Bridging the gaps: using an NHP model to predict single dose radiation absorption in humans

Abstract

Purpose: Design and characterization of a radiation biodosimetry device are complicated by the fact that the requisite data are not available in the intended use population, namely humans exposed to a single, whole-body radiation dose. Instead, one must turn to model systems. We discuss our studies utilizing healthy, unexposed humans, human bone marrow transplant patients undergoing total body irradiation (TBI), non-human primates subjected to the same irradiation regimen received by the human TBI patients and NHPs given a single, whole-body dose of ionizing radiation.

Materials and Methods: We use Bayesian linear mixed models to characterize the association between NHP and human expression patterns in radiation response genes when exposed to a common exposure regimen and across exposure regimens within the same species.

Results: We show that population average differences in expression of our radiation response genes from one to another model system are comparable to typical differences between two randomly sampled members of a given model system and that these differences are smaller, on average, for linear combinations of the probe data and for the model-based combinations employed for dose prediction as part of a radiation biodosimetry device.

Conclusions: Our analysis suggests that dose estimates based on our gene list will be accurate when applied to humans who have received a single, whole-body exposure to ionizing radiation.

Keywords:

• Radiation biodosimetry
• non-human primate model system
• total-body irradiated human bone marrow transplant patient model system
• Bayesian linear mixed models

Disclosure statement

One or more of the investigators conducting this research has a financial interest in the REDI-Dx Biodosimetry Test System. If commercially successful, individual investigator(s) and/or Duke University could benefit financially. The REDI-Dx Biodosimetry Test System is For Investigational Use Only in the USA and CE-IVD marked.

Additional information

Funding

This project has been funded in whole or in part with Federal funds from the Biomedical Advanced Research and Development Authority, Office of the Assistant Secretary for Preparedness and Response, Office of the Secretary, Department of Health and Human Services, under contract numbers HHSO100201000001C and HHSO100201600034C.

Notes on contributors

Edwin S. Iversen

Edwin S. Iversen, Jr is Research Professor in the Department of Statistical Science, Duke University. His primary research focuses on statistical methods for predicting an exposure, disease state or outcome given molecular and/or genetic data; his research interests include risk prediction and classification, model evaluation and refinement, latent/auxiliary variable methods, Bayesian methods and integration of bioinformatics resources and tools into statistical models.

Janice M. McCarthy

Janice M. McCarthy is Medical Instructor in the Department of Biostatistics and Bioinformatics, Duke University. Her research interests are focused in the areas of statistical genetics, biological big data and the analysis of multi-omic datasets.

Kirsten Bell Burdett

Gary Lipton is Bioinformatician, Duke University Cancer Institute.

Gary Lipton

Kirsten Bell Burdett is Bioinformatician, Duke University Cancer Institute.

Gary Phillips

Gary Phillips is Project Planner, Department of Hematologic Malignancies & Cell Therapy, Duke University. He is the program manager and co-investigator for the BARDA biodosimeter contract. His research efforts are focused on the development and commercialization of a gene signature based biodosimeter.

Holly Dressman

Holly Dressman is a Research Professor in the Department of Molecular Genetics and Microbiology and the Director of the Duke Microbiome Shared Resource, a component of the Center for Genomic and Computational Biology. Research efforts have focused on gene expression and microbiome profiling, computational tools for the analysis of gene expression data, as well as database systems for storage and further analysis of expression experiments. In particular, understanding gene expression profiles after radiation exposure has involved studying various model systems that have resulted in co-authorship on four publications.

Joel Ross

Joel Ross is Project Planner, Department of Hematologic Malignancies & Cell Therapy, Duke University. He is the program manager for the Radiation Countermeasures Center of Research Excellence at Duke University. In this role, he is been involved in research and development of radiation injury mitigators and biodosimeters.

Nelson Chao

Nelson Chao is the Donald D. and Elizabeth G. Cooke Cancer Research Professor, Department of Hematologic Malignancies & Cell Therapy, Duke University. His research interests are in two broad areas, clinical hematopoietic stem cell and cord blood transplantation and in the laboratory studies related to graft vs. host disease and immune reconstitution.