Background: Patients (pts) with relapsed/refractory (RR) chronic lymphocytic leukemia (CLL) need therapies that induce prolonged disease control. MEDI-551, an affinity-optimized anti-CD19 antibody, destroys CLL cells by antibody-dependent cellular cytotoxicity. A phase 2 randomized, open-label study (NCT01466153) is evaluating the clinical activity and safety/tolerability of 2 doses of MEDI-551 + bendamustine compared to rituximab + bendamustine in RR CLL patients. Methods: Pts were initially randomized to receive bendamustine + MEDI-551 2 or 4 mg/kg or rituximab (R). Safety assessments include adverse events (AEs) and laboratory parameters. Disease response was determined using 2008 International Working Group criteria. Results: The safety population comprised 124 pts across all arms. Median age was 66y (range 41–81); with deletion (del) (17p): 9%, del (11q): 22%, del (13q): 31%, trisomy 12: 11%. Median number of treatment cycles: 4 (range 1–6). The most common (≥20%) treatment-related AEs with MEDI-551 were infusion-related reaction (IRR), fatigue and nausea; neutropenia and nausea were most common with R; most treatment-related AEs were grade 1/2; grade 3/4 treatment-related AEs are listed in Table 1. 20% of pts receiving MEDI-551and 18% patients receiving R discontinued treatment due to AEs. An interim analysis of the MEDI-551-treated pts (25 per arm) showed response rates of 48 vs. 64% in the 2 vs. 4 mg/kg arms. Conclusions: MEDI-551 in combination with bendamustine has a manageable toxicity profile that appears to be different compared to rituximab in RR CLL patients. MEDI-551 dosed at 4 mg/kg as compared to 2 mg/kg may have a greater efficacy without any increase in toxicity. Clinical trial information: NCT01466153.