Correlation of morphological alterations and functional impairment for different disease stages according to the Gass and Blodi classification with fundus photograph and optical coherence tomography results (left) and photopic (middle) and scotopic (right) fine matrix mapping contour plots (contour steps, 0 .1 log unit) superimposed on the late-phase fluorescein angiogram images. A, Stage 2 (patient 1). In the presence of early features of mild capillary dilatations and diffuse staining in angiography temporal to fixation, no photopic dysfunction compared with age-matched control subjects can be seen. In contrast, there is slight stretching and widening from the center toward eccentricity of scotopic contour plots. B, Stage 3 (patient 5). Despite moderate central visual acuity (20/50), morphological alterations with large cystoid spaces in optical coherence tomography and late-phase intensive staining temporal to the fovea are spatially correlated with a scotoma. Depth and extension of localized dysfunction are greater with scotopic testing compared with photopic testing. C, Stage 4 (patient 9). This patient still has moderate visual acuity (20/50), but a hyperplastic pigmented lesion in the parafovea seen with fundus photography, optical coherence tomography, and angiography results in severe photopic and scotopic scotomata just next to fixation. Photopic thresholds in the area between pigmented lesions are only moderately reduced, indicating that a certain degree of photopic function is preserved. In contrast, scotopic sensitivity dysfunction with 30 dB of loss encompasses both the pigmented lesions and the area in between. Of note, no obvious morphological changes can be seen in the center on the images shown.
Comparison of the distribution of the number of photopic and scotopic test points with a localized loss of 10 dB or more for the tested eyes (compared with those of age-matched control subjects). The difference between photopic and scotopic sensitivity was statistically significant (P = .007). The error bars indicate the entire range of values.
Schmitz-Valckenberg S, Fan K, Nugent A, Rubin GS, Peto T, Tufail A, Egan C, Bird AC, Fitzke FW. Correlation of Functional Impairment and Morphological Alterations in Patients With Group 2A Idiopathic Juxtafoveal Retinal Telangiectasia. Arch Ophthalmol. 2008;126(3):330-335. doi:10.1001/archopht.126.3.330
To correlate functional impairment with morphological alterations in patients with group 2A idiopathic juxtafoveal retinal telangiectasia.
As part of the Macular Telangiectasia Project, a cohort of 10 patients underwent additional functional testing and imaging studies including photopic and scotopic fine matrix mapping, microperimetry, reflectance, and autofluorescence imaging with scanning laser ophthalmoscopy.
From clinical stage 2 to 5, scotopic central function was reduced, which corresponded to depletion of macular pigment density. From clinical stage 3 onward, severe photopic and scotopic scotomata with up to 30 dB of loss were found next to fixation and were not totally confined to abnormalities seen with standard imaging modalities. The number of test points with loss of 10 dB or more was significantly greater for scotopic testing than for photopic testing (P = .007, Wilcoxon signed rank test).
Rod function may be more severely affected than cone function in patients with group 2A idiopathic juxtafoveal retinal telangiectasia, and this may occur early in the disease progression. Severe reduction in retinal sensitivity is spatially confined to morphological alterations seen with scanning laser ophthalmoscopy imaging. The findings imply that idiopathic juxtafoveal retinal telangiectasia is not solely a vascular disease and that early neuronal involvement may be implicated in the pathogenesis of the disease.
Idiopathic juxtafoveal retinal telangiectasia (IJT) is a rare condition causing progressive visual loss. It usually manifests with a slow decrease in vision, metamorphopsia, positive scotoma, and reading difficulties.1- 7
According to the classification introduced by Gass and Blodi2 in 1993, group 2A is the most common type of IJT. Occult, bilateral involvement with juxtafoveolar telangiectasis, minimal exudation, superficial retinal crystalline deposits, and right-angle venules characterize this manifestation of IJT. Late in the course of the disease, hyperplastic pigment plaques and retinal neovascularization may occur. Based on fundus photographs and fluorescein angiography findings, Gass and Blodi divided group 2A IJT into 5 clinical stages according to the disease development. More recent studies of patients with group 2A IJT using optical coherence tomography (OCT) examination have revealed intraretinal cystoid spaces without foveal thickening and disruption of the photoreceptor inner segment–outer segment junction early in disease evolution, whereas later stages showed foveal thinning and outer retinal atrophy.1,8- 13 Preliminary data using 2-wavelength autofluorescence imaging indicate that macular pigment density (MPD) is significantly reduced in the central retina.14
T hese recent findings have provided increasing evidence that group 2A IJT is not a disease limited to the retinal vasculature. However, to date, little is known of the functional consequences of cystoid spaces or the irregular distribution of macular pigment and their relationship with telangiectatic retinal vessels. It is also still unknown why telangiectatic vessels are usually first found in the temporal parafoveal area and which retinal layers and cells are primarily affected. A gradual decrease in central visual acuity has been reported, but no systematic studies on the spatial correlation of morphological alterations and functional impairment are available to our knowledge. Overall, the pathogenesis of group 2A IJT remains unclear and there is no treatment available to halt or alter the progression of this disease.15
In this study, we performed functional testing and imaging studies on a cohort of patients with group 2A IJT to better understand the functional implications of morphological changes with regard to both cone and rod photoreceptor system involvement.
As part of the prospective Macular Telangiectasia Project (MacTel Project, http://www.mactelresearch.com), a cohort of patients with group 2A IJT was recruited from Moorfields Eye Hospital for additional functional and imaging studies. The study followed the tenets of the Declaration of Helsinki and was approved by the local institutional review board and the local ethics committee. Informed consent was obtained from each patient after explanation of the nature and possible consequences of the study.
Only patients with stable fixation and clear media in at least 1 eye to allow central visual function testing and autofluorescence imaging were enrolled in the study. Exclusion criteria were any history of retinal surgery, laser photocoagulation, or other retinal diseases in the study eye. Patients with a history of diabetes were included only if no clinical diabetic retinal microangiopathy was present. One eye was chosen as the study eye. If both eyes were eligible, the eye with the earlier stage according to the Gass and Blodi classification was selected.2 If both eyes were eligible and were at the same disease stage, the right eye was chosen.
Patients were seen as part of their regular 6 months' appointment within the longitudinal MacTel Project. Best-corrected visual acuity was determined using Early Treatment Diabetic Retinopathy Study charts on a logarithmic scale. As part of the current study, a battery of additional functional tests and imaging studies was performed.
The study eye underwent the following protocol: (1) standard visual field examination (Humphrey field analyzer; Carl Zeiss Meditec, Dublin, California) using the 30-2 program (data not shown); (2) photopic fine matrix mapping (FMM); (3) pupil dilatation with 1.0% tropicamide and 2.5% phenylephrine hydrochloride; (4) dark adaptation for 45 minutes; (5) dark-adapted 30-2 visual field examination as described elsewhere16 (data not shown); and (6) scotopic FMM.
The technique of FMM has been described before.16- 21 Briefly, a modified Humphrey field analyzer allowed the assessment of photopic and scotopic retinal function. Whereas photopic retinal sensitivity was measured by standard Humphrey size III target white flashes, scotopic measurements were done after dark adaptation with Humphrey size III blue flashes. Accuracy of fixation was monitored by means of an infrared camera. The measurement encompassed a 9° × 9° field with 100 testing points centered on the fovea.
A numerical matrix of the luminance sensitivity at each of the test locations was calculated by spatial processing of FMM thresholds with a 3 × 3 gaussian (normal) filter. This was subsequently used to generate a contour plot and a 3-dimensional surface plot showing the size and location of luminance sensitivity gradients across a grid (contour steps, 0.1 log unit). Using the center of fixation (obtained with microperimetry) as the foveola and the blind spot as the center of the optic disc, superimposition of luminance sensitivity contour plots and imaging results were achieved with customized software.
Both eyes underwent automated fundus-correlated microperimetry (MP1 Nidek Technologies, Albignasego, Italy) with dilated pupils. The technique has been described previously.22,23 Additional information can be derived from the locus and the behavior of fixation and its stability. Of note, microperimetry is performed in the light-adapted state and is therefore used to test photopic but not scotopic central retinal function.
In this study, a standard red cross as fixation target (2° diameter, thickness of 1 pixel, approximately 2 minarc), a Goldmann size III stimulus with a 100-millisecond projection time, and a 4-2 staircase test strategy were used. A predefined macular test pattern covering the central 20° centered on the fovea included a minimum of 68 stimuli that were at least 4° apart. Additional stimuli were targeted according to localized visible changes to define extension of localized visual impairment more accurately and to detect smaller scotomata. During the study, the test pattern was slightly altered after the introduction of a standardized operation procedure for microperimetry assessment within the MacTel study.
After functional testing, both eyes of each patient underwent confocal scanning laser ophthalmoscopy (SLO) reflectance and autofluorescence imaging with 2 different devices. The maximum retinal irradiance of the lasers used was well below the limits established by the American National Standards Institute and other international standards.24 Using the Heidelberg Retina Angiograph (HRA 2; Heidelberg Engineering, Dossenheim, Germany), the following series of images with a 30° × 30° frame size and 768 × 768 pixel image resolution were acquired: (1) red -free reflectance (λ = 488 nm); (2) blue-light autofluorescence (excitation λ = 488 nm, emission > 500 nm); (3) near-infrared reflectance (λ = 830 nm), and (4) near-infrared autofluorescence (excitation λ = 795 nm, emission > 810 nm). For each series, 64 single images were aligned, the mean image was calculated, and the pixel values were normalized using the commercially available software of the HRA 2.
Using a modified confocal SLO device (Zeiss prototype SM, 30-4024; Carl Zeiss Meditec, Jena, Germany), the following series of images with a 30° × 23° frame size and 768 × 576 pixel image resolution were acquired: blue-light reflectance (excitation λ = 488 nm), blue-light autofluorescence (excitation λ = 488 nm, emission > 505 nm), green-light reflectance (λ = 532 nm), green-light autofluorescence (λ = 532 nm, emission > 550 nm), and red-light reflectance (λ = 635 nm). For each series, 64 single images were aligned and the mean image was calculated using customized image analysis software. The MPD maps were calculated by the logarithm ratio of the normalized aligned red- and blue-light reflectance images and displayed in pseudocolor maps (Matlab 7.0.1; The MathWorks, Inc, Natick, Massachusetts).
As part of the study screening visit, each patient had undergone color fundus photography, fluorescein angiography, and OCT examination (Stratus OCT3; Carl Zeiss Meditec, Dublin). These images were used for comparison with the results obtained from visual function testing and SLO imaging.
Statistical analyses included frequency and descriptive statistics. Photopic and scotopic FMM results were compared with the results of 4 age-matched control subjects. Threshold values were subtracted from the median value of control subjects. The Wilcoxon signed rank test was used to compare the distribution of photopic and scotopic age-matched corrected values.
Ten patients were enrolled in the study. The median age was 65.3 years (range, 43.7-74.4 years). The median visual acuity was 20/50 (range, 20/25-20/100). There were 3 study eyes with stage 2, 3 eyes with stage 3, 3 eyes with stage 4, and 1 eye with stage 5. Further clinical data and classification in clinical stages are provided in the Table.
In eyes with early manifestation (stage 2), photopic FMM and microperimetry showed no localized functional loss of more than 4 dB, whereas abnormalities were detected with scotopic FMM (Figure 1 and eFigure 1]). Owing to the normal photoreceptor distribution with absence of rods at the foveal center and the rapid increase in the number with eccentricity, scotopic FMM in healthy subjects shows a typical threshold peak at the fovea followed by a sharp decrease of thresholds toward the periphery and minimum threshold values at the outer points of the testing grid (eFigure 1A).17,25 In this study, all of the eyes with group 2A IJT, including early manifestations, displayed an extension and widening of increased scotopic thresholds from fixation toward the outer parts of the grid. In clinical stages 2 to 5, no obvious morphological abnormalities on fundus photography, fluorescence angiography, and OCT examination were spatially correlated with this impairment in scotopic function. However, all of the eyes, including those with early manifestation, showed alterations with SLO imaging in the central retina. Healthy subjects typically show markedly decreased blue-light autofluorescence intensities at the fovea due to macular pigment absorption. In all of the 10 eyes, the normal pattern of autofluorescence reduction was not observed. Instead, normal or increased blue-light autofluorescence intensities compared with the background signal were present, particularly in eyes with stages 3 and 4.26 These abnormalities corresponded to scotopic sensitivity loss. Using SLO reflectance images, MPD maps revealed changes of macular pigment distribution compared with healthy subjects. In contrast to the normal peak of MPD at the fovea, significant depletion of macular pigment was seen. In stages 2 to 5, the changes in macular pigment distribution correlated with the extension and widening of scotopic thresholds from the center toward the outer parts of the testing grid.
Very localized and marked scotomata in the temporal parafovea were seen in 2 eyes with stage 3 with a loss of up to 30 dB in photopic as well as scotopic FMM (Figure 1B). Consistent with all of the other eyes, testing points at the edge of the testing grid toward the outer retina were normal or only slightly reduced in microperimetry as well as photopic FMM. Overall, both produced similar results with regard to localization of retinal photopic dysfunction (eFigure 2). However, the comparison of absolute threshold values in decibels was only very limited owing to different setups of both systems.
In the 2 eyes at stage 3 with localized severe reduction in retinal sensitivity temporal to fixation, the scotomata was larger and the overall magnitude of reduction in sensitivity of testing points was higher for scotopic function compared with photopic function. No obvious changes were observed over these retinal areas on fundus photography. Fluorescein angiography showed good spatial correlation with photopic dysfunction by intense hyperfluorescent staining of localized areas in the late phase. By contrast, minimal staining at other retinal locations did not correspond with measurable photopic dysfunction. Examination by OCT showed large intraretinal cystoid spaces corresponding to areas with retinal dysfunction in 2 eyes. The cystoid spaces appeared to be located more in the inner retina in one and more in the outer retina in the other. Interestingly, very small cystoid spaces seen with OCT in other eyes (for example, in Figure 1A) did not correlate with measurable reduction of sensitivity. Blue-light autofluorescence showed marked alterations in these 2 eyes (eFigure 1). In addition to depletion of macular pigment and intense levels of increased blue-light autofluorescence in the fovea that corresponded to a reduction in scotopic sensitivity, the retinal areas temporal to fixation with localized severe scotomata and cystoid spaces in OCT examination were characterized by levels of decreased autofluorescence in both eyes and corresponded to the tips of right-angle venules.
Using photopic FMM and microperimetry in eyes with stages 4 and 5, localized absolute or severe scotomata with up to 30 dB of loss were observed in the nasal and temporal part of the parafovea, whereas testing of the foveal center revealed only a mild degree of photopic dysfunction (Figure 1C). The center of the severe scotomata coincided on fundus photography with hyperpigmentation in stage 4 and neovascularization, which were characterized by edema in stage 5. Fluorescein angiography showed hypofluorescence over pigmented plaques due to absorption phenomena as well as hyperfluorescence with late-phase leakage next to pigmentation and over areas of neovascularization. Localized hyperreflectivity in the inner retinal layer with masking of the underlying tissue was seen over pigmented plaques in OCT examination. However, as with earlier stages, the loss of function was not totally limited to pigment plaques seen with fundus photography, fluorescein angiography, and OCT examination. In other parafoveal areas, localized severe loss of photopic and scotopic retinal sensitivity with up to 32 dB of loss was measured as well. The functional impairment was spatially correlated with increased blue-light autofluorescence and depletion of macular pigment.
Comparing the distribution of photopic and scotopic age-matched corrected values, the number of test points with a loss of 10 dB or more was greater for scotopic testing (median, 51; interquartile range, 33 -65) than for photopic testing (median, 10; interquartile range, 0 -22) (P = .007) (Figure 2).
This study demonstrates that very localized and severe retinal dysfunction correlates with structural changes seen with fundus photography, fluorescein angiography, and OCT scans. This confirms previous speculations of functional impairment over areas with cystoid retinal spaces, retinal atrophy, and a variety of subretinal changes, including neovascularization.3 Furthermore, this observation is important to serve as a validation of the functional tests used in this study. Although sensitivity loss was measured where structural changes exist, central retinal dysfunction was not limited to these changes; losses extended beyond morphological alterations seen on standard imaging techniques.
The size and depth of scotopic scotomata were greater than those of photopic scotomata, suggesting that rod function is more severely affected than cone function. In early stages of the disease, scotopic dysfunction was clearly evident while minimal measurable photopic sensitivity loss was observed. Although no longitudinal data were included, it appears that rod involvement also occurs earlier in the disease process. This is in accordance with histological and clinical findings of preferential vulnerability of the rod system and relative stability of the cone system in other macular diseases.21,27- 32 Rod photoreceptors may serve as an early indicator of impending cone dysfunction and can act as a signal for intervention to maximize cone survival.
In healthy subjects, low-peak optical MPD has been shown to correlate with reduced central scotopic sensitivity.33 In addition, macular pigment has been indicated as a protective factor in the development of age-related macular degeneration.34,35 In our study, macular pigment depletion occurs in the absence of obvious vasculature changes and correlates with reduced scotopic sensitivity. This provides additional evidence that IJT is not solely a vascular disease and that early neuronal involvement takes place in the pathogenesis of the disease. The assumption that outer retinal changes induce vascular abnormalities would be in accordance with the pathogenesis of other retinal disease as has been demonstrated in the Royal College of Surgeons rats, rd and rds mice, or the vldlr −/− mouse model.36- 38
Our findings of marked alterations of autofluorescence intensities and macular pigment distribution using SLO imaging support the preliminary data from Helb et al.14 The distribution of luteal pigment can be assessed by autofluorescence imaging with SLO because the pigment absorbs short-wavelength light and appears as a dark area centered on the fovea.39 The apparent increased autofluorescence centrally in group 2A IJT is likely to be due largely to lack of luteal pigment rather than increased accumulation of lipofuscin in the retinal pigment epithelium. This is supported by 2 histopathological studies40,41 of patients with group 2A IJT that have described the retinal pigment epithelium as being intact. It is possible that photoreceptor loss may lead to macular pigment depletion and subsequent changes in autofluorescence signal given that the luteal pigments are found mostly in these cells.42
Reduction in photopic and scotopic sensitivity just next to fixation may have a profound effect on the actual visual function of the patient and his or her abilities to cope with daily visual tasks, particularly reading. Routine examinations with visual acuity testing might not be sufficient to detect reading difficulties (particularly at dim light illumination) and decreased night vision; hence, it might be worthy to include inquiries about these symptoms in clinical history taking. The functional tests applied in this study permitted accurate investigations of central retinal function but require dark adaption and special equipment. A more practical and widely available test to assess early photoreceptor involvement would be beneficial, particularly to test future therapeutic interventions.
In summary, this study showed localized retinal dysfunction despite good visual acuity in patients with group 2A IJT. The results suggest that rod function is involved earlier and more severely than cone function in the disease process. Dense scotomata next to fixation are not limited to visible angiographic alterations but may involve central retinal areas with SLO abnormalities. Depletion of macular pigment in early stages may reflect important metabolic changes in the early course of the disease. It indicates that the etiology of IJT is not just limited to the retinal vasculature but that neurons are intrinsically involved as well. Histopathological and clinical studies with longitudinal observations using high-resolution imaging and fine-detailed central visual function testing may help in further elucidating the pathogenesis and treatment options of this poorly understood visual impairment.
Correspondence: Fred W. Fitzke, PhD, Department of Visual Science, Institute of Ophthalmology, 11-43 Bath St, London EC1V 9EL, England (firstname.lastname@example.org).
Submitted for Publication: May 25, 2007;final revision received August 20, 2007; accepted August 25, 2007 .
Financial Disclosure: None reported.
Funding/Support: This study was supported by the MacTel Foundation, grant HPRN-CT-2002-00301 from the European Commission FP5, and the Foundation Fighting Blindness.
Previous Presentation: This paper was presented in part at the 2007 ARVO Annual Meeting; May 9, 2007; Fort Lauderdale, Florida.
Additional Contributions: Vy Luong provided technical support and Sally Falk coordinated the study.