Background and purpose: This prospective study sought to determine how the use of combined PET/CT for radiotherapy treatment planning of oesophageal cancer would alter the delineation of tumour volumes compared to CT alone if PET/CT is assumed to more accurately represent true disease extent.
Patients and methods: All patients underwent FDG-PET/CT scanning in the radiotherapy treatment position. For each patient, two separate gross tumour volumes (GTV) were defined, one based on CT images alone (GTV-CT) and another based on combined PET/CT data (GTV-PET). Corresponding planning target volumes (PTV) were generated, and separate treatment plans were then produced. For each patient, volumetric analysis of GTV-CT, PTV-CT and GTV-PET was performed to quantify the proportion of PET-avid disease that was not included in the GTV and PTV (geographic miss) if CT data alone were used for radiotherapy planning. Assessment of the cranial and caudal extent of the primary oesophageal tumour as defined by CT alone vs PET/CT was also compared.
Results: The addition of PET information altered the clinical stage in 8 of 21 eligible patients enrolled on the study (38%); 4 patients had distant metastatic disease and 4 had unsuspected regional nodal disease. Sixteen patients proceeded to the radiotherapy planning phase of the study and received definitive chemoradiation planned with the PET/CT data set. The GTV based on CT information alone excluded PET-avid disease in 11 patients (69%), and in five patients (31%) this would have resulted in a geographic miss of gross tumour. The discordance between CT and PET/CT was due mainly to differences in defining the longitudinal extent of disease in the oesophagus. The cranial extent of the primary tumour as defined by CT vs PET/CT differed in 75% of cases, while the caudal extent differed in 81%.
Conclusions: This study demonstrates that if combined PET/CT is used for radiotherapy treatment planning, there may be alterations to the delineation of tumour volumes when compared to CT alone, with the potential to avoid a geographic miss of tumour.