Comparison of positron emission tomography, computed tomography, and endoscopic ultrasound in the initial staging of patients with esophageal cancer.
INTRODUCTION: Improvement in esophageal cancer staging is needed. Positron emission tomography (PET), computed tomography (CT), and endoscopic ultrasound (EUS) in the staging of esophageal carcinoma were compared. METHODS: PET, CT, and EUS were performed and interpreted prospectively in 75 patients with newly diagnosed esophageal cancer. Either tissue confirmation or fine needle aspiration (FNA) was used as the gold standard of disease. Sensitivity and specificity for tumor, nodal, and metastatic (TNM) disease for each test were determined. TNM categorizations from each test were used to assign patients to subgroups corresponding to the three treatment plans that patients could theoretically receive, and these were then compared. RESULTS: Local tumor staging (T) was done correctly by CT and PET in 42% and by EUS in 71% of patients (P value > 0.14). The sensitivity and specificity for nodal involvement (N) by modality were 84% and 67% for CT, 86% and 67% for EUS, and 82% and 60% for PET (P value > 0.38). The sensitivity and specificity for distant metastasis were 81% and 82% for CT, 73% and 86% for EUS, and 81% and 91% for PET (P value > 0.25). Treatment assignment was done correctly by CT in 65%, by EUS in 75%, and by PET in 70% of patients (P value > 0.34). CONCLUSIONS: EUS had superior T staging ability over PET and CT in our study group. The tests showed similar performance in nodal staging and there was a trend toward improved distant disease staging with CT or PET over EUS. Assignment to treatment groups in relation to TNM staging tended to be better by EUS. Each test contributed unique patient staging information on an individual basis.