Characterization of genes that are associated with age-related macular degeneration (AMD) progression and response to treatment has been a challenge secondary to the variable phenotypes, multiple genes involved, and high incidence of this disease in aging populations. However, despite the challenges, many important genetic associations have been characterized that when used together can predict progression to advanced AMD with similar odds ratios to the association of tobacco smoking with the development of lung cancer. Moreover, there were many important advances reported at this year's Association for Research in Vision and Ophthalmology (ARVO) meeting to allow clinicians to better predict, characterize, and treat AMD patients.

Background

Familial studies showing that there is a 3- to 6-fold higher risk for primary relatives of AMD patients compared to age-matched controls and the high concordance of AMD in monozygotic twins provided the basis to study the genetic linkage of larger AMD populations. The ABCR gene, which was originally linked to Stargardt's disease, was initially studied as a putative AMD-related gene. Although a strong association was not found with AMD and small nucleotide polymorphisms (SNPs) in the ABCR gene, a relatively strong association was found with the complement factor H (CFH) gene. In fact, tobacco smoking both worsens the prognosis of AMD clinically and decreases CFH gene function, further providing support that CFH genetics are tightly linked to AMD progression. Linkages among the APoE and ARMS2 genes were then reported to be associated with AMD progression. Over the past few years, many more linkages of SNPs to AMD have been reported in many gene products so that there are 5 major classes of genes involved in AMD progression: (1) complement pathway genes, CFH, C3, CFB/C2, CFI; (2) cholesterol metabolism genes, APoE, LIPC, CETP; (3) mitochondrial genes, ARMS2, ND2; (4) extracellular matrix regulating genes, TIMP3; and (5) microRNA processing enzymes, DICER1. By combining haplotype odds ratios of multiple SNPs from some of these genes, smoking history, and drusen size, our ability to predict the risk of individuals with AMD to progress to advanced sight-threatening AMD is so good that it approximates the level of correlation between tobacco smoking and the development of lung cancer, odds ratio > 17. Two commercial buccal swab sampling tests, RetnaGene (Sequenom) and Macula Risk (ArcticDx), combine the haplotype ORs of multiple SNPs in the complement and oxygen metabolism pathways. Macula Risk is unique in that it also incorporates smoking history and drusen size. RetnaGene assigns a low, medium, or high risk, whereas Macula Risk assigns a relative risk score of 1-5 (Figure 1). The Macula Risk algorithm will add 7 new SNPs to its algorithm over the next few months.

New Genetic Associations with AMD Progression

There were many novel genetic associations with AMD progression reported this year at ARVO, which will likely permit even better predictability when combined with the SNPs characterized to date. One method to detect genes that are linked to disease is GWAS, genome-wide association studies. DNA is harvested from patients, and microarray technology is used to identify SNPs that segregate with disease states. A large cohort of AMD patients was studied and reported at ARVO, where over 2.5 million SNP's from either 7650 individuals with advanced AMD or 51 812 age-matched controls were analyzed by this method.1 Initially, 32 candidate SNPs were identified with a P < 10-5, and 19 SNPs had significance of P < 10-8. Of the 19 identified highly significant SNPs, 12 have been previously characterized as associated with AMD progression, which provides good corroboration with this and other studies. The 7 novel genes that were identified include COL8A1/FILIP1L, IER3/DDR1, SLC16A8, TGFBR1, RAD51B, MIR548A2, and B3GALTL. These gene products are involved in angiogenic regulation, extracellular signaling, and apoptosis, indicating a physiologic basis to suggest that these newly identified genes are also significant markers for AMD progression. Another AMD-related gene was identified using an exon-based whole genome strategy from eye donors affected with AMD in addition to a custom retinal pigment epithelial (RPE) gene expression profiling strategy, termed CHANGE analysis. The study was directed to identify genes involved in RPE phagocytosis, a known pathophysiologic event for AMD progression. An association with ADAM15, metargidin gene, was characterized where ADAM15 mRNA was preferentially upregulated in AMD diseased eye bank eyes, as well as in mouse models of RPE phagocytosis and smoking exposure.2 The ADAM15 gene product has both metalloprotease and disintegrin functions so that the tissue remodeling role of this gene product is consistent with the linkage of other AMD-associated genes, such as TIMP3, with a similar mechanism of action. Sirt1 represents another gene that has a mechanism known to be associated with AMD progression, NAD+/-dependent deacetylase, which mediates many pathways related to mitochondrial oxidative metabolism.3 Significant reduction of sirt1 expression was observed in donor eyes from AMD patients relative to age-matched controls, and partial blockade of sirt1 activity in human RPE cell lines after exposure to sirt1 siRNA modulated many markers of oxidative metabolism and increased RPE cell apoptosis, thus establishing a link between sirt1 expression and AMD progression. Further evidence that DICER1 inhibition and Alu RNA accumulation was related to AMD progression was provided by eye donor immunohistochemical and semiquantitative PCR studies.4 The mechanism of this toxicity was further elucidated in knockout mice where RPE toxicity was mediated myeloid differentiation factor 88 but not the typical Toll-like receptor RNA sensors. Thus, many new genes involved in AMD progression have been recently characterized to allow a better understanding and predictability of disease progression and to advance the design of novel therapeutic approaches.

Genetic Therapy for Macular Disease

Genetic therapy represents another horizon for the treatment of many common retinal diseases that include AMD. In primate models of retinal and choroidal neovascularization, an adenoviral vector that expresses a soluble VEGF receptor AAV-sFlt is able to successfully target the retinal tissue to inhibit both of these pathological angiogenic processes. Effective transfer of genetic material to deeper tissues in the macula has represented a challenge so that the mechanism of delivery has been limited to subretinal or suprachoroidal delivery to obtain adequate macular tissue load. Modifications of the viral capsid, however, have permitted better targeting to the retinal photoreceptors to hopefully permit genetic delivery by a less invasive and standard intravitreal injection to treat macular disease.5 Although clinical trials of genetic therapy have been limited to aggressive photoreceptor degenerations such as Lebers congenital amaurosis, the positive results from these early-stage clinical trials combined with recent advances in delivery have provided a pathway toward genetic treatment of more common macular diseases.

New Detection Schema

In addition, new detection schema were reported at ARVO that both bolster the current technology for AMD risk assessment, and permit development of new protocols to better understand the pathophysiology of this complex and variable disease. Genes involved in mitochondrial function are known to be related to AMD progression, so that nongenetic markers of oxidative metabolism may also be helpful to predict AMD progression. Combination of analysis of 4 SNPs in the CFH, C3, and ARMS2 genes with 3 proteomic markers of oxidative metabolism CML, CEP, and pentosidine were able to confer a 2-5 times greater risk of AMD progression than analysis of the 4 SNPs alone.6 Moreover, copy number variation of the CFH gene has been linked to advanced AMD.7 Therefore, the potential additional predictive power of many new SNPs associated with AMD progression will likely be further enhanced by additional screening methods such as proteomics, copy number detection, and DNA modifications such as methylation.

Pharmacogenomics of Neovascular AMD

Genetic analyses of patients undergoing anti-VEGF treatment has the potential to help to accomplish 2 of the central issues in the treatment of neovascular AMD: optimizing results anti-VEGF treatment and doing so with the lowest treatment burden. The next advances in treatment of neovascular AMD will likely be contingent upon the ability to better predict the response of antiangiogenic treatment and to develop neovascular inhibitors with novel mechanism of actions. To this end, many groups have reported associations of SNPs to differential anti-VEGF response that are contained within the CFH and APoE genes. This year, more associations were reported. Macula Risk analysis of DNA was performed on a cohort of almost 100 patients who responded similarly to anti-VEGF therapy as described in the CATT and IVAN trials. Although there was no relation of the overall macular risk score to outcome after anti-VEGF injections, the indel deletion of the ARMS2 gene was found to impart visual protection where 63 patients who harbored the indel deletion had an average improvement in the log MAR vision of 0.11 compared to a visual deterioration of 0.09 log MAR in those with the ancestral genotype (Figure 2).8 Analysis of the CATT trial patients, however, has not yet provided statistically significant association of genetic markers with AMD response. Therefore, pharmacogenomics of AMD is a rapidly evolving field, and initial studies of patients with neovascular lesions have provided both interesting results and rationale for larger studies to ultimately allow improved individualized patient care and design of new therapies.

Summary

Ten years ago, we treated neovascular AMD with thermal and photodynamic therapy laser. Little was known about the pathophysiology of dry AMD, and the genetics of AMD were not characterized beyond familial studies. The advent of anti-VEGF therapy and sophisticated genetic screening techniques have led to astounding advances in our knowledge base of AMD progression and treatment. The reports from ARVO this year both advance our understanding of this complex disease and provide framework for continued progress to optimize diagnosis, prognosis, and treatment of our patients.

Alan Franklin, MD, PhD, is with the Retina Specialty Institute in Mobile, AL. He is a Founder and Chief Medical Officer of RFE Pharma. He states that he is a consultant for ArcticDx. Dr. Franklin may be reached at afranklin@retinaspecialty.com.

  1. Schu MC, Chen W, Fritsche LG, Yu Y, Yaspan B. Meta-analysis of genome-wide association studies identifies 19 loci associated with AMD risk. Poster presented at the Annual Meeting of the Association for Research in Vision and Ophthalmology; May 2012; Fort Lauderdale, FL.
  2. Inana G, Murat C, McLaren MJ. A multi-pringed gene expression approach identifies ADAM15 (Metargidin) as a potential therapeutic target for AMD. Poster presented at the Annual Meeting of the Association for Research in Vision and Ophthalmology; May 2012; Fort Lauderdale, FL.
  3. Yoshida T, He S, Spee C, Ryan SJ, Hinton DR. The role of sirt1 in the pathogenesis of dry age-related macular degeneration. Poster presented at the Annual Meeting of the Association for Research in Vision and Ophthalmology; May 2012; Fort Lauderdale, FL.
  4. Hirano Y, Tarallo V, Gelfand S, et al. Alu rna-induced cytotoxicity in age-related macular degeneration is mediated by Myd88, but not by a variety of rna. Poster presented at the Annual Meeting of the Association for Research in Vision and Ophthalmology; May 2012; Fort Lauderdale, FL.
  5. Boye SE. Transduction of the outer retina. Paper presented at the Annual Meeting of the Association for Research in Vision and Ophthalmology; May 2012; Fort Lauderdale, FL.
  6. Crabb JW, Jang G-F, Zhang L, et al; Clinical Genomic and Proteomic AMD Study Group. Carboxymethllysine and pentosidine with genotype as predictors of AMD. Poster presented at the Annual Meeting of the Association for Research in Vision and Ophthalmology; May 2012; Fort Lauderdale, FL.
  7. Cantsilieris S, White SJ, Guymer RH, Baird PN. Assessment of copy number variation in genes associated with in age-related macular degeneration (AMD). Poster presented at the Annual Meeting of the Association for Research in Vision and Ophthalmology; May 2012; Fort Lauderdale, FL.
  8. Franklin AJ, Shuler MF, Gupta S, Myers J, Lauten WB. Arms2 in/del polymorphism predicts response to intravitreal anti-VEGF therapy for choroidal neovascular age-related macular degeneration. Poster presented at the Annual Meeting of the Association for Research in Vision and Ophthalmology; May 2012; Fort Lauderdale, FL.