A45: Neutropenia With Tocilizumab Treatment Is Not Associated With Increased Infection Risk in Patients With Polyarticular‐Course Juvenile Idiopathic Arthritis

Background/Purpose: Tocilizumab (TCZ) treatment has been associated with reduced neutrophil counts in adults with rheumatoid arthritis[1] and in children with systemic juvenile idiopathic arthritis[2]. The aims of this study were to assess changes in neutrophil counts, determine whether neutropenia was associated with increased risk for infection, and investigate variables associated with the development of neutropenia in patients with polyarticular-course juvenile idiopathic arthritis (pcJIA) receiving TCZ for up to 2 years in the CHERISH trial. Methods: One hundred eighty-eight patients (2–17 years) with active pcJIA and inadequate response to methotrexate received open-label (OL) TCZ according to body weight (BW) (BW ≥30 kg, TCZ 8 mg/kg; BW <30 kg, randomly assigned 1:1 to TCZ 8 or 10 mg/kg) every 4 weeks for 16 weeks. Patients who achieved ≥JIA ACR30 response at 16 weeks entered a 24-week, randomized, double-blind withdrawal period and were assigned 1:1 to placebo or continued TCZ. Patients who experienced JIA ACR flare or who completed the withdrawal period entered an OL extension through week 104. Blood cell counts were monitored every 4 weeks. Worst Common Toxicity Criteria (CTC) neutrophil grade and lowest observed neutrophil count (109/L) were identified for each patient. Rates of infection and serious infection (per 100 patient-years [PY]) in periods ±30 days around grade 1/2 and around grade 3/4 neutrophil counts were compared with corresponding rates in periods with normal neutrophil counts. Univariate linear regression analysis was used in patients enrolled in part 3 of the study to investigate the association between patient baseline characteristics and lowest observed neutrophil count from baseline to week 104. Results: At baseline, all patients had neutrophil counts within or above the normal range. Median neutrophil count decreased during the first 16 weeks of treatment and stabilized for the remainder of the study. Throughout the 2-year study, 118 patients (62.8%) had normal neutrophil counts, whereas CTC grades 1, 2, 3, and 4 neutrophil counts were reported in 15 (8.0%), 44 (23.4%), 11 (5.9%), and 0 (0.0%) patients, respectively. Infections and serious infections occurred at rates of 151.4/100 PY and 5.2/100 PY, respectively. All serious infection events (n = 16) occurred during periods of normal neutrophil count. Rates of infection during periods of normal, grade 1/2, and grade 3 neutrophil count were 147.8/100 PY (95% CI, 133.9–162.7), 176.6/100 PY (95% CI, 128.8–236.3), and 340/100 PY (95% CI, 136.7–700.5), respectively, with largely overlapping confidence intervals. Low neutrophil count was not associated with any baseline covariates investigated. Conclusion: In the CHERISH trial, there was no evidence of increased rates of serious infection associated with low neutrophil count. The prevalence of neutropenia in the pcJIA population was lower than in other disease groups (eg, systemic JIA) treated with TCZ.[2]