Article Text

Download PDFPDF

Steroid-sensitive nephrotic syndrome: an evidence-based update of immunosuppressive treatment in children
  1. Nicholas Larkins1,
  2. Siah Kim1,2,
  3. Jonathan Craig1,3,
  4. Elisabeth Hodson1,3
  1. 1Centre for Kidney Research, The Children's Hospital at Westmead, Westmead, New South Wales, Australia
  2. 2Discipline of Paediatrics and Child Health, Sydney Medical School, University of Sydney, Sydney, New South Wales, Australia
  3. 3Sydney School of Public Health, University of Sydney, Sydney, New South Wales, Australia
  1. Correspondence to Dr Nicholas Larkins, Centre for Kidney Research, The Children's Hospital at Westmead, Westmead, NSW 2145, Australia; nlar7569{at}


Nephrotic syndrome is one of the most common paediatric glomerular diseases, with an incidence of around two per 100 000 children per year. Corticosteroids are the mainstay of treatment, with 85%–90% of children going into remission with an 8-week course of treatment. Unfortunately, nephrotic syndrome follows a relapsing and remitting course in the majority, with 90% relapsing at least once. About half will progress to frequently relapsing nephrotic syndrome (FRNS) or steroid-dependent nephrotic syndrome (SDNS). Different initial steroid regimens have been evaluated since the first trials in Europe and America in the 1960s. Most trials have been designed to evaluate the optimal duration of the initial therapy, rather than different cumulative doses of corticosteroid, or the management of relapses. Until recently, these data suggested that an initial treatment duration of up to 6 months reduced the number of children developing a relapse, without evidence of increased steroid toxicity. Recently, three large, well-designed randomised control trials were published, which demonstrated no significant reduction in risk of relapse or of developing FRNS by extended treatment compared with 2 or 3 months. While there are few trial data to guide the treatment of individual relapses in steroid-sensitive nephrotic syndrome (SSNS), there is some evidence that a short course of corticosteroid therapy during upper respiratory tract infection may prevent relapse. In patients with FRNS or SDNS who continue to relapse despite low-dose alternate-day steroids a number of non-corticosteroid, steroid-sparing immunosuppressive agents (cyclophosphamide, ciclosporin, tacrolimus, mycophenolate mofetil, levamisole, rituximab) have been shown to reduce the risk of relapse and of FRNS. However, there are limited head-to-head data to inform which agent should be preferred. In this article, we review recent data from randomised trials to update paediatricians on the current evidence supporting interventions in SSNS.

  • nephrotic syndrome
  • child
  • immunosuppression
  • steroids

Statistics from

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.


Nephrotic syndrome is characterised by a triad of heavy proteinuria (urine protein/creatinine ratio ≥200 mg/mmol or ≥3+ proteinuria on urine dipstick), hypoalbuminaemia (<25 g/L) and oedema. It is one of the most common paediatric glomerular diseases, with an incidence of 1–2 per 100 000 children, and a prevalence of 16 per 100 000 children.1–4 There is clear geographical variation, with a higher incidence reported in South Asian populations (3–9 per 100 000 children).2 ,5 Prior to treatment with corticosteroids there was a high mortality rate from associated sepsis and thrombosis.6 However, the capacity to induce remission with corticosteroids means that the majority of children now outgrow the condition without serious long-term sequelae.7 ,8

This review aims to update general paediatricians on recent developments in the immunosuppressive treatment of steroid-sensitive nephrotic syndrome (SSNS) in children, with a focus on trials published within the last 3 years. The pathogenesis and general management of nephrotic syndrome is beyond the scope of this review and is covered elsewhere.9 ,10

Clinical classification and course

Since most children with nephrotic syndrome do not undergo a kidney biopsy their clinical classification is based upon their response to steroids (table 1).11–15 This approach has been widely used since the International Study for Kidney Disease in Children (ISKDC) trials in the 1960s demonstrated that 93% of children with SSNS have minimal change disease, which carries an excellent long-term prognosis.16 ,17 Although 80%–90% of patients are initially steroid sensitive, a similar proportion will suffer one or more relapses.16 ,17 At some point in their disease course around 50% of children meet the diagnostic criteria for frequently relapsing nephrotic syndrome (FRNS) or steroid- dependent nephrotic syndrome (SDNS).17 ,18 Relapses occur most commonly during periods of increased immune activity, such as during an upper respiratory tract infection (URTI).

Table 1

Clinical characterisation of children with nephrotic syndrome

Immunosuppressive therapy

Treatment of the initial presentation of nephrotic syndrome

The mainstay of immunosuppressive therapy is corticosteroid therapy, normally given as oral prednisolone or prednisone. The initial ISKDC trial used a dose of 60 mg/m2 of prednisone daily for 4 weeks (maximum 80 mg), followed by a dose of 40 mg/m2 on three consecutive days out of seven for 4 weeks (maximum 60 mg).16 Subsequently, the Arbeitsgemeinschaft für Pädiatrische Nephrologie demonstrated that alternate-day prednisone was more effective in maintaining remission than prednisone given on three consecutive days out of seven, so alternate-day therapy is now preferred.19 The 2-month prednisone regimen was associated with a high risk of relapse, so several groups undertook small randomised controlled trials (RCT) to determine if longer courses of prednisone in the initial episode of SSNS were associated with a lower frequency of relapse and of the development of FRNS.

A Cochrane systematic review of RCTs first published in 2000 and updated in 2003, 2005 and 2007 demonstrated that in children with SSNS, increasing the duration of corticosteroid therapy to 3 months or more significantly reduced the risk of relapse by 12–24 months compared with 2 months of therapy (6 RCTs, 422 children: relative risk (RR) 0.70, 95% CI 0.58 to 0.84) and the likelihood of developing FRNS (5 RCTs, 336 children: RR 0.55, 95% CI 0.37 to 0.83).7 ,20 Similarly, the risk of relapse (4 RCTs, 382 children: RR 0.57, 95% CI 0.45 to 0.71) and the number with FRNS (3 RCTs, 350 children: RR 0.55, 95% CI 0.39 to 0.80) were significantly reduced with 6 months of prednisone compared with 3 months. Based on these studies the current KDIGO (Kidney Disease: Improving Global Outcomes) guidelines, published in 2012, recommended at least 3 months of therapy and suggested a benefit of extending therapy for 6 months.14 However, there remained uncertainty about the efficacy of increased durations of steroid therapy, given the small number of patients enrolled and the risk of bias in some studies due to problems in trial design, conduct and reporting.

Recently, the conclusions of the Cochrane systematic review have been challenged by the results of three well-designed RCTs. In 2013, Teeninga et al21 published a randomised placebo-controlled trial, which compared 3 months with 6 months of prednisolone treatment in 150 children presenting with nephrotic syndrome where both groups received similar cumulative doses of prednisolone (3360 mg/m2). The mean follow-up time was 47 months, during which there was no significant difference in the proportion of patients who developed FRNS (3 months vs 6 months: 53% vs 59%, HR 1.08; 95% CI 0.65 to 1.80). There was no difference in the rate of adverse events including severe infections, ophthalmological abnormalities, cushingoid appearance or hypertension (blood pressure ≥95th centile).

Following this, two further groups have compared treatment regimens that included differing lengths of treatment that mandated different cumulative corticosteroid doses.22 In a placebo-controlled trial, Sinha et al23 compared a 3-month regimen of prednisolone (2 mg/kg per day for 6 weeks followed by 1.5 mg/kg on alternate days for 6 weeks; cumulative dose 2792±287 mg/m2) with a 6-month regimen (same initial treatment followed by a gradual steroid taper; cumulative dose 3530±399 mg/m2). At 12 months follow-up, the median number of relapses was 1.27 (IQR: 0–2) and 1.54 (IQR: 0–3) per person-year in the 3- and 6-month groups, respectively, with an incident rate ratio of 0.82 (95% CI 0.64 to 1.05, p=0.13). There was no significant difference in the numbers of children with relapse, frequent relapses or use of steroid-sparing agents between the two groups at 12 and 24 months follow-up. There were no significant differences in the incidence of adverse events between the two groups.

Yoshikawa et al24 compared 2 months of prednisolone (cumulative dose 2240 mg/m2) with 6 months (cumulative dose 3885 mg/m2) in a multicentre, open-label RCT across 90 hospitals in Japan involving 255 patients. There was no significant difference in the time to development of FRNS (primary outcome) (HR 0.86; 90% CI 0.64 to 1.16). There was also no difference in the time to first relapse (HR 0.97; 95% CI 0.82 to 1.31, p=0.86) or number of relapses per person-year (rate ratio 0.94; 95% CI 0.71 to 1.12, p=0.65). The adverse event rate was similar between the two groups.

How can these more recent trials be interpreted in the context of older trials which demonstrated a longer course was more effective? The answer lies in the design and reporting features of RCTs and how these may bias the observed treatment effects. There is very strong and consistent empiric evidence which shows that trials at high or unclear risk of bias for allocation concealment or blinding may overestimate the treatment effect by as much as 40%.25 The recent trials, which ensured adequate allocation concealment and so at low risk of bias, showed no benefit in extending the duration of prednisolone treatment beyond 2 or 3 months in the first episode of SSNS. Earlier trials were at high risk of bias due to inadequate allocation concealment and lack of blinding, which may lead to selection and performance bias, respectively. These new RCTs have now been included in an update of the Cochrane systematic review.26 Subgroup analyses according to risk of bias items showed that, in studies at low risk of bias for allocation concealment or blinding, there was no significant difference in the risk for FRNS. However, in studies at high or unclear risk of bias for allocation concealment or blinding, the risk for FRNS was significantly reduced in children treated with longer durations of prednisolone.27 Further high-quality data are anticipated from the PREDNOS trial due for completion this year (ISRCTN16645249).

Short-term corticosteroid use to prevent relapse with intercurrent illness

Another area of recent investigation has been the short-term use of corticosteroids to prevent relapse during episodes of URTI. The results of several trials suggest a reduction in the rate of relapse in patients with FRNS and SDNS when prednisolone was administered daily for 5–7 days at the onset of a URTI, compared with continuing alternate-day prednisolone or no prednisolone (table 2).28–31 However, small population sizes, variability in study designs and potentially high risk of bias means that there is residual uncertainty around the results of these trials. A large (300 children), placebo-controlled, multicentre trial is underway in the UK, which should clarify the magnitude of benefit and risks associated with this approach (PREDNOS 2, ISRCTN10900733).32

Table 2

Corticosteroid therapy during intercurrent illness

Low-dose alternate-day steroids

Low-dose alternate-day steroids are often used to maintain remission in patients with FRNS and SDNS. If remission can be maintained with a sufficiently low dose to avoid side effects then the use of potentially more toxic steroid-sparing agents may be avoided. While this regimen is commonly used, and is recommended by KDIGO as first-line treatment to reduce the rate of relapse in FRNS and SDNS, there are no randomised data to support its use.14

Steroid-sparing agents

Children with FRNS or SDNS may be prescribed non-corticosteroid immunosuppressive agents when there is a failure to maintain remission with low-dose alternate-day prednisolone and/or significant adverse effects of prednisolone develop. Cyclophosphamide remains the most commonly used steroid-sparing agent, in part due to the possibility of achieving a prolonged period of disease-free remission following therapy.33 The calcineurin inhibitors (CNIs) are similarly effective at achieving remission in FRNS and SDNS.34 However, the use of alkylating agents and CNIs is limited by marrow and gonadal toxicity and by nephrotoxicity, respectively.34 In addition, most patients treated with CNIs will relapse on withdrawal of therapy.35

Recent research has focused on alternative steroid-sparing agents with more acceptable side-effect profiles, such as the CNI, tacrolimus and mycophenolate mofetil (MMF). To date ciclosporin has been the CNI of choice in SSNS (starting dose 5 mg/kg in two divided doses), although tacrolimus (starting dose 0.1 mg/kg in two divided doses) is often used, despite being considerably more expensive than ciclosporin. While there are no trials in children with SSNS, data from a prospective uncontrolled study suggests that the two drugs have a similar efficacy.36 In this 2-year study, 12 of 14 patients in the ciclosporin group relapsed fewer than three times per year, as compared with 17 of 22 patients in the tacrolimus group (p=0.59 in year 1 and 0.75 in year 2). In addition, an RCT of children with steroid-resistant disease found no difference in the rate of remission after 1 year in 41 children randomised to either CNI (HR 1.14; 95% CI 0.84 to 1.55).37 While using tacrolimus avoids the gingival hyperplasia and hypertrichosis seen with ciclosporin, nephrotoxicity remains the most important adverse effect.37 Another adverse effect of tacrolimus is diabetes mellitus, but this has been limited to date to tacrolimus use post kidney transplant.

In recent years MMF (starting dose 1200 mg/m2 in two divided doses) has been used as a steroid-sparing agent in SSNS due to a favourable side-effect profile, though it may be less effective than CNI. In a randomised multicentre open-label crossover trial, comparing MMF and ciclosporin therapy in 60 children with FRNS over a 2-year period, MMF was associated with significantly more relapses per patient in the first year (mean 1.10 vs 0.24, p=0.03).38 This effect was not seen when a single outlier was excluded from the analysis, or in the second year of the trial (potentially due to carry-over effect from year 1 minimising the difference between groups). There was no significant difference in the rate of adverse events. The mean glomerular filtration rate was lower in patients receiving ciclosporin; however, it is unclear how much of this was due to reversible afferent arteriolar vasoconstriction or to chronic CNI toxicity. These results are similar to those of a small (24 children) parallel group trial comparing MMF and ciclosporin, where the relapse rate in the two groups was not significantly different but power was limited (0.83 relapses/year vs 0.08 relapses/year, p=0.08).39 MMF is teratogenic so contraceptive advice is required in adolescent females.40

Rituximab is a monoclonal antibody directed towards the CD-20 antigen on preplasma B cells. Initially developed as a chemotherapeutic agent for B-cell lymphomas, it is now also used in the treatment of autoimmune disorders including SSNS.41 Rituximab has been used in children with CNI and prednisolone-dependent SSNS, and may facilitate the withdrawal of other immunosuppressive agents.42–46 Various dosing regimens have been used but one common strategy is 375 mg/m2 weekly for 1–4 doses until CD-19 depletion is achieved.42 ,44 ,46 In 2014, Iijima et al47 published a trial confirming the benefit of rituximab over placebo in 48 children with FRNS or SDNS. The primary outcome of median relapse-free period was significantly longer in the rituximab (267 days; 95% CI 223 to 374) than in the placebo group (101 days; 95% CI 70 to 155).47 However, all children had relapsed by 19 months indicating that rituximab simply extends the period of remission, rather than changing the ultimate disease course. In this trial, there were no significant increases in adverse events. Nevertheless, there are concerns about side effects that are rare but serious, including opportunistic infections (eg, Pneumocystis jiroveci pneumonia and viral myocarditis), pulmonary fibrosis48–50 and progressive multifocal leukoencephalopathy seen in patients with other diseases treated with rituximab.51 ,52 This means that its use is generally reserved for children who continue to have frequent relapses despite prednisolone, MMF and CNI and/or have serious adverse effects of prednisolone and CNI.41

Other steroid-sparing agents include levamisole and mizoribine.34 ,47 The antihelminth agent levamisole has a favourable side-effect profile, but its use in many countries is limited by availability.53 Several small trials and a meta-analysis suggest a steroid-sparing effect in FRNS and SDNS. This should be clarified by the results of an international multicentre placebo-controlled trial due to be released later this year.34 ,54 Mizoribine is an imidazole nucleotide that inhibits purine synthesis in a manner similar to MMF.55 In a trial of 197 children with FRNS there was no significant reduction in the relapse rate compared with placebo, which has lead KDIGO to recommend that mizoribine not be used in SSNS.14 ,55


The optimal duration of corticosteroid therapy in treating the initial episode of SSNS is becoming clearer, following the publication of recent trials that demonstrate no benefit of extending prednisolone treatment beyond 2–3 months. There remain few data to guide the corticosteroid treatment of individual relapses. In patients who develop FRNS or SDNS, the current 2012 KDIGO guidelines suggest low-dose alternate-day corticosteroids, although there is little evidence to support this recommendation. Low-dose daily or alternate-day prednisolone administered for 7 days at the onset of infection may reduce the risk of relapse.

The number of steroid-sparing drugs available for use in patients with FRNS and SDNS continues to increase, although the number of head-to-head RCTs that might guide the specific choice of agent is limited. Cyclophosphamide remains a reasonable first alternative, as treatment is used for a defined period and the child may achieve a prolonged remission off treatment. However, the benefits and harms of this medication compared with other steroid-sparing agents need to be explained to patients and their families. Ultimately, the initial choice of a steroid-sparing agent will vary according to parent and physician preference. Currently available medications prolong periods of remission but they do not influence the long-term outcome of SSNS.


View Abstract


  • Contributors All authors contributed to writing the paper and vouch for its accuracy.

  • Competing interests None declared.

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