Prospective comparison of optic versus blind endoscopic ultrasound in staging esophageal cancer

Abstract

Esophageal tumours too stenotic to cross with optic endoprobe ultrasound (EUS) may still be staged with the blind endoprobe of 9 mm diameter. The aim of this study was to determine the relative accuracy of both optic and blind endoprobe-defined radiological stages when compared with the histopathological pTN stages. Sixty-seven patients [8 squamous cell carcinoma (SCC), 59 adenocarcinoma (ACA)] with tumours too stenotic to allow optic endoprobe assessment underwent blind endoprobe examination and were compared with 146 patients (48 SCC, 98 ACA) undergoing optic endoprobe assessment. The strengths of agreement between the EUS stage and the histopathological stage were determined by the weighted kappa statistic (Kw). Tumour dilatation was required in 3 (2%) of the patients undergoing optic EUS compared with 20 (30%) of the patients undergoing blind EUS (p = 0.0001). Optic EUS T-stage Kw was 0.612 [95% confidence interval (CI) 0.553-0.671, p = 0.0001] compared with 0.530 (0.426-0.634, p = 0.0001) for blind EUS. Optic EUS N-stage Kw was 0.639 (0.576-0.702, p = 0.0001) compared with 0.666 (0.565-0.737, p = 0.0001) for blind EUS. Patients undergoing blind probe EUS were more likely to have advanced tumour stage than patients undergoing optic probe EUS (p = 0.005). Blind probe EUS facilitated complete radiological staging in 31% of cases that would otherwise have resulted in a designation of failure to cross at EUS, and was as accurate as optic probe EUS in assessing pTN stage.