Towards rapid, high-throughput and cost-effective evaluation of viral vector efficacy: Rapid image pattern analysis from microfluidic cell cultures us

慢病毒载体（LV）的生产特点是昂贵的开发，导致高达500，000美元的高治疗费用每名患者治疗。病毒载体生产过程中关键质量属性的监测（如载体的感染性）至关重要。目前用于病毒载体滴定的金标准依赖于感染性测定。它们允许通过感染细胞来估计样品中包含的感染性颗粒的数量。目前，这些测定是在微孔板中进行的，因此仅限于通过昂贵的过程进行终点监测。此外，对于高通量分析而言，液体体积仍然很大，并且自动化能力有限。为了促进个性化的基因治疗治疗和分散制造，在我们的实验室中开发了一种微流体装置，其以微小的体积操作并提供真实的时间测定数据。在该项目中，微流体装置将用于进行感染性测定，并将其与工业金标准进行基准测试。迄今为止开发的技术使用荧光方法来检测LV滴度，但可以实施新的模式。将开发人工智能算法，实现自动感染性测定。这将联合收割机微流体技术与先进的图像分析程序和机器学习算法相结合，以开发高通量感染性检测，为未来的工作奠定基础，该平台可以最佳地评估患者对病毒载体摄取的变异性。

Lentiviral Vector (LV) manufacture is characterized by expensive development, resulting in high treatment costs of up to $500,000 per patient treatment. Monitoring of critical quality attributes during viral vector manufacture, such as the infectivity of the vector, is vital. Current gold standards for viral vector titering rely on infectivity assays. They allow estimating the number of infectious particles contained in a sample through the infection of cells. Currently, these assays are carried out in micro well plates, thus limited to end-point monitoring through expensive processes. Furthermore, the liquid volumes are still large for high-throughput analysis, and automation capabilities are limited. To facilitate personalized gene therapy treatments and decentralizing manufacturing, a microfluidic device which operates with tiny volumes and which provides real time assay data was developed in our labs. In this project, the microfluidic device will be used to perform infectivity assays and bench-mark them against the industrial gold standard. The technology developed thus far uses a fluorescence approach to detect LV titer but new modalities can be implemented. Artificial Intelligence algorithms will be developed enabling automated infectivity assays. This will combine microfluidic technology with advanced image analysis routines and machine learning algorithms to allow the development of high throughput infectivity assays, laying the foundations for future work where this platform can optimally assess patient to patient variability of viral vector uptake.