RESUMO
Objective To investigate the differences in position and volume between planning target volumes (PTV) based on positron emission tomography?computed tomography (PET?CT) images with an standardized uptake value ( SUV) no less than 2?5, 20% of the maximum SUV ( SUVmax ), or 25% of SUVmax , three?dimensional ( 3D ) CT, and four?dimensional ( 4D ) CT in thoracic esophageal cancer. Methods Eighteen patients with thoracic esophageal cancer sequentially received chest 3DCT, 4DCT, and [18F]fluoro?2?deoxy?D?glucose (FDG) PET?CT scans. PTV3D was obtained by conventional expansion of 3DCT images;PTV4D was obtained by fusion of target volumes from 10 phases of 4DCT images. The internal gross tumor volumes ( IGTV) , IGTVPET2.5 , IGTVPET20%, and IGTVPET25%, were generated based on PET?CT images with an SUV no less than 2?5, 20% of SUVmax , and 25% of SUVmax , respectively. These IGTVs were expanded longitudinally by 3?5 cm and radically by 1 cm to make PTVPET2.5 , PTVPET20%, and PTVPET25%, respectively. Results PTV3D was significantly larger than both PTV4D and PTVPET(P=0?000 -0?044), while there was no significant difference between PTV4D and PTVPET ( P= 0?216 -0?633 ) . The mutual degrees of inclusion ( DIs ) between PTV3D and PTV4D were 0?70 and 0?95, respectively, which were negatively correlated with 3D?Vector ( P=0?039). The mutual DIs between PTVPET2.5, PTVPET20%, and PTVPET25% were 0?74, 0?72, 0?78, 0?73, 0?77, and 0?70, respectively, which showed no correlation with 3D?Vector (P=0?150 -0?822). The mutual DIs between PTV3D and PTVPET were 0?86, 0?84, 0?88, 0?63, 0?67, and 0?59, respectively. Conclusions It is difficult to achieve complete volumetric overlap of PTVs based on 3DCT, 4DCT and PET?CT in thoracic esophageal cancer due to different target volume information. PET scan during free breathing should be used with caution to generate PTVs in thoracic esophageal cancer.
