- Email: [email protected]
Circumscribed choroidal hemangioma: characteristic features with indocyanine green videoangiography
Circumscribed Choroidal Hemangioma: Characteristic Features with Indocyanine Green Videoangiography J. Fernando Arevalo, MD,1,2 Carol L. Shields, MD,1 Jerry A. Shields, MD,1 Philip G. Hykin, MD,1 Patrick De Potter, MD1 Objective: To determine the characteristic features of indocyanine green videoangiography (ICG-V) of circumscribed choroidal hemangioma. Design: Prospective, observational case series. Participants: Twenty-five eyes of 25 consecutive patients with circumscribed choroidal hemangioma. Intervention: Indocyanine green videoangiography and intravenous fluorescein angiography (IVFA) were prospectively performed and reviewed. The specific features on ICG-V were compared with features of IVFA. Main Outcome Measures: The behavior of circumscribed choroidal hemangioma cases was observed with ICG-V and IVFA. Results: On ICG-V, earliest hyperfluorescence of circumscribed choroidal hemangioma was achieved at a mean of 27.6 seconds (range, 13– 62 seconds), whereas maximum hyperfluorescence occurred at 222 seconds (range, 33–707 seconds). In the late frames, all eyes demonstrated a relative decrease in fluorescence, including 18 eyes (72%) that demonstrated “washout” of the dye. Other findings on ICG-V included intrinsic vessels in 19 eyes (76%), a late hyperfluorescent rim in 19 eyes (76%), and late frame hot spots in 14 eyes (56%). On IVFA, the earliest hyperfluorescence was achieved at a mean of 24 seconds (range, 10 – 66 seconds), whereas maximum hyperfluorescence occurred at a mean of 76.3 seconds (range, 21–720 seconds). Increasing hyperfluorescence in the late frames was found in all cases. Other findings included intrinsic vessels in 12 eyes (48%) and hot spots in the late frames in 9 eyes (36%). Conclusions: Circumscribed choroidal hemangioma have specific characteristics on ICG-V that are not visualized with IVFA. We believe that ICG-V may become an important noninvasive tool for the diagnosis of choroidal hemangioma. Ophthalmology 2000;107:344 –350 © 2000 by the American Academy of Ophthalmology. Circumscribed choroidal hemangioma is a well defined benign vascular tumor.1–3 The tumor is almost always unifocal and unilateral. The characteristic fundus appearance of a choroidal hemangioma is that of a subtle red-orange mass with a distinct predilection for the posterior pole.1 Subtle white foci on the tumor surface may be present and probably represent fibrous metaplasia of the overlying retinal pigment epithelium (RPE).1 A surrounding brown rim may be seen in some cases, which represents compressed normal Originally received: August 20, 1998. Accepted: September 15, 1999.
Manuscript no. 98436.
1
Ocular Oncology Service, Wills Eye Hospital, Thomas Jefferson University, Philadelphia, Pennsylvania. 2 Retina Service, Clinica Oftalmologica Centro Caracas, Caracas, Venezuela. Supported by the Macula Foundation, Inc., New York, New York; and the Ocular Oncology Fund, Wills Eye Hospital, and the Eye Tumor Research Foundation, Philadelphia, Pennsylvania. Presented in part as a poster at the American Academy of Ophthalmology annual meeting, Chicago, Illinois, October 1996. Correspondence and reprint requests to Jerry A. Shields, MD, Director, Ocular Oncology Service, Wills Eye Hospital, 900 Walnut Street, Philadelphia, PA 19107.
344
© 2000 by the American Academy of Ophthalmology Published by Elsevier Science Inc.
uvea.1 The visual acuity with circumscribed choroidal hemangioma may be normal or decreased by induced hyperopia from the mass, secondary serous retinal detachment, or macular edema.1 The diagnosis of choroidal hemangioma is usually made on clinical examination, but ancillary tests such as intravenous fluorescein angiography (IVFA), ultrasound, and contrast-enhanced magnetic resonance imaging are helpful in confirming the diagnosis. In some cases, circumscribed choroidal hemangioma can be mistaken for malignant melanoma or choroidal metastasis.4,5 Awareness of the specific clinical features of choroidal hemangioma1–3 and confirmation of the diagnosis using ancillary tests have decreased the frequency of misdiagnosis.6 –10 However, IVFA characteristics of circumscribed choroidal hemangioma may be variable, and, in some cases, it cannot be distinguished from choroidal melanoma.1 Indocyanine green (ICG) is a dye that has several advantages over sodium fluorescein for visualization of circumscribed choroidal hemangioma. It is recognized that ICG provides better visualization of choroidal vessels. The dye absorbs (805 mm) and fluoresces (835 mm) in the near infrared range,11–16 and visualization of fluorescence is posISSN 0161-6420/00/$–see front matter PII S0161-6420(99)00051-2
Arevalo et al 䡠 ICG-V and Choroidal Hemangioma
Figure 1. Circumscribed choroidal hemangioma along the superotemporal vascular arcade. Note the characteristic orange choroidal mass with a surrounding brown rim.
sible through hemorrhage, lipid, RPE, and xanthophyll. Indocyanine green is highly bound to protein, and therefore leaks more slowly through the fenestrations of the choriocapillaris. Recent advances combining digital imaging systems with ICG fundus cameras have allowed high resolution digital ICG videoangiography (ICG-V). Several reports have demonstrated the usefulness of ICG-V in the diagnosis of occult choroidal neovascularizations secondary to agerelated macular degeneration,17–20 other chorioretinal disorders,21–26 and choroidal tumors including choroidal hemangioma.27–29 With regard to choroidal tumors, choroidal hemangioma was identified as a tumor that shows characteristic features on ICG. The purpose of the study was to further investigate ICG-V characteristics of circumscribed choroidal hemangioma to identify specific features that may help in diagnosis.
every 5 to 10 seconds between 15 and 60 seconds, every 30 seconds between 1 and 3 minutes, every 60 seconds between 3 and 10 minutes, and 5 photographs between 10 and 35 minutes. The light intensity was adjusted as little as possible for late frames. Fluorescein angiography was performed by injecting sodium fluorescein 25% (2 ml) into a peripheral arm vein and using a Topcon TRC 50 VT or a TRC 50 IA fundus camera with printing on the Topcon IMAGEnet-H 1024 digital imaging system or on an OIS imaging systems. The IVFA procedure was: first photograph at time 0, second photograph at 5 seconds of injection, followed by photographs every 2 seconds for 10 seconds, every 5 seconds between 15 and 30 seconds, every 10 seconds between 30 and 60 seconds, followed by photographs at 1 to 2 minutes, and photographs at 4.5 minutes and 12 minutes. Parameters analyzed on ICG-V and IVFA included: earliest hyperfluorescence (in seconds); maximum hyperfluorescence (in seconds); intensity of fluorescence (moderate hypofluorescence [⫽ 1], mild hypofluorescence [⫽ 2], isofluorescence [⫽ 3], mild hyperfluorescence [⫽ 4], and moderate hyperfluorescence [⫽ 5]) at 30, 60, 90, and 120 seconds into the study and more than 10 minutes; presence of subretinal fluid (SRF); presence of cystoid macular edema (CME); presence of feeder vessels as well as their location in relation to the optic nerve; presence of intrinsic vessels as well as type (lacy or diffuse), location, and caliber (mm); presence of a hyperfluorescent rim in the late phase; and presence of late focal pinpoint spots of dye accumulation without leakage (termed hot spots). To assist in measuring retinal and choroidal structures, the fluorescein and ICG photographs were viewed with a standard 20-diopter lens (5 ⫻ magnification). Retinal or choroidal structure size was estimated by assuming a retinal vein has a diameter of approximately 125 at the peripapillary area.30 Other lesions, vessels, or subtle retinal changes were measured by comparison to the reference vein. The judgment of fluorescence intensity was made relative to the surrounding normal choroid in both ICG-V and IVFA. To avoid bias, the analysis was performed without details of the clinical findings.
Results Methods Twenty-five consecutive patients with the clinical diagnosis of circumscribed choroidal hemangioma (Fig 1) were prospectively imaged with IVFA and ICG-V at the Ocular Oncology Service of Wills Eye Hospital from March 1992 through October 1995. All patients were initially examined, and a detailed fundus drawing was completed, as well as an A scan and B scan ultrasonography. All 25 patients then underwent ICG-V and IVFA. Indocyanine green (Cardio-Green, Becton Dickinson Microbiology Systems, Cockeysville, MD) was prepared by adding a 25-mg dose to 2 ml aqueous solvent and injecting the solution into a peripheral arm vein after IVFA. Digital ICG-V fundus photographs were taken using a Topcon TRC 50 IA fundus camera (Paramus, NJ) and printed out on the Topcon IMAGEnet-H 1024 digital imaging system (Paramus, NJ) or on an Ophthalmic Imaging Systems (OIS) video camera system (Sacramento, CA). The Topcon camera contained a halogen lamp of 300 W (a continuous light source), and an antireflection coating was applied to the optical system to reduce reflection in the near infrared spectrum. The exciter filter was on the side of the source, and the barrier filter on the side of the film. The ICG-V procedure was: first photograph within 10 seconds of injection, followed by photographs every second for 15 seconds,
The ophthalmoscopic features of the 25 circumscribed choroidal hemangiomas revealed a mean largest diameter of 6.4 mm (range, 4 –9 mm) and mean thickness of 3.1 mm (range, 1.9 – 4.2 mm). The tumor was located within the temporal vascular arcades in 16 eyes (76%). The mean distance from the optic disc was 2.1 mm (range, 0 –9 mm), and the mean distance from the foveola was 1.7 mm (range, 0 – 6 mm). Twenty-two tumors (88%) were orange in color and 3 (12%) were yellow. In 10 cases (40%), a surrounding brown rim was seen. Yellow spots were present on the tumor surface of 11 cases (44%), fibrous metaplasia of the RPE in 6 cases (24%), and RPE clumps were present in 5 cases (20%). Overlying subretinal fluid was present in 10 eyes (40%), overlying cystoid retinal degeneration in 3 cases (12%), surrounding SRF in 17 cases (68%), and CME in 3 eyes (12%).
Indocyanine Green Videoangiography Features Indocyanine green videoangiography findings (Table 1) show that the onset of fluorescence occurred at 27.6 seconds (range, 13– 62 seconds) and achieved a maximum at 222 seconds (range, 33–707 seconds; Figs 2A,B and 3A,B). Subretinal fluid and CME were not visualized with ICG-V. A tumor feeder vessel was seen in only one case (4%), and it was located 4 mm from the optic disc. Intrinsic
345
Ophthalmology Volume 107, Number 2, February 2000 Table 1. Comparison of Intravenous Fluorescein Angiography and Indocyanine Green Videoangiography (25 cases)
Earliest hyperfluorescence (sec) Phase (no. eyes) Prearterial Arterial Laminar venous Feeder Vessels (eyes) Maximum fluorescence (sec) Phase (no. eyes) Laminar venous Full venous Recirculation Late Intrinsic vessels (eyes) Type (no. cases) Lacy Diffuse Location (no. cases) Center Rim Diffuse Caliber (mm) Subretinal fluid (eyes) Cystoid macular edema (eyes) Late hot spots (eyes) Size of spots (mm) Late hyperfluorescent rim (no. of cases)
ICG-V
IVFA
27.6 (range, 13–62)
24 (range, 10–66)
NA NA NA 1 (4 mm from ON) 222 (range, 33–707)
4 16 5 0 76.3 (range, 21–720)
NA NA NA — 19
1 22 1 1 12
10 9
7 5
15 0 4 0.2 (range, 0.125–0.3) 0 0 14 1.1 (range, 0.06–0.2) 19
9 2 1 0.2 (range, 0.06–0.125) 13 2 9 0.1 (range, 0.06–0.125) 0
ICG-V ⫽ indocyanine green videoangiography; IVFA ⫽ intravenous fluorescein angiography; NA ⫽ not applicable; ON ⫽ optic nerve.
vessels were seen in 19 eyes (76%); in 10 (40%) the vessels were lacy, whereas in 9 (36%) they were diffuse (Fig 2A). In 15 cases (60%), intrinsic vessels were seen first in the center of the hemangioma, and their mean caliber was 0.2 mm (range, 0.1– 0.3 mm). In the late frames of the study, all eyes showed decrease of fluorescence including 18 eyes (72%) that demonstrated nearly complete loss of dye with moderate hypofluorescence compared with the surrounding normal choroid. We termed this dye washout (Figs 2C and 3C). Nineteen eyes (76%) showed a late hyperfluorescent rim (Figs 2C and 3C). Hot spots were visualized in the late phases of the angiogram in the 14 eyes (56%) and their mean caliber was 0.1 mm (range, 0.06 – 0.2 mm; Figs 2C and 3C). None of eyes showed SRF and CME as seen on IVFA.
Intravenous Fluorescein Angiography Features The IVFA findings (Table 1) detail that the onset of fluorescence occurred at 24 seconds (range, 10 – 66 seconds) and achieved a maximum at 76.3 seconds (range, 21–720 seconds; Figs 4 and 5). In 20 cases (80%) the earliest fluorescence occurred during the arterial or prearterial phase of the angiogram, with the maximum fluorescence being reached during the full venous phase in 22 (88%) and during the laminar venous, recirculation, and late phase in one case (4%) each. Intrinsic vessels were seen in 12 eyes (48%); in 7 they were of lacy type and in 5 they were diffuse. In 9 cases (36%) intrinsic vessels were seen in the center of the hemangioma, and their mean caliber was 0.2 mm (range, 0.06 – 0.1 mm). Hot spots were visualized in the late phases of the angiogram in 9 eyes (36%), and the mean caliber was 0.1 mm (range, 0.06 – 0.1 mm). Subretinal fluid was seen in 13 eyes (52%), and CME was seen in 2 eyes (8%). None of the eyes showed the washout effect or a late hyperfluorescent rim as seen on ICG-V.
346
Discussion Choroidal hemangioma is the most common vascular tumor of the choroid, and the circumscribed type must be differentiated from other types of amelanotic choroidal tumors such as amelanotic choroidal melanoma, choroidal metastasis, posterior scleritis, neurilemoma, and leiomyoma. Most often, this differentiation can be made clinically with indirect ophthalmoscopy by experienced observers. In difficult cases, IVFA, ICG,27 ocular ultrasonography, contrastenhanced magnetic resonance imaging,31 and radioactive phosphorous (32P) uptake testing32 have been shown to be helpful adjunctive measures in the differential diagnosis of intraocular tumors.1 It was believed that IVFA features of circumscribed choroidal hemangiomas were diagnostic,9,10,33 but in time it was realized that the IVFA patterns of various choroidal tumors were not specific. The most frequent IVFA pattern of circumscribed choroidal hemangiomas is characterized by irregular linear hyperfluorescence of large choroidal vessels within the tumor during the prearterial or early arterial phases. In the arterial or venous phases, there is progressive staining of the extravascular tissue within the tumor, often with pinpoint foci of hyperfluorescence over the tumor. The late phase reveals intraretinal hyperfluorescence secondary to leakage of fluorescein into the cystoid spaces within the retina.1 Advances in infrared imaging technology have improved the resolution of the ICG angiograms over the past years, and recent reports have focused on the use of ICG in
Arevalo et al 䡠 ICG-V and Choroidal Hemangioma scribed choroidal hemangiomas in their series on ICG-V patterns of choroidal tumors. They found that they all achieved maximal fluorescence by nearly 1 minute, which was followed by slow clearance of the fluorescence. The tumor vessels within the choroidal hemangioma were frequently visualized. In the late frames, the tumor demonstrated a washout of dye in two cases and a rim of hyperfluorescence surrounding the tumor in all cases. Sallet et al28 found that in four of their five cases, abnormal intratu-
Figure 2. Same patient as in Figure 1. (A) Early indocyanine green videoangiography (ICG-V) frame shows the onset of fluorescence. Intrinsic vessels are also seen. (B) Maximum fluorescence on ICG-V. (C) Hot spots, hyperfluorescent rim, and washout of dye are seen on late frames on ICG-V.
identifying choroidal neovascularization, especially that which is poorly defined on fluorescein angiography.17–20 Observations on ICG-V findings of three small series of circumscribed choroidal hemangiomas have been recently described.27,29 Shields et al27 studied four cases of circum-
Figure 3. Another patient with a circumscribed choroidal hemangioma inferior to the optic disc. (A) Early indocyanine green videoangiography (ICG-V) frame shows the onset of fluorescence. Intrinsic vessels are also seen. (B) Maximum fluorescence on ICG-V. (C) Hot spots, hyperfluorescent rim, and washout of dye are seen on late frames on ICG-V.
347
Ophthalmology Volume 107, Number 2, February 2000
Figure 5. Same patient as in Figure 3. (A) Intravenous fluorescein angiography (IVFA) arterial phase frame showing the onset of fluorescence. (B) Maximum fluorescence during IVFA. Figure 4. Same patient as in Figures 1 and 2. (A) Intravenous fluorescein angiography (IVFA) arterial phase frame showing the onset of fluorescence. (B) Maximum fluorescence during IVFA.
moral vessels were seen in the early phases. The hyperfluorescence increased within a short period, filling the dark spaces seen in the earlier phases, followed by intense late fluorescence of the whole lesion. Piccolino et al29 recently reported their series of 12 cases of circumscribed choroidal hemangioma, with findings similar to Shields’ previous study. Our study of 25 cases of circumscribed choroidal hemangioma confirms and expands the above findings.27 The ICG-V findings for circumscribed choroidal hemangiomas showed, similar to earlier reports, that the onset of fluorescence (mean, 27.6 seconds) and maximum hyperfluorescence (mean, 222 seconds) were achieved early. The IVFA findings for circumscribed choroidal hemangiomas showed that the onset of fluorescence (mean, 24 seconds) and maximum hyperfluorescence (mean, 76.3 seconds) were also rapid and comparable to ICG-V. Intrinsic vessels were seen in 19 (76%) eyes with ICG-V, whereas they were seen in only 12 eyes (48%) with IVFA. Hot spots were visualized in the late phases of the angiogram in 14 eyes (56%) with ICG-V and in 9 eyes (36%) with IVFA. Hot spot is a term that we use to describe the presence of late focal pinpoint spots of dye accumulation without leakage. We are unsure
348
if the hot spots seen on these two different techniques represent the same phenomenon. It is speculative that the hot spots found on IVFA over circumscribed choroidal hemangioma represent accumulation of fluorescein into cystic spaces of the retinal pigment epithelium or retina. We speculate that hot spots found on ICG-V represent traces of dye remaining in the intrinsic vessels of the choroidal tumor. None of the eyes showed a late hyperfluorescent rim or feeder vessels on IVFA, whereas on ICG-V, a late hyperfluorescent rim was seen in 19 (76%) eyes and feeder vessels were seen in one (4%) case. Intravenous fluorescein angiography was more sensitive in detecting SRF (in 52% eyes) and CME (in 8% eyes) than ICG-V (no SRF or CME detected). The IVFA findings discussed here, although characteristic for choroidal hemangiomas, may occasionally occur with other amelanotic choroidal tumors, limiting the diagnostic usefulness of IVFA. The advantage of IVFA over ICG-V in detecting SRF and CME is also limited by the fact that these findings are readily made by indirect ophthalmoscopy and slit-lamp biomicroscopy. The ICG-V findings, may be complementary to IVFA in diagnostic specificity for choroidal hemangiomas. A constellation of ICG-V findings including early hyperfluorescence of the mass that is maintained up to 2 minutes into the angiogram, washout of dye from the tumor in the late frames, late hyperfluorescent rim
Arevalo et al 䡠 ICG-V and Choroidal Hemangioma
Figure 6. Graph comparing the fluorescence of circumscribed choroidal hemangiomas throughout intravenous fluorescein angiography (IVFA) and indocyanine green videoangiography (ICG-V). The hyperfluorescence with both dyes persisted up to 120 seconds into the angiogram. However, in the late phases of the IVFA, the hyperfluorescence persisted, whereas in the ICG-V, the hyperfluorescence decreased and became mildly hypofluorescent. *Tumor fluorescence graded as compared with surrounding normal choroid, 1 ⴝ Moderate hypofluorescence, 2 ⴝ Mild hypofluorescence, 3 ⴝ Isofluorescence, 4 ⴝ Mild hyperfluorescence, 5 ⴝ Moderate hyperfluorescence.
surrounding the tumor, and hot spots on the surface of the tumor are ICG-V characteristics specific to choroidal hemangiomas. Comparison of fluorescence (Fig 6 and Table 1) with intravenous fluorescein angiography and ICG-V reveals that hyperfluorescence onset occurred early (⬍30 seconds) and increased up to 90 seconds. This hyperfluorescence with both dyes persisted up to 120 seconds. However, in the late phases of the IVFA, the hyperfluorescence persisted, whereas in the ICG-V the hyperfluorescence decreased and became mildly hypofluorescent (Fig 6 and Table 1). Limitations of our study include the fact that ICG-V is still evolving and more information may emerge on other fundus lesions that could have similar appearances. In addition, the recommended technique of dye injection does not include calculation and dilution of Cardio-Green on a weight basis. These factors may contribute to variations in the time of onset, time of maximal hyperfluorescence, and subjective intensity interpretations. Other limitations of this study include the small number of tumors evaluated, the lack of an objective method of measurement of degree of fluorescence, and the lack of a matched control group with choroidal melanoma or metastasis. In summary, our study suggests that ICG-V is superior to IVFA in the assessment of choroidal hemangioma because of specific angiographic features including early hyperfluorescence followed by late hypofluorescence (washout of dye) and a hyperfluorescent rim. Indocyanine green videoangiography may therefore be a significant aid in diagnosing difficult cases and preferable to IVFA in detecting intrinsic vessels and feeder vessels of the tumor.
References 1. Shields JA, Shields CL. Intraocular Tumors: A Text and Atlas. Philadelphia: Saunders; 1992.
2. Anand R, Augsburger JJ, Shields JA. Circumscribed choroidal hemangiomas. Arch Ophthalmol 1989;107:1338 – 42. 3. Augsburger JJ, Shields JA, Moffat KP. Circumscribed choroidal hemangiomas: long-term visual prognosis. Retina 1981;1: 56 – 61. 4. Ferry AP. Lesions mistaken for malignant melanoma of the posterior uvea: a clinicopathologic analysis of 100 cases with ophthalmoscopically visible lesions. Arch Ophthalmol 1964; 72:463–9. 5. Shields JA, Zimmerman LE. Lesions simulating malignant melanoma of the posterior uvea. Arch Ophthalmol 1973;89: 466 –71. 6. Shields JA, McDonald PR. Improvements in the diagnosis of posterior uveal melanomas. Arch Ophthalmol 1974;91:259 – 64. 7. Shields JA, Augsburger JJ, Brown GC, Stephens RF. The differential diagnosis of posterior uveal melanomas. Ophthalmology 1980;87:518 –22. 8. Jarrett WH II, Hagler WS, Larose JH, Shields JA. Clinical experience with presumed hemangioma of the choroid: radioactive phosphorous uptake studies as an aid in differential diagnosis. Trans Am Acad Ophthalmol Otolaryngol 1976;81: 862–70. 9. MacLean AL, Maumenee AE. Hemangioma of the choroid. Am J Ophthalmol 1960;50:3–11. 10. Norton EWD, Gutman F. Fluorescein angiography and hemangiomas of the choroid. Arch Ophthalmol 1967;78:121–5. 11. Baker KJ. Binding of sulfobromophthalein (BSP) sodium and indocyanine green (ICG) by plasma alpha-1 lipoproteins. Proc Soc Exp Biol Med 1966;122:957– 63. 12. Leevy CM, Bender J, Silverberg M, Naylor J. Physiology of dye extraction by the liver: comparative studies of sulfobromophthalein and indocyanine green. Ann NY Acad Sci 1963; 111:161–75. 13. Goresky CA. Initial distribution and rate of uptake of sulfobromophthalein in the liver. Am J Physiol 1964;207:13–26. 14. Kogure K, David NJ, Yamanouchi U, Choromokos E. Infrared absorption angiography of the fundus circulation. Arch Ophthalmol 1970;83:209 –14. 15. Flower RW, Hochheimer BF. Indocyanine green dye fluorescence and infrared absorption choroidal angiography performed simultaneously with fluorescein angiography. Johns Hopkins Med J 1976;138:33– 42. 16. Benson RC, Kues HA. Fluorescence properties of indocyanine green as related to angiography. Phys Med Biol 1978;23:159 – 63. 17. Yannuzzi LA, Slakter JS, Sorenson JA, et al. Digital indocyanine green videoangiography and choroidal neovascularization. Retina 1992;12:191–223. 18. Sorenson JA, Yannuzzi LA, Slakter JS, et al. A pilot study of digital indocyanine green videoangiography for recurrent occult choroidal neovascularization in age-related macular degeneration. Arch Ophthalmol 1994;112:473–9. 19. Yannuzzi LA, Hope-Ross M, Slakter JS, et al. Analysis of vascularized pigment epithelial detachments using indocyanine green videoangiography. Retina 1994;14:99 –113. 20. Guyer DR, Yannuzzi LA, Slakter JS, et al. Digital indocyanine-green videoangiography of occult choroidal neovascularization. Ophthalmology 1994;101:1727–37. 21. Guyer DR, Puliafito CA, Mones JM, et al. Digital indocyanine-green angiography in chorioretinal disorders. Ophthalmology 1992;99:287–91. 22. Dhaliwal RS, Maguire AM, Flower RW, Arribas NP. Acute posterior multifocal placoid pigment epitheliopathy. An indocyanine green angiographic study. Retina 1993;13:317–25. 23. Yuzawa M, Kawamura A, Matsui M. Indocyanine green vid-
349
Ophthalmology Volume 107, Number 2, February 2000
24. 25. 26. 27. 28. 29.
eo-angiographic findings in Harada’s disease. Jpn J Ophthalmol 1993;37:456 – 66. Guyer DR, Yannuzzi LA, Slakter JS, et al. Digital indocyanine green videoangiography of central serous chorioretinopathy. Arch Ophthalmol 1994;112:1057– 62. Ie D, Glaser BM, Murphy RP, et al. Indocyanine green angiography in multiple evanescent white-dot syndrome. Am J Ophthalmol 1994;117:7–12. Yuzawa M, Kawamura A, Matsui M. Indocyanine green video angiographic findings in acute posterior multifocal placoid pigment epitheliopathy. Acta Ophthalmol 1994;72:128 –33. Shields CL, Shields JA, De Potter P. Patterns of indocyanine green videoangiography of choroidal tumours. Br J Ophthalmol 1995;79:237– 45. Sallet G, Amoaku WMK, Lafaut BA, et al. Indocyanine green angiography of choroidal tumors. Graefes Arch Clin Exp Ophthalmol 1995;233:677– 89. Piccolino FC, Borgia L, Zinicola E. Indocyanine green an-
350
30.
31.
32.
33.
giography of circumscribed choroidal hemangiomas. Retina 1996;16:19 –28. Kim JW, Freeman WR, el-Haig W, et al. Baseline characteristics, natural history, and risk factors to progression in eyes with stage 2 macular holes. Results from a prospective randomized clinical trial. Vitrectomy for Macular Hole Study Group. Ophthalmology 1995;102:1818 –28; discussion 1828 –9. De Potter P, Flanders AE, Shields JA, et al. The role of fat-suppression technique and gadopentetate dimeglumine in magnetic resonance imaging evaluation of intraocular tumors and simulating lesions. Arch Ophthalmol 1994;112:340 – 8. Lanning R, Shields JA. Comparison of radioactive phosphorous (32P) uptake test in comparable sized choroidal melanomas and hemangiomas. Am J Ophthalmol 1979;87:769 –72. Novotny HR, Alvis DL. A method of photographing fluorescence in circulating blood in the human retina. Circulation 1961;24:82– 6.
Manuscript no. 98436.
1
Ocular Oncology Service, Wills Eye Hospital, Thomas Jefferson University, Philadelphia, Pennsylvania. 2 Retina Service, Clinica Oftalmologica Centro Caracas, Caracas, Venezuela. Supported by the Macula Foundation, Inc., New York, New York; and the Ocular Oncology Fund, Wills Eye Hospital, and the Eye Tumor Research Foundation, Philadelphia, Pennsylvania. Presented in part as a poster at the American Academy of Ophthalmology annual meeting, Chicago, Illinois, October 1996. Correspondence and reprint requests to Jerry A. Shields, MD, Director, Ocular Oncology Service, Wills Eye Hospital, 900 Walnut Street, Philadelphia, PA 19107.
344
© 2000 by the American Academy of Ophthalmology Published by Elsevier Science Inc.
uvea.1 The visual acuity with circumscribed choroidal hemangioma may be normal or decreased by induced hyperopia from the mass, secondary serous retinal detachment, or macular edema.1 The diagnosis of choroidal hemangioma is usually made on clinical examination, but ancillary tests such as intravenous fluorescein angiography (IVFA), ultrasound, and contrast-enhanced magnetic resonance imaging are helpful in confirming the diagnosis. In some cases, circumscribed choroidal hemangioma can be mistaken for malignant melanoma or choroidal metastasis.4,5 Awareness of the specific clinical features of choroidal hemangioma1–3 and confirmation of the diagnosis using ancillary tests have decreased the frequency of misdiagnosis.6 –10 However, IVFA characteristics of circumscribed choroidal hemangioma may be variable, and, in some cases, it cannot be distinguished from choroidal melanoma.1 Indocyanine green (ICG) is a dye that has several advantages over sodium fluorescein for visualization of circumscribed choroidal hemangioma. It is recognized that ICG provides better visualization of choroidal vessels. The dye absorbs (805 mm) and fluoresces (835 mm) in the near infrared range,11–16 and visualization of fluorescence is posISSN 0161-6420/00/$–see front matter PII S0161-6420(99)00051-2
Arevalo et al 䡠 ICG-V and Choroidal Hemangioma
Figure 1. Circumscribed choroidal hemangioma along the superotemporal vascular arcade. Note the characteristic orange choroidal mass with a surrounding brown rim.
sible through hemorrhage, lipid, RPE, and xanthophyll. Indocyanine green is highly bound to protein, and therefore leaks more slowly through the fenestrations of the choriocapillaris. Recent advances combining digital imaging systems with ICG fundus cameras have allowed high resolution digital ICG videoangiography (ICG-V). Several reports have demonstrated the usefulness of ICG-V in the diagnosis of occult choroidal neovascularizations secondary to agerelated macular degeneration,17–20 other chorioretinal disorders,21–26 and choroidal tumors including choroidal hemangioma.27–29 With regard to choroidal tumors, choroidal hemangioma was identified as a tumor that shows characteristic features on ICG. The purpose of the study was to further investigate ICG-V characteristics of circumscribed choroidal hemangioma to identify specific features that may help in diagnosis.
every 5 to 10 seconds between 15 and 60 seconds, every 30 seconds between 1 and 3 minutes, every 60 seconds between 3 and 10 minutes, and 5 photographs between 10 and 35 minutes. The light intensity was adjusted as little as possible for late frames. Fluorescein angiography was performed by injecting sodium fluorescein 25% (2 ml) into a peripheral arm vein and using a Topcon TRC 50 VT or a TRC 50 IA fundus camera with printing on the Topcon IMAGEnet-H 1024 digital imaging system or on an OIS imaging systems. The IVFA procedure was: first photograph at time 0, second photograph at 5 seconds of injection, followed by photographs every 2 seconds for 10 seconds, every 5 seconds between 15 and 30 seconds, every 10 seconds between 30 and 60 seconds, followed by photographs at 1 to 2 minutes, and photographs at 4.5 minutes and 12 minutes. Parameters analyzed on ICG-V and IVFA included: earliest hyperfluorescence (in seconds); maximum hyperfluorescence (in seconds); intensity of fluorescence (moderate hypofluorescence [⫽ 1], mild hypofluorescence [⫽ 2], isofluorescence [⫽ 3], mild hyperfluorescence [⫽ 4], and moderate hyperfluorescence [⫽ 5]) at 30, 60, 90, and 120 seconds into the study and more than 10 minutes; presence of subretinal fluid (SRF); presence of cystoid macular edema (CME); presence of feeder vessels as well as their location in relation to the optic nerve; presence of intrinsic vessels as well as type (lacy or diffuse), location, and caliber (mm); presence of a hyperfluorescent rim in the late phase; and presence of late focal pinpoint spots of dye accumulation without leakage (termed hot spots). To assist in measuring retinal and choroidal structures, the fluorescein and ICG photographs were viewed with a standard 20-diopter lens (5 ⫻ magnification). Retinal or choroidal structure size was estimated by assuming a retinal vein has a diameter of approximately 125 at the peripapillary area.30 Other lesions, vessels, or subtle retinal changes were measured by comparison to the reference vein. The judgment of fluorescence intensity was made relative to the surrounding normal choroid in both ICG-V and IVFA. To avoid bias, the analysis was performed without details of the clinical findings.
Results Methods Twenty-five consecutive patients with the clinical diagnosis of circumscribed choroidal hemangioma (Fig 1) were prospectively imaged with IVFA and ICG-V at the Ocular Oncology Service of Wills Eye Hospital from March 1992 through October 1995. All patients were initially examined, and a detailed fundus drawing was completed, as well as an A scan and B scan ultrasonography. All 25 patients then underwent ICG-V and IVFA. Indocyanine green (Cardio-Green, Becton Dickinson Microbiology Systems, Cockeysville, MD) was prepared by adding a 25-mg dose to 2 ml aqueous solvent and injecting the solution into a peripheral arm vein after IVFA. Digital ICG-V fundus photographs were taken using a Topcon TRC 50 IA fundus camera (Paramus, NJ) and printed out on the Topcon IMAGEnet-H 1024 digital imaging system (Paramus, NJ) or on an Ophthalmic Imaging Systems (OIS) video camera system (Sacramento, CA). The Topcon camera contained a halogen lamp of 300 W (a continuous light source), and an antireflection coating was applied to the optical system to reduce reflection in the near infrared spectrum. The exciter filter was on the side of the source, and the barrier filter on the side of the film. The ICG-V procedure was: first photograph within 10 seconds of injection, followed by photographs every second for 15 seconds,
The ophthalmoscopic features of the 25 circumscribed choroidal hemangiomas revealed a mean largest diameter of 6.4 mm (range, 4 –9 mm) and mean thickness of 3.1 mm (range, 1.9 – 4.2 mm). The tumor was located within the temporal vascular arcades in 16 eyes (76%). The mean distance from the optic disc was 2.1 mm (range, 0 –9 mm), and the mean distance from the foveola was 1.7 mm (range, 0 – 6 mm). Twenty-two tumors (88%) were orange in color and 3 (12%) were yellow. In 10 cases (40%), a surrounding brown rim was seen. Yellow spots were present on the tumor surface of 11 cases (44%), fibrous metaplasia of the RPE in 6 cases (24%), and RPE clumps were present in 5 cases (20%). Overlying subretinal fluid was present in 10 eyes (40%), overlying cystoid retinal degeneration in 3 cases (12%), surrounding SRF in 17 cases (68%), and CME in 3 eyes (12%).
Indocyanine Green Videoangiography Features Indocyanine green videoangiography findings (Table 1) show that the onset of fluorescence occurred at 27.6 seconds (range, 13– 62 seconds) and achieved a maximum at 222 seconds (range, 33–707 seconds; Figs 2A,B and 3A,B). Subretinal fluid and CME were not visualized with ICG-V. A tumor feeder vessel was seen in only one case (4%), and it was located 4 mm from the optic disc. Intrinsic
345
Ophthalmology Volume 107, Number 2, February 2000 Table 1. Comparison of Intravenous Fluorescein Angiography and Indocyanine Green Videoangiography (25 cases)
Earliest hyperfluorescence (sec) Phase (no. eyes) Prearterial Arterial Laminar venous Feeder Vessels (eyes) Maximum fluorescence (sec) Phase (no. eyes) Laminar venous Full venous Recirculation Late Intrinsic vessels (eyes) Type (no. cases) Lacy Diffuse Location (no. cases) Center Rim Diffuse Caliber (mm) Subretinal fluid (eyes) Cystoid macular edema (eyes) Late hot spots (eyes) Size of spots (mm) Late hyperfluorescent rim (no. of cases)
ICG-V
IVFA
27.6 (range, 13–62)
24 (range, 10–66)
NA NA NA 1 (4 mm from ON) 222 (range, 33–707)
4 16 5 0 76.3 (range, 21–720)
NA NA NA — 19
1 22 1 1 12
10 9
7 5
15 0 4 0.2 (range, 0.125–0.3) 0 0 14 1.1 (range, 0.06–0.2) 19
9 2 1 0.2 (range, 0.06–0.125) 13 2 9 0.1 (range, 0.06–0.125) 0
ICG-V ⫽ indocyanine green videoangiography; IVFA ⫽ intravenous fluorescein angiography; NA ⫽ not applicable; ON ⫽ optic nerve.
vessels were seen in 19 eyes (76%); in 10 (40%) the vessels were lacy, whereas in 9 (36%) they were diffuse (Fig 2A). In 15 cases (60%), intrinsic vessels were seen first in the center of the hemangioma, and their mean caliber was 0.2 mm (range, 0.1– 0.3 mm). In the late frames of the study, all eyes showed decrease of fluorescence including 18 eyes (72%) that demonstrated nearly complete loss of dye with moderate hypofluorescence compared with the surrounding normal choroid. We termed this dye washout (Figs 2C and 3C). Nineteen eyes (76%) showed a late hyperfluorescent rim (Figs 2C and 3C). Hot spots were visualized in the late phases of the angiogram in the 14 eyes (56%) and their mean caliber was 0.1 mm (range, 0.06 – 0.2 mm; Figs 2C and 3C). None of eyes showed SRF and CME as seen on IVFA.
Intravenous Fluorescein Angiography Features The IVFA findings (Table 1) detail that the onset of fluorescence occurred at 24 seconds (range, 10 – 66 seconds) and achieved a maximum at 76.3 seconds (range, 21–720 seconds; Figs 4 and 5). In 20 cases (80%) the earliest fluorescence occurred during the arterial or prearterial phase of the angiogram, with the maximum fluorescence being reached during the full venous phase in 22 (88%) and during the laminar venous, recirculation, and late phase in one case (4%) each. Intrinsic vessels were seen in 12 eyes (48%); in 7 they were of lacy type and in 5 they were diffuse. In 9 cases (36%) intrinsic vessels were seen in the center of the hemangioma, and their mean caliber was 0.2 mm (range, 0.06 – 0.1 mm). Hot spots were visualized in the late phases of the angiogram in 9 eyes (36%), and the mean caliber was 0.1 mm (range, 0.06 – 0.1 mm). Subretinal fluid was seen in 13 eyes (52%), and CME was seen in 2 eyes (8%). None of the eyes showed the washout effect or a late hyperfluorescent rim as seen on ICG-V.
346
Discussion Choroidal hemangioma is the most common vascular tumor of the choroid, and the circumscribed type must be differentiated from other types of amelanotic choroidal tumors such as amelanotic choroidal melanoma, choroidal metastasis, posterior scleritis, neurilemoma, and leiomyoma. Most often, this differentiation can be made clinically with indirect ophthalmoscopy by experienced observers. In difficult cases, IVFA, ICG,27 ocular ultrasonography, contrastenhanced magnetic resonance imaging,31 and radioactive phosphorous (32P) uptake testing32 have been shown to be helpful adjunctive measures in the differential diagnosis of intraocular tumors.1 It was believed that IVFA features of circumscribed choroidal hemangiomas were diagnostic,9,10,33 but in time it was realized that the IVFA patterns of various choroidal tumors were not specific. The most frequent IVFA pattern of circumscribed choroidal hemangiomas is characterized by irregular linear hyperfluorescence of large choroidal vessels within the tumor during the prearterial or early arterial phases. In the arterial or venous phases, there is progressive staining of the extravascular tissue within the tumor, often with pinpoint foci of hyperfluorescence over the tumor. The late phase reveals intraretinal hyperfluorescence secondary to leakage of fluorescein into the cystoid spaces within the retina.1 Advances in infrared imaging technology have improved the resolution of the ICG angiograms over the past years, and recent reports have focused on the use of ICG in
Arevalo et al 䡠 ICG-V and Choroidal Hemangioma scribed choroidal hemangiomas in their series on ICG-V patterns of choroidal tumors. They found that they all achieved maximal fluorescence by nearly 1 minute, which was followed by slow clearance of the fluorescence. The tumor vessels within the choroidal hemangioma were frequently visualized. In the late frames, the tumor demonstrated a washout of dye in two cases and a rim of hyperfluorescence surrounding the tumor in all cases. Sallet et al28 found that in four of their five cases, abnormal intratu-
Figure 2. Same patient as in Figure 1. (A) Early indocyanine green videoangiography (ICG-V) frame shows the onset of fluorescence. Intrinsic vessels are also seen. (B) Maximum fluorescence on ICG-V. (C) Hot spots, hyperfluorescent rim, and washout of dye are seen on late frames on ICG-V.
identifying choroidal neovascularization, especially that which is poorly defined on fluorescein angiography.17–20 Observations on ICG-V findings of three small series of circumscribed choroidal hemangiomas have been recently described.27,29 Shields et al27 studied four cases of circum-
Figure 3. Another patient with a circumscribed choroidal hemangioma inferior to the optic disc. (A) Early indocyanine green videoangiography (ICG-V) frame shows the onset of fluorescence. Intrinsic vessels are also seen. (B) Maximum fluorescence on ICG-V. (C) Hot spots, hyperfluorescent rim, and washout of dye are seen on late frames on ICG-V.
347
Ophthalmology Volume 107, Number 2, February 2000
Figure 5. Same patient as in Figure 3. (A) Intravenous fluorescein angiography (IVFA) arterial phase frame showing the onset of fluorescence. (B) Maximum fluorescence during IVFA. Figure 4. Same patient as in Figures 1 and 2. (A) Intravenous fluorescein angiography (IVFA) arterial phase frame showing the onset of fluorescence. (B) Maximum fluorescence during IVFA.
moral vessels were seen in the early phases. The hyperfluorescence increased within a short period, filling the dark spaces seen in the earlier phases, followed by intense late fluorescence of the whole lesion. Piccolino et al29 recently reported their series of 12 cases of circumscribed choroidal hemangioma, with findings similar to Shields’ previous study. Our study of 25 cases of circumscribed choroidal hemangioma confirms and expands the above findings.27 The ICG-V findings for circumscribed choroidal hemangiomas showed, similar to earlier reports, that the onset of fluorescence (mean, 27.6 seconds) and maximum hyperfluorescence (mean, 222 seconds) were achieved early. The IVFA findings for circumscribed choroidal hemangiomas showed that the onset of fluorescence (mean, 24 seconds) and maximum hyperfluorescence (mean, 76.3 seconds) were also rapid and comparable to ICG-V. Intrinsic vessels were seen in 19 (76%) eyes with ICG-V, whereas they were seen in only 12 eyes (48%) with IVFA. Hot spots were visualized in the late phases of the angiogram in 14 eyes (56%) with ICG-V and in 9 eyes (36%) with IVFA. Hot spot is a term that we use to describe the presence of late focal pinpoint spots of dye accumulation without leakage. We are unsure
348
if the hot spots seen on these two different techniques represent the same phenomenon. It is speculative that the hot spots found on IVFA over circumscribed choroidal hemangioma represent accumulation of fluorescein into cystic spaces of the retinal pigment epithelium or retina. We speculate that hot spots found on ICG-V represent traces of dye remaining in the intrinsic vessels of the choroidal tumor. None of the eyes showed a late hyperfluorescent rim or feeder vessels on IVFA, whereas on ICG-V, a late hyperfluorescent rim was seen in 19 (76%) eyes and feeder vessels were seen in one (4%) case. Intravenous fluorescein angiography was more sensitive in detecting SRF (in 52% eyes) and CME (in 8% eyes) than ICG-V (no SRF or CME detected). The IVFA findings discussed here, although characteristic for choroidal hemangiomas, may occasionally occur with other amelanotic choroidal tumors, limiting the diagnostic usefulness of IVFA. The advantage of IVFA over ICG-V in detecting SRF and CME is also limited by the fact that these findings are readily made by indirect ophthalmoscopy and slit-lamp biomicroscopy. The ICG-V findings, may be complementary to IVFA in diagnostic specificity for choroidal hemangiomas. A constellation of ICG-V findings including early hyperfluorescence of the mass that is maintained up to 2 minutes into the angiogram, washout of dye from the tumor in the late frames, late hyperfluorescent rim
Arevalo et al 䡠 ICG-V and Choroidal Hemangioma
Figure 6. Graph comparing the fluorescence of circumscribed choroidal hemangiomas throughout intravenous fluorescein angiography (IVFA) and indocyanine green videoangiography (ICG-V). The hyperfluorescence with both dyes persisted up to 120 seconds into the angiogram. However, in the late phases of the IVFA, the hyperfluorescence persisted, whereas in the ICG-V, the hyperfluorescence decreased and became mildly hypofluorescent. *Tumor fluorescence graded as compared with surrounding normal choroid, 1 ⴝ Moderate hypofluorescence, 2 ⴝ Mild hypofluorescence, 3 ⴝ Isofluorescence, 4 ⴝ Mild hyperfluorescence, 5 ⴝ Moderate hyperfluorescence.
surrounding the tumor, and hot spots on the surface of the tumor are ICG-V characteristics specific to choroidal hemangiomas. Comparison of fluorescence (Fig 6 and Table 1) with intravenous fluorescein angiography and ICG-V reveals that hyperfluorescence onset occurred early (⬍30 seconds) and increased up to 90 seconds. This hyperfluorescence with both dyes persisted up to 120 seconds. However, in the late phases of the IVFA, the hyperfluorescence persisted, whereas in the ICG-V the hyperfluorescence decreased and became mildly hypofluorescent (Fig 6 and Table 1). Limitations of our study include the fact that ICG-V is still evolving and more information may emerge on other fundus lesions that could have similar appearances. In addition, the recommended technique of dye injection does not include calculation and dilution of Cardio-Green on a weight basis. These factors may contribute to variations in the time of onset, time of maximal hyperfluorescence, and subjective intensity interpretations. Other limitations of this study include the small number of tumors evaluated, the lack of an objective method of measurement of degree of fluorescence, and the lack of a matched control group with choroidal melanoma or metastasis. In summary, our study suggests that ICG-V is superior to IVFA in the assessment of choroidal hemangioma because of specific angiographic features including early hyperfluorescence followed by late hypofluorescence (washout of dye) and a hyperfluorescent rim. Indocyanine green videoangiography may therefore be a significant aid in diagnosing difficult cases and preferable to IVFA in detecting intrinsic vessels and feeder vessels of the tumor.
References 1. Shields JA, Shields CL. Intraocular Tumors: A Text and Atlas. Philadelphia: Saunders; 1992.
2. Anand R, Augsburger JJ, Shields JA. Circumscribed choroidal hemangiomas. Arch Ophthalmol 1989;107:1338 – 42. 3. Augsburger JJ, Shields JA, Moffat KP. Circumscribed choroidal hemangiomas: long-term visual prognosis. Retina 1981;1: 56 – 61. 4. Ferry AP. Lesions mistaken for malignant melanoma of the posterior uvea: a clinicopathologic analysis of 100 cases with ophthalmoscopically visible lesions. Arch Ophthalmol 1964; 72:463–9. 5. Shields JA, Zimmerman LE. Lesions simulating malignant melanoma of the posterior uvea. Arch Ophthalmol 1973;89: 466 –71. 6. Shields JA, McDonald PR. Improvements in the diagnosis of posterior uveal melanomas. Arch Ophthalmol 1974;91:259 – 64. 7. Shields JA, Augsburger JJ, Brown GC, Stephens RF. The differential diagnosis of posterior uveal melanomas. Ophthalmology 1980;87:518 –22. 8. Jarrett WH II, Hagler WS, Larose JH, Shields JA. Clinical experience with presumed hemangioma of the choroid: radioactive phosphorous uptake studies as an aid in differential diagnosis. Trans Am Acad Ophthalmol Otolaryngol 1976;81: 862–70. 9. MacLean AL, Maumenee AE. Hemangioma of the choroid. Am J Ophthalmol 1960;50:3–11. 10. Norton EWD, Gutman F. Fluorescein angiography and hemangiomas of the choroid. Arch Ophthalmol 1967;78:121–5. 11. Baker KJ. Binding of sulfobromophthalein (BSP) sodium and indocyanine green (ICG) by plasma alpha-1 lipoproteins. Proc Soc Exp Biol Med 1966;122:957– 63. 12. Leevy CM, Bender J, Silverberg M, Naylor J. Physiology of dye extraction by the liver: comparative studies of sulfobromophthalein and indocyanine green. Ann NY Acad Sci 1963; 111:161–75. 13. Goresky CA. Initial distribution and rate of uptake of sulfobromophthalein in the liver. Am J Physiol 1964;207:13–26. 14. Kogure K, David NJ, Yamanouchi U, Choromokos E. Infrared absorption angiography of the fundus circulation. Arch Ophthalmol 1970;83:209 –14. 15. Flower RW, Hochheimer BF. Indocyanine green dye fluorescence and infrared absorption choroidal angiography performed simultaneously with fluorescein angiography. Johns Hopkins Med J 1976;138:33– 42. 16. Benson RC, Kues HA. Fluorescence properties of indocyanine green as related to angiography. Phys Med Biol 1978;23:159 – 63. 17. Yannuzzi LA, Slakter JS, Sorenson JA, et al. Digital indocyanine green videoangiography and choroidal neovascularization. Retina 1992;12:191–223. 18. Sorenson JA, Yannuzzi LA, Slakter JS, et al. A pilot study of digital indocyanine green videoangiography for recurrent occult choroidal neovascularization in age-related macular degeneration. Arch Ophthalmol 1994;112:473–9. 19. Yannuzzi LA, Hope-Ross M, Slakter JS, et al. Analysis of vascularized pigment epithelial detachments using indocyanine green videoangiography. Retina 1994;14:99 –113. 20. Guyer DR, Yannuzzi LA, Slakter JS, et al. Digital indocyanine-green videoangiography of occult choroidal neovascularization. Ophthalmology 1994;101:1727–37. 21. Guyer DR, Puliafito CA, Mones JM, et al. Digital indocyanine-green angiography in chorioretinal disorders. Ophthalmology 1992;99:287–91. 22. Dhaliwal RS, Maguire AM, Flower RW, Arribas NP. Acute posterior multifocal placoid pigment epitheliopathy. An indocyanine green angiographic study. Retina 1993;13:317–25. 23. Yuzawa M, Kawamura A, Matsui M. Indocyanine green vid-
349
Ophthalmology Volume 107, Number 2, February 2000
24. 25. 26. 27. 28. 29.
eo-angiographic findings in Harada’s disease. Jpn J Ophthalmol 1993;37:456 – 66. Guyer DR, Yannuzzi LA, Slakter JS, et al. Digital indocyanine green videoangiography of central serous chorioretinopathy. Arch Ophthalmol 1994;112:1057– 62. Ie D, Glaser BM, Murphy RP, et al. Indocyanine green angiography in multiple evanescent white-dot syndrome. Am J Ophthalmol 1994;117:7–12. Yuzawa M, Kawamura A, Matsui M. Indocyanine green video angiographic findings in acute posterior multifocal placoid pigment epitheliopathy. Acta Ophthalmol 1994;72:128 –33. Shields CL, Shields JA, De Potter P. Patterns of indocyanine green videoangiography of choroidal tumours. Br J Ophthalmol 1995;79:237– 45. Sallet G, Amoaku WMK, Lafaut BA, et al. Indocyanine green angiography of choroidal tumors. Graefes Arch Clin Exp Ophthalmol 1995;233:677– 89. Piccolino FC, Borgia L, Zinicola E. Indocyanine green an-
350
30.
31.
32.
33.
giography of circumscribed choroidal hemangiomas. Retina 1996;16:19 –28. Kim JW, Freeman WR, el-Haig W, et al. Baseline characteristics, natural history, and risk factors to progression in eyes with stage 2 macular holes. Results from a prospective randomized clinical trial. Vitrectomy for Macular Hole Study Group. Ophthalmology 1995;102:1818 –28; discussion 1828 –9. De Potter P, Flanders AE, Shields JA, et al. The role of fat-suppression technique and gadopentetate dimeglumine in magnetic resonance imaging evaluation of intraocular tumors and simulating lesions. Arch Ophthalmol 1994;112:340 – 8. Lanning R, Shields JA. Comparison of radioactive phosphorous (32P) uptake test in comparable sized choroidal melanomas and hemangiomas. Am J Ophthalmol 1979;87:769 –72. Novotny HR, Alvis DL. A method of photographing fluorescence in circulating blood in the human retina. Circulation 1961;24:82– 6.