CONTROVERSIES IN CARE

Genetics and AMD: Can We Take Advantage of the Code Yet?

EDITED BY MICHAEL COLUCCIELLO, MD

In the past, we determined that there was a genetic basis for the acquisition of wet AMD (Figure 1), when it was noted that half of the children who have one parent with the disease later develop the condition, whereas only 12% of those without an affected parent do.¹

We are in a different era now. Since the human genome has been sequenced, single-nucleotide polymorphisms (SNPs) have been identified that confer predisposition to disease, including AMD.

SNPs related to genes that code for regulation of steps in the complement cascade system, cholesterol metabolism, protection from oxidative stress in the retinal mitochondria, and extracellular matrix remodeling have been associated with increased AMD risk in individuals.

The complement system includes an amplifying cascade of proteins in the bloodstream that constitute the innate immune system. The system “complements” antibodies and white blood cells to clear pathogens via generation of a membrane attack complex, disrupting their membranes through pore formation. SNPs of complement factor H (several haplotypes)^2,3 and complement factor I (inhibitors of C3 activation) and of C3, C2, and complement factor B (activators of the pathway) have been associated with AMD.⁴

Figure 1. Wet AMD, seen on fundus photography (upper left), a red-free fundus photo (upper right), fluorescein angiography (lower left), and OCT (lower right).

SNPs in enzymes involved in cholesterol metabolism are associated with AMD risk. This association is interesting because drusen contain cholesterol. The loci include the hepatic lipase gene (LIPC), lipoprotein lipase gene (LPL), cholesterol ester transferase gene (CETP), and ATP binding cassette transporter (ABCA1) gene.⁵

The ARMS2 (age-related maculopathy susceptibility 2) protein is localized in the ellipsoid region of the photoreceptors, which is rich in mitochondria. This region is important in energy metabolism and in combating oxidative stress. An ARMS2 gene SNP is associated with AMD.⁶

SNPs in the TIMP3 (tissue inhibitor of matrix metalloproteinase 3) gene also predict wet AMD, and this gene codes for peptidases involved in extracellular matrix degradation.⁵

The more of these SNPs that exist in any given individual, especially with the presence of environmental risk factors, such as tobacco exposure, the greater the risk. This would seem to set up a system whereby we can calculate risk based on the genetic analysis of a buccal smear, but is it worth the cost? That is, how much more accurately will genotyping assist in our counseling of patients over known phenotypical risk factors, such as intermediate and large drusen, macular pigmentation, and late AMD in the fellow eye? Can the presence or absence of SNPs in these regions predict response to different forms of treatment?

We are fortunate to have the perspectives of Mathew MacCumber, MD, PhD, professor and associate chairman for research, Department of Ophthalmology, Rush University Medical Center and Illinois Retina Associates, Milad Hakimbashi, MD, a fellow at Rush and Illinois Retina, and Carl Awh, MD (president of Tennessee Retina PC), share their views on this emerging area.

REFERENCES

1. Klaver CC, Wolfs RC, Assink JJ, et al. Genetic risk of age-related maculopathy. Population-based familial aggregation study. Arch Ophthalmol. 1998;116:1646-1651.

2. Klein RJ, Zeiss C, Chew EY, et al. Complement factor H polymorphism in age-related macular degeneration. Science. 2005;308:385-389.

3. Haines JL, Hauser MA, Schmidt S, et al. Complement factor H variant increases the risk of age-related macular degeneration. Science. 2005;308:419-421.

4. Gold B, Merriam JE, Zernant J, et al. Variation in factor B (BF) and complement component 2 (C2) genes is associated with age-related macular degeneration. Nat Genet. 2006;38:458-462.

5. Chen W, Stambolian D, Edwards AO, et al. Genetic variants near TIMP3 and high-density lipoprotein-associated loci influence susceptibility to age-related macular degeneration. Proc Natl Acad Sci U S A. 2010;107:7401-7406.

6. Kanda A, Chen W, Othman M, et al. A variant of mitochondrial protein LOC387715/ARMS2, not HTRA1, is strongly associated with age-related macular degeneration. Proc Natl Acad Sci U S A. 2007; 104:16227-16232

An Approach to Genetic Testing for AMD

MILAD HAKIMBASHI, MD • MATHEW W. MACCUMBER, MD, PhD

While the majority of patients with AMD have the dry type and maintain relatively good vision, approximately 15% will develop CNV, which can have devastating visual consequences if left untreated. The typical standard of care is to follow patients with nonexudative AMD at six- to 12-month intervals, recommend AREDS supplements, and advise them on home monitoring, such as use of the Amsler grid. However, Amsler grid changes do not have a high predictive value unless a substantial amount of fluid or blood has accumulated in the macula.

Some ophthalmologists see higher-risk patients more frequently, and they may employ testing, such as SD-OCT (Figure 2), to detect subclinical fluid with the hope of treating CNV sooner. However, studies are as yet limited on the benefit of such an approach.

Single-nucleotide polymorphisms in multiple genes have been linked to the development of AMD. The most significant genes code for CFH, found in the RPE, and ARMS2, found in the retinal mitochondria.

Two companies currently offer genetic tests for patients with AMD that rely predominantly on SNPs in these genes: Macula Risk, marketed by ArcticDx, and Retnagene, marketed by Sequenom. Both tests are reimbursed by Medicare and by most private insurers for patients carrying the diagnosis of drusen or AMD.

In response to the proliferation of genetic testing for many ocular diseases, including AMD, the AAO convened a Task Force on Genetic Testing on their use, which included one of the authors (MM) as a member.

The group’s report in the November 2012 issue of Ophthalmology advised against routine genetic testing for AMD “until specific treatment or surveillance strategies have been shown in one or more published clinical trials to be of benefit to individuals with specific disease-associated genotypes.”¹

Pivotal to the value of a genetic test for AMD would be, first, its ability to predict AMD and, second, its ability to predict CNV, because this is the form of the disease that can be successfully treated using injections of anti-VEGF drugs.

To compare the performance of the two available tests for these predictions, Drs. Nancy Holekamp, Arghavan Almony, and MacCumber sent DNA samples from 80 AMD patients and 17 normal individuals to both Sequenom and ArcticDx for the AMD genetic tests available in 2011-2012 (the PIER study). Both companies performed these tests free of charge.

Figure 2. Clinicians may employ OCT in addition to fundus photos in cases of high risk.

As illustrated by the analysis performed by Dr. Edwin Stone, presented at the AAO annual meeting and on his Web site,² genetic testing by both companies was quite good at predicting AMD (P = 1.4 × 10-5 and 2.9 × 10-5, respectively) but weak at predicting the development of CNV (P = 0.81 and 0.74, respectively).

Although not a validation trial, this small study suggested that genetic testing alone would not be of significant value for predicting CNV, but it may be useful for diagnosing AMD in patients with indefinite phenotypes (for example, if macular findings suggestive of polypoidal choroidal vasculopathy or central serous chorioretinopathy were also present).

Similarly, genetic testing alone may give information to family members of AMD patients about their risk of developing AMD before they demonstrate macular findings on clinical examination.

Newer predictive algorithms by both Sequenom and ArcticDX now incorporate phenotype with age, smoking history, and genotype. The tests now appear more predictive of CNV development.

A data analysis comparing the development of CNV in the AREDS database and a multifactorial model presented at the 2012 AAO annual meeting and submitted to Ophthalmology by Sequenom found a C-index value of 0.89 for phenotype alone (employing the AREDS Simplified Severity Scale), 0.90 when age and smoking history were included, and 0.96 when genetic testing was added (1.0 is ideal, P < 0.01 for phenotype vs the full model).³

The new tests from ArcticDX and Sequenom may play a role in the management of some well-informed middle-aged patients with nonexudative AMD, as they might help these patients modify their environmental risk factors, notably smoking and antioxidant consumption, and increase surveillance that might help limit the burden of CNV prior to intervention with anti-VEGF treatment.

The predictive value of the tests will likely increase further as more genetic, environmental, and clinical factors are linked to the development of CNV for individual AMD patients.

However, given the lack of a treatment or validated surveillance strategy proven to alter the natural course of the nonexudative form of the disease significantly, routine testing is unlikely to have a dramatic effect in the clinical outcomes for most AMD patients, and thus is still not warranted, as the AAO Task Force has recommended.

REFERENCES

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

2. Stone EM. Genetic testing for macular degeneration. Available at: http://ivr.uiowa.edu/stone/files/AAO_AMD_Genetic_Testing_text_nf1-2.pdf. Accessed March 1, 2013.

3. Perlee LT, Bansal AT, Gehrs K, et al. Inclusion of genotype with fundus phenotype improves accuracy of predicting choroidal neovascularization and geographic atrophy. Ophthalmology, submitted.

Genetic Testing for AMD

CARL AWH, MD

I am an equity investor and paid consultant to an AMD genetic testing company. I got involved with this industry prior to the commercial availability of the tests because I wanted to learn more about genetic testing and personalized medicine.

I thought that one day a genetic test might be as essential to the complete evaluation of a patient with AMD as is a lipid panel to a complete physical exam. We are far from that day, but the field has grown quickly, and it is encouraging that this subject is now of general interest.

The scope of this article is inadequate to address the potential uses of genetic testing for determining individual response to therapy. Some presented results are promising; others suggest that the genes that predict risk may not be the same ones that predict response to therapy. Likewise, a discussion of the cost-effectiveness of genetic testing for AMD must wait for another day, but formal analyses are under way, and I predict favorable results.

Imagine a screening strategy in which every patient with moderate AMD is evaluated monthly with a dilated retinal exam and OCT. This strategy would lead to earlier detection of many cases of neovascular AMD and, therefore, better vision for treated patients,¹ but it is logistically and economically impractical. It would be akin to performing a cardiac stress test on every adult, rather than on those with a higher than average risk of coronary heart disease (CHD).

The Framingham risk score for CHD is used as a basis for recommending lifestyle changes, diagnostic testing, and preventive medical therapies. Most ophthalmologists are aware of the importance of modifying risk factors for CHD, but few know that current models of AMD risk progression, combining genetic and nongenetic variables, have predictive power equal or superior to the Framingham risk score for CHD.²

The AAO has recommended against routine genetic testing for AMD, and I agree. However, I also think that information provided by the selective use of commercially available genetic tests for AMD offers real and measurable benefits to patients.

The purpose of a genetic test for AMD is not to “identify treatable disease,” as has been implied by the AAO task force statement.

A random static sample of patients with advanced AMD (like the study cited by Drs. MacCumber and Hakimbashi) will contain relatively few patients with elevated genetic risk, because most people (approximately 80%) have normal or low genetic risk.

However, prospective studies have demonstrated a significantly greater incidence of progression to advanced AMD (Figure 3) among patients with elevated genetic risk. A recent study by Yu et al. 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 NV in five years and a 26% probability of developing NV in 10 years.³ Isn’t it reasonable to monitor these high-risk patients more frequently?

Figure 3. Greater incidence of progression to advanced AMD has been tied to genetics.

So now imagine a screening strategy in which patients with higher than normal genetic risk are evaluated more intensively. What is the optimal schedule? Would it be to wait for a “validated surveillance strategy” that holds genetic testing to an unfair standard? Name a “validated surveillance strategy” for any other retinal condition!

Informed and thoughtful clinicians will use the results of genetic tests to influence the ways in which we manage our patients. We will learn by doing, and in the process, we will advance the practical use of this revolutionary technology far more rapidly than would ever be possible in the confines of a pure research setting.

We will identify patients at increased risk and identify treatable disease earlier. Even with current therapies, we will achieve better outcomes. When new preventive therapies arrive, the benefits will be even greater. RP

REFERENCES

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

2. Seddon JM, Reynolds R, Maller J, et al. Prediction model for prevalence and incidence of advanced age-related macular degeneration based on genetic, demographic, and environmental variables. Invest Ophthalmol Vis Sci. 2009;2044-2053.

3. Yu Y, Reynolds R, Rosner B, et al. Prospective assessment of genetic effects on progression to different stages of age-related macular degeneration using multistate Markov models. Invest Ophthalmol Vis Sci. 2012;1448-1556.