Arecent study has estimated that the 16 million people worldwide have retinal vein occlusion (RVO).1 The standards of care for branch retinal vein occlusion (BRVO) and central retinal vein occlusion (CRVO) have long been guided by data from two studies, the Branch Vein Occlusion Study (BVOS) and the Central Vein Occlusion Study (CVOS). The former is from 1984 and the latter from 1995.
As the second leading cause of blindness for patients with retinal vascular disease, there is currently an unmet need for patients with these diseases. Recently, there have been several studies completed that test alternative therapies to laser and observation for BRVO and CRVO.
BRVO
Most RVOs are BRVOs. BRVO involves the distribution
of one retinal vein and is characterized by inner retinal
hemorrhages, retinal edema, cotton-wool spots, dilated
and tortuous vein, and a corresponding artery that is narrowed
and sheathed. BRVO occurs at the artery/vein
crossing where the artery/vein bind together in a common
sheath where the arterial thickening compresses the vein.
The turbulent blood flow leads to thrombus formation.
The risk factors for BRVO include age (most common in people age 60 to 69 years), hypertension, cardiovascular disease, increased body mass index at 20 years of age, and glaucoma. Diabetes is not an independent risk factor. The guidelines set forth by the Branch Vein Occlusion Study (BVOS) are to allow 3 months for improvement of the BRVO. If after 3 months vision is 20/40 or worse, light grid pattern laser spots should be applied to the involved sector of the retina.2
The results of the BVOS showed that of those eyes that were treated with laser according to protocol, the average gain in vision was 1.33 lines and that those who were treated with laser were more likely (65%) to gain two lines of vision than those who were not treated (37%).2
CRVO
CRVO is relatively easy to diagnose and most physicians
can make this determination within their first week of residency.
Patients will often have dilated retinal veins, hemorrhaging,
and edema in all four quadrants of the eye; the
onset of symptoms is usually rapid. Historically, CRVO has
been categorized into ischemic, non-ischemic, and intermediate
based on the findings from the Central Vein
Occlusion Study (CVOS),3,4 data regarding the amount of
nonperfusion in the central retina. We have since learned
that there is a substantial difference in nonperfusion and
that the peripheral retina is often significantly nonperfused
independent of what occurs in the macula.
The pathogenesis of CRVO is thrombosis of the central retinal vein at or posterior to the lamina cribrosa. The artherosclerotic central retinal artery impinges on the central retinal vein, causing turbulent blood flow and leading to thrombosis.
The risk factors for CRVO are hypertension, diabetes, and glaucoma. I often tell my patients, however, that there are four risk factors: high blood pressure, diabetes, age, and bad luck. I further explain that although nothing can be done to prevent the third and the fourth risk factor, the first two should be kept in check. CRVO in young patients, however, requires more extensive workup to determine the cause of disease.
Because some of these patients actually have clotting issues, we recommend that patients younger than 50 years of age see a hematologist for a workup. It is, as seen in Table 1, an extensive workup, and so patients without insurance often receive a more tailored approach. If a patient has any of these diseases, most are treated with an aspirin once a day; however, this may cause more hemorrhaging early on in these CRVO patients.
In the presence of vascular disease, whether it is CRVO or BRVO, I will often make an analogy to a leaky water pipe: if you turn on the pressure, the leakage increases. Thus, in the management of CRVO or BRVO, any increase in hypertension despite treatment will result in more edema and a need for further pharmacologic or laser intervention.
In the CVOS, grid laser was applied; it appeared to decrease macular edema anatomically, but did not improve visual acuity. Thus, the recommendations from CVOS were observation until natural resolution.5
The results of several studies recently have been released: SCORE (Standard Care vs Corticosteroid for Retinal Vein Occlusion) for BRVO and CRVO;6,7 the 12-month reinjection data with the dexamethasone sustained delivery device (Ozurdex, Allergan, Inc.);8 BRAVO (A Study of the Efficacy and Safety of Ranibizumab Injection in Patients With Macular Edema Secondary to Branch Retinal Vein Occlusion);9 and CRUISE (Study of the Efficacy and Safety of Ranibizumab Injection in Patients with Macular Edema Secondary to CRVO).10 As a result, we now have level I evidence for from five major clinical trials for the first time since 1984 for BRVO and 1995 for CRVO.
SCORE DATA
In SCORE-BRVO,6 steroids took an early lead with
improvements in visual acuity at the 4-month mark with
the higher dose (4 mg specially prepared, commercially
unavailable preservative-free triamcinolone acetonide).
Note that by just after 1 year, the paths had crossed and
the grid laser patients continued to improve up to over
11 letters. Not all patients reached the 36 months because
the study was stopped early. The difference between threeline
gainers in the steroid group vs the laser group was not
significant. In regard to three-line losers, the 4-mg triamcinolone
group had a slight edge over the laser group at 1
year. The largest difference between the groups was seen
with the need for intraocular pressure (IOP)-
lowering medications. Of the patients in the 4-mg triamcinolone
group, 41% had to take IOP-lowering medications
vs a minimal amount of patients in the grid laser group.
Cataract progression was as we would expect—higher in
the steroid group. Considering the risk vs the benefits, the
BVOS standard of care grid laser won out over steroids for
BRVO.
The SCORE-CRVO7 study tested steroid vs observation using the same doses as in the BRVO study—1 mg and 4 mg triamcinolone acetonide. The standard of care group lost eight letters very quickly at the 6-month mark, and continued to lose over 10 letters at 24 months. Note there seems to be a protective effect from the steroids with a statistically significant in terms of mean gain change.
There was a clear benefit to steroids in both groups; 27% of those in the 1 mg group 26% of those in the 4 mg group gained three lines in 1 year, compared with 7% in the observation group. In the three-line losers category, the steroid groups lost far less vision than those in the observation groups. When considering the risks (IOP rise, cataract) vs the benefits (fivefold greater chance of gaining three lines), the recommendation of the authors is to treat CRVO with 1 mg triamcinolone acetonide. They chose the lower dose of steroid because of the improved side-effect profile.
DEXAMETHASONE IMPLANT DATA
The dexamethasone intravitreal implant (Ozurdex)
was approved by the US Food and Drug Administration
in June 2009 for macular edema secondary to retinal vein
occlusion. The 6-month trial data showed that 20% to
30% of patients were three-line gainers in the combined
CRVO and BRVO groups.11 The primary effect of the dexamethasone
implant occurs at 1 to 3 months. Patients
implanted with this device had 25% higher IOPs than
those who did not receive the device, but at 6 months,
the dexamethasone patients had relatively low rates of
filter/trabeculectomy.
The data show the implant provide showed that 20% to 30% responded to the high dose of dexamethasone, which is delivered for approximately 1 to 2 months (Figure 1), with the peak effect at 3 months. The effect seems to wear off at the 6-month mark. Even with the effect wearing off, however, the treated patients still fared better than the sham group.
THE ROLE OF ANTI-VEGF
What about the role of antivascular endothelial growth
factor (anti-VEGF) agents in CRVO and BRVO? We all know
that VEGF causes increased vascular permeability in angiogenesis.
VEGF is actually much more potent than histamine
because of vascular leakage, and it stimulates new blood vessel
growth. As early as 1994, Aiello et al12 showed that there
are high levels of VEGF in CRVO and diabetic retinopathy.
Subsequently, Pe'er et al13 showed that upregulation of the VEGF messenger RNA occurs in CRVO in enucleated specimens of neovascular glaucoma and Boyd et al14 demonstrated that increased VEGF levels in rubeosis measured by aqueous taps decrease after PRP. Studies by Campochiaro et al,15 Pieramici et al,16 and Spaide et al17 have all shown that ranibizumab treatment decreased macular edema and improved visual acuity in the short-term for patients with retinal vein occlusion.
At Retina Consultants of Houston, we are performing a trial evaluating ranibizumab (Lucentis, Genentech, Inc.) for patients with preproliferative CRVO who have had marked nonperfusion (RAVE [Rubeosis Anti-VEGF] Trial for Pre-Proliferative CRVO). In our study, we used the criteria from Hayreh et al:17,18 visual acuity worse than 20/200, a large relative pupillary defect (>0.9 log units) loss of I2e isopter on Goldman visual field testing, and reduction in electroretinography.
Per the RAVE protocol, patients received a preinjection anterior chamber tap and then intravitreal injections of ranibizumab for 9 months. Optical coherence tomography (OCT) measurements were taken monthly along with fundus conventional and widefield fluorescein angiography, iris angiography, and Goldman visual fields. We followed the patients for 3 years and reinjected after a 3-month observation period following the initial 9-month treatment protocol on an as-needed basis.
Almost universally, the center subfield retinal thickness decreased (Figure 2) for the first 9 months. To our surprise, the visual acuity also improved in a number of the patients studied (Figure 3). Sixty percent of patients had four-line letter gains at 9 months.
Were we just resetting the clock? When we stopped the anti-VEGF agent, approximately one-third of patients had recurrent edema (Figure 4) and lost the visual acuity that they had gained with treatment. Interestingly, once we entered the as-needed dosing phase, we found that approximately one-third of the patients required recurrent anti-VEGF therapy treatment to control the edema. Another surprising finding was that more than 50% of patients eventually developed neovascularization over the course of 3 years. Most surprising was that one-third of these patients developed neovascularization in the posterior segment, which was not a common finding prior to our use of anti-VEGF agents.
What did we learn? We learned that it is important to determine which patients have severe preproliferation before initiating treatment, so they can be watched diligently. Alternatively, if close monitoring is not an option, these patients might benefit from PRP.
BRAVO/CRUISE DATA
Two large multicenter studies evaluated ranibizumab
for BRVO and CRVO: BRAVO and CRUISE.
The inclusion criteria for BRAVO included visual acuity of 20/40 to 20/400 and for CRUISE it was 20/40 to 20/320. Mean central foveal thickness was required to be over 250 μm at screening and day 0 for both studies. Key exclusion criteria were that patients could not have a prior RVO, or a prior intraocular injection with either steroid or anti-VEGF agent for at least 3 months and systemic anti- VEGF for at least 6 months. We excluded patients with a brisk afferent papillary defect in the study eye, those who gained more than 10 letters of vision between screening and day 0, and those who had experienced a stroke or myocardial infarction 3 months or sooner prior to day 0.
BRAVO design. In the BRAVO trial over half the patients enrolled had less than 3 months duration. The mean best-corrected visual acuity (BCVA) at baseline was between 20/63 and 20/80 and the mean central retinal thickness was 500 µm.
In BRAVO, patients were randomized 1:1:1 to either sham, 0.3 mg ranibizumab, or 0.5 mg ranibizumab. All patients were eligible for rescue laser treatment once during the treatment period (beginning at month 3) and once during the observation period (beginning at month 9) if they met the following criteria: BCVA 20/40 or worse; central foveal thickness greater than or equal to 250 µm and compared with the month 3 visit prior to the current visit; and a BCVA gain of fewer than 5 letters or a decrease in central foveal thickness of less than 50 µm.
During the observation period, all subjects were eligible for as-needed ranibizumab treatment (sham subjects received 0.5 mg ranibizumab and the 0.3 mg and 0.5 mg groups received their assigned doses) if they met the following criteria: BCVA 20/40 or worse or mean central foveal thickness greater than or equal to 250 µm.
We saw no real difference in systemic side effects or ocular side effects between the groups.
Approximately 95% of all subjects completed the study through Month 6. Rescue laser was administered at 3, 4, or 5 months to 54% of the sham groups vs 18% in the 0.3 mg ranibizumab group and 19% in the 0.5 mg ranibizumab group.
BRAVO outcomes. There was a rapid reduction in macular edema in patients randomized to anti-VEGF that we found to be statistically significant at all time points.
The visual acuity in the standard of care arm increased slowly throughout the study period as compared to rapid visual gains in both ranibizumab groups beginning at day 7 and continuing throughout the study period. The differences in three-line gainers was also statistically significant between groups. Only 28.8% of those in the sham group gained three lines of vision at 6 months, compared to 55.2% in the 0.3 mg group and 61.1% in the 0.5 mg group.
CRUISE design. CRUISE was similarly designed with patients being randomized 1:1:1 to either sham, 0.3 mg, or 0.5 mg ranibizumab. At 6 months, patients were randomized to as-needed dosing. There was no rescue laser, however, because the CVOS showed no benefit from laser for macular edema secondary to CRVO.
Fifty percent to 60% of patients had less than 3 months of duration and mean visual acuity was 20/100 in all groups. Mean center foveal thickness was 680 µm.
Patient retention in CRUISE was also good: 88.5% in the sham arm and over 90% in the ranibizumab arms.
CRUISE outcomes. The eyes treated with ranibizumab had rapid resolution of their macular edema with an almost 400 µm mean reduction by day 7. This contrasts with the slow reduction in mean central foveal thickness in the sham arm. By the conclusion of the 6-month mandated treatment period, the central foveal thickness was reduced in both ranibizumab groups was reduced by a mean 434 µm (0.3 mg) and 452 µm (0.5 mg) from baseline; the mean central foveal thickness in the sham group was only reduced by 168 µm.
Although the sham arm had little change in vision over 6 months, the ranibizumab groups had a rapid and robust improvement in visual acuity immediately after the first injection. By day 7, both treatment arms gained over eight ETDRS letters and this improvement continued with the 0.3 mg group gaining 11.9 letters and the 0.5 mg group finishing with an increase of 14.1 letters over baseline. The difference between the sham group and the two ranibizumab arms was statistically significant at day 7 and at every subsequent visit. At month 6, the ranibizumab arms were 11.9 and 14.1 letters better on average than the sham cohort.
Patients treated with ranibizumab were approximately three times more likely to be three-line gainers at 6 months than those in the sham group (Figure 5). Interestingly, more patients in the ranibizumab groups also had less retinal hemorrhaging at month 6 than at baseline compared with sham.
SUMMARY
The results of the BRAVO and CRUISE studies leave
some unanswered questions. Will visual acuity gains
be maintained after 6 months? How many patients
will need ongoing VEGF suppression? Will other anti-
VEGF agents be equivalent or superior to ranibizumab?
Currently, a phase 3 study, COPERNICUS (Controlled
Phase III Evaluation of Repeated Intravitreal Administration
of VEGF Trap-Eye In Central Retinal Vein Occlusion:
Utility and Safety) is recruiting patients, and the results
of this study may answer this last question.
Anti-VEGF agents (specifically ranibizumab in the BRAVO and CRUISE trials) are shown to decrease macular edema secondary to BRVO and CRVO and these decreases are associated with significant visual acuity gains.
The results of the BRAVO and CRUISE trials imply several points regarding the pathophysiology of RVO. First is the implication that edema is VEGF-mediated, which is surprising to me because I had previously believed that Starling's law or osmotic forces would have more of an effect on edema because of the increased venous pressure. Second is the implication that anti-VEGF treatment may decrease retinal hemorrhage, a theory for which we are currently performing more data analysis. Third, these results also imply that all RVOs must be ischemic to some extent. Further, patients with RVO who have significant nonperfusion must be closely monitored if anti-VEGF is not administered indefinitely because unchecked, these patients are likely to develop neovascularization at some point.
David M. Brown, MD, is the director of the Greater Houston Retina Research Center and practices at Retina Consultants of Houston and the Methodist Hospital in Houston, TX. He is a member of the Retina Today Editorial Board. Dr. Brown can be reached at dmbmd@houstonretina.com.
_1773249222.png?auto=compress,format&w=75)
