Azacitidine front-line in 339 patients with myelodysplastic syndromes and acute myeloid leukaemia: comparison of French-American-British and World Health Organization classifications

Background: The MDS-IWG and NCCN currently endorse both FAB and WHO classifications of MDS and AML, thus allowing patients with 20-30 % bone marrow blasts (AML20-30, formerly MDS-RAEB-t) to be categorised and treated as either MDS or AML. In addition, an artificial distinction between AML20-30 and AML30+ was made by regulatory agencies by initially restricting approval of azacitidine to AML20-30. Thus, uncertainty prevails regarding the diagnosis, prognosis and optimal treatment timing and strategy for patients with AML20-30. Here, we aim to provide clarification for patients treated with azacitidine front-line. Methods: The Austrian Azacitidine Registry is a multicentre database (ClinicalTrials.gov: NCT01595295). For this analysis, we selected 339 patients treated with azacitidine front-line. According to the WHO classification 53, 96 and 190 patients had MDS-RAEB-I, MDS-RAEB-II and AML (AML20-30: n = 79; AML30+: n = 111), respectively. According to the FAB classification, 131, 101 and 111 patients had MDS-RAEB, MDS-RAEB-t and AML, respectively. Results: The median ages of patients with MDS and AML were 72 (range 37-87) and 77 (range 23-93) years, respectively. Overall, 80 % of classifiable patients (<= 30 % bone marrow blasts) had intermediate-2 or high-risk IPSS scores. Most other baseline, treatment and response characteristics were similar between patients diagnosed with MDS or AML. WHO-classified patients with AML20-30 had significantly worse OS than patients with MDS-RAEB-II (13.1 vs 18.9 months; p = 0.010), but similar OS to patients with AML30+ (10.9 vs 13.1 months; p = 0.238). AML patients that showed MDS-related features did not have worse outcomes compared with patients who did not (13.2 vs 8.9 months; p = 0.104). FAB-classified patients with MDS-RAEB-t had similar survival to patients with AML30+ (12.8 vs 10.9 months; p = 0.376), but significantly worse OS than patients with MDS-RAEB (10.9 vs 24.4 months; p < 0.001). Conclusions: Our data demonstrate the validity of the WHO classification of MDS and AML, and its superiority over the former FAB classification, for patients treated with azacitidine front-line. Neither bone marrow blast count nor presence of MDS-related features had an adverse prognostic impact on survival. Patients with AML20-30 should therefore be regarded as having 'true AML' and in our opinion treatment should be initiated without delay.