Efficacy and safety of plasma exchange for crescentic and mixed classes of ANCA-associated glomerulonephritis with renal insufficiency Authors: Zhuan’e Yao, Pengbo Wang, Yanting Gao, Peng Zhang Published: 2025 ABSTRACT AbstractBackgroundTherapy regimen of patients with biopsy-proven crescentic and mixed antineutrophil cytoplasmic antibody (ANCA)-associated glomerulonephritis with renal insufficiency is uncertain. Plasma exchange (PE) treatment remains controversial when the kidney is involved in ANCA-associated vasculitis.MethodsPatients with crescentic and mixed classes of ANCA-associated glomerulonephritis and serum creatinine ≥ 250 µmol/L between January 2020 and July 2024 in our center were retrospectively screened. Patients who received PE in addition to standard induction therapy with glucocorticoid and cyclophosphamide (CTX) or mycophenolate mofetil (MMF) were divided into PE group and patients who received standard induction therapy were divided into non-PE group as controls. Primary endpoint was mortality and dialysis independence.ResultsA total of 61 patients with diagnosis of ANCA-associated glomerulonephritis confirmed by renal biopsy with serum creatinine ≥ 250µmol/L were included in this study. 30 patients received standard induction therapy, and 31 were treated with plasma exchange in addition to standard induction therapy. PE plus standard induction therapy showed initial benefit in patient short-term survival rates, mortality at 3 months which occurred in 3 of 31 patients (9.7%) in PE group and in 12 of 30 patients (40.0%) in non-PE group. Mortality at 12 months showed no significant difference in both groups. Addition of PE showed no benefit on dialysis independence during follow-up. Patient dialysis independence rates at 3 months, 12 months were 19 (61.3%) of 31, 18 of 31 (58.1%) in PE group and 14 (46.7%) of 30, 13 of (43.3%) 30 in non-PE group, respectively. Infection events were more common in PE group (incidence rate, 61.3%) than in non-PE group (30%).ConclusionsPE plus standard induction therapy improved short-term survival at 3 months in patients with crescentic and mixed ANCA-associated glomerulonephritis, However, it did not improve the dialysis independence and the 12-month survival rate, and was associated with a higher risk of infection. This indicates that the clinical application of PE requires a balance between its early survival benefits and infection risks, and should be individualized for the patients who are most likely to benefit.Clinical trial numberNot applicableKeywords: ANCA, Glomerulonephritis with crescentic and mixed classes, Renal insufficiency, Plasma exchange Introduction Renal involvement is common in patients with ANCA-associated vasculitis (AAV) (64%–85%), which is often presented as rapidly progressive Glomerulonephritis (GN) with a pauci-immune necrotizing crescentic GN on renal biopsy sample histology [1, 2]. Patients with severe kidney dysfunction are still at a high risk of death and end-stage kidney disease (ESKD) [3]. The role of ANCA in the pathogenesis of renal vasculitis is supported by the induction of neutrophil activation and superoxide release and neutrophil-mediated endothelial cytotoxicity by ANCA in vitro, by the induction of crescentic glomerulonephritis in animal studies by ANCA, and by the close association of ANCA with pauci-immune renal vasculitis [4–9]. The rapid removal of ANCAs by means of plasma exchange(PE)may reduce organ damage from ANCA-associated vasculitis [10]. The use of plasma exchange therapy for remission-induction of AAV with kidney disease is controversial. MEPEX (methylprednisolone versus plasma exchange) trial found a higher rate of renal recovery and dialysis independence at 3 months after plasma exchange than those after the addition of intravenous methylprednisolone for patients with Wegener’s granulomatosis (WG) or microscopic polyangiitis and a creatinine at diagnosis > 500 µmol/L (5.8 mg/dl) [11]. Szpirt WM et al. suggested that PE was recommended for induction therapy in WG patients at creatinine levels > 250 µmol/L (2.85 mg/dl) and they found renal survival after 1, 3 and 12 months, and 5 years was significantly better in the PE groups, however, this study had a small sample size (n = 16, in PE group) and did not involve intravenous methylprednisolone pulse therapy [12]. Moreover, the PEXIVAS trial failed to demonstrate that plasma exchange delayed the time to ESKD or death for patients with AAV presenting with GFR < 50 ml/min per 1.73 m2 or alveolarhemorrhage over a median follow-up of 2.9 years [13]. Kidney histology is a predictive indicator for long-term risk of kidney failure [14]. But renal histology information was not provided in PEXIVAS, limiting the ability to determine whether PE may be beneficial in subgroups of patients based on histological changes. Furthermore, by multivariate Cox regression analysis, the sclerotic class had a significantly worse renal prognosis than the focal or combined mixed/crescentic classes, but no significant differences in outcome were observed in the crescentic class compared with the mixed class or the combined mixed/crescentic class compared with the focal class [15]. The 2021 KDIGO clinical practice guidelines for ANCA-associated vasculitis described that in validation studies of the histopathologic classification by Berden, > 50% normal glomeruli in the focal class were associated with a favorable outcome, whereas > 50% sclerotic glomeruli were associated with a poor outcome [16]. However, regarding the crescentic class (> 50% cellular crescents) and mixed class, discrepancies in outcome have been reported. Therefore, in our study, we aimed to address whether plasma exchange is associated with the improvement survival in patients with a serum creatinine ≥ 250 µmol/L and crescentic or mixed classes of ANCA-associated glomerulonephritis. Materials and methods Patients We reviewed 117 ANCA patients with renal biopsy-proven between January 2020 and July 2024 in our center. We excluded patients if they met any of the following criteria: Sclerotic class (n = 12) and focal class (n = 15):These classes were excluded because they respectively have clear poor prognoses and good renal prognoses, therefore, our study mainly focused on patients with crescentic and mixed classes, in which the benefits of plasma exchange remains the most uncertain and are clinically relevant. Sclerotic class (n = 12) and focal class (n = 15):These classes were excluded because they respectively have clear poor prognoses and good renal prognoses, therefore, our study mainly focused on patients with crescentic and mixed classes, in which the benefits of plasma exchange remains the most uncertain and are clinically relevant. Positive anti-GBM antibodies (n = 3):These patients were excluded because of the different disease mechanisms and treatment plans involved. Positive anti-GBM antibodies (n = 3):These patients were excluded because of the different disease mechanisms and treatment plans involved. Combined with other glomerular disease (n = 6):To ensure the pure analysis of the AAV results, patients combined with other glomerular disease were excluded. Combined with other glomerular disease (n = 6):To ensure the pure analysis of the AAV results, patients combined with other glomerular disease were excluded. Serum creatinine < 250umol/L (n = 17):On the one hand, patients with preserved renal function at presentation were excluded because they typically have a favorable prognosis and are unlikely to be considered for aggressive treatment with plasma exchange in routine practice. On the other hand, based on the article by Szpirt WM et al [12]. , we determined to include patients with an serum creatinine ≥ 250µmol/L Serum creatinine < 250umol/L (n = 17):On the one hand, patients with preserved renal function at presentation were excluded because they typically have a favorable prognosis and are unlikely to be considered for aggressive treatment with plasma exchange in routine practice. On the other hand, based on the article by Szpirt WM et al [12]. , we determined to include patients with an serum creatinine ≥ 250µmol/L Incomplete data (n = 3):Exclude patients who lack key clinical or outcome data. Incomplete data (n = 3):Exclude patients who lack key clinical or outcome data. The remaining 61 patients were included in this study, of which 31 underwent plasma exchange (PE group) and 30 did not (non-PE group) (Figure. 1). It is worth noting that the allocation between the treatment group and the control group was not random. Inclusion flowchart The primary clinical indications for PE include rapid deterioration of renal function, presence of pulmonary hemorrhage, and other life-threatening manifestations of vasculitis. Of course, the final decision on whether to perform plasma exchange is made based on the severity of the patient’s condition, the judgment of the attending physician, and the informed consent of the patient and their family.Typically, patients with more severe manifestations(such as rapidly rising creatinine levels, dependence on dialysis, or presence of pulmonary hemorrhage) were more likely to be offered and to undergo plasma exchange. The study was approved by the Ethics Committee of Shaanxi Provincial People’s Hospital. Before plasma exchange treatment, patients or their surrogate decision makers provided written informed consent. Data collection Demographic, clinical, biochemical data and renal biopsy findings at the time of the ANCA-associated glomerulonephritis diagnosis were recorded. Renal biopsies were reviewed by experienced nephropathologists. ANCA-associated glomerulonephritis (AAGN) severity was categorised based on standard histological classification into crescentic (≥ 50% cellular crescents), focal (≥ 50% normal glomeruli), sclerotic (≥ 50% globally sclerosed glomeruli) and mixed [14]. The record of adverse events, including infection, leucopenia, thrombocytopenia, thrombosis, and hemorrhage, were based on continuous monitoring during follow-up. The study endpoint was mortality and dialysis independence after the therapeutic intervention. Treatments The treatment regimens of ANCA-associated glomerulonephritis is divided into an initial phase, termed “induction,” followed by a “maintenance” phase. Induction treatment included glucocorticoids in combination with cyclophosphamide (CTX) or mycophenolate mofetil (MMF). With regard to the glucocorticoid regimens, patients were treated with daily intravenous methylprednisolone for 3 days (0.5–1.0 g/d), then received oral prednisone (1.0 mg/kg/day for 6-8 weeks, then decrease by 10 mg every two weeks, tapered to 5-10 mg/d for maintenance). The CTX dose was 0.5–1.0 g/m2/month and the cumulative amount is 6–8 g. Oral MMF was prescribed at 1.5 g/day. Furthermore, the maintenance therapy consisted of MMF (1.5 g/d) or azathioprine (1.5-2 mg/kg/d at complete remission until one year after diagnosis then decrease by 25 mg every 3 months). For the small proportion of patients (approximately 13.3% in the non-PE group) who did not receive the aforementioned drug therapy due to contraindications to CTX/MMF, rituximab therapy was administered. Plasma exchange The temporary vascular access was created by inserting of a Prismaflex dual lumen catheter into the femoral vein or internal jugular vein. For plasma exchanges, a plasma separator Haemoselect® L 0.5(B. Braun Avitum Ag, German)was used to separate blood into blood cells and plasma. The replacement fluid mainly consists of 5% albumin, with a plasma exchange volume of 60 ml/kg on each time.Fresh frozen plasma is used for patients at high risk of bleeding, such as those with severe alveolar hemorrhage or who have recently undergone a renal biopsy.The blood flow was set at 100–120 ml/min. During the PE treatment, low molecular weight heparin (60 IU/kg, intravenous bolus) or no anticoagulant was given for anticoagulation according to coagulation function. Seven times plasma exchanges were provided over 14 days as a session. Outcomes assessment The primary endpoint was mortality and dialysis independence. The secondary outcomes was the severe adverse event rates. Adverse events included those requiring admission to infection (pneumonia, and other infection), leucopenia, thrombocytopenia, thrombosis and hemorrhage. To evaluate sustained remission, we used the Birmingham Vasculitis Activity Score for Granulomatosis with Polyangiitis (BVAS/GPA) to assess disease activity (scores range from 0 to 68, with a score of 0 indicating the absence of disease activity and higher scores indicating increasingly active disease) [17]. Definitions The outcomes investigated were kidney survival and patient survival at 3 and 12 months. Kidney survival was defined as dialysis independence, serum creatinine < 250 µmol/L, and patient survival. Survival time was calculated from the beginning of treatment for ANCA-associated glomerulonephritis to the time of death from any cause or censored. Treatment-related complications included infection, leucopenia, thrombocytopenia, thrombosis, and hemorrhage. Specific definitions of adverse events are as follows: Infection requiring hospitalization (mainly pulmonary infection); leucopenia associated with treatment (persistent white blood cell counts less than 4.0 × 109/L); thrombocytopenia associated with treatment (platelet count less than 100 × 109/L); thrombus requiring treatment (deep venous thrombosis; pulmonary embolism and cerebral infarction) and life-threatening bleeding (bleeding of skin and mucous membrane; pulmonary hemorrhage; gastrointestinal bleeding; cerebral hemorrhage). Statistical analysis Continuous variables are shown as the means ± standard deviations. Student’s t-test and Mann-Whitney rank tests were applied to evaluate the difference between groups for normal and non-normally distributed continuous variables, respectively. Categorical variables are shown as percentages and were compared using the chi-square test or Fisher’s exact test. Independent predictors of mortality were evaluated using the Cox regression model. Variables were entered into a multivariate Cox regression model using an enter method, and the variables were derived from adjusted models: model 1 was adjusted for age, male, alveolar hemorrhage present, BVAS, and serum creatinine; model 2 was adjusted for the variables in model 1 plus methylprednisolone IV bolus, and cyclophospamide; model 3 was adjusted for the variables in model 1 plus crescentic class; and model 4 was adjusted for age, male, alveolar hemorrhage present, BVAS, serum creatinine, crescentic class, methylprednisolone IV bolus, and cyclophospamide. The estimated hazard ratios (HR) along with 95% confidence intervals (CI) and P values are reported for all regression covariates. A two-tailed P < 0.05 was considered significant. All statistical calculations were conducted using SPSS 26.0 (SPSS Inc, Chicago, IL). Results Study population The demographics and characteristics of the included patients are displayed in Table 1. Sixty-one patients were included. At the discretion of the physicians, thirty-one of them received plasma exchange as supplement to the standard induction therapy with CYX or MMF and glucocorticoids, while thirty patients were treated with induction therapy alone. Our data revealed that the choice to consider PE was more likely in patients with alveolar hemorrhage present. Patients with alveolar hemorrhage present had a higher proportion in PE group compared to patients in non-PE group (35.5% versus 10.0%, P = 0.018,Table 1). Moreover, there were no significant differences in demographics, clinical parameters, Birmingham Vasculitis Activity Score or histologic lesion between groups before the initial treatment. Table 1Clinical characteristics and histological lesions of the patients with ANCA-associated systemic vasculitis and renal failure new-onsetCharacteristicAll patients(n = 61)PE group(n = 31)Non-PE group(n = 30)P value Clinical characteristics Age, years58.4 ± 13.355.4 ± 2.861.5 ± 9.40.072 Male, n (%)33 (54.1)17 (54.8)16 (53.3)0.906 Alveolar hemorrhage present, n (%)14 (23.0)11 (35.5)3 (10.0)0.018 Dialysis requiring, n (%)34 (55.7)19 (61.3)15 (50.0)0.375 BVAS19.3 ± 3.919.8 ± 4.118.8 ± 3.70.319 Ant-PR3 positive, n (%)13 (21.3)9 (29.0)4 (13.3)0.134 Ant-MPO positive, n (%)49 (80.3)23 (74.2)26 (86.7)0.221 ANCA titer89.2 ± 37.484.8 ± 37.494.5 ± 37.30.348 Serum creatinine, µmol/L653.2 ± 237.7694.8 ± 268.6610.3 ± 196.30.167 C-reactive protein (mg/L)53.5 ± 58.153.9 ± 59.352.8 ± 57.80.985 Histologic Lesion Crescentic class23 (37.7)12 (38.7)11 (36.7)0.869 Mixed class38 (62.3)19 (61.3)19 (63.3)0.869 % normal glomeruli16.1 ± 10.315.4 ± 11.016.9 ± 9.70.591 % crescents43.9 ± 18.945.6 ± 19.242.3 ± 18.70.496 cellular/cellularfibrous43.1 ± 18.644.5 ± 19.141.7 ± 18.30.552 fibrous0.8 ± 0.81.0 ± 3.70.6 ± 2.40.576 % global sclerosis19.9 ± 13.819.0 ± 13.320.9 ± 14.60.599ANCA: antineutrophil cytoplasmic antibody, PE: plasma exchange, Non-PE: non-plasma exchange, BVAS: birmingham vasculitis activity score, PR3 :proteinase 3, MPO: myeloperoxidase Clinical characteristics and histological lesions of the patients with ANCA-associated systemic vasculitis and renal failure new-onset ANCA: antineutrophil cytoplasmic antibody, PE: plasma exchange, Non-PE: non-plasma exchange, BVAS: birmingham vasculitis activity score, PR3 :proteinase 3, MPO: myeloperoxidase Treatment and clinical outcomes At the time of intervention, treatment between patients with and without plasmapheresis was different. A higher frequency of high-dose (bolus) intravenous methylprednisolone administration and cyclophosphamide therapy (93.5% vs. 63.3%, p = 0.004; 93.5% vs. 66.7%, p = 0.008, respectively) were in patients who received plasmapheresis. Moreover, seven patients (one in the PE group and six in the non-PE therapy group) received mycophenolate mofetil at the time of intervention. Treatment information is summarized in Table 2. Table 2Treatment and outcomes of the patients with ANCA-associated systemic vasculitis and renal failureCharacteristicAll patients(n = 61)PE group(n = 31)Non-PE group(n = 30)P value Treatment at the moment of intervention Methylprednisolone IV bolus, n (%)48 (78.7)29 (93.5)19 (63.3)0.004 Cyclophospamide, n (%)49 (80.3)29 (93.5)20 (66.7)0.008 Mycophenolate mofetil, n (%)7 (11.5)1 (3.2)6 (20.0)0.040 Cumulative doses Cumulative dose of prednisone, grams13.18 + 10.214.93 + 3.180.039 Cumulative dose of Cyclophospamide, grams4.89 + 2.964.27 + 2.530.471 Outcomes Follow-up, months17.2(2.7,40.3)20.8(6.5,42.1)11.1(1.2,39.5)0.896 MortalityAt 3 months, n (%)15 (24.6)3 (9.7)12 (40.0)0.006 At 12 months, n (%)23 (37.7)8 (25.8)15 (50.0)0.051 Dialysis independence At 3 months, n (%)33 (54.1)19 (61.3)14 (46.7)0.252 At 12 months, n (%)31 (50.8)18 (58.1)13 (43.3)0.250ANCA antineutrophil cytoplasmic antibody, PE plasma exchange, Non-PE non- plasma exchange, IV intravenous Treatment and outcomes of the patients with ANCA-associated systemic vasculitis and renal failure ANCA antineutrophil cytoplasmic antibody, PE plasma exchange, Non-PE non- plasma exchange, IV intravenous Mortality at 3 months after the intervention was 9.7% in PE group and 40% in non-PE group (p = 0.006) (Table 2). At 12 months, mortality was 25.8% and 50.0%, respectively, without differences between groups (p = 0.051). These findings indicated that the benefit of plasmapheresis was shown in the first three months after the intervention. No difference was observed in the dialysis independence at 3 months and 12 months between PE and non-PE group. Risk factors for the combined event in the study population Univariate Cox regression analyses revealed that plasma exchange had a statistically significantly improved the combined events at 3 months (HR = 0.194, 95% CI 0.055–0.690, P = 0.004) and 12 months (HR = 0.373, 95% CI 0.158–0.883, P = 0.025), respectively. Four multivariate Cox regression models were created, as shown in Table 3. In model 4, PE therapy (HR = 0.113, 95% CI 0.024–0.523, P < 0.01) was associated with a reduced risk of combined events after adjusting for age, male, alveolar hemorrhage present, BVAS, serum creatinine, crescentic class, methylprednisolone IV bolus, and cyclophosphamide at 3 months. This finding was still observed in the other three models. Table 3Effects of plasma exchange on survival of ANCA-associated systemic vasculitis and renal failurePlasma ExchangeAt 3 monthsAt 12 monthsHR (95% CI)P valueHR (95% CI)P valueUnivariate0.194 (0.055–0.690)0.0040.373 (0.158–0.883)0.025Multivariate model 1a0.156 (0.040–0.604)0.0020.228 (0.084–0.616)0.004Multivariate model 2b0.186 (0.044–0.782)0.0220.281 (0.093–0.849)0.024Multivariate model 3c0.121 (0.030–0.492)0.0010.215 (0.078–0.587)0.003Multivariate model 4d0.113 (0.024–0.523)0.0050.215 (0.069–0.668)0.008aModel 1 was adjusted for age, male, alveolar hemorrhage present, BVAS, and serum creatininebModel 2 was adjusted for age, male, alveolar hemorrhage present, BVAS, serum creatinine, methylprednisolone IV bolus, and cyclophospamidecModel 3 was adjusted for age, male, alveolar hemorrhage present, BVAS, serum creatinine, and crescentic classdModel 4 was adjusted for age, male, alveolar hemorrhage present, BVAS, serum creatinine, crescentic class, methylprednisolone IV bolus, and cyclophospamide Effects of plasma exchange on survival of ANCA-associated systemic vasculitis and renal failure aModel 1 was adjusted for age, male, alveolar hemorrhage present, BVAS, and serum creatinine bModel 2 was adjusted for age, male, alveolar hemorrhage present, BVAS, serum creatinine, methylprednisolone IV bolus, and cyclophospamide cModel 3 was adjusted for age, male, alveolar hemorrhage present, BVAS, serum creatinine, and crescentic class dModel 4 was adjusted for age, male, alveolar hemorrhage present, BVAS, serum creatinine, crescentic class, methylprednisolone IV bolus, and cyclophospamide Adverse events A total of 88 adverse events including infection, leucopenia, thrombocytopenia, thrombosis, and hemorrhage were reported for 61 patients (Table 4). There were 19 infection in 31 (61.3%) patients in PE group and 9 infection in 30 (30%) patients in Non-PE group. A higher prevalence of pneumonia events was observed in patients treated with plasmapheresis (51.6% vs. 26.7%, p = 0.046). No significant difference was found for the leucopenia, thrombocytopenia or thrombosis rates between the two groups. Table 4Adverse eventsCharacteristicAll patients(n = 61)PE group(n = 31)Non-PE group(n = 30)P valueInfection events, n (%)28 (45.9)19 (61.3)9 (30.0)0.014 Pneumonia, n (%)24 (39.3)16 (51.6)8 (26.7)0.046 Others, n (%)4 (6.6)3 (9.7)1(3.3)0.317Leucopenia, n (%)6 (9.8)5 (16.1)1 (3.3)0.093Thrombocytopenia, n (%)17 (27.9)11 (35.5)6 (20.0)0.178Thrombosis, n (%)4 (6.6)3 (9.7)1 (3.3)0.317Hemorrhage, n (%)5 (8.2)2 (6.5)3 (10.0)0.614 Adverse events Discussion Our trial in patients with crescentic and mixed classes of ANCA-associated glomerulonephritis showed that plasma exchange can resulted in a lower incidence of death within the 3 months than no plasma exchange, but the effect was not found over longer term follow up. The results suggested that the use of plasma exchange did not significantly improve the recovery to dialysis independence during follow-up. Moreover, infection events were more common in plasma exchange group than in non-plasma exchange group. The use of plasma exchange (PE) for induction treatment of ANCA-associated vasculitis is still controversial. The randomized, controlled trial of patients with severe renal vasculitis (serum creatinine > 500 µmol/L) reported by David R.W. Jayne et al. showed at 3 months, 33 (49%) of 67 after intravenous methylprednisolone compared with 48 (69%) of 70 after plasma exchange were alive and independent of dialysis (95% confidence interval for the difference 18 to 35%; P = 0.02) and patient survival rates were similar in both groups [11]. These results were partially in line with the results reported in our study. In our study, the rate of mortality at 3 months after the intervention was 9.7% in PE group and 40% in Non-PE group (p = 0.006), but at 12 months and at the end of follow-up, there was no significant difference between two groups (p = 0.051 and P = 0.906, respectively). Results from PEXIVAS indicated that routine use of PE during remission-induction did not improve long-term patient survival [13]. Additionally, in a randomized controlled single-centre study, 32 patients with ANCA-positive Wegener’s granulomatosis were randomized to treatment with or without initial PE. There was no difference in mortality between PE and control groups at follow-up [12], but the significantly lower ANCA titers were found in PE groups after 1 month (P < 0.02). And this study also showed that PE is recommended for induction therapy in WG patients at creatinine levels > 250 µmol/L. Similarly, in our cohort, patients presented with a creatinine value of > 250 µmol/L and the results showed that use of PE was not associated with improved patient survival during the long-term follow-up. In accordance with previous study, physical removal of ANCA antibodies, adhesion molecules and cytokines by PE could therefore have a short-term beneficial effect on treatment results [18–20]. Most deaths occurred during the first 3 months, when vasculitis was most active. Thus, effective early therapy is imperative to avoid vasculitic death [11]. We found that PE was associated with a statistically significant reduction in the mortality in the early stage of treatment (within the 3 months). So, the rapid removal of ANCA may have accounted, in part, for the decline of short term mortality in our study. Moreover, we recognize, there were no established treatment guidelines for these patients, therapeutic regimens were determined according to both the doctor’s experience and the patient’s status. In PE group, patients possibly presented a more severe course of the disease with greater organic compromise, including a higher frequency of DAH [21, 22].Our study observed a higher mortality rate, we consider this may be related to the following factors: on the one hand, our study included a higher proportion of patients with severe manifestations, such as 23% of the patients having pulmonary alveolar hemorrhage at the time of consultation, and 55.7% of the patients relying on dialysis before treatment, which are all known risk factors for death.Additionally, our study had a smaller sample size, and in a smaller cohort, a few adverse events might disproportionately increase the mortality rate.Finally, we reported a real-world unselected cohort, which typically includes patients who are excluded from randomized controlled trials (RCTs) due to comorbidities or severity of other diseases, resulting in a higher observed mortality rate.The short-term benefits of plasmapheresis in our study may be due to the reduction of death caused by DAH and cardiovascular events. However, the higher infection rate was observed in PE group, which was the main cause of death in our population. This may then explain why PE did not reduce the risk of death during long-term follow-up. Previous studies have demonstrated the safety of plasma exchange, since catheter-associated infections didn’t occur [23], and serious infections were less frequent in the reduced-dose glucocorticoids group (7 events in 5 patients, 7.2%) than in the high-dose group (20 events in 13 patients, 20.0%) (P = 0.04) [24]. Meanwhile, among patients with newly diagnosed ANCA-associated vasculitis, a reduced-dose glucocorticoid plus rituximab regimen was non-inferior to a high-dose glucocorticoid plus rituximab regimen with regard to induction of disease remission. Together, these observations highlight the recommendations for oral glucocorticoid rapid-tapering regimen for AAV in 2021 KDIGO, which involves a lower glucocorticoid cumulative dose. PE has been advocated as an adjunct therapy in AAV with renal involvement but without a clear consensus. It is worth noting that our study differs from the two major randomized controlled trials in this field.In the MEPEX study [11], the results found a higher rate of renal recovery and dialysis independence after plasma exchange than after the addition of intravenous methylprednisolone for patients with Wegener’s granulomatosis or microscopic polyangiitis and a creatinine at diagnosis > 500µmol/L (5.8 mg/dl), this study included 137 patients with Wegener’s granulomatosis or microscopic polyangiitis and investigated whether the addition of plasma exchange was more effective than intravenous methylprednisolone, which were different from our inclusion criteria and regimens.The MEPEX study only enrolled patients with severe renal impairment (Cr > 500umol/L), whereas our study included patients with lower creatinine levels (Cr ≥ 250umol/L), capturing a broader spectrum of renal impairment severity.The control group in the MEPEX trial received high-dose intravenous methylprednisolone treatment, while the control group in our study received the standard induction treatment regimen (glucocorticoid and CTX/MMF). The primary endpoint in the MEPEX study was dialysis independence at 3 months, while our primary endpoint was mortality and dialysis independence. The PEXIVAS trial was a larger-scale study, the results of the PEXIVAS trial in patients with newly diagnosed or relapsing severe AAV, including those with GN (eGFR < 50 ml/min per 1.73 m2), who were randomized to receive or not receive PE as an adjunct to standard remission-induction therapy, indicated that PE did not reduce the mortality rate or the incidence of ESKD in patients with AAV [13], this is consistent with our long-term research results, which indicated no significant difference in 12-month mortality or dialysis independence.However, our study also differs from the PEXIVAS trial, The PEXIVAS trial enrolled patients with severe active ANCA-associated vasculitis, whereas our study specifically focused on patients with crescentic and mixed histology types.The PEXIVAS trial showed that the reduced-dose regimen of oral glucocorticoids was not inferior to the standard dose regimen, and this regimen was associated with fewer infections. In our study, the PE group received a higher cumulative glucocorticoid dose (13.18 + 10.21 g vs. 4.93 + 3.18 g, p = 0.039), which may explain the higher infection rate observed.A post-hoc analysis ofthe PEXIVAS trial investigated that the addition of PE to standard immunosuppressive therapy led to a significant improvement in early renal function, with the greatest benefit observed at week 4 [relative risk (RR): 1.41; 95% CI: 1.09–1.82].Week 4 renal function recovery was significantly associated with reduced risk of ESKD at week 52 (RR: 0.96, 95% CI: 0.95–0.97 and RR: 0.29, 95% CI: 0.16–0.52, respectively), these findings suggest that PE can improve the early renal function of patients with ANCA-associated glomerulonephritis [25].Furthermore, a prior study showed that AAV patients with progressive disease who received standard induction therapy and plasmapheresis had significant improvement in renal function and similar long-term outcome [26]. Our long-term follow-up of kidney outcomes and mortaltiy rate also confirmed that there was no difference between PE groups and non-PE groups. It is worth noting that although the reduction in 12-month mortality rate with PE in this study did not reach statistical significance (P = 0.051), we observed a strong trend that the mortality rate in the PE group was lower than that in the non-PE group, this trend is consistent with the findings of some previous studies, for example, a meta-analysis of ANCA-associated vasculitis patients, which included 9 trials and 387 patients, showed that the use of PE was associated with a positive signal of reduced mortality [27]. However, because we only studied a relative small number of patients, further studies need to be done to verify whether patients selected for plasmapheresis indeed have a higher dialysis independence rate and a lower mortality rate. Although the higher infection rate observed in PE group, we found that plasmapheresis use was relatively safe and catheter-associated infections didn’t occur, as described in previous trial [22]. In our study, the high incidence of adverse events was primarily attributable to the following factors: Firstly,, infections and cytopenias caused by the use of glucocorticoids, cyclophosphamide, or mycophenolate mofetil; Secondly, delayed clearance of drugs and their metabolites in patients with renal failure; Thirdly, the presence of active vasculitis itself. Patients with ANCA-associated glomerulonephritis have several important risk factors for infections, such as the long-term use of immunosuppression, the placement of central intravenous catheters and uremia/dialysis dependency [28, 29]. In our study, infection events were related to the use of immunosuppressive drugs, we opt to withhold immunosuppressive therapy if the disease is limited to the kidneys or there is little hope to preserve renal function. As such, research to reduce side effects from immunosuppressions while improving disease control is urgently needed. Because of the most deaths occurred during the first 3 months [11], the rapid glucocorticoid tapering regimen combined with plasma exchange for AAV may be more advocated in the future. The differences in efficacy and safety between this study and the main randomized controlled trials (MEPEX and PEXIVAS) need to be carefully considered. Besides the differences in study design and patient baseline characteristics, the ethnicity might be a potential cause for these different findings, Our study population was all Chinese, in contrast to the predominantly Caucasian populations enrolled in the MEPEX and PEXIVAS trials, this disparity could explain the observed differences in outcomes.Unfortunately, as a single-center retrospective study, we lack a control group from different ethnicities, we were unable to directly validate this hypothesis in our data. Future prospective, multiethnic studies are needed to definitively clarify the role of ethnicity factors in AAV therapy outcomes. This study has several limitations that should be considered. First, its retrospective design and single-center setting inevitably introduce selection bias, as the decision to administer PE was made by physicians based on clinical indications rather than randomization, although we used a multivariate regression model to adjust for measurable confounding factors, unmeasured or residual confounding factors may still exist. Second, the relatively small sample size may limit the statistical power to detect significant differences in certain outcomes, particularly those with lower event rates. Third, during the study period, the therapeutic strategies were influenced by individual physicians’ experience and patient status, potentially increasing the treatment heterogeneity. Finally, since all the participants were Chinese, which limited the generalizability of the research results. Conclusions For patients with crescentic and mixed classes of ANCA-associated glomerulonephritis, combined with PE on the basis of standard immunosuppressive therapy can significantly reduce the short-term (3-month) mortality rate, but it does not improve the dialysis independence and long-term survival rate (12 months). Moreover, this treatment strategy is associated with a higher risk of infection. Therefore, in clinical practice, the benefits and risks need to be weighed and individualized treatment should be carried out to achieve precise treatment. Acknowledgements The authors sincerely appreciate the time and effort of all who contributed. Abbreviations Plasma exchange Cyclophosphamide Mycophenolate mofetil Glomerulonephritis End-stage kidney disease Wegener’s granulomatosis Birmingham Vasculitis Activity Score Granulomatosis with polyangiitis ANCA-associated vasculitis Diffuse alveolar hemorrhage Author contributions Zhuan’e Yao and Peng Zhang designed the study, acquired data, interpreted data, drafted and revised the manuscript.Yanting Gao and Pengbo Wang acquired data, interpreted data and revised the manuscript. Peng Zhang designed the study, revised the manuscript, assumed responsibility for the integrity of the data and supervised the research group. All authors read and approved the final manuscript. Funding No funding. Data availability The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request. Declarations Ethics approval and consent to participate The study was approved by the Ethics Committee of Shaanxi Provincial People’s Hospital and was conducted in accordance with the Declaration of Helsinki. The need for informed consent was waived by the Ethics Committee due to the retrospective nature of the study. Consent for publication Not applicable. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Data Availability Statement The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.