Reproducibility of liver position using active breathing coordinator for liver cancer radiotherapy

Purpose: To measure the intrabreath-hold liver motion and the intrafraction and interfraction reproducibility of liver position relative to vertebral bodies using an active breathing coordinator (ABC) in patients with unresectable liver cancer treated with hypofractionated stereotactic body radiation therapy (SBRT). Methods: Tolerability of ABC and organ motion during ABC was assessed using kV fluoroscopy in 34 patients. For patients treated with ABC, repeat breath-hold CT scans in the ABC breath-hold position were acquired at simulation to estimate the volumetric intrafraction reproducibility of the liver relative to the vertebral bodies. In addition, preceding each radiation therapy fraction, with the liver immobilized using ABC, repeat anteroposterior (All) megavoltage verification images were obtained. Off-line alignments were completed to determine intrafraction reproducibility (from repeat images obtained before one treatment) and interfraction reproducibility (front comparisons of the final image for each fraction with the AP) of diaphragm position relative to vertebral bodies. For each image set, the vertebral bodies were aligned, and the resultant craniocaudal (CC) offset in diaphragm position was measured. Liver position during ABC was also evaluated from kV fluoroscopy acquired at the time of simulation, kV fluoroscopy at the time of treatment, and front MV beam's-eye view movie loops acquired during treatment. Results: Twenty-one of 34 patients were screened to be suitable for ABC. The average free breathing range of these patients was 13 mm (range, 5-1 mm). fluoroscopy revealed that the average maximal diaphragm motion during ABC breath-hold was 1.4 mm (range, 0-3.4 mm). The MV treatment movie loops confirmed diaphragm stability during treatment. For a measure of intrafraction reproducibility, an analysis of 36 repeat ABC computed tomography (CT) scans in 14 patients was conducted. The average mean difference in the liver surface position was -0.9 mm, -0.5 mm, and 0.2 mm in the CC, AP, and medial-lateral (ML) directions, with a standard deviation of 1.5 mm, 1.5 mm, and 1.5 mm, respectively. Ninety-five percent of the liver surface had an absolute differences in position between repeat ABC CT scans of less than 4.1 nun, 3.3 turn, and 3.3 rum in the CC, AP, and ML directions, respectively. Analysis of 257 MV All images from patients treated using ABC revealed an average intrafraction CC reproducibility (sigma) of diaphragm relative to vertebral bodies of 1.5 mm (range, 0.6-3.9 rum). The average interfraction CC reproducibility (sigma) was 3.4 mm (range, 1.5-7.9 mm), indicating less day-to-day reproducibility of diaphragm position relative to vertebral bodies. The average absolute intra and interfraction CC offset in diaphragm position relative to vertebral bodies was 1.7 and 3.7 mm, respectively, with 86% of intrafraction and 54% of interfraction absolute offsets 3.0 mm or less. Conclusions: Intrafraction reproducibility of liver position using ABC is good in the majority of screened patients. However, interfraction reproducibility is worse, suggesting a need for image guidance. (C) 2006 Elsevier Inc.