Chimeric antigen receptor (CAR) T-cell therapy is among the most promising breakthroughs in modern oncology. By genetically engineering a patient’s T lymphocytes to recognize and destroy cancer cells, CAR T-cell therapy has redefined what is possible in adoptive cellular immunotherapy.
Within the CERTAINTY consortium—CEllulaR ImmunoTherapy Avatar for personalized cancer treatment—we are investigating how to improve outcomes in hematological cancers, with a focus on multiple myeloma (MM).
MM is a heterogeneous blood cancer caused by uncontrolled plasma cell growth. Its impact is severe, leading to complications such as anemia, kidney dysfunction, or bone lesions. Despite therapeutic advances, MM remains incurable, and patients often undergo multiple lines of therapy.
CAR T-cell therapy offers new hope, but its complexity is immense. Manufacturing is individualized and time-intensive, while follow-up requires both inpatient and outpatient care. Patients face risks of life-threatening toxicities, prolonged immunosuppression, and serious adverse events. This complexity underscores the need for advanced digital approaches that can integrate and analyze data across the entire treatment journey.
High-quality data is the foundation for building models that assess patient status, forecast prognosis, and guide care decisions. This is where virtual twins (VTs) become transformative. In precision medicine, VTs make it possible to simulate multi-organ, multi-scale interactions and inform highly personalized treatment strategies.
What is a Virtual Twin?
In biomedicine, a digital twin typically represents a computational model of a single organ system, updated continuously with patient-specific data. VTs expand on this concept by combining multiple organ-specific models into a unified framework, enabling simulation of complex, multi-organ processes.
A VT integrates data across the full continuum of care, from single-cell multi-omics to clinical outcomes, harmonized into a multi-scale structure.
For CAR T-cell therapy in MM, a VT must replicate immune dynamics, CAR-T–tumor interactions, and potential off-target effects. Achieving this requires the integration of mechanistic cell models, systems biology, and structural biology with artificial intelligence (AI) that powers simulation, learning, and insights.
To be effective, a VT must meet minimum standards, including:
- Continuous updates that reflect evolving therapeutic knowledge
- Strict anonymization and privacy protection to safeguard patients
- Robust infrastructure capable of integrating diverse datasets
Meeting these standards, however, is far from simple.
The Roadblocks: Data, Trust, and Complexity
Building VTs is both a scientific and organizational challenge. Data harmonization is critical, as heterogeneous datasets from multiple sites must be standardized to make machine learning (ML) feasible across institutions.
Adopting a common data format aligned with the FAIR principles (findable, accessible, interoperable, reusable) ensures consistency and enables federated learning, i.e., training models on decentralized datasets without moving them. While federated learning addresses privacy concerns, it adds technical complexity and requires advanced computational expertise.
Equally important are patient trust and data privacy. Transparency, security, and compliance with regulations are non-negotiable, and any breach would jeopardize acceptance. Just as vital, the patient voice must shape VT design. These tools should be created not only for patients but also with their active participation.
Ultimately, the promise of VTs depends on one foundation: clean, harmonized, and secure data. Without it, clinical impact is impossible.
Shaping the Future of MM Care
The rise of VTs represents a new generation of clinical decision-support tools. Their accuracy and utility will always depend on the quality and completeness of the data on which they are built. That is why strong data networks and governance frameworks are essential.
At TriNetX, we are dedicated to advancing this transformation and to advancing MM care. We operate the world’s largest federated network of real-world data (RWD), partnering with healthcare providers across geographies to open new doors in oncology research and patient care. As a data partner within the CERTAINTY consortium, we provide anonymized, standardized datasets capturing the complete treatment journey of MM patients across more than a hundred German sites.
Trained medical experts, supported by programmatic tools grounded in medically relevant rules, conduct rigorous quality controls to ensure VTs are built on reliable data. By systematically collecting key clinical variables across the patient journey, TriNetX lays the foundation for precision-medicine tools at the heart of VTs.
The integration of VTs into oncology care has the potential to transform precision medicine. By combining scientific expertise, high-quality data, and strong governance, we can accelerate innovation, reduce uncertainty, and improve outcomes for patients with MM and beyond.
TriNetX real-world oncology data makes VTs a reality. Let’s accelerate breakthroughs in precision medicine together.
The CERTAINTY project is funded by the European Union (Grant Agreement 101136379). However, views and opinions expressed are those of the author(s) only and do not necessarily reflect those of the European Union or the Health and Digital Executive Agency. Neither the European Union nor the granting authority can be held responsible for them.
About Guillaume Azarias, PhD
As a senior data scientist at TriNetX, Guillaume develops data science tools to accelerate real-world evidence (RWE) generation. His work focuses on improving data quality, harmonizing formats, and applying federated learning to support more effective therapies that benefit patients and help optimize healthcare costs.