The BCR/ABL1 inhibitor nilotinib is increasingly used to treat patients with chronic myeloid leukemia (CML). However, nilotinib apparently induces metabolic changes, including an increase in the fasting glucose level. In addition, vascular adverse events, including progressive atherosclerosis with peripheral arterial occlusive disease (AOD) have been reported in nilotinib-treated CML patients. We reviewed and updated AOD events in our CML patients receiving nilotinib (n=34) and initiated preclinical in vitro and in vivo studies in order to dissect potential targets and mechanisms. After a median observation time (MOT) of 24 months, the frequency of AOD (26.5%) and severe AOD requiring surgical intervention and/or prolonged hospitalization (17.6%) was higher in nilotinib-treated patients compared to risk factor-, observation time-, and age-matched controls (34 imatinib-treated patients with CML, 34 with myelodysplastic syndromes, 34 with JAK2-mutated MPN and 34 with lymphoid neoplasms; <5% AOD, p<0.05). After a MOT of 36 months, the frequency of AOD amounted to 36.1% and the frequency of severe AOD was 19.4%. We next examined the in vitro effects of nilotinib on cultured human umbilical vein endothelial cells (HUVEC), human coronary artery-derived endothelial cells (HCAEC), and the human microvascular endothelial cell line HMEC-1. As determined by 3H-thymidine incorporation, nilotinib was found to inhibit the proliferation of endothelial cells in a dose-dependent manner, with pharmacologically relevant IC50 values obtained in HUVEC (1.0 µM), HCAEC (100 nM), and HMEC-1 (1.0 µM), whereas imatinib showed little effect up to 5 µM. Moreover, nilotinib was found to inhibit the migration of HUVEC in a wound-scratch assay as well as angiogenesis in a tube-formation assay (relative capillary tubes: VEGF+control: 1.8±0.1, VEGF+nilotinib (100 nM): 1.3±0.1, VEGF+imatinib (100 nM): 1.7±0.05; n=3, p<0.01 for VEGF alone vs VEGF+nilotinib). In a mouse model of hindlimb ischemia, nilotinib (75 mg/kg/day p.o. for 28 days) was found to slow blood flow-recovery after induction of ischemia whereas imatinib (100 mg/kg/day p.o. for 28 days) showed no comparable effect (laser Doppler perfusion imaging ratio ischemic/control leg: control mice: 0.81±0.03, imatinib-treated mice: 0.79±0.04, nilotinib-treated mice: 0.68±0.04; n=13/group; p<0.05 for nilotinib vs control and for nilotinib vs imatinib). The decreased blood perfusion was accompanied by an increased rate of limb necrosis (necrosis score: control: 1.15±0.08, imatinib: 1.17±0.05, nilotinib: 1.54±0.18; p<0.05 for nilotinib vs control and nilotinib vs imatinib). Moreover, microvessel density was significantly lower in the affected hind limb in nilotinib-treated mice compared to imatinib-treated mice or control-mice (p<0.05). In addition, we found that nilotinib (between 1-10 µM), but not imatinib (1-10 µM) promotes the expression of pro-atherogenic cytoadhesion molecules (CAM) on HUVEC, including ICAM-1 (CD54), VCAM-1 (CD106) and E-Selectin (CD62E). By contrast, nilotinib (up to 10 µM) showed no effects on expression of plasminogen activators or uPA receptor (CD87) in cultured endothelial cells. As assessed by chemical proteomics profiling and phospho-array analysis, several angiogenesis-related and other endothelial antigens, including Tie-2/TEK, JAK1, BRAF and EPHB2 were identified as molecular targets of nilotinib, whereas imatinib did not bind to these vascular targets in endothelial cells. As assessed by immunohistochemistry using antibodies against KIT and mast cell tryptase, we also found that in our CML patients, nilotinib induces an almost complete depletion of KIT+ mast cells, a cell type that serves as unique source of heparin and uncomplexed tPA and has been implicated as a major repair cell in vascular disorders. However, imatinib was also found to induce mast cell depletion in our patients with CML. Neither nilotinib nor imatinib showed in vitro or in vivo effects on platelet adhesion or platelet aggregation. In conclusion, nilotinib exerts multiple effects on vascular endothelial cells and other perivascular cells such as mast cells, presumably through multiple mechanisms and targets. We hypothesize that these effects may contribute to nilotinib-induced vasculopathy in CML.


Wolf:Bristol-Meyers Squibb: Honoraria; Pfizer: Honoraria; Novartis: Honoraria, Research Funding. Valent:Novartis: Consultancy, Honoraria, Research Funding.

Author notes


Asterisk with author names denotes non-ASH members.

Sign in via your Institution