Systemic inflammation and B cell indices predict rituximab responses in membranous nephropathy

TL;DR

This study finds that B cell levels and a systemic inflammation index can predict how well patients with membranous nephropathy respond to rituximab treatment.

Contribution

The study introduces SIRI and B cell levels as novel biomarkers for predicting rituximab response in membranous nephropathy.

Findings

SIRI ≤1.25 at 3 months independently predicts 6-month response to rituximab.

B cell proportion ≤0.2% at 3 months also predicts 6-month response.

Incorporating SIRI and B cell levels improves prediction accuracy over traditional markers.

Abstract

Membranous nephropathy (MN) is a frequent cause of nephrotic syndrome in adults with variable response to rituximab (RTX) therapy. While traditional markers like proteinuria and anti-phospholipase A2 receptor (PLA2R) antibodies exhibit predictive value, their limitations necessitate more robust biomarkers.

We prospectively analysed 149 MN patients receiving RTX over 12 months. Inflammatory indices such as neutrophil:lymphocyte ratio (NLR), monocyte:lymphocyte ratio (MLR) and systemic inflammation response index (SIRI) together with B cell levels were measured alongside conventional markers at baseline, 3 months and 6 months. Predictive models for 6- and 12-month remission (complete/partial) were developed using multivariate regression and receiver operating characteristics (ROC) analysis.

Non-responders exhibited persistently elevated inflammatory markers (NLR, MLR, SIRI) throughout the entire observation period. Among the three, only SIRI can independently predict the remission of MN. At 3 months, SIRI ≤1.25 {odds ratio [OR] 3.68 [95% confidence interval (CI) 1.39–9.72]} and B cell proportion ≤0.2% [OR 2.90 (95% CI 1.00–8.35)] independently predicted 6-month response. Incorporating these two newly added indicators into the traditional variable model, which includes the levels of proteinuria, albumin and anti-PLA2R antibody at 3 months, markedly enhances prediction accuracy [area under the curve (AUC) 0.86 versus 0.81]. By 6 months, only SIRI ≤0.9 [OR 4.84 (95% CI 1.43–16.40)] and albumin change [OR 1.11 (95% CI 1.03–1.19)] predicted 12-month prognosis, as B cell and anti-PLA2R antibody levels lost significance. The prediction model incorporating SIRI also had better performance (AUC 0.82 versus 0.79).

B lymphocyte levels constitute a robust predictive biomarker for assessing short-term therapeutic response in patients with MN receiving RTX therapy. Furthermore, SIRI emerges as a valuable prognostic indicator capable of predicting both short-term efficacy and long-term renal outcomes. These findings suggest that concurrent monitoring of B lymphocyte levels and SIRI values warrants integration into standardized monitoring frameworks within clinical management protocols.

Graphical Abstract

Figures