Journal Logo


Efficacy and Adverse Events of Mycophenolate Mofetil Versus Cyclophosphamide for Induction Therapy of Lupus Nephritis

Systematic Review and Meta-Analysis

Kamanamool, Nanticha MD, MSc; McEvoy, Mark MSc; Attia, John MD, PhD; Ingsathit, Atiporn MD, PhD; Ngamjanyaporn, Pintip MD; Thakkinstian, Ammarin PhD

Author Information
doi: 10.1097/MD.0b013e3181e93d00
  • Free



Systemic lupus erythematosus (SLE) is a multisystem autoimmune disease characterized by the production of autoantibodies and immune complexes and the development of chronic inflammation. The disease has a prevalence of 1 case per 2000 persons; it predominately affects young women and is more common in African-Americans and Asians than in whites.7 There is significant morbidity and mortality related to the disease itself or its treatment complications, and quality of life is often severely affected. Lupus nephritis (LN) is a common renal manifestation of the disease, with about 30% of patients with SLE developing nephritis by 3 years after diagnosis.20 The immune system plays a major role in the pathogenesis of LN, with the production of autoantibodies leading to immune complex deposition in the kidneys;12 this initiates an inflammatory response and results in lymphocyte and macrophage infiltration in the glomeruli.

There are 6 classes of LN.24 These include minimal mesangial LN (World Health Organization [WHO] class I), mesangial proliferative LN (WHO class II), focal LN (WHO classes III), diffuse segmental (WHO class IV-S) or global LN (WHO class IV-G), membranous LN (WHO class V), and advanced sclerosing LN (WHO class VI). The severity of LN varies from minimal lesions to advanced sclerosis, which may progress to end-stage renal disease. Treatment of LN remains a challenge because of the heterogeneity of the disease and the unpredictable outcomes. The main treatment goal is disease remission, which is associated with longer patient survival5 and reduced renal flares, chronic renal failure, and end-stage renal disease.

Cyclophosphamide (CYC) is an immunosuppressive drug that has been used for the treatment of class III-V LN. Previous researchers have found that a combination regimen of CYC and corticosteroids is more effective than corticosteroids alone, and this combination remains the current standard treatment.3,6,15 Adverse effects of CYC, however, such as severe infection or ovarian failure, have been frequently reported,8 and thus alternative immunosuppressive drugs such as azathioprine and mycophenolate mofetil (MMF) are increasingly used. The efficacy and adverse events of these drugs have been studied and reviewed. However, many trials1,2,4,13,21 comparing MMF with CYC for induction therapy have a small sample size (for example, 20-140), and hence suffer from insufficient power. The small sample size has also limited subgroup analyses, making it difficult to identify which types of patients (for example by age, sex, baseline serum creatinine, 24-hour urine protein, and serum albumin) would receive more or less benefit from alternative agents. Previous reviews9,10,16,17,19,22,25 did not consider other clinical outcomes such as serum creatinine, 24-hour urine protein, and serum albumin, and there have been few randomized trials for induction periods published since the last reviews. We therefore performed a systematic review and meta-analysis with the aims of comparing complete and partial remission, as well as adverse events (infection, leukopenia, and gastrointestinal [GI] symptoms) between MMF and CYC regimens. In addition, we compared other clinical outcomes (serum creatinine, 24-hour urine protein, and serum albumin). Patients and clinical factors were incorporated in analyses if data were available.


Study Search and Selection

One reviewer (NK) electronically searched the MEDLINE database using PubMed (National Library of Medicine, Bethesda, MD) (1951 to December 2009) and Ovid (WoltersKluwer, New York, NY) (1966 to December 2009), and The Cochrane Central Register of Randomized Controlled Trials (CENTRAL-The Cochrane Library issue 4, 2009) (United States Cochrane Center, Baltimore, MD). Search terms used without language restriction were as follows: (mycophenolate mofetil or mycophenolate) and cyclophosphamide and (lupus nephritis or glomerulonephritis), limited to randomized controlled trial. (See Appendix for more details about the search strategies for each database).

Two reviewers (NK and AT) independently screened titles and abstracts. Trials with the following criteria were included: trial included adult patients (aged ≥18 yr) with WHO Class III-V LN; trial compared MMF plus corticosteroids with CYC plus corticosteroids for induction therapy; and trial reported outcomes as complete/partial remission, change in serum creatinine, proteinuria, or serum albumin. For trials with multiple publications, we selected the publication with the most complete information. Disagreements in selection were resolved by discussion and consensus.

Data Extraction and Risk Assessment

Two reviewers (NK and AT) independently performed data extraction. We extracted trial characteristics (for example, study design, sample size, treatment dosage and duration, WHO classification, renal biopsy information) and definitions (complete remission and complete/partial remission). We also extracted numbers of complete remission, complete/partial remissions, and numbers of adverse events between treatment groups. In addition, we extracted mean serum creatinine, 24-hour urine protein, and serum albumin after treatments whenever available.

Outcomes of interest for the current study were complete remission alone, complete or partial remission, and adverse events (for example, infection, leukopenia, and GI symptoms). In addition, continuous outcomes such as the change of serum creatinine, 24-hour urine protein, and serum albumin were considered. Definitions of these outcomes used in the original papers were extracted as described in Table 1.

Definitions and Outcome Measurements Used in Included Trials

Two independent reviewers (AI and AT) assessed risk of bias or methodologic quality of the included trials using Cochrane criteria and Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines,14,18 which considered adequate generation of randomization, allocation concealment, blinding, incomplete outcome data, selective outcome reporting, and possible sources of other bias. Each item was judged as follows: "Yes" or adequate (low risk of bias) if that item was clearly described in detail for the method, "No" (high risk of bias) or inadequate if not described in adequate detail, or "Unclear" if a judgment could not make. Any disagreement was resolved by discussion and consensus.

Data Synthesis

For dichotomous outcomes, risk ratio [RR] with 95% confidence interval [CI] for each trial was estimated. Presence of heterogeneity between studies was assessed using the Cochran Q (heterogeneity chi-square), and I2statistics. The random effects model (DerSimonian and Laird method) was applied for all pooling. To explore sources of heterogeneity, meta-regression was used to assess the association between lnRR of (complete/partial remissions) and covariables (for example, age, sex), baseline serum creatinine, 24-hour urine protein, and serum albumin. When the source of heterogeneity was detected, subgroup or sensitivity analysis was performed accordingly. Publication bias was explored using funnel plots and the Egger regression asymmetry test.

Finally, number needed to treat (NNT) or number needed to harm (NNH) was estimated for dichotomous outcomes. The number of patients who would experience disease recovery or be harmed by complications (for example, infection, leukopenia, and GI symptoms) was estimated if 100 patients were treated.

For continuous outcomes, mean differences of serum creatinine, 24-hour urine protein, and serum albumin between baseline and 6 months after induction period according to 2 treatment groups were summarized using standardized mean difference (SMD). The p value of < 0.05 was considered statistically significant except where the heterogeneity test p < 0.10 was used. All data were analyzed with STATA version 10.0 (StataCorp LP, College Station, TX).


We identified 59 publications from 3 databases, as summarized in Figure 1. Twenty-eight publications were removed due to duplication, which left 31 titles or abstracts screened. Twenty-four papers were excluded for reasons described in Figure 1, leaving 7 trials eligible for review.1,2,4,11,13,21,23 Two of 7 trials were excluded: 1 was available only as an abstract11 and contacting the author for additional data was unsuccessful; another trial2 added tacrolimus in the MMF group, which also has an immunosuppressive effect. Five trials1,4,13,21,23 were finally included in the analysis.

Flow diagram of trial selection. Abbreviations: *CCRCT = Cochrane Central Register of Randomized Controlled Trials, †LN = lupus nephritis, ‡AZA = azathioprine, §CYC = cyclophosphamide.

Trial characteristics are described in Table 2. All 5 trials compared remission at 6 and/or 12 months between MMF plus corticosteroids versus CYC plus corticosteroids. CYC was administered intravenously in 4 trials1,13,21,23 and orally in 1 trial.4

Characteristics of Patients in Included Trials

The MMF and CYC dosages were adjusted according to various criteria, ranging from 1 to 3 g/d for MMF and 0.75-1 g/m2 of body surface area for CYC. Mean age of patients in these trials ranged from 30.8 to 39.9 years, and most patients were female patients with biopsy-proven nephritis WHO class III-V. Serum creatinine at baseline was quite different across trials, ranging from 1.07 to 1.50 mg/dL. The 24-hour urine protein was quite similar (range, 4.10-5.35 g) except for 1 study21 where this was lower than the other trials (mean, 2.48 g). This was also seen with serum albumin, which was similar across all trials (range, 26.12-28.99 g/L).

Risk assessment was performed following PRISMA guidelines14,18 by 2 reviewers and is described in Table 3. The estimated level of agreement was moderately high (Kappa statistic = 0.62). Among the 5 trials, 2 trials4,23 did not adequately describe the method of sequence generation, and 1 trial21 might have bias from other sources-that is, imbalance in baseline proteinuria and use of angiotensin-converting enzyme inhibitors or angiotensin II receptor blocker.

Risk of Bias of Included Trials*

Complete Remission

Five trials1,4,13,21,23 (n = 638 patients) reported complete remission between MMF (n = 317) and CYC (n = 323) groups and thus were included in pooling treatment effects (Table 4). The included treatment effects (RRs) were heterogeneous (I2 = 59.2%, chi-square = 9.80 [degree of freedom {df} = 4], p = 0.044), and thus the DerSimonian and Laird method was applied for pooling. The pooled RR for complete remission was 1.60 (95% CI, 0.87-2.93), which slightly favored MMF although this did not reach statistical significance (Figure 2). We explored the possible source of heterogeneity using meta-regression by fitting mean age, percentage of female patients, baseline serum creatinine, 24-hour urine protein, and serum albumin into the meta-regression model one by one. This suggested that only the percentage of female patients could explain heterogeneity by reducing I2 to be 0, though its coefficient was not significant (coefficient = 0.47, standard error [SE] = 0.33, p = 0.232). The trial by Ginzler et al13 had 90% female patients, which is higher than the other 4 trials (81%-85%). We thus performed a sensitivity analysis with the 4 trials with lower percentages of female patients. The pooled effect was less heterogeneous (chi-square = 4.11 [df = 3], p = 0.250, I2= 26.9%) with a pooled RR of 1.15 (95% CI, 0.74-1.77). This could be interpreted to mean that MMF had approximately 15% higher benefit, but the chance could be from 26% lower to 77% higher benefit compared with CYC group. Using the Egger test on the 5 trials, we found borderline evidence of bias (coefficient = 2.03, SE = 0.64, p = 0.049) from the small study effects. We then performed a funnel plot study, which suggested that the trial by Ginzler et al13 lay outside the pseudo 95% CI and might be the cause of heterogeneity or bias (Figure 3). Eliminating that trial yielded non-significant coefficient of bias (coefficient = 1.56, SE = 0.38, p = 0.054).

Forest plots of remission and adverse outcomes. Abbreviations: *RR = relative risk, †CI = confidence interval.
Funnel plot of complete remission.
Frequencies Between Treatments and Outcomes and Estimated Treatment Effects

Complete or Partial Remission

All 5 trials1,4,13,21,23 (n = 638) included complete or partial remission outcomes (Table 4). There was evidence of heterogeneity among these trials (I2 = 62.7%, chi-squared = 10.72 [df = 4], p = 0.030). The pooled RR was 1.20 (95% CI, 0.97-1.48), which indicated that complete or partial remission was not significantly different between 2 groups (Figure 2).

Adverse Outcomes

Adverse outcomes (infection, leukopenia, and GI symptoms) are described in Table 4. All 5 trials1,4,13,21,23 (n = 638) reported data for infection. Treatment effects on infection were heterogeneous (chi-square = 10.8 [df = 4], p = 0.029, I2 = 63.0%), and the pooled RR was 0. 77 (95% CI, 0.39-1.49). The MMF group tended to have lower infection than the CYC group, but this effect was not significant. Three trials4,13,23 (n = 206 patients) were included in pooling for leukopenia. These trials were homogeneous (chi-square = 0.38 [df = 1], p = 0.825, I2 = 0%) and the pooled RR was 0.65 (95% CI, 0.44-0.96) (Figure 2). This suggests that leukopenia occurred significantly less frequently in patients treated with MMF compared with CYC. Four trials1,4,13,23 (n = 594) were included in pooling for GI symptoms; GI symptoms were not significantly different between the 2 groups (pooled RR, 1.09; 95% CI, 0.74-1.60). Only 2 trials4,13 (n = 179 patients) reported ovarian failures, and data were not sufficient to pool.

Number Needed to Treat/Number Needed to Harm

The NNT/NNH for each outcome was estimated as described in Table 5. For complete remission, the NNT for all 5 trials was 10 (95% CI: NNH = −47, NNT = 3), indicating that if 100 patients were treated with MMF instead of CYC, 10 extra patients would experience complete remission. Excluding the trial by Ginzler et al13 that caused heterogeneity and/or publication bias resulted in an NNT of 28 (95% CI: NNH 16, NNT 5), that is, only 4 extra patients would experience complete remission for every 100 patients treated with MMF. The NNT for complete or partial remission was similar to that for complete remission alone. The NNH for infection and leukopenia were −15 (95% CI: NNH −6, NNT 7) and −9 (95% CI: NNH −6, NNH −74), respectively. This indicates that 7 and 11 fewer patients suffer from infection and leukopenia respectively for every 100 patients treated with MMF.

Estimation of Number Needed to Treat and Number Needed to Harm According to Outcome

Renal Function

Renal function after the induction period is summarized in Table 6. Three trials13,21,23 (n = 174) reported serum creatinine after induction therapy. There was no heterogeneity among these trials (I2 = 0%, chi-square = 0.14, p = 0.933). Applying the SMD method for pooling suggested a statistically not significant pooled SMD of 0.21 (95% CI, −0.09 to 0.51). Three trials13,21,23 (n = 204) were included in pooling for mean 24-hour urine protein and mean serum albumin. There was evidence of heterogeneity in 24-hour urine protein and serum albumin; I2 was 76.9% and 38.9%, chi-square was 8.65 (p = 0.013) and 3.27 (p = 0.195), respectively. The pooled SMDs of 24-hour urine protein and serum albumin were −0.31 (95% CI, −1.32 to 0.69) and 0.78 (95% CI, −0.92 to 2.45), respectively, which were not statistically significant (Table 6).

Mean Difference and Estimated Standardized Mean Difference for Continuous Outcomes


In the current updated systematic review, we have summarized current evidence of clinical outcomes and side effects of MMF and CYC for induction therapy of LN. Our results suggest a potential small advantage of MMF over CYC in complete remission or partial remission, although the results are not significant. Only 1 adverse effect, leukopenia, was significantly less frequent (35% lower) in patients treated with MMF instead of CYC; infection and GI symptoms were similar. Renal function (that is, serum creatinine, 24-hour urine protein, and serum albumin) after induction therapy was also similar between the MMF and CYC groups. Results were consistent after excluding 1 trial13 that caused heterogeneity or publication bias.

The current review provides more complete evidence than previous reviews9,10,16,19,22,25 for a number of reasons. First, we considered more relevant outcomes (that is, complete remission, partial remission, leukopenia, infection, GI symptoms, serum creatinine, 24-hour urine protein, and serum albumin) than previous reviews. Also, some data in the review by Mak et al17 were extracted from abstracts, and thus risk of biases in these studies could not be fully assessed. NNT and NNH for these dichotomous outcomes were also estimated and reported. The current review included a greater number of trials with a larger sample size than previous reviews. Finally, in this review we explored possible causes of heterogeneity and publication bias, and excluded a suspected trial which yielded consistent results of treatment effects on complete remission.

Despite the improvements made in the current review, there may be some limitations. Although all the randomized controlled trials included here are fairly similar in baseline characteristics of patients, definitions of complete and partial remission, and duration of follow-up, there were a few differences that might introduce bias in our analysis. For example, drug dosages and adjustment (dosages of MMF = 1-3 g/d, dosages of intravenous CYC = 0.75-1 g/m2 of body surface area) and routes of administration (oral vs. intravenous CYC) differed among some trials and might contribute to heterogeneity of treatment effects. Performing a meta-analysis of summary data, we were not able to explore what dosages or routes of administration were most beneficial, nor to adjust for these differences. To do this, individual patient data are required.

In summary, results of the current review suggest that there are insufficient data to distinguish a statistical difference in efficacy between MMF and CYC on complete remission, complete/partial remission, infection, and GI symptoms, but there is a definite decreased risk of leukopenia with MMF. Further large scale trials are needed to confirm these results.


A. Search strategies: MEDLINE (PubMed)

("mycophenolate mofetil"[All Fields] OR mycophenolate [All Fields]) AND ("cyclophosphamide" [MeSH Terms] OR "cyclophosphamide" [All Fields] OR "Azathioprine" [All Fields]) AND (("lupus nephritis" [MeSH Terms] OR ("lupus" [All Fields] AND "nephritis" [All Fields]) OR "lupus nephritis" [All Fields]) OR ("glomerulonephritis" [MeSH Terms] OR "glomerulonephritis" [All Fields])) AND "randomized controlled trial" [Publication Type].

B. Search Strategies: MEDLINE (Ovid)

  1. Mycophenolate
  5. Lupus
  7. 1 and 5
  8. 6 and 1
  9. 2 and 5
  10. 6 and 2
  11. 3 and 5
  12. 6 and 3
  13. 4 and 5
  14. 6 and 4
  15. 3 and 7
  16. 8 and 3
  17. 3 and 9
  18. 3 and 10
  19. 1 and 13
  20. 1 and 14
  21. 13 and 2
  22. 2 and 14
  23. 22 or 21 or 18 or 19 or 16 or 17 or 20 or 15
  24. limit 23 to randomized controlled trial

C. Search Strategies: CENTRAL (Issue 4, 2009) Cochrane Central Register of Controlled Trials

  • #1 (Mycophenolate mofetil):ti,ab,kw
  • #2 (Mycophenolate):ti,ab,kw
  • #3 (Cyclophosphamide):ti,ab,kw
  • #4 (Azathioprine):ti,ab,kw
  • #5 (Lupus Nephritis):ti,ab,kw
  • #6 (Glomerulonephritis):ti,ab,kw
  • #7 (#1 AND #5)
  • #8 (#1 AND #6)
  • #9 (#2 AND #5)
  • #10 (#2 AND #6)
  • #11 (#3 AND #5)
  • #12 (#3 AND #6)
  • #13 (#4 AND #5)
  • #14 (#4 AND #6)
  • #15 (#3 AND #7)
  • #16 (#3 AND #8)
  • #17 (#3 AND #9)
  • #18 (#3 AND #10)
  • #19 (#4 AND #7)
  • #20 (#4 AND #8)
  • #21 (#4 AND #9)
  • #22 (#4 AND #10)
  • #23 (#15 OR #16 OR #17 OR #18 OR #19 OR #20 OR #21 OR #22)


1. Appel GB, Contreras G, Dooley MA, Ginzler EM, Isenberg D, Jayne D, Li L-S, Mysler E, Sanchez-Guerrero J, Solomons N, Wofsy D, the Aspreva Lupus Management Study Group. Mycophenolate mofetil versus cyclophosphamide for induction treatment of lupus nephritis. J Am Soc Nephrol. 2009;20:1103-1112.
2. Bao H, Liu Z-H, Xie H-L, Hu W-X, Zhang H-T, Li L-S. Successful treatment of class V+IV lupus nephritis with multitarget therapy. J Am Soc Nephrol. 2008;19:2001-2010.
3. Carette S, Klippel JH, Decker JL, Austin HA, Plotz PH, Steinberg AD, Balow JE. Controlled studies of oral immunosuppressive drugs in lupus nephritis. A long-term follow-up. Ann Intern Med. 1983;99:1-8.
4. Chan T-M, Tse K-C, Tang CS-O, Mok M-Y, Li F-K, for the 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 Nephrol. 2005;16:1076-1084.
5. Drenkard C, Villa AR, Alarcon-Segovia D, Perez-Vazquez ME. Influence of the antiphospholipid syndrome in the survival of patients with systemic lupus erythematosus. J Rheumatol. 1994;21:1067-1072.
6. Felson DT, Anderson J. Evidence for the superiority of immunosuppressive drugs and prednisone over prednisone alone in lupus nephritis. Results of a pooled analysis. N Engl J Med. 1984;311:1528-1533.
7. Feng P-H. Systemic lupus erythematosus: the face of Asia. Ann N Y Acad Sci. 2007;1108:114-120.
8. Fine DM. Pharmacological therapy of lupus nephritis. JAMA. 2005;293:3053-3060.
9. Flanc RS, Roberts MA, Strippoli GFM, Chadban SJ, Kerr PG, Atkins RC. Treatment for lupus nephritis. Cochrane Database Syst Rev. 2004;(1):CD002922-CD002922.
10. Flanc RS, Roberts MA, Strippoli GFM, Chadban SJ, Kerr PG, Atkins RC. Treatment of diffuse proliferative lupus nephritis: a meta-analysis of randomized controlled trials. Am J Kidney Dis. 2004;43:197-208.
11. Flores-Suarez LF, Villa AR. Open randomized trial comparing mycophenolate mofetil versus intravenous cyclophosphamide as induction therapy for severe lupus nephritis. J Am Soc Nephrol. 2004;15:P0257-P0257.
12. Foster MH. T cells and B cells in lupus nephritis. Semin Nephrol. 2007;27:47-58.
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 Med. 2005;353:2219-2228.
14. Higgins JP, Altman DG. Assessing risk of bias in included studies. Cochrane Handbook for Systematic Reviews of Interventions. Version 5.0.1; 2008.
15. Houssiau FA. Management of lupus nephritis: an update. J Am Soc Nephrol. 2004;15:2694-2704.
16. Lee Y, Woo JH, Choi S, Ji J, Song G. Induction and maintenance therapy for lupus nephritis: a systematic review and meta-analysis. Lupus. 2010;19:703-710.
17. Mak A, Cheak AAC, Tan JYS, Su HC, Ho RCM, Lau CS. Mycophenolate mofetil is as efficacious as, but safer than, cyclophosphamide in the treatment of proliferative lupus nephritis: a meta-analysis and meta-regression. Rheumatology (Oxford). 2009;48:944-952.
18. Moher D, Liberati A, Tetzlaff J, Altman DG. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. BMJ. 2009;339:b2535.
19. Moore RA, Derry S. Systematic review and meta-analysis of randomised trials and cohort studies of mycophenolate mofetil in lupus nephritis. Arthritis Res Ther. 2006;8:R182-R182.
20. Nossent JC, Bronsveld W, Swaak AJ. Systemic lupus erythematosus. III. Observations on clinical renal involvement and follow up of renal function: Dutch experience with 110 patients studied prospectively. Ann Rheum Dis. 1989;48:810-816.
21. Ong LM, Hooi LS, Lim TO, Goh BL, Ahmad G, Ghazalli R, Teo SM, Wong HS, Tan SY, Shaariah WAN, Tan CC, Morad Z. Randomized controlled trial of pulse intravenous cyclophosphamide versus mycophenolate mofetil in the induction therapy of proliferative lupus nephritis. Nephrology. 2005;10:504-510.
22. Walsh M, James M, Jayne D, Tonelli M, Manns BJ, Hemmelgarn BR. Mycophenolate mofetil for induction therapy of lupus nephritis: a systematic review and meta-analysis. Clin J Am Soc Nephrol. 2007;2:968-975.
23. Wang J, Hu W, Xie H, Zhang H, Chen H, Zeng C, Liu Z, Li L. Induction therapies for class IV lupus nephritis with non-inflammatory necrotizing vasculopathy: mycophenolate mofetil or intravenous cyclophosphamide. Lupus. 2007;16:707-712.
24. Weening JJ, D'Agati VD, Schwartz MM, Seshan SV, Alpers CE, Appel GB, Balow JE, Bruijn JA, Cook T, Ferrario F, Fogo AB, Ginzler EM, Hebert L, Hill G, Hill P, Jennette JC, Kong NC, Lesavre P, Lockshin M, Looi L-M, Makino H, Moura LA, Nagata M. The classification of glomerulonephritis in systemic lupus erythematosus revisited. J Am Soc Nephrol. 2004;15:241-250.
25. Zhu B, Chen N, Lin Y, Ren H, Zhang W, Wang W, Pan X, Yu H. Mycophenolate mofetil in induction and maintenance therapy of severe lupus nephritis: a meta-analysis of randomized controlled trials. Nephrol Dial Transplant. 2007;22:1933-1942.
© 2010 Lippincott Williams & Wilkins, Inc.