Azacitidine As Front-Line Therapy in AML: Results From Spanish National Registry. Alma Study Investigators

Abstract 3593

Azacitidine (AZA) is currently being used in AML patients (Pts) not deemed candidates for intensive chemotherapy. Most of this usage follows approved drug label by EMA, but off-label usage is not uncommon in Pts with >30% bone marrow blast (BMB) cells. Clinical trials are now comparing AZA vs. conventional care in selected populations, and data from daily practice may shed light on the generalizability of their results.

A retrospective nationwide study was set up in Spain in order to evaluate the population of AML Pts that is receiving AZA as front-line therapy (Rx) in Spain, its patterns of usage, effectiveness and safety in daily practice conditions, as well as the factors linked to overall response rate (ORR) and overall survival (OS). Only Pts treated before Dec/2010 could be included in this study, that was approved by the Spanish Medicines Agency (AEMPS, code ACL-AZA-2011–01). Data were collected from Oct/2011 to Jan/2012 and analyzed as of Jun/2012. ORR was evaluated according to both ELN-2010 criteria for AML as well as IWG-2006 criteria for MDS (in order to assess the impact of PB changes), OS measured from 1^st cycle of AZA to death or last follow-up, and toxicity coded according to NCI CTCAEv3.0.

One-hundred and ten untreated Pts (79 M/31 F, median age 75, range 56–89) were recruited from 22 academic and community sites. Comorbidity was present in 96 Pts (cardiac 43, hepatic 10, renal 3, diabetes 26, other neoplasms 18, etc.), ECOG being ≥2 in 33. Thirty Pts had an antecedent hematological disorder. Five cases had recurrent genetic abnormalities (NPM1-mutated AML in 4), 61 had MDS-related AML, 16 therapy-related AML and 28 AML not otherwise specified. Cytogenetics (Cyto) was available in 95 (86.4%): 48 diploid and 47 abnormal. MRC-2010 cyto category was favourable in 1, intermediate in 64 and adverse in 30. Median WBC at Dx was 3.3 x10E9/L (0.8–172.4), WBC before 1st AZA cycle 2.8 (1.0–175.0), platelet count 56 (7–467), PB blast 4.0% (0–100) and hemoglobin 91 g/L (48–142). Sixty-four Pts (58.2%) had BMB>30% (median 35.0%, range 15.0–98.0%). Median time from Dx to Rx was 19.5 days (0–411). Pts received 745 AZA cycles (median 4, range 1–29; ≥6 cycles 45.4%, ≥12 cycles 18.8%), but 30.9% received ≤2 because of disease progression or toxicity. 7.2% received concomitantly hydroxiurea for WBC control. Route of administration was EV in 5.8%, home administration took place in only 1.5%, and AZA was given as inpatients in 27.3% of the cycles. The no. of days of AZA Rx was 7 in 63.6% of the cycles, week-end off Rx being common place (5–2–2 schedule in 71.9%). Median AZA daily dose was 73.6 mg/sqm (<50 in 2.7%).

During follow-up (median 8.6 months, 0.1–48.7), 56 Pts progressed (39 on and 17 off-Rx). In 18, response was not evaluable (lack of BM assessment or death before 8 weeks). Best ORR according to IWG-2006 was 44.5% and 53.3% in the ITT and evaluable populations, respectively, while ORR according to ELN-2010 was 17.3% and 20.7%, respectively. Complete response rate (including CRm/CRi) was 15.5% and 18.5%, respectively, with both criteria. Platelet count duplication after 1^st cycle predicted ORR with IWG-2006 (82.4% vs. 33.8%, p=0.001, Fisher) but this was less evident with ELN-2010 (41.2% vs., 18.3%, p=0.057, Fisher).

Median OS from 1^st AZA cycle was 8.1 months (CI95% 5.3–10.9, range 0.1–47.9; OS at 12 months 36.7%, OS at 24 months 7%), OS from Dx 9.5 (CI 6.0–10.0, 0.1–48.7) and PFS 7.2 (CI 4.7–9.8; 0.1–29.5). Multivariate analysis showed that the best predictors of OS in our series were: ECOG ≤1, BMB ≤30%, a diploid cyto and WBC before 1^st AZA cycle <10.0xE9/L. GFM score (Park et al, 2008), but not MLD, MDS-related AML or AZA dose, also predicted OS (p<0.001, Logrank). As expected, responders lived longer than non-responders, but discrimination was better for IWG-2006 (HR=2.84) than ELN-2010 (HR=2.32), suggesting that a PB response may also impact survival in AML. Six-hundred thirty-eight AEs were reported (25.9% SAEs and 36.7% grade III-IV), most commonly infectious (25.5%), hematological (19.3%), gastrointestinal (18.3%) and cutaneous (11.8%).