypes and provides clinically meaningful diagnostic information beyond standard variables, supporting its potential to inform biopsy decisions and monitoring in kidney transplant recipients, pending prospective validation.
MeSH
Phenotype, Kidney, Biopsy, Transplants, Allografts