Results of the Randomized Phase II Study Decider (AMLSG 14-09) Comparing Decitabine (DAC) with or without Valproic Acid (VPA) and with or without All-Trans Retinoic Acid (ATRA) Add-on in Newly Diagnosed Elderly Non-Fit AML Patients

Background: Despite the recent approval of DNA-hypomethylating agents (HMAs) for treatment of elderly AML patients (pts) ineligible for induction, their prognosis is still poor, and rational, effective HMA-based combination treatments are under study. Histone deacetylase inhibitors (HDACi) show synergism with HMAs in vitro. ATRA - as single agent clinically ineffective in non-M3 AML - in combination with HMAs also shows in vitro synergistic antileukemic activity in non-M3 AML cells. We previously conducted a non-randomized phase II trial in elderly non-fit AML pts with DAC (3-dy schedule), given alone or combined with ATRA (45 mg/m² dy 4-28, only during course 2), with encouraging results (Lübbert et al., Haematologica 2012). We now expanded this approach to a 4-arm randomized phase II study (2x2 factorial design) asking whether the addition of either VPA (HDACi activity) or ATRA or both to DAC as first-line treatment of elderly AML pts might improve the effect of DAC monotherapy (NCT00867672).

Patients and Methods: Inclusion criteria: newly diagnosed pts >60 yr unfit for induction (reasons for treatment decision prospectively captured) with non-M3 AML (WHO, de novo or after antecedent hematologic disorder [AHD], therapy-associated [t]AML), ECOG performance status (PS) 0-2. Pts with >30,000 WBC/µl were to receive a short course of hydroxyurea. Treatment: DAC 20 mg/m² dy 1-5 (treatment arms A/B/C/D), VPA p.o. continuously (target serum levels: 50-110 mg/l) from dy 6 (arms B/D), ATRA p.o. dy 6-28 (arms C/D) of each 28-dy course (repeated until relapse/progression, prohibitive toxicity, withdrawal or death). Key endpoints: objective response rate (ORR): CR/CRi/PR (ELN criteria), overall survival (OS). Sample size calculation was based on the primary endpoint ORR, assuming an ORR of 25% in arm A (Lübbert et al., Haematologica 2012). For a power of 80% (test in this phase II study at 1-sided alpha=0.1) for an increase of ORR to 40% with VPA or ATRA, 176 pts were necessary, planned sample size 200. Efficacy analyses were performed in the intention-to-treat (ITT) population including all randomized pts for whom treatment was started. VPA was investigated by comparing arms B+D vs arms A+C, ATRA by comparing arms C+D vs arms A+B. ORR was analyzed with logistic regression, OS with Cox regression, without adjustment for prognostic factors. Odds ratios (OR) for the effect on ORR and hazard ratios (HR) for the effect on death with 95% confidence intervals (CI), and two-sided p values of the tests of no treatment effect are presented. Central hematopathological review by an independent morphologist was conducted in a blinded fashion as to treatment arms.

Results: Between 12/2011 and 2/2015, 204 pts were randomized (4 were excluded from the analysis because no treatment was administered). Median age: 76 yrs (interquartile range 72-79, range 61-92), ECOG PS 0/1/2-3: 19/61/20%: 52% had an HCT-CI >3, 16.5% WBC >30.000/µl, 31.5% poor cytogenetics (ELN), 51% had an AHD, 13.5% tAML (characteristics overall balanced across all 4 treatment arms). A median of 3 DAC courses were administered (per arm: 2/3/5.5/4), however 53 pts (26.5%), who were older, with reduced PS and a higher HCT-CI compared to the other 147 pts, received only a single course. The ORR (usually achieved only after >3 courses) was 17.5%, median OS 6.2 mths (arm A: 8.5% and 4.8 CI [2.8,7.6] mths, arm B: 17.5% and 6.1 CI [3.7,7.2] mths, arm C: 26.1% and 8.4 CI [4.0,14.0] mths, arm D: 18% and 7.7 CI [4.6,11.2] mths, respectively). Effect on ORR of VPA vs no VPA (17.8 vs 17.2%): OR 1.06, CI [0.51,2.21], p=0.88; of ATRA vs no ATRA (21.9 vs 13.5%): OR 1.80, CI [0.86,3.79], p=0.12. Effect on OS of VPA vs no VPA (6.2 vs 6.4 mths median OS): HR 0.94, CI [0.70,1.28], p=0.70; of ATRA vs no ATRA (8.2 vs 5.1 months median OS): HR 0.65, CI [0.48,0.88], p=0.006 (after adjustment for PS, HCT-CI, WBC, LDH: HR 0.59, CI [0.43,0.82], p=0.002). Improved survival with ATRA was also seen in pts with poor cytogenetics. Toxicities (predominantly hematologic) did not show relevant differences between the 4 treatment arms.

Conclusions: Based on this ITT analysis of a randomized trial, the addition of ATRA to standard-dose DAC resulted in a higher ORR and in a clinically relevant extension of OS, without additional (hematologic and non-hematologic) toxicity. In contrast, the addition of VPA did not affect ORR or OS.