Room: AAPM ePoster Library
Purpose: To quantify the impact of breathing pattern variations on the implementation of respiration-gated treatment for gastrointestinal cancer.
Methods: Breathing patterns based on external surrogates were collected during four-dimensional computed tomographic simulation scans and respiration-gated treatment sessions from twelve patients with pancreatic cancer and one patient with bile duct carcinoma. Nine patients were treated in 25 fractions and four in 15 fractions. The Varian Real-time Position Management (RPM) system was used to monitor the breathing, represented by the displacement of an external surrogate, namely a plastic block with reflective markers placed on the patient’s abdomen. The respiration-gated treatment was centered at end-expiration with ~40% duty cycle (DC). Amplitude-based gating was used for all patients, with the threshold determined from simulation and adjusted during treatment when necessary. The actual DC and gated motion (GM) were calculated from the motion management waveform files recorded at the treatment console. The DC and GM achieved with fixed thresholds determined from simulation or the first treatment session was retrospectively evaluated. Also evaluated was the GM from phase-based gating with exact DC of 40%.
Results: The actual DC and GM achieved were 50% and 4.6 mm, respectively. The DCs for seven patients (54%) were within 30—50%. The simulation-determined threshold, if unadjusted, would result in DCs beyond 30—50% for ten patients (77%). With the threshold determined at the first treatment session and fixed thereafter, the average DC was 40% and within 30—50% for ten patients (77%); the average GM was 4.4 mm. Phase-based gating resulted in 9 mm average GM.
Conclusion: In the presence of breathing pattern variations between simulation and treatment and between different treatment days, the RPM software was more robust with amplitude-base gating than with phase-based gating. The amplitude threshold was better determined during the first treatment than from simulation.