IgA肾病肾衰竭风险预测方程:中国患者多中心风险评估研究
该多中心回顾性队列研究纳入2155例中国IgA肾病患者,开发并验证了预测终末期肾病的方程。仅含年龄、性别、eGFR、血红蛋白和尿蛋白的临床模型在验证队列10年预测C统计量为0.86,联合牛津分型M和T评分的模型C统计量为0.83。
要点
- 临床五变量模型(年龄、性别、eGFR、血红蛋白、尿蛋白)可有效预测IgA肾病10年ESRD风险。
- 联合牛津分型M和T评分的模型预测效能与纯临床模型相似。
- 新模型优于既往已发表模型,但需在更多种族人群中验证后方可广泛用于临床。
核心结论
- 在中国IgA肾病人群中,利用临床和病理变量可预测肾衰竭风险。
- 新方程在验证队列中10年ESRD预测C统计量临床模型为0.86,联合病理模型为0.83。
研究设计与样本
- 多中心回顾性队列研究,纳入2155例中国IgA肾病患者。
- 发现队列934例,平均随访56.3个月;独立验证队列1221例,平均随访47.8个月。
主要发现
- 联合队列共发生212例ESRD事件。
- 最佳临床预测模型包含5个变量:年龄、性别、eGFR、血红蛋白浓度和尿蛋白排泄量。
- 最佳临床+病理模型包含2个临床变量(年龄和eGFR)和2个病理评分(牛津分型M和T评分)。
- 验证队列10年ESRD预测C统计量:临床模型0.86(95%CI 0.83-0.90),联合模型0.83(95%CI 0.77-0.89)。
- 连续净重分类指数和综合判别改善指数表明新模型优于既往发表模型。
关键数值与推荐
- 临床模型变量:年龄、性别、eGFR、血红蛋白浓度、尿蛋白排泄量。
- 联合模型变量:年龄、eGFR、牛津分型M评分、牛津分型T评分。
- 验证队列10年C统计量:临床模型0.86(95%CI 0.83-0.90),联合模型0.83(95%CI 0.77-0.89)。
临床怎么用
- 可基于年龄、性别、eGFR、血红蛋白和尿蛋白排泄量评估中国IgA肾病患者10年ESRD风险。
- 加入牛津分型M和T评分未显著提升预测效能,临床模型可优先考虑。
局限性
- 回顾性研究设计。
- 发现队列与验证队列疾病严重程度存在差异。
- 存在死亡竞争风险。
- 缺乏在种族多样人群中的验证。
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
BACKGROUND: The clinical course of immunoglobulin A (IgA) nephropathy (IgAN) is highly variable, making it difficult to predict which patients are at risk for rapid progression. The aim of this study was to develop and validate a kidney failure risk prediction equation for adults with IgAN. STUDY DESIGN: Multicenter retrospective cohort study of 2,155 Chinese patients with IgAN. CANDIDATE PREDICTORS: Clinical and histology variables. OUTCOMES: Time to end-stage renal disease (ESRD). ANALYTICAL APPROACH: The association of baseline predictors with the outcome was tested using cause-specific hazards models by treating death as a censoring event. RESULTS: The discovery cohort was composed of 934 IgAN cases with a mean follow-up of 56.3 months. The independent validation cohort was composed of 1,221 additional patients with a mean follow-up of 47.8 months. There were 212 ESRD events in the combined cohort. The best clinical predictive model of ESRD included 5 variables: age, sex, estimated glomerular filtration rate, hemoglobin concentration, and urine protein excretion. The best model combining clinical and histologic data included 2 clinical variables (age and estimated glomerular filtration rate) and 2 pathology scores (M and T scores from the Oxford classification). Both models predicted ESRD well at 10 years in the validation cohort (C statistics of 0.86 [95% CI, 0.83-0.90] and 0.83 [95% CI, 0.77-0.89], respectively). Continuous net reclassification index and integrated discrimination improvement indicated superior performance of the new models compared with previously published models. The performance of the new clinical model was similar to that of the new model that incorporated histologic variables. LIMITATIONS: Retrospective study design, differences in severity of disease between the discovery and validation cohorts, the competing risk of death, lack of validation in ethnically diverse patients. CONCLUSIONS: Kidney failure risk in the setting of IgAN is able to be predicted in a Chinese population using clinical and histologic variables. Additional evaluation of these equations needs to be implemented in more ethnically diverse patients before they can be applied to clinical practice broadly.
引用
- Xie J, Lv J, Wang W, et al. Kidney Failure Risk Prediction Equations in IgA Nephropathy: A Multicenter Risk Assessment Study in Chinese Patients. Am J Kidney Dis. 2018. PMID: 29555434. doi:10.1053/j.ajkd.2018.01.043
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。