- Email: [email protected]
The prognosis of idiopathic choroidal neovascularization in persons younger than 50 years of age
The Prognosis of Idiopathic Choroidal Neovascularization in Persons Younger Than 50 Years of Age Bertil Lindblom, MD, PhD, Tommy Andersson, MD Objective: To examine the prognosis of idiopathic choroidal neovascular membranes in patients younger than 50 years of age at the time of diagnosis. Design: Retrospective noncomparative case series. Participants: Eighteen consecutive patients from a university clinic (20 eyes) with idiopathic choroidal neovascularization in the macular area. All patients diagnosed between 1979 and 1996 were included; follow-up time varied from 5 to 187 months (mean, 55 months). Intervention: Treatment was conservative. All patients were informed about argon laser photocoagulation, but in the majority of cases, patients were not encouraged to undergo such treatment. Treatment was given on patient’s request. Main Outcome Measures: The dimensions of the choroidal neovascular membrane and its distance to the foveola were determined by computerized image analysis and their correlation with final visual acuity determined. Results: Fifteen (75%) of the 20 eyes retained a visual acuity of 20/60 or better. In ten eyes (50%), acuity was 20/25 or better. Six eyes had been treated with laser photocoagulation; only one had a membrane located 200 m or more from the foveola at the time of treatment. The outcome for treated eyes was no better than for untreated eyes, although the two groups were not directly comparable. In no eye did an initially juxtafoveal or extrafoveal membrane continue to grow under the foveola. Conclusions: The prognosis of idiopathic choroidal neovascularization was much better than reported previously for eyes with choroidal membranes associated with age-related macular degeneration. A conservative treatment regimen should be considered. The authors have no reason to believe that laser photocoagulation would have improved the outcome in the nontreated eyes. Ophthalmology 1998;105:1816 –1820 Idiopathic or juvenile choroidal neovascularization is an uncommon disorder affecting primarily individuals younger than the age of 50 years. As the term implies, the condition is not accompanied by other ocular pathology. This is in contrast to choroidal neovascularization in eyes with age-related macular degeneration, high myopia, presumed ocular histoplasmosis syndrome, angioid streaks, or other retinal disorders. The natural history of idiopathic subfoveal choroidal neovascularization in subjects younger than the age of 50 years recently was described.1 The prognosis was favorable compared to that of choroidal neovascularization in eyes with age-related macular degeneration, with several eyes retaining useful vision after many years of follow-up. Similar results were reported in an earlier study.2 The natural history of juxtafoveal or extrafoveal choroidal neovascularization has not been described in de-
Originally received: May 27, 1997. Revision accepted: April 14, 1998. Manuscript no. 97301. From the Institute of Clinical Neuroscience, Department of Ophthalmology, Go¨teborg University, Go¨teborg, Sweden. Reprint requests to Bertil Lindblom, MD, PhD, Institute of Clinical Neuroscience, Department of Ophthalmology, Sahlgrenska University Hospital, S-41345 Go¨teborg, Sweden.
1816
tail, although some information can be extracted from earlier work.3,4 (In this context, a juxtafoveal membrane has its most central border 1–200 m from the foveola while an extrafoveal membrane is 201–2500 m from the foveola.) One reason for the scarcity of reports is the widely used treatment regimen with argon or krypton laser photocoagulation. The treatment of extrafoveal choroidal neovascularization of idiopathic origin has been the subject of several studies, including that of the Macular Photocoagulation Study Group.5,6 In the first study from this group,5 laser photocoagulation was recommended for extrafoveal membranes. This recommendation was renewed in a later study of krypton laser photocoagulation for idiopathic choroidal neovascularization despite the lack of a statistically proven treatment effect.6 The two reports from the Macular Photocoagulation Study Group did not consider the patient’s age. Approximately half of the subjects were older than 50 years. It has been our clinical impression that age is important for the prognosis of idiopathic choroidal neovascularization and that the outcome is more favorable in younger subjects. Therefore, we have for many years been reluctant to treat extrafoveal membranes presenting in patients younger than the age of 50 years. This study was initiated
Lindblom and Andersson 䡠 Idiopathic Neovascularization in an attempt to verify (or disprove) this clinical impression.
Subjects and Methods Subjects By scrutinizing our fundus photograph files, we identified all patients with nontraumatic subretinal neovascularization in the macular area who were younger than 50 years of age at the first examination. The exclusion criteria were: myopia exceeding 6 diopters, signs of other retinal disease such as presumed ocular histoplasmosis or inner punctate choroidopathy, a history of ocular contusion severe enough to cause intraocular bleeding, and signs of age-related maculopathy such as drusen or pigment epithelial detachment.
Table 1. The Underlying Diagnosis in 48 Subjects ⬍50 Years of Age with Nontraumatic Choroidal Neovascularization in the Macular Area Diagnosis
No.
Bilateral
Idiopathic Myopia ⬎ 6 D Punctate inner choroidopathy Presumed ocular histoplasmosis Retinochoroiditis Angioid streaks (pseudoxanthoma elasticum) Toxoplasmosis Best’s vitelliform choroidopathy Diffuse pigment epitheliopathy Heredodegenerative disease Acute posterior multifocal placoid pigment epitheliopathy
19 9 4 4 3
2
3 2 1 1 1
1
Total
48
2
1
1
Methods Fluorescein angiograms and fundus photographs were retrieved from the files of each patient together with relevant clinical data. Data were collected from the following four examinations: 1. The first visit to an ophthalmologist after the presentation of symptoms caused by subretinal neovascularization. 2. The visit at which the patient showed the poorest visual acuity after the diagnosis of subretinal neovascularization. 3. The visit at which the patient showed the best visual acuity after the diagnosis of subretinal neovascularization. 4. The last follow-up visit. In many cases, two of the above-mentioned events coincided. Visual acuity was tested under standardized conditions using an illuminated acuity chart (LIC, Solna, Sweden). This chart has a decimal acuity scale, but measurements were converted to a Snellen fraction except for the computing of mean visual acuity, in which case values were converted to the logarithm of the minimum angle of resolution (logMAR). The luminance of the acuity chart was approximately 500 cd/m2. Test distance was 5 m and best correction was used. Fundus photographs were obtained with a Nikon Retinapan II (Nikon Corp, Tokyo, Japan) or a Canon 60U (Canon Inc, Tokyo, Japan) fundus camera. Fluorescein angiography was performed with a fundus camera in the earlier part of the
study period but was later replaced by a scanning laser ophthalmoscope model 101 (Rodenstock, Ottobrunn, Germany). Fundus photographs and fluorescein angiograms were digitized by a 35-mm film scanner (Nikon Coolscan LS-10; Nikon Corp, Tokyo, Japan) and stored in computer memory together with individual video frames from the scanning laser ophthalmoscope. Images were analyzed with an image analysis program (Optimas Corp, Bothell, WA) using a personal computer. Each image was spatially calibrated. The distance between the center of the optic disc and the foveola was set to 15.6°.7 Distances were then converted to a linear scale by the conversion factor 1 arcmin ⫽ 4.85 m.8 The subretinal neovascular membrane was outlined on the fluorescein angiogram (Fig 1). In eyes with a well-demarcated membrane, the outline was made automatically by the software. In other cases, the automatic procedure appeared influenced by confounding factors such as hemorrhages or pigment epithelial changes surrounding the membrane. In those cases, the membrane was outlined manually. The exact position of the foveola was marked. When needed, this was aided by analysis of the color images. The following measurements were obtained for the choroidal neovascular membrane: area, perimeter, longest axis, width (perpendicular to the longest axis), and distance between the foveola and the nearest membrane edge (this was set to zero if the membrane was subfoveal). In treated eyes in which the choroidal membrane had disappeared, the same measurements were applied to the treatment scar.
Results
Figure 1. Example of measurement of a choroidal neovascular membrane. The membrane is outlined by a white line. A cross marks the foveola, and another cross marks the center of the optic disc.
Forty-seven patients younger than 50 years of age with nontraumatic choroidal neovascularization were identified. The underlying diagnoses are listed in Table 1. Nineteen patients met the inclusion criteria for idiopathic choroidal neovascularization, 2 of whom had bilateral disease. One patient was excluded from further analyses because of insufficient data. He had been treated with argon laser photocoagulation for a neovascular membrane while traveling abroad. Pretreatment photographs were not available. At followup, 10 years after treatment, he had a subfoveal membrane and visual acuity was 20/300. The most relevant measurements for the remaining 18 patients (20 eyes) are listed in Table 2. Membrane length and width did not aid the interpretation of the results and were omitted. Five patients were examined with scanning laser ophthalmoscopy throughout
1817
Ophthalmology Volume 105, Number 10, October 1998 Table 2. Summary of the Result in 18 Subjects with Idiopathic Subretinal Neovascularization Patient No.
Laser Treatment*
1 2 3R 3L 4 5 5REACT 6 7 8 9 10 11 12 13 14 15 16 17 18R 18L Mean SD
Yes No No No Yes No No No No Yes No Yes No Yes No No Yes No No No No
Age at Diagnosis (yrs)
Area at Diagnosis (mm2)
Area Change (%)
Perimeter at Diagnosis (mm)
Perimeter Change (%)
Distance at Diagnosis (mm)
Final Distance (mm)
38 34 20 23 42 35 36 39 39 47 44 50 46 41 38 31 47 45 47 46 46 40 8.0
0.38 0.52 0.29 0.60 0.97 0.26 0.96 0.33 0.61 1.48 0.47 1.42 0.31 2.65 0.41 0.14 0.48 0.23 0.38 0.92 0.61 0.69 0.58
⫺11 ⫺37 331 401 11 ⫺36 ⫺30 ⫺5 ⫺51 ⫺30 ⫺31 ⫺53 ⫺56 ⫺41 ⫺16 ⫺40 18 ⫺10 108 23 77 6.8
2.45 3.28 2.46 3.21 4.17 2.28 4.37 2.60 3.18 4.95 2.69 5.28 2.45 7.18 2.64 1.61 3.06 2.05 2.59 4.03 3.55 3.34 1.30
3 ⫺25 82 121 4 ⫺26 ⫺24 ⫺5 ⫺26 ⫺18 ⫺11 ⫺36 ⫺26 ⫺33 ⫺8 ⫺23 9 ⫺3 47 10 24 ⫺1.73
0.04 0.58 0.03 0.00 0.05 0.17 0.02 0.70 0.12 0.10 0.06 0.00 1.06 0.93 0.27 0.14 1.44 0.47 0.58 0.08 0.40 0.35 0.40
0.06 0.62 0.00 0.00 0.00 0.21 0.11 0.60 0.20 0.15 0.14 0.24 1.32 1.40 0.18 0.17 1.69 0.59 0.44 0.10 0.24 0.40 0.49
Acuity at Diagnosis
Final Acuity
Follow-up (mos)
20/200 20/200 20/30 20/20 20/30 20/40 20/200 20/25 20/20 20/50 20/25 20/200 20/20 20/30 20/125 20/100 20/15 20/25 20/25 20/100 20/30
20/200 20/20 20/50 20/100 HM 20/25 20/30 20/20 20/25 20/50 20/15 20/50 20/20 20/15 20/50 20/15 20/15 20/20 20/20 20/30 20/70
175 187 145 110 20 7 100 63 6 72 71 11 28 10 5 60 32 11 10 14 11 55 58
Area and perimeter ⫽ the size of the subretinal membrane; Distance ⫽ distance from foveola to the nearest border of the subretinal membrane; R ⫽ right; L ⫽ left; REACT ⫽ reactivation; SD ⫽ standard deviation; HM ⫽ hand movements. * If laser treatment has been applied, all measurements refer to the treatment scar.
the follow-up period. Three eyes with involuted choroidal membranes were first examined with fluorescein angiography using a fundus camera and several years later with scanning laser ophthalmoscopy. The results were very similar, suggesting that there was no systematic difference between the methods. Of the 20 eyes, 15 (75%) had a visual acuity of 20/60 or better at the last visit. In ten eyes (50%), visual acuity was 20/25 or better (Fig 2). Two eyes lost six or more lines of Snellen visual acuity: one untreated eye deteriorated from 20/20 to 20/100 and one treated eye from 20/33 to hand movements. The three eyes with a final visual acuity of 20/100 or worse all had a choroidal neovascular membrane less than 200 m from the foveola at first visit. One eye (patient 5) showed a reactivation of the choroidal membrane during the study period. The membrane initially was located 166 m from the foveola and treatment was considered. However, it retracted spontaneously and visual acuity improved to 20/25. Seven months after first diagnosis, reactivation of the membrane reduced visual acuity to 20/200. Six months later, the choroidal membrane was considered inactive and follow-up examinations were discontinued. After another 6 months, the patient noticed successive visual improvement and 8 years later, acuity was 20/30. This was the only case of reactivation in this series of eyes. Figure 3 shows the relation between mean visual acuity and the membrane’s distance to the foveola at the last follow-up examination. Six eyes had undergone argon laser photocoagulation. The central border of the subretinal membrane was less than 200 m from the foveola in four eyes; two had a subfoveal membrane. One of these eyes had a tear of the retinal pigment epithelium develop after treatment and ended up with hand movement vision (patient 4). Two eyes were treated before 1982, and the treatment technique did not meet modern standards. In the other four eyes, treatment was recommended by the ophthalmologist, in two cases
1818
because of growth of the choroidal membrane. Green-only argon laser was used (spot size, 0.1– 0.2 mm; duration, 0.2– 0.3 second). The goal was a distinct white lesion. In general, the size of the
Figure 2. Final Snellen visual acuity plotted against acuity at diagnosis. Data points to the left and below the diagonal line represent improvement. Solid circles represent laser-treated eyes, and open circles represent untreated eyes.
Lindblom and Andersson 䡠 Idiopathic Neovascularization
Figure 3. Mean visual acuity (expressed in logarithm of the minimum angle of resolution) at the last visit in four groups of eyes with idiopathic choroidal neovascularization. The eyes were grouped according to the distance between the choroidal membrane and foveola at the last examination.
Figure 5. Patient 9. Idiopathic choroidal neovascularization in a man who was 44 years old at diagnosis. The examination results after 71 months of follow-up are shown. No treatment had been given. Visual acuity was 20/15.
membrane in the treated eyes was larger than that in the untreated eye (1.23 and 0.44 mm2, respectively; P ⫽ 0.004). Final visual acuity in treated eyes ranged from hand movements to 20/15, but the material was too small to allow any comparisons with the untreated group (Fig 2). In no eye did a subretinal membrane located outside the 200 m radius at the initial examination continue to grow under the fovea. In the only eye in which a membrane extended in a central direction across the 200-m border, visual acuity improved from 20/125 to 20/40 (patient 13). Late pigment epithelial atrophy engaged the foveola in one case (patient 2), but visual acuity still was 20/20 after 15 years of follow-up. In one case (patient 8), widespread pigment epithelial changes developed more than 1 year after laser treatment. We interpreted these changes as “atrophic creep” (Fig 4). Final visual acuity was 20/50. In contrast, a very similar lesion (patient 7) was left untreated and the eye retained a visual acuity of 20/15 (Fig 5). Follow-up time varied from 5 to 187 months (mean, 55 months). In two cases, it was 6 months or less. These two
patients came from another take-up area, and follow-up visual acuity data were requested from their home clinics. Thus, for patient 7, acuity had improved to 20/20 after 68 months of follow-up. For patient 13, visual acuity was unchanged (20/50) after 16 months of follow-up.
Figure 4. Patient 8. Idiopathic choroidal neovascularization in a man who was 47 years old at diagnosis. The macula at the last examination after 72 months of follow-up is shown. Argon laser photocoagulation had been instituted, and widespread retinal atrophy (so-called creeping atrophy) was seen. Visual acuity was 20/50.
Discussion Idiopathic choroidal neovascularization is uncommon in all age groups. The diagnosis is one of exclusion in that a recognizable cause for subretinal neovascularization must be lacking. One reason for studying the younger-than-50 age group separately is that it is difficult to exclude early forms of age-related macular degeneration in older subjects. Indeed, a substantial number of patients enrolled in a study of krypton laser photocoagulation for idiopathic neovascular lesions showed macular drusen during the follow-up period as a sign of age-related disease.6 Another predisposing factor that can be difficult to rule out initially is a chorioretinal scar (histospot) in eyes with presumed ocular histoplasmosis.6 In this study, we aimed to select eyes with truly idiopathic choroidal neovascularization. Therefore, no person older than 50 years of age was included. Further, histoplasmosis is not endemic, and presumed ocular histoplasmosis syndrome is an exceedingly rare disease in our region. A few such patients were found and excluded. The base for spatial calibration of the images was the distance between the optic disc and the fovea, which was set to 15.6°. This is the average angular distance in normal emmetropic eyes.7 This was chosen instead of the more commonly used diameter of the optic disc because the disc diameter varies largely between individuals.9 We have no reason to believe that the different examination techniques used influenced the results since all images were calibrated individually. In the study by the Macular Photocoagulation Study Group,5 the cumulative proportion of eyes losing six or
1819
Ophthalmology Volume 105, Number 10, October 1998 more lines of Snellen visual acuity was 40% after 12 months of follow-up (their Fig 2). In the current study, only two eyes (10%) deteriorated the same way (one treated and one not treated). We believe the reason for this discrepancy is the different age compositions between the two studies. The results of this study cannot be used for evaluation of the effect of laser photocoagulation. The number of treated eyes was too small and photocoagulation was performed by different ophthalmologists. Two eyes were treated before the guidelines of the Macular Photocoagulation Study Group5 were published, and the treatment technique did not meet modern standards. In two other cases, treatment was applied to a subfoveal membrane for which treatment generally is not recommended. Twelve of 20 eyes had a membrane that was closer than 200 m from the foveola and thus nontreatable according to accepted criteria. However, in several cases a membrane inside the 200-m radius was compatible with good vision. From a functional point of view, the critical distance seemed to be approximately 100 m from the foveola (Fig 3). In contrast to an earlier study,1 no eye with a subfoveal membrane had normal visual acuity, but in one eye it was as good as 20/50. This study suffers from many shortcomings. One is that it is retrospective. Another is that it is not a randomized treatment trial. The patients had the opportunity to choose whether treatment should be given or not after being informed about possible treatment benefits and disadvantages. Conversely, because of the rarity of the disease, randomized clinical studies are difficult to perform, and large multicenter studies often suffer from poor statistical power.6 Recommendations for treatment of idiopathic choroidal neovascularization have been based largely on inference from other studies showing a beneficial effect of argon laser treatment for choroidal neovascular membranes in age-related macular degeneration or presumed ocular histoplasmosis syndrome. Treatment recommendations for idiopathic choroidal neovascularization have not usually considered their tendency to heal spontaneously in younger persons. Involution has been reported in both young adults and children.1,2,10 This study suggests that in patients younger than the age of 50 years at diagnosis, the prognosis of idiopathic choroidal neovascularization differs from eyes with age-related macular degeneration, and the prognosis is often favorable. A conservative treatment approach should be considered.
1820
Possible complications of laser photocoagulation include inadvertently administered laser effects to the foveola, tear of the retinal pigment epithelium, and so-called atrophic creep of the retinal pigment epithelium. However, such extension of the laser scar has thus far been described only for age-related choroidal neovascularization.11 If not conclusive, the results of this report are clear enough to suggest the need for a prospective, randomized treatment study. Based on the result from this study, we believe it ethical to refrain from treatment in cases of idiopathic juxtafoveal or extrafoveal choroidal neovascularization presenting in subjects younger than 50 years of age.
References 1. Ho AC, Yannuzzi LA, Pisicano K, De Rosa J. The natural history of idiopathic subfoveal choroidal neovascularization. Ophthalmology 1995;102:782–9. 2. Campochiaro PA, Morgan KM, Conway BP, Stathos J. Spontaneous involution of subfoveal neovascularization. Am J Ophthalmol 1990;109:668 –75. 3. Flage T, Sand AB, Syrdalen P. Haemorrhagic maculopathy in young adults. Acta Ophthalmol (Copenh) 1977;55:489 –96. 4. Laatikainen L, Juhanen L. Management of juvenile disciform lesions of the macula. Acta Ophthalmol (Copenh) 1983;61: 529 – 40. 5. Argon laser photocoagulation for idiopathic neovascularization. Results of a randomized clinical trial. Macular Photocoagulation Study Group. Arch Ophthalmol 1983;101:1358 – 61. 6. Krypton laser photocoagulation for idiopathic neovascular lesions. Results of a randomized clinical trial. Macular Photocoagulation Study Group. Arch Ophthalmol 1990;108: 832–7. 7. Williams TD. Correlation of nerve head and blind spot elliptical features. Am J Optom Physiol Opt 1988;65:29 –36. 8. Le Grand Y; translated by El Hage SG. Physiological Optics. Berlin; New York: Springer–Verlag, 1980; 77 (Springer series in optical sciences; v. 13). 9. Jonas JB, Gusek GC, Naumann GOH. Optic disc, cup and neuroretinal rim size, configuration and correlations in normal eyes. Invest Ophthalmol Vis Sci 1988;29:1151– 8. 10. Goshorn EB, Hoover DL, Eller AW, et al. Subretinal neovascularization in children and adolescents. J Pediatr Ophthalmol Strabismus 1995;32:178 – 82. 11. Morgan CM, Schatz H. Atrophic creep of the retinal pigment epithelium after focal macular photocoagulation. Ophthalmology 1989;96:96 –103.
Originally received: May 27, 1997. Revision accepted: April 14, 1998. Manuscript no. 97301. From the Institute of Clinical Neuroscience, Department of Ophthalmology, Go¨teborg University, Go¨teborg, Sweden. Reprint requests to Bertil Lindblom, MD, PhD, Institute of Clinical Neuroscience, Department of Ophthalmology, Sahlgrenska University Hospital, S-41345 Go¨teborg, Sweden.
1816
tail, although some information can be extracted from earlier work.3,4 (In this context, a juxtafoveal membrane has its most central border 1–200 m from the foveola while an extrafoveal membrane is 201–2500 m from the foveola.) One reason for the scarcity of reports is the widely used treatment regimen with argon or krypton laser photocoagulation. The treatment of extrafoveal choroidal neovascularization of idiopathic origin has been the subject of several studies, including that of the Macular Photocoagulation Study Group.5,6 In the first study from this group,5 laser photocoagulation was recommended for extrafoveal membranes. This recommendation was renewed in a later study of krypton laser photocoagulation for idiopathic choroidal neovascularization despite the lack of a statistically proven treatment effect.6 The two reports from the Macular Photocoagulation Study Group did not consider the patient’s age. Approximately half of the subjects were older than 50 years. It has been our clinical impression that age is important for the prognosis of idiopathic choroidal neovascularization and that the outcome is more favorable in younger subjects. Therefore, we have for many years been reluctant to treat extrafoveal membranes presenting in patients younger than the age of 50 years. This study was initiated
Lindblom and Andersson 䡠 Idiopathic Neovascularization in an attempt to verify (or disprove) this clinical impression.
Subjects and Methods Subjects By scrutinizing our fundus photograph files, we identified all patients with nontraumatic subretinal neovascularization in the macular area who were younger than 50 years of age at the first examination. The exclusion criteria were: myopia exceeding 6 diopters, signs of other retinal disease such as presumed ocular histoplasmosis or inner punctate choroidopathy, a history of ocular contusion severe enough to cause intraocular bleeding, and signs of age-related maculopathy such as drusen or pigment epithelial detachment.
Table 1. The Underlying Diagnosis in 48 Subjects ⬍50 Years of Age with Nontraumatic Choroidal Neovascularization in the Macular Area Diagnosis
No.
Bilateral
Idiopathic Myopia ⬎ 6 D Punctate inner choroidopathy Presumed ocular histoplasmosis Retinochoroiditis Angioid streaks (pseudoxanthoma elasticum) Toxoplasmosis Best’s vitelliform choroidopathy Diffuse pigment epitheliopathy Heredodegenerative disease Acute posterior multifocal placoid pigment epitheliopathy
19 9 4 4 3
2
3 2 1 1 1
1
Total
48
2
1
1
Methods Fluorescein angiograms and fundus photographs were retrieved from the files of each patient together with relevant clinical data. Data were collected from the following four examinations: 1. The first visit to an ophthalmologist after the presentation of symptoms caused by subretinal neovascularization. 2. The visit at which the patient showed the poorest visual acuity after the diagnosis of subretinal neovascularization. 3. The visit at which the patient showed the best visual acuity after the diagnosis of subretinal neovascularization. 4. The last follow-up visit. In many cases, two of the above-mentioned events coincided. Visual acuity was tested under standardized conditions using an illuminated acuity chart (LIC, Solna, Sweden). This chart has a decimal acuity scale, but measurements were converted to a Snellen fraction except for the computing of mean visual acuity, in which case values were converted to the logarithm of the minimum angle of resolution (logMAR). The luminance of the acuity chart was approximately 500 cd/m2. Test distance was 5 m and best correction was used. Fundus photographs were obtained with a Nikon Retinapan II (Nikon Corp, Tokyo, Japan) or a Canon 60U (Canon Inc, Tokyo, Japan) fundus camera. Fluorescein angiography was performed with a fundus camera in the earlier part of the
study period but was later replaced by a scanning laser ophthalmoscope model 101 (Rodenstock, Ottobrunn, Germany). Fundus photographs and fluorescein angiograms were digitized by a 35-mm film scanner (Nikon Coolscan LS-10; Nikon Corp, Tokyo, Japan) and stored in computer memory together with individual video frames from the scanning laser ophthalmoscope. Images were analyzed with an image analysis program (Optimas Corp, Bothell, WA) using a personal computer. Each image was spatially calibrated. The distance between the center of the optic disc and the foveola was set to 15.6°.7 Distances were then converted to a linear scale by the conversion factor 1 arcmin ⫽ 4.85 m.8 The subretinal neovascular membrane was outlined on the fluorescein angiogram (Fig 1). In eyes with a well-demarcated membrane, the outline was made automatically by the software. In other cases, the automatic procedure appeared influenced by confounding factors such as hemorrhages or pigment epithelial changes surrounding the membrane. In those cases, the membrane was outlined manually. The exact position of the foveola was marked. When needed, this was aided by analysis of the color images. The following measurements were obtained for the choroidal neovascular membrane: area, perimeter, longest axis, width (perpendicular to the longest axis), and distance between the foveola and the nearest membrane edge (this was set to zero if the membrane was subfoveal). In treated eyes in which the choroidal membrane had disappeared, the same measurements were applied to the treatment scar.
Results
Figure 1. Example of measurement of a choroidal neovascular membrane. The membrane is outlined by a white line. A cross marks the foveola, and another cross marks the center of the optic disc.
Forty-seven patients younger than 50 years of age with nontraumatic choroidal neovascularization were identified. The underlying diagnoses are listed in Table 1. Nineteen patients met the inclusion criteria for idiopathic choroidal neovascularization, 2 of whom had bilateral disease. One patient was excluded from further analyses because of insufficient data. He had been treated with argon laser photocoagulation for a neovascular membrane while traveling abroad. Pretreatment photographs were not available. At followup, 10 years after treatment, he had a subfoveal membrane and visual acuity was 20/300. The most relevant measurements for the remaining 18 patients (20 eyes) are listed in Table 2. Membrane length and width did not aid the interpretation of the results and were omitted. Five patients were examined with scanning laser ophthalmoscopy throughout
1817
Ophthalmology Volume 105, Number 10, October 1998 Table 2. Summary of the Result in 18 Subjects with Idiopathic Subretinal Neovascularization Patient No.
Laser Treatment*
1 2 3R 3L 4 5 5REACT 6 7 8 9 10 11 12 13 14 15 16 17 18R 18L Mean SD
Yes No No No Yes No No No No Yes No Yes No Yes No No Yes No No No No
Age at Diagnosis (yrs)
Area at Diagnosis (mm2)
Area Change (%)
Perimeter at Diagnosis (mm)
Perimeter Change (%)
Distance at Diagnosis (mm)
Final Distance (mm)
38 34 20 23 42 35 36 39 39 47 44 50 46 41 38 31 47 45 47 46 46 40 8.0
0.38 0.52 0.29 0.60 0.97 0.26 0.96 0.33 0.61 1.48 0.47 1.42 0.31 2.65 0.41 0.14 0.48 0.23 0.38 0.92 0.61 0.69 0.58
⫺11 ⫺37 331 401 11 ⫺36 ⫺30 ⫺5 ⫺51 ⫺30 ⫺31 ⫺53 ⫺56 ⫺41 ⫺16 ⫺40 18 ⫺10 108 23 77 6.8
2.45 3.28 2.46 3.21 4.17 2.28 4.37 2.60 3.18 4.95 2.69 5.28 2.45 7.18 2.64 1.61 3.06 2.05 2.59 4.03 3.55 3.34 1.30
3 ⫺25 82 121 4 ⫺26 ⫺24 ⫺5 ⫺26 ⫺18 ⫺11 ⫺36 ⫺26 ⫺33 ⫺8 ⫺23 9 ⫺3 47 10 24 ⫺1.73
0.04 0.58 0.03 0.00 0.05 0.17 0.02 0.70 0.12 0.10 0.06 0.00 1.06 0.93 0.27 0.14 1.44 0.47 0.58 0.08 0.40 0.35 0.40
0.06 0.62 0.00 0.00 0.00 0.21 0.11 0.60 0.20 0.15 0.14 0.24 1.32 1.40 0.18 0.17 1.69 0.59 0.44 0.10 0.24 0.40 0.49
Acuity at Diagnosis
Final Acuity
Follow-up (mos)
20/200 20/200 20/30 20/20 20/30 20/40 20/200 20/25 20/20 20/50 20/25 20/200 20/20 20/30 20/125 20/100 20/15 20/25 20/25 20/100 20/30
20/200 20/20 20/50 20/100 HM 20/25 20/30 20/20 20/25 20/50 20/15 20/50 20/20 20/15 20/50 20/15 20/15 20/20 20/20 20/30 20/70
175 187 145 110 20 7 100 63 6 72 71 11 28 10 5 60 32 11 10 14 11 55 58
Area and perimeter ⫽ the size of the subretinal membrane; Distance ⫽ distance from foveola to the nearest border of the subretinal membrane; R ⫽ right; L ⫽ left; REACT ⫽ reactivation; SD ⫽ standard deviation; HM ⫽ hand movements. * If laser treatment has been applied, all measurements refer to the treatment scar.
the follow-up period. Three eyes with involuted choroidal membranes were first examined with fluorescein angiography using a fundus camera and several years later with scanning laser ophthalmoscopy. The results were very similar, suggesting that there was no systematic difference between the methods. Of the 20 eyes, 15 (75%) had a visual acuity of 20/60 or better at the last visit. In ten eyes (50%), visual acuity was 20/25 or better (Fig 2). Two eyes lost six or more lines of Snellen visual acuity: one untreated eye deteriorated from 20/20 to 20/100 and one treated eye from 20/33 to hand movements. The three eyes with a final visual acuity of 20/100 or worse all had a choroidal neovascular membrane less than 200 m from the foveola at first visit. One eye (patient 5) showed a reactivation of the choroidal membrane during the study period. The membrane initially was located 166 m from the foveola and treatment was considered. However, it retracted spontaneously and visual acuity improved to 20/25. Seven months after first diagnosis, reactivation of the membrane reduced visual acuity to 20/200. Six months later, the choroidal membrane was considered inactive and follow-up examinations were discontinued. After another 6 months, the patient noticed successive visual improvement and 8 years later, acuity was 20/30. This was the only case of reactivation in this series of eyes. Figure 3 shows the relation between mean visual acuity and the membrane’s distance to the foveola at the last follow-up examination. Six eyes had undergone argon laser photocoagulation. The central border of the subretinal membrane was less than 200 m from the foveola in four eyes; two had a subfoveal membrane. One of these eyes had a tear of the retinal pigment epithelium develop after treatment and ended up with hand movement vision (patient 4). Two eyes were treated before 1982, and the treatment technique did not meet modern standards. In the other four eyes, treatment was recommended by the ophthalmologist, in two cases
1818
because of growth of the choroidal membrane. Green-only argon laser was used (spot size, 0.1– 0.2 mm; duration, 0.2– 0.3 second). The goal was a distinct white lesion. In general, the size of the
Figure 2. Final Snellen visual acuity plotted against acuity at diagnosis. Data points to the left and below the diagonal line represent improvement. Solid circles represent laser-treated eyes, and open circles represent untreated eyes.
Lindblom and Andersson 䡠 Idiopathic Neovascularization
Figure 3. Mean visual acuity (expressed in logarithm of the minimum angle of resolution) at the last visit in four groups of eyes with idiopathic choroidal neovascularization. The eyes were grouped according to the distance between the choroidal membrane and foveola at the last examination.
Figure 5. Patient 9. Idiopathic choroidal neovascularization in a man who was 44 years old at diagnosis. The examination results after 71 months of follow-up are shown. No treatment had been given. Visual acuity was 20/15.
membrane in the treated eyes was larger than that in the untreated eye (1.23 and 0.44 mm2, respectively; P ⫽ 0.004). Final visual acuity in treated eyes ranged from hand movements to 20/15, but the material was too small to allow any comparisons with the untreated group (Fig 2). In no eye did a subretinal membrane located outside the 200 m radius at the initial examination continue to grow under the fovea. In the only eye in which a membrane extended in a central direction across the 200-m border, visual acuity improved from 20/125 to 20/40 (patient 13). Late pigment epithelial atrophy engaged the foveola in one case (patient 2), but visual acuity still was 20/20 after 15 years of follow-up. In one case (patient 8), widespread pigment epithelial changes developed more than 1 year after laser treatment. We interpreted these changes as “atrophic creep” (Fig 4). Final visual acuity was 20/50. In contrast, a very similar lesion (patient 7) was left untreated and the eye retained a visual acuity of 20/15 (Fig 5). Follow-up time varied from 5 to 187 months (mean, 55 months). In two cases, it was 6 months or less. These two
patients came from another take-up area, and follow-up visual acuity data were requested from their home clinics. Thus, for patient 7, acuity had improved to 20/20 after 68 months of follow-up. For patient 13, visual acuity was unchanged (20/50) after 16 months of follow-up.
Figure 4. Patient 8. Idiopathic choroidal neovascularization in a man who was 47 years old at diagnosis. The macula at the last examination after 72 months of follow-up is shown. Argon laser photocoagulation had been instituted, and widespread retinal atrophy (so-called creeping atrophy) was seen. Visual acuity was 20/50.
Discussion Idiopathic choroidal neovascularization is uncommon in all age groups. The diagnosis is one of exclusion in that a recognizable cause for subretinal neovascularization must be lacking. One reason for studying the younger-than-50 age group separately is that it is difficult to exclude early forms of age-related macular degeneration in older subjects. Indeed, a substantial number of patients enrolled in a study of krypton laser photocoagulation for idiopathic neovascular lesions showed macular drusen during the follow-up period as a sign of age-related disease.6 Another predisposing factor that can be difficult to rule out initially is a chorioretinal scar (histospot) in eyes with presumed ocular histoplasmosis.6 In this study, we aimed to select eyes with truly idiopathic choroidal neovascularization. Therefore, no person older than 50 years of age was included. Further, histoplasmosis is not endemic, and presumed ocular histoplasmosis syndrome is an exceedingly rare disease in our region. A few such patients were found and excluded. The base for spatial calibration of the images was the distance between the optic disc and the fovea, which was set to 15.6°. This is the average angular distance in normal emmetropic eyes.7 This was chosen instead of the more commonly used diameter of the optic disc because the disc diameter varies largely between individuals.9 We have no reason to believe that the different examination techniques used influenced the results since all images were calibrated individually. In the study by the Macular Photocoagulation Study Group,5 the cumulative proportion of eyes losing six or
1819
Ophthalmology Volume 105, Number 10, October 1998 more lines of Snellen visual acuity was 40% after 12 months of follow-up (their Fig 2). In the current study, only two eyes (10%) deteriorated the same way (one treated and one not treated). We believe the reason for this discrepancy is the different age compositions between the two studies. The results of this study cannot be used for evaluation of the effect of laser photocoagulation. The number of treated eyes was too small and photocoagulation was performed by different ophthalmologists. Two eyes were treated before the guidelines of the Macular Photocoagulation Study Group5 were published, and the treatment technique did not meet modern standards. In two other cases, treatment was applied to a subfoveal membrane for which treatment generally is not recommended. Twelve of 20 eyes had a membrane that was closer than 200 m from the foveola and thus nontreatable according to accepted criteria. However, in several cases a membrane inside the 200-m radius was compatible with good vision. From a functional point of view, the critical distance seemed to be approximately 100 m from the foveola (Fig 3). In contrast to an earlier study,1 no eye with a subfoveal membrane had normal visual acuity, but in one eye it was as good as 20/50. This study suffers from many shortcomings. One is that it is retrospective. Another is that it is not a randomized treatment trial. The patients had the opportunity to choose whether treatment should be given or not after being informed about possible treatment benefits and disadvantages. Conversely, because of the rarity of the disease, randomized clinical studies are difficult to perform, and large multicenter studies often suffer from poor statistical power.6 Recommendations for treatment of idiopathic choroidal neovascularization have been based largely on inference from other studies showing a beneficial effect of argon laser treatment for choroidal neovascular membranes in age-related macular degeneration or presumed ocular histoplasmosis syndrome. Treatment recommendations for idiopathic choroidal neovascularization have not usually considered their tendency to heal spontaneously in younger persons. Involution has been reported in both young adults and children.1,2,10 This study suggests that in patients younger than the age of 50 years at diagnosis, the prognosis of idiopathic choroidal neovascularization differs from eyes with age-related macular degeneration, and the prognosis is often favorable. A conservative treatment approach should be considered.
1820
Possible complications of laser photocoagulation include inadvertently administered laser effects to the foveola, tear of the retinal pigment epithelium, and so-called atrophic creep of the retinal pigment epithelium. However, such extension of the laser scar has thus far been described only for age-related choroidal neovascularization.11 If not conclusive, the results of this report are clear enough to suggest the need for a prospective, randomized treatment study. Based on the result from this study, we believe it ethical to refrain from treatment in cases of idiopathic juxtafoveal or extrafoveal choroidal neovascularization presenting in subjects younger than 50 years of age.
References 1. Ho AC, Yannuzzi LA, Pisicano K, De Rosa J. The natural history of idiopathic subfoveal choroidal neovascularization. Ophthalmology 1995;102:782–9. 2. Campochiaro PA, Morgan KM, Conway BP, Stathos J. Spontaneous involution of subfoveal neovascularization. Am J Ophthalmol 1990;109:668 –75. 3. Flage T, Sand AB, Syrdalen P. Haemorrhagic maculopathy in young adults. Acta Ophthalmol (Copenh) 1977;55:489 –96. 4. Laatikainen L, Juhanen L. Management of juvenile disciform lesions of the macula. Acta Ophthalmol (Copenh) 1983;61: 529 – 40. 5. Argon laser photocoagulation for idiopathic neovascularization. Results of a randomized clinical trial. Macular Photocoagulation Study Group. Arch Ophthalmol 1983;101:1358 – 61. 6. Krypton laser photocoagulation for idiopathic neovascular lesions. Results of a randomized clinical trial. Macular Photocoagulation Study Group. Arch Ophthalmol 1990;108: 832–7. 7. Williams TD. Correlation of nerve head and blind spot elliptical features. Am J Optom Physiol Opt 1988;65:29 –36. 8. Le Grand Y; translated by El Hage SG. Physiological Optics. Berlin; New York: Springer–Verlag, 1980; 77 (Springer series in optical sciences; v. 13). 9. Jonas JB, Gusek GC, Naumann GOH. Optic disc, cup and neuroretinal rim size, configuration and correlations in normal eyes. Invest Ophthalmol Vis Sci 1988;29:1151– 8. 10. Goshorn EB, Hoover DL, Eller AW, et al. Subretinal neovascularization in children and adolescents. J Pediatr Ophthalmol Strabismus 1995;32:178 – 82. 11. Morgan CM, Schatz H. Atrophic creep of the retinal pigment epithelium after focal macular photocoagulation. Ophthalmology 1989;96:96 –103.