Session Item

We recently reported on quality of life (QOL)-guided radiotherapy including NTCP-based treatment plan optimisation. This method employs NTCP objective functions, prioritised based on their impact on QOL, instead of conventional dose-based objective functions for individual organs at risk (OAR). We recently expanded this method to include simple models for tumour control probability (TCP) and general plan quality (e.g., hotspot reduction, target dose homogeneity), allowing for fully automated planning. The aim of this study was to test the feasibility of this method and to investigate its sensitivity to the use of different schemes for prioritising toxicities during treatment plan optimisation.

Ten patients, representative of our population receiving definitive (chemo)radiotherapy for head and neck cancer, were selected for this study. Fully automated VMAT treatment plans were created using a mix of NTCP and TCP objective functions. TCP objective functions were based on a PTV D98 ≥ 95% of the prescribed dose. Few fixed dose-based objectives were added for critical structures. Three automated plans were created for each patient using different priorities for 20 toxicities including salivary, swallowing and speech problems, pain and general symptoms. Toxicities were prioritised according to three schemes: 1) relative impact according to the published QOL model (weighted toxicity function, with a low priority for salivary toxicities); 2) identical weights for each toxicity; and 3) prioritised prevention of salivary toxicities, in which salivary toxicities, swallowing problems and other toxicities were weighted as 3 : 2 : 1, respectively. For each patient, the manually created clinical VMAT plan was used as reference. General plan quality and beam setup was equal or similar for all plans including the clinical plan. Plans were compared using dose-volume data in combination with NTCP and QOL using published and validated models.

Fully automated NTCP-TCP-based optimisation consistently resulted in superior plans without any hotspots. Compared to the clinical plan, all automated plans resulted in a lower dose in OAR (except for some plans in the parotid glands) and always lower NTCPs for swallowing, speech, pain and general toxicities. The average QOL score in the automated plans was always better than in the clinical plans (Table). Automated plans with prioritised prevention of salivary toxicities had the lowest NTCPs for salivary toxicities, however, at the cost of higher NTCPs for moderate-to-severe general symptoms such as fatigue. Plans optimised based on the published QOL model had the best QOL scores.

Automated NTCP-TCP-based optimisation is feasible and consistently outperforms manually generated VMAT plans in terms of NTCP and QOL scores. The highest QOL score was obtained for plans optimized prioritising the toxicities with the highest relative impact on QOL.