hI-con1™ AND WET AGE-RELATED MACULAR DEGENERATION

Choroidal neovascularization (CNV) is intrinsic to Wet Age-Related Macular Degeneration (wet AMD). In wet AMD, the leakage of fluid and blood from CNV causes rapid vision loss. The current standard of care for wet AMD is Lucentis® and accounts for > 90% of the treatment dollars. Other treatments include Macugen® and Avastin® (used off-label). The mechanism by which these agents work is through the inhibition of vascular endothelial growth factor (“VEGF”), one of a number of vascular growth factors that are known to promote the CNV growth. VEGF was also isolated as Vascular Permeability Factor and it appears that in wet AMD, in addition to inhibiting the formation of new CNV, anti-VEGF agents reduce fluid leakage from existing CNV, which leads to the improvements in vision observed with such agents.

These anti-VEGF agents have several significant limitations. First, their effect is temporary. On average, wet AMD patients receive intraocular injections of anti-VEGF agents every 4-6 weeks for the duration of their lives. At $2000 per injection, the 2009 cost of treating wet AMD well exceeded $1.0B. The second significant limitation is that anti-VEGF agents do not destroy CNV, but merely control further growth and leakage. After treatment with an anti-VEGF agent, CNV are left intact with the potential to leak again. As indicated above, this occurs on average every 4-6 weeks.

A drug that could destroy CNV would have great utility in the treatment of wet AMD. The destruction of CNV could lead to improved vision, reduced treatment frequency and lower lifetime treatment costs. hI-con1 is potentially such a drug.

Efficacy of I-con1 in a Laser-induced Choroidal Neovasculature (CNV) Model in Mice

I-con1 has been shown to trigger destruction of CNV in two different animal models of wet AMD (Bora et al., 2003; Tezel et al., 2007). The early work with I-con1 used the I-con1 gene inserted into a replication-defective adenoviral vector that expressed the protein (Hu et al., 1999; Hu and Garen, 2000). Initially, this vector was used in place of the isolated recombinant protein in early pharmacodynamic testing. Later, Drs. Hu and Garen were able to purify sufficient quantities of I-con1 from CHO cells transfected with a plasmid containing I-con1’s gene. The murine data outlined below describe the results from pharmacodynamic testing using both the purified I-con1 protein and the adenoviral vector.

A krypton red laser was used to generate three spots in each eye surrounding the optic nerve of anesthetized C57BL/6 mice. At each laser spot, the Bruch’s membrane is ruptured and CNV grows in response to this injury. Various treatment routes and regimens with either the adenoviral vector or the recombinant protein were given in the period from Day 1 to Day 14. The laser spots were evaluated for the presence of CNV on Day 7, 14, or 17 after laser treatment, using confocal microscopy. An example of the CNV induced in the laser spots and their prevention by treatment with I-con1 is shown in Figure 1.

FIGURE 1: EFFECTS OF I-con1 TREATMENT IN A MOUSE MODEL OF WET AMD

A and B are confocal micrographs of laser lesion spots in the eyes of mice injected intravenously one day after laser treatment with control adenovirus vector (A) or adenovirus vector containing the gene for I-con1 (B). Six days later, the mice were perfused with FITC-dextran. The mice were then sacrificed and the eyes were removed, immunostained for elastin (a component of Bruch’s membrane) and flat-mounted for microscopy. Green fluorescence denotes blood vessels; red fluorescence denotes Bruch’s membrane.

Intravitreal Treatment

After laser spots were induced on Day 0, intravitreal (IVT) injections (1 µL) that contained 1.5 µg I-con1, 1.5 µg of human IgG1 or PBS were administered on Day 1. The mice were killed on Day 7, perfused with PBS containing 50 mg/L fluorescein-labeled dextran and their eyes examined by confocal microscopy for the presence or absence of CNV. The results are shown in Figure 2.

Laser spots were produced in C57BL/6 mice on Day 0, and IVT injection of 1 µL of PBS or human IgG1 (1.5 µg/µL) as controls or with 1 µL of I-con1 protein (1.5 µg/µL) was administered on Day 1. The mice were killed on Day 7. A Chi-square test of the data showed P < 0.0001 for I-con1 vs. controls.

These results demonstrate that IVT treatment with 1.5 µg of I-con1 one day after laser injury significantly reduced the development of CNV from > 90% CNV positive spots in the controls to 10 % CNV positive spots in the I-con1-treated mice.

Intravenous Injection

Two studies were conducted using a similar protocol to that used for the study summarized in Figure 2 except that I-con1 was administered intravenously (IV) on Day 1 instead of by the IVT route. The results are shown in Figure 3.

Laser spots were produced in C57BL/6 mice on Day 0, and an IV injection of 10 or 50 µg human IgG1 as controls or with 10 or 50 µg of I-con1 protein was administered on Day 1. The mice were killed on Day 7. A Chi-square test of the data showed P < 0.0001 for 50 µg dose.

These data indicate that I-con1, when injected IV the day after production of laser-induced spots in mice prevented the formation of CNV in a dose-related fashion. The incidence of CNV-negative spots was 44% in the three mice injected IV with 10 µg of I-con1, as compared with 17% in the three mice injected with an IgG1 control [p = 0.07, NS] (Figure 3). When the amount of I-con1 was increased to 50 µg, the incidence of CNV-negative spots increased to 89% as compared with 11% for the IgG1 control [p < 0.0001] (Figure 3).

Taken together, these data indicate that I-con1, whether injected IVT or, at much higher doses, systemically via the IV route, prevented the formation of CNV in a laser-damaged Bruch’s membrane in mice. Experiments with the adenoviral vector, reported in this same publication, indicated that the treatment could be delayed for 7 days (by which time the CNV would be formed) and the I-con1 (produced from the vector) still significantly reduced the incidence of CNV at Day 14 or 17. These results suggest that I-con1 could be useful in treating wet AMD.

Efficacy of hI-con1 in a Laser–induced CNV Porcine Model

Once Iconic made sufficient quantities of the purified hI-con1 protein, it set about to conduct additional in vivo studies. Because the pig eye has retinal vascular similarities to human eyes and several cone-dominant regions of the retina that are similar to the human macula, the pig is an excellent model to test hI-con1. The pig is also an appropriate model for wet AMD in that laser-induced lesions do not resorb as quickly as in other animal models (Kiilgaard et al., 2005). This allows for evaluation of the activity of hI-con1 against well-established CNV.

Using the laser model developed by Kiilgaard et al., laser-induced microrupture of the Bruch’s membrane was induced in anesthetized Yucatan pig eyes (74 spots/eye), which resulted in CNV in ~ 80% of the spots by Day 10. At this time, 100 µL of solutions of hI-con1 were administered by IVT injection into both eyes of the pigs; control pigs received 100 µL IVT injections of formulation buffer into both eyes. On Day 14, the pigs were anesthetized, perfused with PBS containing fluorescein-labeled dextran and their eyes removed and processed for analysis of the presence or absence of CNV in each laser spot, as described by Bora et al., 2007. Absence of CNV was defined stringently as the total absence of green fluorescence in the vessels in the spot (see Tezel et al., 2007).

The percentage of laser spots with CNV at different doses of hI-con1 was plotted against dose using a 5-parameter Sigmoidal Weibull curve, which was used to calculate the dose of hI-con1 that reduces the fraction of laser spots with CNV by 50% (ED50). The curve for hI-con1 is shown in Figure 4.

FIGURE 4: EFFECT OF INTRAVITREAL TREATMENT WITH hI-con1
ON LASER-INDUCED CNV IN THE PIG