Identification of Risk Factors for Relapse Following Rituximab Therapy in Children with Steroid-Sensitive Nephrotic Syndrome

Abstract

Background Rituximab (RTX) is a key therapeutic agent for maintaining remission in steroid-dependent nephrotic syndrome (SDNS) and frequently relapsing nephrotic syndrome (FRNS), demonstrating both efficacy and safety. However, relapse risk varies significantly among individuals post-RTX treatment, and reliable biomarkers for predicting relapse remain under investigation. Methods This single-center, retrospective, observational study included 70 patients who received RTX between January 2015 and January 2023. Patients were classified into two groups based on their clinical outcomes: the non-relapse group (no relapse within two years post-RTX) and the relapse group. Cox proportional hazards regression was used to analyze associations between baseline clinical parameters and relapse risk, while Kaplan-Meier survival analysis was performed to estimate recurrence-free survival rates. Results This cohort of 70 pediatric patients (71.4% male; median age: 9.5 years, IQR: 8.5–12.6) exhibited a 24-month post-rituximab relapse rate of 32.9% (23/70). Multivariate analysis identified elevated baseline memory B-cell levels (adjusted HR = 1.103, 95% CI: 1.045–1.164, P < 0.001), reduced baseline NK cell levels (adjusted HR = 0.866, 95% CI: 0.752–0.997, P = 0.045), and higher IgG levels at 3 months post-RTX (adjusted HR = 1.245, 95% CI: 1.080–1.435, P = 0.003) as independent predictors of relapse. Optimal cutoff values demonstrated strong prognostic utility, with memory B-cells ≥ 19.65%BL of baseline lymphocytes (AUC = 0.750; 78.3% sensitivity, 70.2% specificity), NK cells < 6.45%L of lymphocytes (AUC = 0.671; 74.5% sensitivity, 56.5% specificity), and IgG ≥ 5.74 g/L (AUC = 0.747; 73.9% sensitivity, 66.0% specificity) associated with increased relapse risk. Kaplan-Meier survival analysis revealed significantly lower 24-month relapse-free survival in patients with memory B-cells ≥ 19.65%BL at baseline (13.2% vs. 56.3%; P < 0.001), NK cells < 6.45%L of lymphocytes (22.2% vs. 52.0%; P = 0.011), and IgG ≥ 5.74 g/L (14.3% vs. 51.4%; P = 0.001). These findings highlight the potential of these biomarkers in predicting long-term relapse risk following RTX therapy. Conclusion Baseline memory B-cell and NK cell levels (pre-RTX), as well as IgG levels at three months post-RTX, were identified as predictive biomarkers for two-year relapse risk following RTX therapy. These findings may contribute to the development of personalized RTX treatment strategies.