Genetic Testing for AMD

Can the latest risk prediction models improve patient care?

Genetic Testing for AMD

Can the latest risk prediction models improve patient care?


Since the double helix was first described 60 years ago, the pace at which the field of genetics has advanced has been quite remarkable. That Nobel Prize-winning discovery paved the way for the mapping of the human genome, a milestone reached in 2003, as well as the creation of the HapMap in 2005. With these powerful tools, researchers have been able to explore the role that genes play in our health by tracking complex diseases that involve multiple genetic factors. Such diseases — heart disease, cancer and diabetes, for example — have a major impact on individuals and society.

The Human Genome Project has already enabled the identification of more than 1,800 disease-influencing genes, and more than 2,000 genetic tests for human conditions are now available.1 However, in some ways, medical science is struggling to keep up. Much of the newly obtained knowledge has yet to be put to use with targeted clinical treatments or preventive interventions. Furthermore, exactly how heredity and environment interact remains unclear for most complicated diseases. As retina physicians, we find ourselves at a crossroads in this evolving landscape.

Decoding AMD

Through the work of several groups of researchers, nearly 20 genes and variant loci have been linked, some more strongly than others, to an increased risk of AMD. AMD-related single-nucleotide polymorphisms (SNPs) have been found near or within genes responsible for a variety of functions, including extracellular matrix remodeling, oxidative stress protection in the retinal mitochondria, the complement system and cholesterol metabolism.

The discoveries made so far have been used as the basis for commercially available genetic tests designed to identify patients with early or intermediate AMD who are at risk for progressing to advanced disease. Two such tests, Macula Risk NXG (ArcticDx) and RetnaGene AMD (Sequenom), are frequently referenced when discussing what value their information may provide doctors and patients. Both tests are available only to licensed medical professionals for use in patients who have a diagnosis of AMD. ArcticDx specifies that its Macula Risk NXG test is recommended for all AMD patients except those with AREDS stage 1 disease who are under 50 years of age and who don’t have a family history of AMD. Sequenom advises that its RetnaGene AMD test is meant to identify Caucasian patients age 55 and older who are at risk of progressing to wet AMD.

To use the tests, a doctor obtains kits from the testing company. The kits include forms for the doctor and patient to fill out and the materials needed for collecting a DNA sample with a cheek swab. The samples are sent in prepaid envelopes to the respective labs for analysis, and the doctor receives the results in 3 weeks or less, depending on the company. Genetic counselors are available free of charge to help patients and doctors interpret and understand test results. Historically, insurance plans, including Medicare, have covered the costs of testing. Similar to a primary care doctor ordering a lipid panel, the ophthalmologist doesn’t collect a fee for the testing, and the laboratories bill insurers on behalf of the patients.

The Macula Risk NXG test calculates a patient’s 2-, 5- and 10-year risk of progression to advanced AMD based on both genetic and non-genetic factors. The report compares the patient’s risk to the average for those with the same phenotype in the AREDS trial. The report also provides a “lifetime” genetic risk score based on genetic risk only. Doctors may request a recalculation of a patient’s 2- to 10-year risk as age and AMD status change. An online tool that will enable doctors to recalculate risk based on any non-genetic variable is expected to become available in the near future.

The RetnaGene AMD test calculates the risk of progression to advanced disease in 2, 5 and 10 years, and the patient is categorized as high, moderate or low risk. The categories reflect the patient’s risk relative to subjects of the same grade in the AREDS clinical study that was used to validate the test. High risk is assigned to a patient profile found in the top 25% of profiles observed in the AREDS validation cohort with the same stage of disease. Moderate risk reflects a patient profile found in the middle 50% of the AREDS study subjects with the same stage of disease. Low risk is assigned to a patient profile found in the bottom 25% of profiles observed in the AREDS validation cohort with the same stage of disease.

A relatively new development in the area of genetic testing for AMD is the introduction of second-generation tests, which take into account not only genotype but also phenotype based on the AREDS Simplified Severity Scale, and known environmental risk factors for AMD such as smoking history and body mass index (BMI).

What is the Clinical Value of AMD Genetic Tests?

While some retinal specialists have chosen to incorporate a second-generation AMD genetic test into their practices, others have not. (See “Do You Use AMD Genetic Testing in Your Practice?” on page 12.) Many doctors who aren’t utilizing the tests point to the fact that we’ve known for years that AMD phenotype tells us what we need to know about how likely a patient is to progress to advanced disease. They doubt the addition of a genetic test can significantly affect outcomes for most patients because other than reducing risk with the use of AREDS dietary supplements, we have no intervention that prevents the conversion from early and intermediate to advanced AMD. The rationale for potentially alarming patients by introducing hard-to-grasp and inconclusive genetic information has also been questioned, especially because even patients who are considered at high risk by today’s standards may never advance to choroidal neovascularization (CNV) or geographic atrophy (GA).

The gene variants associated with AMD progression haven’t consistently predicted a response to anti-VEGF therapy,2,3 and this has also been cited as a drawback of genetic testing. The American Academy of Ophthalmology (AAO) Task Force on Genetic Testing has made known its position on the issue as well.4 It released a report last year stating that genetic testing for complex diseases will not be relevant to the routine practice of medicine until “clinical trials can demonstrate that patients with specific genotypes benefit from specific types of therapy or surveillance.” Until such benefit can be demonstrated, the report continued, “the routine genetic testing of patients with complex eye diseases, or unaffected patients with a family history of such diseases, is not warranted.”


“A relatively new development in the area of genetic testing for AMD is the introduction of second-generation tests, which take into account not only genotype but also phenotype based on the AREDS Simplified Severity Scale, and known environmental risk factors for AMD, such as smoking history and body mass index (BMI).”

— Carl C. Awh, MD

Analyses of other aspects of genetic testing for AMD, such as cost-effectiveness, are under way, and the results should be enlightening.

Risk Prediction Models Are Improving

I count myself among those who believe that genetic testing for select AMD patients can be a valuable tool for helping us to safeguard vision. I don’t advocate testing every patient who exhibits signs of AMD, nor do I expect the tests to predict response to therapy. The discoveries of the target gene variants were predicated on their association with risk of progression to advanced disease, and it is reasonable that the metabolic pathways that influence response to therapy may be very different. Therefore, in my mind, this is not an issue that casts doubt on this area of science.


“Prospective studies have shown that patients with elevated genetic risk have a significantly greater incidence of progression to advanced AMD.”

— Carl C. Awh, MD

With the inclusion of phenotype and environmental factors, the power of the second-generation genetic tests to predict risk of progression to advanced AMD appears to have increased. Several studies have shown the combination of genetics with clinical assessment of an AMD patient to be a very accurate predictor. For example, using an AMD risk model that included age, gender, education, baseline AMD grade, smoking and BMI, Seddon and colleagues5 plotted risk at 0.757 in the area under the curve (AUC/C-statistic). While baseline grade of AMD was the strongest predictor of risk of progression to advanced AMD in these models, the addition of genetic factors (SNPs in CFH, ARMS2, C2, CFB and C3) increased the AUC to 0.831.5

Another study, conducted by Klein and colleagues,6 confirmed that macular phenotype according to the AREDS Simplified Severity Scale had the greatest predictive value for disease progression. They considered age, family history, smoking, the AREDS Simplified Severity Scale score, the presence of very large drusen, the presence of advanced AMD in one eye, and two polymorphisms in CFH and ARMS2. The AUC was 0.872 with genetic factors included and 0.865 without.6 To determine whether including genetic markers improved the accuracy of predicting conversion from early-stage AMD to CNV or GA compared with prediction using phenotype alone, Perlee and colleagues7 genotyped more than 2,000 patients who were participating in the Age-Related Eye Disease Study for 14 AMD-related SNPs. They also evaluated age, sex, smoking status, BMI, AREDS treatment category and educational level. The CNV prediction models that combined genotype with phenotype (with or without age and smoking) performed the best with a C-statistic of 0.96. The phenotype-alone model based on the simplified severity scale and presence of CNV in the nonstudy eye yielded a C-statistic of 0.89. Similarly, for predicting progression to GA, the model that combined genotype with phenotype demonstrated the highest performance with a 0.94 C-statistic.

Other prospective studies have shown that patients with elevated genetic risk have a significantly greater incidence of progression to advanced AMD. Yu and colleagues8 found that patients with large drusen and low genetic risk have less than a 1% probability of developing GA or CNV in 10 years, while patients with the same phenotype and high genetic risk have a 15% probability of developing CNV in 5 years and a 26% probability of developing CNV in 10 years.8

The AUC/C-statistic values that have been achieved by combining genetic risk factors with phenotype in these AMD studies are equal to or higher than those for the Coronary Heart Disease Risk Score derived from the Framingham Heart Study, which have been reported as 0.6-0.8 in published papers.

An Informed Approach

The approach to genetic testing for AMD that we have chosen to take in our practice is based on the research and predictive models behind the tests, as described above, as well as the known benefits of early detection and treatment of neovascular AMD. Several studies, including a subgroup analysis from the MARINA clinical trial of ranibizumab (Lucentis) for wet AMD, have shown that earlier detection and therapy leads to better final visual acuity.9 Therefore, we recommend a genetic test for patients who have intermediate AMD in order to identify those who are at higher than normal risk of progressing to advanced disease. We then recommend that those patients monitored more frequently, which wouldn’t be practical to suggest for every patient with intermediate AMD.

This strategy is especially useful for patients with intermediate AMD but good vision. Without knowing they’re at higher than normal risk for advanced disease, we would not likely be seeing them more than yearly. And because they aren’t having problems seeing, they are less likely to monitor their vision with the Amsler grid and therefore are less likely to notice changes. Monitoring them more frequently gives us a better chance of detecting a conversion to advanced disease sooner rather than later. Furthermore, a patient with intermediate AMD who learns he or she has a significant genetic risk of advanced AMD may be more likely to make lifestyle changes to reduce the risk, comply with the recommendation to use AREDS dietary supplements, keep follow-up exam appointments, and pay more attention to changes in vision. We’re seeing this effect in our practice. Patients who have the information provided by a genetic test tend to become “better patients.” We have also detected and treated asymptomatic wet AMD in patients who have returned to the office for a visit that was scheduled on a shorter interval on the basis of a genetic test result. One of our partners has submitted a paper detailing our experience with this, and we hope it will be accepted for presentation at the annual meeting of the American Society of Retinal Specialists this summer.

It is crucial for patients who undergo an AMD genetic test to receive appropriate education so they can put the results in the proper context. Although they can take advantage of the genetic counselors available to them through the testing companies, we take the time to counsel them ourselves. We ensure they understand the purpose of the test, which is to determine whether they are at higher than normal risk of developing advanced AMD. We emphasize that above average risk does not mean progression is a certainty, and that below average risk does not mean the disease will not progress.

Importance of Genetic Testing Will Continue to Increase

The predictive power of AMD genetic testing will be even more important for retinal physicians and their patients when treatments targeting specific stages of the disease are developed. In fact, the predictive models will help to inform this development. The most recent research conducted by Dr. Seddon’s group indicates that genes in different pathways influence progression to different stages of AMD, which means we will be able to determine which patients are more likely to move from normal to early disease, from early to moderate and from moderate to severe disease. The group has reported, for example, that when demographic and behavioral factors and other SNPs were controlled for, the CFH, C3, CFB and ARMS2/HTRA1 genes were associated with progression from intermediate drusen to large drusen and from large drusen to GA or CNV.8 They also found the TT genotype of the rs10468017 polymorphism in the hepatic lipase gene to be associated with decreased risk of progression from normal to intermediate drusen and from large drusen to neovascular AMD.8 They recently validated the model used to make these and other determinations in an independent and external cohort.10

Until the time when our ability to specifically predict onset of the various stages of AMD and our ability to treat them are perfectly matched, we can use genetic testing to identify the subset of patients who have an increased risk for vision-threatening disease. We can evaluate them more frequently and educate them aggressively. Targeting our resources in this manner is logistically practical and enables us to detect and treat advanced disease as early as possible. ■

Dr. Awh is an internationally recognized clinician, surgeon and educator. He practices with Tennessee Retina, which is based in Nashville and serves patients in 12 locations throughout Middle Tennessee and Southern Kentucky. Dr. Awh has an equity investment in and is a member of the scientific advisory board for ArcticDx Inc., the company that developed and distributes the Macula Risk test.


1. Research Portfolio Online Reporting Tools [Internet]. Bethesda, Md.: U.S. Department of Health and Human Services. National Institutes of Health. Human Genome Project Fact Sheet [cited 2013 Apr 15]. Available from:

2. Stone EM, Aldave AJ, Drack AV, et al. Recommendations for genetic testing ofinherited eye diseases: report of the American Academy of Ophthalmology Task Force on Genetic Testing. Ophthalmology 2012;119:2408-2410.

3. Lee AY, Raya AK, Kymes SM, Shiels A, Brantley MA Jr. Pharmacogenetics of complement factor H (Y402H) and treatment of exudative age-related macular degeneration with ranibizumab. Br J Ophthalmol 2009;93:610-613.

4. Hagstrom SA, Ying GS, Pauer GJT, et al. Pharmacogenetics for genes associated with age-related macular degeneration in the Comparison of AMD Treatments Trials (CATT). Ophthalmology 2013;120(3):593-599.

5. Seddon JM, Reynolds R, Maller J, Fagerness JA, Daly MJ, Rosner B. Prediction model for prevalence an incidence of advanced age-related macular degeneration based on genetic, demographic, and environmental variables. Invest Ophthalmol Vis Sci 2009;50(5):2044-2053.

6. Klein ML, Francis PJ, Ferris FL 3rd, Hamon SC, Clemons TE. Risk assessment model for development of advanced age-related macular degeneration. Arch Ophthalmol 2011;129(12):1543-1550.

7. Perlee LT, Bansal AT, Gehrs T, et al. Inclusion of genotype with fundus phenotype improves accuracy of predicting choroidal neovascularization and geographic atrophy. Ophthalmology 2013; pii: S0161-6420.

8. Yu Y, Reynolds R, Rosner B, Daly MJ, Seddon JM. Prospective assessment of genetic effects on progression to different stages of age-related macular degeneration using multistate Markov models. Invest Ophthalmol Vis Sci 2012;53(3):1548-1556.

9. Boyer DS, Antoszyk AN, Awh CC, et al. Subgroup analysis of the MARINA study of ranibizumab in neovascular age-related macular degeneration. Ophthalmology 2007;114(2):246-252.

10. Seddon JM, Reynolds R, Yu Y, Rosner B. Validation of a prediction algorithm for progression to advanced macular degeneration subtypes. JAMA Ophthalmol 2013;131(4):448-455.

Do You Use AMD Genetic Testing in Your Practice?


images I began utilizing AMD genetic testing in my practice approximately 2 years ago. In December 2012, I switched to a second-generation test as soon as one became available in my state. The new test is greatly improved as it takes into account the phenotype of the eye, as well as the genotype. I recommend the test for Caucasian patients age 55 or older who have early or intermediate AMD. Along with a thorough examination, fluorescein angiography and OCT imaging, I consider identifying the patients who are at high risk for progressing to CNV based on the presence of AMD-associated single nucleotide polymorphisms to be an integral part of a complete evaluation.

When the test results come in, I ask the patient to schedule an appointment with me during which I personally review the report with him or her. For those identified as high risk, I present a management program and reiterate the importance of modifying the risk factors within their control, monitoring their vision for metamorphopsia and calling the office immediately if they notice it. I discuss the importance of early detection and treatment. AMD screening for their children and siblings is also part of the discussion. Although I routinely educate my AMD patients on these points, the reality of their situation often does not “hit home” until they learn of their genetically elevated risk of potential vision loss. At that point, they’re more likely to become more vigilant in following my recommendations for taking AREDS vitamins, watching their weight, cholesterol and blood pressure, quitting smoking, eating more antioxidant-rich foods, exercising and protecting their eyes from the sun.

I increase surveillance of high-risk patients according to the following guidelines I developed:

Interestingly, I recently asked a patient identified as high risk by a genetic test to return for a follow-up visit in 3 months. When she did, she reported that she hadn’t noticed any changes in her vision. Her visual acuity was 20/25 OU, but OCT revealed she had nascent bilateral CNV with slight subretinal hemorrhage. We began anti-VEGF therapy right away. In light of cases such as this, I think the value of preventative management for dry AMD can’t be underestimated. In my opinion, if we don’t use knowledge of patients’ genetic profile as a tool to help us understand this complex disease and optimize our management protocols, we are shortchanging ourselves — and them.

While I appreciate the American Academy of Ophthalmology’s evaluation of whether we should be using AMD genetic testing in clinical practice at this time, I don’t agree with its conclusion that the testing is not warranted. The Academy’s task force report stated that this type of testing should not be used until a sufficient infrastructure is in place in the medical community for proper patient counseling and we have clinical evidence that it significantly improves outcomes. Unfortunately, it will take some time to collect these data. I agree that applying AMD genetic testing in the vitreoretinal practice is in its infancy, but I don’t feel that should preclude us from using it. I prefer to embrace this new frontier, the impact of which will undoubtedly change over time, and grow with it rather than exclude it. ■


“In my opinion, if we don’t use knowledge of patients’ genetic profiles as a tool to help us understand this complex disease and optimize our management protocols, we are shortchanging ourselves — and them.”

—Pamela Ann Weber, MD

Dr. Weber is the founder of Island Retina, which serves patients in two offices and a surgery center in Suffolk County, N.Y. Dr. Weber reports that she has no financial interest in the products mentioned in this article.

Follow-Up Schedule in Months for Binocular Patients
Simple Severity Grade 4 12 6 4
3 12 6 6
2 12 6 6
1 12 12 12
  Low Moderate High
Risk Category as Predicted by the RetnaGene AMD Test

Follow-Up Schedule in Months for Monocular Patients
(low vision in one eye from advanced AMD)
Simple Severity Grade 4 6 4 3
3 12 6 4
2 12 12 6
  Low Moderate High
Risk Category as Predicted by the RetnaGene AMD Test


images I currently do not use a genetic test for AMD in my practice. I agree with the assessment of the AAO Task Force, which recommended against routine genetic testing for complex ocular diseases.* In my opinion, at this time, the potential harm outweighs the benefits. To properly frame this issue, it’s important that we distinguish between genetic testing and genetic screening. There is great value in genetic testing for certain diseases, those in which the test results allow us to make a definitive diagnosis or provide patients with a more solid understanding of what the course of the disease will be. This is the case for conditions such as birdshot choroidopathy and Leber congenital amaurosis, but it’s not the case for AMD. Screening every AMD patient for contributory genes is premature at this time. While a genetic test can tell us whether a patient carries such genes, it can’t tell us with any certainty what that means for his or her future. With this particular disease, that simply has not been defined.

Any implementation of genetic screening for AMD patients on a large scale should be driven by evidence of its value, which has not yet emerged. Determining the value might be feasible if we had a proven treatment for preventing the conversion from early or intermediate AMD to advanced AMD, which we do not. It’s been suggested that the value of screening lies in more frequent monitoring of patients labeled as high risk in order to detect and treat neovascular disease sooner. However, again, we have no evidence of the effectiveness of that approach, which could create an unnecessary burden for patients and a drain on practice resources. It’s also been suggested that patients who test as higher risk can be more strongly encouraged to make lifestyle changes and take the AREDs dietary supplements. That sounds reasonable enough, but we can take those steps without a genetic test, based on phenotype, well-known environmental risk factors, and very specific clinical findings derived from AREDS, a large, prospective, randomized clinical trial. In other words, the results of a genetic test don’t provide us with any information we can act on with confidence beyond what we would have already known.

We must take into account, too, the downsides to genetic screening for AMD patients. The expense for the healthcare system in the absence of a clear benefit is one important consideration. Another is the potential to create a huge amount of anxiety for patients by telling them there’s a good chance they will go blind when it may or may not occur. Yet another consideration are the repercussions patients could face as far as their insurability or employability. The Genetic Information Nondiscrimination Act has been in place since 2008, but in a field moving as quickly as genetics, we don’t know for sure how protective it may or may not be.


I currently do not use a genetic test for AMD in my practice. I agree with the assessment of the AAO Task Force, which recommended against routine genetic testing for complex ocular diseases.* In my opinion, at this time, the potential harm outweighs the benefits.”

—Pravin U. Dugel, MD

I plan to wait for supporting Level I evidence before I genetically screen AMD patients in my practice. As I see it, the importance and reach of the issues involved require it. If we stop to think about it, we are reminded of the necessity of a thorough study, given the gravity of its implications. Subfoveal surgery, radial optic neurotomy, macular translocation and laser for drusen are only a handful of many interventions that at one time were considered by some to provide obvious benefits. The evidence proved otherwise. What a disservice it would have been for patients had we not insisted on scientific evidence. I am not comparing genetic testing to these discarded procedures concretely, but rather drawing the parallel on how we evaluate evidence — and asking that we be data-driven scientists.

I have no doubt that someday genetic screening will be the most important tool for diagnosing and treating patients with many diseases, including AMD. In the meantime, we should take a step back in the best interest of our patients. If we don’t, we risk derailing the development of a well-conceived and well-designed screening strategy that is based on scientific rigor. ■

Dr. Dugel is Managing Partner of Retinal Consultants of Arizona in Phoenix; Clinical Associate Professor of Ophthalmology, Doheny Eye Institute, Keck School of Medicine at the University of Southern California, Los Angeles; and Founding Member of the Spectra Eye Institute in Sun City, Ariz. He is a shareholder in ArcticDX.