Genomic Science Program
U.S. Department of Energy | Office of Science | Biological and Environmental Research Program

Multiscale Computational Digital Twins for Whole-Body to Subcellular Radiation Effects

Authors:

Greeshma Agasthya1* (agasthyaga@ornl.gov), Paul Inman1, Matthew Andriotty1,2, Zachary Fox1, Arpan Sircar3, Xiao Wang1, Zakaria Aboulbanine1, Jayasai Rajagopal1, Debsindhu Bhowmik1, Christopher Stanley1, Miguel Toro Gonzalez4, Sandra Davern4, Anuj Kapadia1, Rick Stevens5

Institutions:

1Computational Sciences and Engineering Division, Oak Ridge National Laboratory; 2Georgia Institute of Technology; 3Nuclear Energy and Fuel Cycle Division, Oak Ridge National Laboratory; 4Science and Technology Division, Oak Ridge National Laboratory; 5Computing, Environment, and Life Sciences Directorate, Argonne National Laboratory

Goals

The goal of the work is to advance understanding of the contributions of low-dose (LD) radiation to cancer by identifying and modeling the key molecular and cellular mechanisms involved, developing optimal strategies for interrogating these mechanisms experimentally, acquiring and integrating diverse datasets to formulate and test hypotheses, and validating predictive multiscale models of radiation risk. The work will be performed through three broad goals: (1) identifying and modeling molecular and cellular mechanisms of LD radiation damage and repair; (2) identifying and characterizing signatures of LD exposure and LD-induced tumorigenesis; and (3) developing a multiscale modeling and simulation framework for estimating cancer risk from radiation exposures.

Abstract

This work describes the development and implementation of a multiscale computational digital twin framework to assess radiation exposure effects on humans. Using high-performance computing methods, the team assess radiation exposure effects at the whole-body, multicellular, and subcellular scales, seamlessly integrating between the three. The project’s framework includes a population of human digital phantoms along with computing environments to model radiation transport and associated radiation damage. The project includes a multiscale perfusion model for radioisotope delivery and uptake, modules for DNA damage and DNA repair, and multicellular growth models to determine radiation effects on cells and organs. An explanation of the integration of these scales into the framework is provided below.

At the subcellular level, using experimental Hi-C data (Sanders et al. 2020), researchers create 3D chromosome models imported into TOPAS-nbio (a Monte Carlo simulation platform) to estimate DNA double-strand breaks (DSB; Chatzipapas et al. 2020). Mechanistic repair models from the Mechanistic DNA Repair and Survival Model (MEDRAS) predict post-radiation aberrations over time (McMahon and Prise 2021). Validation against in vitro studies and cell experiments with a mouse breast cancer cell line confirms accuracy.

At the multicellular and whole-body scales, the project integrates a tumor growth model in Compucell3D with eXtended CArdiac Torso (XCAT) phantoms for whole-body simulations (Segars et al. 2013). Geant4, a Monte Carlo simulation software, calculates absorbed dose in radiation therapy protocols, feeding back into Compucell3D to assess cell survivability (Allison et al. 2006, 2016; Swat et al. 2012). Validation against recent multicellular models shows promising results, with ongoing efforts to validate against spheroid-based experiments.

For multiscale perfusion modeling, the team has implemented a physiology-based pharmacokinetic (PBPK) model simulating radioisotope distribution at the organ level, feeding into a computational fluid dynamics model for tissue-level perfusion. Work is underway to demonstrate subcellular and multicellular scale radiopharmaceutical perfusion. Preliminary results involve growing spherical tumors with multiple cell types in a controlled nutrient environment, exposing them to Actinium 225 in Geant4 to estimate cell radiation dose and survivability.

In conclusion, the project’s framework offers a robust tool for modeling radiation effects across various radiation dose–exposure scenarios. Importantly, parts of this pipeline are fully automated and optimized for graphics processing unit computation. With potential applications in environmental radiation exposures, occupational exposures, and medical exposures including radiation treatments, the team envisions the multiscale in silico digital twin framework providing a robust method to evaluate outcomes from unwanted exposures as well as the ability to optimize desired exposures (such as radiation treatments) for the benefit of the individual. This framework facilitates predicting cell survival and treatment outcomes across different cancer types, integrating absorbed dose, biodistribution, cell toxicity, and repair mechanisms to determine overall outcomes in the human body.

References

Allison, J., et al. 2006. “Geant4 Developments and Applications,” IEEE Transactions on Nuclear Science 53, 270–78.

Allison, J., et al. 2016. “Recent Developments in Geant4,” Nuclear Instruments and Methods in Physics Research Section A: Accelerators, Spectrometers, Detectors and Associated Equipment 835, 186–225.

Chatzipapas, K. P., et al. 2020. “Ionizing Radiation and Complex DNA Damage: Quantifying the Radiobiological Damage Using Monte Carlo Simulations,” Cancers 12(4), 799.

McMahon, S. J., and K. M. Prise. 2021. “A Mechanistic DNA Repair and Survival Model (Medras): Applications to Intrinsic Radiosensitivity, Relative Biological Effectiveness and Dose-Rate,” Frontiers in Oncology 11.

Sanders, J. T., et al. 2020. “Radiation-Induced DNA Damage and Repair Effects on 3D Genome Organization,” Nature Communications 11, 6178.

Segars, W. P., et al. 2013. “Population of Anatomically Variable 4D XCAT Adult Phantoms for Imaging Research and Optimization,” Medical Physics 40.

Swat, M. H., et al. 2012. “Chapter 13: Multi-Scale Modeling of Tissues Using CompuCell3D.” In Methods in Cell Biology 110, 325–66. Eds. Asthagiri, A. R. and A. P. Arkin. Academic Press.

Funding Information

This work was supported by the BER program by UT-Battelle, LLC under Contract No. DE-AC05-00OR22725 with the U.S. DOE. Oak Ridge National Laboratory’s work on the LUCID (Low-dose Understanding, Cellular Insights, and Molecular Discoveries) program was supported by the U.S. DOE, Office of Science, BER program, under Contract UT-Battelle, LLC-ERKPA71, and Laboratory Directed Research and Development Program of Oak Ridge National Laboratory, managed by UT-Battelle, LLC, for the U. S. DOE. This research used resources from the Compute and Data Environment for Science (CADES) at Oak Ridge National Laboratory.