AIM: To investigate the correlation among tumor markers, curative resection, and recurrence in gastric cancer. median follow-up period was 32.4 mo. Forty-three patients (6.3%) received non-curative operations. The main causes of a non-curative operations were peritoneal dissemination (= 21) followed by direct invasion (= 16) (Table ?(Table1).1). Risk factors for non-curative operation were tumor location, CEA, CA 19-9, and CA 125 in univariate analysis. In a multivariate analysis, area of tumor (HR = 21.303; < 0.001), CA 19-9 positivity (HR = 5.883; < 0.001), and CA 125 positivity (HR = 15.549; < 0.001) were individual risk factors to get a non-curative procedure. (Desk ?(Desk2)2) Recurrences after curative procedure were 124 situations among 636 sufferers, and 64 sufferers had several recurrence site; hematogenous (= 72), peritoneal (= 50), loco-regional (= 46) and faraway lymph node metastases (= 41). The 5-season disease-free survival price after curative procedure was 77.9%, as well as the 5-year disease-free survival rate of CEA, CA 19-9 and CA 125 are 51.3%, 51.7% and 30.8%, respectively (Body ?(Figure1).1). Risk elements of buy 850717-64-5 recurrence had been tumor area, differentiation, lymphovascular invasion, perineural invasion, stage, and CEA, buy 850717-64-5 CA 19-9, and CA 125. Within a multivariate evaluation, gender, stage, and positivity of CA 125 (HR = 2.431; = 0.020) were individual risk elements for recurrence (Desk ?(Desk33). Desk 1 Causes to get a non-curative operations Body 1 Disease free of charge survival curve regarding to positivity of Carcinoembryonic antigen (A), Carbohydrate antigen 19-9 (B) and Carbohydrate antigen 125 (C). Desk 2 Univariate and multivariate evaluation of the chance factors to get a non-curative functions (%) Desk 3 Univariateand multivariate evaluation of prognostic risk elements for disease-free success after curative functions Dialogue Depth of invasion and lymph node metastasis were most important impartial prognostic risk factors in gastric cancer, and currently used TNM stage is useful to predicting survival rate in each stage. However, it was difficult to estimate N stage, to predict survival rate in each stage and to decide treatment of modality in preoperative clinical TNM stage. Many prognostic risk factors of gastric cancer were investigated, and tumor marker is one of the prognostic factors of gastric cancer. CEA and CA 19-9 are the most common tumor markers for predicting prognosis in gastric cancer[14]. The sensitivities of CEA and CA 19-9 in gastric cancer are 16%-58.4% and 34.1%-64.9%[15-17]. Preoperative positivity for CA and CEA 19-9 is usually associated with an unhealthy prognosis[4,5,18], but there's a few record for preoperative CA 125 in gastric tumor. The preoperative CA and CEA 19-9 weren't prognostic aspect for palliative gastric medical procedures, nonetheless it was indie prognostic elements in curative medical procedures[4]. The postoperative CEA positivity in early gastric tumor and postoperative CEA and CA 72-4 positivity in advanced gastric tumor had been indie prognostic elements[18]. Inside our study, the preoperative CEA and CA 19-9 weren't indie prognostic risk elements for recurrence as various other research[19], interestingly preoperative CA 125 was an independent risk factor for recurrence with a HR of 2.431. In early stage (data was not shown), no recurrence observed in the patients with positivity of CA 125 in stage IA (= 2), but there were two cases of recurrences among 4 patients with positivity of CA 125 in stage IB, one is peritoneal carcinomatosis, and the other is usually loco-regional recurrence, respectively. We think that this obtaining may be related with recurrence patterns of gastric malignancy. buy 850717-64-5 The sensitivity of CA 125 is usually 6%-31.6%[20,21], and it is related with peritoneal metastasis[22]. In the Japan Clinical Oncology Group (JCOG9501) trial, recurrence patterns were peritoneal recurrence, regional lymph node recurrence, hepatic, as well FCGR1A as others were: 38.1%, 21.9%, 20.9%, and 19.7%, respectively[23]. In Korean study, recurrent patterns were observed according to time period. Hematogenous recurrence was the most common recurrent pattern in the 1990s,.