Tumour length measured on PET-CT predicts the most appropriate stage-dependent therapeutic approach in oesophageal cancer.

To compare the accuracy of determining the most appropriate treatment approach based on a visual analysis on combined PET-CT, based on a visual analysis on PET (reviewed side-by-side with CT) and based on tumour length measurements (on PET and PET-CT). Tumour length, SUV, and the length-SUV index (length x SUV) were assessed (on PET and PET-CT) in benign oesophageal lesions (reflux oesophagitis; n = 29), in potentially curable stages of oesophageal cancer (Tis; T1-T3NxM0; curable group; n = 52), and in stages of oesophageal cancer best treated with palliative therapy (T4NxMx; TxNxM1; palliative group; n = 30). All lesions were histopathologically proven. Based on a visual analysis, PET-CT (sensitivity: 77%;specificity: 96%) was more accurate than PET (sensitivity: 67%; specificity: 89%) in assessing the appropriate treatment (curative vs. palliative). The length-SUV index was the most accurate quantitative parameter to distinguish palliative from curable stages (sensitivity: 93%; specificity: 90%) and to predict survival. The highest overall accuracy was reached when combining the results of the quantitative (length-SUV index) analysis with those of the qualitative (visual) analysis (sensitivity: 93%; specificity: 96%). Moreover, neither tumour length nor SUV can be used to distinguish reflux oesophagitis from early malignant lesions (T1 stage). Tumour length measured with PET-CT or PET is associated with stage and overall survival of oesophageal cancer and helps to guide the appropriate treatment approach.