Response-Adapted Sequential Azacitidine and Induction Chemotherapy in Patients > 60 Years Old with Newly Diagnosed AML Eligible for Chemotherapy (RAS-AZIC): First Data of the Interim Analysis of the DRKS00004519 Study of the East German Study Group (OSHO)

The prognosis of AML in patients (pts) >60 years (y) is poor and there is no universally accepted standard approach. After induction chemotherapy (IC), a complete remission (CR+CRi) rate of 56.4% and a mortality of 21% at day (d) 90 are expected (Niederwieser et al, ASH 2012). Azacitidine (AZA) yields a one-year survival (OS) rate of 46.5% in pts >65 y and >30% marrow (BM) blasts (Dombret et al, Blood 2015). Yet, AZA and IC are not mutually exclusive. First results of the interim analysis of the DRKS00004519 (RAS-AZIC) study which evaluates priming with AZA followed by AZA or IC in pts >60 y with AML are presented.

Patients and methods: Major inclusion criteria are newly diagnosed AML with marrow blasts >20% by aspiration or biopsy irrespective of WBC count, age >60 y, and eligibility for IC. Secondary AML from an antecedent hematologic disorder and therapy-related AML are not excluded.

In the phase I part of the trial, safety of priming with AZA (75 mg/m2/day s.c) for 7 d followed by IC (Mitoxantrone 10 mg/ m2/day d 1-3 and cytarabine 1 g/ m2/BID d 1, 3, 5, 7) on d 17 was established through a 3+3 design. In the multicenter phase II part (figure), priming was sequentially followed by AZA or IC based on d 15 BM blasts and d 56 response which were identified as early predictors for long-term response to AZA in AML (Al-Ali et al., Leuk Lymph 2012). The primary objective is overall response (OR) which includes CR, CRi, and PR at d 90 according to the International Working Group criteria (Cheson et al, J Clin Oncol 2003). Safety, OS, and mortality were secondary end points. According to the optimal two-stage design (Simon, Control Clin Trials 1989), an expected OR of 61% at the end of trial was estimated. Thus in a first stage analysis, < 19 of consecutive 40 (47.5%) pts not achieving the primary objective would be considered an indication for inferiority of the protocol treatment compared to standard IC. Hence, an interim analysis of the first stage was planned per protocol while accrual continues to include the first 40 pts who received the maximal tolerated priming dose of AZA. The trial is supervised by an independent Data Monitoring Committee. Adverse events (AEs) are reported according to the NCI CTCAE 4.03. All pts gave written informed consent.

Results: Patient characteristics are shown in table 1. Median age is 69 y and median blast count 39%. All pts received priming with AZA. Median BM blasts at d 15 was 27% (range 2.4-88%). To date, data on subsequent therapies are captured in 33 (82.5%) pts. Except for one (withdrew consent on d 27), all pts received protocol assigned treatment based on BM blasts on d 15 [17/32 (53%) continued with AZA, 15/32 (47%) received IC]. Likewise, 26/32 (81%) pts ([2x data not yet available, 2x death (d 34; d 53), and 2x physician‘s choice] received response-adapted treatment on d 56 [12/26 (46%) started AZA maintenance and 14/26 (55%) received IC]. OR at d 90 for the 26 as yet evaluable pts is 80.7% [CR (61.5%, n=16); CRi (11.5%, n=3); PR (7.7%, n=2). No additional mortality till d 90 was reported. For the entire cohort, 85% of pts are alive after a median follow-up of 6 months. AEs were consistent with the known safety profile of AZA and IC. AEs > grade 3 were reported 83x in 20 (50%) pts being mainly infectious and respiratory complications (60%). As expected, the majority (57%) were reported under IC, followed by 19% and 17% under priming and the 2. AZA cycle respectively. AEs led to postponement of the planned therapy in 6 pts (3x IC on d 17, 2x 2. AZA, 1x IC on d 56). The final assessment of the interim analysis is set in the near future.

Conclusions: These data imply that in a heterogeneous disease like AML, integrating epigenetic therapy with chemotherapy in elderly pts in an individualized response-based approach is feasible, induces responses at least comparable to those seen after IC, and is associated with a remarkable low mortality. The definitive response rate and the impact of such an approach on survival will be answered at the end of the trial.

Patient characteristics (n=40)

Median age, range (years)	69 ( 61 – 82)
Male/Female (%)	55 / 45
Baseline ECOG (%) 0/1/2	21 / 63 / 16
Type of AML Do novo/secondary (%)	61 /39
Median WBC (range) x 109/L	6.7 (0.7 - 81.8 )
Median bone marrow blasts (aspiration) (range) (%)	39 (13 – 90)
Cytogenetics (n=35) Normal/intermediate/unfavourable (%)	54 /23 / 23
FLT3+   (n=34) (%)	9
NPM1+ (n=34) (%)	9