PDGF: Ophthalmology's Next Great Target
PDGF: Ophthalmology's Next Great Target
The intraocular concentrations of several growth factors and chemokines including PDGF have been measured in eyes with chorioretinal vascular disorders (AMD, DR, RVOs). Whereas the intraocular concentrations of VEGF are elevated in all of these conditions, the concentrations of PDGF are variable. In eyes with non-proliferative DR, vitreous concentrations of VEGF, PDGF-AA and PDGF-AB were elevated compared with controls, whereas levels of PDGF-BB were decreased. The absolute levels of PDGF-AA and VEGF correlated with the severity of the retinopathy. Aqueous concentrations of PDGF-AA were elevated in patients with DME but were normal in those with BRVO.
In patients that underwent cataract surgery, baseline aqueous levels of IL-6, IP-10, VEGF and macrophage chemotactic protein (MCP)-1 were higher in patients with DME compared with controls. Each patient received an intravitreal injection of 4 mg triamcinolone in 1 eye and 1.25 bevacizumab in the other. Four weeks later, levels of IL-6, IP-10, MCP-1, VEGF and PDGF-AA were decreased in the eyes that received triamcinolone, whereas only VEGF levels were lower in eyes that received bevacizumab. The foveal edema had improved more in the triamcinolone group.
Studies consistently show elevated levels of VEGF and all PDGF dimers in patients with PDR. Vitreous concentrations of PDGF-BB are elevated in patients with PDR but not CRVO. Among factors correlated with VEGF in eyes with PDR, IL-10 and IL-13 were positively correlated and PDGF was negatively correlated (but still elevated). These data suggest that the inflammatory reaction may be more active in eyes with CRVO than in those with PDR.
Increased aqueous concentrations of VEGF and decreased concentrations of PDGF-AA have been measured in patients with AMD, and after ranibizumab injections only the VEGF concentrations decreased. In another study, vitreous PDGF levels were higher in the control group than in the CNVM/AMD group (16.4 + 7.1 vs 2.8 + 1.3 ng/μl) but there was no significant difference in VEGF concentrations (27.4 ± 41.0 vs 39.3 ± 28.9 pg/ml). An insignificant difference in PDGF concentrations between the CNVM/AMD and control groups (1219 vs 1469 ng/ml) was found, but VEGF concentrations were undetectable in both groups. Elevated aqueous VEGF concentrations were measured in patients with intermediate and advanced (but not early) AMD. PDGF levels were depressed more in advanced AMD patients, which suggested an involvement of PDGF in AMD that warrants further investigation.
Several potent non-selective tyrosine kinase inhibitors that interrupt both VEGF and PDGF signaling have been approved for cancer therapy: imatinib, sorafenib, pazopanib, sunitinib, nilotinib and dasatinib. Human RPE cells become less viable and express VEGF and PDGF when exposed to light; sorafenib and pazopanib decrease both VEGF and PDGF-BB while ranibizumab decreases only VEGF. Sorafenib and pazopanib decrease expression of VEGFR1, VEGFR2 and neuropilin 1.
A small number of patients with exudative AMD have been treated with oral sorafenib, a non-selective receptor tyrosine kinase inhibitor that inhibits the activation of VEGFR2, PDGF, FLT3 Ret and c-Kit. A patient with renal cell carcinoma and co-incident, occult CNVM experienced improved macular edema when treated with sorafenib. Three patients with exudative AMD that responded poorly to ranibizumab repeatedly resolved their edema when treated with sorafenib. One explanation for this finding is that non-VEGF pathways, such as PDGF, may be important in some patients with treatment resistant AMD.
The side effects of orally administered receptor tyrosine kinase inhibitors with anti-PDGF effects include rash, fatigue, nausea, pruritus and headache. These may be endured by patients with cancer but those with non-malignant conditions such as AMD are unlikely to tolerate such side effects for years. Therefore, oral RTK inhibitors are not likely to be used for the long-term treatment of exudative AMD. Monoclonal antibodies generally produce fewer side effects than RTK inhibitors, but because of their large molecular weight and poor gastrointestinal absorption, they cannot be taken orally. However, experience with antibodies and fragments shows that a regimen consisting of intravitreal injections may be tolerated for years. Furthermore, intraocular injections are quite safe and result in fewer systemic adverse events than are seen with RTK inhibitors.
Pazopanib, a small RTK inhibitor that blocks several receptors, was evaluated as topical therapy for exudative AMD. In a Phase IIa trial, patients gained an average of +4.3 letters at day 28, though macular thickness did not significantly improve. However, in a 19 patient, 12-week Phase IIb trial, pazopanib failed to improve visual acuity and macular thickness. Because of these disappointing results, the manufacturer (GlaxoSmithKline) decided to halt further AMD testing.
In a Phase I tolerability study, single intravenous doses of an anti-PDGF-D antibody were administered to 40 healthy volunteers. This high affinity drug (KD = 0.18 nM) demonstrated linear pharmacokinetics with a systemic half-life of 15.5–48.1 days.
Clinical testing of specific PDGF inhibitors for ocular disease began with the use of E10030 (fovista, Ophthotec, Princeton, NJ, USA), a 50 kDa aptamer that binds to PDGF-B. A Phase I escalating dose trial enrolled 57 patients with advanced exudative AMD. Patients were given a single dose of 0.03 mg fovista with 3 monthly injections of ranibizumab, or 3 doses of fovista (0.03, 0.3, 1 or 3 mg) together with three doses of ranibizumab. For patients completing the 4-, 8- and 12-week visits, the proportions of patients gaining at least 15 ETDRS letters were 32, 45 and 60%, the mean improvements in vision were +11.1, +13.4 and +15.7 letters, and the mean center point thicknesses were 251, 231 and 229 μm (compared with 395 μm at baseline). Fluorescein angiographic regression of the neovascular lesions occurred in 85% of eyes. None of the patients developed drug-related adverse events.
In a Phase IIb, prospective, superiority trial, 449 patients were randomized to receive monthly ranibizumab monotherapy, or combination therapy consisting of ranibizumab with fovista (0.3 or 1.5 mg) for 24 weeks. The combination group receiving 1.5 mg fovista met the primary superiority end point, with more improvement in vision compared with ranibizumab monotherapy (+10.6 vs +6.5 letters; p = 0.019). Compared with patients receiving monotherapy, those receiving fovista combination therapy enjoyed greater gains at every monthly visit and the differences increased throughout the trial. Compared with ranibizumab monotherapy, the improvements in vision and macular thickness with fovista combination therapy were seen in patients with all CNVM lesion types and these patients experienced greater retraction of the CNVM by fluorescein angiography. Visual benefits due to fovista appeared greater in patients with larger neovascular lesions. Patient receiving sequential injections of ranibizumab and fovista in the combination arms experienced transient increases in intraocular pressure but no other safety concerns were identified.
Clinical Experience
The intraocular concentrations of several growth factors and chemokines including PDGF have been measured in eyes with chorioretinal vascular disorders (AMD, DR, RVOs). Whereas the intraocular concentrations of VEGF are elevated in all of these conditions, the concentrations of PDGF are variable. In eyes with non-proliferative DR, vitreous concentrations of VEGF, PDGF-AA and PDGF-AB were elevated compared with controls, whereas levels of PDGF-BB were decreased. The absolute levels of PDGF-AA and VEGF correlated with the severity of the retinopathy. Aqueous concentrations of PDGF-AA were elevated in patients with DME but were normal in those with BRVO.
In patients that underwent cataract surgery, baseline aqueous levels of IL-6, IP-10, VEGF and macrophage chemotactic protein (MCP)-1 were higher in patients with DME compared with controls. Each patient received an intravitreal injection of 4 mg triamcinolone in 1 eye and 1.25 bevacizumab in the other. Four weeks later, levels of IL-6, IP-10, MCP-1, VEGF and PDGF-AA were decreased in the eyes that received triamcinolone, whereas only VEGF levels were lower in eyes that received bevacizumab. The foveal edema had improved more in the triamcinolone group.
Studies consistently show elevated levels of VEGF and all PDGF dimers in patients with PDR. Vitreous concentrations of PDGF-BB are elevated in patients with PDR but not CRVO. Among factors correlated with VEGF in eyes with PDR, IL-10 and IL-13 were positively correlated and PDGF was negatively correlated (but still elevated). These data suggest that the inflammatory reaction may be more active in eyes with CRVO than in those with PDR.
Increased aqueous concentrations of VEGF and decreased concentrations of PDGF-AA have been measured in patients with AMD, and after ranibizumab injections only the VEGF concentrations decreased. In another study, vitreous PDGF levels were higher in the control group than in the CNVM/AMD group (16.4 + 7.1 vs 2.8 + 1.3 ng/μl) but there was no significant difference in VEGF concentrations (27.4 ± 41.0 vs 39.3 ± 28.9 pg/ml). An insignificant difference in PDGF concentrations between the CNVM/AMD and control groups (1219 vs 1469 ng/ml) was found, but VEGF concentrations were undetectable in both groups. Elevated aqueous VEGF concentrations were measured in patients with intermediate and advanced (but not early) AMD. PDGF levels were depressed more in advanced AMD patients, which suggested an involvement of PDGF in AMD that warrants further investigation.
Several potent non-selective tyrosine kinase inhibitors that interrupt both VEGF and PDGF signaling have been approved for cancer therapy: imatinib, sorafenib, pazopanib, sunitinib, nilotinib and dasatinib. Human RPE cells become less viable and express VEGF and PDGF when exposed to light; sorafenib and pazopanib decrease both VEGF and PDGF-BB while ranibizumab decreases only VEGF. Sorafenib and pazopanib decrease expression of VEGFR1, VEGFR2 and neuropilin 1.
A small number of patients with exudative AMD have been treated with oral sorafenib, a non-selective receptor tyrosine kinase inhibitor that inhibits the activation of VEGFR2, PDGF, FLT3 Ret and c-Kit. A patient with renal cell carcinoma and co-incident, occult CNVM experienced improved macular edema when treated with sorafenib. Three patients with exudative AMD that responded poorly to ranibizumab repeatedly resolved their edema when treated with sorafenib. One explanation for this finding is that non-VEGF pathways, such as PDGF, may be important in some patients with treatment resistant AMD.
The side effects of orally administered receptor tyrosine kinase inhibitors with anti-PDGF effects include rash, fatigue, nausea, pruritus and headache. These may be endured by patients with cancer but those with non-malignant conditions such as AMD are unlikely to tolerate such side effects for years. Therefore, oral RTK inhibitors are not likely to be used for the long-term treatment of exudative AMD. Monoclonal antibodies generally produce fewer side effects than RTK inhibitors, but because of their large molecular weight and poor gastrointestinal absorption, they cannot be taken orally. However, experience with antibodies and fragments shows that a regimen consisting of intravitreal injections may be tolerated for years. Furthermore, intraocular injections are quite safe and result in fewer systemic adverse events than are seen with RTK inhibitors.
Pazopanib, a small RTK inhibitor that blocks several receptors, was evaluated as topical therapy for exudative AMD. In a Phase IIa trial, patients gained an average of +4.3 letters at day 28, though macular thickness did not significantly improve. However, in a 19 patient, 12-week Phase IIb trial, pazopanib failed to improve visual acuity and macular thickness. Because of these disappointing results, the manufacturer (GlaxoSmithKline) decided to halt further AMD testing.
In a Phase I tolerability study, single intravenous doses of an anti-PDGF-D antibody were administered to 40 healthy volunteers. This high affinity drug (KD = 0.18 nM) demonstrated linear pharmacokinetics with a systemic half-life of 15.5–48.1 days.
Clinical testing of specific PDGF inhibitors for ocular disease began with the use of E10030 (fovista, Ophthotec, Princeton, NJ, USA), a 50 kDa aptamer that binds to PDGF-B. A Phase I escalating dose trial enrolled 57 patients with advanced exudative AMD. Patients were given a single dose of 0.03 mg fovista with 3 monthly injections of ranibizumab, or 3 doses of fovista (0.03, 0.3, 1 or 3 mg) together with three doses of ranibizumab. For patients completing the 4-, 8- and 12-week visits, the proportions of patients gaining at least 15 ETDRS letters were 32, 45 and 60%, the mean improvements in vision were +11.1, +13.4 and +15.7 letters, and the mean center point thicknesses were 251, 231 and 229 μm (compared with 395 μm at baseline). Fluorescein angiographic regression of the neovascular lesions occurred in 85% of eyes. None of the patients developed drug-related adverse events.
In a Phase IIb, prospective, superiority trial, 449 patients were randomized to receive monthly ranibizumab monotherapy, or combination therapy consisting of ranibizumab with fovista (0.3 or 1.5 mg) for 24 weeks. The combination group receiving 1.5 mg fovista met the primary superiority end point, with more improvement in vision compared with ranibizumab monotherapy (+10.6 vs +6.5 letters; p = 0.019). Compared with patients receiving monotherapy, those receiving fovista combination therapy enjoyed greater gains at every monthly visit and the differences increased throughout the trial. Compared with ranibizumab monotherapy, the improvements in vision and macular thickness with fovista combination therapy were seen in patients with all CNVM lesion types and these patients experienced greater retraction of the CNVM by fluorescein angiography. Visual benefits due to fovista appeared greater in patients with larger neovascular lesions. Patient receiving sequential injections of ranibizumab and fovista in the combination arms experienced transient increases in intraocular pressure but no other safety concerns were identified.
Source...