Introduction

Renal involvement affects up to 60% of individuals with systemic lupus erythematosus (SLE), and lupus nephritis (LN) continues to be the leading cause of morbidity and mortality in SLE [1,2,3]. Despite decades of advancements in treatment, a considerable number of patients have renal impairment, with 10% developing renal failure after 10 years [4]. Cyclophosphamide (CYC) regimens have long since been regarded as the gold standard for achieving renal remission and avoiding renal flares, because they improve renal outcomes. However, considerable drug-related side effects, such as an increased risk of severe infection and ovarian toxicity, offset these advantages [5].

Mycophenolate mofetil (MMF) is a hypoxanthine nucleotide dehydrogenase inhibitor that specifically decreases T/B lymphocyte proliferation, inhibits antibody generation, controls the immune system, and reduces the formation of circulating immune complexes in renal tissue [6]. Tacrolimus, an effective inhibitor of human T cell proliferation, binds to tacrolimus-binding proteins on T cells and inhibits calcineurin [7]. Immunosuppressive treatments such as tacrolimus are likely to offer therapeutic advantages because of their immunomodulatory effect, as T cell activation is implicated in the etiology of LN [8]. In LN induction therapy, meta-analysis has indicated that MMF seems to be superior to CYC in increasing serum complement C3 and achieving complete remission regardless of ethnicity, as well as having fewer treatment-related side effects [9]. In meta-analysis, tacrolimus was proven to be more efficacious and safer than intravenous CYC as an induction treatment. The general dosing in a tacrolimus regimen was estimated at 3–4 mg twice daily during the induction phase of LN treatment [10]. However, the effectiveness of low-dose tacrolimus in the treatment of LN is uncertain, and it is debatable whether tacrolimus is more effective and safer than MMF in LN therapy. Due to the small number of studies conducted and their small sample sizes, these conclusions are controversial [11,12,13]. Using meta-analysis and systematic review, this study aimed to assess the efficacy and safety of tacrolimus and MMF as induction therapy and low-dose tacrolimus as a treatment for LN.

Materials and methods

Identification of eligible studies and data extraction

We performed an exhaustive search for studies that examined the efficacy and safety of tacrolimus compared with MMF and tacrolimus at low dose in patients with LN. The PubMed, EMBASE, and Cochrane Controlled Trials Register databases were searched to identify available articles (up to June 2022). The following keywords and subject terms were used in the search: “lupus nephritis,” “tacrolimus,” and “mycophenolate mofetil.” The reference lists of all the retrieved articles were reviewed to identify additional studies that were not included in the electronic databases. Randomized controlled trials (RCTs) were included if they met the following criteria: (1) compared tacrolimus with MMF as induction therapy for LN; (2) provided endpoints for efficacy at 6 months after induction therapy and safety during the follow-up period; or (3) RCT or prospective cohort studies including low-dose tacrolimus for LN. The exclusion criteria were as follows: (1) inclusion of duplicate data and (2) lack of adequate data for inclusion.

The efficacy outcomes were as follows: number of patients who achieved (1) complete remission and (2) partial remissions. Complete or partial remission was defined on the basis of the remission criteria used in each trial. The safety outcome was the number of patients who experienced infection, serious infection, or withdrawal due to adverse events (WAE). The following information was extracted from each study: first author, ethnicity, year of publication, kidney biopsy class, number of patients treated with tacrolimus and MMF, efficacy outcome at 6 months after induction therapy, safety results during the follow-up period, and number of patients treated with low-dose tacrolimus. The methodological quality of the RCTs was determined using Jadad scores [14]. The Jadad score ranged from 0 to 5, with higher scores indicating better trial quality. The Newcastle–Ottawa Scale was used to score the quality of each study included in the meta-analysis [15]. Scores ranging from 6 to 9 indicated high methodological quality. This meta-analysis was conducted in accordance with the guidelines provided by the PRISMA statement [16].

Evaluation of statistical associations

The effect size of the study outcomes was represented as an odds ratio (OR) for dichotomous data or standardized mean difference (SMD) for continuous data and the corresponding 95% confidence intervals (95% CIs). We assessed intra- and inter-study variations and heterogeneities using Cochran’s Q‑statistics [17]. The heterogeneity test was used to evaluate the null hypothesis that all studies evaluated the same effect. When a significant Q‑statistic (p < 0.10) indicated heterogeneity across studies, the random-effects model was used for the meta-analysis; otherwise, the fixed-effects model was used. The fixed-effects model assumes that all studies estimate the same underlying effect and considers only intra-study variations. We quantified the impact of heterogeneity using [18]:

$$\begin{array}{cc} I^{2}=100{\%}\times \left(Q-df\right)/Q & [18] \end{array}$$
(1)

where I2 measures the degree of inconsistency between studies and determines whether the percentage of total variation across studies is due to heterogeneity rather than chance. I2 values ranged between 0% and 100%, and 25%, 50%, and 75% were referred to as low, moderate, and high estimates, respectively. Statistical manipulations were performed using the Comprehensive Meta-Analysis Program (Biostat, Englewood, NJ, USA).

Evaluation of publication bias

Funnel plots are normally used to detect publication bias. However, because they require a range of studies with different sizes and subjective judgements, we evaluated publication bias using Egger’s linear regression test [19], which measures funnel plot asymmetry using a natural logarithmic scale of ORs or SMD.

Results

Studies included in the meta-analysis

Electronic and manual searches identified 1094 studies; of these, 12 were selected for full-text review based on the title and abstract details. However, four were excluded because they contained duplicate data or did not contain outcome data (Fig. 1). Thus, eight studies (five RCTs and three prospective cohort studies) including 408 participants (289 for tacrolimus vs. MMF and 119 for low-dose tacrolimus) met the inclusion criteria ([20,21,22,23,24,25,26,27]; Tables 1 and 2). Four studies addressed tacrolimus vs. MMF for LN induction therapy [22, 24, 26, 27], and four studies addressed a low fixed dose of tacrolimus therapy ([20, 21, 23, 25]; Table 3). Oral daily doses of tacrolimus and MMF were taken. Tacrolimus 0.05–0.1 mg/kg and day was titrated to maintain a 12-hour blood concentration of 5–15 ng/ml in trials of tacrolimus vs. MMF. The Jadad scores ranged from 2 to 3, and the quality assessment scores of the prospective cohort ranged between 5 and 6. The relevant features of studies included in the systematic review and meta-analysis are provided in Tables 1, 2 and 3 (Supplementary data).

Table 1 Characteristics of individual studies included in the meta-analysis of induction therapies for lupus nephritis
Full size table
Table 2 Characteristics of the studies included in this systematic review of low-dose tacrolimus therapy in lupus nephritis
Full size table
Table 3 The studies included in this systematic review of tacrolimus at 3 mg/day therapy in lupus nephritis
Full size table
Fig. 1
figure 1

Flowchart of the article selection procedure for research. Search strategy identified 1094 studies and finally found eight that met the inclusion criteria

Full size image

Meta-analysis of the efficacy of tacrolimus vs. MMF in RCTs

The complete remission rate was comparable between tacrolimus and MMF (OR 1.028; 95% CI 0.589–1.796; p = 0.922; Table 4; Fig. 2). The partial remission rate did not differ between tacrolimus and MMF (OR 1.400; 95% CI 0.741–2.646; p = 0.300) (Table 4; Fig. 2). Proteinuria, serum albumin, serum creatinine, creatinine clearance, renal systemic lupus erythematosus disease activity index (SLEDAI), and extrarenal SLEDAI did not differ between tacrolimus and MMF (Table 4). Creatinine clearance was comparable between tacrolimus and MMF groups (tacrolimus vs. MMF: 79.7 ± 32 vs. 71.4 ± 31 mL/min, and 87.8 ± 18.7 vs. 75.6 ± 17.9 mL/min, respectively) [24, 26].

Table 4 Meta-analysis of randomized controlled trials of tacrolimus versus MMF in lupus nephritis
Full size table
Fig. 2
figure 2

Meta-analysis of the efficacy of tacrolimus and mycophenolate mofetil (MMF) in patients with lupus nephritis. a Complete remission, b Partial remissions. The complete and partial remission rates were comparable between tacrolimus and MMF

Full size image

Meta-analysis of the safety of tacrolimus vs. MMF in RCTs

The incidence of infection, severe infection, leukopenia, hyperglycemia, and WAE did not differ between tacrolimus and MMF (Table 4). However, herpes zoster infection was significantly less common in the tacrolimus group than in the MMF group (OR 0.137; 95% CI 0.034–0.546; p = 0.005), while elevation in serum creatinine was considerably higher in the tacrolimus group than in the MMF group (13/103 vs. 0/103; OR 8.148; 95% CI 1.369–48.50; p = 0.021; Table 4; Fig. 3).

Fig. 3
figure 3

Meta-analysis of safety outcomes for tacrolimus versus mycophenolate mofetil (MMF) in patients with lupus nephritis. a Herpes zoster infection, b elevation in serum creatinine. Herpes zoster infection was significantly less common in the tacrolimus group than in the MMF group, while elevation in serum creatinine was higher in the tacrolimus group than in the MMF group (13/103 vs. 0/103)

Full size image

Efficacy and safety of low-dose tacrolimus in LN

Four studies on tacrolimus at low and fixed doses were conducted for the treatment of LN. One RCT included patients receiving tacrolimus (3 mg/day) or placebo therapy for LN [20]. The primary endpoint was the change in LN Disease Activity Index (LNDAI), calculated from the scores for daily urinary protein excretion, urinary red cells, serum creatinine, anti-double-stranded DNA antibody, and serum complement. The LNDAI was decreased by 32.9 ± 31.0% (mean ± SD) in the tacrolimus group (n = 28) and was increased by 2.3 ± 38.2% in the placebo group (n = 35) at 6 months. Significant improvement was observed in the tacrolimus group. Treatment-related adverse events occurred in 92.9% of the tacrolimus group and 80.0% of the placebo group, but the difference was not statistically significant. In patients receiving glucocorticoid therapy for LN, the addition of 3 mg tacrolimus to basal therapy achieved significant improvement compared to placebo. Fei et al. [25] conducted a prospective cohort study to assess the efficacy and safety of low-dose tacrolimus therapy in patients with refractory LN resistant to CYC. A total of 26 patients with LN accompanying persistent proteinuria who were resistant to CYC treatment (> 8 g in less than 6 months) were enrolled. Tacrolimus was initiated at 2 mg/day (if the patient’s weight was < 60 kg) or 3 mg/day (if the patient’s weight was ≥ 60 kg). Mean urinary protein significantly decreased from 6.91 ± 4.50 g at baseline to 1.11 ± 1.10 g at 6 months (p < 0.001). Mean SLEDAI score decreased from 11.42 ± 6.74 at baseline to 3.61 ± 2.73 at 6 months (p < 0.001). A complete or partial response was observed in 88.46% of the patients receiving tacrolimus therapy at 6 months. Tacrolimus was well tolerated at 2–3 mg/day, although one patient developed a severe lung infection. A tacrolimus dosage of 2–3 mg daily appears to be effective and safe. The study by Tanaka et al. [23] was an open-label, prospective, long-term cohort study on tacrolimus once daily at a dose of 3 mg as induction- or reinduction/maintenance treatment in 19 patients with biopsy-proven LN. The median follow-up duration was 42 months. A complete response was achieved in 12 patients (63%) and a partial response was achieved in five patients (26%). The remaining two patients showed no response. Serious adverse effects were not observed. This long-term, low-dose, tacrolimus-based immunosuppressive treatment is beneficial and has low cytotoxicity, suggesting it is an attractive option for the treatment of young patients with LN in daily clinical practice. The study by Tanaka et al. [21] was a prospective cohort study on a once-daily dose of tacrolimus (3 mg/day) in young patients with pediatric-onset, long-standing LN. The U‑protein/U-creatinine ratio gradually decreased after treatment commencement and dropped significantly 24 months after the start of treatment. Complete responses were achieved in eight patients (73%) and partial responses in two patients (18%), but the remaining patients showed no response. Serious adverse effects were not observed.

Heterogeneity and publication bias

Between-study heterogeneity was not found during the meta-analysis of the efficacy and safety of tacrolimus versus MMF, except for serum albumin and creatinine clearance. It was difficult to correlate the funnel plot, which is typically used to detect publication bias, because the number of studies included in the analysis was too small. However, no publication bias was observed (Egger’s regression test, p > 0.1).

Discussion

We systematically reviewed the clinical data from four RCTs that examined the use of low-dose tacrolimus for treatment of LN and merged the clinical data from four RCTs on tacrolimus vs. MMF as induction therapy. Tacrolimus was shown to be as effective and safe as MMF as an induction therapy for LN, with the exception of a reduced risk of herpes zoster infection and a rise in serum creatinine. Tacrolimus at a dose of 3 mg/day was found to be efficacious and safe in patients with LN.

SLE is a heterogeneous autoimmune disorder characterized by autoantibody overproduction and T and B cell abnormalities that contribute to immune complex accumulation in the kidneys. The formation of an immunological complex triggers an inflammatory reaction in glomeruli, resulting in lymphocyte and macrophage infiltration [28]. It was shown that tacrolimus and MMF had comparable efficacies in terms of inducing renal remission. Long-term outcome of an RCT confirmed non-inferiority of tacrolimus to MMF as induction therapy for LN [29]. However, tacrolimus reduces the risk of herpes zoster infection and increases the risk of serum creatinine elevation. Cyclosporin, another calcineurin inhibitor, is an effective and safe treatment for patients with LN [30]. Multitarget therapy such as tacrolimus + MMF showed a higher complete remission rate than monotherapy [31]. However, cases of infection and pneumonia were numerically elevated in the multitarget therapy group compared to the monotherapy group [32].

This meta-analysis differs from a previous meta-analysis by Hannah et al. [33], as the current study included one new study and 42 more patients with LN in the tacrolimus group and 41 more patients in the MMF group. Our meta-analysis is more comprehensive in terms of efficacy and safety. The conclusion of this meta-analysis that tacrolimus is comparable to MMF in terms of effectiveness agrees with previous research; however, our investigation revealed a difference in safety between tacrolimus and MMF.

Due to the limitations of this study, our findings should be considered with care. First, the possibility of publication bias is always a concern. It should be emphasized that it is difficult to rule out publication bias with certainty, particularly when the number of studies considered is low, as in this analysis. Second, variances in clinical characteristics such as race, sex, age, extent of renal impairment, proportion of patients with class III and IV LN, and research quality are likely to skew the meta-analysis findings. Third, because tacrolimus was only studied in Asian patients, further studies are required to determine whether tacrolimus therapy is successful in non-Asian patients with LN. Fourth, the number of studies included and the sample sizes in these studies were small. Especially, the numbers of study participants for the side effects were substantially small (83 vs. 83 for herpes zoster infection and 103 vs. 103 for elevation in serum creatinine). Fifth, the three prospective cohort trials used uncontrolled designs, resulting in a lack of clear evidence for the specific effects of low-dose tacrolimus. However, this meta-analysis and systematic review had some benefits. The number of patients with LN in different studies varied from 11 to 150; nonetheless, 408 individuals were included in this pooled analysis. We generated more accurate data by merging the findings of multiple investigations and performing a systematic review rather than conducting independent research [34,35,36]. This improved statistical power and resolution, enabling collection of more accurate data.

In conclusion, our RCT-based meta-analysis indicated that tacrolimus was comparable to MMF in terms of effectiveness and safety as an induction therapy for LN, with the exception of a lower risk for herpes zoster infection and a greater rate of serum creatinine increase due to tacrolimus. In this systematic review, tacrolimus was shown to be efficacious and safe at a dose of 3 mg/day in patients with LN. Further research is required to assess the long-term effectiveness and safety of tacrolimus therapy in individuals with LN from various ethnic groups.