Benchmarking the Sensitivity and Specificity of the Herringbone Microfluidic Device for Rare Particle Capture

项目摘要 许多新技术都是在研究人员的工作台上开发和验证的，但很少有新技术能够在研究人员的工作台上开发和验证。 有机会用设盲的对照样品测试其性能。缺乏标准可能是由于 技术或处理条件的新奇，或者仅仅是患者体内潜在生物学的不确定性。 这对于液体活检测定特别具有挑战性，因为极度缺乏通用标准，进一步 减缓了将来向诊所的转化。通过我们与RADx-Rad计划的合作， 我们的R 01应用程序中的母技术验证测试：“人字形芯片”或EVHB芯片。 使用他们的盲样是强大的，既验证了我们的内部发现，同时也抓住了 监管机构和其他研究人员的注意。在这项研究中，我们将尽最大努力 复杂的验证小组，一个完整的挑战小组，以评估我们的检测的特异性和灵敏度，测试 超过100个盲样。这项工作将很快完成，因为我们有所有的工具和工作人员来测试这些 样品在一个月的时间。我们将与NIH DCC进行全面的数据协调，以确保一致性， 数据标准化和数据存储到dbGAP数据库。每个样本都有唯一的条形码， 内容将包括一组浓度的纳米级分析物，供我们测试，浓度从 是我们检测极限的1倍到4倍。污染分析物也被添加到这些样品中，以允许特异性 有待确定。我们将通过我们的设备运行每个样本，使用我们的自动系统提取RNA， 量化检测到的RNA转录物的数量和类型，并与DCC共享数据。它将会引起 自动数据揭盲、数据转换和分析，使我们能够快速评估设备 性能除了完成这一关键的设备验证工作外， 建立诊断设备开发的新标准，其中包括多方研发努力。

PROJECT SUMMARY Many new technologies are developed and validated on the bench of the investigator, but few have the opportunity to test their performance with blinded, control samples. The lack of standards can be due to the novelty of technology or processing conditions, or simply the uncertainly of the underlying biology within patients. This is particularly challenging for liquid biopsy assays, as universal standards are desperately lacking, further slowing down future translation to the clinic. Through our work with the RADx-Rad program, we have begun validation testing for the parent technology featured in our R01 application: the ‘herringbone chip’, or EVHB-Chip. The use of their blinded samples has been powerful, both validating our internal findings, while also catching the attention of the regulatory agencies and other researchers. For this proposed study, we will conduct our most complex validation panel yet, a full challenge panel to evaluate specificity and sensitivity of our assay, testing over 100 blinded samples. This work will be done quickly, as we have all the tools and staff in place to test these samples in one month’s time. We will have full data harmonization with the NIH DCC, to ensure consistency, data standardization, and data deposits to the dbGAP database. Each sample is uniquely barcoded and the contents will include a set concentration of nanoscale analytes for us to test with the concentration varying from 1x to 4x our limit of detection. Contaminating analytes are also added to these samples, to allow for specificity to be determined. We will run each sample through our device, extract the RNA using our automated system, quantify the amount of and types of RNA transcripts detected and share the data with the DCC. This will trigger an automatic data unblinding, data transformation and analysis, allowing us to quickly assess device performance. In addition to getting this critical device validation work completed, the resulting publication will establish a new standard for diagnostic device development, which includes multi-party R&D efforts.