Improved detection of individual nodal involvement in squamous cell carcinoma of the esophagus by FDG PET

Because both the number and location of metastatic lymph nodes and the N st age influence survival in esophageal cancer, accurate noninvasive evaluatio n of individual lymph node groups for the presence of metastasis is essenti al for therapeutic planning. Therefore, we investigated the accuracy of FDG PET for evaluating individual lymph groups in esophageal cancer patients a nd compared the results with those of CT and endoscopic sonography (ES). Me thods: Sixty-one consecutive patients with histologically proven primary es ophageal carcinoma were studied prospectively with FDG PET. Thirteen patien ts who were treated nonsurgically were excluded from data analysis. The rem aining 48 patients underwent esophagectomy and lymph node dissection. All 4 8 patients underwent CT scanning, including the lower neck, thorax, and upp er abdomen, with intravenous administration of contrast medium. ES was perf ormed in 45 of the patients but was incomplete in 12 patients because of es ophageal stenosis. The accuracies of FDG PET, CT, and ES were compared with histologic findings. Results: During surgery, a total of 382 lymph node gr oups were dissected in 48 patients, of which 100 node groups in 32 patients were malignant on histologic examination. For assessing metastasis to indi vidual groups, FDG PET showed 57% sensitivity, 97% specificity, and 86% acc uracy, whereas CT showed 18% sensitivity (P < 0.0001), 99% specificity (P = 0.033), and 78% accuracy (P = 0.003). For N staging, FDG PET was correct i n 83% (40/48) of the patients, whereas CT and ES were correct in 60% (29/48 ; P = 0.006) and 58% (26/45; P = 0.003), respectively. Conclusion: FDG PET is more accurate than is CT or ES for evaluating metastasis to individual l ymph node groups and for N staging in esophageal cancer and thus may be hel pful in determining the therapeutic plan.