A statistical comparison of motion mitigation performances and robustness of various pencil beam scanned proton systems for liver tumour treatments.

Abstract

Different scanned proton therapy systems provide different scanning scenarios, directly changing the temporal interference between sequential beam delivery and tumour motion. We aim here to quantify the interplay effects and compare motion mitigation performance among different PBS scanning systems. Using 6 4DCT(MRI) datasets of liver tumours with irregular motions greater than 10 mm, 4D treatments with single- and double-field plans, and assuming various doses and motion mitigation approaches, were simulated for 8 PBS scenarios including spot or raster scanning, layered or volumetric rescanning, gating, constant or varying beam current and cyclotron or synchrotron beam sources. The resulting 4D plans were compared using the homogeneity index (D5-D95 in CTV) and treatment time. Independent of scanning scenario and field dose, neither gating nor rescanning alone could mitigate motion effects completely. Re-gating (rescanning with gating) however was found to be similarly effective for all scanning scenarios, most field doses and both rescan modes, with the difference being mainly in the treatment efficiency. The advantage of cyclotron-based systems together with layer-by-layer beam current variation was demonstrated by the nearly constant treatment time as a function of increased field dose. Independently of PBS scanning dynamics, re-gating is sufficient to achieve acceptable 4D plan quality close to those of the static references.