Investigators calculated perilesional sensitivity and responding sensitivity for eyes in a geographic atrophy clinical trial and compared these to mean retinal sensitivity.
Seeking to improve the design and conduct of clinical trials in geographic atrophy, a team of investigators used data from the Chroma and Spectri trials of intravitreous lampalizumab to develop new endpoints.
Their results were presented at the 2019 Annual Meeting of the Association for Research in Vision and Ophthalmology (ARVO) in Vancouver, BC.
Dolly Chang, MD, PhD, a clinical instructor at Stanford University Byer Eye Institute and assistant medical director at Genentech, spoke with MD Magazine® about the team’s work on developing new endpoints for clinical trials in geographic atrophy.
“What we found is that the most commonly used endpoint, which is the mean retinal sensitivity, is not ideal to detect the changes over time for geographic atrophy,” Chang told MD Mag.
Instead, the team looked at perilesional sensitivity and responding sensitivity. These 2 outcomes provided better correlation with structural changes in the eye and had less variability, she added.
In part 1 of this interview, Chang spoke of the context for this research and the challenges for developing clinical trials in geographic atrophy.
What we found is that the most commonly used endpoint, which is the mean retinal sensitivity, is not ideal to detect the changes over time for geographic atrophy. The reason why is when you include all the points on the microperimetry then we dilute the other responding points in the testing. So, what we did is we calculated 2 other endpoints instead. One of them is perilesional sensitivity, which we focus at the points that are outside the rim around the scotoma. The second endpoint we are looking at is the responding sensitivity, which is looking at all the non-scotomatous points. What we found is that those 2 endpoints provide higher magnitude of change with less variability and better correlation with the structural changes over time.Sure, so, even though this is a better outcome than the mean retinal sensitivity, there's still a lot of work to do. First of all, microperimetry is a very difficult test for patients to do. It takes somewhere—a range between 10 minutes to 30 minutes for a single patient to complete the test. It is highly variable—the test reliability is not high. So, there's definitely a lot more work to do even though we get a slightly better outcome.