Interfractional Reproducibility in Pancreatic Position Based on Four-Dimensional Computed Tomography
Purpose: To assess the interfractional positional variation of the pancreas using four-dimensional computed tomography (4D-CT) and to determine the suitable phase of respiration for dose delivery methods to account for pancreatic tumor motion. Methods and Materials: Fifteen patients with pancreatic cancer were enrolled in this study. For each patient, 4D-CT scans were performed at CT simulation and three times during the course of treatment. Regions of interest were set to the intrapancreatic bile ducts as a surrogate for pancreatic position. The centroids of the regions of interest were calculated at end-inhalation and end-exhalation of the respiration phase. The ranges of respiratory motion and interfractional positional variation were evaluated in the left-right (LR), anterior-posterior (AP), and superior-inferior (SI) directions. Results: The medians of respiratory motion were 1.1 mm (range, 0.0-9.8 mm), 1.5 mm (range, 0.0-7.0 mm), and 5.0 mm (range, 0.0-12.5 mm) in the LR, AP, and SI directions, respectively. The means {+-} SDs of the interfractional positional variation at end-inhalation were 0.9 {+-} 5.1 mm (range, -9.2 to 15.6 mm), -1.9 {+-} 3.9 mm (range, -12.8 to 6.4 mm), and -1.3 {+-} 6.9 mm (range, -15.0 to 13.7 mm) and those at end-exhalation were 0.0 {+-} 3.1 mm (range, -7.0 to 5.3 mm), -1.2 {+-} 3.9 mm (range, -11.2 to 6.7 mm), and 0.1 {+-} 3.2 mm (range, -9.9 to 5.1 mm) in the LR, AP, and SI directions, respectively. The SDs of the interfractional positional variation in the LR and SI directions were significantly larger at end-inhalation than at end-exhalation (LR, p < 0.001; SI, p < 0.001). Conclusions: The ranges of respiratory motion during the course of treatment and the interfractional positional variation were not negligible. The interfractional positional reproducibility was higher at end-exhalation than at end-inhalation under free breathing.