Abstract
BACKGROUND: To improve prognosis of patients with synchronous colorectal liver metastasis (CRLM), we constructed a nomogram model to improve outcome through risk stratification and decision support. METHODS: The 389 CRLM patients (273 training set and 116 validation set at a ratio of 7: 3) receiving systematic chemotherapy and synchronously resection with/without radiofrequency ablation (RFA) were retrospectively investigated. Overall survival (OS) and recurrence free survival (RFS) were mainly endpoint. A normo-gram model was conduct. The receiver operating characteristic (ROC) curve, decision curve analysis (DCA), C-index and calibration curve were performed to assess stablity and efficacy of model. The prognosis was evaluated based on Kaplan-Meier (KM) curve. RESULTS: A total of 389 CRLM patients were included. The median OS and RFS times were 70.20 months (95% CIs: 57.73, 82.68) and 11.70 months (95% CIs: 9.75, 13.65), respectively. These patients were divided into training set and validation set at a ratio of 7: 3. In training set, 1, 3, and 5-year survival rate of OS was 97.38%, 71.18%, and 54.56% as well as RFS was 52.57%, 22.65%, and 21.12%, respectively. Cox model showed that hospital day, R0 resection, RFA, only neoadjuvant chemotherapy and CRS score were independent prognostic factors for CRLM patients. The patients were divided into high-risk group and low-risk group based on cut-off value of score calculated by model. The KM curves were statistically different between two groups (P < 0.01). The ROC curve, DCA and calibration curve showed a good prediction efficacy. the C-index of OS and RFS were 0.72 and 0.68, respectively, which were also verified in the validation set (OS, 0.71; RFS, 0.65). CONCLUSIONS: A good prediction model was developed and validated to assess the prognoses of CRLM patients. Systematic chemotherapy and R0 resection could benefit patients' survival and improve prognosis.