The development of safe, efficacious, and cost-effective local drug delivery technology to treat posterior segment disorders remains a major challenge in ophthalmology.1 Therapeutic targets include diseases or degenerations of the retina and optic nerve. Current local drug delivery techniques to access the posterior segment are intravitreal injections, periocular injections (subconjunctival, subtenon, or juxtascleral), and intravitreal implants.2

Recently, we investigated a method of drug delivery to the suprachoroidal space with the goal of obtaining high local tissue delivery levels of an agent and low systemic levels. In 2002, Einmahl and colleagues from Dr. Behar-Cohen’s lab, first described an injection of polyorthoesters into the suprachoroidal space in rabbit eyes using a rigid cannula.3 Researchers observed the animals for 3 weeks and noted mild pigmentary changes in the choroid. They also described the feasibility of this approach, along with some limitations.

Our novel approach takes advantage of a new technology, microcannulation system (posterior delivery system [PDS]; iScience Interventional, Menlo Park, CA, Figure 1) originally designed for use in Schlemm’s canal and optimized for posterior segment drug delivery. We performed an anterior sclerotomy and cannulated the suprachoroidal space with the microcannula, directing it toward the posterior segment tissues under direct visualization.

MICROCANNULA SYSTEM
The advantage of the microcannula system is that it combines a drug delivery channel using a viscous fluid injection system, a fiberoptic light source (for localization), and with optimal transition properties (ie, pushability). There are depth markers on the cannula itself; we used a wide-angle viewing system to visualize the cannula in the suprachoroidal space. A red flashing tip identified the tip of the cannula, so that the drug may be delivered under direct visualization (Figure 2).

We delineated the suprachoroidal space using corrosion dye casts of both human and pig cadaver eyes. The surgical technique was tested in both a primate and pig model. While pigs do not have a macula, they do have an area centralis that has cone function. When the drug was injected into the pig eye, we saw a bolus of fluid form around the site of intended injection, such as the perimacular region or around the optic nerve. Fortunately, many pigs have a mosaic pigment pattern. In some areas of choroidal depigmentation, we are able to document direct visualization of the cannula and the flashing tip directly in the suprachoroidal space. As the agent was injected the tip of the light became more diffuse. In order to properly localize the tip, we simply turned the light off and on to verify that it was in the proper location. With some injections, the cannula passed into the area between the area centralis and the optic nerve. As it was retracted, the drug bolus was delivered directly in the cannula tract. Therefore, the drug was delivered precisely between the optic nerve and area centralis.

In our report, we injected 94 pigs, and have since performed more than 140 injections to date in our lab. One animal suffered an iatrogenic hemorrhage that occurred with a tissue forcep (rather than the cannula). No hemorrhages occurred in any of the remaining eyes. Additionally, we had one case of endophthalmitis that resulted from an infected suture site, and one case of a wound abscess. We noted scleral ectasia (4/94), choroidal blood flow abnormalities (4/94), and some inflammation (6/94). In one pig eye the optic nerve was rather forcefully impacted with the cannula. Visible injury was noted on red-free imaging, emphasized on infrared imaging, also with indocyanine green and fluorescein angiography, and confirmed with pathology. By 1 month the vascular flow changes completely normalized.

During the injection, a dimpling effect was seen and believed to result from adhesions between the choroid and the sclera, in the area of the short posterior ciliary vessels. This disappeared within 1 week. The agents that we have studied include triamcinolone (1.5 mg or 3 mg in 12 mL of 1% sodium hyaluronate [Healon; Advanced Medical Optics, Santa Ana, CA]) and also bevacizumab (Avastin; Genentech, San Francisco) (studies pending). When the cannula tip is modified or a straight-bore cannula is used, it causes snag crescent shaped lesions in the pigment epithelium. Therefore, the streamlined tip of the device is very important. In control animals, we injected 1% sodium hyaluronate in order to show the separation of the suprachoroidal space from the sclera. After 1 month, this potential space retracts and falls back to its normal position.

PHARMACOKINETICS
The pharmacokinetics from 30 days to 120 days—after a single injection of triamcinolone—showed very little decline in the tissue levels in the choroid confirmed at the posterior vitreous (Figure 3). Serum drug levels were extremely low, with a maximal reading of 387 ±91 picogram/mL serum at 2 hours in the 3-mg group, to less than our detectable limit after 40 days (0.5 picogram/mL detection limit; Figure 4).

CONCLUSIONS
We have described a methodology of drug delivery to the posterior pole (macula and optic nerve) using an advanced microcannula system and demonstrated relative safety by studying the choroidal and retinal blood flow with high-speed cSLO video angiography, fundus exam by cSLO imaging and wide-field fundus photography, and short- and long-term histopathology in the pig model. These data have been used to justify the safety of this methodology for use in humans. Clinical trials are currently investigating the potential use of this technology for treatment of macular disease. As a warning, modification of the tip or aggressive insertion of the cannula should be avoided.

In addition, we have characterized a relatively large potential anatomic space for suprachoroidal drug delivery, as well as demonstrated unique pharmacokinetics, with at least 120 days of sustained delivery of triamcinolone in the posterior tissues and very low systemic levels. Drug delivery via the suprachoroidal space using PDS may be useful in the treatment of posterior segment disease. Further clinical studies are necessary to better define the role of this technology in the management of posterior segment disease.

Timothy Olsen, MD, is Professor of Ophthalmology, Director of Retina, Director of the Minnesota Lions Macular Degeneration Center, and holds the William H. Knobloch Retina Chair at the University of Minnesota. Grant funding for this work has been approved by the University of Minnesota, sponsored projects administration with funds provided by iScience Interventional, Inc. Dr. Olsen has no direct financial interest in any of the products or items mentioned and has not accepted consulting, travel, meeting, or other outside fees for this work. Dr. Olsen may be reached at olsen010@tc.umn.edu.

1. Olsen TW. Drug delivery to the suprachoroidal space. Presented at Retina 2006: Emerging New Concepts. Held in conjunction with the American Academy of Ophthalmology annual meeting. Nov. 10-11, 2006. Las Vegas.
2. Geroski DH, Edelhauser HF. Drug delivery for posterior segment eye disease. Invest Ophthalmol Vis Sci. 2000;41:961-964.
3. Einmahl S, Savoldelli M, D’Hermies F, et al. Evaluation of a novel biomaterial in the suprachoroidal space of the rabbit eye. Invest Ophthalmol Vis Sci. 2002;43:1533-1539.