Photodynamic Therapy for Corneal Neovascularization Using Polymeric Micelles Encapsulating Dendrimer Porphyrins

Abstract

purpose. To investigate the accumulation of new photosensitizers (PSs), dendrimer porphyrin (DP, free DP), and DP encapsulation into polymeric micelles (DP-micelle) and the efficacy of photodynamic therapy (PDT) in an experimental corneal neovascularization model in mice.

methods. Corneal neovascularization was induced by suturing 10–0 nylon 1 mm away from the limbal vessel in C57BL6/J mice. To determine the accumulation of free DP and DP-micelle, 10 mg/kg free DP or DP-micelle was administered by intravenous injection 4 days after suture placement. Mice were killed 1, 4, 24, and 168 hours after the injection of PS. Twenty-four hours after the administration of free DP or DP-micelle, mice were treated with a diode laser of 438-nm wavelength at 10 or 50 J/cm². Fluorescein angiography was performed before and 7 days after irradiation, and the area of corneal neovascularization was quantified.

results. Free DP and DP-micelle accumulated in the neovascularized area 1 hour to 24 hours after administration. Fluorescence of DP was weaker than that of DP-micelle. Neither DP-micelle nor DP could be detected in normal limbal vasculature. In the PDT experiments using PS, mean residual rates of corneal neovascularization were 10.1% in the mice treated with DP-micelle and 21.6% in the mice treated with free DP at 10 J/cm² (P < 0.01). At 50 J/cm², mean residual rates of corneal neovascularization were 10.6% in the mice treated with DP-micelle and 13.7% in the mice treated with free DP (P > 0.05). Although corneal neovascularization in PDT-treated mice exhibited significant regression compared with the control group, significant energy-related vessel regression with increasing laser energy could not be observed.

conclusions. PDT with DP-micelle and free DP can provide efficacious treatment of corneal neovascularization.

Corneal neovascularization is a major sight-threatening condition and is caused by infections, inflammation, degenerative disorders, and long-time contact lens wear. ¹ This major ocular complication can lead to corneal scarring, edema, lipidic deposition, and inflammation that may not only compromise visual acuity but also decrease the success rate of subsequent penetrating keratoplasty. ² In the clinical setting, topical corticosteroids and nonsteroidal anti-inflammatory drugs (NSAIDs) remain the principal primary treatment for suppressing proliferating corneal vessels. ³ However, in corneas in which vessels have been established for extended periods, corticosteroid and NSAID treatment are ineffective. Although laser photocoagulation for corneal neovascularization has been reported, ^{4 5 6 7} this method achieves an inadequate effect because of the high incidence of recanalization and thermal damage to adjacent tissue. ⁴ The discovery of the many factors involved in corneal neovascularization and their mechanisms of action has been followed by efforts to develop new drugs specifically targeting these molecules. For example, therapy targeting vascular endothelial growth factor (VEGF) looks promising for the treatment of corneal neovascularization. ⁸ However, the agents tested thus far have yet to become available clinically. 

Photodynamic therapy (PDT) has been introduced recently as a novel treatment for corneal neovascularization. ⁹ In this therapy, a photosensitizer (PS) is injected systemically and accumulates in newly formed vessels; it is then activated by mild laser excitement to liberate cytotoxic reactive oxygen species (ROS) that selectively occlude the target vessels. Although benzoporphyrin (verteporfin), a PS, is used for choroidal and corneal neovascularization clinically, ⁹ nonspecific binding activities of verteporfin induce skin phototoxicity in bright conditions, and patients must remain in the dark for 48 hours after injection of this drug. Hence, innovative PS should be developed for realizing safe and effective PDT. 

Specific delivery of a PS to the neovasculature site might be a promising way to achieve safe and effective PDT for corneal neovascularization. Drug vehicles such as liposomes can be used for this purpose; however, the self-quenching effect of PS caused by aggregate formation could decrease the efficiency of ROS production. To solve this problem, we have recently developed dendrimer porphyrin (DP) as a novel PS for drug delivery (Fig. 1A) . ¹⁰ It is assumed that the dendritic framework of DP might prevent the interactions of the center dye molecules, thereby achieving efficient ROS production even at extremely high concentrations. Indeed, encapsulation of DP into polymeric micelles (DP-micelle), which are characterized by the polyion complex core surrounded by poly(ethylene glycol) (PEG) palisades (Fig. 1C) , resulted in remarkably increased photocytotoxicity. ^{11 12} We previously reported the review and general introduction of drug delivery of these PSs in corneal neovascularization. ¹³ In this study, to demonstrate the potential usefulness of DP and DP-micelle for PDT of corneal neovascularization, we investigated the accumulation of those PS formulations and their efficacy of PDT in an experimental corneal neovascularization model in mice. 

Materials and Methods

Animals and Experimental Corneal Neovascularization

Eight-week-old male C57 BJ/6 mice were maintained with free access to food and water. All procedures were performed in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. Animals were placed under general anesthesia by the administration (1.5 mL/kg) of a mixture of ketamine hydrochloride (Ketalar; Sankyo, Tokyo, Japan) and xylazine hydrochloride (Celactal; Bayer, Tokyo, Japan). Corneal neovascularization was induced by suturing 10–0 nylon 1 mm away from limbal vessel under microscopy. Erythromycin ophthalmic ointment was instilled immediately after the procedure. 

Photosensitizers

In this study, a third-generation aryl ether dendrimer zinc porphyrin with 32 carboxyl groups on the periphery (DP) and polymeric micelles composed of the DP and PEG-b-poly(l-lysine) (DP-micelle) were used for PDT as a PS formulation (Fig. 1) . The DP-micelle was prepared according to a previous report. ¹¹ Both DP and DP-micelle have a maximum excitation wavelength at 433 nm. DP-micelle showed 130- to 280-fold higher photocytotoxicity against murine Lewis lung carcinoma cells compared with free DP. ¹¹  

Accumulation of DP and DP-Micelle in Corneal Neovascularization Lesions

Four days after suture placement, DP or DP-micelle was administered by intravenous injection at the dose of 10 mg/kg, again under general anesthesia. Mice were killed 1, 4, 24, and 168 hours after the injection of PS. Before the kill, mice received intravenous BS-1 lectin conjugated with FITC (500 μg/g; Vector Laboratories, Burlingame, CA) to trace the corneal neovascularization area. Corneas were excised and flatmounted on glass slides. Accumulations of DP or DP-micelle in vascularized areas were observed by fluorescence microscopy (Leica, Deerfield, IL) using 436-nm excitation wavelength. Fluorescence intensities were calculated using NIH Image software and were normalized by traced vascularized areas. 

Photodynamic Therapy

Twenty-four hours after DP-micelle or free DP, 38 mice were treated with a diode laser (in-house built laser equipment; Topcon, Tokyo, Japan) of 438-nm wavelength at 500 mW/cm² for 20 or 100 seconds, for a total dose of 10 or 50 J/cm². The spot size was 1 mm in diameter. As controls, six mice with corneal neovascularization were irradiated without administration of photosensitizers (total dose, 50 J/cm²). Fluorescein angiography was performed before and 7 days after treatment, and the area of corneal neovascularization was quantified using NIH Image software before and 7 days after irradiation. We defined residual ratio as follows: Residual ratio = (neovascularization area 7 days after irradiation/neovascularization area before irradiation) × 100%. 

Results

Accumulation of DP and DP-Micelle in Corneal Neovascularization Lesions

One hour after the administration of DP-micelle, fluorescence started to accumulate in the neovascularized area and increased until 24 hours after administration (Figs. 2A 2B 2C) . DP-micelle accumulation decreased but continued after 168 hours (data not shown). After DP administration, fluorescence of DP was observed in the corneal neovascularization area, but it was weaker than that of DP-micelle group (Figs. 2D 2E 2F)and disappeared by 168 hours (data not shown). Figure 3shows the time course of normalized fluorescence intensities of DP in the neovascularized area. DP-micelle intensities were significantly higher than those of free DP 1 hour (n = 7/each condition), 4 hours (n = 6/each condition), and 24 hours (n = 6/each condition) after administration (P = 0.032, 0.047, 0.0066; Mann-Whitney U test). Neither DP-micelle nor free DP could be detected microscopically in normal limbal vasculature or other ocular tissue such as iris, retina, and conjunctiva (data not shown). 

Photodynamic Therapy

Figure 4is a series of fluorescence angiographic images of corneal neovascularization before and 7 days after PDT in the control, DP-micelle, and DP groups. As shown in Figure 4A , there was no effect of irradiation in the control (n = 6). Seven days after irradiation at 10 J/cm², the mean residual ratio of corneal neovascularization was 10.1% in the mice treated with DP-micelle (n = 9) and 21.6% in the mice treated with free DP (n = 10; Fig. 5 ). The residual ratio of mice treated with DP-micelle was significantly higher than that of mice treated with free DP (P < 0.01; Mann-Whitney U test). Seven days after irradiation at 50 J/cm², the mean residual area of corneal neovascularization was 10.6% in the mice treated with DP-micelle (n = 10) and 13.7% in the mice treated with free DP (n = 9; P > 0.05; Mann-Whitney U test; Fig. 5 ). Histologic examination after PDT showed no injury on corneal tissue (data not shown). 

Discussion

PDT potentially represents a new approach for the treatment of neovascular disease and tumor. ¹⁴ The successful treatment of choroidal neovascularization by PDT opens the possibility of treating other neovascular diseases of the eye, including the cornea, in a similar manner. ^{15 16 17} Indeed, several past studies revealed that PDT was efficacious for the treatment of corneal neovascularization. ^{18 19 20 21}  

One of ideal characteristics of PS is selectivity to neovasculature. In this study, DP-micelle, a newly developed PS formulation, was accumulated only in pathologic vascularized areas. On the other hand, DP-micelle and free DP were not detectable in normal vessel areas, including other ocular vessels (data not shown). The high selectivity of these PSs to neovascularization should result in minimal side effects in the adjacent normal corneal structure. Indeed, slit lamp examination and histologic observation showed no alterations by PDT to the surrounding structure, including epithelium, stroma, and endothelium (data not shown). 

Macromolecular compounds can accumulate and prolong their retention in the perivascular regions of solid tumors to a greater extent than in normal tissues because newly formed vessels in solid tumors exhibit higher substance permeability than in normal tissues and because lymph systems in tumor tissue are incomplete. This effect is known as the enhanced permeability retention (EPR) effect. ²² We have demonstrated that polymeric micelles with diameters of several tens of nanometers with narrow distribution can accumulate in a solid tumor though the EPR effect. ^{23 24 25}  

As do solid tumors, corneal neovascularization sites appear to have high permeability and incomplete lymph systems. Hence, we assume that DP-micelle can accumulate selectively in corneal neovascularization lesions. 

Another ideal characteristic in PDT is high photocytotoxicity with lower dark cytotoxicity. As reported earlier, the laser energy of PDT for corneal neovascularization with verteporphin, a PS used clinically for choroidal neovascularization, was 150 J/cm² to create long-lasting clinical regression of corneal neovasculature. ¹⁹ This is three times the amount needed to create regression of choroidal neovascularization. PDT with DP-micelle required only 10 J/cm² for long-lasting regression of neovascularization. Most PSs have hydrophobic properties resulting in their self-quenching from aggregation, decreasing photooxidation efficacy to achieve successful PDT. Dendrimer photosensitizers are designed to prevent the aggregation of core PS even in the micellar core. In addition to the EPR effect, this property might have led to a higher neovascularization regression rate in our study than in past studies. 

This study indicates PDT with DP-micelle and free DP can provide efficacious treatment of corneal neovascularization. It is important to observe and examine the long-term effects in the future. We observed 2 months after irradiation with PIC micelle at laser energies of 10 J/cm². As shown in Figure 6 , most of the corneal neovasculature was not recanalized, though a few matured vessels remained over 2 months. In addition, polymeric micelles can encapsulate a variety of drugs, including hydrophobic substances, nucleic acids, and proteins in the core ²³ ; therefore, they have great potential for effective drug delivery targeting to corneal neovascularization. 

 Supported in part by a Grant-in-Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science and Technology of Japan.

 Submitted for publication April 2, 2007; revised June 27, 2007; accepted January 25, 2008.

 Disclosure: K. Sugisaki, None; T. Usui, None; N. Nishiyama, None; W.-D. Jang, None; Y. Yanagi, None; S. Yamagami, None; S. Amano, None; K. Kataoka, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 1.

 

Schematic structure of polymeric micelle encapsulating DP. Chemical structures of poly(ethylene glycol) (PEG)-b-poly(l-lysine) (A) and ionic dendrimer porphyrin (X = COO⁻) (DP) (B). DP-micelle is spontaneously formed through the electrostatic interaction between PEG-b-poly(l-lysine) and DP (C).

Figure 2.

 

Accumulation of DP and DP-micelle to neovascularized area. The flatmounts of the neovascularization area were observed under a fluorescent microscope 1 hour (A), 4 hours (B), and 24 hours (C) after administration of DP-micelle and 1 hour (D), 4 hours (E), and 24 hours (F) hours after injection of DP. Vessels were stained with BS-1 lectin conjugated with FITC. Fluorescence of DP was observed 1 hour after administration (A, D) and enhanced until 24 hours after.

Figure 3.

 

Time course of fluorescence intensity. Fluorescence intensity of DP in cornea after administration of DP-micelle or free DP. Fluorescence of DP was detected in the corneas of the DP-micelle group as early as 1 hour, and intensity reached a peak 24 hours after intravenous administration, whereas the intensity in the free DP group was significantly lower than that in the DP-micelle group.

Figure 4.

 

Neovascularization regression after PDT. (A) Control. No regression was observed without micelle or DP administration. (B) Irradiation at 10 J/cm² with DP-micelle. (C) 50 J/cm² with DP-micelle. (D) Irradiation at 10 J/cm² with DP. (E) Irradiation at 50 J/cm² with DP. (B–E) In each group, significant regression of neovascularization was observed.

Figure 5.

 

Residual ratio of corneal neovascularization 7 days after PDT. Residual ratio of corneal neovascularization 7 days after PDT with DP-micelle or DP or with no administration. Residual ratios were 10.1% with irradiation at 10 J/cm² and micelle, 21.6% with irradiation at 10 J/cm² and DP, 10.6% with irradiation at 50 J/cm² and micelle, and 13.7% with irradiation at 10 J/cm² and DP. Error bars indicate SD.

Figure 6.

 

Long-term effect of PDT. (A) Before irradiation. (B) Seven days after irradiation at 10 J/cm² with micelle. Significant regression of neovascularization was observed. (C) Sixty-three days after irradiation. Little recanalization of neovascularization was observed, but a few matured vessels were not occluded.