1. Introduction
Idiopathic membranous nephropathy (IMN) is an immune-mediated disease with subepithelial immune complex deposition and glomerular basement membrane changes [1]. Patients are mainly manifested by proteinuria, hypoproteinemia, hypertension, and edema [2]. The auto-antigens like M-type phospholipase A2 receptor (PLA2R) [3] and thrombospondin type 1 domain containing 7A (THSD7A) [4] can be detected in most MN patients. As surrogates of immunologic activity, PLA2R and THSD7A provide a new direction for the diagnosis and treatment of IMN [5].
The treatment of IMN is always controversial. Most patients with MN and non-nephrotic proteinuria are sensitive to supportive therapy which includes angiotensin-converting-enzyme (ACE) inhibitors and angiotensin-receptor blockers (ARBs) [6]. However, in patients with the highest risk of disease progression and the decrease of proteinuria is seldom > 30% from pretreatment value, It mainly used alkylating agents and calcineurin inhibitors in treatment [7, 8]. Alkylating agent (mainly cyclophosphamide) is usually used for high-risk patients who failed to respond to 6-month supportive therapy [9] and can effectively prevent IMN from ESRD and death. Calcineurin inhibitors (mainly tacrolimus) can improve therapeutic effects and reduce toxicity to achieve the purpose of reducing the rate of proteinuria and renal function loss [9-12]. However, the efficacy of cyclophosphamide and tacrolimus in IMN treatment has controversy.
RTX is an anti-CD20 monoclonal antibody [13] that targets to trigger B cell to death through apoptosis [14, 15], complement-mediated cytotoxicity [16], and antibody-dependent cellular cytotoxicity [17, 18]. The efficacy of RTX as initial treatment is similar to traditional immunosuppressants treatment (alkylating agents plus steroid combination therapy and calcineurin inhibition therapy) [19-21]. The standard protocol of RTX for MN is intravenous injection of 375 mg/m2 every week for 4 weeks, or two doses of 1000 mg infusions 2 weeks apart or 1×375 mg/m2 intravenously at a 14-day interval [22, 23]. Since the ideal RTX dosage and the long-term side effects of a larger cumulative dose are unknown, and some studies have demonstrated the effectiveness of low-dose regimens versus conventional dosage (standard protocol or two doses of RTX (500 mg each) ) [22], nephrologists argued for the use of low doses of RTX (B cell-driven treatment or the dose of RTX is 1×375 mg/m2 or 2×375 mg/m2 or 100 mg of RTX or two doses of RTX (500mg each) infusion 7 days apart).
Compared to KDIGO (Kidney Disease: Improving Global Outcomes) 2012 guidelines, KDIGO 2020 guidelines emphasize the role of RTX in patients with a moderate (pre-study proteinuria greater than 3.5 g/d and less than 8 g/d) and high (pre-study proteinuria more than 8 g/d) risk of IMN, but evidence regarding the efficacy and safety of RTX in MN treatment remains elusive [9, 24]. There is still not meta-analysis that can provide evidence of the RTX therapeutic advantages by stratifying inclusion studies based on the risk of progression of renal function loss in patients. At the same time, the efficacy of different doses of RTX is controversial and uncertain. To guide the clinical use of RTX, this article used statistical methods to summarize several studies on MN treatment with RTX or conventional therapy and explore the therapeutic effect of different RTX dosages (low dose versus standard dose).
2. Methods
2.1. Data sources and searches
We systematically searched the randomized controlled trials (RCTs), clinical studies, and cohort studies on the efficacy and safety of RTX in the treatment of adult MN from Pubmed, Web of Science, Cochrane Library, Embase, ScienceDirect, and SinoMed from inception until 1 December 2020. Additional articles come from the references of reviews, meta-analyses, systematic reviews, reports, and guidelines. All these articles are English. The appropriate keywords and medical subject headings in logical combinations we used included “Rituximab”, “Rituximab CD20 Antibody”, “Mabthera”, “IDEC-C2B8 Antibody”, “Rituxan”, “GP2013”, “Membranous Nephropathy” “Membranous Glomerulonephritides”, “Primary Membranous Nephropathy”, “Idiopathic Membranous Glomerulonephritis” and “Heymann Nephritis”.
2.2. Inclusion and exclusion criteria
The inclusion criteria contains: a)RCTs, cohort studies, clinical trials, case-control study, and single-center experience; b)All patients in the selected studies were confirmed to have IMN by renal biopsy. Each patient was older than 18 years old, and c) Outcome events and remission (CR or/and PR) need to be reported. The CR was defined as proteinuria < 0.3 g/day accompanied by a normal serum albumin concentration and normal serum creatinine (Scr). The PR was defined as proteinuria < 3.5 g/day and a 50% or greater reduction from peak values, accompanied by an improvement or normalization of the serum albumin concentration and stable Scr. The study needs to evaluate the efficacy of RTX in patients for at least 6 months.
The exclusion criteria contains: a) Any review articles, books, and documents, meta-analysis, meeting records, case reports, ongoing research, and other non-randomized studies were excluded; b) The results of the study were related to the use of RTX in patients with secondary MN or other patients with non-MN, and c) The study looked at RTX in combination with other unsupportive agents for MN.
2.3. Data extraction and quality assessment
Two independent reviewers reviewed the original study reports and sifted and extracted the data. Data includes the type of study, time of publication, author, region or country, study duration, intervention, and control. Additionally, reviewers also extracted demographic, clinical, immunological, and pathological variables of patients. Outcome data is also extracted from the study articles by the observers, mainly including CR, PR, TR. Quality assessment is carried out through the Cochrane Collaboration’s tool. If there is a disagreement between two independent reviewers about the data, then the independent reviewers should discuss and reach a consensus.
2.4. Data analysis
To analyze whether RTX has an advantage over other conventional therapies in the treatment of MN, this meta-analysis using the Stata software (Stata Corporation, Texas, USA). In subgroup analysis, patients are grouped according to their pre-study proteinuria. RTX dose is used as a variable for the analysis of the therapeutic effect of MN. Inconsistency index (I2 ) statistics are used to analyze the heterogeneity of the included studies. To analyze the endpoints of TR, we choose the Mantel-Haenszel method of the random effects model and applied effect index odds ratio (OR), this is because significant heterogeneity (P < 0.1 orI2 > 50%) exists in included studies. To analyze the effects of different doses of RTX on Scr, proteinuria and ALB after treatment, we used the invariance fixed-effect method and invariance random effect method and selected the mean difference (MD) as the continuous outcome. Score statistics and precise binomial method were used to calculate 95% CI for effect sizes. STATA/MP-14 software was used for all data analysis. And all 2-sidedP < 0.05 was considered to indicate statistical significance.
3. Results
3.1. Description of included trials
A total of 1675 relevant pieces of literature were identified in this paper through a database search (Figure. 1), including 281 pieces from Pubmed, 565 pieces from Web of Science, 73 pieces from Cochrane Library, 550 pieces from Embase, 82 pieces from Science Direct, 124 pieces from SinoMed. The Endnote software was used to remove 1102 repeated articles between different databases. Two inspectors removed reviews, Case reports, Meta-analysis, and Off-topic types. Finally, six articles [25-30] were included in the study comparing the therapeutic advantages of RTX with conventional therapy. Twelve studies [30-42] were included in the efficacy studies of different doses of RTX. The studies provided information on 643 patients. Baseline data characteristics of the included studies are summarized in Table 1 and Table 2, respectively.
3.2. Quality assessment
We assessed the quality of the included studies by Cochrane Handbook and assessed the quality of the included studies by giving “author judgments” of ”low risk” or ”unclear risk” or ”high risk” for different types of offset evaluation entries. The results showed that most of the included studies were low-risk or unclear risk related to the offset items, indicating that the quality of the included studies in this paper was good.
3.3. Efficacy of RTX compared with other therapeutic agents in MN (Totally remission)
Forest plots (Figure. 2) show the overall and subgroup efficacy and safety of RTX in the treatment of MN compared to conventional agents. Six articles reported TR and all RTX regimens are uniform dosages (4×375 mg/m2). The included studies showed that there is no significant difference in TR (OR = 1.659, 95% CI 0.668 to 4.119,I2 of 61.9% indicating heterogeneity, P= 0.275) for MN with RTX compared to conventional therapy. In the subgroup analysis, for patients with pre-study proteinuria < 8 g/d, RTX had a higher TR (OR = 2.663, 95% CI 1,361 to 5.210, I2 of 0.0% indicating heterogeneity, P = 0.004) than conventional therapy, however, for patients with pre-study proteinuria > 8 g/d, there was no significant difference in TR (OR = 0.691, 95% CI 0.064 to 7.423,I2 of 83.3% indicating heterogeneity, P= 0.761).
3.4. Effect of RTX dose on the therapeutic of IMN
3.4.1. Proteinuria (g/24 hour)
The post-treatment proteinuria index was evaluated in twelve studies (Figure. 3A). Results of meta-analysis showed RTX significantly reduced proteinuria (MD = -4.885, 95% CI -6.340 to -3.370,I2 of 69.2% indicating heterogeneity, P< 0.0001) in patients with MN. In the subgroup analysis, low-dose RTX was not effective than conventional dosage of RTX in reducing proteinuria (Low dose MD = -3.999, 95% CI -6.177 to -1.820,I2 of 33.3% indicating heterogeneity, P< 0.0001; standard dose MD = -5.220, 95% CI -7.160 to -3.279 , I2 of 78.4% indicating heterogeneity,P < 0.0001).
3.4.2. Serum albumin (g/L)
There are eleven articles reported the lever of ALB (Figure. 3B), which shows that RTX treatment was associated with a statistically significant increase in ALB (MD = 0.884, 95% CI 0.740 to 1.185,I2 of 72.0% indicating heterogeneity, P< 0.0001). In the subgroup analysis, low dose RTX improved ALB (MD = 0.601, 95% CI 0.052 to 1.150, I2 of 83.2% indicating heterogeneity, P = 0.032). However, RTX can be more significantly in increasing ALB (MD = 0.963, 95% CI 0.740 to 1.185, I2 of 53.7% indicating heterogeneity,P < 0.0001) in the standard dose subgroup.
3.4.3. Serum creatinine (mg/dL)
Scr at the end of treatment was reported in nine articles (Figure. 4). Treatment with RTX led to no significant reduction Scr (MD = -0.027, 95% CI -0.196 to 0.142, I2 of 48.8% indicating heterogeneity, P = 0.758). In subgroup analyses, both low dose RTX and standard dose RTX did not significantly reduce Scr (Low dose MD = 0.110, 95% CI -0.050 to 0.271, I2 of 0.0% indicating heterogeneity, P = 0.177. Standard dose MD = -0.176, 95% CI -0.456 to 0.103, I2 of 56.5% indicating heterogeneity, P = 0.216).