Dynamic Blood Dose Estimates for Sarcoma and Their Correlation with Radiation-Induced Lymphocyte Killing
Abstract
Purpose
Radiation-induced lymphocyte killing is correlated with worse clinical outcomes in various cancers, including sarcomas. Models to simulate radiation dose delivery to circulating lymphocytes have been developed, and significant correlations between simulated blood dose estimates and severe lymphopenia and/or adverse clinical outcomes have been reported, but such work remains outstanding for sarcomas, for which treatment site variations complicate blood dose modelling. This work investigates correlations between blood dose estimates and lymphopenia in sarcomas.
Methods
Of 13 sarcoma patients initially included, pre- (ALCPre) and post-radiotherapy active lymphocyte counts (ALCPost) were only available for eight patients (lower extremities: 5, pelvis: 1, upper extremities: 2). For each of these patients, dozens of organs-at-risk (OARs), blood vessels, and tissues were segmented, and simulations of radiation dose delivery to circulating blood during delivery of the patients’ clinical radiotherapy treatment plans were performed using HEDOS, considering temporal aspects like blood flow dynamics and treatment delivery time. Blood dose simulation results were evaluated via the mean blood dose (Dmean) as well as the dose delivered to the 90% and 10% of blood particles receiving the highest dose (D90% and D10%). Linear regression was used to investigate potential correlations between blood dose and lymphocyte killing.
Results
All blood dose metrics exhibited a statistically significant (p<0.05) correlation with radiation-induced lymphocyte killing (ALCPre-ALCPost), with r2=0.547,p=0.036 for D90%, r2=0.614,p=0.021 for Dmean, and r2=0.644,p=0.016 for D10%. At higher blood doses, ALCPre-ALCPost plateaued, potentially due to the high radiosensitivity/low lethal dose of lymphocytes. Neither tumor volume nor integral body dose exhibited a significant correlation with lymphocyte killing.
Conclusion
Radiation-induced lymphocyte killing significantly increased with all evaluated blood dose metrics, despite the inclusion of different treatment sites. Subsequently to dataset expansion, we will therefore pursue a proof-of-principle of blood dose optimization for sarcoma, with the aim of reducing radiation-improved lymphopenia and improving clinical outcomes.