Phase II Trial of Low Dose, Subcutaneous Decitabine in Myelofibrosis

DNA hypermethylation of promoter-specific CpG islands has been implicated in the pathogenesis and progression of myelofibrosis (MF). We have evaluated Decitabine, a DNA methyltransferase inhibitor in patients (pts) with MF, including primary MF (PMF), and MF arising after polycythemia vera or essential thrombocythemia (post-PV or post-ET MF). This was a multicenter, single stage Phase II trial, with a planned accrual of 20 patients. The regimen would be deemed worthy of further investigation if ≥ 5 pts responded. Eligibility criteria included myelofibrosis associated with anemia (hemoglobin <11g/dL) and/or palpable splenomegaly. Decitabine was administered subcutaneously (SQ) at a dose of 0.3mg/kg/d on days 1–5 and days 8–12; cycles were repeated every 6 weeks, in the absence of dose limiting toxicities. Response was determined every 12 weeks as an improvement in cytopenias and/or splenomegaly. A maximum of 9 cycles was allowed; pts who had a complete remission (CR) had treatment discontinued provided they had received at least 4 cycles of therapy. Peripheral blood obtained at baseline and on days 5 and day 12 of the first 2 cycles of therapy was analyzed for potential biomarkers predictive of response to decitabine. Biomarkers evaluated included CD34+ cells measured by flow cytometry, as elevated CD34+ counts have been associated with advanced stage and evolution to blast phase in MF. CXCR4 gene expression levels were measured by REAL-time RT-PCR; reduced expression of CXCR4 on CD34+ cells has been linked to the abnormal stem cell trafficking in this disease, and decitabine has been shown to upregulate CXCR4 levels in primary MF cells in vitro. Hemoglobin F levels were evaluated by HPLC as induction of hemoglobin F levels has been demonstrated with decitabine therapy in primates and in patients with sickle hemoglobinopathy. Twenty-one pts were enrolled on the study. Pt characteristics: M/F:12/9, median age 67 years (range 42–89), median absolute CD34+ cell count 350 × 10⁶/L (1.2–4959), Dupriez score of 2, 1 and 0 in 24%, 57% and 19% respectively, PMF= 76%, post PV-MF=19%; post ET MF=5%; 4 pts had blast phase disease (blast phase- MF) and 12 pts (57%) had transfusion dependent anemia and/or thrombocytopenia. Eight pts (38%) were previously untreated and 38% had abnormal bone marrow karyotype at baseline. Median number of cycles administered was 4 (range 1–9) and four pts remain on treatment. Grade 3/4 neutropenia (ANC) and grade3/4 thrombocytopenia occurred in 95% and 52% of pts respectively. Nine pts have developed febrile neutropenia. Two pts have died of sepsis-related complications while on study, both pts had significant impairment of hematopoiesis at baseline: 1 had blast phase-MF, the other had advanced PMF with a baseline ANC of 50/μL. Drug related non-hematologic toxicities have been infrequent and include gradeI/II fatigue and liver function abnormalities. 19 pts are evaluable for response: A total of 7 pts (37%) have responded, including 2 with blast phase-MF. CR=1 (normalization of counts and transfusion-independence), PR=2 (hemoglobin increase to normal levels, multilineage improvement including ANC and/or platelets). Hematologic improvement in erythroid lineage n=2: (both of these patients achieved red cell transfusion- independence) and platelets n=2: (>50% increase in platelet levels) have also been observed. Median time to response was 2 cycles (range 1–6); median duration of response was 5 months (range 2–15). Two pts are maintaining their responses at 2 and 14 months. All responders who developed disease progression did so while off therapy. Analysis of the effects of decitabine on CD34+ cells revealed a 61% reduction (p<0.001) in mean levels between cycles 1 and 2 of therapy in responders (n=7), in contrast to non-responders (n=12), in whom there was no change (p=0.45). There was no statistically significant change in CXCR4 or hemoglobin F levels over time, but responders to decitabine had higher levels of hemoglobin F (>1%) than non-responders at baseline and post-therapy (p<0.01)

Conclusions: Low dose SQ decitabine is feasible, has clinical activity and deserves further investigation in myelofibrosis. Myelosuppression is common and requires close monitoring. A decline in circulating CD34+ progenitor cells that persists into cycle 2 of therapy may serve as a novel biomarker that predicts response to decitabine in myelofibrosis.