Staging of esophageal carcinoma in a low-volume EUS center compared with reported results from high-volume centers

Background: It is well known that a learning curve exists for performing EUS. Objective: To determine whether the number of EUS investigations performed in a center affects the results of esophageal cancer staging. Design: We compared EUS in the evaluation of T stage and the presence of regional and celiac lymph nodes in a low-volume center where < 50 EUS/endoscopist/y were performed with reported results from 7 high-volume EUS centers. Setting: A reference center for esophageal cancer (> 90 cases/y) but a low-volume center for EUS when it comes to individual endoscopists. Patients: From 1994 to 2003, 244 patients underwent EUS, without specific measures to pass a stenotic tumor or FNA and with postoperative TNM stage as the criterion standard in the low-volume EUS center. In the high-volume centers, 670 EUS investigations for esophageal cancer were performed, if needed, with dilation, and with postoperative TNM stage and/or FNA as the criterion standard. Interventions: Retrospective analysis. Main Outcome Measurements: Sensitivity and specificity of EUS for esophageal cancer staging. Results: In the low-volume center, results of EUS for T3 staging in patients in whom passage of the EUS probe was possible were almost comparable for sensitivity (85% vs 88%-94%) but were lower for specificity (57% vs 75%-90%), whereas both sensitivity (58% vs 75%-90%) and specificity (87% vs 94%-97%) for T1 or T2 stages were lower than those reported in the high-volume centers. in the low-volume center, sensitivities of EUS for regional (45% vs 63%-89%) and celiac (19% vs 72%-83%) lymph nodes were lower, whereas specificities (75% vs 63%-82% and 99% vs 85%-100%, respectively) were comparable with those from high-volume centers. Results in the low-volume EUS center were worse if the EUS probe could not pass the stricture, which occurred in almost 30% of patients. Limitations: Both FNA and dilation before EUS for stenotic tumors were not performed in the low-volume EUS center. Conclusions: The results of EUS performed in a low-volume EUS center compared unfavorably with those reported from high-volume EUS centers. The results of this study suggest that preoperative staging by EUS should be performed by experienced and dedicated EUS endoscopists to optimize staging of esophageal cancer.