Endosonographic examination of gastrointestinal anastomoses with suspected locoregional tumor recurrence

Background: Endoscopic ultrasound is considered one of the best tools for the preoperative staging of esophageal, gastric, and rectal carcinoma. Depending on the individual investigator, the sensitivity of preoperative tumor staging by endosonography of the upper gastrointestinal tract (GEUS) is 80–92% for gastric carcinoma and 86–95% for esophageal carcinoma. However, the sensitivity and specificity of endosonography for the staging of lymph node metastases is less accurate. The accuracy of rectal endosonography (REUS) is ∼90% for tumor assessment and ∼80% for the detection of lymph node metastases. In this study, we address the question of whether endosonography enables the surgeon to distinguish scar tissue, which is rather homogeneous and echo-rich, from changes such as an anastomositis or a locoregional tumor recurrence, which are typically noninhomogeneous and echo-poor. Methods: During a 24-months period, we studied patients enrolled in a special tumor follow-up care program by either upper gastrointestinal (GEUS, n= 37 patients) or rectal endosonography (REUS, n= 49 patients) for exclusion of a locoregional tumor recurrence. In each patient, local tumor recurrence was suspected because of either medical history, clinical examination, or other diagnostic procedures. Results: As in previous studies, our retrospective analysis revealed that endosonography has a high sensitivity in the detection of local tumor recurrences (>90%) for both GEUS and REUS. Conclusion: Endosonography is a highly accurate means of detecting local tumor recurrence. Received: 9 March 1998/Accepted: 9 November 1998