10
QUESTION
WHEN DO I REFER A PATIENT WITH A CENTRAL RETINAL VEIN OCCLUSION, WHAT IS THE WORK-UP, AND WHAT ARE THE TREATMENT OPTIONS?
Richard F. Spaide, MD
Central retinal vein occlusion (CRVO) is a condition manifested by dilated and tortuous retinal veins and intraretinal hemorrhage in all 4 quadrants of the fundus, macular edema, and variable areas of capillary nonperfusion. In the past, grading of the areas of capillary nonperfusion was used to classify eyes with CRVO into 2 distinct types based on perfusion status: (1) ischemic or nonperfused, and (2) non-ischemic or perfused.1,2 The field of retina is one of the few in medicine in which a venous occlusion causing pathological changes is thought to be non-ischemic, highlighting the poor terminology used to describe disease. Curiously, it is not a simple matter to generate animal models that faithfully mimic CRVO. Simple ligation of the central retinal vein in primates did not produce a picture consistent with what is termed CRVO in adult humans. To produce a better approximation of the fundus picture of CRVO, simultaneous arterial occlusion had to be induced as well. Due to the lack of suitable animal models, much of what we have learned about CRVO is based on careful clinical studies of human subjects.
Patients developing CRVO note decreased vision, scotomata, and visual distortion. In the Central Vein Occlusion Study (CVOS), a collaborative study of 725 affected eyes, the baseline acuity was reported in 3 categories: (1) 29% had a good visual acuity of 20/40 or better; (2) 43% had intermediate visual acuity of 20/50 to 20/200; and (3) 28% had poor visual acuity of less than 20/200.1 Approximately, 2/3 of patients who presented with good visual acuity maintained good visual acuity at 3 years.1 Of those presenting with intermediate visual acuity, 19% improved to good acuity, 44% stayed in the intermediate group, and 37% decreased to poor acuity.1 Those initially in the poor visual acuity group had an 80% chance of remaining in that group at 3 years.1 Eyes with 10 or more disc areas of capillary nonperfusion on fluorescein angiography (FA) were classified as nonperfused, while those with less than 10 disc areas were considered to be perfused. In some patients, the classification was not initially possible due to marked intraretinal hemorrhage and these eyes were termed indeterminate. Over the follow-up period, about 1/3 of eyes that were initially classified as perfused became nonperfused, while 5/6 of the eyes that were indeterminate were later classified as nonperfused. The classification of perfused vs nonperfused principally was made to predict which eyes would progress to neovascularization of the iris and angle. In the CVOS, 10% of eyes classified as perfused developed neovascularization of the iris or angle, as did 35% of the nonperfused eyes. Therefore, the classification of perfused vs nonperfused was neither very sensitive nor specific for the development of anterior segment neovascularization.
Initial treatment of decreased visual acuity in CRVO focused on macular edema. The CVOS studied the effect of grid-pattern laser photocoagulation on perfused eyes with macular edema and visual acuity of 20/50 or worse and demonstrated a beneficial anatomic effect (decreased edema on angiography), but visual acuity did not improve.3 This study was done prior to the era of optical coherence tomography (OCT) so it is difficult to estimate the reduction in retinal thickness achieved with laser.
Subsequent therapies attempted to increase collateral vessel development because collateral vessels redirect blood flow around an area of occlusion. These therapies were largely unsuccessful, had low rates of adoption, and are now interesting historical anecdotes. They include high-power laser photocoagulation and vitreous surgical techniques for chorioretinal venous anastomosis formation and radial optic neurotomy. More recently, intravitreal injections of various pharmacologic agents have been studied in eyes with macular edema due to CRVO. Intravitreal triamcinolone4 was associated with cataract formation, elevated intraocular pressure, and a rapid, often large, improvement in visual acuity in eyes with CRVO that was unfortunately short-lived, all of which was compounded by the need for continued injections. A randomized trial sponsored by the National Eye Institute, called the SCORE (Standard Care Versus Corticosteroid for Retinal Vein Occlusion) study, helped determine the efficacy of triamcinolone for macular edema in these eyes.
Subsequently, vascular endothelial growth factor (VEGF) inhibitors were noted to have a beneficial effect upon macular edema and visual acuity, without the risk of cataract or elevated intraocular pressure associated with triamcinolone.5–8 We know the retina has a constitutive secretion of VEGF, and a decrease in blood flow through tissue results in an increase in VEGF production. The following are known to be true about VEGF in these eyes: (1) vitreous samples from patients with CRVO show elevated VEGF; (2) severity of CRVO is related to VEGF level; and (3) intravitreal VEGF injection in primate eyes produced findings that mimic CRVO, including venous dilation and tortuosity, intraretinal hemorrhage, telangiectasis, and areas of nonperfusion. Subsequent histopathologic analysis demonstrated endothelial cell proliferation within capillaries and venules, leading to occlusion and nonperfusion.9 What appears to be regional nonperfusion in ischemic eyes may be the result of elevated VEGF levels, not something intrinsic to CRVO. These findings point to an alternate hypothesis implicating VEGF as an important cause of the clinical findings associated with CRVO, as opposed to simply being a sequelae of the occlusion itself.
Because VEGF may be a driving force in producing pathologic manifestations of CRVO, it is logical to use anti-VEGF agents to treat CRVO. Intravitreal Avastin (bevacizumab) was evaluated in an off-label fashion and benefits included visual acuity improvement, a decrease in macular edema, and a reduction in venous tortuosity, venous dilation, nerve swelling, and intraretinal hemorrhage in select patients. The SCORE2 study compared the efficacy of Eylea (aflibercept) with bevacizumab in eyes with CRVO.10 The trial found that bevacizumab was non-inferior to aflibercept. A gain of at least 15 letters was reported in nearly 2/3 of the eyes and a loss of 15 letters was seen in only 2%.10 The cost of bevacizumab is approximately 1/30 that of aflibercept, but has greater uncertainty in commercially available formulations for intraocular use and is not licensed for treatment of CRVO in some countries.
What Is the Best Way to Handle Central Retinal Vein Occlusion?
The diagnosis of CRVO is generally made by ophthalmoscopic examination and is often confirmed with FA, although this step is not necessary. FA shows that there is a slower than normal arm-to-eye time and delayed arteriovenous filling time, dilated and tortuous veins, dilated capillary segments, leakage from capillaries, staining and leakage from larger veins, and associated blocking defects from intraretinal hemorrhage. At one time, FA was a requisite test to help grade capillary nonperfusion, due to concern of the possibility of neovascular glaucoma. Today, with more mild forms of CRVO, especially when considering modern treatments, FA can be optional in some patients, provided some form of imaging documentation takes place upon diagnosis. Baseline OCT is helpful in providing an objective measurement of macular thickness and for the detection of any subretinal fluid and is used for comparison at future visits. OCT angiography (OCT-A) shows dilated and tortuous vessels with areas of capillary dilation in the superficial vascular plexus and areas of absent flow in the deep capillary plexus, particularly when there is cystoid macular edema.11 Some patients with CRVO may also have evidence of flow problems in a cilioretinal artery, if present. Neither FA nor OCT-A shows an absence of flow in the affected ciliochoroidal vessels in these circumstances, suggesting that either the flow is rapidly restored or that diminution of flow is all that is required to induce associated ischemic retinal whitening.
Once the diagnosis is established, the next step is to consider additional testing that may be needed.12–15 Medical work-up is not indicated for the vast majority of patients because many have well described risk factors and the results of hematologic or clotting evaluations often produce the same frequency of abnormalities as an age-adjusted normal population. Hypertension, diabetes mellitus, and glaucoma are well-known risk factors.12–15 Referral to an appropriate specialist is indicated for patients without regular medical care, unusually severe or bilateral cases, and patients in whom a review of systems indicates other potential medical problems. Questions about past deep vein thrombosis, spontaneous abortions, or family history of clotting abnormalities may provide clues to an underlying thrombophilic disorder, but these are relatively uncommon in retinal vein occlusion patients. There are 2 common scenarios in which CRVO may be found in otherwise healthy younger individuals. The first occurs after prolonged exercise, particularly in patients who may have been dehydrated (Figure 10-1). The second is for a CRVO picture to occur in a patient with pronounced optic nerve edema. This condition has been called thrombophlebitis, because of a suspected associated inflammation. Indeed, inflammatory cells can sometimes be seen in the overlying vitreous by OCT.
Anti-VEGF therapy is a good first-choice treatment given the low risk of side effects and large proportion of treated patients who experience visual improvement. Choice of agents depends on the availability of a dependable source of bevacizumab and country-specific licensing. The keys to successful anti-VEGF treatment are early and repetitive treatment.5–8 In a randomized trial, the patients assigned to delayed treatment had a worse outcome than patients treated immediately.6,8 As there is very little risk to anti-VEGF, and conversely little gained by waiting, prompt treatment is indicated in an eye with edema and decreased visual function (Figure 10-2). I use a treat-and-extend method of administering anti-VEGF agents in eyes with CRVO. Some patients can be weaned from anti-VEGF agents, while others cannot (Figure 10-3).