Phospho-Specific Flow Cytometry of Fixed Whole Blood During AML Clinical Trials to Monitor In Vivo FLT3 Inhibition in Leukemic Blasts,

Abstract

Abstract 3508Although first-generation FLT3 inhibitors may have had limited anti-leukemic effects due to suboptimal target inhibition, newer drugs such as AC220 and KW-2449 have substantially greater in vitro potency and bioavailability. Ex vivo assays such as the plasma inhibition assay (PIA) are useful to estimate free drug bioavailability, but direct confirmation of biochemical FLT3 inhibition in leukemic blasts in vivo has proven more challenging to employ systematically for drug development. Here we report the development of a fixed whole blood intracellular flow cytometry platform to measure real-time signal inhibition during a clinical trial of the second-generation FLT3 inhibitor KW-2449. Methods:Anticoagulated blood samples were aliquoted into FACS tubes within four hours of collection; a subset was exposed to signaling inhibitors (KW-2449, rapamycin × 30 min.) or activators (phorbol ester/PMA or FLT3 ligand/FL × 10 min.) to establish dynamic controls. Following incubation, samples were formaldehyde-fixed, red cells were lysed with the permeabilizing agent triton X-100, and specimens were stored at −20C in glycerol medium. Subjects’ samples from all time points were simultaneously thawed, denatured with ice-cold methanol, and stained with a single cocktail of antibodies. Blasts were identified by CD45 and side scatter (SSC) and confirmed by multiple surface markers (CD33, CD34, CD117, HLA-DR, etc.). Positive gates for phospho-proteins were created by comparing blasts in stimulated and unstimulated conditions and/or autofluorescence (FMO) controls. Results:Despite adequate controls, flow demonstrated limited changes in FLT3-ITD+ blasts' pSTAT5 signal following either FL stimulation or ex vivo KW-2449 treatment of these peripheral blood primary samples. This contrasted with the FLT3-ITD+ cell line Molm14, in which FLT3 inhibition reduced pSTAT5. However, the PI3K/AKT/mTOR downstream target ribosomal protein S6 (S6) was consistently observed to be constitutively phosphorylated in both Molm14 cells and peripheral blood FLT3-ITD+ AML blasts. pS6 in all FLT3-ITD+ samples markedly augmented with ex vivo FL, and decreased following ex vivo KW-2449 treatment. We therefore serially monitored S6 phosphorylation during therapy on a phase 1/2 trial of KW-2449.In this clinical trial, subjects were treated with KW-2449 every 6–8 hours, due to the drug’s relatively short half life. 10 subjects (9 FLT3-ITD+, 1 FLT3-WT) provided serial blood samples for analysis. All FLT3-ITD+ subjects had blasts identifiable by morphology and immunophenotype. Samples with as few as 500 blasts/uL were informative for pS6. In all cases, blasts showed dynamic changes in pS6 in response to ex vivo FL. As previously described using intracellular flow cytometry, pS6 in primary AML samples was heterogeneous, and, at basal state, frequently only demonstrable in a subset of blasts. We observed constitutive S6 phosphorylation in 8/9 subjects’ leukemic cells. The mean percentage of blasts with constitutive pS6 was 21% (median 7%, range 5–70%). To directly quantify FLT3 kinase inhibition in vivo, we serially monitored pS6 in blasts by flow prior to and following their initial oral KW-2449 dose. In 8/8 patients with baseline constitutive S6 phosphorylation, blood obtained two hours following the initial dose showed marked reduction in the percentage of pS6+ blasts to a mean of 3.8% (median 1.3% range 0.1 to 20%). This reflected an 83% mean reduction in the percentage of pS6+ blasts. PIA was performed in 8/9 of FLT3-ITD+ subjects and confirmed that potent FLT3-inhibitory concentrations were present 2 hours after a single dose of KW-2449 (mean reduction from baseline of 79% for pFLT3 and 88% for pSTAT5). Two subjects’ samples were followed serially by flow cytometry throughout the dosing interval. One showed sustained inhibition (consistent with concurrent PIA), while in the other, pS6 returned to baseline within 4–6 hours of the initial dose (concurrent PIA not done). Summary:We confirm that PI3K/AKT/mTOR is a major downstream pathway of FLT3 signaling in primary AML samples. We further demonstrate the feasibility of intracellular flow cytometry for S6 phosphorylation to monitor the biochemical efficacy of FLT3 inhibitors in patients. Studies are underway to correlate biochemical FLT3 inhibition by flow cytometry with clinical response/resistance to KW-2449 and other FLT3 inhibitors. Disclosures:Sato:Kyowa Hakko Kirin Co., LTD: Employment. Akinaga:Kyowa Hakko Kirin Co., LTD: Employment. Rao:Kyowa Hakko Kirin Co., LTD: Employment. Levis:Kyowa Hakko Kirin Co., LTD: Research Funding; Ambit Biosciences: Consultancy. Carroll:Kyowa Hakko Kirin Co., LTD: Research Funding.