Small lung tumors: long-scan-time CT for planning of hypofractionated stereotactic radiation therapy--initial findings.

PURPOSE: To prospectively use long-scan-time computed tomography (CT) to visualize the trajectory of tumor movements or the internal target volume.
MATERIALS AND METHODS: The study was approved by the institutional review board. Written informed consent was obtained from participants after the study and the role of procedures were explained fully. During the planning of stereotactic radiation therapy for 10 patients (nine men, one woman; mean age, 77 years; range, 69-89 years) with small lung tumors (mean volume, 9.0 cm3; range, 3.6-24.9 cm3), fluoroscopic imaging, long-scan-time CT, and thin-section CT were performed. The tumor and the partial-volume-averaging effects that resulted from tumor movement were delineated on each section at long-scan-time CT performed during the patient's steady breathing with scan time of 8 seconds per image. Visualized internal target volume was defined by integrating the sections. A simple model was examined for estimating internal target volume on the basis of respiratory motion and gross target volume delineated on thin-section CT images. Visualized internal target volume and estimated internal target volume were compared quantitatively and graphically. The Mann-Whitney test was used to analyze the relation between gross target volume delineated on thin-section CT images and the ratio of visualized internal target volume to the defined gross target volume.
RESULTS: The correlation coefficient between visualized internal target volume and estimated internal target volume was r = 0.98 (P < .001). The mean relative error +/- standard deviation was 1.9% +/- 19.0 (range, -11.0% to 56.4%). Excluding one case with an irregularly shaped tumor (56.4%), the mean relative error was -4.1% +/- 4.1. In patients with small tumors (defined gross target volume, < or = 10 cm3), the ratio of the visualized internal target volume to the defined gross tumor volume was significantly larger than that in patients with larger tumors (1.2-2.0 vs 1.0-1.2; P < .05). In some cases in which marginal spiculation depicted on thin-section CT images was blurred on long-scan-time CT images, the blurred area was erroneously excluded from the target volume.
CONCLUSION: In most cases, values for visualized internal target volume and estimated internal target volume were similar and long-scan-time CT depicted virtually the entire tumor trajectory. RSNA, 2005