The Uncharted Realm of the Retinal Periphery
How widefield imaging is launching a brand new field of study.
Seenu M. Hariprasad, MD · Ravi D. Patel, MD · John W. Kitchens, MD
Diagnostic imaging has played an increasing role in eye care in recent years. It seems like every time we turn around, retina specialists are being offered some new imaging device boasting an overwhelmingly powerful “Oh, wow!” factor. But precious and few are the devices that change the way we think about a disease or fundamentally alter clinical management. In this article, we will discuss a new technology that may accomplish these very things.
Ultra-widefield fundus fluorescein angiography (UWFFA) has opened up areas of study previously unknown in the vitreoretinal subspecialty. Until recently, the retinal periphery went largely unimaged, mainly because until now there existed no easy way to image it.
Under optimal conditions, traditional angiographic methods employing film or digital fundus cameras can capture 50° views. With seven standard fields, a 75° view of the fundus can be obtained. Finally, with luck and a compliant patient, a contact lens system such as the Ocular Staurenghi 230 SLO Retina Lens (Ocular Instruments, Bellevue, WA) can image out to the 120° range.
Figure 1. Hydroxychloroquine toxicity with “bull's eye” maculopathy and peripheral pigmentary changes that have never been described before in the literature.
None of these modalities can compare to the ultra-wide views of up to 200° provided by the Optos 200Tx scanning laser ophthalmoscope (Optos PLC, Dunfermline, United Kingdom). But perhaps even more valuable than the wider range of field, its images are captured simultaneously. Seven standard fields and contact lens systems capture their images in a successive, piecemeal fashion and then assemble a montage. Because the transit time of fluorescein dye occurs so quickly — 10 to 30 seconds — traditional widefield technologies omit what could be revealing stages of fluorescein dye transit through the fundus.
By contrast, the Optos device permits simultaneous detection of several foci of neovascular dye leakage and an intraretinal comparison of their severity. To understand why this is important, we should remember that each focus may have different levels of dye leakage when viewed at the same time point. Add to this the device's rapid image capture, ease of use, ability to store consecutive digital images, and ability to penetrate media opacity and poor dilation — and its popularity among high-volume retinal practices becomes clear.
Figure 2. Superotemporal BRVO with secondary macular and peripheral ischemia. Note compensatory collateral formation.
Enhanced viewing of the periphery has led to several exciting new clinical findings, most of which have occurred in the past year or two. The technology has unveiled new insights regarding the role of peripheral pathology in retinal vascular, degenerative and inflammatory diseases. Potential new disease markers have emerged, as have the first efforts to quantify peripheral pathology. There is also hope that UWFFA with targeted retinal photocoagulation will help to prolong the effects of anti-VEGF therapy in certain macular diseases with peripheral ischemia.
After surveying leaders in this field (see the box below), we've gathered some of their comments here and cite relevant clinical studies where appropriate.
The Periphery and Retinal Disease
Many researchers believe there may be an association between peripheral nonperfusion and the presence of neovascularization and macular edema secondary to retinal vascular disorders such as diabetic retinopathy and RVO.
The hypothesis is that zones of peripheral nonperfusion may generate biochemical mediators such as VEGF and other growth factors that promote macular edema and retinal neovascularization. “Widefield imaging is particularly interesting in diabetic retinopathy. The degree of poor mid-peripheral perfusion in the diabetic eye, even those that do not appear to have moderate or severe nonproliferative retinopathy, is astounding,” says Paul E. Tornambe, MD, of San Diego. “I believe there is a significant relationship between midperipheral and peripheral poor perfusion and diabetic macular edema. We've overlooked this in the past because the entire midperipheral fundus could not be imaged simultaneously during fluorescein angiography.”
Figure 3. Stargardt disease with classic findings of “bull's eye” maculopathy, numerous pisciform flecks and a dark choroid.
Dr. Tornambe adds that during research he found elevated VEGF levels in patients with DME. He doubts whether posterior pole macular ischemia alone could generate so much of the growth factor. Additionally, he believes cytokines other than VEGF are involved. “In anterior-chamber taps prior to vitrectomy for DME in eyes pretreated with VEGF inhibitors, we found very low VEGF levels — and yet florid DME persists. Other cytokines, such as interluken 6 and 8 and MCP-1, are likely involved and are also elaborated by ischemic midperipheral retina. It appears steroids affect these cytokines,” he says.
There is also evidence that peripheral pathology may be present in age-related macular disease. Led by SriniVas Sadda, MD, a team of retina specialists at the Doheny Eye Institute found that 76% of eyes in patients with AMD had peripheral abnormalities visible on ultra-widefield autofluorescence imaging. Though the findings of this retrospective study are preliminary, they probably deserve additional investigation, Dr. Sadda believes. “At this point, all we know is there are fundus autofluorescence abnormalities in the periphery in a variety of diseases, including AMD,” he notes. “But we are not in a position yet to determine whether this should affect management. We need prospective studies and use in trials. We have noted that peripheral changes and areas of atrophy are very common in patients with AMD. But we don't know yet if they have prognostic implications. A substudy of AREDS2 will explore this question.”
Figure 4. PDR after PRP laser therapy with extensive retinal neovascularization not identified during biomicroscopy but seen on UWFFA imaging.
Figure 5. Sarcoidosis with secondary peripheral ischemia and retinal capillary microangiopathy, periphlebitis and retinal neovascularization.
Figure 6. A nondiabetic young female who presented with proliferative disease. Initially, it was assumed to be a unilateral case until the UWFFA imaging showed retinal capillary microangiopathy and collateralization temporally in the far periphery of the contralateral eye.
An additional conclusion of Dr. Sadda's study was how beneficial autofluorescence in retinal scanning technologies could be for diagnosing AMD. At least one other retina specialist agrees. “I have not used widefield imaging as much in AMD as I have in other conditions,” observes Mathew W. MacCumber, MD, associate professor of ophthalmology at Rush Medical Center in Chicago. “But I think it's worth mentioning that the latest Optos hardware package allows for autofluorescence in the periphery. This opens a whole new area of study.”
Dr. MacCumber, who specializes in retina and inflammatory eye disease, has studied widefield imaging in conjunction with uveitis. “We recently published a paper that compared widefield imaging to conventional fundus photography in patients with CMV [cytomegalovirus] retinitis, which found the Optomap tended to capture greater areas of total retina and peripheral CMV lesions,”1 he points out. “Patients also liked it because it is faster.”
Another recent study found UWFFA to be a useful tool in the detection of incipient inflammation and for following treatment efficacy in patients with intermediate uveitis.2
A variety of strategies for using UWFFA as a basis for improving treatment patterns are currently being discussed. They often include identifying areas of nonperfusion as a means for better targeting and applications of more precise laser therapy — so-called targeted retinal photocoagulation (TRP).3 The hope is that better targeting of laser therapy could reduce the risks of panretinal photocoagulation and possibly enhance the duration and efficacy of anti-VEGF injections.
“My present approach to DME is focal laser to areas of circinate and then targeted laser to areas of poor peripheral nonperfusion,” says Dr. Tornambe. A repeat widefield fluorescein angiography study a month later determines if poorly perfused areas have been adequately treated. Treatments are repeated if necessary. “A few decades ago, PRP was felt to aggravate DME. But I think this is true only if areas of well-perfused retina are also destroyed by heavy laser PRP.”
“Before widefield technology,” Dr. Tornambe continues, “I had no idea extensive disease existed in the midperiphery. Out of sight, out of mind!”
Szilárd Kiss, MD, director of clinical research at Weill Cornell Medical College, says that he uses widefield technology “almost exclusively” on anything other than macular pathology. “It helps in diagnosis, it helps in treatment, it helps in follow-up,” he notes. “If a patient has a large area of nonperfusion, I would offer a pharmacological intervention combined with peripheral laser treatment. Now this strategy is sort of based on clinical experience rather than clinical trial data. But I think that is where we're headed.”
Another retinal specialist who believes that peripheral nonperfusion in some patients is associated with increased VEGF production is David S. Boyer, MD, of Beverly Hills, who says he uses UWFFA to target laser therapy. But like Dr. Tornambe, he is not yet able to say whether this technology has impacted outcomes. “Without a prospective, randomized trial, it is difficult to say. Anti-VEGF injections have improved our results, but the problem is these treatments continue indefinitely. My hope is that with selective PRP laser I can achieve equivalent visual results, plus a reduction in the number of treatment injections.”
On the other hand, some surgeons note that targeting pathology should have its limits. “It is important to remember that PRP laser works in three distinct ways,” says Nancy Holekamp, MD, of St. Louis. “First, it kills ischemic retina tissue to lower VEGF production. Second, PRP kills normal retina tissue so it consumes less available oxygen. Third, PRP allows for oxygen to diffuse from the choroid across laser scars, increasing intraocular oxygen. So, yes, targeted retinal photocoagulation makes sense, directly treating the areas of retinal nonperfusion with PRP. However, we should also remember it may still be necessary to treat healthy areas of retina in some diabetic patients.”
Nevertheless, most retinal specialists we spoke to agree that better targeting of laser therapy will probably prove beneficial. After all, panretinal ablation poses documented risks. The landmark Diabetic Retinopathy Study found that 10% of patients experienced a decline in visual acuity, and 5% showed a constriction of their visual fields.4 Full-scatter photocoagulation may exacerbate macular edema and cause loss of vision. Other rare complications include hemorrhage, choroidal detachments, acute angle closure glaucoma, a decrease in color vision and contrast sensitivity, nyctalopia and lenticular burns.3
“Until we determine a way to improve midperipheral circulation,” concludes Dr. Tornambe, “we must continue to address the ischemic areas with laser ablation. Now that we can image this area ‘en bloc,’ it makes sense to me to ablate only the diseased areas, sooner rather than later. I am treating nonproliferative DR with targeted PRP much more frequently than I did in the past.”
Among patients with branch retinal and hemiretinal vein occlusions (BRVO, HRVO) who develop macular edema, a subset respond transiently to anti-VEGF or steroid therapy and develop recalcitrant macular edema. The answer to this problem may lie in the periphery, one study found.5 Untreated nonperfusion anterior to the globe equator was significantly associated with macular edema in patients with BRVO and HRVO. The study concluded that untreated areas of nonperfusion, many of them found in the periphery, may be the source of biochemical mediators that promote neovascularization and macular edema.
Similar trials are under way. K. Bailey Freund, MD, of New York City, is involved in a study seeking to determine whether TRP can reduce dependence on anti-VEGF injections in patients with CRVO and BRVO. “Speaking anecdotally, I can report that, unfortunately, this has not been a magic bullet so far with CRVO,” says Dr. Freund. “However, with BRVO we appear to be having more luck. In patients with chronic recurrent macular edema at four to six weeks after injections, a little more targeting of the laser therapy appears to be enhancing the benefits of anti-VEGF therapy. But definitive conclusions are still far off.”
Intuitively, it seems like the information provided by UWFFA would be valuable in RVO patients, he says. As an example, he points out a recent study that found eyes with neovascularization on the day of the angiogram correlated with significantly larger areas of nonperfusion.6
Another Potential Disease Marker
Investigations have recently identified the angiographic phenomenon of late leakage from retinal veins and arteries, which they've termed “peripheral vessel leakage” (PVL). It is a characteristic difficult to visualize without a widefield fundus view.5 Also, it frequently occurs in patients with active diabetic retinopathy. Although no one is sure what mechanism causes PVL, there is speculation it might prove a more accurate disease marker than nonperfusion.
Figure 7. A four-year-old with Coat's disease and heavy lipid exudation, pre (left) and post (right) treatment with resolution of exudative changes.
As a recent study put it: “We hypothesize that PVL is a marker for active diabetic retinopathy and may be a sensitive indicator of ischemia in diabetic eyes. Although capillary nonperfusion is a common finding in diabetic retinopathy, it is not always associated with active retinopathy. One possible explanation is that completely non-perfused inner retina may not be able to produce or secrete VEGF. A more relevant angiographic association with active, treatment-threshold diabetic retinopathy may be PVL.”7
“With elevated levels of intraocular VEGF, one will see peripheral vessel leakage,” adds Dr. Holekamp. “PVL in some eyes — and in some areas of the same eye — may precede PDR. Like retinal neovascularization, PVL may be a disease marker of intraocular ischemia, but PVL may be an earlier marker in some eyes.”
One of the big questions in RVO is why some patients go on to develop macular edema while others remain stable. If the hypothesis that untreated retinal nonperfusion is associated with macular edema is valid, then it stands to reason that some method for quantifying nonperfusion and ischemia should help us identify patients at risk for progressing to macular edema.
Attempts to assess the usefulness of quantifying non-perfusion have been met with preliminarily encouraging results.6 Researchers looked at 69 eyes with CRVO and used image analysis software to generate what they called an ischemic index, calculated by dividing the total area of the nonperfused retina by the total image area. The study found a statistically significant relationship between an increased ischemic index and the presence of neovascularization.
Software provided with the Optos device not only identifies areas of nonperfusion but can quantify the amount of nonperfusion as well. Opinions, however, seem mixed about this feature. “The Optos device provides us with information about the level of perfusion and highlights areas of nonperfusion very well. However, it doesn't tell us about the oxygenation status of the retina in these or other areas, which may or may not be correlated,” says Andrew A. Moshfeghi, MD, medical director of the Bascom Palmer Eye Institute. “New technologies are emerging that will provide that information as well. I'm not sure our lack of that information now (ie, the quantitative level of true ischemia) leads me down the wrong therapeutic pathway, but it may help us optimize diabetic treatment approaches in the future in a more proactive and quantitative fashion.”
Also, because the Optomap photograph uses an ellipsoidal mirror, the image is distorted in the same way that a flat map of the spherical earth is: areas in the periphery look much larger than they actually are. The software has not fully compensated for that effect, but the manufacturer is getting closer to a solution.
The retina specialists we spoke to say they most often use the quantifying feature in conjunction with clinical research rather than on routine patients.
Figure 8. Albinism with enhanced visualization of the choroidal circulation through the depigmented RPE layer.
Innovation in technology has greater value if we can apply it in ways that add insight to our understanding of diseases. If new technologies offer promise of improving treatment paradigms and patient outcomes, then they can quickly earn a place in standards of care. We've seen this happen with optical coherence tomography in recent years. We believe that preliminary data suggest UWFFA deserves serious consideration as a technology on the cusp of changing the way we think about retinal disease.
For instance, seven-field FA is limited, notes Dr. Kiss, because “you don't know where the pathology is necessarily, and your photographer certainly doesn't know.” If you capture images at different time points and different quadrants, you're not getting a full picture, he says. “With UWFFA you get the entire view from the injection of the fluorescein all the way to the end of the transit in one sweep.”
With further research and understanding, it is hoped that this new technology may impact treatment paradigms and ultimately improve patient outcomes. RP
1. Mudvari SS, Virasch VV, Singa RM, MacCumber MW. Ultra-widefield imaging for cytomegalovirus retinitis. Ophthalmic Surg Lasers Imaging. 2010;41:311-315.
2. Tsui I, Kaines A, Schwartz S. Patterns of periphlebitis in intermediate uveitis using ultra wide field fluorescein angiography. Sem Ophthalmol. 2009;24:29-33.
3. Reddy S, Hu A, Schwartz SD. Ultra wide field fluorescein angiography guided targeted retinal photocoagulation (TRP). Sem Ophthalmol. 2009;24:9-14.
4. The Diabetic Retinopathy Study Research Group. Photocoagulation treatment of proliferative diabetic retinopathy: clinical application of Diabetic Retinopathy Study (DRS) findings. Ophthalmology. 1981;88:583-600.
5. Prasad PS, Oliver SC, Coffee RE, Hubschman JP, Schwartz SD. Ultra wide-field angiographic characteristics of branch retinal and hemicentral retinal vein occlusion. Ophthalmology. 2010;117:780-784.
6. Tsui I, Kaines A, Havunjian MA, et al. Ischemic index and neovascularization in central retinal vein occlusion. Retina. 2010 Nov 22. [Epub ahead of print]
7. Oliver SC Schwartz SD. Peripheral vessel leakage (PVL): a new angiographic finding in diabetic retinopathy identified with ultra wide-field fluorescein angiography. Sem Ophthalmol. 2010;25:27-33.
|Seenu M. Hariprasad, MD, is associate professor and director of clinical research at the University of Chicago Department of Surgery's section of ophthalmology and visual sciences, where he serves as chief of the vitreoretinal service. He serves as a consultant for Optos. Dr. Hariprasad can be reached at firstname.lastname@example.org.|
Ravi D. Patel, MD, is a first year vitreoretinal surgical fellow at the University of Chicago Department of Surgery's section of ophthalmology and visual sciences. His research interests include ocular imaging technologies and ocular drug delivery. He has no financial disclosures.John W. Kitchens, MD, is an ophthalmologist and vitreoretinal surgeon with Retina Associates of Kentucky, located in Lexington. His focus is in the management of medical and surgical diseases of the retina. His interests include angiogenesis, innovations in retinal imaging and ocular trauma. He serves as a consultant for Optos.
David S. Boyer, MD, is clinical professor of ophthalmology at the Keck School of Medicine at the University of Southern California and practices with Retina Vitreous Associates Medical Group in Beverly Hills. He is a consultant to Optos. Dr. Boyer can be reached at email@example.com.K. Bailey Freund, MD, is a partner in Vitreous-Retina-Macula Consultants of New York. He is also clinical associate professor at New York University School of Medicine and an attending physician at NYU Langone Medical Center, New York Presbyterian Hospital and Manhattan Eye, Ear & Throat Institute. He has no financial disclosures relevant to this article. Dr. Freund can be reached at firstname.lastname@example.org.
Nancy Holekamp, MD, is professor of clinical ophthalmology at Washington University School of Medicine in St. Louis. She has no financial disclosures relevant to this article. Dr. Holekamp can be reached at email@example.com.Szilárd Kiss, MD, is assistant professor of ophthalmology and director of clinical research at Weill Cornell Medical College in New York. He is also an assistant attending physician at New York Presbyterian Hospital. His institution has received research funding from Optos. Dr. Kiss can be reached at firstname.lastname@example.org.
Mathew W. MacCumber, MD is associate professor and associate chairman for research in the ophthalmology department at Rush University Medical Center in Chicago. He has received grant funding from Optos and Heidelberg and is a consultant to Optos. He can be reached at email@example.com.Andrew A. Moshfeghi, MD is medical director of the Bascom Palmer Eye Institute at Palm Beach Gardens and is an assistant professor of ophthalmology at the University of Miami Miller School of Medicine. He has no financial disclosures relevant to this article. Dr. Moshfeghi can be reached at firstname.lastname@example.org.
SriniVas Sadda, MD, is associate professor of ophthalmology at the Keck School of Medicine of USC and the Doheny Eye Institute. He is director of the Medical Retina Unit, Ophthalmic Imaging Unit and Doheny Image Reading Center. He has received research support from Optos. Dr. Sadda can be reached at email@example.com.Paul E. Tornambe, MD, is a retinal specialist with San Diego Retinal Research Foundation in Poway, CA. He has no relevant financial disclosures. Dr. Tornambe can be reached at firstname.lastname@example.org.
Note: Only financial disclosures directly relevant to the content of this article have been included.