Introduction

The Oxford classification is the first universal histological classification system developed by the International IgAN Network and the International Renal Pathology Society [1-3]. This classification system has driven the progression of histological IgA nephropathy (IgAN) evaluation by appraising the following five lesions: mesangial hypercellularity (M), endocapillary hypercellularity (E), segmental sclerosis (S), tubular atrophy/interstitial fibrosis (T), and crescents (C). These pathological lesions were selected due to their reproducibility among pathologists, their correlation with clinical outcomes, and the correlations between them. The features of intrarenal arterial and/or arteriolar sclerosis (AAS), such as arterial intimal thickening (AIT) in the interlobular artery and/or arteriolar hyaline (AH) in the arterioles, were not included in the Oxford classification because AH showed poor reproducibility (intraclass correlation coefficients < 0.4) and the presence of the most severe level of AAS was not statistically associated with a decreased survival from end-stage renal disease (ESRD) or a 50% drop in the glomerular filtration rate (GFR) [1, 2]. A number of previously reported studies have evaluated the Oxford classification [4] and, as a result, it has become popular among nephrologists. Moreover, the Oxford classification revealed pathological features and associations between therapies, such as immunosuppressive treatments and renin–angiotensin system (RAS) inhibitors (RASI). No interactions were detected between RASI and any of the five selected pathological features but the relationship between RASI and AAS has not been demonstrated clearly [1].

However, it is known that AAS are associated with older age [5], dyslipidaemia [6], hypertension [5, 7, 8], hyperuricaemia [6, 7], higher levels of proteinuria [5-7], poor renal function [5, 7], and other severe renal chronic lesions, such as glomerular sclerosis and tubulointerstitial damage [7]. Moreover, several recent studies have reported a relationship between AAS and renal survival [5], interstitial blood flow and hypoxia [9], nocturnal blood pressure [10], and C3 deposition, which result in unfavourable outcomes [11]. These results suggest that AAS may be an independent risk factor for IgAN. Additionally, it is known that AAS induces ischaemic glomerulopathy, increases the activity of RAS, and causes glomerular hypertension and hyperfiltration [12]. A previous report has indicated that RASI are effective in IgAN [13]; however, whether AAS are risk factors for progression in IgAN, and the interactions between RASI and AAS in IgAN, remain unclear in the current literature. Thus, this study aimed to evaluate the relationship between AAS, renal prognosis, and other clinical, laboratory, and histological factors, and to analyse whether RASI is an effective therapy for IgAN patients with AAS.

Materials and methods Study population and design

In this retrospective cohort study, 1,147 patients with IgAN diagnosed by renal biopsy at the Tokyo Women's Medical University between 1974 and 2015 were included. The exclusion criteria were: renal biopsy specimen containing fewer than eight glomeruli; renal biopsy specimen that was not analysed for AAS; the presence of other systemic diseases such as systemic lupus erythematosus, liver cirrhosis, and IgA vasculitis; and a history of prior observation for less than 1 year after undergoing renal biopsy (unless ESRD occurred within 1 year). To analyse the relationship between AAS and renal prognosis, AAS was evaluated in 678 IgAN patients. The patients were classified into two groups: those with AAS (AAS1: n = 338) and those without AAS (AAS0: n = 340). The clinical, laboratory, and histological findings, along with the treatments and renal prognosis, were analysed and compared between the groups. Additionally, risk factors associated with progression were analysed using univariate and multivariate Cox regression analyses. IgAN patients classified as AAS1 underwent evaluation of the impact of RASI initiated during the follow-up period. Accordingly, appropriate adjustments were made for comparing the clinical and histological findings and treatment between the RASI (+) groups and the RASI (−) groups (controls). Renal prognosis was also evaluated (Figure 1).

Algorithm of the study's design.

This study was conducted in accordance with the guidelines of the Declaration of Helsinki and was approved by the Medical Ethics Committee of the Tokyo Women's Medical University (reference #5104-R). Written informed consent to obtain a renal biopsy was obtained from all the participants. All patients were able to access information about this study (aim, plan, study participants, and duration) via the institutional webpage, and they were provided with the opportunity to ask questions and discuss the details and results of this study with us.

Diagnosis of IgAN and histological evaluation of the renal biopsy specimens

IgAN was diagnosed according to the presence of mesangial proliferative changes by light microscopy, mesangial IgA deposition by immunofluorescence, and/or mesangial electron-dense deposits by electron microscopy. Histological findings were graded according to the Oxford classification [1-3]. AAS was recognised as the presence of AIT and/or AH. To reduce sampling errors, AH was considered as the presence of any level of hyalinosis in the arterioles and AIT was considered as the presence of any level of intimal thickness, although they were scored as absent, mild, moderate, and severe based on the Oxford classification. All histological findings were obtained from pathology reports, some of which were repeatedly reviewed by the pathologist for confirmation of the findings.

Clinical and laboratory data

Each patient's sex, age, body mass index (BMI), systolic blood pressure (SBP), diastolic blood pressure (DBP), and mean arterial pressure (MAP) were recorded. Laboratory data, which included the estimated GFR (eGFR), uric acid (UA), total cholesterol (T-Cho), triglyceride (TG), urinary protein excretion (U-Prot), and urinary red blood cells (U-RBC), were evaluated at the time of renal biopsy as baseline data. The eGFR was calculated using the Isotope Dilution Mass Spectrometry adjusted Modification of Diet in Renal Disease (IDMS-MDRD) study equation for Japanese individuals (eGFR = 194 × S-Cr−1.094 × age−0.287 × 0.739 [if female]) [14]. The time taken to progress to ESRD, which was defined as requiring dialysis or renal transplantation, was considered the endpoint. Risk factors associated with the progression to ESRD were also evaluated.

Statistical analysis

Data are expressed as mean ± standard deviation for normally distributed data and median and interquartile range for skewed data and were analysed using JMP® Pro 15.0.0 (SAS Institute Inc., Cary, NC, USA). The unpaired Student's t-test for normally distributed data and the Mann–Whitney U-test for skewed data were used to compare the clinical findings between patients treated with immunosuppressants and those who were not. The chi-square test was used to compare the sex distribution, the number of patients of each U-RBC grade at the time of renal biopsy, the Oxford classification, the prevalence of AAS, and immunosuppressive therapy, including corticosteroid therapy, tonsillectomy, and RASI use, between the groups. The cumulative renal survival rate until ESRD was analysed by the Kaplan–Meier method and compared by the log-rank test. Univariate and multivariate Cox regression analyses were used to evaluate the risk of deterioration to ESRD, which was expressed as a hazard ratio with 95% confidence intervals. Propensity score matching was conducted to adjust for differences in the clinical and histological findings and treatments between the RASI (+) and RASI (−) groups. A logistic regression model was used to estimate the propensity score between both groups. Matching was conducted using 1-1-digit matching (the nearest neighbour matching method). In all analyses, p < 0.05 was considered to be statistically significant.

Results Comparison of the baseline characteristics between IgAN patients with or without AAS

Table 1 demonstrates a comparison of the clinical, laboratory, and histological findings and treatments between the AAS0 and AAS1 groups. The median age, BMI, SBP, DBP, and MAP were significantly higher in the AAS1 group than in the AAS0 group (p < 0.001). Renal function, hyperuricaemia, and dyslipidaemia were significantly more severe and the U-Prot was significantly higher in the AAS1 group than in the AAS0 group (p < 0.001). The histological findings demonstrated that the incidence of the S1, T1, and T2 lesions was significantly higher in the AAS1 group than in the AAS0 group (p < 0.001). The numbers of M, E, and C lesions were similar between the two groups.

Table 1. Baseline characteristics of IgAN patients with AAS0 and AAS1. All patients AAS0 AAS1 P value Clinical findings Age (years) 31.0 (25.0–42.0) 27.0 (23.0–32.0) 40.0 (30.0–49.0) <0.001 Sex (male/female) 274/404 128/212 146/192 0.141 BMI (kg/m2) 21.4 (19.6–23.6) 20.8 (19.2–22.4) 21.8 (20.1–24.5) <0.001 SBP (mmHg) 120.0 (110.0–130.0) 116.0 (107.0–125.0) 123.0 (113.5–137.0) <0.001 DBP (mmHg) 74.0 (66.0–83.0) 71.0 (64.0–80.0) 78.0 (69.0–86.0) <0.001 MAP (mmHg) 89.3 (81.3–98.7) 86.3 (78.8–93.3) 92.7 (84.0–101.7) <0.001 Laboratory findings eGFR (ml/min.1.73 m2) 75.6 (59.0–94.1) 84.2 (68.0–103.3) 66.3 (51.1–81.4) <0.001 UA (mg/dl) 5.5 (4.6–6.8) 5.2 (4.2–6.4) 5.9 (4.8–7.2) <0.001 T-Cho (mg/dl) 193.0 (169.0–225.5) 184.0 (162.2–207.5) 205.0 (179.0–240.0) <0.001 TG (mg/dl) 100.0 (73.0–143.0) 89.5 (66.8–122.5) 108.0 (79.0–165.5) <0.001 U-Prot (g/day) 0.65 (0.29–1.34) 0.5 (0.22–1.09) 0.78 (0.37–1.67) <0.001

U-RBC (counts/HPF)

(<5, 5–25, 26–49, 50–99, ≥100)

70, 296, 112, 82, 115 24, 130, 57, 49, 79 46, 166, 55, 33, 36 <0.001 Histological findings, n M0/M1 340/338 180/160 170/168 0.491 E0/E1 365/313 186/154 179/159 0.648 S0/S1 177/501 109/231 68/270 <0.001 T0/T1/T2 505/134/39 293/36/11 212/98/28 <0.001 C0/C1/C2 335/309/33 177/145/17 158/164/16 0.321 AIT0/AIT1 415/263 340/0 75/263 – AH0/AH1 379/299 340/0 39/299 – The 20-year renal survival rate and risk factors for ESRD in IgAN patients with or without AAS

The renal survival rate was significantly higher in the AAS0 group than in the AAS1 group (84.7 versus 64.0%, p < 0.001; Figure 2). In univariate Cox regression analysis, the presence of AAS was found to be a significant risk factor for the progression to ESRD (Table 2, crude). In multivariate Cox regression analysis, we selected some histological, laboratory, and clinical findings that showed significant differences between the AAS1 and AAS0 groups in Table 1 and in the univariate Cox regression analysis (data not shown); multivariate analysis also revealed that the presence of AAS and other histological findings (S and T scores in the Oxford classification) were significant risk factors for progression to ESRD (Table 2, model 1) and that AAS, along with clinical and laboratory findings, histological findings, and treatments (immunosuppressants, tonsillectomy, and RASI), were significant risk factors for progression to ESRD (Table 2, model 3); however, it was not significant with clinical and laboratory findings (age, BMI, MAP, eGFR, and U-Prot) and histological findings (Table 2, model 2).

Renal survival rates in IgAN patients according to the absence or presence of AAS. Renal survival rate until ESRD was reached (AAS0: 84.7%, AAS1: 64.0%; p < 0.001).

Table 2. Univariate and multivariate Cox regression analyses of the presence of AAS as a risk factor in IgAN patients. AAS1 HR 95% CI P value Crude 3.01 1.85–5.26 <0.001 Model 1 2.07 1.20–3.58 0.009 Model 2 1.72 0.96–3.08 0.067 Model 3 2.23 1.21–4.09 0.010 Model 1: adjusted for S and T scores in Oxford classification. Model-2: model 1 + age, BMI, MAP, eGFR, and U-Prot. Model 3: model 2 + immunosuppressants, tonsillectomy, and RASI. CI, confidence interval; HR, hazard ratio. Comparison of the baseline characteristics and renal survival rate between IgAN patients with AAS who were treated with RASI initiated during the follow-up period and those who were not

In the AAS1 group, 338 IgAN patients were classified into the RASI (+) group (n = 223; RASI initiated during the follow-up period) and the RASI (−) group (n = 115; control group). The RASI (+) group was found to have more severe clinical, laboratory, and histological findings than the RASI (−) group (Table 3, left). Other treatments, such as immunosuppressants and tonsillectomy, were similar between the two groups; however, despite more severe disease, the 20-year renal survival rate was significantly higher in the RASI (+) group than in the RASI (−) group (69.4 versus 46.7%, p = 0.019; Figure 3A). After adjusting for background characteristics using propensity score matching with a logistic regression model for significantly different factors, such as age, sex, BMI, MAP, eGFR, U-Prot, and T-score (calliper a: 0.20, calliper c: 0.23), the baseline characteristics were found to be similar between the two groups (Table 3, right); however, the 20-year renal survival rate was significantly higher in the RASI (+) group than in the RASI (−) group (75.5 versus 44.3%, p = 0.013; Figure 3B). Univariate and multivariate Cox regression analyses revealed RASI as an independent factor that appeared to slow the disease progression in the AAS1 group (see supplementary material, Table S1).

Table 3. Comparison of baseline characteristics between RASI (−) and RASI (+) groups in the IgAN patients in the AAS1 group. Before propensity score matching After propensity score matching RASI (−) group (n = 115) RASI (+) group (n = 223) P value RASI (−) group (n = 79) RASI (+) group (n = 79) P value Clinical findings Age (years) 33.0 (27.0–43.0) 42.0 (33.0–51.0) <0.001 37.0 (29.0–47.0) 37.0 (30.0–46.0) 0.872 Sex (male/female) 40/75 106/117 0.024 35/44 31/48 0.519 BMI (kg/m2) 21.1 (19.3–23.2) 22.3 (20.5–25.4) <0.001 21.6 (20.0–23.4) 21.7 (19.9–24.3) 0.870 SBP (mmHg) 118.0 (109.0–128.0) 126.0 (117.0–140.0) <0.001 120.0 (110.0–131.0) 120.0 (112.0–133.0) 0.734 DBP (mmHg) 72.0 (66.0–84.0) 80.0 (71.0–88.0) <0.001 74.0 (68.0–87.0) 76.0 (68.0–82.0) 0.355 MAP (mmHg) 87.3 (80.0–97.0) 95.8 (87.7–104.3) <0.001 90.5 (82.3–100.3) 90.7 (83.3–99.0) 0.777 Laboratory findings eGFR (ml/min.1.73 m2) 75.4 (55.3–86.8) 63.8 (49.9–78.1) 0.003 66.4 (53.4–81.0) 72.8 (49.9–91.9) 0.674 T-Cho (mg/dl) 201.8 (170.0–248.0) 205.5 (183.1–235.5) 0.617 210.0 (168.0–250.0) 200.0 (177.5–240.5) 0.986 TG (mg/dl) 97.5 (72.5–157.0) 114.0 (82.0–169.0) 0.057 116.0 (78.5–166.5) 105.5 (79.0–159.0) 0.630 U-Prot (g/day) 0.5 (0.28–1.77) 0.94 (0.48–1.67) 0.009 0.62 (0.27–1.73) 0.75 (0.31–2.15) 0.640

U-RBC (counts/HPF)

(<5, 5–25, 26–49, 50–99, ≥100)

11, 52, 20, 17, 15 35, 114, 35, 16, 21 0.090 9, 37, 12, 13, 8 13, 41, 11, 6, 7 0.451 Histological findings, n M0/M1 57/58 113/110 0.847 35/44 44/35 0.152 E0/E1 58/57 121/102 0.504 39/40 43/36 0.524 S0/S1 30/85 38/185 0.053 19/60 13/66 0.234 T0/T1/T2 87/18/10 125/80/18 <0.001 57/15/7 756/20/3 0.306 C0/C1/C2 47/63/5 111/101/11 0.252 33/43/3 43/32/4 0.214 AIT0/AIT1 30/85 45/178 0.220 12/67 10/69 0.646 AH0/AH1 19/96 20/203 0.044 20/59 16/63 0.448 Treatment, n Immunosuppressants (−/+) 43/72 84/139 0.960 36/46 26/53 0.249 Tonsillectomy (−/+) 74/41 150/73 0.592 55/24 48/31 0.242 RASI (−/+) 115/− −/223 — 79/− −/79 — HPF, high-power field; RASI (−) group, RASI never treated during the follow-up period; RASI (+) group, RASI initiated during the follow-up period.

Renal survival rates in patients with IgAN according to whether they received RASI during the follow-up period. Renal survival rates until ESRD in IgAN patients with AAS1 (A) before propensity score matching (RASI (+): 69.4%, RASI (−): 46.7%; p = 0.019) and (B) after propensity score matching (RASI (+): 75.5%, RASI (−): 44.3%; p = 0.013) are shown.

Discussion

This study evaluated whether AAS was an independent risk factor for IgAN and whether RASI prevented progression in IgAN patients with AAS. Initially, we confirmed that IgAN patients with AAS had a higher age, BP, T-Cho, TG, and UA, which were also associated with systemic atherosclerosis. Moreover, AAS was found to be associated with lower renal function; higher concentrations of U-Prot; and S1, T1, and T2 lesions by the Oxford classification. Kaplan–Meier analysis, the log-rank test, and univariate and multivariate Cox regression analyses demonstrated that the presence of AAS was a significant risk factor for progression of the disease. These results indicate that AAS results from several clinical factors related to systemic atherosclerosis and chronic histological findings and that it is important to control these risk factors to prevent the progression of IgAN. The presence of AAS induces microvascular narrowing, which subsequently causes glomerular ischaemia and the activation of RAS [6]. Similarly, AAS induces glomerular hypertension and the impairment of renal autoregulation. All these factors induce systemic hypertension, subsequent glomerular hypertension, and hypertrophy of the remaining glomeruli, which results in glomerulosclerosis, tubulointerstitial damage, renal dysfunction, and proteinuria [15]. Therefore, it was hypothesised that RASI might prevent the progression to ESRD in IgAN patients with AAS. We suspected that RASI initiated at any time during the follow-up period might affect IgAN with AAS and prevent progression to ESRD because AAS are generally considered to cause irreversible chronic lesions. Then, we analysed the effects of RASI initiated during the follow-up period. The results revealed that IgAN patients in the AAS1 group treated with RASI were older, had a higher BP, and suffered from more severe renal dysfunction and tubulointerstitial damage than patients who were not treated with RASI. Despite this, treatment with RASI was found to significantly improve renal survival in IgAN patients with AAS. As expected, after adjusting for background characteristics in both groups, RASI was found to significantly improve renal survival in IgAN patients with AAS. Moreover, in IgAN patients without AAS, the reno-protective effects of RASI were not observed (see supplementary material, Figure S1). All these results support our hypothesis that AAS are risk factors for progression of IgAN, and that RASI is effective in IgAN patients with AAS, which is consistent with the findings of previous studies that reported that AAS induces glomerular ischaemia, increases RAS activity, and impairs autoregulation [7, 11]. Previous reports also support our hypothesis that the reno-protective effects of RASI include the suppression of glomerular hyperfiltration and glomerular hypertension, which are caused by the activation of RAS and increase in angiotensin II, thereby resulting in vasoconstriction, modification of the glomerular basement membrane, suppression of nephrin, induction of profibrogenic and proinflammatory cytokines, induction of chemokines, injury to podocytes, generation of reactive oxygen species, activation of nuclear factor-kappa B, and stimulation of fibroblast proliferation [12, 16]. RASI-induced suppression of RAS activation also reduces glomerular permeability and proteinuria and prevents a progressive decline in renal function. AAS was found to induce glomerulosclerosis and related lesions but our results differ from the original Oxford classification [1, 2] and a previous report [17]; there are very few reports that describe the beneficial effects of RASI on IgAN patients with arteriolar lesions. Kaneko et al reported that AH was associated with renal prognosis on univariate and multivariate analyses, considering other MEST-C scores, but not with clinical findings or RASI [5]. Zhang et al reported that AAS was associated with renal prognosis on univariate analysis, and that RASI prevented ESRD in patients with vascular lesions, including AAS and acute arterial lesions, such as arteriolar endothelial swelling, inflammatory cell infiltration, and thrombosis [18]. These previous reports support the hypothesis and results of this investigation.

Our study has limitations. First, it is a single-centre cohort study undertaken in Japan. AAS are significantly related to obesity, hypertension, dyslipidaemia, and hyperuricaemia, the rates of which vary significantly between countries due to different cultures, lifestyles, and dietary habits. Therefore, this study should be repeated internationally to obtain a global perspective. Second, this study is a retrospective analysis. Although patient characteristics were similar after propensity score matching, large randomised controlled trials are needed to generate more robust data.

In conclusion, our results suggest that AAS are associated with clinical and histological factors that cause disease progression in patients with IgAN. AAS may be an indicator of poor renal prognosis; however, treatment with RASI, initiated at any time during the follow-up period, may effectively slow the rate of patient decline.

Acknowledgement

We thank Editage (www.editage.jp) for editing the draft of this manuscript for English language.