基于无监督学习识别吲哚布芬治疗急性中重度缺血性卒中的应答患者

doi:10.3969/j.issn.1673-5765.2025.04.004

摘要/Abstract

摘要： 目的基于无监督学习识别在降低卒中复发风险方面吲哚布芬疗效不劣于阿司匹林的应答患者。
方法基于吲哚布芬对比阿司匹林治疗急性缺血性卒中（indobufen versus aspirin in acute ischemic stroke，INSURE）研究，通过使用多核学习降维和K-means聚类构建的无监督学习框架，识别吲哚布芬在治疗急性中重度缺血性卒中时主要结局不劣于阿司匹林的患者（即应答患者）；在识别出的患者中，评估吲哚布芬相较于阿司匹林治疗发生次要疗效结局（1年内新发卒中、3个月和1年内新发复合血管事件、3个月和1年内新发缺血性卒中、3个月和1年内功能结局不良）、安全性结局（3个月内中重度出血）和次要安全性结局（1年内中重度出血，3个月和1年内任何出血、死亡）的风险；比较应答患者和非应答患者的基线资料差异。
结果共识别出931例吲哚布芬应答患者（HR 0.70，95%CI 0.46～1.06，P非劣效=0.004）。在应答患者中，吲哚布芬组发生次要疗效结局的风险低于阿司匹林组（1年内新发卒中：HR 0.63，P=0.013；1年内新发缺血性卒中：HR 0.68，P=0.039；1年内新发复合血管事件：HR 0.64，P=0.014）；吲哚布芬组发生1年内任何出血的风险低于阿司匹林组（HR 0.41，P=0.025）。应答患者中的女性、颅内动脉粥样硬化性狭窄、高血压、糖尿病、血脂异常的比例显著高于非应答患者。
结论无监督学习可以用于识别在降低卒中复发风险方面吲哚布芬疗效不劣于阿司匹林的应答患者。应答患者中女性、颅内动脉粥样硬化性狭窄和代谢性疾病的比例高于非应答患者。

文章导读： 本研究基于无监督学习实现了对急性中重度缺血性卒中患者的疗效异质性分类，识别出吲哚布芬的潜在应答患者，强调了无监督学习在抗血小板治疗相关研究，尤其是临床试验数据结果为阴性时的应用价值。

关键词: 无监督学习; 卒中; 缺血性卒中; 替代抗血小板治疗

Abstract: Objective To identify responders with non-inferior efficacy of indobufen compared to aspirin in reducing the risk of stroke recurrence based on unsupervised learning method.
Methods Based on the indobufen versus aspirin in acute ischemic stroke (INSURE) study, an unsupervised learning framework constructed using multiple kernel learning for dimensionality reduction and K-means clustering was developed to identify patients whose primary outcomes of indobufen were non-inferior to aspirin in treating acute moderate-to-severe ischemic stroke (i.e., responders). In the identified patients, the risks of secondary efficacy outcomes (new stroke within 1 year, new complex vascular events within 3 months and 1 year, new ischemic stroke within 3 months and 1 year, and adverse functional outcomes within 3 months and 1 year), safety outcomes (moderate-to-severe bleeding within 3 months), and secondary safety outcomes (moderate-to-severe bleeding within 1 year, any bleeding, and death within 3 months and 1 year) were evaluated for indobufen compared to aspirin. Differences of baseline data between responders and non-responders were compared.
Results A total of 931 indobufen responders were identified (HR 0.70, 95%CI 0.46-1.06, Pnon-inferiority=0.004). Among the responders, the risks of secondary efficacy outcomes were lower in the indobufen group than in the aspirin group (new stroke within 1 year: HR 0.63, P=0.013; new ischemic stroke within 1 year: HR 0.68, P=0.039; new complex vascular events within 1 year: HR 0.64, P=0.014). The risk of any bleeding within 1 year was lower in the indobufen group than in the aspirin group (HR 0.41, P=0.025). The proportions of females, intracranial atherosclerotic stenosis, hypertension, diabetes mellitus, and dyslipidemia in the responders were significantly higher than in the non-responders.
Conclusions Unsupervised learning can be used to identify the responders with non-inferior efficacy of indobufen compared to aspirin in reducing the recurrence risk of stroke. The proportions of females, intracranial atherosclerotic stenosis, and metabolic disorders in responders were higher than those in non-responders.

Key words: Unsupervised learning; Stroke; Ischemic stroke; Alternative antiplatelet therapy

中图分类号:

蒲善宇, 潘岳松, 王拥军. 基于无监督学习识别吲哚布芬治疗急性中重度缺血性卒中的应答患者[J]. 中国卒中杂志, 2025, 20(4): 410-417.

PU Shanyu, PAN Yuesong, WANG Yongjun. Unsupervised Learning-Based Identification of Responders to Indobufen Treatment in Acute Moderate-to-Severe Ischemic Stroke[J]. Chinese Journal of Stroke, 2025, 20(4): 410-417.

参考文献

[1] MARZO A，CRESTANI S，FUMAGALLI I，et al. Endoscopic evaluation of the effects of indobufen and aspirin in healthy volunteers[J]. Am J Ther，2004，11（2）：98-102.
[2] PAN Y S，MENG X，YUAN B S，et al. Indobufen versus aspirin in patients with acute ischaemic stroke in China（INSURE）：a randomised，double-blind，double-dummy，active control，non-inferiority trial[J]. Lancet Neurol，2023，22（6）：485-493.
[3] FEIGIN V L，MARTINS S C，BRAININ M，et al. Twenty years on from the introduction of the high risk strategy for stroke and cardiovascular disease prevention：a systematic scoping review[J/OL]. Eur J Neurol，2024，31（3）：e16157[2025-01-16]. https://doi.org/10.1111/ene.16157.
[4] XU H W，MOHAMED M，FLANNERY M，et al. An unsupervised machine learning approach to evaluating the association of symptom clusters with adverse outcomes among older adults with advanced cancer：a secondary analysis of a randomized clinical trial[J/OL]. JAMA Netw Open，2023，6（3）：e234198[2025-01-16]. https://doi.org/10.1001/jamanetworkopen.2023.4198.
[5] CIKES M，SANCHEZ-MARTINEZ S，CLAGGETT B，et al. Machine learning-based phenogrouping in heart failure to identify responders to cardiac resynchronization therapy[J]. Eur J Heart Fail，2019，21（1）：74-85.
[6] LIN R，QIAN J S，HE H P，et al. Correlation between preoperative Doppler ultrasonography-assessed specific accessory cephalic vein diameter-cephalic vein diameter ratio（r）and early dysfunction of radial artery-cephalic vein arteriovenous fistula：a single-center cross-sectional study[J]. Quant Imaging Med Surg，2023，13（9）：5796-5802.
[7] JUNG S，BI Y T，DAVULURI R V. Evaluation of data discretization methods to derive platform independent isoform expression signatures for multi-class tumor subtyping[J/OL]. BMC Genomics，2015，16（Suppl 11）：S3[2025-01-16]. https://doi.org/10.1186/1471-2164-16-S11-S3.
[8] CABASSI A，KIRK P D W. Multiple kernel learning for integrative consensus clustering of omic datasets[J]. Bioinformatics，2020，36（18）：4789-4796.
[9] STEINLEY D. K-means clustering：a half-century synthesis[J]. Br J Math Stat Psychol，2006，59（Pt 1）：1-34.
[10] ELIGINI S，VIOLI F，BANFI C，et al. Indobufen inhibits tissue factor in human monocytes through a thromboxane-mediated mechanism[J]. Cardiovasc Res，2006，69（1）：218-226.
[11] ELIKOWSKI W，COFTA S，NOWICKI A，et al. Recurrent hemoptysis following thienopyridines and amiodarone administration. Therapeutic dilemma[J]. Pol Arch Med Wewn，2005，114（2）：773-778.
[12] GLÓWKA F K，STRZELECKA D，ZAPALSKI S. Steady-state pharmacokinetics of indobufen enantiomers in patients with obliterative atherosclerosis[J]. Chirality，2001，13（6）：308-312.
[13] PAN Y S，MENG X，CHEN W Q，et al. Indobufen versus aspirin in acute ischaemic stroke（INSURE）：rationale and design of a multicentre randomised trial[J]. Stroke Vasc Neurol，2022，7（5）：457-461.
[14] NTALIANIS E，SABOVČIK F，CAUWENBERGHS N，et al. Unsupervised time-series clustering of left atrial strain for cardiovascular risk assessment[J]. J Am Soc Echocardiogr，2023，36（7）：778-787.

[1]	张栗源, 刘涛, 姜勇, 李子孝, 王拥军, 杨晓萌. 基于脑影像及临床特征的机器学习模型预测缺血性卒中后心房颤动[J]. 中国卒中杂志, 2025, 20(4): 392-400.
[2]	程相鑫, 张烁, 杜松骏, 刘子阳, 周宏宇, 贾伟丽, 李子孝, 刘涛. 基于IMU信号的人工智能上肢多关节运动状态识别系统构建——卒中后人工智能运动功能评估与检测系统建设前导研究[J]. 中国卒中杂志, 2025, 20(4): 401-409.
[3]	蒋兰, 傅新民, 孙梦飞, 李怡萍. 大脑中动脉闭塞性急性缺血性卒中血管内治疗后不良预后的危险因素分析及预测模型开发[J]. 中国卒中杂志, 2025, 20(4): 418-427.
[4]	倪佃丽, 陈晓兵, 张广慧, 彭庆荣. 急性缺血性卒中院前延迟风险预测模型的构建及评价[J]. 中国卒中杂志, 2025, 20(4): 428-434.
[5]	张欣悦, 常红, 赵洁, 李佩佩, 刘梦娆, 李苏爱. 急性缺血性卒中患者运动想象训练中认知负荷与生理指标的相关性研究[J]. 中国卒中杂志, 2025, 20(4): 435-446.
[6]	张永庆, 李娜, 高伊丽, 秦佳文, 俞海萍, 赵婷婷. 1990—2021年中国归因于高LDL-C的卒中死亡负担变化趋势及预测分析[J]. 中国卒中杂志, 2025, 20(4): 447-456.
[7]	王岩, 郝怀宇, 陆强, 张蕾, 沈学延, 魏桂梅. 细胞外水分比率对急性缺血性卒中患者出院结局的影响[J]. 中国卒中杂志, 2025, 20(4): 457-461.
[8]	胡燕琴, 赵陶丽, 李申, 郭东兴, 赵志刚. 静脉应用银杏叶制剂治疗急性缺血性卒中有效性和安全性的meta分析[J]. 中国卒中杂志, 2025, 20(4): 462-469.
[9]	赵乾坤, 李宪东, 徐天策, 陈会生. 基于脑-肠轴探讨针灸治疗缺血性卒中机制的基础研究进展[J]. 中国卒中杂志, 2025, 20(4): 486-492.
[10]	池琦, 赵颖, 韩东倩, 张锶琪, 杜沛洁, 董琬玥, 徐安定, 杨振国, 孟珩. 急性缺血性卒中发病时间评估的研究进展[J]. 中国卒中杂志, 2025, 20(4): 493-499.
[11]	赵开杰, 孙秀亭, 袁曼, 胡婉贞, 张晓燕. 脑侧支循环评估方法的研究进展[J]. 中国卒中杂志, 2025, 20(4): 500-510.
[12]	郭园丽, 高任轲, 杨彩侠, 范文凤, 郭丽娜, 董小方, 吕培华, 高欢欢, 马珂珂. 急性缺血性卒中静脉溶栓意向量表的编制及信效度检验[J]. 中国卒中杂志, 2025, 20(4): 511-518.
[13]	樊燕琴, 曹黎明, 任力杰. 我国社区卒中防治：医防融合视角下的卒中防治现状、挑战与策略分析[J]. 中国卒中杂志, 2025, 20(3): 261-268.
[14]	刘昭, 陈敏霞, 申斗, 池枫, YU Monica, 刘跃华, 杨燕绥. 医防融合背景下卒中高危人群防治策略优化与体系建设[J]. 中国卒中杂志, 2025, 20(3): 270-276.
[15]	夏润禹, 曹黎明, 任力杰. 我国非专科医师对卒中的认识现状、挑战及对策[J]. 中国卒中杂志, 2025, 20(3): 277-282.