Within this review, we would like to discuss our rationale for introducing intravitreal dexamethasone as an adjunct in the treatment of neovascular age-related macular degeneration (AMD). In a step-wise fashion, we first used this agent in combination with verteporfin photodynamic therapy (PDT) (Visudyne; Novartis, Hanover, NJ) alone to determine its safety profile and whether there was any added efficacy compared with PDT alone. We subsequently developed a triple-therapy protocol with the further addition of an anti-vascular endothelial growth factor (VEGF) agent. (Albert J. Augustin, MD, should be duly credited with developing a triple-therapy protocol that also incorporates a monoport vitrectomy.) We will also provide interm results of the PDEX trial, a prospective, randomized trial evaluating the relative efficacy of triple therapy versus a monthly ranibizumab (Lucentis; Genentech; South San Francisco) regimen.
BACKGROUND
AMD is a comprehensive term that encompasses a broad spectrum of macro- and microscopic alterations to the retina and retinal pigment epithelium (RPE) in patients aged >50 years. Subclassification into neovascular and nonneovascular does little to highlight the diverse, multidimensional nature of the processes involved in the development of choroidal neovascular (CNV) membranes. Indeed, biochemical and histological analyses of extracted membranes, as well as numerous animal models, clearly demonstrate that broad-spectrum inflammation is present in the milieu surrounding these membranes. These factors include, but are not limited to, increased production of VEGF, intracellular adhesion molecule, matrix metalloproteinase, monocyte colonizing protein, interleukin-8, and free radical production.
The heterogeneity of the disease process is also evident in vague angiographic classifications (ie, classic, occult, retinal angiomatous proliferation, polypoidal). Diverse manifestations of leakage are seen using ocular coherence tomography (OCT), which can demonstrate fluid under the RPE as well as under and within all layers of the neurosensory retina. It is unclear whether these represent distinct underlying disease processes or developmental stages. With the plethora of available data, it appears then that neovascular AMD is a disease process that is as diverse in its pathophysiology as it is in its clinical presentation.
The treatment for neovascular AMD has evolved over the last 30 years from nonselective destructive laser treatment to today's targeted photoactivated dyes (PDT) and injectable antiangiogenic compounds (ie, pegaptanib [Macugen; OSI/Eyetech, New York, NY], ranibizumab, bevacizumab [Avastin; Genentech]).1-6 A significant improvement was provided by Spaide et al with the addition of intravitreal triamcinolone acetonide, a broad-spectrum antiinflammatory agent, to the PDT treatment protocol.7 Unfortunately, the relative benefits of this treatment strategy had to be weighed against the significant burden of ocular side effects including glaucoma, cataract formation, sterile and infectious inflammation, and visual obscuration from the triamcinolone particles.
The emergence of targeted anti-VEGF agents brought a new era to the management of neovascular AMD and provided hope that improved visual outcomes could be achieved with reduced ocular risks.8,9 While pegaptanib was the first to be US Food and Drug Administration approved, the adoption of bevacizumab and subsequent emergence of ranibizumab provided new gold-standards with respect to visual outcomes. Data from the Minimally Classic/Occult Trial of the Anti-VEGF Antibody Ranibizumab in the Treatment of Neovascular AMD (MARINA) and Anti-VEGF Antibody for the Treatment of Predominantly Classic Choroidal Neovascularization in AMD (ANCHOR) trials suggested that monthly anti-VEGF treatments could provide sustained and significant visual improvement. These extremely favorable results had to be tempered with the fact that they were achieved with rigorous adherence to monthly injections over a 24-month period. Attempts to reduce the treatment frequency or prolong the treatment-free intervals resulted in suboptimal visual outcomes (eg, PIER, PRONTO, VISION).
Although the safety profile as suggested in the ANCHOR and MARINA studies was favorable over the 2-year duration of the study, it is unclear if there will be longer-term consequences. For example, endogenous VEGF appears important in normal retinal vascular homeostasis and injury recovery.
PROTOCOL DEVELOPMENT
In an effort to attain favorable visual outcomes while reducing the overall treatment frequency, we first modified Spaide's protocol and developed a treatment paradigm that incorporated the angioocclusive activity of PDT with the broad-spectrum antiinflammatory activity of a short-acting potent steroidal agent, dexamethasone phosphate. We chose dexamethasone over triamcinolone due to its enhanced antiinflammatory activity, transparent medium, its immediate bioavailability, and shorter duration of action. Intravitreal dexamethasone has been used extensively in the treatment of endophthalmitis, and also as an adjunct during vitrectomy in diabetic patients in doses as high as 0.8 mg without any significant ocular toxicity. It has been shown to inhibit the vascular leakage induced by VEGF in animal models of inflammation and diabetes. It is also effective in preventing the development of laser-induced choroidal neovascular membranes, as well as in preventing photoreceptor apoptosis in various animal models. Other demonstrated effects are its ability to suppress nitric oxide production, and suppress tumor necrosis factor expression.10-17
In our practice, dexamethasone was initially used for neovascular AMD patients in April, 2005, at a time when bevacizumab was not yet generally available for intravitreal use. We performed an independent review board (IRB)-approved review of 105 consecutive charts, from April 2005 through June 2006, of patients receiving dexamethasone in combination with PDT, with and without anti-VEGF supplement. This preliminary analysis was limited by its retrospective nature as well as the lack of exclusionary criteria based on lesion size, previous treatments, and visual acuity.
Reliance on Snellen visual acuities also hampered a more formal visual acuity outcome analysis. As intravitreal dexamethasone had not been previously used in the treatment of neovascular AMD, it was intended primarily to identify potential safety concerns. Additionally, in preparing for an anticipated prospective, randomized, masked study, we wished to have preliminary guidance as to the need and optimal timing of anti-VEGF administration.
RETROSPECTIVE RESULTS
A total of 105 patients were included in the analysis. Fifty-three of these were treatment na•ve (50%). Fifty-five patients (52%) were treated with double therapy and 50 (48%) received combination therapy (all treatments within a 3-month window). Thirty-seven (35%) of the latter group were classified as triple therapy and received all treatments within a 2-week window (Figures 1 and 2). The mean follow-up period for the total 105 patients was 290 days (longest was 564 days). The mean number of retreatments for the total group was 2.35, and 26% were not retreated beyond baseline. The double-therapy group had a mean number of retreatments of 3.15, while 18% were not retreated beyond baseline. In contrast, the combination-therapy group had a mean retreatment number of 1.48 and 34% were not retreated beyond baseline.
The visual acuity outcomes at month 6 and 8 showed an increase of 6.0 and 2.4 letters for all patients, respectively. An increase of 5.7 and 0.0 letters, respectively, was noted in patients receiving double therapy, while there was an increase of 6.6 and 10.5 letters in patients receiving some form of triple therapy. When this was further analyzed by patients receiving triple therapy within a 2-week window, we noted an increase of 6.9 and 16.9 letters at 6 and 8 months, respectively.
The visual acuity outcome at the last follow-up demonstrated an increase of 3.0 letters for all patients (n=105), an increase of 4.9 letters for all treatment-na•ve patients (n=53), and an increase of 0.4 letters for previously treated patients (n=49). A decrease of 2.1 letters was seen in patients with double therapy (n=55) while an increase of 8.7 letters was seen in patients with combination therapy (n=50). Finally, an increase of 9.5 letters was noted in patients who had received triple therapy all within a 14-day window (n=37).
Most importantly, no patients had intraocular pressure (IOP) of greater than 30 mm Hg. IOP rose transiently above 25 mm Hg in 8%, but of the two patients that required IOP lowering medications, neither was on therapy for more than 2 weeks. Now with more than a 2-year follow-up in these intravitreal dexamethasone patients, we have not noted any late developing IOP rise that was often seen in intravitreal triamcinolone acetonide patients. No other serious ocular adverse effects (ie, acute vision loss, uveitis, endophthalmitis, retinal tear or detachment) were noted.
Though interpretation of this data was limited as stated above, the preliminary evaluation suggested that intravitreal dexamethasone was safe; it added a short-term benefit compared with PDT alone, but its effect was not maintained without the addition of an anti-VEGF agent. Furthermore, it appeared that the greatest benefit was obtained when the anti-VEGF agent was added within a 2-week window of the PDT/intravitreal dexamethasone treatment.
STUDY FORMULATION
The PDEX protocol is an IRB-approved, multicentered, prospective, randomized, noninferiority, masked trial comparing the relative merits of group one: same-day triple therapy using reduced fluence Visudyne PDT, intravitreal dexamethasone, and ranibizumab followed by as needed treatment, versus group two (Figure 3): monthly ranibizumab monotherapy for 12 months. The strict inclusion criteria includes only na•ve patients with visions between 20/32 and 20/320. Though all lesion types are included, any lesions with pigment epithelial detachments (PEDs) larger than 50% of the total lesion are excluded in order to minimize the effect of RPE tears on treatment outcomes. Retreatments in group one are given if there is any evidence of subretinal hemorrhage, subretinal, intraretinal or sub-RPE fluid by clinical exam, OCT, or angiogram. Retreatment is initially with ranibizumab, but after three such treatments, the patient is again eligible for triple therapy. Early Treatment Diabetic Retinopathy Study (ETDRS) refractions are performed on the initial visit, and then at 3, 6, 9, and 12 months (Figure 4). OCT scans are performed monthly, and angiograms are performed every 3 months. Preliminary analysis performed at the 3-, 6-, 9-, and 12-month time-points show noninferiority between the two groups (Figure 5). Both groups are demonstrating significant ETDRS letter gain. Group one, however, requires substantially reduced treatments as compared with patients in group two (Figure 6).
This data is in the preliminary evaluation period and may change substantially as more data points are completed. A detailed statistical analysis will be performed when all 60 patients reach the 6-month follow-up and then again when they reach the 12-month follow-up. Thus far, there has been only one ocular serious adverse event, which was noted in a group two patient who developed ranibizumab-induced uveitis with each of her first two doses.
CONCLUSIONS
Our preliminary work suggests that intravitreal dexamethasone may be a safe adjunctive treatment when used in combination with PDT and anti-VEGF agents. Our retrospective data suggests that intravitreal dexamethasone may add to the efficacy of PDT, but that the effect is transient without the further addition of an anti-VEGF agent. Optimal timing of anti-VEGF treatment appears to be within a 2-week window of PDT treatment. Furthermore, very early results from the first prospective randomized trial comparing a regimen of monotherapy to triple therapy suggests at least noninferiority between the two groups with reduced treatment frequency in the triple-therapy group. Obviously this data is from an interim analysis, and any formal conclusions will require study completion.
The use of combination therapy in certainly not unique to ophthalmology, but is rather ubiquitous in the treatment of chronic medical conditions. The combined use of various agents, each acting with its own unique mechanism of action, may allow one to obtain maximal treatment efficacy, while minimizing the toxicity of any one agent. In the particular case of triple therapy, a combination of lower-fluence PDT, a shorter-acting steroid, and reduced frequency of anti-VEGF treatments may allow the physician to bypass the potential toxicities of higher doses of any one of these agents, while still achieving comparable visual results.
In theory, dexamethasone not only ameliorates the inflammation up-regulated by PDT treatment, but also reduces many of the broad-spectrum inflammatory mediators that are already present within the active neovascular complex. The ability of intravitreal dexamethasone to even transiently diminish these factors may reduce the overall burden required from both the PDT and anti-VEGF agent in order to achieve a quiescent lesion. It is also possible that the antiapoptotic effect of dexamethasone seen in animal models may protect the photoreceptors while the active lesion undergoes regression. Furthermore, dexamethasone's own ability to prevent CNV, as seen in animal models, may also impact the neovascular complex directly.
Albert J. Augustin, MD, has played a seminal role in the development of triple-combination therapy. His clinical work and thoughtful publication serve as one of the foundations of the subsequent proliferation of the triple-therapy protocol.18
There have been other authors who have now presented their preliminary work with combination therapy, both double and triple, the former being without the addition of dexamethasone. Investigators such as Henry Hudson, Mark Hughes, Gurav Shah, Joel Pearlman, and others, are involved in ongoing multicenter randomized trials and studies and are evaluating variations in fluence, timing of anti-VEGF addition, loading doses, as well as the relative merits of mono, double, and triple therapies (see sidebar, Ongoing Randomized Trials). These studies, along with the eventual analysis of the full 1-year PDEX results, will hopefully assist the retinal community in defining the optimal treatment protocols for our patients.
CHALLENGES
Challenges remain, such as determining the accurate PDT dosimetry in both na•ve and previously treated patients receiving either double or triple therapy. The optimal timing and dosage of VEGF and/or steroid administration remains to be deciphered from the proliferating plethora of case studies. In addition, possible structural changes to the choroicapillaris following chronic VEGF suppression may further alter the already disrupted retinal milieu, necessitating further refinements in the timing and delivery of combination therapy.
Given the multifactorial nature of neovascular AMD, it is likely that the optimal treatment protocol will need to be individualized for each particular patient. Various biological factors including VEGF and other inflammatory markers may be differentially active in the course of disease progression. Tailoring treatment protocol and duration with this in mind may ultimately provide both the patient and physician with a greater level of satisfaction. Though these are all difficult questions to answer, ultimately our patients will enjoy the fruits of these investigations.
Subhransu Ray, MD, PhD, is a partner at Bay Area Retina Associates, in Walnut Creek, CA. He states that he is a recipient of independent research grants from QLTI and Genentech. He may be reached at lsray01@yahoo.com.
Dr. Ray would like to acknowledge that the data, thoughts, and interpretations expressed within this article are entirely the result of the collaborative and equal efforts of all the partners at Bay Area Retina Associates including Allen Z. Verne, MD; Craig J. Leong, MD; Stewart A. Daniels, MD; and T. Daniel Ting, MD, PhD.
_1773249222.png?auto=compress,format&w=75)
