ObjectiveTo investigate the risk factors and establish the predictive pattern of the metachronous liver metastasis after curative surgery for patientswith gastric cancer. MethodsThe clinicopathologic data of patients who underwent radical gastric cancer surgery and met the inclusion and exclusion criteria from January 1, 2015 and January 1, 2018 in the First Hospital of Lanzhou University were retrospectively analyzed. The risk factors affecting metachronous liver metastasis of gastric cancer were screened out by univariate and multivariate logistic regression analysis. And a nomogram prediction model based on the risk factors screened out was established and its predictive efficiency was evaluated. ResultsA total of 203 patients were collected in this study, of whom 41 (20.4%) developed metachronous liver metastasis of gastric cancer. The results of multivariate logistic regression analysis showed that the tumor diameter ≥5 cm, increasing intraoperative bleeding, carcinoembryonic antigen (CEA) ≥5 μg/L, and lymphovascular invasion increased the risks of metachronous liver metastasis of gastric cancer (all P<0.05). The area under the receiver operating characteristic curve and its 95% confidence interval (95%CI) of the nomogram based on these risk factors in predicting metachronous liver metastasis of gastric cancer was 0.850 (0.793, 0.908), and the consistency index (95%CI) was 0.812 (0.763, 0.859). The calibration curve for predicting the risk of metachronous liver metastasis in gastric cancer by the nomogram was close to the 45° ideal curve and had a stronger calibration (Hosmer Limeshow goodness-of-fit test, χ2=2.116, P=0.347). ConclusionsThe results of this study conclude metachronous liver metastasis of gastric cancer is not low, and the patient with lymphovascular invasion, higher level of CEA (≥5 μg/L), more intraoperative bleeding, and larger tumor diameter (≥5 cm) has a higher risk of metachronous liver metastasis of gastric cancer. The nomogram prediction model established based on these risk factors has a good predictive efficiency and can provide reference for clinicians to identify high-risk patient and take early interventions.