Archetypal Analysis of Deceased Donor Kidneys: A Molecular Approach for Posttransplant Outcomes
Ontology highlight
ABSTRACT: This prospective, single-center study (NCT06171438) included 276 kidneys from 174 deceased donors. Donor procurement biopsies were performed on the backtable and part of the biopsy was stored for future RNA extraction and molecular profiles analysis using Affymetrix microarrays. Three molecular classifiers (the logistic regression model, the Poisson regression and the extreme gradient boosting) with 10-fold cross validation were employed to identify donor categories based on clinical variables (age, BMI, hypertension, marginal kidney) and histological biopsy scores (the vascular fibrous intimal thickening (cv), interstitial fibrosis (ci) and tubular atrophy (ct) and arteriolar hyaline thickening (ah). Archetypal analysis was applied to determine molecular phenotypes and their association with short- and long-term transplant outcomes. Three donor molecular archetypes were identified. Marginal archetype was characterized by older age (median 68 years), extensive chronic morphological changes (IFTA>0 in 66%, ah>1 in 59%, cv>1 in 36%), and the highest Kidney Donor Risk Index (KDRI). Ideal archetype was associated with young age (median 42 years), low KDRI, minimal chronic histological changes (IFTA>0 in 39%, ah>1 in 33%, cv>1 in 13%) and the highest posttransplant eGFR at 12 months (p<0.001). Intermediate archetype 3 was associated with older age (median 63 years) however better eGFR at 12 months follow-up despite similar chronic histological changes (IFTA>0 in 73%, ah>1 in 63%, cv>1 in 56%). Molecular profiling outperforms histology in procurement biopsies and enables the identification of ECD kidneys with better posttransplant kidney allograft function.
ORGANISM(S): Homo sapiens
PROVIDER: GSE290167 | GEO | 2025/12/22
REPOSITORIES: GEO
ACCESS DATA