Pulled from PubMed
Daniel Yoo # 1, Gillian Divard # 1 2, Marc Raynaud 1, Aaron Cohen 3, Tom D Mone 3, John Thomas Rosenthal 4, Andrew J Bentall 5, Mark D Stegall 6, Maarten Naesens 7, Huanxi Zhang 8, Changxi Wang 8, Juliette Gueguen 9, Nassim Kamar 10, Antoine Bouquegneau 11, Ibrahim Batal 12, Shana M Coley 12, John S Gill 13, Federico Oppenheimer 14, Erika De Sousa-Amorim 14, Dirk R J Kuypers 7, Antoine Durrbach 15, Daniel Seron 16, Marion Rabant 17, Jean-Paul Duong Van Huyen 1 17, Patricia Campbell 18, Soroush Shojai 18, Michael Mengel 18, Oriol Bestard 16, Nikolina Basic-Jukic 19, Ivana Jurić 19, Peter Boor 20, Lynn D Cornell 21, Mariam P Alexander 21, P Toby Coates 22, Christophe Legendre 1 23, Peter P Reese 1 24, Carmen Lefaucheur 1 2, Olivier Aubert 1 23, Alexandre Loupy 25 26
In kidney transplantation, day-zero biopsies are used to assess organ quality and discriminate between donor-inherited lesions and those acquired post-transplantation. However, many centers do not perform such biopsies since they are invasive, costly and may delay the transplant procedure. We aim to generate a non-invasive virtual biopsy system using routinely collected donor parameters. Using 14,032 day-zero kidney biopsies from 17 international centers, we develop a virtual biopsy system. 11 basic donor parameters are used to predict four Banff kidney lesions: arteriosclerosis, arteriolar hyalinosis, interstitial fibrosis and tubular atrophy, and the percentage of renal sclerotic glomeruli. Six machine learning models are aggregated into an ensemble model. The virtual biopsy system shows good performance in the internal and external validation sets. We confirm the generalizability of the system in various scenarios. This system could assist physicians in assessing organ quality, optimizing allograft allocation together with discriminating between donor derived and acquired lesions post-transplantation.