It is conceivable that the carcinomatous and sarcomatous components grow separately from the early stage of the tumors, and that the sarcomatous component forms a protruding tumor mass because it has abundant stroma positive for type IV collagen and laminin.

No abstract

No abstract

A total of 63 surgically resected esophageal carcinomas (including 49 superficial esophageal carcinomas) and histologically normal tissue adjacent to the superficial carcinoma (nontumorous epithelium) were examined immunohistochemically for the blood group antigens (BGA) A, B, H, Lewisa, Lewisb, Lewisx, and Lewisy. Deletion of an expected A, B or H antigen occurred in 12 (24.5%) of the 49 superficial carcinomas and three (21.4%) of the 14 advanced carcinomas. Incompatible expression of an unexpected A or B antigens occurred in only one case (1.6%) in the carcinoma. In the clinicopathologic study, there was a significant correlation between immunoreactivity of Lewisa and depth of cancer invasion (chi-square test, P less than 0.05). In the superficial carcinoma, there were significant correlations between immunoreactivity of Lewisx and lymph node status (chi-square test, P less than 0.05), immunoreactivity of Lewisy and prognosis (Z test, P less than 0.05), and incompatible expression of Lewisb for tumor against nontumorous epithelium and histologic variation (chi-square test, P less than 0.01). The functional significance of alternations in BGA expression that may be associated with oncogenesis is not clear. However, immunohistochemical determination of BGA may be a more advantageous marker to predict the patient's clinical course in superficial esophageal carcinoma.

The therapeutic strategy for superficial esophageal cancer should be determined by the depth of invasion. Therefore, we endeavored to develop reliable criteria for endoscopic evaluation of the invasive depth of superficial esophageal cancer. Two hundred and fifty nine superficial squamous cell carcinoma cases were enrolled in this study, designed to investigate the relationship between endoscopic classification and the depth of invasion, and also between lesion size and depth of invasion. Almost all 0-1, 0-I+O-llc, 0-llc+O-lla and 0-111 lesions invaded the submucosal layer, although 0-lla and 0-llb lesions were limited to the mucosal layer. The 0-llc type of lesion showed a wide range of invasive depths from M1 to SM3 ; in M1 cases with minimal depression the surface appears entirely smooth ; in M2 cases the depressed surface is uniformly granular ; and M3 cases have slightly larger granules. With submucosal cancer, the depressed surface showed some irregularity and variability in granule size. In the 0-llc type, 100% of lesions under 1 0 mm and 80% of those under 30 mm in greatest dimension were mucosal cancer. The accuracy of our criteria for endoscopically evaluating the depth of invasion was 80.2%. The accuracy rate, in terms of assignment to group M1 and M2, M3 and SM1, or SM2 and SM3, was 90.6%. (Dig Endosc 1997;9:110-115) words : endoscopic evaluation of depth of invasion, superficial esophageal cancer, depth of invasion, preoperative diagnosis Manuscript