Safety and Efficacy Assessment of Chimeric Ribozyme to Proliferating Cell Nuclear Antigen to Prevent Recurrence of Proliferative Vitreoretinopathy FREE

Proliferative vitreoretinopathy (PVR) is the principal cause of failed retinal reattachment surgery, occurring in 5% to 20% of all patients with retinal detachment (RD).^{1 - 2} It can occur de novo in select cases of chronic RD, vitreous hemorrhage, uveitis, and other conditions. It is characterized by the migration and proliferation of cells that cause membrane formation on or beneath the retina or in the vitreous cavity.³ The cells in the proliferative membranes include retinal pigment epithelial cells, glial cells, fibroblasts, and inflammatory cells.^{3 - 11} The proliferative membranes progress and contract, which eventually causes recurrent or persistent detachment of the retina. Retinal redetachment due to PVR frequently results in permanent and often profound visual impairment.

Current therapies for the treatment of PVR involve vitreoretinal surgical techniques to mechanically peel scar tissue and flatten the retina during pars plana vitrectomy. Surgical adjuncts and instrumentation that aid surgical reattachment include heavy perfluorocarbon liquid and gas, silicone oil (SO), and panoramic visualization. Although modern surgical techniques have improved the reattachment rate up to 90% in patients with RD with PVR, most patients have substantial visual loss.^{12 - 13} Therefore, prophylactic intervention in patients who have undergone retinal reattachment surgery and those at risk for the development of PVR is of utmost importance.

Previous experiments have demonstrated the efficacy of a novel intravitreal drug therapy, VIT100, developed by Immusol, Inc (San Diego, California), to inhibit PVR development in a rabbit dispase model of PVR.¹⁴ VIT100 uses a ribozyme, a small RNA molecule with endoribonuclease activity that hybridizes to complementary sequences of a particular target messenger RNA transcript through Watson-Crick base pairing, to target a cell-cycle–controlling gene to inhibit cell division. Under appropriate conditions, these ribozymes catalytically cleave a sequence-specific target. The target for VIT100 is a sequence in proliferating cell nuclear antigen, a critical cell-cycle division factor in all cell types that is identical in rabbits and humans.^{15 - 16} Delivery of this therapeutic ribozyme to the vitreous cavity and retina in the rabbit transiently inhibits proliferation of a broad range of cells.

This article describes the results of a randomized, double-masked, placebo-controlled, multicenter, phase I and II clinical trial, sponsored by Immusol, Inc, to test the safety and efficacy of a ribozyme to proliferating cell nuclear antigen (VIT100) to prevent the recurrence of PVR in patients with grade C or worse PVR who are undergoing pars plana vitrectomy for repair of RD.

A prospective, randomized, multicenter, double-masked, placebo-controlled clinical trial was undertaken to evaluate the safety and efficacy of VIT100 as an intravitreal injection to prevent recurrence of PVR. In all, 15 centers specializing in vitreoretinal surgery in the United States participated in the trial. The study protocol was reviewed and approved by the institutional review board of each of the 15 centers. Study investigators had been previously trained and certified for participation in clinical trials (Good Clinical Practices). Patient selection, recruitment, and enrollment were initiated and performed by the study investigator of the individual center.

Inclusion criteria included the presence of primary or recurrent RD with grade C or worse PVR (Retina Society classification), defined as the presence of at least 1 fixed retinal fold in at least 1 quadrant, in a patient at least 18 years of age with visual acuity of light perception or better. Notable exclusion criteria included the presence of severe nonproliferative or proliferative diabetic retinopathy or other preexisting ocular vasoproliferative diseases, a history of intraocular inflammatory disease, and hereditary vitreoretinopathies.

Each patient underwent a baseline evaluation that included a detailed medical and ophthalmic history as well as a complete ophthalmic examination. A certified refractionist performed refraction and best-corrected visual acuity (BCVA) measurements using a retroilluminated, modified Ferris-Bailey Early Treatment Diabetic Retinopathy Study (ETDRS) chart. If during visual acuity testing the patient was unable to read any ETDRS chart letters at 1 m, then the patient's ability to count fingers, detect hand motion, or perceive light was evaluated with a standardized low-vision protocol and outcomes were expressed in Snellen equivalents. The investigator performed a detailed slitlamp and funduscopic examination. Intraocular pressure was measured by applanation tonometry or tonopen. The location and extent of RD were carefully detailed and recorded. Blood samples were collected and sent for laboratory evaluation of serum urea nitrogen and creatinine levels.

All of the patients enrolled in this trial met eligibility criteria. Study participation was initiated after informed consent was obtained. Promedica International, Costa Mesa, California, the study monitor, was responsible for patient randomization and medication assignments. Patients were randomized in a 1:1:1 ratio to receive high-dose medication (0.75 mg), low-dose medication (0.15 mg), or placebo (saline solution).

All of the treating physicians and patients remained masked throughout the trial. Treating physicians received the assigned study drug or placebo at the time of surgery from the study coordinator in a masked fashion. High-dose, low-dose, and saline solutions each appeared as identical, clear, colorless solutions within the syringe. Postoperative protocol refraction and BCVA measurements were performed by visual acuity examiners masked to prior measurements and randomization assignment.

All of the patients underwent retinal reattachment surgery with pars plana vitrectomy. Additional intraoperative procedures including scleral buckle (SB) placement or revision, pars plana lensectomy or limbal cataract extraction, intraocular lens implantation or removal, temporary keratoprosthesis and penetrating keratoplasty, retinotomy, and/or gas or SO tamponade were performed at the discretion of the operating surgeon and required the assistance of an anterior segment specialist in certain cases.

The surgical management of the detachment and associated pathological findings was left to the discretion of the treating ophthalmologist. Intraoperative categorization of the degree and extent of proliferation and reattachment of the retina was of critical importance. Use of panoramic visualization and perfluorocarbon liquid was not mandatory. The surgical technique, such as bimanual manipulation of the proliferative tissue or use of retinotomy or retinectomy, was not standardized. The treating physician was urged to attain and document intraoperative retinal reattachment as full or partial under perfluorocarbon liquid with necessary documentation of areas of persistence of traction or proliferation and detachment.

At the conclusion of the procedure, the investigational drug or placebo was injected into the vitreous cavity via a 30-gauge needle, 3.0 mm to 3.5 mm from the limbus. In the gas-filled eye, the injection was performed after air and gas exchange to a gas- or water-tight eye following closure of all of the sclerotomy or anterior segment wounds. In the SO-filled eye, the injection of medication was performed in the same manner. Following either air and SO exchange or perfluorocarbon liquid and SO exchange, the drug or placebo was injected into the vitreous cavity following adequate closure of the eye.

Postoperative study visits were as follows: postoperative day 1 and weeks 1, 2, 4, 8, 12, and 24. Patients were also asked to return when they or the treating physician suspected a severe adverse ocular event. During these study visits, updated ocular and medical symptoms were recorded and patients received a protocol refraction and ETDRS BCVA measurement, ophthalmoscopic examination, and intraocular pressure measurement. Gross visual acuity testing (light perception, detecting hand motion, and counting fingers) was used at postoperative day 1 and week 1 unless it was possible to conduct standard refractive testing. Tonopen tonometry was sufficient for pressure measurements on study visits. Applanation tonometry was performed to verify all of the measurements greater than 30 mm Hg. At weeks 4 and 24, blood was obtained from the patient and analyzed for serum urea nitrogen and creatinine levels.

Successful postoperative retinal reattachment was defined as reattachment not requiring reoperation. In situations in which there was localized traction RD secondary to limited reproliferation, the retina was deemed reattached if the treating physician judged the situation to be stable and not requiring reoperation.

Redetachment occurred in several settings. Reproliferation and recurrent membrane formation causing traction RD judged by the treating physician as being unstable and requiring reoperation were deemed retinal redetachment due to recurrent PVR. Rhegmatogenous RD due to reproliferation with induced traction and resultant retinal break requiring retinal reoperation was deemed RD due to PVR. Rhegmatogenous RD due to either new or unrecognized (intraoperative) retinal break without recurrence of PVR was deemed retinal redetachment.

Surface membrane reproliferation occurred without retinal redetachment. Postoperative recurrent membrane formation (subretinal and preretinal) was noted and monitored during the postoperative period.

Visual acuity was analyzed in a nonparametric manner with distribution of BCVA Snellen equivalents between treatment groups compared in categorical fashion at baseline and again at 24 weeks. Categorical groups included no light perception, light perception, detecting hand motion, counting fingers, worse than 20/200 to 20/400 or better, worse than 20/100 to 20/200 or better, worse than 20/80 to 20/100 or better, worse than 20/40 to 20/80 or better, and 20/40 or better.

Intraocular pressure was monitored during the postoperative period. Ocular hypotony was defined as an applanation intraocular pressure of 5 mm Hg or less at 24 weeks.

Any adverse experience or event (AE) was graded as to severity (mild, moderate, severe, or life threatening) and relationship to study therapy (not related, possibly related, probably related, or definitely related). Investigators were required to report all of the AEs occurring during the course of the study to Immusol, Inc, and to the respective center's institutional review board within 24 hours.

The primary efficacy variable was the failure rate of retinal reattachment surgery due to PVR. The secondary efficacy variable was the failure rate of retinal reattachment surgery due to any other cause.

The sample size was determined as follows: assuming a recurrent RD rate of approximately 35% for patients in the control group and a recurrent RD rate of approximately 10% if intervention proved beneficial in the treated group, a total of 43 patients per group was necessary to display a significant response to the drug. To allow for a potential dropout rate of 15%, the anticipated enrollment was a total of 50 patients per group.

Numerical computations were performed using a spreadsheet package (Excel 2003; Microsoft Corp, Redmond, Washington). Statistical comparison of categorical findings was performed using the χ² test, Fisher exact test, t test, and Kruskal-Wallis test.

For statistical comparison of the primary efficacy variable, the 2 treated groups were separately compared with the placebo group using the Cochran-Mantel-Haenszel test. Further analyses included the dose-response function using modified ridit scores for dosage levels.

A total of 175 eyes from 175 patients were randomized to receive high-dose VIT100 (60 eyes), low-dose VIT100 (58 eyes), or placebo (57 eyes). Patient demographics and baseline ophthalmic findings were statistically balanced aside from a trend toward white men in all of the treatment groups (Table 1). The number of primary and recurrent cases of RD enrolled in the study as well as the extent and location of PVR were similar between the groups. Overall, 91% of enrolled patients had posterior PVR and 63% had anterior PVR, with a mean of 5.0 and 3.8 clock hours affected, respectively. Ninety percent of patients had a macula-off RD. Baseline distribution of BCVA Snellen equivalents was similar between treatment groups (P = .66).

A total of 163 patients (93%) received the study drug or placebo. Nine of 163 patients (6%) were lost to follow-up or withdrew from the study. One of these patients (low-dose group) reached the primary end point of RD due to PVR. One hundred fifty-four patients (88%) reached the study end point of 24 weeks' follow-up. Flow of patient randomization to completion of the trial is summarized in the Figure.

Forty-one eyes (27%) developed recurrent RD due to PVR by 24 weeks, including 18 eyes (33%) in the group receiving 0.75 mg, 13 eyes (24%) in the group receiving 0.15 mg, and 10 eyes (22%) in the placebo group (P = .37) (Table 2). Similarly, 59 eyes (38%) reached the secondary end point of recurrent RD due to any reason by 24 weeks, with 21 eyes (39%) in the high-dose group, 22 eyes (41%) in the low-dose group, and 16 eyes (35%) in the placebo group (P = .83) (Table 2). Of the 95 eyes that did not develop recurrent RD from any cause by 24 weeks, 54 (57%) developed reproliferation without recurrent RD, with no statistically significant difference between the treatment groups (P = .62).

Results of AEs are based on the 163 patients who received the study drug or placebo regardless of study completion status. No deaths occurred during the study. A total of 145 serious AEs were reported in 95 patients. There was no statistically significant difference between treatment groups regarding these events and no serious AEs were considered to be directly attributable to the study drug at either dose. A listing of AEs that occurred in at least 5 patients in any treatment group is presented in Table 3.

An analysis of visual acuity outcomes was performed at study conclusion. The categorical distributions of BCVA Snellen equivalent outcomes at 24 weeks were similar between treatment groups (P = .65).

While all of the study patients received a vitrectomy at the time of retinal reattachment surgery, 29 patients (19%) also underwent SB placement, including 10 (19%) in the high-dose group, 10 (19%) in the low-dose group, and 9 (20%) in the placebo group. One hundred seven patients underwent scleral buckling prior to study entry. In all, 136 of the 154 patients who completed the study had an SB placed. In aggregate, the retinal redetachment rate due to PVR in eyes that received an SB was 25% (n = 34), compared with 38% (n = 7) for patients without an SB (P = .33). Seventeen additional patients in the SB group and 1 in the non-SB group developed recurrent RD due to a cause other than PVR. In all, 51 of 136 eyes in the SB group had detachment and 8 of the 18 eyes in the non-SB group had detachment (P = .75).

When anatomical outcomes for SB and non-SB groups were stratified by drug and placebo, there was no statistical difference across treatment groups.

Fifty-five eyes (36%) were phakic, 71 eyes (46%) were pseudophakic, and 28 eyes (18%) were aphakic at the time of study enrollment. Postoperatively, 18 eyes (12%) were phakic, 66 eyes (43%) were pseudophakic, and 70 eyes (45%) were aphakic. Thirty-seven patients (67%) with eyes that were initially phakic underwent lensectomy at the time of RD repair, and 9 patients (13%) with pseudophakic eyes underwent intraocular lens removal. In aggregate, 6 of 18 patients (33%) with phakic eyes postoperatively developed recurrent RD due to PVR. Thirty-five of the remaining 148 patients (24%) with nonphakic eyes developed recurrent RD due to PVR, including 16 patients (24%) with pseudophakic eyes and 19 patients (27%) with aphakic eyes. There was no difference in the redetachment rate from PVR (P = .73) or from any cause (P = .70) when postoperative lens status was analyzed.

When anatomical outcomes for postoperative lens status groups were stratified by drug and placebo, there was no statistical difference across treatment groups.

Seventy-three patients (47%) underwent retinotomy, including 30 (56%) in the high-dose group, 24 (44%) in the low-dose group, and 19 (41%) in the placebo group (P = .31). The 73 eyes that underwent eventual retinotomy had more extensive preoperative posterior (P = .04) and anterior (P = .003) PVR as measured in mean clock hours when compared with those eyes that did not undergo retinotomy. Presence of retinotomy did not significantly alter anatomical outcomes both in aggregate and when stratified by treatment group.

One hundred ten eyes (71%) received SO and 44 (29%) received gas tamponade as a postoperative extended retinal tamponade. In the SO group, 35 eyes (32%) developed recurrent RD from PVR and an additional 12 eyes (11%) developed recurrent RD from any cause (n = 47). Forty-two of the 44 eyes that received gas tamponade received perfluoropropane. Of these, 41 received a 14% concentration or greater. One of the 2 eyes that received sulfur hexafluoride had recurrent RD due to PVR, and no eyes that received perfluoropropane in a concentration less than 14% had recurrent RD. In eyes that received gas tamponade, 6 (13%) developed recurrent RD due to PVR and an additional 6 eyes (13%) developed recurrent RD from any cause (n = 12). Recurrent RD due to PVR was seen statistically more frequently in the group of eyes that received SO (P = .04). To rule out a drug-tamponade effect, analysis of variance did not reveal an interaction between tamponade and the drug. The retinal redetachment rate due to any cause was similar when comparing the gas and SO groups.

When anatomical outcomes for tamponade groups were stratified by drug and placebo, there was no statistical difference across treatment groups.

Eyes that received SO had more extensive posterior PVR (mean clock hours) when compared with eyes that received postoperative gas tamponade (5.3 vs 4.1 mean clock hours, respectively) (P = .05). Mean clock hours of anterior PVR were similar between the SO and gas groups.

A relaxing retinotomy was performed in 60 of the 110 patients (55%) who received SO and 13 of the 44 patients (30%) who received gas as a postoperative tamponade (P = .01). The extent of retinotomy measured in mean clock hours was 5.3 clock hours in the SO group and 3.6 clock hours in the gas group (P = .05).

Posterior PVR (number of eyes and mean clock hours of involvement) was significantly greater in eyes that eventually received SO (P = .04). Additionally, eyes that received SO and underwent retinotomy had a significantly greater redetachment rate due to PVR than eyes that received gas and underwent retinotomy (P = .01).

Perfluorocarbon liquid was used intraoperatively in 124 patients (81%). There was no significant difference in the retinal redetachment rate due to PVR (P = .49) or any cause (P = .99) between patients who received and did not receive intraoperative perfluorocarbon liquid. Furthermore, there was no significant difference in the redetachment rate for those who received perfluorocarbon liquid across drug and placebo treatment groups.

Hypotony was noted in 14 eyes at study conclusion, with no significant difference between treatment groups (P = .15). Intraocular pressure of 20 mm Hg or greater was noted in 17 eyes at study conclusion, with no significant difference between treatment groups (P = .83).

Final (24-week) intraocular pressure was evaluated comparing tamponade (P = .67), retinotomy status (P = .50), and retinotomy status stratified by tamponade (P = .42). No significant difference in final intraocular pressure was identified when these groups were compared (P = .57).

Proliferative vitreoretinopathy remains the single greatest cause of failure of retinal reattachment surgery. Current management is surgical and necessitates lengthy procedures in which preretinal and subretinal fibroproliferative membranes are dissected off the retinal surface to allow for retinal reattachment. At the conclusion of the procedure, either long-acting perfluorocarbon gas or SO is used as an extended tamponade. With recent innovations such as perfluorocarbon liquid and panoramic viewing, surgeons now obtain long-term retinal reattachment rates as high as 90%.¹³ Unfortunately, however, anatomical reattachment does not necessarily ensure visual function. Visual results are often severely compromised. Given current treatment options, prevention of PVR is an attractive goal. Previous attempts included targeting specific or nonspecific processes in PVR development such as the inhibition of cell types and cell division processes and the reduction of fibrin formation and inflammation.¹⁷ Cytotoxic agents, fluorouracil,^{18 - 19} duanorubicin,^{20 - 21} and radiation²² have exhibited marginal, often temporary benefits in clinical trials. Additionally, steroids and tissue plasminogen activator have been used in an attempt to decrease fibrin production and inflammation, with marginal or no benefit. Low-molecular-weight heparin in combination with fluorouracil has been previously shown to reduce postoperative PVR and the retinal reoperation rate.¹⁹

Overall in this study, 41 patients (27%) developed recurrent RD due to PVR and 59 patients (38%) developed recurrent RD due to any cause following surgical repair with 24 weeks' follow-up. This outcome is consistent with previously reported results from the literature, with single-operation surgical success rates varying from roughly 50% to 80%.^{19 ,23 - 27} More importantly, however, there was no significant difference between treatment arms for recurrence of RD due to PVR (primary efficacy end point) or due to any cause (secondary efficacy end point) at the conclusion of this study. Although previous experiments have established that VIT100 inhibits PVR development in a rabbit dispase model of PVR, our multicenter, double-masked, placebo-controlled, randomized clinical trial failed to show efficacy in human eyes. Our study's results reveal no benefit of VIT100 in limiting the development of recurrent RD due to PVR at either low or high dose in eyes with established PVR. Importantly, the dose-ranging groups and the sham group were similar with regard to the preoperative degree and extent of PVR, the number of vitreoretinal procedures performed prior to study entry, and various other preoperative parameters.

When used as a single-dose intravitreal injection at the conclusion of surgery, VIT100 appears to be safe. The most common serious AE was recurrent RD, an “event” that is really an outcome. Factors that may potentially complicate surgical outcomes in retinal reattachment surgery were further evaluated. Postoperative lens and SB status did not appear to alter ultimate anatomical surgical outcomes when stratifying for the drug and dosage or assessing outcomes independent of the drug. In addition, it does not appear that placement of relaxing retinotomy or use of perfluorocarbon liquid as a temporary intraoperative tamponade altered postoperative anatomical outcomes. It is interesting to note that eyes that underwent retinotomy had significantly more extensive anterior and posterior PVR. It is likely that, in general, more extensive disease necessitated relaxing retinotomy for intraoperative reattachment, consistent with previous reports.¹³

The choice of postoperative tamponade was left to the discretion of the operating surgeon. In all, 71% of the eyes received SO and 29% received gas as a postoperative tamponade, with the overwhelming majority (93%) receiving at least 14% perfluoropropane. Eyes that received SO had more extensive baseline posterior PVR and received a relaxing retinotomy more frequently and, when performed, more extensively. Regardless of retinotomy status, eyes that received SO had recurrent RD from PVR significantly more frequently. One explanation for these findings may be perisilicone surface proliferation and ineffective inferior retinal tamponade by SO. Additionally, more extensive baseline proliferation probably affects ultimate anatomical outcomes.

We feel that the major limitation of this study was the relatively large number of surgeons and sites that enrolled patients with established PVR. Established PVR is a surgical disease with a wide variation in surgeon and institutional outcomes as well as in intraoperative surgical options and approaches. As such, a multicenter surgical trial is fraught with uncontrollable variables and a single-surgeon or single-center surgical outcome study might be preferable. In addition, in the setting of established PVR, the surgeon often removes the surface fibrous proliferation that is accessible and identifiable. Ill-defined and immature but nonetheless established surface membranes may be left in place owing to a lack of recognition and/or difficulty removing them. These membranes often mature postoperatively, becoming more distinct and problematic, sometimes leading to recurrent RD due to PVR. This development may occur regardless of ribozyme inhibition of proliferating cell nuclear antigen. Primarily for this reason, it would be preferable to assess the capacity of a drug such as VIT100 to aid in the prophylaxis or prevention of PVR. The Immusol/VIT100 Study Group feels that the appropriate study would be a prevention study in which eyes at high risk for the development of PVR, such as primary RD with coexisting vitreous hemorrhage, traumatic RD, and giant retinal tear, are treated with pharmacologic therapy at the time of primary retinal reattachment.

Visual analysis in this subset of patients with RD is fraught with difficulties, and interpretation of results should be made with caution. Final visual acuities in most patients are quite low, and many patients have retained SO, some as a final tamponade. There are no data suggesting that the drug impacted vision in either a favorable or unfavorable way.

In summary, this industry-sponsored, multicenter, randomized, placebo-controlled trial compared the safety and efficacy of a ribozyme to proliferating cell nuclear antigen (VIT100) with placebo to prevent the recurrence of PVR in patients with grade C or worse PVR who underwent pars plana vitrectomy for repair of RD. Anatomical results in the high-dose and low-dose VIT100 groups were statistically similar to those in the placebo group. VIT100, therefore, failed to meet the primary efficacy (recurrent RD due to PVR) and secondary efficacy (recurrent RD due to any cause) end points of this study. VIT100 had a similar safety profile when compared with balanced salt solution (placebo).

Correspondence: Naresh Mandava, MD, Rocky Mountain Lions Eye Institute, School of Medicine, University of Colorado, 1675 N Ursula St, Aurora, CO 80045 (naresh.mandava@uchsc.edu).

Submitted for Publication: December 6, 2006; final revision received April 24, 2007; accepted May 11, 2007.

Financial Disclosure: Dr Mandava has received research support from Immusol, Inc, San Diego, California.

Funding/Support: This work was supported by Immusol, Inc, an unrestricted research grant from Research to Prevent Blindness, and a research grant from the Eye Surgery Fund.

Additional Contributions: Shing Lee, ScM, provided help with the statistical analysis.

This article was corrected for missing data on 11/7/2007.

Gary Abrams, MD (principal investigator), Dean Elliot, MD, Robert N. Frank, MD, Raymond Iezzi, MD, James Puklin, MD, Kresge Institute, Wayne State University, Detroit, Michigan; K. V. Chalam, MD (principal investigator), Selena Lin, MD, Department of Ophthalmology, College of Medicine, University of Florida, Jacksonville; Pravin Dugel, MD (principal investigator), Jack O. Sipperley, MD, Scott R. Sneed, MD, Donald W. Park, MD, P. Kumar Rao, MD, Retinal Consultants of Arizona, Phoenix; Ronald Gentile, MD (principal investigator), Richard Rosen, MD, Daniel Will, MD, Estuardo-Alfonso Ponce, MD, Department of Ophthalmology, New York Eye and Ear Infirmary, New York; Julia A. Haller, MD (principal investigator), Peter Campochiaro, MD, James T. Handa, MD, Ingrid Zimmer-Galler, MD, Quan D. Nguyen, MD, MSc, Jennifer Sung, MD, The Wilmer Eye Institute, Johns Hopkins Hospital, Baltimore, Maryland; Tarek Hassan, MD (principal investigator), William Beaumont Hospital Research Institute, Royal Oak, Michigan; Baruch D. Kuppermann, MD (principal investigator), Santosh Patel, MD, Ricardo Carvalho, MD, Gilberto Resende, MD, David Kim, MD, Heikki Kostamaa, MD, Faisal Jehan, MD, Department of Ophthalmology, University of California Irvine Medical Center, Orange; Michael Lambert, MD (principal investigator), Joseph Khawly, MD, Arthur Willis, MD, Vincent Vann, MD, Retina and Vitreous of Texas, Houston; Raj K. Maturi, MD (principal investigator), Donald L. Wilson, MD, John Minturn, MD, Nicholas Hrisomalos, MD, Thomas Ciulla, MD, Midwest Eye Institute, Indianapolis, Indiana; H. Richard McDonald, MD (principal investigator), Robert N. Johnson, MD, Everett Ai, MD, J. Michael Jumper, MD, Sam Yang, MD, West Coast Retina Medical Group, San Francisco, California; Kirk Packo, MD (principal investigator), Pauline T. Merrill, MD, Jack A. Cohen, MD, Mathew W. MacCumber, MD, PhD, Illinois Retina Associates, Rush Presbyterian/St Luke Medical Center, Chicago; Michael B. Rivers, MD (principal investigator), Reginald J. Sanders, MD, Richard A. Garfinkel, MD, The Retina Group of Washington, Fairfax, Virginia; Patrick Rubsamen, MD (principal investigator), Lawrence Halperin, MD, Retina Vitreous Consultants, Fort Lauderdale, Florida; William M. Schiff, MD (principal investigator), Stanley Chang, MD, Gaetano R. Barile, MD, Lucian V. Del Priore, MD, PhD, R. Theodore Smith, MD, PhD, Department of Ophthalmology, Edward S. Harkness Eye Institute, Columbia University, New York, New York; Jay Stallman, MD (principal investigator), Michael S. Jacobson, MD, Scott I. Lampert, MD, Mark J. Rivellese, MD, Haris Amin, MD, Georgia Retina PC, Decatur.