Neovascularization is an intricate process in which abnormal blood vessels grow as a response to vascular endothelial growth factor (VEGF) or downregulation of platelet-derived growth factor (PDGF). An analogy can be made from how neovascularization is powered to the power source of a car engine. There are promoters (gas pedal) and inhibitors (brake pedal). In a car, the gas pedal initiates the fuel injector, which operates on downstream processing, turning on the engine. Currently the only tools that we have to turn off the “gas” in the neovascularization process are anti-VEGF agents such as ranibizumab (Lucentis, Genentech) and bevacizumab (Avastin, Genentech). These agents act in a similar manner to a clogged injector by preventing fuel from reaching the injector; however, there are other “injectors” that are permitted to fire, allowing the signal cascade to continue, albeit in a hindered manner.
KINASE INHIBITORS
The power supply for VEGF is adenosine-tri-phosphate
(ATP). If ATP could be blocked, would the entire neovascularization
process shut off? In an effort to answer this question,
there has been a significant amount of pharmaceutical
development looking at tyrosine kinase inhibitors, which are
theorized to be capable of blocking ATP, inhibiting the pathway
completely, and shutting down this process.
Neovascular AMD is characterized by choroidal neovascularization (CNV) that enters the subretinal space. Aberrant activation of growth factor receptors is a driving force in the growth of many of these blood vessels. As the specific ligand (VEGF, PDGF) binds to its receptor through ATP phosphorylation, a signaling cascade is initiated leading to the pathogenesis of neovascular AMD (Figure 1). Tyrosine kinase inhibitors have the ability to hinder multiple pathways including inflammation and angiogenesis.
PKC412 (Novartis, Inc.) is a kinase inhibitor for diabetic macular edema. In early clinical studies by Campochiaro et al, oral PKC412 proved to be anatomically effective, but caused hepatotoxicity, suggesting that local delivery would be more advantageous.1
My colleagues and I have evaluated another experimental eye drop that is a pro-drug and kinase inhibitor. TG100801 (previously owned by Targegen, which has been acquired by Sanofi-Aventis) is a highly selective kinase inhibitor, which is metabolized by esterases in the eye and converted to TG100572. TG100572 has shown to be a potent inhibitor of kinases that regulate inflammation, angiogenesis, and vascular leak, such as VEGF receptors (VEGFR), PDGF receptors (PDGFR), fibroblast growth receptors, Ephrin receptors, and Src family kinases. In animal models, this drug appears to be better than bevacizumab in terms of preventing leakage. In a laser-induced CNV model, the inhibition appears to be dose dependent.
What happens when TG100801 is applied to humans? We had a patient with subretinal fluid and a pigment epithelial detachment (PED; Figure 2A). After one drop TG100801, the fluid had regressed (Figure 2B) but the PED remained (Figure 2C). We observed on the slit lamp that red, electrostatically charged particles had formed on the cornea upon instillation, which had no effect on visual acuity.
Another kinase inhibitor that is currently under evaluation for retinal disease is pazopanib (GlaxoSmithKline). Pazopanib is a multitargeted tyrosine kinase inhibitor that works via activation of VEGFR, PDGFR, and c-kit receptors to trigger the tyrosine kinase cascade, leading to changes in cellular behavior to prevent VEGF-induced permeability and angiogenesis.
Further research is required for all of these aforementioned tyrosine kinase inhibitors, and one or more may prove promising. Currently, however, there is no one agent that has “knocked our socks off.”
SUMMARY
Targeting the “fuel injector” of neovascularization (tyrosine
kinase) appears to make sense. There is some initial
anatomic validation to tyrosine kinase inhibition, but the
devil lies in the details. Whether this route is the answer that
will help change the way we treat retinal disease in the
future is yet to be determined.
_1773249222.png?auto=compress,format&w=75)
