Long-term follow-up of a combined rituximab and cyclophosphamide regimen in renal anti-neutrophil cytoplasm antibody-associated vasculitis

Abstract

Background: Current guidelines advise that rituximab or cyclophosphamide should be used for the treatment of organ-threatening disease in anti-neutrophil cytoplasm antibody (ANCA)-associated vasculitis (AAV), although few studies have examined the efficacy and safety of these agents in combination.

Methods: We conducted a single-centre cohort study of 66 patients treated with a combination of oral corticosteroids, rituximab and low-dose pulsed intravenous cyclophosphamide followed by a maintenance regimen of azathioprine and tapered steroid for the treatment of biopsy-proven renal involvement in AAV. Patients were followed for a median of 56 months. Case-control analysis with 198 propensity-matched cases from European Vasculitis Study Group (EUVAS) trials compared long-term differences in relapse-free, renal and patient survival.

Results: At entry, the median Birmingham Vasculitis Activity Score (BVAS) was 19 and estimated glomerular filtration rate was 25 mL/min. Cumulative doses of rituximab, cyclophosphamide and corticosteroids were 2, 3 and 4.2 g, respectively, at 6 months. A total of 94% of patients achieved disease remission by 6 months (BVAS < 0) and patient and renal survival were 84 and 95%, respectively, at 5 years. A total of 84% achieved ANCA-negative status and 57% remained B cell deplete at 2 years, which was associated with low rates of major relapse (15% at 5 years). The serious infection rate during long-term follow-up was 1.24 per 10 patient-years. Treatment with this regimen was associated with a reduced risk of death {hazard ratio [HR] 0.29 [95% confidence interval (CI) 0.125-0.675], P = 0.004}, progression to end-stage renal disease (ESRD) [HR 0.20 (95% CI 0.06-0.65), P = 0.007] and relapse [HR 0.49 (95% CI 0.25-0.97), P = 0.04] compared with propensity-matched patients enrolled in EUVAS trials.

Conclusions: This regimen is potentially superior to current standards of care, and controlled studies are warranted to establish the utility of combination drug approaches in the treatment of AAV.

FIGURE 1

Early disease response at 6 and 12 months. (A) BVAS at 0 and 6 months. (B) Oral prednisolone dose (mg/day) during the first year of therapy. Fourteen patients received pulsed intravenous methylprednisolone prior to referral to our centre and in these cases the initial dose of oral prednisolone was reduced. (C) Sequential serum CRP (mg/L). D Serum creatinine (μmol/L). (E) Peripheral B lymphocyte count (cells/μL). (F) ANCA titres as determined by antigen specific assay (1F) at 3, 6 and 12 months. Box and whisker plots represent median (IQR) and range measurements.

FIGURE 2

Long-term outcomes during 5-year follow-up. (A) eGRF (mL/min) at annual intervals during 5-year follow-up. Data censored at point of progression to end-stage renal disease for four patients. Unadjusted Kaplan–Meier survival functions describing (B) overall and ESRD-free survival (the latter censored for death) and relapse-free survival censored for death in (C) MPO-ANCA- and (D) PR3-ANCA-positive patients during the 5-year follow-up. (E) Long-term steroid exposure in the cohort during the 5-year follow-up (censored for relapse). Box and whisker plots represent median (IQR) and range measurements.

FIGURE 3

B-cell and ANCA kinetics and their relationship to relapse during the 5-year follow-up. Kaplan–Meier functions describing (A) the proportion of patients who remained free of peripheral B cells (<10 cells/uL) and (B) the time taken to achieve ANCA-negative testing as determined by antigen-specific assay during the 5-year follow-up. Kaplan–Meier functions describing the incidence of first major or minor relapse in (C) patients who had a return of peripheral B cells (>10 cells/uL) versus those who did not and (D) patients who remained ANCA negative versus those who had a return of ANCA in those patients who became ANCA negative after initial treatment during the 5-year follow-up. (E) Heatmap describing sequential ANCA measurements in the 62 patients who survived beyond 6 months and who were ANCA positive by antigen-specific assay during long-term follow-up. The occurrence of major relapse is indicated by filled circles. (F) Composite Kaplan–Meier curve describing the sequence of B-cell repopulation, return of circulating ANCA and the occurrence of relapse for the entire cohort over the 5-year follow-up. The red plot shows the time to achieve ANCA-negative testing, the blue plot describes the time for return of peripheral B cells, the grey plot describes the time for return of circulating ANCA in those patients who had previously achieved ANCA-negative testing and the black plot describes the time to the first minor or major relapse.

FIGURE 4

Case–control analysis with propensity-matched patients from EUVAS trials. Cox proportional hazards regression models. (A) The adjusted all-cause mortality, (B) ESRD and (C) relapse risk between the EUVAS cohort and combined treatment (‘CycLowVas’) cohort. The covariates in the analysis include age, sex, ANCA specificity, entry eGFR, entry BVAS and dose of cyclophosphamide.