RESUMO
BACKGROUND: Perioperative treatment is an accepted standard approach for treating locally advanced gastric cancer (LAGC). Histopathological tumor regression with < 10% residual tumor is a globally accepted prognosticator in LAGC patients who received neoadjuvant chemotherapy (NAC) and curative surgery. However, despite a response of the primary tumor, a significant percentage of patients dies from recurrence and identification of those at risk for relapse remains challenging. We re-estimated the value of histopathological tumor regression as a prognosticator alongside other factors, especially posttherapy topographical nodal status, ypN-site. PATIENTS AND METHODS: Individual patient data including clinicopathological variables were used from the four JCOG trials investigating NAC (JCOG0001, JCOG0002, JCOG0210, JCOG0405) for analyzing prognosticators in patients with curative surgery excluding those with type 4 AGC by univariable and multivariable Cox regression analyses. RESULTS: Among 85 patients, 5-year overall survival (OS) was 46.0% [95% confidence interval (CI) 35.0-56.4] with a median follow-up of 3.2 years. On univariable analysis, histopathological tumor regression with ≥ 10% residual tumor and ypN-site 2-3 were negatively associated with OS [≥ 10% residual tumor: hazard ratio (HR) 2.60; 95% CI 1.22-5.54; P = 0.014; ypN2-3: HR 3.59; 95% CI 1.60-8.06; P = 0.002). On multivariable analysis, only ypN-site 2-3 was predictive of OS (HR 3.67; 95% CI 1.55-8.69; P = 0.003), whereas histopathological tumor regression with ≥ 10% residual tumor was not (HR 2.24; 95% CI 0.98-5.10; P = 0.055). CONCLUSIONS: ypN-site may have greater impact on OS than histopathological tumor regression in patients who received NAC plus surgery for non-type 4 LAGC.