Search

Despite aggressive chemotherapy and early allogeneic transplantation, acute myeloid leukemia (AML) frequently relapses, so that over-all disease-free survival remains below 50%. Strategies to overcome the chemoresistance of relapse-initiating residual AML blasts are, therefore, warranted. Evidence has been provided that AML cells are sheltered from the insult of chemotherapeutic agents by interacting with bone marrow stroma. Integrin alpha4beta1 (VLA4) mediates adhesion of hematopoietic cells to bone marrow stroma cells and extracellular matrix and has been implicated in cell adhesion-mediated chemotherapy resistance. Based on the evidence thereof provided for ALL blasts, VLA4 is here proposed as a therapeutic target for refractory AML. For this purpose, VLA4 was functionally blocked in vitro and in vivo on patient-derived AML cells using an anti-functional humanized VLA4 antibody, Natalizumab (NZM). VLA4-positive (>90%) patient-derived (primary) AML cells were plated on immobilized human VCAM1 or human stromal cell line HS-5 and treated with control (IgG4) or Natalizumab (NZM) for 2 days. NZM de-adhered 94.0%±7.6 AML cells from its counter receptor VCAM-1, yet only 31.3%±13.8 from HS-5, indicating that stroma cells offer ligands for a wider panel of adhesion receptors besides VLA4. We tested also whether VLA4 blockade is beneficial against AML when combined with chemotherapy. For this purpose, primary AML cells were incubated with NZM and incubated on uncoated tissue culture plates or HS-5 stromal layers in the presence or absence of Ara-C (1µM) for two days. AML cells showed higher viability under Ara-C therapy when incubated with HS-5 cells compared to controls, indicating the chemoprotective effect of the stromal layer. The viability of the AML cells treated with combined Ara-C and NZM was similar to the controls, indicating that HS-5-mediated chemo-protection was completely abrogated by NZM. Significantly more AML cells treated with Ara-C+NZM stained AnnexinV+/7AAD- than after Ara-C+control Ig4 treatment (44.4%±5.6 vs. 29.8%±4.8, p=0.03) indicating increased apoptosis of AML cells. On its own, NZM did not induce apoptosis. Next, we tested NZM as a single agent in our NOD/SCIDIL2Rγ deficient (NSG) xenograft model of primary AML. Luciferase-labeled AML cells were intrafemorally injected into NSG mice (1x105 cells / mouse). NZM (5mg/kg) was given intraperitoneally once per week for 4 weeks. NZM-treated animals survived significantly longer than control Ig-treated animals (Median Survival Time, MST=107 days vs. MST=76 days; *p=0.008 by Log-rank Test.To determine effects of NZM on leukemia cell burden/distribution in different organs, primary AML cells were injected into NSG mice and allowed to engraft for 3 days, subsequently treated with a single dose of NZM or Ig control. 72 hours later, AML cell burden in femurs and spleens of NZM-treated animals was significantly decreased compared to control treated mice, however AML cells were not increased into the peripheral blood, so that whether leukemia cells were selectively killed ormobilized and then retained in non-hematopoietic organs remains to be determined. Further studies addressing molecular mechanisms of increased apoptosis after combined VLA4 blockade and chemotherapy are ongoing. Our data suggest that the paradigm of leukemia cell targeting by VLA4 blockade, previously demonstrated by us for ALL, can also be applied to AML. Disclosures: No relevant conflicts of interest to declare.

Abstract 2565 Chemotherapy drug resistance in acute lymphoblastic leukemia (ALL) remains a major problem, resulting in reduced treatment efficacy and relapse. The bone marrow environment (BME) has been shown to promote resistance of leukemia cells towards chemotherapy, which has been attributed to several proteins, including integrins. Our analysis of 207 children with high-risk (BCR/ABL1−) pre-B ALL revealed that high expression of the laminin-binding integrin VLA6 (alpha6beta1) portends poor clinical outcomes in patients with minimal residual disease (MRD+) on day 29 of induction. In addition, our comparative analysis of pre-B leukemia and normal B-cells revealed that VLA6 is preferentially upregulated on BCR/ABL1+ pre-B ALL blasts. Alterations in adhesion properties have been described for BCR/ABL1+ (p210) chronic myeloid leukemia. The role of integrins and integrin VLA6 in particular for cell adhesion-mediated drug resistance (CAM-DR) in BCR/ABL1+ (p210) ALL has not been addressed. With respect to its role for normal immature hematopoietic cells, contradictory observations have been reported. Therefore, we hypothesized that VLA6-mediated adhesion of ALL cells to the bone marrow stromal niche contributes to drug resistance. We evaluated the role of VLA6 in BCR-ABL1+ leukemia using two of our established models of leukemia, a conditional knockout model of VLA6 in murine BCR-ABL1+ leukemia and a xenograft model of human BCR-ABL1+ leukemia. VLA6fl/fl cells were oncogenically transformed using BCR-ABL1 (p210) and cultured under lymphoid-skewing conditions. Induction of pre- B (B220+ CD19+) ALL was confirmed by flow cytometry. Subsequent transduction with CreERT2 or EmptyERT2 generated leukemia cells in which VLA6 ablation could be induced (CreERT2) or not (EmptyERT2) by addition of Tamoxifen. Conditional ablation of VLA6 in vitro decreased adhesion significantly compared to undeleted controls (19.7%±8.1% vs. 87.7%±6.0%; p=0.00041) and increased apoptosis of murine BCR-ABL1+ leukemia cells as determined by analysis of Annexin V−/7-AAD− viable cells by flow cytometry (VLA6 deleted vs. undeleted: 35.3%±1.1% vs. 75.1%±1.2%; p=0.0001). Moreover, VLA6 deletion sensitized murine ALL to a tyrosine kinase inhibitor (TKI), Nilotinib (p=0.022, 45.6%±2.4% vs. 73.3%±13.0%). To test the effect of VLA6 deletion on leukemic progression in vivo, VLA6 BCR/ABL1+ pre-B (B220+ CD19+) CreERT2+ or control transduced ALL cells were transferred into NOD/SCID mice. 3 days thereafter, VLA6 deletion was induced by Tamoxifen administration to all animals in 2 cycles for 5 days. In vivo deletion of VLA6 in delayed leukemia progression compared to VLA6 competent controls from a median survival time (MST) of 30 days post-leukemia injection to a MST of 43 days post-leukemia injection (p=0.008 Log-rank test). In vivo deletion of VLA6 in combination with Nilotinib treatment delayed leukemia progression compared to VLA6 competent, as Nilotinib-treated control animals have uniformly died of leukemia with a MST of 39.5 days, however the Nilotinib treated VLA6 deleted group is completely alive and is still being monitored (p=0.0025). When VLA6 was ablated before transfer and recipients were observed for leukemia progression, the recipients of VLA6–deficient murine leukemia cells also showed attenuated leukemia progression compared to recipients of VLA6 competent cells. Moreover, we show that VLA6 blockade de-adheres primary ALL cells from their cognate counter receptor laminin in vitro, and sensitizes primary ALL cells to TKI Taken together, modulating the function of VLA6 in ALL offers a new approach to overcome drug resistance in ALL. Given that VLA6 is probably largely redundant for normal immature hematopoiesis, this approach may be preferable over targeting of other integrins in BCR/ABL1+ leukemias on which VLA6 is expressed. Disclosures: No relevant conflicts of interest to declare.