NEW YORK (Reuters Health) – In the workup of patients with early esophageal neoplasia, endoscopic ultrasound is unreliable for identifying submucosal infiltration and lymph node involvement — the deciding factors in choosing between endoscopic therapy or surgical resection.

That conclusion comes from a Dutch study reported in Gastrointestinal Endoscopy.

The authors note that endoscopic ultrasound is often used for locoregional staging of early esophageal neoplasia, with the intention of determining lesion depth and identifying suspicious lymph nodes. The information is important because endoscopic therapy — a much less invasive alternative to surgical resection — is an option only when neoplasia is confined to the mucosal layer.

However, they explain, the value of endoscopic ultrasound is questionable when endoscopic resection is performed as final diagnostic step, as it is at their institution.

To investigate, Dr. Roos E. Pouw, at the Academic Medical Center, University of Amsterdam, and colleagues conducted a retrospective study of 131 of their patients with early esophageal neoplasia to see how often esophageal ultrasound changed the management approach.

In 105 of the patients, ultrasound did not show submucosal invasion or suspicious lymph nodes. However, diagnostic endoscopic resection of the lesions showed that 25 of the 105 patients (24%) had submucosal invasion, poor differentiation or lymphovascular invasion, or deep resection margins positive for cancer.

Among the 26 patients with abnormal endoscopic ultrasound findings, 7 were found not to have submucosal infiltration on diagnostic endoscopic resection and subsequently underwent curative endoscopic treatment; another 3 were found not to have risk factors for lymph node metastasis and were treated endoscopically with no signs of recurrence at 30 months.

In summary, Dr. Pouw and colleagues write,