Long-term survival after liver metastasectomy in gastric cancer: Systematic review and meta-analysis of prognostic factors.

Background: Despite the amelioration of systemic therapy, overall survival (OS) of metastatic gastric cancer (GC) patients remains poor. Liver is a common metastatic site and retrospective series suggest a potential OS benefit from hepatectomy, with interesting 5-year (5 y) and 10-year (10 y) OS rates in selected patients. We aim to evaluate the impact of liver resection and related prognostic factors on long-term outcome in this setting.

Methods: We searched Pubmed, EMBASE, and Abstracts/posters from international meetings since 1990. Data were extracted from publish papers. Random effects models meta-analyses and meta-regression models were built to assess 5yOS and the impact of different prognostic factor. Heterogeneity was assessed using between study variance, I and Cochran's Q. Funnel plot were used to assess small study bias.

Results: Thirty-three observational studies (for a total of 1304 patients) were included. Our analysis demonstrates a 5yOS rate of 22% (95%CI: 18-26%) and 10yOS rate of 11% (95%CI: 7-18%) among patients undergoing radical hepatectomy. A favorable effect on OS was shown by several factors linked to primary cancer (lower T and N stage, no lympho-vascular or serosal invasion) and burden of hepatic disease (≤3 metastases, unilobar involvement, greatest lesion < 5 cm, negative resection margins). Moreover, lower CEA and CA19.9 levels and post-resection chemotherapy were associated with improved OS.

Conclusions: Surgical resection of liver metastases from GC seems associated with a significant chance of 5yOS and 10yOS and compares favourably with results of medical treatment alone. Prospective evaluation of this approach and validation of adequate selection criteria are needed.