Proteomics of human vitreous to investigate mechanisms underlying the variability in anti-VEGF treatment response in neovascular AMD patients

项目摘要 视网膜相关性黄斑变性（AMD）是老年人严重的、不可逆的视力丧失的主要原因。 全世界的成年人大多数AMD患者的视力丧失是由该疾病的晚期新生血管形式引起的 （NVAMD）。玻璃体内注射抗血管内皮生长因子（抗VEGF）可有效保护 许多患者的视力，但有显着的变异性NVAMD患者之间的治疗反应。 导致抗VEGF应答变异性的分子因素代表了知识上的关键差距。 我们的提案通过利用独特而强大的资源直接解决了这一知识差距。我们 拥有从1,100多名患者的办公室收集的去识别玻璃体样本的独家储存库 治疗视网膜疾病，包括许多NVAMD。我们将把这个资源与一个高度敏感的， 基于高通量多重免疫测定蛋白质组学技术鉴定玻璃体相关蛋白 在NVAMD中有抗VEGF治疗反应。我们假设玻璃体炎症水平的变化 和血管生成蛋白影响的基本机制，基础的变化，在回应抗- VEGF治疗。为了验证这一假设，我们将测量参与炎症的蛋白质水平， 在NVAMD患者的玻璃体样品中的血管生成，在NVAMD患者的治疗之前和整个过程中， 治疗我们确定了83名接受标准三种治疗的NVAMD初治患者， 每月一次玻璃体内贝伐珠单抗负荷剂量，然后根据视觉需要每月注射一次 视力、眼底检查和OCT评估。使用视力和OCT测量，每个患者的 主要（三次负荷剂量后1个月;第3个月）和次要（治疗开始后6个月; 第6个月）抗VEGF治疗应答被分类为良好、部分、差或无应答。 在目标1中，我们将使用Olink Proteomics的多重免疫测定来测量733种相关蛋白质的水平 在这些NVAMD患者的初始治疗之前收集的玻璃体样品中， 贝伐单抗注射液。为了比较良好+部分应答者和不良+无应答者，我们将 对应答者状态（良好/部分应答或差/无应答）与基线水平进行logistic回归， 在主要和次要响应时间点，对每种蛋白进行和不进行协变量校正。 在目标2中，我们将确定贝伐珠单抗诱导的玻璃体中的 炎症和血管生成蛋白与NVAMD患者的临床结果相关。其中58名 对于来自目标1的患者，我们在第1个月（治疗后一个月）收集了额外的玻璃体样本 启动）、第3个月和第6个月。我们将量化相同的733炎症和血管生成蛋白的水平， 在这些纵向玻璃体样本中。我们将比较蛋白质的倍数变化和蛋白质的变化趋势 在每个缓解时间点，良好/部分缓解者和不良/无缓解者之间随时间的变化。

PROJECT SUMMARY Age-related macular degeneration (AMD) is the leading cause of severe, irreversible vision loss in older adults worldwide. Most vision loss in AMD is caused by the advanced neovascular form of the disease (NVAMD). Intravitreal anti-vascular endothelial growth factor (anti-VEGF) injections are effective in preserving vision for many patients, but there is marked variability among NVAMD patients in treatment response. Molecular factors contributing to this variability in anti-VEGF response represent a critical gap in knowledge. Our proposal addresses this knowledge gap directly by leveraging a unique and powerful resource. We have an exclusive repository of deidentified vitreous samples collected in-office from more than 1,100 patients treated for retinal diseases, including many with NVAMD. We will couple this resource with a highly sensitive, high-throughput multiplex immunoassay-based proteomics technology to identify vitreous proteins associated with anti-VEGF treatment response in NVAMD. We hypothesize that variation in vitreous levels of inflammatory and angiogenic proteins impacts the fundamental mechanisms that underlie the variability in response to anti- VEGF treatment. To test this hypothesis, we will measure levels of proteins involved in inflammation and angiogenesis in vitreous samples collected from NVAMD patients both prior to and throughout the course of treatment. We have identified a cohort of 83 treatment-naïve NVAMD patients who received the standard three monthly loading doses of intravitreal bevacizumab, followed by monthly injections as needed based on visual acuity, fundus examination, and OCT assessment. Using visual acuity and OCT measurements, each patient’s primary (one month after three loading doses; Month 3) and secondary (6 months after treatment initiation; Month 6) anti-VEGF treatment responses were classified as Good, Partial, Poor, or Non-Response. In Aim 1, we will use Olink Proteomics’ multiplex immunoassays to measure levels of 733 proteins involved in inflammation and angiogenesis in vitreous samples collected from these NVAMD patients prior to their initial bevacizumab injection. To compare between Good+Partial Responders and Poor+Non-Responders, we will perform logistic regression of responder status (Good/Partial or Poor/Non-Response) against baseline levels of each protein with and without adjustment for covariates at both primary and secondary response timepoints. In Aim 2, we will determine the longitudinal bevacizumab-induced changes in vitreous levels of inflammatory and angiogenic proteins that correlate with clinical outcomes in NVAMD patients. For 58 of the patients from Aim 1, we have additional vitreous samples collected at Month 1 (one month after treatment initiation), Month 3, and Month 6. We will quantify levels of the same 733 inflammatory and angiogenic proteins in these longitudinal vitreous samples. We will compare the protein fold-changes and the trend of protein changes over time between Good/Partial Responders and Poor/Non-Responders at each response timepoint.