E-selectin, a cytoadhesive glycoprotein, is expressed on venular endothelial cells and mediates leukocyte localization to inflamed endothelium, the first step in inflammatory cell extravasation into tissue. Constitutive marrow endothelial E-selectin expression also supports bone marrow hematopoiesis via NF-kappa B-mediated signaling. Correspondingly, E-selectin interaction with E-selectin ligand (sialyl Lewisx) on acute myeloid leukemia (AML) cells leads to chemotherapy resistance in vivo. Uproleselan (GMI-1271) is a carbohydrate analog of sialyl Lewisx that blocks E-selectin binding. A Phase 2 trial of MEC chemotherapy combined with uproleselan for relapsed/refractory AML showed a median overall survival of 8.8 months and low (2%) rates of severe oral mucositis. Clinical trials seek to confirm activity in AML and mitigation of neutrophil-mediated adverse events (mucositis and diarrhea) after intensive chemotherapy. In this review we summarize E-selectin biology and the rationale for uproleselan in combination with other therapies for hematologic malignancies. We also describe uproleselan pharmacology and ongoing clinical trials.

The interaction of acute myeloid leukaemic (AML) blasts with the bone marrow (BM) microenvironment is a major determinant governing disease progression and resistance to treatment. The constitutive expression of E-selectin in the vascular compartment of BM, a key endothelial cell factor, directly mediates chemoresistance via E-selectin ligand/receptors. Despite the success of hypomethylating agent (HMA)-containing regimens to induce remissions in older AML patients, the development of primary or secondary resistance is common. We report that following treatment with 5-azacitidine, promoter regions regulating the biosynthesis of the E-selectin ligands, sialyl Lewis X, become further hypomethylated. The resultant upregulation of these gene products, in particular α(1,3)-fucosyltransferase VII (FUT7) and α(2,3)-sialyltransferase IV (ST3GAL4), likely causes functional E-selectin binding. When combined with the E-selectin antagonist uproleselan, the adhesion to E-selectin is reversed and the survival of mice transplanted with AML cells is prolonged. Finally, we present clinical evidence showing that BM myeloid cells from higher risk MDS and AML patients have the potential to bind E-selectin, and these cells are more abundant in 5-azacitidine-non-responsive patients. The collective data provide a strong rationale to evaluate 5-azacitidine in combination with the E-selectin antagonist, uproleselan, in this patient population.

Abstract CXC chemokine receptor 4 (CXCR4)/CXC motif ligand 12 (CXCL12) and E-(endothelial)-selectin/E-selectin ligands (E-selectin-L) axes play critical roles in leukemia cell homing to the bone marrow niche and are closely associated with resistance to FLT3-targeted therapy in FLT3-mutant acute myeloid leukemia (AML) patients. Hence, it is imperative to co-target CXCR4/E-selectin/FLT3 in FLT3 mutant AML. Herein, we determined whether FLT3 inhibition modulates CXCR4/E-selectin-L levels and whether co-targeting CXCR4/E-selectin enhances the anti-leukemia effects and reduces bone marrow niche-mediated resistance in FLT3-targeted therapy. Our results demonstrate that CXCR4/E-selectin-L are transcriptionally upregulated by FLT3 inhibition. Concomitant blockage of CXCR4/E-selectin with the dual inhibitor GMI-1359 disrupts leukemia cell homing and migration to bone marrow niches. Combination treatment with GMI-1359 and quizartinib significantly reduced leukemia cell burden and extended mouse survival in a patient derived xenograft AML mouse model. These findings provide pre-clinical rationale for combined CXCR4/E-selectin/FLT3 targeting in FLT3-mutant AML.

The efficacy and safety of rivipansel, a predominantly E-selectin antagonist, were studied in a phase 3, randomized, controlled trial for vaso-occlusive crisis (VOC) requiring hospitalization (RESET). A total of 345 subjects (204 adults and 141 children) were randomized and 320 were treated (162 with rivipansel, 158 with placebo) with an IV loading dose, followed by up to 14 additional 12-hourly maintenance doses of rivipansel or placebo, in addition to standard care. Rivipansel was similarly administered during subsequent VOCs in the Open-label Extension (OLE) study. In the full analysis population, the median time to readiness for discharge (TTRFD), the primary end point, was not different between rivi-pansel and placebo (-5.7 hours, P = .79; hazard ratio, 0.97), nor were differences seen in secondary end points of time to discharge (TTD), time to discontinuation of IV opioids (TTDIVO), and cumulative IV opioid use. Mean soluble E-selectin decreased 61% from baseline after the loading dose in the rivipansel group, while remaining unchanged in the placebo group. In a post hoc analysis, early rivipansel treatment within 26.4 hours of VOC pain onset (earliest quartile of time from VOC onset to treatment) reduced median TTRFD by 56.3 hours, reduced median TTD by 41.5 hours, and reduced median TTDIVO by 50.5 hours, compared with placebo (all P < .05). A similar subgroup analysis comparing OLE early-treatment with early-treatment RESET placebo showed a reduction in TTD of 23.1 hours (P = .062) and in TTDIVO of 30.1 hours (P = .087). Timing of rivipansel administration after pain onset may be critical to achieving accelerated resolution of acute VOC.

The bone marrow (BM) microenvironment has been established as a major factor limiting efficacy to leukemia therapy. Targeting the SDF-1a/CXCR4 axis, important for mesenchymal stem cell (MSC)/ acute myeloid leukemia (AML) interactions, has met with limited success. The adhesion molecule E-selectin is expressed on the surface of endothelial cells (ECs) in the perivascular niche where leukemia stem cells (LSCs) survive and develop therapy resistance. The binding of E-selectin and its ligands has been well investigated, but its role in AML therapy is not established. We aimed to better define the role of E-selectin in AML survival and drug resistance. In vivo studies with AML patient-derived xenograft (PDX) models demonstrated that pharmacological blockade of E-selectin sensitized therapy-resistant LSCs to Bcl-2–targeted therapy. Single-cell proteomics using CyTOF revealed that therapy resistant AML blasts expressing high levels of E-selectin ligand and the substantial proportion of them were eliminated by co-targeting E-selectin and Bcl-2. Concomitant inhibition of E-selectin and Bcl-2 markedly reduced LSCs in the BM and conveyed a significant survival. E-selectin inhibition was required to protect ECs in the BM from damage induced by venetoclax combined with a hypomethylating agent (Ven/HMA). Moreover, the dual inhibition of E-selectin/CXCR4 supported the proliferation of normal hematopoietic cells and progenitors. Taken together, results provide evidence that E-selectin inhibition or dual E-selectin/ CXCR4 inhibition profoundly reduces survival of AML cells in the BM and protects non-malignant HSCs and BM niche component cells from targeted therapy–induced apoptosis, thereby improving the therapeutic ratio of Ven/HMA therapy in AML.Funding Information: This work was supported by funding from National Cancer Institute (3UM1CA186688-05S2), a corporate research funding (GlycoMimetics, Inc.), and the Paul and Mary Haas Chair in Genetics, (all to M.A.). Declaration of Interests: M.A. received research funding from GlycoMimetics, Inc. W.E.F. and J.L.M. are employees of GlycoMimetics, Inc. M.A. is a consultant of GlycoMimetics, Inc. The remaining authors declare no competing financial interests.Ethics Approval Statement: All animal experiments were performed in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals under protocols approved by MD Anderson’s Institutional Animal Care and Use Committee.

The inability to generate robust anti-tumor responses with therapeutics affecting T cell function has been ascribed to multiple parameters such as the tumor fibrotic compartment and the degree/makeup of host infiltrative cells. These parameters are affected through pathways involving the interaction of galectin-3 and E-selectin with cognate ligands. In the current investigations we assessed the impact of a dual E-selectin and galectin-3 glycomimetic antagonist, GMI-1757, in combination with anti-PD-L1 on tumor progression using an orthotopic pancreatic carcinoma model. Eight days following the orthotopic injection of 5x105 luciferase-enabled Pan02 carcinoma cells into the pancreas of female C57BL/6 mice, mice began daily IP treatment for 11 days with control diluent (C) or 50 mg/kg GMI-1757. Beginning on study day 15 C- or GMI-1757-treated mice were given 10 mg/kg anti-PD-L1 or an isotype control antibody (LTF-2). Sixteen mice were allocated into each of 4 groups: group 1 = C + LTF-2; group 2 = GMI-1757 + LTF-2; group 3 = C + anti-PD-L1; and group 4 GMI-1757 + PD-L1. Bioluminescence imaging (BLI) intensities were determined over the course of the study (Day 0 to 26) to assess anti-tumor activity, and tissue was excised at study conclusion to assess fibrosis and immune infiltration. Tumor growth progressed in mice treated with LTF-2 in combination with C or GMI-1757 (final %T/C of 100 and 60.7, respectively), and appeared arrested in mice treated with C and anti-PD-L1 antibody (%T/C = 18.3). In contrast, treatment of tumor-bearing mice with GMI-1757 in combination with anti-PD-L1 antibody resulted in tumor regressions in 5/8 mice with a final %T/C = 0.60. Using a histopathologic scoring system, the extent of fibrosis within the primary pancreatic tumors was dependent on treatment with GMI-1757. Tumor fibrosis was moderate to marked in groups 1 and 3, and minimal to mild in groups 2 and 4. These results were confirmed by morphometric analysis of Masson9s Trichrome stained sections where the percent tumor fibrosis area was approximately 29 and 19% in groups 1 and 3, respectively, and 10 and 5% in groups 2 and 4, respectively. Notably, tumor infiltrating immune cells were enhanced with GMI-1757 treatment as compared to C (moderate to marked vs. minimal). The immune phenotypes comprising the increased mononuclear cell infiltration into orthotopic Pan02 tumors continues to be investigated. In summary, we report on the use of a dual E-selectin and galectin-3 antagonist to attenuate fibrosis and enhance mononuclear cell infiltration in an orthotopic model of pancreatic carcinoma. The shifts in fibrotic response and cellular infiltration obtained with GMI-1757 treatment in this model creates a more robust antitumor effect when combined with anti-PD-L1 treatment compared to anti-PD-L1 treatment alone. The impact of GMI-1757 to combine with immune modulators where fibrosis and restricted host cell infiltration negatively impact tumor response continues to be investigated. Citation Format: William E. Fogler, Theodore A. Smith, Ji-Won Lee, Silvia Locatelli-Hoops, David Stewart, John Peterson, John L. Magnani. A novel glycomimetic compound (GMI-1757) with dual functional antagonism to E-selectin and galectin-3 attenuates fibrosis, facilitates mononuclear cell infiltration and optimizes anti-PD-L1 therapeutic activity in a pancreatic adenocarcinoma model [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr LB164.

Uproleselan (GMI-1271) is a novel E-selectin antagonist that disrupts cell survival pathways, enhances chemotherapy response, improves survival in mouse xenograft and syngeneic models, and decreases chemotherapy toxicity in vivo. A phase 1/2 study evaluated the safety, tolerability, and antileukemic activity of uproleselan (5-20 mg/kg) with MEC (mitoxantrone, etoposide, and cytarabine) among patients with relapsed/refractory (R/R) acute myeloid leukemia (AML). Among the first 19 patients, no dose-limiting toxicities were observed. The recommended phase 2 dose (RP2D) was 10 mg/kg twice daily. An additional 47 patients with R/R AML were treated with uproleselan at the RP2D plus MEC. At the RP2D, the remission rate (complete response [CR]/CR with incomplete count recovery [CRi]) was 41% (CR, 35%), and the median overall survival (OS) was 8.8 months. In a separate cohort, 25 newly diagnosed patients age ≥60 years received uproleselan at the RP2D plus cytarabine and idarubicin (7 + 3). In these frontline patients, the CR/CRi rate was 72% (CR, 52%), and the median OS was 12.6 months. The addition of uproleselan was associated with low rates of oral mucositis. E-selectin ligand expression on leukemic blasts was higher in patients with relapsed vs primary refractory AML and in newly diagnosed older patients with high-risk cytogenetics and secondary AML. In the R/R cohort, E-selectin expression >10% was associated with a higher response rate and improved survival. The addition of uproleselan to chemotherapy was well tolerated, with high remission rates, low induction mortality, and low rates of mucositis, providing a strong rationale for phase 3 randomized confirmatory studies. This trial was registered at www.clinicaltrials.gov as #NCT02306291.

FLT3 inhibitors (FLT3i) have had transient success treating FLT3-mutant acute myeloid leukemia (AML) patients, especially those with FLT3 internal tandem duplication (ITD) mutations that account for one-third of adult AML cases (Daver et al., 2021). However, FLT3i are typically ineffective in eliminating leukemia stem cells in the protective bone marrow (BM) microenvironmental “niche” (Borthakur et al., 2011; Cortes et al., 2013; Zhang et al., 2008). Cytokines and chemokines such as CXCR4 and E-selectin ligands play a critical role in leukemia cell protection in the BM niche. Indeed, interactions of leukemic cells with their vicinal support cells, including mesenchymal stem cells (MSCs) and endothelial cells (ECs), in the BM niche is mediated mainly through the CXCR4/SDF-1 and E-selectin/HECA-452/CD44 axes (Erbani et al., 2020; Peled and Tavor, 2013). Therefore, evaluation of the effects of FLT3i on CXCR4 and E-selectin signaling in leukemia cells could enhance our understanding of AML FLT3i resistance mechanisms. To this end, we investigated the levels of CXCR4 and E-selectin ligands on FLT3-ITD-mutated AML cells in vitro and in vivo during FLT3i treatment (e.g., quizartinib or sorafenib), and evaluated the anti-leukemia effects of CXCR4/E-selectin blockade with the dual inhibitor GMI-1359. We first checked the effects of FLT3i on the levels of CXCR4 and E-selectin ligands, as well as CD44, in vitro in human MOLM14 AML cells, which harbor FLT3-ITD mutations. All of these were upregulated, as measured by flow cytometry, following exposure to quizartinib (p < 0.001) or sorafenib (p < 0.01) for 96 h. The mRNA levels were also increased roughly 2-fold, as measured by qPCR, suggesting transcriptional regulation was involved in the upregulation. Further, the upregulation of CXCR4 and E-selectin ligands and CD44 was time dependent (from 2 to 96 h). FLT3i profoundly suppresses activation of ERK, AKT, and Stat5 (Zhang et al., 2008). Therefore, we tested if the suppression of each signaling pathways individually could upregulate of CXCR4. Unexpectedly, 72-h suppression of MEK/ERK signaling with selumetinib or pimasertib also upregulated CXCR4 in MOLM14 cells. No effects in this regard were observed by suppressing AKT/mTOR or Stat5 with AZD8055 or STAT5-IN-1, respectively. Additionally, in Dox-inducible NRAS (G12D)-mutated MOLM13 AML cells which also harbor FLT3 ITD mutations, ERK activation by doxycycline downregulated CXCR4 levels implying the MEK/ERK signaling pathway was associated with the suppression of CXCR4. Furthermore, under BM microenvironment-mimicking, co-culture using human MSCs/ECs and MOLM14 cells, blockade of CXCR4 and/or E-selectin signaling using the CXCR4 antagonist plerixafor, the E-selectin antagonist GMI-1271, or the CXCR4 and E-selectin dual inhibitor GMI-1359 showed that GMI-1359 markedly abrogated BM protection and sensitized MOLM14 cells to quizartinib-induced apoptosis. We further validated the effect of GMI-1359 in a PDX model of AML which were from a patient who relapsed from sorafenib+E6201+DAC in clinic and showed resistant to quizartinib ex vivo. The combination of GMI-1359 with quizartinib profoundly reduced leukemia burden and extended survival of the PDX mice compared to the vehicle or the single-agent treatments (median survival was 158 days vs. 82.5, 79 and 128 days, respectively, in combination group vs. vehicle, quizartinib and GMI-1359; p < 0.0001) [Figure 1,2]. Our results suggest that FLT3i can upregulate CXCR4 and E-selectin ligands and CD44 in FLT3-ITD leukemia cells, which is mediated, at least in part, via suppression of MEK/ERK signaling. GMI-1359 sensitized AML cells to quizartinib-induced apoptosis in vitro and statistically significantly extended AML PDX mouse survival in vivo. These findings provide a pre-clinical rational for using GMI-1359 to prevent or overcome FLT3i resistance when treating FLT3-mutant AML patients.

Abstract Pre-clinical and clinical data suggest that the bone marrow (BM) environment provides breast cancer (BC) cells a protective haven against chemotherapeutic insult, endocrine therapies, and immune recognition. Novel agents that intercept cross-talk between BC cells and the host could therefore improve the depth of response to therapies and potentially increase overall survival in metastatic BC. We have previously shown that bone metastatic BC cells reside in perivascular niches expressing high levels of SDF-1α (CXCL12) and E-selectin, two molecules for which emerging data have demonstrated a significant role in metastatic BC progression. Our prior pre-clinical studies have shown that E-selectin inhibition specifically prevents circulating BC cells from homing to the BM, whereas CXCR4 blockade mobilizes micrometastases from the marrow into circulation, where they may be sensitized to chemotherapeutic cell kill. Given the normal role of these molecules in host immune responses, E-selectin/CXCR4 blockade has additional potential to impact the tumor immune microenvironment. GMI-1359 is a small molecule, glycomimetic compound with dual inhibitory activity against E-selectin and CXCR4. We hypothesize that GMI-1359 can block E-selectin and CXCR4 binding to E-selectin ligands and SDF-1, disrupting the protective effects of the bone microenvironment on tumor cells. A single-center, phase 1b, open-label, single and multiple ascending dose study (NCT04197999) is therefore evaluating GMI-1359 in patients with HR+ metastatic breast cancer with bony metastases who are stable or minimally progressive on endocrine therapy, with or without a CDK4/6 inhibitor. To date, 2 patients have been enrolled and received single monthly doses of 3.5, 5.0 and 7.0 mg/kg followed by 3 daily doses of 7 mg/kg. Both patients have completed treatment with no dose limiting toxicities observed. Coincident with GMI-1359 administration at all dose levels, we observed mobilization of CD34+ cells coupled with a reduction in elevated serum sE-selectin levels, demonstrating the dual functionality of the compound. Initial peripheral blood immunophenotype data on one patient revealed a redistribution of myeloid derived suppressor cells and a shift in macrophage polarization from M2 to M1. Notably, we observed similar immune alterations in ongoing preclinical studies in the E0771 syngeneic BC mouse model. We found that orthotopically-engrafted mice treated with single agent 1359 had a significant increase in the ratio of CD8/Tregulatory cells isolated from the primary tumor and BM. We also observed a decrease in primary tumor M-MDSCs and lung metastasis-associated macrophages. In summary, our preliminary clinical data suggest that GMI-1359 is well-tolerated and elicits on-target effects expected from disruption of the host microenvironment. Ongoing clinical and preclinical work will define the efficacy of this strategy to enhance responses to chemo and immune therapies. Citation Format: P. Kelly Marcom, Trevor T. Price, Andrew S. Murray, William E. Fogler, John L. Magnani, Helen Thackray, Eric Feldman, Dorothy Sipkins. Development of GMI-1359, a novel agent targeting tumor-microenvironment cross-talk in bone metastatic cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr CT144.

BACKGROUND:This study evaluated E-selectin inhibition with GMI-1271 (Uproleselan [GMI]) alone and in combination with the standard of care low-molecular-weight heparin (LMWH) to improve vein recanalization, decrease vein wall inflammation and protect against adverse bleeding in a primate model. We sought to examine this novel treatment of venous thrombosis. METHODS:Using a well-documented primate animal model, iliac vein thrombosis was induced by balloon occlusion of the iliac vein for 6 hours. Starting on day 2 after thrombosis, animals began treatment in two phases. In phase one, nontreated controls received no treatment (n = 5) vs animals treated with the E-selectin inhibitor GMI, 25 mg/kg, subcutaneous (SC), once daily (n = 4) for 21 days (previously published data). In phase two, animals were treated with GMI plus a combination of LMWH 1.5 mg/kg or 40 mg (GMI + LMWHc) SC once daily (n = 8) for 19 days; and animals treated with LMWH 1.5 mg/kg or 40 mg (LMWHc) SC once daily (n = 6) for 19 days. Animals were evaluated by magnetic resonance venography for vein recanalization and inflammation by gadolinium extravasation, duplex ultrasound, coagulation tests (thromboelastography, bleeding time, prothrombin time, activated partial thromboplastin time, fibrinogen) and complete blood count at baseline, days 2, 7, 14, and 21 at euthanasia. Statistical analysis included using unpaired t test with Welch's correction for direct comparisons and one-way analysis of variance for comparison between the groups. RESULTS:Percent vein recanalization by magnetic resonance venography was highest in the GMI alone group followed by GMI + LMWHc, both significantly different from control. On ultrasound examination, animals treated with GMI alone had no decrease in open vein lumen by day 21, whereas decreases were observed in groups GMI + LMWHc (-26%), LMWHc (-27%), and controls (-80%). Vein wall inflammation decreased significantly in all treated groups. Intimal fibrosis and intimal thickness was best preserved in the GMI alone group. An analysis of total vein wall collagen revealed a trend in all treatment groups of decreasing vein wall collagen. No clinically significant bleeding events were noted in any group. The LMWH groups trended to have prolonged coagulation test values, whereas E-selectin inhibition with GMI did not cause clinically significant changes in coagulation measures. CONCLUSIONS:Treatment with E-selectin inhibition results in improved vein recanalization, a decrease in vein wall inflammation and vein wall intimal thickness and fibrosis, with no changes in markers of coagulation. E-selectin inhibition with GMI alone is superior to E-selectin inhibition combined with LMWH, LMWH alone, and no treatment in this deep vein thrombosis model of iliac vein thrombosis.

Acute myeloid leukemia (AML) is characterized by the heterogeneous clonal expansion of undifferentiated myeloid cells in the bone marrow (BM). AML cells compete with normal hematopoietic cells and rewire the BM microenvironment into niches that selectively support leukemia stem cells (LSC). The leukemic niche produces soluble factors that facilitate the retention of LSC and provide protection from cytotoxic and targeted agents. The vascular adhesion molecule, E-selectin is expressed on endothelial cells (EC) in the perivascular niche where therapy-resistant AML cells have an increased affinity to E-selectin compared to normal hematopoietic stem cells (HSC) (Winkler et al., 2020). We previously demonstrated (Chang et al., ASH 2020) that E-selectin blockade by the pharmacological antagonist, GMI-1271 (uproleselan; GlycoMimetics, Inc) sensitized therapy-resistant LSC to Bcl-2 targeted therapy. Efficacious eradication of LSC in the BM however requires blocking multiple receptors and/or associated signaling pathways. A more optimal dislodgement of LSC from the BM could be attained by combining an E-selectin antagonism with blockade of the CXCR4/SDF-1α axis. The dual antagonist of E-selectin and CXCR4, GMI-1359 (GlycoMimetics, Inc.), has been tested in a phase 1 clinical trial (NCT02931214). Previously, we showed that GMI-1359 in combination with a FLT3-ITD inhibitor, improved survival in a xenograft model of FLT3-ITD + AML (Zhang et al., 2016). Hence, we hypothesized that co-targeting E-selectin/CXCR4 more efficiently mobilizes AML cells from BM niches and synergizes with the anti-leukemia activity of venetoclax/hypomethylating agent (Ven/HMA).

The endothelial cell adhesion molecule E-selectin is a key component of the bone marrow hematopoietic stem cell (HSC) vascular niche regulating balance between HSC self-renewal and commitment. We now report in contrast, E-selectin directly triggers signaling pathways that promote malignant cell survival and regeneration. Using acute myeloid leukemia (AML) mouse models, we show AML blasts release inflammatory mediators that upregulate endothelial niche E-selectin expression. Alterations in cell-surface glycosylation associated with oncogenesis enhances AML blast binding to E-selectin and enable promotion of pro-survival signaling through AKT/NF-κB pathways. In vivo AML blasts with highest E-selectin binding potential are 12-fold more likely to survive chemotherapy and main contributors to disease relapse. Absence (in Sele −/− hosts) or therapeutic blockade of E-selectin using small molecule mimetic GMI-1271/Uproleselan effectively inhibits this niche-mediated pro-survival signaling, dampens AML blast regeneration, and strongly synergizes with chemotherapy, doubling the duration of mouse survival over chemotherapy alone, whilst protecting endogenous HSC.

Acute myelogenous leukemia (AML) is characterized by an accumulation of abnormal white blood cells. Internal tandem duplications in the fms-like tyrosine kinase 3 (FLT3-ITD) account for 30% of adult AML cases and confer poor prognosis (Nakao et al., Leukemia 1996). FLT3 inhibitors like sorafenib efficiently eliminate circulating leukemia blasts, but frequently not in the bone marrow (BM), which suggests a protective effect of the BM niche for leukemic stem cell survival (Zhang et al., JNCI 2008). The homing of AML cells in BM is mediated chiefly by the adhesion to E-selectin on endothelial cells (ECs) and by CXCR4-directed cellular migration to stromal CXCL12 (SDF1) sources (Chien et al., Blood 2013; Peled and Tavor, Theranostics 2013). In many respects, BM homing signals are shared between leukemia and hematopoietic stem cells (HSCs). Our previous study demonstrated that targeting E-selectin/CXCR4 with the dual E-selectin/CXCR4 antagonist GMI-1359 markedly reduced leukemia cell adhesion to ECs and mesenchymal stem cells, reduced the BM-mediated protection of leukemic cells during FLT3-targeted therapy in vitro, and effectively reduced leukemia cellularity in the BM in vivo (Zhang et al., Can Res suppl 2016). Further, GMI-1359 combined with cytarabine/daunorubicin provided a profound survival benefit in mice with FLT3-mutated leukemia (Zhang et al., Blood suppl 2015). In the present study, we sought to evaluate dual E-selectin/CXCR4 blockage in the context of FLT3 inhibition by sorafenib in vivo, and to better understand the underlying mechanism. We compared expression levels of E-selectin ligands and CXCR4 in FLT3 inhibitor-sensitive Ba/F3-FLT3-ITD cells and their inhibitor-resistant counterparts Ba/F3-FLT3-ITD+D835Y and Ba/F3-FLT3-ITD+F691L. Resistant cells expressed 1.7 to ~5.6-fold higher levels of total E-selectin ligand detected by a soluble E-selectin reagent, and 10-fold higher levels of CXCR4. In addition, BM-mimetic hypoxia culture profoundly upregulated the cell surface expression of E-selectin ligands and CXCR4 on leukemia cells. We evaluated anti-leukemia effects of co-targeting E-selectin/CXCR4 and FLT3 with GMI-1359 and sorafenib in a patient-derived AML xenograft (PDX) model harboring FLT3-ITD and WT1 mutations. We observed that addition of GMI-1359 to sorafenib greatly reduced leukemia cellularity compared to sorafenib alone, and as much as by 92%, 82%, 69% and 45% in, respectively, liver, lung, spleen and BM (Fig. 1) as compared with vehicle-treated mice (p < 0.05). As expected, the number of circulating leukemia cells transiently increased. The GMI-1359/sorafenib combination improved mouse survival (median survival 138.5 versus 109, 87 and 126 days for the GMI-1359/sorafenib versus vehicle, GMI-1359 and sorafenib, respectively, p < 0.001). Using intravital 2-photon microscopy, we observed AML cell behavior in calvarial BM and their response to acute GMI-1359 bolus infusion. Remarkably, AML cell mobility began to increase in the BM microenvironment as soon as 20 min after treatment (Fig. 2), followed by intravasation and cellular outflow through the BM capillary vasculature over the next 2-4 hours. Moreover, although BM homing signals are thought to be shared between leukemia and HSCs, the combination therapy improved hematopoiesis parameters compared to sorafenib alone. In particular, this important effect was associated with increased numbers of megakaryocytes (2.1-fold), myelocytes (2.1-fold), and erythrocytes (7.1-fold) in BM (p < 0.01). The underlying mechanism(s) of hematopoiesis protection by GMI-1359 are under investigation. Conclusion: Co-inhibition of E-selectin/CXCR4 enhances the anti-leukemia efficacy of FLT3 inhibition and preserves hematopoiesis in the BM in a PDX model of AML. Disclosures Fogler: GlycoMimetics: Current Employment, Current equity holder in publicly-traded company, Patents & Royalties. Magnani:GlycoMimetics, Inc.: Current Employment, Current equity holder in publicly-traded company, Membership on an entity's Board of Directors or advisory committees, Patents & Royalties. Zal:Daiichi-Sankyo: Research Funding; Moleculin Biotech, Inc.: Research Funding. Andreeff:Daiichi-Sankyo; Breast Cancer Research Foundation; CPRIT; NIH/NCI; Amgen; AstraZeneca: Research Funding; Centre for Drug Research & Development; Cancer UK; NCI-CTEP; German Research Council; Leukemia Lymphoma Foundation (LLS); NCI-RDCRN (Rare Disease Clin Network); CLL Founcdation; BioLineRx; SentiBio; Aptose Biosciences, Inc: Membership on an entity's Board of Directors or advisory committees; Amgen: Research Funding; Daiichi-Sankyo; Jazz Pharmaceuticals; Celgene; Amgen; AstraZeneca; 6 Dimensions Capital: Consultancy.

Abstract The involvement of the adhesion molecule E-selectin and its interactions with E-selectin ligands as pertains to hematopoietic stem cells (HSC) and transplantation has been investigated (Winkler et al. 2012; Winkler et al. 2014). These preclinical studies focused on the role of E-selectin and the use of uproleselan, an E-selectin antagonist, during HSC mobilization in harvesting procedures of donors to accelerate recovery in transplant recipients. However, the impact of E-selectin and uproleselan administration on transplant recipients was less clear. In the current investigations we assessed the survival outcome of bone marrow depleted mice when reconstituted with HSC in combination with uproleselan. Lethally-irradiated, bone marrow depleted C57BL/6 mice were reconstituted with bone-marrow harvested from a congenic strain. Twenty-four hours post irradiation (6Gy x2), cohorts of mice (n=10/group) were injected i.v. with 1 × 106 cells (study day 0) from congenic donors using three i.p. dosing regimens with 40 mg/kg uproleselan. These regimens were: (a) bid on study days 0 and1; (b) bid on study days 1 and 2; and (c) bid on study day 1 only. Control groups in this study included irradiated mice alone (expected survival = 0%), non-irradiated mice alone (expected survival = 100%), and irradiated, reconstituted mice (no uproleselan). The survival of mice was determined over the course of the study (Day 0 to 30). Additional parameters of evaluation included sinusoidal obstructive syndrome such as hepatic veno-occlusive disease known to be E-selectin dependent and a complication of HSC transplantation. Treatment with uproleselan as part of the transplant regimen significantly increased the median survival time (MST) of mice compared with the control group – the MST of mice treated with uproleselan and HSC was >30 days with 80-90% of mice alive at study completion. In contrast, the MST of irradiated mice (no transplant) was 11.5 days with no survivors at study conclusion. The MST of mice irradiated and transplanted with congenic HSC was 9 days with 40% survival on day 30. The impact of uproleselan on survival represented a >233.3% increase in life span. Flow cytometric analysis in all surviving mice on day 30 showed that the mean percentage of CD45.1+ cells from donor congenic mice was approximately 90% (blood and bone marrow) indicating that all surviving mice were successfully reconstituted. In summary, we report on a novel therapeutic use of inhibitors of E-selectin, such as uproleselan, which results in the increased survival of mice when combined with HSC transplantation for reconstitution of depleted and compromised bone marrow. The impact on increased host survival could extend to the use of peripheral blood and stem cell transplantations as a therapeutic option in various malignancies where curative intent is intended. Citation Format: William E. Fogler, Dylan Daniel, Sheri Barnes, Alden Wong, David Draper, John L. Magnani. Enhanced survival of lethally-irradiated mice with HSC reconstitution in combination with the E-selectin antagonist, GMI-1271 (uproleselan) [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 486.

Abstract Acute myelogenous leukemia (AML) is characterized by an accumulation of abnormal white blood cells. Internal tandem duplications in the Fms-like tyrosine kinase 3 (FLT3-ITD) account for 30% of adult AML cases and confer poor prognosis (Kottaridis et al., 2003; Thiede et al., 2002). FLT3 inhibitors like sorafenib efficiently eliminate leukemia blast in the peripheral blood (PB), but frequently not in the bone marrow (BM) (Zhang et al., 2008). This suggests a protective effect of the BM on leukemic stem cells, which is mediated by E-selectin and CXCL12 expression in endothelial cells (ECs) and mesenchymal stem cells (MSCs) in the BM vascular and endosteal niches (Horacek et al., 2013; Peled and Tavor, 2013). Our previous study demonstrated that targeting E-selectin/CXCR4 with the dual E-selectin/CXCR4 antagonist GMI-1359 markedly reduced leukemia cell adhesion to ECs and MSCs and eliminated the BM-mediated protection of leukemic cells during FLT3-targeted therapy in vitro, and effectively reduced leukemia cell in the BM in vivo (Zhang et al., 2016). Further, GMI-1359 combined with cytarabine/daunorubicin provided a profound survival benefit in FLT3-mutated leukemia cell MV4-11-bearing mice (Zhang et al., 2015). In the preset study, we compared expression levels of E-selectin ligands and CXCR4 in FLT3 inhibitor-sensitive Ba/F3-FLT3-ITD cells and their inhibitor-resistant counterparts Ba/F3-FLT3-ITD+D835Y and Ba/F3-FLT3-ITD+F691L. The resistant cells expressed 1.7 to ~5.6-fold higher E-selectin ligand and 10-fold higher CXCR4 levels. In addition, BM-mimetic hypoxia culture profoundly upregulated P-selectin glycoprotein ligand-1 (PSGL-1) and CXCR4 levels on leukemia cells. PSGL-1 is also a high-efficiency ligand for E-selectin. We then evaluated anti-leukemia effects of co-targeting E-selectin/CXCR4 and FLT3 with GMI-1359 and sorafenib in a patient-derived AML xenograft model (harboring FLT3-ITD and WT1 mutations). Leukemia cell engraftment was evaluated using FACS and immunohistochemistry by hCD45 staining. We observed that GMI-1359 efficiently mobilized leukemic cells into PB and in combination with sorafenib enhanced the anti-leukemia activity and significantly reduced leukemia cell infiltration by 92%, 82%, 69% and 45% in liver, lung, spleen and BM, respectively, as compared with vehicle-treated mice (p < 0.05). The GMI-1359/sorafenib combination improved mouse survival (median survival were 109, 87, 126 and 138.5 days for the vehicle, GMI-1359, sorafenib and combination groups, respectively). Importantly, we observed that the combination protected normal hematopoiesis by increasing the number of BM megakaryocytes, myelocytes, and erythrocytes. The underlying mechanism(s) for this effect is under investigation. Our findings suggest that co-targeting E-selectin/CXCR4/FLT3 exerts remarkable protection of normal hematopoiesis in addition to more efficiently reducing leukemia cells. Citation Format: Weiguo Zhang, Kyung Hee Chang, Mahesh Basyal, Yannan Jia, Lauren Ostermann, William E. Fogler, John L. Magnani, Michael Andreeff. Combined targeting of E-selectin/CXCR4 and FLT3 by GMI-1359 and sorafenib effectively reduces leukemia cell burden and protects normal hematopoiesis in a patient-derived AML xenograft model [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 6038.