Tests may foretell age-related macular degeneration risk

Genotyping of patients to determine the risk of progression to advanced age-related macular degeneration is going to be a critical part of current/ongoing clinical trials and those in the future.

Orlando, FL-Genotyping of patients to determine the risk of progression to advanced agerelated macular degeneration (AMD) is going to be a critical part of current/ongoing clinical trials and those in the future.

Ophthalmologists should learn about the test kits that are commercially available to help patients best who are interested in learning about their individual risks of developing AMD. Ivana K. Kim, MD, described considerations involved in genetic testing for AMD during Retina Subspecialty Day at the annual meeting of the American Academy of Ophthalmology.

"Over the past 6 years, great progress has been made in gaining an understanding of the genetics involved in AMD," said Dr. Kim, assistant professor of ophthalmology, Retina Service, Massachusetts Eye and Ear Infirmary, Harvard Medical School, Boston.

An important question in AMD therapy, however, involves the potential for genetic testing for individual patients.

"Are we really ready to start this?" Dr. Kim posed to the audience.

Several tests for genetic risk factors of AMD are already commercially available (Macula Risk, ArcticDx; ARUP Laboratories; deCODE Com plete Scan, deCODEhealth; RetnaGene AMD, Sequenom; and 23andMe). Each test uses varying numbers of genetic markers and different models to predict a patient's risk of developing AMD, according to Dr. Kim. One test (Retna-Gene AMD) uses an independently validated model for specifically predicting an individual's risk of neovascular AMD. Patients are now able to purchase some of these tests directly.

Can physicians determine the value of doing genetic testing on specific individuals? The Centers for Disease Control and Prevention has developed a model (ACCE) that considers analytic validity; clinical validity; clinical utility; and associated ethical, legal, and social implications for evaluating scientific data on emerging genetic tests, she said.

Analytic validity

This involves the technical accuracy of a test. With current technology, the accuracy and reliability of genotyping is very high, Dr. Kim noted. Any errors that are introduced most likely are related to sample handling rather than the test itself.

Clinical validity

The consideration with clinical validity is whether a test is able to discriminate accurately between risk groups, specifically, what individuals are at high and low risk for developing AMD. Plotting the frequency of true positives versus false positives generates a receiver operating curve.

"The area under the curve (AUC) is a measure of the discriminatory power of a test," Dr. Kim said. "A perfect test has an AUC of 1. The recommended AUC for screening patients at increased risk of disease is greater than 0.75. The current models for AMD risk have an AUC of greater than 0.8."

Does genotyping add to the ability to predict the risk of developing advanced AMD? It may not be necessary. She explained that the macular phenotype represented by the Age-Related Eye Disease Study (AREDS) simple scale score really has the greatest predictive value, as reported by Klein et al. (Arch Ophthalmol., online Aug. 8, 2011).

Investigators evaluated models that included epidemiologic factors such as age, family history, and smoking; phenotype manifested by the simple scale score and the presence of very large drusen and unilateral advanced AMD; and genotype represented by CFH-rs1061170 and ARMS2rs10490924. The AUC with genetic risk factors was 0.872, which was only slightly higher than the AUC without genetic risk factors (AUC = 0.865).

"This suggested that we can predict with high accuracy individuals at high risk of developing advanced AMD without necessarily doing genetic testing," Dr. Kim said. A positive predictive value is another measure of clinical validity (how many individuals who are identified as high risk actually develop advanced AMD).

"This correlates with disease prevalence," she said. "We will have a better positive predictive value if the test is applied to older patients with early-stage disease, rather than testing young unaffected individuals."