| Objective: The influence of oncological behaviors such as TNM stage and degree of differentiation on the prognosis of pancreatic ductal adenocarcinoma（PDAC）has been confirmed,but the influence of inflammation and nutritional status on the prognosis of PDAC is still unclear.This study used the inflammatory index Lymphocyte to monocyte ratio（LMR）and nutritional index Albumin to fibrinogen ratio（AFR）,combined with the clinicopathological characteristics of patients to conduct statistical modeling,aiming to predict the long-term survival of PDAC patients after radical surgery.Methods: The clinical data of PDAC patients after radical surgery in multicenter were retrospectively collected.Cut-off values were taken using X-tile software,and survival analysis was performed using the Kaplan-Meier method.Machine learning algorithms and Cox regression model were used to screen the risk factors affecting the prognosis of patients,and the prediction models of postoperative 1-and 2-year survival were constructed respectively,and the prediction performance of each model was evaluated.Finally,the optimal model was selected.Results: A total of 284 patients were enrolled in this study.The cutoff values for LMR and AFR were 4.9 and 10.8,respectively.Age ≥65 years old,high TNM stage,CA125>35.00U/ml,CA19-9>27.00U/ml,no postoperative chemotherapy,AFR<10.8,LMR<4.9,and CK19 positive are the Independent risk factors that affect the prognosis of patients after radical PDAC surgery（P<0.05）.Among the serum markers,LMR and AFR showed the greatest prognostic power.The predictive performance of the Cox nomogram model in the prognosis of PDAC patients after radical surgery is better than that of the machine learning algorithm model.Conclusions: Patients with LMR≥4.9 and AFR≥10.8 had better long-term survival,and the Cox nomogram model combined with clinicopathological characteristics could better predict the long-term survival of PDAC patients after radical surgery... |
PDF Full Text Request