For this month's Retina Minute, I have the honor of interviewing my dear friend, Professor Richard Gale from Hull York Medical School in the United Kingdom, who was involved in the QUASAR study for Bayer.
Professor Gale, thanks for coming today. Tell us about the study and what the results were.
Dr. Singer, it is always a great pleasure and honor to speak to you about the latest scientific developments, and I am pleased to be involved in the QUASAR study. QUASAR was a Bayer-sponsored study to investigate high-dose aflibercept (8 mg; Eylea; Regeneron) compared with 2 mg in retinal vein occlusion (RVO). We presented the primary endpoint just recently at Angiogenesis.
This was a multicenter, randomized, double-masked study in patients with all forms of RVO, including branch RVO (BRVO), hemi RVO, and central (CRVO). They were randomized at baseline 1:1:1 into aflibercept 2 mg every 4 weeks (2q4), 8 mg with initially three monthly doses followed by Q8 week doses (8q8 3), and aflibercept 8 mg at five initial monthly doses followed by Q8 week dosing (8q8 5). The primary endpoint was at week 36 and measured the change in best corrected visual acuity from baseline as the noninferiority endpoint, with a few interesting secondary endpoints. The end of the study is at week 64, but the study has not completed yet, so the Angiogenesis presentation was for the results of the week 36 primary efficacy endpoint.
The participants had the opportunity to enter a treat-and-extend regimen at different time points. The 2q4 group received four weekly doses all the way through to week 32, and then could enter an interval extension. The 8q8 3 group could also enter an extension up to week 32, but the 8q8 5 group could not extend until week 40, which is after the results we are talking about today. The criteria for extension was a loss of less than 5 letters or a central retinal thickness (CRT) of less than 320 µm (Heidelberg or 300 µm Topcon / Cirrus). Intervals could also be shortened to a minimum of 4 weeks if there was loss of best corrected visual acuity of greater than five letters or there was an increase of greater than 50 µm in CRT from a reference visit, which was typically at the end of the loading phase or week 20 for 2q4. This is a slightly complex design, but it is a good design to test the power of the molecule and different dosages to suppress VEGF in this high-VEGF disease state.
QUASAR had typical inclusion criteria. Participants were at least 18 years of age, with best corrected visual acuity between 73 and 24 letters, and CRT greater than 300 µm on the Cirrus or Topcon OCTs, and greater than 320 µm on the Heidelberg OCT. The trial also had typical exclusion criteria: namely, no concomitant disease that could cause a substantial reduction in best corrected visual acuity. Glaucoma had to be controlled with intraocular pressure of greater than 25 mmHg excluded, no intraocular inflammation within 12 weeks of inclusion, and no uncontrolled blood pressure of 160 over 95. QUASAR was also a truly global study with 237 sites across 27 countries, and almost 95% of patients have completed through week 36 so far.
The baseline demographics were well-balanced with a mean age of around 65 years across the groups. Patients were just slightly predominantly male and predominantly White, while approximately 30% were Asian. Best corrected visual acuity was around 54 to 55 ETDRS letters at baseline with a CRT of around 610 to 650 µm across the groups.
QUASAR met its primary endpoint; in particular, noninferiority of the 8 mg groups compared with the 2 mg groups with a visual acuity gain of around 17 to 18 letters. That result was reflected in the CRT reductions, which were robust and rapid in all groups with an observed a reduction of around 370 µm. There were minor fluctuations at the end of the loading phase in the 8q8 3 group.
One of the important secondary endpoints studies the proportion of patients who stayed on the Q8 week dosing. Through the study, approximately 90% of the 8 mg groups stayed on that 8 mg as a minimum. In fact, at the end of week 36, we began to see that patients were starting to be assigned to longer treatment intervals and nearly 70% of the 8q8 3 individual groups were assigned to 12 weeks or more.
How long could patients have potentially been assigned to, given that we are only in the first 36 weeks of the trial? Could they go 4 months?
There was no limit on the extension. Within this first 36 weeks, the longest patients could be assigned was 12 weeks, but if we take that forward to the next 4-week extension, they could potentially be extended out to 20 weeks. It is great to see these initial efficacy results, but I am waiting to see the outcome at week 64 to observe the potential for further extension and how the different groups perform in comparison with one another. We need to remember this is a high VEGF-state disease. It is, in this respect, perhaps quite different from diabetic macular edema and neovascular age-related macular degeneration. This study has really put the molecule and dosages through their paces to see what the extension could be.
Was there any difference between BRVO and CRVO?
We do not know yet because these data are yet to be analyzed. We have seen in other anti-VEGF therapy studies where CRVO and BRVO have had numerical differences in the vision gains. QUASAR was powered for combining all the RVO groups, but there is powering to analyze how CRVO and BRVO perform too. I look forward to those results and perhaps presenting them to you at a later date.
How was the safety?
Safety was as expected in that there were no new safety signals. We know that there are two key things that people are talking about with high-dose aflibercept: One is around the slightly higher volume of 70 µm injected in the 8 mg groups rather than the typical 50 µm volume in the 2 mg group. The most important aspect in my opinion is that we saw no sustained raise in intraocular pressure across the study and that the rate of paracentesis was low: 0.7% to 1% in the 8 mg and 2 mg groups respectively. Of course, we know that when a paracentesis is performed, it is subjective between clinicians.
The second safety aspect is the rate of intraocular inflammation. Again, there were no new safety signals. In fact, numerically intraocular inflammation was slightly lower in the 2 mg group compared with the 8 mg groups with no occlusive retinal vasculitis events.
We are waiting for when 8 mg aflibercept gets a license for RVO. It will be interesting to understand how it will perform in the real world in terms of effectiveness, durability, and safety. I look forward to analyzing some audit data when it is available to us.
Professor Gale, thank you so much for spending the time and showing us this new indication for a second-generation anti-VEGF, which obviously is well-suited for the high-VEGF load that RVOs bring.
