We investigated how cerebrospinal-fluid protein patterns relate to clinical manifestations of systemic lupus erythematosus. Proteomic measurements from 29 patients passing quality control were analysed using patient clustering and protein-network analysis, then related to clinical and neurocognitive data. The analysis identified groups with distinct molecular and clinical profiles, including differences in renal involvement and indicators of neuronal injury. Protein modules provided a complementary view of the relationships between biological processes and disease manifestations. By combining proteomics, statistical learning and clinical interpretation, the study explores heterogeneity within lupus and generates hypotheses for further investigation; it does not establish a clinically validated prediction tool.