Reinhard Laubenbacher, PhD
Today, nearly every industry that deals with complex technologies uses sophisticated computational infrastructure to perform in silico simulations, employing “digital twins” to forecast how individual pieces of equipment perform under ever-changing real-world conditions. For instance, a digital twin of a jet engine is continuously updated with operational data streamed over the “internet of things” while the engine is in flight. Any deviation between prediction and actual engine performance is an early warning of a potential problem, which can be addressed before it becomes a serious issue. In medicine, more limited medical “digital twins” are already helping to provide better care to patients. Digital twins of major blood vessels allow the early diagnosis of potential abnormalities and aid in the design of interventions if and when they become necessary. A digital twin of a computational “artificial pancreas,” for instance, can largely automate the administration of insulin to control and reduce the long-term consequences of Type I diabetes.
The immune system plays a key role in many diseases, beyond the response to pathogens, including cancer, diabetes, and a host of autoimmune diseases, so immune digital twin (IDT) technology could have a high impact on many key areas of human health. For instance, our response to the SARS-CoV-2 pandemic could have been enhanced greatly by the ability to repurpose drugs targeted to an individual patient’s immune system characteristics. At the same time, the immune system is immensely complex, and poses many challenges, from collection of appropriate data to modeling approaches. The time is right for a concerted research effort to develop key components of immune digital twin technology. The hypothesis underlying this workshop is that a sufficiently well-developed computational immune system model, targeted to one or two suitably chosen medical use cases, can serve as a driver and coordinating framework for research, and can guide the collection and use of appropriate data for model validation and personalization. Therefore, the workshop will primarily be focused on delineating a roadmap for the development of a computational immune system model in a biomedical context that can serve as a blueprint for other uses of IDT technology. Immune digital twin models will have a wide range of applications, from drug development, e.g., the creation of synthetic populations for drug trials, repurposing of drugs, diagnostics and prognostics, as well as treatment optimization. This technology will enable us to account for heterogeneity among patient populations, and will help address health inequities.
Workshop Aim 1. Assemble a group of 30 scientists with expertise in immunology, clinical science, and mathematical and computational modeling, to discuss the main challenges facing the development of digital twin technology, determine the most promising immediate application areas, outline a conceptual map of an immune system model appropriate for these application areas, and determine the key data needs to calibrate, validate, and personalize such models.
Workshop Aim 2. Prepare drafts of two documents on the topic of the workshop, one a manuscript for a perspective-type article for publication, the other a more detailed “Roadmap” document that can form the basis for an IDT project.
The IDT technology proposed here has wide applicability to the personalized treatment of diseases and conditions in which the immune system plays a role. This includes in particular the large population of veterans with health conditions. Beyond this relevance to the mission of the Veterans Administration to provide healthcare to this population, an additional important aspect of IDTs is that they could be adapted for individualized risk assessment and performance prediction for soldiers faced with adverse conditions during deployment, such as contaminated environments or extreme physical conditions.