Abstract 924 Bendamustine (B) is an alkylating agent recently approved for relapsed and refractory indolent non-Hodgkin lymphoma NHL. In vitro studies suggest synergy between bendamustine and rituximab (R), and two phase II trials of this combination demonstrated tolerability and high response rates in indolent and mantle cell (MCL) NHL. The proteasome inhibitor bortezomib (Velcade; V) has significant single agent activity in indolent and mantle cell NHL. We have completed enrollment of a multicenter phase II trial combining bortezomib with the bendamustine/rituximab combination. Eligible pts had relapsed or refractory indolent or mantle cell NHL. Treatment consisted of B 90 mg/m2 day 1 and 4; R 375 mg/m2 day 1 and V 1.3 mg/m2 day 1, 4, 8, 11. Six 28-day cycles of therapy were planned. 31 patients (23 male) were enrolled; median age was 64 yrs (range 44-84). Histology included 16 follicular NHL, 7 MCL, 3 marginal zone NHL, 3 SLL and 2 lymphoplasmacytic NHL. Patients were heavily pretreated with a median of 4 prior regimens, including anthracycline containing chemotherapy (n=19), purine analog chemotherapy (n=6), ASCT (n=6), and radioimmunotherapy (n=9). 10 pts were refractory to rituximab and 25 pts had advanced stage disease at time of study. One pt never received therapy due to transformation. One pt died of sepsis after the first cycle of therapy. Common expected toxicities included thrombocytopenia, fatigue, fever, anemia, neutropenia and infusion reactions. Peripheral neuropathy was reported in 18 pts, including 2 with grade 3 neurotoxicity; additionally 14 patients reported constipation. There were 5 pts who developed varicella zoster reactivation (no prophylaxis given). Fourteen pts completed 6 cycles; reasons for premature withdrawal include progressive disease (PD), (n=4); cytopenias (n=4), infection (n=2), death (n=1) and nausea (n=1); 4 patients currently remain on treatment as of August 2009. Of 25 pts evaluable for response, overall response rate was 84% (95%CI 65%-94%). Best response following therapy was CR/Cru: 13 (52%); PR: 8 (32%); SD: 1(4%); PD: 3 (12%). Interestingly, 11/11 pts with FL responded to treatment (including 7 (64%) CR/CRu), and 5/7 pts with MCL responded to treatment. With a median follow-up of 13 months, 4 patients have died (3 of PD). There is no association between prior ASCT, radioimmunotherapy or purine analog chemotherapy and premature study withdrawal. There was also no association between rituximab sensitivity and response to BVR. Correlative laboratory studies are ongoing to determine predictors of toxicity and response. In summary, in this heavily pretreated population (as compared with prior studies of BR, including 33% rituximab-refractory pts), the BVR regimen is highly active, with over half of evaluable pts achieving CR/CRu. It appears more toxic than BR alone, with expected additive toxicities from V. Prophylaxis against varicella zoster reactivation is indicated when using this regimen. Further follow-up will determine whether the high CR/CRu rate corresponds to prolonged PFS. These promising results warrant additional evaluation of this regimen in de novo disease. Ultimately, a randomized trial will be necessary to determine the degree to which V adds efficacy to the BR combination. Disclosures: Friedberg:cephalon: Research Funding; millenium: Research Funding; genentech: LymphoCare Advisory Board. Off Label Use: bendamustine/bortezomib/rituximab combination therapy.