Abstract:Objective To analyze the influencing factors for recurrence in elderly patients with hepatocellular carcinoma (HCC) after radical resection and construct a prediction model. Methods Clinical data of 136 elderly HCC patients undergoing radical resection in our hospital from January 2019 to April 2023 were collected and retrospectively analyzed. All of them were reexamined once a month after surgery. According to the recurrence status at one year after surgery, they were divided into a recurrent group and a non-recurrent group. Geriatric nutritional risk index (GNRI) and intestinal Bifidobacterium/Enterobacter ratio were calculated. The general data were compared between two groups. SPSS statistics 23.0 was used for data analysis. Depending on data type, t test or Chi-square test was used for comparison between groups. Multivariate logistic regression analysis was performed to screen the risk factors for recurrence at one year after radical resection in elderly patients with HCC, and a prediction model for one-year recurrence was constructed. Receiver operating characteristic (ROC) curve was plotted to assess the predictive efficiency of the model. Results At one year after radical resection, all the patients followed the doctor′s advice and received regular re-examinations, without loss of follow-up or refusal to check. During the period, recurrence was observed in 28 cases (20.59%). The preoperative intestinal Bifidobacterium/Enterobacter ratio, preoperative GNRI and microvascular invasion were statistically different between the two groups (all P<0.05). Multivariate logistic regression analysis suggested that preoperative intestinal Bifidobacterium/Enterobacter ratio <1 (OR=2.680,95%CI 1.387-5.180, P<0.05), preoperative GNRI ≤98 (OR=3.473,95%CI 2.043-5.903, P<0.05), and microvascular invasion (OR=3.022,95%CI 1.642-5.562, P<0.05) were risk factors for recurrence at one year after radical resection in elderly HCC patients. Based on these risk factors, a prediction model was constructed as Y=-3.256+0.986× preoperative intestinal Bifidobacterium/Enterobacter ratio (assignment:imbalance=1, normal=0)+1.245× preoperative GNRI (assignment:≤98=1, > 98=0)+1.106× microvascular invasion (assignment:positive=1, negative=0). ROC curve analysis showed that the area under the curve (AUC) of the predictive model was 0.855 (95%CI 0.776-0.934), with a sensitivity of 79.17% and a specificity of 84.62%. Conclusion The recurrence of HCC at one year after radical resection in elderly patients is not only related to microvascular invasion, but also to preoperative intestinal flora imbalance and nutritional status. The predictive model based on these three risk factors of preoperative intestinal Bifidobacterium/Enterobacter ratio <1, preoperative GNRI ≤98 and presence of microvascular invasion has high predictive value on recurrence at one year after surgery, and it is beneficial to guiding clinical diagnosis and treatment.