5′ Azacitidine in Combination with Valproic Acid Induces Complete Remissions in Patients with Advanced Acute Myeloid Leukaemia but Does Not Eradicate Clonal Leukaemic Stem/Progenitor Cells.

Demethylating drugs (e.g. 5′ Azacitidine) and histone deacetylation inhibitors are a relatively new class of clinically active agents for acute myeloid leukaemia (AML) and high-risk myelodysplasia (MDS) that have engendered a lot of interest. However, their mechanisms of action (in terms of achieving a clinical response), and how best to use them, both remain unclear. These agents rarely cure patients; in most cases the disease relapses. To better understand the impact of these agents on haemopoiesis, we have treated 45 patients with high risk AML/MDS (30 relapsed/refractory AML, 6 de novo AML, 9 high risk MDS) with azacitidine (75 mg/m2 5 days every 28 days), sodium valproate, all-trans retinoic acid and theophylline and correlated clinical response with immunophenotypic and clonogenic analysis of committed and multi-potent leukemic stem/progenitor cells. The median age of the cohort was 66 years (range 32–85 years). Clinical responses were assessed using Cheson criteria. Responses were observed in 15 patients; 7 achieved a complete remission (CR) or CRi and 8 a partial response (>50% reduction in bone marrow blasts). The median time to achievement of maximal response was 2 cycles (range 1–6). Cytogenetic data in 40/45 patients studied showed abnormalities in 60% (24/40) of patients of which 50% (12/24) were adverse risk. 40% (6/15) patients achieving CR, CRi or PR had adverse risk cytogenetics. The most common grade 3/4 haematological toxicities were thrombocytopenia (51%), leukopenia (42%) and neutropenia (42%). Common non-haematologic toxicities were fatigue (31%), neutropenic sepsis (20%), chest infection (16%) and diarrhoea (13%). Detailed immunophenotypic analysis, clonal assays (methyl cellulose and cobblestone-area forming colony assays, CAFC) and FISH analysis were performed on stem cell enriched (CD34+CD38−), common myeloid progenitor (CMP), granulocyte-monocyte progenitor (GMP) and megakaryocyte erythroid progenitor (MEP) compartments at diagnosis and post-treatment. In all diagnostic samples there were abnormal immunophenotypic CD34+CD38-, CMP, GMP and MEP (myeloid progenitor) compartments. There was an anticipated failure of normal colony and CAFC growth from sorted CD34+CD38- and myeloid progenitor populations. In responding, but not non-responding, patients there was partial or complete restoration of normal immunophenotypic populations accompanied by normal haemopoietic colony output. However, we extended these analyses to look at the kinetics of change in clonally abnormal cells that remain after treatment in total bone marrow mononuclear cells and FACS sorted cells CD34+CD38- and myeloid progenitor populations from one informative AML patient with an abnormal karyotype at diagnosis, who had a complete morphological response. At diagnosis, cytogenetic FISH analysis on sorted cells showed that 90–100% cells were clonally abnormal in the CD34+CD38- and myeloid progenitor compartments. At the beginning of cycle 3, there was

• complete haematological response (normal blood count and bone marrow myelogram and morphology):

• karyotypic remission in bone marrow mononuclear cells (50 cells examined);

• restoration of normal colony output; and

• normal immunophenotypic pattern of stem/progenitor populations.

However, in sorted stem/progenitor populations, clonally abnormal cells were still detected by FISH at ~15% level. These abnormal cells provide a cellular substrate for future loss of complete clinical response. This was seen by cycle 6. Now the patient developed

• peripheral cytopenias;

• reduction in normal colony output;

• increased bone marrow dysplasia but no excess of blasts.

The bone marrow mononuclear karyotype was still normal. However, now there was an increase to 100% clonally abnormal cells in the GMP compartment whereas the CD34+CD38- compartment there were still 15% clonally abnormal cells. Though information from additional patients under study will be critical to understand the generality of these findings, these data provide the first insights into the impact of treatment and the cellular kinetics of relapse. In conclusion, this is the first study to document the clinical and biological impact on stem/progenitor and peripheral haemopoiesis of this class of agent, or indeed of any drug in AML and MDS. The findings from studies like this will have general implications on how to assess the clinical and biological impact of therapies in AML and MDS.