Fast, Automated, Knowledge-Based Treatment Planning for Selecting Patients for Proton Therapy Based on Normal Tissue Complication Probabilities

Roni Hytönen,Marije R Vergeer,Reynald Vanderstraeten,Timo K Koponen,Christel Smith,Wilko F.A.R Verbakel

doi:10.1016/j.adro.2022.100903

Abstract

PurposeSelecting patients who will benefit from proton therapy is laborious and subjective. We demonstrate a novel automated solution for creating high-quality knowledge-based plans (KBPs) using proton and photon beams to identify patients for proton treatment based on their normal tissue complication probabilities (NTCP). Methods and MaterialsTwo previously validated RapidPlan PT models for locally advanced head and neck cancer were used in combination with scripting to automatically create proton and photon KBPs for 72 patients with recent oropharynx cancer. NTCPs were calculated for each patient based on the KBPs, and patient selection was simulated according to the current Dutch national protocol. ResultsThe photon/proton KBP exhibited good correlation between predicted and achieved organ-at-risk mean doses, with a ≤5 Gy difference in 208/196 out of 215 structures relevant for the head and neck cancer NTCP model. The proton KBPs yielded on average 7.1/6.1/7.6 Gy lower dose to salivary/swallowing structures/oral cavity than the photon KBPs. This reduced average grade 2/3 dysphagia and xerostomia by 7.1/3.3 and 5.5/2.0 percentage points, resulting in 16 of 72 patients (22%) being indicated for proton treatment. The entire automated process took <30 minutes per patient. ConclusionsAutomated support for decision making using KBP is feasible and fast. The planning solution has potential to speed up the planning and patient-selection process significantly without major compromises to the plan quality.

Highlights

Recent increases in proton treatment capacity and advancement in radiotherapy treatment modalities have made it possible to treat more patients with proton therapy[1]
We investigated which fraction of patients would be eligible for proton therapy, using the original RapidPlan models and the model with the adapted oral cavity (OC) objectives
In case the dose-volume histograms (DVHs) estimates are used only for the proton plan, while the actual DVHs from knowledge-based plans (KBPs) are used for the photon plan, the number of patients indicated erroneously for proton therapy drops to six. In this proof-of-principle study we demonstrated that commercially available scripting and knowledge-based planning tools can be used to create a fast, fully automated decision support pipeline for the selection between photon and proton treatments for head-and-neck cancer (HNC)

Summary

Introduction

Recent increases in proton treatment capacity and advancement in radiotherapy treatment modalities have made it possible to treat more patients with proton therapy[1]. One objective way to select the patients that can benefit most from proton therapy is by using a model-based indication methodology. The methodology could be based on the estimated reduction in normal-tissue complication probability (NTCP)[2,3,4,5,6]. This selection process necessitates creation of high-quality photon and proton treatment plans for individual patients, which can be laborious, time consuming, and prone to bias. We demonstrate a novel automated solution for creating high-quality knowledge-based plans (KBPs) using proton and photon beams to identify patients for proton treatment based on their normal tissue complication probabilities (NTCP)

Methods

Results

Discussion

Conclusion