A novel nomogram based on the identification of sTREM2 as a biomarker to predict developing neuropsychiatric systemic lupus erythematosus in lupus patients.

Objectives: This study aimed to identify potential biomarkers and construct a nomogram able to predict the development of neuropsychiatric systemic lupus erythematosus (NPSLE) among SLE patients.

Methods: Using bioinformatics analysis, TREM2 was identified as an upregulated gene in NPSLE, participating in various pathological pathways of NPSLE. This study included 80 NPSLE patients and three matched SLE controls with no neuropsychiatric events (non-NPSLE controls) for each of the NPSLE patients. Both serum and cerebrospinal fluid (CSF) concentrations of soluble TREM2 (sTREM2) were assessed. The diagnostic capability of sTREM2 for NPSLE was evaluated using the receiver-operating characteristic curve (ROC curves). The study subsequently integrated a substantial volume of clinical data. Following missing data imputation, patients were randomly allocated to either the training set or the validation set. The Boruta algorithm and Multiple analyses were utilized for constructing the nomogram. Diagnostic performance was assessed using ROC curves, the Hosmer-Lemeshow test, and clinical decision curves.

Results: sTREM2 levels were notably elevated in both serum and CSF of NPSLE patients compared to non-NPSLE controls. Serum TREM2 concentrations correlated with NPSLE severity and neuropsychiatric state. Notably, higher SLE Disease Activity Index (SLEDAI), increased systemic lupus international collaborating clinics (SLICC)/ACR damage index (SDI), prolonged activated partial thromboplastin time (APTT), a higher serum B cells, and elevated serum sTREM2 levels emerged as significant predictors for NPSLE.

Conclusions: sTREM2 presents as a promising biomarker for NPSLE diagnosis. The nomogram that includes serum sTREM2 level as one of the predictors is effective for distinguishing NPSLE from non-NPSLE patients.