Multiple approaches have been used to deliver drugs to the posterior pole. Of these approaches (Table 1), the two that are used most commonly are intravitreal injections and the implantation of sustained delivery devices into the vitreous cavity.1 Both of these approaches, however, can be associated with multiple serious complications including endophthalmitis, retinal detachment, vitreous hemorrhage, and ocular trauma. Additionally, drugs delivered into the vitreous have the disadvantage of poor access to the outer retina and choroid, the anatomic location of diseases such as age-related macular degeneration (AMD) (Figure 1).

Periocular delivery through a subconjunctival, suprachoroidal, or juxtascleral method offers the advantage of a noninvasive approach. These methods may be especially suitable for the long-term treatment of disease such as diabetes, nonneovascular AMD, uveitis, and especially in prophylactic therapy to prevent progression and vision loss.

CURRENT STATE-OF-THE-ART Method?
Juxtascleral administration of anecortave acetate, for example, has been reported to result in therapeutic levels of the drug in the choroid and the retina for up to 4 to 6 months after administration. A subconjunctival injection of water-soluble dexamethasone appears to enter the eye and is detectable in the subretinal fluid up to 20 hours after a presurgical injection.2

Researchers have injected microspheres containing the antiangiogenic compound PKC 412 subconjunctivally in a porcine model of laser-induced choroidal neovascularization (CNV) Figure 2.3 Panel A in Figure 2 shows the experimental CNV control lesion while panel C shows the barely detectable lesion after therapy.

In these experiments, the drug was detected in the retina 20 days after injection. For true long-term therapy, however, the use of a sustained delivery device may be essential. We have been exploring this technique and have studied the long-term safety of these devices in rabbits. For reservoir devices made of silicon or polyvinyl alcohol, indeed, these implants are highly tolerable for longer than 1 year.

TYPES OF SUBCONJUNCTIVAL IMPLANTS
There have been several studies on various types of subconjunctival implants. Of great importance are the limitations of this approach, which have been demonstrated. The release of lipophilic fluorescein in a subconjunctival implant shows that the drug penetrates a small anatomic area with subsequent poor levels in the retina (Figure 3). This tells us that the limitation of this approach might not be technical but scientific, namely understanding the question of why drugs do not enter the eye from an episcleral location.

To study this question we have been using MRI to examine drug distribution from a subconjunctival location. MRI has the advantage of allowing us to follow the drug in a live animal. In addition, MRI has the potential to allow us to extend our animal data and study ocular distribution in patients as well.

We placed subconjunctival implants containing the MRI contrast agent gadolinium in rabbits and studied drug distribution in both live animals and animals posteuthanasia. Previous work had demonstrated that the sclera was not an important barrier to drug delivery. We found that no drug penetrated the posterior segment in the live animal (Figure 4), and a small amount of drug entered the anterior chamber. To our surprise, however, once the animal was euthanized, periocular drug appeared to pour into the eye. These results led us to the conclusion that the main barrier to drug delivery was actually physiologic and not anatomic in nature.

To verify this result we performed direct drug level detection in rabbits following subconjunctival injections of triamcinolone. The drug was not detected in the vitreous in the live animal, but following euthanasia the amount increased in the vitreous, confirming the importance of these physiologic barriers.

PHYSIOLOGIC BARRIERS
What are the main physiologic barriers to ocular drug penetration? There are three important ones: the episclera and conjunctival lymphatics and blood vessels, and the blood flow of the choroid. To understand more fully the relative contribution of these mechanisms, surgical elimination of each barrier was performed in rabbits. To eliminate the role of choroidal blood flow, cryotherapy was performed, and cuts in the conjunctiva and episclera were performed to negate the superficial elimination pathways.

Following each of these steps, triamcinolone was placed over the area and drug levels were measured in the vitreous. Only the animals undergoing the surface elimination of superficial lymphatics and conjunctival blood flow had a change in vitreous levels of triamcinolone, suggesting that this pathway—the episcleral lymphatic and blood flow pathway—may be playing a more important role than we had previously thought.

In fact, when we went back to our original gadilinium implant studies in the live animal to try to find out where the drug went when it did not go into the eye, we could localize the drug in the buccal lymph node, confirming the robust lymphatic clearance of drug from around the eye.

Do these observations have any clinical relevance as we develop transcleral drug delivery modalities? While a trial in humans is still being planned, a trial in horses provides some potential clues. In a trial treating chronic equine uveitis, sustained drug delivery implants containing cyclosporine were initially placed in a subconjunctival location. Using this approach, however, no clinical response was noted.

To our surprise, when the same implant was placed into the suprachoroidal space, a dramatic response ensued, leading to an almost complete resolution of this disease. These results have important implications for future human trials.

It is important to note that a critical aspect of transcleral delivery is the drug compound itself. There are properties of a drug that lend themselves to transcleral delivery, including activity at low levels and lipophilicity. Newer polymer technologies, now under development, may allow for future transcleral delivery of peptides, aptamers, and indeed proteins.

The future of transcleral drug delivery is exciting and clearly represents one of the main important goals of treatment for our patients.

Karl G. Csaky, MD, PhD, is Associate Professor of Ophthalmology and Director of the Ocular Unit of the Duke Clinical Research Institute at Duke University Medical Center in Durham, NC. Dr. Csaky disclosed that he has no financial interest in the products or companies mentioned. He may be reached at karl.csaky@duke.edu.

1. Csaky KG. Transcleral delivery of drugs. Presented at Retina 2006: Emerging New Concepts. Held in conjunction with the American Academy of Ophthalmology 2006 annual meeting. Nov. 10-11, 2006. Las Vegas.
2. Weijtens, O, Shoemaker, RC, Lentjes, EG, et al. Dexamethasone concentration in the subretinal fluid after a subconjunctival injection, a peribulbar injection, or an oral dose. Ophthalmology. 2000; 107: 1932-1938.
3. Saishin Y, Silva RL, Saishin Y, et al. Periocular injection of microspheres containing PKC412 inhibits choroidal neovascularization in a porcine model. Invest Ophthalmol Vis Sci. 2003;44:4989-4993.