Associations of Complement Factor H (CFH) and Age-Related Maculopathy Susceptibility 2 (ARMS2) Genotypes with Subtypes of Polypoidal Choroidal Vasculopathy

Abstract

Purpose.: To clarify whether complement factor H (CFH) and age-related maculopathy susceptibility 2 (ARMS2) genotypes are associated with subtypes of polypoidal choroidal vasculopathy (PCV), such as polypoidal choroidal neovascularization (CNV) and typical PCV.

Methods.: Two hundred eighty-seven patients were categorized as having polypoidal CNV (85 patients) or typical PCV (202 patients) on the basis of indocyanine green angiographic findings. In total, 277 subjects without age-related macular degeneration (i.e., free of PCV and CNV), served as controls. I62V (rs800292) in the CFH gene and A69S (rs10490924) in the ARMS2 gene were genotyped, and case–control studies were performed in subjects with these PCV subtypes.

Results.: The polypoidal CNV group included no subjects homozygous for the A/A genotype of rs800292, whereas 7% of the typical PCV group had this genotype. Case–control studies of polypoidal CNV and typical PCV showed significant differences in all distributions of rs10490924 between these two groups. In contrast, the distributions of rs10490924 did not differ between the typical PCV and control groups. Logistic regression analysis with adjustment for confounding factors showed the distributions of rs10490924 to differ significantly between the controls and polypoidal CNV cases (P = 2.1 × 10⁻¹⁰; OR, 10.87). The T/T genotype was significantly more common in the polypoidal CNV than in the typical PCV group (P = 3.6 × 10⁻¹⁴; OR, 19.61).

Conclusions.: PCV may be genetically divisible into polypoidal CNV and typical PCV. The rs800292 variant of the CFH gene is a potential marker for typical CNV. The rs10490924 variant of the ARMS2 gene was shown to be associated with polypoidal CNV. Typical PCV was not associated with this variant.

Polypoidal choroidal vasculopathy (PCV), characterized by a branching vascular network with polypoidal lesions detected by indocyanine green angiography (IGA), ¹ is included among the forms of exudative age-related macular degeneration (AMD) in Japan. ² Our group has reported on subtypes of PCV and has categorized PCV as polypoidal choroidal neovascularization (CNV) and PCV in the narrow sense (also referred to as typical PCV). ³ In the first type, both feeder and draining vessels are visible on IGA and network vessels are numerous. In the second group, neither feeder nor draining vessels are detectable, and the number of network vessels is small. Okubo et al. ⁴ also reported that PCV can be divided into two types; the small–short and large–long types, but the clinical features in their report differed from those described by our group. They defined the small–short type as having small lesions and a short period of disturbed vision. The large–long type was defined as having large lesions and prolonged vision disturbance. Visual prognosis and clinical course differed between these two types of PCV. 

Two different histopathologic features have also been reported; one is related to choroidal vasculature abnormalities and the other to CNV. Vascular endothelial growth factor (VEGF) was identified in vascular endothelial cells in some studies but not in others. ^{5 –8}  

PCV is reportedly associated with rs800292 in the complement factor H (CFH) gene ^9,10 and the age-related maculopathy susceptibility 2 (ARMS2) gene. ¹¹ There are reports showing I62V (rs800292) in the CFH gene to have a stronger association with PCV than does Y402H in the CFH gene in Japanese populations. ^10,12 Furthermore, A69S (rs10490924) in the ARMS2 gene was reported to be strongly associated with AMD while showing no association with subtypes of PCV Japanese populations. ^{11 –13}  

The present study was conducted to investigate whether there is an association between rs800292 or rs10490924 and subtypes of PCV. To our knowledge, this is the first study to examine associations of the CFH and ARMS2 genes with PCV subtypes. 

Materials and Methods

Subjects

Two hundred eighty-seven patients with PCV were enrolled in the study at Nihon University Surugadai Hospital in Tokyo between 2008 and 2010 (200 men, 87 women; mean age, 69.9 ± 8.9 years). PCV was diagnosed on the basis of the presence of polypoidal lesions on IGA. Classification into two types of PCV (i.e., polypoidal CNV and typical PCV) was based on whether both feeder and draining vessels were seen on IGA. Eighty-five patients were diagnosed as having polypoidal CNV; 202 had typical PCV. Radiation-associated choroidal neovasculopathy was not included in the present study, because it is rare, and environmental rather than genetic factors are the main cause of the disease. Information on hypertension, diabetes mellitus, and smoking was obtained from medical histories collected for each patient. Smokers were defined as current or former smokers, whereas nonsmokers were defined as subjects with no previous or current smoking history. 

In total, 277 subjects without AMD (i.e., free of polypoidal CNV and PCV; 111 men, 166 women; mean age 72.9 ± 8.7 years), served as the control. There were no remarkable findings on fundus examinations of the control. Both groups were enrolled in Japan, and informed consent was obtained from each individual as per the protocol approved by the Human Studies Committee of Nihon University. This investigation was performed according to the guidelines of the Declaration of Helsinki. 

Genotyping

DNA was extracted from peripheral blood leukocytes by the phenol and chloroform extraction method. ^14,15 Genotyping was performed using an SNP genotyping assay (TaqMan; Applied Biosystems Inc. [ABI], Foster City, CA), performed using the Taq amplification method. ^14,15  

Plates were read in the endpoint analysis mode of the software (SDS 7700, ver. 1.6.3 software; ABI). Genotypes were determined visually based on the dye-component fluorescent emission data depicted in the X–Y scatterplot of the system software. Genotypes were also determined automatically by the signal-processing algorithms of the software. ^14,15  

Statistical Analysis

Data are shown as the mean ± SD. Differences between the PCV subtype and control groups were assessed by analysis of variance (ANOVA), followed by Fisher's protected least significant difference (PLSD) test. Hardy-Weinberg equilibrium was assessed by χ² analysis. The overall distribution of alleles was analyzed using 2 × 2 contingency tables. Genotype distributions were compared between the patient groups and the controls using a two-sided Fisher's exact test and multiple logistic regression analysis. Statistical significance was set at P < 0.05. 

Results

The clinical features of PCV patients and the control group are shown in Table 1. There were no significant differences in any parameters between the polypoidal CNV and typical PCV patient groups. 

Table 1.

 

Characteristics of Study Participants

	Case							Control
	Total PCV	P vs. Control	Polypoidal CNV	P vs. Control	P vs. Typical PCV	Typical PCV	P vs. Control
Subjects, n	287		85			202		277
Age, y (mean ± SD)	69.9 ± 8.9	<0.0001*	70.1 ± 8.9	<0.0001*	0.427	70.9 ± 8.9	<0.0001*	72.9 ± 8.7
Male/female	200/87	<0.0001*	63/22	<0.0001*	0.427	137/65	<0.0001*	111/166
Hypertension, %	39	0.342	39	0.499	0.954	39	0.843	43.0
Diabetes mellitus, %	9	<0.0001*	9	0.014*	0.824	8	<0.0001*	19.5
Smoker, %	33	<0.0001*	38	<0.0001*	0.415	31	0.115	17.7

  P-values reflect comparisons between each of the case groups and control group, calculated using Fisher's exact test.

*   P < 0.05.

Distributions of genotypes and alleles of the two variants are shown in Table 2. Both variants in the control group were in Hardy-Weinberg equilibrium (data not shown, P > 0.05). There were significant differences in all genotype models and allele distributions of rs800292 (CFH gene) between the two case groups and between each case group and the controls. Most notably, the polypoidal CNV group had no AA homozygous subjects, whereas 7% of patients in the typical PCV group had this variant. For rs10490924 (ARMS2 gene), there were significant differences in all genotype models and allele distributions between the polypoidal CNV and control groups. However, there were no significant differences in any genotype models or allele distributions of rs10490924 (ARMS2) between the typical PCV and control groups. 

Table 2.

 

Genotype and Allele Distributions in PCV Patients and Control Group

	Total PCV Patients			Polypoidal CNV			Typical PCV			Control
	n	%	P	n	%	P	n	%	P	n	%
rs800292 (CFH)
Genotype
G/G	164	57%	<0.0001*	53	62%	<0.0001*	111	55%	0.0001*	100	36%
G/A	109	38%		32	38%		77	38%		141	51%
A/A	14	5%		0	0%		14	7%		36	13%
Dominant model
G/G	164	57%	<0.0001*	53	62%	<0.0001*	111	55%	<0.0001*	100	36%
GA/AA	123	43%		32	38%		91	45%		177	64%
Recessive model
AA	14	5%	0.0007*	0	0%	0.0005*	14	7%	0.032*	36	13%
GA/GG	273	95%		85	100%		188	93%		241	87%
Allele
G	437	76%	<0.0001*	138	81%	<0.0001*	299	74%	<0.0001*	341	62%
A	137	24%		32	19%		105	26%		213	38%
rs10490924 (ARMS2)
Genotype
G/G	80	28%	<0.0001*	8	9%	<0.0001*	72	36%	0.9611	99	36%
G/T	126	44%		24	28%		102	50%		142	51%
T/T	81	28%		53	62%		28	14%		36	13%
Dominant model
GG	80	28%	0.0448*	8	9%	<0.0001*	72	36%	0.9826	99	36%
GT+TT	207	72%		77	91%		130	64%		178	64%
Recessive model
TT	81	28%	<0.0001*	53	62%	<0.0001*	28	14%	0.7835	36	13%
GT+GG	206	72%		32	38%		174	86%		241	87%
Allele
G	286	50%	<0.0001*	40	24%	<0.0001*	246	61%	0.8802	340	61%
T	288	50%		130	76%		158	39%		214	39%

  P-values for comparison between cases and controls, calculated using Fisher's exact test.

*   P < 0.05.

The results of logistic regression analysis, with adjustment for confounding factors, including age, sex, and risk factors, are shown in Table 3. This analysis was performed for the dominant or recessive genotype models showing significant results, as presented in Table 2. Susceptibility genotypes were those with high frequencies in patient groups in case–control studies. The distributions of rs10490924 differed significantly between the controls and the polypoidal CNV group (P = 2.1 × 10⁻¹⁰; OR, 10.87). The T/T genotype was significantly more frequent in the polypoidal CNV than in the typical PCV group (P = 3.6 × 10⁻¹⁴; OR, 19.61). As no polypoidal CNV patients had the AA genotype in CFH, it was not possible to calculate an OR by using the G/A+A/A model. 

Table 3.

 

Logistic Regression Analysis with Adjustment for Confounding Factors

	Susceptibility Genotype	Total PCV Patients			Polypoidal CNV						Typical PCV
		P vs. Control	OR	95% CI	P vs. Control	OR	95% CI	P vs. Typical PCV	OR	95% CI	P vs. Control	OR	95% CI
rs800292 (CFH)
Dominant model	G/G genotype	8.44 × 10−6 *	2.47	1.66–3.69	8.7 × 10−4 *	2.95	1.56–5.58	0.760	0.91	0.51–1.64	1.0 × 10−4 *	2.36	1.53–3.65
Recessive model	G/A+G/G genotype	2.6 × 10−3 *	3.03	1.46–6.23	0.000	∞	—	0.000	∞	—	0.050	2.08	0.99–4.31
rs10490924 (ARMS2)
Dominant model	G/T+T/T genotype	0.1160	1.41	0.92–2.15	1.7 × 10−4 *	6.68	2.48–17.99	3.3 × 10−5 *	7.19	2.84–18.18	1	1.00	0.64–1.57
Recessive model	T/T genotype	5.2 × 10−4 *	2.47	1.48–4.12	2.1 × 10−10 *	10.87	5.21–22.70	3.6 × 10−14 *	19.61	9.17–43.47	0.752	1.11	0.59–2.06

 Logistic regression analysis performed for each genotype with adjustment for confounding factors (age, sex, hypertension, diabetes mellitus and smoking).

  P-values for genotype were calculated using Fisher's exact test.

*   P < 0.05.

Discussion

In the present study, the rs800292 (CFH gene) was associated with both subtypes of PCV analyzed. We did not examine the Y402H variant in the present study because the minor allele frequency is very low in the Japanese population (0.057%). Furthermore, each subtype was independently associated with CFH. Therefore, we were not able to rule out the possibility of a susceptibility variant for typical PCV. No subjects in the polypoidal CNV group had the A/A genotype, whereas 7% of typical PCV group patients had this variant. The A/A genotype may be associated with a risk of developing typical PCV. The Y402H variant in the CFH gene is reportedly associated with ischemic heart disease related to arteriosclerosis. ¹⁶ Histopathologic findings of typical PCV suggested arteriosclerosis to likely be the underlying pathology. ⁷ These reports suggest that the CFH gene is associated with typical PCV via pathophysiological factors associated with arteriosclerosis. This remains speculative, however. Major limitations of this study are the small sample size and that one of the genotypes (A/A) was not isolated in the CNV group. Further study, with a larger number of subjects, is clearly needed to assess the roles of arteriosclerosis-related pathophysiological factors in PCV. 

It has been well established that the CFH and ARMS2 genes are associated with AMD, as broadly defined by degenerative maculopathy, and PCV in many cohorts. In particular, A69S in the ARMS2 gene is strongly associated with AMD and PCV in the Japanese population. Hayashi et al. ¹² reported the A69S variant to be associated with all three subtypes of AMD. Their results also indicated a strong association with retinal angiomatous proliferation (RAP), as well as associations with typical AMD (tAMD) and, finally, PCV, which raises the possibility that PCV is not a single disease in terms of its genetic background(s). Recently, Sakurada et al. ¹⁷ reported the A69S variant to be associated with subretinal hemorrhage, serous pigment epithelial detachment (PED), and hemorrhagic PED, as well as bilateral lesions and age at PCV onset. Their report also supports a possible correlation of PCV with genetic background. 

On the other hand, considering the characteristic IGA and histopathologic findings of PCV, our group reported that PCV could be classified into three groups: typical PCV, which represents choroidal vasculature abnormalities; polypoidal CNV involving deformation of CNV under the retinal pigment epithelium (RPE); and radiation-associated choroidal neovasculopathy. ³ Our research goals, designed to test our hypothesis, include determining which genetic variants are associated with these forms of PCV. 

In the present study, there was a significant difference between the typical PCV and polypoidal CNV groups in the distribution of rs10490924. Although rs10490924 was strongly associated with polypoidal CNV, there was no association with typical PCV. In many previous reports, genetic variants in the ARMS2 gene were found to be associated with PCV. These reports did not estimate associations with the subtypes of PCV described herein and may actually have reflected an association only with polypoidal CNV. Histopathologic features of PCV previously reported by our group support this assumption. ⁷ One study showed aqueous VEGF levels in eyes with PCV to be higher than those in normal eyes, but significantly lower than levels in eyes with tAMD. ¹⁸ This result may also support the existence of distinct varieties of PCV. Although IGA findings of polypoidal CNV appeared to be consistent with CNV, the histopathologic and IGA features of typical PCV showed choroidal vasculature abnormalities. These observations suggest polypoidal CNV to be genetically and histopathologically close to tAMD, a representative form of CNV. Furthermore, typical PCV showed no association with CNV. 

In conclusion, the present study is the first to examine the associations between variants in the CFH and ARMS2 genes and PCV subtypes. We found the A/A genotype of rs800292 in the CFH gene to be a potential marker for typical PCV, whereas the T/T genotype of rs10490924 in the ARMS2 gene may be a marker for polypoidal CNV. The rs10490924 variant showed no association with typical PCV. These results suggest that polypoidal CNV has a genetic background different from that of typical PCV. Further studies are needed to examine the effects of various treatments on PCV subtypes. 

Footnotes

 Supported by the Research Committee on Chorioretinal Degenerations and Optic Atrophy and by The Ministry of Health and Welfare of Japan (MY).

Footnotes

 Disclosure: K. Tanaka, None; T. Nakayama, None; R. Mori, None; N. Sato, None; A. Kawamura, None; Y. Mizutani, None; M. Yuzawa, None

The authors thank all the patients who participated in the study.