Abstract
INTRODUCTION: Diffuse large B-cell lymphoma (DLBCL) is the most common subtype of non-Hodgkin lymphoma and carries a poor prognosis when it involves the central nervous system (CNS), a condition known as secondary CNS lymphoma (SCNSL). Although the CNS International Prognostic Index (CNS-IPI) is used to estimate SCNSL risk, its limited sensitivity highlights the need for more reliable biomarkers to improve risk stratification and enable earlier intervention. METHODS: We evaluated pretreatment levels of interleukin-10 (IL-10) and interleukin-6 (IL-6) in both peripheral blood (PB) and cerebrospinal fluid (CSF), and compared clinical characteristics between DLBCL patients with and without SCNSL. RESULTS: Fifty-six newly diagnosed DLBCL patients who received at least two treatment cycles were included. Compared to patients without CNS relapse, those with SCNSL exhibited distinct clinical features: more frequent B symptoms, increased bone marrow involvement, and lower B-cell lymphoma 6 (BCL6) immunohistochemical positivity. Biochemically, SCNSL patients showed elevated serum IL-10 levels, higher serum IL-10/IL-6 ratios, increased CSF IL-10 concentrations, and markedly elevated CSF IL-10/IL-6 ratios. On multivariate analysis, a serum IL-10/IL-6 ratio ≥2.30 independently predicted both progression-free survival (PFS) (HR = 7.300, p = 0.010) and SCNSL development (OR = 43.200, p = 0.001). Notably, time to CNS relapse did not significantly differ between high- and non-high-risk groups defined by CNS-IPI (χ² = 1.654, p = 0.198). However, incorporating the serum IL-10/IL-6 ratio into the CNS-IPI yielded a refined scoring model-CNS-IPI-ratio-where the high-risk group had a significantly shorter median time to CNS relapse compared to the non-high-risk group (22.32 vs. 40.35 months; χ² = 5.680, p = 0.017). Although survival outcomes were similar between high- and non-high-risk groups based on the NCCN-IPI alone (p > 0.05), adding the serum IL-10/IL-6 ratio identified significantly poorer outcomes in high-risk patients (PFS: p = 0.046; OS: p = 0.023). Furthermore, SCNSL patients demonstrated significantly higher CSF IL-10/IL-6 ratios compared to non-SCNSL patients and controls (67.88 vs. 0.74-0.79, p < 0.05), with this ratio strongly correlating with CSF lactate dehydrogenase (LDH) levels (r = 0.625, p = 0.006). CONCLUSION: In DLBCL, an elevated serum IL-10/IL-6 ratio at diagnosis independently predicts disease progression and SCNSL risk. Incorporating this biomarker enhances the prognostic utility of both CNS-IPI and NCCN-IPI models. Additionally, the markedly elevated CSF IL-10/IL-6 ratio in SCNSL patients suggests potential diagnostic value for CNS involvement, warranting further investigation.