To the Editor

Dear Editor,

With regard to the paper published recently in J. Gastroenterol. entitled “Risk of metastasis in adenocarcinoma of the esophagus: a multicenter retrospective study in a Japanese population” by R. Ishihara et al. [1], I consider the methodology employed to have been fundamentally flawed in view of the fact that the dataset they analyzed was derived from both endoscopically resected (ER) and surgically resected specimens. The authors state that the ER specimens were cut into slices 2 mm thick, whereas the surgical specimens were cut into slices 5 mm thick. This difference in the thickness of the slices obtained would have led to obvious differences in the assessment of cancer invasion depth. The degree of lymphatic and venous invasion would also have differed for this reason. I consider that the authors should have analyzed the ER and surgical specimens as two independent groups.

In this paper, the authors considered sm1 adenocarcinoma of the esophagus to be cancer with an invasion depth of between 1 and 500 μm in the submucosa. On the basis of the Japanese Classification, Japanese pathologists usually consider sm1 squamous cell carcinoma to be cancer with an invasion depth of between 1 and 200 μm in the submucosa [2]. Based on the difference in depth of sm1 between adenocarcinoma and squamous cell carcinoma, the authors maintain that carcinomas of different histologic types require a different definition of sm1. The Japanese Classification [2] of malignant neoplasms includes more than 10 well-established histologic types, including basaloid squamous carcinoma, carcinosarcoma, neuroendocrine carcinoma, and malignant melanoma. Do clinicians need more than 10 definitions of sm1 cancer in the esophagus based on histologic type?