Abstract
Background: A growing number of epidemiological evidence supports that the coagulation cascade and tumor-associated inflammatory indicators are related to the recurrence and survival of diffuse large B-cell lymphoma (DLBCL). Plasma fibrinogen and neutrophil-lymphocyte ratio (F-NLR) is a novel hallmark that is proposed as an adverse prognosticator in a variety of malignancies. The purpose of this study was to investigate the prognostic impact of the F-NLR score in DLBCL.
Methods: In this retrospective study, 231 patients with DLBCL treated between October 2013 to January 2020 at Shandong Provincial Hospital were included. The critical values of fibrinogen and NLR were determined according to receiver operating characteristic (ROC) curve analysis. The included population was stratified into three groups according to the F-NLR score. The prognostic factors associated with progression-free survival (PFS) and overall survival (OS) were screened by multivariate Cox regression analyses and log-rank test.
Results: Within the 231 patients analyzed, 55.4% male and 44.6% female were enrolled in this study. There were 138 (59.7%) patients under 60 years old, with mean age of 55.23±15.02 years old. A total of 184 (79.7%) patients had advanced stage (III or IV) disease, and the great majority of people were defined as non-GCB (64.1%) subtypes and had extranodal involvement (86.6%). The cutoff value was identified as 3.12 for NLR (AUC: 0.628), 4.05 g/L for fibrinogen concentrations (AUC: 0.616), and the AUC value was 0.676 for the F-NLR (Fig 1) by OS outcome. The median follow-up period for all patients was 22 months, ranging from 1 to 90 months. After median follow-up time, 101 (43.7%) patients relapsed or progressed and 77 (34.2%) patients died during or after first-line therapy in total.
In the further univariate and multivariate analyses, we demonstrated that F-NLR (HR=1.953, 95%CI=1.404-2.717, P=0.000) was an independent prognostic factor of DLBCL survival. We calculated the β for OS value of NLR and fibrinogen according to Cox regression equation. Since the HR for OS of each marker is approximation, we assigned the same weight to each factor in the prognostic scoring model (F-NLR). The model was simplified to 3 groups with F-NLR score of 0: neither hyperfibrinogenemia nor high NLR (low risk; 39.8% of pts), F-NLR score of 1: one of these hematological abnormalities (intermediate risk; 38.5% of pts), or F-NLR score of 2: both hyperfibrinogenemia and high NLR (high risk; 21.7% of pts).
As a result, the PFS and OS were significantly worse in DLBCL pts with an F-NLR score of 2 than those with an F-NLR score of 0-1 (P=0.002 Fig 2A, P=0.000 Fig 2B). Moreover, the F-NLR score is significantly associated with clinical outcomes in elderly pts with advanced stage (III or IV) or extranodal involvement through subgroup analysis.
Conclusion: This study validates that the F-NLR score, a new inflammation-based grading system, is a promising clinical predictor for DLBCL especially elderly patients with advanced stage (III or IV) or extranodal involvement.
No relevant conflicts of interest to declare.
