A novel serum biomarker panel (ICAM2, VSIG4, NrCAM, and P-cadherin) for the non-invasive prediction of active lupus nephritis
Ting Yu, Xiaohui Tang, Na Hu, Hongbo LiBackground: Lupus nephritis (LN) is a serious and dynamic manifestation of systemic lupus erythematosus (SLE). The gold standard for diagnosis is the renal biopsy. However, it is invasive and not suitable for serial monitoring. Standard serological markers, such as anti-dsDNA and complement components, are not sufficiently sensitive or specific to accurately reflect intrarenal inflammation, underscoring the need for novel, non-invasive liquid biomarkers. We hypothesised that a panel of serum proteins indicative of endothelial activation (ICAM2), macrophage complement processing (VSIG4), podocyte integrity (NrCAM), and tubular epithelial stress (P-cadherin) could accurately predict active LN. Methods: In this cross-sectional study, 168 SLE patients were prospectively enrolled and divided into 3 groups: active LN (n=68), inactive LN (n=50), and non-LN SLE controls (n=50). Serum concentrations of ICAM2, VSIG4, NrCAM, and P-cadherin were measured by specific, validated enzyme-linked immunosorbent assays (ELISAs). A combined biomarker panel score was derived by multivariate logistic regression. We analysed the discriminative power of active LN versus all non-active LN controls using receiver operating characteristic (ROC) curves and compared it with that of conventional biomarkers. Results: A clear molecular profile was identified. Serum levels of ICAM2, VSIG4 and P-cadherin were significantly higher, and NrCAM was significantly lower in active LN patients compared to either the inactive LN or non-LN SLE groups (p<0.001 for all). The novel four-biomarker panel showed excellent predictive accuracy for active LN, with an area under the ROC curve (AUC) of 0.921 (95% CI 0.876-0.966). This panel was better than single novel biomarkers, anti-dsDNA (AUC 0.732) and complement C3 (AUC 0.695). With the optimal cutoff, the panel had a sensitivity of 85.3% and a specificity of 88.0%. The panel maintained strong discriminative performance in both proliferative class III/IV LN and membranous class V LN in the exploratory histology-stratified analysis, with no statistically significant difference in AUC between subgroups. Conclusion: Serum biomarkers, including ICAM2, VSIG4, NrCAM, and P-cadherin, are potent and non-invasive predictors of active lupus nephritis. This panel covers several pathogenic pathways, including vascular inflammation, complement-mediated macrophage activation, podocyte injury, and tubular stress. It thus provides a comprehensive "molecular biopsy" of the kidney. This panel is a promising complementary tool to improve the monitoring and management of LN.