To investigate the expression profile of cerebrospinal fluid (CSF) lipocalin-2 (LCN2) in neuropsychiatric systemic lupus erythematosus (NPSLE), and to evaluate its associations with clinical parameters and its potential diagnostic value.

This was a single-center retrospective case-control study. CSF LCN2 levels were measured by enzyme-linked immunosorbent assay (ELISA) in patients with NPSLE, age- and sex-matched patients with multiple sclerosis (MS) and healthy controls (HCs). Clinical characteristics, laboratory findings, and CSF parameters were retrospectively collected for the NPSLE and MS groups. Differences in CSF LCN2 levels among the three groups were analyzed using the Kruskal-Wallis test followed by pairwise comparisons. Within the NPSLE group, associations between CSF LCN2 and clinical variables were assessed using Spearman correlation analysis and linear regression analysis. Receiver operating characteristic (ROC) curve analysis was performed to evaluate the discriminatory performance of CSF LCN2 for NPSLE.

A total of 64 subjects were included, including 27 patients with NPSLE, 21 patients with MS, and 16 HCs. CSF LCN2 levels differed significantly among the three groups (P = 0.0076). Pairwise comparisons showed that CSF LCN2 levels were significantly higher in the NPSLE group than in the HC group (P = 0.0133) and the MS group (P = 0.0069), whereas no significant difference was observed between the HC and MS groups.

Within the NPSLE group, CSF LCN2 levels were positively correlated with CSF protein, QAlb, and erythrocyte sedimentation rate (ESR), but negatively correlated with serum albumin. No significant associations were found with disease duration, modified Rankin Scale (mRS) score, CSF white blood cell count, CSF glucose, intrathecal IgG synthesis rate, or serum C-reactive protein (CRP). Univariable linear regression showed that serum albumin was significantly and inversely associated with ln(CSF-LCN2). Exploratory multivariable linear regression suggested that the associations of CSF LCN2 with CSF protein and serum albumin remained directionally consistent, although they did not reach independent statistical significance.

ROC curve analysis showed that CSF LCN2 had moderate discriminatory performance for NPSLE. The area under the curve (AUC) was 0.729 for distinguishing NPSLE from HCs, 0.730 for distinguishing NPSLE from MS, and 0.730 for distinguishing NPSLE from non-NPSLE controls (MS + HCs).

CSF LCN2 is significantly elevated in patients with NPSLE and is associated with blood-brain barrier dysfunction and inflammatory activity, suggesting that it may be involved in the neuroimmune inflammatory process of NPSLE. CSF LCN2 shows moderate discriminatory performance for NPSLE and may serve as a candidate adjunctive biomarker for diagnosis and disease activity assessment. Further validation in larger prospective studies is warranted.