Protocol T Provides Additional Guidance for Clinical Practice

Take-home points from a direct comparison of anti-VEGF agents for the treatment of DME.

BY CHARLES C. WYKOFF, MD, PHD

One-year results were recently published from “Aflibercept, Bevacizumab, or Ranibizumab for Diabetic Macular Edema”¹ (Protocol T), a clinical trial conducted by the Diabetic Retinopathy Clinical Research Network. Protocol T is an excellent trial that has broken new ground — it is the only study on this scale to directly compare three anti-VEGF medications for the treatment of patients with decreased visual acuity due to center-involved DME.

What Have We Learned?

We can pull several important points from the study report.

➤ All three medications led to excellent vision benefits in the patient population (Table 1) as a whole. The benefits were achieved with averages of 9 or 10 injections during the year. It’s worth noting that this is more injections than patients in this population typically receive in 1 year on average in clinical practice.

In addition, the re-treatment criteria used in the trial don’t precisely parallel how most patients are managed in clinical practice. In the first 6 months of the trial, patients received an injection each month unless protocol-defined stability was achieved, visual acuity was 20/20 and OCT showed no significant central DME, which was rare. After 6 months, the modified prn dosing strategy dictated that patients did not receive a treatment if visual acuity and OCT status hadn’t changed substantially over two consecutive injections, even if the macula continued to have substantial DME. In contrast, in clinical practice, if a patient’s progress hits a “plateau,” the treating physician would likely shift to a different strategy, perhaps treating more frequently, switching to a different anti-VEGF agent or adding a steroid, rather than simply not treating.

➤ Based on a pre-specified subgroup analysis, eyes with initial visual acuity worse than 20/40 achieved more substantial visual acuity improvement on average with Eylea than with Avastin or Lucentis, 18.9, 11.8 and 14.2 letters respectively.

➤ The anatomic effect of Avastin on retinal thickening was inferior to the effect of Lucentis or Eylea regardless of initial visual acuity, an important distinction because we know chronic edema limits the visual outcomes patients can ultimately achieve. On average, central subfield thickness decreased by 169±138 μm with Eylea, 147±134 μm with Lucentis and 101±121 μm with Avastin (P < .001 for both Eylea vs. Avastin and Lucentis vs. Avastin; P = .036 for Eylea vs Lucentis). The greater reduction in retinal thickness produced by Eylea was associated with a lower rate of supplemental laser treatment. The percentage of patients who received focal macular laser treatment was 37% with Eylea, 46% with Lucentis and 56% with Avastin.

➤ Cost considerations are important. Ideally, patients would receive the medication that would work best for them, and according to the Protocol T results, the more expensive agents produced the best anatomical results in all subgroups, and Eylea treatment led to significantly better visual outcomes in patients with worse initial vision. The cost issue is a nearly insurmountable challenge for an individual physician. Many would argue this issue should be addressed on a national level as has been done in many developed countries. The United States is a rarity in the developed world in not considering pricing when approving new pharmaceuticals. For example, in British Columbia, Canada, the government endorses splitting of Lucentis and Eylea vials, allowing for more than one treatment per vial.

➤ A concern that one or more of the anti-VEGF agents causes more intraocular inflammation than the others was not borne out in this trial. Rates among the three agents were similar.

➤ Both Eylea and Avastin contain an Fc domain, which theoretically leads to a longer systemic half-life. Lucentis does not contain an Fc domain. Differences in systemic exposure to anti-VEGF activity between the medications based on this difference is often debated and is a potentially clinically relevant issue. While the Protocol T trial wasn’t powered to detect small differences in safety, no differences were observed among the three agents in the rates of serious systemic adverse events, Anti-Platelet Trialists’ Collaboration events or deaths.

➤ The Avastin used in the trial was repackaged by a single compounding pharmacy and stored in glass vials. It also was independently tested for purity and potency. Because these conditions aren’t met in many clinical practices, it’s unclear how real-world results with Avastin might differ.

➤ The 0.3-mg dose of Lucentis, used only in the United States, was used in the trial. Therefore, the results may not be applicable in other countries, where the 0.5-mg dose is used in the management of DME. However, the 0.5-mg dose of Lucentis was found to have a worse side effect profile, especially when looking at rates of death.

➤ All patients in the trial had visual acuity of 20/32 or worse. Therefore, the results may not directly apply to patients who have center-involved DME but visual acuity of better than 20/32.

Study Year 2 in Progress

The Protocol T trial has provided us with very good 1-year data regarding the treatment of DME, a chronic and complex disease. Longer-term data is needed, and the trial will continue for an additional year. We look forward to learning about the results achieved and further guidance that will emerge in year 2. ■

Reference