A Proof of Principle Clinical Trial in Myelodysplastic Syndromes of Non-Cytotoxic Differentiation Therapy with Decitabine,

Abstract 3830

In myeloid malignancies, genetic abnormalities in key apoptosis pathway genes (eg., p53) are associated with poor responses to conventional cytotoxic therapy. In pre-clinical models, non-cytotoxic, DNA methyltransferase 1 (DNMT1) depleting regimens of the deoxycytidine analogue decitabine relieve aberrant epigenetic repression of key late-differentiation genes and induce cell cycle exit by p53-independent differentiation pathways (CEBPE, MXD1, p27/CDKN1B) (Ng et al, Leukemia 2011). To translate these observations into practice, a clinical trial is being conducted in MDS, using decitabine at minimum doses required to deplete DNMT1 (0.1–0.2 mg/kg [5–10 mg/m²]), administered by the subcutaneous (SC) route to avoid high peak levels that cause apoptosis, and using a metronomic schedule (1-3X/week for ≥1y) to increase exposure time for S-phase specific depletion of DNMT1. To evaluate mechanism of action, correlative studies include quantification of pH2AX (DNA damage marker) and DNMT1 levels in bone marrow by flow-cytometry, and immunohistochemical evaluation by ImageQuant of p27/CDKN1B and KI67 expression, expected to vary directly and inversely respectively with terminal differentiation. A two-stage Simon design was used, and results from the first stage (n=15, patient characteristics table 1 ) are reported (median follow-up 330 days, range 142–180). Anti-emetics were not required, and there were no administration related adverse events. Neutropenic fever (NF) occurred in 11 patients, 7 of whom did not have NF prior to therapy (median time to nadir 40 days). By IWG criteria, complete hematologic and cytogenetic remissions (CR) with persistent dysplasia occurred in 2 subjects, hematologic improvement (HI) occurred in 4 subjects (overall response rate, ORR=40%), and stable disease in 7. Median response duration for HI/CR is 243 days, with 5 of 6 responses ongoing (range 74–292). Complete cytogenetic responses occurred even in patients with highly complex chromosome abnormalities (table 1 ). Bone marrow cell pH2AX expression decreased non-significantly from pre-treatment to week 6 to week 12 (median values 1.2, 0.5 and 0.4% respectively, p=0.27 Wilcoxon test), with a >3-fold reduction in mean percentage of cells expressing DNMT1 in the same period (Turkey-Kramer test p<0.001). Consistent with differentiation-mediated cell cycle exit, median p27/CDKN1B expression increased from 26.8 to 66.2 to 78.7% of bone marrow cells (p<0.001), with a concomitant decrease in median KI67 expression from 65.8 to 46.2 to 28.7% (p<0.001) (table 1 ). ORR of only 40% despite major p27 and cytogenetic responses and stable disease in almost all subjects (table 1 ), suggested that relief of cytopenia may require a threshold of normal stem cell reserve and a supportive marrow microenvironment. Accordingly, in HI/CR versus other subjects, median duration of disease was 855 versus 1350 days (p=0.15), and median pre-treatment bone marrow cellularity was 65 versus 30% (p=0.14). In conclusion, this study provides clinical proof of principle that a decitabine regimen rationalized for non-cytotoxic epigenetic-differentiation effects is active in myeloid malignancy, correlates with molecular markers of terminal differentiation, and has potentially important safety and efficacy advantages over cytotoxic therapy that warrant further evaluation and optimization.

All subjects: baseline and response characteristics