Application of Multivariate Modeling for Radiation Injury Assessment: A Proof of Concept (Radiation Injury Algorithms)
Abstract
The focus of this study was to formulate a multivariate algorithm using classical CBC and serum chemistry blood parameters for utility in predicting severe hematopoietic Acute Radiation Syndrome (H-ARS) injury (i.e., Response Category three or RC3) in a Rhesus monkey totalbody irradiation (TBI) model. Multivariate Radiation Injury Estimation algorithms were formulated for estimating a H-ARS RC3 condition, which was induced by a 6.5-Gy TBI dose. An archived blood dataset was examined from a radiation study involving 24 nonhuman primates (NHP) (Macaca mulatta) given 6.5-Gy 60Co -rays (0.4 Gy min-1) TBI. Blood biosampling was performed prior to irradiation (d 0) and on d 7, 10, 14, 21, and 25 postirradiation. Changes in CBC and serum chemistries were identified for multivariate modeling. A correlation matrix was then formulated with the RC3 (radiation dose 6.5 Gy) designated as the "dependent variable." Independent variables were identified based on their radio-responsiveness and relatively low multi-collinearity using stepwise- linear regression analyses. Final candidate independent variables included CBC counts (absolute number of neutrophils, lymphocytes, and platelets) in formulating the "CBC" RC3 estimation algorithm. Additionally, the formulation of a diagnostic CBC and serum chemistry CBC-SCHEM RC3 algorithm expanded upon the CBC algorithm model with the addition of hematocrit and the serum enzyme levels of aspartate aminotransferase, creatine kinase, and lactate dehydrogenase. Both algorithms estimated RC3 spanning 7 to 25 days post-irradiation with over 90 % predictive power (CBC: 91 % 1.01, P = 0.00001, n = 92; CBC-SCHEM: 93 % 0.88, P = 0.00001, n = 92). Only the CBC-SCHEM RC3 algorithm however, met the critical three assumptions of Linear-Least-Squares demonstrating slightly greater precision for radiation injury estimation, but with significantly decreased prediction error (t > 108, P =0.00001) indicating increased statistical robustness.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jan 01, 2014
- Accession Number
- ADA606231
Entities
People
- David J. Villa Vilmar
- David L. Bolduc
- G. David Ledney
- Rolf Buenger
- William F. Blakely
Organizations
- Armed Forces Radiobiology Research Institute