Idiopathic nephrotic syndrome (INS) is the most common glomerular disorder of childhood, with an incidence of 2 per 100,000 child population in the UK. The disease presents at a median age of 2 to 3 years and is twice as common in boys than in girls.^1,2 There is ethnic variability in the disease incidence, with a fourfold to sixfold higher incidence in the UK South Asian population.^1,3,4

The onset of INS is characterised by the acute onset of heavy proteinuria, resulting in the development of hypoalbuminaemia and generalised oedema. There is not infrequently a delay in diagnosis, with the child having being treated for allergy prior to eventual presentation to a paediatrician or paediatric nephrologist as an emergency. The disease pathogenesis is poorly understood; however, both in vitro and in vivo experiments have identified the immune system to be dysregulated at the time of disease onset.⁵ The presence of nephrotic syndrome places the child at increased risk of a number of complications, including thromboembolic disease and infection, particularly with Streptococcus pneumoniae. Prior to the development of adequate antibiotic and remission-inducing therapy, the mortality rate from INS was of the order of 50%, the majority of deaths being related to infection.^6,7

In excess of 90% of children who present with INS will respond to a course of high-dose corticosteroid therapy.⁸ For this reason, the majority are treated empirically with a course of corticosteroids without a renal biopsy being performed. Those who respond to the treatment are given the diagnostic label of having steroid-sensitive nephrotic syndrome (SSNS). Only those with atypical features at presentation (age < 12 months or over 12 years, persistent hypertension or impaired renal function, gross haematuria, low plasma C3, hepatitis B or C virus positivity) and those who do not respond to this initial course of corticosteroid therapy undergo renal biopsy.⁹ This is in contrast to practice in adult patients with nephrotic syndrome, in whom the causes of nephrotic syndrome are diverse and biopsy at presentation is routinely performed to establish a histological diagnosis and to guide subsequent therapy. Little emphasis is placed upon histological diagnosis in children with SSNS, as it has been shown that corticosteroid sensitivity rather than histology is the key prognostic indicator.¹⁰ Those children who respond to corticosteroids generally have a good long-term prognosis with a low risk of developing chronic kidney disease; in contrast, those who are corticosteroid unresponsive suffer significant morbidity, and around 50% will progress to end-stage kidney failure, necessitating dialysis and kidney transplantation over a 15-year period.¹¹ The majority of these children will have focal segmental glomerulosclerosis on renal biopsy. In an early seminal study, conducted by the International Study of Kidney Disease in Children (ISKDC),² a large cohort of children underwent renal biopsy at presentation prior to the commencement of corticosteroid therapy. The majority of those who responded to corticosteroids were noted to have minimal change disease (MCD) histology, so called because the appearance of the kidney tissue at light microscopic level is essentially normal. Somewhat confusingly, in much of the published literature, the terms MCD and SSNS are used interchangeably, although this is not strictly correct, as a small number of children with MCD do not respond to corticosteroids and, similarly, a small number of corticosteroid-sensitive children have a histological diagnosis other than MCD.

Following initial successful treatment with corticosteroids, around 80% of children with SSNS develop disease relapses necessitating further courses of high-dose prednisolone, and around 50% develop frequently relapsing nephrotic syndrome (FRNS), which is defined as two or more relapses within the first 6 months following presentation or four relapses within any 12-month period, or steroid-dependent nephrotic syndrome (SDNS), which is defined as relapses occurring within 14 days of discontinuation of corticosteroid therapy.¹² Similar to the presenting episode, nephrotic syndrome relapses are associated with a risk of significant complications, including sepsis, thrombosis, dyslipidaemia and malnutrition.⁷ The treatment of relapses with repeated courses of high-dose prednisolone is associated with major adverse effects, including hip avascular necrosis, growth failure, hypertension, obesity, diabetes and behavioural problems.^13,14 Furthermore, children frequently have to miss school during relapses, resulting in impaired academic performance and parental absence from work.

When complications of repeated courses of corticosteroids develop, or when they are expected, a range of immunosuppressive strategies are employed in an attempt to reduce the frequency of disease relapses. These include the use of long-term, low-dose, alternate-day prednisolone, as well as a range of non-corticosteroid immunosuppressive agents, including levamisole, cyclophosphamide, ciclosporin, tacrolimus, mycophenolate mofetil and rituximab.

Early follow-up studies^8,15 suggested that the long-term prognosis for children with SSNS was excellent, with all retaining normal kidney function and over 90% achieving long-term remission, with complete cessation of relapses by the end of puberty. However, subsequent studies have reported higher rates of relapsing disease persisting beyond childhood, with 19% of UK patients suffering ongoing relapses into early adult life.¹⁶ In the majority of patients who achieve permanent remission during childhood, relapse occurs at 13–16 years of age.¹⁷ However, by this time, the majority of affected children will have received a significant cumulative corticosteroid dose, and many will have been exposed to other immunosuppressive agents. The role of the paediatric nephrologist is to maintain the child with SSNS as being well and free from relapses, while at the same time minimising the adverse effects of exposure to corticosteroids and other immunosuppressive therapies, thus ensuring that they emerge as healthy adults free from relapses and with no significant long-term treatment-related morbidity.

The ideal initial corticosteroid regimen for use at presentation of childhood INS should rapidly induce urinary remission (defined as 3 consecutive days of zero or trace proteinuria) with resolution of oedema. It must be sufficient to prevent frequent relapses necessitating the use of alternative immunosuppressive agents, although not so intensive that serious corticosteroid-related AEs develop. The first standardised corticosteroid treatment regimen was introduced by the ISKDC in the 1960s and consisted of a dose of 60 mg/m² of prednisone (maximum 80 mg) given daily for 4 weeks followed by 40 mg/m² (maximum 60 mg) on 3 consecutive days out of 7 for a total of 4 weeks.¹⁸ Many centres made a minor modification whereby 40 mg/m² was administered on alternate days rather than on 3 days out of 7 during the second 4-week period. Centres in the UK adopted the use of prednisolone rather than prednisone, as this was and remains the corticosteroid in routine use in UK paediatric practice, although children in the USA and other parts of Europe have continued to receive prednisone. These two agents are very closely related, with prednisone being metabolised to the active prednisolone following absorption.

Following the introduction of the ISKDC regimen, there has been significant debate regarding the optimal prednisolone regimen at the time of presentation of SSNS, and a number of randomised controlled trials (RCTs)^19–28 have investigated whether giving a more or less intensive corticosteroid regimen at first presentation of INS affects the number of children suffering both disease relapses and adverse effects of corticosteroid therapy. One single RCT¹⁹ has shown that, compared with the ISKDC regimen, a less intensive regimen comprising a daily dose of 60 mg/m² of prednisone only until the urine was negative for 3 consecutive days (urinary remission, median time 14 days) followed by 4 weeks of 40 mg/m² of prednisone on alternate days resulted in a higher rate of disease relapse. No further studies have investigated this therapeutic strategy. By 2005, when the PREDnisolone in NephrOtic Syndrome (PREDNOS) study was being planned, a total of six RCTs^20–25 had compared 2 months of prednisolone using the ISKDC regimen with a variety of different regimens of ≥ 3 months in duration. These regimens intensified the initial prednisolone regimen by increasing the duration of both the daily and alternate daily prednisolone phases. A Cochrane review²⁹ of these six studies concluded that intensification of the initial corticosteroid therapy at disease presentation significantly reduced the rate of relapse at 12–24 months [risk ratio 0.7, 95% confidence interval (CI) 0.58 to 0.84]. There was an inverse linear relationship between treatment duration and risk of relapse [relative risk (RR) 1.26–0.112 duration; p = 0.03]. Furthermore, there was a significant reduction in the number of frequent relapsers and the mean relapse rate per participant per year. In addition to these six studies comparing the ISKDC regimen with longer duration corticosteroid regimens, a further four studies^23,26–28 compared 3 months’ treatment with prednisolone with 6 months’ treatment; two were published only in abstract form.^27,28 Longer therapy duration resulted in a significantly decreased risk of relapse at 6 and 12 months (risk ratio 0.48, 95% CI 0.35 to 0.64, and risk ratio 0.57, 95% CI 0.45 to 0.71, respectively). Furthermore, the number of participants who developed FRNS was also lower in the 6-month group than in the 12-month group (risk ratio 0.55, 95% CI 0.38 to 0.80). Further analysis suggested that the benefits of more intensive corticosteroid therapy were more likely to be related to the increased duration of treatment than the higher cumulative dose; however, collinearity between treatment duration and dose prevented the Cochrane group²⁹ from drawing definitive conclusions.

However, significant concerns were raised on a number of issues relating to the six studies contributing to the meta-analysis comparing 2 months of treatment with treatment for ≥ 3 months. The total number of participants was small, at only 520 participants across all six studies, and concerns were expressed about the quality of a number of these trials. One was (and remains) unpublished,²² with data available only in abstract form. None was placebo controlled or blinded in any way, and only two were at low risk of bias for allocation concealment. Trials with inadequate allocation concealment can exaggerate the efficacy of the experimental treatment by 30–40%, and meta-analysis of low-quality trials may overestimate the benefit of therapy.^30,31 Furthermore, only one of these trials was analysed on an intention-to-treat (ITT) basis; however, this same study indicated, in the discussion, that parents could exert some influence on which treatment group their child was allocated to, implying that the randomisation process was flawed.²³ It was also unclear whether or not there was a clinically useful reduction in the incidence of steroid-dependent disease and the use of second-line immunosuppressive agents. The studies also reported somewhat different corticosteroid-related adverse events, making interpretation of the impact that increased duration of corticosteroid therapy had on adverse effect profile difficult. Therefore, the authors of the Cochrane review²⁹ (Dr Elisabeth Hodson and Professor Jonathan Craig) concluded that further well-designed and adequately powered RCTs were required to establish the optimum dose and duration of treatment, and were consulted from an early stage regarding the design of the PREDNOS study.

There has continued to be great debate regarding what the ideal corticosteroid regimen at disease presentation should be, and there is considerable variation in the treatment regimens used. Kidney Disease Improving Global Outcomes guidelines³² published in 2013 supported the conclusions of the Cochrane review,²⁹ recommending that INS in children be treated initially with 60 mg/m² or 2 mg/kg of prednisone or prednisolone for at least 12 weeks (4–6 weeks daily followed by 40 mg/m² or 1.5 mg/kg every other day), followed by a slow tapering of dose over the next 2–5 months. Despite these recommendations, in the UK the majority of centres have continued to use the 8-week ISKDC regimen, as is the case in Canada, Nigeria and South Korea. In contrast, Germany, the Netherlands, Spain and other European nations have adopted a longer treatment regimen, as proposed by the Arbeitsgemainschaft für Pädiatrische Nephrologie (APN, the German Society for Paediatric Nephrology) and investigated in its RCT.²⁰ This consists of 6 weeks of daily prednisolone at a dose of 60 mg/m² followed by 6 weeks of alternate daily prednisolone at 40 mg/m². In France, a longer 18-week course of prednisone is in routine use.³³ A questionnaire survey reported significant heterogeneity in the regimens used in centres in the USA, with 13% using the ISKDC regimen, 7% using the APN regimen and many using either of these with a subsequent corticosteroid taper.³⁴ This genuine clinical equipoise confirmed the importance of conducting a high-quality RCT to determine whether or not extending the course of prednisolone beyond that recommended by the ISKDC was associated with improved clinical outcomes in UK children. We chose time to first relapse as our primary outcome measure and, following consultation with the British Association for Paediatric Nephrology and our patient group advisers from the Nephrotic Syndrome Trust (NeST) and the Renal Patient Support Group, we selected secondary outcome measures that were felt to be of clinical importance, including the incidence of FRNS and SDNS and the need for alternative, potentially more potent, immunosuppressive therapies. Given the paucity of high-quality information on the adverse effect profiles of standard course (SC) and extended course (EC) treatment courses of prednisolone, we also aimed to collect comprehensive adverse effect data, including abnormal behaviour, which was assessed through the use of the Achenbach Child Behaviour Checklist (ACBC). Abnormal behaviour is one of the most commonly reported adverse events (AEs) in routine clinical practice; however, it was rarely reported on in previously conducted clinical trials. Finally, we aimed to perform a detailed cost-effectiveness analysis to determine the relative cost and efficacy of the two regimens in quality-adjusted life-year (QALY) terms.

Since the commencement of the PREDNOS study, three further studies have reported their findings.^33,35,36 A well-conducted double-blind placebo-controlled RCT³³ performed in the Netherlands aimed to ascertain whether the apparently better outcomes associated with prolonged prednisolone treatment occurred as a result of the increased duration of treatment or the higher cumulative dose of prednisolone administered. One hundred and fifty Dutch children were randomised to receive 3 months of prednisolone followed by 3 months of placebo or 6 months of prednisolone; both groups received equal cumulative doses of prednisolone (3360 mg/m², the same dose as that administered in the APN regimen) and were followed up for a median of 47 months. One hundred and twenty-six children commenced trial medication. A primary end point of the development of FRNS was selected and no difference was detected (45% with 3 months of prednisolone vs. 50% with 6 months of prednisolone). There was no difference in the number of participants who developed relapses (77% vs. 80%), the number requiring alternative immunosuppressive agents or the number of AEs. The authors concluded that the reduced relapse rate associated with longer prednisolone regimens observed in previous studies most likely occurred as a result of the increased cumulative prednisolone dose administered rather than the lengthening of the duration of the treatment course.

More recently, two high-quality studies,^35,36 published alongside one another in Kidney International with an accompanying editorial by Hoyer,³⁷ reported outcomes that differed significantly from those reported in the Cochrane review.¹³ Sinha et al.³⁵ from New Delhi and four other Indian centres enrolled 181 children aged 1–12 years presenting for the first time with INS. Participants were treated with a dose of prednisone 2 mg/kg daily for 6 weeks followed by 1.5 mg/kg on alternate days for a further 6 weeks, and were then randomised in a double-blind manner to receive either placebo or prednisone in decreasing doses for a further 3 months. The total dose of prednisone received was 3530 mg in the 6-month group and 2792 mg in the 3-month group. There was no difference between the two groups in the chosen primary end point [the number of relapses per 12 months of follow-up (1.26 vs. 1.54, respectively; p = 0.21)] or the percentage of participants with relapses or frequent relapses. There was no significant difference in the mean time to first relapse. The authors concluded that extending initial prednisolone treatment from 3 to 6 months did not influence the course of illness in children with SSNS. The second study³⁶ randomised 255 Japanese children presenting with INS to either the ISKDC regimen (total dose of 2240 mg/m²) or a 6-month prednisolone regimen comprising 4 weeks of daily prednisolone followed by 20 weeks of tapering alternate-day prednisolone (total dose of 3885 mg/m²). Median follow-up was 36.7 months in the 2-month group and 38.2 months in the 6-month group. The chosen primary end point was the time to development of FRNS and was similar in both groups [hazard ratio (HR) 0.86, 90% CI 0.64 to 1.16]. The time to first relapse was also similar in both groups, as was the number of relapses per year. The frequency and severity of AEs were similar in both groups, despite the 6-month group receiving a significantly higher median cumulative dose of prednisolone over 2 years. Yoshikawa et al.³⁶ concluded that prolongation of the initial corticosteroid regimen from 2 to 6 months did not improve patient outcomes.

Following the publication of these studies, in 2015 the Cochrane group performed an update of their systematic review and meta-analysis.¹³ They reported that the addition of these three well-designed studies had changed the conclusion of their review. They noted that studies of long versus shorter duration of corticosteroid treatment had heterogeneous treatment effects, with the older studies that were rated as having a higher risk of bias tending to overestimate the effect of longer-course therapy compared with more recently published studies rated as having a low risk of bias. Among the studies rated as having a low risk of bias, the group found that there was no significant difference in the risk of FRNS between those given prednisolone for 2 or 3 months and those receiving treatment for longer durations or a higher total dose, indicating that there is no benefit of increasing the duration of prednisolone beyond 2 or 3 months in the initial episode of SSNS.¹³