Macula 2011: Highlights From the Conference

Wills Eye's 50-year-old Retina Service hosts.

Andrew E. Mathis, PhD, Medical Editor

The Macula 2011 conference was held in Philadelphia in late winter, offering retinal physicians presentations on a number of conditions and medical and surgical techniques for treating retinal disease, with Wills Eye Hospital hosting the event and celebrating the 50-year anniversary of its Retina Service. Continuing on from our March 2011 article of highlights, here we recap several additional presentations from the conference.

CRAIG GREVEN ON TUBULIN INHIBITION

There was only one session presenting preclinical data, and that presentation was made by Craig M. Greven, MD, FACS, who demonstrated tubulin inhibition in a primate model of choroidal neovascularization. Dr. Greven, who is professor and chair of the Department of Ophthalmology at Wake Forest University in Winston-Salem, NC, began his presentation by first describing tubulin as a protein dimer of a and b subunits that form microtubules.

Microtubules, in turn, are “critically important in intracellular functions, such as mitosis, transport and cytokinesis, and they help form the cell cytoskeleton,” Dr. Greven explained. Among the tubulin inhibitors currently in use are several compounds used in oncology, as well as combretastatin, which is also an oncology drug but is now being tested in AMD.

One problem with tubulin inhibitors, Dr. Greven said, is that there is significant toxicity with most, leaving only a narrow therapeutic window for their use due to the adverse effects they have on normal cells.

Dr. Greven turned then to discussing OC-10X (a quinazolanone), which is OcuCure Therapeutics's (Roanoke, VA) topical tubulin inhibitor currently undergoing testing in primates. Given its high lipid solubility and low molecular weight (~300 Da), OC-10X is being formulated as a topical eyedrop capable of reaching therapeutic levels in the retina, vitreous and choroid. By inhibiting tubulin, it prevents the proliferation of vascular endothelial cells, working independently of VEGF and other growth factors.

With his next slide, Dr. Greven demonstrated on a molecular level how OC-10X affects both cell division and cytoskeletal structure. Furthermore, he showed the audience how OC-10X inhibited angiogenesis in an in vivo chick chorioallantoic membrane assay (Figure 1). Dr. Greven also demonstrated the site on the cell where OC-10X binds, showing that it does this in a distinct place from where other tubulin inhibitors bind.

Figure 1. In an in vivo chick chorioallantoic membrane assay, OC-10X inhibited angiogenesis: A. Normal chick assay; B. OC-10X assay.

Dr. Greven turned to the animal trials themselves. First, he reviewed data from the Retina Society's 2008 annual meeting, at which OC-10X's ability to inhibit angiogenesis in a laser-induced rat model of CNV had been shown. Next, he explained the design of the primate trial, in which eight eyes from eight cynomolgus monkeys with laser-induced CNV were given an intravitreal injection of 500 µg OC-10X. (Dr. Greven noted the difficulty in giving eyedrops to primates by way of explanation.) The animals were sacrificed at four weeks and their CNV examined. There was a 43% reduction in CNV in the OC-10X–injected eyes compared to the control eyes (P=.025).

Next, anesthetized monkeys were administered OC-10X as eyedrops every 30 minutes for three doses, and sacrificed after four hours. Again, OC-10X demonstrated its ability to penetrate the eye to the level of the vitreous and retina (Figure 2). Dr. Greven noted here that OC-10X's peak retina concentration may ultimately prove to be even higher than the results of the primate trial indicated. There were only trace amounts of the drug in the blood plasma of the monkeys and no accumulation in the peripheral organs.

Figure 2. OC-10X has a strong ability to penetrate the eye to the level of the vitreous, choroid and retina.

The last set of data Dr. Greven presented dealt with longer-term toxicity following 16 months of chronic topical administration of OC-10X. Using several measurement tools (including intraocular pressure, ERG and histologic evaluation), no toxicity was observed.

While several stages of clinical trials for OC-10X clearly remain, Dr. Greven summarized his presentation by restating the positive results so far. Tubulin inhibition, he said in conclusion, may represent a new approach to the management of ocular neovascularization.

OMESH GUPTA ON TREAT AND EXTEND

One of the earliest presentations of the conference was given by Omesh Gupta, MD, MBA, assistant professor of clinical ophthalmology at Temple University Hospital, also in Philadelphia. Dr. Gupta talked about “treat-and-extend” (TAE) regimens for using ranibizumab to treat age-related macular degeneration.

Dr. Gupta began by reviewing the results from the MARINA and ANCHOR trials, both of which tested ranibizumab in fixed-dosage regimens. “However,” he said, “as we all know, nearly 92% of respondents to the ASRS PAT Survey reported treating neovascular AMD in an individualized fashion.” While everyone knows that individualized dosing offers greater convenience and safety, as well as a reduced cost for the patient, there are essentially only two approaches to individualized treatment currently being used in the United States: PRN (or “treat and observe”) and TAE. These approaches are used, according to the PAT Survey results, by 56% and 44% of physicians, respectively.

Next, Dr. Gupta detailed the results of three PRN ranibizumab trials from 2009, the results of which ranged from a gain of 7.3 ETDRS letters of visual acuity to only a 0.7-letter gain. The rates at which the individual trials demonstrated efficacy were correlated with the numbers of follow-up appointments and treatments given.

Dr. Gupta compared those results to data from a French trial comparing the one-year outcomes from PRN treatment to TAE outcomes. This study, whose authors published their findings in the January 2011 issue of Retina, found significantly better outcomes using the TAE method (+10.8 ±8.8 letters gained) vs the PRN methods (+2.3 ±17.4 letters gained). However, the TAE cohort was given significantly more injections.

Within this context, Dr. Gupta and his colleagues undertook a study of both ranibizumab and bevacizumab designed to evaluate the visual outcomes, number of injections, and direct medical cost of the TAE regimen. After reviewing the inclusion criteria, Dr. Gupta described the regimen itself, which consisted of monthly injections until the macula appeared dry on optical coherence tomography, followed by the treatment interval being extended by two weeks unless exudation recurred. If this happened, then the interval was instead reduced by two weeks.

Before discussing the overall results, however, Dr. Gupta showed a case study from the trial. At baseline, this patient had visual acuity of 20/400 and central retinal thickness of 516 µm. Four weeks after the first treatment, VA had improved to 20/100, and the macular thickness was 282 µm but still not dry. Therefore, the next follow-up was set for four weeks later, and at that appointment, the patient presented with a dry macula and VA of 20/40 (Figure 3).

Figure 3. From baseline (top), the patient improved to 282 µm retinal thickness (center). Four weeks after that (bottom), the macula was dry.

At that point, the patient underwent an extension of the time between follow-ups, so that the longest interval was 10 weeks (Figure 4). However, at the appointment following that 10-week interval, exudation recurred, so the interval was decreased to eight weeks. The patient rebounded quickly, however, and the interval was again extended to 10 weeks. At the last follow-up of the study, the patient's interval had reached 12 weeks (Figure 5).

Figure 4. The same patient then was seen after six weeks (top), eight weeks (center), and 10 weeks (bottom), but at the last of these three visits, exudation recurred.

Figure 5. In the same patient, the macula dried (top) when the interval was reduced to eight weeks, and was then in creased to 10 weeks (center) and 12 weeks (bottom).

Returning to the study at large, Dr. Gupta detailed the measurements used for the economic impact for TAE treatment, and then he turned to the results from all 166 eyes of 159 patients studied. The mean follow-up, with a minimum of six months, was 1.5 years. Baseline characteristics were similar to those seen in MARINA, ANCHOR and PrONTO.

Then, Dr. Gupta showed the mean visual acuity outcomes graphs for ranibizumab and bevacizumab. In direct comparison to the other three ranibizumab trials, the results from Dr. Gupta's study were comparable, with more than 30% of patients gaining three ETDRS lines or more of visual acuity with both drugs and only 4% of patients losing vision in both arms.

Compared to one another, patients treated with bevacizumab had a longer mean interval of days between treatments than patients treated with ranibizumab (90.4 vs 79.8, P=.019). And while the mean number of injections over the first year did not differ significantly, the number of injections in the second year was less in the bevacizumab group than in the ranibizumab group (5.6 vs 7.45, P=0.039).

Dr. Gupta concluded by recapping the trial outcomes, which indicated that significant vision improvement, reduction in number of visits and injections, and cost savings could all be achieved with TAE dosing.

JUSTIN GOTTLIEB ON AVASTIN IN AMD

Following Dr. Gupta's presentation was another study looking at the use of bevacizumab in wet AMD. Justin L. Gottlieb, MD, who is on the faculty of the Department of Ophthalmology and Visual Sciences at the University of Wisconsin–Madison, described his department's experiences using bevacizumab in this capacity.

The first part of Dr. Gottlieb's presentation provided a short history of neovascular AMD treatment, with specific emphasis on anti-VEGF treatments, from the first purification and cloning of VEGF-A in 1989 through the 2006 FDA approval of ranibizumab. He also reviewed the molecular format of the three approved anti-VEGF treatments for wet AMD (Figure 6). “You can see how far we've come,” Dr. Gottlieb said. He then very briefly reviewed the MARINA and ANCHOR data that helped get ranibizumab approved.

Figure 6. The molecular formats of pegaptanib and ranibizumab.

“So what about Avastin?” Dr. Gottlieb asked then. He showed slides from a case study of a 74-year-old woman with VA of 20/400 in the left eye treated with photodynamic therapy; bevacizumab was initiated at that point as a “salvage therapy.” Two months later, subretinal fluid has resolved and VA was 20/80.

At this point, Dr. Gottlieb began discussing the UW study directly. First, he addressed the issue of why the study was necessary, explaining that 85% of the patients being treated with anti-VEGF agents in the UW Retina Service receive bevacizumab. In addition, he said, faculty has a “gut feeling” that bevacizumab was working, although they only had case reports and no controlled studies.

The primary study question of the UW trial was whether there was visual acuity improvement, stabilization, or decrease in eyes with exudative AMD treated with bevacizumab monotherapy. As secondary questions, they considered whether there were reductions in maximal retinal thickness and subretinal fluid thickness.

Regarding its design, the study was a retrospective case series, and Dr. Gottlieb reviewed the inclusion criteria and other methodology. Then, he said, “This was kind of a messy study,” and proceeded to detail what he called “important caveats” regarding the patient cohort, eg, there was no treatment protocol, lesion criteria or follow-up protocol. “These are the patients we see every day in clinic and treat with Avastin,” Dr. Gottlieb explained.

Turning to the results, Dr. Gottlieb first detailed the visual acuity results, demonstrating an overall mean improvement from 20/160 to 20/125 at 48 weeks (Figure 7). The measurements of mean maximum retinal thickness and subretinal fluid followed, with Dr. Gottlieb explaining why both measurements were deemed necessary by showing an OCT image of how disparate these measurements can be. Both measurements decreased, as did the injection rates. Finally, Dr. Gottlieb showed the visual acuity improvement curves based on initial visual acuity (Figure 8).

Figure 7. Visual acuity improved from 20/160 to 20/125 at 48 weeks.

Figure 8. Depending on baseline visual acuity, mean final VA was either >20/160 or <20/40.

The conclusions the UW Retina Service was able to draw included that VA changes due to bevacizumab treatment were not age-dependent, nor was the reduction of maximum retinal thickness. And while the improvement of VA was not statistically significant for baseline VA better than 20/200, reduction of maximum retinal thickness was significant for baseline VA better than 20/200, implying that bevacizumab is effective in this group.

“I feel comfortable with continuing my use of Avastin for exudative AMD,” Dr. Gottlieb said in closing, “as a monotherapy in eyes without previous therapy.” Conceding that he did not know how bevacizumab compares to ranibizumab, he reminded the audience of the imminent CATT trial results.

HARRY FLYNN ON ENDOPHTHALMITIS

Harry W. Flynn, Jr., MD, professor and J. Donald M. Gass Distinguished Chair of Ophthalmology at the Bascom Palmer Eye Institute (BPEI) in Miami, discussed endophthalmitis following intravitreal injections in the anti-VEGF era by reviewing BPEI's own experience in this area.

Dr. Flynn began his presentation with a case study of a patient at BPEI who developed endophthalmitis four days after an injection of ranibizumab, leaving visual acuity of hand motions (Figure 9). The patient was treated with a combination of vancomycin, ceftazidine and dexamethasone, and the infection was identified as coagulase-negative Staphylcoccus. Six months after treatment, the patients had VA of 20/50 and resumed injections of ranibizumab (Figure 10).

Figure 9. Four days after being injected with ranibizumab, this patient presented with a visual acuity of hand motions.

Figure 10. After six months, the same patient had recovered visual acuity of 20/60 and continued with ranibizumab injections.

Turning then to the BPEI data, Dr. Flynn reviewed the injection technique used there, which includes no preinjection antibiotics, but does use povidone-iodine perioperatively and (by some but not all physicians) topical antibiotics postoperatively, the latter since 2008 at the discretion of the individual physician. Disposable specula are used as part of an injection kit.

Dr. Flynn also reviewed the numbers of injections given at BPEI since 2005, with bevacizumab accounting for more injections that ranibizumab, although the numbers have grown closer together since 2009 (Figure 11). Dr. Flynn noted the purpose of the BPEI study, which was to identify the rates of culture-proven endophthalmitis following intravitreal injections since 2005 (the “anti-VEGF era”), as well as to characterize the treated cases of endophthalmitis over the same period.

Figure 11. Over five years, bevacizumab injection rates have soared at BPEI, while ranibizumab has leveled off and pegaptanib has sunk to zero.

This retrospective study covered January 1, 2005, through December 31, 2010, and studied the clinical data of patients treated for endophthalmitis following intravitreal injection of an anti-VEGF agent. Also reviewed were the intravitreal injection logs to provide, as Dr. Flynn said, the denominator for the total number of cases of postinjection endophthalmitis.

In the identified five-year period, BPEI administered over 60,000 intravitreal injections of anti-VEGF agents and identified only 12 clinically diagnosed and treated endophthalmitis cases following those injections. Seven were in cases of injections of bevacizumab, and five were in cases of ranibizumab injections; there were no cases consequent to injection with pegaptanib. Of these 12 cases, seven were culture-positive. The mean age of the patients in whom the infections were identified was 80.5 years old. Faculty members had administered the injections in nine of the 12 cases (fellows constituted the other three cases).

Thus, the overall rate of treated cases was 0.02%, while the rate of culture-positive cases was 0.01%. For ranibizumab, treated cases constituted 0.027%, while the culture-positive cases were 0.011%. The respective percentages for bevacizumab were 0.017% and 0.008%, with no statistically significant difference between the bevacizumab and ranibizumab cohorts.

Regarding the species of infection in the 12 cases of endophthalmitis, five of the seven culture-positive cases were Streptococcus infections, and four of those five patients had very poor treatment outcomes, with only one recovering good visual acuity ( 20/40), while the rest had VAs of either light perception or no light perception.The other culture-positive cases were S. epidermis and Bacillus anthracite infections, with final VAs of 20/70 and 20/60, respectively.

At this point, Dr. Flynn reviewed two of the Streptococcus cases, both of which began when diagnosed with VA of hand motions, and with the first case degenerating to light perception and the second to no light perception. Dr. Flynn also reviewed the culture-negative cases in the BPEI series, only one of which had a final VA worse than 20/80, which was count fingers at one month, before turning to another case study.

This case was of an 80-year-old pseudophakic male wet AMD patient with a baseline VA of 20/50. The patient received three prior injections of bevacizumab and four injections of ranibizumab, and he presented two days after his last injection with pain and decreased VA to hand motions. Ultimately, at one month following the injection, the patient underwent pars plana vitrectomy. The patient's infection was identified as Streptococcus intermedius, and his final VA was light perception at 25 months. Over the course of treatment, this patient also developed a central retinal vein occlusion.

In another case, an 83-year-old male patient with wet AMD and a VA of 20/100 presented one day following an injection of ranibizumab with floaters, pain and decreased VA. Over a six-month course of treatment, it was found that the patient was resistant to fluoroquinolones, so he was treated with vancomycin and returned at 39 months to a VA of 20/70.

Before closing, Dr. Flynn reviewed several studies similar to the BPEI study, including a meta-analysis presented at last year's AAO meeting by Colin A. McCannel, MD, which identified an endophthalmitis rate of 0.049% in over 100,000 intravitreal injections. Of these 54 cases, one-third were coagulase-negative non-Streptococcus infections, while eight were S. viridans cases. The take-home point here was that the rate of endophthalmitis due to Streptococcus species was three to four times higher following anti-VEGF injections than following vitrectomy or cataract surgery.

To reduce this risk, Dr. Flynn recommended no talking over or around the patient, as bacteria from the doctor or nurse can be transmitted to the needle this way. In closing, Dr. Flynn reviewed the data and underscored that, while uncommon, endophthalmitis can be devastating, particularly when caused by Streptococcus species.

NEXT YEAR IN NEW YORK

The Macula 2012 is scheduled to be held in New York next year. Please keep an eye on our Calendar section near the back of the magazine for more information once it's announced. RP