Neutrophil-to-lymphocyte ratio, Factor VIII and Antithrombin III: inflammatory-clotting biomarkers in glioma.

Abstract

One of the key difficulties in glioma treatment is our limited ability to consistently assess cancer response or progression either by neuroimaging or specific blood biomarkers. An ideal biomarker could be measured through non-invasive methods such as blood-based biomarkers, aiding both early diagnosis and monitoring disease evolution. This is a single-center, case-control, 10-year retrospective, longitudinal study. We evaluated routine coagulation factors in 138 glioma patients (45 Females/93 Males; median [range] age, 56.4 [27-82] years; 64 non-recurrent/74 recurrent) and, for comparison, in 56 relapsing-remitting MS patients (41 Females/15 Males; 40.8 [25-62] years, 35 stable/21 active) and 23 controls (16 Females/7 Males; 41.7 [24-62] years) as well as Neutrophil-to-lymphocyte ratio (NLR) in subgroups of 127 glioma patients, 33 MS patients and 23 healthy controls. Secondly, we assessed whether these indicators could be predictive of overall (OS) and progression-free survival (PFS) in glioma patients. NLR, d-dimer, Antithrombin III and Factor VIII were significantly higher in glioma patients compared to both MS patients and controls (p<0.0001 for all). ROC curves confirmed that either NLR, Antithrombin III or Factor VIII were moderately accurate biomarkers (0.7<AUC<0.9) for glioma patients compared to other two groups whereas d-dimer was a moderately accurate marker for glioma only when compared to controls. In multivariable analysis, NLR ≥ 4.3 (median) (HR 1.53 [95 % CI 1.04-2.26], p=0.03) together with the Karnofsky Performance Status (KPS) ≥ 80 (median) (0.46 [0.31-0.69], p<0.0001) and use of steroids (1.75 [1.19-2.57], p=0.004) resulted independent predictors of OS while only KPS was independently associated with PFS. Our study showed increased levels of either NLR, Antithrombin III, Factor VIII, or d-dimer in glioma patients compared to MS patients and controls, where the first three represented moderately accurate biomarkers for this cancer. Among these markers, only NLR was found to be predictive for OS along with severe disability and steroid therapy.