We appreciate the interest of Dr Sakurada and assoicates in our article.

Our study was a screening study for genes significantly associated with age-related macular degeneration (AMD). As Drs Sakurada and Iijima stated, there have been several reports suggesting that pseudoexfoliation syndrome (XFS) could be one of the risk factors for AMD. However, the small sample of our study was a limitation, and we could not exclude the possibility of false positives.

However, AMD is a complex disorder that is genetically associated with many susceptible loci. Two major genes, the complement factor H ( CFH ) gene and ARMS2 / HTRA1 , are significantly associated with a distinct component of the AMD phenotype in 2 different biological pathways. ARMS2 is located in the ellipsoid, a mitochondria-concentrated part of human photoreceptor cells, and HTRA1 is a serine protease expressed in retinal pigment epithelium and drusen. For the ARMS2 gene, the genotype frequency of the previously reported single nucleotide polymorphism (SNP), p.Ala69Ser variant, was statistically higher in the exudative type of AMD and the polypoidal choroidal vasculopathy (PCV) groups than in the control group (exudative AMD vs control, P = 3.1 × 10 ⁻⁸ ; PCV vs control, P = 6.9 × 10 ⁻³ ). The distribution of ARMS2 SNPs in the AMD subjects in our study is in agreement with earlier findings.

It has been suggested that the exudative type of AMD results from a defect in the elastic lamina in the Bruch membrane, which would then stimulate the development of a choroidal neovascularization (CNV). This was supported by the recent report that the elastic lamina of the Bruch membrane was more fragmented and less continuous in a LOXL1 knockout mouse. In addition, the CNVs induced by laser photocoagulation in these mice were grew more aggressively. It was suggested that the elastin gene ( ELN ) may be involved in the phenotypic expression of neovascular AMD and PCV. Nevertheless, the factors that underlie why some subjects develop the more aggressive exudative AMD, while others the dry AMD, are not well understood. Thus, the purpose of our study was to investigate the gene that influences the phenotype of the AMD and not the prevalence of AMD. It is correct that a recent genome-wide association study (GWAS) did not reveal a positive association between exudative AMD and LOXL1 gene locus mapped on 15q24.1, but in general GWAS cannot detect modifier genes.

As for LOXL1 SNP rs1048661, the allele frequency of the T variant was not significantly different between cases with AMD from the controls (major allele frequency T = 0.620 in exudative AMD and T = 0.507 in controls). However, in the T/T homozygote in exudative AMD case, 6 out of 22 cases showed the massive subretinal hemorrhage due to neovascularization. Thus, there is the possibility that the LOXL1 locus could be a modifier locus affecting the integrity of the Bruch membrane and lead to more aggressive CNV growth.

Again, we thank Drs Sakurada and Iijima for their interest and comments.