Graefe's Archive
Graefe's Arch Clin Exp Ophthalmol (1992)230:11 19
for Clinical and Expedmenlal
Ophthalmology © Springer-Verlag 1992
Classification of retinal pigment epithelial detachments associated with drusen M. Elizabeth Hartnett 1' 3 , , John J. Weiter 1' 2, 3, Armando Garsd 3, and Alex E. Jalkh 1' 2, 3 1 Eye Research Institute, 2 Retina Associates, and 3 Harvard Medical School, Boston, Mass., USA Received July 9, 1990 / Accepted April 15, 1991
Abstract. To u n d e r s t a n d the natural history o f retinal pigment epithelial ( R P E ) detachments (PEDs) associated with drusen, we retrospectively reviewed 125 P E D s that h a d not u n d e r g o n e laser treatment over an average follow-up period o f 28 m o n t h s . We identified six types o f R P E d e t a c h m e n t s : pseudo-vitelliform, confluent drusen, serous, vascular, and h e m o r r h a g i c lesions as well as P E D s associated with a retinal vascular abnormality. We correlated the type o f d e t a c h m e n t with visual acuity and a n a t o m i c outcome. The p o o r e s t results were obtained for P E D s associated with neovascularization (vascular, hemorrhagic, and retinal vascular a b n o r m a l i t y types). Eyes presenting with pseudo-vitelliform lesions or confluent drusen h a d the best visual prognosis and a n a t o m i c results. Obvious subretinal new vessels developed in 16 (34%) o f the serous P E D s over an average follow-up period o f 25 m o n t h s . Subretinal neovascularization was rare in the pseudo-vitelliform a n d confluent drusen groups. The p r o p o s e d classification system m a y be useful in predicting the natural course o f PEDs.
Introduction M a c u l a r degeneration is the leading cause o f central visual loss in patients aged > 6 0 years [10, 12]. The hallm a r k o f m a c u l a r degeneration in the elderly is drusen, a l t h o u g h their presence alone is n o t usually the cause o f visual disturbance. However, severe central visual loss is associated with subretinal new vessels, pigment epithelial d e t a c h m e n t (PED), or large areas o f retinal atrophy. Treatment o f subretinal new vessels located outside the foveal avascular zone with argon-laser p h o t o c o a g u lation has been effective [13, 23]. The efficacy o f krypton-laser treatment o f subfoveal and juxtafoveal subretinal neovascularization is being evaluated [14]. However, the m a n a g e m e n t o f P E D s associated with drusen and * Present address and address f o r offprint requests: M.E. Hartnett, Erie County Medical Center, 462 Grider Street, Buffalo, NY 14215, USA
m a c u l a r degeneration remains controversial [16, 21] in part because the clinical appearance o f P E D s varies [4-6, 15, 18] and because it is frequently difficult to determine whether subretinal neovascularization has occurred. We developed a classification system that appears to be helpful in distinguishing P E D s with relatively g o o d prognoses f r o m those that are m o r e likely to develop neovascularization and result in p o o r visual function.
Patients and methods The charts of 156 patients presenting with macular PEDs were reviewed retrospectively. The subjects had been referred to Retina Associates, a private Boston-based office practice, over a 20-year period (1966-1985). Of the 247 PEDs, 130 were treated with photocoagulation and then excluded from the study, except for 8 displaying retinal vascular abnormalities that were included for classification purposes but not for statistical analysis. Another 14 untreated PEDs that were included in the classification were also excluded from statistical analysis, as they were not followed after the initial evaluations. Thus, a total of 125 PEDs were entered into our classification and 103 were analyzed statistically. All patients had been routinely examined every 2-12 months as determined by the requirements of acute and long-term followup. The final examination represented the last time a patient was seen before data collection. All subjects displayed evidence of drusen and a PED measuring _>1 disc diameter in one or both eyes. At each visit, best-corrected bilateral visual acuities were recorded in decimal notation (e.g., 20/100 = 0.2, 20/20 = 1). Fundus photographs and fluorescein angiograms were examined stereoscopically. Subretinal new vessels, retinal pigment epithelium (RPE) tears, large subretinal hemorrhages, and vitreous hemorrhages were recorded. The final visual acuity was determined. The anatomic outcomes included retinal atrophy, stable PED appearance, and neovascularization beneath the RPE (hereafter referred to as subretinal neovascularization) followed by fibrovascular scar formation. We divided the PEDs into six categories according to their appearance on funduscopy and fluorescein angiography (Table 1): 1. Pseudo-vitelliform lesions [8] (Fig. 1) resembled the yellow "eggyolk" form of Best's disease. On fluorescein angiography, the yellow subretinal material blocked early fluorescence and stained in the late phase of dye transit. 2. In cases of confluent drusen (Fig. 2), numerous confluent exudative drusen appeared in the fovea, were often associated with pigment mottling, and stained late on angiography.
12
13 toreceptors. A vascular PED was defined as hemorrhagic when the hemorrhage occupied at least two-thirds of the lesion. 6. Retinal vascular abnormalities (Fig. 6) are newly recognized PEDs that are associated with an outer retinal angiomatous lesion displaying either a feeding retinal arteriole or an emptying venule. The angiomatous lesion fills with the retinal vasculature and occurs deep in the outer retina. This retinal vascular abnormality may ultimately lead to a chorioretinal anastomosis. On angiography, these lesions appeared in the center of the PED; they fluoresced in the early angiogram and leaked in the late phase of dye transit. As none of these eyes had undergone laser treatment before diagnosis, the lesions did not represent post-laser chorioretinal anastomoses. The clinical diagnoses of the fellow eyes were noted. If the second eye developed a PED during the study period, it was entered separately into the data.
Statistical methods PED-induced changes in visual acuity were fitted to a generalized linear regression model using interactions [17]. To minimize age confounding, this analysis was restricted to patients whose final visit occurred within 1 year of PED diagnosis (ti~, ~
Graefe's Arch Clin Exp Ophthalmol (1992)230:11 19
for Clinical and Expedmenlal
Ophthalmology © Springer-Verlag 1992
Classification of retinal pigment epithelial detachments associated with drusen M. Elizabeth Hartnett 1' 3 , , John J. Weiter 1' 2, 3, Armando Garsd 3, and Alex E. Jalkh 1' 2, 3 1 Eye Research Institute, 2 Retina Associates, and 3 Harvard Medical School, Boston, Mass., USA Received July 9, 1990 / Accepted April 15, 1991
Abstract. To u n d e r s t a n d the natural history o f retinal pigment epithelial ( R P E ) detachments (PEDs) associated with drusen, we retrospectively reviewed 125 P E D s that h a d not u n d e r g o n e laser treatment over an average follow-up period o f 28 m o n t h s . We identified six types o f R P E d e t a c h m e n t s : pseudo-vitelliform, confluent drusen, serous, vascular, and h e m o r r h a g i c lesions as well as P E D s associated with a retinal vascular abnormality. We correlated the type o f d e t a c h m e n t with visual acuity and a n a t o m i c outcome. The p o o r e s t results were obtained for P E D s associated with neovascularization (vascular, hemorrhagic, and retinal vascular a b n o r m a l i t y types). Eyes presenting with pseudo-vitelliform lesions or confluent drusen h a d the best visual prognosis and a n a t o m i c results. Obvious subretinal new vessels developed in 16 (34%) o f the serous P E D s over an average follow-up period o f 25 m o n t h s . Subretinal neovascularization was rare in the pseudo-vitelliform a n d confluent drusen groups. The p r o p o s e d classification system m a y be useful in predicting the natural course o f PEDs.
Introduction M a c u l a r degeneration is the leading cause o f central visual loss in patients aged > 6 0 years [10, 12]. The hallm a r k o f m a c u l a r degeneration in the elderly is drusen, a l t h o u g h their presence alone is n o t usually the cause o f visual disturbance. However, severe central visual loss is associated with subretinal new vessels, pigment epithelial d e t a c h m e n t (PED), or large areas o f retinal atrophy. Treatment o f subretinal new vessels located outside the foveal avascular zone with argon-laser p h o t o c o a g u lation has been effective [13, 23]. The efficacy o f krypton-laser treatment o f subfoveal and juxtafoveal subretinal neovascularization is being evaluated [14]. However, the m a n a g e m e n t o f P E D s associated with drusen and * Present address and address f o r offprint requests: M.E. Hartnett, Erie County Medical Center, 462 Grider Street, Buffalo, NY 14215, USA
m a c u l a r degeneration remains controversial [16, 21] in part because the clinical appearance o f P E D s varies [4-6, 15, 18] and because it is frequently difficult to determine whether subretinal neovascularization has occurred. We developed a classification system that appears to be helpful in distinguishing P E D s with relatively g o o d prognoses f r o m those that are m o r e likely to develop neovascularization and result in p o o r visual function.
Patients and methods The charts of 156 patients presenting with macular PEDs were reviewed retrospectively. The subjects had been referred to Retina Associates, a private Boston-based office practice, over a 20-year period (1966-1985). Of the 247 PEDs, 130 were treated with photocoagulation and then excluded from the study, except for 8 displaying retinal vascular abnormalities that were included for classification purposes but not for statistical analysis. Another 14 untreated PEDs that were included in the classification were also excluded from statistical analysis, as they were not followed after the initial evaluations. Thus, a total of 125 PEDs were entered into our classification and 103 were analyzed statistically. All patients had been routinely examined every 2-12 months as determined by the requirements of acute and long-term followup. The final examination represented the last time a patient was seen before data collection. All subjects displayed evidence of drusen and a PED measuring _>1 disc diameter in one or both eyes. At each visit, best-corrected bilateral visual acuities were recorded in decimal notation (e.g., 20/100 = 0.2, 20/20 = 1). Fundus photographs and fluorescein angiograms were examined stereoscopically. Subretinal new vessels, retinal pigment epithelium (RPE) tears, large subretinal hemorrhages, and vitreous hemorrhages were recorded. The final visual acuity was determined. The anatomic outcomes included retinal atrophy, stable PED appearance, and neovascularization beneath the RPE (hereafter referred to as subretinal neovascularization) followed by fibrovascular scar formation. We divided the PEDs into six categories according to their appearance on funduscopy and fluorescein angiography (Table 1): 1. Pseudo-vitelliform lesions [8] (Fig. 1) resembled the yellow "eggyolk" form of Best's disease. On fluorescein angiography, the yellow subretinal material blocked early fluorescence and stained in the late phase of dye transit. 2. In cases of confluent drusen (Fig. 2), numerous confluent exudative drusen appeared in the fovea, were often associated with pigment mottling, and stained late on angiography.
12
13 toreceptors. A vascular PED was defined as hemorrhagic when the hemorrhage occupied at least two-thirds of the lesion. 6. Retinal vascular abnormalities (Fig. 6) are newly recognized PEDs that are associated with an outer retinal angiomatous lesion displaying either a feeding retinal arteriole or an emptying venule. The angiomatous lesion fills with the retinal vasculature and occurs deep in the outer retina. This retinal vascular abnormality may ultimately lead to a chorioretinal anastomosis. On angiography, these lesions appeared in the center of the PED; they fluoresced in the early angiogram and leaked in the late phase of dye transit. As none of these eyes had undergone laser treatment before diagnosis, the lesions did not represent post-laser chorioretinal anastomoses. The clinical diagnoses of the fellow eyes were noted. If the second eye developed a PED during the study period, it was entered separately into the data.
Statistical methods PED-induced changes in visual acuity were fitted to a generalized linear regression model using interactions [17]. To minimize age confounding, this analysis was restricted to patients whose final visit occurred within 1 year of PED diagnosis (ti~, ~
