Abstract 696


We have previously demonstrated that PPAR-γ agonists (such as pioglitazone) negatively regulate the level of Stat5A and Stat5Bgene expression in normal CD34+ bone marrow progenitors (Prost, Le Dantec et al. 2008).


We investigated whether targeting Stat5 expression with pioglitazone may impact clonogenic activity of Chronic Myelogenous Leukemia (CML) cells in vitro and result in molecular response improvement in vivo.

Patients and methods.

Preliminary in vitro studies tested the ability of pioglitazone to impact viability and clonogenicity of CD34+ primary cells from CML patients (pts). We conducted clonogenic and LTC-IC studies. Cultivated Ph+ -CD34+cells were characterized by facs and analyzed by CSFE assay. BCR-ABL and Stat5 expression were quantified by real time PCR. Control experiments were conducted using lentiviral vectors and siRNA assay for Stat5 and PPAR-g.

CML pts were eligible in the ACTIM trial (EudraCT 2009–011675-79) if they were treated by imatinib for at least 2 years with a stable daily dose for at least 3 months and in major molecular response without having achieved CMR4.5 (defined by a BCR-ABL/ABL IS ratio ≤ 0.0032%). After inclusion, pts received imatinib (no dose modification) and started pioglitazone (Actos®) 30 mg/d during 2 months and 45 mg/d thereafter for 12 months. BCR-ABL transcript level was monitored every 3 months during the study period. Primary objective was the proportion of patients achieving a confirmed undetectable level of BCR-ABL transcript. Secondary objectives included cumulative incidence of CMR4.5 and safety. A companion biologic study evaluated imatinib through levels, Stat5 expression in bone marrow at baseline, months 6 and 12 and clonogenic activity of bone marrow mononuclear cells at baseline, months 6 and 12.


From the in vitro studies we first demonstrated that pioglitazone at pharmacological doses inhibited cell growth of the Bcr-Abl positive cell line K562 through the activation of the PPAR-g/STAT5 pathway. We next showed that PPAR-g / STAT5 pathway induced a clonogenic defect in CD34+ cells from CML patients. Moreover, the activation of the PPAR-g / STAT5 pathway also induced a clonogenic and a proliferative defect in CML LTC-IC. We then confirmed that imatinib induced a selection of insensitive quiescent CML cells and showed that this effect was abrogated by the activation of the PPAR-g / STAT5 pathway.

Twenty seven pts were enrolled in the clinical trial and 24 were evaluable (1 was excluded in CMR4.5, 1 pt was not in MMR and 1 pt had consent withdrawal). Median age was 61.6 years (24.1–79) and median follow-up after inclusion was 13 months (9.8–21). All evaluable pts started pioglitazone as planned. Seven pts (29.2%) discontinued pioglitazone before 12 months, 6 following investigator decision after the warning of the French ministry of health regarding the risk of bladder cancer and 1 after its own decision. No pt discontinued due to adverse events. Discontinuations occurred between month 3 and month 9. Median duration of pioglitazone therapy was 11.2 months (2.6–15.4) median daily dose was 39.9 mg. No interaction was observed between imatinib and pioglitazone in term of through level before (median 850 ng/ml) and after (median 927 ng/ml) pioglitazone initiation (p=ns). Main adverse events were weight gain and worsening fluid retention in 3 pts. Three pts (14%) obtained a confirmed undetectable level of BCR-ABL transcript. The one year cumulative incidence of CMR4.5 was 57%. Stat5 mRNA quantification was significantly diminished in pt samples at M6 and M12 compared to the baseline values and a reduction of the clonogenic potential was also observed in bone marrow cells at M6 and M12. We collected “control pts” with similar characteristic (n=20). None of these pts obtained a confirmed CMR and the cumulative incidence of CMR4.5 in this control group was 27% as compared to 57% in the pioglitazone group (p=0.02).


We have extended our in vitro results showing that PPAR-γ agonists resulted in Stat5 down regulation in CML CD34+ cells and preferentially reduced their clonogenic and long term potency in CFCs and LTC-IC assays. We now demonstrated that these effects translate in vivo by the achievement of MMR in more than half of the pts treated with the combination of pioglitazone and imatinib suggesting that it may be possible to target quiescent CML cells in vivo and supporting the concept of stem cell pool erosion.


Off Label Use: pioglitazone in CML. Roy:Novartis, BMS: Speakers Bureau. Legros:Novartis, BMS: Research Funding, Speakers Bureau. Coiteux:Novartis, BMS: Speakers Bureau. Mahon:Novartis, BMS: Consultancy, Research Funding.

Author notes


Asterisk with author names denotes non-ASH members.