As retina specialists, we rely heavily on both static and dynamic imaging platforms to inform diagnoses and guide our management decisions. Despite our dependence on imaging, however, most patients could not care less precisely what their imaging shows; they care about what they can see, and what they will be able to see moving forward.
Our imaging serves to provide us with biomarkers that allow indirect assessment of visual function. The National Institutes of Health defines a biomarker as, “a characteristic that is objectively measured and evaluated as an indicator of normal biological processes, pathogenic processes, or pharmacologic responses to a therapeutic intervention.”
As our imaging capabilities advance, so does our knowledge of potential biomarkers. On page 16, Chris Or, MD, and Nadia Waheed, MD, MPH, describe more than a dozen biomarkers of age-related macular degeneration, based on optical coherence tomography and OCT-angiography, being investigated for their potential value in predicting disease progression to either geographic atrophy and/or neovascular AMD.
An appreciation of the prognostic significance of our imaging advances has defined entirely new categories of AMD, including nascent GA and nonexudative neovascular AMD.
Such nomenclature expansion is impacting other disease states, too, including early stages of diabetic retinopathy harboring quantifiable pathology detectable with OCT-A, but not fundus photography.
Imaging characteristics hold substantial value if they can be reliably correlated with disease progression and/or functional endpoints. Traditionally, the Food and Drug Administration has used changes in central vision to guide approval of new pharmaceutical agents designed to treat vitreoretinal diseases. More recently, structural biomarkers have also been employed to seek drug approval, including change in DR severity using the
Diabetic Retinopathy Severity Scale, release of vitreomacular traction by OCT and change in GA area by autofluorescence.
Moving forward, our field will certainly identify additional anatomic endpoints that can be correlated strongly enough with functional endpoints to be utilized as approvable endpoints. This is particularly needed in early disease states before significant central vision is lost, such as early diabetic macular edema. Developments in artificial intelligence, including deep-learning algorithms, appear capable of advancing prognostication towards the ultimate threshold of personalized medicine or, “an n of 1,” as Anthony P. Adamis, MD, of Genentech recently eloquently predicted.
The philosophy of, “Show me the money,” instilled in our culture by the 1996 movie “Jerry Maguire,” seems particularly relevant. If we listen to our patients, and the FDA, they will tell us, sometimes more sub