RESUMO
BACKGROUND AND STUDY AIMS: The high-frequency endosonography miniprobe has been advocated prior to endoscopic treatment of superficial esophageal neoplasm. We conducted a retrospective study, which summarized our clinical experience with various miniprobe techniques in the assessment of early squamous cell carcinoma (SCC) and superficial adenocarcinoma on Barrett's mucosa (SAB). PATIENTS AND METHODS: A total of 106 superficial and naive esophageal lesions were included in this series (52 SAB, 54 SCC). The results of the miniprobe staging were compared with pathology in order to determine the accuracy of this technique in differentiating mucosal from submucosal infiltration. Secondary aims were a comparison between the frequencies (20- vs. 30-MHz) and coupling methods (water-filled lumen vs. balloon-sheathed catheter) used. RESULTS: Overall accuracy, sensitivity, and specificity to differentiate T1sm from T1 m tumors were 73.5 %, 62 %, and 76.5 %, respectively. Overdiagnosis occurred in 18.6 % of the lesions. No statistically significant difference was found in the accuracy of the miniprobe staging between the two lesion types (SCC and SAB) ( P = 0.65), and results were significantly poorer in the distal part of the esophagus ( P = 0.00039). No significant difference was found between the two frequencies ( P = 0.51). Water-filled lumen proved to be superior to the balloon-sheathed catheter ( P = 0.015). CONCLUSION: High-frequency miniprobe examination has a limited accuracy in the detection of submucosal invasion in early esophageal cancer. Further improvements in acoustic coupling and ultrasound technique are required to improve the miniprobe accuracy before its implementation into clinical routine.