在血浆中特异性检测肿瘤源性外泌体（tEVs）受到非肿瘤外泌体和非外泌体颗粒的干扰，因此精确识别tEVs并直接在临床血浆中对其表面蛋白进行分析仍是一个未解决的需求。在此，我们提出了一种动态免疫分析技术，用于单个tEV表面蛋白的分析（DISEP），这是一种基于表面等离子共振显微镜（SPRM）的动力学分析技术，用于特异性单个tEV表面蛋白的分析。DISEP采用一对低亲和力寡核苷酸探针来分别标记EV表面蛋白并功能化SPRM生物传感器界面。使用寡核苷酸标记的tEVs与血浆中的非特异性颗粒具有不同的结合动力学特性，从而可以准确计数单个EV的光学等离子计数。我们证明了DISEP在350倍背景血浆颗粒中识别目标EV的高敏感性（每微升4677个EV）。对于临床血浆样本进行分析，以区分胰腺癌患者（n = 40）和健康供体（n = 45）。使用一组生物标志物签名（EpCAM，HER2和GPC1），DISEP仅需要每个供体的10μL原始样本就可以将肿瘤患者进行分类，曲线下面积为0.98。DISEP为早期癌症诊断提供了一种高度特异性的EV检测和表面蛋白分析策略。

Specific detection of tumor-derived EVs (tEVs) in plasma is complicated by nontumor EVs and non-EV particles. To accurately identify tEVs and profile their surface protein expression at single tEV resolution directly with clinical plasma is still an unmet need. Here, we present a Dynamic Immunoassay for Single tEV surface protein Profiling (DISEP), a kinetic assay based on surface plasmon resonance microscopy (SPRM) for specific single tEV profiling. DISEP adopts a pair of low-affinity oligonucleotide probes to respectively label EV surface proteins and functionalize an SPRM biosensor interface. tEVs labeled with the oligonucleotide probes possess distinctive binding kinetics from nonspecific particles in plasma, which permits accurate digital plasmonic counting of single EVs. We demonstrate DISEP for recognizing target EVs among 350-fold background plasma particles with high sensitivity (4677 EVs per μL). Clinical plasma samples were analyzed to discriminate between pancreatic cancer patients (n = 40) and healthy donors (n = 45). With a panel of biomarker signatures (EpCAM, HER2, and GPC1), DISEP only requires 10 μL primary sample from each donor to classify tumor patients with an area under the curve of 0.98. DISEP provides a highly specific EV detection and surface protein profiling strategy for early cancer diagnosis.