Multivariate Analysis of Prognosis in Patients with Pancreatic Cancer

Abstract

: Objective Pancreatic cancer is a highly malignant digestive system tumor. The prognosis of pancreatic cancer may be affected by many factors. But there is a lack of research on the multivariate analysis of prognostic factors of pancreatic cancer. This study aimed to explore the factors that multivariate influence the prognosis of pancreatic cancer. Methods We conducted a retrospective cohort study using electronic medical record database from The First Hospital of Jilin University from January 2014 to December 2016. Clinical information, such as TNM stage, serum tumor markers, serum albumin and the patient-generated subjective global assessment (PG-SGA), were collected. Ka-plan-Meier method was used to calculate survival rate. The survival curve was compared with the Log-rank test. A multivariate analysis was performed for each prognostic factor using the Cox proportional hazards model. The significance level was 0.05. Results 1. The median overall survival time (OS) of 493 patients was 10 months. The 1-year, 2-year, and 3-year cumulative survival rates were 41.5%, 20.6%, and 9.9%, respectively. Univariate survival analysis revealed that TNM stage ( P < 0.001), treatment regimen ( P < 0.001), CA125 ( P < 0.001), the level of serum albumin ( P = 0.014), and nutritional status ( P = 0.006) were significantly correlated with prognosis of pancreatic cancer. COX proportional hazards regression model showed that TNM stage ( P = 0.038), Carbonhydrate antigen 125 (CA125) ( P = 0.003) and the level of serum albumin were independent factors affecting the median OS. The prognosis of patients with Ⅳ stage, elevated levels of serum albumin group and elevated levels of CA125 group were poor. 2. Seventy-three patients with pancreatic cancer underwent surgical resection. The median postoperative survival time was 28 months. The 1-year, 2-year, and 3-year cumulative survival rates were 74.4%, 47.7%, and 21.4%, respectively. Univariate analysis revealed that TNM stage ( P = 0.032), pathological type ( P = 0.012), hypoglycemia ( P = 0.040), CA125 ( P < 0.001), were significant ly correlated with prognosis of pancreatic cancer. COX proportional hazards regression model showed that TNM stage ( P = 0.008), pathological type ( P = 0.045) and CA125 ( P = 0.017), were independent prognostic factors after resection of pancreatic cancer. The prognosis of patients with advanced stage, non-ductal adenocarcinoma and elevated levels of CA125 group was poor. Conclusions 1. TNM stage, CA125 and serum albumin are independent prognostic factors for the overall survival of patients with pancreatic cancer. 2. TNM stage, pathological type, and CA125 are independent prognostic factors for the postoperative survival time in patients with pancreatic cancer after resection.