- Email: [email protected]
Nonfluorescent Malignant Melanoma of the Choroid Diagnosed with the Radioactive Phosphorus Uptake Test
N O N F L U O R E S C E N T MALIGNANT MELANOMA O F T H E CHOROID DIAGNOSED W I T H T H E RADIOACTIVE P H O S P H O R U S UPTAKE TEST JERRY A. SHIELDS, M.D., WILLIAM H. ANNESLEY, JR., M.D., AND JOSEPH A. TOTINO,
M.D.
Philadelphia, Pennsylvania lems except for myopia until June 1973, when she noted the gradual onset of blurred vision in her right eye. Seen by her local ophthalmologist in August 1973, she had normal intraocular pressures and a refractive error of — 2.0O sphere, both eyes. There was a pigmented macular lesion in the right eye. She was referred to the Retina Service of the Wills Eye Hospital for further evaluation and intravenous fluorescein angiography. The patient had recently found that she could see better with her right eye if she removed her glasses. Her visual acuity was R.E. : 6/60 with correction and 6/9 without correc tion when a +2.00 spherical lens was placed in front of her correction, and L.E. : 6/24 without correction and 6/6 with correction. The results of her physical examination, intraocular pressures, and slit-lamp ex aminations were normal. Ophthalmoscopy of the right eye revealed a round pigmented macular lesion that measured l.S disk diameters in size and about 1 mm in elevation. There were no clinically apparent changes in the overlying retinal pigment epithelium or sensory retina, but the lesion was surrounded by a yellow halo. Perimetry studies performed with the Topcon perimeter using We studied the case of an extremely small 4 mm, 1 mm, and % mm white tests objects with illumination intensity 4 revealed no defect corre choroidal melanoma that failed to show sig sponding to this lesion. Intravenous fluorescein angi nificant fluorescence or to produce a visual ography was performed by rapidly injecting 5 ml field defect, but which had a positive 32 P of 10% fluorescein sodium (Fluorescite) intrave nously. There was no significant fluorescence of the test. It illustrated the great accuracy of the lesion in either the early or late angiograms. 32 P test, when properly performed, in diag Because no field defect was demonstrated and the nosing small malignant intraocular tumors. 13 lesion did not fluoresce, she was followed for about Pathologic study of the eye demonstrated a five months and no significant change was noted in the diameter of the lesion. Because it appeared to be possible cause for nonfluorescence of a cho slightly more elevated, however, the patient was roidal melanoma that had been suggested,1* referred for fluorescein angiography and a " P test. The patient was examined on March 13,1974. Her but not previously documented histologically. visual acuity was unchanged but was subjectively As a result, this case may add to our under worse. The macular lesion was unchanged (Fig. 1), standing of why choroidal melanomas usu although some observers felt that it was slightly more elevated. Once again, no visual field defect ally demonstrate fluorescence. could be demonstrated, and the lesion showed no significant fluorescence (Fig. 2). B-scan ultrasound CASE REPORT demonstrated a tumor pattern with "choroidal exca A 57-year-old white woman had no ocular prob- vation"" (Fig. 3). An incisional **P test was done by the routine technique using two instruments.1*'" From the Oncology Unit, Retina Service, Wills With the Geiger-Müller instrument," the uptake Eye Hospital, Philadelphia, Pennsylvania. This study was 125%, and 92% with the solid state semiconduc was supported in part by the Retina Research and tor detector." This is an unequivocally positive Development Foundation, Philadelphia, and the test.»"» A number of consultants examined the patient, Lions Club of Pennsylvania. Reprint requests to Jerry A. Shields, M.D., On and most agreed that the lesion was a small malignant cology Unit, Retina Service, Wills Eye Hospital, melanoma. After we listed the various alternatives, Philadelphia, PA 19130. including simple observation of the lesion, photo634
Several recent reports have presented the difficulties that may occur in the clinical diagnosis of malignant melanomas of the uveal tract. 1-4 The use of indirect ophthalmoscopy and certain ancillary tests appears to alleviate some of these diagnostic prob lems.5 Two of the more widely publicized ancillary tests are intravenous fluorescein angiography and the radioactive phosphorus uptake test ( 3 2 P ) . Most reports indicate that choroidal melanomas usually show a rather characteristic pattern of fluorescence with the former procedure.6"8 Occasional reports, however, have indicated that such tumors may sometimes fail to fluoresce.8'9 The 3ZP test has recently been reintroduced into oph thalmology and is now considered to be a most helpful diagnostic adjunct.10"12
VOL. 79, NO. 4
NONFLUORESCENT MALIGNANT MELANOMA
635
Fig. 1 (Shields, Annesley, and Totino). Right eye with pigmented macular lesion of about 1.5 disk diameter with surrounding yellow halo.
Fig. 3 (Shields, Annesley, and Totino). B-scan ultrasonogram showing lesion in posterior pole measuring 1 mm in elevation (arrow). The dark area at the base of the tumor represents choroidal excavation.
Fig. 2 (Shields, Annesley, and Totino). Top, Fluorescein angiogram during arteriovenous phase with normal choroidal flush over the tumor and slight hypofluorescence of the yellow halo. Bottom, Late angiogram (about 12 minutes) with disap pearance of the choroidal flush and no late staining of the tumor.
coagulation, or enucleation, the patient elected to have her eye enucleated. Gross findings—The specimen consisted of a firm right eye measuring 25 X 25 χ 24 mm with 2.5 mm of optic nerve attached. The clear cornea measured 12 X 10 mm. No transillumination shadow was dem onstrated. The sectioned globe revealed an ovalshaped mass in the macula measuring 2 mm in di ameter and 1 mm in elevation. It was pigmented centrally. The periphery of the lesion, adjacent to the normal choroid, was yellow (Fig. 4). Microscopic findings—The examination showed that a basophilic tumor arose from the choroid in the macular region (Fig. 5). It showed moderate pigmentation, particularly in the superficial portion of the lesion (Fig. 6). The overlying retinal pigment epithelium (RPE) was intact (Fig. 7). The choriocapillaris was slightly attenuated due to encroach ment by the tumor. The tumor consisted mainly of closely arranged spindle-shaped, slightly pigmented cells with spindle-shaped nuclei and prominent nu cleoli (Fig. 8). In the superficial peripheral portion of the tumor, the cells were larger and had clear cytoplasm (Fig. 9), that was mildly positive with the oil red 0 stain. The sensory retina over the tumor
636
AMERICAN JOURNAL OF OPHTHALMOLOGY
Fig. 4 (Shields, Annesley, and Totino). Section of gross globe with small pigmerrted tumor in pos terior pole. Note the light color of the periphery of the tumor. was normal, with no destruction of the photoreceptor cells (Fig. 10). The histologie diagnosis was malignant melanoma of the choroid, predominantly spindle B-cell type, with areas of lipoidal degeneration.
APRIL, 1975
ing the unusual visual symptoms. There was no demonstrable visual field defect. Perhaps if more refined techniques, such as static perimetry, had been utilized, a field defect could have been found.17 Histologically, however, the normal retina over lying the lesion correlated with the normal field. A normal visual field in a histologically proven malignant melanoma of the choroid is extremely rare. In contrast, even benign choroidal nevi may sometimes have associ ated visual field defects.17'18 The perimetry study, therefore, has definite drawbacks in differentiating benign from malignant cho roidal tumors. There have been previous reports of mela nomas that failed to fluoresce. Nadel, O'Connor, and Lincoff9 reported a nonfluorescent choroidal melanoma in which the lack of fluorescence was attributed to hemorrhage overlying the tumor. Pettit and associates8 reported two nonfluorescent cho roidal melanomas. One was due to hemor-
DISCUSSION
This case illustrates several important clinical points about small malignant mela nomas of the choroid, especially in the pa tient's presenting symptoms, the lack of a visual field defect and the failure of the lesion to fluoresce. It also emphasizes the great accuracy of the 3"P test, when properly performed, in diagnosing small malignant tumors of the posterior uvea. With regard to her symptoms, the patient presented with progressive unilateral hyperopia that was corrected by removing her minus correction or by adding a plus cor rection. This has been observed in central serous chorioretinopathy where an intact sensory retina is elevated by serous subretinal fluid. In most malignant melanomas of the choroid, the overlying R P E and photoreceptors are destroyed. In this case, however, they were intact, possibly explain
Fig. 5. (Shields, Annesley, and Totino). Histo logie section with section through center of tumor. Note the small size of the lesion, as compared to the entire globe (hematoxylin-eosin).
VOL. 79, NO. 4
NONFLUORESCENT MALIGNANT MELANOMA
637
Fig. 6 (Shields, Annesley, and Totino). Tumor with intact RPE and more dense pigmentation in superficial portions of tumor (hematoxylin-eosin, X64).
rhage over the surface of the tumor and the second was presumably due to extensive pig mentation in the superficial portion of the tumor. Other reports have indicated that occasional nonfiuorescence of a melanoma may be because the tumor was anteriorly located and the fluorescein pattern could not be adequately observed or photographed.7'18 Hayreh, 14 in attempting to explain fluo rescein patterns, studied the histologie sec tions of 50 malignant melanomas and found that the R P E was degenerated or destroyed in every instance. One of us (J.A.S.) has examined a number of choroidal melanomas and has made similar observations. Hayreh speculated that one of the reasons for the fluorescence observed with choroidal mela nomas was this destruction of the RPE. Our case is unusual in that the malignant melanoma was associated with an intact and normal RPE. The fact that this lesion failed to fluoresce appears to confirm Hayreh's speculation. The nonfiuorescence may have been partly related to the more dense pig
mentation in the superficial portions of the tumor (Fig. 6), but this is unlikely, because we have seen many deeply pigmented mela nomas which showed typical fluorescence. Fluorescein angiography, although useful in
Fig. 7. (Shields, Annesley, and Totino). Super ficial portion of tumor showing intact RPE and only minimal encroachment of choriocapillaries by tumor (hematoxylin-eosin, X256).
638
AMERICAN JOURNAL OF OPHTHALMOLOGY
Fig. 8. (Shields, Annesley, and Totino). Highpower photomicrograph of central portion of tumor with typical spindle melanoma B-cells (hematoxylineosin, X640).
most cases, may have definite drawbacks in the diagnosis of malignant melanomas of the choroid, since they may occasionally show no fluorescence and benign simulating lesions may occasionally fluoresce.18 B-scan ultrasound demonstrated a solid echo corresponding to the lesion, associated with choroidal excavation. Coleman and as sociates15 pointed out that such a finding suggests a choroidal melanoma. We have performed B-scan ultrasound on more than ISO patients with choroidal tumors and agree that such a finding suggests a mela noma, but that it is not diagnostic. Many choroidal melanomas do not demonstrate this feature, and in such eases it is even more difficult to differentiate a small melanoma from a hemangioma or certain other lesions. Another interesting feature of this case was the surrounding yellow halo along the margins of the tumor. We have observed such halos on several occasions and have found that pigmented lesions having them are, for some reason, nonprogressive or slowly progressive. In this case, the halo
APRIL, 1975
corresponded to lipoidal degeneration of tumor cells in the superficial portions of the lesion. We do not know the significance of this finding. The oil red 0 stain was only mildly positive, probably because frozen sec tions were not available and the tissue had been fixed in formalin and embedded in paraffin. Perhaps the most important aspect of this case is the fact that the 32 P test was un equivocally positive, even before the lesion produced a visual field defect or visible fluorescence with angiography. This em phasizes the great accuracy of this procedure in diagnosing small malignant tumors of the choroid. False-positive 32 P tests are ex tremely rare, and a positive test is, there fore, highly suggestive of a malignant tu mor.10"13 We have found that false-negative results are also uncommon if great care is taken to accurately localize the lesion with indirect ophthalmoscopy and transillumination when the test is performed.11-13 This patient underwent an enucleation for !' T W . " ■si- r - - : v iSL*.
A.V -Jfa-- ■-- -■>
Vf i'
V<î-, «s
Fig. 9. (Shields, Annesley, and Totino). Highpower photomicrograph of superficial peripheral portion of tumor with large cells with clear cyto plasm. Note the intact RPE over the tumor (hematoxylin-eosin, X640).
VOL. 79, NO. 4
NONFLUORESCENT MALIGNANT MELANOMA
639
3 Fig. 10 (Shields, Annesley, and Totino). Photomicrograph of retina in macular region over the tumor with normal layers and intact photoreceptors (hematoxylineosin, X64).
a small malignant melanoma of the choroid. We do not recommend enucleation in all these cases, but rather offer the patient sev eral alternatives on the basis of our present knowledge of the prognosis of choroidal melanomas. When enucleation is performed for small melanomas, the overall prognosis is better than when enucleation is performed for large melanomas.20-21 Once the diag nosis of malignant melanoma of the choroid is clearly established, the choice between enucleation, photocoagulation, cryotherapy, and other approaches to eradicate the tumor should be made jointly by the physician, pa tient, and family. We feel that the 32 P test, when properly performed, is the most ac curate method now available for confirming the clinical suspicion of choroidal melanoma. SUMMARY
A 57-year-old white woman presented with a small malignant melanoma of the cho roid in the macular area of the right eye. The lesion failed to produce a visual field defect or to show fluorescence with angiography, but the radioactive phosphorus uptake ( 3 2 P) test was positive. Histologically, the retinal pigment epithelium was intact over the tumor and the sensory retina was normal. This unusual finding was the most likely explana tion for the normal visual field. The intact retinal pigment epithelium over this nonfluorescent melanoma suggests that the de struction of the retinal pigment epithelium occurring with most choroidal melanomas may partially explain why these tumors usually demonstrate fluorescence. This case
emphasizes the great accuracy of the 32 P test for diagnosing small malignant tumors of the choroid, even before other popular diagnostic modalities indicate the malignancy of the lesion. ACKNOWLEDGMENTS
We thank Jay L. Federman, M.D., and Ms. Thea Fischer for reviewing the manuscript; Ramon L. Font, M.D., and Merlyn Rodrigues, M.D., for as sistance with the pathology ; Donald L. Morozin and Terranee L. Tomer for photographic assistance; and Ms. Jean Douglas for editorial help with the manuscript. REFERENCES
1. Ferry, A. P. : Lesions mistaken for malignant melanoma of the posterior uvea. Arch. Ophthalmol. 72:463, 1964. 2. Shields, J. A., and Zimmerman, L. E. : Lesions simulating malignant melanoma of the posterior uvea. Arch. Ophthalmol. 89:466,1973. 3. Zimmerman, L. E. : Problems in the diagnosis of malignant melanoma of the choroid and ciliary body. The Arthur J. Bedell Lecture, 1972. Am. J. Ophthalmol. 75:917,1973. 4. Shields, J. A., Green, W. R., and McDonald, P. R. : Uveal pseudomelanoma due to posttraumatic pigmentary migration. Arch. Ophthalmol. 89:519, 1973. 5. Shields, J. A., and McDonald, P. R. : Improve ments in the diagnosis of posterior uveal melanomas. Arch. Ophthalmol. 86:466,1973. 6. Norton, E. W. D., Smith, J. L., Curtin, V. T., and Justice, J., Jr. : Fluorescein fundus photography. An aid in the differential diagnosis of posterior ocu lar lesions. Trans. Am. Acad. Ophthalmol. OtolaJ-yngol. 68:755,1964. 7. Edwards, W. C, Leyden, W. E., and McDonald, R., Jr. : Fluorescein angiography of malignant mela noma of the choroid. Am. J. Ophthalmol. 68:797, 1969. 8. Pettit, T. H., Barton, A., Foos, R. Y., and Christensen, R. Ë. : Fluorescein angiography of ma lignant melanomas. Arch. Ophthalmol. 83:27, 1970.
640
AMERICAN JOURNAL OF OPHTHALMOLOGY
9. Nadel, A., O'Connor, P., and Lincoff, H. : Nonfluoresccnt melanoma. Am. J. Ophthalmol. 70: 748,1970. 10. Hagler, W. S., Jarrett, W. H., II, and Hum phrey, W. T. : The radioactive phosphorus uptake test in the diagnosis of uveal melanoma. Arch. Oph thalmol. 83 :S48, 1970. 11. Shields, J. A., Carmichael, P. L., Federman, J. L., and Sarin, L. K. : An appraisal of 200 consecu tive 31P tests for uveal melanomas. Presented at the Association for Research in Vision and Ophthal mology meeting, Sarasota, Florida, April 25,1974. 12."Hagler, W. S., Jarrett, W. H., II, Schnauss, R. H., LaRose, J. H„ and Palms, J. M. : The diag nosis of malignant melanoma of the ciliary body or choroid. Use of the radioactive phosphorus uptake test. S. Med. J. 65 :49,1972. 13. Shields, J. A., Sarin, L. K., Federman, J. L., Men^heha-Manhart, O., and Carmichael, P. L. : Sur gical approach to the "P test for posterior uveal melanomas. Ophthalmol. Surg. 5:13, 1974. 14. Hayreh, S. S.: Choroidal melanomata. Fluorescein a'.iaiioaranhic a'xl histopathological study. Br. J. Ophthalmol. 54:145,1970. 15. Coleman, D. J., Abramson, D. H., Jack, R. L.,
APRIL, 1975
and Franzen, L. A. : Ultrasonic diagnosis of tumors of the choroid. Arch. Ophthalmol. 91:344, 1974. 16. Ruiz, R. S. : New radioactivity detection probe and sealer for phosphorus 32 testing of ocular lesions. Trans. Am. Acad. Ophthalmol. Otolaryngol. 76 :535, 1972. 17. Flindall, R. J., and Drance, S. M. : Visual field studies of benign choroidal melanomata. Arch. Ophthalmol. 81:41, 1969. 18. Shields, J. A., and Font, R. L. : Melanocyîoma of the choroid clinically simulating a malignant mela noma. Arch. Ophthalmol. 87:396, 1972. 19. Snyder, W. B., Allen L., and Frazier, O. : Fluorescein angiography of ocular tumors. Trans. Am. Acad. Ophthalmol. Otolaryngol. 71:820, 1967. 20. Flocks, M., Gerende, J. H., Zimmerman, L. E : The size and shape of malignant melanomas of the choroid and ciliary body in relation to the prognosis and histologie characteristics. A statistical study of 210 tumors. Trans. Am. Acad. Ophthalmol. Oto laryngol. 59:740,1955. 21. Warren, R. M. : Prognosis of malignant mel anomas of the choroid and ciliary body. In Biodi, F. (ed.) : Current Concepts in Ophthalmology, vol. 4. St. Louis, C. V. Mosby, 1974.
M.D.
Philadelphia, Pennsylvania lems except for myopia until June 1973, when she noted the gradual onset of blurred vision in her right eye. Seen by her local ophthalmologist in August 1973, she had normal intraocular pressures and a refractive error of — 2.0O sphere, both eyes. There was a pigmented macular lesion in the right eye. She was referred to the Retina Service of the Wills Eye Hospital for further evaluation and intravenous fluorescein angiography. The patient had recently found that she could see better with her right eye if she removed her glasses. Her visual acuity was R.E. : 6/60 with correction and 6/9 without correc tion when a +2.00 spherical lens was placed in front of her correction, and L.E. : 6/24 without correction and 6/6 with correction. The results of her physical examination, intraocular pressures, and slit-lamp ex aminations were normal. Ophthalmoscopy of the right eye revealed a round pigmented macular lesion that measured l.S disk diameters in size and about 1 mm in elevation. There were no clinically apparent changes in the overlying retinal pigment epithelium or sensory retina, but the lesion was surrounded by a yellow halo. Perimetry studies performed with the Topcon perimeter using We studied the case of an extremely small 4 mm, 1 mm, and % mm white tests objects with illumination intensity 4 revealed no defect corre choroidal melanoma that failed to show sig sponding to this lesion. Intravenous fluorescein angi nificant fluorescence or to produce a visual ography was performed by rapidly injecting 5 ml field defect, but which had a positive 32 P of 10% fluorescein sodium (Fluorescite) intrave nously. There was no significant fluorescence of the test. It illustrated the great accuracy of the lesion in either the early or late angiograms. 32 P test, when properly performed, in diag Because no field defect was demonstrated and the nosing small malignant intraocular tumors. 13 lesion did not fluoresce, she was followed for about Pathologic study of the eye demonstrated a five months and no significant change was noted in the diameter of the lesion. Because it appeared to be possible cause for nonfluorescence of a cho slightly more elevated, however, the patient was roidal melanoma that had been suggested,1* referred for fluorescein angiography and a " P test. The patient was examined on March 13,1974. Her but not previously documented histologically. visual acuity was unchanged but was subjectively As a result, this case may add to our under worse. The macular lesion was unchanged (Fig. 1), standing of why choroidal melanomas usu although some observers felt that it was slightly more elevated. Once again, no visual field defect ally demonstrate fluorescence. could be demonstrated, and the lesion showed no significant fluorescence (Fig. 2). B-scan ultrasound CASE REPORT demonstrated a tumor pattern with "choroidal exca A 57-year-old white woman had no ocular prob- vation"" (Fig. 3). An incisional **P test was done by the routine technique using two instruments.1*'" From the Oncology Unit, Retina Service, Wills With the Geiger-Müller instrument," the uptake Eye Hospital, Philadelphia, Pennsylvania. This study was 125%, and 92% with the solid state semiconduc was supported in part by the Retina Research and tor detector." This is an unequivocally positive Development Foundation, Philadelphia, and the test.»"» A number of consultants examined the patient, Lions Club of Pennsylvania. Reprint requests to Jerry A. Shields, M.D., On and most agreed that the lesion was a small malignant cology Unit, Retina Service, Wills Eye Hospital, melanoma. After we listed the various alternatives, Philadelphia, PA 19130. including simple observation of the lesion, photo634
Several recent reports have presented the difficulties that may occur in the clinical diagnosis of malignant melanomas of the uveal tract. 1-4 The use of indirect ophthalmoscopy and certain ancillary tests appears to alleviate some of these diagnostic prob lems.5 Two of the more widely publicized ancillary tests are intravenous fluorescein angiography and the radioactive phosphorus uptake test ( 3 2 P ) . Most reports indicate that choroidal melanomas usually show a rather characteristic pattern of fluorescence with the former procedure.6"8 Occasional reports, however, have indicated that such tumors may sometimes fail to fluoresce.8'9 The 3ZP test has recently been reintroduced into oph thalmology and is now considered to be a most helpful diagnostic adjunct.10"12
VOL. 79, NO. 4
NONFLUORESCENT MALIGNANT MELANOMA
635
Fig. 1 (Shields, Annesley, and Totino). Right eye with pigmented macular lesion of about 1.5 disk diameter with surrounding yellow halo.
Fig. 3 (Shields, Annesley, and Totino). B-scan ultrasonogram showing lesion in posterior pole measuring 1 mm in elevation (arrow). The dark area at the base of the tumor represents choroidal excavation.
Fig. 2 (Shields, Annesley, and Totino). Top, Fluorescein angiogram during arteriovenous phase with normal choroidal flush over the tumor and slight hypofluorescence of the yellow halo. Bottom, Late angiogram (about 12 minutes) with disap pearance of the choroidal flush and no late staining of the tumor.
coagulation, or enucleation, the patient elected to have her eye enucleated. Gross findings—The specimen consisted of a firm right eye measuring 25 X 25 χ 24 mm with 2.5 mm of optic nerve attached. The clear cornea measured 12 X 10 mm. No transillumination shadow was dem onstrated. The sectioned globe revealed an ovalshaped mass in the macula measuring 2 mm in di ameter and 1 mm in elevation. It was pigmented centrally. The periphery of the lesion, adjacent to the normal choroid, was yellow (Fig. 4). Microscopic findings—The examination showed that a basophilic tumor arose from the choroid in the macular region (Fig. 5). It showed moderate pigmentation, particularly in the superficial portion of the lesion (Fig. 6). The overlying retinal pigment epithelium (RPE) was intact (Fig. 7). The choriocapillaris was slightly attenuated due to encroach ment by the tumor. The tumor consisted mainly of closely arranged spindle-shaped, slightly pigmented cells with spindle-shaped nuclei and prominent nu cleoli (Fig. 8). In the superficial peripheral portion of the tumor, the cells were larger and had clear cytoplasm (Fig. 9), that was mildly positive with the oil red 0 stain. The sensory retina over the tumor
636
AMERICAN JOURNAL OF OPHTHALMOLOGY
Fig. 4 (Shields, Annesley, and Totino). Section of gross globe with small pigmerrted tumor in pos terior pole. Note the light color of the periphery of the tumor. was normal, with no destruction of the photoreceptor cells (Fig. 10). The histologie diagnosis was malignant melanoma of the choroid, predominantly spindle B-cell type, with areas of lipoidal degeneration.
APRIL, 1975
ing the unusual visual symptoms. There was no demonstrable visual field defect. Perhaps if more refined techniques, such as static perimetry, had been utilized, a field defect could have been found.17 Histologically, however, the normal retina over lying the lesion correlated with the normal field. A normal visual field in a histologically proven malignant melanoma of the choroid is extremely rare. In contrast, even benign choroidal nevi may sometimes have associ ated visual field defects.17'18 The perimetry study, therefore, has definite drawbacks in differentiating benign from malignant cho roidal tumors. There have been previous reports of mela nomas that failed to fluoresce. Nadel, O'Connor, and Lincoff9 reported a nonfluorescent choroidal melanoma in which the lack of fluorescence was attributed to hemorrhage overlying the tumor. Pettit and associates8 reported two nonfluorescent cho roidal melanomas. One was due to hemor-
DISCUSSION
This case illustrates several important clinical points about small malignant mela nomas of the choroid, especially in the pa tient's presenting symptoms, the lack of a visual field defect and the failure of the lesion to fluoresce. It also emphasizes the great accuracy of the 3"P test, when properly performed, in diagnosing small malignant tumors of the posterior uvea. With regard to her symptoms, the patient presented with progressive unilateral hyperopia that was corrected by removing her minus correction or by adding a plus cor rection. This has been observed in central serous chorioretinopathy where an intact sensory retina is elevated by serous subretinal fluid. In most malignant melanomas of the choroid, the overlying R P E and photoreceptors are destroyed. In this case, however, they were intact, possibly explain
Fig. 5. (Shields, Annesley, and Totino). Histo logie section with section through center of tumor. Note the small size of the lesion, as compared to the entire globe (hematoxylin-eosin).
VOL. 79, NO. 4
NONFLUORESCENT MALIGNANT MELANOMA
637
Fig. 6 (Shields, Annesley, and Totino). Tumor with intact RPE and more dense pigmentation in superficial portions of tumor (hematoxylin-eosin, X64).
rhage over the surface of the tumor and the second was presumably due to extensive pig mentation in the superficial portion of the tumor. Other reports have indicated that occasional nonfiuorescence of a melanoma may be because the tumor was anteriorly located and the fluorescein pattern could not be adequately observed or photographed.7'18 Hayreh, 14 in attempting to explain fluo rescein patterns, studied the histologie sec tions of 50 malignant melanomas and found that the R P E was degenerated or destroyed in every instance. One of us (J.A.S.) has examined a number of choroidal melanomas and has made similar observations. Hayreh speculated that one of the reasons for the fluorescence observed with choroidal mela nomas was this destruction of the RPE. Our case is unusual in that the malignant melanoma was associated with an intact and normal RPE. The fact that this lesion failed to fluoresce appears to confirm Hayreh's speculation. The nonfiuorescence may have been partly related to the more dense pig
mentation in the superficial portions of the tumor (Fig. 6), but this is unlikely, because we have seen many deeply pigmented mela nomas which showed typical fluorescence. Fluorescein angiography, although useful in
Fig. 7. (Shields, Annesley, and Totino). Super ficial portion of tumor showing intact RPE and only minimal encroachment of choriocapillaries by tumor (hematoxylin-eosin, X256).
638
AMERICAN JOURNAL OF OPHTHALMOLOGY
Fig. 8. (Shields, Annesley, and Totino). Highpower photomicrograph of central portion of tumor with typical spindle melanoma B-cells (hematoxylineosin, X640).
most cases, may have definite drawbacks in the diagnosis of malignant melanomas of the choroid, since they may occasionally show no fluorescence and benign simulating lesions may occasionally fluoresce.18 B-scan ultrasound demonstrated a solid echo corresponding to the lesion, associated with choroidal excavation. Coleman and as sociates15 pointed out that such a finding suggests a choroidal melanoma. We have performed B-scan ultrasound on more than ISO patients with choroidal tumors and agree that such a finding suggests a mela noma, but that it is not diagnostic. Many choroidal melanomas do not demonstrate this feature, and in such eases it is even more difficult to differentiate a small melanoma from a hemangioma or certain other lesions. Another interesting feature of this case was the surrounding yellow halo along the margins of the tumor. We have observed such halos on several occasions and have found that pigmented lesions having them are, for some reason, nonprogressive or slowly progressive. In this case, the halo
APRIL, 1975
corresponded to lipoidal degeneration of tumor cells in the superficial portions of the lesion. We do not know the significance of this finding. The oil red 0 stain was only mildly positive, probably because frozen sec tions were not available and the tissue had been fixed in formalin and embedded in paraffin. Perhaps the most important aspect of this case is the fact that the 32 P test was un equivocally positive, even before the lesion produced a visual field defect or visible fluorescence with angiography. This em phasizes the great accuracy of this procedure in diagnosing small malignant tumors of the choroid. False-positive 32 P tests are ex tremely rare, and a positive test is, there fore, highly suggestive of a malignant tu mor.10"13 We have found that false-negative results are also uncommon if great care is taken to accurately localize the lesion with indirect ophthalmoscopy and transillumination when the test is performed.11-13 This patient underwent an enucleation for !' T W . " ■si- r - - : v iSL*.
A.V -Jfa-- ■-- -■>
Vf i'
V<î-, «s
Fig. 9. (Shields, Annesley, and Totino). Highpower photomicrograph of superficial peripheral portion of tumor with large cells with clear cyto plasm. Note the intact RPE over the tumor (hematoxylin-eosin, X640).
VOL. 79, NO. 4
NONFLUORESCENT MALIGNANT MELANOMA
639
3 Fig. 10 (Shields, Annesley, and Totino). Photomicrograph of retina in macular region over the tumor with normal layers and intact photoreceptors (hematoxylineosin, X64).
a small malignant melanoma of the choroid. We do not recommend enucleation in all these cases, but rather offer the patient sev eral alternatives on the basis of our present knowledge of the prognosis of choroidal melanomas. When enucleation is performed for small melanomas, the overall prognosis is better than when enucleation is performed for large melanomas.20-21 Once the diag nosis of malignant melanoma of the choroid is clearly established, the choice between enucleation, photocoagulation, cryotherapy, and other approaches to eradicate the tumor should be made jointly by the physician, pa tient, and family. We feel that the 32 P test, when properly performed, is the most ac curate method now available for confirming the clinical suspicion of choroidal melanoma. SUMMARY
A 57-year-old white woman presented with a small malignant melanoma of the cho roid in the macular area of the right eye. The lesion failed to produce a visual field defect or to show fluorescence with angiography, but the radioactive phosphorus uptake ( 3 2 P) test was positive. Histologically, the retinal pigment epithelium was intact over the tumor and the sensory retina was normal. This unusual finding was the most likely explana tion for the normal visual field. The intact retinal pigment epithelium over this nonfluorescent melanoma suggests that the de struction of the retinal pigment epithelium occurring with most choroidal melanomas may partially explain why these tumors usually demonstrate fluorescence. This case
emphasizes the great accuracy of the 32 P test for diagnosing small malignant tumors of the choroid, even before other popular diagnostic modalities indicate the malignancy of the lesion. ACKNOWLEDGMENTS
We thank Jay L. Federman, M.D., and Ms. Thea Fischer for reviewing the manuscript; Ramon L. Font, M.D., and Merlyn Rodrigues, M.D., for as sistance with the pathology ; Donald L. Morozin and Terranee L. Tomer for photographic assistance; and Ms. Jean Douglas for editorial help with the manuscript. REFERENCES
1. Ferry, A. P. : Lesions mistaken for malignant melanoma of the posterior uvea. Arch. Ophthalmol. 72:463, 1964. 2. Shields, J. A., and Zimmerman, L. E. : Lesions simulating malignant melanoma of the posterior uvea. Arch. Ophthalmol. 89:466,1973. 3. Zimmerman, L. E. : Problems in the diagnosis of malignant melanoma of the choroid and ciliary body. The Arthur J. Bedell Lecture, 1972. Am. J. Ophthalmol. 75:917,1973. 4. Shields, J. A., Green, W. R., and McDonald, P. R. : Uveal pseudomelanoma due to posttraumatic pigmentary migration. Arch. Ophthalmol. 89:519, 1973. 5. Shields, J. A., and McDonald, P. R. : Improve ments in the diagnosis of posterior uveal melanomas. Arch. Ophthalmol. 86:466,1973. 6. Norton, E. W. D., Smith, J. L., Curtin, V. T., and Justice, J., Jr. : Fluorescein fundus photography. An aid in the differential diagnosis of posterior ocu lar lesions. Trans. Am. Acad. Ophthalmol. OtolaJ-yngol. 68:755,1964. 7. Edwards, W. C, Leyden, W. E., and McDonald, R., Jr. : Fluorescein angiography of malignant mela noma of the choroid. Am. J. Ophthalmol. 68:797, 1969. 8. Pettit, T. H., Barton, A., Foos, R. Y., and Christensen, R. Ë. : Fluorescein angiography of ma lignant melanomas. Arch. Ophthalmol. 83:27, 1970.
640
AMERICAN JOURNAL OF OPHTHALMOLOGY
9. Nadel, A., O'Connor, P., and Lincoff, H. : Nonfluoresccnt melanoma. Am. J. Ophthalmol. 70: 748,1970. 10. Hagler, W. S., Jarrett, W. H., II, and Hum phrey, W. T. : The radioactive phosphorus uptake test in the diagnosis of uveal melanoma. Arch. Oph thalmol. 83 :S48, 1970. 11. Shields, J. A., Carmichael, P. L., Federman, J. L., and Sarin, L. K. : An appraisal of 200 consecu tive 31P tests for uveal melanomas. Presented at the Association for Research in Vision and Ophthal mology meeting, Sarasota, Florida, April 25,1974. 12."Hagler, W. S., Jarrett, W. H., II, Schnauss, R. H., LaRose, J. H„ and Palms, J. M. : The diag nosis of malignant melanoma of the ciliary body or choroid. Use of the radioactive phosphorus uptake test. S. Med. J. 65 :49,1972. 13. Shields, J. A., Sarin, L. K., Federman, J. L., Men^heha-Manhart, O., and Carmichael, P. L. : Sur gical approach to the "P test for posterior uveal melanomas. Ophthalmol. Surg. 5:13, 1974. 14. Hayreh, S. S.: Choroidal melanomata. Fluorescein a'.iaiioaranhic a'xl histopathological study. Br. J. Ophthalmol. 54:145,1970. 15. Coleman, D. J., Abramson, D. H., Jack, R. L.,
APRIL, 1975
and Franzen, L. A. : Ultrasonic diagnosis of tumors of the choroid. Arch. Ophthalmol. 91:344, 1974. 16. Ruiz, R. S. : New radioactivity detection probe and sealer for phosphorus 32 testing of ocular lesions. Trans. Am. Acad. Ophthalmol. Otolaryngol. 76 :535, 1972. 17. Flindall, R. J., and Drance, S. M. : Visual field studies of benign choroidal melanomata. Arch. Ophthalmol. 81:41, 1969. 18. Shields, J. A., and Font, R. L. : Melanocyîoma of the choroid clinically simulating a malignant mela noma. Arch. Ophthalmol. 87:396, 1972. 19. Snyder, W. B., Allen L., and Frazier, O. : Fluorescein angiography of ocular tumors. Trans. Am. Acad. Ophthalmol. Otolaryngol. 71:820, 1967. 20. Flocks, M., Gerende, J. H., Zimmerman, L. E : The size and shape of malignant melanomas of the choroid and ciliary body in relation to the prognosis and histologie characteristics. A statistical study of 210 tumors. Trans. Am. Acad. Ophthalmol. Oto laryngol. 59:740,1955. 21. Warren, R. M. : Prognosis of malignant mel anomas of the choroid and ciliary body. In Biodi, F. (ed.) : Current Concepts in Ophthalmology, vol. 4. St. Louis, C. V. Mosby, 1974.