Long Term Followup and Patterns of Failure in Patients with Acute Myeloid Leukemia (AML) and High Risk Myelodysplastic Syndrome (MDS) Treated On Studies Combining a Hypomethylating Agent and the Histone Deacetylase Inhibitor (HDACi) Valproic Acid.

Abstract 2074

Poster Board II-51

Hypomethylating agents such as 5-aza-2'-deoxycitidine (decitabine) or 5-azacytidine (vidaza) are well tolerated, produce durable responses, and have become the standard of care in the treatment of MDS. Combination epigenetic therapy using hypomethylating agents and HDACi has also been shown to have significant activity, excellent tolerability, and low induction mortality in the treatment of elderly patients (pts) with high-risk MDS and AML. However, longer term followup from these studies is not available. We conducted 2 clinical trials of combination epigenetic therapy using the HDACi valproic acid (VPA). The 1st study (Blood 2006;108:3271-9), in which VPA was combined with decitabine, included 14 pts age older than 60 years (yrs) who had not received prior therapy. In the 2nd study (Blood 2007;110:2302-8), VPA was combined with 5-azacytidine and all-trans-retinoic-acid and included 33 previously untreated pts older than 60 yrs. Here, we present the long term follow-up of these 47 pts. with a 216 week (wk) duration of followup. Pt characteristics are as follows: 31 pts were male (66%) and the median age of the group was 71 yrs (range 39-84). Forty-four (94%) had AML, 2 (4%) had MDS, and 1 (2%) had chronic myelomonocytic leukemia (CMML). Cytogenetics were diploid or –Y in 21 (45%), -5/-7 in 10 (21%), +8 as sole abnormality in 2 (4%), 11q in 1 (2%), complex in 3 (6%), other in 8 (17%), and not done (ND) in 2 (4%). At the start of therapy, the median WBC was 6.4 (0.7-79.1), median bone marrow blast count was 35% (8-85), and the pts received a median of 3 (1-14) cycles. Of 46 evaluable pts, 19 (41%) achieved a response including 16 CR (35%), 1 CRp (2%), and 2 PR (4%). The median duration of response (CRD) was 36 wks (1-120) and the median overall survival (OS) was 39 wks (2-216). Interestingly, of the 4 pts with erythroleukemia (AML-M6), 3 (75%) achieved CR, with median OS of 65 wks and median CRD of 42 wks. The 4th is still alive as of last follow-up with an OS of 216+ wks. We then analyzed the characteristics of the responding vs. non-responding pts (Table 1). Of the 19 responding pts, all have died at last followup. There were 3 (14%) deaths in CR. The patterns of failure in responders were: 7 (37%) with progressive AML, 6 (32%) with prolonged cytopenias and unable to receive more therapy, 3 (16%) stopped therapy by pt choice, 1 (5%) died of infection and sepsis while in CR, and 2 (11%) died in CR of unrelated causes. Of 13 responding pts who were candidates for further therapy after failing, salvage regimens were as follows: ‘high intensity therapy’ (eg. high-dose ara-C-based induction) in 4 (31%), ‘low intensity therapy’ (oral vidaza, revlimid, sapacitabine, low-dose araC) in 6 (46%), and no further therapy or supportive care in 3 (23%). Only 1 of 4 (25%) pts receiving high intensity therapy responded and had a short CR duration of 3 months. Combination therapy with a hypomethylating agent and VPA has a high response rate with durable remissions in untreated elderly pts with AML. There is a suggestion that this combination is very active in pts with AML-M6 and should be explored further. Non-responding pts generally had a higher WBC and higher bone marrow blast count at the start of therapy. Responders received more cycles of therapy and had significantly longer OS. Most pts who failed therapy had either progression of their AML or prolonged cytopenias that precluded further therapy and led to the associated complications of infections and bleeding. Pts who relapse after combination epigenetic therapy have a poor prognosis, low response to high intensity therapy, and should be considered for investigational approaches.