We thank Dr. Fatourou and Roukos for their interest and remarks in response to our article regarding endoscopic resection for undifferentiated early gastric cancer [1].

Since 1999, a national cancer screening program has recommended upper gastrointestinal endoscopy to the normal population older than 40 years for early detection of gastric cancer in Korea. The proportion of early gastric cancer indicated for endoscopic resection has increased. Although conventional endoscopic resection was limited to complete resection only for large or deep tumors confined to the mucosa without the risk of lymph node metastasis, newly developed endoscopic submucosal dissection has enabled complete resection of larger and deeper tumor without any technical barrier.

With the progress of technical advances, Japan has proposed expanding the indications for endoscopic resection of early gastric cancer [2]. Although tumor of undifferentiated histology has the risk of much larger size and deeper invasion than expected, some portion of an early lesion could possibility be indicated for complete endoscopic resection as a minimally invasive treatment.

In a retrospective study with postoperative pathologic review, undifferentiated early gastric cancer smaller than 2.5 cm and confined to the mucosa did not show any lymph node metastasis in a large series [3]. As expected, the complete resection rate for undifferentiated early gastric cancer was significantly lower than for differentiated histology, which explains the possibility of diffuse infiltration of tumor with undifferentiated histology and the risk of lymph node metastasis.

Although endoscopic submucosal dissection has improved the complete resection of early gastric cancer more than conventional mucosal resection, expanding the indication to tumor of undifferentiated histology has the risk of incomplete resection or lymph node metastasis and should be accompanied by long-term follow-up evaluation. We think the indication for endoscopic resection of undifferentiated gastric cancer should be limited to small tumors confined to the mucosa.