The oral HDAC inhibitor pracinostat (SB939) is efficacious and synergistic with the JAK2 inhibitor pacritinib (SB1518) in preclinical models of AML

Abstract

Acute myeloid leukemia (AML) is currently treated with aggressive chemotherapy that is not well tolerated in many elderly patients, hence the unmet medical need for effective therapies with less toxicity and better tolerability. Inhibitors of FMS-like tyrosine kinase 3 (FLT3), JAK2 and histone deacetylase inhibitors (HDACi) have been tested in clinical studies, but showed only moderate single-agent activity. High efficacy of the HDACi pracinostat treating AML and synergy with the JAK2/FLT3 inhibitor pacritinib is demonstrated. Both compounds inhibit JAK-signal transducer and activator of transcription (STAT) signaling in AML cells with JAK2(V617F) mutations, but also diminish FLT3 signaling, particularly in FLT3-ITD (internal tandem duplication) cell lines. In vitro, this combination led to decreased cell proliferation and increased apoptosis. The synergy translated in vivo in two different AML models, the SET-2 megakaryoblastic AML mouse model carrying a JAK2(V617F) mutation, and the MOLM-13 model of FLT3-ITD-driven AML. Pracinostat and pacritinib in combination showed synergy on tumor growth, reduction of metastases and synergistically decreased JAK2 or FLT signaling, depending on the cellular context. In addition, several plasma cytokines/growth factors/chemokines triggered by the tumor growth were normalized, providing a rationale for combination therapy with an HDACi and a JAK2/FLT3 inhibitor for the treatment of AML patients, particularly those with FLT3 or JAK2 mutations.

Keywords: AML; FLT3 inhibitor; HDAC inhibitor; JAK2 inhibitor; in vivo combination.

Figure 1

Pracinostat downregulates JAK and FLT3 signaling in JAK2^V617F and FLT-ITD cell lines, and shows synergy in combination with pacritinib. (a–d) Western blot analyses from 25 or 50 μg (pFLT3 and FLT3 blots only) of total cell lysate of the indicated cell lines with JAK2^V617F, JAK2, or FLT3-ITD or FLT3 wt are shown. (a, b) Cells were treated with the indicated concentrations of pracinostat (SB939) for 24 h. (c, d) Cells were treated for 24 h with pracinostat with pacritinib (SB1518) added in the indicated concentrations for the last 2 h of incubation before lysis.

Figure 2

Pracinostat potently inhibits cell proliferation in AML cell lines and primary AML cells. (a) IC₅₀ of cell lines tested in 48 h cell proliferation assays (CellTiterGlo); FAB: French–American–British classification of AML cells. Results show mean±s.d. from at least three rounds of experiments, each performed in triplicates. (b) Primary AML blasts from 16 patients were expanded and then treated on d12–d13 with dimethyl sulfoxide or pracinostat serially diluted in nine steps from 10 μℳ to 1.5 nℳ for 48 h. Results show means from two rounds of blast expansion/proliferation assays. Gray-shaded icons depict blast cells carrying the FLT3-ITD.

Figure 3

Pracinostat is efficacious in two s.c. and an orthotopic model of AML. (a) Female BALB/c nude mice (n=10 per group) were inoculated with 1 × 10⁷ MV4-11 cells s.c. into the right flank, and treatment as indicated was started on study d9. (b) Female BALB/c nude mice (n=5 per group) were inoculated s.c. with 5 × 10⁶ HEL92.1.7 cells, treatment was started on d18. (c) Female SCID mice (n=6 per group) were injected with 1 × 10⁷ HL-60 cells intravenously and scored daily for symptoms of paralysis. Treatment with pracinostat (SB939) 125 mg/kg three times per week was initiated on d30, before the first animal showed symptoms of paralysis. Daily disease scores are shown in (c); blood counts of naive mice, vehicle- or pracinostat-treated mice are shown in (d). ‘Naive mice' refers to age-matched SCID mice that were not injected with HL-60 cells. Statistical significance was determined using analysis of variance/Bonferroni, ***P<0.001 and **P<0.01 significance compared with vehicle-treated animals.

Figure 4

Pracinostat combined with pacritinib is efficacious and synergistic in vivo in two different models of human AML. Female SCID-beige mice (n=12 per group) were inoculated with 5 × 10⁶ SET-2 cells and treatment was started on d33 post-implantation (a, b). For the combination group, both pacritinib (SB1518) and pracinostat (SB939) were administered in half the volume (as a 20 mg/kg solution) every other day; when both compounds were given simultaneously, all other doses were given at 10 mg/kg, at least 8 h apart for 19 days (d0–d18). There was one non-treatment-related death in the vehicle group on study d11 (gavaging error). Female SCID mice (n=12 per group) were inoculated with 5 × 10⁶ Molm-13 cells (c, d). Treatment was started on d11, animals were dosed every other day for 8 days, in contrast to (a), animals were dosed with pacritinib only daily (in the evenings), pracinostat dosing remained unchanged, all dosing was done in a 10 mg/kg solution. On the last day, mice were killed and excised tumors were weighed. The doses showing significant TGI versus vehicle (on the last day of the study) using analysis of variance/Dunnet's post test, are indicated with **P<0.01 or ***P<0.001. (e) Tumors from the MOLM-13 efficacy study (c, d) were harvested on the last day 3 h post pacritinib dosing and immediately lysed. Western blot analyses for pFLT3, FLT3, pSTAT5 and β-actin are shown from the lysates of three randomly selected tumors from each treatment group of (c).

Figure 5

Pharmacokinetic analysis of pracinostat and pacritinib administered in combination. Female BALB/c nude mice were dosed with 150 mg/kg pacritinib (SB1518) plus 75 mg/kg pracinostat (SB939) simultaneously either once (‘day 1') or for a total of 14 days (i.e., 7 doses of SB939, administered every other day in the morning, plus SB1518 for 14 times, every day morning and evening, ‘day 14'). Plasma was collected at 10, 30 and 60 min post-dose on d1/d14 (n=5) as well as 4, 8 and 24 h post-dose (n=3). Pharmacokinetic parameters were calculated by a non-compartmental method, using WinNonlin software, and compared with values obtained in previous experiments in the same mouse strain. * Indicates that values were extrapolated from BALB/c nude mice dosed with 50 mg/kg pacritinib or pracinostat. Results for pracinostat are shown in (a), and for pacritinib in (b).

Figure 6

Pracinostat and pacritinib have synergistic effects on AML-induced plasma cytokines/growth factors/chemokines. Plasma levels of a panel of 32 cytokines/growth factors/chemokines were analyzed using the MAP Mouse Cytokine/Chemokine panel from Millipore. Plasma was collected from SCID-beige mice bearing s.c. SET-2 tumors on d18, 3 h post-dose, after chronic dosing with pacritinib and pracinostat from the mice used in the efficacy study described in Figure 4.