Periodic Reporting for period 1 - AIDPATH (Artificial Intelligence-driven, Decentralized Production for Advanced Therapies in the Hospital)
Okres sprawozdawczy: 2021-01-01 do 2022-06-30
The AIDPATH automated closed manufacturing platform integrates a modular approach to CAR-T Cell manufacturing to enable flexible and adaptable production of patient-tailored therapies. Semi-automated devices lack of sufficient scalability and are not equipped with the necessary process analytics to fully characterize and control the process and constant culture monitoring capability. Using the Aglaris Facer Bioreactor technology enables close culture control and provides a scale-out pathway for multiple autologous products in a parallel culture significantly increasing the cost-efficiency of the whole platform.
To address and compensate for the inherent variation in the patient starting material and production parameters itself will be optimized with the aid of relevant, experimentally derived process models. Data generated using extensive automated QC equipment on the platform during manufacturing as well as off-line available data from clinical trials and experiments will be used as a basis for these models.
Metabolomic monitoring of the cells can suggest early quality control markers, which aid to adapt the bioprocess to increase the probabilities of successful admissions.
AIDPATH will apply AI technology to integrate patient-specific data and biomarkers in CAR-T therapy and apply flexible manufacturing schemes to obtain CAR-T cell products with optimal fitness and anti-tumor potency. AI technology will also be applied in pre-and in-process controls to optimize scheduling and resource planning to reduce cost and hospital resource utilization and augment patient access.
To support the AI solutions developed within the project a data architecture will be developed to establish a continual learning and optimization process by bundling data from various sources and to achieve patient monitoring and optimized CAR-T production.
The AIDPATH consortium analyzed manufacturing SOPs and developed production sequence diagrams and material flow mappings for the automated CAR-T cell platform. We also prepared a flowchart of the CAR T cell manufacturing and delivery process from the patient to the patient, indicating the decision-making points and which criteria are currently used, and the various bioprocess protocols involved according to the information provided by the partners.
Concepts for the automation have been evaluated and the design of the final platform was finalized. Procurement of the devices, and components has begun and software development for the integration and automation has been started.
The adaptation of the Facer bioreactor and cartridge for integration in the platform is being tested and the control software is adapted.
For the integration of the platform into the hospital, an IT architecture was defined including all software components (control software, data management system, AI models). The first components of the software architecture such as a patient data management system have been deployed.
We undertook experimental work including the generation of initial high-throughput datasets governed by Design of Experiments (DoE) for CAR-T manufacture using perfusion and generated an initial understanding of key process parameters for CAR-T processes, with a specific focus on the metabolites and gaseous concentrations. In collaboration with numerous partners, we have undertaken a series of independent, parallel medium screening studies with 4 different medium formulations.
Media and cells were collected for metabolomic analysis. Also, sampling protocols for metabolic sampling have been developed and standardized across the consortium.
In addition to the experimental work, we determined the main concepts of using AI in CAR-T cell manufacturing and therapy. We could establish five AI use cases including a problem description, a solution concept, and the main functionalities, a review of the related AI techniques, the relevant software technologies, the input and output requirements, the key performance indicators, and a validation concept.
We collected numerous datasets for AI development such as data on T-cell expansion in various bioreactors, historical data on past CAR-T therapy manufacturing runs at FCRB and UKW, data on clinical patients, clinical patient follow-up data, and hospital resource management data.
The AIDPATH consortium also established processes for Stakeholder engagement and held a workshop surveying stakeholder trust and acceptance of AI technology in healthcare and medicine.
Additionally, a regulatory Roadmap has been developed and first steps such as first meetings on regulatory guidance with the ITF and PEI have been held.
The Project will deliver important innovations:
Demonstrating the Smart Manufacturing Hospital with end-to-end AI-optimized processes increases the efficiency of the whole CAR-T manufacturing process from vein-to-vein
A modular and adaptable CAR-T cell manufacturing concept at the point of care results in a wider application of CAR-T cell therapy in Europe by the increased availability of therapies
Accurate mirroring and prediction of CAR-T manufacturing outcomes using novel digital Twin technology leading to improved manufacturing processes
Increased understanding of and optimized bioprocesses by metabolomic analysis of clinical CAR-T cells
Development of novel methods for resource management optimization in Smart Manufacturing Hospitals
Improved cell expansion devices using AI-optimization and advanced process control contribute to the safety and efficacy of CAR-T therapies
We hope that delivering these innovations will enable hospitals to become smart manufacturing hospitals that offers
Benefits for patients with cancer by treating them with affordable high-quality CAR-T cell therapies with shorter vein-to-vein time
Benefits for clinicians and hospitals by making more efficient use of the hospital resources and intensification of data integration
Hospital overarching use and optimization of treatment data for AIDPATH platforms on multiple locations
Increased engagement of healthcare policymakers, investors, and the stakeholder community