AIM: To evaluate clinicopathologic parameters and the clinical significance related lymphovascular invasion (LVI) by immunohistochemical staining (IHCS) in endoscopic submucosal dissection (ESD). METHODS: Between May 2005 and May 2010, a total of 348 lesions from 321 patients (mean age 63 ± 10 years, men 74.6%) with early gastric cancer (EGC) who met indication criteria after ESD were analyzed retrospectively. The 348 lesions were divided into the absolute (n = 100, differentiated mucosal cancer without ulcer ≤ 20 mm) and expanded (n = 248) indica-tion groups after ESD. The 248 lesions were divided into four subgroups according to the expanded ESD indication. The presence of LVI was determined by factor Ⅷ-related antigen and D2-40 assessment. We compared LVI IHCS-negative group with LVI IHCSpositive in each group. RESULTS: LVI by hematoxylin-eosin staining (HES) and IHCS were all negative in the absolute group, while was observed in only the expanded groups. The positive rate of LVI by IHCS was higher than that of LVI by HES (n = 1, 0.4% vs n = 11, 4.4%, P = 0.044). LVI IHCS-positivity was observed when the cancer invaded to the mucosa 3 (M3) or submucosa 1 (SM1) levels, with a predominance of 63.6% in the subgroup that included only SM1 cancer (P < 0.01). In a univariate analysis, M3 or SM1 invasion by the tumor was significantly associated with a higher rate of LVI by IHCS, but no factor was significant in a multivariate analysis. There were no cases of tumor recurrence or metastasis during the median 26 mo follow-up. CONCLUSION: EGCs of the absolute group are immunohistochemically stable. The presence of LVI may be carefully examined by IHCS in an ESD expanded indication group with an invasion depth of M3 or greater. AIM: To evaluate clinicopathologic parameters and the clinical significance related lymphovascular invasion (LVI) by immunohistochemical staining (IHCS) in endoscopic submucosal dissection (ESD). METHODS: Between May 2005 and May 2010, a total of 348 lesions from 321 patients 63 ± 10 years, men 74.6%) with early gastric cancer (EGC) who met indication criteria after ESD were analyzed retrospectively. The 348 lesions were divided into the absolute (n = 100, differentiated mucosal cancer without ulcer ≤ 20 mm) and expanded The presence of LVI was determined by factor Ⅷ-related antigen and D2-40 assessment. We compared LVI IHCS- negative group with LVI IHCSpositive in each group. RESULTS: LVI by hematoxylin-eosin staining (HES) and IHCS were all negative in the absolute group, while was observed in only the expanded groups. The positive rate of L VI by IHCS was higher than that of LVI by HES (n = 1, 0.4% vs n = 11, 4.4%, P = 0.044). LVI IHCS-positivity was observed when the cancer invaded to the mucosa 3 (M3) or submucosa 1 (SM1) levels, with a predominance of 63.6% in the subgroup that included only SM1 cancer (P <0.01). In a univariate analysis, M3 or SM1 invasion by the tumor was significantly associated with a higher rate of LVI by IHCS, but no factor was significant in a multivariate analysis. There were no cases of tumor recurrence or metastasis during the median 26 months follow-up. CONCLUSION: EGCs of the absolute group are immunohistochemically stable. The presence of LVI may be carefully examined by IHCS in an ESD expanded indication group with an invasion depth of M3 or greater.