视网膜标志物在血管性认知障碍中的研究进展

Advancements in Research on Retinal Biomarkers in Vascular Cognitive Impairment

摘要: 血管性认知障碍(Vascular Cognitive Impairment, VCI)是由脑血管病变及其危险因素引发的认知功能受损，其早期诊断主要由传统神经影像学及认知量表的侵入性、高成本和敏感性不足而受限。近年来，视觉器官作为中枢神经系统的“可视窗口”，因其与中枢神经系统的胚胎同源性及病理机制相似性，视网膜指标成为非侵入性生物标志物研究的热点。光学相干断层扫描血管成像(Optical Coherence Tomography Angiography, OCTA)技术的突破，结合人工智能算法，实现了视网膜微血管参数的精准量化，为VCI的早期筛查、亚型鉴别及疗效评估提供了新工具。研究显示，VCI患者的视网膜微血管特征与大脑病变密切相关：深层毛细血管丛(Deep Retinal Capillary Plexus, DCP)密度降低与脑白质病变体积负相关，静脉迂曲度可能与脑白质病变有关。在临床应用中，多模态模型可显著提升诊断效能，并支持亚型鉴别。此外，视网膜参数还可动态监测卒中后认知衰退及药物疗效。尽管视网膜标志物在高血压、糖尿病等高风险人群的VCI预测中展现出巨大潜力，但要解决技术标准化、多中心验证及多模态数据整合，仍具有挑战性。未来，需开发基于深度学习的社区筛查系统、解析视网膜-脑轴的分子机制，以及推动超分辨OCTA和靶向治疗策略的临床转化。视网膜成像技术为VCI的精准诊疗提供了重要窗口，有望纳入国际诊疗标准并助力于个体化干预。

Abstract: Vascular Cognitive Impairment (VCI), caused by cerebrovascular pathologies and their risk factors, leads to cognitive decline. Early diagnosis remains challenging due to the invasiveness, high cost, and limited sensitivity of traditional neuroimaging and cognitive assessments. Recent advances highlight the retina—a “visible window” of the central nervous system owing to shared embryonic origins and pathological mechanisms—as a promising source of non-invasive biomarkers. Optical Coherence Tomography Angiography (OCTA), combined with AI algorithms, enables precise quantification of retinal microvascular parameters, offering novel tools for early VCI screening, subtype classification, and treatment monitoring. Key findings include reduced deep capillary plexus (DCP) density correlating with white matter lesion volume, and venous tortuosity predicting annual lesion progression. Multimodal models integrating retinal vascular metrics and demographic data enhance diagnostic accuracy and subtype differentiation. Retinal parameters also track post-stroke cognitive decline and drug responses (e.g., cholinesterase inhibitor efficacy). While retinal biomarkers show potential in high-risk populations (hypertension, diabetes), challenges persist in standardization, multicenter validation, and multimodal data integration. Future priorities include AI-driven community screening systems, molecular exploration of the retina-brain axis, and clinical translation of super-resolution OCTA and targeted therapies. Retinal imaging holds promise for refining VCI diagnostics and personalized interventions, potentially reshaping global clinical standards.