Phase 2 Trial of the Histone Deacetylase Inhibitor Valproic Acid as a Monotherapy or in Combination with All-Trans Retinoic Acid in 24 Patients with Acute Myeloid Leukemia.

Valproic acid (VPA) has been shown to inhibit histone deacetylase activity, and to synergize with ATRA in the differentiation induction of leukemic myeloid blast cells in vitro. We applied VPA to 20 patients (16 sAML/MDS, 2 de-novo-AML, 2 sAML/OMF) too old or physically unfit to receive intensive chemotherapy. VPA monotherapy was targeted to reach serum concentrations of 50–100mg/ml. ATRA was added (80mg/m2/d in two divided doses, every other week) in some of the patients who did not respond or who relapsed. To enhance responses, we treated an additional 4 patients (2 sAML/MDS, 1 sAML/ET, 1 de novo AML) with VPA+ATRA from the start. Median age was 70 years (51–84). Median bone marrow blast count was 30% (10–80). 5 patients had only 10–15% marrow blasts but were included because they showed treatment failure or relapse after chemotherapy and were unable to receive further cytotoxic treatment. Median treatment duration was 99 days (20–396) for VPA and 79 days (18–339) for ATRA. Responses according to international working group (IWG, Cheson et al., 2003) criteria were observed in 5 patients (25%) on VPA monotherapy (4PR, 1CR). Of the responding patients two have ongoing responses (CR, PR) for 12 and 13 months, respectively. 1 patient reaching PR discontinued VPA when her physical condition had improved sufficiently to allow further chemotherapy. 1 patient relapsed after 2 months and was switched to VPA+ATRA, without response. 1 patient died of infectious complications. 8 additional patients showed stable disease without increases in peripheral blast count. Responses lasted for a median of 4 months (2–13). Among the 4 patients receiving VPA+ATRA from the start, 1 (25%) achieved PR. When he stopped VPA after 3 months because of side effects, he continued on ATRA, achieving a CRi (CR with incomplete recovery of platelets) lasting for 8 months. 4 of 14 nonresponders were switched to VPA+ATRA, but none of them showed a response. Response to VPA treatment was not associated with FAB subtype or karyotype. Median bone marrow blast count was 28 (13–45)% in responders, 30 (10–75)% in patients with stable and 41 (25–80)% in patients with progressive disease. Since our patients mainly had secondary AML, we also analyzed our results according to the proposals of the IWG for MDS (Cheson et al., 2000). Among patients receiving VPA monotherapy 1 patient had a major trilineage response. 2 patients showed a minor erythroid and one a minor neutrophil response. In the second group of patients one had a major erythroid response. Concerning side effects, VPA caused tremor in four cases, leading to cessation of treatment in two. Regarding ATRA, grade 1–2 skin toxicity was observed in 4, grade 1–2 gastrointestinal toxicity in 2, and pleural effusion in 1 patient. In summary, we observed responses according to IWG criteria in 25% of our patients (6/24). The best responses to VPA or VPA+ATRA in AML patients occurred in patients with low blast count, mainly in patients who showed relapsed or refractory disease shortly after intensive chemotherapy. These data indicate that VPA might be most effectively applied after or in addition to intensive chemotherapy.