Establishment and validation of a nomogram based on coagulation parameters to predict the prognosis of pancreatic cancer

Abstract

Abstract Background In recent years, multiple coagulation and fibrinolysis (CF) indexes have been reported to be significantly related to the progression and prognosis of some cancers. Objective The purpose of this study was to comprehensively analyze the value of CF parameters in prognosis prediction of pancreatic cancer (PC). Methods The preoperative coagulation related data, clinicopathological information, and survival data of patients with pancreatic tumor were collected retrospectively. Mann Whitney U test, Kaplan-Meier analysis, and Cox proportional hazards regression model were applied to analyze the differences of coagulation indexes between benign and malignant tumors, as well as the roles of these indexes in PC prognosis prediction. Results Compared with benign tumors, the preoperative levels of some traditional coagulation and fibrinolysis (TCF) indexes (such as TT, Fibrinogen, APTT, and D-dimer) were abnormally increased or decreased in patients with pancreatic cancer, as well as Thromboelastography (TEG) parameters (such as R, K, α Angle, MA, and CI). Kaplan Meier survival analysis based on resectable PC patients showed that the overall survival (OS) of patients with elevated α angle, MA, CI, PT, D-dimer, or decreased PDW was markedly shorter than other patients; moreover, patients with lower CI or PT have longer disease-free survival. Further univariate and multivariate analysis revealed that PT, D-dimer, PDW, vascular invasion (VI), and tumor size (TS) were independent risk factors for poor prognosis of PC. According to the results of modeling group and validation group, the nomogram model based on independent risk factors could effectively predict the postoperative survival of PC patients. Conclusion Many abnormal CF parameters were remarkably correlated with PC prognosis, including α Angle, MA, CI, PT, D-dimer, and PDW. Furthermore, only PT, D-dimer, and PDW were independent prognostic indicators for poor prognosis of PC, and the prognosis prediction model based on these indicators was an effective tool to predict the postoperative survival of PC.