Intravitreal Aflibercept for Diabetic Macular Edema

We review the data on which approval of the newest anti-VEGF drug for DME was based.

JOHN F. PAYNE, MD • W. LLOYD CLARK, MD

Diabetic retinopathy is the most common cause of vision loss among working-age individuals.¹ Approximately one-third of patients with DR have diabetic macular edema, accounting for 7.5% of the diabetic population.²

A major cause of vision loss, DME results from microvascular changes secondary to hyperglycemia and inflammatory mediators associated with diabetes mellitus. Various cytokines and growth factors, such as VEGF, are produced as a response to local hypoxia secondary to capillary nonperfusion.³ Aiello and colleagues demonstrated that VEGF levels were increased in the ocular fluid of patients with DR and other retinal disorders, including DME.⁴

Focal laser photocoagulation was established as the standard of care for treatment of DME when the Early Treatment Diabetic Retinopathy Study (ETDRS) demonstrated a 50% reduction in moderate vision loss with laser therapy, compared to observation.⁵ The Diabetic Retinopathy Clinical Research Network (DRCRnet) used a modified ETDRS laser protocol and found that approximately 15% of patients gained ≥15 letters at one year, a rate similar to the ETDRS and other laser control arms in several subsequent clinical trials.⁶

INTRAVITREAL THERAPY EMERGES

Over the past few years, intravitreal injections of anti-VEGF agents have progressively replaced focal laser photocoagulation for the primary treatment of center-involving DME.⁷

The DRCRnet demonstrated that intravitreal ranibizumab (Lucentis, Genentech, South San Francisco, CA), combined with prompt or deferred laser, was more effective through one year than prompt laser alone for the treatment of center-involved DME.⁶ At year 3, the results suggested that focal/grid laser therapy at the initiation of intravitreal ranibizumab was no better, and possibly worse, for visual outcomes than deferring laser treatment for 24 weeks or more.⁷

The RISE and RIDE studies were two phase 3 clinical trials demonstrating that monthly treatment with intravitreal ranibizumab resulted in significant visual acuity improvement in a large percentage of patients when administered with PRN laser.⁸

Aflibercept (Eylea, Regeneron, Tarrytown, NY) is an intermediate-sized fusion protein that binds to all isomers of the VEGF-A family, similar to ranibizumab and bevacizumab (Avastin, Genentech).

Aflibercept differs from ranibizumab and bevacizumab in that it binds VEGF-A more tightly than its native receptors in a strict 1:1 fashion, and it also binds other VEGF family members, such as placental growth factor.⁹ As a soluble circulation trap for VEGF, aflibercept binds to the target, clears it from the vitreous, and inhibits the binding and activation of the VEGF receptors.³

CLINICAL TRIALS

All of the clinical trial data publicly available concerning aflibercept for DME come from trials sponsored by Regeneron. The findings from these trials will be discussed in some detail here.

Phase 1

In 2009, a phase 1 study was conducted by Do and colleagues to assess the safety, tolerability and bioactivity of a single 4-mg intravitreal injection of aflibercept in five patients with DME over a six-week period.¹⁰

The investigators were able to establish that a single 4-mg injection of aflibercept was well tolerated without any ocular toxicity. The median visual improvement in these patients was 9 letters after four weeks and 3 letters after six weeks.

Phase 2: The DA VINCI Trial

The primary purpose of the DME and VEGF Trap-Eye: INvestigation of Clinical Impact (DA VINCI) study was to determine whether different doses and dosing regimens of intravitreal aflibercept were superior to focal laser photocoagulation over a 24-week study duration in eyes with DME.⁹

A total of 220 eyes were randomly assigned in a 1:1:1:1:1 ratio to one of five treatment groups in one eye only: 0.5 mg of aflibercept every four weeks; 2.0 mg of aflibercept every four weeks; 2.0 mg of aflibercept for three initial monthly doses and then every eight weeks; 2.0 mg of aflibercept for three initial monthly doses and then on an as-needed basis; or modified ETDRS macular laser photocoagulation.

The primary endpoint, mean change in VA, was measured at week 24, but treatment was extended through one year. Baseline demographic and disease characteristics were fairly well matched across all of the groups.

At the week 24 primary endpoint, eyes treated with aflibercept gained between 8.5 and 11.4 letters, compared to a 2.5-letter gain in the laser group. The change in visual acuity from baseline to week 24 was statistically significantly greater in each of the aflibercept treatment groups, compared with the laser group (P=.0085). It should be noted that the study was not powered to detect differences among the aflibercept treatment groups, and no statistically significant differences were observed.

At one year, the difference between aflibercept and laser persisted. There were visual gains of between 9.7 and 13.1 letters in the aflibercept treatment groups, compared to a 1.3-letter loss in the laser treatment group.

Similar findings were seen in the change in macular thickness on optical coherence tomography. In the aflibercept groups, central subfield thickness decreased by 165 µm to 227 µm, compared to a 58-µm decrease in the laser group. These anatomical differences also persisted at one year.

In terms of patient safety, there were no significant differences in serious adverse events in patients receiving aflibercept compared to laser or among the different aflibercept groups. While three patients in the aflibercept groups experienced arterial thromboembolic events, there was an imbalance in the previous history of cardiac events at randomization.

The baseline incidence of cardiac disease in the groups randomized to aflibercept was twice that of the laser group (39.4% vs 18.2%). The investigators concluded that this disparity may have contributed to the higher rate of nonocular serious adverse events in the aflibercept group.

Although DA VINCI was not powered to detect differences among the aflibercept treatment groups, important signals were identified. First, monthly treatment with 2.0 mg appeared to have the greatest treatment effect, necessitating its inclusion in any future registration trial. In contrast, the 0.5-mg dose appeared to have a reduced clinical effect. Finally, favorable clinical outcomes with dosing every eight weeks supported the inclusion of this strategy for future studies.

Phase 3: The VISTA and VIVID Trials

Based on the positive data in DA VINCI, two large phase 3 clinical trials, using intravitreal injections of aflibercept for DME, were organized. The VISTA and VIVID trials were parallel, double-masked trials comparing the safety and efficacy of focal laser photocoagulation (with sham intraocular injections) to intravitreal aflibercept dosed either every four weeks or every eight weeks, after five initial monthly doses.¹¹ These trials enrolled 872 eyes with vision loss due to center-involved DME, and approximately 90% of the subjects completed the week 52 primary endpoint visit.

The VISTA and VIVID studies were noteworthy in that they were the first head-to-head trials comparing anti-VEGF blockade alone to laser therapy alone. Figure 1 shows the primary efficacy endpoint at one year. The mean change in VA was similar in the monthly (+10.5 letters, VISTA; +12.5 letters, VIVID) and every other month (+10.7 letters, VISTA and VIVID) groups and was superior to macular laser (+0.2 letters, VISTA; +1.2 letters, VIVID).

Figure 1. The mean change in best-corrected visual acuity from baseline to week 52 in the two aflibercept treatment groups compared to macular laser photocoagulation.

Other secondary endpoints, including the change in central subfield thickness on OCT (Figure 2), percentage of patients gaining ≥15 letters (Figure 3), and two-step improvement in DR score (Figure 4), favored both treatment schedules of aflibercept over laser photocoagulation by a statistically significant margin. Furthermore, the percentage of eyes in the laser group losing ≥15 letters of vision was 9.1% (VISTA) and 10.6% (VIVID), compared to <1% in the aflibercept groups.

Figure 2. The mean change from baseline in central subfield thickness on OCT at each study visit through week 52.

Figure 3. The percentage of patients gaining ≥15 letters from baseline at week 52 in the aflibercept and laser photocoagulation groups.

Figure 4. The percentage of patients with a ≥2-step improvement in DRSS score from baseline to week 52.

During enrollment for VISTA, off-label use of anti-VEGF agents had gained popularity in the United States, primarily with the use of bevacizumab. As a result, approximately 43% of study eyes in VISTA had been previously treated with anti-VEGF agents (with a ≥3-month washout period).

Because of the large number of subjects receiving prior anti-VEGF treatment, it was possible to perform post-hoc subgroup analysis to determine whether prior treatment status was a predictor of response to aflibercept.

Figure 5 shows the treatment effects in the subgroups of patients treated with an anti-VEGF agent prior to enrollment compared to treatment-naïve subjects. Subjects assigned to either dosing regimen of aflibercept demonstrated similar benefits to therapy, regardless of prior treatment

Figure 5. The mean change in best-corrected visual acuity at week 52 based on prior anti-VEGF exposure.

Additionally, there was no increased rate of death, stroke, or myocardial infarction in the monthly aflibercept treatment group at the 52-week endpoint. The safety outcomes at one year were similar across all of the treatment groups.

The percentage of treatment-emergent nonocular serious adverse events, defined by the Anti-Platelet Trialists’ Collaboration, was 2.8% in the laser group vs 3.1% and 3.5% in the aflibercept treatment groups.

DISCUSSION

In July 2014, the FDA approved the use of intravitreal aflibercept for the treatment of DME. Aflibercept is also approved for treatment of neovascular AMD and cystoid macular edema following RVO.

The recommended dosage of aflibercept for DME treatment is 2.0 mg given every eight weeks after five initial monthly injections. It should be noted that it may be given as frequently as every four weeks, but no additional benefit was found when dosed every four weeks, compared to every eight weeks.

An interesting finding of the VISTA and VIVID trials was that there was a significant improvement in the level of DR at week 52 in the aflibercept treatment groups, compared to the laser group (Figure 4).

These findings were similar to results observed in the RISE and RIDE clinical trials evaluating ranibizumab for DME.¹² At the conclusion of RISE and RIDE, 38.9% to 39.2% of eyes demonstrated a two-step regression in DR severity, and 13.2% to 15.0% demonstrated a three-step regression.

Campochiaro and colleagues recently analyzed the degree of retinal nonperfusion in the RISE and RIDE studies. They concluded that monthly injections with ranibizumab could slow, but not completely prevent, retinal capillary closure in patients with DME.¹³ While the data from these clinical trials are exciting, future studies are needed to better determine the degree of benefit of VEGF blockade with regard to DR regression.

In the near future, data will be available from the DRCRnet-sponsored Protocol T. This is a head-to-head trial comparing intravitreal injections of aflibercept, ranibizumab, and bevacizumab for the treatment of center-involved DME.

This trial employs a novel treatment schedule, which allows for less than monthly dosing of all drugs based on disease activity. The results of Protocol T will likely be met with much interest in the retina community. RP

REFERENCES