Abstract: FR-PO024
Uncovering Subgroups of Diabetic Deceased Donor Kidney Transplant Recipients with Differing Outcomes Using Consensus Cluster Analysis
Session Information
- AI, Digital Health, Data Science - II
November 03, 2023 | Location: Exhibit Hall, Pennsylvania Convention Center
Abstract Time: 10:00 AM - 12:00 PM
Category: Augmented Intelligence, Digital Health, and Data Science
- 300 Augmented Intelligence, Digital Health, and Data Science
Authors
- Tangpanithandee, Supawit, Mayo Clinic Minnesota, Rochester, United States
- Thongprayoon, Charat, Mayo Clinic Minnesota, Rochester, Minnesota, United States
- Miao, Jing, Mayo Clinic Minnesota, Rochester, Minnesota, United States
- Jadlowiec, Caroline, Mayo Clinic, Phoenix, Arizona, United States
- Mao, Shennen, Mayo Clinic, Jacksonville, Florida, United States
- Mao, Michael A., Mayo Clinic, Jacksonville, Florida, United States
- Leeaphorn, Napat, Mayo Clinic, Jacksonville, Florida, United States
- Pattharanitima, Pattharawin, Thammasat University Faculty of Medicine, Khlong Nueng, Pathum Thani, Thailand
- Krisanapan, Pajaree, Mayo Clinic Minnesota, Rochester, Minnesota, United States
- Cheungpasitporn, Wisit, Mayo Clinic Minnesota, Rochester, Minnesota, United States
Background
Diabetic donor kidney transplants have inconsistent clinical outcomes, possibly due to population heterogeneity. This study aimed to use an unsupervised machine learning approach to cluster recipients of diabetic deceased donor kidney transplants and identify subgroups with higher risk of inferior outcomes and associated variables.
Methods
The study analyzed recipient-, donor-, and transplant-related characteristics of 7,876 recipients of diabetic deceased donor kidney transplants from 2010 to 2019 in the OPTN/UNOS database. Consensus cluster analysis was performed to identify important characteristics of each assigned cluster and compare posttransplant outcomes between the clusters.
Results
The analysis identified three clinically distinct clusters. Cluster 1 (N=2,903) recipients were characterized by the oldest age (64±8 years) and highest rate of comorbid diabetes mellitus (55%). They were more likely to receive kidney allografts from older donors (58±6.3 years) with hypertension (89%), meeting ECD status (78%), having a high rate of cerebrovascular death (63%) and carrying a high-KDPI (KDPI ≥85%) (77%). Cluster 2 (N=687) recipients were younger (49±13 years) and all were re-transplant patients with higher PRA (88 [IQR 46, 98]), receiving kidneys from younger (44±11 years), non-ECD deceased donors (88%) with low number of HLA mismatch (4 [IQR 2, 5]). The cluster 3 cohort was characterized by first-time kidney transplant recipients (100%) who received kidney allografts from younger (42±11 years), non-ECD deceased donors (98%). Compared to cluster 3, cluster 1 had a higher incidence of primary non-function, delayed graft function, patient death, and death-censored graft failure, while cluster 2 had a higher incidence of delayed graft function and death-censored graft failure but comparable primary non-function and patient death.
Conclusion
An unsupervised machine learning approach identified three clinically distinct clusters of diabetic donor kidney transplant patients with differing outcomes. The study suggests opportunities to improve utilization of high KDPI kidneys coming from diabetic donors in recipients with survival-limiting comorbidities, such as those observed in cluster 1.