Rituximab-induced long-term remission in childhood-onset, uncomplicated, frequently relapsing or steroid-dependent nephrotic syndrome: a randomized, placebo-controlled trial and a follow-up study

Abstract

Rituximab maintains remission of complicated frequently relapsing or steroid-dependent nephrotic syndrome (FRNS/SDNS) by depleting peripheral B cells, but most patients eventually experience relapses after B cell recovery. We performed a multicenter, double-blind, randomized, placebo-controlled trial to assess rituximab's efficacy and safety for childhood-onset uncomplicated FRNS/SDNS (without prior treatment with glucocorticoid-sparing immunosuppressive agents) with a follow-up study to assess rituximab's long-term effect after B cell recovery. Patients were randomly assigned to receive either rituximab (375 mg/m², maximum 500 mg, once weekly for 2 weeks) or placebo. The primary endpoint was the relapse-free period. Of 43 randomized patients, 40 received the intervention (18 rituximab, 22 placebo). The relapse-free period during the 1-year trial was significantly longer in the rituximab vs. placebo groups (median: 285 vs. 81 days; p < 0.001). Infusion reactions were more frequent in the rituximab group (p < 0.001), with no difference in adverse events incidence between the groups. Interestingly, the follow-up study demonstrated markedly higher 3-year cumulative relapse-free survival probability without further treatments in the rituximab vs. placebo groups (38% vs. 9%). A mini-systematic review with meta-analyses supported the findings. Rituximab is effective and well-tolerated, potentially leading to long-term remission with substantially high rates after B cell recovery for childhood-onset uncomplicated FRNS/SDNS.Trial registration JSKDC10, Clinical Trials Registry ID: jRCT1091220380; JSKDC10 follow-up study, Clinical Trials Registry ID: jRCT1050230024.

Keywords: Children; Long-term remission; Rituximab; Uncomplicated frequently relapsing or steroid-dependent nephrotic syndrome.

Fig. 1

Study design. Candidates with a relapse of nephrotic syndrome were screened for eligibility to this study, and simultaneously they were treated with the standard prednisolone regimen for relapse. When the patients showed complete remission, the eligible patients were randomized into rituximab group or placebo group. When patients experienced treatment failure, their allocation code was urgently disclosed. If a patient with Treatment failure (1) was treated with the placebo, he or she could then choose to begin the standard immunosuppressive treatments deemed the best choice by the investigators or enter the open-label phase (rituximab treatment). Patients with Treatment failure (1) in the rituximab group and those with Treatment failure (2) or (3) in both groups were treated with the standard immunosuppressive treatments deemed the best choice by the investigators.

Fig. 2

Flow diagram. The solid lines indicate the blinded phase, and the dashed lines indicate the open-label phase.

Fig. 3

Kaplan–Meier curves for the primary outcome. (A) The relapse-free period during the trial was markedly longer in the rituximab group compared with that in the placebo group (median: 285 vs. 81 days, hazard ratio (95% confidence interval): 0.27 (0.12–0.59); p < 0.001). (B) Post-hoc analysis including three patients excluded from FAS. The data of three patients excluded from FAS were censored at the time an event occurred that warranted their exclusion from the FAS. The Kaplan–Meier curves of the primary analysis, which included the three cases excluded from the FAS, matched those of the original analysis exactly, except for the vertical bars indicating censoring at day 9 (placebo, 1 patient) and day 13 (rituximab, 2 patients), respectively and the number at risk at time 0.

Fig. 4

Kaplan–Meier curves during the blinded phase and follow-up period. The relapse-free period including the follow-up period was still longer in the rituximab group than in the placebo group (HR (95% CI): 0.28 (0.13–0.61)). The cumulative 2- and 3-year relapse-free survival probability in the rituximab group was 44% (95% CI 22–65%) and 38% (95% CI 17–60%), respectively, whereas in the placebo group, both probabilities were 9% (95% CI 2–25%).

Fig. 5

Meta-analyses of the three trials to evaluate the outcomes at 12, 24, and 36 months after the intervention. All of the pooled risk ratios were obtained from meta analyses yielding the fixed-effect models since the analyses include only three trials from two research teams (details of the explanation have been included in the supplementary material (Supplementary mini-systematic review). Meta-analyses of the three trials showed marked beneficial effects of rituximab in children with uncomplicated FRNS/SDNS during up to 36 months of follow-up.