Mechanistic Modeling of Radium-223 Treatment of Bone Metastases

Abstract

Despite the effectiveness of 223RaCl2 for treating patients with symptomatic bone metastatic disease, its mechanisms of action are still unclear. Even established dosimetric approaches differ considerably in their conclusions. In silico tumor models bring a new perspective to this situation because they can quantitatively simulate the interaction of α-particles with the target(s). Here, we investigated 3 different mathematical models of tumor growth that consider the radiation effect of radium-223 (223Ra) treatments and compared the results with clinical data. The well-established Gompertz growth model was applied to simulate metastatic tumor burden. On the basis of published measurements of 223Ra uptake, we have incorporated the radiation effect of α-particles into the model and investigated 3 radium distribution scenarios-uniform exposure, exposure of only an outer layer, and exposure of a constant volume of the tumor. For each scenario, the times for various tumor stages to progress to the first symptomatic skeletal event were calculated. Uniform and outer-layer exposure scenarios showed very poor agreement with the Kaplan-Meier patient curves from clinical data. However, the constant-volume effect predicted outcomes very similar to the observed clinical results, suggesting, depending on the dose rate, that relatively small fractions of the cell population see damage from 223Ra. The commonly used assumption of uniform 223Ra distribution does not accurately reflect clinical responses. The suggestion that only a subpopulation of the tumor might be affected by 223Ra shows a pressing need to further study the tumor and drug kinetics to schedule more effective treatments in the future.