Currently, the strongest data that are available for therapeutic agents for branch retinal vein occlusion (BRVO) and central retinal vein occlusion (CRVO) lie with ranibizumab (Lucentis, Genentech) and the intravitreal dexamethasone implant (Ozurdex, Allergan), the 2 US Food and Drug Administrationapproved therapies for these indications.

This article reviews the available data on ranibizumab and the dexamethasone implant and also addresses the subgroup data that may help clinicians determine factors that may have affected outcomes in the BRAVO (ranibizumab for BRVO), CRUISE (ranibizumab for CRVO), and GENEVA (dexamethasone implant).

Ranibizumab

In the BRAVO and CRUISE studies, patients were randomized to either sham or 1 of 2 doses of ranibizumab. Injections or sham were given monthly for 6 months, and rescue laser was available to BRVO patients at month 3. The primary endpoint at 6 months was change in vision from baseline to 6 months. After this point, all patients were eligible for as-needed (prn) ranibizumab and were followed for the second half of the study for safety and longer-term data (Figure 1). The purpose of the BRAVO and CRUISE studies was to evaluate the efficacy and safety of intravitreal injections of ranibizumab in patients with macular edema following BRVO and CRVO, respectively.

During the BRAVO study, focal or grid laser for macular edema, the standard of care, was allowed after 3 months in the first 6 months, and again beginning at month 9 in the second 6 months. Patients who were treated with ranibizumab received an injection every 4 weeks in the first 6 months, and then prn in the second 6 months if visual acuity was 20/40 or worse, or if the central foveal thickness on optical coherence tomography was greater than 250 μm.

Top-line Data: BRAVO and CRUISE

The top-line data from BRAVO and CRUISE show that BRVO patients had better vision at all time points in both the treatment and sham groups compared with CRVO patients, suggesting a less aggressive natural history compared with CRVO. The BRVO patients treated with ranibizumab gained 18 letters compared with 7 in the sham group at 6 months.1 CRVO patients also fared better in the active treatment group, gaining 14 letters compared with barely 1 letter at 6 months.2 In both studies, the sham groups did recover some vision once they were allowed prn treatment in the second half of the study; however, the sham groups did not catch up to the treatment arms and did not achieve similar vision gains with treatment in the second 6 months (Figure 2).

BRAVO and CRUISE Subanalyses

Subanalyses were performed to determine what factors influenced outcomes in both BRAVO and CRUISE. The subanalyses included baseline/month 6 characteristics influencing outcomes including age, baseline vision, time since diagnosis, prior treatment for RVO, smoking status, gender, baseline/month 6 central field thickness on optical coherence tomography (OCT), baseline lens status, and whether rescue laser was needed in the treatment period; the impact of treating or not treating at month 6; a review of overall outcomes; and a review of adverse events.

The graphs in Figure 3 show gains in vision at 6 and 12 months for BRVO patients. The solid bars represent the sham vs active drug group at month 6, and the striped bars show the month 12 vision after the prn phase was initiated. Although the sham to prn patients improved vision after the 6 month prn therapy was initiated, the patients who gained the most vision during the study were those who had worse vision at baseline and had more fluid on OCT. Figure 4 is a similar presentation of vision data that is sorted by months since diagnosis and age at baseline. Patients who were treated within 3 months of diagnosis or who were younger than 65 years of age had an average benefit of 5 letters. The difference was less significant in patients who were observed or who crossed over to therapy after month 6.

Figure 5 shows the 6- and 12-month data for CRUISE. Patients also gained 1 more line of vision if they started with worse vision or more fluid on OCT. Note that patients who were started with sham drug did not gain vision as robustly as those who were receiving active treatment from the beginning of the study. In contrast to BRAVO, in the CRVO patients in Figure 6, the time since diagnosis made a less significant difference, but age younger than 65 years still provided at least a 1-line of vision advantage, and even more so in the sham-toactive drug patients.

Figure 7 is a waterfall plot depiction that shows each individual patient's change in vision from month 6 to 7 in the BRAVO study. The peaks at each end are the extremes of change, but the average ranges from a loss of 2.8 letters to a gain of 0.4 letters. The waterfall plot shows that by skipping the ranibizumab dose at month 6, many patients lost some of the improvements in visual acuity they achieved during the first 6 months of treatment at month 7. By continuing ranibizumab injection at month 6, some patients gained further improvement in visual acuity at month 7. These results, however, should be interpreted with caution, as the analysis did not adjust for possible confounding factors, such as baseline/month 6 visual acuity, impact of rescue laser prior to month 6, and/or selection bias in patients who received 0.5-mg ranibizumab at month 6 due to meeting prespecified criteria for prn treatment.

Similarly, Figure 8 shows the distribution of OCT retinal thickness changes between months 6 and 7. The peaks at each end are the extremes of change, but the averages of the data show a minimal difference between groups. The effect of holding a ranibizumab dose was more pronounced in anatomic outcomes than in visual outcomes. Most patients experienced an increase in central field thickness at 1 month after not receiving ranibizumab dose (at month 7), with a mean increase of 64 μm. By continuing ranibizumab injection at month 6, some patients achieved additional reduction in central field thickness, with a mean decrease of 14 μm at month 7.

Figure 9 is a waterfall plot from the CRUISE study showing that by skipping ranibizumab dose at month 6, most patients at month 7 lost the improvements in visual acuity they achieved during the first 6 months of treatment. Patients with CRVO had a more substantial difference between those treated and those not treated at month 6. Those treated still had a smaller letter gain in vision, while those not treated had a loss of 1.5 lines of vision. Looking at CRVO patients' visual acuity from month 6-7, there were some patients in the middle who did not have any changes in visual acuity, but more patients who were treated at month 6 gained vision. The difference was a statistically significant gain of nearly 2 letters vs loss of approximately 7 letters.

The change in OCT for the CRVO group was more significant than in the BRVO group. Most patients had similar OCTs from month 6 to month 7, but the averages of those treated vs those not treated was significant. Untreated CRVO patients had a mean gain of 200 um of macular fluid in just 1 month. The effect of holding ranibizumab dose in anatomic outcomes was prominent. The majority of patients who skipped a ranibizumab injection at month 6 lost their improvements in central field thickness over 6 months of treatment, with a mean increase of 200 μm at month 7. Most patients who continued to receive ranibizumab at month 6 experienced further reduction in central field thickness, with a mean decrease of 19 μm.

Predictors of Frequency of PRN Dosing

The BRAVO and CRUISE subanalyses used a statistical model to predict the number of prn injections given. A logistic regression analysis was used to evaluate baseline factors that may help predict a greater need for prn injections during the second half of the year. The baseline and month 6 factors evaluated in a stepwise regression model of prn injections included the following: age (years) at baseline; best corrected visual acuity (BCVA) at baseline and month 6; rescue laser at months 3, 4, or 5 (yes vs no); prior therapy for RVO (yes vs no); months from diagnosis to day 0; smoking (current vs nonsmoker) status at baseline; gender; OCT central field thickness at baseline and month 6; baseline lens status (phakic vs pseudophakic vs unreported); and baseline weekly alcohol consumption (none vs 1 vs ≥2).

The analyses found that patients with BRVO who required rescue laser in the treatment period had an odds ratio nearly 3 times that of patients who did not. Prior therapy increased the odds but was associated with a wide confidence interval.

When the same model was applied to the CRVO data, the signals were not as strong except for alcohol consumption. This factor was not well balanced across groups, and therefore analyses may not have been adequately powered to demonstrate a clinically meaningful effect.

BRAVO and CRUISE Subanalyses Conclusions

The subanalyses overall confirmed our instincts— patients who required fewer therapies were more likely to have a more mild RVO. Patients who had edema from a vein occlusion had a better and faster response to anti-VEGF treatment delivered on a frequent and regular basis. Additionally, younger patients and those who had very poor visual acuity and more fluid at baseline fared better in terms of visual acuity improvement. In BRVO, earlier treatment produced better results. Finally, CRVO patients had a more unpredictable course than BRVO patients, suggesting that these patients should be monitored more frequently and regularly.

Intravitreal Dexamethasone Implant

With the dexamethasone intravitreal implant, the steroid implant is preloaded into a single-use, specially designed applicator to facilitate injection of the rod-shaped implant directly into the vitreous. The kinetic profile of steroid release is biphasic, and the drug concentrations peak at 2-6 weeks after implantation. Dexamethasone is detectable in the vitreous for at least 6 months after injection. The duration of effect of the dexamethasone depends on individual patient characteristics.

The GENEVA study combined patients with BRVO and CRVO and had inclusion criteria similar to those in BRAVO and CRUISE. It had 3 arms: sham, low-dose (0.35 μm) dexamethasone, and high-dose (0.7 μm) dexamethasone. The primary endpoint was 6 months, and if patients met the retreatment criteria, they could receive a second injection at that time. OCTs were taken at baseline and then every 3 months.

Figure 10 illustrates the mean change from baseline BCVA throughout the study. Statistically significant visual acuity improvement peaked at 2 months with approximately 2 lines or 10 letters. This improvement did drop, but remained significant at month 6.

Top-line Data: GENEVA

The proportion of patients achieving at least a 15-letter improvement from baseline BCVA was significantly greater in the dexamethasone high-dose group than the sham group from day 30-90 (Figure 11). The greatest response was seen at day 60, when approximately 29% of patients in the dexamethasone high-dose group achieved at least a 15-letter improvement from baseline as compared with 11% in the sham group. In the dexamethasone high-dose group, the proportion of patients achieving at least a 15-letter improvement remained high on day 180. In fact, at 21.5%, this represents only a 0.3% decrease between day 90-180. However, the difference from the sham group was no longer statistically significant.3

In this study, 15% of patients did not meet the criteria for a second dexamethasone implant injection. Statistically, 15-20% of vein occlusions will resolve, so ongoing therapy may be unnecessary (Figure 12).

GENEVA Subanalysis

Figure 13 shows a cross-trial comparison among GENEVA, SCORE-BRVO, and BRAVO. At first glance at the top-line data, it appears that the anti-VEGF treatment is superior because the magnitude of visual acuity improvement is better.

The duration of BRVO and secondary edema was longer in GENEVA than in BRAVO by an average of over 2 months.

To assess the effect of duration of macular edema (≤90 days or >90 days duration at study entry) on visual acuity in subjects with BRVO following treatment with the dexamethasone implant, a post-hoc analysis of pooled data from the GENEVA clinical trials was performed. For inclusion in this analysis, patients were required to be aged 18 years or older, with decreased BCVA due to macular edema involving the fovea secondary to BRVO. Visual acuity was between 20/200 and 20/50, and retinal thickness was greater than or equal to 300 μm. The duration of macular edema in the subanalysis was between 6 weeks and 12 months. All patients with BRVO were treated with intravitreal 0.7-mg dexamethasone implant and were broken down into 2 groups: those with macular edema for less than 3 months (n=50) and those with macular edema for more than 3 months (n=241).

Figure 14 shows the percentage of BRVO subjects who gained 3 or more lines (15 letters or more) over the course of 6 months. The more acute vein occlusions are represented in yellow. The outcomes of patients who were treated earlier vs those who had vein occlusions for over 90 days support an early treatment approach in macular edema. This is even more clear in Figure 15, which illustrates the effect that each month of duration of BRVO with macular edema has on the ability to achieve 15-letter BCVA improvement at 6 months. Overall, earlier intervention improves the chances of a better prognosis. The chances of gaining 3 lines or more are worse if the duration is 12 months vs just 1 month.

Conclusions

Macular edema duration is a significant predictor of BCVA improvement, particularly in BRVO. Patients with branch vein occlusion did better if the duration of their edema was less than 90 days. These results suggest a greater clinical benefit in these types of patients. The implication of this is that, from a public health standpoint, it is best to inform patients with BRVO that earlier treatment is better.

Carl D. Regillo, MD, is the Director of the Retina Service of Wills Eye Institute and a Professor of Ophthalmology at Thomas Jefferson University, Philadelphia. Dr. Regillo states that he receives research grant support from Genentech, Regeneron, Inc., and Allergan, Inc.; and is a consultant to Genentech, Regeneron, Inc., and Allergan, Inc. He may be reached at +1 800 331 6634; or via email at cregillo@aol.com.

  1. Campochiaro PA, Heier JS, Feiner L, et al. Ranibizumab for macular edema following branch retinal vein occlusion: six-month primary end point results of a phase III study. Ophthalmology. 2010;117(6):1102-1112 e1.
  2. Brown DM, Campochiaro PA, Singh RP, et al. Ranibizumab for macular edema following central retinal vein occlusion: six-month primary end point results of a phase III study. Ophthalmology. 2010;117(6):1124-1133 e1.
  3. Haller JA, Bandello F, Belfort R, Jr., et al. Randomized, sham-controlled trial of dexamethasone intravitreal implant in patients with macular edema due to retinal vein occlusion. Ophthalmology. 2010;117(6):1134-1146 e3.