The Specific Aurora Kinase Inhibitor AZD1152 Significantly Affects the Growth of Human Leukaemic Cells in an In Vivo AML Model.

AZD1152 is a specific Aurora kinase inhibitor with selectivity for Aurora B kinase, designed to target cell division in proliferating tumour cells. It has been shown in model systems that inhibition of Aurora B kinase with AZD1152 reduces histone H3 phosphorylation (phH3) and inhibits cytokinesis, thus inducing cellular multi-nucleation and polyploidy, leading to apoptosis and cell death.

Here, we present results of preliminary studies on the in vivo action of AZD1152 on the growth and development of both a human acute myeloid leukaemia (AML) cell line (HL-60) and primary AML samples in the NOD/SCID xeno-transplantation model. Two weeks of AZD1152 at 25 mg/kg/day is well tolerated in NOD/SCID mice. HL-60 cells were injected iv into sub-lethally irradiated NOD/SCID mice (n=24) and allowed to engraft in the bone marrow. Three weeks later, when malignant cell growth was well established in the bone marrow, mice were split into treatment groups. Either vehicle (n=12) or AZD1152 (25 mg/kg/day via minipump, n=12) was administered for 7 days. One AZD1152-treated group was sacrificed immediately post-treatment (n=9) and another was left for a further 2 weeks before bone marrow analysis for HL-60 cell content (defined by human CD45+/CD33+/CD19− expression). The mean percentage HL-60 cell content in the marrow of untreated mice was 64.5 ± 19.0 (range 14.0–86.6, n=11), whereas in mice that had received AZD1152 and were sacrificed at the end of dosing period, the mean percentage was 0.29 ± 0.74 (range 0–2.26, n=9; p<0.001). In AZD1152-treated mice that were left for 2 weeks post-treatment, the mean percentage of HL-60 cell content in the bone marrow was 0.008 ± 0.74 (n=3).

Similar experiments were performed with two primary AML samples. AML-1 is from a patient with French-American-British (FAB)-type M1 AML and AML-2 is from a FAB-type M2 AML patient. Ten million cells from both AML samples were injected into NOD/SCID mice (n=12 for AML-1; n=6 for AML-2). Ten weeks later, mice were split into treatment groups and dosed with either control vehicle or 25 mg/kg/day of AZD1152 via minipumps for 7 days. All mice were sacrificed after 1 week of treatment in these initial experiments. The mean percentage engraftment in untreated mice was 59.7 ± 36.0 (range 6.2–84.0, n=4) for AML-1 and 21.8 ± 22.5 (range 5.9–37.7, n=2) for AML-2. Whereas, in mice that had received 25 mg/kg/day of AZD1152 for 7 days, the percentage of engraftment was 1.8 ± 1.7 (range 0.38–4.53, n=5; p<0.01) and 0.07 ± 0.07 (range 0.12–0.014, n=2) for AML-1 and AML-2, respectively.

These data indicate that AZD1152 has a profound effect on the engraftment of human AML cell lines and primary AML samples in an in vivo model. Furthermore, initial results may indicate a long-term effect of AZD1152 treatment in vivo. AZD1152 is currently undergoing Phase I clinical trials in AML patients.