Abstract: Introduction: Despite 40 years of intense clinical research, there remain no new approved treatments or standard of care for patients with relapsed or refractory (R/R) acute myeloid leukemia (AML). New safe and effective salvage treatments are urgently needed. Vosaroxin is a first-in-class anticancer quinolone derivative that is active in AML. Vosaroxin is minimally metabolized, evades P glycoprotein receptormediated efflux and has activity independent of p53 status. VALOR is a rigorously designed and conducted phase 3, adaptive design, randomized, double-blind, placebo-controlled trial evaluating vosaroxin plus cytarabine (vos/cyt) vs placebo plus cytarabine (pla/cyt) in patients with R/R AML (NCT01191801). Methods: Patients were randomized 1:1 to receive cytarabine (1 g/m2 IV over 2 hr, d 1-5) plus either vosaroxin (90 mg/m2 IV over 10 min d 1 and 4; 70 mg/m2 in subsequent cycles) or placebo. Up to 2 induction and 2 consolidation cycles were administered. Eligible patients had refractory disease (persistent disease after induction, or first complete remission [CR1] 〈 90 d) or were in first relapse (early relapse: CR1 of 90 d to 12 mo; late relapse: CR1 of 12 mo to 24 mo). Patients had received 1-2 cycles of prior induction chemotherapy including at least 1 cycle of anthracycline (or anthracenedione) and cytarabine. Randomization was stratified by disease status (refractory, early relapse, late relapse), age ( 〈 60, ≥ 60 years), and geographic location (US, non-US). Primary efficacy and safety endpoints were overall survival (OS) and 30- and 60-day mortality; secondary endpoints were complete remission (CR) rate and incidence of adverse events (AEs). Results: Between Dec 2010 and Sept 2013, 711 patients were randomized to receive vos/cyt (n = 356) or pla/cyt (n = 355) at 124 sites; per the adaptive design, a prespecified 1-time sample size increase of 225 patients was implemented after the interim analysis. At the final analysis, median OS was 7.5 mo (95% CI: 6.4-8.5) with vos/cyt vs 6.1 mo (95% CI: 5.2-7.1) with pla/cyt (HR = 0.866 [95% CI: 0.73-1.02]; 2-sided unstratified log-rank P = 0.06) (Figure). The OS difference was statistically significant in a preplanned analysis accounting for the stratification factors at randomization (2-sided stratified log-rank P = 0.02). Overall, 29.5% of patients underwent allogeneic stem cell transplant (ASCT), including 45.8% of patients 〈 60 years and 20.2% of patients ≥ 60 years. Transplant rates were comparable between the 2 treatment arms (30.1% with vos/cyt and 29.0% with pla/cyt). In a predefined analysis censoring for subsequent ASCT, median OS was improved with vos/cyt (6.7 mo vs 5.3 mo with pla/cyt; HR = 0.81 [95% CI: 0.67-0.97]; P = 0.02; stratified P = 0.03) (Figure). In predefined subgroup analyses, OS benefit was greatest in patients aged ≥ 60 years (7.1 mo with vos/cyt vs 5.0 mo with pla/cyt; HR = 0.75; P = 0.003) (Figure) and those with early relapse (6.7 mo vs 5.2 mo; HR = 0.77; P = 0.04). OS with vos/cyt vs pla/cyt was 9.1 mo vs 7.9 mo in patients 〈 60 years (HR = 1.08; P = 0.60); 6.7 mo vs 5.0 mo in patients with refractory disease (HR = 0.87; P = 0.23); and 14.1 mo vs 12.3 mo in patients with late relapse (HR = 0.98; P = 0.96), respectively. A CR was achieved in 30.1% of patients treated with vos/cyt vs 16.3% treated with pla/cyt (P = 0.00001). Thirty-day and 60-day all-cause mortality was similar in the 2 arms (30-day: 7.9% vs 6.6%; 60-day: 19.7% vs 19.4% with vos/cyt vs pla/cyt, respectively). Most common serious AEs were febrile neutropenia (11.3% with vos/cyt vs 7.4% with pla/cyt), sepsis (8.7% vs 4.3%), pneumonia (7.6% vs 4.9%), bacteremia (8.5% vs 2.9%), and stomatitis (3.4% vs 1.4%). Serious and non-serious cardiac, renal, neurologic, and hepatic AEs were comparable between treatment groups. Conclusion: Vos/cyt demonstrated improved OS and higher CR rates in patients with R/R AML without increased early mortality. In the primary OS analysis, the overall clinical benefit associated with vosaroxin may be underestimated, particularly in younger patients, due to the confounding effect of high transplant rates, a methodological limitation of AML trials. Vosaroxin-containing therapy had acceptable tolerability. VALOR results represent one of the largest datasets available in this setting, and the OS benefit was confirmed by a robust sensitivity analysis. These data support the use of this combination as a new option for salvage therapy in patients with R/R AML. Figure 1 Figure 1. Disclosures Ravandi: Sunesis: Consultancy, Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding. Sayar:Sunesis: Membership on an entity's Board of Directors or advisory committees, Research Funding. Strickland:Sunesis: Membership on an entity's Board of Directors or advisory committees. Schiller:Sunesis: Membership on an entity's Board of Directors or advisory committees, Research Funding. Erba:Sunesis: Consultancy; Seattle Genetics: Consultancy; Novartis: Consultancy; Incyte: Consultancy; Celgene: Consultancy; Amgen: Consultancy. Pigneux:Sunesis: Consultancy. Horst:Sunesis: Research Funding. Recher:Sunesis: Consultancy; Celgene: Consultancy, Research Funding; Chugai: Research Funding. Klimek:Sunesis: Membership on an entity's Board of Directors or advisory committees, Research Funding. Craig:Sunesis: Equity Ownership. Fox:Sunesis: Consultancy, Equity Ownership. Ward:Sunesis: Employment, Equity Ownership. Smith:Sunesis: Employment, Equity Ownership. Acton:Sunesis: Consultancy. Mehta:Sunesis: Consultancy. Stuart:Sunesis: Consultancy, Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding.

Abstract: TPS6637 Background: The promising phase 2 activity/tolerability profile of vosaroxin with cytarabine in first relapsed or refractory AML (N=69) with median overall survival (OS) 7.1 mo, combined CR 28% (CR 25%), median LFS 25 mo and 30-day all‑cause mortality 3%, supported phase 3 trial. VALOR (NCT01191801), a phase 3, randomized, controlled, double-blind trial, evaluates vosaroxin and cytarabine versus placebo and cytarabine in patients with first relapsed or refractory AML and incorporates an adaptive design. Primary objective is OS; secondary/tertiary objectives include CR rates, safety, EFS, LFS, and transplantation rate. Key eligibility criteria: persistent AML or first relapsed AML after 1 or 2 induction cycles that include at least 1 regimen of cytarabine with an anthracycline/anthracenedione; adequate cardiac, hepatic, renal function; and adults with no upper age restriction. Methods: VALOR is recruiting patients at over 100 sites in 14 countries in North America, Europe, and Australia/NZ. Enrollment opened Dec 2010; 183 patients enrolled through Dec 2011. Adaptive design: Base case assumed 40% improvement in median OS (hazard ratio, HR=0.71), 90% power, 2-sided alpha of 0.05. At the interim analysis (50% events), DSMB may recommend a 50% sample size increase from 450 to 675 evaluable patients if results indicate a larger sample size is required to reduce the risk of failing to confirm a clinically meaningful OS benefit (Mehta, Pocock, Stat Med 2010). In consideration of FDA and EMA guidance on Data Monitoring Committees and adaptive design with respect to trial integrity, operational bias, firewalls, and data confidentiality and archival, VALOR uses the Access Control Execution System (ACES ) to store, share, and archive confidential DSMB reports in a secure environment with an audit trail, and to facilitate communication between the DSMB and trial sponsor. The DSMB periodically reviews a combination of blinded and unblinded analyses while the study team remains blinded as specified in the DSMB charter. The DSMB recommended VALOR continue as planned after reviewing safety data in Dec 2011.