Increased Plasma FLT3 Ligand Levels Following Chemotherapy May Interfere with the Clinical Efficacy of FLT3 Inhibitors.

Abstract 937

FLT3 mutations are relatively common in AML and are associated with a poor prognosis. A large number of different small molecule inhibitors of FLT3 tyrosine kinase activity have been identified and many are being studied in clinical trials. There are currently several trials in the planning stage or actively accruing in which the FLT3 inhibitor is administered along with a chemotherapy regimen. While FLT3 inhibitors have some degree of clinical activity when given as monotherapy, their efficacy in the setting of chemotherapy is not well understood. We examined in vivo FLT3 inhibition in AML patients treated with chemotherapy followed by the FLT3 inhibitor lestaurtinib (CEP-701), comparing newly-diagnosed AML patients (enrolled on the MRC AML15 trial) with relapsed patients enrolled on the Cephalon 204 trial. We noted that in vivo FLT3 inhibition by lestaurtinib (as measured by a plasma inhibitory activity assay) was less effective in the relapsed patients (204 trial) compared with the newly-diagnosed patients (AML15 trial). It has been noted by others that FLT3 ligand (FL) levels increase dramatically following myelosuppressive therapy. We hypothesized that FL could be interfering with the ability of lestaurtinib to inhibit FLT3 autophosphorylation, particularly in the patients enrolled on the 204 trial. Using an ELISA assay, we measured plasma FL levels in the two patient populations at Day 15 following the start of intensive chemotherapy (34 patients from the AML15 trial, 78 patients from the 204 trial). FL levels averaged 2.7 pg/mL in newly-diagnosed AML and 15.6 pg/mL in AML patients at first relapse (prior to therapy). FL levels rose to a mean of 402 pg/mL (range 3-3183 pg/mL) on day 15 of induction therapy for the AML15 trial (newly diagnosed) patients, while they rose to a mean of 1174 pg/mL (range 18.6-3818) in the 204 trial (relapsed) patients. In many patients on the 204 (relapsed) trial, FL levels remained elevated at day 42 of therapy (mean 641 pg/mL, range 4.4-5767) pg/mL). Following successive courses of consolidation chemotherapy, FL levels continued to rise higher with each course of chemotherapy in patients on AML15, with a mean of 3151 pg/mL after the 4^th course. On comparing individual samples from both trials, we noted that samples with similar levels of drug but different levels of FL showed different degrees of FLT3 inhibition, with FL levels seeming to exert a negative influence on inhibition. To explore this phenomenon in vitro, we tested the influence of exogenous FL on FLT3 inhibition and cytotoxic effects exerted by lestaurtinib and 4 other FLT3 inhibitors- AC220, KW-2449, PKC-412, and sorafenib. In culture medium using FLT3 mutant cell lines (Molm14, TF/ITD), the IC50 for cytotoxic effect of all inhibitors shifted upwards 2-4-fold in the presence of exogenous FL at concentrations of 1000-3000 pg/mL (the levels we observed in patients following chemotherapy). A similar degree of shift was observed in immunoblot assays of FLT3 autophosphorylation using the same cell lines. In primary samples from AML patients that expressed predominantly the mutant allele (i.e., hemi- or homozygous FLT3/ITD), exogenous FL had similar effect as that seen in the cell lines: a modest effect on FLT3 autophosporylation, and a 2-4-fold shift upward of the cytotoxicity IC50 of FLT3 inhibitors. However, a somewhat different pattern was observed in samples from patients with heterozygous FLT3/ITD mutations (i.e., both mutant and wild type allele present). The addition of exogenous FL increased the degree of FLT3 autophosphorylation, and dramatically increased cell survival in culture medium. The enhanced survival was abrogated by FLT3 inhibition. These findings have important implications regarding trial design for FLT3 mutant AML. It is possible that any FLT3 inhibitor administered concomitantly with chemotherapy could be affected by this phenomenon, with only the most potent inhibitors being able to overcome the effects of FL in vivo. Furthermore, newly diagnosed FLT3/ITD patients (in whom the mutations are commonly heterozygous) might be predicted to respond differently to FLT3 inhibition according to the context in which the inhibitor is given- i.e., in the presence of absence of chemotherapy. Finally, we might speculate that FL is an important driver in FLT3 mutant AML, and may have a strong influence on prognosis. This hypothesis should be tested prospectively by examining FL levels in ingoing clinical trials in which FLT3 mutant AML patients are enrolled.