Background: Neutropaenia is reported as a complication of rituximab (RTX) therapy for B cell malignancies with an incidence of 3-27%. [1] Data in rheumatic diseases are more limited and the optimal management of these patients has not been defined.

Objectives: The aims of this study were to determine the (1) incidence of rituximab-associated neutropaenia; (2) rates of infection; (3) time for recovery; (4) safety of retreatment with RTX, as a basis for management guidelines.

Methods: We conducted an observational study on all patients treated with RTX in a single centre between 2003 and 2014. Each cycle consisted of 2x1000mg (with a small percentage receiving half dose if in remission), repeated either on clinical relapse or pre-emptively. RAN was defined as an absolute neutrophil count <2.0×10⁹/L occurring at least 4 weeks after RTX excluding chronic neutropaenia or alternative plausible explanation (e.g. Felty's syndrome).

Results: Rituximab-associated neutropaenia was identified in 23 patients (2.5%) from a cohort of 912, and in 36 cycles (1.2%) out of 3062 administered. 20 patients were female; median (IQR) age 61 (54-68,2) years. 19 had rheumatoid arthritis (RA) (2.72% of all RA patients in this cohort), 1 SLE, 1 cryoglobulinemia, 1 Sjogren's syndrome and 1 GPA. 21 patient received concomitant methotrexate, 1 SLE patient received cyclophosphamide and 1 RA patient received hydroxychloroquine. Neutropaenia occurred at median 17 weeks (range 4-31) following RTX infusion and median 3nd cycle (range 1-9). The majority of neutropenic episodes were transient; neutrophil count was normal on the first repeat test in 24 (66.7%) of the episodes. The frequency of mild (>1.0×10⁹/L), moderate (0.5-1.0×10⁹/L) and severe (<0.5×10⁹/L) neutropaenia were 18 (50%), 6 (16.7%) and 12 (32.3%) episodes respectively. Of these, 10 infections requiring antibiotics were recorded in severe neutropenia cases, most commonly chest infection (6 also required granulocyte-colony stimulating factor (GCSF)). No case of neutropaenia >0.5 x 10⁹/L was associated with infection. Irrespective of the degree of neutropaenia, all patients responded to RTX for the original indication. 26 (72.2%) of the episodes had complete B cell depletion as assessed by highly sensitive FACS. Of the patients who were retreated with RTX 2x1000mg: 11/19 had no recurrence of neutropaenia, 8/19 had mild neutropaenia in next cycle, and 4 patients had again recurrence of mild neutropenia. All the subsequent episodes were without associated infection or requirement for GCSF.

Conclusions: This is the largest cohort analysed for rituximab-associated neutropaenia. It can be concluded: (1) at <3%, it is less common in rheumatic disease than lymphoma; (2) monitoring alone is appropriate unless there is evidence of infection, when GCSF may be required; (3) the majority of the neutropenia cases recovered promptly; (4) counts >0.5 x109/L had no infections; (5) on retreatment, mild neutropaenia recurred in less than half with few consequences, with no evidence for neutropaenia becoming more severe on repeat cycles; (6) it therefore appears retreatment with monitoring is appropriate with caution only in severely neutropenic patients. The aetiology of rituximab-associated neutropaenia needs further investigation.

Disclosure of Interest: J. Ferreira: None declared, M. Y. Md Yusof: None declared, S. Das: None declared, E. Vital Grant/research support from: Roche and GSK., P. Emery Grant/research support from: Abbott, BMS, Pfizer, MSD and Roche., Consultant for: BMS, Abbott, Pfizer, MSD, Novartis, Roche and UCB.

DOI: 10.1136/annrheumdis-2015-eular.3881