文献速递临床生化临床免疫分子诊断

IgA肾病多组学体液生物标志物的分析验证、标准化与临床化学转化

Multi-omics body-fluid biomarkers in immunoglobulin a nephropathy: Analytical Validation, standardization, and clinical chemistry translation of diagnostic and prognostic Analytes.
PubMed 文献 · Clin Chim Acta · 2026-06-15

本文综述了反映IgA肾病多打击发病机制的多组学血清和尿液生物标志物,涵盖半乳糖缺陷型IgA1的多种检测方法及基于质谱和适配体的多肽分类器。机器学习分类器可整合这些分析物以辅助个体化诊断、预后和治疗监测,但临床评估受分析前变异、实验室间标准化不完整、缺乏参考物质、阈值模糊及敏感性和特异性不一致等因素限制。传统尿液标志物如白蛋白尿、蛋白尿、血尿和eGFR可能需要与多组学标志物联合以改善预后分类。

要点

  • 多组学血清和尿液标志物可反映IgA肾病多打击发病机制,补充肾活检。
  • 半乳糖缺陷型IgA1可用凝集素ELISA、KM55抗体、质谱糖谱分析和NMR等方法检测。
  • 机器学习分类器整合多组学分析物有助于个体化诊断、预后和治疗监测。

核心结论

  • 多组学体液标志物有潜力补充肾活检用于IgA肾病的诊断和预后,但需标准化IFCC可溯源检测和统一系列检测算法。
  • 传统尿液标志物如白蛋白尿、蛋白尿、血尿和eGFR可能需要与多组学标志物联合以改善预后分类。

研究设计与样本

  • 本文为综述,评估反映IgA肾病多打击发病机制的血清和尿液多组学标志物。

主要发现

  • 半乳糖缺陷型IgA1可通过凝集素ELISA、KM55单克隆抗体、位点分辨质谱糖谱分析、NMR及LC-MS代谢和脂质指纹技术检测。
  • 多肽分类器可通过SWATH、DIA和适配体蛋白质组学技术生成。
  • 机器学习分类器如人工神经网络、XGBoost和随机生存森林可整合分析物用于个体化诊断、预后和治疗监测。

临床怎么用

  • 在常规实验室使用中,多组学标志物应使用标准化IFCC可溯源检测并统一系列检测算法,才能有意义地补充肾活检。

局限性

  • 临床评估受分析前变异、实验室间标准化不完整、缺乏参考物质、生物标志物阈值模糊及敏感性和特异性不一致等因素限制。
  • 需要在不同种族群体以及儿童和成人人群中进一步验证。

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

IgA nephropathy (IgAN) is the most prevalent primary glomerulonephritis worldwide. Although optimized supportive therapy is administered to these patients, a substantial proportion still progresses to end-stage kidney disease. The advancement of these noninvasive laboratory tools will complement renal biopsies. This review focuses on evaluating the multi-omics serum and urinary biomarkers that reflect the multi-hit pathogenesis of IgAN, such as abnormal O-glycosylation of immunoglobulin A1 (IgA1), autoantibody production, deposition of immune complexes in the mesangium, complement system activation, podocyte dysfunction, and tubulointerstitial fibrosis. Galactose-deficient IgA1 can be measured using various methodologies, including lectin-based enzyme-linked immunosorbent assay (ELISA), KM55 monoclonal antibody assays, site-resolved mass spectrometric glycoprofiling, nuclear magnetic resonance (NMR), and liquid chromatography-mass spectrometry (LC-MS)-based metabolite and lipid fingerprinting techniques. Additionally, peptide classifiers are generated using sequential window acquisition of all theoretical spectra (SWATH), data-independent acquisition (DIA), and aptamer-based proteomics techniques. Machine learning classifiers, such as artificial neural networks, XGBoost, and random survival forests, enable the integration of these analytes, facilitating individualized diagnosis, prognosis, and therapy monitoring within the framework of guideline-directed therapy. However, the clinical evaluation of these biomarkers is constrained by several factors, including preanalytical variation, incomplete interlaboratory standardization, absence of reference materials, ambiguous biomarker thresholds, and inconsistent sensitivity and specificity. Additionally, further validation is required across different ethnic groups and in both pediatric and adult populations. Therefore, traditional markers of urinary involvement such as albuminuria, proteinuria, hematuria, and estimated glomerular filtration rate (eGFR) may need to be combined with multi-omics markers for better prognostic classification. These biomarkers can only meaningfully complement renal biopsy in routine laboratory use if standardized International Federation of Clinical Chemistry (IFCC)-traceable assays are used and serial testing algorithms are harmonized.

来源:Clin Chim Acta,查看原文。版权归原作者所有。

引用

  1. Rathore V, Asad M, Saxena S, et al. Multi-omics body-fluid biomarkers in immunoglobulin a nephropathy: Analytical Validation, standardization, and clinical chemistry translation of diagnostic and prognostic Analytes. Clin Chim Acta. 2026. PMID: 42297320. doi:10.1016/j.cca.2026.121183

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。