Résumé
Inhibitors of VEGF signaling target tumor angiogenesis and inhibit tumor growth; however, the effects of antiangiogenic therapy are often transient. Cessation of treatment or dose interruption can induce rapid regrowth of tumor vessels and tumor relapse. Preclinical studies have demonstrated that upon withdrawal of VEGFR inhibitors, tumor vessels can rapidly repopulate the tumor on the tracks of the empty basement membrane sleeves (1,2). This rapid vessel regrowth is mediated, at least in part, by the interaction of VEGF produced by the tumor and VEGF receptors expressed on endothelial cells. Neutralization of the VEGF ligand combined with VEGFR inhibition may provide more complete VEGF signaling blockade, however the combination of bevacizumb, a monoclonal antibody to VEGF ligand, and sunitinib, a mulitkinase/pan-VEGFR inhibitor, has resulted in significant clinical toxicity (3). As an alternative approach, we reasoned that inhibition of VEGF production would decelerate the regrowth of tumor vessels following VEGFR inhibition. We sought to reduce VEGF production in the tumor by targeting the mechanism for VEGF translation using a selective inhibitor of mTOR. The mTOR kinase is a critical signaling hub which regulates multiple cellular networks including cap-dependent translation. The translation of many cellular growth factors, including VEGF isoforms, is regulated by the 4E-BP1 complex, which is in turn regulated by mTOR. ATP-competitive, selective inhibitors of mTORC1/mTORC2, OXA-01 and OSI-027 (ASP4786) can attenuate 4E-BP1 phosphorylation more completely than rapamycin, an allosteric inhibitor of mTORC1. In RIP-TAg mouse pancreatic tumors, OXA-01 dramatically reduced VEGF production whereas rapamycin, was less effective. Interestingly, OXA-01-mediated reduction in VEGF was associated with decreased vessel growth and normalization of the vascular architecture, but not vascular regression. In contrast, treatment with a selective inhibitor of VEGF receptors, OSI-930, resulted in substantial vascular regression but no decrease in tumor VEGF levels. Upon discontinuation of OSI-930 tumor vessels rapidly regrew, and this regrowth was diminished by OXA-01 treatment but not with rapamycin. Another key mTORC2 activity is regulation of the Akt signaling cascade, a central mediator of cellular survival. Inhibition of mTORC1/mTORC2 with OXA-01 induced tumor cell apoptosis in vivo, and this was enhanced when combined with the VEGFR inhibitor, OSI-930. In order to explore the clinical feasibility of this total VEGF blockade approach, we tested the combination of OSI-027, an mTORC1/mTORC2 inhibitor in clinical trials, and sunitinib, an approved multikinase/VEGFR2 inhibitor, in human tumor xenografts. The combination decreased the growth of human tumor xenografts to a greater degree than either single agent. Together these data demonstrate one mechanism for co-targeting angiogenic ligand production as well as inhibition of VEGFR signaling to provide maximal blockade of VEGF signaling and to limit vessel regrowth after cessation of the VEGFR inhibitor.