Pharmacogenomics, an evolving research discipline within ophthalmology, investigates genotype- phenotype correlations in an attempt to explain interpatient variability in response to medications. To date, most ophthalmologic pharmacogenomics research has concerned the treatment of open-angle glaucoma and age-related macular degeneration (AMD).

OPEN-ANGLE GLAUCOMA
Both primary open-angle glaucoma (POAG) and steroid-induced ocular hypertension have been studied using pharmacogenomic approaches. Pharmacologic strategies to reduce intraocular pressure (IOP) remain the leading treatments for POAG and allied disorders. However, an unpredictable rate of nonresponse to medication is a long-standing clinical challenge. For example, in one randomized clinical trial, the rate of nonresponse (defined as reduction of IOP by less than 15% from baseline) was reported as 28% with timolol and 18% with latanoprost.1 The precise mechanisms for this variability remain largely undetermined. Pharmacogenomics has been suggested as an approach to investigate this clinically important phenomenon.2,3

The topical β-blockers comprise multiple nonselective agents (β1- and β2-antagonists), including timolol, and one β1-selective agent, betaxolol. The β1-adrenergic receptor gene, ADRB1, contains two single nucleotide polymorphisms.4 The β2-adrenergic receptor gene, ADRB2, contains four polymorphisms.5 In a prospective, nonrandomized clinical trial, 48 normal volunteers were treated with betaxolol for 6 weeks. The Arg389 homozygote genotype independently correlated with higher baseline IOP and a greater magnitude of response to betaxolol, even after adjusting for baseline IOP.6

In a prospective clinical trial, 89 normal volunteers were treated with timolol. No statistically significant association was found between the clinical effectiveness of timolol and three polymorphisms in the β2-adrenergic receptor gene.7

Latanoprost is a highly selective agonist against the prostaglandin F (FP) receptor.8 In a prospective, nonrandomized clinical trial, 100 normal volunteers were treated with latanoprost for 1 week. The polymorphisms rs3753380 and rs3766355 showed statistically significant associations with the clinical efficacy of latanoprost.9

More recently, the phenomenon of steroid-induced glaucoma has been studied with pharmacogenomic approaches. The etiology of the steroid response has never been fully explained, but a genetic determinant has long been suspected.10 Intravitreal triamcinolone acetonide (IVTA) is used to treat a variety of retinal diseases, including exudative age-related macular degeneration11 and macular edema secondary to diabetes mellitus, 12 retinal vein occlusion,13 and other causes. Clinically significant elevation of IOP has been reported in about 40% of patients.14

The gene MYOC is expressed in the trabecular meshwork, and its expression has been shown to be steroidinduced. Variations in MYOC may occur in about 5% of patients with open-angle glaucoma. However, there is no statistically significant association between MYOC mutations and steroid-induced glaucoma.15 In cadaver eyes, timolol reduced MYOC expression in some cases, but timolol does not affect MYOC induction by dexamethasone.16

Glucocorticoid receptors are present in the trabecular meshwork, suggesting a possible etiology of steroidinduced glaucoma.17 In a pilot study, 52 patients were treated with IVTA for a variety of retinal diseases. There were no statistically significant associations between six common polymorphisms in the glucocorticoid receptor gene and the magnitude of IOP elevation following treatment with IVTA.18

AGE-RELATED MACULAR DEGENERATION
The primary treatment for nonexudative AMD is vitamin supplementation to prevent progression to advanced disease.19 For patients with exudative AMD, the MARINA and ANCHOR trials demonstrated, respectively, the efficacy of ranibizumab in the treatment of minimally classic or occult choroidal neovascularization20 and the superiority of ranibizumab to photodynamic therapy with verteporfin in patients with predominantly classic CNV.21 Bevacizumab is also widely used to treat exudative AMD.22

A polymorphism in the complement factor H gene (CFH) has been associated with an increased risk of AMD.23-26 In a retrospective analysis of a subgroup of patients from the Age-Related Eye Disease Study (AREDS), 876 patients with AREDS categories 3 and 4 were genotyped for polymorphisms in CFH. In patients with the CFH TT genotype, treatment with antioxidants plus zinc was associated with a 68% risk reduction. In patients with the CFH CC genotype, treatment with antioxidants plus zinc was associated with only an 11% risk reduction. These differences in risk reduction were statistically significant.27

In a retrospective cohort study of 86 patients with exudative AMD treated with bevacizumab, patients with the CFH TC and TT genotypes experienced more favorable visual outcomes than did patients with the CFH CC genotype, even after adjusting for age, pretreatment visual acuity, and CNV lesion size.28

In a retrospective study of 156 patients with exudative AMD treated with ranibizumab, there were no differences in visual outcomes related to the CFH genotype, but a statistically significant genetic “dose-response” effect was noted with regard to the number of injections required. Over a 9-month period, patients with the CFH TT genotype required a mean of 3.3 injections, patients with the CFH TC genotype required a mean of 3.8 injections, and patients with the CFH CC genotype required a mean of 3.9 injections.29

In a retrospective study of 69 patients with exudative AMD treated with verteporfin, patients with the CFH TT genotype experienced poorer visual outcomes than did patients with the CFH TC or CC genotypes, even after adjusting for age, pretreatment visual acuity, and CNV lesion type.30 These results appear to contrast with the results from a previous series of 27 patients treated with verteporfin, in which the CFH CC genotype was associated associated with a greater degree of posttreatment visual loss than the TC genotype. However, in this series, the number of patients with the TT genotype (2) was too small for statistical analysis.31

IMPLICATIONS FOR THE PRACTICING RETINAL SPECIALIST
Interpatient variability in response to medications remains an important daily clinical challenge in all specialties of medicine. For example, although the majority of patients with elevated IOP will respond to topical β-blockers or prostaglandin analogues, there is no reliable way to identify the nonresponders in advance. Similarly, although only a minority of patients treated with IVTA will develop a steroid response, again there is no reliable way to identify these patients prior to treatment. Therefore, all treated patients must be followed to determine the response, which at the very least leads to additional clinic visits and exposure to additional medications in at least some patients. Pharmacogenomics may provide some assistance with this problem. For example, knowledge of an individual patient's risk of developing a steroid response might influence the decision to treat a patient with IVTA or one of the longer-acting corticosteroid preparations.

Based on currently available data, there appears to be a complex relationship between CFH variants and response to interventions to treat AMD. The CFH CC polymorphism appears to be associated with relatively poorer outcomes following treatment with antioxidants plus zinc to prevent progression, and with relatively poorer outcomes following treatment with bevacizumab for CNV. However, this polymorphism may be associated with relatively better visual outcomes following treatment with verteporfin. Additional studies are necessary to validate all of these initial findings, but pharmacogenomics appears to hold great promise to improve patient care.

Stephen G. Schwartz, MD, MBA, is an Associate Professor of Clinical Ophthalmology, Bascom Palmer Eye Institute, University of Miami Miller School of Medicine.

M. Elizabeth Fini, PhD, is Vice Dean for Research; Keck School of Medicine of USC; Director, Institute for Genetic Medicine; and Professor of Cell & Neurobiology and Ophthalmology, University of Southern California, Los Angeles.

The authors state that they are co-holders of a patent pending entitled “Molecular targets for modulating intraocular pressure and differentiation of steroid responders versus non-responders.”