Abstract
BACKGROUND: Diffuse large B-cell lymphoma (DLBCL) is the most common aggressive non-Hodgkin lymphoma and often carries a poor prognosis. Current prognostic systems such as the International Prognostic Index (IPI) and National Comprehensive Cancer Network (NCCN)-IPI do not incorporate patients' nutritional, immune, or inflammatory status, which may affect outcomes. METHODS: We retrospectively analyzed 423 newly diagnosed DLBCL patients and collected 12 clinical indicators reflecting nutritional, immune, and inflammatory status. Patients were randomly divided into training and validation sets in a 7:3 ratio. A LASSO-Cox regression model was applied to identify variables for constructing the Nutritional and Immune-Inflammatory Scoring System (NII). The independent prognostic value of NII was evaluated using univariable and multivariable Cox regression analyses. Its added prognostic value was further assessed in combination with the IPI and NCCN-IPI. RESULTS: We developed the NII, including Nutritional Risk Screening 2002 (NRS2002), Geriatric Nutritional Risk Index (GNRI), systemic immune-inflammation index (SII), lactic dehydrogenase to albumin ratio (LAR), β2-microglobulin (β2-MG), and CD8(+) T cells. A high NII (≥ 6) effectively identifies high-risk DLBCL patients and serves as an independent prognostic factor beyond other clinical characteristic, IPI, and NCCN-IPI. DLBCL patients with a high NII (≥ 6) exhibit significantly adverse clinical features, including older age, lower frequency of the non-GCB subtype, advanced Ann Arbor stage (III/IV), poor performance status (ECOG PS ≥ 2), involvement of ≥ 2 extranodal sites, presence of B symptoms, elevated lactate dehydrogenase (LDH) levels, and classification into higher-risk groups according to IPI and NCCN-IPI. Combining NII with IPI or NCCN-IPI significantly improves the assessment of patient prognosis compared to using IPI or NCCN-IPI alone. CONCLUSION: The NII score, integrating readily available nutritional, immune, and inflammatory markers, enhances prognostic accuracy in DLBCL and complements conventional scoring systems. This simple tool may aid in early identification of high-risk patients and guide personalized treatment.