Radiotherapy volume delineation using 18F-FDG-PET/CT modifies gross node volume in patients with oesophageal cancer.

PURPOSE: Evidence supporting the use of 18F-FDG-PET/CT in the segmentation process of oesophageal cancer for radiotherapy planning is limited. Our aim was to compare the volumes and tumour lengths defined by fused PET/CT vs. CT simulation.

MATERIALS AND METHODS: Twenty-nine patients were analyzed. All patients underwent a single PET/CT simulation scan. Two separate GTVs were defined: one based on CT data alone and another based on fused PET/CT data. Volume sizes for both data sets were compared and the spatial overlap was assessed by the Dice similarity coefficient (DSC).

RESULTS: The gross tumour volume (GTVtumour) and maximum tumour diameter were greater by PET/CT, and length of primary tumour was greater by CT, but differences were not statistically significant. However, the gross node volume (GTVnode) was significantly greater by PET/CT. The DSC analysis showed excellent agreement for GTVtumour, 0.72, but was very low for GTVnode, 0.25.

CONCLUSIONS: Our study shows that the volume definition by PET/CT and CT data differs. CT simulation, without taking into account PET/CT information, might leave cancer-involved nodes out of the radiotherapy-delineated volumes.