Incorporating a FLT3 inhibitor into AML therapy

Abstract

7060 Background: Activating mutations of FLT3 occur in 30% of adults with AML and are associated with worse survival. Lestaurtinib (CEP-701), an oral FLT3 kinase inhibitor, has activity as monotherapy in relapsed/refractory AML patients with FLT3 mutations. Using the results of pre-clinical models, a trial was designed in which patients with AML and FLT3 mutations in first relapse were randomized 1:1 to receive chemotherapy alone or chemotherapy followed by treatment with lestaurtinib. The primary endpoint was complete remission (CR), with a target accrual of 220 patients. We present here correlative data from this trial. Methods: Efficacy of FLT3 inhibition was measured using the plasma inhibitory activity (PIA) assay for FLT3. This assay method has been previously published (Blood 108(10):3477–83). Prior to initiation of chemotherapy, leukemia cells of the patients randomized to lestaurtinib were tested for in vitro sensitivity to lestaurtinib using a cytotoxicity assay. Plasma samples were obtained at baseline and at trough time points on Days 15 and 42. PIA assay and in vitro cytotoxicity results were correlated with the clinical results from the first 40 patients enrolled. In addition, plasma levels of lestaurtinib and alpha-1 acid glycoprotein (AAG; the principle plasma binding protein for this drug) were measured. Results: Accrual is ongoing. Twenty-three of 35 patients treated with lestaurtinib achieved a PIA result of 85% or greater (i.e., successful FLT3 inhibition). Of the first 40 patients enrolled, all those who achieved a FLT3 PIA results of 85% or greater and whose pretreatment leukemia cells were sensitive in vitro to lestaurtinib achieved a clinical response. Conversely, patients with insensitive cells or inadequate PIA did not respond. Ten of 17 patients randomized to lestaurtinib showed response (5 CR, 3 CRp, 2 PR). Four of 17 patients randomized to receive chemotherapy alone responded (2 CR, 2 CRp). Conventional pharmacokinetic analysis showed no correlation whatsoever with response, due to the high degree to which lestaurtinib is bound to AAG. Conclusions: These results demonstrate that the PIA assay for FLT3 is able to accurately predict response to lestaurtinib, and provides important evidence that targeting FLT3 remains a valid therapeutic approach for AML with FLT3 activating mutations. No significant financial relationships to disclose.