ESTRO 2024 - Abstract Book

S4229

Physics - Intra-fraction motion management and real-time adaptive radiotherapy

ESTRO 2024

The KIM method was implemented at seven radiotherapy centres in Australia on standard Elekta and Varian linear accelerators. The KIM method converted 2D positions of gold markers from the kilovoltage x-ray system into a real time 6DoF position of the tumour at 5.5 Hz (Elekta) or 7 Hz (Varian) during treatment. Commissioning tests were performed using KIM to measure static and patient-measured liver traces replicated by a 6DoF programmable motion platform 9 supporting a phantom with implanted gold markers. Following KIM commissioning, 28 liver cancer patients were treated with KIM combined with external surrogate guidance (Elekta ABC 10 or Varian RPM 11 ). Patients received treatments in one of four treatment schedules: 3×17 Gy, 5×10 Gy, 4×10 Gy, 5×6.5 Gy either in end-exhale breath-hold (EEBH) (21 patients), deep-inspiration breath-hold (DIBH) (3 patients), or using free-breathing (FB) (4 patients) techniques with a 5 mm PTV margin. During treatment, if the patient motion was outside the KIM tolerance in three consecutive breaths, a couch shift was performed to correct for the motion. The KIM trigger threshold was determined from a pre-treatment motion assessment session and varied from 3 mm up to 15 mm. The external surrogates were used for patient motion monitoring and gating. The KIM localisation accuracy was measured using a kV/MV triangulation method. The gating accuracy was quantified by the percentage of the treatment the tumour motion was outside the treatment tolerance while the treatment beam was ON. The dosimetric accuracy was measured using the motion-included dose reconstruction method 12,13 . The gating accuracy and dosimetric accuracy were assessed for the as-treated KIM combined with external surrogate guidance method, and compared to the standard-of-care scenario assuming external surrogate guidance alone. The treatment time was determined including the entire treatment duration starting from the pre-treatment CBCT. The KIM localisation accuracy was 0.2±0.9 mm (left-right), 0.3±0.6 mm (superior-inferior) and 1.2±0.8 mm (anterior posterior) for translations and -0.1˚±0.8˚ (left-right), 0.6˚±1.2˚ (superior-inferior) and 0.1˚±0.9˚ (anterior-posterior) for rotations. For breath-hold patients, intrafraction translation greater than 3 mm, 5 mm and 10 mm was observed 40% and 12% and 3% of the time. For free breathing patients, the peak-to-peak motion varied across patients ranging from 15 mm up to 35 mm. With KIM+external surrogate, the tumour motion was always within treatment tolerance during treatment beam on; the gating accuracy for the external surrogate alone guidance was 30%±10% for the ABC system (Fig. 1) and 15±5% for the RPM system. Intra-fractional rotation greater than 5˚ was observed 23% of the time. Dosimetric accuracy with KIM + external surrogate showed that all patients received dose within 5% of the plan. In 3 out of 28 patients, >5% dose error would have occurred to the GTVD100 (Fig. 2) and in one patient >5% overdosing to the stomach would have occurred. In 2 out of 24 breath-hold patients, the PTV margin had to be increased to 15 mm due to the irregularity in patient motion reproducibility detected by KIM during treatment. With external surrogate alone, this motion would have remained undetected 60% of the time during those fractions. The treatment time with KIM + external surrogate guidance for EEBH, DIBH and FB patients were 27±12 min, 20±14 min and 14±3 min, respectively. Results: