Comparison Of The Reproducibility Of Tumor Positions In The Liver And Upper Abdomen With Those In The Lung With Exhalation Breath-Holding Methods

One of the problems of the breath-holding irradiation method is the reproducibility of tumor position during planning and treatment. Movement of abdominal lesions is influenced by the volume and peristalsis of the gastrointestinal tract as well as breathing. The purpose of this study was to compare the reproducibility of tumor position in the liver and upper abdomen with that in the lung under exhalation breath hold. Eight lung tumors, 7 liver tumors, and 6 other upper abdominal tumors treated during the same period were included. The upper abdominal tumors included 2 pancreatic cancers, 1 lymph node metastasis, 1 reconstructed stomach cancer and 2 adrenal gland metastases. No fiducial marker was used in the lung lesions. Embolization coils were used as fiducial markers in all 7 lesions of the liver. Embolization coils were used in 4 lesions and a linear fiducial marker was used in 2 of the upper abdominal lesions. Planning CT and CBCTs were performed under exhalation breath hold using the Real-Time Position Management Respiratory Gating System. CBCTs were performed before each treatment. The centers of the lung tumors, and those of embolization coils and fiducials in the liver and upper abdominal lesions were recorded on planning CT and 4 sessions of CBCTs for each case. On bony structure matching, the differences between the centers on planning CT and those on 4 CBCTs were calculated, and compared among the lung, liver and upper abdominal lesions. The differences in the left and right, anterior and posterior, and superior and inferior directions were up to 3.2 and 3.0 mm, 4.6 and 2.6 mm, and 4.1 and 8.1 mm, respectively, in the lung lesions, up to 4.8 and 2.9 mm, 6.4 and 2.9 mm, and 3.7 and 6.7 mm, respectively, in the liver lesions, and up to 13.3 and 6.6 mm, 9.4 and 7.8 mm, and 5.7 and 10.1 mm, respectively, in the upper abdominal lesions. The differences over 5 mm were observed once in the inferior direction in the lung lesions, once in the anterior and once in the inferior direction in the liver lesions, and five times in the left, once in the right, twice in the anterior and the posterior, and twice in the inferior and once in the superior directions in the upper abdominal lesions. Between the liver and lung lesions, there was a significant difference in the LR direction (p=0.042). Between the upper abdominal and lung lesions, there was a significant difference in the LR direction (p=0.00092). No significant differences were shown in all directions between the liver and upper abdominal lesions. The exhalation breath-holding method on bony structure matching is insufficient from the standpoint of the reproducibility of tumor position in the liver and upper abdomen compared with that in the lung. Marker matching methods with appropriate margins might be necessary to improve dose distribution to the tumors and circumjacent organs.