AB0370 UTILITY OF CRP AND ESR IN THE DIAGNOSIS OF GIANT CELL ARTERITIS RELAPSE IN A PHASE 2 TRIAL OF MAVRILIMUMAB

Abstract

Background:No universally accepted definition of flare currently exists in giant cell arteritis (GCA). Although relapses are defined mostly on clinical grounds (recurrence of GCA-related signs/symptoms), C-reactive protein (CRP) and erythrocyte sedimentation rate (ESR) help clinicians assess disease activity. In fact, >70% of patients on glucocorticoids (GCs) alone have increased CRP or ESR when the disease is active. In contrast, tocilizumab, given its IL-6-blockade effect in the liver, rapidly reduces CRP and ESR levels, rendering them unreliable for disease activity monitoring. Mavrilimumab – a GM-CSF receptor α inhibitor with demonstrated efficacy in a Phase 2 GCA trial1 – downregulates inflammation upstream of IL-6. We hypothesized that mavrilimumab would not interfere with the utility of CRP and ESR in monitoring disease activity and in identifying GCA relapse.Objectives:To analyze the relationship between CRP/ESR and clinical disease activity in GCA patients treated with mavrilimumab.Methods:New-onset and relapsing GCA patients with active disease were recruited. GC-induced remission (no GCA symptoms and CRP <1 mg/dL or ESR <20 mm/hr) was required by baseline. Patients were randomized 3:2 to mavrilimumab 150 mg or placebo subcutaneously every 2 weeks plus a protocol-defined 26-week prednisone taper. The primary efficacy endpoint was time to relapse by Week 26. Relapse (adjudicated) was defined as recurrent GCA-related signs/symptoms, including new/worsening vasculitis on imaging, concurrent with CRP ≥1 mg/dL and/or ESR ≥30 mm/hr. CRP and ESR were also measured periodically during the trial.This post hoc analysis assessed the association of recurrent GCA-related signs/symptoms with concurrent CRP or ESR elevation post-randomization by treatment arm. We also assessed the proportion of patients with CRP or ESR elevation without GCA-related signs/symptoms up to Week 26.Results:Seventy patients were enrolled (mavrilimumab, N=42; placebo, N=28). The association of CRP or ESR elevation with unequivocal GCA-related signs/symptoms post-randomization was consistent regardless of treatment arm: 8/8 in the mavrilimumab group and 13/13 in the placebo group (Table 1). During relapse, median (range) CRP was 1.8 (1.4 – 8.4) mg/dL (mavrilimumab group) and 1.8 (1.1 – 9.0) mg/dL (placebo group). Corresponding ESR values were 39.5 (30 – 102) mm/hr (mavrilimumab group) and 49 (31 – 101) mm/hr (placebo group). Four mavrilimumab recipients had self-limited, equivocal GCA-related signs/symptoms without concurrent CRP or ESR elevation; all 4 completed the prespecified GC taper by Week 26 without need for rescue GCs, so relapse was not confirmed. At least 1 elevated CRP or ESR value in the absence of GCA-related signs/symptoms was observed in 58.8% of mavrilimumab recipients and 93.3% of placebo recipients by Week 26.Conclusion:The observed association of CRP or ESR elevation with GCA-related signs/symptoms is consistent with the upstream mechanism and supports the utility of the stringent protocol definition of relapse. The frequency and magnitude of CRP and ESR elevations at relapse were similar in both treatment groups, suggesting that CRP and ESR remain useful in assessments of disease activity in mavrilimumab-treated patients. CRP and ESR elevations without GCA-related signs/symptoms occurred more often in placebo recipients.