PARP Inhibitor Veliparib and Busulfan in a Xenograft Model of Myeloproliferative Neoplasm

Despite new targeted therapies, patients with high risk myeloproliferative neoplasms (MPNs) have an increased chance of transforming into acute myeloid leukemia (AML) and can only be cured with allogeneic hematopoietic stem cell transplant (HSCT). JAK2^V617F mutation is common in MPNs and is associated with genomic instability, with baseline DNA double strand breaks and homologous recombination (HR) activity as a result of stalled replication forks. Poly(ADP-ribose) polymerase-1 (PARP-1) detects disrupted replication forks and recruits HR repair enzymes to restart DNA replication. Because busulfan, which is used in preparative regimens for HSCT, also leads to stalled replication forks through DNA strand crosslinking, we hypothesized that the PARP-1 inhibitor veliparib and busulfan may lead to synergistic cytotoxicity in MPN cells.

We first treated two JAK2^V617F positive MPN cell lines with increasing doses (from 0.1 to 100μM) of veliparib in liquid cultures and measured cell proliferation. SET2 and HEL cells were relatively sensitive to veliparib (IC50 of 11.3μM and 74.2μM respectively). We next treated the cell lines with increasing doses of busulfan in combination with a fixed subtherapeutic dose of veliparib (4μM) for 48 hours. With combination treatment, the busulfan IC50 decreased from 27μM to 4μM in SET2 cells and from 45.1μM to 28.1μM in HEL cells. The combination was synergistic with a mean combination index (CI) of 0.55 for SET2 and 0.40 for HEL cells. Combination treatment resulted in a higher fraction of cells in G2M arrest than with veliparib alone (p=0.03) or busulfan alone (p=0.04). Similarly, when the combination treatment was tested on CD34+ cells obtained from 5 patients with JAK2^V617F mutated and 2 CALR mutated PMF in a standard clonogenic (CFU) assay, it caused a greater inhibition of colony formation than busulfan or veliparib alone (p=0.001). Veliparib alone did not show any effect on colony formation from healthy control CD34+ cells.

Finally we utilized a xenograft model with NOD/SCID/IL-2Rγ^null (NSG) mice to test the in-vivo effect of combined low doses of busulfan and veliparib in JAK2^V617F MPNs. In order to establish disease, mice were injected with 5x10⁶ SET2 cells via tail vein 12 hours after sub-lethal irradiation. Then, 14 days after SET-2 injection 4 groups of mice (n=5 each) were treated for up to 3 weeks with intra-peritoneal vehicle (control); or a subtherapeutic dose of veliparib (3mg/kg daily) for 5 days a week; or busulfan (25mg/kg weekly); or a combination of both drugs. Death from leukemia was documented by marrow and spleen cell immunophenotype using non-cross reactive anti-human CD33 and CD34 antibodies. Veliparib alone did not improve survival vs control. On the contrary, survival was increased by the combination treatment vs busulfan alone (p=0.02).

Here we show that treatment with the PARP-1 inhibitor veliparib and busulfan elicits a synergistic cytotoxicity in MPN cells both in-vitro and in-vivo. Our data provide the rationale for testing novel pre-transplant conditioning regimens by combining veliparib with alkylating agents, such as busulfan or melphalan, in HSCT for high risk MPNs or MPN-AML.