A prospective randomized comparison of curved array and radial echoendoscopy in patients with esophageal cancer.

Background: Both curved array and radial scanning echoendoscopy are used for locoregional staging of cancer arising in the esophagus or cardia. The accuracy of TNM staging of these malignancies by curved array and radial EUS was compared in a prospective, randomized study.

Methods: Patients with cancer of the esophagus or cardia were examined by both curved array and radial echoendoscopy in randomized order by the same endosonographer in an unblinded fashion. The staging results and the examination time for the two echoendoscopies were compared and statistically analyzed, and finally compared with surgical and histopathologic staging.

Results: A total of 104 patients underwent EUS; 36 had surgical resection of the tumor, 26 surgical exploration without resection, and 42 did not undergo surgery. Comparison of the TNM staging results for the two echoendoscopies gave high kappa values (T, 0.77; N, 0.75; M, 0.89), indicating excellent agreement. The accuracy of curved array echoendoscopy and radial echoendoscopy by component of the TNM staging system were, respectively, T, 72% and 73%; N, 70% and 77%; and M, 61% and 57%. Mean procedure time for the curved array and radial examinations was, respectively, 15 and 12 minutes (p<0.01).

Conclusion: There was no clinically relevant difference between curved array and radial echoendoscopy in the staging of cancer of the esophagus or cardia. The choice of echoendoscope for TNM staging in patients with these malignancies is, therefore, a question of personal preference.