Current Best Clinical Practices—Management of Retinal Vein Occlusion

Abstract

Retinal vein occlusion (RVO) is the second most common cause of vision loss from retinal vascular diseases in adults in the United States. Visual loss arises as a result of a host of factors, including macular ischemia and macular edema. Primary antivascular endothelial growth factor therapy is the current standard of care, with level I evidence demonstrating sustained visual gains up to 2 years after treatment in both branch and central RVO. Prompt antivascular endothelial growth factor therapy is important because delays in treatment yield lesser visual gains. Steroid therapy also improves visual outcomes in RVO but with higher rates of adverse effects, including cataract formation and ocular hypertension. Although the treatment burden can be high, these drugs have collectively revolutionized treatment outcomes in this disease state, providing improved visual outcomes over previous laser therapies.

Introduction

Retinal vein occlusion (RVO) is the second most common cause of vision loss from retinal vascular disease in adults in the United States following diabetic retinopathy. ¹ The type of RVO is characterized based on the site of venular occlusion. A central RVO (CRVO) ensues after obstruction of the retinal vein posterior to the optic nerve head, commonly within the lamina cribrosa. A hemi-RVO (HRVO) occurs as a result of an occlusion at the level of the optic disc of 1 of the 2 trunks of the central retinal veins that drain the superior or inferior retinal hemisphere. A branch RVO (BRVO) occurs secondary to occlusion of a branch of the central retinal vein, usually at the site of arteriovenous crossing. ²

The pathogenesis of CRVO and HRVO may be due to complete or partial venous occlusion caused by external structures (eg, an adjacent sclerotic artery), primary venous wall disease (eg, inflammatory disease), or hemodynamic disturbances (eg, hypercoagulable states), which commonly leads to thrombus formation. The occlusion results in increased intraluminal venous pressure with subsequent intraretinal hemorrhages, edema, and ischemia. This, in turn, triggers a complex molecular cascade resulting in upregulation of vascular endothelial growth factor (VEGF) and a host of additional proinflammatory mediators. ^3,4

Elevated intraocular pressure and glaucoma are ophthalmic risk factors for CRVO and HRVO, whereas systemic risk factors for RVO include arterial hypertension, older age, diabetes mellitus, dyslipidemia, ischemic heart disease, peripheral vascular disease, cerebrovascular stroke, blood dyscrasias, hypercoagulable states, smoking, and systemic inflammatory disease. ⁵ RVO is common after age 65 years, particularly in those with a history of cardiovascular disease, hypertension, diabetes, or glaucoma. Systemic management involves optimizing blood pressure, lipid, and glycemic control. However, if RVO occurs in a young patient, with bilateral RVO, or with a personal or family history of thrombosis, a systemic evaluation for coagulation disorders may be indicated. ⁶ Although there is no cure for RVO, management is aimed at treating complications that cause vision loss, including macular edema (ME), retinal neovascularization, and anterior segment neovascularization. There is no treatment as of 2019 for macular or retinal ischemia, an unmodifiable cause of vision loss due to RVO.

Methods

Grid and Scatter Laser in RVO

In pre–anti-VEGF era therapies, laser was the sole treatment to address ME, with differential responses noted for BRVO vs CRVO. Whereas the Branch Vein Occlusion Study in 1984 demonstrated a modest improvement of vision in eyes receiving grid laser photocoagulation (0.1 second, 100 μm) vs observation (63% of treated eyes gained ≥ 2 lines of vision at 3 years, compared with 36% of untreated eyes), ⁷ the Central Vein Occlusion Study did not demonstrate a similar benefit of macular grid laser over observation. ⁸ This finding is mostly of historical reference now because laser photocoagulation for ME has largely been replaced by the criterion standard of anti-VEGF therapy followed by second-line intravitreal steroid therapy.

What remains relevant today is the timing of panretinal and sector photocoagulation to address neovascular complications. The Branch Vein Occlusion Study evaluated laser scatter photocoagulation for the prevention of neovascularization and vitreous hemorrhage in BRVO. It and found that laser treatment applied preemptively decreased the risk of neovascularization from 31% to 19% in eyes with more than 5 disc diameters of retinal ischemia, but because 69% of patients who would not develop neovascularization would be unnecessarily treated, the study recommended applying laser only after the development of neovascularization. ⁹ Similarly, the Central Vein Occlusion Study found that panretinal photocoagulation was beneficial only if applied after the development of anterior segment neovascularization. ¹⁰ For more than 30 years, these pivotal studies have shaped our approach away from prophylactic photocoagulation to adjunctive laser photocoagulation when neovascularization is present.

Anti-VEGF Therapy in RVO

Anti-VEGF therapy has led to better treatment of RVO compared with laser, demonstrating significant reductions of ME, improved visual acuity (VA), and decreased neovascular complications. ^11,12 Given the efficacy and favorable side effect profile, it is now first-line therapy worldwide. ^13,14 Ranibizumab and aflibercept are the 2 most commonly used agents approved by the United States Food and Drug Administration (FDA) for the treatment of ME in RVO. Compounded, off-label bevacizumab is also widely administered.

Ranibizumab: BRAVO and CRUISE Studies

The BRAVO ¹¹ and CRUISE ¹² randomized, controlled trials led to FDA approval of monthly intravitreal ranibizumab 0.5 mg for the treatment of ME secondary to BRVO and CRVO, respectively. This was because of significant visual gains with monthly therapy of 0.5-mg ranibizumab vs sham therapy. In CRUISE, patients receiving monthly 0.5-mg ranibizumab gained 14.9 Early Treatment of Diabetic Retinopathy Study letters at 6 months relative to 0.8 letters in patients receiving sham injections. ¹² Similarly, in BRAVO, patients receiving monthly 0.5-mg ranibizumab gained 18.3 letters at 6 months compared with 7.3 letters in patients receiving sham treatment with rescue laser 90 days after enrollment. ¹¹

Patients in the BRAVO and CRUISE studies were followed for an additional year in an open-label extension study (HORIZON). ¹⁵ Contrary to the monthly therapy that was used in the FDA registration trials, these patients were followed every 3 months and treated on an as-needed basis. The authors reported reduced follow-up and fewer intravitreal injections of ranibizumab in the second year, resulting in a decline in vision in patients with CRVO (–4.1 letters) but stability in those with BRVO (–0.7 letters). Patients with RVO likely benefit from an individual follow-up and management paradigm that involves more frequent follow-up and treatment than every 3 months. ¹⁵

Aflibercept: COPERNICUS, GALILEO, and VIBRANT Studies

The approval of ranibizumab was followed shortly by the approval of aflibercept 2 mg for the treatment of ME secondary to RVO. The parallel COPERNICUS and GALILEO studies demonstrated rapid and durable improvement of VA in patients receiving monthly aflibercept 2 mg for the treatment of ME secondary to CRVO. ^16,17 Similarly, the VIBRANT study, which randomly assigned patients to monthly aflibercept 2 mg through 6 months vs macular laser photocoagulation, resulted in FDA approval of aflibercept for the treatment of ME secondary to BRVO.

Delay in Treatment

In all the ranibizumab and aflibercept registration trials, patients in the sham group (observation or laser) were eligible to switch to the study drug at 24 weeks. ^{11,12,15 -17} Although all patients demonstrated improvements in letters gained and reductions in central macular thickness, patients uniformly demonstrated inferior VA results at 52 weeks relative to patients who were initiated on anti-VEGF therapy at the outset. From the lessons learned in these registration trials, we recognize that it is imperative to begin immediate therapy for patients with ME secondary to RVO.

Off-Label, Compounded Bevacizumab

In addition to the use of ranibizumab and aflibercept, off-label bevacizumab has been used for the treatment of ME secondary to RVO. ^{18 -20} In the United States and worldwide, patients are frequently started first on bevacizumab because of the significant cost difference and similar efficacy when dosed similarly to the FDA-approved anti-VEGF therapies. ^18,19 However, this medication requires compounding, which raises concern about sterility, silicone oil droplets from prefilled syringes, and shelf-life stability of a compounded injectable. ^21,22 Despite these challenges, the medication has demonstrated an acceptable safety profile ²² and is used by more than 70% of the US and international population per the 2018 Preferences and Trends Survey. ²³

Comparative Effectiveness Studies of Anti-VEGF Treatments: SCORE2 and LEAVO Studies

In 2018, the National Eye Institute–funded SCORE2 (Study of Comparative Treatments for Retinal Vein Occlusion 2) trial evaluated the efficacy of bevacizumab vs aflibercept for the management of ME secondary to CRVO and HRVO. ²⁴ The study concluded that intravitreal monthly bevacizumab was noninferior to (no worse than) aflibercept with respect to the primary outcome of VA at 6 months. Anatomically, there was more residual ME in the bevacizumab arm, but this did not result in a visually significant difference. ²⁵

A preplanned secondary subgroup analysis examined the outcomes at 12 months after a switch in treatment at 6 months for patients who were protocol-defined as poor responders, in which those in the monthly aflibercept arm were switched to a dexamethasone implant (DEX) (Ozurdex, Allergan, Inc). Those in the monthly bevacizumab arm were switched to 3 monthly aflibercept injections and then to treat-and-extend aflibercept. The sample size was limited at 35 patients, but there were statistically significant improvements in both VA and central subfield thickness in those who responded poorly to bevacizumab and were switched to aflibercept. ²⁶ Steroids are discussed separately as follows.

Furthermore, the LEAVO study was a multicenter, phase 3, double-masked, randomized, controlled noninferiority trial comparing the clinical efficacy and cost-effectiveness of intravitreal therapy with ranibizumab, aflibercept, and bevacizumab for ME due to CRVO. The study was supported by the National Institute for Health Research Health Technology Assessment Program and supported by National Institute for Health Research Moorfields Biomedical Research Centre. Patients in the study were randomized 1:1:1 to receive intravitreal bevacizumab, aflibercept, or ranibizumab given by four mandated monthly injections followed by therapy every 4 to 8 weeks per pre-specified re-treatment criteria until week 96. The primary outcome was change in best-corrected VA from baseline to 100 weeks. ²⁷

In all arms, there was substantial and sustained improvement in VA at weeks 52 and 100. With respect to noninferiority for VA improvement after 100 weeks of treatment, bevacizumab was inferior to ranibizumab at 100 weeks, bevacizumab was inferior to aflibercept at weeks 52 and 100, and aflibercept was noninferior to ranibizumab but not superior. The mean gain in best-corrected VA at 100 weeks was 15.1 letters for aflibercept, 12.5 letters for ranibizumab, and 9.8 letters for bevacizumab. The authors concluded that bevacizumab may not be interchangeable with aflibercept or ranibizumab. ²⁸

Treatment Schedules

The FDA registration trials that have approved ranibizumab and aflibercept require fixed dosing with frequent follow-up that may be time-intensive for the patient and physician alike. In search of finding a balance between maximizing therapeutic benefit and minimizing treatment burden, clinician practice patterns have evolved to different dosing regimens. ²⁹ Retina specialists may apply an as-needed or treat-and-extend paradigm that was first described for neovascular age-related macular degeneration. ³⁰ The decision to opt into one treatment paradigm vs the other is frequently driven by clinician and patient preference, because all treatment paradigms demonstrate favorable evidence when a strict follow-up paradigm is used.

The SHORE study evaluated patients originally enrolled in BRAVO and CRUISE. ³¹ After 7 monthly ranibizumab treatments, patients were randomly assigned to as-needed injections vs continued monthly treatment. At month 15, there were no differences in VA between patients receiving as-needed treatments vs continued monthly therapy. Unlike the HORIZON study, however, these patients were followed monthly to ensure timely retreatment if the ME was to return. ³¹

In SCORE2, 293 participants who responded well to bevacizumab or aflibercept during the first 6 months of the clinical trial were subsequently randomly assigned to either continued monthly treatment or a treat-and-extend schedule, carrying forward their respective anti-VEGF agents. ³² This was a preplanned subgroup analysis that showed that 1 or 2 fewer injections in the treat-and-extend arms provided similar visual outcomes compared with monthly treatment during the 6-month extension period of the trial. Although additional prospective treat-and-extend studies are limited to small cohorts of patients, they uniformly demonstrate improved vision with decreased injection burden. ^33,34

Collectively, these studies support the diversity of treatment paradigms used by clinicians. The critical elements that support the success of these management strategies over the ones applied in HORIZON, however, include monthly reevaluation in as-needed treatment strategies and optical coherence tomography–driven treatment extensions when using a treat-and-extend paradigm.

Results

Long-Term Anti-VEGF Outcomes

The data evaluating long-term outcomes in patients with RVO are limited. The RETAIN study, however, evaluated patients out to 5 years and found that 50% of 34 patients with BRVO and 56% of 32 patients with CRVO required ranibizumab injections 4 years after therapeutic onset. ³⁵ Similarly, another recent prospective study with a mean follow-up of 58 months for 50 eyes with BRVO and 78 months for 40 eyes with CRVO showed that initial VA gains peaked, and then declined by the final visit, primarily because of persistent or recurrent edema. ³⁶ This underscores the importance of long-term follow-up and continued treatment to achieve enduring visual outcomes by controlling ME.

Steroid Therapy in RVO

Besides anti-VEGF agents, intravitreal corticosteroids are used in the treatment of ME for RVO. ^13,14 Because adverse effects include cataract formation and ocular hypertension, intravitreal corticosteroids are generally recommended as a second-line option after anti-VEGF therapy. Commonly used corticosteroids include off-label, preservative-free triamcinolone acetonide (TA) and the on-label DEX.

The SCORE study was a landmark National Eye Institute–funded, randomized, clinical trial comparing 2 doses of TA, 1 mg and 4 mg, vs the standard of care at the time (macular grid laser for BRVO and observation for CRVO) in 411 individuals with BRVO and 271 with CRVO. ^37,38 These studies demonstrated opposing results, favoring macular grid laser photocoagulation over TA in BRVO but TA over observation in CRVO.

The primary outcome for both BRVO and CRVO cohorts was the proportion of patients with a gain of 15 or more letters in VA at 1 year. This was met in 29%, 26%, and 27% of eyes receiving laser photocoagulation, 1-mg TA, and 4-mg TA, respectively, for BRVO. Given the similar efficacy between groups, coupled with the adverse side effects of ocular hypertension (41% required intraocular pressure–lowering medication) and cataract progression in 35% of eyes, the authors suggested that TA was not an ideal treatment for ME secondary to BRVO. ³⁷ Contrarily for eyes with CRVO, the proportion of eyes gaining 15 letters or more at 1 year was 7%, 27%, and 26% for observation, 1-mg TA, and 4-mg TA, respectively. Given the higher rates of ocular hypertension in the 4-mg TA group relative to the 1-mg TA group, the authors recommended the use of 1-mg TA for the treatment of ME secondary to CRVO. ³⁸

The GENEVA study was taking place during the late 2000s also, which led to FDA approval of DEX in 2009 for the treatment of ME secondary to BRVO and CRVO. GENEVA was a 6-month, randomized, clinical trial comparing sham vs a single injection of 1 of 2 doses of DEX (0.35 mg or 0.7 mg) in 1267 eyes with ME secondary to BRVO (66%) or CRVO (34%). ³⁹ The primary outcome was distinct from SCORE and anti-VEGF trials because it assessed time to reach a 15-letter gain in VA relative to baseline. The time to achieve this improvement was significantly faster in both DEX-implant groups relative to sham for both BRVO and CRVO. The mean increase in VA was greatest between 30 to 90 days in both groups. This improvement in VA was not seen at 180 days, however. ³⁹ Collectively, this indicates therapeutic efficacy is maximal around 1 to 3 months with waning effects thereafter.

An open-label extension of the GENEVA study demonstrated that a repeat injection at 6 months improved VA up to 3 months after initial injection. ⁴⁰ Not surprisingly, similar to the SCORE study, ocular adverse effects included cataract and ocular hypertension that were more common at 1 year than 6 months. ^39,40

In the SCORE2 extension study mentioned previously, poor responders in the aflibercept arm were converted to DEX injections at 6 months. ²⁶ Most participants responded well to aflibercept, so this represented only 14 patients. There were no statistically significant changes in vision or central subfield thickness, but the inadequate sample size from this SCORE2 subgroup analysis is of limited value.

Despite the higher adverse-effect profile of steroids relative to anti-VEGF agents, there remains a role of steroids in the management of ME secondary to RVO. Benefits include an enduring reduction of central retinal thickness and sustained visual gain out to 3 months. ³² Given this, it is reasonable to repeat a DEX injection at 3 months to maintain visual and anatomic gains, rather than follow the treatment paradigm applied in GENEVA. For patients who travel from afar, demonstrate suboptimal responses to anti-VEGF therapy, or wish to minimize treatment burden, this is an excellent treatment option as long as the ocular hypertension and cataract risks are appropriately discussed and managed, whether for TA or DEX. The contraindications for DEX should be remembered, which include advanced glaucoma and a disrupted posterior capsule (ie, aphakic eyes, secondary intraocular lens including anterior chamber intraocular lens).

Conclusions

Treatment paradigms in RVO have evolved tremendously over the last decade. There is level I evidence supporting the use of all available anti-VEGF agents for RVO with ME. Ranibizumab and aflibercept are FDA approved, and bevacizumab use remains off label. Prompt therapy is indicated in both BRVO and CRVO with vision loss due to ME because delays in initiation of treatment may limit visual gains. In a search to maximize vision outcomes while minimizing a patient's treatment burden, retina specialists have evolved treatment to include as-needed and treat-and-extend paradigms, with comparable results in smaller clinical trials. Approximately half of patients will require anti-VEGF therapy 5 years after disease onset, underscoring the importance of regular and long-term follow-up. Whereas clinical trials provide high-level evidence supporting primary anti-VEGF therapy as an ideal treatment strategy for RVO, real-world outcomes may not always replicate clinical trial results. Therefore, an individualized treatment plan at the treating retina specialist’s discretion may be needed for some patients, and can include a steroid-driven approach, combination anti-VEGF and steroid therapy, or even macular or peripheral laser.