ISSN 1006-298X      CN 32-1425/R

Chinese Journal of Nephrology, Dialysis & Transplantation ›› 2025, Vol. 34 ›› Issue (4): 315-320.DOI: 10.3969/j.issn.1006-298X.2025.04.003

Previous Articles     Next Articles

Predictive model for membranous nephropathy remission based on multi-domain antibodies and analysis of risk factors

  

  • Online:2025-08-28 Published:2025-08-28

Abstract: Objective: To construct a model for predicting the probability of renal response at 6 months after immunosuppressive therapy in patients with membranous nephropathy (MN).
Methodology: A total of 115 MN patients confirmed by renal pathology were enrolled, and serum samples from 69 patients were collected 6 months after immunosuppressive therapy to measure phospholipase A2 receptor (PLA2R) and domain-specific antibody levels. LASSO regression was used to screen variables, and a multivariate logistic regression model was constructed. The predictive ability of the model was evaluated using ROC curves.
Results: LASSO regression identified lnCTLD1-IgG-M0 and lnCTLD1-IgG4-M0 as independent risk factors affecting renal response in MN patients after 6 months of treatment. The composite predictive model outperformed conventional PLA2R antibody indices in predictive capability.
Conclusion: lnCTLD1-IgG-M0 and lnCTLD1-IgG4-M0 are the main risk factors for predicting renal remission in MN patients. The prediction model based on PLA2R domain-specific antibodies exhibits excellent discriminative and calibration abilities, providing a valuable predictive tool for clinical practice to improve renal outcomes in MN patients.

Key words: font-family:Inter, -apple-system, BlinkMacSystemFont, ", font-size:16px, background-color:#F9FAFB, ">membranous nephropathy, phospholipase A2 receptor antibody, epitope spreading