Response to First Course of Intensified Immunosuppression in Genetically Stratified Steroid Resistant Nephrotic Syndrome

Visual Abstract

Abstract

Background and objectives

Intensified immunosuppression in steroid-resistant nephrotic syndrome is broadly applied, with disparate outcomes. This review of patients from the United Kingdom National Study of Nephrotic Syndrome cohort aimed to improve disease stratification by determining, in comprehensively genetically screened patients with steroid-resistant nephrotic syndrome, if there is an association between response to initial intensified immunosuppression and disease progression and/or post-transplant recurrence.

Design, setting, participants, & measurements

Pediatric patients with steroid-resistant nephrotic syndrome were recruited via the UK National Registry of Rare Kidney Diseases. All patients were whole-genome sequenced, whole-exome sequenced, or steroid-resistant nephrotic syndrome gene-panel sequenced. Complete response or partial response within 6 months of starting intensified immunosuppression was ascertained using laboratory data. Response to intensified immunosuppression and outcomes were analyzed according to genetic testing results, pattern of steroid resistance, and first biopsy findings.

Results

Of 271 patients, 178 (92 males, median onset age 4.7 years) received intensified immunosuppression with response available. A total of 4% of patients with monogenic disease showed complete response, compared with 25% of genetic-testing-negative patients (P=0.02). None of the former recurred post-transplantation. In genetic-testing-negative patients, 97% with complete response to first intensified immunosuppression did not progress, whereas 44% of nonresponders developed kidney failure with 73% recurrence post-transplant. Secondary steroid resistance had a higher complete response rate than primary/presumed resistance (43% versus 23%; P=0.001). The highest complete response rate in secondary steroid resistance was to rituximab (64%). Biopsy results showed no correlation with intensified immunosuppression response or outcome.

Conclusions

Patients with monogenic steroid-resistant nephrotic syndrome had a poor therapeutic response and no post-transplant recurrence. In genetic-testing-negative patients, there was an association between response to first intensified immunosuppression and long-term outcome. Patients with complete response rarely progressed to kidney failure, whereas nonresponders had poor kidney survival and a high post-transplant recurrence rate. Patients with secondary steroid resistance were more likely to respond, particularly to rituximab.

Introduction

Approximately 10%–15% of children with nephrotic syndrome are resistant to steroids (1,2), and most of these receive intensified (or second-line) immunosuppression. Response is often disappointing and there are significant side effects. Between 30% and 40% progress to kidney failure within 10 years, requiring dialysis and transplantation (3). Disease recurrence is common and associated with poor long-term outcome. Unfortunately, our ability to predict the disease course, treatment response, and risk of post-transplantation recurrence for individual patients is limited.

There are many proposed risk factors for post-transplantation recurrence, including age at diagnosis, rate of progression to kidney failure, biopsy result, ethnicity, and previous recurrence (4,5). However, the most informative factors remain secondary steroid resistance for increased risk (6) or a monogenic cause of disease as a protective feature. Mutations have been identified in >70 genes, causing podocyte defects, and are responsible for approximately 30% of childhood steroid-resistant nephrotic syndrome (7–21). Patients with genetic disease are usually resistant to immunosuppression and progress more rapidly to kidney failure, but do not recur after transplantation (7,22).

Post-transplantation recurrence is thought to be immune-mediated. It is hypothesized that a plasma circulating factor, derived from immune cell dysfunction, acts on the podocyte and disrupts glomerular permeability. However, its identity remains elusive (23–25). We have previously shown that secondary steroid resistance can be used as a marker for circulating factor disease and is associated with a high risk of post-transplantation recurrence (6). Strikingly, 93% of patients with secondary steroid resistance recurred post-transplantation compared with 30% with primary steroid resistance.

The latter study lacked detailed genetic analyses. We propose that comprehensively genetically stratifying patients with steroid-resistant nephrotic syndrome, then correlating response to intensified immunosuppression with progression to kidney failure and recurrence, will identify subgroups that are useful for clinical prognostication and management. Our results identify two distinct groups of genetic-testing-negative patients: one that responds to intensified immunosuppression and has a good long-term outcome, and one that is multidrug-resistant with rapid progression, very poor kidney survival, and high post-transplant recurrence risk.

Materials and Methods

Patient Cohort

Cases were taken from the UK National Registry of Rare Kidney Diseases (RaDaR), a Renal Association initiative set up in 2010 that collates clinical data from patients with rare kidney diseases (26). Data are collected both retrospectively and prospectively via an online portal and include demographics, family history, consanguinity, pattern of steroid resistance, medications, transplantation, and recurrence. Cases were selected in January 2018, at which point there were 2457 patients with idiopathic nephrotic syndrome enrolled (see Supplemental Appendix 1 for inclusion/exclusion criteria). Patients included in this analysis had steroid-resistant nephrotic syndrome with age of onset <18 years and were screened for disease-causing mutations. Steroid-resistant nephrotic syndrome is defined as failure to respond to 4 weeks of high-dose oral prednisolone. The cohort consisted of 271 patients (Figure 1) with a date of diagnosis ranging from 1995 to 2017. A total of 188 patients had whole-exome sequencing (7), nine had whole-genome sequencing, and 74 had clinical gene-panel testing. Further, 68 patients who had clinical gene-panel testing underwent testing at Bristol Genetics Laboratory with next-generation sequencing of 37 (27) or 70 genes associated with steroid-resistant nephrotic syndrome (28). The remaining six patients who had clinical gene-panel testing had testing in other locations, with results documented in RaDaR. Patients were considered to have monogenic disease if a mutation was found in one of the known “nephrotic” genes (7,27). Follow-up data were inputted to the RaDaR registry on at least a 6-monthly basis. Local clinical teams were contacted individually to provide specific items of missing data.

Figure 1.

Summary of patient selection. Steroid-resistant nephrotic syndrome includes patients with primary, presumed, and secondary steroid resistance. A total of 184 patients were part of the original whole-exome sequencing cohort, which has been previously described by Bierzynska et al. (7). RaDaR, National Registry of Rare Kidney Diseases.

Clinical Data Retrieval

Demographic, clinical, and long-term outcome data were extracted from the RaDaR database. Only medications started before kidney failure, and only the first course of each medication, were included (Supplemental Table 1). Complete response was defined as urine protein-to-creatinine ratio <200 mg/g, urine albumin-to-creatinine ratio <30 mg/g, or negative/trace dipstick proteinuria within 6 months of starting therapy. Partial response was defined as urine protein-to-creatinine ratio >200 mg/g or dipstick ≥1+ but plasma albumin >2.5 g/dl. If a medication was stopped within 6 months, only laboratory data obtained when the patient was receiving the medication were used. If two medications were started simultaneously or within 1 month, the same response outcome was assigned to both. For management of missing data, see Supplemental Appendix 1.

Data Analysis

Proportions of patients achieving complete and partial response were calculated for the whole cohort and stratified by genetic disease, pattern of steroid resistance, and biopsy results. Particular attention was given to genetic-testing-negative patients with post-transplant recurrence because they are most likely to have circulating factor disease. To minimize bias from the order in which clinicians chose to use medications, outcomes for the first intensified immunosuppression drug used per patient were analyzed separately.

Statistical Analyses

Data analysis was performed using GraphPad Prism 7 with Fisher exact test, chi-squared analysis, or Mann–Whitney U test.

Results

Patient Characteristics

Demographic features of the 271 patients and the treatments received are shown in Tables 1 and 2. In total, 186 patients (69%) received intensified immunosuppression. Completeness of response data were 91%. A total of 346 intensified immunosuppression treatments with responses available were given to 178 patients.

Table 1.

Demographic characteristics of the patient cohort

Characteristic		Total Cohort	Genetic-Testing Negative	Monogenic Disease
Total patients		271	190	81
Male, n (%)		137 (51)	98 (52)	39 (48)
Age at onset, yr, n (%)	0–0.25	36 (13)	8 (4)	28 (35)
	0.25–1	10 (4)	4 (2)	6 (7)
	1–5	129 (47)	107 (56)	22 (27)
	6–12	74 (27)	55 (29)	19 (24)
	13–18	22 (8)	16 (8)	6 (7)
Median age at onset, yr/ interquartile range		4.8/2.4–9.3	4.9/2.6–9.4	3.2/1.6–8.4
Family history positive/number with data available (%)		39/252 (15)a	16/180 (9)	23/72 (32)
Consanguinity/number with data available (%)		25/245 (10)	9/175 (5)b	16/70 (23)
Ethnicity (% of patients where data available)	White	186 (72)	130 (70)	56 (76)
	Asian	18 (7)	15 (8)	3 (4)
	Pakistani	17 (7)	10 (5)	7 (10)
	Black African/Caribbean	13 (5)	10 (5)	3 (4)
	Mixed	8 (3)	8 (4)	0 (0)
	Indian	10 (4)	7 (4)	3 (4)
	Bangladeshi	2 (1)	2 (1)	0 (0)
	Other	6 (2)	4 (2)	2 (3)
	No ethnicity data available	11	4	7
First biopsy findings (% of patients where data available)	FSGS	128 (55)	103 (60)	25 (40)
	Minimal change disease	56 (24)	45 (26)	11 (18)
	Mesangial hypercellularity	16 (7)	9 (5)	7 (11)
	Finnish type	6 (3)	0 (0)	6 (10)
	Kidney failure	5 (2)	0 (0)	5 (8)
	Diffuse mesangial sclerosis	4 (2)	2 (1)	2 (3)
	Focal global glomerulosclerosis	2 (1)	1 (1)	1 (2)
	Other	17 (7)	11 (6)	6 (10)
	No biopsy data available/not biopsied	37	19	18
Pattern of steroid resistance (%)	Presumed	54 (20)	15 (8)	39 (48)
	Primary	179 (66)	138 (73)	41 (51)
	Secondary	38 (14)	37 (19)	1 (1)

Percentages are calculated for column totals.

^aPatients were deemed to have a positive family history if they had an affected first-degree relative, or an affected cousin in a consanguineous family. In the genetic-testing-negative group this includes seven siblings from three families. The monogenic disease group includes 13 siblings from six families.

^bOne patient from a consanguineous family had no mutations identified in known nephrotic genes but is under investigation for a novel gene candidate.

Table 2.

Number of treatments received and availability of response data

Groups	Subgroups	No. of Patients	No. of Treatments	No. of Patients with Response Data	No. of Treatments with Response Data
Total cohort		271	—	—	—
Not receiving ACEi/ARB or intensified immunosuppression (%)	Total	52 (19)	—	—	—
Reason for no ACEi/ARB or intensified immunosuppression	Congenital nephrotic syndrome	24	—	—	—
	CKD/kidney failure at presentation	13	—	—	—
	Syndromic	3	—	—	—
	Familial	1	—	—	—
	No medication data	11	—	—	—
Total receiving treatments (%)	ACEi/ARB or intensified immunosuppression	219/271 (81)	540	202/219 (92)	480/540 (89)
Grouped by patients (%)	ACEi/ARB only	33/219 (15)	—	—	—
	Intensified immunosuppression only	86/219 (39)	—	—	—
	ACEi/ARB and intensified immunosuppression	100/219 (46)	—	—	—
Grouped by treatments (%)	All ACEi/ARB	133/219 (61)	160/540 (30)	112/133 (84)	134/160 (84)
	All intensified immunosuppression	186/219 (85)	380/540 (70)	178/186 (96)	346/380 (91)

ACEi, angiotensin-converting enzyme inhibitor; ARB, angiotensin II receptor blocker.

Response and Outcomes in Patients with Monogenic Disease

Of the 271 patients, 81 (30%) had monogenic disease. Further, 26 (32%) of these were treated with intensified immunosuppression. Complete response to first intensified immunosuppression was seen in 4% of monogenic patients compared with 25% of genetic-testing-negative patients (P=0.02; no data for n=0/26 monogenic patients and n=3/152 genetic-testing-negative patients). There was a significant difference in combined complete and partial response between monogenic and genetic-testing-negative patients for all intensified immunosuppression treatments (35% versus 53%; P=0.04; no data for n=5/45 monogenic patient treatment episodes and n=29/335 genetic-testing-negative patient treatment episodes), but not for first intensified immunosuppression only (35% versus 46%; P=0.29). One monogenic patient had complete response and eight showed partial response to first-administered intensified immunosuppression (Table 3). None of the 21 monogenic patients (26%) who received a transplant had responded to intensified immunosuppression and none recurred post-transplantation. This is significantly different to transplanted genetic-testing-negative patients, who had a 68% recurrence rate (21 of 31) post-transplantation.

Table 3.

Patients with monogenic disease who responded to immunosuppression

Patient	Gene	Sex	Age at onset, yr	Resistance to Steroids	First Biopsy	CKD Stage	Extrarenal Phenotype	Length of Follow-Up, yr	Medication for which There Was Response	Response
7656	WT1	M	3	Primary	Diffuse mesangial sclerosis	2	Denys–Drash syndrome	1.6	MMF, tacrolimus	Complete, complete
495	NPHS1	F	2	Primary	Minimal change disease	2	No	17.8	MMF	Partial
514	SMARCAL1	M	7	Primary	FSGS	1	No	0.4	Cyclosporin	Partial
687	CRB2	F	0	Presumed	Minimal change disease	1	No	4.0	Levamisole	Partial
729	NPHS2	M	7	Primary	Other	2	Asthma	2.8	Tacrolimus	Partial
731	MAGI2	M	0	Primary	Minimal change disease	1	Pyloric stenosis, polydactyly, thrombocytosis	11.7	Cyclosporin	Partial
770	COL4A3	F	7	Primary	FSGS	1	No	2.5	Cyclosporin	Partial
811	WT1	F	3	Primary	Alport syndrome	1	Chronic cough and diarrhea, Frasier syndrome	1.1	Tacrolimus	Partial
900	LMX1B	F	14	Primary	FSGS	1	Delayed puberty	1.1	Tacrolimus	Partial

M, male; MMF, mycophenolate mofetil; F, female. For correlation of UK National Registry of Rare Kidney Diseases numbers with previously published identification numbers, see Supplemental Table 3.

Response to Immunosuppression in Genetic-Testing-Negative Patients

Of the 190 genetic-testing-negative patients, 152 received 306 intensified immunosuppression treatments with response data available (no data for n=29/335 treatment episodes). The average number of treatments was two per patient (range one to five). Overall, the complete response rate was 28%. Complete response was highest for rituximab (39%, 16 of 41; no data for n=1) and lowest for cyclophosphamide (19%, eight of 43; no data for n=5), but this was not statistically significant (P=0.05). The combined complete and partial response rate was highest for tacrolimus (59%, 48 of 82; no data for n=4), and this was significant compared with cyclophosphamide (40%; P=0.04; Figure 2A). Response data were available for first-administered intensified immunosuppression in 149 patients (no data for n=3 patients). Cyclosporin was the first treatment in 66 patients (44%), tacrolimus in 35 patients (23%), and cyclophosphamide in 32 patients (21%). Cyclosporin and tacrolimus had similar levels of complete response (27% and 31%, respectively; Figure 2B). This was higher than cyclophosphamide (13%), but not statistically significant (P=0.13 versus cyclosporin; P=0.08 versus tacrolimus).

Figure 2.

Response to intensified immunosuppression, and associated outcomes. immunosuppression medications and kidney survival in genetic-testing-negative patients. Response to (A) all or (B) first-administered intensified immunosuppression medications. The number of treatments with response data available is given in parentheses. No data were available for 29 of 335 treatment episodes (first intensified immunosuppression in three patients). (C) Kidney survival analyzed by response to first intensified immunosuppression treatment. Numbers in the table represent the number of patients at risk for each time point.

Association between Response to First Intensified Immunosuppression and Likelihood of Kidney Failure in Genetic-Testing-Negative Patients

Characteristics and long-term outcomes for genetic-testing-negative patients stratified by response to first intensified immunosuppression treatment are shown in Table 4. The median follow-up time was 5.2 years (range 0.1–22.2 years). Strikingly, 97% of patients (36 of 37) with complete response showed no progression to kidney failure (median follow-up 5.0 years, range 0.1–15.2 years) (Figure 3). Nonresponders had significantly quicker progression to kidney failure, as shown in the Kaplan–Meier survival curves (P<0.001; Mantel–Cox test) (Figure 2C). The 5-year kidney-failure-free survival rates were 97%, 87%, and 59% for patients with complete, partial, and no response, respectively. The corresponding 10-year rates were 97%, 74%, and 27%. Frequency of transplantation was significantly higher in nonresponders (P<0.001). In total, 31 genetic-testing-negative patients received a transplant; 26 (84%) were nonresponders to first intensified immunosuppression, four had partial response, and only one complete response. The overall recurrence rate was 68% (21 of 31), with 73% (19 of 26) recurrence in nonresponders. Strikingly, kidney survival for genetic-testing-negative patients with no response to first intensified immunosuppression is the same as for patients with monogenic disease (Figure 2C). This supersedes previous data where comparison between monogenic and (all) nonmonogenic steroid-resistant nephrotic syndrome shows worse kidney survival for monogenic patients (2,7).

Table 4.

Characteristics and long-term outcomes of genetic-testing-negative patients stratified by response to first intensified immunosuppression treatment

Characteristic		Total with Outcomes	Complete	Partial	No	P Value
No. of patients		149	37	32	80
First-line intensified immunosuppression treatment, n (% of column total)	Cyclosporin	66 (44)	18	17	31	0.53a
	Tacrolimus	35 (23)	11	5	19
	Mycophenolate mofetil	6 (4)	2	0	4
	Cyclophosphamide	32 (21)	4	8	20
	Rituximab	5 (3)	1	1	3
	Levamisole	4 (3)	1	1	2
	Azathioprine	1 (1)	0	0	1
Age at onset in years, n (% of column total)	0–0.25	0 (0)	0 (0)	0 (0)	0 (0)	0.03b
	0.25–1	1 (1)	0 (0)	0 (0)	1 (1)
	1–5	89 (60)	26 (70)	24 (75)	39 (49)
	6–12	47 (32)	11 (30)	5 (16)	31 (39)
	13–18	12 (8)	0 (0)	3 (9)	9 (11)
Pattern of steroid resistance, n (% of column total)	Presumed steroid resistance	1 (1)	0 (0)	0 (0)	1 (1)	0.15c
	Primary steroid resistance	117 (79)	26 (70)	24 (75)	67 (84)
	Secondary steroid resistance	31 (21)	11 (30)	8 (25)	12 (15)
Number (%) who developed kidney failure		41 (28)	1 (3)	5 (16)	35 (44)	<0.001d
Number (%) transplanted		31 (21)	1 (3)	4 (13)	26 (33)	<0.001d
Number (% of those transplanted) with post-transplant recurrence		21/31 (68)	1/1 (100)	1/4 (25)	19/26 (73)	0.30e

Treatment response data were unavailable for the first intensified immunosuppression treatment in three genetic-testing-negative patients (all three received cyclosporin). All data were complete for age, steroid resistance, kidney failure, transplant, and post-transplant recurrence. Percentages (shown in parentheses) are calculated for column totals. P values are for the comparison between complete, partial, and no response.

^aChi-squared analysis, 8 degrees of freedom (df). Azathioprine and levamisole excluded from analysis.

^bChi-squared analysis, 4 degrees of freedom (df). “0–0.25,” “0.25–1,” and “1–5” combined into one group.

^cChi-squared analysis, 2 degrees of freedom (df). “Presumed” and “Primary steroid resistance” combined into one group.

^dChi-squared analysis, 2 degrees of freedom (df).

^eFisher exact test. Complete and partial response combined into one group.

Figure 3.

Patient stratification by genetic testing and response to intensified immunosuppression. The three stratified patient groups are indicated in the gray highlighted boxes.

Response Stratified by Pattern of Steroid Resistance

Genetic-testing-negative patients with primary steroid resistance (120 patients) were analyzed as a distinct subgroup from secondary steroid resistance (32 patients). Secondary steroid-resistant patients had significantly higher complete response than those with genetic-testing-negative primary steroid resistance (43% versus 23%; P=0.001; no data for n=29/334 treatment episodes). The combined complete and partial response rate was also significantly higher (65% versus 48%; P=0.01). The highest complete response rate was to rituximab with 64% (nine of 14, no data for n=1) in secondary steroid-resistant patients compared with 26% (seven of 27) in genetic-testing-negative primary steroid resistance (P=0.02). Secondary steroid-resistant patients also had a significantly higher complete response rate to cyclophosphamide (38% versus 10%; P=0.04; no data for n=3/16 secondary steroid-resistant patients and n=2/32 primary steroid-resistant patients). When considering only first intensified immunosuppression, there was no significant difference in response between the two groups (complete response 36% versus 22%, P=0.16; combined complete and partial response 61% versus 42%, P=0.07; no data for n=3/152). A total of 22% of genetic-testing-negative primary steroid-resistant patients received a transplant, with 69% (18 of 26) recurrence rate; 16% of secondary steroid-resistant patients were transplanted, with 60% (three of five) recurrence rate.

Characteristics and Outcomes in Patients Treated with Rituximab

Complete response to intensified immunosuppression was generally <30%. The exception was rituximab in patients with secondary steroid resistance (64%), so these patients have been analyzed in more detail (Figure 4). In total, 46 patients received rituximab. Treatment was performed according to center-based decisions. Six were simultaneously treated with other intensified immunosuppression (one mycophenolate mofetil, one cyclosporin, three tacrolimus, one with both tacrolimus and mycophenolate mofetil). These six patients all had primary steroid resistance and showed a varied response (one complete, three partial, and two nonresponders). Four monogenic patients received rituximab, and none responded. Forty two genetic-testing-negative patients were given rituximab (no data for n=1). Complete response was 39% (16 of 41) and partial response was 12% (five of 41); this was not statistically significant compared with other medications. No complete responders progressed to kidney failure (zero of 16, median follow-up 7.6 years), whereas 50% (ten of 20) of nonresponders developed kidney failure (median follow-up 5.4 years), with an 83% (five of six) post-transplant recurrence rate. There was no significant difference between responders and nonresponders in terms of median age of diagnosis or time between diagnosis and treatment. Fifteen secondary steroid-resistant patients received rituximab (no data for n=1). Complete response occurred in 64% (nine of 14) and none of these developed kidney failure (median follow-up 7.2 years). Of the four nonresponders, two developed kidney failure and both recurred after transplantation. Twenty seven genetic-testing-negative primary steroid-resistant patients received rituximab. Complete response occurred in 26% (significantly lower than in secondary steroid-resistant patients, P=0.02) and no response occurred in 59%. Further, 50% of nonresponders developed kidney failure and three out of the four patients who were transplanted suffered recurrence.

Figure 4.

Outcome of patients according to Rituximab response.

To validate these findings, we examined the comprehensive database of Necker Hospital, Paris, for outcomes of pediatric patients with steroid-resistant nephrotic syndrome who were treated with rituximab (29). In 82 genetic-testing-negative, primary steroid-resistant patients with no response to first intensified immunosuppression treatment, 60 (73%) showed no response to rituximab. Further, 29 of these 60 (48%) patients developed kidney failure and eight of the 17 (47%) patients who were transplanted suffered disease recurrence. None of the 22 patients with complete (10) or partial response (12) progressed to kidney failure. No patients with secondary steroid resistance were treated with rituximab.

Response Stratified by First Biopsy Findings

A total of 170 patients receiving intensified immunosuppression had biopsy results available (not biopsied/no data for n=16). FSGS was seen in 100 (59%) first biopsies, minimal change disease in 45 first biopsies (26%), and mesangial hypercellularity in 11 first biopsies (6%). Other findings are detailed in Supplemental Table 2. There was no significant difference in response to intensified immunosuppression or clinical outcome on the basis of biopsy findings. When considering only FSGS and minimal change disease, there was no significant difference in complete response (24% versus 27%; P=0.66) or in combined complete and partial response (49% versus 51%; P=0.70; no data for n=27/300 treatment episodes).

Discussion

Our ability to predict long-term outcome and risk of post-transplantation recurrence in pediatric steroid-resistant nephrotic syndrome is limited. There is now emerging literature regarding the differences between responders and nonresponders to intensified immunosuppression (1,2). However, these studies lack comprehensive genetic stratification and their clinical message has not yet been widely appreciated. We present data from a large, national cohort of pediatric patients with steroid-resistant nephrotic syndrome, with an emphasis on full genetic screening. This allows us to reinforce and validate the findings of recent literature regarding response to intensified immunosuppression, and adds substantial new information regarding clinical outcomes, individual therapies, biopsy findings, and post-transplantation recurrence.

We focused our stratification on complete versus no response, as partial response is more susceptible to natural variation, incomplete recording, and confounding by hemodynamic factors (e.g., angiotensin-converting enzyme inhibition). Nevertheless, it is interesting that most partial responders had good outcomes, suggesting that they are in the same or similar immune mechanistic category as complete responders. We defined complete response as occurring within 6 months of starting treatment. To a certain extent, this is arbitrary and may underestimate later responders. One German study of 231 patients with steroid-resistant nephrotic syndrome reported 60% complete response to cyclosporin A in genetic-testing-negative patients, with 18% of those achieving complete response doing so beyond 6 months (30). However, the response rate in this cohort remains high even when late responders are accounted for; approximately 49% showed complete response within 6 months compared with 28% in our study. Our findings are very similar to that shown by the largest steroid-resistant nephrotic syndrome cohort to date, where complete response to calcineurin inhibitors was 29.9%, and fits within the wider literature (2,31–33). We believe the 6-month period will capture the majority of responders as well as minimize the potential for confounding.

Our results confirm that monogenic disease is a distinct subgroup that responds very poorly to immunosuppression. There is no convincing evidence of complete response to intensified immunosuppression in monogenic disease, albeit several reported cases (22,30,34). Here, one monogenic patient (out of 26) demonstrated complete response. This patient has a WT1 mutation, with Denys–Drash syndrome. Case reports suggest response to immunosuppression, usually cyclosporin A, in WT1-associated nephrotic syndrome (35–37). Therefore, in this specific mutation, we cannot rule out a direct effect of intensified immunosuppression on podocytes. None of the 21 monogenic patients who were transplanted suffered post-transplantation recurrence, compared with 68% (21 of 31) of genetic-testing-negative patients. This is consistent with previous literature reporting that monogenic patients do not generally recur after transplantation. Patients were considered to have monogenic disease only if a mutation was found in a known “nephrotic” gene, regardless of family history. Of the genetic-testing-negative patients who received intensified immunosuppression, seven had an affected first-degree relative. One had presumed steroid resistance, family consanguinity (no family history), and a variant of unknown significance in WT1 (7). It is possible that some of these patients may have monogenic disease attributable to currently undiscovered gene mutations.

The genetic-testing-negative subgroup is more heterogenous, making it harder to predict treatment response and outcome. There was no significant difference in response or clinical outcome on the basis of first biopsy finding, suggesting this is a poor discriminator. However, there was a clear association between response to first intensified immunosuppression and long-term outcome: responders (complete and partial) rarely developed kidney failure, whereas nonresponders had a high likelihood of kidney failure and post-transplantation recurrence.

These results are generally consistent with those published from the PodoNet registry, where 27% of nongenetic patients responded completely to intensified immunosuppression, although only 43% of the cohort had gene-panel sequencing, and correlation between intensified immunosuppression and post-transplant recurrence was not performed (1,2). Our study only included patients with comprehensive genetic screening, thereby reflecting current clinical practice. A total of 97% of patients with complete response did not develop kidney failure and the 10-year kidney-failure-free survival rate was 97%. The one patient with complete response who developed kidney failure was steroid-sensitive at initial treatment (secondary steroid resistance), and treated with cyclosporin. They suffered post-transplant recurrence. In contrast, 44% of nonresponders to first intensified immunosuppression progressed to kidney failure (10-year kidney-failure-free survival 28%) and the post-transplantation recurrence rate was 73%. One advantage of thorough stratification is that we show the time to kidney failure in the nonresponders is as rapid as the monogenic group (Figure 2C). This is in contrast to the previous, less-stratified studies by us and Trautmann et al., where monogenic patients had more rapid progression than nonmonogenic patients.(2,7).

We have previously shown that secondary steroid resistance is associated with post-transplantation recurrence. In this study, there was no difference in recurrence between secondary steroid resistance and genetic-testing-negative primary steroid resistance, but the numbers are too small to be conclusive. The most striking difference was the response to rituximab. A total of 64% of secondary steroid-resistant patients showed complete response, compared with 26% (12% Paris cohort) in genetic-testing-negative primary steroid resistance (P=0.02). No patients with complete response to rituximab progressed to kidney failure (zero of 16, United Kingdom cohort; zero of ten, Paris cohort). The number of secondary steroid-resistant patients treated with rituximab is currently small, but it raises an interesting new observation that should be explored further as the numbers in these cohorts increase. The clinical implication in our view is that all genetic-testing-negative patients resistant to their first intensified immunosuppression should be treated with rituximab, as a proportion (26% of genetic-testing-negative primary steroid resistance, 64% of secondary steroid resistance) will respond and not progress. This is also consistent with the 11 currently published observational studies and one randomized, controlled trial for multidrug-resistant steroid-resistant nephrotic syndrome, where in total, approximately 30% of almost 200 treated patients achieved complete remission (38–49).

Some medications in secondary steroid-resistant patients may have commenced when they were still steroid-sensitive or already in remission. Our response criteria do not distinguish these patients, and this could contribute to the high response rate to rituximab in the secondary steroid resistance group. Of the 14 secondary steroid-resistant patients receiving rituximab (with response data available), five were steroid-resistant when starting treatment, four were steroid-sensitive, and five were unknown. Complete response occurred in all steroid-sensitive patients and three of the five steroid-resistant patients. None of the steroid-sensitive patients progressed to kidney failure (median follow-up 8.2 years).

In summary, we have identified three subgroups of steroid-resistant nephrotic syndrome, which likely represent mechanistically different disease (Figure 3). First, patients with monogenic disease show poor or no response to immunosuppression and generally do not recur after transplantation. Second, genetic-testing-negative patients who respond to first intensified immunosuppression rarely progress to kidney failure. Third, those who fail first intensified immunosuppression usually become multidrug-resistant, with high likelihood of rapid progression to kidney failure and 73% risk of post-transplantation recurrence. These patients mostly have immune-mediated circulating factor disease, given the very high rate of post-transplant recurrence. A smaller proportion will be monogenic, with currently undiscovered gene mutations. Response to rituximab identifies a subset of patients whose disease may be B cell–mediated, although the drug could also have a direct action on podocytes (50). Response is high in secondary steroid-resistant patients and associated with good long-term outcome.

We propose that stratification according to genetic testing, steroid response, and response to early immunosuppression can be valuable in guiding treatment and transplantation decisions. Our data continue to support comprehensive genetic testing in all steroid-resistant children. In those who test negative, immunosuppression including calcineurin inhibitors with or without steroids and other medications should be used with a trial of rituximab in nonresponders. Failure indicates the patient is likely to be multidrug-resistant, and the chance of progression to kidney failure and post-transplant recurrence becomes high.

Disclosures

Prof. O. Boyer has received consulting fees from Novartis and lecture fees from Octapharma. Prof. M. Saleem sits on Pfizer and Retrophin advisory boards and receives grant funding from Evotec AG and Union Chimique Belge. All remaining authors have nothing to disclose.

Funding

This study was supported by the Elizabeth Blackwell Institute for Health Research at the University of Bristol, Medical Research Council’s Global Challenges Research Fund, Kidney Research UK, the Medical Research Council’s Stratified Medicine Initiative, the National Institute for Health Research’s Rare Diseases Translational Research Collaboration, and Nephrotic Syndrome Trust, and Wellcome Trust’s Institutional Strategic Support Fund.

Supplemental Material

This article contains the following supplemental material online at http://cjasn.asnjournals.org/lookup/suppl/doi:10.2215/CJN.13371019/-/DCSupplemental.

Supplemental Appendix 1. Detailed methods, methodology for the Paris cohort, and inclusion and exclusion criteria for the RaDaR cohort.

Supplemental Table 1. Terms used for free-text medication search.

Supplemental Table 2. “Other” findings on kidney biopsy.

Supplemental Table 3. Correlation of RaDaR numbers with previously published identification numbers.