- Email: [email protected]
Anterior segment indocyanine green angiography in anterior scleritis and episcleritis
Anterior Segment Indocyanine Green Angiography in Anterior Scleritis and Episcleritis Yan Guex-Crosier, MD, Jacques Durig, MD Objective: To evaluate the pattern of anterior segment indocyanine green (ICG) angiography in episcleritis and scleritis. Design: Prospective comparative (paired-eye) observational case series. Participants: Twenty subjects presenting clinical diseases compatible with episcleritis or scleritis. Methods: Anterior segment ICG angiography was performed according to a standard protocol in subjects presenting either episcleritis or scleritis. Photographs of the anterior segment were taken in the early phase (up to 3 minutes after dye injection), intermediate phase (10 –12 minutes) and late phase (30 – 45 minutes). The inflamed zones were compared with the same regions of the controlateral eye. The amount of protein ICG exudation was scored by a masked observer as follows: zero for no exudation, one for slight exudation, two for moderate exudation, and three for severe exudation. Main Outcome Measures: Evaluation of dye leakage, which reflects protein exudation, with anterior segment ICG angiography in episcleritis and scleritis. Results: Twenty subjects with a mean age of 43 ⫾ 15 years (7 male, 13 female) were enrolled in the study. Thirteen subjects had anterior scleritis (7 nodular, 5 diffuse, and 1 scleromalacia perforans), and 7 subjects had episcleritis. Only 1 out of 7 subjects with episcleritis showed a slight ICG leakage (a score of one), whereas all subjects with scleritis had ICG leakage scores of one or more (P ⫽ 0.0005, Fisher exact test). Conclusions: ICG angiography of the anterior segment of the eye is a good clinical test to differentiate episcleritis from scleritis. Ophthalmology 2003;110:1756 –1763 © 2003 by the American Academy of Ophthalmology.
Scleral fibroblasts produce type I, type III, and type VIII collagen fibers that intercede with elastin fibers.1 This particular organization of collagen fibers thus confers to scleral tissue its rigidity. When scleral tissue reaches maturity after the first years of life, it loses its elasticity and cannot be stretched. Scleral fibroblasts have a very low metabolic requirement, which is supplied mainly by vascularized adjacent structures such as the episclera and the choroid. Scleral tissue does not contain any blood vessels, with the exception of perforating arteries that cross through the sclera to reach the ciliary body and choroid. Vascularization of the anterior segment of the eye depends on the anterior ciliary arteries and on the long posterior ciliary arteries,2 which subsequently subdivide into capillaries. In contrast to retinal vessels, the superficial and deep plexi of episcleral Originally received: March 1, 2002. Accepted: March 5, 2003. Manuscript no. 220166. From Jules Gonin Eye Hospital, University of Lausanne, Lausanne, Switzerland. Presented in part at the Association for Research in Vision and Ophthalmology meeting, Fort Lauderdale, Florida, May 1999. The authors have no financial interest in any product, drug, instrument, or piece of equipment mentioned in this article. Proprietary interest: none. Correspondence and reprint requests to Yan Guex-Crosier, MD, Jules Gonin Eye Hospital, 15 av. de France, CH-1004, Lausanne, Switzerland.
1756
© 2003 by the American Academy of Ophthalmology Published by Elsevier Inc.
vessels consist of fenestrated capillaries. These histologic characteristics give to the scleral vessels their high permeability to fluorescein. In systemic diseases, inflammation occurs primarily around blood vessels.3– 8 Focal zones of inflammation are characterized by cellular infiltration and local edema, resulting from exudation of proteins from fenestrated capillaries. During systemic diseases, inflammatory mediators are produced in the proximity of episcleral and scleral vessels, and upregulation of leukocyte adhesion molecule expression has also been observed.9 The extent of inflammation depends, obviously, on the extent of the disease. The local liberation of large amounts of inflammatory mediators in scleritis can also affect adjacent tissues, such as choroidal vessels and orbital structures, and can result in a secondary myositis. Orbital involvement is more frequently seen in posterior scleritis, in which ultrasonography and computerized tomography help in diagnosis by demonstrating a thickened posterior sclera.10 –12 The poor stretching capacities of collagen bundles and the high prevalence of pain receptors within scleral tissues are responsible for the intense pain that is pathognomonic of scleritis. When the anterior segment of the eye is involved, intense ocular redness can be observed. Anterior scleritis is typically defined as simple, nodular, or necrotizing scleritis, with or without surrounding inflammation.1 Scleromalacia perforans is a specific ISSN 0161-6420/03/$–see front matter doi:10.1016/S0161-6420(03)00567-0
Guex-Crosier and Durig 䡠 Anterior Segment ICG in Scleritis scleral disease that results from severe vascular occlusion and vasculitis and occurs primarily in subjects with longstanding histories of rheumatoid arthritis. Depending on the tissue layer involved, two conditions have been clinically described: episcleritis and scleritis. Focal ocular redness is described classically as bright red in episcleritis, and bluish or purple– blue in scleritis. In scleritis, both superficial and deep episcleral vessels are dilated and do not bleach after topical 10% phenylephrine application. Furthermore, pain is more intense in scleritis than in episcleritis. A mild to severe photophobia associated with tearing can be present with either condition, and mild anterior uveitis can be observed. However, in some cases, the differential diagnosis between episcleritis and scleritis can be difficult. Differentiation of episcleritis from scleritis is of the utmost importance because, in most cases, episcleritis, in contrast to scleritis, is a self-limited disease that is rarely associated with systemic complications.13 Anterior segment fluorescein angiograms have already been used to characterize ocular inflammation in episcleritis and scleritis.14 But because fluorescein is not a proteinbound molecule, after injection, it diffuses freely through the fenestrated capillaries. The presence of both fenestrated and nonfenestrated capillaries of the conjunctiva and episclera is responsible for a patchy leakage during the early phase of anterior segment angiography. During the first transit of dye, the superficial vascular plexus is filled, dilated vessels can be seen, and rapid filling of the vessels can be observed both in episcleritis and scleritis. Unfortunately, the rapid superficial leakage of dye that occurs even in noninflamed tissues in the presence of fenestrated capillaries precludes good visualization of the deep layers. The patchy leakage of conjunctiva and diffuse leakage of the nonfenestrated capillaries of episclera impede a good visualization of the inflamed zone during the late phase of angiography. The recent development of infrared cameras for posterior pole angiography has allowed for the use of indocyanine green (ICG) angiography in clinical investigations. ICG dye has its peak excitation (805 nm) and peak fluorescence spectra in the near-infrared region, where the pigment epithelium, melanin, macular xanthophylls, and choroidal pigment are relatively transparent. These properties allow deeper tissular penetration. The ICG molecule has some unique properties that allow the study of protein exudation. ICG has a higher molecular weight than does fluorescein, and ⬎98% of the molecule is bound by proteins. In contrast, fluorescein is dissolved primarily as a free molecule in plasma. During fluorescein angiography of the posterior pole of the eye, fluorescein leakage is impeded by the blood– brain barrier, a barrier that is present also in nonfenestrated capillaries of the retina. Thus, in external ocular structures, fluorescein leaks spontaneously from fenestrated capillaries of noninflamed tissues, producing a patchy leakage.14 Thus, the ICG molecule is almost completely protein bound in the blood, rendering it relatively impermeable to fenestrated capillaries (such as the choriocapillaries). The dye does not leak extensively from fenestrations of choroidal vessels. An increase of permeability resulting in focal leakage can be observed in ocular inflammation. Therefore,
Table 1. Watson’s Criteria of Episcleritis versus Scleritis Episcleritis Color Pain 10% phenylephrine test
Red Slight Blanching of the superficial episcleral and conjunctival vessels Narrow bright slit Inner reflection of the beam examination light seem undisturbed, whereas outer reflection seem displaced forward
Scleritis Purple Severe No blanching of the scleral tissue Distortion and tortuousity of both anterior and posterior lines
the fluorescence of the ICG molecule through infrared wavelength allows visualization of deep scleral leakage, whereas fluorescein is rapidly blocked. These properties of ICG suggest that images obtained from external green angiographies could be useful in differentiating episcleritis from scleritis. The aim of this study was to analyze the pattern of anterior segment ICG angiography in scleral inflammations such as episcleritis and scleritis.
Subjects and Methods From November 1997 to March 2000, 20 consecutive subjects presenting clinical diseases compatible with the diagnosis of episcleritis or scleritis were involved in a prospective study. The subjects were seen at the Jules Gonin Eye Outpatient Clinic. Clinical evaluation of all subjects was performed by the same clinician, a senior fellow at the Jules Gonin Eye Hospital (YGC). The diagnosis of episcleritis compared to scleritis was made on a clinical basis according to the following four clinical criteria, suggested by P. Watson1 (Table 1): 1. The ocular pain (scored from 0 –10) according to an analogical scale, 0 corresponding to the absence of pain and 10 to the highest degree of pain that the subject can imagine. 2. The 10% phenylephrin test. The use of 10% epinephrine produces a vasoconstriction of the inflamed superficial episcleral vessels. Persistence of inflammation can be seen in scleral vessels (positive ⫽ 1, negative ⫽ 0). 3. The episcleritis has in daylight a red color, whereas scleritis seems blue or necrotic in necrotizing scleritis (red ⫽ 0, purple ⫽ 1). 4. In episcleritis, only the outer reflection line of the slit-lamp beam seems distorted, and the posterior line seems undisturbed. In scleritis, distortion and tortuosity of both anterior and posterior lines are present (absence of posterior distortion ⫽ 0, presence of posterior distortion ⫽ 1). Each of these criteria was scored as 0 if negative (or compatible with an episcleritis) and 1 if positive, except for pain, which was scored from 0 to 10. The overall score consisted of the sum of the four criteria. A global score greater than or equal to four was considered to result of a scleritis. A score of less than four was considered to correspond to an episcleritis. This analogical scale was given because the diagnosis of episcleritis compared to scleritis may be difficult in some borderline cases. Once the diagnosis of episcleritis compared to scleritis was established, subjects were investigated further according to the following protocol. While a bolus of 10% fluorescein was injected, external frames were obtained of the
1757
Ophthalmology Volume 110, Number 9, September 2003 Table 2. Scoring of Exudation during the Late Phase of External Indocyanine Green Angiography (30 – 45 minutes) Score
Exudation
0 1 2 3
None Slight, barely visible Moderate Severe, larges zones of hyperfluorescence
anterior segment of the eye with a Topcon TRC 50 Digital Imaging System Imagenet (Topcon Corporation, Tokyo, Japan). A bolus of 25 mg of ICG (Cardiogreen, Becton Dickinson, Franklin Lakes, NJ), diluted in 7.5 ml of 0.9% NaCl, was then injected, and frames were obtained at the following three time points. During the early phase (up to 3 minutes after dye injection), frames were taken from the inflamed part of the eye. During the intermediate phase (10 –12 minutes), nine pictures were taken from the 9 primary eye positions. The zone corresponding to the site of acute inflammation in episcleritis or scleritis was compared with the same portion of the eye in the controlateral eye. During the late phase (30 – 45 minutes), nine additional frames were taken. Because ICG is highly protein bound, “hot spots” corresponding to dye leakage exudation may be visible during the late phase of external ICG angiography. ICG leakage, which reflects the amount of protein exudation, was scored from zero to three in a masked analysis by a neutral observer (JD) who did not know the clinical diagnosis. The following scale was used to grade dye leakage: 0 ⫽ no exudation, 1 ⫽ mild exudation, 2 ⫽ moderate exudation, and 3 ⫽ severe
exudation (Table 2). A standard clinical workup was performed with all subjects to detect any possible underlying systemic disease. For statistical analysis, the Fisher exact test was used to compare leakage scores in the two groups (episcleritis and scleritis). Receiver operating characteristic (ROC) analysis was used to compare the sensitivity and the specificity of the ICG leakage; a leakage score of 1 or more was considered as suggestive of the presence of a scleritis. Systat software (Systat Inc., Evanston, IL) was used for statistical analysis.
Results Twenty subjects (7 male, 13 female) presenting inflammations of the anterior segment of the eye that were compatible with either episcleritis or scleritis were enrolled in a prospective open study from November 1997 to March 2000. The mean age of the subjects was 43 ⫾ 15 years. Two additional subjects were investigated according to the same protocol as negative controls; one presented with a conjunctival injection secondary to ocular rosacea, and the other was suffering from acute anterior HLA-B27-related uveitis. The clinical inflammation score is summarized in Table 3; subjects presenting overall clinical scores of less than four were considered as presenting episcleritis. On a clinical basis, 13 subjects had anterior scleritis and 7 had episcleritis. Among subjects with scleritis, 7 had the nodular form, 5 had the diffuse form and 1 had scleromalacia perforans. Seven subjects had episcleritis: 4 nodular and 3 simple.
Table 3. Clinical Score and Indocyanine Green (ICG) Score in Episcleritis and Scleritis Color Blue (score ⴝ 1 if positive)
10% Phenylephrine Test (score ⴝ 1 if positive)
Distorted Posterior Line (score ⴝ 1 if positive)
Overall Clinical Score (score 0–14)
ICG Leakage score (0–3)
Type
Pain (score 0–10)
Nodular episcleritis
1.5
0
0.5
0
2
0
Nodular episcleritis Nodular episcleritis Diffuse episcleritis Diffuse episcleritis Diffuse episcleritis Diffuse episcleritis Nodular scleritis
3 1.5 1.5 2 2 1 4
0 0 0 0 0 0 1
0 0 0 0 0.5 0 1
0 1 1 0 0 0 1
3 2.5 2.5 2 2.5 1 7
1 0 0 0 0 0 1
10
1
1
1
13
3
4 7 6 3 5
1 1 1 0 1
1 1 1 0.5 1
1 1 1 0.5 1
7 10 9 4 8
2 3 1 1 3
Diffuse scleritis
6
1
1
1
9
3
Diffuse scleritis
10
1
1
1
13
1
0 0.5 1 1
1 0.5 1 0
1 0.5 1 1
5 6.5 12 4
1 2 2 3
Nodular scleritis Nodular Nodular Nodular Nodular Nodular
scleritis scleritis scleritis scleritis scleritis
Diffuse scleritis Diffuse scleritis Diffuse scleritis Necrotizing scleritis
3 5 9 2
Patients with overall clinical scores of inflammation under four were considered as presenting episcleritis.
1758
Medical Workup Serum rheumatoid factor — — (case 1) — — — — Anticardiolipid positive Anticardiolipid positive Neurosarcoidosis — — Rheumatoid arthritis Anticardiolipid (case 2) Wegener ANCA positive — — — Systemic lupus erythematosus (case 3)
Guex-Crosier and Durig 䡠 Anterior Segment ICG in Scleritis
Figure 1. A–D, External indocyanine green angiography of the anterior segment. The subject presents a simple anterior episcleritis (case 1). During the early phase (A), a rapid filling of the vessels is seen. No leakage is visible during the intermediate phase (B) and late phase (score leakage ⫽ 0) (C). The image in the late phase is comparable to the controlateral eye (D). E, F, Image capture 50 seconds (E) and 10 minutes (F) after fluorescein injection.
All subjects were investigated by anterior segment ICG angiography. External angiographies of the eye were scored from zero to three according to the classification shown in Table 2. Only one out of 7 subjects with episcleritis showed a slight ICG leakage (a score of one), whereas all subjects with scleritis had ICG leakage scores of one or more (P ⫽ 0.0005, Fisher exact test). ROC analysis showed that for a cutoff value of one, the sensitivity of the test was 100% and the specificity was 85.7%. Severe ICG leakage (an angiographic score greater than three) was seen only in subjects presenting with anterior scleritis associated with a systemic disease, such as rheumatoid arthritis or sarcoidosis. All subjects presenting scleritis scores of two or more were treated with systemic steroids or immunosuppressive drugs. A clear decrease in ICG dye leakage was seen after such treatment and disappearance of symptoms. In all subjects enrolled in the early, intermediate, and late phases, angiograms were obtained of the controlateral eyes. None of the controlateral eyes showed ICG exudation, except for one subject with a severe bilateral scleritis associated with rheumatoid arthritis. One subject had severe scleromalacia perforans in the right eye, which was associated with systemic lupus erythematosus; this case is described in detail below.
Case Reports Case 1: Simple Anterior Episcleritis A 30-year-old white woman presented with a 10-day history of ocular pain in the temporal portion of the left eye (pain score of 1.5 in the analogic scale). Visual acuity was 10/10 in each eye. No cell or flare was present in either eye. The ocular pressure was 12 mmHg in each eye. A slight red scleral injection (clinical score ⫽ 0) was present in the nasal part of the left sclera but disappeared after application of 10% phenylephrine drops (clinical score ⫽ 0). No distortion of the posterior line could be seen by slit-lamp
examination (clinical score ⫽ 0). The overall clinical score was 1.5. Funduscopy was normal in both eyes. Results of complete medical workup were normal. The ocular pain gradually abated after therapy with topical diclofenac sodium application 4 times a day and topical 0.1% fluorometholone 3 times a day. Anterior segment ICG angiography did not reveal any hot spots during the late time point frames (leakage score ⫽ 0) (Fig 1).
Case 2: Simple Anterior Scleritis A 55-year-old man presented with a superotemporal lesion in the left eye. Pain had been present for 4 days (clinical pain score ⫽ 6). Slit-lamp examination of the anterior segment revealed the presence of a purple lesion (clinical score ⫽ 1), both the anterior and posterior lines of the slit-lamp beam were distorted (clinical score ⫽ 1), a slight rupture of the blood– ocular barrier, with a 2⫹ Tyndall effect, but no visible cell. The lesion did not disappear after application of 10% phenylephrine (clinical score ⫽ 1). The sum of the clinical scores was nine, corresponding to a scleritis diagnosis. A systemic laboratory workup revealed a positive rheumatoid factor. Rapid filling of episcleral and scleral vessels could be observed during the early frames of fluorescein angiography, and 2 hot spots were clearly visible during the intermediate and late phases of the external ICG angiography (leakage score ⫽ 3) (Fig 2C), whereas no exudation was present in fluorescein angiography (Fig 2E, F). Therapy consisted of hourly topical 1% prednisolone acetate and 4 times daily diclofenac sodium. Systemic nimesulide, a specific cox-2 inhibitor, was also administered. Systemic and topical therapies were progressively tapered, and ultimately stopped, 2 months later. After the resolution of the apparent inflammation, external ICG angiography was repeated, at which time no leakage was observed (Fig 3).
Case 3: Necrotizing Scleritis A 68-year-old white woman presented severe pain in the right eye. She had a necrotizing scleritis in the temporal region with ulcer-
1759
Ophthalmology Volume 110, Number 9, September 2003
Figure 2. A–D, External indocyanine green (ICG) angiography of the anterior segment in a subject presenting a simple anterior scleritis (case 2). During the early phase, scleral and episcleral vessels are filled with ICG. During the intermediate phase at 10 minutes (B) and late phase at 30 minutes (C), progressive leakage of ICG appears (hot spot; score leakage ⫽ 3), whereas no leakage is seen in the controlateral eye in the late phase (D). E, F, The external fluorescein angiography.
ation of the sclera at the limbus. At that time, she was treated with 4 mg/day of methylprednisolone for lupus alopecia. Ocular pain was scored as two on the analogic scale. The necrotic area seemed dark blue (clinical score ⫽ 1) Visual acuity was limited to 5/60 in the right eye with S-1, C-2 at 140°. A round scleral ulceration with a black aspect (clinical score ⫽ 1) of 5.4 mm in diameter, 3.8 mm away from the limbus, was present in the temporal region of the right eye and corresponded to scleral necrosis (clinical score ⫽ 1; scleral necrosis) (Fig 4); this was surrounded by irregular dilated vessels. A 1⫹ Tyndall with 1⫹ cell was present in the right eye. The subject had a 2⫹ nuclear sclerosis, and the fundus revealed inferior choroidal thickening adjacent to the ora serrata. In the left eye, the subject had 3/10 visual acuity with S-1.5, C-1 at 90°. The anterior segment was without inflammation; tonometry was 13 mmHg in each eye. Systemic workup revealed anti SS-A (Ro) antibodies with 26 UI/ml, and negative anti-SS-B (Le). Tests for anti-DNA antibodies were negative, and the rheumatoid factor was negative. Anterior segment angiography revealed the presence of a large avascular zone that corresponded to the necrotic sclera. A delay in scleral perfusion was present in the inferotemporal region of the eye. ICG angiography of the anterior segment revealed a 3⫹ leakage score of the vessels in the inferotemporal region (Fig 5).
Discussion ICG external angiography seems useful in the diagnosis of episcleritis compared to scleritis. This small pilot study clearly demonstrated a difference in dye leakage between episcleritis and scleritis. ICG has been widely used as a tracer to measure organ perfusion15 and in the investigation of inflammation of the superficial and deep plexi in scleritis.16,17 Fluorescein dye has been very useful in the investigation of retinal diseases, because the blood– ocular barrier
1760
is impermeable to fluorescein dye unless inflammation is present. In the outer section of the eye, the capillary vessels are fenestrated. When fluorescein is used to detect vascular abnormalities in episcleritis or scleritis, a rapid filling of the vessels is seen during angiography of the outer segment of the eye, but the dye leaks very rapidly out of the episcleral plexus. Investigation of the anterior segment of the eye by external ICG angiography was necessary to show that leakage is almost specific to deep scleral inflammation and was almost absent in episcleritis. This probably can be explained by the different degrees of severity of the inflammatory process in episcleritis and in scleritis. The latter is a more severe disease that can lead to vision-threatening complications such as keratitis, secondary glaucoma, uveitis, cataract, optic neuritis, macular edema, subretinal mass, annular ciliochoroidal detachment, serous retinal detachment, and even perforation of the globe.18,19 ICG was shown to be useful to quantify the severity of inflammation because the amount of dye leakage reflects the degree of protein exudation during the inflammatory process. In healthy eyes (or controlateral eyes, in this study), no dye leakage is present during the intermediate and late phases of angiograms. Episcleritis is primarily a self-limited disease that can resolve spontaneously within 2 to 3 weeks, whereas scleritis is a more severe process that can lead to visual loss or scleral perforation. Although these two entities can usually be clinically differentiated according to the criteria previously defined by Watson,1 differentiating between severe episcleritis and slight scleritis may be difficult. Slight dye leakage was seen in only one out of 7 episcleritis cases, whereas it was present in all scleritis cases. The
Guex-Crosier and Durig 䡠 Anterior Segment ICG in Scleritis
Figure 3. A–D, External indocyanine green angiography performed in case 2, after 2 months of local and systemic therapy. No more leakage appears during the intermediate (B) and late (C) phases. During the late stage, the image is comparable to the controlateral eye (D).
severity of dye leakage was particularly high in scleritis cases presenting overlying systemic diseases (Table 3). This test will probably be useful in the management of severe scleritis cases that need to be treated by systemic immunosuppressive therapy (such as cyclophosphamide or methotrexate), but this point needs to be verified by larger studies. Our aim was to demonstrate that ICG dye leakage is characteristic of scleritis and is almost absent in episcleritis (ICG leakage score ⫽ 0 in 6 out of 7 subjects) and in normal eyes. This point could be largely verified (P ⫽ 0.0005, Fisher exact test). With a cutoff value of one for the leakage score, the sensitivity is approximately 100% and the specificity is 85.7% (ROC analysis). The use of standard time points during ICG angiography is helpful to compare dye leakage during the intermediate and late phases. In all subjects with severe dye leakage, hot spots could be seen in the blue–purple zones of severe inflammation. Protein exudation that was detected by external ICG angiography reflects the changes in reflectivity that could be observed by ultrasound biomicroscopy in severe scleritis.20 Necrotizing scleritis is a severe form of scleritis in which an area of necrosis is surrounded by acutely congested episcleral vessels.21 Ocular or systemic complications are
observed in ⬎60% of such cases.18 The extent of the disease can clearly be delimited by external ICG angiography. The necrotic area is surrounded by zones of hyperpermeability of vascular episcleral channels (Fig 5). In necrotizing scleritis, the activity of the inflammation seems to be very low using the slit-lamp examination (clinical score of four in case 3); a recent progression of the zone of necrosis is a sign that an increase in immunosuppressive therapy is necessary. A clear dye leakage was visible in the anterior segment ICG angiographies (ICG dye score of three). This high score reflects the persistence of an acute inflammatory process in the scleral and episcleral vessels. In this case, the immunosuppressive therapy should be increased to avoid further progression of the zone of necrosis. In 2 subjects with scleritis, a decrease of ICG dye exudation was clearly visible after recovery from scleritis. Monitoring of the severity of the disease by measuring protein exudation by external ICG angiography could be a good method for the follow-up of the disease and the tapering of antiinflammatory therapy. This point could be verified in a larger group of subjects. Three subjects (negative controls) were also tested: two had severe ocular rosacea and one, with HLA-B27 uveitis, presented intense ocular pain during an acute onset of
1761
Ophthalmology Volume 110, Number 9, September 2003
Figure 4. Anterior necrotizing scleritis in a 68-year-old white woman presenting a systemic lupus erythematosus. The choroid is visible (black area).
anterior uveitis. No dye leakage was seen in any of these three subjects. The leakage of dye during ICG external angiographies seems to be specific to scleral inflammation
but not to ocular inflammation. ICG angiography of the external segment of the eye thus seems to be very useful in grading the intensity of ocular inflammation in scleritis.
Figure 5. A–D, External indocyanine green (ICG) angiography in case 3 presenting a necrotizing scleritis. During the early phase (A), the zones of necrosis seem hypofluorescent. In the adjacent zones, vascular filling can be seen, and some hot spots of dye leakage appear in the late phase (score leakage ⫽ 3) (C). No leakage is visible in the controlateral eye (D). E, F, Fluorescein angiography shows a large zone of hypofluorescence (necrosis). The adjacent zone (where inflamed vessels are seen in ICG picture) seems hypofluorescent (ischemia).
1762
Guex-Crosier and Durig 䡠 Anterior Segment ICG in Scleritis External ICG angiography thus seems to be a valuable tool for the evaluation of disease activity in episcleritis and scleritis. Acknowledgments. The authors thank Guy van Melle, MD, Institute for Social and Preventive Medicine, Lausanne, for statistical assistance.
References 1. Watson P. Diseases of the sclera and episclera. In: Tasman W, Jaeger EA, eds. Duane’s Clinical Ophthalmology, Vol. 2, 19th ed. Philadelphia: JB Lippincott, 1995:1– 45. 2. Meyer PA. Patterns of blood flow in episcleral vessels studied by low-dose fluorescein videoangiography. Eye 1988;2(pt 5): 533– 46. 3. Hoang-Xuan T, Foster CS, Rice BA. Scleritis in relapsing polychondritis. Response to therapy. Ophthalmology 1990;97: 892– 8. 4. Sainz de la Maza M, Foster CS, Jabbur NS. Scleritis associated with systemic vasculitic diseases. Ophthalmology 1995; 102:687–92. 5. Denk PO, Thiel HJ, Zierhut M. Episcleritis and scleritis. An overview of modern diagnostic and therapeutic concepts [in German]. Klin Monatsbl Augenheilkd 1997;211:140 –50. 6. Di Girolamo N, Lloyd A, McCluskey P, et al. Increased expression of matrix metalloproteinases in vivo in scleritis tissue and in vitro in cultured human scleral fibroblasts. Am J Pathol 1997;150:653– 66. 7. Kitahara T, Hiromura K, Maezawa A, et al. Case of propylthiouracil-induced vasculitis associated with antineutrophil cytoplasmic antibody (ANCA); review of literature. Clin Nephrol 1997;47:336 – 40. 8. Messmer EM, Foster SC. Vasculitic peripheral ulcerative keratitis. Surv Ophthalmol 1999;43:379 –96. 9. Sangwan VS, Merchant A, Sainz de la Maza M, Foster CS.
10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21.
Leukocyte adhesion molecule expression in scleritis. Arch Ophthalmol 1998;116:1476 – 80. Gedde SJ, Augsburger JJ. Posterior scleritis as a fundus mass. Ophthalmic Surg 1994;25:119 –21. Perri P, Mazzeo V, De Palma P, et al. Posterior scleritis: ultrasound findings in two cases. Ophthalmologica 1998; 212(Suppl 1):110 –2. Beiran I, Scharf J, Miller B. Role of computerized tomography in the differential diagnosis of posterior scleritis. Ann Ophthalmol 1999;31:254 – 6. Akpek EK, Uy HS, Christen W, et al. Severity of episcleritis and systemic disease association. Ophthalmology 1999;106: 729 –31. Watson PG, Bovey E. Anterior segment fluorescein angiography in the diagnosis of scleral inflammation. Ophthalmology 1985;92:1–11. Flower RW, Fryczkowski AW, McLeod DS. Variability in choriocapillaris blood flow distribution. Invest Ophthalmol Vis Sci 1995;36:1247–58. Aydin P, Akova YA, Kadayifc¸ ilar S. Anterior segment indocyanine green angiography in scleral inflammation. Eye 2000; 14:211–5. Yannuzzi LA, Slakter JS, Sorenson JA, et al. Digital indocyanine green videoangiography and choroidal neovascularization. Retina 1992;12(3):191–223. Watson PG, Hayreh SS. Scleritis and episcleritis. Br J Ophthalmol 1976;60(3):163–91. Sainz de la Maza M, Jabbur NS, Foster CS. An analysis of therapeutic decision for scleritis. Ophthalmology 1993;100: 1372– 6. Heiligenhaus A, Schilling M, Lung E, Steuhl KP. Ultrasound biomicroscopy in scleritis. Ophthalmology 1998; 105:527–34. Nguyen QD, Foster CS. Scleral patch graft in the management of necrotizing scleritis. Int Ophthalmol Clin 1999;39: 109 –31.
1763
Scleral fibroblasts produce type I, type III, and type VIII collagen fibers that intercede with elastin fibers.1 This particular organization of collagen fibers thus confers to scleral tissue its rigidity. When scleral tissue reaches maturity after the first years of life, it loses its elasticity and cannot be stretched. Scleral fibroblasts have a very low metabolic requirement, which is supplied mainly by vascularized adjacent structures such as the episclera and the choroid. Scleral tissue does not contain any blood vessels, with the exception of perforating arteries that cross through the sclera to reach the ciliary body and choroid. Vascularization of the anterior segment of the eye depends on the anterior ciliary arteries and on the long posterior ciliary arteries,2 which subsequently subdivide into capillaries. In contrast to retinal vessels, the superficial and deep plexi of episcleral Originally received: March 1, 2002. Accepted: March 5, 2003. Manuscript no. 220166. From Jules Gonin Eye Hospital, University of Lausanne, Lausanne, Switzerland. Presented in part at the Association for Research in Vision and Ophthalmology meeting, Fort Lauderdale, Florida, May 1999. The authors have no financial interest in any product, drug, instrument, or piece of equipment mentioned in this article. Proprietary interest: none. Correspondence and reprint requests to Yan Guex-Crosier, MD, Jules Gonin Eye Hospital, 15 av. de France, CH-1004, Lausanne, Switzerland.
1756
© 2003 by the American Academy of Ophthalmology Published by Elsevier Inc.
vessels consist of fenestrated capillaries. These histologic characteristics give to the scleral vessels their high permeability to fluorescein. In systemic diseases, inflammation occurs primarily around blood vessels.3– 8 Focal zones of inflammation are characterized by cellular infiltration and local edema, resulting from exudation of proteins from fenestrated capillaries. During systemic diseases, inflammatory mediators are produced in the proximity of episcleral and scleral vessels, and upregulation of leukocyte adhesion molecule expression has also been observed.9 The extent of inflammation depends, obviously, on the extent of the disease. The local liberation of large amounts of inflammatory mediators in scleritis can also affect adjacent tissues, such as choroidal vessels and orbital structures, and can result in a secondary myositis. Orbital involvement is more frequently seen in posterior scleritis, in which ultrasonography and computerized tomography help in diagnosis by demonstrating a thickened posterior sclera.10 –12 The poor stretching capacities of collagen bundles and the high prevalence of pain receptors within scleral tissues are responsible for the intense pain that is pathognomonic of scleritis. When the anterior segment of the eye is involved, intense ocular redness can be observed. Anterior scleritis is typically defined as simple, nodular, or necrotizing scleritis, with or without surrounding inflammation.1 Scleromalacia perforans is a specific ISSN 0161-6420/03/$–see front matter doi:10.1016/S0161-6420(03)00567-0
Guex-Crosier and Durig 䡠 Anterior Segment ICG in Scleritis scleral disease that results from severe vascular occlusion and vasculitis and occurs primarily in subjects with longstanding histories of rheumatoid arthritis. Depending on the tissue layer involved, two conditions have been clinically described: episcleritis and scleritis. Focal ocular redness is described classically as bright red in episcleritis, and bluish or purple– blue in scleritis. In scleritis, both superficial and deep episcleral vessels are dilated and do not bleach after topical 10% phenylephrine application. Furthermore, pain is more intense in scleritis than in episcleritis. A mild to severe photophobia associated with tearing can be present with either condition, and mild anterior uveitis can be observed. However, in some cases, the differential diagnosis between episcleritis and scleritis can be difficult. Differentiation of episcleritis from scleritis is of the utmost importance because, in most cases, episcleritis, in contrast to scleritis, is a self-limited disease that is rarely associated with systemic complications.13 Anterior segment fluorescein angiograms have already been used to characterize ocular inflammation in episcleritis and scleritis.14 But because fluorescein is not a proteinbound molecule, after injection, it diffuses freely through the fenestrated capillaries. The presence of both fenestrated and nonfenestrated capillaries of the conjunctiva and episclera is responsible for a patchy leakage during the early phase of anterior segment angiography. During the first transit of dye, the superficial vascular plexus is filled, dilated vessels can be seen, and rapid filling of the vessels can be observed both in episcleritis and scleritis. Unfortunately, the rapid superficial leakage of dye that occurs even in noninflamed tissues in the presence of fenestrated capillaries precludes good visualization of the deep layers. The patchy leakage of conjunctiva and diffuse leakage of the nonfenestrated capillaries of episclera impede a good visualization of the inflamed zone during the late phase of angiography. The recent development of infrared cameras for posterior pole angiography has allowed for the use of indocyanine green (ICG) angiography in clinical investigations. ICG dye has its peak excitation (805 nm) and peak fluorescence spectra in the near-infrared region, where the pigment epithelium, melanin, macular xanthophylls, and choroidal pigment are relatively transparent. These properties allow deeper tissular penetration. The ICG molecule has some unique properties that allow the study of protein exudation. ICG has a higher molecular weight than does fluorescein, and ⬎98% of the molecule is bound by proteins. In contrast, fluorescein is dissolved primarily as a free molecule in plasma. During fluorescein angiography of the posterior pole of the eye, fluorescein leakage is impeded by the blood– brain barrier, a barrier that is present also in nonfenestrated capillaries of the retina. Thus, in external ocular structures, fluorescein leaks spontaneously from fenestrated capillaries of noninflamed tissues, producing a patchy leakage.14 Thus, the ICG molecule is almost completely protein bound in the blood, rendering it relatively impermeable to fenestrated capillaries (such as the choriocapillaries). The dye does not leak extensively from fenestrations of choroidal vessels. An increase of permeability resulting in focal leakage can be observed in ocular inflammation. Therefore,
Table 1. Watson’s Criteria of Episcleritis versus Scleritis Episcleritis Color Pain 10% phenylephrine test
Red Slight Blanching of the superficial episcleral and conjunctival vessels Narrow bright slit Inner reflection of the beam examination light seem undisturbed, whereas outer reflection seem displaced forward
Scleritis Purple Severe No blanching of the scleral tissue Distortion and tortuousity of both anterior and posterior lines
the fluorescence of the ICG molecule through infrared wavelength allows visualization of deep scleral leakage, whereas fluorescein is rapidly blocked. These properties of ICG suggest that images obtained from external green angiographies could be useful in differentiating episcleritis from scleritis. The aim of this study was to analyze the pattern of anterior segment ICG angiography in scleral inflammations such as episcleritis and scleritis.
Subjects and Methods From November 1997 to March 2000, 20 consecutive subjects presenting clinical diseases compatible with the diagnosis of episcleritis or scleritis were involved in a prospective study. The subjects were seen at the Jules Gonin Eye Outpatient Clinic. Clinical evaluation of all subjects was performed by the same clinician, a senior fellow at the Jules Gonin Eye Hospital (YGC). The diagnosis of episcleritis compared to scleritis was made on a clinical basis according to the following four clinical criteria, suggested by P. Watson1 (Table 1): 1. The ocular pain (scored from 0 –10) according to an analogical scale, 0 corresponding to the absence of pain and 10 to the highest degree of pain that the subject can imagine. 2. The 10% phenylephrin test. The use of 10% epinephrine produces a vasoconstriction of the inflamed superficial episcleral vessels. Persistence of inflammation can be seen in scleral vessels (positive ⫽ 1, negative ⫽ 0). 3. The episcleritis has in daylight a red color, whereas scleritis seems blue or necrotic in necrotizing scleritis (red ⫽ 0, purple ⫽ 1). 4. In episcleritis, only the outer reflection line of the slit-lamp beam seems distorted, and the posterior line seems undisturbed. In scleritis, distortion and tortuosity of both anterior and posterior lines are present (absence of posterior distortion ⫽ 0, presence of posterior distortion ⫽ 1). Each of these criteria was scored as 0 if negative (or compatible with an episcleritis) and 1 if positive, except for pain, which was scored from 0 to 10. The overall score consisted of the sum of the four criteria. A global score greater than or equal to four was considered to result of a scleritis. A score of less than four was considered to correspond to an episcleritis. This analogical scale was given because the diagnosis of episcleritis compared to scleritis may be difficult in some borderline cases. Once the diagnosis of episcleritis compared to scleritis was established, subjects were investigated further according to the following protocol. While a bolus of 10% fluorescein was injected, external frames were obtained of the
1757
Ophthalmology Volume 110, Number 9, September 2003 Table 2. Scoring of Exudation during the Late Phase of External Indocyanine Green Angiography (30 – 45 minutes) Score
Exudation
0 1 2 3
None Slight, barely visible Moderate Severe, larges zones of hyperfluorescence
anterior segment of the eye with a Topcon TRC 50 Digital Imaging System Imagenet (Topcon Corporation, Tokyo, Japan). A bolus of 25 mg of ICG (Cardiogreen, Becton Dickinson, Franklin Lakes, NJ), diluted in 7.5 ml of 0.9% NaCl, was then injected, and frames were obtained at the following three time points. During the early phase (up to 3 minutes after dye injection), frames were taken from the inflamed part of the eye. During the intermediate phase (10 –12 minutes), nine pictures were taken from the 9 primary eye positions. The zone corresponding to the site of acute inflammation in episcleritis or scleritis was compared with the same portion of the eye in the controlateral eye. During the late phase (30 – 45 minutes), nine additional frames were taken. Because ICG is highly protein bound, “hot spots” corresponding to dye leakage exudation may be visible during the late phase of external ICG angiography. ICG leakage, which reflects the amount of protein exudation, was scored from zero to three in a masked analysis by a neutral observer (JD) who did not know the clinical diagnosis. The following scale was used to grade dye leakage: 0 ⫽ no exudation, 1 ⫽ mild exudation, 2 ⫽ moderate exudation, and 3 ⫽ severe
exudation (Table 2). A standard clinical workup was performed with all subjects to detect any possible underlying systemic disease. For statistical analysis, the Fisher exact test was used to compare leakage scores in the two groups (episcleritis and scleritis). Receiver operating characteristic (ROC) analysis was used to compare the sensitivity and the specificity of the ICG leakage; a leakage score of 1 or more was considered as suggestive of the presence of a scleritis. Systat software (Systat Inc., Evanston, IL) was used for statistical analysis.
Results Twenty subjects (7 male, 13 female) presenting inflammations of the anterior segment of the eye that were compatible with either episcleritis or scleritis were enrolled in a prospective open study from November 1997 to March 2000. The mean age of the subjects was 43 ⫾ 15 years. Two additional subjects were investigated according to the same protocol as negative controls; one presented with a conjunctival injection secondary to ocular rosacea, and the other was suffering from acute anterior HLA-B27-related uveitis. The clinical inflammation score is summarized in Table 3; subjects presenting overall clinical scores of less than four were considered as presenting episcleritis. On a clinical basis, 13 subjects had anterior scleritis and 7 had episcleritis. Among subjects with scleritis, 7 had the nodular form, 5 had the diffuse form and 1 had scleromalacia perforans. Seven subjects had episcleritis: 4 nodular and 3 simple.
Table 3. Clinical Score and Indocyanine Green (ICG) Score in Episcleritis and Scleritis Color Blue (score ⴝ 1 if positive)
10% Phenylephrine Test (score ⴝ 1 if positive)
Distorted Posterior Line (score ⴝ 1 if positive)
Overall Clinical Score (score 0–14)
ICG Leakage score (0–3)
Type
Pain (score 0–10)
Nodular episcleritis
1.5
0
0.5
0
2
0
Nodular episcleritis Nodular episcleritis Diffuse episcleritis Diffuse episcleritis Diffuse episcleritis Diffuse episcleritis Nodular scleritis
3 1.5 1.5 2 2 1 4
0 0 0 0 0 0 1
0 0 0 0 0.5 0 1
0 1 1 0 0 0 1
3 2.5 2.5 2 2.5 1 7
1 0 0 0 0 0 1
10
1
1
1
13
3
4 7 6 3 5
1 1 1 0 1
1 1 1 0.5 1
1 1 1 0.5 1
7 10 9 4 8
2 3 1 1 3
Diffuse scleritis
6
1
1
1
9
3
Diffuse scleritis
10
1
1
1
13
1
0 0.5 1 1
1 0.5 1 0
1 0.5 1 1
5 6.5 12 4
1 2 2 3
Nodular scleritis Nodular Nodular Nodular Nodular Nodular
scleritis scleritis scleritis scleritis scleritis
Diffuse scleritis Diffuse scleritis Diffuse scleritis Necrotizing scleritis
3 5 9 2
Patients with overall clinical scores of inflammation under four were considered as presenting episcleritis.
1758
Medical Workup Serum rheumatoid factor — — (case 1) — — — — Anticardiolipid positive Anticardiolipid positive Neurosarcoidosis — — Rheumatoid arthritis Anticardiolipid (case 2) Wegener ANCA positive — — — Systemic lupus erythematosus (case 3)
Guex-Crosier and Durig 䡠 Anterior Segment ICG in Scleritis
Figure 1. A–D, External indocyanine green angiography of the anterior segment. The subject presents a simple anterior episcleritis (case 1). During the early phase (A), a rapid filling of the vessels is seen. No leakage is visible during the intermediate phase (B) and late phase (score leakage ⫽ 0) (C). The image in the late phase is comparable to the controlateral eye (D). E, F, Image capture 50 seconds (E) and 10 minutes (F) after fluorescein injection.
All subjects were investigated by anterior segment ICG angiography. External angiographies of the eye were scored from zero to three according to the classification shown in Table 2. Only one out of 7 subjects with episcleritis showed a slight ICG leakage (a score of one), whereas all subjects with scleritis had ICG leakage scores of one or more (P ⫽ 0.0005, Fisher exact test). ROC analysis showed that for a cutoff value of one, the sensitivity of the test was 100% and the specificity was 85.7%. Severe ICG leakage (an angiographic score greater than three) was seen only in subjects presenting with anterior scleritis associated with a systemic disease, such as rheumatoid arthritis or sarcoidosis. All subjects presenting scleritis scores of two or more were treated with systemic steroids or immunosuppressive drugs. A clear decrease in ICG dye leakage was seen after such treatment and disappearance of symptoms. In all subjects enrolled in the early, intermediate, and late phases, angiograms were obtained of the controlateral eyes. None of the controlateral eyes showed ICG exudation, except for one subject with a severe bilateral scleritis associated with rheumatoid arthritis. One subject had severe scleromalacia perforans in the right eye, which was associated with systemic lupus erythematosus; this case is described in detail below.
Case Reports Case 1: Simple Anterior Episcleritis A 30-year-old white woman presented with a 10-day history of ocular pain in the temporal portion of the left eye (pain score of 1.5 in the analogic scale). Visual acuity was 10/10 in each eye. No cell or flare was present in either eye. The ocular pressure was 12 mmHg in each eye. A slight red scleral injection (clinical score ⫽ 0) was present in the nasal part of the left sclera but disappeared after application of 10% phenylephrine drops (clinical score ⫽ 0). No distortion of the posterior line could be seen by slit-lamp
examination (clinical score ⫽ 0). The overall clinical score was 1.5. Funduscopy was normal in both eyes. Results of complete medical workup were normal. The ocular pain gradually abated after therapy with topical diclofenac sodium application 4 times a day and topical 0.1% fluorometholone 3 times a day. Anterior segment ICG angiography did not reveal any hot spots during the late time point frames (leakage score ⫽ 0) (Fig 1).
Case 2: Simple Anterior Scleritis A 55-year-old man presented with a superotemporal lesion in the left eye. Pain had been present for 4 days (clinical pain score ⫽ 6). Slit-lamp examination of the anterior segment revealed the presence of a purple lesion (clinical score ⫽ 1), both the anterior and posterior lines of the slit-lamp beam were distorted (clinical score ⫽ 1), a slight rupture of the blood– ocular barrier, with a 2⫹ Tyndall effect, but no visible cell. The lesion did not disappear after application of 10% phenylephrine (clinical score ⫽ 1). The sum of the clinical scores was nine, corresponding to a scleritis diagnosis. A systemic laboratory workup revealed a positive rheumatoid factor. Rapid filling of episcleral and scleral vessels could be observed during the early frames of fluorescein angiography, and 2 hot spots were clearly visible during the intermediate and late phases of the external ICG angiography (leakage score ⫽ 3) (Fig 2C), whereas no exudation was present in fluorescein angiography (Fig 2E, F). Therapy consisted of hourly topical 1% prednisolone acetate and 4 times daily diclofenac sodium. Systemic nimesulide, a specific cox-2 inhibitor, was also administered. Systemic and topical therapies were progressively tapered, and ultimately stopped, 2 months later. After the resolution of the apparent inflammation, external ICG angiography was repeated, at which time no leakage was observed (Fig 3).
Case 3: Necrotizing Scleritis A 68-year-old white woman presented severe pain in the right eye. She had a necrotizing scleritis in the temporal region with ulcer-
1759
Ophthalmology Volume 110, Number 9, September 2003
Figure 2. A–D, External indocyanine green (ICG) angiography of the anterior segment in a subject presenting a simple anterior scleritis (case 2). During the early phase, scleral and episcleral vessels are filled with ICG. During the intermediate phase at 10 minutes (B) and late phase at 30 minutes (C), progressive leakage of ICG appears (hot spot; score leakage ⫽ 3), whereas no leakage is seen in the controlateral eye in the late phase (D). E, F, The external fluorescein angiography.
ation of the sclera at the limbus. At that time, she was treated with 4 mg/day of methylprednisolone for lupus alopecia. Ocular pain was scored as two on the analogic scale. The necrotic area seemed dark blue (clinical score ⫽ 1) Visual acuity was limited to 5/60 in the right eye with S-1, C-2 at 140°. A round scleral ulceration with a black aspect (clinical score ⫽ 1) of 5.4 mm in diameter, 3.8 mm away from the limbus, was present in the temporal region of the right eye and corresponded to scleral necrosis (clinical score ⫽ 1; scleral necrosis) (Fig 4); this was surrounded by irregular dilated vessels. A 1⫹ Tyndall with 1⫹ cell was present in the right eye. The subject had a 2⫹ nuclear sclerosis, and the fundus revealed inferior choroidal thickening adjacent to the ora serrata. In the left eye, the subject had 3/10 visual acuity with S-1.5, C-1 at 90°. The anterior segment was without inflammation; tonometry was 13 mmHg in each eye. Systemic workup revealed anti SS-A (Ro) antibodies with 26 UI/ml, and negative anti-SS-B (Le). Tests for anti-DNA antibodies were negative, and the rheumatoid factor was negative. Anterior segment angiography revealed the presence of a large avascular zone that corresponded to the necrotic sclera. A delay in scleral perfusion was present in the inferotemporal region of the eye. ICG angiography of the anterior segment revealed a 3⫹ leakage score of the vessels in the inferotemporal region (Fig 5).
Discussion ICG external angiography seems useful in the diagnosis of episcleritis compared to scleritis. This small pilot study clearly demonstrated a difference in dye leakage between episcleritis and scleritis. ICG has been widely used as a tracer to measure organ perfusion15 and in the investigation of inflammation of the superficial and deep plexi in scleritis.16,17 Fluorescein dye has been very useful in the investigation of retinal diseases, because the blood– ocular barrier
1760
is impermeable to fluorescein dye unless inflammation is present. In the outer section of the eye, the capillary vessels are fenestrated. When fluorescein is used to detect vascular abnormalities in episcleritis or scleritis, a rapid filling of the vessels is seen during angiography of the outer segment of the eye, but the dye leaks very rapidly out of the episcleral plexus. Investigation of the anterior segment of the eye by external ICG angiography was necessary to show that leakage is almost specific to deep scleral inflammation and was almost absent in episcleritis. This probably can be explained by the different degrees of severity of the inflammatory process in episcleritis and in scleritis. The latter is a more severe disease that can lead to vision-threatening complications such as keratitis, secondary glaucoma, uveitis, cataract, optic neuritis, macular edema, subretinal mass, annular ciliochoroidal detachment, serous retinal detachment, and even perforation of the globe.18,19 ICG was shown to be useful to quantify the severity of inflammation because the amount of dye leakage reflects the degree of protein exudation during the inflammatory process. In healthy eyes (or controlateral eyes, in this study), no dye leakage is present during the intermediate and late phases of angiograms. Episcleritis is primarily a self-limited disease that can resolve spontaneously within 2 to 3 weeks, whereas scleritis is a more severe process that can lead to visual loss or scleral perforation. Although these two entities can usually be clinically differentiated according to the criteria previously defined by Watson,1 differentiating between severe episcleritis and slight scleritis may be difficult. Slight dye leakage was seen in only one out of 7 episcleritis cases, whereas it was present in all scleritis cases. The
Guex-Crosier and Durig 䡠 Anterior Segment ICG in Scleritis
Figure 3. A–D, External indocyanine green angiography performed in case 2, after 2 months of local and systemic therapy. No more leakage appears during the intermediate (B) and late (C) phases. During the late stage, the image is comparable to the controlateral eye (D).
severity of dye leakage was particularly high in scleritis cases presenting overlying systemic diseases (Table 3). This test will probably be useful in the management of severe scleritis cases that need to be treated by systemic immunosuppressive therapy (such as cyclophosphamide or methotrexate), but this point needs to be verified by larger studies. Our aim was to demonstrate that ICG dye leakage is characteristic of scleritis and is almost absent in episcleritis (ICG leakage score ⫽ 0 in 6 out of 7 subjects) and in normal eyes. This point could be largely verified (P ⫽ 0.0005, Fisher exact test). With a cutoff value of one for the leakage score, the sensitivity is approximately 100% and the specificity is 85.7% (ROC analysis). The use of standard time points during ICG angiography is helpful to compare dye leakage during the intermediate and late phases. In all subjects with severe dye leakage, hot spots could be seen in the blue–purple zones of severe inflammation. Protein exudation that was detected by external ICG angiography reflects the changes in reflectivity that could be observed by ultrasound biomicroscopy in severe scleritis.20 Necrotizing scleritis is a severe form of scleritis in which an area of necrosis is surrounded by acutely congested episcleral vessels.21 Ocular or systemic complications are
observed in ⬎60% of such cases.18 The extent of the disease can clearly be delimited by external ICG angiography. The necrotic area is surrounded by zones of hyperpermeability of vascular episcleral channels (Fig 5). In necrotizing scleritis, the activity of the inflammation seems to be very low using the slit-lamp examination (clinical score of four in case 3); a recent progression of the zone of necrosis is a sign that an increase in immunosuppressive therapy is necessary. A clear dye leakage was visible in the anterior segment ICG angiographies (ICG dye score of three). This high score reflects the persistence of an acute inflammatory process in the scleral and episcleral vessels. In this case, the immunosuppressive therapy should be increased to avoid further progression of the zone of necrosis. In 2 subjects with scleritis, a decrease of ICG dye exudation was clearly visible after recovery from scleritis. Monitoring of the severity of the disease by measuring protein exudation by external ICG angiography could be a good method for the follow-up of the disease and the tapering of antiinflammatory therapy. This point could be verified in a larger group of subjects. Three subjects (negative controls) were also tested: two had severe ocular rosacea and one, with HLA-B27 uveitis, presented intense ocular pain during an acute onset of
1761
Ophthalmology Volume 110, Number 9, September 2003
Figure 4. Anterior necrotizing scleritis in a 68-year-old white woman presenting a systemic lupus erythematosus. The choroid is visible (black area).
anterior uveitis. No dye leakage was seen in any of these three subjects. The leakage of dye during ICG external angiographies seems to be specific to scleral inflammation
but not to ocular inflammation. ICG angiography of the external segment of the eye thus seems to be very useful in grading the intensity of ocular inflammation in scleritis.
Figure 5. A–D, External indocyanine green (ICG) angiography in case 3 presenting a necrotizing scleritis. During the early phase (A), the zones of necrosis seem hypofluorescent. In the adjacent zones, vascular filling can be seen, and some hot spots of dye leakage appear in the late phase (score leakage ⫽ 3) (C). No leakage is visible in the controlateral eye (D). E, F, Fluorescein angiography shows a large zone of hypofluorescence (necrosis). The adjacent zone (where inflamed vessels are seen in ICG picture) seems hypofluorescent (ischemia).
1762
Guex-Crosier and Durig 䡠 Anterior Segment ICG in Scleritis External ICG angiography thus seems to be a valuable tool for the evaluation of disease activity in episcleritis and scleritis. Acknowledgments. The authors thank Guy van Melle, MD, Institute for Social and Preventive Medicine, Lausanne, for statistical assistance.
References 1. Watson P. Diseases of the sclera and episclera. In: Tasman W, Jaeger EA, eds. Duane’s Clinical Ophthalmology, Vol. 2, 19th ed. Philadelphia: JB Lippincott, 1995:1– 45. 2. Meyer PA. Patterns of blood flow in episcleral vessels studied by low-dose fluorescein videoangiography. Eye 1988;2(pt 5): 533– 46. 3. Hoang-Xuan T, Foster CS, Rice BA. Scleritis in relapsing polychondritis. Response to therapy. Ophthalmology 1990;97: 892– 8. 4. Sainz de la Maza M, Foster CS, Jabbur NS. Scleritis associated with systemic vasculitic diseases. Ophthalmology 1995; 102:687–92. 5. Denk PO, Thiel HJ, Zierhut M. Episcleritis and scleritis. An overview of modern diagnostic and therapeutic concepts [in German]. Klin Monatsbl Augenheilkd 1997;211:140 –50. 6. Di Girolamo N, Lloyd A, McCluskey P, et al. Increased expression of matrix metalloproteinases in vivo in scleritis tissue and in vitro in cultured human scleral fibroblasts. Am J Pathol 1997;150:653– 66. 7. Kitahara T, Hiromura K, Maezawa A, et al. Case of propylthiouracil-induced vasculitis associated with antineutrophil cytoplasmic antibody (ANCA); review of literature. Clin Nephrol 1997;47:336 – 40. 8. Messmer EM, Foster SC. Vasculitic peripheral ulcerative keratitis. Surv Ophthalmol 1999;43:379 –96. 9. Sangwan VS, Merchant A, Sainz de la Maza M, Foster CS.
10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21.
Leukocyte adhesion molecule expression in scleritis. Arch Ophthalmol 1998;116:1476 – 80. Gedde SJ, Augsburger JJ. Posterior scleritis as a fundus mass. Ophthalmic Surg 1994;25:119 –21. Perri P, Mazzeo V, De Palma P, et al. Posterior scleritis: ultrasound findings in two cases. Ophthalmologica 1998; 212(Suppl 1):110 –2. Beiran I, Scharf J, Miller B. Role of computerized tomography in the differential diagnosis of posterior scleritis. Ann Ophthalmol 1999;31:254 – 6. Akpek EK, Uy HS, Christen W, et al. Severity of episcleritis and systemic disease association. Ophthalmology 1999;106: 729 –31. Watson PG, Bovey E. Anterior segment fluorescein angiography in the diagnosis of scleral inflammation. Ophthalmology 1985;92:1–11. Flower RW, Fryczkowski AW, McLeod DS. Variability in choriocapillaris blood flow distribution. Invest Ophthalmol Vis Sci 1995;36:1247–58. Aydin P, Akova YA, Kadayifc¸ ilar S. Anterior segment indocyanine green angiography in scleral inflammation. Eye 2000; 14:211–5. Yannuzzi LA, Slakter JS, Sorenson JA, et al. Digital indocyanine green videoangiography and choroidal neovascularization. Retina 1992;12(3):191–223. Watson PG, Hayreh SS. Scleritis and episcleritis. Br J Ophthalmol 1976;60(3):163–91. Sainz de la Maza M, Jabbur NS, Foster CS. An analysis of therapeutic decision for scleritis. Ophthalmology 1993;100: 1372– 6. Heiligenhaus A, Schilling M, Lung E, Steuhl KP. Ultrasound biomicroscopy in scleritis. Ophthalmology 1998; 105:527–34. Nguyen QD, Foster CS. Scleral patch graft in the management of necrotizing scleritis. Int Ophthalmol Clin 1999;39: 109 –31.
1763