Clinical characteristics of the study population

A total of 172 patients who met inclusion criteria were analyzed in this research. We summarized patients’ characteristics, clinical data, and kidney biopsy histological evaluations in the “Overall” column of Table 1. In general, the mean age of this population was 43.94 ± 12.98 years old. There were 105 male patients, making up 61.05% of the population. The mean arterial pressure (MAP) was 102.01 ± 12.82 mmHg, medium estimated glomerular filtration rate (eGFR) was 56.00 (47.50) mL/min·1.73 m2, and medium 24-h proteinuria excretion was 1.25 (1.76) g/24 h (Table 1). Other demographic characteristics and clinical information in detail are listed in Table 1.

Table 1 Clinical characteristics of IgA nephropathy patients by kidney interstitial inflammatory cell infiltration score

Patients were divided into two group by tubulointerstitial inflammatory cell infiltration score. Scoring 0 to 1 indicated none to mild inflammatory cell infiltration and were included in the G0 group. Scoring 2 to 3 indicated moderate to severe inflammatory cell infiltration and were included in the G1 group. Two groups presented with a similar male proportion, 60.00% and 62.67%, respectively. Also, they presented with similar age, 43.86 ± 12.98 years and 44.07 ± 13.09 years, respectively. As shown in Table 1, moderate-to-severe inflammatory cell infiltration group (group G1) manifested with significantly lower eGFR, serum albumin level and hemoglobin level and higher serum uric acid and 24-h proteinuria excretion than those in non-to-mild inflammatory cell infiltration group (G0). Mean arterial pressure (MAP) was higher in group G1 (104.16 ± 13.34 mmHg) than in group G0 (100.62 ± 12.32 mmHg) but did not reach statistical significance.

Lee classification presented significant difference between two groups. There were 19 patients graded Lee I-II in G0 group while there was only 1 patient in the G1 group. There were 57 patients graded Lee III while there were only 2 patients in Lee III in the G1 group. There were 29 patients graded Lee IV-V in the G0 group while there were 64 patients graded Lee IV-V in the G1 group. Mesangial proliferation (M1) showed no difference between two groups. There was more proportion of endothelial lesions (E1), segmental sclerosis (S1), tubular lesions (T1-T2), and crescents formation (C1-C2) in G1 group. Details are listed in Table 2.

Table 2 Lee and Oxford MEST-C classification of IgA nephropathy patients by kidney interstitial inflammatory cell infiltration score

We also explored the relationship between renal inflammatory cell infiltration and traditional inflammation biomarkers and cytokines (Table 3). The highly sensitive C reaction protein (hs-CRP) showed no difference between group G0 and G1. The neutrophilic granulocyte percentage (NEUT%) in group G1 was 62.47 ± 7.90%, which was significantly higher than it was 57.20 ± 7.25% in group G0. The lymphocyte percentage (LYMPH%) in group G1 was 26.88 ± 7.18%, which was significantly lower than it was 31.88 ± 6.83% in group G0. Therefore, the neutrophilic granulocyte-to-lymphocyte ratio (NLR) in group G1, which was 2.28 (1.27), was significantly higher than it, which was 1.80 (0.92), in group G0. The platelet count and platelet-to-lymphocyte ratio showed no difference between two groups. Among the six cytokines analyzed in the current study, TNF-α, sIL-2Rα, and IL-6 significantly elevated in group G1 than group G0. The logistic regression indicated that only sIL-2Rα showed significant correlation with tubulointerstitial inflammatory cell infiltration. Then we focused on the subsequent analysis of sIL-2Rα.

Table 3 The complete blood count test and serum cytokines levels of IgA nephropathy patients by kidney interstitial inflammatory cell infiltration scoreCorrelation between serum sIL-2Rα and clinical parameters in IgA nephropathy

We first explored the relationship between serum sIL-2Rα and eGFR. Levels of serum sIL-2Rα were negatively correlated with eGFR at the time of renal biopsy (p < 0.0001, r2 = 0.435; Fig. 1A). Proteinuria is one of the clinical features of IgA nephropathy (IgAN) and can predict disease prognosis. Serum sIL-2Rα was positively correlated with 24-h proteinuria excretion at the time of renal biopsy (p < 0.0001, r2 = 0.193; Fig. 2B). Patients were then divided into groups according to tertiles of serum sIL-2Rα levels (T1, serum sIL-2Rα ≤ 421 U/mL; T2, 422 ≤ serum sIL-2Rα ≤ 577 U/mL; T3, serum sIL-2Rα ≥ 581 U/mL). As shown in Table 6, hemoglobin and serum albumin levels were significantly lower in the higher tertile (T3) than those in the lower tertile (T1) and middle tertile (T2) (both p = 0.000; Table 6). Serum uric acid was higher in the T3 group than in the T1 or T2 groups (p = 0.000; Table 6). In summary, elevated serum sIL-2Rα was associated with severe clinical manifestations in IgAN.

Fig. 1

Serum sIL-2Rα was associated with eGFR and proteinuria in IgAN patients. A Correlation between serum sIL-2Rα level and eGFR; B correlation between serum sIL-2Rα level and ln [proteinuria]. The proteinuria indicated 24-h proteinuria excretion (g/24 h). Each dot represented a value from an individual patient. Coefficients of determination (r2) and p-values were shown

Fig. 2

Serum sIL-2Rα level reflected histological injury in IgAN patients. A–E Association between serum sIL-2Rα and Oxford classification. *p < 0.05; **p < 0.01; ****p < 0.0001. F Correlation between serum sIL-2Rα and global glomerulosclerosis. G Correlation between serum sIL-2Rα and segmental glomerulosclerosis. H Correlation between serum sIL-2Rα and tubulointerstitial score. I Correlation between sIL-2Rα and MEST-C score. Each dot represented a value from an individual patient. Coefficients determination (r2) and p value were shown. J Quantitative analysis of serum sIL-2R levels in IgAN patients with differing Lee’s pathological grades. ***:p < 0.001; ****p < 0.0001

Serum sIL-2Rα and histology

To confirm whether serum sIL-2Rα can reflect disease activity in IgAN, we studied IgAN biopsy histological data and explored the relationship between serum sIL-2Rα levels and IgAN histological characteristics. Mesangial proliferation is one of the basic pathological manifestations. Specimens scored M1 presented higher serum sIL-2Rα levels than M0 (p = 0.022; Fig. 2A). Similar situations were also observed in specimens scored E1 (p = 0.049; Fig. 2B), S1 (p = 0.001; Fig. 2C), and C1-2 (p = 0.002; Fig. 2E). Glomerulosclerosis in IgAN reflects glomerulus injury. Global glomerulosclerosis percentage (p < 0.0001, r2 = 0.311; Fig. 2F) and segmental glomerulosclerosis percentage (p = 0.0002, r2 = 0.079; Fig. 2G) were both positively correlated with serum sIL-2Rα levels. T score in Oxford MEST-C classification represents tubulointerstitial injury in IgAN. Specimens that classified as T1-2 manifested significantly higher serum sIL-2Rα levels than T0 (p < 0.0001; Fig. 2D). And serum sIL-2Rα was also positively correlated with tubulointerstitial score (p < 0.0001, r2 = 0.373; Fig. 2H). MEST-C score and Lee Grade comprehensively present the degree of disease activity. Specimens that graded with Lee IV-V manifested higher serum sIL-2Rα than Lee I and Lee II-III (p < 0.001, Fig. 2J). Serum sIL-2Rα was positively correlated with MEST-C score (p < 0.0001, r2 = 0.245; Fig. 2I).

Association between serum sIL-2Rα and tubulointerstitial inflammatory cell infiltration

Serum sIL-2Rα was secreted by activated T cells and reflected inflammatory status in vivo. We explored the relationship between serum sIL-2Rα and inflammatory cell infiltration in IgAN. Inflammatory cell infiltration can reflect the immune response to injuries. We used inflammatory cell infiltration score to evaluate cell infiltration extent. In the present study, moderate-to-severe inflammatory cell infiltration represented with a remarkable higher serum sIL-2Rα level than none-to-mild group (p < 0.0001; Fig. 3A). Serum sIL-2Rα was positively correlated with the infiltration score (p < 0.0001, r2 = 0.313; Fig. 3B).

Fig. 3

Serum sIL-2Rα predict tubulointerstitial inflammatory cell infiltration. A Association between serum sIL-2Rα levels and infiltration score. 0–1 referred to non-mild tubulointerstitial inflammatory cell infiltration. 2–3 referred to moderate-severe tubulointerstitial inflammatory cell infiltration. ****p < 0.0001. B Correlation between serum sIL-2Rα levels and infiltration score in IgAN patients. Coefficient determination (r2) and p value were shown. C–E ROC curve for discrimination between patients with or without moderate-severe tubulointerstitial inflammatory cell infiltration at the time of renal biopsy according to different biomarkers: C hemoglobin, D sIL-2Rα, and E integrated curve of sIL-2Rα, hemoglobin, eGFR < 60 mL/min·1.73 m2, and 24-h proteinuria excretion > 1.0 g/24 h. AUCs were shown under each curve, respectively

We subsequently analyzed the serum sIL-2Rα’s predictive value on tubulointerstitial inflammatory cell infiltration in IgAN, because of its non-invasiveness. And inflammatory cell infiltration was recognized as important factors when considering individual treating strategy. As showed in Table 4, serum sIL-2Rα levels (stratified to per 100 U/mL increase) was still independent risk factors of medium-to-severe inflammatory cell infiltration after being adjusted for age, gender, MAP, BMI, eGFR < 60 mL/min·1.73 m2, 24-h proteinuria excretion > 1.0 g/24 h, hemoglobin levels, and ACEI/ARB administration at time of biopsy. Model 3 shows the odds of being moderate-to-severe inflammatory cell infiltration will increase 0.290 times when serum sIL-2Rα level increases every 100 U/mL (OR = 1.290, 95% CI: 1.015, 1.640, p = 0.038; Table 4). Detailed variables and analysis data of Table 4 were listed in Table S1 of supplementary materials.

Table 4 Univariable and multivariable OR of sIL-2Rα for interstitial inflammatory cell infiltration

In this logistic regression model, eGFR < 60 mL/min·1.73 m2, 24-h proteinuria excretion > 1.0 g/24 h, and hemoglobin levels were also strongly independent risk factors, therefore we calculated the area under curve (AUC) of the receiver operating characteristic (ROC) curve, respectively (Table 5). The curve combined serum sIL-2Rα levels, eGFR < 60 mL/min·1.73 m2, 24-h proteinuria excretion > 1.0 g/24 h, and hemoglobin levels as predictive factors had the highest AUC 0.859 (0.801, 0.918). The different ROC curve models containing different factors are summarized in Table 5.

Table 5 Receiver’s operation characteristics (ROC) curve for diagnosis of patients with or without moderate-severe inflammatory cell infiltration in IgANSix-month disease prognosis analysis

We studied the association between serum sIL-2Rα and IgAN disease progression at 6 months after renal biopsy. Patients were divided into three tertile groups according to serum sIL-2Rα levels. Clinical characters of three tertile groups are summarized in Table 6. Lee classification and MEST-C classification of three tertile groups are summarized in Table 7. Serum sIL-2Rα in the higher tertile was a strong independent predictor of kidney function loss at 6 months after renal biopsy after adjustment in model 3 (adjusted gender, age, MAP, BMI, eGFR < 60 mL/min·1.73 m2, 24-h proteinuria excretion > 1.0 g/24 h and use of immunosuppressants during follow-up, Table 8). Patients with serum sIL-2Rα levels in the higher tertile (T3 sIL-2α ≥ 581 U/mL) indicated a 4.161-fold greater risk of kidney function loss 6 months after renal biopsy than their serum sIL-2Rα levels in the T1 and T2 groups.

Table 6 Characteristics of IgAN patients by serum sIL-2Rα tertile at time of biopsyTable 7 Lee and Oxford MEST-C classification of IgAN patients by serum sIL-2Rα tertile at time of biopsyTable 8 Univariable and multivariable OR of sIL-2Rα for predicting kidney function loss for 6 months after kidney biopsy

Continuous proteinuria excretion greater than 1 g/24 h is one of the strong predictors of kidney function loss in IgAN. Therefore, we further explored the association between serum sIL-2Rα levels and IgAN remission status at 6 months after renal biopsy. In model 1 (adjusted gender, age, MAP, and BMI), serum sIL-2Rα was an independent predictor of remission failure at 6 months after biopsy (Table 9). Every 100 U/mL increase in serum sIL-2Rα means a 1.448-fold greater risk of remission failure. When adjusted for either eGFR < 60 mL/min·1.73 m2 (model 2) or 24-h proteinuria excretion > 1.0 g/24 h (model 3), serum sIL-2Rα was still independent predictor of remission failure. However, this significance lost in model 4 (covariant in model 1 plus eGFR < 60 mL/min·1.73 m2 and 24-h proteinuria excretion > 1.0 g/24 h). This result indicated that deficient eGFR and proteinuria excretion were still strongest predictors for IgAN proteinuria remission failure, which were well proved in previous studies.

Table 9 Univariable and multivariable OR of sIL-2Rα for predicting proteinuria remission failure for 6 months after kidney biopsy

Detailed variables and analysis data of Tables 8 and 9 were listed in Tables S2 and S3 of supplementary materials.