Individual patient information to select patients for different radiation techniques.

BACKGROUND AND PURPOSE: Proton therapy is an emerging technique in radiotherapy which results in less dose to the normal tissues with similar target dose than photon therapy, the current standard. Patient-level simulation models support better decision making on which patients would benefit most. MATERIALS AND METHODS: A simulation model was developed tracking individual patients' status regarding the primary tumour and multiple complications. As a proof of principle, the model was populated based on information from a cohort of 1013 head and neck cancer patients. Dose-volume parameters for photon and proton radiation treatment plans were then fed into the model to compare outcomes in terms of length and quality of life and select patients that would benefit most. RESULTS: The illustrative model could adequately replicate the outcomes of photon therapy in the cohort. Improvements from proton therapy varied considerably between patients. The model projects medium-term outcomes for specific individuals and determines the benefits of applying proton rather than photon therapy. CONCLUSIONS: While the model needs to be fed with more and especially recent data before being fully ready for use in clinical practice, it could already distinguish between patients with high and low potential benefits from proton therapy. Benefits are highest for patients with both good prognosis and high expected damage to adjacent organs. The model allows for selecting such patients a priori based on patient relevant outcomes.