Aims To investigate an association between genotype for three single nucleotide polymorphisms strongly associated with the development of age-related macular degeneration (AMD) and the early response to treatment with intravitreal ranibizumab for neovascular AMD.
Methods Best corrected visual acuity letter score was recorded at baseline and each subsequent visit. Age, sex, smoking history, lesion type and the number of injections were also recorded. Genotypes were obtained for rs11200638 in HTRA1, rs1061170 in CFH and rs1413711 in VEGF. Data were analysed with treatment response at month 6 as both a binary (>5 letter improvement vs ≤5 letter gain) and a linear trait.
Results This initial study cohort consisted of 104 Caucasian neovascular AMD patients treated with intravitreal ranibizumab. Trends towards a more favourable outcome were seen with the higher AMD risk genotypes in CFH and VEGF in both the linear and binary models and in HTRA1 in the linear model alone. For CFH, mean letter score change after 6 months was +1.6, +5.9 and +7.2 letters for the TT, TC and CC genotypes and a >5 letter gain was seen in 34.6%, 56.6% and 56%, respectively. For VEGF, mean letter score change after 6 months was +1.3, +5.8 and +7.4 letters for the TT, TC and CC genotypes and a >5 letter gain was seen in 40%, 55.8% and 51.9%, respectively. For HTRA1, mean letter score change was +2.2, +7.5 and +2.9 letters for the GG, GA and AA genotypes.
Conclusions This study reports preliminary evidence suggesting that the higher AMD risk genotypes in CFH, VEGF and HTRA1 may influence the short-term response to treatment with ranibizumab for neovascular AMD.
- age-related macular degeneration
- treatment medical
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
Age-related macular degeneration (AMD) is the most common cause of visual impairment in the Western world, accounting for over 50% of certifiable sight impairment in the UK alone.1 Epidemiological and association studies have shown that both genetic and environmental factors play a major role in the development of AMD.2 Association studies have implicated variants in a number of genes, with the largest effects observed with alleles of the complement factor H (CFH), HTRA1 and Apolipoprotein E genes.3–6 Association with AMD has also been reported for polymorphisms in the VEGF gene.7 8
Given the greater understanding of the genetic changes underlying AMD, we hypothesised that intravitreal ranibizumab (Lucentis, Novartis Pharmaceuticals, Camberley, UK) therapy may be more effective in patients with certain combinations of AMD susceptibility alleles than with others. In this study, we therefore compared genotypes for CFH, HTRA1 and VEGF, and other potential variables, with visual acuity response after 6 months of intravitreal ranibizumab therapy for neovascular AMD.
This study was approved by the Leeds (East) Research Ethics Committee and registered as a clinical trial with the Medicines and Healthcare Regulatory Authority. Participants were recruited from the intravitreal injection clinics at St James's University Hospital, Leeds. Inclusion and exclusion criteria are detailed in box 1. Written informed consent was taken prior to enrolment.
Main inclusion and exclusion criteria
Age 65 years or more
Choroidal neovascularisation (CNV) secondary to age-related macular degeneration and involving the foveal centre
CNV occupying at least 50% of the total lesion area
Greatest linear diameter ≤5400 μm
No subfoveal atrophy or fibrosis
Recent disease progression
Minimum of 6 months follow-up after first intravitreal ranibizumab injection
Other ocular disease that may influence the treatment outcome
CNV secondary to pathological myopia, inflammatory disease, angioid streaks or trauma
Peripapillary CNV extending to the foveal centre or polypoidal choroidal vasculopathy
Cataract surgery in the study eye any time from 3 months before until 6 months after the initiation of intravitreal ranibizumab therapy
Failure to complete 6 months of intravitreal ranibizumab therapy
Baseline stereoscopic colour fundus, fluorescein and indocyanine green (when available) angiogram images were reviewed and graded by a single observer (MM) to identify lesion characteristics.9 The baseline choroidal neovascularisation phenotype was classified as being classic only (100% classic), predominantly classic (50–99%), minimally classic (1–49%) and occult only or retinal angiomatous proliferation (RAP). (Given the uncertainty surrounding response to treatment, data from eyes with RAP lesions were also analysed as a subgroup and compared with eyes with other baseline lesion characteristics.) Information regarding sex, smoking history and the number of injections before the month 6 visit was also recorded.
At baseline and subsequent visits, best corrected visual acuity letter score was measured using an Early Treatment Diabetic Retinopathy Study (ETDRS) chart and recorded for both eyes. Optical coherence tomography (OCT) examination was performed using either Stratus OCT (Carl Zeiss Meditec, Dublin, California, USA) or Spectralis OCT (Heidelberg Engineering, Heidelberg, Germany). Intravitreal ranibizumab therapy was administered in accordance with the European Medicines Agency marketing authorisation and standard UK practice, with three consecutive monthly injections in a loading phase and further injections as required in a maintenance phase. Retreatment in the maintenance phase was prompted by standard criteria, namely a decrease in best corrected visual acuity of ≥5 letters and/or persistent or recurrent subretinal fluid, intraretinal cysts or thickening on OCT or new subretinal haemorrhage on clinical examination.
DNA was extracted from peripheral blood lymphocytes and genotyped for single nucleotide polymorphisms (SNPs) in the CFH (Y402H, rs1061170), VEGF (+674, rs1413711) and HTRA1 (−512, rs11200638) genes. Amplimers of 220, 246 and 348 bp in length were generated by the PCR using the PCRx Enhancer System (Invitrogen, Paisley, UK). Briefly, a 20 μl reaction volume contained 100 μM dNTPs, 1.5 mM MgSO4, 1 × PCRx Amplification buffer, 0.15 units Taq DNA polymerase, 100 ng genomic DNA and 20 pmol of each of the F and R primers. The HTRA1 PCR also used 2× Enhancer solution. For the CFH polymorphism, the primers used were CFH-F (dtgg tcc tta gga aaa tgt ta) and CFH-R (dgaa cat gct agg att tca gag tag tg); for the VEGF polymorphism, the primers consisted of VEGF-F (dtgg cta cag gcc tcc aag ta) and VEGF-R (dtca ccc att ccc atg aca c) and, for the HTRA1 polymorphism, the primers were HTRA1-F (dtag gct ctc tgc gaa tac gg) and HTRA1-R (dggg gaa agt tcc tgc aaa tc). The reactions were performed in a DNA thermal cycler for 2 min at 95°C followed by 40 cycles of 94°C denaturation for 30 s, an annealing temperature of 58°C for the CFH primers, 56°C for the VEGF primers and 56.5°C for the HTRA1 primers for 45 s and an extension step of 72°C for 45 s with a final extension of 72°C for 5 min. The PCR products were digested with ExoSAP-IT (GE Healthcare, Chalfont St Giles, UK), sequenced using the BigDye Terminator V.3.1 Cycle Sequencing Kit and resolved on an ABI3130xl Genetic Analyzer (Applied Biosystems, Warrington, UK). The sequencing primers were CFH-R, VEGF-F and HTRA1-F, respectively. Chromatograms were viewed using Sequencing Analysis 5.2 software (Applied Biosystems) to identify SNPs.
To test the statistical significance of potential predictor variables, multiple regression analyses were conducted, using both a linear model with actual letter score change and a binary logistic model with a gain of >5 ETDRS letters. The following potential predictor variables were included in both models: sex (male or female), smoking status (current smoker, ex-smoker and never smoked), number of injections received (<4 or ≥4 in the first 6 months), lesion type (RAP or other angiographic classification), baseline acuity (<54 letters or ≥54 letters), VEGF genotype (TT, TC or CC), HTRA1 genotype (AA, GA or GG) and CFH genotype (TT, TC or CC). All p values reported refer to significance tests of the potential predictor variables derived from these multiple regression models. The predetermined level of statistical significance was p<0.01.
The baseline demographics, lesion characteristics and ETDRS letter score for the initial cohort of 104 Caucasian participants without prior treatment are described in table 1. Mean baseline acuity was 51.5 ETDRS letters (SD =15.7), and subjects received an average of 4.0 injections (SD =0.9) before the month 6 visit. The mean change in acuity at the month 6 visit was a gain of +5.2 letters (SD =13.6). Visual loss of ≥15 letters was avoided in 98 eyes (94.2%). A gain of at least 1 letter was seen in 68 eyes (65.4%) and gains of ≥5 letters and ≥15 letters were seen in 57 (54.8%) and 27 eyes (26.0%), respectively.
The association of sex, smoking status, lesion type, number of injections and baseline visual acuity with ETDRS letter score change at month 6 is shown in table 2. Of these variables, only baseline visual acuity had a significant association (p<0.01) with visual acuity change at month 6 when included in the multiple regression analyses. Eyes with baseline acuity <54 letters had a mean change of +9.9 letters while those with 54 letters or more at baseline had a change of +0.4 letters. A gain of >5 ETDRS letters was seen in 65% of the eyes with baseline acuity <54 letters and only 37% of those with 54 or more letters at baseline.
The CFH Y402H, VEGF and HTRA1 genotype frequencies, baseline ETDRS letter score, letter score change after 6 months and number of injections within the first 6 months are shown in table 3. In the multivariate analysis, non-significant trends, suggesting a more favourable visual acuity outcome with possession of one or more high-risk alleles, were seen for CFH, VEGF and HTRA1. For the CFH genotype, mean letter score change after 6 months was +1.6 letters for the lower risk TT genotype and +5.9 and +7.2 for the higher risk TC and CC genotype (p=0.18, 0.71 and 0.17 for TC, CC and combined TC or CC genotypes compared with TT). A gain of >5 ETDRS letters was seen in 34.6%, 56.6% and 56% of eyes with the TT, TC and CC genotypes (p=0.04, 0.4 and 0.06 for TC, CC or combined TC or CC genotypes compared with TT).
For VEGF, mean ETDRS letter score change at month 6 was a gain of +1.3, +5.8 and +7.4 letters for the lower risk TT and higher risk TC and CC genotypes, respectively (p=0.11, 0.11 and 0.09 for the TC, CC and combined TC or CC genotypes compared with TT). A gain of >5 letters was seen in 40%, 55.8% and 51.9% of eyes with the TT, TC and CC genotypes (p=0.08, 0.38 and 0.13 for TC, CC and combined TC or CC genotypes compared with TT).
For HTRA1, mean ETDRS letter score change at month 6 was a gain of +2.2, +7.5 and +2.9 letters for the lower risk GG and higher risk GA and AA genotypes, respectively (p=0.03, 0.55 and 0.07 for GA, AA and combined GA or AA genotypes compared with GG). No association was seen in the binary visual acuity change model.
In the ANCHOR and MARINA studies, the response to treatment seemed to be consistent despite baseline variation in lesion size, visual acuity and lesion characteristics.10–12 Although current evidence still suggests that regular monthly treatment is associated with the optimum visual acuity response, a number of other studies have suggested that a proportion of patients may not need such regular dosage after the initial loading phase to maintain the early gain in visual acuity.13 14 In the PIER study, 40% of patients maintained their initial gain with quarterly treatment after the initial loading phase although 60% lost their initial gain.13 We hypothesised that the variable acuity responses seen in these trials may be related to genotype. The results of our pilot study suggest that the visual acuity change after 6 months of ranibizumab therapy may indeed be influenced by polymorphisms in three genes that are known to increase the risk of developing neovascular AMD.
Prior studies have investigated a possible association between the CFH genotype and response to verteporfin photodynamic therapy (PDT) and both intravitreal bevacizumab and ranibizumab therapies. For PDT, the results are conflicting. Goverdhan reported greater loss of visual acuity after PDT in patients with the CC and CT CFH genotypes and Feng reported a trend towards an association between a negative PDT response and the CC genotype.15 16 However, Brantley identified a worse outcome with the TT CFH genotype and Seitsonen found a lower proportion of PDT responders in the TT genotype group.17 18 With intravitreal bevacizumab, Brantley and Nischler reported a trend towards reduced visual acuity and a lower percentage of subjects with improved acuity for the CC genotype.19 20 Lee identified that patients with the CC genotype required one more ranibizumab injection over the first 9 months, but the CFH genotype was not significantly associated with post-treatment visual acuity at either 6 or 9 months, after adjusting for pretreatment acuity.21 In our study, we did not find an association between CFH genotype and the number of injections in the first 6 months, but we did find a trend towards a better visual acuity response with the higher risk TC or CC genotypes.
Although the VEGF gene does not appear to be a major genetic contributor to the development of neovascular AMD, the +674 C/T polymorphism has been most strongly associated with neovascular AMD in a Caucasian population.7 8 McKay has also suggested that VEGF polymorphisms may influence the development of certain anatomic types of neovascular AMD.22 Given the role of VEGF in the angiogenesis and hyperpermeability that characterise neovascular AMD, this polymorphism was considered to be another potential factor that might influence the response to blockage of the biological effect of VEGF within the eye. In the linear acuity change model alone, there was a trend suggesting a better outcome with the higher risk TC or CC genotypes. To our knowledge, no other group has investigated the influence of this polymorphism on the response to ranibizumab therapy for AMD. However, Immonen has recently identified a potential link between other intronic and promoter site VEGF polymorphisms and the anatomic response to PDT.23 These SNPs cover the entire VEGF gene and will therefore detect the segregation of common haplotypes containing the +674 polymorphism.
Polymorphisms in the promoter region of the HTRA1 (high-temperature requirement factor A-1) gene have been shown to increase susceptibility to AMD, especially the neovascular form, in Caucasian and other populations.3 4 Possession of the high-risk A allele is associated with increased levels of the HTRA1 protein in drusen, retinal pigment epithelium and choroidal neovascular membranes of eyes with AMD.3 4 Chowers found no association between the HTRA1 promoter genotype and the visual acuity outcome or number of injections after PDT.24 In our series, a trend towards better visual acuity outcome after 6 months of intravitreal ranibizumab therapy was seen for the increasing AMD risk HTRA1 genotypes, using the linear model alone. Although the total number of injections was similar for all three HTRA1 genotypes, there was a large gain in visual acuity in the individuals who were heterozygous for the high AMD risk A allele and a more modest gain in homozygotes. The apparent advantage of heterozygotes over homozygotes for either HTRA1 genotype is unexpected. However, heterozygote advantage is a documented phenomenon in many biological systems and is not unprecedented in autoimmune disease, which some consider to be the basis of AMD.25 A similar, but non-significant, trend was also seen in the binary acuity outcome model. In this study, the allele frequency of the A allele was 0.48, higher than in other published Caucasian AMD series.3 24 26 The higher frequency of this allele in our study population of AMD patients from Yorkshire may have provided the opportunity to identify this effect.
In this pilot study, we have identified preliminary evidence of associations between visual acuity outcome after 6 months of intravitreal ranibizumab and polymorphisms in the CFH, VEGF and HTRA1 genes. For each gene, a trend towards a better outcome was seen with the polymorphisms considered to increase the risk of developing neovascular or advanced AMD. Although the mechanism by which these polymorphisms contribute to AMD susceptibility is not known, these findings may imply that neovascular AMD is not one disease but rather several different diseases with a common end point. There may be a common form of AMD that results from the action of one or more high-risk alleles at HTRA1, CFH and VEGF. This form of AMD may be characterised by high levels of intraocular VEGF and may respond favourably to ranibizumab therapy. Another form of AMD, secondary to other susceptibility genes, may not be characterised by high intraocular VEGF and may respond differently to treatment.27
Our original hypothesis was that genotype may be associated with either visual acuity change or the number of additional injections required. The 6-month time point provided the opportunity to investigate both these effects.13 28 The initial response to treatment for the entire group is in keeping with data from multicentre clinical trials, suggesting that the association with genotype is real and not the result of selection bias. In addition, the allele frequencies for the CFH and VEGF genes are similar to other published series and there are no minor alleles with a frequency below 10%. Despite these strengths, there are potential limitations to the findings of our study. As the treatment was given with a loading phase of three consecutive injections, followed by a maintenance phase with repeated injection only as required, the results may not be applicable to patients treated with regular monthly injections. The modest number of subjects in this initial analysis means that we are liable to detect only large associations between genotype and visual outcome and that none of the potential associations met the predetermined level of significance. The study population indicates that the results are only applicable to other Caucasian populations and that the high risk HTRA1 A allele was particularly common in the local population. The association with visual acuity was not supported by an increased number of injections.
In this study, we have identified trends suggesting a more favourable visual acuity outcome after 6 months of intravitreal ranibizumab therapy with the higher AMD risk CFH, VEGF and HTRA1 genotypes. These findings may imply that the mechanisms underlying the more common form(s) of AMD, due largely to the action of risk alleles at the loci tested in this study, are those best targeted by anti-VEGF therapy. Neovascular AMD in individuals without these high-risk alleles may in part be due to mechanisms that respond less favourably to anti-VEGF therapy.
A poster including some of these data has been presented at the ARVO meeting, May 2010.
Funding This project was supported by a grant from Yorkshire Eye Research and by an investigator-initiated research grant from Novartis Pharmaceuticals UK. The funding organisations had no role in either the design or the conduct of the research.
Competing interests None.
Ethics approval This study was conducted with the approval of the Leeds East Research Ethics Committee.
Provenance and peer review Not commissioned; externally peer reviewed.