Kidney Week

Abstract: FR-OR088

Development and Internal Validation of a Risk Calculator for Postdonation eGFR Decline in Living Kidney Donors in Japan

Session Information

• The Changing Landscape of Kidney Transplantation: Clinical Innovation, Outcomes, and Health Policy
  October 23, 2026 | Location: Mile High Ballroom 4A, Convention Center
  Abstract Time: 05:20 PM - 05:30 PM

Category: Transplantation

• 2002 Transplantation: Clinical

Authors

• Hoshino, Akiko, Gunma Daigaku Daigakuin Igakukei Kenkyuka Igakubu, Maebashi, Japan

Akiko Hoshino, MD, MPH

• Ikeuchi, Hidekazu, Gunma Daigaku Daigakuin Igakukei Kenkyuka Igakubu, Maebashi, Gunma Prefecture, Japan

Hidekazu Ikeuchi, MD, PhD

• Hiromura, Keiju, Gunma Daigaku Daigakuin Igakukei Kenkyuka Igakubu, Maebashi, Gunma Prefecture, Japan

Keiju Hiromura, MD, PhD

Background

Living donor kidney transplantation accounts for a higher proportion of kidney transplants in Japan than in other countries. Although donor safety is paramount and the incidence of end-stage renal disease among living donors is low (~0.05%), long-term renal risk stratification remains limited, particularly in Asians. Using a nationwide Japanese living donor registry, we previously identified higher body mass index (BMI), hypertension, and lower predonation estimated glomerular filtration rate (eGFR) as independent predictors of postdonation kidney function decline. We aimed to develop a web-based calculator to estimate individualized risk of postdonation eGFR decline.

Methods

We developed and internally validated a prediction model using data from 13,095 living kidney donors in the Japan Society for Transplantation registry between 2009 and 2024 with predonation eGFR ≥70 mL/min/1.73 m². The endpoint was postdonation eGFR <45 mL/min/1.73 m². We incorporated the multivariable Cox proportional hazards model previously used to identify predonation risk factors into a web calculator to estimate individualized 5-year absolute risk and visualize survival curves. We assessed performance using bootstrap validation, including optimism-corrected concordance index (C-index) and bootstrap-corrected calibration slope. We additionally stratified donors by predicted 3-year risk to assess the model's ability to identify clinically distinct groups and performed sensitivity analysis excluding events within 3 months after donation to evaluate robustness beyond the early postoperative period.

Results

Among 13,095 donors, 4,643 (35.5%) were men. Mean age was 61.4±11.2 years, predonation eGFR was 102.2±10.1 mL/min/1.73 m², and median follow-up was 36 months. During follow-up, 523 donors (4.0%) developed postdonation eGFR <45 mL/min/1.73 m². The model showed good discrimination (C-index 0.865) and calibration (slope 0.990). Observed 3-year risks in the low-, intermediate-, and high-risk groups were 0.3%, 1.0%, and 8.6%, respectively. Similar performance was observed after excluding events within 3 months after donation (C-index 0.862; slope 0.986).

Conclusion

We developed and internally validated a web calculator to estimate individualized risk after donation. The model demonstrated good calibration and stable performance, supporting individualized donor counseling and postdonation follow-up.