文献速递肾移植

遗体供肾质量公平供者评估模型(EDAM)的构建与验证

Equitable Donor Assessment Model of Deceased Donor Kidney Quality.
PubMed 文献 · Kidney Int Rep · 2026-01-21

研究基于UNOS 2010-2020年122,646例遗体肾移植数据,用Fine-Gray竞争风险模型构建仅反映供者内在移植物失败风险的EDAM评分。EDAM区分度良好(C-index=0.69)且校准优秀,可将近半数中高KDPI(≥0.20)肾脏重分类为低风险(<0.80),其10年移植物存活与常规KDPI<0.20器官相同。

要点

  • EDAM仅基于供者因素评估移植物失败风险,不受受者生存影响。
  • EDAM将近半数中高KDPI肾脏重分类为低风险,10年存活等同低KDPI。
  • EDAM区分度C-index=0.69,跨美国区域校准良好。

核心结论

  • EDAM提供以供者为中心的公平肾质量评估框架,可在不损害结局前提下安全扩大低风险供肾池。

研究设计与样本

  • 基于UNOS 2010-2020年数据(N=122,646),采用Fine-Gray回归建模死亡删失的移植物失败,并以死亡为竞争事件。
  • 供者系数校正受者与移植协变量(含HLA错配和缺血时间),推导供者特异性亚风险比(SHR)构成EDAM评分。
  • 性能用Harrell C指数及跨5个地理对齐美国超级区域的内部-外部交叉验证评估。

主要发现

  • 供者年龄较高、糖尿病、高血压、卒中为死因、蛋白尿、CMV血清阳性及肌酐升高独立增加移植物失败风险。
  • EDAM区分度稳健(C-index=0.69),跨区域校准优秀(合并斜率=1.02;截距=-0.002)。
  • 按5个数据驱动类别分层,移植物失败累积发生率呈梯度上升(Gray's P < 0.05)。
  • 近半数被常规KDPI判为中高(≥0.20)的肾脏被EDAM重分类为低风险(<0.80),其10年移植物存活与常规KDPI<0.20器官相同。

关键数值与推荐

  • EDAM低风险阈值为<0.80,常规KDPI中高阈值为≥0.20。
  • EDAM区分度C-index=0.69,校准合并斜率=1.02、截距=-0.002。
  • 5个EDAM类别间移植物失败累积发生率梯度上升,Gray's P < 0.05。

临床怎么用

  • 可考虑用EDAM对常规KDPI中高(≥0.20)的供肾重新评估,识别可安全用于低风险受者的器官。
  • EDAM评分聚焦供者内在质量,有助于减少受者因素混杂带来的分配不公平。

局限性

  • 分类阈值在美国体系内数据驱动得出,需在国际队列中外部验证。

以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。

原文

INTRODUCTION: The Kidney Donor Profile Index (KDPI) guides organ allocation but blends donor and recipient influences, potentially misclassifying organ quality and contributing to inequity. We developed the Equitable Donor Assessment Model (EDAM), a donor-focused index that isolates intrinsic graft-failure risk independent of recipient survival. METHODS: Using the United Network for Organ Sharing (UNOS) data (2010-2020; N = 122,646), we modeled death-censored graft failure with death as a competing event using Fine-Gray regression. Donor coefficients were adjusted for recipient and transplant covariates-including human leukocyte antigen (HLA) mismatch and ischemia time-to derive donor-specific subhazard ratios (SHRs) for the EDAM score. Performance was evaluated using Harrell's C-index and internal-external cross-validation across 5 geography-aligned US super-regions. RESULTS: Higher donor age, diabetes, hypertension, stroke as cause of death, proteinuria, cytomegalovirus (CMV) seropositivity, and elevated creatinine independently increased graft-failure risk. EDAM demonstrated robust discrimination (C-index = 0.69) and excellent calibration across US regions (pooled slope = 1.02; intercept = -0.002). Stratification into 5 data-driven categories showed a graded rise in cumulative incidence of graft failure (Gray's P < 0.05). Nearly half of kidneys classified as moderate-to-high KDPI (≥ 0.20) were reclassified as low-risk by EDAM (< 0.80) and achieved identical 10-year graft survival to conventional KDPI < 0.20 organs. CONCLUSION: EDAM provides an equitable, donor-centric framework for assessing kidney quality. Although these categorical thresholds were derived in a data-driven manner within the US system and require external validation in international cohorts, EDAM's ability to safely expand the low-risk pool without compromising outcomes suggests it could significantly refine allocation policy and enhance fairness in kidney transplantation.

来源:Kidney Int Rep,查看原文。版权归原作者所有。

引用

  1. Ali H, Daoud A, Mohammed M, et al. Equitable Donor Assessment Model of Deceased Donor Kidney Quality. Kidney Int Rep. 2026. PMID: 41777292. doi:10.1016/j.ekir.2026.103790

供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。