10 Nov Age-Related Macular Degeneration: Computerized Imaging Predicts Risk Progression
MedicalResearch.com Interview with:
Daniel L. Rubin, MD, MS
Assistant Professor of Radiology and Medicine (Biomedical Informatics)
Department of Radiology | Stanford University
Stanford, CA 94305-5488
Medical Research: What is the background for this study? What are the main findings?
Dr. Rubin: Age-Related Macular Degeneration is the leading cause of blindness and central vision loss among adults older than 65. An estimated 10-15 million people in the United States suffer from the disease, in which the macula — the area of the retina responsible for vision — shows signs of degeneration. While about one of every five people with AMD develop the so-called “wet” form of the disease that can cause devastating blindness. In wet AMD, abnormal blood vessels accumulate underneath the macula and leak blood and fluid. When that happens, irreversible damage to the macula can quickly ensue if not treated quickly. Until now, there has been no effective way to tell which individuals with AMD are likely to convert to the wet stage. Current treatments are costly and invasive — they typically involve injections of medicines directly into the eyeball — making the notion of treating people with early or intermediate stages of Age-Related Macular Degeneration a non-starter. In our study, we report on a computerized method that analyzes images of the retina obtained with a test called spectral domain optical coherence tomography (SD-OCT), and our method can predict, with high accuracy, whether a patient with mild or intermediate Age-Related Macular Degeneration will progress to the wet stage. Our method generates a risk score, a value that predicts a patient’s likelihood of progressing to the wet stage within one year, three years or five years. The likelihood of progression within one year is most relevant, because it can be used to guide a recommendation as to how soon to schedule the patient’s next office visit. In our study, we analyzed data from 2,146 scans of 330 eyes in 244 patients seen at Stanford Health Care over a five-year period. Patients were followed for as long as four years, and predictions of the model were compared with actual instances of conversion to wet AMD. The model accurately predicted every occurrence of conversion to the wet stage of AMD within a year. In approximately 40% of the cases when the model predicted conversion to wet AMD within a year, the prediction was not borne out, however. We are currently refining the model to reduce the frequency of these false positives.
Medical Research: What should clinicians and patients take away from your report?
Dr. Rubin: It is important to be able to stratify Age-Related Macular Degeneration patients into risk categories so that the degree of close follow up can be established. Visual outcomes can be improved when conversion to wet AMD occurs by catching that conversion as early as possible and instituting prompt treatment. It is not feasible to screen all Age-Related Macular Degeneration patients very frequently, so it’s currently up to the patient to detect changes in their vision that may indicate conversion to wet AMD, and often by the time patients are seen by ophthalmologists, the disease has advanced. Using our approach, it may be possible to identify the subset of patients who should be evaluated more frequently by ophthalmologists to catch conversion to wet Age-Related Macular Degeneration at the earliest possible stage and to institute specific treatment for wet AMD.
Medical Research: What recommendations do you have for future research as a result of this study?
Dr. Rubin: Our results to date, while promising, have been obtained on data from one institution. We are now undertaking work to validate our methods on data from other institutions, obtained using different SD-OCT scanners. We are also improving the computational method to improve the reliability of its risk prediction.
Quantitative SD-OCT Imaging Biomarkers as Indicators of Age-Related Macular Degeneration Progression
Luis de Sisternes, Noah Simon, Robert Tibshirani, Theodore Leng, and Daniel L. Rubin
IOVS November 2014 55:7093–7103; published ahead of print October 9, 2014, doi:10.1167/iovs.14-14918