Institute of Clinical Epidemiology and Biometry (F.A.M., K.U., P.U.H.), University of Würzburg, Germany.
Interdisciplinary Center for Clinical Research (F.A.M.), Würzburg University Hospital, Germany.
Stroke. 2021 Jan;52(2):543-551. doi: 10.1161/STROKEAHA.120.029232. Epub 2021 Jan 12.
Outcome prognostication in ischemic stroke patients remains challenging due to limited predictive properties of existing models. Blood-based biomarkers might provide additional information to established prognostic factors. We intended to identify the most promising prognostic biomarkers in ischemic stroke, their incremental prognostic value, and whether their predictive value differs among etiologies.
We searched MEDLINE (Ovid) and Institute for Scientific Information Web of Knowledge for articles reporting the predictive performance of blood-based biomarkers measured up to 7 days after ischemic stroke and reporting functional outcome or death at least 7 days after stroke. This work updates a previous systematic review (up to January 2007), follows the Preferred Reporting Items for Systematic Reviews and Meta-Analyses statement and was registered (International Prospective Register of Systematic Reviews PROSPERO 2018; https://www.crd.york.ac.uk/PROSPERO/; Unique identifier: CRD42018094671).
Two hundred ninety-one articles published between January 2007 and August 2018 comprising 257 different biomarkers met inclusion criteria. Median sample size was 232 (interquartile range, 110-455); 260 (89%) articles reported regression analyses with 78% adjusting for stroke severity, 82% for age, 67% for both, and 9% for none of them; 37% investigated discrimination, 5% calibration, and 11% reclassification. Including publications from a previous systematic review (1960-January 2007), natriuretic peptides, copeptin, procalcitonin, mannose-binding lectin, adipocyte fatty acid-binding protein, and cortisol were the biomarkers most consistently associated with poor outcome in higher-quality studies showing an incremental value over established prognostic factors. Other biomarkers were less consistently associated with poor outcome or were reported in lower quality studies. High heterogeneity among studies precluded the performance of a meta-analysis.
The number of reports on prognostic blood-based biomarkers in ischemic stroke increased 3.5-fold in the period January 2007 to August 2018. Although sample size increased, methodological flaws are still common. Natriuretic peptides and markers of inflammation, atherogenesis, and stress response are the most promising prognostic biomarkers among identified studies.
由于现有模型的预测性能有限，缺血性脑卒中患者的预后预测仍然具有挑战性。基于血液的生物标志物可能提供额外的信息，以补充现有的预后因素。我们旨在确定缺血性脑卒中最有前途的预后生物标志物，评估其增量预后价值，并探讨其预测价值是否因病因不同而有所差异。
我们在 MEDLINE（Ovid）和 Institute for Scientific Information Web of Knowledge 中检索了截至 2007 年 1 月的文献，这些文献报告了缺血性脑卒中后 7 天内测量的基于血液的生物标志物的预测性能，并报告了脑卒中后至少 7 天的功能结局或死亡。本研究更新了一项先前的系统评价（截至 2007 年 1 月），遵循系统评价和荟萃分析的首选报告项目声明，并在国际前瞻性注册系统评价注册处（PROSPERO 2018 年；https://www.crd.york.ac.uk/PROSPERO/；唯一标识符：CRD42018094671）进行了注册。
2007 年 1 月至 2018 年 8 月期间发表的 291 篇文章包含了 257 种不同的生物标志物，符合纳入标准。中位数样本量为 232（四分位距，110-455）；260 篇（89%）文章报道了回归分析，其中 78%的文章调整了卒中严重程度，82%的文章调整了年龄，67%的文章同时调整了卒中严重程度和年龄，9%的文章未调整任何因素；37%的文章研究了区分度，5%的文章研究了校准度，11%的文章研究了再分类。包括之前系统评价的出版物（1960 年至 2007 年 1 月），利钠肽、 copeptin、降钙素原、甘露糖结合凝集素、脂肪细胞脂肪酸结合蛋白和皮质醇是在显示出优于现有预后因素的增量预后价值的高质量研究中与不良结局最相关的生物标志物。其他生物标志物与不良结局的相关性较差，或者仅在质量较低的研究中报道。研究之间的高度异质性使得无法进行荟萃分析。
2007 年 1 月至 2018 年 8 月期间，缺血性脑卒中预后相关的基于血液的生物标志物的报道数量增加了 3.5 倍。尽管样本量有所增加，但仍存在常见的方法学缺陷。在已确定的研究中，利钠肽和炎症、动脉粥样硬化形成和应激反应标志物是最有前途的预后生物标志物。
