Search

A 32-year-old man with atypical retinitis pigmentosa developed bilateral neovascularization of the optic disk. One year after the diagnosis, vitreous hemorrhage occurred in one eye. Panretinal photocoagulation of this eye resulted in some regression of the neovascularization.

Prognostic indicators for the development of metastasis following proton beam irradiation of uveal melanomas were evaluated for 510 patients treated from 1975 to 1984. Thirty-three patients developed metastasis (6.5%) from 3 to 51 months following treatment. The primary site of metastasis was the liver in 28 cases (85%). Both demographic and clinical factors were considered. The three leading predictors of survival without metastasis after proton beam irradiation in order of importance were: (1) largest diameter of the tumor; (2) location of the anterior margin of the tumor; and (3) age at treatment. Worse prognosis was associated with largest tumor diameter greater than 15.0 mm, tumor involvement of the ciliary body and age at treatment older than 59 years.

Anatomical details in the ocular fundus can be seen with increased contrast when appropriate monochromatic illumination is used. This technique permits more accurate visualization and documentation than can be achieved with white light. This article reports the optimal spectral ranges for visualization of the different fundus structures. It describes and illustrates the appearance of individual structures and areas of the fundus under different spectral illuminations. There is a detailed discussion of the optical properties of the fundus layers that provides a background for interpreting the data and a baseline for examination of pathologic conditions.

Proton beam irradiation is one of the radiotherapeutic techniques currently used to manage uveal melanomas. Although this therapeutic modality has been in use for a decade and although nearly 500 patients have been so treated, there are only two published reports of the pathologic examination of these eyes. Key features found on pathologic examination of our patient's enucleated eye included vascular alterations in the tumor's blood supply, lymphocytic infiltrates, and lipoidal degeneration of tumor cells. The interval from therapy to enucleation in our patient was much longer than in the several previously reported cases, a factor allowing for the development of a more cumulative radiation effect on the tumor. The extent and degree of radiation-induced necrosis in our patient's tumor were more profound than in previous reports.

Four patients underwent cataract extraction with implantation of intraocular lenses. Preoperative ultrasonography did not include a comprehensive diagnostic B-scan analysis. Postoperative examination revealed uveal melanomas. The clinical presentations as well as the histopathologic findings of the two enucleated eyes are discussed. These cases emphasize the necessity for B-scan ultrasonography prior to cataract surgery when media opacity prevents adequate visualization of the fundus.

Enucleation was performed after proton treatment in 57 of 1006 (5.7%) uveal melanoma patients treated with proton beam therapy at the Harvard Cyclotron Laboratory between July 1975 and December 31, 1986. Only 2% of 99 patients with small tumors and 4% of 566 patients with intermediate size tumors underwent enucleation after treatment; 10% of 341 patients with large tumors lost the treated eye. No eyes were removed after 52 months, with 89% of enucleations performed during the first 36 months after treatment. Eye retention rates at 60 months were 89.1 +/- 3.0% for the entire group, and 97 +/- 3.7%, 92.7 +/- 3.1%, and 78.3 +/- 7.0% in patients with small, intermediate, and large tumors, respectively. Significantly greater enucleation rates were observed in patients with large tumors than in those with intermediate tumors (p = less than .0001), in patients with tumor height greater than 8 mm relative to those with tumors less than or equal to 8 mm, p = (less than .0001), with tumor diameter greater than 16 mm compared to less than or equal to 16 mm, (p = less than .0001), and with tumor involvement of the ciliary body compared to involvement of the choroid only (p = less than .0001). Possible strategies to decrease the likelihood of enucleation in patients at apparently increased risk of losing the eye after conservative therapy, that is, those with large tumors involving the ciliary body, might include a lower total dose, a more protracted treatment course, or a lower radiation dose and adjuvant treatment with chemotherapy and/or immunotherapy, with hyperthermia, or with other radiation sensitizers.

Although the underlying cause or causes of uveal melanoma have yet to be elucidated, important insights may be gained by examining the epidemiologic features of the disease. Uveal melanoma is an uncommon cancer with an incidence of only six cases per million population per year. It is most often diagnosed in the sixth decade and is somewhat more common in males. Apart from sporadic reports of family clusters, uveal melanoma is not considered an inherited disease. Whether some environmental exposure triggers the development of uveal melanoma remains an open question. Sunlight has been proposed as an environmental risk factor because sunlight is known to cause melanoma of the skin and both diseases are rare in nonwhite races. Unlike cutaneous melanoma, however, rates have not been increasing over time and do not vary by latitude. This paper evaluates the available evidence for sunlight and other potential risk factors for uveal melanoma, highlighting areas requiring further research.

Eighty-three choroidal neovascular membranes (CNVMs) associated with pathologic entities other than age-related macular degeneration (study group) were compared with 64 CNVMs associated with macular degeneration (comparison group). Nine CNVMs (11%) in the study group, compared with 28 CNVMs (44%) in the comparison group, were occult membranes. Of the remaining well-defined CNVMs, 16 (22%) in the study group, compared with 21 (58%) in the comparison group, were subfoveal. Of the well-defined subfoveal membranes, four (25%) in the study group, compared with 17 (81%) in the control group, were large (greater than 1500 micron in size). All large subfoveal membranes occurred in patients over 55 years of age. Thus, in this study the majority of CNVMs associated with pathologic entities other than age-related macular degeneration were well defined and not subfoveal at initial presentation. The majority of CNVMs associated with macular degeneration were occult or subfoveal. Membranes in patients over 55 years of age, regardless of cause, were commonly found to be occult or subfoveal at presentation.

One hundred twenty-eight consecutive aphakic glaucomatous eyes of 91 patients were examined by fluorescein angiography. Macular edema was present in 16 (28%) of 56 eyes currently being treated with epinephrine and 9 (13%) of 72 eyes not currently being treated with epinephrine. This difference is statistically significant at the 95% confidence level. Epinephrine therapy was discontinued in seven eyes with macular edema. A follow-up of these patients for six months to one year with serial anglograms showed resolution of edema in six of seven eyes. In one eye without macular edema, administration of epinephrine was associated with appearance of edema. Discontinuing epinephrine therapy was associated with resolution of edema. No significant correlation was found between use of other antiglaucomatous medications and macular edema. It is evident from this study that there is statistical support for the concept of an epinephrine-induced macular edema.

Clinical evaluation of uveal melanomas by nuclear magnetic resonance (NMR) techniques depends on ascertaining how these tumors characteristically appear in NMR images and spectra. The authors have determined NMR characteristics of suspected uveal melanomas by phosphorus-31 (31P) NMR spectroscopy of freshly enucleated human eyes. Nuclear magnetic resonance examination was performed at 8.45 Tesla within 90 minutes after enucleation. Enucleated eyes were maintained at 4 degrees C in tissue culture medium during the 30 minutes required for transport. Nuclear magnetic resonance spectra were obtained within 10 minutes, a clinically acceptable time, using a two-turn 31P surface coil. Spectral parameters included 10-kHz spectral width, 1024 data points, and 0.5-second recycle delay. Phosphorus-31 NMR spectroscopy allowed differentiation of choroidal melanomas from normal ocular structures. Differentiating features include significant peaks in tumor spectra due to the phosphodiesters glycerol 3-phosphoryl ethanolamine (GPE) and glycerol 3-phosphorylcholine (GPC), and the phosphomonoesters phosphorylethanolamine (PE) and phosphorylcholine (PC). These preliminary data are encouraging and suggest that clinical trials at the lower magnetic field strengths available in NMR imaging systems seem feasible and warrant investigation.

Of 84 patients with unilateral disciform macular degeneration, 36 were followed up for one to four years (average, 22 months). The fellow eye in 13 cases developed disciform lesions during that follow-up period. A significantly greater number of drusen were found in the fellow eyes of patients with unilateral disciform macular degeneration than in the eyes of a comparable control group of patients (P less than .001). Furthermore, the eyes with unilateral disciform macular degeneration that developed disciform lesions in the other eye were characterized as a group by a significantly larger number of drusen than those that did not. We postulate that drusen are precursors of disciform macular degeneration in fellow eyes and are probably manifestations of the same exudative process.

Proton irradiation was used in the treatment of uveal melanomas in 36 eyes. The average follow-up period was 16 months. One patient developed metastatic disease and died. No eye has been enucleated and tumor regression has been observed in all 22 eyes with a follow-up of more than 12 months. This type of treatment offers definite advantages over previously used methods, can be used for the treatment of relatively large melanomas, and should be considered before enucleation.

A malignant ciliary body melanoma received proton-beam irradiation. After an apparent failure of the tumor to respond, the eye was enucleated. A predominantly epithelioid cell tumor appeared viable by light microscopy, and a low degree of mitotic activity persisted, despite therapy. The tumor cells, however, displayed degenerative changes ultrastructurally, presumably results of the radiotherapy. These consisted of numerous cytoskeletal filaments, lipid vacuoles, prominent phagolysosomes, and nuclear convolutions and fragmentations. The mitochondria were fewer in number in the present tumor than typically encountered in epithelioid cells. A rare leptomeric structure was discovered, probably an organizational modification of the cytoplasmic filaments. The tumor's capillaries showed radiation-induced changes in terms of thickened basement membranes and perivascular fibrin deposition. The foregoing features are indicative of cellular and metabolic injury from the radiotherapy, but these were evidently not sufficiently injurious to sterilize the tumor.

A Q-switched neodymium-YAG laser was used to cut experimental vitreous membranes in rabbit eyes. We were able to cut vitreous membranes up to 4 mm from the retinal surface using single pulse energies of up to 4 mJ without fluorescein angiographic evidence of retinal injury. When cutting membranes between 1.5 and 3 mm from the retinal surface, injury was documented in five of seven cases. The most frequently observed fundus injury was choroidal hemorrhage. The neodymium-YAG laser may be useful in cutting avascular membranes in the posterior segment of the eye.

Criteria for the Callender classification of uveal melanoma are subjective and lead to variations among observers of the same tumor. Two other measures for cell type-the number of epithelioid cells per high-power field (Epi/HPF) and the inverse SD of nucleolar area (ISDNA)-were evaluated as prognostic factors for survival following enucleation for uveal melanoma in 232 patients. Variables significantly related to the number of Epi/HPF were related to ISDNA. Larger, more anteriorly located tumors had a high number of Epi/HPF and a low ISDNA. The two measures were also correlated. Using Kaplan-Meier survival analysis, both measures were related to tumor deaths. Worse prognosis was associated with a higher number of Epi/HPF and a lower ISDNA. Considered together in Cox' proportional hazards analysis, both measures independently predicted outcome. Separately, each measure predicted uveal melanoma-related deaths, in combination with largest tumor dimension, invasion of tumor to the line of transection, and location of the anterior tumor margin. Although ISDNA is the more objective measure of cell type, its practicality compared with other measures of cell type must also be considered.

We successfully treated a patient with uveal effusion syndrome and abnormal sclera with a partial-thickness sclerectomy. Part of the sclera was immediately cultured, and the excised sclera and the cultured cells were examined by electron microscopy. The sclera demonstrated increased glycosaminoglycan-like deposits between the scleral fibers. The cultured scleral cells showed large intracellular glycogen-like deposits, which were not seen in cells cultured from two control scleras. These findings may be the result of a metabolic defect, which causes a thick, impermeable sclera in some cases of uveal effusion.

Although uncommon, a wide variety of ocular manifestations can be seen in polyarteritis nodosa. These occur as a result of the arteritis or secondary to the associated renal induced hypertension. A case of biopsy documented polyarteritis nodosa is reported in which the patient presented with bilateral iritis, vitritis, and a retinal vasculitis involving both the retinal arteries and veins, a feature not described previously. Patients with this potentially fatal disorder may initially present with ocular involvement; thus ophthalmologists should be familiar with the clinical features of this disease.

The authors examined the relationship of change in tumor height after proton beam irradiation with the risk of metastasis in 700 patients treated for uveal melanoma before July 1, 1986. Rates of change in tumor height were computed for each patient using follow-up ultrasonographic measures during the first 2 years after treatment. Risk of metastasis was evaluated separately in the first 2 years after treatment when tumor decline was assessed (concurrent metastasis), and 2 years or later after treatment (subsequent metastasis). Using Cox proportional hazards models to adjust for other known risk factors, tumors regressing rapidly were significantly more likely to metastasize concurrently with their regression (P less than 0.05). The opposite was found for subsequent metastasis: tumors with a slow annual decline were more likely to metastasize after 2 years after treatment (P less than 0.05). Substantial changes also were noted in the ability of previously described prognostic factors (largest tumor diameter, tumor height, ciliary body involvement, and patient's age) to predict early and late metastasis. Characteristics placing individuals at high risk of metastasis may change markedly with time after treatment.

The incidence of metastasis and prognostic factors for metastasis in 780 consecutive patients with uveal melanomas treated with proton beam irradiation were evaluated. Metastasis developed in 64 patients (8%). The median time from treatment to the diagnosis of metastasis was 2.1 years (range, 3 months to 7.3 years). The liver was primarily involved in 58 (90%) patients. The 5-year cumulative probability of metastasis developing was 20%. Prognostic factors for metastasis developing were quite comparable to those found for patients treated by enucleation and included largest tumor diameter, involvement of the ciliary body, older age, and extrascleral extension. Surgical localization, tumor height, and elevated liver enzymes before treatment were not important factors in the development of metastasis.

We observed a young woman who showed growth of a choroidal melanoma over the course of 2 pregnancies, with subsequent enucleation of the eye. This is the first such documented case. In addition 4 other women with uveal malignant melanomas presented during pregnancy. This observed number of pregnancies (5) was greater than the expected number (2.1) among women of childbearing age who underwent enucleation with subsequent analysis in our pathology laboratory. However, this difference was not statistically significant. Further, more females 44 years of age or younger underwent enucleation for malignant melanoma than men of comparable age. Evaluation of these cases led us to propose that there is a subset of patients whose ocular melanomas are hormonally responsive. We therefore analysed uveal melanoma and choroidal tissue from 7 patients, including one of the pregnant women, for the presence of oestrogen receptors. No specific oestrogen binding was found. The possibilities that other hormones are involved or an immunological mechanism is operative are discussed.

Proton magnetic resonance imaging (1H MRI) has emerged as a clinically useful tool for the diagnosis of intraocular tumors. During the last four years 1H MRI characteristics, including spin-lattice relaxation times (T1) and spin-spin relaxation times (T2), have been established for several types of tumors. The introduction of surface coils to the imaging process has significantly improved the quality of intraocular MR images, leading some clinicians to suggest that 1H MR images are preferable to CT scans. Another MRI technique, in which sodium-23 (23Na) is imaged rather than protons, is now under development as tool for intraocular diagnosis. The potential of 23Na MRI depends upon the high concentration and "visibility" of sodium in the vitreous body, and upon the apparent differences in sodium behavior in normal cells vs. tumor cells. The metabolism of normal ocular tissues and intraocular tumors may be probed noninvasively with phosphorus-31 MR spectroscopy (31P MRS). Much progress has been made during the last few years in understanding the appearance of 31P MR spectra of many types of healthy and diseased cells and tissues. Clinical application of this technique to the diagnosis and monitoring of intraocular tumors following conservative treatment will be dependent upon the development of spectroscopy techniques that collect information from the volume of interest (tumor) only.

Fifteen owl monkeys (30 eyes) that received proton irradiation of discrete areas of the retina and choroid have been observed for 3 1/2 years or more after irradiation. The dose delivered to the retina varied from a single dose of 1,600 to 4,750 proton rads and from 10,250 to 18,250 rads delivered in five fractions over a period of one month. Fundus examination, fluorescein angiography, and light and electron microscopy revealed chorioretinal changes at the area of irradiation. The part of the retina adjacent to the chorioretinal scar showed preservation of normal retinal vascular architecture on trypsin digest studies of the retina and on electron microscopic examination, even 1 mm from the edge of the scar. The results of this study suggest that proton beam irradiation may markedly reduce delayed ocular complications, especially radiation retinopathy, which is common to other forms of radiotherapy in the treatment of intraocular tumors.

Therapy of choroidal metastases depends on several factors, but conventional external beam irradiation using photons is the most common treatment. Proton beam irradiation has recently been used for selected lesions. The advantage of this method is that the dose can be localized to the tumor more accurately. Other therapies include chemotherapy, photocoagulation, or, in extreme cases, enucleation. The author discusses these treatment modalities, complications, and prospects for future therapies.

We reviewed 242 cases of choroidal and ciliary body melanoma that were treated by enucleation to determine the importance of juxtapapillary location as a prognostic factor for tumor-related deaths. Analysis involved Kaplan-Meier survival curves and Cox proportional hazards analysis. Patients with juxtapapillary tumors had a worse prognosis than those with tumors in other locations (64% vs 78% five-year survival), but the difference was not statistically significant. Juxtapapillary location was not an independent risk factor on Cox multivariate analysis. Risk factors included number of epithelioid cells per high-power field, largest tumor dimension, location of the tumor's anterior margin, and invasion to the line of transection. In 63 cases (26%), the tumor was in contact with the optic disc. For these juxtapapillary tumors, we also evaluated the prognostic importance of six variables relevant to juxtapapillary location. On univariate analysis, tumor height at the disc margin, subarachnoid space invasion, and postlaminar optic nerve invasion were significantly related to survival, while disc compression by tumor, prelaminar optic nerve invasion, and extent of disc perimeter surrounded by tumor were not. On Cox multivariate analysis, however, none of these six variables was an independent risk factor. Risk factors for juxtapapillary tumors included number of epithelioid cells per high-power field and location of the tumor's anterior margin.