Analysis of the sensitivity of tumor control probability in molecular radiotherapy to uncertainties in the dose rate curves

Abstract

In this work, we have investigated the sensitivity of the effectiveness (TCP) of molecular radiotherapy (MRT) treatment to uncertainties of the dose rate curves that may appear when reconstructing those curves. We generated different dose rate curves from experimental data, imposing the constraint of equal dose for each of them. Then, we computed TCPs and looked for correlations between metrics measuring the differences between the dose rate curves and differences in TCP. Finally, according to these results, we estimated the range of tolerable uncertainties in the dose rate curves. The study was performed for different radiopharmaceuticals and different radiosensitive parameters that can affect the dose rate response ($\alpha/\beta$, sub-lethal repair rate). The best correlation between differences in the dose rate curves and TCP was found for a metric that computes averaged linear differences between the curves. With this metric, we quantified differences in dose rate curves that would lead to differences in TCP of 0.02, a parameter denoted $m_{1,\: 0.02}$ that is a surrogate of the dependence of the TCP on the dose rate profile. The results showed that the sensitivity of the TCP to dose rate variations decreases (i.e. larger values of $m_{1,\: 0.02}$) with increasing $\alpha/\beta$ and sub-lethal damage repair rate of the tumor cells, and increasing biological half-life of the dose rate curves. The radiobiological effect of a MRT treatment on a tumor depends on the absorbed dose and the dose rate profile. Ideally, both magnitudes should be measured with accuracy in order to progress towards the optimization of individualized MRT treatments. Our study suggests that this would be more important for tumors with low $\alpha/\beta$ and moderately slow sub-lethal damage repair treated with fast-decaying radiopharmaceuticals.