Periodic Reporting for period 1 - CERTAINTY (A CEllulaR immunoTherapy virtuAl twin for personalIsed cancer treatmeNT)
Période du rapport: 2023-12-01 au 2025-05-31
A virtual twin (VT) is a computer-based system that defines a digital representation of a patient based on their entire relevant data and is frequently updated. When comparing this digital representation with those of other patients and their treatment outcome, patient-specific predictions can be derived. This supports treatment guidance in personalised medicine based on the patient’s current condition. VTs are particularly suitable to assist clinical decision making for patients that receive engineered adoptive cellular immunotherapies (eACIs), like chimeric antigen receptor (CAR) T cells. These “living drugs” exert complex and longitudinally adapting interactions with the patient that can be modeled with VTs. CAR T cell therapy is a novel and innovative treatment option for patients with blood cancers like lymphoma or multiple myeloma (MM). Here, a patient’s T cells are genetically modified outside of the body to express a CAR recognising a tumor antigen. Upon infusion of the CAR T cells, they recognise the tumor cells and kill them. While CAR T cell therapy overall leads to unprecedented response rates with manageable risk, some patients do not respond and/or develop severe toxicities. The underlying mechanisms leading to the differences in response and toxicities are insufficiently understood. VTs can support patients and their doctors to develop individual treatment strategies with CAR T cells, shifting the trial-and-error approach towards a personalised and risk-adapted decision support, implementing true precision medicine. In the project CERTAINTY – a cellular immunotherapy virtual twin for personalised cancer treatment, we develop a VT for MM patients that are eligible for CAR T cell therapy. In this pioneering work, we set up a ground-breaking ecosystem of computational models using comprehensive real-world-data that comprise various biological scales, with a focus on multi-modal molecular data. This allows the incorporation of the unique features of eACIs as “living drugs” into the VT. We implement an incremental and co-creational development process that includes all relevant stakeholders. The VT prototype we develop in CERTAINTY can serve as a blueprint and can be adapted to eACIs beyond CAR T cells and other indications. Including expertise in social sciences and humanities supports CERTAINTY in achieving an overarching view on the VT development, crucial for the compliance by design approach we follow to develop a trusted, usable, and interoperable in silico ecosystem. To this end, we include legal, ethical, socio-cultural, as well as economic expertise and furthermore involve patients and the public as relevant stakeholders.
In the first 18 months of the project, we laid the legal and technical foundation for data sharing: joint controller agreement, data management plan, “FAIR cookbook”, data harmonisation toolkit, recommendation for data exchange standards. Based on this, the first real-world datasets have been harmonised. We set up a federated learning platform and the VT infrastructure based on inputs from relevant stakeholders following the first stakeholder workshop. The first components of the VT are tested and available. We set up in silico dynamic models for individual cell types as well as in vitro personalised 2D models for the bone marrow niche. We analysed MM patient samples prior to and after CAR T cell therapy to longitudinally elucidate the molecular characteristics of patients that underwent CAR T cell therapy.
In CERTAINTY, we develop the first VT prototype that enables patient-specific modeling approaches in eACIs (Weirauch et al. Design Specifications for Biomedical Virtual Twins in Engineered Adoptive Cellular Immunotherapies. NPJ Digital Medicine. In press). Current clinical decision support systems are population-based models, as they do not continuously update model parameters with each patient’s individual data. Furthermore, no VT currently supports decision-making tasks across the entire journey of patients eligible for ACI. But the novel field of eACIs is evolving quickly and significant growth is expected for the upcoming years. With the VT prototype developed in CERTAINTY, we are innovating the use of real-world-data in AI-driven use cases for CAR T cell therapy and will provide unprecedented insights into patient eligibility and outcome. For example, we characterised more than one hundred peripheral blood samples from more than 50 patients at three time points with single-cell multiomics (Rade et al. A longitudinal single-cell atlas to predict outcome and toxicity after BCMA-directed CAR T cell therapy in multiple myeloma. Preprint: https://www.researchsquare.com/article/rs-6165798/v1(s’ouvre dans une nouvelle fenêtre)). With this, we drive the development of innovative AI-based complementary diagnostics for ACIs. This promotes Europe-wide data acquisition and harmonisation following the FAIR principles, that make data findable, accessible, interoperable, and reusable, as well as supports increased availability of ACIs. The VT we develop in CERTAINTY will be fully integrated into the ecosystem supported under the Digital Europe Programme (EDITH-CSA) and will serve as its reference model for eACIs, with setting up a trusted and secure federated network that breaks down data silos and builds a foundation for extending VT application to a wider range of therapies and indications. From the deep insights we generate in CERTAINTY on the mode-of-action of CAR T cells, also drug developers, drug manufacturers, reimbursement agencies, and health technology assessment can benefit, enabling the identification of needs and optimisation of processes.