Session Item

Respiratory motion during proton pencil beam scanning (PBS) treatment of oesophageal cancer may compromise target coverage due to geometric misses and interplay effects. We investigate the effect of respiration for two target coverage strategies.

The study included 26 oesophageal cancer patients (pts). All pts had a ten-phase 4DCT for planning (pCT) and at fraction (F) ten (surveillance, sCT). Retrospectively, PBS plans with two oblique posterior fields separated by 30 degrees were created using the mv-phase. Two strategies for robust optimization (RO) were tested: RO (3mm isocentre shifts, 3% density uncertainty) of the CTV (IMPT_RO), and CTV RO combined with coverage of PTV margins 5mm AP/LR, 8mm CC (IMPT_PTVRO). All plans fulfilled V95%_CTV>99.5%. Dose calculations with interplay effects were performed by simulating treatment delivery at one F and distributing the spots into the 4DCT phases where they were delivered. The phase-specific doses were accumulated at the mv-phase (4D interplay dose), mimicking worst case scenario of delivering only one F. Additionally, the full plan was recalculated in all 4DCT phases and accumulated in the mv-phase to emulated motion blurring without interplay effects (4D dose). Deformable dose propagation was used for dose accumulation (MIM Software).

Fig 1 (left panels) shows box plots of V95%_CTV in each of the ten phases for all pts for IMPT_RO and IMPT_PTVRO plans, showing better coverage for the IMPT_PTVRO. V95%_CTV exceeded 99% in all phases in 18 pts (IMPT_RO) and 23 pts (IMPT_ROPTV). In two pts, V95%_CTV<95% for IMPT_RO in at least one phase. Due to inter-fractional anatomical changes, the CTV coverage decreased on sCT leading to only 9 and 17 pts maintaining V95%_CTV>99% in all phases (Fig 1, right panels). The CTV DVHs for pCT and sCT for planned dose, 4D dose and 4D interplay dose for pt 3 in Fig 2 (upper panels) show a clear coverage decrease on sCT due to inter-fractional changes. Furthermore, interplay effects of a single F decreased V95%_CTV. The 4D dose maintained V95%_CTV>99% for all pts (IMPT_ROPTV) and for 24 pts (IMPT_RO). The 4D interplay dose maintained V95%_CTV>99% for only 14 pts(IMPT_ROPTV) and 10 pts(IMPT_RO), but V95%_CTV>95% for 24 and 21 pts. The homogeneity index (HI= (D98%_CTV -D2%_CTV)/Dp) increased when accounting for interplay (Fig 2, lower panels). On pCT, HI is lower for IMPT_ROPTV than for IMPT_RO for all pts when the 4D dose is considered, but when accounting for interplay, 6 pts had the lowest HI for IMPT_RO. In mean HI increased from 0.04 and 0.06 (4D dose) to 0.096 and 0.104 (4D interplay dose) for IMPT_RO and IMPT_ROPTV on pCT. On sCT mean HI for IMPT_PTVRO and IMPT_RO increased from 0.06 and 0.12 (4D dose) to 0.12 and 0.13 (4D interplay dose).

IMPT_ROPTV is more robust toward respiratory motion than IMPT_RO yielding high CTV coverage in all phases. However, for a few pts dose coverage was compromised by respiration and it is important to evaluate the effect of the motion for all patients treated with IMPT.