Editorial: Multiscale cancer modeling, in silico oncology and digital (virtual) twins in the cancer domain

Georgios S. Stamatakos^1*Maria Angeles Perez Anson²Ravi Radhakrishnan³

• ¹In Silico Oncology and In Silico Medicine Group, Institute of Communication and Computer Systems, School of Electrical and Computer Engineering, National Technical University of Athens, Athens, Greece
• ²University of Zaragoza, Zaragoza, Aragon, Spain
• ³University of Pennsylvania, Philadelphia, Pennsylvania, United States

The following five papers in this collection provide diverse novel approaches, new findings, new insights and new protocols for the development of verifiable cancer digital twins.Perez-Benito et al present a novel framework to create a personalized prostate cancer model by integrating clinical MRI data, such as the prostate and tumor geometry, the initial distribution of cells and the vasculature, so that a full representation of the whole prostate is obtained. On top of the personalized model construction, their approach simulates and predicts temporal tumor growth in the prostate through the Finite Element Method, coupling the dynamics of tumor growth and the transport of oxygen, and incorporating cellular processes such as proliferation, differentiation, and apoptosis. In addition, their approach includes the simulation of the prostate-specific antigen (PSA) dynamics, which allows to evaluate tumor growth through the PSA patient's levels. To obtain the model parameters, a multi-objective optimization process is performed to adjust the best parameters for two patients simultaneously. This framework is validated by means of data from four patients with several MRI follow-ups. The diagnosis MRI allows the model creation and initialization, while subsequent MRI-based data provide additional information to validate computational predictions. The model predicts prostate and tumor volumes growth, along with serum PSA levels.Stamatakos et al present a novel mechanistic multiscale model as the core of a digital twin of prostate tumor growth and response to external radiotherapeutic schemes, based on a discrete entity and discrete event simulation approach. Following technical verification, an adaptation to clinical data approach is delineated. Multiscale data has been provided by the German clinical study HypoFocal-SBRT. Additionally, a sensitivity analysis has been performed. The impact of model parameters such as cell cycle duration, dormant phase duration, apoptosis rate of living and progenitor cells, fraction of dormant stem and progenitor cells that reenter cell cycle, number of mitoses performed by progenitor cells before becoming differentiated, fraction of stem cells that perform symmetric division, fraction of cells entering the dormant phase following mitosis, alpha and beta parameters of the linear-quadratic radiobiological model and oxygen enhancement ratio has been studied. A qualitative agreement of the model behavior with experimental and clinical knowledge has set the basis for the next steps towards its thorough clinical validation, its technological integration and its eventual certification and clinical translation. A brief historical review of the formal emergence of in silico medicine in 2002 and cancer digital twins in 2007 is also included.Hadjicharalambous et al introduce a data-driven in silico modeling procedure to simulate the biomechanics of prostate brachytherapy through the use of Finite Elements. Comprehensive magnetic resonance and transrectal ultrasound images acquired prior, during and post brachytherapy are employed for model personalisation, while the therapeutic procedure is simulated via sequential insertion of multiple catheters in the prostate gland. The medical imaging data are also employed for model evaluation, thus, demonstrating the potential of the proposed in silico procedure to be utilised pre-and intra-operatively in the clinical setting.Meyerheim et al delineate a study that has used T2-weighted MRI scans, chemotherapy treatment plans, and post-surgical histological profiles from three patients enrolled in the SIOP 2001/GPOH clinical trial in order to clinically adapt a nephroblastoma digital twin i.e. the Nephroblastoma Oncosimulator. Each patient represents a distinct clinically assessed risk group. The paper investigates the clinical adaptation of the digital twin to these datasets. The goal is to derive appropriate value distributions of the model input parameters that enable accurate prediction of tumor volume reduction in response to preoperative chemotherapy. Distributions of the total cell kill ratio for one patient of each risk group: low, intermediate and high risk are derived. Statistically significant differences are observed between the high-risk group and both the low-and intermediate-risk groups.Kemkar et al provide a literature review that discusses recent advancements in computational methodologies for precision oncology, highlighting the potential of