Value of high-frequency miniprobes and conventional radial endoscopic ultrasound in the staging of early Barrett's carcinoma: a prospective randomized trial with a cross-over design

Aims: High-frequency miniprobes (HFPs) and conventional radial endoscopic ultrasound (crEUS) are considered to be valuable tools in the staging of early esophageal carcinomas. However, there is some controversy on whether HFPs are superior in the T staging of early esophageal neoplasias or whether the same level of accuracy can be achieved by the sole use of crEUS.

Objective: To compare the accuracy of T staging and N staging with HFP and crEUS.

Design: Prospective randomized trial (ratio 1:1) with a cross-over design.

Methods: Patients (Pts) referred for endoscopic treatment for early Barrett's carcinoma (BC) were prospectively included and randomly assigned to either HFP (Fujinon, 20MHz) or crEUS (Pentax EG-3630) examinations as the initial diagnostic method. Afterwards, all of the pts were reexamined with the alternative procedure. The examination time and quality of each method with regard to T and N staging were documented using three grades (good, intermediate, or poor). The staging results with each method were also documented prospectively.

Results: A total of 43 pts (mean age 64.7 years; 34 men, nine women) were included. Assessment of the T category was not possible with HFP in 7% of pts (3/43) or with crEUS in 33% of pts (14/43)(p=0.004), due to positioning problems. A total of 22 mucosal BC and 17 submucosal BC were confirmed at histology (ER or surgical specimen). Histological confirmation was not possible in 4 pts. The T category was correctly assessed with HFP in 56% (22 of 39) and with crEUS in 44% (17 of 39). Histological confirmation of the lymph-node staging was possible in 11 pts after surgery; 27% (3/11) had positive lymph nodes. N staging was correct in 91% (10/11). The subjective quality of the assessment of T stage was significantly better with HFP than with crEUS (p=0.004). In contrast, crEUS was highly superior in assesing the lymph node status.

Conclusions: The accuracy of assessment of the T category is unsatisfactory with both HFP and crEUS in pts with early BC. N staging can only be provided by crEUS, and had a high level of accuracy in the present series. With crEUS, lesions can be significantly evaluated with adequate quality less frequently than with HFP.