Purpose: To develop a novel 4D dynamic liver blood flow model, capable of accurate dose estimation to circulating blood during fractionated radiotherapy.
Methods: Adult male and female liver phantoms with detailed vascular trees including the hepatic arterial, portal venous and hepatic venous trees were developed. A discrete time Markov Chain approach was applied to determine the spatiotemporal distribution of 10⁵ blood particles (BP) in the human body, based on reference cardiac output and blood volume values. For BPs entering the liver, an explicit Monte Carlo simulation was implemented to track the propagation of individual BPs through the vascular trees and time-dependent radiation fields. The model was evaluated for photon (VMAT, IMRT) and proton (passive SOBP and active PBS) treatments and the impact of total delivery time on circulating blood was investigated and quantified using mean dose(μdose), V0Gy, V0.5Gy.
Results: The model tracks over 10⁴ BPs in the liver simultaneously along 1996 distinct vascular pathways, accumulating dose from time-dependent radiation fields with a 0.1s time resolution. The simulations of the four modalities estimate μdose(VMAT) (2.44Gy)>μdose(IMRT) (2.42Gy)>μdose(SOBP) (2.20Gy)>μdose(PBS) (2.08Gy) after 15 fractions, related to the differences in mean liver dose. However, V0Gy and V0.5Gy are linked to the total delivery time, where V0Gy is lowest for VMAT (37%, 60s delivery), and highest for PBS (81%, 255s delivery). V0.5Gy shows the opposite trend, highest for VMAT (15%) and lowest for PBS (0.6%), demonstrating the tradeoff between low dose to large fractions of circulating blood and high dose to a small fraction depending on delivery speed.
Conclusion: The 4D dynamic liver blood flow model enables realistic dose estimation to circulating blood during radiotherapy. The model allows us to study the impact of treatment modalities and delivery time on the dose to circulating blood, which will provide insights into radiation-induced lymphopenia and guide radiotherapy planning.
Funding Support, Disclosures, and Conflict of Interest: This work was supported by NCI R01 CA248901, NCI R21 CA248118, and NCI R21 CA241918
Not Applicable / None Entered.
Not Applicable / None Entered.