Identification of Risk Factors for Renal Insufficiency in Primary Membranous Nephropathy

Summary

Objectives:

To identify possible evidence linking certain characteristics associated with primary membranous nephropathy (PMN) to renal failure progression in patients diagnosed with PMN and to establish probable evidence of an association between PMN and other risk factors leading to end-stage renal failure.

Methods:

A historical investigation was carried out including 436 subjects with PMN, with or without renal insufficiency, who were treated at the No. 82 Division Defense Medical Center of the People's Liberation Army and Liaocheng People's Medical Center during January 2020 and December 2023. Binary logistic regression study was applied to determine possible hazard elements. Binary logistic regression analysis was employed to identify potential risk factors, while the diagnostic capabilities of statistically significant continuous variables were assessed through detector performance feature receiver operating characteristic (ROC) plots.

Results:

Dichotomous logit regression study detected age, blood uric acid level, 24-hour urinary protein level, crescent formation, and renal tubular atrophy/interstitial fibrosis as independent risk factors for renal insufficiency, while hemoglobin level was a protective factor. The ROC analysis showed that age, 24-hour urine protein, and uric acid level had areas under the curve of 0.711 (p < 0.05), 0.652 (p < 0.05), and 0.60 (p < 0.05), with optimal threshold values of 52.5 years, 4.815 g/24 hour, and 446.5 μmol/L.

Conclusion:

Advanced age, hyperuricemia, anemia, high proteinuria, crescent formation, and renal tubular atrophy/interstitial fibrosis are significant risk factors for renal insufficiency in PMN patients. These findings may help to facilitate risk categorization and referral patterns.

Introduction

Membranous nephropathy (MN) is characterized by immune complex deposits at the outer surface of the glomerular basement membrane (GBM), diffuse thickening of the GBM, and subepithelial immune activation. The MN can be classified as primary (PMN) or secondary to other conditions. The etiology of PMN is incompletely elucidated but it is largely believed to be due to an overzealous immune response against multiple visceral epithelial antigenic markers such as type M phospholipid hydrolase A2 receptor molecule (PLA2R) leading to consequent podocyte and glomerular filtration barrier injury and resultant proteinuria. Approximately one third of patients with PMN will achieve late phase kidney condition by 10 years exposing both the patient and society to disruptions. Several studies have sought to investigated the factors determining the rate of deterioration of persistent renal disease [1,2]. There is less data on the factors affecting prognostications of PMN with consequent effects on the approach to timing interventions on a more favorable footing.

We performed a retrospective study of 436 patients diagnosed with PMN. Patients were stratified according to estimated glomerular filtration rate (eGfr) valuation: with normal renal function (eGfr ≥ 90 mLs/min/1.73 m²) and renal insufficiency (eGfr < 90mLs/min/1.73 m²). Our objective was to evaluate whether clinical and laboratory features differ between these 2 groups, and which features are independently associated with renal insufficiency, potentially to assist in stratifying prognosis and selecting personalized treatment options.

Methods

Historical review was performed for the clinical files of 517 subjects identified with MN who received treatment at the No. 82 Group Military Division Defense Medical Center of the National Citizens’ Emancipation Force or Liaocheng Public Hospital, in the era from January 2020 to December 2023. We included patients who: i) had MN confirmed by renal biopsy; ii) renal biopsies specimens showed at least 8 glomeruli by light microscopy; iii) had sufficient clinical, pathological, and laboratory information; and iv) had no significant coexisting conditions such as depression and low blood pressure. Exclusion criteria were: i) Microalbuminuria secondary to diabetic renal disease, lupus, viral liver disease, cancer or drugs (n = 61); ii) steroid and immunosuppressant treatment prior to renal biopsy (n = 12); iii) fewer than 8 glomeruli seen under light microscopy in renal biopsy specimens (n = 8). After exclusion of 81 patients, 436 primary membranous nephropathies were incorporated within the concluding assessment.

The 436 participants were categorized into individuals with healthy kidney function and individuals with kidney dysfunction by estimation of the eGFR. As per Kidney Disease - Improving Global Outcomes (KDIGO), an eGFR < 90 mL/min/1.73 m² is designated as dysfunction insufficiency and an eGFR ≥ 90 mL/min/1.73 m² is defined as optimal kidney performance [3,4]. In our study, 320 subjects had normal renal function comprising 190 males and 130 women. The average age was 49.28 ± 13.21 years, ranging from 15 to 76 years. The eGFR value obtained was 111.03 ± 15.48 mL/min/1.73 m², serum creatinine (SCr) was 60.39 ± 13.04 μmol/L from serum and blood urea nitrogen (BUN) was 4.70 ± 1.50 mmol/L. There were 116 patients who fell into the category of renal insufficiency, of which 84 were made up of males and 32 females. The average age stood at 58.96 ± 9.88 years, spanning from 32 to 77 years. The corresponding eGFR for this group was 72.33 ± 15.83 mL/min/1.73 m², SCr levels were 98.55 ± 32.73 μmol/L, and BUN levels reached 6.54 ± 2.89 mmol/L.

Clinical and laboratory test results, including subject's age and sex, eGFR, number of erythrocytes (red blood cell [RBC]), number of leukocytes (white blood cell [WBC]), platelets (PLT), number of urine red blood cells (URBC), and the subject's body mass index (BMI), as well as blood analysis results such as SCr, blood uric acid (UA), hemoglobin (Hb), total cholesterol (TC), triglycerides (TG), high density lipoprotein (HDL), low density lipoprotein cholesterol (LDL), blood serum albumin (ALB) and total proteins within 24 hour urinary sample (UTP) were obtained from participants in this study. The estimated glomerular filtration rate was calculated employing the chronic kidney disease epidemiology collaboration (CKD-EPI) eGFR formula [5].

For assessment, renal biopsy samples were examined by light, electron and fluorescent microscopic techniques to evaluate the extent of glomerulosclerosis includes focal segmental glomerular sclerosis, crescentic formations, tubular atrophy (TA) with interstitial fibrosis (IF), and small vessel abnormalities. Pathological features were scored according to the Oxford classification system for immunoglobulin A (IgA) nephropathy [6]. Scarring is characterized by mesangial cell hyperplasia, which may be followed by mesangial matrix expansion, capillary closure, and hyaline alterations at the affected site. Glomerulosclerosis was categorized as follows: i) no sclerosis (S0); ii) < 25% sclerosis (S1); iii) or ≥ 25% sclerosis (S2). Segmental sclerosis was recorded as present or absent. Crescentic lesions are defined by the extensive proliferation and accumulation of epithelial cells in the glomerular capsule, forming crescent-shaped bodies around capillary tufts and which was recorded as presentor absent. Tubular atrophy is characterized by reduced volume of tubular epithelial cells with a thicker basement membrane and narrower tubular lumen, while dilated tubules lumina may be seen in cases of obstruction. Another common renal tubulointerstitial lesion is renal interstitial fibrosis, which involves increased accumulation of collagen fibres within the renal interstitium. Tubular and renal interstitial lesions often occur together, with renal interstitial fibrosis being a key pathological change in this context. Tubular atrophy and interstitial fibrosis were classified based on the percentage of the affected area, as 0–25% (T0), 26–50% (T1), and > 50% (T2), in accordance with the Oxford classification criteria for IgA nephropathy. Small arterial abnormalities, such as small arterial sclerosis and hyaline degeneration, typically involve thickening of the muscular layer in these vessels, resulting in narrowed lumens or uniform thickening of the arterial walls.

Statistical analysis

Statistical computations were carried out employing statistical analysis tool (release 27.0; IBM, Armonk, NY, USA), and prior to analysis, the dispersion features of every smooth factor were assessed using the Shapiro–Wilk test to determine normality. Gaussian-distributed observations were contrasted across the 2 cohorts employing Student's t-test, and the results were expressed as the mean ± standard deviation (± s), whereas a Wilcoxon rank-sum method was utilized to contrast abnormally spread values, and these outcomes were exhibited in the form of midpoint and interquartile range (Q1, Q3). Count data were tested using the χ² test and are expressed as frequencies and percentages. Binary logit model study was performed to determine the distinct risk predictors contributing to renal insufficiency. Receiver operating characteristic plots were generated via GraphPad Prism software to determine the ability of risk factors to predict renal insufficiency in patients with PMN. A 2-sided p-value < 0.05 was deemed statistically significant.

The study followed the guidelines specified in the Declaration of Helsinki and secured approval by the ethical review board of the No. 82 Division Defense Medical Center of the People's Liberation Army with the reference number 2023121. All procedures were carried out in line with relevant guidelines and regulations, serving the purpose of patients’ diagnosis and evaluation. Given that the research adopted a retrospective design, the ethical board of the Chinese People's Liberation Army 82nd Division Defense Medical Center did not require the acquisition of informed consent.

3. Results

Comparison of baseline patient characteristics between the groups

There were no notable variations within BMI, WBC count, PLT count, URBC count, TG level, or HDL level between the 2 groups (p > 0.05). The ALB and Hb levels were significantly lower in the renal insufficiency group than in the normal renal function group (p < 0.05). The TC, LDL, UA, UTP, and PLA2R antibody levels, as well as age and the proportion of males, were significantly higher in the renal insufficiency group than in the normal renal function group (p < 0.05) shown in Table 1.

Table 1.

Comparison of clinical data between the 2 groups n(%), x¯ ± s, or M(Q1,Q3).

Clinical factors	Normal renal function group (n = 320)	Renal insufficiency group (n = 116)	t/χ2/Zvalue	P value
Sex
Male	190(59.4)	84(72.4)	6.199	0.013
Female	130(40.6)	32(27.6)
Age	49.28 ± 13.21	58.96 ± 9.88	–8.222	<0.001
BMI	26.45(24.08,29.15)	26.35(24.21,29.12)	–0.491	0.624
UA (μmol/L)	364.53 ± 89.63	399.91 ± 103.69	–3.490	<0.001
WBC(× 109/L)	6.39(5.17,7.55)	6.55(5.54,7.78)	–1.372	0.170
PLT(× 109/L)	256.0(209.0,296.0)	242.5(200.75,299.25)	–1.035	0.301
Hb (g/L)	138.29 ± 18.16	129.49 ± 19.60	4.376	<0.001
ALB (g/L)	25.18 ± 6.83	22.34 ± 6.94	3.809	<0.001
URBC (/HPF)	28.90(13.80,75.53)	20.55(11.13,54.80)	–1.278	0.201
UTP (g/24h)	3.98(2.08,6.57)	5.84(3.42, 8.68)	–4.839	<0.001
TC(mmol/L)	6.95 (5.76, 8. 59)	7.41 (6.02, 9.36)	–2.256	0.024
TG(mmol/L)	2.17 (1.58, 2.87)	2.35 (1.60, 3.50)	–1.242	0.214
LDL(mmol/L)	4.21 (3.24, 5.52)	4.80 (3.38, 6.29)	–2.143	0.032
HDL(mmol/L)	1.27 (1.09, 1.52)	1.23 (1.00, 1.42)	–1.626	0.104
PLA2R antibody (RU/ml)	37.91 (6.31, 104.24)	48.81 (8.78, 183.43)	–2.759	0.006

BMI: body mass index, UA: uric acid, WBC: white blood cells, PLT: platelet, Hb: hemoglobin, ALB: albumin, URBC: urine blood cells, UTP: total proteins within 24 hour urinary sample, TC: total cholesterol, TG: triglycerides, LDL: low density lipoprotein, HDL: high density lipoprotein, PLA2R: phospholipid hydrolase A2 receptor molecule.

Comparison of pathological indicators between the groups

There was no significant difference in the presence of segmental sclerosis across the 2 cohorts (p > 0.05). Nonetheless, the rate of glomerulosclerosis, crescent formation, lesions in small vessels, and tubular atrophy/interstitial fibrosis (TA/IF) remained substantially greater in the kidney dysfunction cohort than in the normal renal function group (p < 0.05) (Table 2).

Table 2.

Analysis of pathologic findings among the 2 cohorts n (%).

Item	Normal renal function group (n = 320)	Renal insufficiency group (n = 116)	χ2	P-value
Glomerulosclerosis
S0	135 (42.2)	35 (30.2)	–2.971	0.003
S1	171 (53.4)	66 (56.9)
S2	14 (4.4)	15 (12.9)
Segmental glomerulosclerosis
Yes	20 (6.2)	13 (11.2)	2.990	0.084
No	300 (93.8)	103 (88.8)
Crescent formation
Yes	6 (1.9)	12 (10.3)	15.431	<0.001
No	314 (98.1)	104 (89.7)
TA/IF
T0	158 (49.4)	33 (28.4)	–5.662	<0.001
T1	136 (42.5)	43 (37.1)
T2	26 (8.1)	40 (34.5)
Small artery disease
Yes	78 (24.4)	49 (42.2)	13.165	<0.001
No	242 (75.6)	67 (57.8)

TA/IF: tubular atrophy/interstitial fibrosis.

Risk factors for renal insufficiency in patients with PMN

Univariate analysis identified 13 significant factors that were subjected to binary logistic regression analysis, including age, UA level, UTP level, crescent formation, and TA/IF, were separately correlated with kidney dysfunction in participants with PMN (p < 0.05). The hemoglobin level was determined to be a protective factor (p < 0.05) (Table 3).

Table 3.

Binary logistic regression analysis of induced renal insufficiency.

Item	B-value	SE	Wald-value	OR	95% CI	P value
Sex	–0.576	0.385	2.238	0.562	0.264~1.196	0.135
Age	0.093	0.015	39.159	1.097	1.066~1.129	<0.001
UA	0.007	0.002	16.932	1.007	1.004~1.010	<0.001
Hb	–0.038	0.009	17.124	0.962	0.945~0.980	<0.001
UTP (g/24h)	0.133	0.040	11.218	1.142	1.057~1.235	<0.001
Alb	0.017	0.030	0.322	1.017	0.959~1.080	0.570
Tc	0.194	0.149	1.683	1.214	0.906~1.626	0.194
LDL	–0.196	0.175	1.253	0.822	0.584~1.158	0.263
PLA2R antibody (RU/ml)	0.001	0.000	1.303	1.001	1.000~1.001	0.254
Glomerulosclerosis(S0)	–0.143	0.314	0.207	0.867	0.468~1.605	0.649
Glomerulosclerosis(S1)	0.746	0.60	1.546	2.109	0.651~6.835	0.214
Crescent formation	2.546	0.787	10.456	12.752	2.726~59.659	<0.001
TA/IF (T0)	0.303	0.328	0.851	1.353	0.712~2.574	0.356
TA/IF (T1)	1.762	0.387	20.757	5.825	2.729~12.430	<0.001
Small artery disease	0.046	0.328	0.020	1.047	0.551~1.990	0.887

UA: uric acid, HB: hemoglobin, ALB: albumin, UTP: 24h urine sample, TC: total cholesterol, LDL: low density lipoprotein, PLA2R: phospholipid hydrolase A2 receptor molecule, TA/IF: tubular atrophy/interstitial fibrosis.

Predictive value of identified risk factors

To assess the diagnostic value of these variables, diagnostic ROC curves were constructed, considering the continuous factors that had been identified as independent risk factors for renal insufficiency in patients with PMN, namely age, UTP level, and UA level. The area under the ROC curve for age was 0.711 (p < 0.05), with a critical value of 52.5 years and a Youden index of 0.331. These analyses yielded a sensitivity of 78.4% and a specificity of 54.7% for the prediction of renal insufficiency. This suggests that age >52.5 years is an important predictor of renal insufficiency among subjects having PMN. The area under the ROC curve measured 0.652 (p < 0.05), with a cut-off level of 4.815 g/24 hour and Youden index of 0.273. When analyzing the diagnostic capabilities, the values of 66.4% sensitivity and 60.9% specificity were obtained for forecasting renal impairment. This suggests that a UTP level >4.815 g/24 hour constitutes a crucial marker of renal dysfunction in individuals with PMN. For UA levels, the area under the ROC curve was 0.60 (p < 0.05), with a critical value of 446.5 μmol/L and a Youden index of 0.207; the responsiveness and precision for forecasting renal insufficiency were 37.9% and 82.8%. These figures indicate that a UA level exceeding 446.5 μmol/L serves as a crucial indicator for renal insufficiency in patients with PMN (Fig. 1).

Fig. 1.

Receiver operating characteristic analysis of age, urine 24-hour sample and uric acid (UA) to predict renal insufficiency in primary membranous nephropathy. UTP: total proteins within 24 hour urinary sample, AUC: area under the curve.

Discussion

The PMN represents a frequent form of nephrotic syndrome affecting adults, and its occurrence has notably risen in China over recent years, especially among younger populations [7]. A pivotal diagnostic method for polymorphous nephritis (PMN) involved analysis of renal biopsy specimens through histopathological examination which is often conclusive. Characteristic changes of MN are seen including diffuse extensive thickening of the glomerular basement membrane (GBM), revealed under light microscopy, granular deposits of IgG along GBM with or without C3 detected via fluorescence microscopy, and regularly distributed electron-dense deposits beneath the GBM. Extensive fusion of podocyte foot-processes are often noted with electron microscopy. The PMN are diagnosed in the absence of other systemic diseases, infections, medications and tumours. Yet, the therapeutic responses in PMN are not fully encouraging, and persistent proteinuria may progress to renal failure which then eventually necessitates either renal replacement therapy or kidney transplantation. Thus, it is important to study the factors associated with renal insufficiency in patients with PMN so as to identify those individuals, at high risk for developing end-stage renal disease (ESRD), so that measures can be instituted to halt the progression of disease in such patients.

Creatinine is a metabolite of creatine, which in human muscle tissue occurs in the form of an energy reserve substance. It has a relative molecular weight of 113 Da, is non-teratogenic, is not metabolised by the kidneys, does not bind to proteins in the circulation, freely passes the glomerulus and is secreted by renal tubules. The SCr levels are indicative to renal function and are directly associated with adverse outcomes. The SCr level is influenced by muscle mass, dietary intake, sex, and age. The eGFR, on the other hand, compensates for those effects on SCr measurement, providing a more accurate assessment, and forms the basis for chronic kidney disease (CKD) staging. Hence, eGFR was used in assessing renal impairment with a cut-off of 90 mL/min/1.73m². Serum creatinine values also showed renal function status with values of 60.39 ± 13.04 and 98.55 ± 32.73 μmol/L in the respective normal and impaired kidney function groups. In patients with PMN, a correlation analysis was conducted to explore the associations between various factors and renal insufficiency, including age, UA level, Hb level, UTP level, crescent formation, and TA/IF. This is largely consistent with previous studies [8,9].

The rise in the proportion of elderly individuals worldwide has emerged as a significant concern in the field of public health. With the growing proportion of senior citizens worldwide, there is a corresponding rise in the prevalence of kidney disorders among this demographic. A single-center retrospective observational analysis of 74 subjects diagnosed with seropositive pauci-immune glomerulonephritis and severe proteinuria (urine protein-to-creatinine ratio ≥3.5 g/d) revealed that older patients experienced a significantly faster decline in eGFR (risk ratio, 1.04; 95% confidence limits: 1–1.1) [10]. This is congruent with the findings of our research, which identified age was independently associated with the development of renal insufficiency in patients with PMN. For every additional year of age, the risk of renal insufficiency increased by 9.7%. ROC curve evaluation indicated that age exceeding 52.5 years was a notable indicator of renal insufficiency in patients with PMN. With advancing years, glomerulosclerosis, immune complex deposition (IF), and arteriolar wall thickening tend to occur more frequently, and immune complex deposition has also been found to be linked to the progression of ESRD [11].

A growing body of research suggests that elevated blood UA concentrations serve as a standalone determinant in the progression of CKD [12]. Aligning with this accumulating body of findings, our analysis revealed that UA levels also exhibit an independent correlation with renal impairment in individuals afflicted with PMN. Specifically, each increment of 1 μmol/L in serum UA concentration was linked to a 0.7% higher probability of developing renal insufficiency. In conventional understanding, once the concentration of uric acid in the blood surpasses the threshold solubility of 420 μmol/L, monosodium urate crystals will form and accumulate in various tissues throughout the body. These deposits within the kidneys may lead to conditions such as uric acid nephrolithiasis, aseptic interstitial nephritis, or obstructive nephropathy. Nowadays, it is generally accepted that UA contributes to the impairment of vascular endothelium, which may be brought about by suppressing the expression of endothelial nitric oxide synthase and the generation of nitric oxide, stimulating the renin-angiotensin system, upregulating the renin receptors on endothelial cells, increasing the expression of nicotinamide adenine dinucleotide phosphate (NADPH) oxidase 4, and producing reactive oxygen species that trigger oxidative stress and other modifications in endothelial function [13,14,15]. Furthermore, UA reduces the synthesis of E-cadherin while enhancing its breakdown, and simultaneously induces a phenotypic switch in renal tubular together with vascular units through reactive challenge as well as elevated carbohydrate coat loss, that in turn leads to renal tubulointerstitial fibrosis and subsequently impacts renal function [16,17].

Chronic kidney disease frequently gives rise to anemia, a condition often linked to reduced erythropoietin synthesis in the body. Nevertheless, there is limited research available regarding how anemia influences the progression of CKD. In our research, a reduction in hemoglobin levels was observed to be strongly linked to renal impairment in patients with PMN, while an elevation in such levels demonstrated a protective effect against the development of renal insufficiency. Tissue ischemia and hypoxia causing renal tubular interstitial injury are associated with the advancement of CKD. Anemia can induce organ oxygen deficit with related oxidative stress and change the activity of the adrenergic nervous system in the kidney, consequently acting as a contributing factor for the worsening of CKD [18].

In our research, we noted that proteinuria in patients with PMN is a risk factor for developing renal insufficiency. We found that for each increase of 1 gram of UTP, there was a 14.2 % higher risk of developing renal insufficiency. Elevated proteinuria may cause glomerular pressure and glomerular perfusion to increase in more than 1 way, increasing glomerular permeability and causing damage. Infections can cause tubular necrosis and proteinuria is related to renal tubular damage as it can cause apoptosis of renal tubular epithelial cells, resulting in cell injury and functional disruption of renal tubules.

Patients exhibiting crescentic glomerulonephritis were 12.75 times more likely to develop acute renal dysfunction when compared to patients who did not exhibit any evidence of crescentic deposition. Clinically, membranous lesions accompanied by a small amount of crescent formation were observed; no secondary factors were identified during renal biopsy [19]. In clinical settings, the study excluded cases where glomerular lesions were associated with anti-GBM antibodies or anti-neutrophil cytoplasm antibodies, as these would be classified separately as anti-GBM disease and anti-neutrophil cytoplasm antibody-related vasculitis. Formation of crescents is a histomorphological marker of severe glomerular injury [20], often accompanied by capillary endothelial cell injury, basement membrane disruption, renal capsule epithelial cell proliferation, and infiltration of various inflammatory cells, which can impair glomerular filtration and lead to proteinuria and decreased renal function. Numerous investigations have demonstrated that the presence of a limited number of crescent structures can exert a negative impact on the clinical outcome of patients with PMN [21].

Renal function impairment is significantly associated with the extent of tubulointerstitial damage in the kidney. Renal biopsy investigations have demonstrated that, even when the primary damage begins in the glomeruli, renal tubulointerstitial fibrosis correlates more strongly with CKD progression than does glomerular damage [21]. Research indicates that chronic tubulointerstitial lesions serve as an independent risk factor in the transition of PMN to ESRD, and they also mediate this progression across diverse CKD types, through mechanisms such as tubular backleak, flawed tubuloglomerular feedback, capillary rarefaction, and the formation of a tubular glomeruli [22,23,24]. Our research established the extent of TA/IF as a notable risk factor for renal insufficiency in patients with PMN.

One notable drawback of this research lies in its use of a cross-sectional and retrospective approach, as well as the absence of follow-up information, along with the failure to assess how treatment affects outcomes. In addition, this research provides insight into potential risks that may cause kidney damage in people with PMN but additional studies should be conducted to see if those same potential risk factors actually do cause kidney damage.

In conclusion, the development of renal impairment in patients with PMN relates to aspects such as patient age, concentrations of UA, Hb and UTP, the development of crescents, and the presence of TA/IF. The latter, together with male gender, older age, increased UA, anemia, proteinuria and crescent progression have been associated with a higher chance of reduced eGFR. Predicting who is at highest risk may permit timely intervention which in turn could preserve renal function in patients with PMN.

Acknowledgment

The authors would like to acknowledge Editage (www.editage.cn) for the English language review.

Disclosure statement

The authors declare that no AI-assisted technologies were used in the writing of this manuscript. The data used to support the study are included within the supplementary information files. For data analysis and image creation, the statistical software SPSS and GraphPad Prism were employed.

Disclosure

The authors declare that they have no affiliations with or involvement in any organization or entity with any financial interests in the subject matter or materials discussed in this manuscript.

This work was supported by the Scientific Research Project of the Baoding City Science & Technology Bureau (2341ZF188).