Rituximab With or Without Avacopan in ANCA-Associated Vasculitis With Severe Renal Involvement Authors: Cristián Juanet, Isabel Hassi, Anila Cara, Cristina García Majado, Jose Arriola-Montenegro, Trisha Laxamana, Gian Marco Berti, Maria J. Vargas-Brochero, Charat Thongprayoon, Jose A. Meade-Aguilar, Samuel D. Falde, Ulrich Specks, Fernando C. Fervenza, Ladan Zand Published: 2026 ABSTRACT IntroductionAvacopan has shown efficacy as a glucocorticoid-sparing therapy in anti-neutrophil cytoplasmic antibody (ANCA)-associated vasculitis (AAV), but its benefit in patients with severe renal involvement treated with rituximab remains uncertain.MethodsWe conducted a retrospective comparative cohort study at Mayo Clinic including patients with AAV and estimated glomerular filtration rate (eGFR) < 30 ml/min per 1.73 m2 at induction treated with rituximab plus avacopan (n = 30) or rituximab alone (n = 120) between 2010 and 2025. Coprimary outcomes were time to renal remission (no > 25% eGFR decline and hematuria ≤ 10 red blood cells per high-power field [RBC/HPF]), analyzed by Kaplan-Meier and Cox proportional hazards regression, and longitudinal eGFR trajectory, analyzed by linear mixed-effects models.ResultsBaseline characteristics were well balanced. Median eGFR was 15.0 and 16.0 mL/min/1.73 m2 and dialysis was required at baseline in 13.3% and 12.5% in the avacopan and rituximab groups, respectively. Avacopan significantly reduced cumulative prednisone exposure at 12 months (1.8 vs 4.3 g; P < 0.001) and shortened time to prednisone discontinuation (3.3 vs 7.3 months; log-rank P < 0.001). On adjusted Cox regression, renal remission did not differ significantly between groups (hazard ratio [HR], 0.74, 95% confidence interval [CI], 0.45–1.21; P = 0.232). Longitudinal eGFR trajectories did not differ significantly between groups. One patient (3.3%) discontinued avacopan because of drug-induced liver injury, with transaminase normalization after discontinuation.ConclusionsIn our cohort of patients with AAV and severe renal involvement treated with rituximab-based induction, avacopan did not improve renal remission rates or longitudinal eGFR trajectory in comparison to glucocorticoids but achieved significant and faster glucocorticoid reduction with an acceptable safety profile. Rapidly progressive glomerulonephritis is one of the most severe manifestations of AAV, and despite advances in induction therapy, a substantial proportion of patients fail to recover kidney function.1,2 Optimal treatment strategies for patients with severe renal impairment remain incompletely defined.3 Avacopan, an oral selective complement C5a receptor inhibitor, was approved by the US food and drug administration for the treatment of AAV in October 2021 after the results of the ADVOCATE trial, which reported noninferiority to prednisone taper at week 26 and superiority in sustained remission at week 52.4 Clinical guidelines subsequently endorsed avacopan primarily as a glucocorticoid-sparing agent.3,5 In the ADVOCATE trial, eGFR change favored avacopan over prednisone, with subgroup and post hoc analyses showing significantly greater eGFR recovery in patients with baseline eGFR < 30 ml/min per 1.73 m2.4,6,7 However, whether avacopan provides additional renal benefit beyond standard glucocorticoid tapering in patients with severe kidney involvement treated with rituximab-based regimens remains uncertain. The ADVOCATE trial excluded patients with an eGFR < 15 ml/min per 1.73 m2 and did not include rituximab maintenance beyond the initial 4 weekly infusions, potentially introducing imbalance between treatment arms and limiting the interpretation of 1-year outcomes.4 Furthermore, the available real-world evidence is limited by small sample sizes and heterogeneity in induction regimens and glucocorticoid exposure, resulting in mixed findings.8, 9, 10, 11, 12, 13 We conducted a retrospective comparative cohort study of patients with active AAV and severe renal involvement, defined by eGFR < 30 ml/min per 1.73 m2, treated with rituximab-based induction with or without avacopan. Our primary aim was to assess whether the addition of avacopan was associated with improved renal outcomes compared with rituximab alone, and secondarily to evaluate its glucocorticoid-sparing effect and safety profile in this severe subgroup. Methods Study Design and Patient Population This was a retrospective comparative cohort study of adults with active AAV and severe renal involvement treated at Mayo Clinic sites in Rochester, MN; Jacksonville, FL; and Scottsdale, AZ, between January 1, 2010, and December 31, 2025. Potentially eligible patients were identified using Mayo Data Explorer, an electronic medical record search tool, through the following 2 parallel queries: (1) patients with a diagnosis of AAV who received rituximab, and (2) patients who received avacopan regardless of diagnosis.14 Records were manually reviewed to confirm eligibility. Patients were eligible if they met all of the following criteria: diagnosis of AAV with positive ANCA serology (myeloperoxidase antineutrophil cytoplasmic antibody myeloperoxidase-ANCA and/or proteinase 3-ANCA); active kidney involvement attributed to AAV; severe acute kidney involvement defined as eGFR < 30 ml/min per 1.73 m2 at induction therapy start; and treatment with rituximab-based induction immunosuppression with or without avacopan. For patients in the avacopan group, avacopan must have been initiated before achieving renal remission. Exclusion criteria included negative ANCA serology, eGFR ≥ 30 ml/min per 1.73 m2 at induction, age < 18 years, chronic kidney disease (CKD) stage 5, kidney dysfunction not attributable to active ANCA glomerulonephritis, positive anti-GBM antibodies, concomitant renal parenchymal disease, drug-induced AAV, and concomitant use of cyclophosphamide, plasma exchange, mycophenolate mofetil, or azathioprine as part of the induction regimen (i.e., nonprotocol immunosuppression). Prior immunosuppressive therapy was permitted under prespecified conditions (Supplementary Table S1). A total of 2 cohorts were defined according to the induction regimen received—rituximab plus avacopan and rituximab without avacopan. In both cohorts, rituximab maintenance therapy was administered at the discretion of the treating physician, given the retrospective nature of the study. Patients who required addition of a nonprotocol immunosuppressive agent during follow-up were censored at that timepoint. Study data were manually collected and managed using REDCap (Research Electronic Data Capture) tool. Data were collected at baseline and at predefined follow-up timepoints at 1, 3, 6, 9, 12, 18, and 24 months. To account for variability in clinical visit timing, allowable time windows were applied: ±2 weeks for the 1-month visit, ±1 month for the 3-month visit, and ±2 months for visits at 6 months or later. When more than 1 visit fell within a given window, the visit closest to the scheduled time point was selected. Definitions, Outcomes and Data Collection The index date was defined as the date of induction therapy initiation. Severe renal involvement at presentation was assessed within a ±7-day window around the index date. End of follow-up was defined as the earliest of the following: (1) completion of the 24-month study period; (2) last available clinical assessment; (3) initiation of an immunosuppressive agent other than rituximab, glucocorticoids, or avacopan; (4) kidney transplantation; (5) death; or (6) loss to follow-up. The 2 coprimary outcomes were time to renal remission and longitudinal eGFR trajectory. Renal remission was defined as stabilization or improvement of kidney function (no decline in eGFR > 25% from baseline and no dialysis requirement) combined with resolution of hematuria (≤ 10 RBC/HPF). A parallel analysis using a ≤ 15% eGFR decline threshold was performed as a sensitivity analysis. Estimated GFR was calculated using the 2021 Chronic Kidney Disease Epidemiology Collaboration creatinine equation; values reported at landmark timepoints reflect directly measured eGFR, as opposed to the model-estimated marginal means derived from the linear mixed-effects model, and the term “eGFR” is used throughout the manuscript for simplicity. For patients requiring dialysis at baseline, the last available predialysis eGFR was used as the baseline value. During follow-up, dialysis-dependent patients were assigned an eGFR of 5 ml/min per 1.73 m2 for eGFR analyses and were considered not to have achieved renal remission or eGFR stabilization at that time point. Regarding missing data, for the renal remission outcome, missing hematuria data were handled as follows: if the eGFR criterion was met but hematuria data were unavailable, the patient was classified as missing for that time point; if the eGFR criterion was not met, the patient was classified as not having achieved remission regardless of hematuria status. For the longitudinal eGFR outcome, all follow-up visits included an eGFR measurement. Secondary outcomes included landmark kidney outcomes at 6 and 12 months (renal remission rates, eGFR, eGFR stabilization, hematuria resolution, and proteinuria), dialysis requirement and recovery (defined as ≥ 60 consecutive days without dialysis), progression to end-stage kidney disease (defined as dialysis dependence for ≥ 90 consecutive days), cumulative immunosuppressive doses at follow-up, time to prednisone dose reduction to ≤ 5 mg/d and to discontinuation, and safety outcomes. Urinary outcomes were not assessed in dialysis-dependent patients. Additional data collected included clinical (i.e., clinical manifestations attributable to AAV requiring intensification of immunosuppressive therapy) and immunological (i.e., ANCA reappearance after prior negativization) relapse rates, as well as longitudinal inflammatory markers (erythrocyte sedimentation rate and C-reactive protein). Kidney biopsy reports were reviewed, and data were extracted when available, including the normal glomerular score15 and Berden classification.16 Statistical Methods Continuous variables are presented as medians (interquartile range [IQR]) and were compared using the Mann-Whitney U test. Categorical variables are presented as n (%) and were compared using Fisher exact test for binary variables and the Pearson χ2test for multinomial variables. Time to prednisone dose reduction to ≤ 5 mg/d and to discontinuation were analyzed using the Kaplan-Meier method, with groups compared using the log-rank test and median time reported with 95% CI. Time to renal remission was analyzed using the Kaplan-Meier method with log-rank testing and Cox proportional hazards regression adjusted for age, ANCA serotype, baseline eGFR, and dialysis requirement at baseline. A histologic model, restricted to patients who underwent kidney biopsy performed during the current episode, additionally included Berden classification and interstitial fibrosis and tubular atrophy (IFTA) category. To account for the limited number of avacopan patients at risk beyond 12 months, Cox models were performed with follow-up truncated at 12 months as the primary analysis and at 24 months as a sensitivity analysis. Time-varying Cox models with avacopan modeled as a time-dependent covariate were additionally performed to account for its variable initiation timing and treatment duration. Results from the time-varying analysis were visualized using Simon-Makuch curves. In addition, a Fine-Gray competing risk regression was performed treating death as a competing event for renal remission. Longitudinal eGFR was analyzed using linear mixed-effects models with follow-up visit as a categorical variable, patient-specific random intercepts, and a treatment group-by-time interaction term, adjusted for age, ANCA serotype, and dialysis requirement at baseline. A histologic model, restricted to patients who underwent kidney biopsy performed during the current episode, additionally included Berden classification and IFTA category. Sensitivity analyses included models with additional random slopes and models with time as a continuous variable with patient-specific random intercepts and slopes. Statistical significance of fixed effects was assessed using Type III F-tests. As a sensitivity analysis for potential confounding by calendar period, the primary Cox and linear mixed-effects models were repeated in a contemporaneous rituximab-alone subgroup (index date on or after September 7, 2021, the index date of the first patient treated with avacopan). Predictors of renal remission at 6- and 12-month landmarks were evaluated by Firth penalized likelihood logistic regression with p-values calculated by likelihood ratio test. Univariable analyses were performed for all candidate variables; histologic variable analyses were restricted to patients with a kidney biopsy available during the current episode. For multivariable models, a marginal type II likelihood ratio test was used, and 2 models were performed: a clinical model, including age, baseline eGFR, dialysis requirement, and treatment group; and a histologic model, additionally including normal glomerular percentage, crescentic Berden class, and moderate or severe tubulointerstitial inflammation, restricted to patients with an available Berden classification during the current episode. Throughout all analyses, rituximab was used as the reference group and avacopan as the group of interest. No formal adjustment for multiplicity was applied across the 2 coprimary outcomes and sensitivity analyses, consistent with the exploratory nature of this retrospective study. A two-sided p value < 0.05 was considered statistically significant. Statistical analyses were performed using R version 4.1.3 and BlueSky Statistics version 10. Ethics The study was approved by the Mayo Clinic Institutional Review Board and conducted in accordance with the Declaration of Helsinki and institutional policies. Results Patient Characteristics Between January 2010 and December 2025, 1539 patients were screened and 150 treatment episodes met the inclusion criteria—30 in the rituximab plus avacopan group and 120 in the rituximab group (Figure 1). For patients treated with avacopan, the earliest index date was September 7, 2021. Median follow-up was 11.1 months (IQR, 5.5–18.6) in the avacopan group and 21.4 months (IQR 6.8–24.1) in the rituximab group. Reasons for the end of follow-up are detailed in Supplementary Table S2.Figure 1Screening flowchart. ANCA, anti-neutrophil cytoplasmic antibody; AAV, ANCA-associated vasculitis; eGFR, estimated glomerular filtration rate; CKD, chronic kidney disease; GN, glomerulonephritis; GBM, glomerular basement membrane; CYC, cyclophosphamide; PLEX, plasma exchange; MMF, mycophenolate mofetil; AZA, azathioprine. ∗Patients may have met more than 1 exclusion criterion; numbers reflect sequential application of criteria in the order listed. ∗∗One patient contributed 2 distinct disease episodes, both meeting inclusion criteria and analyzed independently. Screening flowchart. ANCA, anti-neutrophil cytoplasmic antibody; AAV, ANCA-associated vasculitis; eGFR, estimated glomerular filtration rate; CKD, chronic kidney disease; GN, glomerulonephritis; GBM, glomerular basement membrane; CYC, cyclophosphamide; PLEX, plasma exchange; MMF, mycophenolate mofetil; AZA, azathioprine. ∗Patients may have met more than 1 exclusion criterion; numbers reflect sequential application of criteria in the order listed. ∗∗One patient contributed 2 distinct disease episodes, both meeting inclusion criteria and analyzed independently. Baseline characteristics were balanced between groups and are summarized in Table 1. Median age was 69.5 years in the avacopan group and 72.0 years in the rituximab group, with a predominance of female sex (66.7% vs 55.8%) and myeloperoxidase-ANCA positivity (66.7% vs 65.0%) in both groups; 3 patients (2.0%) were idiopathic double-positive for myeloperoxidase- and proteinase 3-ANCA (1 in the avacopan group and 2 in the rituximab group). Rates of relapsing disease (23.3% vs 18.3%) and extrarenal involvement (50.0% vs 45.0%) were similar. Kidney disease severity at baseline was comparable, with median eGFRs of 15.0 and 16.0 ml/min per 1.73 m2 and dialysis requirement at baseline in 13.3% and 12.5% of patients in the avacopan and rituximab groups, respectively. Kidney biopsy during the current episode was available in 73.3% and 85.8% of patients, with no statistically significant differences in histopathologic characteristics between groups, including Berden classification (P = 0.602), IFTA category (P = 0.807), and normal glomerular percentage (P = 0.073).Table 1Demographic and clinical characteristics at baselineCharacteristicsRituximab + Avacopan (N = 30)Rituximab (N = 120)P-valueDemographicsAge (y), median (IQR)69.5 (63.2–78.5)72.0 (66.0–76.0)0.698Sex (female), n (%)20 (66.7%)67 (55.8%)0.309Race, n (%)0.798White29 (96.7%)114 (95.0%)Asian1 (3.3%)1 (0.8%)Other/Unknown0 (0%)5 (4.2%)ComorbiditiesChronic hypertension, n/N (%)17/30 (56.7%)80/119 (67.2%)0.291Diabetes mellitus, n (%)4 (13.3%)16 (13.3%)1.000Pre-existing CKD, n/N (%)a12/29 (41.4%)44/111 (39.6%)1.000Pre-existing CKD stage, n/N (%)1.000Stage 2 (eGFR 60–89)1/12 (8.3%)4/42 (9.5%)Stage 3 (eGFR 30–59)8/12 (66.7%)29/42 (69.0%)Stage 4 (eGFR 15–29)3/12 (25.0%)9/42 (21.4%)Disease characteristicsANCA-MPO, n (%)b20 (66.7%)78 (65.0%)0.687Vasculitis disease statusRelapsed, n (%)7 (23.3%)22 (18.3%)0.606Prior kidney involvement in relapsing cases, n/N (%)6/7 (85.7%)14/22 (63.6%)0.382Constitutional symptoms, n (%)27 (90.0%)92 (76.7%)0.134Extrarenal involvement, n (%)15 (50.0%)54 (45.0%)0.685Pulmonary, n (%)14 (46.7%)37 (30.8%)0.131ENT, n (%)4 (13.3%)13 (10.8%)0.748Skin, n (%)1 (3.3%)5 (4.2%)1.000Mucous membranes/Ocular, n (%)0 (0%)3 (2.5%)1.000Neurologic, n (%)0 (0%)9 (7.5%)0.205Gastrointestinal, n (%)0 (0%)3 (2.5%)1.000Inflammatory markersESR elevated, n/N (%)21/23 (91.3%)71/78 (91.0%)1.000CRP elevated, n/N (%)23/27 (85.2%)74/94 (78.7%)0.589Kidney parameterseGFR (ml/min per 1.73 m2), median (IQR)15.0 (11.0–19.5)16.0 (10.5–21.7)0.526Dialysis requirement, n (%)4 (13.3%)15 (12.5%)1.000Hematuria > 10 RBC/HPF, n (%)27 (90.0%)111 (92.5%)0.707Hematuria severity, median (IQR)c4 (2.2–7.0)5 (3.0–6.3)0.798Proteinuria g/g Cr or per 24 h, median (IQR)d1.4 (1.1–3.0)1.3 (0.8–2.1)0.269Kidney biopsyKidney biopsy performed (current episode), n (%)22 (73.3%)103 (85.8%)0.135Berden classification, n/N (%)e0.602Focal3/21 (14.3%)28/100 (28.0%)Crescentic5/21 (23.8%)18/100 (18.0%)Mixed10/21 (47.6%)41/100 (41.0%)Sclerotic3/21 (14.3%)13/100 (13.0%)IFTA category, n/N (%)0.807 < 25%11/22 (50.0%)58/101 (57.4%) 25–50%9/22 (40.9%)35/101 (34.7%) > 50%2/22 (9.1%)8/101 (7.9%)Normal glomeruli (%), median (IQR)27.8 (7.3–37.5)34.5 (18.2–50.0)0.073Necrotizing arteritis, n/N (%)1/21 (4.8%)12/100 (12.0%)0.462Moderate/severe tubulointerstitial inflammation, n/N (%)6/21 (28.6%)21/100 (21.0%)0.564ANCA, anti-neutrophil cytoplasmic antibody; CKD,chronic kidney disease; CRP, C-reactive protein; eGFR, estimated glomerular filtration rate; ENT, ear, nose and throat; ESR, erythrocyte sedimentation rate; HPF, high power field; IFTA, interstitial fibrosis and tubular atrophy; IQR, interquartile range; MPO, myeloperoxidase; PR3, proteinase 3; RBC, red blood cells.n/N (%) is shown when data were not available for all patients.aPre-existing CKD defined according to KDIGO criteria, based on the last available data before the current AAV episode and/or previously documented diagnosis in the clinical record. Patients with unknown or missing status were excluded from the denominator (RTX n = 9, RTX + avacopan n = 1).bThe remaining patients had PR3-ANCA, except for 1 patient in the avacopan group and 2 patients in the rituximab group who were double positive for MPO- and PR3-ANCA.cHematuria severity was graded on an ordinal scale—0 = <3 RBC/HPF; 1 = 3–10; 2 = 11–20; 3 = 21–30; 4 = 31–40; 5 = 41–50; 6 = 51–100; 7 = >100 RBC/HPF.dProteinuria was collected in order of preference: 24-h urine collection when available, followed by urine protein-to-creatinine ratio (UPCR) and predicted proteinuria. Data available in 29 of 30 patients treated with avacopan and 106 of 120 patients treated with rituximab.eBerden classification was not assigned in 1 patient in the avacopan group and 3 patients in the rituximab group who had biopsy-confirmed necrotizing arteritis without demonstrable glomerulonephritis. Table 1Demographic and clinical characteristics at baselineCharacteristicsRituximab + Avacopan (N = 30)Rituximab (N = 120)P-valueDemographicsAge (y), median (IQR)69.5 (63.2–78.5)72.0 (66.0–76.0)0.698Sex (female), n (%)20 (66.7%)67 (55.8%)0.309Race, n (%)0.798White29 (96.7%)114 (95.0%)Asian1 (3.3%)1 (0.8%)Other/Unknown0 (0%)5 (4.2%)ComorbiditiesChronic hypertension, n/N (%)17/30 (56.7%)80/119 (67.2%)0.291Diabetes mellitus, n (%)4 (13.3%)16 (13.3%)1.000Pre-existing CKD, n/N (%)a12/29 (41.4%)44/111 (39.6%)1.000Pre-existing CKD stage, n/N (%)1.000Stage 2 (eGFR 60–89)1/12 (8.3%)4/42 (9.5%)Stage 3 (eGFR 30–59)8/12 (66.7%)29/42 (69.0%)Stage 4 (eGFR 15–29)3/12 (25.0%)9/42 (21.4%)Disease characteristicsANCA-MPO, n (%)b20 (66.7%)78 (65.0%)0.687Vasculitis disease statusRelapsed, n (%)7 (23.3%)22 (18.3%)0.606Prior kidney involvement in relapsing cases, n/N (%)6/7 (85.7%)14/22 (63.6%)0.382Constitutional symptoms, n (%)27 (90.0%)92 (76.7%)0.134Extrarenal involvement, n (%)15 (50.0%)54 (45.0%)0.685Pulmonary, n (%)14 (46.7%)37 (30.8%)0.131ENT, n (%)4 (13.3%)13 (10.8%)0.748Skin, n (%)1 (3.3%)5 (4.2%)1.000Mucous membranes/Ocular, n (%)0 (0%)3 (2.5%)1.000Neurologic, n (%)0 (0%)9 (7.5%)0.205Gastrointestinal, n (%)0 (0%)3 (2.5%)1.000Inflammatory markersESR elevated, n/N (%)21/23 (91.3%)71/78 (91.0%)1.000CRP elevated, n/N (%)23/27 (85.2%)74/94 (78.7%)0.589Kidney parameterseGFR (ml/min per 1.73 m2), median (IQR)15.0 (11.0–19.5)16.0 (10.5–21.7)0.526Dialysis requirement, n (%)4 (13.3%)15 (12.5%)1.000Hematuria > 10 RBC/HPF, n (%)27 (90.0%)111 (92.5%)0.707Hematuria severity, median (IQR)c4 (2.2–7.0)5 (3.0–6.3)0.798Proteinuria g/g Cr or per 24 h, median (IQR)d1.4 (1.1–3.0)1.3 (0.8–2.1)0.269Kidney biopsyKidney biopsy performed (current episode), n (%)22 (73.3%)103 (85.8%)0.135Berden classification, n/N (%)e0.602Focal3/21 (14.3%)28/100 (28.0%)Crescentic5/21 (23.8%)18/100 (18.0%)Mixed10/21 (47.6%)41/100 (41.0%)Sclerotic3/21 (14.3%)13/100 (13.0%)IFTA category, n/N (%)0.807 < 25%11/22 (50.0%)58/101 (57.4%) 25–50%9/22 (40.9%)35/101 (34.7%) > 50%2/22 (9.1%)8/101 (7.9%)Normal glomeruli (%), median (IQR)27.8 (7.3–37.5)34.5 (18.2–50.0)0.073Necrotizing arteritis, n/N (%)1/21 (4.8%)12/100 (12.0%)0.462Moderate/severe tubulointerstitial inflammation, n/N (%)6/21 (28.6%)21/100 (21.0%)0.564ANCA, anti-neutrophil cytoplasmic antibody; CKD,chronic kidney disease; CRP, C-reactive protein; eGFR, estimated glomerular filtration rate; ENT, ear, nose and throat; ESR, erythrocyte sedimentation rate; HPF, high power field; IFTA, interstitial fibrosis and tubular atrophy; IQR, interquartile range; MPO, myeloperoxidase; PR3, proteinase 3; RBC, red blood cells.n/N (%) is shown when data were not available for all patients.aPre-existing CKD defined according to KDIGO criteria, based on the last available data before the current AAV episode and/or previously documented diagnosis in the clinical record. Patients with unknown or missing status were excluded from the denominator (RTX n = 9, RTX + avacopan n = 1).bThe remaining patients had PR3-ANCA, except for 1 patient in the avacopan group and 2 patients in the rituximab group who were double positive for MPO- and PR3-ANCA.cHematuria severity was graded on an ordinal scale—0 = <3 RBC/HPF; 1 = 3–10; 2 = 11–20; 3 = 21–30; 4 = 31–40; 5 = 41–50; 6 = 51–100; 7 = >100 RBC/HPF.dProteinuria was collected in order of preference: 24-h urine collection when available, followed by urine protein-to-creatinine ratio (UPCR) and predicted proteinuria. Data available in 29 of 30 patients treated with avacopan and 106 of 120 patients treated with rituximab.eBerden classification was not assigned in 1 patient in the avacopan group and 3 patients in the rituximab group who had biopsy-confirmed necrotizing arteritis without demonstrable glomerulonephritis. Demographic and clinical characteristics at baseline ANCA, anti-neutrophil cytoplasmic antibody; CKD,chronic kidney disease; CRP, C-reactive protein; eGFR, estimated glomerular filtration rate; ENT, ear, nose and throat; ESR, erythrocyte sedimentation rate; HPF, high power field; IFTA, interstitial fibrosis and tubular atrophy; IQR, interquartile range; MPO, myeloperoxidase; PR3, proteinase 3; RBC, red blood cells. n/N (%) is shown when data were not available for all patients. Pre-existing CKD defined according to KDIGO criteria, based on the last available data before the current AAV episode and/or previously documented diagnosis in the clinical record. Patients with unknown or missing status were excluded from the denominator (RTX n = 9, RTX + avacopan n = 1). The remaining patients had PR3-ANCA, except for 1 patient in the avacopan group and 2 patients in the rituximab group who were double positive for MPO- and PR3-ANCA. Hematuria severity was graded on an ordinal scale—0 = <3 RBC/HPF; 1 = 3–10; 2 = 11–20; 3 = 21–30; 4 = 31–40; 5 = 41–50; 6 = 51–100; 7 = >100 RBC/HPF. Proteinuria was collected in order of preference: 24-h urine collection when available, followed by urine protein-to-creatinine ratio (UPCR) and predicted proteinuria. Data available in 29 of 30 patients treated with avacopan and 106 of 120 patients treated with rituximab. Berden classification was not assigned in 1 patient in the avacopan group and 3 patients in the rituximab group who had biopsy-confirmed necrotizing arteritis without demonstrable glomerulonephritis. Immunosuppressive Regimens and Glucocorticoid Exposure In the avacopan group, avacopan was initiated a median of 11 days (IQR, 2.5–26.5) after the index date, with 83.3% of patients initiating treatment within 30 days. Median treatment duration was 6.2 months (IQR, 4.2–12.7), with 36.7% of patients discontinuing it before end of follow-up. Cumulative oral prednisone dose was significantly lower in the avacopan group—1.8 g (IQR, 1.4–2.2) vs 3.8 g (IQR, 2.9–5.2) at 6 months and 1.8 g (IQR, 1.4–2.2) vs 4.3 g (IQR, 3.0–5.6) at 12 months (all P < 0.001). Time to prednisone reduction to ≤ 5 mg/d was significantly shorter in the avacopan group (median 2.2 months, 95% CI, 1.8–3.3 vs. 4.2 months, 95% CI, 3.9–4.9; log-rank P < 0.001; Table 2 and Supplementary Figure S1), as was time to prednisone discontinuation (3.3 months, 95% CI, 2.3–4.5 vs 7.3 months, 95% CI, 5.7–8.3; log-rank P < 0.001; Supplementary Figure S2). Rituximab cumulative dose at 12 months was higher in the avacopan group (median 3.4 g vs 3.0 g, P = 0.040), reflecting higher rates of rituximab maintenance (80.0% vs 51.2%, P = 0.024). More detail on immunosuppressive regimens is presented in Supplementary Table S3.Table 2Oral prednisone exposure by treatment groupPrednisone outcomeRituximab + Avacopan (N = 30)Rituximab (N = 120)P-valueOral prednisone, n (%)28 (93.3%)119 (99.2%)0.102Cumulative dose at 3 mo, median g (IQR)1.6 (1.2–2.1)2.9 (2.3–4.0)< 0.001aCumulative dose at 6 mo, median g (IQR)1.8 (1.4–2.2)3.8 (2.9–5.2)< 0.001aCumulative dose at 12 mo, median g (IQR)1.8 (1.4–2.2)4.3 (3.0–5.6)< 0.001aTime to prednisone ≤ 5 mg/d, mo, median (95% CI)2.2 (1.8–3.3)4.2 (3.9–4.9)< 0.001aTime to prednisone discontinuation, mo, median (95% CI)3.3 (2.3–4.5)7.3 (5.7–8.3)< 0.001aCI, confidence interval; IQR, interquartile range.Median times to dose reduction to ≤5 mg/day and prednisone discontinuation estimated by the Kaplan-Meier method.aStatistically significant (P < 0.05). Table 2Oral prednisone exposure by treatment groupPrednisone outcomeRituximab + Avacopan (N = 30)Rituximab (N = 120)P-valueOral prednisone, n (%)28 (93.3%)119 (99.2%)0.102Cumulative dose at 3 mo, median g (IQR)1.6 (1.2–2.1)2.9 (2.3–4.0)< 0.001aCumulative dose at 6 mo, median g (IQR)1.8 (1.4–2.2)3.8 (2.9–5.2)< 0.001aCumulative dose at 12 mo, median g (IQR)1.8 (1.4–2.2)4.3 (3.0–5.6)< 0.001aTime to prednisone ≤ 5 mg/d, mo, median (95% CI)2.2 (1.8–3.3)4.2 (3.9–4.9)< 0.001aTime to prednisone discontinuation, mo, median (95% CI)3.3 (2.3–4.5)7.3 (5.7–8.3)< 0.001aCI, confidence interval; IQR, interquartile range.Median times to dose reduction to ≤5 mg/day and prednisone discontinuation estimated by the Kaplan-Meier method.aStatistically significant (P < 0.05). Oral prednisone exposure by treatment group CI, confidence interval; IQR, interquartile range. Median times to dose reduction to ≤5 mg/day and prednisone discontinuation estimated by the Kaplan-Meier method. Statistically significant (P < 0.05). Time to Renal Remission Unadjusted Kaplan-Meier analysis showed a nonsignificant trend toward lower cumulative probability of renal remission (25% eGFR definition) in the avacopan group at 6 months (62.5% vs 76.0%) and 12 months (73.3% vs 88.9%; log-rank P = 0.063; Figure 2); results over 24 months are shown in Supplementary Figure S3. Estimated median time to remission was 4.5 months (95% CI, 2.9–not estimable) in the avacopan group and 3.2 months (95% CI, 2.9–3.4) in the rituximab group; among patients who achieved remission, however, the observed median was identical in both groups (2.9 months). Results using the 15% eGFR definition were consistent, with cumulative remission probabilities of 59.1% vs 76.0% at 6 months and 69.4% vs 87.3% at 12 months (log-rank P = 0.026).Figure 2Kaplan–Meier curves for time to renal remission (25% eGFR definition), follow-up truncated at 12 months. Shaded areas represent 95% CI. Kaplan–Meier curves for time to renal remission (25% eGFR definition), follow-up truncated at 12 months. Shaded areas represent 95% CI. To account for potential confounders, an adjusted Cox proportional hazards model was fitted (age, ANCA serotype, baseline eGFR, and dialysis requirement); renal remission (25% eGFR definition) did not differ significantly between groups (HR, 0.74 for avacopan; 95% CI, 0.45–1.21; P = 0.232). Results using the 15% eGFR definition were consistent (P = 0.082). Higher baseline eGFR was independently associated with higher remission rates (HR, 1.05 per mL/min/1.73 m2; 95% CI, 1.01–1.08; P = 0.014). In the biopsy subgroup, results were consistent after additional adjustment for Berden classification and IFTA (HR, 0.68 for avacopan; 95% CI, 0.37–1.24; P = 0.209). Crescentic Berden class was associated with lower remission rates compared to focal class (HR, 0.49; P = 0.053 in the 12-month model; HR, 0.50; P = 0.049 in the 24-month model). In the time-varying Cox analysis, avacopan use was also not significantly associated with renal remission (HR, 0.90 for avacopan; 95% CI, 0.55–1.47; P = 0.663) (Supplementary Figure S4). On Fine-Gray competing risk analysis treating death as a competing event, results remained consistent with the primary Cox model, with no significant association between avacopan and renal remission (subdistribution HR, 0.84; 95% CI, 0.52–1.36; P = 0.479; Gray test P = 0.295). Longitudinal eGFR Trajectories In the primary linear mixed-effects model (categorical time, random intercepts, adjusted for age, ANCA serotype, and baseline dialysis requirement), eGFR improved significantly over time in both groups (P < 0.001 for time effect). There was no statistically significant overall difference in eGFR between groups (P = 0.090), and the treatment group-by-time interaction was not significant (P = 0.279) (Figure 3). In the avacopan group, adjusted estimated marginal mean eGFR rose from 12.5 ml/min per 1.73 m2 (95% CI, 5.2–19.7) to 26.2 ml/min per 1.73 m2 (95% CI, 17.9–34.4); in the rituximab group, it rose from 13.7 ml/min per 1.73 m2 (95% CI, 8.0–19.3) at baseline to 30.8 ml/min per 1.73 m2 (95% CI, 24.9–36.6) at 12 months. Findings were consistent across all 8 sensitivity analyses, regardless of whether time was modeled as categorical or continuous, random slopes were included, or histopathologic covariates were added. Across models, baseline dialysis requirement and age were independently and inversely associated with longitudinal eGFR, whereas ANCA serotype, Berden classification, and IFTA category were not significant.Figure 3Adjusted estimated marginal mean eGFR trajectories by treatment group over 24 months of follow-up. Values derived from a linear mixed-effects model with patient-specific random intercepts, follow-up time modeled as a categorical variable, and adjusted for age, ANCA serotype, and dialysis requirement at baseline. Shaded areas represent 95% CI. Number of patients with available data at 1 month: Rituximab n = 107, Rituximab + avacopan n = 28. EMM, estimated marginal mean. Adjusted estimated marginal mean eGFR trajectories by treatment group over 24 months of follow-up. Values derived from a linear mixed-effects model with patient-specific random intercepts, follow-up time modeled as a categorical variable, and adjusted for age, ANCA serotype, and dialysis requirement at baseline. Shaded areas represent 95% CI. Number of patients with available data at 1 month: Rituximab n = 107, Rituximab + avacopan n = 28. EMM, estimated marginal mean. In the sensitivity analysis comparing contemporaneous cohorts, baseline characteristics were balanced between groups, except for a higher prevalence of elevated erythrocyte sedimentation rate in the avacopan group (70.0% vs 40.0%; P = 0.024). Results were consistent with the full-cohort analysis—adjusted Cox HR 0.79 (95% CI, 0.43–1.47; P = 0.461; log-rank P = 0.378), and no significant group effect (P = 0.107) or group-by-time interaction (P = 0.383) for longitudinal eGFR. Outcomes at Landmark Time Points At the 6-month follow-up visit, renal remission (25% eGFR definition) was achieved by 59.3% of patients treated with avacopan vs 86.9% of those treated with rituximab (P = 0.004). This difference was no longer apparent at 12 months (80.0% vs. 82.6%; P = 0.726). eGFR at 6 months was lower in the avacopan group (median 24.5 vs 31.0 ml/min per 1.73 m2; P = 0.018), with a correspondingly smaller gain from baseline (median ΔeGFR +10.5 vs +16.0 ml/min per 1.73 m2; P = 0.017). By 12 months, eGFR and gain from baseline were similar between groups (median 26.0 vs 33.0 ml/min per 1.73 m2, P = 0.583; median ΔeGFR +15.5 vs +15.5 ml/min per 1.73 m2, P = 0.684). Resolution of hematuria (≤ 10 RBC/HPF) at 6 months was also less frequent in the avacopan group (77.3% vs 93.6%; P = 0.039), whereas rates at 12 months were similar (100% vs. 93.5%; P = 1.000) (Supplementary Table S4). Proteinuria did not differ significantly at either time point. Full landmark data are presented in Table 3. Inflammatory markers (C-reactive protein and erythrocyte sedimentation rate) during follow-up are described in Supplementary Table S5.Table 3Kidney outcomes at landmark time points of 6 and 12 months of follow-upOutcomeRituximab + Avacopan (N = 30)Rituximab (N = 120)P-value6 moRenal remission (25% eGFR definition), n/N (%)16/27 (59.3%)73/84 (86.9%)0.004bRenal remission (15% eGFR definition), n/N (%)15/27 (55.6%)71/84 (84.5%)0.003beGFR, ml/min per 1.73 m2, median (IQR)24.5 (11.8–33.8)31.0 (25.0–41.4)0.018bChange in eGFR from baseline, ml/min per 1.73 m2, median (IQR)10.5 (−0.2–16.5)16.0 (8.2–23.2)0.017beGFR > 30 ml/min per 1.73 m2, n/N (%)10/28 (35.7%)49/90 (54.4%)0.129eGFR stabilization (< 25% decline from baseline), n/N (%)22/28 (78.6%)84/90 (93.3%)0.035beGFR stabilization (< 15% decline from baseline), n/N (%)21/28 (75.0%)82/90 (91.1%)0.046bDialysis requirement, n/N (%)5/28 (17.9%)6/90 (6.7%)0.128Hematuria ≤ 10 RBC/HPF, n/N (%)17/22 (77.3%)73/78 (93.6%)0.039bProteinuria, g/g Cr or per 24 h, median (IQR)a0.9 (0.4–2.2)0.6 (0.3–1.2)0.35812 moRenal remission (25% eGFR definition), n/N (%)12/15 (80.0%)57/69 (82.6%)0.726Renal remission (15% eGFR definition), n/N (%)12/15 (80.0%)57/69 (82.6%)0.726eGFR, ml/min per 1.73 m2, median (IQR)26.0 (21.0–38.8)33.0 (22.0–44.0)0.583Change in eGFR from baseline, ml/min per 1.73 m2, median (IQR)15.5 (4.2–22.0)15.5 (6.2–24.8)0.684eGFR > 30 ml/min per 1.73 m2, n/N (%)7/16 (43.8%)44/74 (59.5%)0.278eGFR stabilization (< 25% decline from baseline), n/N (%)13/16 (81.2%)66/74 (89.2%)0.405eGFR stabilization (< 15% decline from baseline), n/N (%)13/16 (81.2%)66/74 (89.2%)0.405Dialysis requirement, n/N (%)3/16 (18.8%)7/74 (9.5%)0.375Hematuria ≤ 10 RBC/HPF, n/N (%)12/12 (100.0%)58/62 (93.5%)1.000Proteinuria, g/g Cr or per 24 h, median (IQR)a0.9 (0.4–1.2)0.4 (0.2–1.2)0.265eGFR, estimated glomerular filtration rate; HPF, high power field; IQR, interquartile range; RBC, red blood cells.aProteinuria denominators—6 mo Avacopan n = 18, Rituximab n = 56; 12 mo Avacopan n = 10, Rituximab n = 44.bStatistically significant (P < 0.05). Table 3Kidney outcomes at landmark time points of 6 and 12 months of follow-upOutcomeRituximab + Avacopan (N = 30)Rituximab (N = 120)P-value6 moRenal remission (25% eGFR definition), n/N (%)16/27 (59.3%)73/84 (86.9%)0.004bRenal remission (15% eGFR definition), n/N (%)15/27 (55.6%)71/84 (84.5%)0.003beGFR, ml/min per 1.73 m2, median (IQR)24.5 (11.8–33.8)31.0 (25.0–41.4)0.018bChange in eGFR from baseline, ml/min per 1.73 m2, median (IQR)10.5 (−0.2–16.5)16.0 (8.2–23.2)0.017beGFR > 30 ml/min per 1.73 m2, n/N (%)10/28 (35.7%)49/90 (54.4%)0.129eGFR stabilization (< 25% decline from baseline), n/N (%)22/28 (78.6%)84/90 (93.3%)0.035beGFR stabilization (< 15% decline from baseline), n/N (%)21/28 (75.0%)82/90 (91.1%)0.046bDialysis requirement, n/N (%)5/28 (17.9%)6/90 (6.7%)0.128Hematuria ≤ 10 RBC/HPF, n/N (%)17/22 (77.3%)73/78 (93.6%)0.039bProteinuria, g/g Cr or per 24 h, median (IQR)a0.9 (0.4–2.2)0.6 (0.3–1.2)0.35812 moRenal remission (25% eGFR definition), n/N (%)12/15 (80.0%)57/69 (82.6%)0.726Renal remission (15% eGFR definition), n/N (%)12/15 (80.0%)57/69 (82.6%)0.726eGFR, ml/min per 1.73 m2, median (IQR)26.0 (21.0–38.8)33.0 (22.0–44.0)0.583Change in eGFR from baseline, ml/min per 1.73 m2, median (IQR)15.5 (4.2–22.0)15.5 (6.2–24.8)0.684eGFR > 30 ml/min per 1.73 m2, n/N (%)7/16 (43.8%)44/74 (59.5%)0.278eGFR stabilization (< 25% decline from baseline), n/N (%)13/16 (81.2%)66/74 (89.2%)0.405eGFR stabilization (< 15% decline from baseline), n/N (%)13/16 (81.2%)66/74 (89.2%)0.405Dialysis requirement, n/N (%)3/16 (18.8%)7/74 (9.5%)0.375Hematuria ≤ 10 RBC/HPF, n/N (%)12/12 (100.0%)58/62 (93.5%)1.000Proteinuria, g/g Cr or per 24 h, median (IQR)a0.9 (0.4–1.2)0.4 (0.2–1.2)0.265eGFR, estimated glomerular filtration rate; HPF, high power field; IQR, interquartile range; RBC, red blood cells.aProteinuria denominators—6 mo Avacopan n = 18, Rituximab n = 56; 12 mo Avacopan n = 10, Rituximab n = 44.bStatistically significant (P < 0.05). Kidney outcomes at landmark time points of 6 and 12 months of follow-up eGFR, estimated glomerular filtration rate; HPF, high power field; IQR, interquartile range; RBC, red blood cells. Proteinuria denominators—6 mo Avacopan n = 18, Rituximab n = 56; 12 mo Avacopan n = 10, Rituximab n = 44. Statistically significant (P < 0.05). In the univariable analysis, lower baseline eGFR, dialysis requirement at baseline, avacopan treatment, lower normal glomerular percentage, crescentic Berden class, and moderate or severe tubulointerstitial inflammation were associated with lower odds of renal remission at 6 months (Supplementary Table S6). In the multivariable logistic regression, avacopan treatment was independently associated with lower odds of renal remission at 6 months (odds ratio [OR], 0.19; 95% CI, 0.06–0.53; P = 0.002; Supplementary Table S7) after adjustment for age, baseline eGFR, and dialysis requirement; this association was no longer significant at 12 months (OR, 0.82; 95% CI, 0.22–3.64; P = 0.778). In the histologic multivariable model, avacopan treatment (OR, 0.10; 95% CI, 0.02–0.38; P < 0.001) and crescentic Berden class (OR, 0.14; 95% CI, 0.03–0.61; P = 0.009) were independent predictors of lower renal remission at 6 months. Dialysis Requirement, Dialysis Recovery, and Progression to End-Stage Kidney Disease Dialysis was required at any point during the study period in 9 of 30 patients treated with avacopan (30.0%) and 21 of 120 patients treated with rituximab (17.5%), including 4 and 15 patients who required dialysis at baseline, respectively (Supplementary Table S8). Among those who required dialysis, recovery was achieved in 3 of 9 patients treated with avacopan (33.3%) and 8 of 21 patients treated with rituximab (38.1%; P = 1.000). Progression to end-stage kidney disease occurred in 6 of 30 patients treated with avacopan (20.0%) and 10 of 120 patients treated with rituximab (8.3%; OR, 2.73, 95% CI, 0.74–9.27; P = 0.093). A composite outcome of end-stage kidney disease or death while on dialysis was observed in 6 patients treated with avacopan (20.0%) and 12 patients treated with rituximab (10.0%; OR, 2.24; 95% CI, 0.62–7.25; P = 0.204). These findings should be interpreted in light of the substantially shorter follow-up in the avacopan group, which limits the comparability of cumulative event rates between cohorts. Relapses Clinical relapse rates were similar between groups (10.0% avacopan vs 9.2% rituximab), with renal involvement observed in 66.7% and 81.8% of relapsing patients, respectively (Supplementary Table S9). ANCA negativity was achieved at least once during follow-up in 20.0% and 22.5% of patients, with comparable immunologic relapse rates among seroconverters (33.3% vs 37.0%). Safety Infections requiring hospitalization occurred in 13.3% of patients treated with avacopan and 12.6% of patients treated with rituximab. When adjusted for individual follow-up time, infection rates were similar between groups (Rituximab + Avacopan: 12.54 per 100 patient-years; 95% CI, 3.42–32.11; Rituximab: 11.14 per 100 patient-years; 95% CI, 6.60–17.61; rate ratio, 1.13; 95% CI, 0.28–3.42; P = 0.775). Severe hypogammaglobulinemia (IgG < 300 mg/dl) was identified in 2.9% of patients treated with rituximab and 0% of patients treated with avacopan; however, IgG monitoring was available in only a minority of patients during follow-up (22/30 in the avacopan group and 34/120 in the rituximab group), limiting interpretation. Drug-induced liver injury (alanine aminotransferase [ALT] and/or aspartate aminotransferase [AST] ≥3× upper limit of normal) occurred in 3 patients treated with rituximab (2.5%) and 1 patient treated with avacopan (3.3%). All 3 cases in the rituximab group were severe and unrelated to immunosuppression. In the avacopan group, 1 patient developed liver injury after 3.4 months of therapy (peak ALT 4.2× upper limit of normal, AST 3.4× upper limit of normal), leading to avacopan discontinuation; hematuria had already resolved at that time point, and transaminase levels normalized thereafter. An additional patient experienced a transaminase elevation below the prespecified hepatotoxicity threshold (ALT 2.9× upper limit of normal) at 0.9 months, also leading to early discontinuation; follow-up laboratory data were not available. Both patients were White. Death occurred in 21 of 120 patients treated with rituximab (17.5%) and 1 of 30 patients treated with avacopan (3.3%). This numerical difference in mortality should be interpreted with caution given the longer follow-up in the rituximab cohort. Safety outcomes are summarized in Supplementary Table S10. Discussion In this cohort of patients with active AAV and severe renal impairment (eGFR < 30 ml/min per 1.73 m2) treated with rituximab-based induction, the addition of avacopan was associated with significantly lower glucocorticoid exposure but no significant difference in cumulative renal remission or longitudinal eGFR trajectory. The differences in renal outcomes at landmark time points should be interpreted with caution, given that those analyses were limited to patients with available data at each time point. Of note, the avacopan group showed a statistically nonsignificant trend toward a less favorable histopathologic profile at baseline—including less focal and more crescentic Berden class, a lower normal glomerular percentage, and a higher prevalence of moderate or severe tubulointerstitial inflammation—which may have contributed to the observed 6-month landmark differences and early divergence in longitudinal eGFR trajectories. These findings confirm the glucocorticoid-sparing effect of avacopan consistently reported in the literature, while raising questions about whether its addition confers meaningful renal benefit beyond glucocorticoid reduction in this particularly severe subgroup—a population largely underrepresented in prior studies, including the ADVOCATE trial.4,8, 9, 10, 11, 12, 13 This is especially relevant given the high cost of avacopan and its potential for serious adverse effects, including hepatotoxicity. Our renal outcome results somewhat contrast with those reported in the ADVOCATE trial and subsequent post hoc analyses. In the primary publication, eGFR recovery was significantly greater with avacopan in patients with baseline eGFR < 30 ml/min per 1.73 m2 at both week 26 and week 52 (least-squares mean change 13.7 vs. 8.2 ml/min per 1.73 m2 at week 52).4 Post hoc analyses confirmed significantly greater eGFR gains in the eGFR 15–29 subgroup at week 52 (13.5 vs. 7.8 ml/min per 1.73 m2),6 and in the eGFR 15–20 subgroup at week 26 (11.9 vs. 6.1) and week 52 (16.1 vs 7.7 ml/min per 1.73 m2),7 although when restricted to patients treated with rituximab, the eGFR advantage in the < 30 subgroup was nonsignificant (8.7 vs. 6.6 ml/min per 1.73 m2 at week 52).17 Avacopan showed a nonsignificant trend toward higher remission rates in the eGFR 15–29 subgroup at week 26 (70.6% vs. 68.1%) and week 52 (68.6% vs. 59.6%); remission rates for this subgroup were not separately reported among patients treated with rituximab.6,17 Several methodological differences may explain the divergence in eGFR outcomes between our cohort and the ADVOCATE post hoc analyses. First, our cohort included patients with eGFR < 15 ml/min per 1.73 m2, a population excluded from the ADVOCATE trial.4 Second, rituximab was used exclusively as induction therapy in our study, whereas in ADVOCATE induction regimens were heterogeneous; notably, in the subgroup with eGFR between 15 and 20 ml/min per 1.73 m2, 56% of patients in the avacopan arm received cyclophosphamide.7 Third, rituximab maintenance was administered in a substantial proportion of our cohort (80.0% in the avacopan group and 51.2% in the rituximab group), whereas maintenance rituximab was not used in ADVOCATE.4 In addition, the renal remission definition used in our study differs from that of ADVOCATE. We did not incorporate Birmingham Vasculitis Activity Score, as it is difficult to apply retrospectively and includes proteinuria, which may reflect chronic kidney damage rather than active glomerular inflammation. Glucocorticoid discontinuation was also not included as a remission criterion, given the uncertainty regarding the optimal timing of glucocorticoid withdrawal in this more severe patient population. One hypothesis is that the broader immunosuppressive effect of glucocorticoids may confer greater early benefit in patients with more severe renal involvement, in whom rapid suppression of glomerular inflammation is critical.18, 19, 20, 21 Avacopan selectively blocks the C5a–C5aR1 axis, which, whereas central to neutrophil-mediated glomerular injury in AAV, represents only 1 component of the inflammatory cascade.22, 23, 24 This selective mechanism may be insufficient during the early phase of treatment, before the full clinical effect of rituximab-induced B-cell depletion has been established.25 However, this assumption cannot be confirmed or excluded in the absence of a randomized controlled design. Real-world studies without a comparative cohort have generally reported favorable renal outcomes with avacopan8,10, 11, 12; however, studies including comparator groups have not shown superior renal outcomes compared with standard therapy.9,13 Notably, a study with a design most similar to ours—including patients with severe renal involvement (eGFR < 30 ml/min per 1.73 m2), available kidney biopsy data, and a glucocorticoid-based comparator cohort— found similar kidney response rates at 6 and 12 months between avacopan and glucocorticoid-based regimens after adjustment for histopathologic findings, regardless of the eGFR end point used; however, hematuria outcomes were not reported, limiting the assessment of renal remission.13 This study also contributes prognostic data on histopathologic predictors of renal recovery. Crescentic Berden class was associated with lower probability of renal remission, suggesting that a higher burden of active glomerular inflammation poses a greater barrier to early recovery. Whether more aggressive or alternative induction strategies may improve outcomes in this histologic subgroup remains an important unanswered question. Overall, we observed an acceptable safety profile with avacopan use. Drug-induced liver injury has emerged as an important safety concern with avacopan, particularly in Asian patients26; however, our cohort had limited representation of non-White patients, which may partly explain the lower hepatotoxicity rates observed. This study has several important strengths. First, unlike most prior real-world reports, it included only patients treated with rituximab-based induction, excluding cyclophosphamide and plasma exchange, allowing a more homogeneous assessment of outcomes in this specific therapeutic context. Second, our cohort extended beyond the ADVOCATE trial population by including patients with eGFR < 15 ml/min per 1.73 m2, incorporating kidney biopsy data in the majority of patients, and including rituximab maintenance in a substantial proportion. Finally, by restricting inclusion to patients who initiated avacopan before achieving renal remission, our study provides a more informative assessment of its potential contribution to renal recovery. The main limitations of this study are its retrospective design and relatively small sample size, particularly in the avacopan group. However, to our knowledge, this represents the largest reported cohort of patients with AAV and severe renal involvement treated exclusively with rituximab-based induction, with or without avacopan. In addition, although baseline characteristics were well balanced between groups, the later treatment period of the avacopan cohort may have introduced bias related to evolving treatment practices, potentially contributing to the observed differences in cumulative glucocorticoid and rituximab exposure between groups. Regarding the latter, this may also reflect a tendency to prescribe maintenance rituximab preferentially to patients with worse early renal recovery. The regulatory landscape surrounding avacopan has evolved substantially during the preparation of this manuscript. The ADVOCATE trial was retracted by the New England Journal of Medicine in June 2026,27 after an food and drug administration investigation that identified manipulation of primary end point data. Subsequently, the food and drug administration has proposed withdrawal of avacopan's marketing authorization,28 with a final decision pending, and the European Medicines Agency has recommended revoking its authorization in the European Union.29 Before these developments, secondary end points and post hoc analyses4,6,7,17 suggested that patients with severe renal involvement might derive benefit from avacopan, although this remained incompletely characterized, particularly in those receiving rituximab-based induction. Our real-world findings add evidence in this specific population, showing that while adding avacopan reduces glucocorticoid exposure, it does not improve kidney outcomes. Conclusion In patients with AAV and severe renal involvement treated with rituximab-based induction, the addition of avacopan was associated with significantly lower glucocorticoid exposure and faster discontinuation but did not improve renal remission rates or longitudinal eGFR trajectory compared with rituximab alone. These findings raise questions about the additive renal benefit of avacopan in this particularly severe subgroup and highlight the need for prospective randomized trials to define its optimal role in patients with severe AAV. Disclosure US reports grants/contracts from Amgen, AstraZeneca, Genentech, GSK, NorthStar Medical Radioisotopes, Novartis, and NS-Pharma, and consulting fees from Amgen, AstraZeneca, Argenx, CSL Vifor, Boehringer Ingelheim, Melodia Therapeutics, and Novartis. FF reports unrestricted grants from Amgen and Genentech. LZ reports unrestricted grants from Genentech. The remaining authors declare no competing interests. Data Availability Statement The data may be shared by contacting the corresponding author and will be available as deidentified participant data. Data will be shared after approval of the proposal by the corresponding author and a signed data access agreement. Author Contributions CJ, IH, and LZ conceptualized the study. CJ, AC, CGM, JAM, TL, GMB, MJV, CT, and JMA curated the data. CJ performed the formal analysis. CJ, IH, and LZ drafted the manuscript. CJ, IH, AC, MJV, SF, US, FF, and LZ reviewed and edited the manuscript.