Multimodal imaging in a 44-year-old patient. A, Near-infrared reflectance composite image; arrow marks the position of the optical coherence tomographic (OCT) image. B, Near-infrared reflectance. C, Fundus color. D, Fundus autofluorescence. E, Fluorescein angiography at 10 minutes. F, Indocyanine green angiography at 10 minutes. G and H, Spectral-domain OCT. B-F and H represent a 2-fold magnification of the dashed squares in A and G, respectively.
Nine-field composite near-infrared reflectance images of 4 representative patients illustrating the topographic association between peau d’orange (blue area) and reticular pseudodrusen (yellow area).
A-D, ICG-A images at 30° field of view 10 to 15 minutes after dye injection. The area with RPD is marked in yellow. E-H, ICG-A images at 55° field of view approximately 30 minutes after dye injection to illustrate the extent of decreased ICG-fluoresence in the same eyes shown in the upper row.
A, Fundus color. B, Near-infrared reflectance. C, Indocyanine green angiography (ICG-A) at 10 minutes. D, Fundus autofluorescence. E and F, Spectral-domain optical coherence tomography (SD-OCT). Dotted lines indicate nasal margin of peau d’orange (left) and temporal margin of the RPD (right). Arrows indicate the position of the optical coherence tomographic sections (right corresponds to E and left to F). eTable 4 in the Supplement gives the criteria for differentiation.
eTable 1. Results of Skin Biopsy and Genetic Analysis of Investigated Patients With Pseudoxanthoma Elasticum
eTable 2. Demographic Data of Investigated Patients With Pseudoxanthoma Elasticum
eTable 3. Phenotypes of Investigated Eyes of Patients With Pseudoxanthoma Elasticum
eTable 4. Differentiation of Reticular Pseudodrusen and Peau d’Orange in Pseudoxanthoma Elasticum
eFigure. Topographic Distribution of Reticular Pseudodrusen in Pseudoxanthoma Elasticum
Gliem M, Hendig D, Finger RP, Holz FG, Charbel Issa P. Reticular Pseudodrusen Associated With a Diseased Bruch Membrane in Pseudoxanthoma Elasticum. JAMA Ophthalmol. 2015;133(5):581-588. doi:10.1001/jamaophthalmol.2015.117
Reticular pseudodrusen (RPD) are frequently associated with age-related macular degeneration and considered to be an independent risk factor for disease progression, but the pathophysiologic mechanisms are only incompletely understood. Therefore, it may be helpful to identify the associations of RPD with other diseases that have defined pathophysiologic mechanisms.
To describe the phenotype, prevalence, and topographic distribution of RPD in patients with pseudoxanthoma elasticum (PXE) and their association with a diseased Bruch membrane.
Design, Setting, and Participants
In this single-center, prospective, cross-sectional case series, 57 consecutive patients with PXE from a university referral center whose diagnosis has been confirmed by genetic testing and/or skin biopsy were studied from March 1, 2013, through February 28, 2014.
Main Outcomes and Measures
Phenotypic characteristics of RPD were evaluated with multiple imaging techniques. The RPD were defined as irregular networks of round to oval lesions that appear hyporeflective on near-infrared reflectance, hypoautofluorescent on fundus autofluorescence, and as subretinal deposits on spectral-domain optical coherence tomographic images. The presence of RPD was judged based on characteristic findings in at least 2 of the 3 imaging modalities.
A total of 57 patients were examined, and 15 patients were excluded mainly because of large central atrophy or fibrosis. In the remaining 42 patients with PXE, RPD were detected in 22 patients (52%; 95% CI, 38%-67%). Prevalence of RPD was highest in the fifth decade at 67% (10/15; 95% CI, 42%-85%). The RPD were most frequently located within the superior quadrant and least frequently located within the central macula. The RPD were always located central to areas with peau d’orange and within an area of hypofluorescence on late-phase indocyanine green angiographic images.
Conclusions and Relevance
These data suggest that RPD have a high prevalence in eyes of patients with PXE. Although RPD in patients with PXE occur at a younger age, their distribution and phenotype appear to be similar to RPD associated with age-related macular degeneration. The association with diseased Bruch membrane in PXE suggests a pathogenetic role of Bruch membrane alterations for the development of RPD.
Reticular pseudodrusen (RPD) were first described in 1990 by Mimoun et al1 as “pseudo-drusen visible en lumière bleu” based on their visibility in blue light. Within the past decade, interest in RPD has considerably increased mainly because of their frequent association with age-related macular degeneration (AMD),2- 5 which is the most frequent cause of severe vision loss and blindness in the western developed countries.6 It is assumed that RPD are a risk factor for progression to advanced atrophic or neovascular forms of AMD.7- 9 Furthermore, RPD may also directly contribute to loss of vision by inducing outer retinal atrophy.10
Despite the obvious association with AMD, there is yet very limited knowledge about the exact pathogenesis of RPD. There is some evidence of associated changes of inner choroidal layers, indicating vascular nutritive factors,11- 14 and histologic studies15,16 support an associated disturbance of lipid turnover.
Because of its localization between the choroid and the retinal pigment epithelium (RPE), the Bruch membrane (BM) might also play a role in the pathogenesis of RPD. The BM is a multilayer extracellular matrix that amongst others consists of 2 collagen-rich layers that encase a central layer, which is dominated by elastic fibers. The BM has a narrow interaction with the adjacent inner choroid and the RPE17 and has important functions for the integrity of the outer retina (ie, passively as a scaffold for adjacent cells but also as a regulator of diffusion processes between the choroid and the RPE). Changes of BM occur with increasing age and are also implicated in the pathophysiology of AMD with both focal (eg, drusen) and diffuse changes.18
A suitable model disease to study consequences of pathologic alteration in BM is pseudoxanthoma elasticum (PXE), which is a rare multisystem disorder of autosomal recessive inheritance (OMIM 264800). The disease has an estimated prevalence of 1:25 000 to 1:100 00019 and is associated with mutations in the ABCC6 gene (OMIM 603234), leading to calcification and fragmentation of connective tissue rich in elastic fibers.20 Because of its high content of elastic fibers, the BM becomes thickened and calcified in PXE. Subsequently, chorioretinal atrophy and choroidal neovascularization may develop, mimicking phenotypic characteristics of AMD.21 To date, there is some evidence of lesions resembling RPD in PXE,21,22 suggesting a possible role of the BM in the pathophysiology of RPD. Therefore, we investigated the phenotype, prevalence, and topographic localization of RPD and their relation to BM calcification in patients with PXE using various imaging modalities to further clarify the association of diseased BM and the presence of RPD.
Patients were recruited for this single-center, cross-sectional prospective case series from the retinal clinic of the Department of Ophthalmology, University of Bonn, from March 1, 2013, through February 28, 2014. Institutional review board approval (Ethikkommission, Medizinischen Fakultät, Rheinische Friedrich–Wilhelms–Universität Bonn) and written patient consent were obtained. The study adhered to the Declaration of Helsinki.
All patients underwent a complete ophthalmologic examination, including best-corrected visual acuity, dilated ophthalmoscopy, and fundus photography. Confocal scanning laser ophthalmoscopy, spectral-domain optical coherence tomography (SD-OCT), and, in selected cases, fluorescein angiography (FA) and/or indocyanine green angiography (ICG-A) were performed with the same instrument (Spectralis HRA+ OCT; Heidelberg Engineering). Refractive error (spherical equivalent) was measured by autorefractometry (Nidek ARK-560A; Nidek Co).
Inclusion criteria were the diagnosis of PXE based on genetic testing as described previously23 and/or histopathologic findings in skin biopsy specimens and the presence of characteristic ocular fundus alterations. Exclusion criteria were an unstable fixation and severe media opacities that would not allow for high-quality retinal imaging, widespread central retinal disease reaching beyond the outer ring of an Early Treatment Diabetic Retinopathy Study grid centered to the fovea, any additional retinal disease, prior retinal surgery, or any other previous treatment, including laser photocoagulation or photodynamic therapy. Pretreatment with intravitreal vascular endothelial growth factor inhibitors was not an exclusion criterion. Details on the imaging protocol and image analysis can be found in the eMethods in the Supplement.
According to previous studies,2,4,7,24- 28 RPD were defined as yellowish to whitish round structures that formed an interlacing network on fundus color images. On near-infrared (NIR) reflectance and fundus autofluorescence (AF) images, RPD were defined as an irregular network of round to oval hyporeflective or hypoautofluorescent lesions, sometimes with higher reflectivity or autofluorescence in the center of the lesion, respectively. On SD-OCT, accumulations between the band most likely representing the ellipsoid zone and the RPE-BM complex were considered RPD. For a definite diagnosis of RPD, characteristic findings had to be present in at least 2 of the 3 most sensitive imaging modalities for the detection of RPD (fundus AF, NIR reflectance, and SD-OCT).
Statistical analysis was performed using SPSS statistical software, version 20.0 (IBM Corp). The mean subfoveal choroidal thickness and distance between peau d’orange and the optic disc was compared between groups using the 2-tailed t test after data distribution was confirmed to be gaussian. Categorical variables were assessed using the Pearson χ2 test. The 95% CIs of categorical variables were assessed using a modified Wald method.
A total of 57 consecutive patients with PXE were examined for this study. Genetic testing was performed in 53 patients. A total of 42 patients had 2 mutations, 10 patients had 1 mutation, and 1 patient had no detected disease-causing mutation in ABCC6. Histopathologic examination of a skin biopsy specimen was conducted in 33 patients, including the patient without a detected disease-causing mutation, and revealed changes characteristic for PXE in all patients tested (eTable 1 in the Supplement).
Fifteen patients were excluded from further analysis because reliable conclusion on the presence or absence of RPD was not possible (13 because of bilateral widespread fibrosis and/or atrophy at the posterior pole and 2 because of poor imaging quality). Of the remaining 42 patients, 22 (52%; 95% CI, 38%-67%) revealed evidence of RPD in one or both eyes (total of 38 eyes) on at least 2 imaging modalities (NIR reflectance, fundus AF, and/or SD-OCT) (Figure 1A). The RPD were detected in all 3 imaging modalities applied in 33 (87%) of the 38 eyes (95% CI, 72%-95%). Visibility was highest on NIR reflectance (38 [100%] of 38; 95% CI, 89%-100%) followed by SD-OCT (37 [97%] of 38; 95% CI, 85%-99%) and AF images (35 [92%] of 38; 95% CI, 78%-98%). In 9 of the 42 patients, only one eye was evaluated for the presence of RPD because of widespread fibrosis and/or atrophy in the fellow eye.
The mean (SD) age of the 22 patients with PXE and RPD was 48.6 (6.7) years (range, 41-63 years). A difference in sex distribution and presence of choroidal neovascularization or atrophy between patients with PXE with and without RPD was not identified, although there was a tendency toward a higher rate of atrophy in patients with RPD (eTable 2 and eTable 3 in the Supplement).
Figure 1B illustrates the distribution of all patients (n = 57) with or without RPD and with bilateral late disease stages across different age groups. There were no RPD in patients younger than 40 years. The frequency of RPD in at least one eye was highest (10 [67%] of 15; 95% CI, 42%-85%) between 40 and 50 years of age, decreasing thereafter (3 [20%] of 15; 95% CI, 6%-45%) in those older than 60 years. The proportion of patients without RPD was stable, ranging from 25% to 30% across groups older than 40 years. The proportion of probands with bilateral widespread central atrophy or fibrosis not allowing image analysis for RPD increased with age, with 8 (53%) of 15 (95% CI, 30%-75%) affected by such late disease manifestations in those older than 60 years.
On NIR reflectance images, RPD appeared as round to oval hyporeflective lesions (Figure 2A and B and eTable 4 in the Supplement). Sometimes RPD exhibited a target-like structure with reflectance in the center of the lesion ranging from normal to increased. Frequently, RPD were visible as white to yellowish dots on fundus photographs (Figure 2C). On fundus AF images (Figure 2D), RPD presented as round to oval hypoautofluorescent lesions with a network-like appearance of the interlacing areas comparable to that visible on NIR reflectance images. Infrequently, RPD also had dots of increased AF. On 10-minute FA and ICG-A images (Figure 2E and F), lesions were hypofluorescent with a comparable distribution as observed on NIR reflectance or fundus AF images. The SD-OCT images revealed hyperreflective accumulations, which seemed to be located above the RPE, leading to an undulation of the ellipsoid band in these areas (Figure 2G and H). Sometimes there was discontinuity of the ellipsoid band and intrusion of the subretinal material toward the outer nuclear layer.
Topographically, RPD were most frequently located in the outer superior region (35 [92%] of 38 eyes; 95% CI, 78%-98%) or the middle superior area (27 [71%] of 38 eyes; 95% CI, 55%-83%). The RPD were observed less commonly in the central subfield (1 [3%] of 38 eyes; 95% CI, 0%-15%) and the inner nasal area (3 [8%] of 38 eyes; 95% CI, 2%-22%) (eFigure in the Supplement).
A characteristic finding in PXE is the so-called peau d’orange, which likely represents a transition zone between central calcification of BM and more peripheral uncalcified areas.29,30 Differentiation between peau d’orange and RPD may be difficult in selected cases; however, specific differences guide differentiation between those 2 phenotypic findings (eTable 4 in the Supplement). Localization of RPD was linked to the localization of peau d’orange. Eyes with peau d’orange in the central fundus typically had no RPD (Figure 3A). Increasing eccentricity of peau d’orange (blue area in Figure 3) was associated with presence and increasing eccentricity of RPD (yellow area in Figure 3B-D). Peau d’orange was always located more peripherally than RPD and thus precedes occurrence of RPD in a disease process that spreads centrifugally over time. To quantitatively investigate the topographic relation between peau d’orange and RPD, the horizontal distance between the temporal margin of the optic disc and the most central border of peau d’orange temporally was measured and compared between probands with and without RPD (eyes with advanced atrophy and/or fibrosis were not included in this analysis). This distance was larger in the group with RPD (mean, 7.2 mm; 95% CI, 6.9-7.5 mm; range, 5.6-8.1 mm) compared with patients without RPD (mean, 4.8 mm; 95% CI, 3.9-5.8 mm; range, 2.5-8.1 mm) (P < .001). The RPD were never observed in patients with peau d’orange visible within less than 5.6 mm to the margin of the optic disc.
Another consistent finding in patients with PXE is a reduced fluorescence central to peau d’orange on late-phase ICG-A recordings.29 Although this phenomenon is typically fully developed approximately 30 minutes after dye injection, it was visible on images recorded at approximately 10 minutes in all 23 patients examined by ICG-A. The RPD were visible as darker spots within the area of reduced fluorescence at 10 minutes (Figure 4A-D) but not at 30 minutes. Patients without RPD typically have the least extensive hypofluorescence (Figure 4A, as compared with Figure 4F-H). If present, RPD were always located within the area of reduced ICG fluorescence and never in more eccentric areas with preserved late-phase ICG fluorescence (Figure 4B-D).
The findings in various imaging modalities suggest a centrifugal spread of subsequent PXE-related fundus changes. The most peripheral alteration is peau d’orange, which is best visible on fundus photographs or NIR reflectance images and has no drusen-like changes on OCT images (Figure 5A, B, and F). Central to peau d’orange is an area of decreased ICG-A late-phase fluorescence (Figure 5C). The most central fundus changes are RPD (Figure 5A-E), which may be followed by chorioretinal atrophy. Angioid streaks, which may occur within calcified BM, mostly extend from the area around the optic nerve head and do not cross peau d’orange.
Choroidal thickness was significantly thinner in patients with PXE and RPD (mean, 200.1 µm; 95% CI, 169.6-230.5 µm) compared with those without RPD (mean, 270.1 µm; 95% CI, 226.8-313.4 µm; P = .006). There was no significant difference between patients with or without RPD with regard to mean (SD) age (50.7 [6.3] vs 47.5  years; P = .40) and refractive error (mean [SD] spherical equivalent) (−0.7 [1.3] vs 0 [1.6] D; P = .20).
Pseudoxanthoma elasticum is a monogenetic disease with a defined pathophysiology that results in an ocular phenotype that is driven by progressive calcification and thickening of the BM.21 Thus, PXE appears suitable to specifically investigate the consequences of alterations in the BM.
On the basis of this study, RPD are a common finding in patients with PXE, with a peak prevalence between 40 and 50 years of age. At older ages, RPD become less common and appear to be absent or rare in younger patients. Despite the cross-sectional analysis of this study, these results suggest a sequence of fundus changes in which the BM alterations and/or associated changes culminate in a critical damage of the BM, which then leads to the development of RPD. The decreasing prevalence of RPD with age may be explained by the development of chorioretinal atrophy or secondary choroidal neovascularization development with subsequent scarring, leading to reduced detectability or disappearance of RPD.
Ocular alterations related to PXE typically initially occur at the central fundus from where they spread centrifugally, resulting in the most pronounced phenotype in the papillomacular area and the least changes in the periphery.29 This process also allows the observation of different disease stages from the periphery to the central fundus. Peau d’orange, which marks the transition from calcified to uncalcified BM, was localized peripherally to the RPD. A similar topographic relation was identified for the transition zone between normal and centrally reduced late-phase ICG-A fluorescence, the latter being a result from reduced ICG staining of altered BM and/or RPE.29 Thus, specific PXE-related fundus changes are localized eccentric to the RPD and appear to precede their development.
On the basis of large epidemiologic studies, RPD in patients with AMD or otherwise healthy probands are only rarely observed before the age of 50 years, and the peak prevalence is typically in those older than 65 years. In those older than 75 years, the frequency of RPD is estimated to be approximately 5% in healthy probands3 and approximately 30% to 50% in patients with AMD depending on included phenotypes and sensitivity of the imaging modalities used.2,4,31 Although epidemiologic data are mostly relying on fundus photography with a generally lower detection rate, making comparison with this study difficult, one can conclude that RPD in AMD-affected or otherwise healthy eyes occur considerably later compared with those eyes with PXE.
The phenotype24,25,27 and distribution of RPD in patients with PXE with preference of the superior quadrant and relative foveal sparing3,32 were similar to RPD in older patients and in patients with AMD, indicating common pathogenic pathways with a contribution of diseased BM. Aging is associated with changes in the BM, and such changes appear to be aggravated and to occur earlier in patients with AMD. Reported changes include thickening,33 calcification and fragmentation of elastic fibers,34 lipid accumulation,35,36 and deposition of advanced glycation end products.37 Severe pathologic changes of BM typically occur in patients with PXE typically earlier in life.38 Thus, the different age at onset of RPD in healthy controls and patients with AMD and PXE might reflect a continuum of BM pathologic changes, with a minimal late-onset phenotype due to normal aging and pronounced early alterations associated with PXE.
Patients with PXE have no evidence of more frequent or earlier occurrence of sub-RPE drusen as they are characteristic of AMD. Their pathogenesis appears to be driven rather by a dysregulated complement activation39 and/or oxidative stress40 than by the pathologic mechanisms of the BM. Different pathophysiologic pathways for these 2 lesion types are also in line with a previously reported 2-compartment biogenesis model15 and differences in the lipid composition of RPD and sub-RPE drusen.41
The findings of this study suggest an important contribution of BM pathologic changes for the pathophysiology of RPD. These age- and disease-related alterations of BM might change its properties (eg, for the attachment of RPE cells or the diffusion processes between the choroid and the RPE as shown for the aging BM).42- 44 These alterations subsequently might disturb various physiologic processes, including the removal of waste products of photoreceptors and the RPE, the supply of the RPE and outer retina with nutrients and oxygen, and the recycling or degradation of shed outer segments and retinoids.18,45 The formation of RPD might eventually result from such impaired physiologic mechanisms of the choroid-BM-RPE complex.
Previous studies have suggested an important role of the choroid for the development of RPD, including reduced choroidal thickness12- 14 or an association with choroidal watershed zones.11 In patients with PXE, a markedly reduced choroidal thickness has been reported46 that is associated with the presence of RPD. Possible explanations may include an independent effect of choroidal alterations on RPD development. Then, various causes of choroidal pathologic changes should lead to the development of RPD. Alternatively, the choroidal alterations might also be secondary to BM disease. In this scenario, RPD could result directly from BM impairment or from secondary changes of the choroid or RPE. Combined effects might certainly also be involved. Histopathologic investigations of eyes of patients with PXE have not yet focused on choroidal changes and the presence of RPD, but limited evidence suggests a characteristic loss of the choriocapillaris underneath a thickened and calcified BM.21,38,47 Investigation of other model diseases for primary RPE, BM, and choroidal alterations, respectively, as well as histopathologic studies of such diseases for the presence of RPD and choroidal changes, will be valuable to further elucidate the pathogenesis of RPD.
The main limitation of the study is the relatively low number of included patients across age groups with a wide variety of different phenotypes, including eyes with and without choroidal neovascularization or chorioretinal atrophy. This limitation led to relatively wide CIs but is an inherent challenge when reporting findings in rare diseases.
The findings suggest an association of BM pathologic changes with the development of RPD. Moreover, the study reveals that RPD not only are found in AMD and in advanced age but also represent a specific phenomenon that results from impaired choroid-BM-RPE interactions, independent of the underlying disease. These findings might lead to a better understanding not only of the pathophysiologic mechanisms of RPD but also of multifactorial diseases, such as AMD, to find more specific treatments to prevent vision loss.
Submitted for Publication: March 23, 2014; final revision received January 5, 2015; accepted January 5, 2015.
Corresponding Author: Peter Charbel Issa, MD, DPhil, Department of Ophthalmology, University of Bonn, Ernst-Abbe-Str 2, 53127 Bonn, Germany (firstname.lastname@example.org).
Published Online: March 12, 2015. doi:10.1001/jamaophthalmol.2015.117.
Author Contributions: Dr Gliem had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Gliem, Finger, Holz, Charbel Issa.
Acquisition, analysis, or interpretation of data: Gliem, Hendig, Holz, Charbel Issa.
Drafting of the manuscript: Gliem.
Critical revision of the manuscript for important intellectual content: Hendig, Finger, Holz, Charbel Issa.
Statistical analysis: Gliem.
Obtained funding: Charbel Issa.
Administrative, technical, or material support: Hendig, Holz, Charbel Issa.
Study supervision: Charbel Issa.
Conflict of Interest Disclosures: Dr Holz reported serving as a consultant for Heidelberg Engineering. No other disclosures were reported.
Funding/Support: This study was supported by ProRetina Deutschland, Aachen, Germany; grant O-137.0018 from the BONFOR research program of the University of Bonn, Bonn, Germany (Dr Gliem); and grant 529923 from the National Health and Medical Research Council Centre for Clinical Research Excellence, Canberra, Australia (Dr Finger). The Department of Ophthalmology, University of Bonn, receives imaging devices from Heidelberg Engineering. The Centre for Eye Research Australia receives operational infrastructure support from the Victorian government.
Role of the Funder/Sponsor: The funding sources had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and the decision to submit the manuscript for publication.
Correction: This article was corrected on April 6, 2015, to fix an error in the byline.