AMD first presents in the so-called dry form, which may be detected in its early and intermediate stages, with drusen of variable sizes and numbers visualized in the macula in 1 or both eyes. Individuals with dry AMD may develop GA (ie, the late stage of dry AMD) at any point; however, it typically takes several years for dry AMD to progress to GA. Thus, patients are usually older when they are diagnosed with GA. 

There can be overlap between wet and dry AMD. Patients with dry AMD may develop macular neovascularization or wet AMD at any time in their disease course. When this occurs, the wet manifestations become superimposed on the background of dry AMD. Moreover, with advancing age, macular atrophy may develop in patients who are being treated for wet AMD, causing vision loss or symptoms related to atrophy despite the adequate management of the neovascularization with anti-VEGF agents.

GA lesions typically spare the foveal center until late in the disease course, and visual acuity may be unaffected (eg, starting and remaining at 20/20) while the areas of atrophy continue to expand and progress. It is not until the lesions begin to involve the foveal center that visual acuity drops. 

Nonetheless, patients can be symptomatic even prior to foveal center encroachment; these individuals may experience difficulty with reading, driving, and functioning in low-light conditions. This visual dysfunction can be significant, but we are not yet able to measure it adequately in the clinic. Reading speed, low luminance visual acuity, and standard best-corrected visual acuity are all ways to measure visual function, or aspects of visual function, but there is no single test that can completely tell us the whole picture of how GA affects vision. In clinical trials, we can measure GA lesion growth and the visual impact of expanding atrophic lesions; however, more precision and sensitivity in our measurements of visual function would be welcome in both clinical trials and in the clinic. 

At AAO 2022, intriguing data presented by Ursula M. Schmidt-Erfurth, MD, hinted at the possibility of improving our tools to track AMD, and I think that this might shift the way that we look at the progression of dry AMD and GA. GA is classically defined by the loss of retinal pigment epithelium (RPE) and choriocapillaris. We measure a patient's RPE loss with fundus autofluorescence or by looking at optical coherence tomography (OCT; ie, the "B-scan") cross-sectional images. However, researchers have been applying deep learning tools to map RPE and photoreceptor (PR) loss over time, generating interesting findings. Using OCT, one can discern PR loss, or outer retinal changes that are independent of RPE loss. In the natural history of GA, the rate of PR loss is, evidently, greater than the rate of RPE loss. PR loss not only exceeds but also precedes RPE loss in GA, so this type of an artificial intelligence–assisted imaging approach might enable us to detect GA earlier. PR loss represents an area where there will be some degree of visual dysfunction.

Such PR mapping might also be useful as we look to select patients for, and assess therapeutic responses to, emerging GA treatments. For instance, Dr Schmidt-Erfurth and colleagues showed the effect of treatment with the C3 inhibitor pegcetacoplan on PR maintenance via artificial intelligence–based OCT analysis. The effect was pronounced and statistically significant, and there was a trend toward greater inhibition of PR loss than RPE loss, so it is possible that PR loss could be a measurement that is more sensitive for assessing response.

Also at the conference, Karl G. Csaky, MD, discussed the role of novel functional assessments for following AMD, including microperimetry. This technique is a type of visual field test that is used to create a light sensitivity map in specific parts of the retina. There is quite a bit of intra- and interpatient variability, so interpretation can be a challenge, but microperimetry may give us some additional insights on how intervening in GA can affect visual function over time in treated vs nontreated eyes. For instance, in the 24-month data presentation from the DERBY and OAKS trials at AAO 2022, Charles C. Wykoff, MD, PhD, included a discussion of post hoc perilesional microperimetry results, which revealed a signal of functional preservation with pegcetacoplan treatment.

Regarding how the natural history of AMD plays out in the clinic, our impression is that GA is likely underdiagnosed or unrecognized in its earliest stages. Some of this may relate to having no available treatments heretofore. As we look forward to potentially having therapies for GA, it is important to educate the primary eye care community about these treatments for dry AMD that can slow progression and about the importance of GA detection at its earliest stages. Once the visual acuity has dropped and the GA is relatively severe, the relative benefits of treatment are decreased.

Early intervention is going to be important in that we are delaying the onset of symptoms and/or vision loss by slowing atrophy, so it does make sense to diagnose and treat patients early. Of course, we would also consider the patient's life expectancy in view of the natural history of the disease. If GA is just starting to develop in a person with advanced age, significant health problems, and a short life expectancy, early intervention may not be in their best interest. Thus, having a treatment available would not necessarily mean that we would treat all patients with early GA whom we encounter. We will need to weigh the risks vs benefits of treatment for each individual patient and assess the burden of treatment (ie, monthly vs every-other-month injections on an ongoing and regular basis).