70th Annual Meeting of the Radiation Research Society, Tuscon, AZ, September 15-18, 2024
USTUR faculty were authors on two presentations at the 2024 Radiation Research Society meeting in Tuscon, AZ.
Can misclassification of disease change the conclusion of significant dose-response associations?
Xirui Liu (USTUR), Stacey McComish (USTUR), Joey Zhou (USTUR) and Sergey Tolmachev (USTUR)
Objective: Misclassification of causes of death on death certificates can occur. This study aims to quantify the probability of death certificate misclassification errors changing the conclusion of dose-response associations.
Methods: This study utilizes a real dose distribution from Rocky Flats nuclear workers. A logistic function was used to generate outcome probabilities for each dose record, assuming an odds ratio of 1.85 with a baseline cancer rate of 26%. Health outcomes corresponding to each case were then generated based on these probabilities. This process was repeated 1,000 times to create 1,000 datasets. From these, a statistically non-significant dataset with a p-value slightly exceeding 0.05 was selected. Over- and under-misclassification rates of 30% were randomly applied to this initially selected dataset to simulate a real-world
misclassified dataset. This process was repeated 20,000 times. Subsequently, odds ratios and p-values of the misclassified datasets were calculated to evaluate the effect of misclassification.
Results: There is a 10.0% chance that misclassification errors cause the association to become statistically significant between the radiation dose and the outcome.
Conclusion: In general, it is believed that if misclassification of disease had been incorporated into epidemiological studies with significant dose-response associations, it would have resulted in findings that were more significant. However, these findings demonstrated that in a notable percentage of the time this belief is incorrect. [USTUR-0675-24A]
Calibration of a novel microdosimetry system for in vitro applications of actinide radiopharmaceuticals
Lydia J. Wilson (Thomason Jefferson University), Brian Miller (University of Arizona), George Tabatadze (USTUR) and Firas Mourtada (Thomason Jefferson University)
Alpha-emitting radionuclides show promise for targeted radiopharmaceutical therapy (TRT). The high linear energy transfer and short range of alphas result in highly localized DNA damage for effective cancer management. However, the distributions of radioactivity and absorbed dose on the cellular scale remain elusive, hindering progress toward understanding α-TRT radiobiologic response, evaluating novel α-TRT drug efficacy, and optimizing administration for personalized treatments. A novel real-time α camera, the ionizing radiation quantum imaging detector (iQID), has successfully mapped the micro-distribution of α-emitting radionuclides in mouse and human tissues. This study characterized the pixel size, spatial resolution, sensitivity, and background rate of the iQID for use in whole-animal α-TRT experiments.
The scintillator-based α detection system maps individual α emissions on a 10 ⨯ 10 cm2 detector area. We measured pixel size using calibrated image templates. Line spread functions (LSF) measured with a 5 μm ⨯ 5 mm laser-drilled collimator and a 5 mCi 210Po source (5.4 MeV α) quantified intrinsic spatial resolution. We measured the LSF at 5 locations on the detector surface: center and 4 edges in a ⨯ pattern to evaluate uniformity. Decays counted from a 50.2 ± 1.1 Bq source (243Am, 242Pu, and 239Pu; 5.16 – 5.43 MeV α) over 48 hours quantified detection efficiency. Finally, we evaluated the background rate over 60
hours.
Pixels measured 53.7 μm/pixel. The full width at half maximum of LSFs averaged 43.0 ± 5.2 μm. The mean radioactivity over 48 h was 54.34 ± 0.03 Bq, within 10% of the source-certificate value. The mean background rate over 60 hours revealed a lower limit of detection (LLD) of 4.04 mBq/cm2.
Imaging results with a novel α camera for α-TRT experiments showed that it can image and identify α-emission events with nearly 100% efficiency, low LLD, and spatial resolution approaching the cellular scale. Future work will evaluate energy discrimination and build Artificial Intelligence tools to reconstruct whole-animal volumetric activity concentrations. The iQID is a promising tool for α-TRT experiments, filling a critical need in the field. With advanced microdosimetric understanding, novel applications of α-TRT can offer optimal cancer management to patients with previously intractable disease. [USTUR-0673-24A]