Terminal Differentiation of FLT3/ITD AML Blasts In Patients Treated with the FLT3 Inhibitor AC220

AML is characterized by abnormal proliferation of myeloid cells that have a block in differentiation. FLT3/ITD mutations are relatively common in AML, and previous in vitro studies have demonstrated that signaling from ITD-mutated FLT3 blocks myeloid differentiation through repression of CEBP/a. As part of an ongoing phase 2 trial, we treated 6 patients with FLT3/ITD AML who were refractory to either primary induction therapy or salvage therapy after relapse with the highly potent and selective FLT3 inhibitor AC220. At the start of therapy, all 6 patients had circulating blasts (mean 9864 blasts/mm3; median 2970) and the median blast percentage in the bone marrow was 71.5%. Western blotting revealed a high level of sustained in vivo FLT3 inhibition in all patients. By Day 8, no patient had detectable blasts in the peripheral blood. After 14 days of treatment with AC220, all 6 patients displayed striking differentiation to the myelocyte stage within the bone marrow. By light microscopic evaluation of bone marrow aspirates, myelocytes (promyelocytes, myelocytes, and metamyelocytes) increased from a median of 10.5% pre-treatment to 52% after 2 weeks. Most patients were neutropenic on Day 1 of treatment (mean 574, median 560 neutrophils/mm3), but rose to a mean of 3275 neutrophils/mm3 after 4–8 weeks of treatment (median time to peak 34 days). By Day 28 of treatment, marrows were most often still hypercellular, but consisted primarily of fully differentiated neutrophils. Marrow blasts were markedly reduced or absent by Day 28 in all 6 cases (mean 2.3%, median 1.5%). In all 6 patients the FLT3/ITD mutation originally detected at the beginning of treatment was present in the marrow and peripheral blood despite the absence of circulating blasts after the first week of therapy. The FLT3 mutant allelic ratio did not change between pre-therapy and Day 28. Neutrophils were isolated to homogeneity (confirmed by cytospin) from peripheral blood by double ficoll density centrifugation. Using genomic DNA obtained from these purified neutrophils, we confirmed by PCR that the FLT3/ITD mutation was present, at a similar ratio as compared with the pre-treatment blasts. However, there was no detectable expression of FLT3 either by RNA (quantitative PCR) or protein (western blotting and flow cytometry) in these neutrophils. The isolated neutrophils morphologically resemble normal neutrophils by light microscopy, and by flow cytometry they express the differentiation antigen CD15 and CD11b, and have lost expression of immature markers such as cKIT and CD34. Stimulation of these neutrophils by endotoxin results in normal respiratory burst activity, as measured by reduction of nitroblue tetrazolium. They also express lactoferrin and MMP-9, proteins typically expressed in mature neutrophils. Clinically, lung nodules and fever occurred in 3 of the 6 patients within 14 days of the peak neutrophil count. They were not treated with steroids, but rather with antibiotics, and in all cases resolved. Other patients on this trial have developed Sweet's syndrome during the neutrophil surge. CEBPa transcript levels in Molm14 cells (an AML cell line with a FLT3/ITD mutation) rose 3–5-fold over baseline following treatment with AC220. This is consistent with our previously published findings, and suggests at least one mechanism for the observed release of the differentiation block observed in the AC220-treated patients. These clinical and correlative laboratory results suggest that effective, sustained in vivo FLT3 inhibition in AML patients with FLT3/ITD mutations induces terminal differentiation in blasts in many ways similar to that seen with all trans retinoic acid in acute promyelocytic leukemia. Furthermore, these findings demonstrate the direct link between the growth factor receptor pathway and control of differentiation, and provide new insight into mechanisms of leukemogenesis.

Levis:Ambit Biosciences, Inc: Consultancy.