Introduction

Approximately 80% of children presenting with an initial episode of nephrotic syndrome respond to steroids (steroid sensitive) whereas another 20% do not respond and are considered as steroid resistant [1]. Of children with steroid-sensitive nephrotic syndrome, 50–60% become frequently relapsing to steroid-dependent nephrotic syndrome (SDNS) on follow-up. Children with SDNS and steroid-resistant nephrotic syndrome (SRNS) pose therapeutic challenges, albeit of somewhat different nature. The standard therapy for these conditions has not changed in more than two decades with the exception of the introduction of cyclosporine A (CSA) [2, 3, 4]. Although useful, CSA therapy is associated with a significant side-effect profile, the most concerning of which are the potential for nephrotoxicity and development of CSA dependence. Clearly, better alternative therapies with improved efficacy and fewer side effects are desirable.

Steroid-dependent nephrotic syndrome

The majority of children with SDNS show minimal change in lesions on histologic examination, and a small number may show focal segmental sclerosis (FSGS), mesangial proliferation (MesPGN), and other histologic lesions [5]. Therapies employed in the past to treat SDNS have included long-term alternate-day prednisone, oral cyclophosphamide, chlorambucil, levamisole, CSA, and tacrolimus [6, 7]. Recently, mycophenolate mofetil (MMF) was shown to be useful in the management of these children [8, 9, 10, 11].

Long-term alternate-day prednisone in tapering doses has been the mainstay of therapy for a few decades. Arbeitsgemeinschaft fur Padiatrische Nephrologie (APN) was first to show that prolongation of steroid therapy for the initial episode to 12 weeks rather than 8 weeks, as proposed by the International Study of Kidney Disease in Children (ISKDC), is usually associated with more durable remissions [12]. Recently, Hiraoka et al. have also shown that prolonged therapy for the initial episode of nephrotic syndrome with 4 weeks of daily prednisone and then tapering doses of alternate-day prednisone for 6 months was associated with more cumulative remissions in children younger than 4 years of age [13]. While long-term alternate prednisone therapy has the best track record in the management of frequently relapsing steroid-sensitive nephrotic syndrome, it is usually not the most acceptable therapy to the patients. Long-term steroid therapy is complicated by side effects such as excessive weight gain, cushingoid features, altered body habitus, hypertension, growth retardation, glucose intolerance, cataracts, mood changes, decreased bone mineralization in growing children, acne, and hirsutism. Cosmetic features are particularly unacceptable to teenagers in an ever-growing body-image-conscious society. This becomes particularly difficult in steroid-dependent children. Thus, alternate therapies are desireable.

Oral cyclophosphamide therapy for 8–12 weeks can induce long-term remission in 25–60% of children with frequently relapsing and SDNS [14]. The results are less beneficial in children with steroid-dependent type [15]. Side effects include marrow suppression and increased risk of infections. Usually, there are no long-term side effects of limited cyclophosphamide therapy. Repeated courses are fraught with dangers of marrow suppression, gonadal toxicity, hemorrhagic cystitis, alopecia, and increased risk of malignancy. Gonadal toxicity is more marked in males. These side effects are dependent upon the cumulative dose of cyclophosphamide used [16]. Therefore, repeated courses of cyclophosphamide are not desirable and should be avoided. Chlorambucil has similar efficacy and side-effect profile [17].

Since early 1990, CSA has emerged as a useful therapy in children with frequently relapsing and SDNS. About 80% of these children respond to CSA, and steroids can be tapered off. A minority of children may need a small dose of steroids in addition to CSA to keep them in remission. However, CSA withdrawal is usually associated with disease relapses in 90% of children, necessitating reinstitution of cyclosporine [18]. Long-term CSA therapy is associated with a number of side effects that include nephrotoxicity, hypertension, cosmetic side effects (hirsutism, gum hyperplasia), tremors, glucose intolerance, dyslipidemia, hyperkalemia, increased risk of infections, and increased potential for malignancy. Nephrotoxicity is related to the dose and duration of CSA therapy and the extent of proteinuria [19]. In limited uncontrolled studies, tacrolimus (FK 506) was shown to be effective in the treatment of SDNS [20, 21]. Tacrolimus has a similar side effect profile to CSA except for reduced cosmetic side effects and slightly decreased incidence of nephrotoxicity, hypertension, and dyslipidemia. However, its use is associated with an increased incidence of diabetes compared to CSA [22].

Levamisole has a marginally beneficial effect on the course of children with frequently relapsing and SDNS. It has steroid sparing properties with a rather safe side effect profile. However, most children tend to relapse upon discontinuation of the drug [23, 24, 25]. Vasculitis is a known side effect of the drug that is usually reversible on drug discontinuation [26]. This drug is unfortunately not available in the US market. Azathioprine was not found to be useful in the treatment of frequently relapsing and SDNS in early trials in the 1970s [27]. There was renewed interest in the drug, but limited beneficial effects were seen [28].

MMF is a prodrug of mycophenolic acid that is rapidly absorbed following oral administration and is hydrolyzed to form mycophenolic acid (MPA). MPA, the active metabolite of MMF, is a potent, selective, uncompetitive, and reversible inhibitor of inosine monophosphate dehydrogenase (IMPDH), an enzyme required for de novo purine synthesis (Fig. 1). MPA inhibits T- and B-lymphocyte proliferation, as these cells are critically dependent upon de novo purine synthesis for their proliferation whereas other cell types can utilize salvation pathways for purine synthesis. MPA also prevents glycosylation of lymphocyte and monocyte glycoproteins that are involved in the intercellular adhesion of leukocytes to endothelium [29]. MPA may induce lymphocyte apoptosis and alter cell-surface adhesion molecules and cytokine gene expression [30].

Fig. 1
figure 1

Inhibition of de novo purine synthesis by mycophenolic acid (MPA), the active metabolite of mycophenolate mofetil (MMF). PPRP phosphoribosyl triphosphate, ADA adenosine deaminase, IMPDH inosine monophosphate dehydrogenase

Full size image

Since its FDA approval in 1995 for use in renal transplantation, considerable interest has arisen regarding the potential benefits of MMF in the treatment of immune-mediated glomerular diseases, including childhood nephrotic syndrome [8, 9, 10, 11, 31, 32, 33, 34, 35, 36, 37]. We recently reported a study on the beneficial effects of MMF in 19 children (ten with minimal change, three with FSGS, and six not biopsied) with SDNS [10]. Patients with two consecutive relapses of nephrotic syndrome while receiving prednisolone at a dose of 1.5 mg/kg on alternate days or within 15 days of its discontinuation were defined as SDNS. The initial therapy of SDNS comprised long-term treatment with tapering doses of alternate-day prednisolone. Patients requiring prednisolone dosage of more than 0.5 mg/kg on alternate days or/and features of steroid toxicity were treated using levamisole or cyclophosphamide. Those who continued to show SDNS despite therapy with levamisole and/or cyclophosphamide were considered for treatment with MMF. These children were a group of very difficult nephrotics who were previously treated with a long course of alternate-day steroids (19/19), cyclophosphamide (15/19, four with more than one course), and levamisole (16/19) before treatment with MMF. On an average, these children had 6.6 relapses/year preceding MMF therapy. During 12 months of MMF therapy, the relapse rate decreased to two relapses per year, and cumulative doses of steroids decreased from 0.7 mg/kg per day to 0.3 mg/kg/day (p<0.01). Five patients had transient abdominal pain. None had diarrhea, vomiting, leukopenia, or thrombocytopenia. The frequency of serious systemic infections before, during, and after MMF therapy was similar. Therapy was well tolerated, and no patient had to stop therapy because of side effects. Unfortunately, the effect was not always sustained. The number of relapses increased to 4.2 per year following cessation of MMF, and the cumulative dose of steroids increased to 0.4 mg/kg per day.

This is an interesting preliminary study, and a number of important questions arise:

  1. 1.

    Effect of MMF in different racial groups: This study was carried out in India with a rather homogeneous Asian population. How the results will translate to the other racial groups remains to be seen. People of African descent need a higher dose of MMF to achieve protection from rejection in allograft recipients [38]. It is possible, therefore, that children of African descent may require a higher dose in treatment of frequently relapsing and SDNS.

  2. 2.

    Optimal duration of MMF therapy: How long one should treat with MMF is not known. In this study, treatment with MMF was carried out for 12 months. It is possible that prolonged therapy for a few years may have a more sustained effect (we know 12 months is not enough) and needs to be investigated. How long is too long for a sustained effect to be aquired needs to be determined since unnecessary prolongation of therapy may cause undue risks associated with immunosuppression. Monitoring for viral infections and other parameters associated with excess immunosuppression may minimize this risk.

  3. 3.

    Timing of MMF therapy in relation to the onset of nephrotic syndrome: In this study, patients were treated with alternate therapies for over a year prior to institution of MMF. Whether treatment with MMF earlier in the course of nephrotic syndrome will alter the course of steroid dependence is not known. The repeated episodes of massive proteinuria induce fibrotic changes in the tubulointerstitial compartment that may not be beneficial for long-term renal survival [39].

In another study of seven children with frequently relapsing nephrotic syndrome [six with minimal-change nephrotic syndrome (MCNS) and one with FSGS] in Europe, MMF was found to be beneficial [11]. These children were steroid sensitive and had been previously treated with CSA for a mean of 67.4 months. All had evidence of CSA nephrotoxicity as determined by decrease in glomerular filtration rate (GFR), decreased effective renal plasma flow (ERPF), and changes on renal biopsy. In five of six children with MCNS, no relapses were seen whereas one patient with low MPA levels showed a relapse. CSA could be discontinued in all six children with MCNS with an improvement in GFR and ERPF. These children have been on MMF therapy for a mean of 25.4 (15.3–39) months with no significant side effects. As expected, the renal histologic changes were not reversible. In one patient with FSGS, CSA therapy could be reduced successfully with institution of MMF, and the child went into complete remission that persisted over a follow-up of 28 months. At least in this pilot study, MMF seems to be promising in sustaining remission in children with frequently relapsing MCNS and partially reversing decline in ERPF and GFR. This study raises an interesting question of adjusting MMF dose by monitoring MPA levels, at least in some children not showing a desired response.

Barletta et al. treated ten children with CSA-dependent nephrotic syndrome with MMF. Four of these children had evidence of CSA toxicity, and eight showed evidence of tubulointerstitial fibrosis. They were able to completely taper CSA in five children after 1–2 years and both CSA and MMF in one patient. In a group of four children with SDNS where CSA was not used, MMF did not decrease frequency of relapses [34]. In this small, single-center, uncontrolled experience, MMF therapy in patients with CSA-dependent nephrotic syndrome appeared to be effective in reducing CSA exposure and significantly decreasing the frequency of relapses whereas MMF was not useful in children with SDNS in whom CSA was not used.

Ulinski et al. from France recently reported their results in nine children with SDNS and SRNS who were treated with MMF [35]. These children were on treatment with CSA and had shown a decrease in their GFR as measured by inulin clearance. MMF was introduced progressively until the full dose of 1 gm/1.73 m2 twice daily was reached. CSA treatment was stopped after introduction of MMF. Steroid dose was reduced if possible. After a median follow-up of 261 days, no adverse effects of MMF, such as diarrhea or hematologic abnormalities, were noted. After switch from CSA to MMF, the children with SDNS remained in remission and children with SRNS had no significant change in their residual proteinuria. The GFR improved from 76.9+4.8 to 119.9+5.9 (p<0.001). Oral steroid dose could be reduced from a median of 0.85 mg/kg per day pre-MMF to 0.29 mg/kg per day. This single-center, small, preliminary study shows beneficial effects of MMF in this group of patients showing decreased GFR on treatment with CSA; MMF therapy resulted in improvement of GFR, had a steroid-sparing effect, and maintained remission without causing any adverse effects. Since the follow-up interval was rather short, long-term efficacy and side effects of MMF therapy in this group of patients cannot be commented upon.

Recommendations

As MMF holds therapeutic promise in a small number of uncontrolled trials, prospective, randomized, controlled trials are needed to compare the efficacy and safety of MMF with CSA in frequently relapsing and SDNS. In our opinion, until more data become available, clinicians may want to use MMF in frequently relapsing and SDNS before trying CSA. If useful, it will obviate CSA nephrotoxicity and other undesired side effects. If not found to be helpful after 4–6 months, it is unlikely to be beneficial and MMF should be discontinued. As appropriate duration of treatment with MMF is undetermined, it will need to be tailored in individual patients. Given the accumulating data regarding efficacy and safety of MMF therapy in other autoimmune diseases, it seems likely that long-term use of MMF may be safe in children with SDNS and offer a less-toxic alternative to current therapies [40].

Steroid-resistant nephrotic syndrome

Children with SRNS pose the most difficult therapeutic challenge. These children are at risk of complications of unremitting nephrotic syndrome as well as developing end-stage renal disease (ESRD). Most children with SRNS have FSGS as the underlying lesion. Some may have MCNS, mesangial proliferative glomerulonephritis, or other lesions [1]. The rate of progression to ESRD is faster in African American and Hispanic children compared with Caucasians [41, 42]. Treatment options include high doses of pulse steroids with a combination of cyclophosphamide or chlorambucil. This regimen has met with variable success rates of inducing remission ranging from 10% to 70% [43, 44, 45, 46, 47, 48]. Many patients exhibit features of severe steroid toxicity. This regimen has been found to be less effective in African American children [48]. Prolonged pulses of IV cyclophosphamide over a few months induced remission in 25–60% of children with SRNS [48, 49, 50]. CSA reduces proteinuria in 50–70% children with SRNS and is used most commonly [4, 5].

MMF has been used in a few children with SRNS with FSGS [32]. It was shown to decrease proteinuria, increase serum albumin, and decrease serum cholesterol although complete remission was not achieved. These patients were pretreated with three pulses of Solumedrol prior to initiation of MMF therapy. Angiotensin-converting enzyme (ACE) inhibitors decrease proteinuria by about 50% in children with SRNS [52, 53]. Based on these uncontrolled findings, the National Institutes of Health has initiated a prospective, randomized, multicentric, controlled trial to compare CSA with a combination of pulse oral dexamethasone and MMF in steroid-resistant children and young adults with FSGS. Both groups will receive low-dose alternate-day prednisone and an ACE inhibitor. This study should answer questions regarding the efficacy of MMF in combination with steroids and ACE inhibitors in SRNS when compared with CSA. Unfortunately, this study does not include patients with SRNS who do not have an established diagnosis of FSGS. It is possible that some patients with FSGS may need a combination of CSA and MMF. However, this study is not intended to answer that question.

Future concepts for the use of MMF

Pharmacokinetic monitoring

Although initial reports suggested inter-individual variability of MPA to be rather low, more recent data have shown extensive variability at fixed-dose treatment [54]. Clinical trials have shown a strong correlation between plasma MPA levels and the likelihood of developing an acute rejection after kidney transplantation [55]. Pharmacokinetic data on MMF are limited in children, and most data are centered on pediatric renal transplant recipients [56]. Area under the curve (AUC) 0–12 h of free MPA in children receiving a dose of 1,200 mg/m2 per day in two divided doses corresponds to adults receiving 2 gm/day of MMF. Pharmacokinetics of MMF will need to be evaluated, especially in children with nephrotic syndrome. Hypoalbuminemia may alter the free MPA level and therefore affect toxicity and efficacy [57]. Glucocorticoids stimulate hepatic glucuronyl transferase, the key enzyme for the metabolism for MPA. Therefore, concomitant administration of glucocorticoids may affect MPA levels [58].

Pharmacodynamic monitoring

Large interindividual differences exist in baseline IMPDH enzyme activity. It is conceivable that the variability in the baseline IMPDH activity is caused by genetic differences in the IMPDH gene. The in vitro assessment of IMPDH activity and the analysis of cell-surface markers in activated T cells may be some of the newer pharmacodynamic approaches to further optimize MMF therapy [55].

Conclusions

In summary, early experience at single centers with MMF has shown benefit in children with frequently relapsing SDNS and SRNS, and its use may provide a steroid-sparing effect. However, of equal importance is a more benign side effect profile of MMF compared with currently available alternate choices of prednisone and CSA. Its use is not associated with abnormalities of lipid and carbohydrate metabolism, organ toxicity (nephrotoxicity, hepatotoxicity, and neurotoxicity), or body disfigurement. Future studies will have to address a number of questions and concerns: (1) Results will need to be validated in large controlled studies. (2) Optimal duration of MMF therapy and when and how to withdraw treatment in relapse-free patients will need to be determined. (3) Its efficacy in different racial groups and histologic subtypes will need to be studied. (4) The best way to monitor its efficacy and tailor its dose in individuals with varying circumstances, such as hypoalbuminemia, renal insufficiency, and use of concomitant drugs, needs to be determined. Clearly, the therapeutic potential of MMF needs more intensive investigation. It is not the only answer to treatment of childhood nephrotic syndrome but represents a new alternative with the potential to reduce side effects of existing agents that have been used for decades.