FLT3 Inhibition and Retinoid Signaling Overcome Stromal Protection to Target FLT3/ITD-Expressing Leukemia Stem Cells in the Bone Marrow Microenvironment

Abstract

While novel FLT3 tyrosine kinase inhibitors (TKIs) are increasingly efficient, when used as monotherapy they achieve only limited clinical responses in patients with FLT3/ITD acute myeloid leukemia (AML). Leukemia stem cells (LSCs) that share characteristics with normal hematopoietic stem cells have been implicated in persistence of minimal residual disease (MRD) and resistance to chemotherapy. Elimination of LSCs is paramount for any curative therapy in AML. Retinoic acid (RA) pathways are essential for normal and malignant stem cell homeostasis (Ghiaur G et al. 2013, Su M et al. 2015). We and others have previously reported that combining all-trans RA (ATRA) with FLT3 inhibitors leads to synergistic FLT3/ITD+ cell killing (Ma H et al. 2013, Chi et al. 2015). However, bone marrow (BM) mesenchymal cells were shown to inactivate ATRA via expression of CYP26 and thus protect LSCs from pro-differentiation effects. Here we investigated whether inhibition of CYP26 via talarozole and/or use of the CYP-resistant retinoid tamibarotene can overcome stromal protection and restore the full cytotoxic effect of the TKIs against FLT3/ITD+ LSCs in the stem cell niche.The combinatorial effect of FLT3 TKIs, retinoids, and talarozole was assessed in FLT3/ITD+ AML patient samples and cell lines in liquid as well as stroma co-culture. Our studies reveal synergistic activity against FLT3/ITD+ cells between FLT3 TKIs and ATRA in liquid culture, with combination index (CI) values of 0.1-0.7, leading to a significant induction of apoptosis. This treatment was rendered inefficient by co-culture in the presence of BM stroma (1.7- to 3-fold reduction in AnnexinV+ cells, P < 0.001). Inhibition of stromal CYP26 via talarozole or by-passing stromal CYP26 via a CYP26 resistant retinoid restored the TKI's ability to induce apoptosis of FLT3/ITD+ cell lines and patient samples cultured on stroma to a similar level as achieved with a FLT3 inhibitor alone in liquid culture. Colony-forming unit (CFU) assays further demonstrated decreased clonogenicity of FLT3/ITD+ cells on stroma co-culture upon treatment with TKI, ATRA, and talarozole (85% reduction in CFUs vs. 57% without talarozole, P < 0.001) or FLT3 TKI plus tamibarotene (86% reduction vs. 50% with ATRA, P < 0.01). Similar to treatment with CYP26 inhibitor, genetic deletion of CYP26 in the BM stroma also partially rescued the combinatorial effect between ATRA and TKIs otherwise lost in stroma co-culture conditions.Using various mouse models of FLT3/ITD leukemia (Molm14 and patient sample xenografts, transgenic FLT3/ITD;NUP98HOXD13), we have observed that as with AML patients, treatment with FLT3 TKIs alone is not sufficient to eliminate MRD and cure the disease. Addition of ATRA to the in vivo treatment with sorafenib (a FDA approved TKI with activity against FLT3) greatly decreased the level of engraftment of leukemia cells in mice, and significantly increased median survival compared to either drug alone. Furthermore, there was significant depletion of the LSCs as measured by limiting dilution transplantation of BM from mice treated with sorafenib and ATRA in combination. Nevertheless, elimination of the last bastion of LSCs remained elusive even after combination therapy with ATRA and sorafenib in these mice. Our studies suggest that this effect may be due to the enhanced metabolism of retinoids in the BM, thereby diminishing the combinatorial effect of retinoids and FLT3 TKIs. Therefore, the use of CYP-resistant retinoids or CYP inhibitors in combination with TKIs may improve the cure rate of FLT3-mutant AML in mouse models. We propose the existence of RA-low microenvironments in the BM where the combinatorial activity with TKIs against FLT3/ITD is impaired and thus, persistence of MRD is possible. Our findings support the development of a clinical trial of FLT3 TKIs, retinoids, and/or CYP inhibition in relapsed/refractory FLT3-mutant AML patients. DisclosuresAplan:NIH Office of Technology Transfer: Patents & Royalties.