A-D, Multimodal imaging of a 30-year-old man with multifocal choroiditis in the left eye. A fundus autofluorescence (FAF) image at presentation using the Optos system showed multiple hypoautofluorescent spots and a peripapillary zonal hyperautofluorescent area (A), colocalizing with an area of disruption of both the ellipsoid and retinal pigment epithelium–photoreceptor interdigitation zones in the corresponding spectral-domain optical coherence tomographic (SD-OCT) image (B). A FAF image 7 months later (C) showed resolution of the peripapillary zonal hyperautofluorescent area concomitant with near-complete restoration of both the ellipsoid and the digitation zones in the corresponding SD-OCT image (D). E-H, Multimodal imaging of a 50-year-old woman with multiple evanescent white dot syndrome. A 30° FAF image using the confocal scanning laser ophthalmoscope system showed multiple hyperautofluorescent spots (white arrow) (E), which corresponded to areas of focal disruption (arrow) of both the ellipsoid and retinal pigment epithelium–photoreceptor interdigitation zones on a horizontal SD-OCT scan (F). After bleaching, the FAF signal of the surrounding retinal areas increased more than the FAF signal of the spots, resulting in a markedly decreased difference in autofluorescence level between the pathological spots and the relatively normal-appearing surrounding retinal tissue (G). However, on the corresponding SD-OCT images before (F) and after (H) bleaching, the retinal structure looked identical. One hour later, the hyperautofluorescent spots reappeared just as in E (image not shown). The green arrows in A, C, E, and G indicate the levels of the SD-OCT scans in B, D, F, and H, respectively.
Multimodal imaging of a 35-year-old woman with central serous chorioretinopathy in the left eye, at presentation (A-D) and after complete resolution of the subretinal fluid (E-H). Transit phase of fluorescein angiography (A) and midphase of indocyanine green angiography (B) both showed the focal leak point with typical inkblot diffusion. C, The spectral-domain optical coherence tomographic scan at the level of the arrow in A showed subretinal fluid. D, The spectral-domain optical coherence tomographic thickness map showed an area of retinal thickening (red) around the focal leak point corresponding to the subretinal fluid. E, After resolution of the subretinal fluid, the 55° fundus autofluorescence image showed an area of increased fundus autofluorescence superior to the optic disc, corresponding to the area where the subretinal fluid was previously detected. The bleaching experiment was performed twice: first temporal to the hyperautofluorescent area (E) and then including most of the abnormal hyperautofluorescent area (F). The second time, the lesion became much less distinguishable from the bleached background, presumably due to a lack of bleaching in this retinal area that shows a disruption of both the ellipsoid and retinal pigment epithelium–photoreceptor interdigitation zones on the corresponding spectral-domain optical coherence tomographic image (arrows) (G) (taken at the level of the arrow in E). The same area corresponds to retinal thinning (blue) in the corresponding retinal thickness map (H).
Multimodal imaging of the left eye of a 50-year-old man with resolved central serous chorioretinopathy. The infrared reflectance image (same in A-C), the fundus autofluorescence (FAF) image (same in D-F), and the spectral-domain optical coherence tomographic (OCT) transverse sections (G-I) and B scans (J-L) were acquired with the confocal scanning laser ophthalmoscope (Heidelberg Retina Angiograph; Heidelberg Engineering). The segmentation for the transverse OCT was between the superior border of the ellipsoid zone and the superior border of the retinal pigment epithelial band. The level of segmentation is shown by the red lines in the OCT B scans in M-O. E and H, Note that there were 2 autofluorescent patterns within the previously detached retinal area (dotted lines). A-I, First, there are focal granular hyperautofluorescent spots (arrows) appearing bright in the infrared image, hyperautofluorescent in the FAF image, and bright in the transverse OCT image. J-O, These granular hyperautofluorescent spots correspond to hyperreflective deposits at the level of the retinal pigment epithelium in the corresponding OCT B scans (white lines). Second, there is a diffuse hyperautofluorescent area in the FAF image (E, dotted line) colocalizing precisely with a hyporeflective area in the transverse OCT image (H, dotted line). This demonstrates that the diffuse hyperautofluorescence in the FAF image colocalizes precisely with the disruption of both the ellipsoid and interdigitation zones in the corresponding OCT B scans.
Freund KB, Mrejen S, Jung J, Yannuzzi LA, Boon CJF. Increased Fundus Autofluorescence Related to Outer Retinal Disruption. JAMA Ophthalmol. 2013;131(12):1645-1649. doi:10.1001/jamaophthalmol.2013.5030
Fundus autofluorescence (FAF) imaging is able to map metabolic changes at the level of the retinal pigment epithelium (RPE) noninvasively in vivo. However, the observed autofluorescence signal is a summation of not only the autofluorescence originating from the RPE but also that from more anterior ocular structures including the overlying neuroretina.1 Retinal photopigments within the photoreceptor outer segments have absorption properties, and illumination with blue light decreases the optical pigment density by photobleaching.2 After blue light irradiation, there is photoisomerization of the opsin proteins from the 11-cis to all-trans conformation in the photoreceptor outer segments. This photoisomerization to the all-trans configuration causes a decrease in the optical density of the photopigment in the outer segments of the photoreceptors, resulting in a temporary loss of light absorption properties. Theelen et al3 have shown that in healthy and diseased eyes, the illumination of a retinal area with blue light produces a relative hyperautofluorescence as compared with the surrounding nonilluminated area. The mechanism involved is a window defect due to the relative loss of photopigment density in the outer segments of the photoreceptors after bleaching of these photopigments.
When the outer retinal structure is damaged, the photopigment density will be reduced. There is a quantitative deficiency in visual pigment in both rods4 and cones5 in areas of both active and resolved serous retinal detachments in central serous chorioretinopathy, with rhodopsin levels reduced to 20% to 30% of normal.4 We investigated whether outer retinal structural damage with photopigment loss in various retinal diseases may also lead to an increased autofluorescence signal through a window defect that increases the excitation of and unmasks the autofluorescence emitted from the underlying relatively preserved RPE.
Patients with various retinal diseases with focal outer retinal structural loss, an underlying RPE that appeared intact as detected by focal disruption of the ellipsoid zone, and an intact RPE band on corresponding spectral-domain optical coherence tomographic (SD-OCT) images were evaluated with FAF imaging. The FAF imaging was performed with the confocal scanning laser ophthalmoscope (Heidelberg Engineering). One patient was imaged with the Ultra-Widefield (200°) Retinal Imaging System (Optos). Two patients were evaluated before and after bleaching. For the bleaching condition, the patients were dark adapted for 20 minutes. Following this, a 55° confocal short (blue) wavelength–autofluorescent image was acquired. A retinal area of 30° × 30° adjacent to the hyperautofluorescent region of interest was then illuminated with the excitation light source for 60 seconds to achieve light adaptation of this area. Subsequently, another 55° image was acquired. Transverse SD-OCT images segmented between the superior border of the ellipsoid zone and the superior border of the RPE band were acquired and correlated with the FAF images.
Six patients were included: 2 patients with multiple evanescent white dot syndrome, 1 with multifocal choroiditis, and 3 with resolved central serous chorioretinopathy. The retinal areas with a disruption of both the ellipsoid and interdigitation zones and intact RPE band on SD-OCT imaging corresponded to areas of hyperautofluorescence. After bleaching, the qualitative analysis showed that the FAF signal of the background increased more than the FAF signal of the outer retinal diseased areas, therefore decreasing the contrast and making the originally hyperautofluorescent lesions almost disappear (Figure 1 and Figure 2). In the patient with multifocal choroiditis, a zonal peripapillary hyperautofluorescent area corresponded precisely to a disruption of the ellipsoid zone in the corresponding SD-OCT image (Figure 1). Over time, the resolution of this zonal hyperautofluorescent retinal area was concomitant with the restoration of the ellipsoid zone in the corresponding SD-OCT image (Figure 1). The diffuse hyperautofluorescent areas in the FAF images colocalized precisely with a hyporeflective area in the transverse SD-OCT images, demonstrating that the diffuse hyperautofluorescence in the FAF image colocalizes with the disruption of both the ellipsoid and interdigitation zones in the corresponding SD-OCT B scans (Figure 3).
Outer retinal disruption may result in increased autofluorescence due to a window defect as a result of photopigment loss. This mechanism could help explain hyperautofluorescence in a variety of settings including inflammatory entities such as multiple evanescent white dot syndrome and multifocal choroiditis, the active margin of acute zonal occult outer retinopathy, both acute and resolved central serous chorioretinopathy, rhegmatogenous retinal detachment after repair, the hyperautofluorescent ring in retinitis pigmentosa6 and autoimmune retinopathy,7 and at the edge of geographic atrophy where there is loss of photoreceptors over an intact RPE. Recognizing that increased FAF can occur in the absence of increased fundus fluorophores may help clinicians detect and more precisely monitor early photoreceptor damage. In some patients, these findings may provide an anatomical correlate of persistent suboptimal visual function despite relatively normal retina-RPE structure.
There may be other causes of hyperautofluorescent signals in these eyes with various diseases and complex pathophysiological mechanisms, but we believe that optical pigment density reduction as a result of outer retinal disruption should be considered as a possible explanation for the presence of a hyperautofluorescent signal in cases of retinal diseases that tend to affect the photoreceptors prior to RPE involvement.
Corresponding Author: K. Bailey Freund, MD, Vitreous Retina Macula Consultants of New York, 460 Park Ave, Fifth Floor, New York, NY 10022 (firstname.lastname@example.org).
Published Online: October 17, 2013. doi:10.1001/jamaophthalmol.2013.5030.
Author Contributions: Dr Freund had full access to all of 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: Freund, Mrejen, Yannuzzi.
Acquisition of data: Freund, Mrejen, Jung, Yannuzzi.
Analysis and interpretation of data: Freund, Mrejen, Jung, Boon.
Drafting of the manuscript: Freund, Mrejen, Yannuzzi, Boon.
Critical revision of the manuscript for important intellectual content: Freund, Mrejen, Jung, Boon.
Obtained funding: Yannuzzi.
Administrative, technical, ormaterial support: Jung.
Study supervision: Freund, Mrejen, Boon.
Conflict of Interest Disclosures: None reported.
Funding/Support: This work was supported by The Macula Foundation, Inc.
Role of the Sponsor: The funding organization had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; or preparation, review, or approval of the manuscript.