A probabilistic evaluation of the Dutch robustness and model-based selection protocols for Head-and-Neck IMPT: A...

Area of Science:

• Medical Physics
• Radiation Oncology
• Oncology

Background:

• Proton therapy (PT) protocols, including robustness evaluation (RE) and model-based selection (MBS), are standardized across three Dutch proton centers.
• This study addresses inter-patient and inter-center variations in target dose and the robustness of the MBS protocol for head-and-neck cancer (HNC) patients.

Purpose of the Study:

• To investigate variations in target dose delivered by the RE protocol across Dutch proton centers.
• To assess the robustness of the MBS protocol for patient selection in intensity-modulated proton therapy (IMPT) for HNC.

Main Methods:

• Evaluated clinical treatment plans of 100 HNC patients from three Dutch proton centers.
• Utilized Polynomial Chaos Expansion (PCE) for probabilistic evaluation of 100,000 fractionated treatments per plan.
• Assessed dose to the clinical target volume (CTV), tumor control probability (TCP), organs at risk (OARs), and normal tissue complication probabilities (NTCP).

Main Results:

• RE protocol doses for CTV70.00 were consistent across centers.
• Minor dose variations (1 GyRBE) for CTV54.25 at one center did not impact modeled TCP.
• Differences in expected versus nominal NTCP were minimal (<0.3 percentage points).

Conclusions:

• Standardized RE and MBS protocols yield comparable TCP and NTCP results.
• Significant inter-patient and inter-center variations persist due to institutional clinical practices.
• The MBS approach is confirmed as a robust protocol for qualifying patients for PT.