A combined marker based on plasma D-dimer and serum albumin levels in patients with nasopharyngeal carcinoma is associated with poor survival outcomes in a retrospective cohort study

Abstract

Background: Activation of the clotting-fibrinolytic system in cancer patients is common and results in an unfavorable clinical outcome. This study aimed to investigate the role of pretreatment plasma D-dimer levels and the combination of D-dimer and albumin (DA) on the prediction of survival prognosis in patients with nasopharyngeal carcinoma (NPC).Methods: The study comprised 511 patients with NPC. Pretreatment plasma D-dimer and serum albumin levels were measured. DA was classified as a new biomarker where D-dimer and albumin levels were combined and was grouped by the cutoff value of both. The correlations of plasma D-dimer levels with clinicopathological features and survival outcome were calculated using the Chi-square test. Kaplan-Meier estimates were performed to analyze the survival functions and were compared using log-rank tests. Cox proportional hazard regression analysis was used to assess the effects of D-dimer and DA on distant overall survival (OS) and distant metastasis-free survival (DMFS).Results: The median follow-up period was 45.2 months (range 2.1-79.8). Elevated plasma D-dimer levels were positively associated with age at diagnosis (P = 0.034), platelet levels (P = 0.043), and Epstein Barr Virus (EBV) DNA copy number (P = 0.035). Additionally, multivariate analysis demonstrated that elevated plasma D-dimer levels were strongly associated with a poorer OS (HR 2.074, 95% CI 1.190-3.612, P = 0.010), but not DMFS. After adjustment for other variables, DA stratification acted as an independent prognostic marker for OS (P = 0.038) and DMFS (P = 0.031) in patients with NPC, when combined with albumin levels.Conclusions: Increased plasma D-dimer levels accurately predict poor OS and may be an effective independent prognostic factor in patients with NPC. Moreover, in conjunction with serum albumin, DA may serve as a factor in predicting OS and DMFS.