Abstract: SA-PO835
Distinct Characteristics of High Sensitized Kidney Transplant Recipients in the United States by Machine Learning Consensus Clustering
Session Information
- Transplantation: Clinical - Pretransplant Assessment and Living Donors
November 05, 2022 | Location: Exhibit Hall, Orange County Convention Center‚ West Building
Abstract Time: 10:00 AM - 12:00 PM
Category: Transplantation
- 2002 Transplantation: Clinical
Authors
- Thongprayoon, Charat, Mayo Clinic Minnesota, Rochester, Minnesota, United States
- Mao, Michael A., Mayo Clinic in Florida, Jacksonville, Florida, United States
- Mao, Shennen, Mayo Clinic in Florida, Jacksonville, Florida, United States
- Jadlowiec, Caroline, Mayo Clinic Arizona, Scottsdale, Arizona, United States
- Acharya, Prakrati C., Texas Tech University Health Sciences Center El Paso, El Paso, Texas, United States
- Leeaphorn, Napat, Saint Luke's Health System, Kansas City, Missouri, United States
- Kaewput, Wisit, Phramongkutklao College of Medicine, Bangkok, Thailand
- Pattharanitima, Pattharawin, Thammasat University Hospital, Khlong Nueng, Pathum Thani, Thailand
- Cheungpasitporn, Wisit, Mayo Clinic Minnesota, Rochester, Minnesota, United States
Background
Our study aimed to characterize highly sensitized kidney transplant patients using unsupervised machine learning approach.
Methods
We used the OPTN/UNOS database from 2010 to 2019 to perform consensus cluster analysis based on recipient-, donor-, and transplant-related characteristics in 7,458 kidney transplant patients with pre-transplant panel reactive antibody (PRA) ≥98%. We identified each cluster’s key characteristics using the standardized mean difference of >0.3. We compared the posttransplant outcomes among the assigned clusters
Results
Consensus cluster analysis identified two clinically distinct clusters of highly sensitized kidney transplant patients. Cluster 1 patients were older (mean age 45 vs 54 years), more male (59% vs. 9%), had more kidney retransplant (98 vs. 3%), but less diabetic kidney disease (3% vs 29%), compared to cluster 2. While patient survival was comparable between two clusters, cluster 1 had lower death-censored graft survival but higher acute rejection compared to cluster 2.
Conclusion
Unsupervised machine learning approach characterized highly sensitized kidney transplant patients into 2 clinically distinct clusters based on age, sex, kidney retransplant status, and diabetes, with differing posttransplant outcomes.