Study of Vosaroxin or Placebo in Combination With Cytarabine in Patients With First Relapsed or Refractory AML

The study includes additional objectives to ones listed above as Outcome Measures. These additional objectives also compared treatment groups in the following:

CR + CRp rate, defined as CR + CRp based on modified IWG response criteria.

Combined CR rate (CR+CRp+CRi).

Percentage of patients who have post-treatment (subsequent) transplantation.

Percentage of patients who received subsequent non-protocol therapy (including transplantation).

Safety and tolerability.

In keeping with FDA guidance for adaptive trial designs, the study incorporated an independent DSMB (Drug Safety Monitoring Board) to address potential uncertainty concerning the true treatment affect between the treatment groups and to address a deterioration of power from a small difference. Sunesis remained blinded and had no involvement in the interim data analysis, interpretation, or adaptive design. Based on the results of the interim data analysis the DSMB recommended an increase in the target number of deaths from 375 in 450 patients to 562 in 675 patients which based on a 5% dropout rate increased enrollment from 475 to 712.

The primary analysis was performed when the target number of deaths had been achieved based on a permuted block randomization procedure, stratified by disease status (refractory, first relapse with duration of first CR or CRp ≥ 90 days and < 12 months, or first relapse with duration of first CR or CRp ≥ 12 months and ≤ 24 months), age (< 60 years or ≥ 60 years), and geographic location (US or outside US). The study included periods of screening, treatment / hematologic recovery, post-treatment follow-up, and long-term follow-up for survival.

Follow-up was monthly during the first year, every 2 months during the second year, and every 3 months thereafter until death, withdrawal of consent, or loss to follow-up, whichever occurred first. Long-term follow-up began for all patients when the required number of deaths for primary analysis had been met; thereafter, survival data were collected every 4 months until death, withdrawal of consent, or loss to follow-up, whichever occurred first.

The long term follow-up for this study continues at this time and the September 2014 date reflects database lock for primary analyses reflected in the Results Section. During long term follow-up Sunesis is not collecting Adverse Events.