Person:

Kim, Dae Hyun

CONTEXT: The clinical validity of mucin expression in gastric cancer is debated. Whereas several reports demonstrate a correlation between mucin expression and prognosis, others deny such an association. OBJECTIVES: This survival analysis study aims to elucidate the prognostic significance of mucin expression in gastric cancer. DESIGN: A retrospective survival analysis was done with 412 cases of gastric cancer characterized on the basis of MUC immunohistochemistry using MUC2, MUC5AC, MUC6, and CD10 antibodies; the cases were divided into those with a gastric, an intestinal, or a null mucin phenotype based on the predominant mucin. RESULTS: There was no association between mucin expression and survival when considering overall gastric cancers or the advanced gastric cancer subtype. However, early gastric cancers with a gastric mucin phenotype showed longer survival than those with an intestinal mucin phenotype (P = .01) or a null phenotype (P = .01). In particular, MUC5AC-positive early gastric cancers resulted in longer survival than did those that did not express MUC5AC (P = .009). The loss of MUC5AC expression was identified as an independent, poor prognostic factor in early gastric cancers using the Cox regression proportional hazard model (hazard ratio, 3.50; P = .045). CONCLUSIONS: MUC5AC expression is significantly associated with patient survival and can be used to predict outcomes in the gastric cancers, especially in the early gastric cancers.

After endoscopic resection of early gastric cancer (EGC), it is imperative to accurately determine whether follow-up surgery is indicated, since this technique is used as a first line of treatment. Herein, we developed a scoring system to indicate the risk of lymph node metastasis in submucosal EGC (smEGC), and present a novel method to measure depth of submucosal invasion. In our series, 15.9% of the smEGC presented with lymph node metastasis. A nodal prediction index, based on the variables extracted from the univariate analysis and defined as nodal prediction index = (2.128 × lymphovascular tumor emboli) + (1.083 × submucosal invasion width ≥ 0.75 cm) + (0.507 × submucosal invasion depth ≥ 1000 μm) + (0.515 × infiltrative growth pattern), yielded an area under the receiver operating characteristic curve of 0.809 (P =.000, 95% CI = 0.713-0.096) in a training group, and showed comparable result in validation group (0.886, P =.000, 95% CI = 0.796-0.977). Depth of invasion was statistically higher in the metastatic group when measured from the lowest point of an imaginary line in continuity with the adjacent muscularis mucosa to the point of deepest tumor penetration, but not when using the classic measurement method. The area under the receiver operating characteristic curve of the alternative measurement method was 0.652 (P =.013, 95% CI = 0.550-0.754) compared to 0.620 for the classic measurement method (P =.0480, 95% CI = 0.509-0.731). In deciding whether surgery is indicated after endoscopic submucosal dissection for smEGCs, we recommend to test our alternative method of measuring submucosal invasion and to evaluate our nodal prediction index as an adjunct tool.

BACKGROUND: Given the increasing use of endoscopic resection as a therapeutic modality for cases of early gastric cancer (EGC), it is very important to define strict criteria for the use of endoscopic mucosal resection and endoscopic submucosal dissection. To date, the criteria are almost entirely based on Japanese literature evaluating the risk of lymph node (LN) metastasis in patients with EGC. OBJECTIVE: To analyze our own experience with the factors affecting LN metastasis and to reappraise the extended criteria for endoscopic submucosal dissection. DESIGN: Retrospective, single-center study. SETTING: University teaching hospital. PATIENTS: This study involved 478 patients who underwent gastrectomy with LN dissection (n = 270, mucosal [m] EGC; n = 208, submucosal [sm] EGC). INTERVENTION: Gastrectomy with LN dissection. MAIN OUTCOME MEASUREMENTS: LN metastasis. RESULTS: Overall, 12.6% (60/478) of patients with EGCs presented with LN metastasis (mEGC, 3.0% [8/270], smEGC, 25.0% [52/208]). Increased size, macroscopic type (elevated), depth of invasion, and lymphovascular invasion were associated with LN metastasis. In 270 cases of mEGC, there was no relationship between clinicopathologic features and LN metastasis. In the smEGC group, size, depth of invasion, and lymphovascular emboli were associated with an increased risk of LN metastasis. Significantly, LN metastasis was noted in EGCs falling within established extended endoscopic submucosal dissection criteria, that is, intestinal-type mucosal cancer of any size without ulcer and no lymphovascular emboli (2/146 [1.4%]) or < or =3 cm with no lymphovascular emboli and irrespective of the presence of ulceration (2/126 [1.6%]) or intestinal-type submucosal cancer (sm1, <500 microm) without lymphovascular invasion and measuring < or =3 cm in size (3/20 [15.0%]). LIMITATIONS: Retrospective review of a single-center study. CONCLUSION: We recommend that more centers survey their experiences of LN metastasis in cases of EGC to refine the criteria for endoscopic submucosal dissection as a therapeutic modality of intestinal-type EGC.

Despite wide acceptance of the chronic gastritis-intestinal metaplasia-dysplasia-carcinoma sequence, especially for intestinal-type gastric adenocarcinoma, the precise nature of the subtle precursor lesions of gastric cancer remains to be delineated. For example, pit dysplasia with surface foveolar maturation is not well defined, nor is its prevalence and biological characteristics well characterized. We have evaluated the surrounding gastric mucosa of 414 gastric cancers for the presence of gastric pit dysplasia. We investigated its relationship with various clinicopathological and immunophenotypic features of gastric adenocarcinoma, as well as the severity and extent of any surrounding gastritis and intestinal metaplasia. p53 expression and Ki-67 proliferation index were also evaluated. We have found that 21.0% (n=87) of gastric cancer cases showed pit dysplasia in adjacent gastric mucosa. Gastric cancers with pit dysplasia were significantly associated with older age, male sex, body/fundic location, and intestinal histologic type (P<0.05). Interestingly, gastric mucin-containing intestinal metaplasia (incomplete intestinal metaplasia) was highly associated with adenocarcinoma with pit dysplasia (P=0.000). In addition, MUC6 expression in gastric adenocarcinoma was associated with pit dysplasia (P=0.036). p53 overexpression and increased Ki-67 proliferation index were more evident in gastric pit dysplasia compared with adjacent gastric mucosa. We suggest that gastric pit dysplasia is an important candidate precursor of gastric adenocarcinoma and may represent another morphologic step in the pathogenesis of gastric adenocarcinoma, especially of intestinal type. More detailed prospective studies are needed to determine the precise significance of these findings.