Impact of hybrid FDG-PET/CT on gross tumor volume definition of cervical esophageal cancer: reducing interobserver variation

Intensity-modulated radiation therapy is being increasingly used to treat cervical esophageal cancer (CEC); however, delineating the gross tumor volume (GTV) accurately is essential for its successful treatment. The use of computed tomography (CT) images to determine the GTV produces a large degree of interobserver variation. In this study, we evaluated whether the use of [18F]-fluoro-2-deoxy-D-glucose positron emission tomography (FDG-PET)/CT fused images reduced interobserver variation, compared with CT images alone, to determine the GTV in patients with CEC. FDG-PET/CT scans were obtained for 10 patients with CEC, imaged positioned on a flat tabletop with a pillow. Five radiation oncologists independently defined the GTV for the primary tumors using routine clinical data; they contoured the GTV based on CT images (GTVCT), followed by contouring based on FDG-PET/CT fused images (GTVPET/CT). To determine the geometric observer variation, we calculated the conformality index (CI) from the ratio of the intersection of the GTVs to their union. The interobserver CI was compared using Wilcoxon's signed rank test. The mean (±SD) interobserver CIs of GTVCT and GTVPET/CT were 0.39 ± 0.15 and 0.58 ± 0.10, respectively (P = 0.005). Our results suggested that FDG-PET/CT images reduced interobserver variation when determining the GTV in patients with CEC. FDG-PET/CT may increase the consistency of the radiographically determined GTV in patients with CEC.