Comparative evaluation of static and dynamic 4D dose recalculations in pencil beam scanning proton therapy for oesophageal cancer.

Background and purpose: The robustness of pencil beam scanned (PBS) proton plans to respiratory motion is often assessed in clinical practice by static 4D dose recalculations on selected 4D computed tomography (4DCT) phases. These capture anatomical variation but neglect interplay effects from sequential beam delivery. This study investigates these effects by comparing static and dynamic 4DDC for esophageal cancer patients. Materials and methods: PBS proton plans following the PROTECT trial protocol were created for ten esophageal cancer patients from the open-access DIR-Lab 4DCT dataset. Plan robustness was evaluated by static and dynamic 4DDC, where the static approach accumulated the computed dose in individual 4DCT phases, while dynamic incorporated the temporal delivery sequence to capture interplay effects. The two 4DDCs were compared by their compliance to the dose restrictions for target volumes and organs at risk (OARs). Results: Static 4DDC consistently predicted higher target coverage than dynamic approach. Discrepancies were most pronounced in patients with substantial target motion (≳10 mm). However, dose metrics for the OARs showed high agreement between the two methods. Compliance with the clinical constraint on target coverage (V95% >97 %) was achieved in 100 % and 70 % of static and dynamic 4D recalculations. Rescanning improved the compliance of target coverage to 90 %. Conclusion: Protocol-based static 4DDC tended to overestimate target coverage robustness to respiratory motion. Although differences were minor in most cases, patients with large motion can have significant discrepancies, underscoring the importance of implementing dynamic 4DDC in PBS proton planning for esophageal cancer.