Pharmacologic blockade of angiopoietin-2 is efficacious against model hemangiomas in mice.

Hemangioma of infancy is the most common neoplasm of childhood. While hemangiomas are classic examples of angiogenesis, the angiogenic factors responsible for hemangiomas are not fully understood. Previously, we demonstrated that malignant endothelial tumors arise in the setting of autocrine loops involving vascular endothelial growth factor (VEGF) and its major mitogenic receptor vascular endothelial growth factor receptor 2. Hemangiomas of infancy differ from malignant endothelial tumors in that they usually regress, or can be induced to regress by pharmacologic means, suggesting that angiogenesis in hemangiomas differs fundamentally from that of malignant endothelial tumors. Here, we demonstrate constitutive activation of the endothelial tie-2 receptor in human hemangioma of infancy and, using a murine model of hemangioma, bEnd.3 cells; we show that bEnd.3 hemangiomas produce both angiopoietin-2 (ang-2) and its receptor, tie-2, in vivo. We also demonstrate that inhibition of tie-2 signaling with a soluble tie-2 receptor decreases bEnd.3 hemangioma growth in vivo. The efficacy of tie-2 blockade suggests that either tie-2 activation or ang-2 may be required for in vivo growth. To address this issue, we used tie-2-deficient bEnd.3 hemangioma cells, which, surprisingly, were fully proficient in in vivo growth. Previous studies from our laboratory and others have implicated reactive oxygen-generating nox enzymes in the angiogenic switch, so we examined the effect of nox inhibitors on ang-2 production in vitro and on bEnd.3 tumor growth in vivo. We then inhibited ang-2 production pharmacologically using novel inhibitors of nox enzymes and found that this treatment nearly abolished bEnd.3 hemangioma growth in vivo. Signal-transduction blockade targeting ang-2 production may be useful in the treatment of human hemangiomas in vivo.