Clinical and genetic characteristics of pachydrusen in patients with exudative age-related macular degeneration

We investigated the clinical and genetic characteristics of patients with unilateral exudative age-related macular degeneration (AMD), including typical AMD, polypoidal choroidal vasculopathy, and retinal angiomatous proliferation, in whom pachydrusen was seen. Patients with unilateral exudative AMD with at least a 12-month follow-up period were included. According to the fellow eye condition, 327 consecutive patients were classified into 4 groups: Group 0: no drusen (42.8%), Group 1: pachydrusen (12.2%), Group 2: soft drusen (30.3%), Group 3: pseudodrusen with or without soft drusen (14.7%). Development of exudative AMD in the fellow eye was retrospectively studied for a 60-month period and this inter-group comparisons were performed. Genotyping was performed for ARMS2 A69S and CFH I62V. The thickness of the choroid in the fellow eyes increased significantly in Group 1 than in other groups (all P < 1.0 × 10−7). The development of exudative AMD in the fellow eye was significantly less frequent in Group 1 than in Groups 2 or 3 (P = 0.022 and 0.0015, respectively). Risk allele frequency of ARMS2 A69S was significantly lower in Group 1 than in Group 2 and 3 (all P < 1.0 × 10−4). Patients with pachydrusen have genetic and clinical characteristics distinct from those of soft drusen and pseudodrusen.


Results
presents the genetic and clinical characteristics of patients with unilateral exudative AMD classified into 4 groups according to the drusen types in the fellow eye. The concordance rate between the graders was 96.4% (318/330). A total of 94 patients showed scattered drusen in their fellow eye and late phase ICGA images of 94 eyes were evaluated. Patients in Group 1 (pachydrusen group) were significantly younger than those in Group 2 (soft drusen group) and Group 3 (pseudodrusen group) (p = 3.4 × 10 −4 and 2.4 × 10 −8 , respectively). T allele frequency of A69S of ARMS2 gene was significantly lower in Group 1 than in Group 2 and Group 3 (p = 2.4 × 10 −8 and 2.4 × 10 −5 , respectively). Prevalence of polypoidal choroidal vasculopathy (PCV) was significantly higher in Group 1 (85%) than in the other groups.
The follow-up periods to involvement of the fellow eye were 43.8 ± 16.3, 47.1 ± 15.5, 42.8 ± 15.7, and 38.1 ± 16.6 months in Groups 0, 1, 2, and 3, respectively, which were not significantly different from one another (p = 0.061, analysis of variance). The number of patients having developed exudative AMD in the fellow eye was 5, 0, 11 and 9 in Groups 0, 1, 2, and 3, respectively. Of 5 fellow eyes developing exudative AMD in Group 0, one eye and four eyes were PCV and typical AMD, respectively. Of 11 fellow eyes developing exudative AMD in Group 2, three eyes, three eyes and five eyes were PCV, typical AMD, and RAP, respectively. Of 9 fellow eyes developing exudative AMD in Group 3, one eye, four eyes, and four eyes were PCV, typical AMD, and RAP, respectively. Kaplan Meier estimator demonstrated a significantly lower risk of developing exudative AMD in the fellow eye for patients in Group 1 compared with those in Groups 2 and 3 (p = 0.022 and 0.0015, respectively, log-rank test). There were no significant differences of the risk between Group 0 and Group 1 (p = 0.21) (Fig. 2).

Discussion
In the present study, we investigated clinical and genetic characteristics of pachydrusen in comparison with soft drusen and pseudodrusen using patients with unilateral exudative AMD.
Soft drusen are mainly seen in the central macula, but are sometimes scattered around the macula; therefore, it is difficult to differentiate pachydrusen from soft drusen in some cases. Soft drusen contain lipid and many inflammatory materials 21 . It has been reported that hypofluorescent spots on late phase indocyanine green angiography (ICGA) indicate neutral lipid accumulation in Bruch's membrane and hypofluorescent spots represent "lipid wall" between the basal lamina of RPE and inner collagenous layer of BrM 22 . On the other hand, Tsujikawa et al. 23 reported that punctate hyperfluorescent spots were seen in 38 of 41 eyes (93%) with active central serous  www.nature.com/scientificreports www.nature.com/scientificreports/ chorioretinopathy (CSC) on late phase ICGA. Although they did not demonstrate punctate hyperfluorescent spots on color fundus images with late phase ICGA, it is considered that they may be pachydrusen. We used the late phase ICGA images to differentiate pachydrusen from scattered soft drusen in the present study.
In the present study, Group 1 patients with pachydrusen were significantly younger than those with soft drusen (Group 2) and pseudodrusen (Group 3). Pachydrusen might be early-onset and male-dominant drusenoid deposits while pseudodrusen have been reported to be late-onset and female-dominant 26 .
It has been reported that ARMS2 A69S and CFH I62V are two major genetic variants associated with AMD in the Japanese population and that the risk variant of ARMS2 A69S is associated with fellow eye involvement in patients with unilateral exudative AMD 27,28 . In the present study, risk (T) allele frequency of ARMS2 A69S was significantly lower in Group 1 (pachydrusen group) than in Group 2 (soft drusen group) or Group 3 (pseudodrusen group), and the follow-up study revealed that the risk of developing exudative AMD in the fellow eye was significantly lower in Group 1 than in Group 2 or Group 3. Interestingly, risk allele frequency of ARMS2 A69S was also lower in Group 1 than Group 0 (no drusen group) although it did not reach statistical significance. The results suggest that pachydrusen might not pose a risk of developing exudative AMD but conversely, might have a protective role against developing exudative AMD.
Subfoveal choroidal thickness in the fellow eye was the greatest in Group 1 patients among all 4 groups. It has been reported that choroidal thickness is associated with age, axial length and genetic factors including ARMS2 and CFH variants 29,30 . The patients in Group 1 were the youngest and had the least risk allele frequency of ARMS2 A69S and CFH I62V, which could explain why subfoveal choroidal thickness was the greatest in Group 1 in this study. The axial length was not measured in any of the subjects in the present study.  www.nature.com/scientificreports www.nature.com/scientificreports/ The frequency of PCV in Group 1 (pachydrusen) was 85%, which was significantly higher than the 47.5% in Group 2 (soft drusen) and 14.3% in Group 3 (pseudodrusen). This finding is consistent with recent reports that drusen subtypes and/or subfoveal choroidal thickness are associated with the type of exudative AMD 12,31 . In the present study, Group1 (pachydrusen group) account for 12.2% among fellow eyes with unilateral exudative AMD and 17.9% among fellow eyes with unilateral PCV. Cheung et al. 32 reported that pachydrusen were seen in 25.5% of exudative AMD on the basis of color fundus photography and Lee et al. 33 reported that pachydrusen were seen in 49.3% of PCV on the basis of multimodal imaging including color fundus photography, SD-OCT, and ICGA. The figure in this study is lower than recent reports 32,33 . Further large studies would be needed to reveal the prevalence of pachydrusen in Asians.
A limitation of the present study is the relatively small sample size of patients with pachydrusen and patients with pseudodrusen. Therefore, the difference in the prevalence of CFH I62V variants between Group 1 and Group 3 did not reach statistical significance. In the present study, we did not subdivide Group3 into pseudodrusen with or without soft drusen. Large cohort study will reveal whether they are distinct entities and further studies would be necessary to elucidate the clinical and genetic differences among drusen subtypes.
In summary, pachydrusen have different clinical and genetic characteristics from soft drusen and pseudodrusen. They might be associated with pachychoroid and protection against exudative AMD.

Method
The medical charts of 330 consecutive patients with unilateral exudative AMD who first visited the Macular Clinic of the Department of Ophthalmology at University of Yamanashi Hospital between August 2011 and August 2017 were retrospectively reviewed. This retrospective study was approved by the institutional review board of the University of Yamanashi and followed the tenets of the Declaration of Helsinki.
All patients underwent comprehensive ophthalmic examination in both eyes including best-corrected visual acuity (BCVA) assessment using the Landolt chart, slit-lamp biomicroscopy with or without a 78D lens intraocular pressure measurement, color fundus photography covering the posterior retina within 45 degrees, fluorescein and indocyanine green angiography (FA/ICGA), near-infrared reflectance (NIR), fundus autofluorescence (FAF) (HRA-2; Heidelberg Engineering, Dossenheim, Germany), and spectral-domain optical coherence tomography (SD-OCT) (Spectralis version 5.4 HRA + OCT). The subjects were classified into 4 groups according to the type of drusen seen in the fellow eye. Group 0: no drusen within 45 degrees in the posterior retina. Group 1: pachydrusen. Group 2: soft drusen. Group 3: pseudodrusen with or without soft drusen. Late phase ICGA was employed to differentiate pachydrusen, which exhibits punctate hyperfluorescent spots 23 , from soft drusen showing hypofluorescence 22 (Figs 3 and 4). Late phase ICGA images were defined as ICGA images 10 minutes after dye injection. Eyes were diagnosed to have pseudodrusen if they showed a characteristic reticular pattern in at least one imaging modality including color fundus photography, FAF and NIR or subretinal drusenoid deposits with SD-OCT. (Fig. 5) The classification was confirmed by two independent graders (Y. F. and Y. S.) who was masked to the diagnosis of the affected eye. If their assessment differed, the final judgment was made by the third grader www.nature.com/scientificreports www.nature.com/scientificreports/ (H. I.). Since three patients were judged to have both soft drusen and pachydrusen in the fellow eye, they were excluded in this study. There were no patients who had both pseudodrusen and pachydrusen in the fellow eye.
The subfoveal choroidal thickness in the fellow eyes was measured as the vertical distance between the outer border of the retinal pigment epithelium and the choroidoscleral border, using SD-OCT images. A peripheral blood sample was collected before FA/ICG. Genomic DNA was purified using PUREGENE DNA Isolation Kit (Gentra Systems, Minneapolis, USA). Genotyping of ARMS2 A69S (rs10490924) and CFH I62V (rs800292) was performed using Taqman genotyping assays with the 7300/7500 Real-Time PCR System (Applied Biosystems, Foster City, USA), as previously described 34 . Written informed consent was obtained from all participants.