Staging laparoscopy improves treatment decision-making for advanced gastric cancer

AIM: To evaluate the clinical value of staging laparoscopy in treatment decision-making for advanced gastric cancer (GC). METHODS: Clinical data of 582 patients with advanced GC were retrospectively analyzed. All patients underwent staging laparoscopy. The strength of agreement between computed tomography (CT) stage, endoscopic ultrasound (EUS) stage, laparoscopic stage, and final stage were determined by weighted Kappa statistic (Kw). The number of patients with treatment decision-changes was counted. A chi(2) test was used to analyze the correlation between peritoneal metastasis or positive cytology and clinical characteristics. RESULTS: Among the 582 patients, the distributions of pathological T classifications were T2/3 (153, 26.3%), T4a (262, 45.0%), and T4b (167, 28.7%). Treatment plans for 211 (36.3%) patients were changed after staging laparoscopy was performed. Two (10.5%) of 19 patients in M1 regained the opportunity for potential radical resection by staging laparoscopy. Unnecessary laparotomy was avoided in 71 (12.2%) patients. The strength of agreement between preoperative T stage and final T stage was in almost perfect agreement (Kw = 0.838; 95% confidence interval (CI): 0.803-0.872; P < 0.05) for staging laparoscopy; compared with CT and EUS, which was in fair agreement. The strength of agreement between preoperative M stage and final M stage was in almost perfect agreement (Kw = 0.990; 95% CI: 0.977-1.000; P < 0.05) for staging laparoscopy; compared with CT, which was in slight agreement. Multivariate analysis revealed that tumor size (>= 40 mm), depth of tumor invasion (T4b), and Borrmann type (III or IV) were significantly correlated with either peritoneal metastasis or positive cytology. The best performance in diagnosing P-positive was obtained when two or three risk factors existed. CONCLUSION: Staging laparoscopy can improve treatment decision-making for advanced GC and decrease unnecessary exploratory laparotomy.