Skip to main content
NCBI home page
Clinical and Translational Science logoLink to Clinical and Translational Science
. 2026 Sep 28;19(10):e70730. doi: 10.1111/cts.70730

Rituximab as Rescue or Add‐On Therapy for Relapsing or Refractory Lupus Nephritis: A Systematic Review and Meta‐Analysis

Xiaolong Wang 1, Shuang Liang 1, Xiaowei Sun 1, Fuling Wang 1,2, Guichun Tang 1, Weizhu Deng 1, Meng Li 1,2, Qian Wang 1,✉, Zheyi Dong 1,✉
PMCID: PMC13620286  PMID: 42806593

ABSTRACT

Relapsing and refractory lupus nephritis (LN) remains difficult to manage despite advances in immunosuppressive and targeted therapies. Rituximab (RTX) is often used as rescue or add‐on therapy, but renal outcomes in this specific setting have not been quantitatively synthesized. PubMed, Embase, and Web of Science were searched from inception to March 30, 2026. Studies of RTX‐based therapy in relapsing or refractory LN were included if renal outcomes were reported. Primary outcomes were complete renal response (CRR), partial renal response (PRR), and overall renal response (ORR) at 6, 12, and 24 months; secondary outcomes were changes in proteinuria and serum creatinine. Random‐effects models were used. Nineteen studies involving 445 patients were included. Pooled CRR rates were 0.33 (95% CI 0.21–0.46), 0.41 (95% CI 0.28–0.54), and 0.42 (95% CI 0.26–0.60) at 6, 12, and 24 months, respectively. Corresponding PRR rates were 0.29 (95% CI 0.16–0.45), 0.23 (95% CI 0.07–0.46), and 0.19 (95% CI 0.07–0.34), and ORR rates were 0.67 (95% CI 0.52–0.80), 0.70 (95% CI 0.51–0.86), and 0.66 (95% CI 0.43–0.86). Proteinuria decreased by −2.69 g/day‐equivalent (95% CI −3.23 to −2.16), and serum creatinine decreased by −11.79 μmol/L (95% CI −19.86 to −3.72). Sensitivity analyses were generally stable, and Egger's test detected no significant publication bias. In conclusion, RTX‐based rescue or add‐on therapy was associated with moderate renal response rates and reductions in proteinuria and serum creatinine in this difficult‐to‐treat population. Further prospective comparative studies are needed to define its role.

Keywords: Lupus nephritis, meta‐analysis, refractory lupus nephritis, relapsing lupus nephritis, rituximab

Study Highlights

  • What is the current knowledge on the topic?
    • ○
      Relapsing or refractory lupus nephritis remains a difficult clinical problem. Rituximab is often considered when standard immunosuppressive treatment is insufficient, but the available evidence is scattered and its role in this setting is still not well defined.
  • What question did this study address?
    • ○
      This study examined renal outcomes after rituximab‐based rescue or add‐on therapy in patients with relapsing, refractory, or relapsing/refractory lupus nephritis.
  • What does this study add to our knowledge?
    • ○
      In 19 studies including 445 patients, rituximab‐based therapy was associated with moderate renal response rates. Overall renal response was maintained at approximately two‐thirds during follow‐up, while complete renal response increased from about one‐third at 6 months to about two‐fifths at 12 and 24 months. Proteinuria and serum creatinine also decreased after treatment, with continuous renal outcomes reported using clarified units and measurement approaches.
  • How might this change clinical pharmacology or translational science?
    • ○
      These findings suggest that rituximab may be a reasonable escalation option for selected patients with difficult‐to‐treat lupus nephritis. However, because most available data come from observational or single‐arm studies, prospective studies are still needed to define which patients are most likely to benefit, how rituximab should be combined with background therapy, and how long responses can be sustained.

Abbreviations

CI

confidence interval

CRR

complete renal response

CYC

cyclophosphamide

LN

lupus nephritis

MMF

mycophenolate mofetil

ORR

overall renal response

PRISMA

Preferred Reporting Items for Systematic Reviews and Meta‐Analyses

PRR

partial renal response

RTX

rituximab

1. Introduction

Lupus nephritis (LN) remains one of the most serious manifestations of systemic lupus erythematosus and continues to be a major cause of chronic kidney injury and long‐term morbidity [1, 2]. In recent years, LN management has become more structured, and treatment strategies have expanded beyond the conventional regimens of cyclophosphamide or mycophenolate mofetil, with increasing use of multitarget approaches and newer biologic or targeted agents [3, 4]. Nevertheless, achieving durable renal remission remains difficult in a substantial proportion of patients, and a pressing need exists to refine treatment goals beyond short‐term response [5, 6].

This challenge is particularly relevant in patients with relapsing or refractory LN. In routine practice, these patients often have ongoing proteinuria, incomplete biochemical recovery, or continued disease activity despite standard immunosuppressive therapy [6, 7]. Real‐world analysis has shown that the burden of immunosuppressant‐treated LN remains considerable, with frequent changes in treatment strategy and substantial healthcare burden [7]. A recent review of randomized controlled trials in LN also noted that most available trial evidence still centers on its initial management, whereas patients with sustained, difficult‐to‐treat disease remain underrepresented and important evidence gaps persist [8].

Rituximab (RTX) has remained of particular interest in this setting. As an anti‐CD20 monoclonal antibody, RTX targets B cell depletion and has a clear biologic rationale in LN [9, 10]. It is used as rescue therapy or add‐on therapy in LN patients with refractory or resistant disease, as recent cohort studies continue to report [11, 12]. However, available evidence remains fragmented, mostly consisting of retrospective cohorts, single‐center experiences, registry‐based studies, or broader reviews of B cell‐targeted biologics. Thus, a need exists for dedicated, high‐quality evidence focused on difficult‐to‐treat LN populations [8, 10].

The broader LN treatment landscape has continued to evolve. Recent clinical trials have reinforced the role of newer agents such as voclosporin and obinutuzumab within the setting of contemporary LN management [13, 14]. However, the clinical position of RTX in difficult‐to‐treat LN remains insufficiently defined. Although published systematic reviews and meta‐analyses have examined RTX in LN more broadly, there is still a lack of focused synthesis addressing its role in difficult‐to‐treat disease, particularly when used as an escalation strategy after inadequate response to conventional therapy [15, 16]. Given its continued clinical use in this setting and the fragmented nature of the available evidence, a dedicated systematic review and meta‐analysis is warranted.

2. Methods

This systematic review and meta‐analysis evaluated RTX‐based therapy in patients with relapsing or refractory LN. The review was conducted in accordance with the Cochrane Handbook for Systematic Reviews of Interventions and reported following the Preferred Reporting Items for Systematic Reviews and Meta‐Analyses (PRISMA) statement [17, 18]. The protocol was prospectively registered in PROSPERO (CRD420261351337).

2.1. Literature Search

A systematic search was carried out in PubMed, Embase, and Web of Science from database inception to March 30, 2026. The search strategy combined terms related to lupus nephritis, rituximab, and difficult‐to‐treat disease states, including refractory, relapsing, rescue, add‐on, salvage, inadequate response, treatment failure, and related expressions. Only studies with human subjects published as full‐length articles in peer‐reviewed journals were considered in the analyses. To avoid missing relevant reports, the reference lists of eligible studies and related reviews were also screened manually. The full search strategies for each database are presented in Table S1.

2.2. Eligibility Criteria

Studies were eligible if they enrolled patients with relapsing LN, refractory LN, or relapsing/refractory LN; evaluated RTX‐based therapy used as rescue therapy, add‐on therapy, or another prespecified RTX‐based regimen; and reported extractable renal outcomes, including complete renal response (CRR), partial renal response (PRR), overall renal response (ORR), change in proteinuria, or change in serum creatinine. Original clinical studies were considered eligible regardless of whether they were prospective, retrospective, registry‐based, or randomized in design. Studies were excluded if they involved newly diagnosed LN without separately extractable data for the target population, included pediatric‐only populations, evaluated RTX‐based regimens outside the predefined framework, or did not provide extractable renal efficacy data relevant to the present analysis. When randomized controlled trials met the eligibility criteria, only the RTX treatment arm was extracted if the study contributed single‐arm data to the meta‐analysis.

2.3. Study Selection and Data Extraction

Two reviewers independently screened titles and abstracts, assessed full texts for eligibility, and extracted data using a predefined form. Any disagreement was resolved by discussion, consulting a third reviewer when needed.

The following information was extracted from each study: first author, publication year, country, study design, sample size, age, sex distribution, LN type, main pathological class, previous treatment, RTX treatment strategy, follow‐up duration, and renal outcome data. For renal response outcomes, the number of patients from each study achieving CRR, PRR, and ORR was extracted at 6, 12, and 24 months whenever available. If a study did not provide an exact 24‐month assessment but reported a near, corresponding long‐term follow‐up, those data were used in the 24‐month analysis. For continuous outcomes, the extracted data included the reported proteinuria measure, baseline and follow‐up values, mean change and SD when available, and the corresponding units. Baseline and follow‐up serum creatinine values, mean change and SD, and units were also extracted. Study‐specific renal response definitions were accepted as reported in the original publications. Data that were nonstandard or not comparable with the prespecified framework were not entered into the pooled analyses.

2.4. Quality Assessment

Methodological quality was assessed using the non‐comparative version of the Methodological Index for Non‐Randomized Studies (MINORS) [19]. This approach was chosen because the present meta‐analysis synthesized single‐arm data. The same method was applied to randomized controlled trials when only the RTX arm was included. Each item was scored as 0, 1, or 2, with a maximum total score of 16. Quality assessment was performed independently by two reviewers, and discrepancies were resolved by discussion.

2.5. Statistical Analysis

Pooled proportions with 95% confidence intervals (CIs) were calculated for CRR, PRR, and ORR using random‐effects models. Freeman–Tukey double arcsine transformation was applied for proportional outcomes. For continuous variables, pooled mean changes with 95% CIs were synthesized using random‐effects models. Proteinuria values were harmonized as g/day‐equivalent before pooling, and serum creatinine values were standardized to μmol/L. For serum creatinine, the primary analysis used absolute mean change, and relative change from baseline was calculated as a supportive post hoc analysis when baseline values were available. Statistical heterogeneity was evaluated using Cochran's test and the I 2 statistic [20]. Because clinical and methodological differences across studies were anticipated, random‐effects models were used throughout.

Exploratory subgroup analyses were performed according to LN type, treatment method, and combination regimen. Leave‐one‐out sensitivity analysis was conducted to assess the stability of each pooled estimate. Publication bias was evaluated using Egger's test [21]. All tests were two‐sided, and p < 0.05 was considered statistically significant. Statistical analyses were performed using Stata software (version 14.0; Stata Corporation, College Station, TX).

3. Results

3.1. Study Selection, Characteristics, and Quality Assessment

The PRISMA flow diagram is shown in Figure 1. A total of 965 records were identified through the database search, including 297 from PubMed, 355 from Embase, and 313 from Web of Science. After duplicate removal and title and abstract screening, 41 full‐text articles were further assessed for eligibility. Of these, 22 were excluded, and 19 studies were ultimately included in the qualitative and quantitative analyses.

FIGURE 1.

FIGURE 1

Open in a new tab

PRISMA flowchart presenting the study selection process for this systematic review and meta‐analysis.

The 19 included studies enrolled 445 total patients with lupus nephritis who received RTX‐based regimens [12, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39]. Most studies were observational cohorts, whereas two were randomized controlled trials from which only single‐arm data were extracted. The included populations mainly comprised refractory LN, relapsing LN, or relapsing/refractory LN. RTX was administered either as rescue therapy or add‐on therapy, with RTX + Cyclophosphamide (CYC)–based, RTX + Mycophenolate mofetil (MMF)–based, and other/mixed regimens represented across studies. Previous treatment mainly consisted of conventional immunosuppression, most commonly CYC and/or MMF. Follow‐up ranged from 6 months to long‐term observation. Detailed study characteristics are summarized in Table 1.

TABLE 1.

Characteristics of the included studies.

Study Country Study design Sample size (n) Age, years Female, % LN type Main LN class Previous treatment RTX strategy Follow‐up Renal response, n (CRR/PRR/ORR)
6 months 12 months 24 months/last
C. Alejandro Arce‐Salinas (2012) [22] Mexico Prospective cohort 8 28.5 (median, 22–64) 75 Refractory Mixed III/IV/V CYC 6/8 or MMF 2/8; AZA/MMF maintenance Rescue; other/mixed 104 weeks NR NR 2/2/4
Catharina Lindholm (2008) [23] Sweden Retrospective cohort 17 37.0 ± 11.1 88.2 Refractory Mixed II/III/IV CYC 14/17 or MMF 3/17 Add‐on; RTX + CYC‐based Median 24 months (7–58) 2/9/11 NR NR
Chrysanthi Staveri (2025) [24] Greece Multicenter retrospective cohort 8 28.6 ± 12.7 100 Refractory Unclassified/mixed CYC 5/8; MMF 6/8 Rescue; other/mixed ≥ 10 years NR 4/2/6 NR/NR/NR
Esra F. Senturk (2025) [25] Turkey Retrospective cohort 47 42.3 ± 11.1 72.3 Refractory Unclassified/mixed CYC 35/47; MMF 31/47; ±AZA/CSA Rescue; other/mixed Majority ≥ 24 months NR NR NR
Fausta Catapano (2010) [26] United Kingdom Retrospective cohort 11 NR NR Relapsing/refractory Unclassified/mixed ≥ 1 prior IS; LN agents NR Rescue; other/mixed NR NR NR 4/6/10
Iva Gunnarsson (2007) [27] Sweden Prospective cohort 7 30.0 (19–43) 100 Refractory Mixed III/IV CYC 7/7; ±AZA/MMF/CSA/MTX Rescue; RTX + CYC‐based 6 months 3/1/4 NR NR
Jin Zhang [28] China RCT (single‐arm extraction) 42 38.7 ± 6.2 73.8 Refractory Unclassified/mixed High‐dose corticosteroids; no prior CYC Add‐on; RTX + CYC‐based 12 months NR 27/8/35 NR
John N. Boletis (2009) [29] Greece Prospective cohort 10 25.8 ± 6.3 100 Relapsing Mixed III/IV IV CYC induction; MMF/AZA maintenance Rescue; RTX + MMF‐based Median 38 months NR NR 6/0/6
Julián Barbosa‐Arana (2024) [30] Colombia Registry‐based cohort 46 34 (median, IQR 13) 95.7 Refractory Proliferative III/IV ± V CYC 43/46; MMF 41/46 Rescue; other/mixed 12 months NR NR/NR/27 NR
Luca Iaccarino (2015) [31] Italy Registry‐based cohort 68 Overall cohort 37.3 ± 12.4; LN subgroup NR NR Refractory Unclassified/mixed ≥ 1 prior IS; LN agents NR Rescue; other/mixed 27.3 ± 18.5 months NR 21/43/64 NR
Nantakarn Pongtarakulpanit (2020) [32] Thailand Retrospective cohort 13 Overall cohort 36.8 ± 15.2; LN subgroup NR NR Refractory Unclassified/mixed CYC 8/13; MMF 8/13 Rescue; other/mixed 6 months 3/5/8 NR NR
RJ Davies (2013) [33] United Kingdom Prospective cohort 18 30.9 ± 7.1 77.8 Refractory Mixed classes IV CYC 18/18; MMF 17/18 Rescue; RTX + CYC‐based 6 months 11/2/13 NR NR
Rudra Prosad Goswami (2019) [34] India Retrospective cohort 14 23.2 ± 7.6 NR Relapsing/refractory Predominantly proliferative LN CYC 10/14; MMF 8/14 Rescue; other/mixed 6 months 10/4/14 NR NR
Ruth Pepper (2009) [35] United Kingdom Prospective cohort 18 36.9 ± 12.2 83.3 Relapsing Mixed classes CYC 8/18; MMF 8/18; AZA 9/18 Add‐on; RTX + MMF‐based 12 months NR 6/6/12 NR
Sandeep Yadav (2024) [12] India Retrospective cohort 24 28 (IQR 24.5–42) 91.6 Relapsing/refractory Mixed classes CYC/MMF induction; MMF/AZA/CYC maintenance Add‐on; other/mixed Median 29 months 5/17/22 6/8/14 12/5/17
Shahzad Alam (2024) [36] India Prospective cohort 25 Adults > 18 years; mean NR NR Relapsing Mixed III/IV CYC + MMF failure Rescue; other/mixed 6 months 6/7/13 NR NR
Su Jin Choi (2022) [37] South Korea Retrospective cohort 22 39 (IQR 24–50) 90.9 Relapsing/refractory Mixed classes CYC 17/22; MMF 21/22; ±AZA/TAC Rescue; other/mixed 12 months 8/2/10 12/0/12 NR
Thórunn Jónsdóttir (2013) [38] Sweden Prospective cohort 25 35 (mean, 18–71) 92 Refractory Mixed III/IV/V CYC and/or MMF (most); ±AZA/CSA/MTX Rescue; RTX + CYC‐based Mean 36 months (9–95) NR NR 16/6/22
Yemil Atisha‐Fregoso [39] United States RCT (single‐arm extraction) 22 32.3 ± 11.4 81.8 Relapsing/refractory Proliferative III/IV ± V Prior CYC or MMF Rescue; RTX + CYC‐based 96 weeks 5/4/9 7/2/9 4/2/6
Open in a new tab

Note: Renal response data are presented as the number of patients with CRR/PRR/ORR. RCT (single‐arm extraction) indicates a randomized controlled trial from which only single‐arm data were extracted.

Abbreviations: AZA, azathioprine; CRR, complete renal response; CSA, cyclosporine A; CYC, cyclophosphamide; IS, immunosuppressant; LN, lupus nephritis; MMF, mycophenolate mofetil; MTX, methotrexate; NR, not reported; ORR, overall renal response; PRR, partial renal response; RTX, rituximab; TAC, tacrolimus.

The methodological quality of the included studies was moderate to good, with the MINORS total scores ranging from 10 to 13, indicating that most studies had clearly stated aims, clinically appropriate endpoints, and adequate follow‐up, although prospective sample size calculation was generally unavailable (Table 2).

TABLE 2.

Methodological quality assessment using the MINORS tool.

Study Clearly stated aim Inclusion of consecutive patients Prospective collection of data Endpoints appropriate to the aim of the study Unbiased assessment of the study endpoint Follow‐up period appropriate to the major endpoint Loss to follow‐up less than 5% Prospective calculation of the study size Total
C. Alejandro Arce‐Salinas (2012) [22] 2 1 2 2 2 2 1 0 12
Catharina Lindholm (2008) [23] 2 1 0 2 2 2 1 0 10
Chrysanthi Staveri (2025) [24] 2 2 0 2 2 2 1 0 11
Esra F. Senturk (2025) [25] 2 2 0 2 2 2 1 0 11
Fausta Catapano (2010) [26] 2 1 0 2 2 2 1 0 10
Iva Gunnarsson (2007) [27] 2 1 2 2 2 2 1 0 12
Jin Zhang (2015) a [28] 2 2 2 2 2 2 1 0 13
John N. Boletis (2009) [29] 2 1 2 2 2 2 1 0 12
Julián Barbosa‐Arana (2024) [30] 2 2 1 2 2 2 1 0 12
Luca Iaccarino (2015) [31] 2 2 1 2 2 2 1 0 12
Nantakarn Pongtarakulpanit (2020) [32] 2 1 0 2 2 2 1 0 10
RJ Davies (2013) [33] 2 1 2 2 2 2 1 0 12
Rudra Prosad Goswami (2019) [34] 2 1 0 2 2 2 1 0 10
Ruth Pepper (2009) [35] 2 1 2 2 2 2 1 0 12
Sandeep Yadav (2024) [12] 2 1 0 2 2 2 1 0 10
Shahzad Alam (2024) [36] 2 1 2 2 2 2 1 0 12
Su Jin Choi (2022) [37] 2 1 0 2 2 2 1 0 10
Thórunn Jónsdóttir (2013) [38] 2 1 2 2 2 2 1 0 12
Yemil Atisha‐Fregoso a [39] 2 2 2 2 2 2 1 0 13
Open in a new tab

Note: Scores range from 0 to 16 for non‐comparative studies.

Abbreviation: MINORS, Methodological Index for Non‐Randomized Studies.

a

These studies were randomized controlled trials; however, they were assessed using MINORS because only single‐arm data were included in this meta‐analysis.

3.2. Complete Renal Response

Data from nine studies contributed to the analysis of CRR at 6 months [12, 23, 27, 32, 33, 34, 36, 37, 39]. The pooled CRR was 0.33 (95% CI 0.21–0.46), with substantial heterogeneity (I 2 = 63.79%, Figure 2A). At 12 months, 7 studies [12, 24, 28, 31, 35, 37, 39] yielded a pooled CRR of 0.41 (95% CI 0.28–0.54; I 2 = 65.70%, Figure 2B). At 24 months, 6 studies [12, 22, 26, 29, 38, 39] yielded a pooled CRR of 0.42 (95% CI 0.26–0.60; I 2 = 61.31%, Figure 2C). Overall, CRR was achieved in approximately one‐third of patients by 6 months and two‐fifths by 12 to 24 months after treatment inception.

FIGURE 2.

FIGURE 2

Open in a new tab

Forest plots of CRR after RTX‐based therapy in relapsing or refractory LN. (A) CRR at 6 months; (B) CRR at 12 months; (C) CRR at 24 months.

3.3. Partial Renal Response

Data from nine studies contributed to the analysis of PRR at 6 months [12, 23, 27, 32, 33, 34, 36, 37, 39]. The pooled PRR was 0.29 (95% CI 0.16–0.45), with substantial heterogeneity (I 2 = 74.56%, Figure 3A). At 12 months, 7 studies [12, 24, 28, 31, 35, 37, 39] yielded a pooled PRR of 0.23 (95% CI 0.07–0.46; I 2 = 89.91%, Figure 3B). At 24 months, 6 studies [12, 22, 26, 29, 38, 39] yielded a pooled PRR of 0.19 (95% CI 0.07–0.34; I 2 = 59.34%, Figure 3C). Overall, pooled PRR estimates were consistently lower than pooled CRR estimates across all assessed time points.

FIGURE 3.

FIGURE 3

Open in a new tab

Forest plots of PRR after RTX‐based therapy in relapsing or refractory LN. (A) PRR at 6 months; (B) PRR at 12 months; (C) PRR at 24 months.

3.4. Overall Renal Response

Data from 10 studies contributed to the analysis of ORR at 6 months [12, 23, 27, 30, 32, 33, 34, 36, 37, 39]. The pooled ORR was 0.67 (95% CI 0.52–0.80), with substantial heterogeneity (I 2 = 75.08%, Figure 4A). At 12 months, 7 studies [12, 24, 28, 31, 35, 37, 39] yielded a pooled ORR of 0.70 (95% CI 0.51–0.86; I 2 = 84.98%, Figure 4B). At 24 months, 6 studies [12, 22, 26, 29, 38, 39] yielded a pooled ORR of 0.66 (95% CI 0.43–0.86; I 2 = 79.47%, Figure 4C). Overall, approximately two‐thirds of patients achieved an overall renal response during follow‐up.

FIGURE 4.

FIGURE 4

Open in a new tab

Forest plots of ORR after RTX‐based therapy in relapsing or refractory LN. (A) ORR at 6 months; (B) ORR at 12 months; (C) ORR at 24 months.

3.5. Changes in Proteinuria and Serum Creatinine

Details of the original reported measures and standardized values for the continuous renal outcomes are provided in Table S2. Proteinuria data were available from nine studies. Eight studies reported 24‐h or daily urinary protein excretion, whereas one study reported protein‐to‐creatinine ratio data that were converted to g/day‐equivalent for pooling. The pooled mean change in proteinuria was −2.69 g/day‐equivalent (95% CI −3.23 to −2.16), with substantial heterogeneity (I 2 = 86.40%, Figure 5A). This finding was unchanged in direction and similar in magnitude after excluding the protein‐to‐creatinine ratio study (−2.76 g/day; 95% CI −3.34 to −2.19), suggesting that the proteinuria result was not driven by this study. Serum creatinine data were available from seven studies. The pooled mean absolute change in serum creatinine was −11.79 μmol/L (95% CI −19.86 to −3.72), with moderate heterogeneity (I 2 = 58.50%, Figure 5B). In a supportive analysis accounting for baseline values, the pooled relative change in serum creatinine was −12.8% (95% CI −20.0% to −5.7%; I 2 = 44.3%). Overall, RTX‐based therapy was associated with reductions in both proteinuria and serum creatinine.

FIGURE 5.

FIGURE 5

Open in a new tab

Forest plots of changes in renal parameters after RTX‐based therapy in relapsing or refractory LN. (A) Change in proteinuria, g/day‐equivalent; (B) absolute change in serum creatinine, μmol/L.

3.6. Subgroup Analyses

Exploratory subgroup analyses were performed according to LN type, treatment method, and combination regimen. Most between‐subgroup comparisons were not statistically significant. However, the treatment method was associated with significant subgroup effects for CRR at 6 months (p = 0.026) and PRR at 6 months (p < 0.001). In addition, the LN type showed significant subgroup differences for ORR at 12 months (p < 0.001) and change in proteinuria (p = 0.001), while the combination regimen was significantly associated with change in serum creatinine (p = 0.014). Detailed subgroup results are summarized in Table S3, and the corresponding forest plots for subgroup analyses with significant between‐group differences are shown in Figure S1. Given the limited number of studies utilized for analysis in several subgroups, these findings should be interpreted cautiously.

3.7. Sensitivity Analyses

Leave‐one‐out sensitivity analyses showed that the pooled estimates were generally stable (Table S4). For CRR at 6 months, pooled estimates ranged from 0.29 to 0.36, and I 2 ranged from 45.4% to 68.0%. For CRR at 12 months, pooled estimates ranged from 0.34 to 0.44, with I 2 ranging from 12.4% to 70.9%, suggesting that heterogeneity was partially driven by individual studies. For PRR at 6 months, pooled estimates ranged from 0.24 to 0.33, and heterogeneity decreased after the omission of Sandeep Yadav [12]. ORR estimates varied only modestly across the leave‐one‐out analyses. For the continuous outcomes, the pooled mean change in proteinuria remained consistently negative across the leave‐one‐out analyses, ranging from −2.84 to −2.48, indicating that the direction of effect was consistent. The serum creatinine outcome appeared somewhat less robust: the pooled estimate ranged from −14.25 to −8.91, and one leave‐one‐out analysis removed statistical significance, although the direction of effect remained unchanged.

3.8. Assessment of Publication Bias

Publication bias was assessed using Egger's test, with p > 0.05 considered to indicate no statistically significant small‐study effects. No significant evidence of publication bias was detected for any of the assessed outcomes (Table S5). These findings should, however, be interpreted with caution because several analyses were based on a limited number of studies.

4. Discussion

In this systematic review and meta‐analysis of 19 studies involving 445 total patients, RTX used as rescue or add‐on therapy was associated with moderate renal response rates in relapsing LN, refractory LN, and relapsing/refractory LN. Across follow‐up, pooled ORR remained around two‐thirds, whereas pooled CRR increased from approximately one‐third at 6 months to about two‐fifths at 12 and 24 months. Proteinuria and serum creatinine also decreased after treatment. Taken together, these findings provide a quantitative summary of renal outcomes observed after RTX‐based escalation in selected patients with difficult‐to‐treat LN.

The pattern of renal response over time is clinically relevant. ORR remained relatively stable from 6 to 24 months, whereas CRR increased between 6 and 12 months and then showed little further change. This is keeping with the clinical course of LN, in which broader improvement may occur earlier, while CRR often requires more sustained disease control. Recent cohort studies have shown that renal response status is closely linked to long‐term kidney outcome and renal survival, which supports the use of these response endpoints in the present analysis [40, 41].

The continuous renal outcomes help place these findings in a clinically relevant context. Proteinuria showed a marked pooled reduction and remained directionally stable in the leave‐one‐out analyses, making it the most consistent signal in the dataset. Most studies reported proteinuria as 24‐h or daily urinary protein excretion, whereas one study reported protein‐to‐creatinine ratio data. The sensitivity analysis excluding this study yielded a similar estimate, supporting the robustness of the proteinuria finding. Serum creatinine also improved overall, although the result was somewhat less robust in the sensitivity analyses. Because absolute changes in serum creatinine depend on baseline renal function, a supportive relative‐change analysis was added and showed a consistent direction of effect. This difference has a plausible clinical explanation. Proteinuria is often more responsive to short‐ and intermediate‐term changes in inflammatory activity, whereas serum creatinine is altered not only by active nephritis but also by baseline renal reserve, chronic scarring, and irreversible nephron loss. Long‐term studies of LN similarly suggest that residual proteinuria, renal function, and chronic damage remain major determinants of later kidney disease outcomes [42, 43].

The subgroup analyses should be interpreted with caution. Significant subgroup effects were observed for treatment method in the 6‐month CRR and PRR analyses, for LN type in the 12‐month ORR and proteinuria analyses, and for the combination regimen in the serum creatinine analysis. These findings suggest that the effect of RTX may vary according to the clinical setting and the background treatment regimen. Additionally, several subgroup comparisons were based on small numbers of studies, with some categories represented by only one study. For that reason, these findings are better regarded as exploratory rather than confirmatory.

The present results are best understood against the background of current LN management. Treatment strategies have broadened substantially in recent years, with increasing use of multitarget approaches and newer targeted agents [44]. Despite these advances, relapsing and refractory disease remains a major therapeutic challenge, and recent reviews continue to emphasize the lack of uniformity in the definition and management of refractory LN [45]. In routine practice, RTX is often considered when standard therapy has failed or when further escalation with conventional regimens is not desirable; thus, the value of the present study lies in its focus on that specific setting rather than on LN as a whole.

This study has several strengths. It focused on relapsing, refractory, and relapsing/refractory LN rather than pooling all active LN populations together. It also examined renal response at clinically meaningful time points and included both categorical response outcomes and continuous renal parameters. In addition, the main findings were supported by sensitivity analyses showing generally stable effect directions.

However, several limitations should be considered when interpreting these findings. First, most included studies were observational and single‐arm in design, and only two randomized trials contributed RTX‐arm data. The evidence base therefore remains largely non‐comparative and vulnerable to selection bias, residual confounding, and confounding by indication. For this reason, the pooled estimates should be interpreted as renal response rates observed after RTX‐based escalation rather than as comparative efficacy estimates against standard immunosuppression or newer targeted agents. Second, moderate to substantial heterogeneity was present in several pooled analyses, likely reflecting differences in patient phenotype, previous treatment exposure, background immunosuppression, renal response definitions, and follow‐up duration. Third, several subgroup analyses were based on a limited number of studies and should be regarded as exploratory. Long‐term renal outcomes and safety outcomes were also not consistently available for quantitative synthesis. Taken together, these limitations mean that the present findings provide supportive but not definitive evidence for the role of RTX in difficult‐to‐treat LN.

Future studies should focus more specifically on relapsing and refractory LN rather than embedding such patients within broader active LN cohorts. Longer follow‐up is also important for clarifying the durability of renal response, the risk of flare, and the place of RTX in treatment de‐escalation strategies. More consistent reporting of these issues would help define the place of RTX more clearly in future treatment strategies.

5. Conclusion

In patients with difficult‐to‐treat LN, including refractory LN and relapsing/refractory LN, RTX‐based rescue or add‐on therapy was associated with moderate renal response rates and reductions in proteinuria and serum creatinine. Examples of RTX strategies included rescue therapy after inadequate response to conventional immunosuppression and add‐on therapy combined with background agents such as CYC‐ or MMF‐based regimens. The available evidence primarily reflects RTX use after conventional immunosuppression and does not address RTX use after newer regimens such as belimumab. Overall renal response remained relatively high across follow‐up, whereas complete renal response increased modestly over time. Exploratory subgroup analyses suggested that treatment method, LN type, and combination regimen may influence some outcomes, but these findings should be interpreted cautiously because the evidence was mainly derived from observational, single‐arm studies. Further prospective comparative studies are needed to define the role of RTX in difficult‐to‐treat LN.

Author Contributions

X.W., S.L., and X.S. wrote the manuscript; Z.D. and Q.W. designed the research; X.W., S.L., X.S., F.W., G.T., W.D., and M.L. performed the research; X.W., S.L., F.W., and W.D. analyzed the data.

Funding

The authors have nothing to report.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Table S1: The search strategy of each database.

Table S2: Measurement details and standardized values for continuous renal outcomes.

Table S3: Summary of overall estimates and exploratory subgroup analyses.

Table S4: Summary of leave‐one‐out sensitivity analyses.

Table S5: Egger's test for small‐study effects across outcomes.

Figure S1: Forest plots of subgroup analyses with significant between‐group differences. (A) CRR at 6 months by treatment method; (B) PRR at 6 months by treatment method; (C) ORR at 12 months by LN type; (D) change in proteinuria by LN type; (E) change in serum creatinine by combination regimen.

CTS-19-e70730-s001.docx (283.6KB, docx)

Acknowledgments

We thank Medjaden Inc. for scientific editing and proofreading of this manuscript.

Contributor Information

Qian Wang, Email: 794317255@qq.com.

Zheyi Dong, Email: shengdai26@163.com.

References

  • 1. Rovin B. H., Ayoub I. M., Chan T. M., et al., “Executive Summary of the KDIGO 2024 Clinical Practice Guideline for the Management of Lupus Nephritis,” Kidney International 105 (2024): 31–34. [DOI] [PubMed] [Google Scholar]
  • 2. Parodis I., Rovin B. H., Tektonidou M. G., et al., “Lupus Nephritis,” Nature Reviews Disease Primers 11 (2025): 69. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3. Sammaritano L. R., Askanase A., Bermas B. L., et al., “2024 American College of Rheumatology (ACR) Guideline for the Screening, Treatment, and Management of Lupus Nephritis,” Arthritis & Rhematology 77 (2025): 1115–1135. [DOI] [PubMed] [Google Scholar]
  • 4. Mok C. C., So H., Hamijoyo L., et al., “The 2024 APLAR Consensus on the Management of Lupus Nephritis,” International Journal of Rheumatic Diseases 28 (2025): e70021. [DOI] [PubMed] [Google Scholar]
  • 5. Sales G. T. M., Macedo N. J., Reis Neto E. T. D., and Kirsztajn G. M., “Current Treatment of Lupus Nephritis: An Overview of the New Guidelines,” Jornal Brasileiro de Nefrologia 47 (2025): e20250092. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6. De Vriese A. S., Sethi S., and Fervenza F. C., “Lupus Nephritis: Redefining the Treatment Goals,” Kidney International 107 (2025): 198–211. [DOI] [PubMed] [Google Scholar]
  • 7. Garal‐Pantaler E., Schultze M., Georgiou M. E., Pignot M., Gairy K., and Hunnicutt J. N., “Real‐World Burden of Immunosuppressant‐Treated Lupus Nephritis: A German Claims Database Analysis,” Rheumatology and Therapy 11 (2024): 113–127. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8. Nordmann‐Gomes A., Cojuc‐Konigsberg G., Hernández‐Andrade A., et al., “Lupus Nephritis Randomised Controlled Trials: Evidence Gaps and Under‐Represented Groups,” Lupus Science & Medicine 11 (2024): e001331. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9. Mo S., Li Y., He J., and Lin L., “Progress of Rituximab in the Treatment of Systemic Lupus Erythematosus and Lupus Nephritis,” Frontiers in Medicine (Lausanne) 11 (2024): 1472019. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10. Mok C. C., “Anti‐CD20 Therapy in Lupus Nephritis: A Revisit,” Rheumatology and Immunology Research 6 (2025): 66–69. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11. Tanaka Y., Nakayamada S., Yamaoka K., Ohmura K., and Yasuda S., “Rituximab in the Real‐World Treatment of Lupus Nephritis: A Retrospective Cohort Study in Japan,” Modern Rheumatology 33 (2023): 145–153. [DOI] [PubMed] [Google Scholar]
  • 12. Yadav S., Balakrishnan C., Mangat G., and Kothari J., “Rituximab as Add‐On Therapy in Patients With Resistant Lupus Nephritis Who Have Failed Induction or Maintenance Therapy With Other Agents: A Real‐World Experience From a Single Center in Mumbai,” Lupus 33 (2024): 88–95. [DOI] [PubMed] [Google Scholar]
  • 13. Rovin B. H., Teng Y. K. O., Ginzler E. M., et al., “Efficacy and Safety of Voclosporin Versus Placebo for Lupus Nephritis (AURORA 1): A Double‐Blind, Randomised, Multicentre, Placebo‐Controlled, Phase 3 Trial,” Lancet 397 (2021): 2070–2080. [DOI] [PubMed] [Google Scholar]
  • 14. Furie R. A., Rovin B. H., Garg J. P., et al., “Efficacy and Safety of Obinutuzumab in Active Lupus Nephritis,” New England Journal of Medicine 392 (2025): 1471–1483. [DOI] [PubMed] [Google Scholar]
  • 15. Tang Z., Huang Y., Lin Y., et al., “Efficacy and Safety of Rituximab in Patients With Lupus Nephritis: A Systematic Review and Meta‐Analysis,” Clinical Nephrology 103 (2025): 53–62. [DOI] [PubMed] [Google Scholar]
  • 16. Zhao X., Yang S. Q., Li M., and Wang Y. G., “Effectiveness and Safety of B Cell‐Targeting Biologics in the Treatment of Lupus Nephritis: A Systematic Review and Network Meta‐Analysis,” Renal Failure 46 (2024): 2416605. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17. Page M. J., McKenzie J. E., Bossuyt P. M., et al., “The PRISMA 2020 Statement: An Updated Guideline for Reporting Systematic Reviews,” BMJ 372 (2021): n71. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18. Page M. J., Moher D., Bossuyt P. M., et al., “PRISMA 2020 Explanation and Elaboration: Updated Guidance and Exemplars for Reporting Systematic Reviews,” BMJ 372 (2021): n160. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19. Slim K., Nini E., Forestier D., Kwiatkowski F., Panis Y., and Chipponi J., “Methodological Index for Non‐Randomized Studies (Minors): Development and Validation of a New Instrument,” ANZ Journal of Surgery 73 (2003): 712–716. [DOI] [PubMed] [Google Scholar]
  • 20. Higgins J. P. and Thompson S. G., “Quantifying Heterogeneity in a Meta‐Analysis,” Statistics in Medicine 21 (2002): 1539–1558. [DOI] [PubMed] [Google Scholar]
  • 21. Egger M., Davey Smith G., Schneider M., and Minder C., “Bias in Meta‐Analysis Detected by a Simple, Graphical Test,” BMJ 315 (1997): 629–634. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22. Arce‐Salinas C. A., Rodríguez‐García F., and Gómez‐Vargas J. I., “Long‐Term Efficacy of Anti‐CD20 Antibodies in Refractory Lupus Nephritis,” Rheumatology International 32 (2012): 1245–1249. [DOI] [PubMed] [Google Scholar]
  • 23. Lindholm C., Börjesson‐Asp K., Zendjanchi K., Sundqvist A. C., Tarkowski A., and Bokarewa M., “Longterm Clinical and Immunological Effects of Anti‐CD20 Treatment in Patients With Refractory Systemic Lupus Erythematosus,” Journal of Rheumatology 35 (2008): 826–833. [PubMed] [Google Scholar]
  • 24. Staveri C., Lykoura C., Melissaropoulos K., and Liossis S. C., “Favorable Outcomes for Patients With Refractory Systemic Lupus Erythematosus Treated With Rituximab as Evidenced With a Follow‐Up of ≥ 10 Years: A Real‐World Evidence Study,” Rheumatology International 45 (2025): 127. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25. Senturk E. F., Ogun H., Durucan I., Yurttas B., and Ugurlu S., “Rituximab Treatment in Lupus Nephritis Resistant to Conventional Therapy: A Single‐Centre Experience,” Mediterranean Journal of Rheumatology 36 (2025): 229–235. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26. Catapano F., Chaudhry A. N., Jones R. B., Smith K. G., and Jayne D. W., “Long‐Term Efficacy and Safety of Rituximab in Refractory and Relapsing Systemic Lupus Erythematosus,” Nephrology, Dialysis, Transplantation 25 (2010): 3586–3592. [DOI] [PubMed] [Google Scholar]
  • 27. Gunnarsson I., Sundelin B., Jónsdóttir T., Jacobson S. H., Henriksson E. W., and van Vollenhoven R. F., “Histopathologic and Clinical Outcome of Rituximab Treatment in Patients With Cyclophosphamide‐Resistant Proliferative Lupus Nephritis,” Arthritis and Rheumatism 56 (2007): 1263–1272. [DOI] [PubMed] [Google Scholar]
  • 28. Zhang J., Zhao Z., and Hu X., “Effect of Rituximab on Serum Levels of Anti‐C1q and Antineutrophil Cytoplasmic Autoantibodies in Refractory Severe Lupus Nephritis,” Cell Biochemistry and Biophysics 72 (2015): 197–201. [DOI] [PubMed] [Google Scholar]
  • 29. Boletis J. N., Marinaki S., Skalioti C., Lionaki S. S., Iniotaki A., and Sfikakis P. P., “Rituximab and Mycophenolate Mofetil for Relapsing Proliferative Lupus Nephritis: A Long‐Term Prospective Study,” Nephrology, Dialysis, Transplantation 24 (2009): 2157–2160. [DOI] [PubMed] [Google Scholar]
  • 30. Barbosa‐Arana J., López‐López J. D., Guerra‐Zarama S., et al., “Outcomes With the Use of Rituximab in Patients With Refractory Lupus Nephritis in a Colombian Cohort,” Revista Colombiana de Reumatología (English Edition) 31 (2024): 143–149. [Google Scholar]
  • 31. Iaccarino L., Bartoloni E., Carli L., et al., “Efficacy and Safety of Off‐Label Use of Rituximab in Refractory Lupus: Data From the Italian Multicentre Registry,” Clinical and Experimental Rheumatology 33 (2015): 449–456. [PubMed] [Google Scholar]
  • 32. Pongtarakulpanit N., Pisitkun P., and Ngamjanyaporn P., “Efficacy and Safety of Rituximab Biosimilar in Refractory Lupus,” Lupus Science & Medicine 7 (2020): e000442. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33. Davies R. J., Sangle S. R., Jordan N. P., et al., “Rituximab in the Treatment of Resistant Lupus Nephritis: Therapy Failure in Rapidly Progressive Crescentic Lupus Nephritis,” Lupus 22 (2013): 574–582. [DOI] [PubMed] [Google Scholar]
  • 34. Goswami R. P., Sircar G., Sit H., Ghosh A., and Ghosh P., “Cyclophosphamide Versus Mycophenolate Versus Rituximab in Lupus Nephritis Remission Induction: A Historical Head‐To‐Head Comparative Study,” Journal of Clinical Rheumatology 25 (2019): 28–35. [DOI] [PubMed] [Google Scholar]
  • 35. Pepper R., Griffith M., Kirwan C., et al., “Rituximab Is an Effective Treatment for Lupus Nephritis and Allows a Reduction in Maintenance Steroids,” Nephrology, Dialysis, Transplantation 24 (2009): 3717–3723. [DOI] [PubMed] [Google Scholar]
  • 36. Alam S., Mazumder M. A., Sharma M., Mahanta P. J., Parry M. A., and Doley P. K., “Rituximab Treatment in Resistant Lupus Nephritis: A Single‐Center Prospective Study,” Clinical Nephrology 102 (2024): 1–7. [DOI] [PubMed] [Google Scholar]
  • 37. Choi S. J., Ahn S. M., Oh J. S., et al., “Initial Preserved Renal Function as a Predictor of Favorable Renal Response to Rituximab in Refractory or Relapsing Lupus Nephritis: A Single‐Center Cohort Study in Korea,” Journal of Rheumatic Diseases 29 (2022): 22–32. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38. Jónsdóttir T., Zickert A., Sundelin B., Henriksson E. W., Van vollenhoven R. F., and Gunnarsson I., “Long‐Term Follow‐Up in Lupus Nephritis Patients Treated With Rituximab‐Clinical and Histopathological Response,” Rheumatology 52 (2013): 847–855. [DOI] [PubMed] [Google Scholar]
  • 39. Atisha‐Fregoso Y., Malkiel S., Harris K. M., et al., “Phase II Randomized Trial of Rituximab Plus Cyclophosphamide Followed by Belimumab for the Treatment of Lupus Nephritis,” Arthritis and Rheumatology 73 (2021): 121–131. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40. Yap D. Y. H., Xu X., Juliao P. C., et al., “Long‐Term Kidney Outcome of Lupus Nephritis by Renal Response Status,” Kidney International Reports 9 (2024): 3532–3541. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41. Urowitz M., Georgiou M. E., Touma Z., et al., “Renal Response Status to Predict Long‐Term Renal Survival in Patients With Lupus Nephritis: Results From the Toronto Lupus Cohort,” Lupus Science & Medicine 11 (2024): e001264. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42. Farinha F., Barreira S., Couto M., et al., “Risk of Chronic Kidney Disease in 260 Patients With Lupus Nephritis: Analysis of a Nationwide Multicentre Cohort With up to 35 Years of Follow‐Up,” Rheumatology (Oxford, England) 64 (2025): 1201–1209. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43. Polycarpo S. B. and Kirsztajn G. M., “Renal Outcomes in the Long‐Term Follow‐Up of Lupus Nephritis,” Jornal Brasileiro de Nefrologia 47 (2025): e20240118. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44. Avasare R., Drexler Y., Caster D. J., Mitrofanova A., and Jefferson J. A., “Management of Lupus Nephritis: New Treatments and Updated Guidelines,” Kidney360 4 (2023): 1503–1511. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45. Ding T., Fan Y., Zhang X., and Zhang Z., “Refractory Lupus Nephritis: Current Understanding and Emerging Treatment Strategies,” Clinical and Experimental Medicine 26 (2025): 78. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Table S1: The search strategy of each database.

Table S2: Measurement details and standardized values for continuous renal outcomes.

Table S3: Summary of overall estimates and exploratory subgroup analyses.

Table S4: Summary of leave‐one‐out sensitivity analyses.

Table S5: Egger's test for small‐study effects across outcomes.

Figure S1: Forest plots of subgroup analyses with significant between‐group differences. (A) CRR at 6 months by treatment method; (B) PRR at 6 months by treatment method; (C) ORR at 12 months by LN type; (D) change in proteinuria by LN type; (E) change in serum creatinine by combination regimen.

CTS-19-e70730-s001.docx (283.6KB, docx)

Articles from Clinical and Translational Science are provided here courtesy of Wiley

RESOURCES