Lupus nephritis is a common and severe manifestation of systemic lupus erythematosus (SLE) that can lead to ESRD and death (1). In the past 20 yr, treatment of lupus nephritis has advanced significantly, and induction therapies that combine cyclophosphamide and corticosteroids have improved renal outcomes compared with treatment with steroids alone (2,3). These benefits, however, continue to be tempered by significant drug-related toxicity (4–7).
In an attempt to limit therapy-related toxicity, mycophenolate mofetil (MMF), an antirejection drug that is used in solid-organ transplantation, has been studied. Evidence from early observational studies suggested that MMF may be efficacious in inducing remission of lupus nephritis (8–11). More recently, randomized, controlled trials (RCT) comparing MMF and cyclophosphamide as induction agents in lupus nephritis have also been completed (12–15) with similar results. Conclusions based on these trials, however, are limited given their small sample sizes. In addition, none of the studies was powered to detect a difference in the more clinically relevant long-term outcomes of death or ESRD. Given the increasing popularity of MMF in the treatment of SLE, we performed a systematic review and meta-analysis of RCT to assess the relative risk for failure to induce remission and the composite end point of death or ESRD for MMF compared with cyclophosphamide in lupus nephritis.
Materials and Methods
We searched PubMed (1996 through March 2006), MEDLINE (1966 through March 2006), EMBASE (1980 through March 2006), and CENTRAL (Cochrane Controlled Clinical Trials Register 1996 through March 2006) using the OVID search engine. Two comprehensive search themes were developed and combined using the Boolean operator “and.” The first theme, lupus nephritis, was created by using the Boolean search term “or” to search for the following terms, text words, and database-specific thesaurus terms: “Lupus nephritis” or “glomerulonephritis” or “nephritis” or “systemic lupus erythematosus” or “lupus” or “SLE.” The second theme, MMF, was created using a text word and database-specific thesaurus terms search for “mycophenolate mofetil” or “mmf” or “mycophenolic acid” or “cellcept.” Citations identified from electronic databases were supplemented by reviewing the reference lists of all identified research and review articles, by reviewing the conference proceedings of the American Society of Nephrology from 2001 to 2005, and by contacting experts in the field.
Study Selection and Data Abstraction
Two individuals (M.J. and M.W.) independently evaluated articles for eligibility in a two-stage procedure. In the first stage, all identified abstracts were reviewed. Those that met the inclusion criteria or those for which there was uncertainty as to eligibility were selected for full-text review in the second stage. Articles that were selected by either individual were reviewed by both M.J. and M.W. in the second stage and evaluated on both inclusion and exclusion criteria. Inclusion criteria consisted of (1) study design (controlled trial of three or more patients), (2) study population (patients with biopsy-proven lupus nephritis), (3) intervention (MMF), and (4) and outcome (trials with explicit definition and reporting of renal remission or change in proteinuria or serum creatinine or GFR). Studies were eligible for inclusion whether published or unpublished and irrespective of language. Exclusion criteria were use of MMF for maintenance therapy but not induction therapy or studies that included children.
The same reviewers independently extracted data from all primary studies that fulfilled the inclusion criteria, with disagreement resolved by consensus. Data extracted included study design; details of treatment protocol; and baseline demographic, clinical, laboratory, and renal biopsy information. Data on the following outcomes were extracted when reported: Proteinuria; serum creatinine, GFR, complete renal remission, partial renal remission, treatment failure, mortality, progression to ESRD, and treatment-related toxicities. When outcomes were not reported, we attempted to contact authors for additional information.
The methodologic quality of the included studies was independently scored by M.J. and M.W., with disagreements resolved by consensus. Criteria that were used to assess quality included method of randomization; allocation concealment, blinding, use of placebo, reporting of losses to follow-up, intention-to-treat analysis, and important baseline differences between groups. An overall quality score was determined for each study as described by Jadad et al. (16).
The primary outcome was the relative risk (RR) for failure to induce remission (total or partial), as defined in the original articles on the basis of the degree of proteinuria, renal function compared with baseline, and urinalysis. The secondary outcome was the RR for the composite outcome of death or ESRD. We also tabulated the frequency of adverse events by treatment group.
Quantitative Data Synthesis
We used the fixed-effects model of Mantel-Haenszel (17) to estimate the pooled RR with 95% confidence intervals (CI) for study outcomes, using data from all eligible RCT. Given the small number of RCT, we performed a sensitivity analysis that included all controlled studies irrespective of treatment allocation. For studies with no reported events in one group, statistical methods of approximation were used to calculate RR (17). Two studies reported on the outcomes of death and ESRD at multiple time points (13,15,18). Analyses were undertaken using data from both the original and the extended study duration for the purposes of assessing the secondary outcomes. The presence of heterogeneity across trials was evaluated using a Q statistic for heterogeneity. We tested for publication bias using funnel plots, Begg's test for asymmetry and an Egger's test. All statistical analyses were performed using Stata version 9.0 (Stata Corp., College Station, TX).
Role of the Funding Source
No external funding was obtained. All authors had full access to all study data, and all took full responsibility for the decision to submit for publication.
Identification of Studies
Progress through stages of the systematic review is summarized in Figure 1. A total of 1158 unique citations were identified by our search strategy, 1117 of which were excluded on initial screening. Of the 41 studies that underwent further review, 4 RCT (13,14,18,19) and two nonrandomized trials (12,20) were identified for systematic review. Five of these studies were published in peer-reviewed journals, and one was published in abstract form (19). There was agreement between reviewers on the selection of 39 of the 41 studies selected for review (κ = 1.0) and complete agreement for the studies that were selected for meta-analysis and their validity assessment.
A total of 268 patients were included in the four RCT, 133 of whom were treated with MMF and 135 of whom were treated with cyclophosphamide (Table 1). The mean age of patients ranged from 28 to 42 yr, with most patients of female gender (range 79 to 94%). The most common ethnic group was Asian (42%), followed by black (30%). Baseline mean (SD) serum creatinine measurements ranged from 93.3 (46.0) to 112.7 (65.8) μmol/L in the MMF group and from 94.0 (40.2) to 113.1 (36.9) μmol/L in the cyclophosphamide group; baseline mean (SD) proteinuria ranged from 1.8 (1.2) to 6.2 (4.1) g/24 h in the MMF group and from 3.0 (1.8) to 4.4 (3.6) g/24 h in the cyclophosphamide group. A total of 238 patients had proliferative lupus nephritis (World Health Organization [WHO] class III/IV), and 30 patients showed pure membranous lupus nephritis (WHO class V). All trials used adjunctive therapy with corticosteroids. One RCT used oral cyclophosphamide as a comparator (18), and three RCT compared MMF with intravenous cyclophosphamide (13,14,19). Dosing of MMF varied across studies, ranging from 1 to 3 g/d.
Definitions of failure to induce remission differed only slightly between trials and in all four studies were based on absolute levels and changes in proteinuria, serum creatinine, and activity of urine sediment (Table 2).
Characteristics of the nonrandomized trials included in the sensitivity analysis are shown in Table 3. The mean baseline proteinuria and serum creatinine measurements were of similar values to the RCT. Notably, in the study by Hu et al. (12), patients were treated with MMF only when cyclophosphamide failed.
Trial quality was limited, as demonstrated by the Jadad scores reported in Table 1. No trials were blinded, and only one trial used allocation concealment (13). Only one trial performed an intention-to-treat analysis (13); however, for two other trials no patients were lost to follow-up, and in the remaining trial, only one patient in each arm was lost to follow-up (18). The only potentially important difference between treatment groups that we identified for any trial was the greater baseline proteinuria among MMF recipients in the study by Ong et al. (14). In studies that reported extended follow-up periods, no mention of loss to follow-up was made.
Quantitative Data Synthesis
Three of the four studies that were included in the meta-analysis showed no significant difference between MMF and cyclophosphamide for the RR for failure to induce remission (14,18,19), whereas one study showed a significant reduction in the RR for failure to induce remission in patients who received MMF (13). The pooled RR for failure to induce remission for MMF compared with cyclophosphamide, using a fixed-effects model, was 0.70 (95% CI 0.54 to 0.90; P = 0.004), suggesting a significant treatment benefit for MMF (Figure 2). The Q statistic did not detect significant heterogeneity between trials (P = 0.76). Because of the lack of heterogeneity, a random-effects model was done only as a sensitivity analysis and did not differ significantly from the fixed-effects model. Inclusion of the non-RCT in a sensitivity analysis did not alter the summary statistic (Figure 3) with a pooled RR for failure to induce remission for MMF compared with cyclophosphamide of 0.68 (95% CI 0.53 to 0.85; P = 0.001). There was no evidence of heterogeneity (P = 0.93).
Analysis of the composite outcome of death or ESRD at the end of the prespecified study duration showed a nonsignificant lower risk for patients who were treated with MMF compared with cyclophosphamide (pooled RR 0.35; 95% CI 0.10 to 1.22; P = 0.10; Figure 4). Analysis of the composite outcome of death or ESRD using the extended study follow-up showed a significant lower risk for patients who were treated with MMF compared with cyclophosphamide, with a pooled RR of 0.44 (95% CI 0.23 to 0.87; P = 0.02; Figure 5). No heterogeneity was detected in either analysis, with P = 0.41 and 0.31, respectively.
Treatment-related adverse events for RCT between 6 and 12 mo duration are summarized in Table 4. In general, too few events were observed in most categories, resulting in several empty cells. For this reason, we chose to subject only the RR for infection to meta-analysis. The RR for infection for patients who were treated with MMF compared with cyclophosphamide was 0.64 (95% CI 0.39 to 1.06; P = 0.085). No patients who were treated with MMF developed amenorrhea compared with six of 106 patients who were treated with cyclophosphamide. The proportions of patients who experienced other adverse effects seemed grossly equal between the MMF- and cyclophosphamide-treated groups. There was no evidence of publication bias by Egger's (P = 0.87) or Begg's test (P = 1.0), although the number of identified trials was small and the utility of such tests to assess the possibility of bias in this situation is extremely limited.
In this meta-analysis of four studies with 268 patients who were treated for lupus nephritis, we found a significant reduction in the risk for treatment failure for MMF compared with cyclophosphamide. In addition, our findings suggest that the risk for death or ESRD may be lower with MMF compared with cyclophosphamide, an important clinical finding not previously demonstrated. There was no evidence of an increased risk for adverse events associated with MMF.
Given concern about infertility and other toxicities with cyclophosphamide treatment, MMF has been considered an important alternative agent for the treatment of lupus nephritis. However, there has been speculation that MMF may not be as effective for inducing remission in lupus nephritis or, in the long term, may lead to an increased likelihood of relapse compared with cyclophosphamide (21). The study by Ginzler et al. (13) in 2005 was the only individual study to demonstrate a significant benefit of MMF compared with cyclophosphamide in inducing remission of lupus nephritis. Our meta-analysis extends the results of individual randomized trials and suggests a 30% reduction in the risk for treatment failure for MMF. Therefore, the benefit of MMF may translate not only to improved remission rates but also to a reduction in cyclophosphamide-associated toxicity for patients with lupus nephritis.
The composite outcome of death or ESRD is more clinically relevant than treatment failure for lupus nephritis (21); however, individual clinical trials to date have not been powered to detect a difference in these clinical outcomes. Results from our meta-analysis suggest an almost 60% reduction in the composite end point of death or ESRD for MMF compared with cyclophosphamide when the results of extended follow-up are considered, although these findings were NS when outcomes were assessed at the end of the prespecified study duration. This finding is particularly impressive given that a recent meta-analysis of 226 patients who were treated with either cyclophosphamide and glucocorticoids or glucocorticoids alone was unable to find a mortality or ESRD benefit for cyclophosphamide-treated patients (2). This finding, however, is based predominantly on only two studies, one of which did not report loss to follow-up during the extended observation and was based on relatively few events, so caution must be taken in accepting the result despite its statistical significance.
Any systematic review must assess the suitability of identified trials for pooling in a meta-analysis. In this study, all trials that were included in the primary analysis were comparable in several respects: Duration of follow-up; definitions of treatment failure; and patient age, gender, and renal function. Despite these similarities, there are also several differences, including the dosages of MMF used and the route of cyclophosphamide administration. These differences may play a role in the significant variability in the absolute rates of treatment failure between studies (range 3 to 48% for MMF groups and 6 to 80% for cyclophosphamide groups). The role of chance, however, may account for a significant amount of this variability given the small size and short duration of each individual trial. This highlights the need for adequately powered RCT and meta-analyses. The trials also differed in racial mix, with two studies including only Asian patients (14,18), one including only Hispanic patients (19), and one including a mixed but predominantly black population (13). The lack of heterogeneity in the RR for treatment failure among these studies, despite their differences in racial mix, is reassuring in that our study results are generalizable across races. However, there are inadequate data to examine directly treatment allocation–race interactions to make any final conclusions on this topic.
Although our conclusions are based on a small number of randomized trials, they are strengthened by the homogeneity of study results and lack of any obvious publication bias. The reduction in death and ESRD may be driven by use of MMF as a maintenance agent as well as an induction agent, thereby representing a difference between MMF and other maintenance immunosuppressive agents. However, that the results favor MMF for both outcomes studied (failure to induce remission and the composite of death or ESRD) strengthens the evidence for the efficacy of MMF as an induction agent.
The results of this meta-analysis should be interpreted in light of the study limitations as well as the limitation of each of the primary studies. First, the generalizability of these study results is limited. Patients who had lupus nephritis and were included in these trials had relatively preserved renal function, and no randomized trials have shown MMF to be equivalent or superior to cyclophosphamide in patients with severe renal impairment. Although additional clinical trials directed at patients with severe renal impairment are warranted, it should be recognized that the majority of patients who present with proliferative lupus nephritis have relatively preserved kidney function. There were also very few patients with WHO class V (membranous) lupus nephritis in these studies; therefore, the results cannot be extrapolated to this population with any confidence. Second, although we did not detect evidence of publication bias, the use of statistical tests to detect such bias with so few available studies in the first place is unhelpful. Finally, the overall low quality of the available studies is a limitation. It is encouraging, however, to find that the results of lower quality studies are comparable with those of the most recent and methodologically sound study. In addition, although these results are statistically significant, all results, particularly the composite of death or ESRD, are based on a relatively small number of events and may therefore be susceptible to random error. Small meta-analyses (i.e., analyses of fewer than several hundred events) should be regarded with caution even in the presence of statistically significant results (22). Given these limitations, another large, high-quality RCT with long-term follow-up is warranted (23).
We have demonstrated that compared with cyclophosphamide, MMF reduces the risk for failure to induce remission of lupus nephritis and may reduce the risk for death or ESRD. Although these data support the use of MMF as the first-line induction agent for lupus nephritis, there is still a need for further high-quality studies with long-term follow-up to confirm these findings. Although these results are applicable to patients with reasonably preserved renal function, those with severe renal impairment deserve further study.
D.J. has received grants/research support from Aspreva and Roche Pharmaceuticals UK.
M.W. is supported by the Kidney Research Scientist Core Education National Training Program and the Alberta Heritage Foundation for Medical Research.
This study appeared in abstract form at the annual meeting of the American Society of Nephrology; November 14 through 19, 2006; San Diego, CA.
Published online ahead of print. Publication date available at www.cjasn.org.
See the related editorial, “Role of Mycophenolate Mofetil in the Treatment of Lupus Nephritis,” on pages 879–882.
Access to UptoDate on-line is available for additional clinical information at http://www.cjasn.org/
1. Rosner S, Ginzler EM, Diamond HS et al.: A multicenter study of outcome in systemic lupus erythematosus. II. Causes of death. Arthritis Rheum25 :612– 617,1982
2. Flanc RS, Roberts MA, Strippoli GF, Chadban SJ, Kerr PG, Atkins RC: Treatment of diffuse proliferative lupus nephritis: A meta-analysis of randomized controlled trials. Am J Kidney Dis43 :197– 208,2004
3. Cross J, Jayne D: Diagnosis and treatment of kidney disease. Best Pract Res Clin Rheumatol19 :785– 798,2005
4. Boumpas DT, Austin HA 3rd, Vaughn EM, Klippel JH, Steinberg AD, Yarboro CH, Balow JE: Controlled trial of pulse methylprednisolone versus two regimens of pulse cyclophosphamide in severe lupus nephritis. Lancet340 :741– 745,1992
5. Boumpas DT, Austin HA 3rd, Vaughan EM, Yarboro CH, Klippel JH, Balow JE: Risk for sustained amenorrhea in patients with systemic lupus erythematosus receiving intermittent pulse cyclophosphamide therapy. Ann Intern Med119 :366– 369,1993
6. Austin HA, Boumpas DT: Treatment of lupus nephritis. Semin Nephrol16 :527– 535,1996
7. Gourley MF, Austin HA 3rd, Scott D, Yarboro CH, Vaughan EM, Muir J, Boumpas DT, Klippel JH, Balow JE, Steinberg AD: Methylprednisolone and cyclophosphamide, alone or in combination, in patients with lupus nephritis. A randomized, controlled trial. Ann Intern Med125 :549– 557,1996
8. Gaubitz M, Schorat A, Schotte H, Kern P, Domschke W: Mycophenolate mofetil for the treatment of systemic lupus erythematosus: An open pilot trial. Lupus8 :731– 736,1999
9. Lazarov V, Rapondjievava A, Shipkova M, Michailova A, Naumova E, Tiskov I, Nikolov D, Oellerich M: Treatment with mycophenolate mofetil of patients with immune nephropathies. Nephrol Hemodial Transplant5 :18– 21,1999
10. Dooley MA, Cosio FG, Nachman PH, Falkenhain ME, Hogan SL, Falk RJ, Hebert LA: Mycophenolate mofetil therapy in lupus nephritis: Clinical observations. J Am Soc Nephrol10 :833– 839,1999
11. Cross J, Jayne D: Mycophenolate mofetil in lupus nephritis. Lupus9 :647– 650,2000
12. Hu W, Liu Z, Chen H, Tang Z, Wang Q, Shen K, Li L: Mycophenolate mofetil vs cyclophosphamide therapy for patients with diffuse proliferative lupus nephritis. Chin Med J (Engl)115 :705– 709,2002
13. Ginzler EM, Dooley MA, Aranow C, Kim MY, Buyon J, Merrill JT, Petri M, Gilkeson GS, Wallace DJ, Weisman MH, Appel GB: Mycophenolate mofetil or intravenous cyclophosphamide for lupus nephritis. N Engl J Med353 :2219– 2228,2005
14. Ong LM, Hooi LS, Lim TO, Goh BL, Ahmad G, Ghazalli R, Teo SM, Wong HS, Tan SY, Shaariah W, Tan CC, Morad Z: Randomized controlled trial of pulse intravenous cyclophosphamide versus mycophenolate mofetil in the induction therapy of proliferative lupus nephritis. Nephrol10 :504– 510,2005
15. Chan TM, Li FK, Tang CS, Wong RW, Fang GX, Ji YL, Lau CS, Wong AK, Tong MK, Chan KW, Lai KN: Efficacy of mycophenolate mofetil in patients with diffuse proliferative lupus nephritis. N Engl J Med343 :1156– 1162,2000
16. Jadad AR, Moher D, Klassen TP: Guides for reading and interpreting systematic reviews: II. How did the authors find the studies and assess their quality? Arch Pediatr Adolesc Med152 :812– 817,1998
17. Egger M, Smith GD, Altman D: Meta-analysis in Stata. In: Systematic Reviews in Health Care: Meta-Analysis in Context, London, BMJ Publishing Group,2001 , pp347– 372
18. Chan TM, Tse KC, Tang CS, Mok MY, Li FK; Hong Kong Nephrology Study Group: Long-term study of mycophenolate mofetil as continuous induction and maintenance treatment for diffuse proliferative lupus nephritis. J Am Soc Nephrol16 :1076– 1084,2005
19. Flores-Suarez LF, Villa AR: Open randomized trial comparing mycophenolate mofetil versus intravenous cyclophosphamide as induction therapy for severe lupus nephritis. J Am Soc Nephrol15 :PO257 ,2004
20. Lin YK, Liu D, Wei G: A comparison of response between mycophenolate mofetil and cyclophosphamide therapy for lupus nephritis. J Clin Derm31 :636– 638,2002
21. Balow JE: Choosing treatment for proliferative lupus nephritis. Arthritis Rheum46 :1981– 1983,2002
22. Flather MD, Farkouh ME, Pope JM, Yusuf S: Strengths and limitations of meta-analysis: Larger studies may be more reliable. Control Clin Trialsl8 :568– 579,1997
23. Sinclair A, Appel G, Dooley M, Ginzler E, Isenberg D, Jayne R, Wofsy D: Protocol for the Aspreva Lupus Management Study (ALMS). J Am Soc Nephrol16 :FPO882 ,2005