GTV spatial conformity between different delineation methods by ¹⁸FDG PET/CT and pathology in esophageal cancer

Purpose To find optimal threshold of length and GTV delineation for esophageal cancer using ¹⁸FDG PET/CT. Materials and methods Sixteen patients with esophageal carcinoma underwent surgery. For each patient, six GTVs were defined. GTV_CT was based on CT data alone. GTV_20%, GTV_40%, GTV_2.5 and GTV_40%M were generated by PET/CT, using SUV_bgd + 20%(SUV_max(slice) - SUV_bgd), SUV_bgd + 40%(SUV_max(slice) - SUV_bgd), 2.5 and 40%SUV_max(total) as thresholds. GTV_path was derived from pathology. Lengths of GTVs were recorded as L_CT, L_20%, L_40%, L_2.5 L_40%M and L_path, respectively. The former five GTVs lengths were compared with GTV_pathL_path by means of a conformity index CI/CI′, which is the square of intersection of two GTVs lengths divided by their product. Results Mean L_CT, L_20%, L_40%, L_2.5, L_40%M and L_path were 6.30 ± 2.69, 5.55 ± 2.48, 6.80 ± 2.92, 6.65 ± 2.66, 4.88 ± 1.99 and 5.90 ± 2.38 cm. Mean CI_path^′, CI _20%, ^′, CI_{40% path}, ^′, CI_2.5, ^′ and CI_{40%M path}, ^′ were 0.68 ± 0.16, 0.84 ± 0.17, 0.76 ± 0.14, 0.78 ± 0.15 and 0.80 ± 0.11. CI_20%, _path^′ and CI _{40%M path}^′ was significantly superior to CI _{CT path}^′ (P < 0.05). Mean GTV_CT, GTV_20%, GTV_40%, GTV_2.5, GTV_40%M and GTV_path were 29.16 ± 18.56, 18.75 ± 12.37, 12.52 ± 8.08, 22.69 ± 14.84, 9.18 ± 5.96 and 28.16 ± 17.02 cm³. Mean CIs increased significantly from CI_40%path(0.27 ± 0.09) and CI_40%M&path(0.28 ± 0.08) < CI_20%&path(0.52 ± 0.16) and CI_2.5&path(0.52 ± 0.20) < CI_CT&path(0.77 ± 0.17). Conclusions The SUV_bgd + 20%(SUV_max(slice) - SUV_bgd) method optimally estimated gross tumor length, but only reached an unsatisfactory CI for GTV. Due to possible motion factor enveloped in PET images and lack of histopathologic transverse reference, the information from both PET and CT should be referred to complementarily when delineating GTV.