BackgroundThe optimal treatments for gastric cancer with liver metastases (GCLM) remain controversial. This study aimed to evaluate the efficacy of hepatectomy, RFA and TACE as local treatments for GCLM.MethodsFrom 2001 to 2015, 119 consecutive patients who received multidisciplinary treatments based on curative gastrectomy and local treatments (hepatectomy, RFA and TACE) for liver metastases were enrolled in this retrospective cohort study. Patients were divided into Group A (46, hepatectomy) and Group B (73, either or both RFA and TACE). Propensity score matching analysis was employed.ResultsThe propensity model revealed that hepatectomy was associated with significantly longer OS compared with either or both RFA and TACE (P=0.021). The 1-, 3- and 5-year OS rates were 80.5%, 41.5% and 24.4%, respectively in Group A; and 85.4%, 21.9% and 12.2%, respectively in Group B. Subgroup analyses indicated that hepatectomy was associated with significantly longer long-term survival compared with TACE (P=0.033) and RFA (P=0.010). TACE had a similar efficacy as RFA (P=0.518), but with significantly lower costs (P=0.014) in for patients with metachronous GCLM.ConclusionHepatectomy is the optimal local treatment for GCLM when surgical R0 resection is intended. TACE attained a similar prognosis as RFA with relatively high cost-effectiveness, particularly for patients with metachronous GCLM.