Unexpected discrepancy between the planned gating window and the actual irradiated gating window in phase-based gating radiotherapy during regular respiration

We observed unexpected gating window discrepancies in asymmetric respiratory patterns. Therefore, the present study was conducted to test the hypothesis that asymmetric breathing causes this discrepancy. Six regular respiratory patterns were generated using the QUASAR software. 4-dimensional computed tomography (4DCT) was acquired with each breathing pattern using a QUASAR respiratory motion phantom with a 1.2 cm spherical acrylic ball embedded in a cedar insert. A Real-time Position Management system was used for respiratory signal acquisition during the 4DCT. Each treatment plan was created using gating windows ranging from 30% to 70%. TrueBeam STx was used for gated beam delivery, and the stored waveform data files and gated kV images were analyzed after irradiation. Two different concepts of phase were used to analyze the waveform data files. The respiratory phase defined prospectively by the TrueBeam gating system during treatment was called "phasepro", and the phase defined retrospectively by waveform data analysis after treatment was called "phaseretro". The phasepro values were close to the planned gating window (30% or 70%) for in all breathing patterns. However, the phaseretro values differed from the planned gating window in the asymmetric patterns. The gating window was either pulled forward or pushed back from the treatment plan depending on the direction of the asymmetry. Furthermore, this discrepancy increased as the asymmetry of the respiratory pattern increased. In some asymmetric breathing models, the tumor location on the gated kV image was outside the internal target volume contour. This study confirmed that the gating window discrepancy between 4DCT and the TrueBeam gating system occurs with asymmetric respiration pattern, even with regular breathing. Therefore, if a patient's breathing is not exactly symmetrical, appropriate measures should be taken against this unexpected systematic error.