PEER REVIEWED

Anti-VEGF Therapy for Macular Edema Secondary to CRVO

A review of the data and ongoing head-to-head clinical trials

LUKE NICHOLSON, MD • SOBHA SIVAPRASAD, MD • PHILIP HYKIN, MD

Central retinal vein occlusion is one of the commonest retinal vascular diseases. The prevalence in the adult population is 0.4%, increasing to 1% at 80 years or older, and it remains a significant cause of visual loss.¹

The most common cause of visual impairment in CRVO is macular edema. In addition, consequences of significant retinal nonperfusion may result in significant visual loss, requiring urgent attention. The burden of CRVO on the population has led to numerous studies investigating its natural history and possible future treatments.

NATURAL HISTORY

One of the seminal papers on CRVO is the Central Vein Occlusion Study, which reported the natural history of CRVO, based on presenting visual acuities. Patients presenting with VA of 20/40 or better, although having a good chance of maintaining a good level of vision, had a 45% chance of worsening vision.²

The outcomes of those presenting VAs of 20/50 to 20/200 were variable, with only 19% improving and the rest either maintaining the presenting VA or worsening.² Patients with poor presenting VA of 20/200 or worse had poor outcomes, with 80% not improving.²

GRID LASER

Grid laser photocoagulation was considered following its success in managing diabetic macular edema. Despite angiographic evidence of decreased macular edema, the visual outcome was not favorable, with no significant effect on the final VA.^3,4

INTRAVITREAL STEROID THERAPY

The failure of grid laser led to trials of intravitreal steroid agents. Initially, triamcinolone acetonide was introduced as a drug to treat macular edema from CRVO. Early evidence came from case reports and case series with promising outcomes.^5-7

Good level evidence in the form of a prospective randomized, controlled trial, the SCORE study, evaluated the safety and efficacy of intravitreal triamcinolone as a first-line treatment for this condition.

The SCORE study was designed to study the safety and efficacy of a preservative-free preparation of triamcinolone in treating macular edema.⁸ This large National Eye Institute-funded trial compared observation with two doses of triamcinolone, 1 mg and 4 mg, given four times monthly. Ophthalmic assessments were performed every four months with the primary outcome reported at 12 months.

The study found 25% to 27% of patients in the treatment arm managed to gain 15 or more letters, compared to only 6.8% in the observation arm. For the first time, a therapeutic option was made available to improve the outcome of this previously irreparable condition.

However, the mean VA change was minimal with 1- to 2-letter loss but a significant improvement compared to the control arm, which had a mean loss of 12 letters. The results were confounded by cataract development and progression in 26% to 33% of patients receiving treatment.

GENEVA

The GENEVA trial compared an intravitreal dexamethasone sustained-release implant (Ozurdex, Allergan, Irvine, CA) to observation.^9,10 GENEVA also included patients with branch RVO but also reported substudy results of patients with CRVO.

Interestingly, the best response was noted at the two-month mark, with 29% achieving a 15-letter or more gain in VA in the 0.7-mg dexamethasone arm. This initial improvement declined to 18% at both 90 and 180 days. The associated mean gains in VA were approximately 8 letters at 60 days, declining to 5 and 0 letters at 90 and 180 days, respectively.

In the 12-month data from the GENEVA trial, in which open-label retreatment with 0.7-mg dexamethasone was given to all patients (including the initial sham group) after six months, visual outcomes show a similar trend.

However, patients receiving sham treatment in the first six months did not experience similar visual gains to patients treated from the start of the study. Dose-related cataract development and progression were also observed in 10% to 30% of patients at 12 months.

ANTI-VEGF THERAPY

The introduction and success of monoclonal antibodies and fusion proteins targeting VEGF in neovascular age-related macular degeneration has led to its use in other indications, such as myopic choroidal neovascularization and DME.^11-13

Several studies have evaluated the three intravitreal anti-VEGF agents, namely bevacizumab (Avastin, Genentech, South San Francisco, CA), ranibizumab (Lucentis, Genentech), and aflibercept (Eylea, Regeneron, Tarrytown, NY) in CRVO.^14-17

CRUISE and CRYSTAL. The CRUISE and CRYSTAL trials investigated the use of intravitreal ranibizumab treatment for macular edema secondary to CRVO with different treatment regimens. In CRUISE, patients with brisk relative afferent pupillary defect (RAPD) were excluded, and only two patients of the 392 included had more than 10 disc areas of retinal nonperfusion.

The outcomes reported suggested a significant benefit, with 46.2% to 47.7% achieving 15 letters or more at six months with monthly treatment of either 0.3 or 0.5 mg of intravitreal ranibizumab.¹⁴

At 12 months of the CRUISE study with a PRN treatment regime introduced in the second six months, the benefits continued to improve in 47.0% to 50.5% of patients achieving a 15-letter or more gain in the treated group. The mean gain in vision at 12 months was 12.9 letters.

The sham group did receive treatment on a PRN basis after six months and showed a mean improvement of 7.3 letters.¹⁸ The six-month delay resulted in inferior VA gains, compared to the patients who were initiated on prompt treatment.

Moreover, this group was initiated on PRN dosing, so the treatment regime was not as intensive compared to that given to patients receiving treatment in the first six months. Therefore, the results in this group indicated that patients should be treated promptly.

VIEW and more. There have been limited prospective, randomized, controlled trials investigating the efficacy of bevacizumab for this specific indication. Epstein et al reported six-month outcomes from a prospective, randomized, controlled trial with patients receiving bevacizumab every six weeks, compared to observation.

The mean visual gains were 7.5 letters at six weeks, 11.4 at 12 weeks, 13.9 at 18 weeks, and 14.1 at the final visit. This slower response may reflect the six-week dosing interval used in this study. Despite this fact, 60% of patients on treatment achieved a 15-letter or better vision gain, compared to 20% in the control group at six months.¹⁵

Following the promising results from the VIEW studies in AMD, the efficacy of intravitreal aflibercept for CRVO was evaluated in two large, multicenter, randomized clinical trials: COPERNICUS and GALILEO.^16,17,19 These two trials investigated the use of monthly aflibercept treatment for six months, and the primary outcome was at six months.

The sham arm in the COPERNICUS trial received aflibercept in the second six months, unlike in the GALILEO study, in which the sham period was extended to 12 months. The mean VA gain with monthly treatment was 17.3-18.0 letters, and 56.1% to 60.2% achieved a visual gain of 15 letters or more.^16,17,20,21

Table 1 (next page) summarizes the visual outcomes using intravitreal steroids, ranibizumab, aflibercept, and bevacizumab at both six and 12 months, along with the results of the corresponding control groups. A direct comparison of the superiority or inferiority of any one drug compared to another is not possible given the differing eligibility criteria, baseline characteristics, and retreatment criteria used (Table 2).

The proportion of nonperfused eyes recruited in CRUISE was very low because a brisk RAPD was an exclusion criterion. This criterion differed significantly from COPERNICUS, in which 32% were nonperfused at baseline.

The study by Epstein et al appeared to have a higher percentage of eyes with poorer VA at baseline. The effects of these variations are unclear. A poor baseline VA may result in a greater VA gain, as reported in the COPERNICUS subgroup analysis.

Furthermore, the cost for each drug varies, with bevacizumab available at $50 per dose, while ranibizumab and aflibercept cost $1,850 to $1,950, respectively.²³ Therefore, the ideal option, both clinically and from a cost-effective perspective, is a difficult decision given individualized studies performed on a single drug of choice.

The CRAVE study attempted to compare the efficacy of ranibizumab with bevacizumab in the treatment of macular edema secondary to RVO. It found no significant difference in VA gain or decrease in central retinal thickness.²⁴ However, this study was not limited to CRVO, and it included all forms of RVO.

A large, prospective, randomized, head-to-head study comparing the efficacy of these drugs in the context of macular edema from CRVO is required. Two large multicenter, prospective, randomized noninferiority trials are currently recruiting for this specific reason.

SCORE2. The SCORE2 study is currently ongoing in the United States. The team is comparing four weekly intravitreal bevacizumab with four weekly intravitreal aflibercept injections in a targeted 362 participants.

The primary objective of the SCORE2 study is to determine whether the VA change at 6 months of bevacizumab is noninferior to aflibercept in the treatment of macular edema from CRVO. The inferiority margin is set at 5 letters.

Secondary objectives include the change in central subfield thickness between the two arms at month 6. Other than the usual parameters studied, the SCORE2 group has also set out to assess the outcomes at month 12 with regard to different dosing strategies after month 6 for participants who respond well. These different dosing regimens include randomization to continued monthly treatment or to a treat-and-extend regimen with two-week increments.

For participants who respond poorly to aflibercept after 6 months, this group will receive the intravitreal dexamethasone implant, while patients who respond poorly to bevacizumab will receive aflibercept.

With recent literature suggesting that anti-VEGF modifies the progression or even promotes retinal reperfusion in CRVO, there is an additional need to study the area of nonperfusion, which is being studied in SCORE2, along with the risk of neovascular conversion.²⁵

The cost-effectiveness of aflibercept and bevacizumab and the economic implications will also be studied in the SCORE2 study.

LEAVO. In the United Kingdom, a multicenter, phase 3, double-masked, randomized, controlled noninferiority trial comparing the clinical and cost-effectiveness of intravitreal therapy with ranibizumab vs aflibercept vs bevacizumab for macular edema due to CRVO — the LEAVO study — is currently ongoing with a targeted recruitment of 459 patients.

No exclusions are being made of ischemic or nonperfused eyes, with a lower limit of VA of 19 letters. Patients are randomized to receive bevacizumab, ranibizumab, or aflibercept, and the participants will be followed up for 24 months.

The treatment regimen is a mandated course of four injections given four weeks apart, followed by an as-needed phase, guided by retreatment criteria of VA gain of 6 letters or more, VA loss of 6 letters or more associated with an increase in central subfield thickness, central subfield thickness of more than 320 µm, and central subfield thickness increase of 50 µm or more, compared to the lowest measurement. The primary objective of the study is to determine whether bevacizumab is noninferior to ranibizumab or aflibercept in treating visual impairment due to macular edema from CRVO.

The secondary objectives are to determine the differences among the three arms for change in VA at 12 months, proportions with ≥15- and ≥10-letter improvements at 12 and 24 months, proportions with ≥73 letters at 12 and 24 months, changes in central subfield thickness and macular volume on optical coherence tomography, the relative cost-effectiveness, and changes in the area of nonperfusion.

CONCLUSION

The results of the two large clinical trials, SCORE2 and LEAVO, will enhance our understanding of the relative clinical and cost-effectiveness of the various drugs that are available for macular edema secondary to CRVO. RP

REFERENCES