Fig 1.
The receiver operating curve (ROC) analysis of pretreatment plasma D-dimer levels.
The optimal cut-off value for D-dimer levels to predict mortality is 1.2μg/mL (sensitivity = 52.6%, specificity = 93.5%, accuracy = 75.0%).
Table 1.
The clinical characteristics of patients with NK/T-cell lymphoma.
Table 2.
Correlation between pretreatment plasma D-dimer levels and clinicopathological factors in patients with NK/T-cell lymphoma.
Fig 2.
Overall survival (OS) in patients with high (≥ 1.2 μg/mL, dashed line) or low (<1.2 μg/mL, solid line) pretreatment plasma D-dimer levels.
Table 3.
Univariate and multivariate analysis of prognostic factors for overall survival in patients with NK/T-cell lymphoma.
Fig 3.
Overall survival (OS) according to Ann Arbor stage and D-dimer (D-D) levels.
(a) OS differed significantly (P < 0.001) between patients with Ann Arbor stage I-II (blue line) or III-IV (red line) disease. (b) Among patients with stage I-II disease, OS was significantly worse in those with D-D levels of ≥ 1.2μg/mL (dashed blue line) compared with those with D-D levels of < 1.2μg/mL (solid blue line, P < 0.001). OS were similar between stage I-II patients with D-D levels of ≥ 1.2μg/mL (dashed blue line) and stage III-IV patients (solid red line, P = 0.785).