Dasatinib Plus Smoothened (SMO) Inhibitor BMS-833923 in Chronic Myeloid Leukemia (CML) with Resistance or Suboptimal Response to a Prior Tyrosine Kinase Inhibitor (TKI): Phase I Study CA180323

Background: Dasatinib is a potent inhibitor of BCR-ABL and SRC-family kinases, but may not eradicate noncycling early progenitor cells (Konig, Cancer Res 2008). Hedgehog signaling is implicated in CML progenitor cell self-renewal (Dierks, Cancer Cell 2008; Long, J Exp Clin Cancer Res 2011), and may be inhibited by SMO antagonists (Zhao, Nature 2009). BMS-833923 is a potent orally-available SMO inhibitor (SMOi) with activity in selected solid tumors. In a phase 1 pharmacokinetic (PK) and pharmacodynamic study, we determine if addition of BMS-833923 to dasatinib improves control of CML.

Methods: In study CA180323 (NCT 01218477), dasatinib plus BMS-833923 was evaluated using a “Rolling-Six” design. Patients (pts) with chronic phase (CP) CML and resistance or suboptimal response to any prior TKI or with either advanced phase (accelerated phase/blast phase [AP/BP]) CML or Ph+ acute lymphoblastic leukemia (Ph+ALL) resistant to imatinib or nilotinib were treated with dasatinib (100 mg/d in CML-CP, 140 mg/d in CML-AP/BP or Ph+ ALL). After 4 weeks, SMOi was added at dose levels (DLs) of 100 mg/d (DL1), 200 mg/d (DL2) or 50 mg/d (DL-1). Dose-limiting toxicities (DLT, definition also included grade 2 events requiring >7 days interruption) were defined in the first 4 weeks of combination therapy (weeks 5–8). For the purpose of SMO escalation or reduction, adverse events occurring at any time were also considered; pts discontinuing for disease prior to week 8 were replaced. PK was performed for both agents. Serial replating of marrow colony-forming cells (CFC), with fluorescence in situ hybridization (FISH) for BCR-ABL in secondary progeny, was performed to detect reduction of early Ph+ progenitors.

Results: Twenty-seven pts were enrolled (19 CML-CP, 4 CML-AP/BP, 4 Ph+ ALL), of whom 5 discontinued before week 8 (4 with early progression and 1 ineligible), leaving 22 pts evaluable for DLT. DLT was reported in 2 of 12 pts at DL1, (1 CML-CP pt with confusion, somnolence, fatigue, nausea and 1 Ph+ ALL pt with anorexia, dysgeusia, asthenia), and 1 of 2 pts at DL2 (CML-CP pt with joint stiffness, muscle cramps, and peripheral sensory neuropathy). Although a maximum tolerated dose was not exceeded, frequent interruption and dose reduction led to exploration of DL-1; no DLT was experienced in 8 CML-CP pts at DL-1. Adverse events included previously reported toxicities of SMOi: dysgeusia (67%), alopecia (67%), anorexia or nausea with weight loss (42%), muscle spasms (33%), and fatigue (33%) were the most common, leading to discontinuation in 7 pts (32%). Both onset and resolution of alopecia were delayed >2 months, consistent with a stem cell effect. No drug-drug PK interactions were identified. Minor efficacy could be attributed to addition of SMOi; 3 of 10 pts with CML-CP resistant to dasatinib had prolonged (47–51 weeks) combination treatment and 1 attained complete cytogenetic response. No pt with CML-Adv or Ph+ALL showed benefit. Replating CFC assays with FISH on secondary progeny did not demonstrate preferential reduction of early BCR-ABL progenitors by addition of SMOi.

Conclusions: Despite observation of expected toxicities due to SMO inhibition, no evidence of efficacy was shown for the combination. Extensive replating assays failed to demonstrate selective reduction of CML progenitors. Although SMOi was reasonably tolerable at a dose of 50 mg/d, the lack of efficacy and observed toxicities do not support use in early stage CML-CP. Translation of preclinical work on CML stem cell eradication to the clinical setting remains challenging.