Published in

BMJ Publishing Group, BMJ Surgery, Interventions, & Health Technologies, 1(5), p. e000137, 2023

DOI: 10.1136/bmjsit-2022-000137

Links

Tools

Export citation

Search in Google Scholar

Distinct phenotypes of kidney transplant recipients aged 80 years or older in the USA by machine learning consensus clustering

This paper is made freely available by the publisher.
This paper is made freely available by the publisher.

Full text: Download

Green circle
Preprint: archiving allowed
Green circle
Postprint: archiving allowed
Green circle
Published version: archiving allowed
Data provided by SHERPA/RoMEO

Abstract

ObjectivesThis study aimed to identify distinct clusters of very elderly kidney transplant recipients aged ≥80 and assess clinical outcomes among these unique clusters.DesignCohort study with machine learning (ML) consensus clustering approach.Setting and participantsAll very elderly (age ≥80 at time of transplant) kidney transplant recipients in the Organ Procurement and Transplantation Network/United Network for Organ Sharing database database from 2010 to 2019.Main outcome measuresDistinct clusters of very elderly kidney transplant recipients and their post-transplant outcomes including death-censored graft failure, overall mortality and acute allograft rejection among the assigned clusters.ResultsConsensus cluster analysis was performed in 419 very elderly kidney transplant and identified three distinct clusters that best represented the clinical characteristics of very elderly kidney transplant recipients. Recipients in cluster 1 received standard Kidney Donor Profile Index (KDPI) non-extended criteria donor (ECD) kidneys from deceased donors. Recipients in cluster 2 received kidneys from older, hypertensive ECD deceased donors with a KDPI score ≥85%. Kidneys for cluster 2 patients had longer cold ischaemia time and the highest use of machine perfusion. Recipients in clusters 1 and 2 were more likely to be on dialysis at the time of transplant (88.3%, 89.4%). Recipients in cluster 3 were more likely to be preemptive (39%) or had a dialysis duration less than 1 year (24%). These recipients received living donor kidney transplants. Cluster 3 had the most favourable post-transplant outcomes. Compared with cluster 3, cluster 1 had comparable survival but higher death-censored graft failure, while cluster 2 had lower patient survival, higher death-censored graft failure and more acute rejection.ConclusionsOur study used an unsupervised ML approach to cluster very elderly kidney transplant recipients into three clinically unique clusters with distinct post-transplant outcomes. These findings from an ML clustering approach provide additional understanding towards individualised medicine and opportunities to improve care for very elderly kidney transplant recipients.