Figure 1 illustrates the clinical practice flowchart for rituximab use in patients with steroid-sensitive nephrotic syndrome.

Recommendation 1: rituximab is recommended for the treatment of children with FRNS/SDNS, and it can achieve better efficacy in FRNS/SDNS − than in FRNS/SDNS + (1D)

Rationale: the recommendation is justified in terms of disease relapse, steroid reduction and time to steroid withdrawal after RTX treatment in children with FRNS/SDNS.

Relapse rate

The one-year relapse rate of RTX treatment in children with FRNS/SDNS was lower than that of placebo (decrease of 95%) and CTX, CNI, or MMF (decrease of 68%).

A meta-analysis of 10 studies [including seven RCTs [18,19,20,21,22,23,24], one nonrandomized-controlled trial (NRCT) [25], and two cohort studies [26, 27]] with a one-year follow-up on the relapse rate of FRNS/SDNS treated with RTX revealed a statistically significant 82% reduction compared with the control group [odds ratio (OR) = 0.18, 95% CI 0.07–0.44] (Supplementary Fig. 1). In the FRNS/SDNS + subgroup, the relapse rate related to RTX decreased by 81% compared with that in the control group (OR = 0.19, 95% CI 0.05–0.68), and the difference was statistically significant. In the FRNS/SDNS − subgroup, the relapse rate associated with RTX was 85% lower than that in the control group (OR = 0.15, 95% CI 0.03–0.68), and the difference was statistically significant. Compared with the placebo or no treatment subgroup, the rate of relapse with RTX decreased by 94% (OR = 0.06, 95% CI 0.02–0.18), and the difference was statistically significant. Compared with those in the CTX, CNI, or MMF subgroups, the rate of relapse with RTX decreased by 68% (OR = 0.32, 95% CI 0.13–0.82), and the difference was statistically significant.

The average one-year relapse rate following RTX treatment was 41% in children with FRNS/SDNS (47% in FRNS/SDNS + and 22% in FRNS/SDNS −).

A meta-analysis of 21 studies [eight CSRs [28,29,30,31,32,33,34,35] and the RTX groups in seven RCTs [18,18,19,20,21, 23, 24, 36], one non-randomized studies of interventions (NRSI) [25], and five cohort studies [26, 27, 29, 37, 38]] on the relapse rate at 12 months (Supplementary Fig. 3) revealed an overall relapse rate of 41% (95% CI 31%–52%), a rate of 47% (95% CI 36%–59%) in the FRNS/SDNS + subgroup, and 22% (95% CI 9%–37%) in the FRNS/SDNS − subgroup; there was a statistically significant difference in the data between the two subgroups (χ2 = 26.926, P < 0.001).

The meta-analysis (Supplementary Fig. 4) of 12 studies (five CSRs [28, 30, 31, 33, 35] and the RTX groups in four RCTs [19,20,21, 24] and one NRSI [25] and two cohort studies [27, 29]) on FRNS/SDNS after 1–2 doses of RTX with a follow-up of 12 months revealed an overall relapse rate of 33% (95% CI 22%–45%), a rate of 14% (95% CI 7%–21%) in the FRNS/SDNS − subgroup, and 42% (95% CI 30%–55%) in the FRNS/SDNS + subgroup, with a statistically significant difference between the two subgroups (χ2 = 27.494, P < 0.001).

There was no difference in the two-year relapse rate of the FRNS/SDNS group between the RTX treatment group and the continuous immunosuppressive therapy group, but there were higher two- and three-year relapse rates in the FRNS/SDNS + group than in the FRNS/SDNS − group after RTX treatment.

Pooling the three FRNS/SDNS studies on RTX (one RCT [19], two cohort studies [27, 39]), the meta-analysis for the 24-month follow-up relapse rate (Supplementary Fig. 5) revealed no statistically significant difference between the RTX group and the steroid alone group or the CNI or CTX control subgroup (OR = 0.65, 95% CI 0.31–1.35), no such difference in the FRNS/SDNS + subgroup (CNI or CTX as a control), and no such difference in the FRNS/SDNS − subgroup (placebo as a control).

The meta-analysis of the relapse rate at the 24-month follow-up in 10 studies (five CSRs [28, 30,31,32, 40] and the RTX groups in two RCTs [19, 21] and three cohort studies [27, 37, 39]) for RTX treatment in FRNS/SDNS patients (Supplementary Fig. 6) revealed a rate of 72% (95% CI 61%–83%), 75% (95% CI: 63%–86%) in the FRNS/SDNS + subgroup, and 54% (95% CI 33%–75%) in the FRNS/SDNS-subgroup, with a statistically significant difference between the two subgroups (χ2 = 4.376, P = 0.036).

Another meta-analysis of the relapse rate at the 36-month follow-up in four studies (two CSRs [28, 40] and the RTX group in one RCT [21]) for RTX treatment in FRNS/SDNS (Supplementary Fig. 7) reported a rate of 85% (95% CI 76%–91%), 92% (95% CI 84%–96%) in the FRNS/SDNS + subgroup, and 47% (95% CI 21%–73%) in the FRNS/SDNS − subgroup, with a statistically significant difference between the two subgroups (exact test, P < 0.001).

In a multicenter RCT, 120 patients with SDNS were randomized to receive RTX or tacrolimus treatment, followed up for one year (RITURNS trial [20]) and then for another two years (RITURNS II trial [41]). Once relapse occurred, all the patients in the RTX group received the second course of RTX treatment, 44 of whom received RTX plus MMF and 15 of whom did not receive MMF. The 56 patients who relapsed in the tacrolimus group switched to RTX treatment, 52 of whom switched to RTX + MMF and four to RTX alone. The two-year relapse-free survival rates (n = 19) was 9% in the RTX monotherapy arm and 67% in the RTX + MMF arm (n = 96), respectively, with a statistically significant difference (P < 0.001).

Steroid reduction

RTX treatment significantly improved one-year cumulative steroid dose reduction in FRNS/SDNS pediatric patients compared with other immunosuppressants (RTX vs. tacrolimus, − 0.15 mg/kg/day; RTX vs. placebo, − 0.26 mg/kg/day).

Pooling the four studies (three RCTs [18, 20, 23], and one cohort study [26]), the meta-analysis for the 12-month cumulative steroid dose revealed that RTX significantly reduced the mean 12-month cumulative steroid dose by 0.16 mg/kg/day compared with that of the tacrolimus or placebo control [mean decrease (MD) =  − 0.16, 95% CI − 0.20 to − 0.12], with a statistically significant difference (Supplementary Fig. 8); decreased the cumulative steroid dose by 0.15 mg/kg/day compared with that of the tacrolimus control subgroup (MD =  − 0.15, 95% CI − 0.18 to − 0.11), with a statistically significant difference; and significantly reduced cumulative steroid dose by 0.26 mg/kg/day compared with that of the placebo subgroup (MD =  − 0.26, 95% CI − 0.36 to − 0.16), with a statistically significant difference (Supplementary Fig. 9). There was a 0.14 mg/kg/day reduction in the FRNS/SDNS + subgroup compared with tacrolimus or placebo control (MD =  − 0.14, 95% CI − 0.21 to − 0.08), with the difference statistically significant (Supplementary Fig. 8); and a 0.17 mg/kg/day in the FRNS/SDNS − subgroup, compared with tacrolimus or placebo controls (MD =  − 0.17, 95% CI − 0.28 to − 0.07), with the difference statistically significant (Supplementary Fig. 9).

For FRNS/SDNS + , the meta-analysis of three RCTs [18, 20, 23] and two cohort studies [26, 38] comparing the cumulative steroid dose before and after intervention in the RTX groups over 12 months revealed a decrease of 0.35 mg/kg/day (MD =  − 0.35, 95% CI − 0.38 to − 0.31), with a statistically significant difference (Supplementary Fig. 10).

One RCT on RTX in FRNS/SDNS − patients [22] reported no difference between the RTX and CNI control groups in terms of the median cumulative steroid dose [0.11 vs. 0.11 mg/kg/day at the 12-month follow-up (P = 0.15)].

One NRSI in FRNS/SDNS − patients [25] revealed a statistically significant reduction in the one-year cumulative steroid dose by 0.78 ± 0.23 mg/kg qod at the 12-month follow-up in the RTX treatment group.

A case series report on RTX treatment of FRNS/SDNS + patients with a mean follow-up period of 17 (13–21) months [42] reported a significant reduction in the steroid dose (mg/kg/day) by 63% after 12 months of RTX exposure compared with the baseline.

Among the 65 articles retrieved for this guideline, the steroid molecules used to achieve remission before rituximab was prescribed were as follows: methylprednisolone was mentioned in five articles, prednisolone in 25, and prednisone in 16. Furthermore, 19 articles provided only a general mention of “steroids”, lacking molecular specificity.

Steroid withdrawal and duration of steroid therapy

The meta-analysis of two studies (one RCT [43] and one NRSI [25]) with RTX therapy for FRNS/SDNS revealed that the steroid withdrawal rate in the RTX group at the three-month follow-up was 16-fold greater than that in the CNI or CTX control group (OR = 15.71, 95% CI 5.72–43.16) (Supplementary Fig. 11).

The meta-analysis of steroid withdrawal rates during the 12-month follow-up in two studies (one RCT [20] and one cohort study [39]) on RTX intervention for FRNS/SDNS revealed no statistically significant difference between RTX and CNI or CTX control (OR = 1.93, 95% CI 0.87–4.29) (Supplementary Fig. 12).

The meta-analysis of six-month steroid withdrawal rates from three studies (two CSRs [29, 34] and the RTX arm in one cohort study [37]) in FRNS/SDNS + patients revealed a rate of 63% (95% CI 51%–73%) (Supplementary Fig. 13).

In a cohort study on RTX for the treatment of FRNS/SDNS [27], no statistically significant difference in the time to steroid withdrawal was found between the RTX group and the CNI control group during follow-up for 12–24 months.

One case series report on RTX treatment (n = 17) in FRNS/SDNS + patients [29] reported 100% steroid withdrawal at the three-month follow-up. Another case series report on RTX intervention (n = 81) in FRNS/SDNS + patients [44] reported a median time to steroid withdrawal of 66 (26–409) days for 69 patients during follow-up for 13–90 months. In a case series on RTX exposure (n = 101) in FRNS/SDNS + patients [38], the average time to steroid withdrawal in 90 patients was 4.8 ± 2.0 months. In another case series report on RTX intervention (n = 37) in FRNS/SDNS + patients [30], 35 patients discontinued steroid treatment after a median of 1.3 (0.37–6) months. The RCT on RTX in the treatment of FRNS/SDNS + [23] and the comparison between RTX and placebo or no treatment control revealed a statistically significant difference in steroid-free duration one year after intervention (140.5 ± 91.3 days, RTX vs. 80.2 ± 98.5 days, controls; P = 0.02), although the two groups had no statistically significant difference in this parameter before intervention.

The time to steroid withdrawal in the above-mentioned studies is summarized as follows: within three months for five studies [20, 25, 29, 39, 43], within 3–6 months for three studies [23, 38, 44], and within at least 6–9 months for one study [34].

Median time to first relapse

The median time to first relapse was about 10 months in RTX-treated children with FRNS/SDNS + .

The meta-analysis of nine studies (six CSRs [28, 30,31,32,33, 44] and the RTX arms in three cohort studies [26, 37, 39]) on RTX treatment in FRNS/SDNS + patients for outcome follow-up at ≥ 12 months revealed that the median time to first relapse was 9.89 (95% CI 7.14–12.65) months, with I2 = 82.88% and P < 0.001 (Supplementary Figs. 14–15).

The meta-analysis of relapse-free survival at the 12-month follow-up in three studies (two RCTs [19, 20], and one NRSI [