Abstract 3655 Obinutuzumab (GA101) is a type II glycoengineered, humanized anti-CD20 monoclonal antibody that has increased antibody-dependent cellular cytotoxicity and direct cell death activity but lower complement-dependent cytotoxicity compared with type I anti-CD20 antibodies such as rituximab and ofatumumab. GA101 is in clinical development for the treatment of lymphoma and chronic lymphocytic leukemia. The Phase I/II study BO20999 has evaluated the efficacy and safety of GA101 monotherapy in patients with relapsed/refractory aggressive non-Hodgkin9s lymphoma (aNHL). Here, we report updated Phase II results including progression-free survival (PFS) and best overall response (BOR). Patients (n = 40) were randomized to receive GA101 (D1, D8 and D22, then 3-weekly for total of 9 infusions) at either a high dose (1,600 mg on D1 and D8, then 800 mg thereafter; 1,600/800 mg cohort; n = 19), or a flat low dose of 400 mg (400/400 mg cohort; n = 21). Baseline patient characteristics were similar for both cohorts (Table 1). The median observation time for all patients was 9.5 months (0.3–26.1 months). BOR rates are given in Table 2, with 8/25 diffuse large B-cell lymphoma (DLBCL) patients (32%) and 4/15 mantle cell lymphoma (MCL) patients (27%) responding to GA101. Among the patients with rituximab-refractory disease, a response was observed in 1/13 patients (7.7%) and 4/12 patients (33.3%) treated in the 400/400 mg and 1,600/800 mg cohorts, respectively. Of these, 4 patients (1,600/800 mg cohort) had a response duration > 6 months, with 2 patients having an ongoing response (response duration: 9.8, 16.5+, 19.5 and 20.0+ months). Median PFS for patients with DLBCL (Figure 1) was 1.9 months (range: 0.3–15.7 months) for the 400/400 mg cohort and 2.7 months (range: 0.2–22.3) months) for the 1,600/800 mg cohort (hazard ratio: 0.70; 95% CI: 0.30–1.66). For the DLBCL subgroup, response duration was 3.1, 3.1+, 5.8, 16.5+ and 19.5 months for the 5 responders in the 1,600/800 mg cohort, compared with 6.3, 8.6 and 9.8 months for the 3 responders in the 400/400 mg cohort. Individual response data indicated that 2 MCL patients had an ongoing response for ≥ 20 months (20.0 and 20.4 months). GA101 was well tolerated in both cohorts. Infusion-related reactions (IRRs; all grades) were the most common adverse event (AE), occurring in 81% of patients in the 400/400 mg cohort and 68% of patients in the 1600/800 mg cohort. Grade 3/4 AEs occurring in >5% of patients across both cohorts included IRRs (10%), tumor lysis syndrome (10%), cardiac failure (not treatment-related; 10%), anemia (14%) and thrombocytopenia (14%) in the 400/400 mg cohort and IRRs (5%) and anemia (5%) in the 1,600/800 mg cohort. In conclusion, GA101 shows encouraging single-agent efficacy in these heavily pretreated patients with relapsed/refractory aNHL (DLBCL or MCL). A Phase III trial of rituximab plus CHOP vs GA101 plus CHOP in first-line DLBCL has recently started. Disclosures: Morschhauser: Roche: Honoraria; Celgene: Consultancy, Honoraria. Cartron: Roche: Consultancy, Honoraria; GSK: Honoraria; LFB: Honoraria. Milpied: Roche: Honoraria, Membership on an entity9s Board of Directors or advisory committees. Wenger: Roche: Employment. Wassner-Fritsch: Roche: Employment. Asikanius: Roche: Employment. Salles: Roche: Consultancy, Honoraria.