Abstract BACKGROUND AND AIMS Rituximab is increasingly used to induce and maintain remission in ANCA-associated vasculitis (AAV). However, it remains unclear what the optimal dosing regimen is, especially in the setting of maintenance. Clinical trials have shown that vasculitis flares are unlikely as long as B-cell depletion in peripheral blood is sustained, with some evidence supporting a rituximab redosing strategy guided by B-cell-count monitoring. A better understanding of B-cell kinetics following rituximab treatment can facilitate personalized treatment approaches. This study aims to (1) describe B-cell repopulation in AAV patients treated with rituximab induction, and (2) identify clinical features predictive of B-cell repopulation. METHOD This is a single-centre, retrospective study, including AAV patients on no baseline immunosuppression who were treated with rituximab induction between December 2009 and March 2018 in the Nephrology Unit of Spedali Civili, Brescia. Only patients with available longitudinal B-cell monitoring for at least 6 months following rituximab were considered. B-cell counts were determined as CD19 + cells in peripheral blood using flow cytometry (Beckman Coulter Navios©). B-cell repopulation following rituximab was defined as CD19 + cell count >10 cells/μL, with a documented CD19 + count ≤ 10 cells/μL within the preceding year. For each patient, CD19 + cell counts were followed up to the earliest event among the following: (1) B-cell repopulation, (2) retreatment with rituximab or other induction agents, or (3) most recent observation available. Estimated glomerular filtration rate (eGFR) was calculated using the CKD-EPI equation. Flares were defined as active disease retreated with rituximab or cyclophosphamide and classified as major if one or more major items of the BVAS score v3 were present. Predictors of time to B-cell repopulation were assessed using uni- and multivariate analysis with log-rank test and Cox proportional hazards regression, respectively. RESULTS The study cohort included 64 patients. The main clinical characteristics are summarized in Table 1. During a median follow-up of 33.4 months (range 7.5–101.7 months), B-cell repopulation was observed in 36 patients (56.2%). Median time from rituximab induction to B-cell repopulation was 39.8 months [95% confidence interval (95% CI) 33.8–69.7]. Gender, ANCA subtype, rituximab dose, steroid pulses, plasma exchange and methotrexate/azathioprine maintenance were not significantly associated with time to B-cell repopulation. In contrast, B-cell repopulation was significantly slower in patients who were older (median time to repopulation 51.6 months in patients ≥ mean age of 66.8 years old versus 33.9 months in < mean age; P = 0.044), who had microscopic polyangiitis (median 51.6 months), as opposed to granulomatosis with polyangiitis (median 33.9 months; P = 0.012) and who had lower eGFR at the time of rituximab (69.7 months in eGFR < median of 18.3 mL/min/1.73 m2 versus 28.4 months in eGFR ≥ median; P = 0.001, Figure 1). Of these three predictors, only eGFR remained significantly associated with time to B-cell repopulation in a multivariate Cox regression model (hazard ratio for B-cell repopulation in eGFR ≥ median: 2.74, 95% CI 1.31–5.72). During sustained B-cell depletion, six vasculitis flares were observed, of which one was major (rise in creatinine > 30%) and the rest were minor (n = 4 worsening of urine abnormalities, n = 1 retro-orbital pain). CONCLUSION A substantial proportion of AAV patients treated with a single course of rituximab shows sustained B-cell depletion. Renal function emerged as the main independent predictor of B-cell repopulation, with delayed B-cell repopulation in people with kidney impairment. Only one major vasculitis flare was observed during sustained B-cell depletion. These data support the idea that longitudinal monitoring of B-cell counts can be a useful biomarker to guide rituximab dosing for maintenance in AAV, especially in patients with kidney impairment.
Read full abstract