In Silico Trial to Estimate Personalized RT Dose in Head and Neck Cancer

Abstract

<h3>Purpose/Objective(s)</h3> As current radiotherapy (RT) treatment schedules are not personalized for individual patients, with the prescribed dose being uniform for particular subtypes and stages of cancer, despite highly variable responses between patients, we performed an <b>in silico</b> trial to determine optimal personalized RT dose for head and neck cancer patients in order to minimize excess dose with the objective of minimizing toxicity and improving QOL without sacrificing tumor control. <h3>Materials/Methods</h3> Weekly tumor volume data were collected from two cancer centers for n = 39 head and neck cancer patients that received 66-70 Gy in 2-2.12 Gy daily fractions or with accelerated fractionation. Clinical outcome data, i.e., locoregional control (LRC) and disease-free survival (DFS), were also collected. Tumor volume reduction was connected to LRC by means of a volume reduction threshold associated with LRC. Due to limited patient numbers, the <b>in silico</b> trial was performed in an iterative leave-one-out fashion as follows: (1) a simple 2 parameter model of tumor growth and response to RT was calibrated to tumor volume data from 38 patients, (2) the calibrated model parameters were combined with the left-out patient's tumor volume data from weeks 1-4 of RT to simulate tumor volumes forward sampling from a patient-specific distribution of radiosensitivity parameters, (3) minimum cumulative radiation dose required for local tumor control was estimated by calculating the dose at which > 95% the simulated tumor volume trajectories were below the volume reduction threshold associated with LRC. <h3>Results</h3> The optimal minimum radiation dose required for LRC was calculated for each patient. We found that 87% of the patients (34/39) received a higher total dose than estimated as necessary by our model with an average overdose of 37 Gy/patient, while the remaining patients were estimated to have received too little dose, with an average underdose of 47 Gy/patient. Notably, our results showed that the current one-size-fits-all approach results in no patient receiving their optimal RT dose. <h3>Conclusion</h3> Here we show a method to determine a personalized minimum RT dose to achieve LRC. Such simulations and estimates may allow radiation oncologists to identify candidates for dose de-escalation, dose escalation, or additional concurrent treatments, such as chemotherapy.