July 2024
Volume 65, Issue 8
Open Access
Genetics  |   July 2024
ANO2 Genetic Variants and Anti-VEGF Treatment Response in Neovascular AMD: A Pharmacogenetic Substudy of VIEW 1 and VIEW 2
Author Affiliations & Notes
  • Robyn H. Guymer
    Centre for Eye Research Australia, Royal Victorian Eye and Ear Hospital, East Melbourne, Australia
    The University of Melbourne, Melbourne, Australia
  • Rufino Silva
    Faculty of Medicine, University of Coimbra (FMUC-UC), Coimbra, Portugal
    Unidade Local de Saude de Coimbra (ULS-Coimbra), Coimbra, Portugal
    Association for Innovation and Biomedical Research on Light and Image (AIBILI), Coimbra, Portugal
    Clinical and Academic Centre of Coimbra (CACC), Coimbra, Portugal
  • Mercedeh Ghadessi
    Bayer Pharmaceuticals US LLC, Whippany, NJ, United States
  • Sergio Leal
    Bayer Consumer Care AG, Basel, Switzerland
  • Isabella Gashaw
    Bayer AG, Berlin, Germany
  • Amy Damask
    Regeneron Pharmaceuticals Inc., Tarrytown, NY, United States
  • Charles Paulding
    Regeneron Pharmaceuticals Inc., Tarrytown, NY, United States
  • Kay D. Rittenhouse
    Bayer Consumer Care AG, Basel, Switzerland
  • Correspondence: Mercedeh Ghadessi, Data Science and Analytics, Bayer Pharmaceuticals US LLC, 100 Bayer Blvd., Whippany, NJ 07981, USA; [email protected]
  • Footnotes
     Current affiliation: IG, *Boehringer Ingelheim Pharma GmbH & Co. KG Binger Str. 173, Ingelheim am Rhein 55216, Germany.
  • Footnotes
     Current affiliation: KDR, Retired.
Investigative Ophthalmology & Visual Science July 2024, Vol.65, 17. doi:https://doi.org/10.1167/iovs.65.8.17
  • Views
  • PDF
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Robyn H. Guymer, Rufino Silva, Mercedeh Ghadessi, Sergio Leal, Isabella Gashaw, Amy Damask, Charles Paulding, Kay D. Rittenhouse; ANO2 Genetic Variants and Anti-VEGF Treatment Response in Neovascular AMD: A Pharmacogenetic Substudy of VIEW 1 and VIEW 2. Invest. Ophthalmol. Vis. Sci. 2024;65(8):17. https://doi.org/10.1167/iovs.65.8.17.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

Purpose: This analysis investigated potential associations between gene variants and clinical end points in the VIEW 1 and 2 randomized clinical trials of intravitreal aflibercept and ranibizumab in neovascular age-related macular degeneration (AMD).

Methods: A genome-wide association analysis was conducted in a subgroup of patients from VIEW 1 and 2 consenting to the optional pharmacogenetic analysis.

Results: Data were pooled from 780 samples from patients representative of the overall VIEW 1 and 2 populations. After Bonferroni correction for multiplicity and statistical adjustment for baseline risk factors, no significant associations were found between previously identified prognostic AMD gene variants and treatment response according to key prespecified VIEW 1 and 2 end points. Genome-wide, there were no significant genetic associations in patients experiencing gains of ≥15 Early Treatment of Diabetic Retinopathy Study letters after 1 or 2 years of treatment. A cluster of variants in ANO2 (encoding anoctamin 2, a calcium-activated chloride channel expressed on photoreceptor cells) on chromosome 12 reached the level of significance for loss of ≥5 letters after 1 year of treatment (P < 5 × 10–8), with the ANO2 rs2110166 SNP demonstrating highly significant association (P = 1.99 × 10–8). Carriers of the ANO2 rs2110166 TT genotype showed a robust increase in visual acuity versus baseline compared with a small decrease in those with the TC genotype.

Conclusions: None of the potential prognostic candidate genes were associated with the clinical end points for treated patients. Preliminary analyses suggest an association of ANO2 with retinal function, with a potential impact on vision of approximately one line over at least the first year. Further investigation of the function of ANO2 in retinal pathophysiology is merited.

Age-related macular degeneration (AMD) is a major clinical concern, given its rising incidence owing to increasing life expectancy, population growth, and associated risk factors.1 Of the two types of late-stage AMD, neovascular (wet/exudative) AMD (nAMD) is associated with rapid and severe vision loss.2 However, anti-vascular endothelial growth factor (VEGF) therapy has led to great improvement in prognosis for patients with nAMD, resulting in a considerable decrease in severe permanent vision loss in the vast majority of patients.3 
However, variability in the response to anti-VEGF treatment has been observed and a small proportion of patients experience substantial vision loss (≥15 Early Treatment of Diabetic Retinopathy Study [ETDRS] letters) despite treatment.4 Research efforts have aimed to identify biomarkers that are either prognostic for the development or progression of AMD (prognostic biomarkers) or that predict likely response to nAMD treatment (predictive biomarkers), or both.5 For example, older age is associated with worse treatment outcomes with anti-VEGF therapy in nAMD, as is poorer visual acuity and larger choroidal neovascularization (CNV) lesions at presentation.6 AMD is recognized as a complex genetic disease, with a higher risk of disease in first-degree relatives of affected individuals.7 Although multiple prognostic gene variants have been proposed to be associated with the development, recurrence, or progression of AMD, predictive biomarkers that reliably predict response to treatment in nAMD have not been established. 
Linkage studies have led to the discovery of prognostic genetic variants at the complement factor H (CFH) and age-related maculopathy susceptibility 2/high-temperature requirement A-1 (ARMS2/HTRA1) loci as the strongest genetic contributors to AMD susceptibility, findings that were confirmed in genome-wide association studies (GWAS).2 The International AMD Genomics Consortium has identified 52 common and rare single-nucleotide polymorphisms (SNPs) across 34 loci (including CFH, ARMS2/HTRA1, and VEGFA) that are associated independently with AMD.8 In another GWAS, 10 independent loci were identified with statistical significance for early AMD; 8 of them overlapped with known advanced AMD loci (near ARMS2/HTRA1, CFH, C2, C3, CETP, TNFRSF10A, VEGFA, and APOE).9 
In terms of predictive genetic biomarkers, a recent meta-analysis identified 33 publications assessing genetic associations of anti-VEGF therapy response in AMD10; nine SNPs of four genes (HTRA1, ARMS2, CFH, and OR52B4) were significantly associated with the anti-VEGF therapy response in the meta-analysis, which included only one study assessing the response to intravitreal aflibercept (IVT-AFL). The association of genotype with response to anti-VEGF therapy was further explored by pooling the results of the well-characterized VIEW 1 and VIEW 2 studies, both phase 3, multinational, randomized, controlled trials in patients with nAMD in which anti-VEGF treatment was protocol-determined,11,12 to increase the power of the analysis. The combined number of patients treated with anti-VEGF agents in VIEW 1 and VIEW 2 allowed for the exploration of their genetic background to provide some insight into genes driving response to treatment. The results of VIEW 1 and 2 have been reported previously.11,12 
The purpose of this pharmacogenetic analysis was to identify any associations between genetic variants and clinical end points in patients with nAMD treated with anti-VEGF agents (IVT-AFL or ranibizumab) in the VIEW 1 and 2 randomized controlled trials, where treatment was mandated monthly or every 2 months in year 1, after which it was pro re nata (PRN), but at least every 12 weeks.12 
Methods
Study Design and Population
The study design of and patient population enrolled in VIEW 1 (NCT00509795) and VIEW 2 (NCT00637377) have been published previously (summarized in Supplementary Methods).11,12 In summary, patients (n = 2419) with active, subfoveal, CNV lesions (or juxtafoveal lesions with leakage affecting the fovea) secondary to AMD were assigned randomly to IVT-AFL 0.5 mg monthly (every 4 weeks), 2 mg every 4 weeks, 2 mg every 2 months after 3 initial monthly doses (every 8 weeks), or ranibizumab 0.5 mg every 4 weeks from baseline to week 52. During weeks 52 to 96, patients received their original dosing assignment PRN with defined retreatment criteria and mandatory dosing at least every 12 weeks (capped PRN regimen). The study protocols were approved by the institutional review board or ethics committee for each clinical site, and all patients signed a written consent form before initiation of the study-specific procedures. The VIEW 1 and 2 studies were conducted in compliance with regulations of the Health Insurance Portability and Accountability Act and the tenets of the Declaration of Helsinki.12 
The primary end point of the VIEW 1 and 2 studies was the proportion of patients maintaining vision (losing <15 letters on an ETDRS chart) at week 52. Key secondary outcomes were mean change from baseline in best-corrected visual acuity (BCVA) as measured by ETDRS letter score at week 52, percentage of patients who gained ≥15 ETDRS letters at week 52, and mean change from baseline in CNV area at week 52.11 
Pharmacogenetic Substudy
Patients could decide to participate in an optional exploratory pharmacogenetic study independent of their decision to participate in the main clinical studies, and separate informed consent was obtained. Patients who agreed to enroll in the exploratory pharmacogenetic substudy had one additional blood sample (10 mL) drawn for DNA analysis before the administration of IVT-AFL or ranibizumab at day 1. In this analysis, data were pooled from consenting patients from VIEW 1 and VIEW 2 treated with either IVT-AFL or ranibizumab. 
Candidate Gene Association Analyses
Targeted analysis was performed on previously published prognostic AMD gene variants as predictive biomarkers through potential associations with prespecified clinical end points from VIEW 1 and 2: (1) the proportion of patients with BCVA gain from baseline of ≥15 letters at week 52; (2) the proportion of patients with presence of intraretinal fluid/cysts (IRF) at week 52; and (3) the proportion of patients receiving ≥7 injections in year 2. A total of 33 candidate genes identified from the literature (Supplementary Table S1) were assessed for a potential association: ABCA1; ADAMTS9; APOC1; APOE; ARMS2/HTRA1; B3GALTL; C3; CCL2; CETP; CFB; CFH; COL8A1; COL10A1; DDR1; FILIP1L; FLT1; HSPH1; IER3; IL8; KDR; LIPC; MAGI1; MEX3C/DCC; MIR548A2; OR52B4; RAD51B; SCN1A; SERPINF1; SERPINF2; SLC16AB; TNFRFS10A; TGFBR1; and VEGFA
Genotyping, Quality Control, and Imputation
Genome-wide SNP genotyping was performed on the Illumina HumanOmniExpressExome-8 v1.2 (VIEW 1) and v1.3 (VIEW 2) platforms at Illumina, Inc. Laboratories (Illumina Inc., San Diego, CA, USA). The IIlumina genotyping array clustering and genotype calling were quality controlled using GenomeStudio v2011.1 and procedures described by Illumina13 and Guo et al.14 Imputation (the only factor that could be modified based on the known factors that affect the power of detection in genetic studies, Supplementary Fig. S1) was performed using IMPUTE2 on all autosomal chromosomes using the 1000 Genomes Project phase 3 data.15 VIEW 1 and 2 were processed separately through quality control procedures and missing SNPs on either Illumina platforms were imputed to increase the overlap between the two studies and increase the power of analysis. 
SNP quality control was performed using PLINK (https://zzz.bwh.harvard.edu/plink/) and associated methodology (Supplementary Methods).14,16 Variants with <97% call rates or with a minor allele frequency of <0.01 were excluded, as were those with deviations from the Hardy–Weinberg equilibrium (< 10−5). Sample quality control was also carried out, whereby patients were removed owing to low sample call rate, high relatedness or inbreeding coefficient. Principal component analysis (PCA) and multidimensional scaling were used to identify underlying ancestry to adjust for population structure as a confounding factor and reduce bias. 
Primary and Exploratory End Points
Primary and exploratory end points for the GWAS discovery analysis are listed in Box 1. The numbers of SNPs and samples progressing through the quality control processes varied depending on the specific end point. Adjustments for multiple testing were only made for different SNPs for GWAS and not for different hypotheses, because these analyses were exploratory and hypothesis generating. 
 
Box 1. End Points Evaluated in the GWAS Discovery Analysis
 
  • Primary end points  
    • Proportion of patients with BCVA gain from baseline of ≥15 letters at week 52
    • Proportion of patients with presence of IRF at week 52
    • Proportion of patients receiving ≥7 injections in year 2
  • Exploratory end points
  • BCVA-related end points  
    • Proportion of patients with BCVA gain of ≥5 letters at week 52
    • Proportion of patients with BCVA loss of ≥5 letters at week 52
    • Proportion of patients with BCVA loss of ≥15 letters at week 52
    • Visual acuity score after week 52 as a continuous variable
    • Visual acuity score after week 96 as a continuous variable
    • Proportion of patients with BCVA loss of ≥5 letters from week 12 to week 52 and 96
    • Proportion of patients with BCVA loss of ≥5 letters from peak BCVA to week 52 and 96
    • Early responder defined as patients with a gain in BCVA from baseline to week 4 that is >50% of the gain from baseline to week 12
  • Other end points  
    • Change in CNV area at week 52 as a continuous variable
    • Proportion of patients receiving ≤3 injections in year 2
    • Proportion of patients with atrophy at week 52
    • Proportion of patients with fibrosis at week 52
 

IRF, intraretinal fluid/cysts.

Statistical Analysis
Patients with missing outcomes were excluded from the analysis. For binary outcomes, univariable tests of association were undertaken using Pearson χ2 test or Fisher exact test to test for the significance of each SNP separately. In significant markers identified as candidates for further interrogation, Cochran–Armitage trend tests were used to evaluate the trend or ordering in effect of the genotype categories (A/A, A/a or a/A, and a/a). For continuous outcomes, ANOVA models were run under the assumptions of a normally distributed outcome variable and homogeneity in variance across genotype groups. 
The significance level for univariate analysis displayed in Manhattan plots was set to 5 × 10–8 after applying a Bonferroni correction for multiple comparisons. Owing to the small sample size, candidate regions were identified considering a lower P value boundary of 5 × 10–6
Association analyses were performed with and without P value adjustments for clinical baseline risk factors relevant to the end point of interest (e.g., baseline visual acuity, central retinal thickness, total lesion size, and age group) and PCA components. Only the first four PCA components were used for adjustment because no major population structure or genomic inflation was detected; no batch effects were observed by Illumina, and the original studies did not detect any multicenter effect. 
Results
In total, 1217 and 1240 patients from VIEW 1 and VIEW 2, respectively, received either IVT-AFL or ranibizumab.11 Of these, approximately 30% of patients from VIEW 1 and approximately 34% of patients from VIEW 2 consented to be included in the pharmacogenetic analysis, resulting in 780 samples available for analysis. Demographics and baseline characteristics of these subsets of patients were generally representative of the overall VIEW 1 and VIEW 2 populations. 
Candidate Gene Association Analyses
No associations were found between variants in the 33 tested candidate genes (Supplementary Table S1) and the three clinical end points of the proportion of patients with (1) BCVA gain of ≥15 letters at week 52, (2) presence of IRF at week 52, or (3) receipt of ≥7 injections in year 2 (see Supplementary Figs. S2, S3, and S4 for ARMS2 and HTRA1, CFH, and VEGFA, respectively). 
GWAS Discovery Analysis
After SNP genotyping, quality control and imputation, 735 patients and 6,801,133 SNPs that were common between the two VIEW 1 and VIEW 2 datasets remained for analysis (Supplementary Fig. S5Table). GWAS analysis was carried out to investigate potential associations between gene variants and primary and exploratory clinical end points in the VIEW 1 and 2 studies. The number of patients and SNPs that remained after merging genotype and clinical data varied according to the specific end point under analysis, depending on the number of missing values. 
Table.
 
Demographics and Baseline Characteristics of Patients Included in the GWAS Discovery Analysis
Table.
 
Demographics and Baseline Characteristics of Patients Included in the GWAS Discovery Analysis
No significant associations between any of the approximately 7 million SNPs and the primary end points defined for the GWAS (the proportion of patients with BCVA gain of ≥15 letters at week 52, presence of IRF at week 52, or receipt of ≥7 injections in year 2) were found (data not shown). A number of noncoding SNPs on chromosome 12 were in the candidate region for BCVA gain of ≥15 letters at week 52, but were shown to be nonsignificant after adjustment for baseline factors (visual acuity, central retinal thickness, lesion size, and age). 
Among the exploratory end points (Box 1), after quality control and relevant adjustments for baseline visual acuity, central retinal thickness, lesion size, and age (Supplementary Fig. S7), the only SNPs reaching genome-wide significance (P = 5 × 10−8) were located on the anoctamin 2 (ANO2) gene on chromosome 12 for the end point of losing ≥5 ETDRS letters (Fig. 1). 
Figure 1.
 
Genome-wide Manhattan plot for associations with losing ≥5 ETDRS letters during anti–VEGF therapy in nAMD. The x axis represents the entire complement of chromosomes against which transformed (negative log 10) P values for each SNP in the analysis are plotted on the y axis. Observed SNPs are shown in bright red, and imputed SNPs are shown in dark red.
Figure 1.
 
Genome-wide Manhattan plot for associations with losing ≥5 ETDRS letters during anti–VEGF therapy in nAMD. The x axis represents the entire complement of chromosomes against which transformed (negative log 10) P values for each SNP in the analysis are plotted on the y axis. Observed SNPs are shown in bright red, and imputed SNPs are shown in dark red.
Further regional plot analysis (Fig. 2) identified the ANO2 rs2110166 SNP as highly significantly associated with losing ≥5 ETDRS letters during anti-VEGF therapy in nAMD (P = 1.99 × 10–8). In the pharmacogenetic study population, 50 patients had the rs2110166 TC genotype and 684 patients had the TT genotype; only one patient had the CC genotype. The minor allele frequency was 3%. At week 52, 73 of 684 patients (11%) with the ANO2 rs2110166 TT genotype had a loss of ≥5 letters compared with 21 of 50 patients (42%) with the ANO2 rs2110166 TC genotype. 
Figure 2.
 
Regional (LocusZoom) plot of ANO2 and surrounding regions for associations with losing ≥5 ETDRS letters during anti-VEGF therapy in nAMD. The blue line on the graph depicts recombination rates. Highly significant SNPs fall in the region with high recombination rates. The color spectrum (red to dark blue) denotes pairwise correlation (r2) between rs2110166 and each SNP in the region. cM, centimorgan; Mb, megabase.
Figure 2.
 
Regional (LocusZoom) plot of ANO2 and surrounding regions for associations with losing ≥5 ETDRS letters during anti-VEGF therapy in nAMD. The blue line on the graph depicts recombination rates. Highly significant SNPs fall in the region with high recombination rates. The color spectrum (red to dark blue) denotes pairwise correlation (r2) between rs2110166 and each SNP in the region. cM, centimorgan; Mb, megabase.
At baseline, the median BCVA score was comparable in the TT and TC genotype groups (approximately 58 ETDRS letters) (Fig. 3). A robust increase in BCVA score from baseline was observed at year 1 in those patients with the TT genotype (Fig. 3). The mean ± SD BCVA ETDRS letters score at week 52 increased to 63.7 ± 17.2 in patients with the rs2110166 TT genotype (n = 684) compared with a small decrease to 53.0 ± 20.2 in patients with the TC genotype (n = 50). BCVA score at the end of the second year of treatment was 61.9 ± 18.5 (n = 589) and 50.9 ± 19.4 (n = 45), respectively. Data on the single patient with the CC genotype were not included in Figure 3 for data privacy reasons; however, it was noted that the BCVA score in this patient decreased from baseline at the end of year 1. 
Figure 3.
 
BCVA letter score values at baseline and end of year 1 in patients treated with IVT-AFL or ranibizumab according to rs2110166 TC (red markers) or TT (blue markers) genotype. There was only one patient with the CC genotype. Data show median and 95% confidence intervals represented by the notches. The lower and upper edges of the box indicate the 25th and 75th percentiles, respectively; any data point below the lower and above the upper whiskers are identified as outliers.
Figure 3.
 
BCVA letter score values at baseline and end of year 1 in patients treated with IVT-AFL or ranibizumab according to rs2110166 TC (red markers) or TT (blue markers) genotype. There was only one patient with the CC genotype. Data show median and 95% confidence intervals represented by the notches. The lower and upper edges of the box indicate the 25th and 75th percentiles, respectively; any data point below the lower and above the upper whiskers are identified as outliers.
Although not reaching the significance level of P < 5 × 10−8, some other BCVA-related end points (Supplementary Table S2) showed consistent trends in differences between the ANO2 rs2110166 TT and TC genotypes, with less favorable outcomes with the TC genotype. 
Discussion
To our knowledge, this GWAS is the first in nAMD to identify a new SNP in the ANO2 gene predictive of response to anti-VEGF treatment. Carriers of the rs2110166 TT genotype, but not the TC genotype, had a robust gain in BCVA after 1 year of treatment, and there was only one patient with the CC genotype. In the candidate gene analysis of two randomized controlled trials, there was no association between 33 candidate genes, including ARMS2/HTRA1, CFH, and VEGFA, that previously had been shown to be associated with AMD development or progression and/or predictive of response to anti-VEGF treatment in nAMD. 
Evidence of the influence of genetic polymorphisms in the response to anti-VEGF treatment in nAMD is inconsistent.6 A degree of variability is likely due to the fact that results have emerged from various study types (observational and clinical), with heterogeneous patient populations treated with different anti-VEGF agents, with different dose regimens, and with response to treatment defined using different functional and anatomic end points.6,10,17 The largest body of evidence on genetic variants associated with treatment response is related to ARMS2/HTRA1, CFH, and VEGFA,6,10,1719 but even here results are conflicting. For example, meta-analyses have reported an association between anti-VEGF treatment response in nAMD and the A69S polymorphism (rs10490924) in the ARMS2 gene,20 the Y402H polymorphism (rs1061170, a T-to-C transition at amino acid position 402) in the CFH gene,21,22 and the rs833061 polymorphism in the VEGFA gene.23 However, specific alleles for CFH (rs1061170), ARMS2 (rs10490924), and HTRA1 (rs11200638) did not predict response to anti-VEGF therapy in the Comparison of Age-Related Macular Degeneration Treatments Trials (CATT) randomized controlled trial.24 Also, data in the Inhibition of VEGF in Age-Related Choroidal Neovascularization (IVAN) and CATT randomized controlled trials did not support a pharmacogenetic association between the two VEGFR2 SNPs, rs4576072 and rs6828477, and response to anti-VEGF therapy.25 
In this pharmacogenetic subgroup analysis of patients in VIEW 1 and VIEW 2, there was no association between the three primary end points reflecting visual outcomes, anatomic outcomes, and treatment requirement or burden and ARMS2/HTRA1, CFH, and VEGFA (nor with 30 other candidate genes). Of the 12 additional exploratory end points, the only SNP reaching the GWAS statistical significance threshold after Bonferroni correction for multiplicity was located on the ANO2 gene for the end point of losing ≥5 ETDRS letters. This result was apparent before and after adjustments for baseline characteristics, most importantly, baseline visual acuity, which is known to be associated with response to treatment. A three-line (15-letter) change in BCVA has traditionally been accepted to represent a clinically significant impact in interventional trials assessing the effectiveness of drugs where there is an inherent benefit:risk ratio to be considered.26 However, a loss of ≥5 or >5 letters has often been used to guide treatment decisions in randomized controlled trials,27,28 and a margin of four letters is commonly used to reflect noninferiority in randomized controlled trials comparing active treatments.29,30 
The lack of identification of specific high-risk alleles of genes (including ARMS2/HTRA1, CFH, and VEGFA) that predict response to anti-VEGF therapy may not be surprising, given the well-acknowledged heterogeneity in results from different studies examining this topic. There are a number of aspects of our study that may contribute to this disparity. The patients in our analysis were treated within a very controlled experimental environment in a clinical trial, and had predefined visits. This is in contrast with the majority of the studies—frequently real-world prospective observational studies—where patients are generally older and less likely to adhere to their treatments or attend their clinical appointments regularly. In addition, the patients in the clinical trial received aflibercept, while most studies to date have reported on analyses with bevacizumab or ranibizumab; the different treatments may result in different outcomes, even though they are all anti-VEGF agents. Another factor that may contribute significantly to disparity in results is the great heterogeneity in end points and assessments determining positive or negative response to anti-VEGF treatment in different studies. Indeed, this was observed in our analysis, where different SNPs seemed to be linked more closely to different end points. 
The anoctamin family of calcium-activated chloride channels consists of 10 members with different cellular functions (ANO1‒ANO10).31 ANO2 (TMEM16B) is found in the retinal pigment epithelium (RPE) in rats and primates32 and is also expressed in photoreceptor synaptic terminals in rats33 and mice.34 In addition, ANO2 contributes to calcium-dependent chloride conductance in the RPE in mice.32 This factor may also be of relevance because chloride channels play an essential role in the RPE, providing membrane conductance for chloride, which is important for transepithelial transport and volume regulation.32 In addition, immunohistochemical staining of the human retina shows moderate positivity for ANO2 in the outer plexiform layer of the retina, according to the Human Protein Atlas35 based on data from the Eye Genotype Expression database,36 with mRNA expression enhanced in photoreceptor cells and microglial cells.35 Interestingly, ANO2 has been identified as an autoimmune target in multiple sclerosis, where decreased macular volume and peripapillary retinal nerve fiber layer thinning are observed.3739 
As in other studies, which determined that the majority of the prognostic variants associated with AMD were noncoding,40 this analysis also found many noncoding SNPs associated with the different anti-VEGF treatment response end points. It is accepted that genetic variation in the noncoding regions of the genome can increase the susceptibility to diseases by disrupting various genomic regulatory elements (e.g., promoters, enhancers, silencers, and insulator regions).41 However, future research is required to understand the role and/or mechanism of action of these noncoding variants in impacting nAMD disease pathogenesis and physiology, as well as treatment response. 
The current study was unique in that the two randomized controlled trials being assessed were almost identical in nature, with all patients treated under controlled and rigorously monitored conditions, and selected for participation according to study-relevant inclusion and exclusion criteria. In addition, over the first year, treatment was fixed, without variation based on treatment response (patients received one of four fixed-dose treatment regimens). This strategy resulted in data of a much higher quality with considerably lower bias compared with real-world data from observational studies. To our knowledge, there are no other studies with access to data of similar quality. The decreased variability in various parameters, including different dosing regimens, is especially important in the evolving environment of individualized therapy captured within other GWAS analyses, especially those outside of randomized controlled trials. However, although VIEW 1 and 2 were large randomized controlled trials, the number of patients included in this pharmacogenetic analysis (n = 735) was relatively small compared with more robust GWAS-sized datasets, and it is possible that results with smaller effect size or rare variants may not be detected. Thus, some genetic variants that are in the candidate region or showing trends toward statistical significance and may be of importance can only be verified in studies with greater statistical power. Additionally, given the lack of association between the ANO2 SNP and anatomic end points, such as the presence of intraretinal fluid, we should acknowledge that VIEW 1 and 2 used time-domain rather than spectral-domain optical coherence tomography, which may have limited the detection of subtle but potentially important changes that may have been associated with the ANO2 SNP. Additionally, this analysis did not consider subretinal fluid, which can reflect nAMD disease activity.42 
The results of this analysis have not yet been validated in another comparable dataset; however, this GWAS shows a novel finding supporting further research into the molecular functions of the ANO2 gene in the retina. Our findings show that even patients with the TC genotype, who had less of a response to the anti-VEGF therapy, still had a stabilized disease. Thus, more studies are warranted to understand the impact of this finding in the type and route of treatment for this subset of patients. Routine genetic testing for AMD is not currently recommended in clinical practice; it remains to be determined whether knowledge of ANO2 gene variants will have therapeutic implications or guide treatment patterns in the future. 
Acknowledgments
The authors thank all patients and investigators of the VIEW 1 and VIEW 2 studies for their participation in these trials. The authors would also like to thank David Bauer for support in coordinating genotyping efforts, Anke Weispfenning for assistance with knowledge transfer and regulations, and Jens Hooge for data analyses. Medical writing and editorial support for the preparation of this manuscript (under the guidance of the authors) was provided by Sarah Feeny, BMedSci, and Karen Yee, PhD, of ApotheCom (UK), funded by Bayer Consumer Care AG, Basel, Switzerland, in accordance with Good Publication Practice (GPP) standards (Ann Intern Med 2022;175:1298–1304). 
Supported by Bayer Consumer Care AG, Switzerland. The VIEW studies were sponsored by Regeneron Pharmaceuticals Inc, USA, and Bayer AG, Germany. Medical writing and editorial support for the preparation of this manuscript, under the guidance of the authors, was provided by Sarah Feeny and Karen Yee of ApotheCom, UK, and was funded by Bayer Consumer Care AG, Basel, Switzerland, in accordance with Good Publication Practice (GPP) standards (Ann Intern Med 2022;175:1298–1304). The authors alone are responsible for the content and writing of the paper. 
Disclosure: R.H. Guymer, Apellis (C, F, R), Bayer (C), Novartis (C), Roche Genentech (C); R. Silva, Allergan (C), Alimera (C), Bayer (C, F, R), Novartis (C), Novo Nordisk (C, F), Roche (C, F, R), Thea (C), AbbVie (R); M. Ghadessi, Bayer (E, I); S. Leal, Bayer (E, I); I. Gashaw, Bayer (E); A. Damask, Regeneron (E); C. Paulding, Regeneron (E); K.D. Rittenhouse, Bayer (E) 
References
Wang Y, Zhong Y, Zhang L, et al. Global incidence, progression, and risk factors of age-related macular degeneration and projection of disease statistics in 30 years: a modeling study. Gerontology. 2022; 68: 721–735. [CrossRef] [PubMed]
Swaroop A, Chew EY, Rickman CB, Abecasis GR. Unraveling a multifactorial late-onset disease: from genetic susceptibility to disease mechanisms for age-related macular degeneration. Annu Rev Genomics Hum Genet. 2009; 10: 19–43. [CrossRef] [PubMed]
Schmidt-Erfurth U, Chong V, Loewenstein A, et al. Guidelines for the management of neovascular age-related macular degeneration by the European Society of Retina Specialists (EURETINA). Br J Ophthalmol. 2014; 98: 1144–1167. [CrossRef] [PubMed]
Mitchell P, Holz FG, Hykin P, et al. Efficacy and safety of intravitreal aflibercept using a treat-and-extend regimen for neovascular age-related macular degeneration: the ARIES study. Retina. 2021; 41: 1911–1920. [CrossRef] [PubMed]
FDA-NIH Biomarker Working Group. BEST (Biomarkers, EndpointS, and other Tools) resource. Available at: https://www.ncbi.nlm.nih.gov/books/NBK338449/: US Food and Drug Administration; Bethesda, MD: National Institutes of Health (US); 2016. [PubMed]
Finger RP, Wickremasinghe SS, Baird PN, Guymer RH. Predictors of anti-VEGF treatment response in neovascular age-related macular degeneration. Surv Ophthalmol. 2014; 59: 1–18. [CrossRef] [PubMed]
Seddon JM, Ajani UA, Mitchell BD. Familial aggregation of age-related maculopathy. Am J Ophthalmol. 1997; 123: 199–206. [CrossRef] [PubMed]
Fritsche LG, Igl W, Bailey JN, et al. A large genome-wide association study of age-related macular degeneration highlights contributions of rare and common variants. Nat Genet. 2016; 48: 134–143. [CrossRef] [PubMed]
Winkler TW, Grassmann F, Brandl C, et al. Genome-wide association meta-analysis for early age-related macular degeneration highlights novel loci and insights for advanced disease. BMC Med Genom. 2020; 13: 1–18. [CrossRef]
Wang Z, Zou M, Chen A, et al. Genetic associations of anti-vascular endothelial growth factor therapy response in age-related macular degeneration: a systematic review and meta-analysis. Acta Ophthalmol. 2022; 100: e669–e680. [PubMed]
Heier JS, Brown DM, Chong V, et al. Intravitreal aflibercept (VEGF trap-eye) in wet age-related macular degeneration. Ophthalmology. 2012; 119: 2537–2548. [CrossRef] [PubMed]
Schmidt-Erfurth U, Kaiser PK, Korobelnik JF, et al. Intravitreal aflibercept injection for neovascular age-related macular degeneration: ninety-six-week results of the VIEW studies. Ophthalmology. 2014; 121: 193–201. [CrossRef] [PubMed]
Illumina Inc. Infinium® genotyping data analysis. A guide for analyzing Infinium genotyping data using the GenomeStudio® Genotyping Module. 2014; Available at: https://www.illumina.com/Documents/products/technotes/technote_infinium_genotyping_data_analysis.pdf.
Guo Y, He J, Zhao S, et al. Illumina human exome genotyping array clustering and quality control. Nat Protoc. 2014; 9: 2643–2662. [CrossRef] [PubMed]
IGSR: The International Genome Sample Resource. Available at: http://www.internationalgenome.org.
Purcell S, Neale B, Todd-Brown K, et al. PLINK: a tool set for whole-genome association and population-based linkage analyses. Am J Hum Genet. 2007; 81: 559–575. [CrossRef] [PubMed]
Strunz T, Pollmann M, Gamulescu MA, Tamm S, Weber BHF. Genetic association analysis of anti-VEGF treatment response in neovascular age-related macular degeneration. Int J Mol Sci. 2022; 23: 6094. [CrossRef] [PubMed]
Kozhevnikova OS, Fursova AZ, Derbeneva AS, et al. Association between polymorphisms in CFH, ARMS2, CFI, and C3 genes and response to anti-VEGF treatment in neovascular age-related macular degeneration. Biomedicines. 2022; 10: 1658. [CrossRef] [PubMed]
Sorenson CM, Gurel Z, Song YS, Peterson KD, Blodi BA, Sheibani N. Thrombospondin-1, BIM and CFH polymorphisms and response to anti-VEGF treatment in neovascular age-related macular degeneration patients. PLoS One. 2024; 19: e0297135. [CrossRef] [PubMed]
Zhang J, Liu Z, Hu S, Qi J. Meta-analysis of the pharmacogenetics of ARMS2 A69S polymorphism and the response to advanced age-related macular degeneration. Ophthalmic Res. 2021; 64: 192–204. [CrossRef] [PubMed]
Hong N, Shen Y, Yu CY, Wang SQ, Tong JP. Association of the polymorphism Y402H in the CFH gene with response to anti-VEGF treatment in age-related macular degeneration: a systematic review and meta-analysis. Acta Ophthalmol. 2016; 94: 334–345. [CrossRef] [PubMed]
Chen G, Tzekov R, Li W, Jiang F, Mao S, Tong Y. Pharmacogenetics of complement factor H Y402H polymorphism and treatment of neovascular AMD with anti-VEGF agents: a meta-analysis. Sci Rep. 2015; 5: 14517. [CrossRef] [PubMed]
Wu M, Xiong H, Xu Y, et al. Association between VEGF-A and VEGFR-2 polymorphisms and response to treatment of neovascular AMD with anti-VEGF agents: a meta-analysis. Br J Ophthalmol. 2017; 101: 976–984. [CrossRef] [PubMed]
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: 593–599. [CrossRef] [PubMed]
Hagstrom SA, Ying GS, Maguire MG, et al. VEGFR2 gene polymorphisms and response to anti-vascular endothelial growth factor therapy in age-related macular degeneration. Ophthalmology. 2015; 122: 1563–1568. [CrossRef] [PubMed]
Csaky KG, Richman EA, Ferris FL. Report from the NEI/FDA ophthalmic clinical trial design and endpoints symposium. Invest Ophthalmol Vis Sci. 2008; 49: 479–489. [CrossRef] [PubMed]
Busbee BG, Ho AC, Brown DM, et al. Twelve-month efficacy and safety of 0.5 mg or 2.0 mg ranibizumab in patients with subfoveal neovascular age-related macular degeneration. Ophthalmology. 2013; 120: 1046–1056. [CrossRef] [PubMed]
Guymer RH, Markey CM, McAllister IL, Gillies MC, Hunyor AP, Arnold JJ. Tolerating subretinal fluid in neovascular age-related macular degeneration treated with ranibizumab using a treat-and-extend regimen: FLUID study 24-month results. Ophthalmology. 2019; 126: 723–734. [CrossRef] [PubMed]
Dugel PU, Koh A, Ogura Y, et al. HAWK and HARRIER: phase 3, multicenter, randomized, double-masked trials of brolucizumab for neovascular age-related macular degeneration. Ophthalmology. 2020; 127: 72–84. [CrossRef] [PubMed]
Heier JS, Khanani AM, Quezada Ruiz C, et al. Efficacy, durability, and safety of intravitreal faricimab up to every 16 weeks for neovascular age-related macular degeneration (TENAYA and LUCERNE): two randomised, double-masked, phase 3, non-inferiority trials. Lancet. 2022; 399: 729–740. [CrossRef] [PubMed]
Schreiber R, Kunzelmann K. Expression of anoctamins in retinal pigment epithelium (RPE). Pflugers Arch. 2016; 468: 1921–1929. [CrossRef] [PubMed]
Keckeis S, Reichhart N, Roubeix C, Strauss O. Anoctamin2 (TMEM16B) forms the Ca(2+)-activated Cl(-) channel in the retinal pigment epithelium. Exp Eye Res. 2017; 154: 139–150. [CrossRef] [PubMed]
Dauner K, Mobus C, Frings S, Mohrlen F. Targeted expression of anoctamin calcium-activated chloride channels in rod photoreceptor terminals of the rodent retina. Invest Ophthalmol Vis Sci. 2013; 54: 3126–3136. [CrossRef] [PubMed]
Stohr H, Heisig JB, Benz PM, et al. TMEM16B, a novel protein with calcium-dependent chloride channel activity, associates with a presynaptic protein complex in photoreceptor terminals. J Neurosci. 2009; 29: 6809–6818. [CrossRef] [PubMed]
The Human Protein Atlas: ANO2. Available at: https://www.proteinatlas.org/ENSG00000047617-ANO2/tissue.
Ratnapriya R, Sosina OA, Starostik MR, et al. Retinal transcriptome and eQTL analyses identify genes associated with age-related macular degeneration. Nat Genet. 2019; 51: 606–610. [CrossRef] [PubMed]
Wicki CA, Hanson JVM, Schippling S. Optical coherence tomography as a means to characterize visual pathway involvement in multiple sclerosis. Curr Opin Neurol. 2018; 31: 662–668. [CrossRef] [PubMed]
Ayoglu B, Mitsios N, Kockum I, et al. Anoctamin 2 identified as an autoimmune target in multiple sclerosis. Proc Natl Acad Sci USA. 2016; 113: 2188–2193. [CrossRef] [PubMed]
Burkholder BM, Osborne B, Loguidice MJ, et al. Macular volume determined by optical coherence tomography as a measure of neuronal loss in multiple sclerosis. Arch Neurol. 2009; 66: 1366–1372. [CrossRef] [PubMed]
Singh N, Swaroop A, Ratnapriya R. Making biological sense of genetic studies of age-related macular degeneration. Adv Exp Med Biol. 2021; 1256: 201–219. [CrossRef] [PubMed]
Ignatieva EV, Matrosova EA. Disease-associated genetic variants in the regulatory regions of human genes: mechanisms of action on transcription and genomic resources for dissecting these mechanisms. Vavilovskii Zhurnal Genet Selektsii. 2021; 25: 18–29. [PubMed]
Schmidt-Erfurth U, Waldstein SM. A paradigm shift in imaging biomarkers in neovascular age-related macular degeneration. Prog Retin Eye Res. 2016; 50: 1–24. [CrossRef] [PubMed]
Figure 1.
 
Genome-wide Manhattan plot for associations with losing ≥5 ETDRS letters during anti–VEGF therapy in nAMD. The x axis represents the entire complement of chromosomes against which transformed (negative log 10) P values for each SNP in the analysis are plotted on the y axis. Observed SNPs are shown in bright red, and imputed SNPs are shown in dark red.
Figure 1.
 
Genome-wide Manhattan plot for associations with losing ≥5 ETDRS letters during anti–VEGF therapy in nAMD. The x axis represents the entire complement of chromosomes against which transformed (negative log 10) P values for each SNP in the analysis are plotted on the y axis. Observed SNPs are shown in bright red, and imputed SNPs are shown in dark red.
Figure 2.
 
Regional (LocusZoom) plot of ANO2 and surrounding regions for associations with losing ≥5 ETDRS letters during anti-VEGF therapy in nAMD. The blue line on the graph depicts recombination rates. Highly significant SNPs fall in the region with high recombination rates. The color spectrum (red to dark blue) denotes pairwise correlation (r2) between rs2110166 and each SNP in the region. cM, centimorgan; Mb, megabase.
Figure 2.
 
Regional (LocusZoom) plot of ANO2 and surrounding regions for associations with losing ≥5 ETDRS letters during anti-VEGF therapy in nAMD. The blue line on the graph depicts recombination rates. Highly significant SNPs fall in the region with high recombination rates. The color spectrum (red to dark blue) denotes pairwise correlation (r2) between rs2110166 and each SNP in the region. cM, centimorgan; Mb, megabase.
Figure 3.
 
BCVA letter score values at baseline and end of year 1 in patients treated with IVT-AFL or ranibizumab according to rs2110166 TC (red markers) or TT (blue markers) genotype. There was only one patient with the CC genotype. Data show median and 95% confidence intervals represented by the notches. The lower and upper edges of the box indicate the 25th and 75th percentiles, respectively; any data point below the lower and above the upper whiskers are identified as outliers.
Figure 3.
 
BCVA letter score values at baseline and end of year 1 in patients treated with IVT-AFL or ranibizumab according to rs2110166 TC (red markers) or TT (blue markers) genotype. There was only one patient with the CC genotype. Data show median and 95% confidence intervals represented by the notches. The lower and upper edges of the box indicate the 25th and 75th percentiles, respectively; any data point below the lower and above the upper whiskers are identified as outliers.
Table.
 
Demographics and Baseline Characteristics of Patients Included in the GWAS Discovery Analysis
Table.
 
Demographics and Baseline Characteristics of Patients Included in the GWAS Discovery Analysis
×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×