Individualizing Treatment of Steroid-Resistant Nephrotic Syndrome

The management of steroid-resistant nephrotic syndrome, which comprises about 10%–15% of nephrotic syndrome in children, is challenging. The condition is defined by the lack of complete remission of proteinuria despite 4–6 weeks of therapy with prednisone either at onset of disease (initial, primary) or following a subsequent relapse (late, secondary). Steroid resistance is believed to have a genetic or immune etiology. More than 80 monogenic disorders, involving structural and regulatory proteins of the glomerular filtration barrier, are associated with initial steroid resistance (1). Immune perturbations are postulated to result in generation of a “circulating factor” that increases the permeability of the glomerular filtration barrier, resulting in nephrotic-range proteinuria. All patients with steroid resistance should undergo a kidney biopsy to delineate the histology, usually FSGS or minimal change disease (1). Therapy comprises a combination of calcineurin inhibitors (CNIs), prednisone, and angiotensin-converting enzyme inhibitors (1). Concerns in management include nonresponse to intensified immunosuppression, progressive kidney disease, and 30% risk of post-transplant FSGS (1,2). Late resistance, which comprises 5%–10% of patients with nephrotic syndrome, is not associated with a genetic cause but has higher risk of allograft recurrence (2). Given the heterogeneity in pathogenesis, response to therapy, and outcomes, identification of the underlying cause is likely to translate into individualized therapies.

Multicenter collaborative studies, using next-generation sequencing panels, show a monogenic etiology of steroid resistance in 24%–30% of patients (3–5). Mutations in other genes may occasionally result in phenocopies, such as in Alport syndrome, Dent disease, Lowe syndrome, and cystinosis (6). The proportion of patients with a genetic etiology varies with ethnicity and indications for performing these investigations. Recent guidelines from the International Pediatric Nephrology Association emphasize that all patients with initial steroid-resistant nephrotic syndrome should undergo these studies (1). We prefer a focused approach, with studies limited to patients with onset of disease during infancy, family history of similar illness, syndromic features, or nonresponse to CNI or before transplant. Although international collaborative studies have helped define the genetic basis of the illness and provide genotype-phenotype relationships, they lack comprehensive data on clinical features and course (3,4,6,7).

In a study published in this issue of CJASN, Mason et al. (5) review data on 271 children with steroid-resistant nephrotic syndrome enrolled in the United Kingdom National Registry of Rare Kidney Diseases (RaDaR), aiming to stratify the disease by etiology and outcomes. Kidney biopsies and screening for variations in 37–70 genes, using a next-generation approach, were done in all and related where possible to response to immunosuppression, disease course, and recurrent FSGS. A monogenic cause was identified in 30% of patients; the remaining patients were labeled genetics negative, and they were stratified for initial or late resistance and response to intensified immunosuppression. Most patients with genetic cause showed absence of complete remission with immunosuppressive therapy and progressive kidney failure but had no risk of recurrent FSGS. However, one half of 152 genetics-negative patients who received intensive immunosuppression showed complete or partial remission. Patients with late resistance had negative genetic studies but better response to almost all forms of immunosuppressive therapy. Patients showing remission had excellent prognosis, with >90% kidney survival at 10 years.

These findings endorse prior reports that 5%–35% of patients with monogenic etiology compared with 41%–76% of patients with nongenetic etiology show complete or partial response to therapy with CNI (odds for nonresponse: odds ratio, 3.82; 95% confidence interval, 2.32 to 6.49) (4–6). The rates for kidney failure were 44%–62% and 23%–50%, respectively (odds ratio, 3.05; 95% confidence interval, 2.28 to 4.08) (4–6). Although few patients with WT1 mutations may show partial response to CNI (4,5), there is consensus that patients with confirmed genetic etiology should not receive immunosuppressive medications (1). Genetic screening might enable specific therapy in few patients, including those with defects in coenzyme Q10 pathway and certain phenocopies (Fabry disease and cystinosis), and allow counseling on outcomes (Alport syndrome, Denys–Drash syndrome, and Dent disease) (1,3,4,6).

Findings by Mason et al. (5) confirm that response to immunosuppressive therapy predicts kidney survival, with significantly higher risk of kidney failure in patients with persistent nephrotic-range proteinuria (4,5,7). Results from this study and other multicenter observation cohorts indicate that the rates of complete and partial remission with CNI therapy are 40%–50% (4,5,7). However, data from randomized, controlled trials estimate the efficacy of cyclosporin and tacrolimus at 62% (eight trials; n=156) and 83% (three trials; n=94), respectively (8). Although it is possible that effectiveness of therapy with CNI in real-world situations is indeed lower, there is risk that limited or single-point data collection in large databases might not be accurate. Prospective cohorts, with detailed and accurate phenotyping, are required to better ascertain these outcomes.

Rituximab is an important therapeutic strategy for steroid-resistant and CNI-refractory nephrotic syndrome (9). Although its mechanism of action is unclear, rituximab is reported to induce remission in 30%–50% of patients, with higher rates for minimal change disease (53%–64%) compared with FSGS (20%–42%) (9,10). In this report, rituximab successfully achieved remission in 51% of patients with nongenetic disease, with response occurring twice as often among late resistance (71%) compared with initial resistance (41%). A similar trend of better response in patients with late resistance has been reported previously (9). Nevertheless, these results are indeed superior to response rates of 19%–29% in the same paper from Necker Hospital, Paris (5); from our center (10); and from a randomized, controlled trial (8).

Multicenter registries have been increasingly used to provide real-life observational data on the natural history of rare diseases. However, they do have limitations due to selection and information bias, risk of attrition, and confounding. It is possible that patients with milder forms of steroid resistance lacking a genetic diagnosis were not included in the RaDaR database. Similarly, 6-month data entries for patients with steroid resistance might not reflect sustained remission or inform if they were elicited in the context of a prior steroid-sensitive illness. Although there is consensus that initial therapy in steroid resistance should comprise CNI, treatment in this report also included levamisole, cyclophosphamide, and rituximab, which are associated with different responses, affecting the significance and relevance of the findings. Similarly, there is limited information on whether rituximab was used in the context of steroid resistance, CNI dependence, or CNI resistance, all of which have different outcomes (10). An important previous observation, from the RaDaR and the Midwest Pediatric Nephrology Consortium (2), that patients with late resistance are at high risk of recurrent FSGS was not confirmed in this report due to the small number of patients with comprehensive genetic screening (5).

Mason et al. (5) thus provide important evidence that outcomes in patients with steroid-resistant nephrotic syndrome depend on the presence or absence of genetic variations and response of disease to subsequent treatment. This report includes 70% of nationwide incident children with the illness, making the findings generalizable to the United Kingdom. In the context of similar data from other observational cohorts (4,7), these findings seem valid for other geographic regions. Figure 1 depicts a consensus view on the approach to the management of patients with the illness. It is suggested that patients who fail to respond to CNI and rituximab might have undiscovered mutations. Post-transplant FSGS represents a model for circulating factor disease, the identity of which continues to be the “holy grail” in nephrotic syndrome research. Although CNIs occupy center stage for therapy, areas for research include its duration of therapy, optimal management of CNI-refractory disease, and prevention and therapy of recurrent FSGS.

Figure 1.

Management of initial and late steroid-resistant nephrotic syndrome. A monogenic etiology accounts for 25%–30% of patients with initial resistance. These patients do not respond to immunosuppression and require monitoring for progressive disease. Patients with nongenetic initial resistance and those with late steroid resistance who show remission of proteinuria with calcineurin inhibitors (CNIs) or rituximab have excellent long-term outcomes. Patients failing to respond to these therapies are at risk for kidney failure and recurrent FSGS. ACE, angiotensin-converting enzyme; Treg, T regulatory cells. *Strategies that have shown limited benefit in CNI-refractory disease include rituximab coadministration of CNI and mycophenolate mofetil (MMF); therapies targeted to CD80 (abatacept), TNF (adalimumab), and melanocortin-1 receptor (adrenocorticophin); galactose; and lipid apheresis.

Disclosures

All authors have nothing to disclose.

Funding

None.