The concept of local delivery of pharmacologic agents to the eye has been important for some time. The presence of the blood-retinal barrier renders systemic therapy relatively ineffective. Fortunately, we have an effective posterior drug delivery system available to us—intravitreal injection—and we have been successful with this technique.

Now with the current era of effective pharmacotherapy, however, we are being a bit overwhelmed with the frequency with which we have to give intravitreal injections. Therefore, we are interested in looking at longer-term drug delivery systems. Two of these have already been licensed, both by Bausch & Lomb (Rochester, NY), have been approved by the US Food and Drug Administration, and are extremely effective: Ganciclovir implant for cytomegalovirus retinitis and a fluocinolone acetonide reservoir implant for uveitis. A smaller version of the fluocinolone reservoir implant (Medidur; Alimera Sciences, Alpharetta, GA) is in phase 3 clinical trials. Allergan (Irvine, CA) has a biodegradable dexamethasone implant in phase 3 trials, and a helical coil containing triamcinolone (SurModics, Eden Prairie, MN) is in phase 1 trials. Another new technology is an encapsulated cell implant (Neurotech, Lincoln, RI) that is in phase 1 trial.

RESERVOIR STYLE IMPLANTS: RETISERT AND MEDIDUR
Retisert. Bausch & Lomb’s Retisert reservoir-style implant for fluocinolone is quite remarkable; it is designed to last approximately 1,000 days (Figure 1). The implant contains a total of 0.59 mg of drug, and it delivers a 0.5 µg per day of fluocinolone with a terrific benefit. But we also see, unfortunately, some significant ocular side effects. The concept that such a small amount of drug can be effective, however, highlights the efficacy of drug delivery systems.

There have been uveitis and diabetic macular edema (DME) trials with the Retisert technology, but currently the label indication is for uveitis only. While a benefit is seen in both conditions, the ocular side effects were significant enough that the indication was limited to uveitis.

The type of efficacy we see is not near misses or close calls: 1-year postimplant the reccurence rate for uveitis was 5.4% compared with 46% in the fellow eye. At 3 years that curve begins to close together, but it is still statistically significant. In this case of DME with an extensive cystoid component (Figure 2), complete resolution of the fluid was seen 6 weeks later and the preexisting laser spots that were previously invisible in the face of edema were visible (Figure 3).

The trouble is the ocular side effects of steroids, and the question of whether all steroids behave the same. In our series we saw almost a 50% rate of glaucoma—a significant trade off which seems acceptable in uveitis patients, but not in DME patients.

Additionally, cataracts are almost ubiquitous by the time you get to 3 years in the study: 93% of eyes needed cataract surgery compared to 20% in the control group.

Medidur. The Medidur (Alimera) implant also contains fluoricinolone, but it is much smaller (Figure 4). The implant is a surgical cut down that must be performed through a 3.5-mm incision in the operating room with a 25-ga needle. It is a reservoir implant, but it is not sutured to the eye wall, and it is allowed to float freely in the vitreous space. A phase 3 trial is underway with some limited safety data on 20 patients; 900 patients will be enrolled in a larger phase 3 trial of DME.

POSDUREX BIODEGRADABLE IMPLANT
The Posurdex biodegradable dexamethasone implant (Allergan) is a different concept. It’s biodegradable system is not reservoir based, it uses polymer and drug mixed together and the drug releases over time. As the polymer degrades the drug is released, so finally all that is left is the polymer which completely degrades.

Phase 2 trials tested a version of this implant through a 20-ga incision, surgically inserted in the operating room. The version being tested in phase 2b and 3 trial is inserted in the office setting with no surgical cut down and a 20-ga biplanar pass; it is also longer and thinner than the original version.

We have seen strong efficacy with this technology as well. In a large study of patients with persistent macular edema, 18% of eyes had a three-line improvement in vision out to 180 days (Figure 5). This study, in press with Archives of Ophthalmology, included eyes that had failed all prior therapy. Figure 6 and Figure 7 show a patient before treatment and then again 2 1/2 months after treatment.

EFFECTIVE
These therapies work, and we have found that sequestering the drug into the eye is very effective. The concern with the fluoricinolone implant—and we are looking at this very carefully with the Posurdex implant—is the increase in intraocular pressure (IOP) and the incidence of cataract. So far with 6-month follow-up, we are seeing some evidence of an increase in IOP with the fluorocinolone implant, but to a lesser degree than with triamcinolone. In the current phase 3 trials patients received repeated injections for up to 3 years, therefore we will be getting data to understand the long-term consequence of this technology.

NOVEL HELICAL DESIGN
The I-vation technology from SurModics (Eden Prairie, MN) consists of a helical coil with an eluting polymer containing triamcinolone. This system provides a way of obtaining a controlled release of triamcinolone, the steroid that we are most familiar with and have used most commonly. It is implanted through a 25-gauge needlestick and it is self-anchoring within the sclera (Figure 8 and Figure 9). A prospective, randomized, double-masked multicenter trial is underway with this technology, using two formulations in 30 patients with DME. Enrollment is complete, and this trial will have a 3-year follow-up.

ENCAPSULATED CELL TECHNOLOGY
Encapsulated Cell Technology (Neurotech, Lincoln, RI) ophthalmic implant is different and exciting, but also controversial. This technology uses ARPE-19 cells, a human retinal pigment epithelium (RPE) cell line retinal cell biologists use in their lab. These cells are commercially available and have been modified to produce ciliary neurotrophic factor (CNTF). They can be designed to produce almost any growth factor. Neurotech claims, for example, the cells can produce rhuFab V2, a ranibizumab-like (Lucentis; Genentech, San Francisco) compound. The controversy is that these RPE cells produce other compounds as well and the question is whether these are being disproportionately stimulated. Of course, RPE cells in our own eyes produce growth factors all the time as well.

The Encapsulated Cell Technology implant is 6 mm long and is surgically placed inside the eye (Figure 10). The CNTF elutes over time (Figure 11), and the explants continue to secrete CNTF (Figure 12). Postexplant the cells still appear viable. In this solution to the problem, how can complex proteins remain stable at 37º Centigrade for a prolonged period of time inside the eye, they are “baked fresh daily,” so to speak.

MICROMACHINED DRUG DELIVERY SYSTEM
Our Ocular Drug Delivery Group at the University of California, Irvine is working on the Micromachined Drug Delivery System (Figure 13). The goal of this pro-ject is to use micromachining technology to build a passive, programmable, pulsatile drug delivery system, with many pulses, yet small enough for ocular implants. There are null zones in our polymer and then zones loaded with drug, this enables drug levels to cycle up and down in a preprogrammed fashion, also allowing native expression of endogenous cytokines and growth factors.

CONCLUSION
There are two fundamental approaches and philosophies with regard to ocular drug delivery. One is the longer-acting reservoir implants that offer good long-term control of disease but with potential for drug or suppressive side effects from chronic suppression of growth factors. The other option provides for shorter-acting biodegradable inserts that potentially expose the eye to less drug or suppressive side effects but may not control disease as well. Depending on what drug you pick for this method, however, we might get such good suppression of the cytokine cascade that the eyes can then have a prolonged drug holiday with a slow mounting of the cytokine cascade before clinical disease is manifested—we would treat at that point.

The hope is that all of these technologies will see the light of day and that different approaches will have preferential usage in different diseases or with different drugs.

Baruch D. Kuppermann, MD, PhD, is Associate Professor of Ophthalmology and Biomedical Engineering, at the University of California, Irvine. Dr. Kuppermann disclosed that he is a consultant for Allergan, Bausch & Lomb, Neurotech, Acuity Pharmaceuticals, Genentech, Genaera Corporation, OSI/Eyetech, Alcon, Regeneron, TargeGen, and Ista. He may be reached at bdkupper@uci.edu or 949-824-6256.

Kuppermann BD. Implant delivery of corticosteroids and other pharmacologic agents. 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.