Azacitidine (AZA) Combined with Idarubicin in Higher Risk MDS - Results of a Phase I/II Study By the Groupe Francophone Des Myelodysplasies (GFM)

Background: Hypomethylating agents, especially AZA, have become the reference treatment of higher risk MDS, but the median survival of about 2 years obtained with AZA remains modest, and must be further improved. In addition, AZA yields only about 30% of marrow responses (including CR+PR+marrow CR). Intensive chemotherapy combining Idarubicin (IDA) and AraC yields 30 to 50 % CR in higher risk MDS (Beran and all, cancer 2001) and IDA, as single agent, induces about 30% CR in elderly AML patients (Carella, haematologica 1990). We designed a phase I/II study evaluating the safety and efficacy of the combination of AZA and IDA (1 day during each AZA cycle) in higher risk MDS patients (NCT01305135).

Methods: Main Inclusion criteria were: (1) IPSS int-2 or high MDS, or CMML with WBC < 13 G/l and marrow blasts > 10%, or AML with 20-30% marrow blasts (2) Age ≥ 18 years (3) Performance Status (PS) <=2 (4) no prior treatment, except ESAs. Patients received AZA 75 mg/m2/d SC x7d every 4 weeks combined, on day 8 of each cycle (for the first 9 cycles), with IDA 5 mg/m2 (IV) in a first cohort of 10 patients, escalated to 10 mg/m2 IV in a second cohort of 10 patients, followed by an extension study in 21 patients with the IDA 10 mg/m2 schedule. The primary endpoint of the study was response after 6 cycles according to IWG 2006 criteria.

Results: Between Dec 2010 and Jan 2014, 41 patients (from 13 centers) were enrolled, including 13 women and 28 men with a median age of 74 years [IQR 70; 76]. At inclusion, WHO classification was RCMD in 1 pt, CMML in 2 pts, RAEB-1 in 10 pts, RAEB-2 in 13 pts, AML in 12 pts and unclassified MDS in 3 pts. Median marrow blast % was 9.5 [IQR: 6-19.9] and karyotype according to IPSS was favorable in 12 (29%), intermediate in 9 (22%) and unfavorable in 18 pts (44%) (2 cytogenetic failures). IPSS was int-2 and high in 56% and 44%, respectively. PS was 0 in 39%, 1 in 55% and 2 in 6% pts.

10 patients received 5 mg/m2 of idarubicin (cohort 1) and 31 received 10 mg/m2 (cohort 2). 375 cycles of AZA were administered (219 of them with AZA+IDA, as IDA was used only for the first 9 cycles), with a median number of 6 cycles/patient (median 6 in the IDA 5 mg/m2 cohort and 4 in the 10 mg/m2 cohort (p=0.9).

Of the 41 patients enrolled, 20(48.8%, 95%CI: 32.9-64.9) achieved response (6 CR, 7 PR, 4 mCR and 3 stable disease with HI) with no difference between the two cohorts (50% vs 48%) and a marrow response rate (CR + PR + mCR) of 41,5%. Thirteen of the 22 patients with abnormal karyotype were evaluable for cytogenetic response: 5 achieved cytogenetic response (4 complete, 1 partial), 1 in cohort 1 and 4 in cohort 2. With a median follow up of 14 months, 9 of the 20 responders had relapsed. Median response duration was 11 months [3.2-42.7], with no difference between the two cohorts.

Median OS was 14.3 months [IC95%: 12.5; NA] and 2y OS was 24.8%, with no significant difference between the 2 cohorts (p=0.43).

By univariate analysis no baseline parameter including gender, karyotype, marrow blast %, IPSS and IDA dosage (5 or 10 mg/m2), had any significant impact on response or survival.

45 SAEs were reported in 26 patients, including febrile neutropenia (n=25), bleeding (n=7) and 2 non-clinically significant reductions in left ventricular ejection fraction (1 transient, and 1 persisting without symptoms). The number of infections per cycle [9/85 (10%) in the IDA 5 mg/m2 arm and 38/281 (14%) in the IDA 10 mg/m2 arm] and the number of bleeding events (9% vs 17%) did not significantly differ between the two cohorts.

Conclusion: In our experience, Idarubicin (on day 8 of each cycle) can be combined to Azacitidine without any additional toxicity. The marrow response rate obtained with the combination (41.5%) may be higher than with AZA alone. We are currently comparing in higher risk MDS patients this AZA-IDA combination versus AZA alone (and other combinations of AZA with other drugs) in a prospective randomized GFM trial.