Patients with T315I positive CML are resistant to most tyrosine kinase inhibitors (TKIs). Ponatinib (Iclusig) is approved for CML patients with the T315I ABL kinase polymorphism. However, ponatinib treatment is associated with vascular events (myocardial infarction, stroke, coronary artery stenosis, limb ischemia and occlusion, and venous thrombosis) in~29% of patients. The mechanism(s) for these events has not been characterized. We developed a murine model to examine TKIs influence on arterial thrombosis risk. C57BL/6 mice, 18-22 weeks of age and treated with ponatinib by gavage for 14 days at 15 mg/kg PO BID, had significantly shorter carotid artery occlusion times induced by photochemical activation of Rose-Bengal compared to vehicle-treated mice (10.4 ± 2.9 min versus 32.3 ± 4.8 min, p < 0.0001). Mice were treated with ponatinib for 14 days at the 3 mg/kg PO BID, a dose that yields plasma concentrations similar to patients at 45 mg po daily, also had significantly shorter vessel occlusion times compared to control (18.7 ± 3.7 min versus 32.3 ± 4.8 min, p<0.0001). No difference in time to carotid artery occlusion was observed between imatinib at 180 mg/kg PO BID treatment compared to control (32.7 ± 5.6 min versus 32.3 ± 4.8 min, p = 0.85) or nilotinib at 29 mg/kg PO BID treatment compared to vehicle-treated mice (32.8 ± 5.5 min versus 33.8 ± 5.1 min, p = 0.71). These studies show that uniquely ponatinib treatment is prothrombotic. Plasma of ponatinib-treated animals had normal PT, aPTT, and complete blood counts (WBC, RBC, Hgb, Hct, MCV, MCH, MCHC and platelet counts) assays. Also contact activation- and tissue factor-initiated thrombin generation times (TGT) showed no difference between treated and untreated mouse plasma.
We next examined the mechanism(s) of ponatinib-induced thrombosis. Ponatinib at 3 mg/kg PO BID daily inhibited p-LynY396 in murine platelets. Lyn is a negative regulator of platelet GPVI. Collagen-related peptide (CRP)-induced expression on murine platelets of the activated heterodimeric complex of α2bβ3 integrins (the JON/A epitope) and the alpha granule constituent P-selectin (CD62) when examined by flow cytometry ex vivo were significantly higher at 3 μg/ml in ponatinib-treated versus untreated mice (p< 0.03). The CRP concentration needed to induce platelet activation in ponatinib-treated mice was significantly lower than untreated mice (p<0.0001, 2-way ANOVA). These data suggested that platelets from ponatinib-treated mice are more GPVI actable. Additional studies with α-thrombin also show ponatinib-treatment makes more active platelets. The threshold for α-thrombin-induced expression of the JON/A epitope also was significantly lower (p<0.0125) at 0.075 and 0.1 nM in ponatinib-treated platelets versus untreated platelets. Likewise, α-thrombin-induced platelet membrane expression of P-selectin also was significantly lower (p<0.025) at 0.075 and 0.1 nM in ex vivo studies of ponatinib-treated platelets. Next, we examined vessel wall for changes in ponatinib-treated mice. Aortic sections showed increased caspase 3 staining in vessel adventitia and surrounding adipose tissue in treated mice, a sign of apoptosis. Also genes involved in vessel anti-thrombosis were altered in 3 mg/kg PO BID ponatinib-treated mice. Expression of mRNA of both COX2 and eNOS and their vasculo-protective transcriptional regulators, Sirt1 and KLF4, respectively, were significantly decreased (p<0.05) in the vessel wall of ponatinib-treated mice.
We then sought agents to blunt the prothrombotic changes with ponatinib treatment. Since PPAR-γ agonism elevates Sirt1, and vessel wall Sirt1 is reduced in treated mice, we determined if pioglitazone treatment, a PPAR-γ agonist thiazolidinedione, corrects thrombosis risk after ponatinib in vivo. When ponatinib-treated mice were given oral pioglitazone (10 mg/kg/day po), their short times to thrombosis significantly lengthened (49±6.9 min, p<0.0251) to values like untreated mice. Additionally, neither lisinopril nor atorvastatin ameliorated the ponatinib's prothrombotic effect in vivo. In sum, ponatinib uniquely induces a prothrombotic state due increased platelet activation and reduced vessel wall anti-thrombosis. The effect of ponatinib on platelets may arise in part from its inhibition of p-Lyn. In a murine model of arterial thrombosis, ponatinib's prothrombotic effect is ameliorated by PPAR-γ agonism with pioglitazone.
No relevant conflicts of interest to declare.
Asterisk with author names denotes non-ASH members.