The array sections were then incubated in a detection kit in accordance with the manufacturer’s instructions. Slides from the immunohistochemical analysis were independently reviewed by two investigators, A-1155463 price who recorded the staining as negative or positive. All cells in all the cores were evaluated. Unequivocal nuclear staining in >5% of tumor cells was considered as positive response; nuclear staining in <5% of tumor cells was considered as negative response. Statistical analysis The following variables were examined: age, gender, tumor type, lymph node status, pathologic stage, and EBV expression. For all statistical tests, two categories were analyzed in pairs as

positive versus negative and present versus absent. We analyzed categorical variables using the Fisher’s exact test, McNemar test and the Mann-Whitney AZD5363 molecular weight rank-sum test. The follow-up time was calculated using the potential follow-up method. Overall patient survival was defined as the time between the date of surgical diagnosis to the date of last follow-up (censored) or the date of patient death (event). The end of follow-up date of this study was December 31, 2006. Censored cases included those cases (n = 6) in which the last follow-up date occurred before December 31, 2006. Patients who deceased of causes other than gastric cancer were

not included in the study. We analyzed the AP26113 cell line differences in survival times between patient subgroups using the log-rank test. Survival probabilities were calculated (using the Kaplan-Meier method) and compared (using the log-rank test) [23]. We performed Cox proportional hazards regression analysis [24] using SAS software (SAS Institute, Cary, NC) to determine the association between the clinicopathologic variables and overall patient survival. First, we analyzed the association between possible prognostic factors (including

age, gender, stage, and node classification) and death, considering one factor at a time. Second, multivariate Cox analysis was performed on backward (stepwise) procedures that always forced EBV into the model, along with all variables that satisfied an entry level of P < 0.05. As the model continued to add factors, independent factors did not exceed an exit level of P > 0.05. Results Clinicopathologic data Clinicopathologic features of the study subjects are summarized MTMR9 in Table 1. Our study consisted of 88 female (37%), and 147(63%) male. One hundred eighteen (50%) patients were older than 65 years, while the other 117 (50%) were 65 years or younger. Eighty-three tumors (35%) were intestinal type, and 152 (65%) were diffuse type. One hundred thirty-one patients (56%) had stage I-II disease, and the remaining 104 patients (44%) had stage III or IV disease. Sixty patients (27%) had nodal involvement and 165 (73%) had no nodal metastases. Table 1 Clinicopathologic features and EBV expression in gastric cancer     EBV Expression     Negative Positive Total p Gender Female 87 (37%) 1 (0%) 88 (37%) 0.