BackgroundIndications for adjuvant chemotherapy for advanced gastric cancer are determined based on the pathological diagnosis. However, macroscopic diagnoses have been reported as predictors of peritoneal recurrence and survival. This study investigated the predictability of peritoneal recurrence and survival based on macroscopically (sT) and pathologically (pT) diagnosed serosal invasion to identify more sensitive predictors of peritoneal recurrence. MethodsThis study included 396 patients who underwent R0 resection without adjuvant chemotherapy with S-1 in the JCOG0110 study. Tumor depth limited to the subserosa (SS) was defined as serosal invasion negative (T-), while tumors with serosal invasion (SE, SI) were defined as serosal invasion positive (T+). The predictability of peritoneal recurrence based on sT and pT was evaluated using the Fine and Gray model. Cox regression analyses were performed for overall survival (OS) and relapse-free survival (RFS) with sT or pT as covariates. FindingsA total of 150 patients (37.9%) were sT+ and 82 (26.3%) were pT+. Sixty-two patients (15.7%) were sT+/pT+, 88 (22.2%) were sT+/pT-, 20 (5.1%) were sT-/pT+, and 226 (57.1%) were sT-/pT-. Both sT and pT were found to be independent predictors of peritoneal recurrence, OS, and RFS. The 5-year RFS rates of sT+/pT+, sT+/pT-, sT-/pT+, and sT-/pT-patients were 45.2%, 63.6%, 55.0%, and 81.8%, respectively. ConclusionIntraoperatively diagnosed macroscopic serosal invasion showed a similar predictive value for peritoneal recurrence as pathologically diagnosed serosal invasion. The establishment of a novel staging system incorporating macroscopic serosal invasion is recommended.