• Journal of Gastric Cancer
• /
• v.9 no.4
• /
• pp.177-185
• /
• 2009

Purpose: Preoperative clinical staging of gastric cancer is very important for determining the treatment plans and predicting the prognosis. The previous reports regarding the accuracy of computed tomography or endoscopic ultrasound for the preoperative staging of gastric cancer have shown various outcomes. We analyzed the diagnostic performance of CT and EUS, which are important staging tools for the staging of TN gastric cancer. Materials and Methods: We retrospectively analyzed 1,174 patients who underwent gastrectomy for gastric cancer at Seoul National University Bundang Hostpital from May, 2003 to December, 2007. We derived the Kappa value to examine the agreement of the preoperative staging obtained from CT and EUS with the pathological staging. Results: The mean age of the 1,174 patients was $59.31{\pm}11.98$ years. Six hundred thirty seven patients had early gastric cancer and 536 had advanced gastric cancer. The diagnostic performance between CT and EUS for the T staging showed no significant difference between CT and EUS for the kappa values. The kappa values showed moderate agreement at 0.4039 (P=0.021) and 0.4201 (P=0.026), respectively. This suggests that there is no difference between the two examinations for the overall T staging. Analysis of the discrimination of mucosal and submucosal lesions with EUS showed an accuracy of 58.92% and a Kappa value of 0.206 (P<0.001), suggesting fair agreement and a lower diagnostic performance than expected. To differentiate lesions with stages higher than or equal to T2 or T3 from the lesion with stages lower than T2 or T3, respectively, adoption of the higher stage from the CT staging or the EUS staging showed a larger AUC of 0.84 than that from either stage alone. The CT-derived node stage had the higher diagnostic performance (68.55%) than that of the EUS-derived node stage (60.82%) for the node staging. Conclusion: The CT-derived stage and EUS-derived stage showed comparable results for determining the T stage of gastric cancer. Yet the higher stage of the two stages from CT and EUS most accurately discriminated between those lesions with stages higher than T2 and those lesions with stages lower than T2.

• Journal of Gastric Cancer
• /
• v.5 no.1
• /
• pp.40-46
• /
• 2005

Purpose: Gastrointestinal stromal tumors (GISTs) are mesenchymal neoplasms of the gastrointestinal tract. GISTs are positive for the expression of c-Kit protein at immunohistochemistry, and their clinical presentations vary. This retrospective study was performed to evaluate the clincopathologic characteristics of GISTs and to define the prognostic factors. Materials and Methods: 40 patients who underwent a complete resection of a GIST during the period $1996\~2003$ at the Department of Surgery, Korea University College of Medicine, were studied. We divided them into low- and high-risk. groups by using tumor size and mitotic count: 23 cases were low risk, and 17 were high risk. Clinicopathologic features, immunohistochemical findings, and prognoses were compared between the low- and the high-risk groups. Results: The mean age of the 40 patients was $61.3\pm11.1$years, and the male-to-female ratio was 1：1.1. There was no significant difference in age and sex between the groups. A comparative analysis revealed tumor size, mitotic count, clinical symptoms, preoperative pathologic diagnosis, ulceration, and necrosis to be variables that had statistically significant differences between the high- and the low-risk groups. In the univariate analysis, tumor size, mitotic count, ulceration, necrosis, and abnormal endoscopic ultrasound findings were associated with disease-free survival, but in the multivariate analysis, mitotic activity was the only independent factor associated with disease-free survival. 8 patients had recurrences during the follow-up period, and four of them were treated with STI-571 (imatinib mesylate, $Gleevec^{(R)}$). The treated patients have survived until now; however, two of non-treated patients died from disease progression. Conclusion: Based on this study, tumor size, ulceration, and necrosis are significant factors affecting survival, and mitotic activity may be a useful prognostic marker. STI-571 may be used in an adjuvant setting because the drug has shown anticancer activity in patients with recurrence or metastasis.