Activity of Azacitidine in 26 Unselected, Consecutive CMML Patients Included in the Austrian Azacitidine Registry (AAR) of the AGMT-Study Group

Abstract 1715

Twelve participating centers in Austria included 257 unselected, consecutive patients with MDS/CMML or AML, who received azacitidine (AZA) between 02/2007 and 07/2011, in the nationwide Austrian Azacitidine Registry (AAR) of the AGMT-study group. Approval by the national Ethics Committee was obtained. The AAR includes 26 patients with CMML (10 with CMML-1, 16 with CMML-2) and is currently the 2^nd largest report on CMML patients treated with AZA, the largest comprising 36 evaluable patients (Costa R. et al., Cancer, June 2011, pp2690). Myeloproliferative CMML (MP-CMML), defined as WBC>13.000/μl at diagnosis, was present in 16/26 patients, whereas 10/26 had myelodysplastic CMML (MD-CMML). Thus, the AAR includes patients for whom AZA is currently not approved by the FDA/EMEA.

The AAR includes a high proportion of very old CMML patients (31% 75–79a, 15% >80a). Median age was 75a for both males (n=15) and females (n=11). Patients with various levels of PS were included (ECOG-0 (n=6), ECOG-1 (n=14), ECOG-2 (n=8) and ECOG-3 (n=4)). CMML-patients of the AAR suffered from renal insufficiency (n=8), mild liver disease (n=8), coronary artery disease (n=5), diabetes mellitus (n=5), a prior/concomitant solid tumor (n=3), COPD (n=2) and/or venous thromboembolic disease (n=1), respectively. There was no relevant difference in RBC transfusion dependence (TD) for MP-CMML vs. MD-CMML. However, PLT-TD at diagnosis was lower for MP-CMML (23%) than for MD-CMML (67%). In the 19 patients in whom pretreatment cytogenetics were performed, 15, 4 and 0 could be grouped into IPSS good, intermediate and poor risk categories, respectively. Only 57% of all included CMML-patients were treatment-naïve, whereas the rest were pretreated with ICT (10%), ESA (14%), intensive chemotherapy (21%) and other substances (17%) (mainly hydroxyurea and anagrelide) (numbers add up to more than 100% as several patients were pretreated with more than 1 agent). Thus, this registry more accurately reflects a real-life treatment scenario, than most clinical trials that have strict inclusion/exclusion criteria.

Most AZA-cycles were applied s.c. (98%), whereas 2% were applied i.v. Median and mean number of AZA-cycles was 5.5 and 9 (SD 10.61; range 1–43), respectively. The FDA-approved d1-7 schedule was used in 62% of all AZA cycles, whereas the non-approved alternative schedules d1-5, 5-2-2 and ‘others’ were applied in 21%, 12% and 6% of all cycles, respectively. Concerning the number of patients however, 20/26 predominantly received the d1-7 schedule. The FDA-approved target dose (75mg/m² over 7 days) was achieved in 52% of all cycles and in 17/26 of patients, respectively.

Reasons for termination of treatment with AZA were death (33%), disease progression (22%), transformation to AML (11%), no response (6%), toxicity (6%) and other reasons (22%).

Any kind of HI was noted in 12/26 (46%) of patients. When looking at each lineage separately, 8/26 had HI-ery, 7/26 HI-PLT and 9/26 HI-neutrophils. 6/14 patients who were RBC-TD and 3/7 patients who were PLT-TD prior to AZA-treatment achieved transfusion independence (TI). In patients, in whom (repetitive) bone marrow aspirations/biopsies were performed for response evaluation (n= 12), the following best marrow responses were observed: CR in 4/12 (33%), marrow CR 1/12 (8%), PR in 2/12 (16%), SD in 4/12 (33%) and primary PD in 1/12 (8%). The OR rate observed prior to 07/2011 was 44% (CR + marrow CR + PR + HI). At the time of writing, 9 patients had received <=2 cycles. When limiting response analysis to patients who received >2 AZA cycles, which is required for achievement of hematologic response by the IWG-criteria, ORR was 66%. These are amongst the highest response rates observed for CMML to date.

The median OS was 12.7 months (95%CI 6.9; 19.4). Median OS for patients demonstrating HI was significantly higher than for those who did not (19.4 vs. 5.6 months, p<0.0001). Detailed OS analyses and statistical analyses of various factors known or thought to influence prognosis are currently being analyzed, including MP-CMML vs. MD-CMML, CMML-1 vs. CMML-2, comparison of IPSS cytogenetic risk groups, as well as treatment-naïve vs. pretreated patients.

In conclusion, in this population of partly pretreated, comorbid, very old patients with CMML, AZA was well tolerated and yielded substantial clinical and hematological benefit.