3.1 Search results and trial characteristics
Initial literature search identified 1627 relevant articles, of which 980 were from PubMed, 537 from EMBASE, 90 from Cochrane library and 20 from WOS. Using endnote software, 107 repeated studies were removed. 1486 studies were excluded after screening the titles and abstracts and remaining 34 articles were acquired in full-text form and 8 studies were found appropriate for inclusion in this systemic meta-analysis (Fig. 1) [6, 10-16]. The clinical and therapeutic data for 588 patients diagnosed with MN are provided by included studies. The characteristics of included studies are summarized in table 1.
3.2 Quality assessment
Quality of the included studies was assessed according to NOS score and Cochrane Handbook. The NOS score of these studies range from 6 to 8 and “low bias” base on the Cochrane handbook, indicating high methodological quality of included studies (as shown in Table 2).
3.3 Efficacy of RTX in MN patients
3.3.1 Efficacy of RTX on TR and CR in MN
As Fig.2 shown, data from pooled 8 studies showed no significant difference in the efficiency between RTX and control treatment to induce TR (OR=1.90, 95%CI 0.70 to 5.16; I2 of 82% indicating heterogenicity, P=0.21). However, CR rate in RTX group is significantly higher than in control group (OR=2.60, 95%CI 1.72 to 3.94; I2 of 33% indicating no heterogenicity, P<0.0001) (Fig. 3).
3.4 Efficacy of high-dose and low-dose RTX protocols in MN patients
3.4.1 Efficacy of high-dose and low-dose RTX in MN patients compared with control
Two studies reported TR with high-dose RTX. Pooled data from these studies indicated that high-dose RTX exhibited significantly higher TR (OR=5.49, 95%CI 2.59 to 11.60; I2 of 0% indicating no heterogenicity, P<0.0001) than control group (Fig 4).
Similarly, two studies reported TR with low-dose RTX treatment. But pooled analysis of data revealed no significant difference between low-dose rituximab group and control group (OR=2.05, 95%CI 0.79 to 5.33; I2 of 0% indicating no heterogenicity, P=0.14) (Fig. 5).
3.4.2 Comparison between high-dose and low-dose RTX’s efficiency in MN patient
3.4.2.1 TR and CR
Three studies compared TR and CR rates between high-dose and low-dose RTX treatment in MN. These studies totally covered 119 patients, 66 of whom were treated with high-dose RTX and 53 with low-dose RTX. As shown in Fig. 6 and 7, there is no significant difference between high dose and low dose RTX group on TR and CR. These findings suggest the overall equivalence between high dose and low dose RTX treatment groups.
After further stratification by follow-up period, we noticed that TR rate was higher in high-dose RTX group than in low-dose RTX group in those studies followed up ≤1year. But in studies followed up >1year, there was no significant difference between high-dose and low-dose RTX groups in the TR rate (Fig. 6). The CR rates was not dramatically different between high-dose and low-dose RTX groups in studies with follow-up period ≤1year or >1year (Fig. 7). These data implied that the longer follow-up period is needed to evaluate the response of low-dose RTX in patients with MN.
3.4.2.2 Biochemical indicators
Proteinuria (g/24hr): Two studies reported 24-hour urinary protein at the end of treatment with high dose and low dose RTX respectively. As shown in Fig. 8, the statistical analysis showed no significant difference (MD=0.14, 95%CI -0.08 to 0.35; I2 of 0% indicating no heterogenicity, p=0.22) between the two groups.
Serum albumin(g/dl): Two studies evaluated serum albumin level after different RTX regiments treatment. Pooled analysis of these data revealed that there was no significant difference between high-dose and low-dose RTX regiment groups (MD=0.26, 95%CI 0.15 to 0.37; I2 of 0% indicating no heterogenicity, p=0.33) (Fig. 9).
Serum creatinine (mg/dl): Only one study assessed the serum creatine (Scr) levels after different dosage RTX treatment (MD=0.60, 95%CI 0.07 to 1.13) and the result showed no significant difference between groups (Fig. 10).
PLA2R-Antibody-depleted patients: Two studies tested the serum PLA2R antibody levels in patients with IMN at end of treatment. Forty-one patients were assigned to high-dose RTX treatment group and forty patients to low-dose RTX group. The fixed- effect model was used for evaluation because of minimal heterogeneity. As shown in Fig. 11, pooled data indicated that there were more PLA2R antibody depleted patients in high-dose RTX treatment group when compared with low-dose RTX group (MD=3.20, 95%CI 1.09 to 9.38; I2 of 0% indicating no heterogenicity, p=0.03).
3.4.2.3 Safety and serious adverse effects (SAE)
Most studies reported only mild infusion reaction which could be resolved spontaneously with reducing the drug infusion rate, temporary discontinuation or with minor supportive treatment. These results suggest that both high dose and low dose RTX were well-tolerated in most patients. Here, to understand the adverse events in different RTX dose regiments with accuracy, only serious side events which required hospitalization and life-threatening (grade 3 or higher) were included. The statistical analysis showed that there was a tendency for patients receiving RTX treatment to have less SAEs than patients in the control group, but the difference was not significant (Fig.12). Only one study compared SAE between high-dose and low-dose RTX in patients with primary MN. And one patient on high-dose protocol had a severe adverse reaction. There was no such report in patients assigned to low-dose RTX treatment (Fig. 13).