Optimization and Validation of an indicator cell assay for blood-based diagnosis of lung cancer

项目摘要。指示细胞分析平台(ICAP)是一种廉价的基于血液的分析，可以 用于各种疾病的早期发现、疾病分期、预后和治疗反应 疾病的威胁。ICAP使用培养的标准化细胞作为生物传感器，利用细胞的能力 对血清中存在的疾病信号做出敏锐的反应，而不是传统的依赖于 直接检测血液中的分子。开发ICAP需要将培养的细胞暴露于来自 正常或患病的受试者，测量全局差异反应模式，并使用它来训练可靠的 基于少量基因表达的疾病分类器。部署ICAP涉及到测量 仅使用具有成本效益的工具表达分类器基因。我们已经演示了ICAP， 在肌萎缩侧索硬化症小鼠模型中对疾病的症状检测和早期检测 人类的阿尔茨海默氏症，我们目前正在验证这一点。 在这里，我们正在开发一种诊断肺癌(LC)的ICAP。LC的血液生物标志物是至关重要的 需要与现有成像工具结合使用，以提高诊断准确性。我们的目标是发展 一种ICAP，用于通过成像识别出不确定的肺结节(IPN)的患者 从数据中自信地将其归类为恶性或良性。为了临床实用，ICAP需要区分 良恶性结节1)高灵敏度和阴性预测值(NPV)最大限度地减少 检测结果为阴性的恶性肿瘤患者的数量，以及2)将提供 对正确诊断为良性结节的患者的经济影响和可操作结果。我们有 演示了LC ICAP的概念验证，并获得了96%的净现值、92%的敏感性和52%的特异性 鉴别非小细胞肺癌和良性结节(独立样本)。临床应用的潜力 实用性高，恶性肿瘤漏诊风险低(8%FNR)，52%的患者具有可操作的结果 良性结节，以及比市场上和正在开发的其他检测方法更好或相似的性能。 对于第二阶段，我们建议用来自独立站点的更大的队列来优化和验证检测 让我们把化验商业化作为一种临床试验。我们的目标是：1)优化实验、技术和 ICAP的计算参数；2)使用最优条件训练和测试改进的ICAP分类器， 3)用盲独立样本对该分类器进行验证。我们的目标是实现临床实用和更大的 性能优于开发中的竞争测试，≥95%的敏感性和60%的特异性，90%的净现值。 我们的最终目标是开发一种可以在患者通过成像发现IPN时提供给患者的测试。我们的 简单的基于血液的化验将使用连续变量给患者疾病的概率。我们将发展 患者和医生可以用来评估风险和决定治疗的数据的可视化描述。带着我们的 合作者，马塞斯博士，我们将努力改进最好的临床方法。

Project Summary. The Indicator Cell Assay Platform (iCAP) is an inexpensive blood-based assay that can be used for early detection of disease, disease stage stratification, prognosis and response to therapy for a variety of diseases. The iCAP uses cultured, standardized cells as biosensors, capitalizing on the ability of cells to respond to disease signals present in serum with exquisite sensitivity, as opposed to traditional assays that rely on direct detection of molecules in blood. Developing the iCAP involves exposing cultured cells to serum from normal or diseased subjects, measuring a global differential response pattern, and using it to train a reliable disease classifier based on the expression of a small number of genes. Deploying the iCAP involves measuring only expression of classifier genes using cost-effective tools. We have demonstrated the iCAP by pre- symptomatic detection of disease in an amyotrophic lateral sclerosis mouse model, and early detection of Alzheimer's disease in humans, which we are currently validating. Here, we are developing an iCAP for diagnosis of lung cancer (LC). Blood biomarkers of LC are critically needed for use in combination with existing imaging tools to improve diagnostic accuracy. Our goal is to develop an iCAP for use on patients who have indeterminate pulmonary nodules (IPNs) identified by imaging that cannot be confidently classified as malignant or benign from the data. For clinical utility, the iCAP needs to distinguish malignant from benign nodules with 1) High sensitivity and negative predictive value (NPV) to minimize the number of patients with malignant tumors that have negative test results, and 2) A specificity that will provide economic impact and actionable results for patients correctly identified with benign nodules. We have demonstrated proof of concept for the LC iCAP and achieved 96% NPV, 92% sensitivity and 52% specificity in distinguishing non-small cell lung cancer from benign nodules (with independent samples). Potential for clinical utility is high with low risk of missing malignant tumors (8% FNR), actionable results for 52% of patients with benign nodules, and performance that is better or similar to other assays on the market and in development. For Phase II, we propose to optimize and validate the assay with larger cohorts from independent sites to position us to commercialize the assay as a clinical test. We aim to: 1) Optimize experimental, technical, and computational parameters of the iCAP, 2) Train and test an improved iCAP classifier using optimal conditions, and 3) Validate the classifier with blind independent samples. Our goal is to achieve clinical utility and greater performance than competing tests in development with ≥95% sensitivity and >60% specificity, with >90% NPV. Our ultimate goal is to develop a test that can be offered to patients at the time of finding an IPN by imaging. Our simple blood-based assay will give patients a probability of disease using a continuous variable. We will develop a visual depiction of the data that patients and doctors can use to assess risk and decide treatment. With our collaborator, Dr. Massion, we will work to refine the best clinical approach.