To subclassify pseudodrusen based on their appearance in multimodal imaging.
Retrospective, observational series.
The color fundus photographs and infrared scanning laser ophthalmoscope (IR-SLO) images of patients with pseudodrusen were evaluated along with spectral-domain optical coherence tomography (SD OCT) by masked readers. Distinct types of pseudodrusen could be differentiated.
There were 140 eyes of 93 patients with a mean age of 82.4 years. Multimodal imaging analysis showed 3 subtypes of pseudodrusen. One principal type was an orderly array of whitish discrete accumulations principally located in the perifovea, termed dot pseudodrusen. They appeared as hyporeflective spots, often with a target configuration, in IR-SLO images. The second type was interconnected bands of yellowish-white material forming a reticular pattern, called ribbon pseudodrusen, which were located in the perifovea. This subtype was faintly hyporeflective in IR-SLO imaging. Dot pseudodrusen were detected more commonly with IR-SLO imaging than in color photography ( P = .014) and ribbon pseudodrusen were seen more frequently in color than in IR-SLO images ( P < .001). An uncommon third type of pseudodrusen, yellow-white globules primarily located peripheral to the perifoveal region, appeared hyper-reflective in IR-SLO and were called peripheral pseudodrusen. All 3 types were seen as subretinal drusenoid deposits by SD OCT.
Pseudodrusen may be classified into at least 3 categories, each with optimal methods of detection and only 1 that formed a reticular pattern. These findings suggest pseudodrusen could contain differing constituents and therefore may vary in conferred risk for progression to advanced age-related macular disease.
Numerous studies have shown an association between pseudodrusen and manifestations of advanced forms of age-related macular degeneration (AMD), including geographic atrophy, choroidal neovascularization (CNV), and outer retinal atrophy. Population-based studies have classified pseudodrusen as soft drusen and therefore do not provide an opportunity to evaluate the independent risk pseudodrusen confer. Review of clinic-based studies has shown a remarkable variation in the proportion of eyes demonstrating pseudodrusen, and consequently the estimates of risk also vary. While the large range of reported prevalence may be attributable in part to differing patient samples and to improved recognition of pseudodrusen during clinical examination and in fundus imaging, there may be a larger problem with phenotype recognition.
In the first publication by Mimoun and associates in 1990, pseudodrusen were described as being a yellowish interlacing network that was best recorded with blue-light fundus photography. Arnold and associates modified the name in 1995 to “reticular pseudodrusen.” The pseudodrusen they visualized could be round or oval yellow spots that could join to form branches and an ill-defined interlacing network—hence the name reticular, which was derived from the Latin rete, or net. The use of infrared confocal scanning laser ophthalmoscopy for detecting pseudodrusen was introduced by Schmitz-Valckenberg and associates, but they described pseudodrusen as an array of round or oval irregularities approximately 150–250 μm in diameter showing decreased near-infrared reflectance. The lesions were single or could be clumped, but areas between these discrete lesions were said to exhibit no marked changes. In a subsequent publication Schmitz-Valckenberg and associates diagnosed the presence of pseudodrusen in infrared scanning laser ophthalmoscopic images if there was a regular complex of uniform round or oval irregularities with a diameter ranging between 50 and 400 μm. On the other hand, they made the diagnosis in color images if there were “yellow-pale or pale light ill-defined networks of broad, interlacing ribbons.” This raises the question: are the lesions imaged in color (or its subset, blue channel photography) the same as those imaged by infrared light, or is there more than 1 phenotype of pseudodrusen, each with differing imaging characteristics?
To answer this question, we reviewed patients with the known diagnosis of pseudodrusen who were part of previous Institutional Review Board (IRB)-approved studies, most of which have already been published. The color and infrared scanning laser ophthalmoscopic images of the patients within the region subtended by the Early Treatment Diabetic Retinopathy Study (ETDRS) grid overlay were evaluated. Along the way we realized that there was another presentation of pseudodrusen located outside of the ETDRS grid. From this investigation it appears that there are at least 3 different forms of pseudodrusen and all 3 types appear as subretinal deposits when imaged by spectral-domain optical coherence tomography (SD OCT).
Patients and Methods
This study retrospectively reviewed a compilation of patients with the diagnosis of pseudodrusen who were part of previous IRB-approved studies conducted by the senior author (R.F.S.), most of which have already been published; the present study was approved by the Western IRB and complied with the Health Insurance Portability and Accountability Act of 1996. The patients could have no history of laser photocoagulation of the macula, rhegmatogenous or tractional retinal detachment, high myopia, acquired vitelliform detachment, or any retinal dystrophy or tapetoretinal degeneration.
Color fundus photographs were obtained with a Topcon ImageNet camera (Topcon America, Paramus, New Jersey, USA) and were viewed in Topcon ImageNet (version 2.55; Topcon America). Histogram stretching was used to standardize the images. To fully evaluate the color photographic information the blue channel was evaluated. This same viewing module was used to view the blue channel of the color fundus photographs by selecting the commands Utilities>RGB channels. The 3 principal channels (red, green, and blue) comprising the color image were then displayed along with the original color photograph and histogram stretching to standardize the blue color channel. An Early Treatment Diabetic Retinopathy Study (ETDRS) grid was overlaid on the color fundus photograph centered on the geometric center of the macula. The SD OCTs of the eyes were obtained with the Heidelberg Spectralis (version 1.6.1) as viewed with the contained Heidelberg software (Spectralis Viewing Module 184.108.40.206; Heidelberg Engineering, Heidelberg, Germany). Coincident with the OCT imaging was the capture of an infrared scanning laser ophthalmoscope (IR-SLO) image (820 nm high-resolution mode: 1536 × 1536 pixels) that was evaluated by deselecting the option “Show Scan Positions” under “HRA Image.” In each patient of the case group, 31 B-scans were obtained within a 20 × 25-degree rectangle to encompass the macula, including the area to the temporal arcades. The distance between scans was 240 μm. The number of averaged images per section was dictated by image quality and the ability of the patient to maintain fixation, but typically was at least 10.
Each class of imaging was reviewed by 2 reviewers masked to the outcome of each other, and if there were discrepancies there was open adjudication with the senior author. Nonexudative AMD was diagnosed if the patient had 1 or more soft drusen >125 μm or more than 5 intermediate drusen (≥63 and <125 μm) or any focal hyperpigmentation but did not have evidence of either geographic atrophy or choroidal neovascularization. The diagnosis of neovascular AMD was based on fluorescein angiography. Because of the nature of the study, eyes were graded for pseudodrusen on a per-modality basis as opposed to a per-eye basis. In color imaging pseudodrusen were considered to be present if they were visible in the color photograph and if 5 or more drusen were brighter in the blue than in the green channel of the color photograph. In the infrared imaging pseudodrusen were thought to be present if there were 5 or more hyporeflective spots unrelated to the presence of an alternate ocular pathologic process. OCT was used in a confirmatory sense and was not primarily evaluated as a diagnostic testing modality in this study. Subretinal drusenoid deposits, the histologic and OCT visualization of subretinal material, was considered to be consistent with the diagnosis of pseudodrusen if there were 5 or more collections visible. Enhanced depth imaging OCT was performed for the choroidal thickness measurements. Digital calipers were placed at the outer edge of the hyper-reflective retinal pigment epithelium (RPE) line and the inner border of the hyper-reflective surface located behind the large choroidal vessels, which is the scleral/choroidal interface.
Definitions and Data Analysis
Pseudodrusen distribution was evaluated according to the ETDRS grid. The presence of pseudodrusen in the center circle and the 4 sectors (superior, temporal, inferior, and nasal) was recorded. Separate categories were created to summarize these data. The term “Any Subfield” was considered to be true if any 1 of the 5 regions contained pseudodrusen. The term “Preponderance of Subfields” was considered to be true if 3 or more of the 5 subfields contained pseudodrusen. Since there is no accepted gold-standard fundus photographic method to detect pseudodrusen, their presence, and that of their phenotypic subtypes, was considered to be true in any given eye if they were detected by either color photography or IR-SLO.
The data obtained were analyzed with frequency and descriptive statistics. χ 2 testing was used for categorical analysis. The statistical analyses were performed with IBM SPSS software version 20 (IBM SPSS, Inc, Chicago, Illinois, USA). For all tests a P value less than .05 was considered significant.
There were 93 patients with a mean age of 82.4 years (median 83.7, interquartile range [IQR] 78.1–87.5 years); 23 were male and 70 were female. Forty-six of 186 eyes were excluded because there were no pseudodrusen in the evaluated eye (12 eyes), scarring and atrophy occupying the ETDRS grid (18 eyes), or inadequate color photographs (16 eyes). Of the remaining 140 eyes, 132 had involvement principally within the ETDRS grid and 8 outside the grid; 63 eyes had a history of CNV, all of which were treated. The mean visual acuity of the eyes with no CNV was 20/36 (logMAR 0.255) as compared with those with CNV, 20/64 (logMAR 0.507). Pseudodrusen detected in the ETDRS grid by any means were present in all eyes, consistent with the entry criteria. Pseudodrusen detected by color fundus photography were found in Any [ETDRS] Subfield in 101 of 132 eyes (76.5%) and by IR-SLO imaging in Any Subfield in 115 (87.1%). Pseudodrusen detected in the Preponderance of Subfields occurred in 56 of 132 eyes (42.4%) using color fundus photography and 78 eyes (59.1%) using IR-SLO imaging. The difference in correlated proportions was not significant for the designation Any Subfield ( P = .059), but the difference was significant for the classification Preponderance of Subfields ( P = .005, McNemar test).
Detailed review of the multimodal imaging information showed 3 subtypes of appearance of pseudodrusen ( Table ). The most common were seen as an orderly array of discrete dot-like accumulations principally located in the perifoveal area. For the purposes of this study this type of pseudodrusen was termed “dot” pseudodrusen. Dot pseudodrusen were considered to be relatively white spots in a color photograph. In IR-SLO imaging the dot pseudodrusen appeared as hyporeflective structures that sometimes had a central round area of slightly greater reflectivity, giving a target appearance ( Figure 1 ). Among eyes with dot pseudodrusen, they were more commonly detected using IR-SLO imaging than with color fundus photography for both the designations Any Subfield and Preponderance of Subfields ( P = .014 and P = .044, respectively, χ 2 test). Dot pseudodrusen were seen in 127 of the evaluated eyes (96.1%). A second phenotype appeared as interconnected ribbons or bands of material located most prominently in the perifoveal region. For the purposes of this study this phenotype was termed ribbon pseudodrusen, although the pattern caused by the interlocking bands did suggest a reticular pattern, a sweeping term sometimes applied to all appearances of pseudodrusen ( Figure 2 ). Ribbon pseudodrusen were considered to be an interconnected network of material that creates the appearance of broad interlacing ribbons in a color photograph. This form was seen in 53 eyes (40.2%). The ribbon pseudodrusen was much more commonly seen in color photographs than in IR-SLO images for both the designations Any Subfield and Preponderance of Subfields ( P < .001 each, χ 2 test). Both the dot and ribbon patterns were seen together in 48 eyes. Representative images of the appearance of dot and ribbon pseudodrusen using fundus photographs, infrared scanning laser ophthalmoscopy, and spectral-domain OCT are shown in Figure 3 . Sector representation of the percentage of detection of dot and ribbon pseudodrusen in color and infrared images can be seen in Figure 4 . Along the way we realized that there was another presentation of pseudodrusen located outside of the ETDRS grid. The third phenotype of pseudodrusen was small, irregularly spaced, and frequently confluent globules principally located peripheral to the perifoveal region and therefore located outside of the ETDRS grid ( Figures 5 and 6 ). The third phenotype of pseudodrusen, considered to be yellow, small individual globules, can be subconfluent in a color photograph. The third phenotype of pseudodrusen was seen in 8 eyes. The majority of the deposits were located outside of the vascular arcades. The material was easily visualized in both color and IR-SLO images. Unlike other forms of pseudodrusen, these deposits were hyper-reflective in the IR-SLO images. The material was seen to be situated in the subretinal space in OCT images. This presentation is not to be confused with the more common conventional drusen found in the periphery, which are below the RPE and are not hyper-reflective in IR-SLO images. In addition, the representative case of this type showed regression of pseudodrusen, which was reported recently.
|Color photography (when visible)||Discrete dots, can be confluent||Interlocking ribbons, can be confluent||Small individual globules, can be subconfluent|
|IR-SLO (when visible)||Discrete hyporeflective dots, often with target configuration||Faint hyporeflective ribbons||Hyper-reflective spots|
|OCT||Subretinal accumulation of material, typically forming sharp peaks||Subretinal accumulation of material, typically forming broad, rounded elevations||Subretinal accumulation of material, typically forming rounded elevations|
|Distribution||Perifoveal with decreasing size in the more central parafoveal areas||Perifoveal with decreasing visibility in the more central parafoveal areas||Zone peripheral to the perifovea extending to outside the vascular arcades|
|Ocular correspondence||Bilateral symmetry||Bilateral symmetry||Too few cases to make definitive statement; appears bilaterally symmetrical|