A multicenter case-control study was conducted in patients with corticosteroid-resistant NS caused by idiopathic FSGS, diagnosed by renal biopsy. The study included adult and pediatric patients treated at 4 hospital centers between January 2012 and March 2019. The inclusion criteria were steroid resistance, histological diagnosis of FSGS by renal biopsy, absence of secondary etiologies according to a systematic study and protocol, absence of a family history of nephropathy, and negative results in the genetic study. In addition, only patients who had not yet received treatment with steroids, immunosuppressants, angiotensin II receptor blockers, or statins at the time of blood and urine sample collection were included. The exclusion criteria were secondary etiologies such as a reduction in renal mass; morbid obesity; HIV-associated nephropathy; heroin or cocaine use; parvovirus B19 infection; the use of analgesics, bisphosphonates, or interferon; vesicoureteral reflux; and obstructive sleep apnea. A genetic study was performed to rule out pathogenic mutations in all patients under 35 years of age with corticosteroid resistance using a previously described and validated renal panel (Bullich et al., 2018).

The design consisted of 2 phases: 1) a case‒control observational phase to analyze the response to CNIs and associated variables and 2) a prospective follow-up of the cohort of treated patients to evaluate the evolution of the variables of interest after treatment as a function of the outcome.

Idiopathic NS (INS) was defined by 24 h urinary protein excretion values >3.5 g/day in adults and >40 mg/m2/h in children, associated with hypoalbuminemia <3.5  g/dL and edema. After diagnosis, both clinical follow-up and treatment were carried out following the recommendations indicated in the Kidney Disease: Improving Global Outcomes (KDIGO)15 guidelines at the discretion of each participating center. To be included in the study, the criteria for corticosteroid resistance were explicitly defined and were the same for all the centers. In pediatric patients, corticosteroid resistance was defined as a lack of complete remission (reduction in the protein/creatinine ratio in urine to <0.2  mg/mg) after 8 weeks of therapy with prednisone at a standard dosage of 60 mg/m2/day. In adult patients, corticosteroid resistance was defined as the lack of remission (reduction in proteinuria to <0.3 g/day, stable serum creatinine levels, and serum albumin levels > 3.5 g/dL) after a minimum exposure of 16 weeks to prednisone at a daily dosage of 1 mg/kg or at a dosage every other day of 2 mg/kg. Patients with partial remission to corticosteroids were not included; only those patients who, according to the definition, did not achieve remission were considered resistant to corticosteroids, regardless of whether they presented a decrease in proteinuria without reaching the threshold established for partial remission. The type of CNI indicated (tacrolimus or cyclosporine) was freely selected by each of the participating centers. The guidelines for treatment in terms of dose, trough levels, and controls were the same in all centers and were also based on the recommendations of the KDIGO guidelines.15 A response to treatment with CNIs was considered if, after the start of treatment, total or partial remission of proteinuria was achieved, according to the criteria described in the KDIGO guidelines for pediatric and adult patients. Pediatric patients were considered those <18 years of age.

The control group included healthy subjects paired by age and sex selected from an electronic database containing the demographic data of 450 healthy blood donors and healthy individuals of pediatric age obtained from preoperative studies of dermatological processes or minor local surgeries. All participants in the study signed the corresponding informed consent form prior to inclusion. In the case of pediatric controls, blood samples were obtained after informed consent was obtained from their parents or guardians. The study was carried out in accordance with the parameters established by the Declaration of Helsinki and the local law on the protection of personal data (LOPD 15/1999). The study protocol was approved by the independent bioethics committees of the participating centers (internal code 2551).

All blood, serum, and DNA samples were obtained before any immunosuppressive or immunomodulatory treatment was started. Blood samples were obtained in standard tubes without additives, centrifuged at 1500 g for 10 min, and stored at −80 °C until use. Serum creatinine levels were measured by the compensated method traceable to isotope dilution mass spectrometry (IDMS) (Hitachi Modular P-800 Roche Diagnostics, Germany). The estimated glomerular filtration rate (eGFR) was calculated using the Schwartz equation in children and the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) formula in adults. Serum levels of CD25 were measured by enzyme-linked immunosorbent assay (ELISA) (Quantikine R&D Systems, Inc., MN, USA). All ELISA analyses were performed in the same laboratory using the same equipment and in duplicate. In addition, to analyze the reproducibility of the measurements, in a sample of 15 patients, 3 or more determinations were performed during the nephrotic flare-up phase before starting any treatment. In all cases, the coefficients of variation of the repeated measurements were <10%.

The analysis of lymphocyte activation was performed using multichannel flow cytometry (Cytoflex Beckman Coulter, USA) including the following: subpopulations of T lymphocytes (helper (CD4 DR+), cytotoxic (CD8 DR+) and regulatory T cells), soluble IL-2 receptor (CD25), and CD3 lymphocytes with CD25 expression. For this analysis, monoclonal antibodies conjugated with fluorophores specific for each marker (Abyntek Biopharma, Spain) were used.

All renal biopsies were analyzed centrally in a blinded manner by 2 experienced pathologists who confirmed the morphological pattern of FSGS on a case-by-case basis. Renal biopsies were stained with hematoxylin and eosin, periodic acid-Schiff-methenamine, and Masson’s trichrome for morphological analysis, and immunofluorescence studies were performed with antibodies against immunoglobulin A (IgA), IgG, IgM, C3, fibrinogen, and light chains. Samples were also processed for examination by electron microscopy.

Quantitative variables are expressed as the mean ± standard deviation, and qualitative variables are expressed as proportions. Comparisons of means between groups for independent data were performed using Student’s t-test or the Mann‒Whitney U test in the case of a nonnormal distribution. Comparisons of 2 intragroup means were performed using Student’s t-test for paired data or its equivalent and the Wilcoxon test for nonparametric variables. The association between quantitative variables was analyzed using the Pearson or Spearman correlation coefficient as a function of the distribution of the variable. Comparisons between categorical variables were performed using the chi-square test. To analyze the normality of the distribution of the data, the Kolmogorov‒Smirnov or Shapiro‒Wilk test was used. A p value was accepted as significant when <0.05. After the patients were classified according to their response to CNI treatment, univariate analysis was performed to analyze the variables associated with remission. The cutoff values for the quantitative variables associated with the response to treatment were calculated by receiver operating characteristic (ROC) curve analysis, and the Youden index was used to select the optimal value. Statistical calculations were performed using SPSS®, version 20.0.

A total of 39 patients with corticosteroid-resistant FSGS were included; 15 (38.5%) were pediatric patients (15 ± 2 years), and 24 (61.5%) were older than 18 years (38 ± 17 years). The control group included 23 healthy controls, of whom 26.1% were pediatric, with an average age of 17 ± 1 years, and 73.9% were adults, with an average age of 34 ± 12 years.

Table 1 summarizes the demographic, clinical, and biochemical characteristics of both groups. No statistically significant differences were observed in terms of age, sex, or the eGFR between patients with FSGS and healthy controls. However, compared with controls, patients with FSGS showed significantly higher levels of soluble CD25, expression of IL2 receptors in CD3 T lymphocytes (CD3-IL2+), and DR in CD4 and CD8 lymphocytes, as well as a higher CD4/CD8 ratio (p < 0.01).

Table 2 summarizes the correlation matrix between the variables related to the activation of T lymphocytes studied in the total group of patients with FSGS. The level of soluble CD25 was moderately associated with the expression of IL2 in CD3 lymphocytes and weakly associated with DR expression in CD4 and CD8 lymphocytes. A significant correlation was also observed between the total number of CD3 lymphocytes and the number of CD8 DR+ lymphocytes, and a strong negative correlation was detected between the number of CD4 DR+ lymphocytes and the number of CD8 DR+ lymphocytes. In the healthy control group, there were no significant associations between any of the variables studied (data not shown).

The minimum duration of treatment with calcineurin inhibitors was 6 months, following the recommendations of the clinical guidelines, with a mean of 11.7 ± 5.35 months (range: 6–24 months), without significant differences in the duration between cyclosporine A and tacrolimus (p = 0.92). However, when the response or resistance to CNI treatment was analyzed, statistically significant differences were observed: responding patients received more months of treatment, whereas in patients without any response, the treatment was interrupted relatively early to avoid side effects and due to a lack of efficacy (Table 3).

Of the 39 patients, 19 (48.7%) responded to CNI treatment; among them, 12 (63.2%) achieved a partial response, and 7 (36.8%) achieved a complete response. Compared with nonresponders, responders had significantly higher levels of soluble CD25, absolute numbers of CD3 lymphocytes, and expression of IL-2 in CD3 lymphocytes and lower expression of DR in CD8 lymphocytes (p <  0.05; Table 3).

When patients who responded to CNI treatment were analyzed by partial remission and complete remission (Table 4), patients with complete remission were significantly younger, whereas no significant differences were found in the other variables studied (Table 4).

The analysis of CD25 levels revealed statistically significant differences between responsive FSGS patients and controls (477 ± 84 vs. 262 ± 62 pg/mL, respectively; p < 0.01). In contrast, no significant differences were found between nonresponders and controls (290 ± 86 vs. 262 ± 62 pg/mL, respectively; p = 0.31).

The pediatric cohort included a total of 15 patients, 52.4% of whom were male. The mean age was 14.7 ±  2.1 years. When demographic, clinical, and biochemical characteristics were compared between the pediatric and adult populations, no significant differences were observed. Similarly, no differences were detected in the serum levels of soluble CD25, in the proportion of CD3 lymphocytes with IL-2 expression, or in the rest of the lymphocyte subpopulations analyzed.

Regarding the response to CNI treatment, 53.3% (n = 8) of the pediatric patients met the response criteria. Among these patients, 87.5% (n = 7) achieved total remission of NS, which represents a statistically significant difference compared with the adult population, where none of the patients achieved complete remission (p < 0.01).

A binary logistic regression model was constructed based on the statistically significant variables of the bivariate study, and only the CD25 variable demonstrated significance in the final model.

ROC curve analysis showed an AUC of 0.937 and revealed a cutoff of 324 pg/mL for soluble CD25, with a sensitivity of 94%, specificity of 75%, positive predictive value of 78%, and negative predictive value of 94% (Fig. 1).

Figure 1.

ROC curve of soluble CD25.

After an average follow-up of 13.7 ± 5.3 months, both groups (responders and nonresponders) showed a significant decrease in the eGFR and an increase in the serum creatinine level. The loss of renal function was greater in nonresponders (−6.97 [p25: −1.92; p75: −4.8]   ml/min/month) than in responders (−3.39 [p25: −0.3; p75: −2.2]   ml/min/month; p < 0.01).

In responsive patients, a statistically significant reduction in proteinuria and the cholesterol level and an increase in the serum albumin level were observed. In addition, these patients presented statistically significant decreases in the levels of soluble CD25, the number of CD3 and CD3-IL2+ lymphocytes, and the CD4/CD8 ratio. In contrast, they showed a significant increase in the number of CD8 DR+ and regulatory T cells. In patients who did not respond to treatment, a significant reduction in the level of soluble CD25, the level of CD3 IL-2+, and the level of CD8 DR+ and a significant reduction in the CD4/CD8 ratio were detected, but the other variables did not change (Table 5).

The reduction in the soluble CD25 level was significantly greater in patients who responded to treatment than in those who did not respond (mean reduction: −169.5 ± 99.2 vs. −20.15 ± 8.9 pg/mL, respectively, p < 0.01). When the reduction in the soluble CD25 level was analyzed according to partial or total remission, no statistically significant difference was detected.