ALBERT

Description: Abstract 745 Background: The optimum chemotherapy combination prior to autologous stem cell transplantation (ASCT) for patients with relapsed or refractory aggressive non-Hodgkin lymphomas (NHL) has not been defined. We previously established the safety and efficacy of outpatient treatment with GDP in a phase II study (Crump et al, Cancer 2004) leading to this phase III trial testing the hypothesis that GDP is as effective and less toxic than standard DHAP, and potentially associated with better quality of life (QoL) and less resource utilization. Methods: Patients (pts) with relapsed/refractory aggressive NHL were stratified by IPI score at relapse (0, 1 vs 2,vs 〉=3 risk factors), immunophenotype (B vs T cell), disease status following initial treatment (response duration 〈 1yr vs duration 〉 1yr vs no response/PD) and prior treatment with rituximab (R); and were randomized to 2–3 21-day cycles of G 1000 mg/m2 day 1 & 8, D 40 mg day 1–4, P 75 mg/m2day 1 or standard DHAP. Pts with CR, PR (or SD, per institutional policy) proceeded to PBSC collection and ASCT. The protocol was amended 11/2005 to include R with GDP or DHAP for pts with CD20+ lymphoma. Co-primary endpoints were response rate (RR) after 2 cycles (CR/CRu/PR) according to 1999 International Workshop criteria, and transplantation rate; FDG-PET was not used to determine response. Using a non-inferiority design, the study was powered to exclude a 10% difference in RR between GDP and DHAP (α 0.05, power 80%). Other endpoints included toxicity, event-free (EFS) and overall survival (OS), QoL and an economic analysis. FACT-G, FACT-CNS and FACT-LYM instruments and validated lymphoma-specific questions were used to assess QoL. Results of a second randomization to post-ASCT rituximab consolidation vs observation for pts with CD20+ lymphoma will be reported subsequently. Results: From 8/2003 to 11/2011, 619 pts were enrolled at 52 centres in Canada, Italy, Australia and the US (GDP 310 DHAP 309). Pt characteristics: median age 55 yrs (28.4% 〉60 y); histology: DLBCL 71%, PTCL 4%, ALCL 4%, transformed 14%; IPI risk factors: 0,1: 38%, 2: 29%; 〉3: 33%; elevated LDH 59%; prior R treatment 67%; refractory to initial therapy 31%; all baseline characteristics were balanced between treatment arms. In the intention to treat (ITT) population, RR for GDP was 45.2% and for DHAP 44.0%. The upper boundary of the one-sided 95.6% confidence interval for the difference in RRs was 5.67% and did not cross the pre-specified 10% non-inferiority boundary, meaning protocol-stated criteria for declaring GDP non-inferior to DHAP were met (p=0.005); results were robust when assessed using a per-protocol analysis. 362/554 pts (65.3%) with B cell NHL received R with protocol therapy: RRs were similar between arms in pts with and without prior R exposure, and with and without R added to GDP or DHAP. The transplantation rate was 51.8% for GDP vs 49.3% for DHAP (per protocol analysis; p = 0.49); 6 pts in each arm could not mobilize adequate PBSCs. At a median follow-up of 53 months, 4 year EFS was 25.6% and 26.1% (HR 0.99, p=0.95) and 4 year OS 39.0% and 39.1% (HR 1.03, p=0.78) for GDP and DHAP, respectively. Pts receiving GDP experienced less grade 3–4 toxicity (47 vs 61%, p=0.0003), including febrile neutropenia (9 vs 23%, p