Scleritis as an Initial Manifestation of Choroidal Malignant Melanoma Eng-Yiat Yap, MD, Dennis M. Robertson, MD, Helmut Buettner, MD Purpose: The purpose of this article is to present the unusual circumstances in which malignant melanoma of the choroid can masquerade as scleritis, thus confounding its diagnosis. Methods: Three cases of plaque-like malignant melanomas of the choroid are reported that, on initial examination, had features of scleritis. The events leading to their eventual correct diagnosis also are presented. Results: In each case, there was ocular pain, blurred vision, anterior chamber and/ or vitreous cavity cellular reaction, and an exudative retinal detachment associated with an ill-defined, relatively flat variably pigmented choroidal mass. In all three cases, the inflammatory component responded promptly to corticosteroid treatment and was accompanied by visual improvement. In two eyes, shrinkage of the choroidal mass accompanied the corticosteroid treatment, lending support to a working diagnOSis of scleritis. By demonstrating expansion of the choroidal masses, examination of sequential fundus photographs influenced the decision to enucleate the eyes for presumed malignant choroidal melanoma. Conclusion: Clinicians should be alert to the circumstances in which malignant melanomas of the choroid can masquerade as scleritis. Careful evaluation by ophthalmoscopy, ultrasonography, fundus photography, and subsequent sequential examination is necessary to arrive at the correct diagnosis. Ophthalmology 1992;99: 1693-1697
The majority of malignant choroidal melanomas can be diagnosed accurately with information obtained by indirect ophthalmoscopy, ultrasonography, and fluorescein angiography. In some cases, however, differentiating a melanoma from simulating lesions can be difficult. Posterior scleritis sometimes mimics a choroidal melanoma, \-4 and eyes with this condition occasionally have been removed because of the mistaken diagnosis of malignant melanoma. Although these cases have heightened awareness that scleritis can appear as a choroidal melaOriginally received: March 19. 1992. Revision accepted: June II. 1992. From the Department of Ophthalmology. Mayo Clinic and Mayo Foundation. Rochester. Supported in part by the Higher Manpower Development Plan of the Ministry of Health, Singapore (Dr. Yap). by Research to Prevent Blindness, Inc, New York. New York. NIH grant EY06253. Bethesda. Maryland. and by the Mayo Foundation. Rochester. Minnesota. Reprint requests to Dennis M. Robertson, MD. Department of Ophthalmology, Mayo Clinic. Rochester. MN 55905.
noma, the reverse situation in which a melanoma can appear with manifestations of a scleritis, and thus escape recognition, is less well known. We report three cases of plaque-like malignant melanomas of the choroid that initially had features ofsc1eritis.
Case Reports Case 1. A 58-year-old man was first seen by us in September 1968 with a 5-month history of 3 recurring episodes of retrobulbar discomfort, redness, and blurred vision in his right eye. Orally administered corticosteroids had been associated with resolution of symptoms in each of the first two episodes. The most recent symptoms, present for several weeks, prompted his referral with the diagnosis of a retinal detachment in his right eye. On examination, the visual acuity with -10.00 spherical correction in both eyes was 20/50 in the right eye and 20/25 in the left. There was a superior defect in the right visual field. The right eye was mildly proptotic with chemosis and scleral and
1693
Ophthalmology
Volume 99, Number 11, November 1992
episcleral injection. In addition to a small hypopyon, grayish white cells diffusely infiltrated the vitreous. There was an inferior serous retinal detachment with shifting fluid and a nonpigmented yellow-white vascularized choroidal lesion, 3 X 5 disc diameters in base dimension, beneath the retinal detachment inferior nasal to the disc. A diagnosis of malignant choroidal melanoma with secondary inflammation was made, and the patient was scheduled for enucleation. During the subsequent 3 days, while undergoing a preoperative medical evaluation, the proptosis receded and the pain lessened, as did the severity of the episcleritis and scleritis. The ocular media cleared, allowing a better view of a receding retinal detachment. The clinical improvement, together with the history of previous response to systemic corticosteroids, suggested that the findings better fit the diagnosis of primary scleritis rather than scleritis secondary to a choroidal melanoma. After daily treatment with 40 mg of orally administered prednisone, the cellular reaction in the anterior chamber resolved, the vitreous cells diminished, the retina reattached, the yellowwhite choroidal lesion shrunk, and concentric lines of pigmentation formed around its base. These findings were believed to support the diagnosis of scleritis. The oral prednisone was tapered and discontinued 2 months later. The patient returned a year later, 38 months after the initial presentation, relating that he had had 4 additional recurrent episodes of discomfort and redness of the eye. Each time, selfadministered steroids had resulted in prompt symptomatic improvement. The choroidal mass now measured approximately 8 X 8disc diameters, and there was pigmentation on its surface and along its border. The episcleritis, scleritis, and inferior exudative retinal detachment had returned, and visual acuity had decreased to 20/200. After treatment with orally administered corticosteroids, the choroidal mass lesion enlarged, and the eye was enucleated. Results of histopathologic examination showed a malignant melanoma of the choroid made up of epithelioid
Figure 1. Case 2. A, photograph shows marked injection of superior sclera and episclera of the left eye. B, 5 days after topically and systemically administered corticosteroids, the injection had decreased.
1694
and spindle B cells. Extrascleral extension of the tumor was noted, with inflammatory cells present in the posterior sclera and episclera. A general medical evaluation demonstrated lung and liver metastases, which caused the patient's death 2 months later. B-scan ultrasonography was not available during the time this patient was evaluated. Case 2. A 55-year-old man was first seen by us in April 1980. He gave a history of recent redness of his left eye associated with blurred central vision and loss of the temporal visual field. A chorioretinal scar had been described previously in this eye. Visual acuity was 20/20 in both eyes. The left upper lid was swollen, the globe was tender, and the vessels were injected in the superior episclera and sclera (Fig IA). There was a mild cellular reaction in the anterior chamber and vitreous. A plaquelike, shallowly elevated, gray choroidal mass lesion measuring approximately 9 X 9 disc diameters was present superonasally. Patches of pigment covered the surface of the lesion. It was surrounded by a serosanguineous retinal detachment (Fig 2A). Auorescein angiography showed a pattern of patchy hyperfluorescence and late staining of the choroidal mass but no intrinsic circulation. B-scan ultrasonography showed the presence of a shallowly elevated, solid choroidal lesion (Fig 3A). However, it did not show an echolucent zone behind the globe indicating retrobulbar edema. With the possible diagnosis of posterior scleritis, but not ruling out a choroidal melanoma, the patient was treated with 1% prednisolone acetate drops administered topically 4 times a day and with 80 mg prednisone given orally once a day. Five days later, all external signs of inflammation had disappeared (Fig I B), and the serosanguineous retinal detachment was resolving (Fig 2B). The prednisone was tapered, and, by June 1980, the subretinal hemorrhage and detachment had resolved. B-scan ultrasonography suggested that the choroidal mass lesion had decreased in thickness (Fig 3B).
Yap et al . Choroidal Malignant Melanoma On follow-up in October 1980, however, the choroidal mass had increased in both elevation and base diameter, now clearly favoring the diagnosis of malignant melanoma of the choroid (Figs 2C and 3C). The eye was enucleated, and the tumor proved to be a malignant choroidal melanoma comprised of spindle B
and epithelioid cells. Melanoma cells extensively infiltrated the sclera and also were found to extend into emissary channels. Large atypical inflammatory cells, representing plasma cells, plasmacytoid cells, and lymphocytes, were present in the episcleral connective tissue.
A
B
c Figure 2. Case 2. A, appearance of the choroida11esion and serosanguineous retinal detachment on initial presentation. B, choroidal lesion 5 days after treatment with corticosteroids. The serosanguineous retinal detachment had regressed, allowing a clearer view of choroidal detail. C, 6 months after initial presentation, the choroidal lesion had expanded.
Figure 3. Case 2. A, B-scan ultrasonogram shows an elevated, solid choroidal lesion with high internal reflectivity on initial presentation. B, Bscan ultrasonogram shows decrease in size of the choroidal lesion after treatment with corticosteroids. C, B-scan ultrasonogram shows increase in thickness of the choroidal lesion 6 months after initial presentation.
1695
Ophthalmology Volume 99, Number 11, November 1992 In February 1986, metastatic melanoma was discovered in the lung and liver. The patient died of metastatic disease 3 months later, approximately 6 years after his initial presentation. Case 3. A 36-year-old woman was referred to us in July 1989 with the diagnosis of posterior scleritis in the left eye, for which treatment with orally administered corticosteroids had been initiated 2 weeks earlier. Both eyes had normal vision and were quiet externally. White cells were present in the left posterior vitreous. A relatively flat, irregularly pigmented choroidal lesion, approximately 15 mm in base diameter, was visible temporal to the macula. There was an inferior temporal exudative retinal detachment with shifting fluid. B-scan ultrasonography showed irregular choroidal thickening with a maximum thickness of 2.4 mm but no echolucent zone behind the globe. The findings were believed to be consistent with a plaque-like melanoma, but posterior scleritis remained a possible diagnosis because of the inflammatory component that had recently been suppressed with corticosteroids. Over the ensuing 6 months, the choroidal lesion remained unchanged. Although there were two episodes of recurrent inflammation characterized by loss of central vision, vitreous cellular reaction, exudative retinal detachment involving the macula, and intraretinal exudates inferior to the macula, these symptoms and findings partially resolved spontaneously in the first episode and after long-acting corticosteroids were administered subconjunctivally during the second episode. By April 1990, 9 months after presentation, the retinal detachment had disappeared. There were only a few cells in the vitreous cavity. However, comparison of the sequential fundus photographs confirmed that the choroidal mass had expanded. The eye was enucleated with a presumptive diagnosis ofmalignant melanoma of the choroid. Results of histopathologic examination showed a malignant choroidal melanoma composed predominantly of spindle B cells with foci of epithelioid cells and sparse balloon cells. Periemissary invasion was recognized, but no intraretinal or extrascleral extension was seen. The patient remains well 2 years after enucleation.
Discussion Numerous reports have detailed the difficulty that clinicians sometimes face in differentiating malignant choroidal melanoma from other simulating lesions. Studies published between 1964 and 1973 showed that the suspected diagnosis of malignant melanoma of the choroid was incorrect for 5.6% to 19.7% of the enucleated eyes. 5- 8 More recently, the rate of misdiagnosis of malignant choroidal melanomas has decreased noticeably. In 1979, Robertson and Campbe1l 9 reported a misdiagnosis rate of 2.7%, and in 1990 the Collaborative Ocular Melanoma Study found a misdiagnosis rate of only 0.48% among 413 enucleated eyes.1O These data suggest that ophthalmologists are becoming more expert at accurately diagnosing choroidal melanoma, despite the presence of strict inclusion criteria for the Collaborative Ocular Melanoma StUdy. Despite growing experience and diagnostic advances, some mimicking choroidal lesions will continue to be difficult to differentiate from malignant melanoma. Inflammatory processes, for example, are known to mimic malignant melanoma. As early as 1957, Reese and Jones!! wrote that inflammatory processes comprised over one third of the conditions that mimicked melanoma.
1696
Fraunfelder and Watson! reported on a series of 30 eyes with a diagnosis of scleritis in which 1 case was mistakenly believed to be a melanoma. Feldon and coworkers 2 presented four case reports of brawny scleritis in which two eyes were mistakenly enucleated for suspected melanoma. Shields et al 3 found that posterior scleritis was mistaken for malignant melanoma in 6 of their 400 (1.5%) cases. Finger et al 4 reported a single case of posterior scleritis mimicking malignant melanoma. Whereas most ophthalmologists have become aware that posterior scleritis can mimic malignant uveal melanoma, little has been written about the reverse scenario, wherein malignant uveal melanoma can mimic a primary scleritis. In 1973, Zimmerman!2 commented that uveal melanomas occasionally produced clinically confusing complications, including uveitis and panophthalmitis. In 1979, Fraser and Font I3 reported that 22 of 450 eyes (4.9%) with uveal malignant melanoma initially presented with inflammation. Eight had an episcleritis (usually overlying a ciliary body melanoma) and the remaining 14 had some form of uveitis, endophthalmitis, or panophthalmitis. Those with panophthalmitis had necrotic choroidal melanomas. Our three patients with choroidal melanoma had ocular findings consistent with scleritis, which confounded an accurate diagnosis early in the course of the disease. Although in the first patient an initial diagnosis of malignant melanoma was made and enucleation was scheduled, the dramatic improvement of the retrobulbar discomfort, episcleritis and scleritis, anterior chamber and vitreous reaction, and proptosis after administration of systemic corticosteroids led to the mistaken diagnosis of posterior scleritis. The clinical condition waxed and waned in accordance with repeated corticosteroid treatment during a 3-year follow-up period before the eye was enucleated. The patient eventually died of lung and liver metastasis 3 years and 4 months after initial presentation. Our second patient also presented with clinical evidence of episcleritis and scleritis, cells in the anterior chamber and vitreous, and an exudative retinal detachment with choroidal thickening. The presence of subretinal blood was considered consistent with either recurrent scleritis or choroidal melanoma. Ultrasonography of the choroidal lesion was not considered diagnostic. After treatment with systemic corticosteroids, the inflammatory reaction regressed, and the choroidal lesion decreased in thickness. When the choroidal lesion was observed to expand on follow-up, the clinical diagnosis favored malignant melanoma, and enucleation was performed. Approximately 6 years after initial presentation, the patient died of metastatic disease. The third patient, a young woman, presented with ocular pain, a cellular reaction in the vitreous, and a secondary retinal detachment in association with an irregular, shallowly elevated choroidal lesion. Ultrasonography was inconclusive. A favorable response to corticosteroids supported the diagnosis of an inflammatory condition, although a high suspicion of choroidal melanoma remained. Frequent follow-up examinations, including color fundus
Yap et al . Choroidal Malignant Melanoma photography, fluorescein angiography, and ultrasonography, eventually showed expansion of the choroidal lesion, and the eye was enucleated for presumed malignant melanoma. At present, the patient is well. As demonstrated by these three cases, patients with malignant choroidal melanoma can have features of episcleritis or scleritis accompanied by pain, proptosis, anterior chamber and vitreous cellular reaction, and an exudative retinal detachment. All of these expressions can respond dramatically to corticosteroid treatment, which may confound the clinical picture. In each of these cases, the choroidal lesion was relatively flat or plaque-like. The thickest ultrasonographic tumor measurement in our third case was 2.4 mm. Choroidal lesions of this thickness may be difficult to profile with ultrasonography because too little tissue is available to produce diagnostic tissue reflectivity patterns. Benson et al 14 in 1979 and Benson l5 in 1988 listed features considered helpful in differentiating posterior scleritis from malignant melanoma. They concluded that A scan and B scan ultrasonography showing high internal reflectivity within the nodular elevations, low amplitude echoes or echolucent zone behind the globe identifying retrobulbar edema, and scleral thickening confined to an area of intense inflammation were practically diagnostic of scleritis and served to rule out primary and metastatic choroidal neoplasms. In neither of our two cases in which the eyes were studied ultrasonographically did we observe echolucent zones behind the globe identifying retrobulbar edema, yet histopathology demonstrated retrobulbar inflammatory cell infiltrates consistent with scleritis in one of these two cases. The absence of an echolucent zone behind the globe in this one case suggests that this ultrasonographic finding is either inconsistent or is dependent on a greater degree of inflammation than present in these cases. Proponents of fine needle biopsy might argue that the diagnosis in the second and third cases could have been made sooner if a fine needle biopsy of these plaque-like lesions had been performed. We continue to believe that the management of malignant uveal melanoma is not an emergency and that the observation interval in each of these cases was justified and preferable over fine needle biopsy. In addition to the possibility of causing local complications and retrieving material that may be difficult to interpret, the long-term effect on survival by fine needle biopsy of malignant uveal melanoma remains unknown. As demonstrated by these cases, eyes presenting with irregularly thickened choroidal mass lesions and evidence ofscleritis/episcleritis may contain plaque-like melanomas that can cause a secondary scleritis. If a diagnosis is not
reasonably certain on presentation, these eyes should be followed closely so that an accurate diagnosis can be made as early as possible. Ultrasonographic studies may help to identify features consistent with a primary scleritis. Serial fundus photography demonstrating expansion of the choroidal lesion may help in making the diagnosis of melanoma.
References I. Fraunfelder Fr, Watson PG. Evaluation of eyes enucleated
for scleritis. Br J Ophthalmol 1976;60:227-30. 2. Feldon SE, Sigelman J, Albert DM, Smith TR. Clinical manifestations of brawny scleritis. Am J Ophthalmol 1978;85:781-7. 3. Shields JA, Augsburger JJ, Brown GC, Stephens RF. The differential diagnosis of posterior uveal melanoma. Ophthalmology 1980;87:518-22. 4. Finger PT, Perry HD, Packer S, et al. Posterior scleritis as an intraocular tumour. Br J Ophthalmol 1990;74:121-2. 5. 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. 6. Blodi FC, Roy PE. The misdiagnosed choroidal melanoma. Can J Ophthalmol 1967;2:209-11. 7. Howard GM. Erroneous clinical diagnoses of retinoblastoma and uveal melanoma. Trans Am Acad Ophthalmol Otolaryngol 1969;73: 199-203. 8. Shields JA, Zimmerman LE. Lesions simulating malignant melanoma of the posterior uvea. Arch Ophthalmol 1973;89: 466-71. 9. Robertson DM, Campbell RJ. Errors in the diagnosis of malignant melanoma of the choroid. Am J Ophthalmol 1979;87:269-75. 10. The Collaborative Ocular Melanoma Study Group. Accuracy of diagnosis of choroidal melanomas in the Collaborative Ocular Melanoma Study. COMS report no. I. Arch Ophthalmol 1990; 108: 1268-73. II. Reese AB, Jones IS. The differential diagnosis of malignant melanoma of the choroid. Arch Ophthalmol 1957;58:47782. 12. Zimmerman LE. Problems in the diagnosis of malignant melanomas of the choroid and ciliary body. The 1972 Arthur J . Bedell Lecture. Am J Ophthalmol 1973;75:917-29. 13. Fraser DJ Jr, Font RL. Ocular inflammation and hemorrhage as initial manifestations of uveal malignant melanoma: incidence and prognosis. Arch Ophthalmol 1979;97: 1311-4. 14. Benson WE, Shields JA, Tasman W, Crandall AS. Posterior scleritis: a cause of diagnostic confusion. Arch Ophthalmo) 1979;97: 1482-6. 15. Benson WE. Posterior scleritis. Surv Ophthalmo) 1988;32: 297-316.
1697
The majority of malignant choroidal melanomas can be diagnosed accurately with information obtained by indirect ophthalmoscopy, ultrasonography, and fluorescein angiography. In some cases, however, differentiating a melanoma from simulating lesions can be difficult. Posterior scleritis sometimes mimics a choroidal melanoma, \-4 and eyes with this condition occasionally have been removed because of the mistaken diagnosis of malignant melanoma. Although these cases have heightened awareness that scleritis can appear as a choroidal melaOriginally received: March 19. 1992. Revision accepted: June II. 1992. From the Department of Ophthalmology. Mayo Clinic and Mayo Foundation. Rochester. Supported in part by the Higher Manpower Development Plan of the Ministry of Health, Singapore (Dr. Yap). by Research to Prevent Blindness, Inc, New York. New York. NIH grant EY06253. Bethesda. Maryland. and by the Mayo Foundation. Rochester. Minnesota. Reprint requests to Dennis M. Robertson, MD. Department of Ophthalmology, Mayo Clinic. Rochester. MN 55905.
noma, the reverse situation in which a melanoma can appear with manifestations of a scleritis, and thus escape recognition, is less well known. We report three cases of plaque-like malignant melanomas of the choroid that initially had features ofsc1eritis.
Case Reports Case 1. A 58-year-old man was first seen by us in September 1968 with a 5-month history of 3 recurring episodes of retrobulbar discomfort, redness, and blurred vision in his right eye. Orally administered corticosteroids had been associated with resolution of symptoms in each of the first two episodes. The most recent symptoms, present for several weeks, prompted his referral with the diagnosis of a retinal detachment in his right eye. On examination, the visual acuity with -10.00 spherical correction in both eyes was 20/50 in the right eye and 20/25 in the left. There was a superior defect in the right visual field. The right eye was mildly proptotic with chemosis and scleral and
1693
Ophthalmology
Volume 99, Number 11, November 1992
episcleral injection. In addition to a small hypopyon, grayish white cells diffusely infiltrated the vitreous. There was an inferior serous retinal detachment with shifting fluid and a nonpigmented yellow-white vascularized choroidal lesion, 3 X 5 disc diameters in base dimension, beneath the retinal detachment inferior nasal to the disc. A diagnosis of malignant choroidal melanoma with secondary inflammation was made, and the patient was scheduled for enucleation. During the subsequent 3 days, while undergoing a preoperative medical evaluation, the proptosis receded and the pain lessened, as did the severity of the episcleritis and scleritis. The ocular media cleared, allowing a better view of a receding retinal detachment. The clinical improvement, together with the history of previous response to systemic corticosteroids, suggested that the findings better fit the diagnosis of primary scleritis rather than scleritis secondary to a choroidal melanoma. After daily treatment with 40 mg of orally administered prednisone, the cellular reaction in the anterior chamber resolved, the vitreous cells diminished, the retina reattached, the yellowwhite choroidal lesion shrunk, and concentric lines of pigmentation formed around its base. These findings were believed to support the diagnosis of scleritis. The oral prednisone was tapered and discontinued 2 months later. The patient returned a year later, 38 months after the initial presentation, relating that he had had 4 additional recurrent episodes of discomfort and redness of the eye. Each time, selfadministered steroids had resulted in prompt symptomatic improvement. The choroidal mass now measured approximately 8 X 8disc diameters, and there was pigmentation on its surface and along its border. The episcleritis, scleritis, and inferior exudative retinal detachment had returned, and visual acuity had decreased to 20/200. After treatment with orally administered corticosteroids, the choroidal mass lesion enlarged, and the eye was enucleated. Results of histopathologic examination showed a malignant melanoma of the choroid made up of epithelioid
Figure 1. Case 2. A, photograph shows marked injection of superior sclera and episclera of the left eye. B, 5 days after topically and systemically administered corticosteroids, the injection had decreased.
1694
and spindle B cells. Extrascleral extension of the tumor was noted, with inflammatory cells present in the posterior sclera and episclera. A general medical evaluation demonstrated lung and liver metastases, which caused the patient's death 2 months later. B-scan ultrasonography was not available during the time this patient was evaluated. Case 2. A 55-year-old man was first seen by us in April 1980. He gave a history of recent redness of his left eye associated with blurred central vision and loss of the temporal visual field. A chorioretinal scar had been described previously in this eye. Visual acuity was 20/20 in both eyes. The left upper lid was swollen, the globe was tender, and the vessels were injected in the superior episclera and sclera (Fig IA). There was a mild cellular reaction in the anterior chamber and vitreous. A plaquelike, shallowly elevated, gray choroidal mass lesion measuring approximately 9 X 9 disc diameters was present superonasally. Patches of pigment covered the surface of the lesion. It was surrounded by a serosanguineous retinal detachment (Fig 2A). Auorescein angiography showed a pattern of patchy hyperfluorescence and late staining of the choroidal mass but no intrinsic circulation. B-scan ultrasonography showed the presence of a shallowly elevated, solid choroidal lesion (Fig 3A). However, it did not show an echolucent zone behind the globe indicating retrobulbar edema. With the possible diagnosis of posterior scleritis, but not ruling out a choroidal melanoma, the patient was treated with 1% prednisolone acetate drops administered topically 4 times a day and with 80 mg prednisone given orally once a day. Five days later, all external signs of inflammation had disappeared (Fig I B), and the serosanguineous retinal detachment was resolving (Fig 2B). The prednisone was tapered, and, by June 1980, the subretinal hemorrhage and detachment had resolved. B-scan ultrasonography suggested that the choroidal mass lesion had decreased in thickness (Fig 3B).
Yap et al . Choroidal Malignant Melanoma On follow-up in October 1980, however, the choroidal mass had increased in both elevation and base diameter, now clearly favoring the diagnosis of malignant melanoma of the choroid (Figs 2C and 3C). The eye was enucleated, and the tumor proved to be a malignant choroidal melanoma comprised of spindle B
and epithelioid cells. Melanoma cells extensively infiltrated the sclera and also were found to extend into emissary channels. Large atypical inflammatory cells, representing plasma cells, plasmacytoid cells, and lymphocytes, were present in the episcleral connective tissue.
A
B
c Figure 2. Case 2. A, appearance of the choroida11esion and serosanguineous retinal detachment on initial presentation. B, choroidal lesion 5 days after treatment with corticosteroids. The serosanguineous retinal detachment had regressed, allowing a clearer view of choroidal detail. C, 6 months after initial presentation, the choroidal lesion had expanded.
Figure 3. Case 2. A, B-scan ultrasonogram shows an elevated, solid choroidal lesion with high internal reflectivity on initial presentation. B, Bscan ultrasonogram shows decrease in size of the choroidal lesion after treatment with corticosteroids. C, B-scan ultrasonogram shows increase in thickness of the choroidal lesion 6 months after initial presentation.
1695
Ophthalmology Volume 99, Number 11, November 1992 In February 1986, metastatic melanoma was discovered in the lung and liver. The patient died of metastatic disease 3 months later, approximately 6 years after his initial presentation. Case 3. A 36-year-old woman was referred to us in July 1989 with the diagnosis of posterior scleritis in the left eye, for which treatment with orally administered corticosteroids had been initiated 2 weeks earlier. Both eyes had normal vision and were quiet externally. White cells were present in the left posterior vitreous. A relatively flat, irregularly pigmented choroidal lesion, approximately 15 mm in base diameter, was visible temporal to the macula. There was an inferior temporal exudative retinal detachment with shifting fluid. B-scan ultrasonography showed irregular choroidal thickening with a maximum thickness of 2.4 mm but no echolucent zone behind the globe. The findings were believed to be consistent with a plaque-like melanoma, but posterior scleritis remained a possible diagnosis because of the inflammatory component that had recently been suppressed with corticosteroids. Over the ensuing 6 months, the choroidal lesion remained unchanged. Although there were two episodes of recurrent inflammation characterized by loss of central vision, vitreous cellular reaction, exudative retinal detachment involving the macula, and intraretinal exudates inferior to the macula, these symptoms and findings partially resolved spontaneously in the first episode and after long-acting corticosteroids were administered subconjunctivally during the second episode. By April 1990, 9 months after presentation, the retinal detachment had disappeared. There were only a few cells in the vitreous cavity. However, comparison of the sequential fundus photographs confirmed that the choroidal mass had expanded. The eye was enucleated with a presumptive diagnosis ofmalignant melanoma of the choroid. Results of histopathologic examination showed a malignant choroidal melanoma composed predominantly of spindle B cells with foci of epithelioid cells and sparse balloon cells. Periemissary invasion was recognized, but no intraretinal or extrascleral extension was seen. The patient remains well 2 years after enucleation.
Discussion Numerous reports have detailed the difficulty that clinicians sometimes face in differentiating malignant choroidal melanoma from other simulating lesions. Studies published between 1964 and 1973 showed that the suspected diagnosis of malignant melanoma of the choroid was incorrect for 5.6% to 19.7% of the enucleated eyes. 5- 8 More recently, the rate of misdiagnosis of malignant choroidal melanomas has decreased noticeably. In 1979, Robertson and Campbe1l 9 reported a misdiagnosis rate of 2.7%, and in 1990 the Collaborative Ocular Melanoma Study found a misdiagnosis rate of only 0.48% among 413 enucleated eyes.1O These data suggest that ophthalmologists are becoming more expert at accurately diagnosing choroidal melanoma, despite the presence of strict inclusion criteria for the Collaborative Ocular Melanoma StUdy. Despite growing experience and diagnostic advances, some mimicking choroidal lesions will continue to be difficult to differentiate from malignant melanoma. Inflammatory processes, for example, are known to mimic malignant melanoma. As early as 1957, Reese and Jones!! wrote that inflammatory processes comprised over one third of the conditions that mimicked melanoma.
1696
Fraunfelder and Watson! reported on a series of 30 eyes with a diagnosis of scleritis in which 1 case was mistakenly believed to be a melanoma. Feldon and coworkers 2 presented four case reports of brawny scleritis in which two eyes were mistakenly enucleated for suspected melanoma. Shields et al 3 found that posterior scleritis was mistaken for malignant melanoma in 6 of their 400 (1.5%) cases. Finger et al 4 reported a single case of posterior scleritis mimicking malignant melanoma. Whereas most ophthalmologists have become aware that posterior scleritis can mimic malignant uveal melanoma, little has been written about the reverse scenario, wherein malignant uveal melanoma can mimic a primary scleritis. In 1973, Zimmerman!2 commented that uveal melanomas occasionally produced clinically confusing complications, including uveitis and panophthalmitis. In 1979, Fraser and Font I3 reported that 22 of 450 eyes (4.9%) with uveal malignant melanoma initially presented with inflammation. Eight had an episcleritis (usually overlying a ciliary body melanoma) and the remaining 14 had some form of uveitis, endophthalmitis, or panophthalmitis. Those with panophthalmitis had necrotic choroidal melanomas. Our three patients with choroidal melanoma had ocular findings consistent with scleritis, which confounded an accurate diagnosis early in the course of the disease. Although in the first patient an initial diagnosis of malignant melanoma was made and enucleation was scheduled, the dramatic improvement of the retrobulbar discomfort, episcleritis and scleritis, anterior chamber and vitreous reaction, and proptosis after administration of systemic corticosteroids led to the mistaken diagnosis of posterior scleritis. The clinical condition waxed and waned in accordance with repeated corticosteroid treatment during a 3-year follow-up period before the eye was enucleated. The patient eventually died of lung and liver metastasis 3 years and 4 months after initial presentation. Our second patient also presented with clinical evidence of episcleritis and scleritis, cells in the anterior chamber and vitreous, and an exudative retinal detachment with choroidal thickening. The presence of subretinal blood was considered consistent with either recurrent scleritis or choroidal melanoma. Ultrasonography of the choroidal lesion was not considered diagnostic. After treatment with systemic corticosteroids, the inflammatory reaction regressed, and the choroidal lesion decreased in thickness. When the choroidal lesion was observed to expand on follow-up, the clinical diagnosis favored malignant melanoma, and enucleation was performed. Approximately 6 years after initial presentation, the patient died of metastatic disease. The third patient, a young woman, presented with ocular pain, a cellular reaction in the vitreous, and a secondary retinal detachment in association with an irregular, shallowly elevated choroidal lesion. Ultrasonography was inconclusive. A favorable response to corticosteroids supported the diagnosis of an inflammatory condition, although a high suspicion of choroidal melanoma remained. Frequent follow-up examinations, including color fundus
Yap et al . Choroidal Malignant Melanoma photography, fluorescein angiography, and ultrasonography, eventually showed expansion of the choroidal lesion, and the eye was enucleated for presumed malignant melanoma. At present, the patient is well. As demonstrated by these three cases, patients with malignant choroidal melanoma can have features of episcleritis or scleritis accompanied by pain, proptosis, anterior chamber and vitreous cellular reaction, and an exudative retinal detachment. All of these expressions can respond dramatically to corticosteroid treatment, which may confound the clinical picture. In each of these cases, the choroidal lesion was relatively flat or plaque-like. The thickest ultrasonographic tumor measurement in our third case was 2.4 mm. Choroidal lesions of this thickness may be difficult to profile with ultrasonography because too little tissue is available to produce diagnostic tissue reflectivity patterns. Benson et al 14 in 1979 and Benson l5 in 1988 listed features considered helpful in differentiating posterior scleritis from malignant melanoma. They concluded that A scan and B scan ultrasonography showing high internal reflectivity within the nodular elevations, low amplitude echoes or echolucent zone behind the globe identifying retrobulbar edema, and scleral thickening confined to an area of intense inflammation were practically diagnostic of scleritis and served to rule out primary and metastatic choroidal neoplasms. In neither of our two cases in which the eyes were studied ultrasonographically did we observe echolucent zones behind the globe identifying retrobulbar edema, yet histopathology demonstrated retrobulbar inflammatory cell infiltrates consistent with scleritis in one of these two cases. The absence of an echolucent zone behind the globe in this one case suggests that this ultrasonographic finding is either inconsistent or is dependent on a greater degree of inflammation than present in these cases. Proponents of fine needle biopsy might argue that the diagnosis in the second and third cases could have been made sooner if a fine needle biopsy of these plaque-like lesions had been performed. We continue to believe that the management of malignant uveal melanoma is not an emergency and that the observation interval in each of these cases was justified and preferable over fine needle biopsy. In addition to the possibility of causing local complications and retrieving material that may be difficult to interpret, the long-term effect on survival by fine needle biopsy of malignant uveal melanoma remains unknown. As demonstrated by these cases, eyes presenting with irregularly thickened choroidal mass lesions and evidence ofscleritis/episcleritis may contain plaque-like melanomas that can cause a secondary scleritis. If a diagnosis is not
reasonably certain on presentation, these eyes should be followed closely so that an accurate diagnosis can be made as early as possible. Ultrasonographic studies may help to identify features consistent with a primary scleritis. Serial fundus photography demonstrating expansion of the choroidal lesion may help in making the diagnosis of melanoma.
References I. Fraunfelder Fr, Watson PG. Evaluation of eyes enucleated
for scleritis. Br J Ophthalmol 1976;60:227-30. 2. Feldon SE, Sigelman J, Albert DM, Smith TR. Clinical manifestations of brawny scleritis. Am J Ophthalmol 1978;85:781-7. 3. Shields JA, Augsburger JJ, Brown GC, Stephens RF. The differential diagnosis of posterior uveal melanoma. Ophthalmology 1980;87:518-22. 4. Finger PT, Perry HD, Packer S, et al. Posterior scleritis as an intraocular tumour. Br J Ophthalmol 1990;74:121-2. 5. 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. 6. Blodi FC, Roy PE. The misdiagnosed choroidal melanoma. Can J Ophthalmol 1967;2:209-11. 7. Howard GM. Erroneous clinical diagnoses of retinoblastoma and uveal melanoma. Trans Am Acad Ophthalmol Otolaryngol 1969;73: 199-203. 8. Shields JA, Zimmerman LE. Lesions simulating malignant melanoma of the posterior uvea. Arch Ophthalmol 1973;89: 466-71. 9. Robertson DM, Campbell RJ. Errors in the diagnosis of malignant melanoma of the choroid. Am J Ophthalmol 1979;87:269-75. 10. The Collaborative Ocular Melanoma Study Group. Accuracy of diagnosis of choroidal melanomas in the Collaborative Ocular Melanoma Study. COMS report no. I. Arch Ophthalmol 1990; 108: 1268-73. II. Reese AB, Jones IS. The differential diagnosis of malignant melanoma of the choroid. Arch Ophthalmol 1957;58:47782. 12. Zimmerman LE. Problems in the diagnosis of malignant melanomas of the choroid and ciliary body. The 1972 Arthur J . Bedell Lecture. Am J Ophthalmol 1973;75:917-29. 13. Fraser DJ Jr, Font RL. Ocular inflammation and hemorrhage as initial manifestations of uveal malignant melanoma: incidence and prognosis. Arch Ophthalmol 1979;97: 1311-4. 14. Benson WE, Shields JA, Tasman W, Crandall AS. Posterior scleritis: a cause of diagnostic confusion. Arch Ophthalmo) 1979;97: 1482-6. 15. Benson WE. Posterior scleritis. Surv Ophthalmo) 1988;32: 297-316.
1697
