Background: Repeat endoscopic mucosal resection (EMR) to cure locally recurrent early gastric cancer (EGC) is difficult to perform because the initial EMR causes submucosal fibrosis; however, ESD allows submucosal dissection through the fibrosis and provides an en bloc specimen.
Objective: The objective was to determine the safety and efficacy of ESD and compare the results to historical controls.
Design: This was a retrospective, case-control study.
Patients: We used our prospectively entered database (1993-2003) to identify 64 patients who had locally recurrent EGC after EMR.
Main outcome measurements: We defined the resections as curative when the lateral and vertical margins were free of cancer and repeat endoscopy showed no recurrent disease.
Results: Among 46 patients who underwent ESD, 41 (89.1%) en bloc resections were achieved compared to none in 18 conventional procedures (P < .0001). The specimen of 1 lesion (2.4%) out of 41 en bloc resections was histologically nonevaluable, compared with 10 lesions (43.4%) in 23 piecemeal resections (P < .0001). Three residual tumors (27.3%) were found in the 11 nonevaluable specimens. Three perforations occurred during ESD; all were successfully treated endoscopically with endoclips.
Limitations: The limitation of the study was retrospective design.
Conclusions: ESD provides high en bloc resection rate for locally recurrent EGC after previous EMR. In turn, en bloc resections allow precise histological staging to be assessed and prevent residual disease and recurrence.