Article Text


Extended report
Tacrolimus versus mycophenolate mofetil for induction therapy of lupus nephritis: a randomised controlled trial and long-term follow-up
  1. Chi Chiu Mok1,
  2. King Yee Ying2,
  3. Cheuk Wan Yim3,
  4. Yui Pong Siu1,
  5. Ka Hang Tong1,
  6. Chi Hung To1,
  7. Woon Leung Ng3
  1. 1Department of Medicine, Tuen Mun Hospital, Hong Kong
  2. 2Department of Medicine, Princess Margaret Hospital, Hong Kong
  3. 3Department of Medicine, United Christian Hospital, Hong Kong
  1. Correspondence to Dr Chi Chiu Mok, Department of Medicine, Tuen Mun Hospital, Hong Kong; ccmok2005{at}


Objective To compare the efficacy of tacrolimus (TAC) and mycophenolate mofetil (MMF) for the initial therapy of lupus nephritis (LN).

Study design This is an open randomised controlled parallel group study.

Methods Adult patients with biopsy-confirmed active LN (class III/IV/V) were randomised to receive prednisolone (0.6 mg/kg/day for 6 weeks and tapered) in combination with either TAC (0.06–0.1 mg/kg/day) or MMF (2–3 g/day) for 6 months. Good responders were shifted to azathioprine for maintenance. The primary outcome was the rate of complete renal response (CR) at 6 months and the secondary outcomes included partial renal response, renal flares and decline of renal function over time.

Results 150 patients (92% women; aged 35.5±12.8 years; 81% class III/IV) were randomised (76 MMF, 74 TAC). At month 6, the rate of CR was 59% in the MMF and 62% in the TAC group (treatment difference: 3.0% (−12%, 18%); p=0.71). Major infective episodes occurred in 9.2% patients treated with MMF and in 5.4% patients treated with TAC (p=0.53). Maintenance therapy with azathioprine was given to 79% patients. After 60.8±26 months, proteinuric and nephritic renal flares developed in 24% and 18% of patients in the MMF group and 35% (p=0.12) and 27% (p=0.21) in the TAC group, respectively. The cumulative incidence of a composite outcome of decline of creatinine clearance by ≥30%, development of chronic kidney disease stage 4/5 or death was 21% in the MMF and 22% in the TAC group of patients (p=0.35).

Conclusions TAC is non-inferior to MMF, when combined with prednisolone, for induction therapy of active LN. With azathioprine maintenance for 5 years, a non-significant trend of higher incidence of renal flares and renal function decline is observed with the TAC regimen.

Trial registration number Hospital Authority Research Ethics Committee Clinical Trial Registry (HARECCTR0500018; Hong Kong) and US (NCT00371319).

Statistics from


Renal disease in systemic lupus erythematosus (SLE) carries significant morbidity and mortality.1–5 Up to 26% of patients with diffuse proliferative lupus nephritis (LN) develop end-stage renal failure6–9 and the mortality increases by eightfold as compared with the general population.5 Thus, intensive immunosuppression (induction) has to be given to dampen kidney inflammation timely, followed by maintenance therapy to consolidate the efficacy and reduce the risk of renal flares.

The conventional induction therapy for severe LN is cyclophosphamide (CYC) in combination with corticosteroids.10–12 However, toxicities such as infection, marrow suppression, cystitis and ovarian failure limit its use.11 ,13 ,14 Mycophenolate mofetil (MMF) has emerged to be the first-line therapy because randomised controlled trials (RCTs) have shown that it is as effective as CYC,15 ,16 but ovarian toxicities and leucopenia are less common. Despite this, both CYC and MMF are not ideal in terms of efficacy and alternative regimens have to be explored.

The calcineurin inhibitors such as ciclosporin A (CSA) and tacrolimus (TAC) have long been used in organ transplantation. Recent RCTs and meta-analysis revealed that TAC was more favourable than CSA in renal transplantation in terms of acute graft rejection, treatment failures and long-term graft survival.17–19 Moreover, the incidence of hypertension, hyperlipidaemia, gingival hyperplasia and hirsutism was lower with TAC.20

TAC has been used successfully in LN, including the pure membranous subtype and refractory disease, in uncontrolled series.21–24 Small RCTs showed similar efficacy of TAC to CYC for the initial treatment of proliferative LN.25 ,26 Low-dose combination of TAC and MMF has also been shown to be more effective than intravenous pulse CYC in mixed proliferative and membranous LN (MLN).27 More recently, this regimen has been used successfully in refractory LN.28 The mechanisms of action of TAC in LN have not been fully elucidated. In vitro and animal studies have shown that TAC reduces production of interleukin 2 (IL-2) and other cytokines such as tumour necrosis factor α, interferon γ, IL-6 and IL-10 by activated T cells.29 By inhibiting T cell activation, signals for B cell activation, class-switching and immunoglobulin production are indirectly attenuated.30 The calcineurin inhibitors were also suggested to stabilise the actin cytoskeleton in kidney podocytes, leading to amelioration of proteinuria.31

We conducted the current RCT to compare the efficacy between TAC and MMF for induction therapy of active LN and evaluate longitudinally the rate of renal flares and change in renal function.

Patients and methods

Inclusion criteria of patients for this study were (1) age ≥18 years, (2) fulfilment of four or more American College of Rheumatology (ACR) criteria for SLE,32 (3) biopsy-proven active LN (International Society of Nephrology (ISN) / Renal Pathology Society (RPS) class III/IV/V) within 4 weeks of entry and (4) serum creatinine (SCr) <200 µmol/L. Exclusion criteria were (1) refusal to be randomised; (2) preference for treatment with conventional regimens such as CYC and (3) planning for pregnancy within 12 months after randomisation.

Treatment protocol and randomisation

All patients were given oral prednisone (0.6 mg/kg/day for 6 weeks, then tapered by 5 mg/day every week to <10 mg/day), which was continued indefinitely as maintenance therapy. Participants were randomised by computer-generated blocks of four in a 1 : 1 ratio to one of the following treatment arms: (1) MMF (2 g/day initially, augmented to up to 3 g/day if clinical response was suboptimal at month 3), in two divided doses for 6 months or (2) TAC for 6 months (initial dosage 0.1 mg/kg/day in two divided doses, reduced to 0.06 mg/kg/day if clinical response was satisfactory at month 3). A central research assistant was responsible for the allocation of the treatment regimens according to the block sequence. New ACE inhibitors or angiotensin receptor blockers were prohibited in the first 6 months, and in those receiving these drugs at entry, their doses were kept constant.

Dosage adjustment of study drugs

Dosage of MMF or TAC was adjusted according to patients’ tolerability and adverse events (AEs; eg, leucopenia for MMF, nephrotoxicity for TAC and intercurrent infections). TAC would be withheld when there was persistent increase in SCr level by >40% compared with baseline or >30% compared with the last visit for consecutive 2 weeks. In case of SCr improvement after temporary suspension, TAC would be resumed at a lower dose and titrated up according to renal function. If the SCr did not improve by 4 weeks, the patient would be withdrawn. Trough serum TAC level was assayed to achieve a target of >5 ng/mL.

Outcomes of interest

The primary outcome of this study was the proportion of patients who achieved complete renal response (CR) at 6 months. Secondary outcomes included (1) rates of partial response (PR) and non-response (NR) at 6 months, (2) rate of renal flares during the maintenance phase and (3) a composite outcome of renal function deterioration (decline of creatinine clearance (CrCl) by ≥30% or progression to chronic kidney disease (CKD) stage 4 or 5 (ie, CrCl <30 mL/min) or mortality over time (see online supplementary table S1 for the definitions for renal response and renal flares). We also assessed the rates of CR at month 6 according to the ACR SLE renal response criteria.33

Maintenance and salvage treatment

Patients who achieved CR or good PR after MMF or TAC treatment at 6 months were shifted to azathioprine (AZA; 2 mg/kg/day) indefinitely for maintenance. For suboptimal responders to either MMF or TAC, salvage/reinduction therapy would be given in the form of oral CYC (2 mg/kg/day) or monthly intravenous pulse CYC (0.5–1 g/m2) for 6 months (unless intolerant to or reluctant for CYC), with high-dose prednisolone (for 6 weeks and tapered) with or without intravenous pulse methylprednisolone. Patients who had renal flares would be retreated as decided by the attending physicians.

Patients were followed up every 2 months in the first 6 months and then 3 monthly. Blood counts, SCr and albumin, CrCl and urine protein-to-creatinine ratio (uP/Cr) estimation (according to a 24 h urine sample) and lupus serology (anti-dsDNA and complements) were performed in each visit. Urine sediments were measured at baseline, month 6 and when renal flares were suspected.

Disease activity of SLE was assessed by the Safety of Estrogens in Lupus Erythematosus National Assessment (SELENA)-SLE Disease Activity Index (SLEDAI), a validated global index used in the prospective SELENA trials.34

Sample size calculation and statistical analyses

This study was designed as a non-inferiority trial. The non-inferiority margin was set at 12.5% for the primary outcome (CR at 6 months), meaning that the lower bound of the two-sided 95% CI for the difference CR rates between TAC and MMF (as reference) should exceed −12.5%. Analysis for the primary outcome was performed on the intent-to-treat population. A previous study of MMF in Asian patients with LN reported a CR rate of 58%.35 Our pilot study of TAC showed that 67% of patients with LN could achieve CR after 6 months.23 Assuming that the CR rates of MMF and TAC at 6 months would differ by 10%, a sample size of 140 patients was needed to yield a power of 80% to establish non-inferiority of TAC to MMF, with a one-sided α level of 0.025.

Values in this study are expressed as mean±SD. Comparison between MMF and TAC groups at baseline was performed by the independent Student t test for continuous variables and χ2 test for categorical variables. Between-group differences at different time points were compared with adjustment of baseline values by one-way analysis of covariance. Within-group (between various time points and baseline) comparison was made by the paired Student t test. Correlation between two variables was studied by Spearman's rank correlation.

The cumulative incidence of renal flare and renal function deterioration/mortality over time was evaluated by Kaplan–Meier's analysis. Difference between the MMF and TAC group was compared by the log-rank test. For deceased patients or those who were lost to follow-up, data were censored at their last visits.


Study population

Between 2005 and 2012, 150 Chinese patients with active LN were recruited and randomised (76 to MMF and 74 to TAC; 92% women; mean age 35.5±12.8 years). All except one patient completed the 6-month induction phase (see online supplementary figure S1 for patient enrolment and disposition).

At baseline, 100 (67%) patients had CrCl <90 mL/min and 64 (43%) patients had nephrotic syndrome. The distribution of the histological classes of LN was IVG±V (33%), IVS±V (13%), III±V (36%) and pure V (19%). The clinical and renal parameters of the patients at entry were similar between the two groups (table 1).

Table 1

Clinical characteristics and renal parameters at baseline

In patients who were randomised to MMF, 58 (76%) received a dosage of 2 g/day, 7 (9%) received 2.5 g/day and 11 (14%) received 3 g/day. In patients treated with TAC, the mean trough drug level achieved was 7.8±3.9 ng/mL (median 7.0).

Renal response and change in SLE disease activity at 6 months

Significant improvement in uP/Cr, serum albumin, lupus serology (C3, anti-dsDNA titre), urinary sediments, renal and extra-renal SELENA-SLEDAI scores and lipid profile were observed in both groups of patients (p<0.001 in all), with no significant differences between MMF and TAC at month 6 (table 2). Although no significant change in CrCl was observed in patients treated with TAC (p=0.78), MMF-treated patients had significant improvement of CrCl (p<0.001).

Table 2

Change in clinical parameters in the first 6 months

Table 3 shows the rates of renal response at month 6. CR occurred in 59% of patients in the MMF arm and 62% patients in the TAC arm (difference 3% (95% CI −12% to 18%)). Using the ACR renal response criteria, 11% of MMF-treated patients and 14% of TAC-treated patients achieved CR, respectively (difference 3% (−8% to 14%); p=0.59).

Table 3

Overall renal response at month 6

In the TAC group, the mean serum TAC level was 7.86±4.3, 7.60±3.2 and 8.53±3.9 ng/mL, respectively, in those who achieved CR, PR and NR. No significant correlation was observed between the TAC level and the improvement in uP/Cr at 6 months compared with baseline (ρ=−0.14, p=0.25).

Subgroup data on pure MLN

Table 4 shows the change in uP/Cr and CrCl in the first 6 months in patients with pure MLN. The improvement in uP/Cr was more profound in the TAC than in the MMF group. Numerically, more patients treated with TAC achieved CR or PR than MMF (100% vs 75%; p=0.09). A quarter of patients were refractory (NR) to MMF but none of the patients treated with TAC had NR.

Table 4

Renal response of pure membranous lupus nephritis

Adverse events in the first 6 months

Table 5 summarises the AEs experienced by our patients during the first 6 months. Major infective episodes developed in 7 (9.2%) and 4 (5.4%) patients treated with MMF and TAC, respectively (p=0.53). There was one death in the MMF group (uncontrolled sepsis). Herpes zoster infection was significantly more common in patients treated with MMF than in patients treated with TAC (18% vs 3%; p=0.003). Diarrhoea was more commonly reported in MMF-treated patients, whereas more alopecia, diabetes mellitus, leg cramps and neurological symptoms, such as tremor, were reported in TAC-treated patients. Neurological symptoms resolved completely on reduction of TAC dosage. Reversible increase in SCr by 30% was exclusively observed in TAC-treated patients. No patients were withdrawn because of AEs except for the one who died.

Table 5

Adverse events in the first 6 months

Relapse of LN and renal function decline

AZA maintenance was given to 59 (78%) MMF-treated (dose 82.5±24 mg/day) and 60 (81%) TAC-treated patients (dose 86.5±21 mg/day; p=0.32). Other patients were reinduced with alternative immunosuppressive regimens (high-dose prednisolone with CYC (N=20), low-dose combination of MMF and TAC (N=5), cross-over to TAC (N=4) or cross-over to MMF (N=2)). The need for salvage treatment for patients with suboptimal response to induction therapy was similar in both groups.

After a follow-up of 60.8±26 months, proteinuric and nephritic renal flares occurred in 24% and 18% of patients treated initially with MMF and 35% and 27% in those treated with TAC, respectively (see online supplementary table S2). The cumulative risk of having a renal flare of patients treated with MMF followed by AZA was 8% at 1 year, 28% at 3 years and 38% at 5 years, whereas the corresponding figures for patients treated with TAC followed by AZA was 9% in 1 year, 33% in 3 years and 54% in 5 years (p=0.13; figure 1A). For those who achieved CR after induction therapy, the mean time to first renal flare was 28.9±17 months in the MMF group and 30.1±18 months in the TAC group (difference 1.3 (−13.7 to 11.2) months).

Figure 1

(A) Kaplan–Meier plot of the cumulative probabilities of having a renal flare according to treatment arm. (B) Kaplan–Meier plot of the cumulative probabilities of having a composite outcome of decline of creatinine clearance by ≥30%, development of chronic kidney disease stage 4/5 or death according to treatment arm. MMF, mycophenolate mofetil; TAC, tacrolimus.

Six patients died in the MMF group (uncontrolled sepsis in two, suicide in one, cancer in one and sudden cardiac death in two) and five patients died in the TAC group (uncontrolled sepsis in three, haemorrhagic stroke in one and pulmonary haemorrhage in one). The cumulative incidence of a composite outcome of decline of CrCl by ≥30%, development of CKD stage 4/5 or death at 5 years was 21% in patients treated with MMF and 22% in those treated with TAC (p=0.35; figure 1B).


This open randomised study showed that TAC was non-inferior to MMF, in conjunction with high-dose corticosteroids, for induction therapy of active LN. As a lower dose of prednisolone and shorter duration of therapy were used when compared with previous studies,15 ,16 ,25–27 the efficacy observed at 6 months was likely a combined effect of prednisolone with the study drugs. With AZA maintenance for 5 years, there was a trend of higher incidence of renal flares and renal function decline in the TAC than in the MMF group of patients.

Our results are in line with two small RCTs, which showed that the short-term efficacy of TAC was similar to pulse CYC in LN.25 ,26 In fact, lower doses of TAC have also been shown to improve residual proteinuria after induction treatment in LN.36 ,37

Pure MLN is well recognised to have a more delayed response to treatment than its proliferative counterpart. Pooling of data from two RCTs showed that MMF was equally effective with CYC for pure MLN.38 Open-label studies have also shown that CSA and TAC were effective for pure MLN.22 ,39 A multitarget regimen consisting of corticosteroid, MMF and TAC was shown to be more effective than CYC in mixed proliferative and MLN.27 The rationale of this combination is the efficacy of TAC on the membranous component of LN. In our study, TAC appeared to be more effective than MMF in reducing proteinuria in pure MLN. This was unlikely because of the drop in CrCl as we had assessed the uP/Cr ratio instead of the absolute amount of proteinuria. However, the sample size of this subgroup was not powered to detect a difference between MMF and TAC.

The relationship between serum TAC level and its efficacy is intriguing. Our previous study23 failed to show a correlation between clinical response and TAC level. In the present study, which involved much more patients, again we could not demonstrate better efficacy with higher TAC levels. In fact, NR patients had achieved higher TAC levels than those with CR or PR. Despite this, it is prudent to monitor the TAC level for dosage titration to minimise toxicity and for surveillance of drug compliance.

In our study, 12% of TAC-treated patients had transient increase in SCr and no improvement of CrCl at month 6, whereas in MMF-treated patients, an improvement in CrCl was observed. This could possibly be related to a relatively high dosage of TAC being used. In fact, the mean trough TAC level achieved in our patients (7.8 ng/mL) was higher than those reported in renal transplantation studies (<7.5 ng/mL) at 1 year.17 ,19 Although the proportion of patients who had deterioration in CrCl by 30% at 5 years was not significantly higher in the TAC than in the MMF group of patients, the difference might become significant on longer follow-up. As the exposure to TAC was only 6 months, whether the unfavourable trend on long-term renal function in this group was related to the numerically higher number of renal flares remains to be determined. Nevertheless, the hazard of nephrotoxicity, especially in those with impaired renal function at baseline, should be taken into account when TAC is used for LN. Close monitoring of renal function and the TAC level is necessary.

Patients in our study were maintained on AZA after successful induction therapy. We have previously shown that AZA maintenance in LN was associated with a better outcome in terms of renal function deterioration and mortality.7 However, whether MMF is more cost-effective in preventing renal flares than AZA as maintenance therapy remains unclear as conflicting evidence has been reported in different studies.40 ,41

The use of MMF during pregnancy is linked with an increased incidence of first-trimester pregnancy loss and fetal malformations that involve the ear and face such as cleft palate and lip.42 Therefore, MMF is contraindicated during pregnancy. On the contrary, TAC is not linked to congenital malformations in solid organ transplantation43 and has been successfully used to treat LN flares during pregnancy.44

The strength of our study is its randomised design and a sufficiently large sample to confirm non-inferiority of TAC to MMF for LN. We followed up our patients for 5 years and reported the long-term outcome in terms of renal flares and renal function decline. The major limitation of our study is the open-label design and the shifting to AZA after 6 months. As some patients may have delayed response to therapy, extension of induction-consolidation treatment to 12 months for the assessment of efficacy might be a better option. The definition of CR with a proteinuria of <1 g/day when the study was designed a decade ago was a bit loose when compared with recent consensus recommendation (proteinuria <0.5 g/day with normal or near normal renal function).45 In addition, the lack of monitoring of the trough mycophenolic acid level for dosage titration of MMF might have undermined its efficacy when compared with TAC. Finally, our results cannot be extrapolated to other ethnic populations. Despite these caveats, this is the largest study to show non-inferiority of TAC to MMF for induction therapy of active LN. TAC/corticosteroid combination should be considered as an alternative regimen to conventional CYC or MMF for the initial therapy of active LN, especially for those who are intolerant, contraindicated or refractory to the latter agents.


View Abstract
  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

    Files in this Data Supplement:


  • Handling editor Tore K Kvien

  • Declaration We did not receive any grants, sponsor or support from any organisation including pharmaceutical companies, in particular Roche and Astella, for the submitted work. All authors do not have any conflicts of interests to be declared. None of the authors have financial relationships with any organisations that might have an interest in the submitted work in the previous 3 years. There are no other relationships or activities that could appear to have influenced the submitted work.

    The lead author affirms that the manuscript is an honest, accurate and transparent account of the study being reported; that no important aspects of the study have been omitted and that any discrepancies from the study as planned have been explained.

  • Contributors CCM: study design, patients’ assessment, data collection and analysis (guarantor of the submitted work). KYY, KHT, CWY, YPS, CHT, WLN: patients’ assessment and follow-up, data collection. The investigators have full access to all of the data and take responsibility for the integrity of the data and the accuracy of the data analysis.

  • Competing interests None.

  • Patient consent Obtained.

  • Ethics approval Research and Ethics Committee of Tuen Mun Hospital, Princess Margaret Hospital and United Christian Hospital.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Data sharing statement We will present additional unpublished data from the study if available and deemed necessary by the editorial office.

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.