Background Following remission induction with rituximab (RTX) in ANCA-associated vasculitis (AAV), duration of response is highly variable [1,2]. Repeat cycles may be fixed-interval or on clinical relapse. The former may lead to hypogammaglobulinaemia and severe infection leading to discontinuation in up to 1/3 patients . However, the latter may allow severe flares. Hence biomarkers that can guide best time to retreat are needed.
Objectives To investigate whether B cell biomarkers may have value in guiding retreatment decisions in patients with response to RTX.
Methods We conducted an observational study of patients with AAV treated after a cycle of RTX in a single centre between January 2006 and September 2013 – a total of 162 patient-years follow up. Each cycle consisted of 2×1000mg infusions repeated on clinical relapse. B cells were measured using highly sensitive flow cytometry (HSFC) as previously described  at baseline, week 2, week 6 and week 26 after RTX. 12 healthy controls were used to measure normal B cell subsets and 26-week B cell subsets were compared to our previously published data in RTX-treated RA and SLE.
Results 35 patients (18 female, median age 56) received a first cycle (C1). ANCA positivity were PR3: 25, MPO: 5 and immunofluorescence only: 4. 34 (97%) patients were resistant to conventional immunosuppressants, including relapse or non-response after cyclophosphamide in 31 (89%). 22 (63%) continued oral immunosuppressants.
Prior to RTX, active AAV was characterised by naïve- and memory-B lymphopenia compared to healthy controls at baseline. This dysregulation was more marked in patients with raised CRP. Overall response rate to C1 was 94%. Median time to relapse was 81 weeks (range 31-209). Naïve, memory and plasmablast numbers at 26 weeks were significantly lower in AAV compared to RA or SLE (all p<0.05).
We found a significant association between repopulation of naïve B cells at 6 months and later clinical relapse (more than 18 months); (p=0.010) as shown in Figure 1. No association between memory B cell (p=0.399) or plasmablast repopulation (p=0.262) and relapse (in contrast to our finding in SLE ). Patients with undetectable (<0.0001×109 cells/L) naïve B cells at 6 months had lower naïve B cells and significantly higher CRP (p=0.015) at baseline, but no difference in age, disease duration, Birmingham Vasculitis Activity Score (BVAS) or ANCA titres. Of patients with detectable naïve B cells, 100% did not experience a relapse within the next 26 weeks and 83% within the next 52 weeks. The same trend followed subsequent cycles; for cycle 2 and 3 (n=12 and n=8 respectively, 100% did not relapse in the following 26 weeks).
Conclusions Naïve B-lymphopenia is a B cell specific marker of disease activity in AAV. Evaluation of naïve B cell numbers in very early repopulation using HSFC may have value in guiding retreatment decisions and warrants validation in larger cohorts.
Dass et al EULAR 2011.
Md Yusof et al EULAR 2013.
Besada et al Rheumatology 2013.
Dass et al Arthritis Rheum 2008.
Vital et al Arthritis Rheum 2011.
Disclosure of Interest None declared