40 may 2014 may Doheny Eye Institute in California. If traction goes inward, macular pucker may result. may pucker macular inward, goes traction If California. in Doheny Eye Institute the and Institute VMR MD, at the Sebag, Jerry explained hole form, may a macular outward, is tion trac of tangential direction the If AMD). of DR and exacerbation as well (as occur can syndrome VMT macula, the in occurs traction When . on the traction creating PVD), (partial-thickness split can vitreous PVD. The anomalous in resulting not weakened, has (VMA) macula to the ous or PVD. detachment, vitreous posterior called is story. This the of end that’s the and floaters, causing typically retina, of the free pulls vitreous the people, most in is it as synchronized, is process two-step this When retina. the and cortex vitreous posterior the tween “Velcro” of the be a weakening is there time, same At the explained. Dr. Holekamp retina,” on the pulling solids remaining the with eye, the in around sloshing to start liquefied has gel of the enough of floaters.” we’re full old, it’s “When and 100 years 100 liquid percent Louis. St. in Institute Vision Pepose MD, at Holekamp, the M. Nancy life-long,” said is over time liquefying gel of the process 100 but gel, the percent and clear optically is that gel avitreous with born “We all are vitreolysis. role of pharmacologic current the particularly approaches, best on the differ viewpoints (AMD). degeneration macular age-related exudative and Pathophysiology. Velcro. like mid-60s, “By our surface, on its everywhere retina to the adheres vitreous the agel, As diseases, interface vitreomacular of these treatment the in advances many have been there Although and VMA is also associated with retinal tears and detachment, diabetic retinopathy (DR), retinopathy diabetic detachment, and tears retinal with associated also is VMA and of PVD combination Moreover, problematic (VMA). this adhesion vitreomacular persistent with (PVD) detachment vitreous posterior anomalous problem: fundamental same of the manifestations to be now understood are puckers macular some hole, and macular drome, syn (VMT) traction one vitreomacular another, from different very to be thought nce In some instances, the vitreous gel has liquefied, but the adhesion of the vitre of the adhesion but the liquefied, has gel vitreous the instances, some In BY GABRIELLE WEINER, WEINER, GABRIELLE BY decide on their course of treatment. consider when helping patients Four experts weigh in on what to CONTRIBUTING WRITER CONTRIBUTING 1

3,4 CONTENT VIDEO & ADDITIONAL &ADDITIONAL VIDEO - - - - AVAILABLE AVAILABLE 2 EXTRA

alfred t. kamajian eyenet 41 The consequent structural changes in the macula can cause metamorphopsia, , central defects, and image size disparity. About 1.5 percent of the population is estimated to have The most important symptom to watch for in vitreomacular vitreomacular interface disease, although that figure interface diseases is metamorphopsia. The most import- is expected to rise with the growing number of older ant physical finding is evidence of VMA with disruption in people and more widespread use of optical coherence the macula. “Just having VMA [with no disruption] can be tomography (OCT).5 normal,” said Dr. Holekamp. But as Dr. Sebag pointed out, It is worth noting that macular pucker is different “When a patient has VMA with structural changes in the from VMT. The latter has anomalous PVD with sep- macula, that’s VMT.” Depending on the severity of symp- aration peripherally but full-thickness vitreous cortex toms and the OCT images, referral to a retina specialist is pulling on the macula, usually in an axial or oblique likely indicated. direction. Macular pucker results from anomalous PVD with vitreoschisis, which is a split in the poste- rior vitreous cortex that leaves the outermost layer of vitreous attached to the macula while the rest of the vitreous cortex detaches away from the retina.1 There Thanks to advances in imaging technologies, clini- can also be minor damage to the retina, stimulating cians are identifying vitreomacular interface diseases an immune response. The subsequent proliferation of earlier than ever before. And the fact that mild cases cells in the macular area can form a layer of scar tis- can now be picked up by OCT raises questions about sue that tightens, creating traction on the macula and treatment. Although surgery remains the standard of causing it to pucker. Although a pucker occasionally care, ophthalmologists may wonder if they should try disintegrates (particularly in people under age 50), the something else first. “The answer is largely driven by condition is permanent in the majority of patients. symptoms and ,” said Dr. Thompson. Terminology confusion. There is no universally “As physicians, we have to collect data before mak- accepted nomenclature or classification system for ing a treatment decision. We look at vision, look at vitreomacular interface diseases. “It’s difficult to com- the OCT scan, and talk to patients about how much pare outcomes when we’re defining things differently they’re bothered by symptoms,” said Dr. Holekamp, from one another,” said John T. Thompson, MD, at who emphasized three points: 1) “We don’t treat the Wilmer Eye Institute in Baltimore. “Other than scans; we treat people. Just because we see an awful- with OCT pictures, there’s no way to accurately com- looking OCT doesn’t mean a patient needs interven- municate a patient’s condition.” tion if his or her vision is good.” 2) “Most epiretinal Just listing some of the terms used for macular membranes [macular puckers] don’t get worse. Only pucker highlights the problem: a minority of patients with epiretinal membranes go (ERM), epimacular membrane, preretinal membrane, on to require intervention.” 3) “These are mechanical cellophane , and retinal wrinkle, for problems, so they need mechanical solutions. If inter- instance. Andreas K. Lauer, MD, at Oregon Health & vention is appropriate, the options are either pharma- Science University in Portland, uses the term epiret- cologic vitreolysis or surgery.” inal membrane when talking with colleagues and macular pucker when talking with patients, given that the Academy’s patient education materials use When to watch. When there is evidence of disruption macular pucker. Dr. Sebag advocates use of the term to the macula on the OCT scan but a patient is asymp- premacular membrane rather than ERM because it is tomatic, Dr. Lauer recommends watchful waiting. He more specific, and he prefers to use macular pucker to gives these patients an to monitor their describe the effects of the premacular membrane on status at home and schedules them for a follow-up the macula. dilated exam about six months later. “Only when pa- There are a few classification systems floating tients are symptomatic and are willing to accept the around, according to Dr. Thompson, but the most risks of intervention do I treat vitreomacular interface recent is an OCT-based anatomic classification system diseases.” Similarly, Dr. Holekamp watches patients proposed by the International Vitreomacular Traction with VMT or ERM if they have good vision, if their Study (IVTS) Group6 (see “Classifying VMA”). Given OCT has only mild changes, or if they simply aren’t the current level of understanding and our diagnostic bothered by it. capabilities, this proposed classification system is ex- When not to watch. However, Dr. Holekamp never pected to help clinicians select the most appropriate watches anyone with a macular hole. “Ten out of 10 treatment and is potentially useful in the execution retinal specialists would recommend treatment for and analysis of clinical studies, the experts agreed. macular hole,” she said. “Most would say vitrectomy;

42 may 2014 some might say pharmacologic vitreolysis.” When the patient is under age 65, is phakic, does And Dr. Sebag argued that watchful waiting is no not have a macular pucker, and has an adhesion less longer an option. “We now have pharmacologic vitre- than 1,500 µm, the probability of success goes up to 60 olysis, which can be administered while waiting. If it to 80 percent. “These criteria are not obligatory; they works, great; if not, we can always move on to surgery. just help the physician and patient weigh the probabil- My preferred choice is to inject and then wait.” ity of success,” Dr. Sebag said, adding that the success of injecting ocriplasmin for a macular hole is directly related to the hole’s size. Data show that the drug Historically, treatment for symptomatic VMA has works fairly well for holes 250 to 400 µm in diameter been surgical. But in January 2013, when the vitreo­ and does best for holes under 250 µm, he said. No lytic agent ocriplasmin (Jetrea) came to market, the holes larger than 400 µm have resolved with ocriplas- era of pharmacologic vitreolysis began. min to date. Just as in surgery, the objective of pharmacologic Safety concerns. Clinical trial data showed ocriplas- vitreolysis is to relieve traction on the macula by lys- min to be safe, but the postmarketing experience has ing the anomalous VMA. “Ocriplasmin makes sense uncovered some complications, including temporary when there is a small area of adhesion with VMT or separation of the retina from the retinal pigment ep- a small or medium full-thickness macular hole with ithelium, with disruption of the ellipsoid layer, and VMT,” said Dr. Thompson. In such cases, success rates electrophysiological changes as detected by electro- are around 40 percent for good candidates, according retinography. “It’s not a common phenomenon, but to Drs. Thompson and Lauer; Dr. Holekamp’s esti- it’s something that needs to be studied more,” said Dr. mate is 50 percent. At $4,000 per shot, cost is a major Thompson. consideration, given that some retina specialists are Dr. Holekamp concurred. “Ocriplasmin is expen- still having reimbursement problems with ocriplas- sive and works under ideal circumstances only some min. (While the drug is cheaper than surgery, the suc- of the time. Throw into the equation that there may cess rate of surgery approaches 100 percent.) be complications that weren’t described in the clinical Patient selection. Dr. Sebag, who was an investiga- trials but that we may now be seeing, and I say to my- tor in the phase 2 and 3 trials of ocriplasmin, pointed self that I have concerns. I’ll let our profession figure out that when the trials reported 27.1 percent efficacy, out the true risks and efficacy and then adjust my oc- that rate was for all participants. “We didn’t know riplasmin use accordingly.” that there were favorable characteristics at that time, In contrast, Dr. Sebag is of the opinion that the but we were able to identify retrospectively certain safety concerns are overblown. “Personally, I hav- features that incrementally increased the likelihood of en’t experienced any untoward events,” he said. “As success,” he said. a consultant to the company, I have been provided

A B C

D E F

These OCT scans tissues. B) Broad VMA with no de- with intrafoveal pseudocyst. illustrate VMA and VMT according tectable change in foveal contour F) Broad VMT with distortion of to the IVTS Group. In scans A, B, of underlying retinal tissues. the foveal surface and elevation and D, white arrows mark sites C) Focal VMA with concomitant of the foveal floor. The VMT is also of vitreous attachment. A) Focal wet age-related macular degenera- associated with an epiretinal mem- VMA with no detectable change in tion. D) Focal VMT with distortion brane and macular pucker (white

ophthalmology foveal contour of underlying retinal of the foveal surface. E) Focal VMT arrowheads).

eyenet 43

CLINICAL STAGES ATTRIBUTES COMMENTS

VMA Vitreous adhesion to central macula with no Has been called stage 0 in the past when con- demonstrable retinal morphologic changes. tralateral eye has FTMH; normal appearance on clinical examination; no symptoms. VMT Vitreous adhesion to central macula with May or may not have yellow changes in central demonstrable changes by OCT but no macula on examination; can be referred to as full-thickness tissue dehiscence; may in- impending macular hole if FTMH is present in clude the following: tissue cavitation, cystoid contralateral eye. changes in macula, loss of foveal contour, elevation of fovea above the RPE. Small FTMH Hole < 250 µm, may be round or have a flap Visual acuity may be relatively good; optimal size adherent to vitreous; operculum may or may for successful repair by pharmacologic vitreoly- not be present. sis; very high probability of success with vitrecto- my surgery.

Medium FTMH Hole > 250 but < 400 µm; may be round or High probability of success with vitrectomy sur- have a flap adherent to vitreous; operculum gery. may or may not be present. Large FTMH Hole > 400 µm; vitreous more likely to be Slightly less probability of successful closure fully separated from macula. with vitrectomy surgery. FTMH, full-thickness macular hole; OCT, optical coherence tomography; RPE, retinal pigment epithelium; VMA, ; VMT, vitreo­macular traction. SOURCE: Adapted from Duker JS et al. Ophthalmology. 2013;120(12):2611-2619.

with data on all the adverse events that occurred in Dr. Sebag said that his patients are similarly keen to the study, and the transient nature of these adverse try the drug. “Overwhelmingly, they choose the injec- events and the minuscule number of people affected tion because they understand that they’re not compro- impressed me. It’s not something that I think should mising themselves—meaning, if it doesn’t work, we do thwart our implementation of this approach. Of the surgery, and the surgery proceeds as it would have course, safety is of paramount importance; thus, we if they’d never had the injection, with the same likeli- should collect our experiences to see how implementa- hood of success.” Like Dr. Lauer, Dr. Sebag estimated tion in the real world compares to the clinical trials.” that it works in about one in three patients. Dr. Sebag further emphasized that ocriplasmin “With careful patient selection, some people can is the first pharmacologic vitreolysis drug brought avoid surgery with ocriplasmin, and that’s a signifi- to market. “There are going to be problems, just like cant advance,” Dr. Thompson said. “It also might be there were for the first antibiotic when it was intro- helpful in the secondary complications that relate to duced. But it’s the right way to go. It’s going to help VMT in patients with diabetic retinopathy and per- patients, lower health care costs, and be a wonderful haps neovascular AMD. That would be invaluable.” new paradigm. As we learn more, we will refine our approach, and we’ll have better success rates.” Current use. “We all became so excited when oc- In surgery, the vitreous is removed to gain access riplasmin came to market, but I’m more cautious to the site of VMA so that the surgeon can remove now,” said Dr. Lauer. He considers using ocriplasmin the adhesion from the macula. Vitrectomy has been in patients who are phakic, have a focal area of VMT, around for 40 years, but advances in surgical equip- have no concurrent eye problems or history of pre- ment and techniques have transformed it into a safer vious retinal laser treatment, and are not diabetic. and more effective procedure. Two-thirds of his VMT patients come in asking for Small-gauge vitrectomy. Today, many retina sur- ocriplasmin, as they don’t want surgery and are eager geons have adopted sutureless techniques with small- to try an injection first. He has had success in a third gauge instruments under local anesthesia. The sur- of his patients. gical and postoperative recovery time has decreased

44 may 2014

Retina specialists have yet to come to a consensus on whether the tra- ditional recommendation of face- down positioning after macular hole surgery should be followed or aban- for a week or even longer.” large macular holes,” he said. In the doned. Dr. Lauer counts himself among latter instance, he recommends a Dr. Thompson noted that some the group of surgeons moving away 14-day period. surgeons are advocating no (or very from face-down positioning. “I’m For her part, Dr. Holekamp said, short periods of) face-down position- not very strict about it. When I can “I position my folks face down for ing, “and the reported success rates get a good gas fill in patients who’ve a week, even although I believe are pretty good—but the success already had cataract surgery, I sim- most holes close within three to four rates are a little better if the patient ply ask them to lie on their sides days.” She added, “What we prac- stays face down.” He added, “There at night and remain upright during tice and what we know scientifically are some OCT devices where you the day. Typically, the macular hole are sometimes slightly different. can image through the gas bubble, will heal in the first week. In phakic One of the most amazing things and studies have shown that about patients, I encourage them to do about the ocriplasmin clinical trial two-thirds of macular holes close face-down positioning to reduce the is that macular holes closed about within one to two days; but one-third degree of cataract formation, but I’m one-third of the time without a gas don’t close, even up to a week af- not overly compulsive about it. I am bubble. But with gas tamponade and ter surgery. So if you want to catch compulsive about face-down posi- face-down positioning, they closed all-comers, then it’s useful to have tioning if there’s not a good enough 94 percent of the time. It’s hard to the gas bubble against the macula gas fill after surgery; likewise for change a winning game.”

as a result. If there’s one surgical tip that Dr. Lauer have at least a mild degree of ERM; and, by peeling would give, it’s to not spend too much time in the eye. the ILM, I know I’ve removed the associated ERM that “I strongly recommend either 23- or 25-gauge instru- could potentially cause problems with VMT later on.” ments to reduce operating time, invasiveness, postop- Dr. Holekamp said that she selectively peels the erative discomfort, and the risk of complications.” ILM in medium-to-large macular holes and moder- ILM peeling. There is continuing debate for and ate-to-difficult macular pucker cases, but she usually against peeling the internal limiting membrane does not do so in VMT cases. (ILM), with studies supporting both sides. “I make an Chromodissection. The challenge of membrane effort to remove the ILM in all macular hole cases,” peeling can be reduced with chromodissection—the said Dr. Lauer. Dr. Sebag concurred, saying, “Studies staining of membranes to facilitate their removal. have proven that ILM removal results in better out- “Peeling the ILM can be difficult, so my advice is to comes in macular hole cases.” use indocyanine green [ICG] stain,” said Dr. Thomp- In macular pucker, Drs. Lauer and Sebag tend to son. “It can be helpful in making certain that you’ve remove the premacular membrane only. “Sometimes, removed the ILM successfully. Some surgeons prefer in the course of removing the epiretinal [premacular] to dust the ILM with triamcinolone, which is a per- membrane, there’s an adhesion between the two, and fectly good method, but the triamcinolone only shows a part of the ILM will come with it. If there appears you where you’ve removed the ILM; it doesn’t allow to be a partial peeling of the ILM, I may remove it all. you to see the ILM to grab it as easily.” Dr. Sebag not- But I do not deliberately remove the ILM in all epiret- ed, “I believe that ICG alters the ILM on a molecular inal membrane cases. There are studies indicating that level, making it easier to peel.” when you remove the ILM, there can be some injury Dr. Lauer uses ICG for macular hole surgery to to the ganglion cells,” said Dr. Lauer.7 visualize the ILM. But because some studies have Dr. Thompson is an advocate for peeling the ILM raised the specter of ICG toxicity, he uses trypan blue in VMT syndrome, ERM, and macular hole surgery or triamcinolone for macular pucker and VMT. Dr. almost all the time. “That way I’m assured I’ve re- Holekamp simply uses triamcinolone for everything.

andreas k. lauer, md moved the VMT,” he said. “Many patients with VMT Gas tamponade. “For VMT and ERM, gas tampon-

eyenet 45 ade is not needed, but if a surgeon feels it’s necessary, VMA. “Sometimes the injected air/gas is sufficient to two to three days is adequate,” said Dr. Thompson. disrupt the adhesion. There’s no real downside except For macular holes, gas tamponade to temporarily for a transient elevation in intraocular pressure,” he seal off the hole is the standard of care. Most surgeons said. In terms of surgery, the main focus will be im- use a short-acting tamponade, such as sulfur hexa- provements in safety and postop recovery time. Along

fluoride (SF6), although some surgeons are now using those lines, 27-gauge instruments will be available even shorter-acting agents, such as sterile air. Early shortly, he said. research has reported similar rates of hole closure be- Dr. Holekamp hopes that the vitreous itself will

tween air and SF6 and a reduction in the time patients get more attention in the future. “We’re doing a lot of needed to spend in the face-down position.8 However, research on trying to create a PVD with an injection, evidence is currently inadequate, and air tamponade is but maybe we should be doing research on trying to not recommended. prevent the vitreous from liquefying in the first place,” To help ensure that the gas bubble covers the mac- she said. After all, vitreous liquefaction causes nuclear ular hole, Dr. Thompson prescribes seven to 10 days of cataract, 15 to 20 percent of open-angle glaucoma, face-down positioning after surgery, but this practice retinal tears and detachments, bleeding in diabetics, is still up for debate (see “Face Down—or Not?”). VMT, macular puckers, and macular holes. “If we could figure out how to prevent vitreous liquefaction, we wouldn’t have these problems. To me, it’s a final For Dr. Sebag, the future lies in pharmacologic vitre- frontier of ophthalmic research.” n olysis, and the future of pharmacologic vitreolysis lies in prevention. “The big bang will come when we learn 1 Sebag J. Graefes Arch Clin Exp Ophthalmol. 2004;242(8):690- to identify the right patients for inducing a pharmaco- 698. logic PVD that will prevent anomalous PVD. The two 2 Sebag J. Retina. 2009;29(7):871-874. diseases that will be most impacted by this preventive 3 Sebag J et al. Trans Am Ophthalmol Soc. 2009;107:35-44. approach will be diabetic retinopathy and exudative 4 Wang MY et al. Retina. 2009;29(5):644-650. AMD. The potential for savings in human and eco- 5 Jackson TL et al. Retina. 2013;33(8):1503-1511. nomic terms is staggering.” 6 Duker JS et al. Ophthalmology. 2013;120(12):2611-2619. Dr. Thompson noted that he anticipates improve- 7 Pichi F et al. Int Ophthalmol. 2013 Jul 18. [Epub ahead of print.] ments to drugs like ocriplasmin, improved drug deliv- 8 Hasegawa Y et al. Graefes Arch Clin Exp Ophthalmol. 2009; ery systems to enhance the effect at the vitreomacular 247(11):1455-1459. interface, and a better understanding of how to use the drugs. He also hopes to learn more about sim- MORE ONLINE. For a slideshow of images and more ply injecting an air or gas bubble to lyse a very small thoughts on this topic, see this article at www.eyenet.org.

NANCY M. HOLEKAMP, ophthalmology resi- a consultant to Alcon and Thrombo- MD, is director of reti- dency program, and Genics. na services at the Pe- associate professor of JOHN T. THOMPSON, pose Vision Institute ophthalmology at Ore- MD, is assistant pro- and clinical professor gon Health & Science fessor of ophthalmol- of ophthalmology University in Portland. ogy at the Wilmer Eye and visual sciences Financial disclosure: Institute and clinical at Washington University School of Receives research grant support from associate professor Medicine in St. Louis, Mo. Financial Allergan, the National Institutes of of ophthalmology disclosure: Is a consultant to Alimera Health, and Oxford Biomedica. at the University of Maryland in Sciences, Allergan, Genentech, Re- JERRY SEBAG, MD, Baltimore. He is also the current generon, and Sequenom; is a speak- FACS, FRCOPHTH, president of the American Society er for Genentech and Regeneron; FARVO, is professor of of Retina Specialists and cofounder receives research funds from Allergan clinical ophthalmology of Retina Specialists, which is a and Notal Vision; and has equity in at the Doheny Eye In- private practice in the greater Balti- Katalyst. stitute in Los Angeles more area. Financial disclosure: Is ANDREAS K. LAUER, MD, is chief and founding director a consultant to Genentech and re- of the vitreoretinal division, vice- of the VMR Institute in Huntington ceives grant support from Genentech chair for education, director of the Beach, Calif. Financial disclosure: Is and Regeneron.

46 may 2014