Effects of immunosuppressive therapy on renal prognosis in primary membranous nephropathy

Abstract

Background

Immunosuppressive therapy plays a crucial role in treating membranous nephropathy, with previous studies highlighting its benefits for patients with primary membranous nephropathy (PMN). Guidelines suggest that the management of membranous nephropathy should be tailored to individual risk levels. However, there is a lack of real-world studies examining the effects of immunosuppressive therapy on renal outcomes in PMN patients. This study aimed to investigate the relationship between immunosuppressive therapy and renal prognosis in PMN patients.

Methods

This was a real-world retrospective study including patients diagnosed with PMN in Shenzhen Second People’s Hospital and Hechi People’s Hospital. Univariate and multivariate Cox regression analysis and Kaplan-Meier survival analysis were used.

Results

After propensity score-matching, 464 PMN patients were included and they were assigned to conservative and immunosuppressive group in a 1:1 ratio. Immunosuppressive therapy was the protective factor of renal composite outcome (HR = 0.65, p < 0.01). Separately, the effect was significant in moderate- and high-risk but not in low-risk patients. Key influencing factors including age, blood pressure, albumin and total cholesterol levels, with slight differences among patients at different risk.

Conclusions

This study demonstrates the efficacy of immunosuppressive therapy in non-low-risk PMN patients. The key factors affecting renal prognosis in patients with different risk levels are emphasized to help provide individualized treatment.

Introduction

Immune suppression is essential in the treatment of autoimmune diseases, such as primary membranous nephropathy (PMN) [1]. PMN is characterized by the deposition of immune complexes in the glomerular basement membrane, leading to proteinuria and renal dysfunction [2]. The pathogenesis of MN involves activation of the immune system, participation of B and T cells, and production of autoantibodies against podocyte antigens, including phospholipase A2 receptor (PLA2R) and thrombospondin type-1 domain-containing 7 A (THSD7A) [3,4]. Although numerous target antigens associated with PMN have been continuously identified in recent years, PLA2R remains the most reliable biomarker for diagnosis, risk stratification, prediction of spontaneous remission, and evaluation of response to immunosuppressive therapy in PMN, being detected in 74–78% of cases [5,6].

Immunosuppressive therapy plays a critical role in the management of PMN to address underlying immune dysregulation [7,8]. In the late 20th century, randomized controlled trials conducted on PMN demonstrated that high-dose corticosteroids were effective in mitigating renal deterioration. However, the associated adverse effects resulted in this treatment not being preferred [9]. Subsequent studies revealed that the combination of alkylating agents (including chlorambucil and cyclophosphamide) with corticosteroids improved renal outcomes in PMN patients [10,11,12], surpassing the efficacy of corticosteroids alone [13], and this combination is now recommended in the KDIGO guidelines [14]. Additionally, calcineurin inhibitors (CNIs) have been increasingly recommended for PMN patients due to their anti-proteinuric effects [15,16,17,18,19]. A However, the adverse effects of immunosuppressive agents, which include infections, leukopenia, cardiovascular events, and malignancies, have also been reported [20,21]. A high relapse rate has been noted among PMN patients treated with CNIs [16,18,22]. Consequently, the optimal therapeutic regimen for MN remains a subject of debate due to the varying efficacy and safety profiles of the available agents. Previous research has even indicated renal deterioration following immunosuppressive therapy and highlighted severe adverse effects [23]. The emergence of rituximab (RTX) appears to offer a new avenue for the treatment of PMN, particularly in patients experiencing relapse or refractory disease [24,25,26,27]. Although the safety of RTX has been proved, its efficacy in some specific patients remained controversial [28].

This real-world study therefore aims to investigate the effects of immunosuppressive agents on renal outcomes in PMN patients. Additionally, we will examine the factors influencing the prognosis of PMN patients with varying risk stratifications.

Materials and methods

Data collecting

This retrospective cohort study gathered data from electronic medical records at Shenzhen Second People’s Hospital and Hechi People’s Hospital. Information on demographics, comorbidities, comedications, and laboratory test results was collected. Raw data were inputted into a spreadsheet, scrutinized for inconsistencies, and then subjected to data cleaning and anonymization procedures to ensure accuracy. Approval for this study was obtained from the Medical Ethics Committee of Shenzhen Second People’s Hospital (ethical number: 201408186) and Hechi People’s Hospital (ethical number: 2020022). The data collection process strictly adhered to ethical standards and legal regulations to uphold the research’s integrity and ethical conduct.

Study participants

In this real-world study, participants were retrospectively recruited continuously from 2008 to 2023 at Shenzhen Second People’s Hospital and from 2020 to 2023 at Hechi People’s Hospital. Eligible participants were patients diagnosed with PMN through kidney biopsy and screening of secondary factors including gastrointestinal endoscope, detection of common infectious diseases (such as TB and HBV) and PET-CT. Exclusion criteria included: (1) age < 14 years old; (2) baseline estimated glomerular filtration rate (eGFR) < 30 mL/min/1.73 m² (in order to ensure the stability of renal function and facilitate the observation of treatment efficacy); (3) with unknown treatment; (4) use of any immunosuppressants within 12 months before kidney biopsy. Patients were divided into two groups according to the therapy. Conservative group was defined as PMN patients treated with supportive treatments including renin-angiotensin-system inhibitors (RASi), other hypotensive drugs, antiplatelets/anticoagulant drugs, urate-lowering drugs, lipid regulating agents and sodium-glucose cotransporter 2 inhibitors (SGLT2i). Immunosuppressive group was defined as PMN patients with immunosuppressive therapy (involving rituximab, tacrolimus, cyclosporin, cyclophosphamide, steroids, tripterygium glycosides, and mycophenolate mofetil) combined with/without supportive treatments. They were monitored until conversion of therapy, or reaching the study’s endpoint. Participant confidentiality was rigorously maintained throughout the study, with personal information anonymized and securely stored.

Variables and clinical outcomes

Groups were categorized based on the therapeutic regimens. Comorbidity denotes additional health conditions or diseases present in the participants. This study focused solely on analyzing diabetes mellitus, hypertension. Comedication was described as medications used prior to the kidney biopsy. The estimated glomerular filtration rate was determined using the EPI equation. Proteinuria was assessed through the urine protein-to-creatinine ratio or 24-hour urine protein quantification. The study’s outcome was defined as achieving the renal composite outcome, indicated by a combination of an eGFR < 15 mL/min/1.73 m² or a 30% decline, or transitioning to hemodialysis. Kidney transplant was not observed in our cohort, so it was not a part of the composite outcome. In addition, a 50% decline was also analyzed as an outcome.

Risk stratification: based on baseline data (last data within 30 days before renal biopsy).

• (1) Low-risk: normal eGFR, proteinuria < 3.5 g/d and serum albumin > 30 g/L;

• (2) Moderate-risk: normal eGFR and proteinuria ≥ 3.5 g/d, not fulfilling high-risk criteria;

• (3) High-risk: satisfy any of the following:

• ① eGFR < 60 mL/min/1.73 m²;

• ② proteinuria > 8 g/d;

• ③ normal eGFR, proteinuria ≥ 3.5 g/d, and serum albumin < 25 g/L (or PLA2R antibody titers [PLA2Rab] > 50 RU/ml, urinary β₂-microglobulin > 250 mg/d) [all patients who did not undergo PLA2Rab testing had data on urinary β₂-microglobulin].

Statistical analysis

All data analysis was conducted using Empower Software (http://www.empowerstats.com, X&Y Solutions, Inc., Boston, MA, USA) and R version 4.2.2 (2022-10-31) software. Propensity score matching (PSM) was employed to establish a 1:1 matched cohort, and propensity scores were calculated using logistic regression. Age, gender, body mass index, blood pressure, comorbidities, albumin, eGFR, proteinuria and comedications (RASi, antiplatelets, urate-lowering drugs, diuretics, calcium channel blockers, β-blockers, lipid regulating agents, calcitriol, calcium supplements, and proton-pump inhibitors) were used to construct regression model. To balance adequate sample size and eliminate confounding factors and selection bias between the two groups, the 1:1 matched pairs of recipients of the two treatments were selected with a caliper of absolute value of propensity score difference at 0.01. However, data after PSM were not used to explore influencing factors of renal prognosis in patients with different risk levels.

In the assessment of balance, all statistical tests were two-tailed, with a significance level set at P < 0.05 indicating statistically significant differences. Continuous variables following a normal distribution were presented as mean ± standard deviation (Mean ± SD) and compared using independent samples t-tests between groups. Non-normally distributed continuous variables were presented as median and interquartile range [M (Q1, Q3)] and compared using the Mann-Whitney U test. Categorical data were presented as frequencies and proportions [n (%)] and compared using Pearson’s χ2 test or Fisher’s exact test. Additionally, Standardized Mean Differences (SMD) were used to compare differences between groups. SMD values < 0.10 were considered indicative of acceptable balance between groups, values between 0.10 and 0.34 indicated small differences, values between 0.35 and 0.64 indicated moderate differences, values between 0.65 and 1.19 indicated large differences, and values ≥ 1.20 indicated very large differences between groups [29].

Results

Characteristics of participants

A total of 829 PMN patients were initially enrolled in the study conducted between 2008 and 2023 at Shenzhen Second People’s Hospital and Hechi People’s Hospital. According to the exclusion criteria, 79 patients were eliminated. 750 patients entered the study cohort including 390 patients treated with conservative therapy and 360 patients treated with immunosuppressive therapy. Among total participants, 61.40% patients were male, with an average age of 48.28 ± 14.05 years old (not shown in the Tables). Patients with immunosuppressive therapy were generally younger, had combined nephrotic syndrome, and used more comedications. Following propensity score-matching (PSM), 232 patients were selected for each treatment group (Table 1; Fig. 1). Table 2 showed the baseline data of the included patients from the perspective of risk stratification.

Table 1 Baseline characteristics of the participants stratified by initial therapy

Fig. 1

Flowchart of the participants

Table 2 Baseline characteristics of PMN patients in different risk levels

Table 3 details the immunosuppressive treatment regimens administered to the study cohort, involving combination therapies with steroids and ACEI/ARB. Cyclosporin was the most frequently prescribed regimen at 36.21%, followed by tacrolimus at 25.00%. 28 (12.07%) were treated with steroids alone. This single-steroids treatment regimen was mostly used in patients before 2012, who underwent monotherapy for approximately three months. If there was no significant improvement in proteinuria, immunosuppressants would be subsequently added. The number of individuals using RTX, cyclophosphamide (CTX), and tripterygium glycosides was comparable. Only 3 (1.29%) patients used mycophenolate mofetil (MMF) which was also before 2012.

Table 3 Immunosuppressive treatment regimens of patients

Association between immunosuppressive therapy on renal survival

A total of 457 patients were included in the follow-up analysis, with 225 in the conservative group and 232 in the immunosuppressive group. Among these patients, 177 (38.73%) reached the renal composite outcome. The overall median survival time was 77.51 months, with the conservative group exhibiting a median survival time of 45.00 months and the immunosuppressive group showing a substantially higher median of 121.10 months. The survival rates at the first, third, and fifth year were significantly higher in the immunosuppressive group compared to the conservative group (Table 4).

Table 4 Survival rates (Kaplan-meier estimates)

Kaplan-Meier (KM) analysis demonstrated that patients receiving immunosuppressive therapy had a lower cumulative incidence of the renal composite outcome than those undergoing conservative therapy (p < 0.001) (Fig. 2A). Notably, improved renal outcomes were observed in moderate- (p = 0.013) and high-risk (p = 0.003) PMN patients treated with immunosuppressants (Fig. 2C-D). Conversely, immunosuppressive therapy did not significantly influence renal prognosis in low-risk populations (Fig. 2B).

Fig. 2

Cumulative incidence of renal composite outcomes in total (A), low-risk (B), moderate-risk (C), and high-risk (D) patients

In the analysis of risk factors affecting renal prognosis, we first conducted univariate Cox regression, identifying age, gender, BMI, blood pressure, comorbidities, certain comedications, hemoglobin levels, albumin levels, blood lipid profiles, renal function, proteinuria, and PLA2Rab titers as significant risk factors for the renal composite outcome (Table 5). Many of these factors were adjusted for in the subsequent multivariate Cox regression analysis. As indicated in Table 6, immunosuppressive therapy was significantly associated with improved renal outcomes in all patients (p = 0.0241), primarily reflecting the prevention of a decline in renal function exceeding 30% (p = 0.0305). However, the protective effect was less pronounced for patients experiencing greater than 50% decline in renal function, end-stage renal disease (ESRD), and those requiring hemodialysis. Nevertheless, the significant protective effect persisted in high-risk patients.

Table 5 Univariate cox analysis of renal composite outcome

Table 6 Association between immunosuppressive therapy and renal composite outcome (after PSM)

Subgroup analysis was conducted to explore differences and influencing factors among various patient categories. Univariate Cox analysis was employed to compare characteristics and outcomes across subgroups, with results for each subgroup illustrated in Fig. 3. Male patients without comorbidities, but of older age, lower albumin levels, and higher proteinuria levels, appeared to benefit more from immunosuppressive therapy. A significant interaction effect was observed exclusively within the proteinuria groups. Additionally, a sub-analysis was performed for moderate and high-risk patients treated with calcineurin inhibitors (CNIs; tacrolimus and cyclosporin) compared to those treated with CTX or RTX. After adjustment, no significant differences were observed in the relationship between treatment regimen and renal survival, whether assessing declines greater than 30% (p = 0.422) or greater than 50% (p = 0.106).

Fig. 3

Subgroup analysis of immunosuppressive therapy on PMN patients. DM, diabetes mellitus; HTN, hypertension; Alb, albumin

Influencing factors of renal composite outcome in different risk levels

In addition to thoroughly examining the factors influencing renal prognosis and emphasizing the impact of immunosuppressive therapy on renal outcomes, we conducted a preliminary analysis of the determinants affecting renal prognosis in different risk-stratified populations using univariate Cox regression (Table 7). Although the population meeting the criteria for nephrotic syndrome is not classified within the low-risk group, baseline levels of albumin (HR = 0.92; p < 0.001), proteinuria (HR = 1.00; p < 0.01), and lipid levels (HR = 1.19; p < 0.01) emerged as significant factors influencing renal prognosis. Blood pressure also proved to be an important factor. With the exception of PLA2Rab titers and the use of stomach protectants, other factors did not demonstrate significant effects in moderate-risk patients. In contrast, high-risk patients were notably influenced by age, BMI, blood pressure, comorbidities, comedications, and renal function. Regarding immunosuppressive therapy, it was identified as a protective factor against the progression of renal function decline in non-low-risk patients.

Table 7 Analysis of the factors influencing the renal composite outcome in different risk levels

Discussion

This retrospective study investigated the association between immunosuppressive therapy and composite renal outcomes in PMN patients. Analyzing data from 464 participants after propensity score matching (PSM), we found that immunosuppressive therapy significantly improves renal prognosis. In addition to therapy, key risk factors for renal composite outcomes included age, blood pressure, albumin levels, blood lipid profiles, renal function, proteinuria, PLA2Rab titers, and the use of certain medications, with slight variations observed across the three risk levels.

Previous studies conducted over 20 years ago indicated the benefits of early immunosuppressive treatment for most PMN patients, in contrast to conservative approaches, which were associated with poorer prognoses [30]. Since then, the landscape of treatment has evolved considerably, marked by the increased utilization of various immunosuppressive agents in managing MN. Monoclonal antibody therapy, particularly RTX, has gained prominence due to its proven efficacy and safety demonstrated in large randomized controlled trials [31]. RTX has shown notable efficacy in patients with PLA2R-associated MN, resulting in reduced proteinuria and improved eGFR and albumin levels [32]. While some patients do not respond to RTX treatment, challenges remain in determining the optimal dosage based on PLA2R antibody titers [33]. Emerging anti-CD20 monoclonal antibodies, such as ofatumumab, ocrelizumab, and obinutuzumab, have demonstrated enhanced antibody-dependent cellular cytotoxic activities and efficacy in relapsed, refractory patients, as well as in those with anti-RTX antibodies [34,35,36,37,38,39,40]. Until now, most evidence supporting these agents has come from case reports or case series, with only a limited number of randomized controlled trials or retrospective studies conducted, primarily focusing on pediatric nephrotic syndrome [41,42]. Consequently, no patients in our cohort received second- or third-generation CD20 monoclonal antibodies.

In addition to the continuous introduction of new therapeutic drugs, treatment regimens are evolving alongside the accumulation of clinical experience. In terms of monotherapy, RTX achieved a 60% rate of complete or partial remission compared to 52% with cyclosporine at 12 months. Furthermore, these advantages were even more pronounced at 24 months due to a lower relapse rate [31]. A real-world study revealed higher remission rates in PMN patients receiving multitarget therapy combining RTX, cyclophosphamide, and prednisone [43]. Similarly, a randomized controlled trial demonstrated comparable efficacy between multitarget therapy and cyclical alternating treatment involving corticosteroids and cyclophosphamide, but with fewer side effects [44]. Reports have also underscored the positive impact of multitarget therapy on secondary and rare forms of MN [45,46]. Overall, immunosuppressive therapy, particularly the use of monoclonal antibodies, proves beneficial for most PMN patients by enhancing renal function and offering new treatment possibilities for MN.

In clinical practice, the stratification of patients’ risk levels and the subsequent selection of treatment regimens cannot be rigidly confined to established guidelines. In this study, we primarily stratified patients based on baseline data. Some patients classified as intermediate to high risk opted for conservative therapy due to economic constraints or concerns regarding potential adverse drug reactions. Conversely, some patients designated as low risk according to established criteria were found, through comprehensive clinical assessment, to still possess a risk of disease progression. This discrepancy elucidates why certain low-risk patients in our cohort received immunosuppressive therapy, whereas some high-risk patients did not.

A review of our historical medical records revealed that several treatment regimens not recommended by the guidelines, such as MMF or steroid monotherapy, were exclusively utilized prior to 2012. MMF has emerged as a substitute for CTX due to its reduced adverse effects [47]. Some previous studies suggested its efficacy in achieving remission in a shorter timeframe [48,49], however, more recent evidence has indicated a higher rate of primary nonresponse and relapse as follow-up time extends [50]. Additionally, the use of steroids alone was based on a treatment protocol in which certain patients were initially treated with steroids for three months. If there was no remission of proteinuria during this period, immunosuppressants were subsequently added. The 28 patients in our cohort who were treated with steroids alone demonstrated acceptable efficacy by the end of the three months, thus not requiring the addition of immunosuppressants.

Our study confirmed the benefits of immunosuppressive therapy in PMN patients. The immunosuppressive group achieved higher survival rates at the first, third, and fifth years. To further clarify the extent to which immunosuppressive therapy protects renal function, we subdivided our study endpoints into different degrees of renal function decline. A significant protective effect was observed in patients exhibiting greater than 30% decline in eGFR, however, this effect was less pronounced in those with greater than 50% decline in eGFR, as well as in cases of ESRD and those requiring hemodialysis. When analyzed according to risk stratification, high-risk patients demonstrated a significantly lower cumulative incidence of ESRD and hemodialysis. This indicates that the use of immunosuppressive agents in PMN, particularly among moderate to high-risk patients, can effectively delay the progression of renal dysfunction. This finding supports the guidelines’ recommendation for the use of immunosuppressants in populations categorized as moderate and high risk.

We also analyzed the prognostic factors for renal outcomes across different risk stratifications. Blood pressure and indicators related to nephrotic syndrome were identified as risk factors in low-risk patients. Although patients meeting the diagnostic criteria for nephrotic syndrome were eliminated from this group, albumin, blood lipids, and proteinuria levels remained important for renal function protection. As mentioned in the guidelines, moderate-risk patients exhibit greater treatment flexibility, allowing for the selection of either conservative or immunosuppressive therapy based on individual circumstances. PLA2R antibody titers emerged as a significant factor guiding treatment decisions. In high-risk patients, age appeared to have a more influential impact on renal prognosis than clinical markers. Elderly patients may find it difficult to tolerate immunosuppressants due to increased susceptibility to infections and other adverse effects. The presence of diabetes may further elevate the risk of infections, including severe infections in these patients. In some high-risk patients with renal impairment, renin-angiotensin system inhibitors and other antihypertensives may reduce renal perfusion, thereby exacerbating the progression of kidney dysfunction. Therefore, antihypertensive agents should be used with caution in this population. Consistent with the results of multivariate Cox regression analysis, immunosuppressive therapy emerged as a protective factor in moderate- and high-risk patients.

Our study presents several notable advantages. Firstly, this cohort includes the largest sample size of PMN patients compared to any previous real-world studies. We utilized PSM to minimize the influence of confounding factors between groups. Secondly, with the emergence of new therapeutic regimens and drugs, we reaffirmed the protective effect of immunosuppressive therapy on PMN patients, particularly those not classified as low risk. Finally, we explored the factors that influence renal prognosis across varying levels of disease progression risk. However, our study also has some limitations. Its retrospective design may introduce information bias and residual confounding, despite the use of PSM. Additionally, data on adverse effects were inadequately documented in our prior medical records and were therefore not included in our analysis. Furthermore, due to the rapid adjustments of medications based on patients’ conditions in clinical practice, some drugs had limited follow-up observation time, hindering the assessment of their long-term effects on renal function. Future research should involve larger prospective cohorts to validate these findings and investigate the interplay of various influencing factors.

In conclusion, in the non-low-risk PMN population, the use of immunosuppressants may modestly slow the decline of renal function. In high-risk groups, these agents can significantly reduce the likelihood of progression to end-stage renal disease and the need for hemodialysis. It is crucial to emphasize the importance of blood pressure control and the appropriate use of comedications in managing PMN patients. Individualized treatment plans should take into account the patient’s risk level and relevant influencing factors to optimize renal prognosis.