Design of formulation processes for improved thermostability of viral vaccines

病毒载体疫苗是目前在COVID-19第三期临床试验中评估的最先进的候选疫苗之一。在制造之后，热稳定性是确保这些救生产品在需要时的有效性和安全交付的关键质量属性。基于对细胞培养产生的病毒的广泛的科学知识和技术诀窍，我们建议为这项研究计划指导我们的努力，以研究和整合过程开发周期，病毒疫苗的液体和固体配方，这一过程开发领域仍然主要是经验主义的。因此，一种新型的混合模型将使冷冻干燥机能够实时优化运行。该研究计划的长期目标是确定关键参数并揭示其作用机制，以整合专家知识，以数字辅助设计有效病毒疫苗的液体和固体配方，提高热稳定性和成本效益。我们的目标是1)设计增强的液体配方，并揭示包膜病毒（如水疱性口炎病毒（VSV）和慢病毒）在赋形剂存在下的配方机制，以保持其生物活性；2)设计包膜病毒和非包膜病毒（如VSV和腺病毒载体疫苗）冷冻干燥的固体配方和机制模型，为机器学习提供大量高质量数据；3)结合机械模型和数据驱动模型，采用数字孪生方法建立冻干过程的混合模型。数字孪生体通过传感器和反馈控制与物理冷冻干燥机连接，加速病毒疫苗冷冻干燥过程的设计和优化运行。与该领域的专家（来自麦吉尔大学的共同主管）合作，我们建议结合机器学习（ML）算法的最新进展，以协助设计实验并加速设计稳健的疫苗配方和流程。实验、结果验证和基于数字孪生的冷冻干燥过程优化将在我们已经运行的生物安全2+级实验室中进行，并由我们经过验证的过程分析技术提供支持。这项研究将促进对包膜和非包膜病毒配方的了解，从而发现新的赋形剂和增强的过程。通过将设计方法的重点放在提高VSV载体埃博拉疫苗的热稳定性上，我们将促进这种拯救生命的产品以及使用相同VSV抗原递送平台（HIV-1、SARS-CoV-2）的其他产品的递送和储存。至关重要的是，该研究项目将有助于培训16名HQPs，以满足加拿大疫苗生物制造行业对合格人员的高需求。

Viral vectored vaccines are among the most advanced vaccine candidates currently evaluated in COVID-19 phase-3 clinical trials. Following their manufacturing, thermostability is a critical quality attribute to ensure the efficacy and safe delivery of these life-saving products where needed. Building on an extensive scientific knowledge and know-how of cell-culture produced viruses, we propose for this research program to direct our efforts to study and integrate in the process development cycle, liquid and solid formulations of viral vaccines, an area of process development that remained largely empirical. As a result, a novel hybrid model will enable the freeze-dryer optimal operations in real-time. The long-term objective of this research program is to identify the critical parameters and uncover their mechanisms of action towards integration of expert knowledge for digitally assisted design of liquid and solid formulations of effective viral vaccines with improved thermostability and cost-effectiveness. We aim to 1) Design enhanced liquid formulations and uncover mechanisms involved in the formulation of enveloped viruses such as Vesicular Stomatitis Virus (VSV) and lentivirus in presence of excipients to preserve their biological activity; 2) Design solid formulations and develop mechanistic model for freeze-drying of enveloped and non-enveloped viruses such as VSV and adenovirus vectored vaccines generating a large pool of quality data for Machin Learning; 3) Develop a Hybrid model for freeze-drying process using a digital twin approach by combining mechanistic and data-driven models. The digital twin is connected to the physical freeze-dryer by sensors and feed-back control to accelerate the design and optimal operation of freeze-drying processes of viral vaccines. In collaboration with experts in the field (co-supervisors from McGill), we propose to combine recent advancements in Machine Learning (ML) algorithms to assist the design of experiments and accelerate the design of robust vaccine formulations and processes. Experimentation, validation of results and digital twin-based optimization of freeze-drying process will be executed in our already functional biosafety level-2+ laboratory and supported by our validated process analytical technologies. This research will advance knowledge on formulation of enveloped and non-enveloped viruses leading to discovery of novel excipients and enhanced processes. By focusing the design approach on improving the thermostability of the VSV-vectored Ebola vaccine, we will facilitate the delivery and stockpiling of this life-saving product as well as other products using the same VSV delivery platform of antigens (HIV-1, SARS-CoV-2). Critically, this research program will contribute to training 16 HQPs to respond to the high demand of qualified personnel expressed by the vaccine biomanufacturing industry in Canada.