CURRENT OPHTHALMOLOGY

Year : 1997  |  Volume : 45  |  Issue : 3  |  Page : 143-161

Current concepts in the management of adult intraocular tumours

MP Shanmugam, P De Potter, L Gopal, J Biswas, MP Bhende
Sankara Nethralaya, Medical Research Foundation, Chennai, India

Correspondence Address:
M P Shanmugam
Sankara Nethralaya, Medical Research Foundation, Chennai
India

Source of Support: None, Conflict of Interest: None

Check

2

PMID: 9475017

The management of intraocular tumours has come a long way since the days when enucleation was the only modality of treatment available to the ophthalmologist. Despite the fact that enucleation is still necessary in certain situations, the emphasis is currently on conservative management, thereby saving the eye and some amount of useful vision wherever possible. This review highlights the current trends in the management of adult intraocular tumours with emphasis on newer diagnostic modalities such as computed tomography, magnetic resonance imaging, and fine needle aspiration biopsy. Only those tumours that are most likely to be seen in clinical practice are included.

Keywords: Adult intraocular tumours, choroidal malignant melanoma, metastatic tumour, choroidal haemangioma, management

Click here to view

Click here to view

Click here to view

Click here to view

Click here to view

Click here to view

I. DIAGNOSTIC AIDS

B. Ultrasonography

Choroidal melanoma

The typical acoustic properties of choroidal melanoma [Figure - 7] include collar button shape,[10],[11] low to moderate internal reflectivity,[11] regular internal structure, and internal blood flow or vascularity.

Additional findings of importance include sound attenuation,[11] choroidal excavation, associated exudative retinal detachment, changes in internal reflectivity, and status of adjacent scleral/ orbital interface.

Extrascleral extension typically appears as one or more nodules near the sclera, adjacent to the tumour base, and often displayed with high resolution B scans. The minimum size for detection should be 1.0-1.5 mm. In larger nodules, internal reflectivity, blood flow and thickness can be measured. Small areas may be detected but it may be difficult to diagnose them confidently. Serial follow ups can be useful to document growth. Small melanomas can be associated with large extraocular tumour growth. Therefore, it is important to repeat ultrasound at follow-up even if clinically there is no apparent change in size of the intraocular lesion.

Scleral infiltration, with or without extraocular extension can be demonstrated as replacement of a portion of the sclera by tumour tissue. Extension through the sclera within small blood vessels has been observed, as also extension into the orbit via a vortex vein. Congested blood vessels, localized inflammation in Tenon's space and extraocular muscles could simulate an extraocular growth.

Diffuse melanomas are rare lesions which are challenging to detect and diagnose.[10] They are only mildly elevated but can be quite extensive. The surface is irregular and bumpy with indistinct margins. The edges of the base cannot be accurately determined, and also their low elevation may prevent assessment of the internal reflectivity. Reflectivity can vary from low to medium amplitude. Vascularity may also be difficult to assess. These tumours are associated with a high incidence of extrascleral extension. The differential diagnosis includes metastasis to the choroid, Vogt Koyanagi Harada syndrome, and reactive lymphoid hyperplasia. The ultrasonographic differential diagnosis of malignant melanoma is shown in [Table - 1].

Echographic changes following radiation include:

Irregular structure and higher reflectivity as the lesions decrease in size due to necrosis.[10],[11], [12]

Loss of internal vascularity.

Initial enlargement of tumours 3 to 4 weeks after treatment, followed by eventual decrease in size. This is due to irradiation induced edema.

Thickening of adjacent extraocular muscles.

Decreased scleral reflectivity.

Inflammation of Tenon's space corresponding to the location of the plaque can simulate extrascleral extension after it is removed.[13]

It is important to review previous ultrasounds during serial examinations to ensure that there is consistency in measurement.

Correlation between histopathology and ultrasound findings may be useful as the current commercially available instruments do not routinely help to clinically evaluate the prognosis. However, in about 50% of cases histological characteristics influence the amount of reflection in the A mode.[11] Ultrasound spectral analysis is still in the experimental stage that could predict the histologic pattern of the tumour to a certain extent. [14, 15]

Other uses of ultrasound in melanomas include localising tantalum rings which identifies the tumour for charged particle irradiation,[10] and also as an useful adjunct to transillumination to verify the position of the radioactive plaque in relation to the tumour.[10],[16] This is most helpful in tumours near the optic nerve, and also in mushroom shaped tumours with overhang where the entire base cannot be visualized ophthalmoscopically. If the plaque is not well positioned over the tumour it can be moved and ultrasound repeated to ensure adequate coverage. It is reliable in tumours with sloping indistinct borders.

Doppler ultrasonography has also been described as useful in the follow-up of conservatively treated melanomas.[11],[17]

Choroidal metastasis

The typical echographic appearance of choroidal metastasis is diffuse thickening of the ocular coat. The characteristics include:

Mild to moderate elevation.

Rarely highly elevated; irregular bumpy surface contour, often with central excavation.

More extensive echographically than clinically appreciable.

Medium to high internal reflectivity, but structure can be irregular as a result of varied histologic architecture.

Internal vascularity minimal or absent.

Secondary retinal detachment more elevated and extensive than melanomas of similar size.

Simultaneous occurrence of orbital metastases may be noted; rarely have vitreous/subretinal haemorrhage

The atypical presentations include:

Metastatic oat cell carcinoma of lung: low internal reflectivity and internal vascularity.

Metastatic adenocarcinoma: strong acoustic interfaces.

Cutaneous melanoma: cannot be differentiated from primary choroidal melanoma.

Large metastases: bullous choroidal detachment.[18]

Choroidal haemangioma:

Circumscribed haemangiomas [Figure - 8] have the following echographic characteristics:

Dome Shaped, slightly elevated in the posterior pole, usually temporal to optic nerve head.

Associated retinal detachment.

Strongly reflective, nearly concentric widening of the ocular coats

Regular internal structure.

Despite abundant vasculature no significant internal flow due to slow circulation within the cvities of dilated blood vessels.

Calcification is seen occasionally on the surface or within the lesion in long-standing cases; in such cases, ultrasound is not diagnostic.[10]

Diffuse type [Figure - 9] choroidal hemangiomas are less elevated, and may be mistaken for non-specific thickening of the retinochoroid layer.

C. Computed Tomography (CT) and Magnetic Resonance Imaging (MRI)

The diagnosis of intraocular lesions is usually made with assurance by indirect ophthalmoscopy. However, imaging studies are often required tohelp document the lesion, to differentiate it from other similar conditions, and to evaluate the intraocular tissues when opaque media precludes view of the fundus.

The role of CT has been limited in the evaluation of intraocular tumours and simulating lesions due to its poor tissue definition. Because of its safety, soft tissue resolution, and its multiplanar capabilities, MRI appears to be the most helpful and non-invasive imaging study of choice for evaluation of selected intraocular lesions. Information obtained from MR images allows us to identify tissue compounds such as melanin, methemoglobin, deoxyhemoglobin, and proteinaceous fluid. The use of orbital surface coil, gadolinium-DTPA (Gd-DTPA) as contrast agent, and fat suppression techniques are most critical in detecting and evaluating intraocular lesions.

Ct

With CT, uveal melanoma appears as an elevated, solid, moderately dense uveal mass which demonstrates minimal to moderate enhancement after contrast injection. Associated secondary retinal detachment usually appears as a semilunar moderate density which shifts in location with respect to the position of the eye and does not enhance.[19][20][21][22][23] The differential diagnosis between uveal melanoma and other uveal tumours such as choroidal metastasis, choroidal haemangioma, and uveal leiomyoma is almost impossible on CT.[24]Choroidal osteoma appears as a minimally elevated choroidal mass with linear calcification at the level of the choroid and usually without subretinal fluid.[25] Extraocular extension of posterior uveal melanoma may be easily identified on CT if MR studies are contraindicated.

MRI Choroidal melanoma

The MRI characteristics of uveal melanoma have been mainly attributed to the paramagnetic properties of the melanin.[26] Melanin causes enhancement of proton relaxation on MRI that leads to an increase signal intensity on Tl-weighted images. [27,28] Therefore, on Tl-weighted images, choroidal melanoma appears with a hyperintense signal with respect to the hypointense vitreous [Figure - 10]. On T2-weighted images, choroidal melanoma becomes hypointense with respect to the hyperintense vitreous [Figure - 11]. Previous studies have shown that 72-95% of posterior uveal melanomas evaluated at 1.5 Tesla demonstrate these characteristic MRI appearances.[22],[27][28][29][30][31][32] The presence of tumour necrosis containing water (free or in extracellular fluid), the presence of blood or altered blood products (methemoglobin, deoxyhemoglobin, hemosiderin), and iron content may explain the varying combination of signal intensities such as decreased signal on Tl-weighted images and /or increased signal on T2-weighted images.[27],[30][31][32][33] The combination of T1 hyperintensity and T2 hypointensity has also been observed with other intraocular lesions as explained later. Associated serous retinal detachment exhibits hyperintense signal on both T1- and T2-weighted image due to its proteinaceous content.[22],[31][32][33] On Gd-DTPA enhanced Tl-weighted images, posterior uveal melanoma demonstrates moderate to marked enhancement.[30],[31],[34],[35] Larger tumours show more heterogeneous enhancement than smaller tumours.[31] Postcontrast T1-weighted images with fat suppression techniques appear to be most helpful in detecting and delineating small posterior uveal melanoma with a thickness more than 1.8 mm.[30] Because secondary subretinal fluid does not enhance after Gd-DTPA injection, postcontrast T1-weighted images are helpful in differentiating uveal melanomas from retinal detachment.[30],[31],[34],[35] Contrast enhanced MRI with fat suppression technique appears as sensitive as ultrasonography in detecting extraocular extension of posterior uveal melanoma.[27],[8]

Choroidal metastasis

Choroidal metastasis has been reported to demonstrate varying signal intensities such as a combination of T1 and T2 hyperintensities or T1 hyperintensity and T2 hypointensity.[22],[27],[32],[36] Most often, choroidal metastasis shows a hyperintense signal on T1-weighted images and hypointense signal on T2-weighted images with respect to the vitreous.[20][29][30][31],[37] These MRI features suggest caution in differentiating choroidal metastasis from amelanotic choroidal melanoma which exhibits the same MRI characteristics mentioned later.

Choroidal haemangioma

The MRI features of circumscribed choroidal haemangioma are mainly related to the high vascular flow within the lesion. Choroidal haemangioma exhibits an isointense to slightly hyperintense signal on Tl-weighted images with respect to the vitreous. Because of its high content of extracellular free-water (stagnant blood), choroidal haemangioma becomes hyperintense on T2-weighted images and therefore isointense with respect to the vitreous.[20],[22],[29][30][31],[36] Based on our experience, the T1and T2 hyperintensities features of choroidal haemangioma are most helpful in differentiating circumscribed choroidal haemangioma from other amelanotic choroidal tumours. Moreover, choroidal melanoma demonstrates marked enhancement on postcontrast T1-weighted images. The associated exudative T2-weighted images does not enhance after Gd-DTPA injection.[20],[30],[31]

Intraocular lesions with thickness >1.8 mm that can show the combination of T1 hyperintensity and T2 hypointensity on noncontrast MRI include uveal melanoma, uveal metastasis, uveal leiomyoma, uveal melanocytoma, choroidal osteoma, subacute choroidal/ retinal haemorrhagic detachment, retinoblastoma, retinal capillary haemangioma, retinal gliosis (focal), and inflammatory granuloma.[31]

D. Fine Needle Aspiration Biopsy (FNAB)

FNAB was first proposed as an adjunct in the management of intraocular tumours by Jackobiec et al.[38] The advantage of FNAB is that it offers histopathologic correlation to the clinical diagnosis when faced with anatypical presentation of intraocular tumours. It is recommended in (i) cases in which the diagnosis is difficult, distinction between malignant and benign not clear, and where the therapeutic decision will be made based on the cytologic findings; (ii) patients with metastatic disease to the choroid but with no known primary; and (iii) patient refuses recommended therapy until confirmed histopathologically.[39]

Technique

The technique varies with the location of the lesion and the suspected diagnosis. In the anterior segment, free floating cells or a pseudohypopyon can be aspirated by passing a 30 or 27 gauge needle through the clear cornea just inside the limbus into the inferior anterior chamber. For a solid iris tumour, the needle is passed tangentially avoiding the pupil. Posterior segment rumours can be biopsied transvitreally or subretinally depending on the height of associated retinal detachment. A tumour in the ciliary body, retina or choroid with minimal associated retinal detachment is approached through a sclerotomy placed 180° away. The needle is guided through the vitreous into the retina in case of a subretinal tumour under indirect ophthalmoscopic control. When a subretinal mass lesion is associated with a bullous retinal detachment, the needle is guided tangentially through the subretinal space via a suitably placed posterior sclerotomy. In the extremely rare case when FNAB is indicated in a patient of possible retinoblastoma, a clear corneal approach is employed. The needle is guided through a clear corneal wound 2-3 mm within the limbus through the root of the iris, zonules and into the vitreous and the tumour.

A 25 gauge one and a half inch needle attached to a short disposable plastic tubing and a disposable syringe is employed. Once the needle is in the tumour, assistant aided gentle suction is employed to aspirate the cells into the needle. The needle is withdrawn after releasing the suction force and immersed in balanced salt solution and the contents of the needle aspirated into the syringe with the balanced salt solution. Gentle pressure on the globe usually limits the amount of intraocular haemorrhage.

Tumours that have involved the vitreous jelly such as leukemias and reticulum cell sarcoma can be biopsied using the needle technique, or by using a vitreous cutter. This approach is not to be utilized for biopsy of vitreous seeds in friable tumours such as retinoblastoma.

The reported complications of this procedure are intraocular haemorrhage and retinal break formation.[39][40][41] Intraocular haemorrhage as a complication of FNAB occurs virtually in all patients. This haemorrhage is usually of limited quantity and is controlled by gentle pressure on the globe. The haemorrhages clear adequately with time. The retinal breaks created when a subretinal tumour is biopsied transvitreally almost never lead to a rhegmatogenous retinal detachment as they get closed by the blood clot. Even in the presence of minimal subretinal fluid overlying the mass lesion when the break is created, the detachment seldom becomes rhegmatogenous.

The major deterrent to wide use of FNAB is the fear of needle track dissemination of tumour cells. Despite experimental evidence of needle track dissemination, clinical evidence suggests otherwise.[42],[43] Shields et al have not reported any case of needle track dissemination in their large series of 140 patients.[40] The factors which may limit the incidence of tumour dissemination are that the vitreous "wipes" off the tumour cells sticking on to the needle surface, and the prompt treatment of the malignancy with appropriate therapy which usually follows the diagnosis with FNAB.

FNAB is not to be embarked upon without the presence of a skilled cytologist, as a part of the team. Despite the ease of obtaining the specimen, improper interpretation could cause disastrous results. Shields et al have reported that FNAB has a sensitivity rate of 84% and a specificity of 98%, making it a useful adjunct in difficult cases.[40] It is a specialized technique which is not to be used for routine diagnosis of intraocular tumours due to the real risk of tumour dissemination and other potentially vision threatening complications.

II. Treatment A. Factors influencing treatment choice

Malignant melanoma

Eyes with salvageable vision should be conservatively treated. Eyes with severe secondary glaucoma are generally those with large diffuse tumours that rarely respond to conservative methods of management necessitating enucleation.[44],[45]

Small tumours (2mm thick, <10 mm diameter) are generally watched periodically for signs of growth. Medium sized tumours (3-5 mm high, 10-15 mm diameter) can be periodically watched, photocoagulated, treated with plaque therapy, or enucleated based on presence of other clinical features. Large melanomas (5-10 mm high, >15 mm diameter) also need active management. Extra large melanomas (>10 mm high) generally need enucleation.[46][47][48]

Diffusely growing tumours generally will need to be enucleated, while nodular tumours can be conservatively managed depending on other factors. Small degrees of extrascleral extension can also be managed conservatively.[48]

Tumours situated near the equator are likely to benefit from photocoagulation, local resection, or radiotherapy, while in tumours close to the fovea, photocoagulation is best avoided. Circumpapillary tumours will need enucleation.

Dormant tumours can be watched periodically for evidence of activity. Dormant tumours have minimal subretinal fluid, are sessile, and have sparse orange pigment overlying them.[48] Older patients with slow growing tumours are conservatively managed while aggressive tumours in younger patients are more aggressively managed.

Dormant small and medium sized tumours, and slow growing small and medium tumours in sick and one eyed patients are generally observed periodically.[48]

Choroidal metastasis

Chroidal metastasis may be watched periodically, treated with chemotherapy, irradiated or surgically resected, depending on the degree of activity, location, extent, of the tumour, degree of control obtained with chemotherapy, site of the primary, and patient's symptoms.

Some metastatic tumours of the choroid may be inactive, or may have regressed spontaneously following treatment of the primary. They do not have an associated retinal detachment and are flat with pigment clumping. These tumours need periodic observation.

Others

Solitary choroidal haemangiomas that are asymptomatic need be periodically observed until exudative retinal detachment involving the fovea develops. Von Hippel's angiomas may also be watched if there be no exudation or accumulation of subretinal fluid.

B. Enucleation

Choroidal malignant melanoma

Large tumours producing visual loss that cannot be managed by conservative methods, small or medium sized tumours with optic nerve invasion, posterior uveal melanomas with total retinal detachment or severe glaucoma need enucleation. A minimal manipulation enucleation is advised to prevent haematogenous dissemination of the tumour. A "no touch technique" has been advocated as also cannulation of the anterior chamber to avoid sudden fluctuations in the intraocular pressure thereby avoiding egress of tumour cells from the eye.[49][50][51] The current concern is about high rate of mortality following enucleation, despite lack of evidence of metastatic disease at the time of enucleation (30% in the first 5 years following enucleation and 50% within 10 years).[52] The presumed reason for this is that the antitumour antibody levels fall following enucleation, while it increases after local treatment such as photocoagulation or irradiation.[53] This decreased antibody response may not prevent onset of metastatic disease.

Choroidal metastastis

Eyes with large metastatic tumours of the choroid with bullous retinal detachment and intractable painful glaucoma may be difficult to control by any modality other than enucleation.[54] When the ocular tumour is characteristic of metastastic disease, but with no identified primary, and if the involved eye is painful and blind, enucleation of the eye may serve two purposes: relief from pain and diagnosis of the primary site. This is also useful in appropriate treatment of the fellow eye in the presence of bilateral metastasis.

C. Local resection

Ciliary body and choroidal malignant melanoma

Tumours close to the ora not extending more than 4 clock hours of pars plicata and choroidal melanoma not greater than 15 mm in diameter centered at the equator can be considered for partial lamellar sclerouvectomy if the tumour has not involved the retina, and penetrating sclerochorioretinovitrectomy if the retina and sclera are involved.[55],[56]

The technique involves localisation of the tumour on the sclera, partial thickness dissection of the scleral bed, resection of a thin layer of the sclera with the tumour, leaving the overlying retina intact if there is no involvement. Preoperative barrage laser photocoagulation or cryotheraphy 2-3 mm surrounding the tumour is done, if retinal involvement is noted, in an attempt to keep the retina attached postoperatively. In such a situation, the overlying retina is also removed during tumour removal with open sky vitrectomy. The sclera, if involved, is also removed and replaced with donor scleral graft.

Complications of this procedure include retinal detachment, cataract formation, vitreous haemorrhage, expulsive haemorrhage, chronic cystoid macular oedema, anterior segment ischaemia, tumour recurrence, hypotony, wound leak, and preretinal fibrosis. In a large series of 81 posterior uveal melanomas which were resected, 91% were alive at a mean of 4 year follow-up with 26% having 6/9 vision or better following surgery.[48] This group of patients can also be effectively managed by radiotherapy. Resection is generally preferred for anterior tumours with a small base and greater height.

Internal resection is a technique which involves preoperative photocoagulation around the tumour followed by vitrectomy and internal removal of the tumour piece-meal in posteriorly located tumours that are not amenable to lamellar sclerouvectomy.[57] The concern is that of promoting metastasis with this technique. In one study of 23 patients with a variable follow-up, three patients developed metastatic disease in the postoperative period and two patients developed tumour recurrence.[58] Studies with long term results are awaited.

Choroidal metastasis

In selected cases of metastatic tumours from primaries of low malignancy and a probable single metastatic foci, such as bronchial carcinoid, local resection may be considered as in one case of iris and ciliary body metastasis of bronchial carcinoid.[59][60][61]

Capillary angiomas of the retina

Selected cases of Von Hippels angiomas have been subjected to eye wall resection, with attendant decrease in subretinal exudation.[62] Internal resection of these tumours with vitrectomy, endodiathermy, retinectomy, laser photocoagulation, and silicon oil tamponade has also been attempted in a few cases.

D. Photocoagulation

Choroidal malignant melanoma

Photocoagulation has been tried for malignant melanoma of choroid[63] but not with the same success as for retinoblastoma. The classification of choroidal melanoma by the Collaborative Ocular Melanoma Study (COMS)[64] is useful in discussing the role of photocoagulation.

Meyer Schwickerath was the first to use xenon arc photocoagulation for treatment of small malignant melanoma of the choroid.[65] The indications for use of this modality are not clear. Previous reports showed frequent complications such as preretinal fibrosis, macular pucker, tractional retinal detachment, and vitreous haemorrhage. Shields[66] recommended treatment with laser photocoagulation in case of small nasal melanomas and temporal peripheral small melanoma located away from macula outside the arcade. The tumours should be <3 mm in height and 9-10 mm in basal diameter without secondary retinal detachment. They should not be closer than one disc diameter from disc. Some authors feel that argon is less effective compared to xenon or krypton for melanoma of the choroid.[65] Less than 5% of melanomas meet the criteria laid down for use of photocoagulation.

For treatment with xenon,[65] retrobulbar anesthesia is administered. A 6° aperture and green I/II option is used. In the first sitting the tumour is surrounded by 2 rows of burns. This is aimed at preventing rhegmatogenous retinal detachment from spreading if a break occurs during treatment of tumour itself subsequently, reducing the likelihood of marginal miss during treatment, and reducing blood supply to the tumour. In the second sitting, moderate to heavy treatment is given to the entire tumour. This is done approximately 6-8 weeks after the first sitting. The setting in xenon would be green I-IV. Repeat treatment is done if need be up to 4 sessions. If 4-5 sessions do not yield the desired response, other modalities of treatment are opted for.

For treatment with argon or krypton laser,[67] 500 µ spot size, 0.2 second duration, and 500-1,200 mW power are used. Successful laser treatment is indicated by a flat or depressed gliotic scar with pigmentation. Total obliteration of the tumour may sometimes take up to 2 years. If 4-5 sessions of treatment have failed to achieve any results, one should resort to alternative modes of treatment. Amelanotic tumours are relatively resistant to treatment.

Complications of photocoagulation include retinal break, especially during the subsequent stages of treatment (retreatment), which could cause rhegmatogenous retinal detachment, vitreous haemorrhage during 2nd or 3rd sittings of treatment, choroidoretinovitreal neovascularisation near the edge of treatment, branch retinal vein occlusion, secondary macular edema, and corneal and iris burns.

The combination of tunable dye laser and photoactivated dye has been tried to treat some malignant melanomas.[68] Haemtoporphyrin dye derivative is the most commonly used dye. The efficacy of this treatment has not been well evaluated.

In a rabbit model Chong et al[69] have shown that ICG enhanced diode laser treatment seems to destroy all tumour cells even in the depths of the tumour. The authors implanted green amelanotic melanoma into iris of pigmented rabbits. Incomplete tumour necrosis was seen with only diode laser versus total tumour destruction following ICG enhanced diode laser. The chromophore in the deep blood vessels selectively absorbs diode and facilitates tumour cell destruction in the depths of the tumour. The clinical application of this technique of treatment remains to be evaluated.

Metastatic tumours

There is no role of photocoagulation in the management of metastatic tumours.

Vascular posterior segment tumours

Treatment of angiomatosis retinae can be done with photocoagulation. Large tumours are resistant to photocoagulation treatment. Angiomas located near the optic disc are extremely difficult to treat and eradicate. In general, small tumours of <1.5 mm diameter show little growth and can be watched. Tumours <2.5 mm diameter and <1 mm in elevation can be treated with argon laser or xenon. Slightly thicker lesions may need other modalities of treatment.

In the first sitting, the tumour is surrounded with 500 μ spot size burns of long duration. The burns are placed in contiguous fashion. Multiple sittings may be needed to achieve total destruction of the tumour. In the "feeder vessel technique", heavy intensity 500 µ burns can be applied to the feeder vessels. This can succeed in closing the vessel and help eradicate the tumour. This technique may work for tumours <2.5 disc diameter in size. The burn diameter must be larger than the vessel diameter to avoid rupture of blood vessel. Temporary occlusion of central retinal artery by raising the intraocular pressure and then coagulating the feeder vessel has also been described.

Johnston et al[70] treated angioma involving temporal aspect of optic disc with photocoagulation along with creation of chorioretinal adhesion to barricade the macula from angioma. The vision has been maintained at 6/6 for 16 years. They recommend early treatment when tumour is still small.

Complications include gliosis that occurs around the tumour following photocoagulation, which may make it difficult to treat subsequently. Vitreous and retinal haemorrhage can occur during feeder vessel technique. As mentioned before, large tumours fail to respond to treatment. Diode laser cannot be used to treat angiomas in view of its lack of absorption by reddish lesions. ICG enhanced diode laser treatment may be a possibility for angiomatous lesions also.

Choroidal haemangioma

Choroidal haemangioma of the localised variety is amenable to treatment [Figure - 12]. If there is no secondary retinal detachment, there is no indication for treatment. Both argon laser and dye yellow laser can be used in the treatment of choroidal haemangioma. The presumed advantage of dye yellow laser is the greater absorption potential by the abnormal tissue of this wave length. It has been found that treatment does not always lead to obliteration of the tumour mass. Treatment of the surface of the tumour leads to resolution of the secondary retinal detachment even when the tumour mass is not destroyed. Hence, the goal of laser therapy is to achieve resolution of secondary retinal detachment. A 500 µ spot size is useful.

E. Cryotherapy

Von Hippel Lindau disease

Photocoagulation sometimes fails to reach the full thickness of the angioma if it is quite thick. Therefore, cryotherapy can be used as a primary or adjunctive modality of treatment to double freeze thaw the lesion. In general, only small tumours without vitreous and retinal changes do well with this treatment.

Choroidal malignant melanoma

Cryotheraphy has not been very popular in the treatment of malignant melanoma. Klein at al have reported successful treatment of 4 eyes with tumours 1.7-3.8 mm thick and 1.7-11.4 mm in diameter with triple freeze thaw cryoptheraphy.[71]

F. Radiotherapy

Choroidal malignant melanoma

Pre-enucleation external beam radiotherapy had been advocated to decrease mortality rate following enucleation, but it's value is currently debatable.[72] Post-enucleation irradiation to the orbit decreased the mortality in a study but this was not a prospective randomized trial.[73] This issue of benefit of pre-enucleation irradiation is currently being evaluated by the COMS study.[64]

In external beam particle radiotherapy, the charged particle radiation consists of usage of helium ions and protons, and the benefit lies in the fact that relatively large doses can be delivered to the tumour specifically. The high energy charged particles have minimal scatter and a well defined, finite, energy dependent range. The entire tumour can be treated uniformly, and due to the inherent Bragg peak effect, unnecessary radiation to the normal tissue is avoided.[74]

Uveal melanomas up to 24 mm in diameter and 14 mm in height have been treated with charged particle irradiation. Tumours involving the fovea, optic nerve head, those with small extrascleral extension are also manageable. The technique involves localisation of the tumour with transillumination and indirect ophthalmoscopy, and suturing four 2.5 mm tantalum rings as localisation markers. The treatment is then planned using a computer, thereby choosing the correct direction of the beam which entails minimum irradiation to normal structures but also complete treatment of the tumour and 1.5 mm of surrounding normal tissue. Approximately 70 Cobalt-60 gray equivalents are delivered in 5 sittings over 7-10 days.

In one study 58% of the patients treated with protons had 6/36 or better vision at a mean 5.4 year followup,[75] while in another study 55% of helium ion treated patients had 6/60 or better vision at 5.6 year followup.[76] The cumulative probability of developing metastasis is 20% with charged particle radiation.[76],[77] The five year survival rate following proton beam radiation has been reported to be 85%, comparable with that of plaque therapy (82%).[78] Complications include those of the surgery such as transient diplopia and intratumour haemorrhage, and those related to radiation such as epilation, epiphora due to punctal occlusion, dry eye, cataract formation, rubeosis iridis, neovascular glaucoma, radiation vasculopathy, papillopathy, maculopathy, macular oedema, neovascularisation of the retina, and vitreous haemorrhage.

The disadvantage of the current charged particle radiation systems are that they are

expensive to install and maintain.

Choroidal haemangioma

Laser photocoagulation has been advocated in the management of solitary choroidal haemangiomas associated with secondary retinal detachment.[79],[80] In special situations such as juxtapapillary and subfoveal tumours photocoagulation by itself may cause irreversible visual loss. In the presence of a bullous retinal detachment, photocoagulation may not be feasible and may need multiple sessions. Other techniques advocated are drainage of sub retinal fluid followed by laser photocoagulation or vitrectomy with gas tamponade and laser photocoagulation. [79, 80] Diffuse haemangiomas associated with or without Sturge Weber syndrome rarely respond to laser photocoagulation. Low dose external beam radiotherapy has been used successfully in the management of such refractile tumours.[81][82][83]

Solitary and diffuse haemangiomas can be treated with 2,000 cGy to 3,000 cGy of external beam radiation after localising the tumour on a CT or MRI. The largest series of localised and diffuse choroidal haemangioma treated with irradiation is by Alberti and Gerber.[81] Out of the 22 eyes of 21 patients treated, regression of the tumour occurred in all eyes with resorption of the exudative retinal detachment in all the 19 eyes where it was present. Vision improved in 8 eyes, and in the 3 eyes with glaucoma, the intraocular pressure could be brought under control. Cataract and radiation retinopathy as complications of this therapy may occur but are rare due to low dosage of radiation involved. This modality can be employed in eyes that are not suitable for photocoagulation thereby atleast preventing a painful blind eye. Visual recovery is usually small and may not occur in all the eyes, as there would be some degree of amblyopia in the eyes with macular tumours associated with macular degeneration due to long standing retinal detachment.

Choroidal metastasis

The presence of exudative retinal detachment involving the fovea is an indication for radiotherapy. Considering that most of the metastatic tumours involve posterior choroid, irradiation of the posterior segment with a temporal or anteromedial beam is advocated. A dose of 3,000 cGy to 4,000 cGy is delivered over a 4 week period.[84] The technique may have to be modified in the presence of anterior segment involvement or it is advisable to irradiate the whole eye. Some authors advocate adjuvant irradiation for better tumour control even if the tumour responds to chemotherapy.[85] The tumour decreases in thickness with attendant decrease in the associated retinal detachment and improvement in visual acuity. The survival depends on the nature of systemic malignancy and in general have a poor prognosis.

G. Chemotherapy

Choroidal metastasis

In the presence of a known primary that is amenable to chemotherapy, choroidal metastases also respond to the specific treatment of the primary. If the exudative retinal detachment subsides and the patient is asymptomatic, it is a good practice to observe periodically for regrowth or recurrence of the retinal detachment.[86],[87]

Posterior uveal malignant melanoma

There are sporadic reports of treatment of uveal melanomas with various chemotherapeutic agents such as bichloroethyl nitrourea, adjuvant treatment with vinblastine, thiotepa, methotrexate, and deticene, but conclusive proof of benefit is lacking.[88],[89] Experimental studies with fluorouracil are also documented.[90] There has been an attempt at adjuvant chemotherapy with no-touch enucleation with an average follow-up time of 6.75 years. The patients have been treated with hydroxyurea, BCNU and DTIC with reported five year survival rate of 90%. Long term results are awaited.[91]

H. Brachytherapy

Uveal melanoma

The current relative indications for treating a posterior uveal melanoma with plaque radiotherapy are generally as follows: (i) selected small melanomas that are documented to be growing or that show obvious signs of activity on the first visit; (ii) most medium-sized and some large choroidal and ciliary body melanomas in an eye with potential salvageable vision; and (iii) almost all actively growing melanomas that occur in the patient's only useful eye. If a melanoma exceeds 15 mm in diameter and 10 mm in thickness, one should anticipate visual morbidity from radiation therapy, and enucleation should be strongly advised.[92] The surgical technique of radioactive plaque application has been described in detail in the literature.[92],[93] More recently, specially designed notched plaques have been effectively employed to surround the optic nerve and irradiate those tumours with a juxtapapillary location.[92],[93]

The visual outcome of an eye treated with brachytherapy depends mainly on the tumour size and location as well as on the development of radiation retinopathy and papillopathy.[94][95][96] Patients with tumours located near the fovea or optic disc, and patients with larger tumours have the worse visual outcome.[95],[96] In their series of 77 patients with posterior uveal melanoma managed by Cobalt-60 plaque radiotherapy, Cruess et al found that eyes receiving a radiation dose in excess of 5,000 cGy to the fovea and/or optic disc commonly lose a substantial amount of vision within 2-3 years.[97] This toxic dose to these sites is generally given for management of medium and large choroidal melanomas with margins closer than 5 mm to the fovea/optic disc.[97] In a more recent series of 93 patients with choroidal melanoma touching the optic disc, De Potter et al[98] found an incidence of 87% of radiation retinopathy and 52% of radiation papillopathy after a mean interval of 21 and 27 months, respectively, after plaque therapy. The significant predictors of radiation retinopathy were history of diabetes mellitus and the use of notched-designed radioactive plaque. The factors predictive of development of radiation papillopathy were the older age of patient (>45 years old), diabetes mellitus, mushroom-shaped tumour configuration, and nasal location of the tumour. Using life table analysis, the proportion of patients who experienced a decrement of at least three lines of vision was zero by 0-20 months, 22% by 20-30 months, 45% by 30-40 months, and 72% by 50-60 months. These results were also confirmed by Lommatzsch et al that consisted of 106Ru/106Rh plaques for juxtapapillary choroidal melanomas.[99]

Local tumour relapse after plaque radiotherapy has been reported to occur in up to 16% of cases.[95],[96],[98],[99] Local tumour recurrence constitutes an important post-treatment clinical indicator of tumour's greater malignant potential and the patient's increased risk of melanoma-specific mortality.[98] In two series of 93 patients with juxtapapillary choroidal melanoma managed by plaque radiotherapy local tumour recurrence was documented in 14 cases (15%).[98],[99] The younger age of the patient (<35 years old) and the superior and inferior location of the tumour were predictive for local tumour recurrence.[98]

In the published literature there is no statistically significant difference in survival between patients treated with plaque radiotherapy (Cobalt-60, Iodine-125) and those treated with enucleation.[100][101][102][103] De Potter and associates found that in patients younger than 50 years old, the method of treatment (enucleation versus Iodine-125 plaque radiotherapy) did not significantly affect the risk of metastasis. The 5-year survival of relapse-free patients (87%) was significantly better than that of patients whose tumour locally recurred (58%).[98] A rapid local tumour regression behavior has been identified as an unfavorable prognostic sign for survival.[104][105][106]

Combined plaque radiotherapy and laser photocoagulation or thermotheraphy have been used recently to increase the likelihood of complete local tumour destruction particularly in patients with tumour adjacent to the optic disc.[107] Preliminary results have shown that hyperthermia in combination with plaque radiotherapy may decrease the adverse effects of radiation to the retina, optic disc, and choroid.[108] Custom designed plaque radiotherapy appears to be an effective alternative method of controlling non-resectable diffuse iris melanoma.[109] In the series of Shields,[109] 14 patients with non-resectable iris melanoma were treated with Iodine-125 plaque radiotherapy. The mean length of treatment was 96 hours in order to give a mean dose of 29,300 cGy to the base (corneal endothelium) and 10,600 cGy to the apex of the iris melanoma. Tumour control was achieved in 93% of the patients. Despite large dose of radiation given transcorneally, the cornea tolerated it very well without corneal melting. Cataract developed in 6 patients. All but one patient had tumour control and retention of the eye.

Uveal metastasis

Standard external beam radiotherapy in a patient with choroidal metastasis, typically requires 3-4 weeks to complete, which may be a substantial portion of the patient's remaining life expectancy, in contrast to brachytherapy which does not take more than 3-4 days. In a series of 36 patients with uveal metastasis, plaque radiotherapy was used as primary or secondary treatment, after failure of the uveal tumour to respond to external beam radiotherapy, chemotherapy, or hormonal therapy.[110] The mean time for treatment was 86 hours and the mean therapeutic dose was 6,880 cGy to the tumour apex and 23,560 to the tumour base. Regression of the uveal metastasis was documented in 94% of the cases. Plaque radiotherapy salvaged five of the six eyes that had failed prior external beam radiotherapy.[110]

Capillary angioma of the retina

A selected group of solitary angiomas associated with exudative retinal detachment have been treated with local irradiation by application of plaques.[111] There are reports of treating these with charged particle radiation also.

III. ADVANCES IN PATHOLOGIC INTERPRETATION

Choroidal malignant melanoma

In pathologic diagnosis of intraocular tumours mostly routine light microscopy, and occasionally special stain, for example, Fontana-Mason silver stain is needed to identify inconspicuous melanin granules. Most cases of malignant melanoma of the choroid can be diagnosed fairly well by routine light microscopy. Amelanotic melanoma can cause diagnostic dilemma. If definitive diagnosis cannot be established by routine and special stains, a panel of monoclonal antibodies[112] can be used to identify the intraocular tumour [Table - 2]. Immunohistochemistry is needed to differentiate amelanotic melanoma from leiomyoma and peripheral nerve sheath tumours. Positive staining with antibodies against muscle-specific antigen and smooth-muscle antigen, and negative staining with HMB-45 and S-100 protein establishes the diagnosis of leiomyoma. One can distinguish amelanotic melanoma from peripheral nerve sheath tumour using neural marker leu-7 which will be positive, and HMB-45 which will be negative. Positive staining with S-100 and HMB-45 indicates malignant melanoma of the choroid. Histopathologic assessment of prognosis of uveal malignant melanoma can be done by the following features:

Cell type

Callender classified uveal malignant melanoma into six groups which includes spindle A, Spindle B, epithelioid, mixed (mixture of epithelioid and spindle), fascicular, and necrotic. Subsequently, investigators of Armed Force Institute of Pathology (AFIP) simplified Callender's classification by deleting fascicular and necrotic type. Now AFIP classification consists of four types of cells in melanotic tumour [Table - 3]. However, at present controversy exists whether to call an uveal melanoma as epithelia cell type in the presence of few spindle cells as all uveal melanomas contain atleast few spindle cells. Current practice at the AFIP is to designate all epithelia cell melanoma as mixed while assessing the prognosis. Pathology committee of Collaborative Ocular Melanoma Study (COMS) designated a new category: intermediate epithelia cell which are similar to epithelioid cell but often smaller in size and have indistinct borders. Cell type is still considered one of the most important prognostic factors in malignant melanoma of the choroid. In a recent study Seregard et al[113] have found immunoreactivity to a monoclonal antibody PC-10 for proliferating cell nuclear antigen to correlate with the prognosis in uveal melanoma. Ten year survival proportion was 84% for the low PC-10 count group in contrast to 40% for patients having tumour with high PC-10 counts.[113]

Cytomorphometric features

In recent years, Gamel and coworkers[114] have found that the most important prognostic feature in malignant melanoma of the choroid is the size and variability of the nucleolus rather than the shape of the cell as suggested by Callender. They suggested measuring the standard deviation of the area of 200 nucleoli (SDNA) is a reliable and reproducible predictor of prognosis. However, such measurement is time consuming (about half an hour for each case), needs well-trained technical personnel, and a semi-automated image analysis system. Silver staining for nucleolar organizing regions can be done for such evaluation.

Nuclear organising regions (NOR)

NOR are the genes that code for ribosomal RNA. It has been shown that RNA content correlates with prognosis in patients with uveal melanoma. A recent study has indicated that higher mean NOR occurs in uveal melanomas in comparison to benign nevi. A significant correlation between number of NOR in uveal melanoma and computerized cytomorphometric analysis was found.[115]

Tumour size

Tumour size has been found to be an important predictor of prognosis of uveal melanoma. Largest tumour dimension has been found to be the most important prognostic indicator. However, limitations in measuring the size of an irregular tumour exist. Volume of the tumour (length x width) is reported to be strongly associated with the prognostic outcome.[116]

Mitotic activity

Mitotic figures in uveal melanoma are relatively less. One needs to count atleast 40 high power fields using 40x objective. Mitotic activity was found to correlate with prognosis.[117]

Tumour infiltrating lymphocytes

In approximately 5-12% of all uveal melanomas, lymphocytic infiltration are seen and in 74% of the tumours the infiltrating lymphocytes were predominantly T-cells.[118] Lymphocyte infiltration has been found to be associated with higher mortality.[118][119][120] An abnormal ganglioside profile on the surface of the melanoma cells was associated with mixed cell type and lymphocytic infiltration. An immune mechanism caused by antigencity of malignant melanoma has been suggested.[120],[121]

Vascular pattern

Several studies[120],[122],[123] have indicated that microvascular pattern of uveal malignant melanoma is another important indicator of prognosis. Nine vascular patterns seen on 10x objective field of the tumours are: "normal", "silent" with no vessels, "straight" with randomly distributed vessels, "parallel" with oriented straight vessels without cross linking, "parallel with cross linking" between the vessels, "arcs" or incomplete loops, "arcs with branching", "loops" of vessels that completely surround a lobule of tumour, and "networks" composed of atleast three back to back loops. Presence of networks were most strongly related with outcome. But both loops or parallel with cross linking pattern had survival functions similar to that associated with networks. Patients with these patterns have poor survival (40%) compared to their absence (80% survival). In another study of Japanese patients, the prognosis correlated with network pattern and closed loop pattern.[123]

Age and race

In a recent study, the average patient age was found to be younger in the Japanese (55.2 years) than in the white population.[123] Much younger age was been noted by us (unpublished data) in 60 patients of uveal melanoma. In the Kaplan-Meir survival curves comparing younger and older patients, the long term survival rates were similar.

Multivariate statistical analysis of various prognostic factors has been assessed by Mclean et al[117] which indicated best predictor for patient outcome are cell type and largest dimension of the tumour.[124]

Gamel-Boag's log-normal model is a statistical model that can be used to correlate prognostic features to the cure rate and median survival time. Using this model Gamel found larger tumours were associated with a much lower probability of cure and a much shorter median survival time.[124]

Detection of subclinical metastasis is still a challenge in malignant melanoma.[125] Forty percent of patients of uveal melanoma die within 10 years due to metastasis despite enucleation, and in only 1-2% of these casesmetastasis is detected prior to enucleation. In recent years, monoclonal antibodies have been used to detect subclinical metastasis for several cancers like breast carcinoma and uveal melanomas. There are two monoclonal antibodies, Mab8-1H and ME491, which have been found to detect melanoma associated antigen (MAA). Using these antibodies elevated serum MAA levels have been found in patients with documented metastatic choroidal melanoma. Subclinical metastasis of uveal melanoma has also been detected using monoclonal antibody.

Two candidate genes that encode uveal melanoma-specific antigens are currently being investigated.[126],[127] One encodes uveal melanoma associated antigens defined by monoclonal antibody Mab8-1H. The other is the tyrosinase gene which is tissue specific and can detect melanocytes from the peripheral blood. Use of polymerase chain reaction to detect subclinical metastasis in uveal melanoma is under investigation. Detection of malignant melanoma cells in peripheral blood by means of reverse transcriptase and polymerase chain reaction has been done in cutaneous malignant melanoma.

Metastatic tumours

Metastatic tumours to the eye usually involve the choroid. However, less frequently the iris and the ciliary body are involved by metastatic tumours. The most common site of primary in males are the prostate and lungs, in females, the breast. Another common site of primary is the gastrointestinal tract. Routine histology often indicates the morphologic pattern suggestive of the cell of origin of the primary site. Special stains and immunohistochemistry with a panel of primary antibodies, for example, prostate specific antigen can indicate the primary site.

 

References

1.	Shields JA, Annesly WH, Totino JA. Nonfluorescent malignant melanoma of the choroid diagnosed with the P-32 test. Am J Ophthalmol 1975;79:634-40.
2.	Leff SR, Augsburger JJ, Shields JA. Focal fluorescence of choroidal melanoma.Br J Ophthalmol 1986;70:104-106.   [PUBMED]
3.	Edwards WL, Layden WE, Macdonald Jr R. Fluorescein angiography of malignant melanoma of choroid. Am J Ophthalmol 1969;68:797-808.
4.	Pettit TL, Barton A, Foos RY, Christensen RE. Fluorescein angiography of choroidal melanomas. Arch Ophthalmol 1970;83:27-38.
5.	Nadel A, O'Connor P, Lincoff H. Nonfluorescent choroidal melanoma. Am J Ophthalmol 1970;70:748-52.   [PUBMED]
6.	Davis DL, Robertson DM. Fluorescein angiography of metastatic choroidal tumors. Arch Ophthalmol 1973;89:97-99.   [PUBMED]
7.	Shields JA, Shields CL. Intraocular Tumours: A Text and Atlas. Philadelphia: W.B. Saunders; 1992. p 223.
8.	Anand R, Augsburger JJ, Shields JA. Circumscribed choroidal haemangiomas. Retina 1981;1:56-61.
9.	MacLean AL, Maumenee E. Haemangioma of the choroid. Am J Ophthalmol 1960;50:3-11.
10.	Byrne SF, Green RL. Ultrasound of the Eye and Orbit. St. Louis: CV Mosby;1992.
11.	Shammas HJ: Atlas of Ophthalmic Ultrasonography and Biometry. St. Louis: CV Mosby; 1984.
12.	Saornil MA, Egan KM, Gragoudas ES, Seddon JM, Walsh SM, Albert DM. Histopathology of proton beam irradiated vs enucleated uveal melanomas. Arch Ophthalmol 1992;110:1112-18.   [PUBMED]
13.	Daubert GL, Puklin JE. Echographic patterns simulating extrascleral extension of malignant melanoma following plaque removal. In: Sampaolesi R, editor. Ultrasonography in Ophthalmology. Dordrecht: Kluwer Academic Publishers; 1990. p 345-49.
14.	Coleman DJ, Lizzi FL, Silverman RH, Ellsworth RM, Haik BH, Abramson DH. Regression of uveal malignant melanomas following Cobalt-60 plaque: correlates between acoustic spectrum analysis and tumour regression. Retina 1995;5:73-78.
15.	15. Zarate JO, Sampolesi R. Morphological parameters of intraocular tumours taking part in echographical tracings. In: Sampolesi R, editor. Ultrasonography in Ophthalmology. Dordrecht: Kluwer Academic Publishers; 1990. p 281-91.
16.	Harbour JW, Murray TG, Byrne SF, Hughes JR, Gendron EK, Ehlies FJ. Intraoperative echographic localisation of Iodine-125 episcleral radioactive plaques for posterior uveal melanomas. Retina 1996;16:129-34.
17.	Berger RW, Guthoff R, Helmke K, Winkler P. Doppler ultrasonography in.the follow-up of malignant melanoma of choroid. In: Sampolesi R, editor. Ultrasonography in Ophthalmology. Dordrecht: Kluwer Academic Publishers; 1990. p 327-31.
18.	Sneed RS, Byrne SF, Meiler WF, Nicholson DH, Olsen K, Hughes RJ. Choroidal detachment associated with malignant choroidal tumours. Ophthalmology 1991;98:963-70.
19.	Char DH, Stone RD, Irvine AR, Crawford JB, Hilton GF, Lonn LT. Diagnostic modalities in choroidal melanoma. Am J Ophthalmol 1980;89:223-30.
20.	De Potter P, Gonzalez CF, Flanders AE, Shields JA, Shields CL. Imaging studies of intraocular tumours. In: Alberti WE, Sagerman RH, editors. Radiotherapy of Intraocular and Orbital Tumours. Berlin: Springer-Verlag; 1993. p 295-309.
21.	Mafee MF, Peyman GA. Retinal and choroidal detachments: the role of magnetic resonance imaging and computed tomography. Rad Clin North Am 1987;25:487-507.
22.	Peyman GA, Mafee MF. Uveal melanoma and simulating lesions: the role of magnetic resonance imaging and computed tomography. Rad Clin North Am 1987;25:471-86.
23.	Shields JA, Shields CL. Intraocular Tumours: A Text and Atlas. Philadelphia: W.B. Saunders; 1992. p 155-69.
24.	Mafee MF, Peyman GA, Me Kusick MA. Malignant melanoma and similar lesions studied by computed tomography. Radiology 1985;156:403-8.
25.	Schulman JA, Peyman GA, Mafee MF, Lawrence L, Bauman AE, Goldman A. The use of magnetic resonance imaging in the evaluation of retinoblastoma. J Ped Ophthalmol Strab 1986;23:144-47.
26.	Sealy RC. Radicals in melanin biochemistry. Methods Enzymol 1984;105:479-83.   [PUBMED]
27.	Chambers RB, Davidorf FH, McAdoo JF, Chakeres DW. Magnetic resonance imaging of uveal melanomas. Arch Ophthalmol 1987;105:917-21.   [PUBMED]
28.	Gomori JM, Grossman RI, Shields JA, Augsburger JJ, Joseph PM, DeSimeone D. Choroidal melanomas: correlation of NMR spectroscopy and MR imaging. Radiology 1986;158: 443-45.   [PUBMED]
29.	De Potter P, Flanders AE, Shields JA, Shields CL. Magnetic resonance imaging of intraocular rumours. Int Ophthalmol Clin 1993;33:37-45.   [PUBMED]
30.	De Potter P, Flanders AE, Shields JA, Shields CL, Gonzales CF, Rao VM. The role of fat suppression technique and gadolinium dimeglumine in the MRI evaluation of intraocular tumours and simulating lesions. Arch Ophthalmol 1994;112:340-48.   [PUBMED]
31.	De Potter P, Shields JA, Shields CL. Magnetic Resonance Imaging of the Eye and Orbit. Philadelphia: Lippincott; 1995.
32.	Mafee MF, Peyman GA, Peace JH, Cohen SB, Mitchell MW. Magnetic resonance imaging in the evaluation of and differentiation of uveal melanoma. Ophthalmology 1987;93:341-48.
33.	Kolodny NH, Gragoudas ES, D'Amico DJ, Kohler SJ, Seddon JM, Murphy EJ. Proton and sodium-23 magnetic resonance imaging of human ocular tissues: a model study. Arch Ophthalmol 1987;105:1532-36.
34.	Bond JB, Haik BG, Mihara F, Gupta KL. Magnetic resonance imaging of choroidal melanoma with and without gadolinium contrast enhancement. Ophthalmology 1991;98:459-66.   [PUBMED]
35.	Mihara F, Gupta KL, Murayama S, Lee N, Bond JB, Haik BG. MR imaging of malignant uveal melanoma: role of pulse sequence and contrast agent. Am J Neuro Rad 1991;12:991-96.
36.	Wilms G, Dralands G, Van Fraeyenhoven L, Decrop E, Demaerel P, Van Eyck L. Magnetic resonance imaging in lesions of the eye globe. J Beige Rad 1989;72:165-72.
37.	De Potter P, Shields JA, Shields CL, Yannuzzi LA, Fisher YE, Rao VM. Unusual MRI findings in metastatic carcinoma to the choroid and optic nerve: a case report. Int Ophthalmol 1992;16:39-44.   [PUBMED]
38.	Jackobiec FA, Coleman DJ, Chattock A. Ultrasonically guided needle biopsy and cytologic diagnosis of solid intraocular tumours. Ophthalmology 1979;86:1662-78.
39.	Augsburger JJ, Shields JA. Fine needle biopsy of solid intraocular tumours: indications, instrumentation and techniques. Ophthalmic Surg 1984;15:34-40.   [PUBMED]
40.	Shields JA, Shields CL, Ehya H, Eagle RC, De Potter P. Fine needle aspiration biopsy of suspected intraocular rumours. Ophthalmology 1993;100:1677-84.
41.	Midnea E, Segato T, Piermarocchi S, Boccato P. Fine needle aspiration biopsy in ophthalmology. Surv Ophthalmol 1985;29:410-22.
42.	Karcioglu ZA, Gordon RA, Karcioglu GL: Tumour seeding in ocular fine needle aspiration biopsy. Ophthalmology 1985;92:1763-67.
43.	Glasgow BJ, Brown HH, Zargoza AM, Foos RY. Quantitation of tumour seeding from fine needle aspiration of ocular melanomas. Am J Ophthalmol 1988;105:538-46.   [PUBMED]
44.	Shields CL, Shields JA, Shields MB, Augsburger JJ. Prevalence and mechanisms of secondary intraocular pressure elevation in eyes with intraocular tumours. Ophthalmology 1987;94:839-46.   [PUBMED]
45.	Shields JA, Shields CL, Shields MB. Glaucoma associated with intraocular tumours. In: Ritch R, Shields MB, Krupin X editors. The Glaucomas. St Louis: CV Mosby; 1989. Vol 2. p 1111-24.
46.	Barr CC, Sipperly JO, Nicholson DH. Small melanomas of the choroid. Arch Ophthalmol 1978;96:1580-82.
47.	Zimmerman LE, Mc Lean IW. Do growth and onset of symptoms of uveal melanomas indicate subclinical metastasis. Ophthalmology 1984;91:685-91.
48.	Shields JA, Shields CL, Donoso LA. Management of posterior uveal melanoma. Surv Ophthalmol 1991;36:161-95.   [PUBMED]
49.	Fraunfelder FT, Boozman FW, Wilson RS, Thomas AH. No touch technique for intraocular malignant tumours. Arch Ophthalmol 1977;95:1616-20.   [PUBMED]
50.	Migdal C. Effect of the method of enucleation on the prognosis of choroidal melanoma. Br J Ophthalmol 1983;67:385-88.   [PUBMED]
51.	Kramer KK, La Piana FG, Whitmore PV. Enucleation with stabilization of intraocular pressure in treatment of uveal melanomas. Ophthalmol Surg 1980;11:39-43.
52.	Zimmerman LE. Malignant melanoma of the uveal tract. In: Spencer WH, Font RL, Green WR, editors. Ophthalmic Pathology: An Atlas and Textbook. 3rd ed. Philadelphia: W.B. Saunders; 1986. p 2113.
53.	Federman JL, Felberg NT, Shields JA. Effect of local treatment on antibody levels in malignant melanoma of choroid.Trans Ophthalmol Soc UK 1977;97:436-39.   [PUBMED]
54.	Michelson JB, Felberg NT, Shields JA. Subretinal fluid examination of LDH, PGI and CEA in a case of metastatic bronchogenic carcinoma of the choroid. Cancer 1978;41:2301-4.
55.	Shields JA, Shields CL. Surgical approach to lamellar sclerouvectomy for posterior uveal melanomas: the Schoenberg lecture. Ophthalmic Sur 1988;19:774-80.
56.	Peyman GA, Juarez CP, Diamond JG, Raichand M. Ten years experience with eye wall resection of uveal melanomas. Ophthalmology 1984;91:1720-24.   [PUBMED]
57.	Peyman GA, Charles H. Internal eye wall resection in the management of uveal melanoma. Can J Ophthalmol 1988;23:219-23.
58.	Peyman GA, Schulman JA. Intravitreal Surgery: Principles and Practice. 2nd ed. Connecticut: Appleton Lange;1994. p 777-811.
59.	Bell RM, Bullock JD, Albert DM. Solitary choroidal metastasis from bronchial carcinoid. Br J Ophthalmol 1975;59:155-63.   [PUBMED]
60.	Fu YS, Me Williams N, Stratford TP, Kay S. Bronchial carcinoid with choroidal metastasis in an adolescent. Cancer 1974;33:707-15.
61.	Rodrigues MM, Shields JA. Iris metastasis from a bronchial carcinoid tumour.Arch Ophthalmol 1978;96:77-83.
62.	Peyman GA, Schulman JA, Raichand M. Diagnosis and therapeutic surgery of the uvea. II. Results. Ophthalmic Surg 1987;18:310-16.
63.	Char DH. Clinical Ocular Oncology. New York: Churchill Livingston; 1989. p 115-17.
64.	Straatsma BR, Fine SL, Earle JD, Hawkins BS, Diener-West M, MCLaughlin JA, et al. Enucleation versus plaque irradiation for choroidal melanoma. Ophthalmology 1988;95:1000-4.
65.	Meyer-Schwickerath G. Photocoagulation of choroidal melanoma. Doc Ophthalmol 1980;50:57-61.   [PUBMED]
66.	Shields JA. The expanding role of laser photocoagulation for intraocular tumours: the 1993 H. Christian Zweng Memorial lecture. Retina 1994;14:310-22.   [PUBMED]
67.	Karlin DB. Lasers in Ophthalmic Surgery. Cambridge: Blackwell Science Inc; 1995. p 190-96.
68.	L'esperance FA. Ophthalmic Lasers. 3rd ed. St Louis: CV Mosby; 1989. p 460-65.
69.	Chong LP, Ozler SA, Ligget E. Indocyanine green enhanced diode laser treatment of melanoma in a rabbit model. Retina 1993;13:251-59.
70.	Johnston PB, Lotery AJ, Logan WC. Treatment and long term follow up of a capillary angioma of the optic disc. Invest Ophthalmol Vis Sci 1995;19:129-32.
71.	Klein ML, Wilson DJ. Cryotherapy for the treatment of small growing choroidal melanomas. Abstract. Invest Ophthalmol Vis Sci 1996;37:S242.
72.	Char DH, Philips TL, Andejeski Y, Crawford JB, Kroll MA. Failure of pre-enucleation radiation to decrease uveal melanoma mortality. Am J Ophthalmol 1988;106:21-26.
73.	Sobanski J, Zeydler-Gredzielewska L, Szusterowska-Martinowa E. Decreased mortality of patients with intraocular malignant melanoma after enucleation of the eyeball followed by orbit x-ray irradiation. Pol Med J 1972;11:1512-16.
74.	Gragoudas ES, Char DH. Charged particle irradiation of uveal melanomas. In: Albert DM, Jakobiec FA, editors. Principles and Practice of Ophthalmology. Philadelphia: W.B. Saunders; 1994. p 3233-43.
75.	Gragoudas ES, Seddon JM, Egan KM, Glynn RJ, Munzerider J, Austin- Seymour M, et al. Long term results of proton beam irradiated uveal melanomas. Ophthalmology 1987;94:349-53.
76.	Char DH, Castro JR, Kroll SM, Irvine AR, Quivey JM, Stone RD. Five year follow-up of helium ion therapy for uveal melanoma. Arch Ophthalmol 1990;108:209-14.   [PUBMED]
77.	Gragoudas ES, Seddon JM, Egan KM, Glynn RJ, Goietein M, Munzenrider J, et al. Metastasis from uveal melanoma after proton beam irradiation. Ophthalmology 1988;95:992-99.
78.	Egger E, Zografas L, Perret C, Gailloud C. Proton beam irradiation of choroidal melanomas at the PSI: technique and results. In: Alberti WE, Sagerman RH, editors. Radiotherapy of Intraocular and Orbital Tumours. Berlin: Springer-Verlag; 1993. p 58-72.
79.	Anand R, Augsburger JJ, Shields JA. Circumscribed choroidal haemangiomas. Arch Ophthalmol 1989;107:1338-42.   [PUBMED]
80.	Shields JA, Shields CL. Intraocular Tumours: A Text and Atlas. Philadelphia: W.B. Saunders; 1992. p 239-59.
81.	Alberti WE. Clinical features and management of choroidal haemangiomas including those occurring in association with Sturge Weber syndrome. In: Alberti WE, Sagerman RH, editors. Radiotherapy of Intraocular and Orbital Tumours. Berlin: Springer-Verlag; 1993. p 87-92.
82.	Scott TA, Augsburger J, Brady LW, Herbabdez C, Woodleigh R. Low dose irradiation for diffuse choroidal haemangiomas associated with bullous non-rhegmatogenous retinal detachment. Retina1991;11:389-93.
83.	Plowman PN, Harnett AN. Radiotherapy in benign orbital disease. I. Complicated ocular angiomas. Br J Ophthalmol 1988;72:286-88.   [PUBMED]
84.	Jaeger EA, Frayer WC, Southard ME, Kramer S. Effect of radiation therapy on metastatic choroidal tumours. Trans Am Acad Ophthalmol Otolaryngol 1971;75:94-101.   [PUBMED]
85.	McCormick B, Harrison LB. Radiation therapy of choroidal metastasis. In: Alberti WE, Sagerman RH, editors. Radiotherapy of Intraocular and Orbital Tumours. Berlin: Springer-Verlag; 1993. p 93-97.
86.	Stephen RF, Shields JA. Diagnosis and management of cancer metastatic to uvea: a study of 70 cases. Ophthalmology 1979;86:1336-49.
87.	Letson AD, Davidorf FH, Bruce RA. Chemotherapy for treatment of choroidal metastases from breast carcinoma. Am J Ophthalmol 1982;93:102-6.
88.	Stark WJ, Rosenthal AR, Mullins GM, Green WR. Simultaneous bilateral uveal melanomas responding to BCNU therapy. Trans Am Acad Ophthalmol Otolaryngol 1971;75:70-83.
89.	Sellami M, Weil M, Dhermy P, Aron-Rosa D, Jacquillat C. Adjuvant chemotherapy in ocular malignant melanoma. Oncology 1986;43:221-23.   [PUBMED]
90.	Olsen KR, Blumencranz M, Hernandez E, Hajek A, Hartzer M. Fluorouracil therapy of intraocular green melanoma in the rabbit. Arch Ophthalmol 1988;106:812-15.
91.	Fraunfelder FT, Wilson DJ, Moseley HS. Adjuant chemotherapy and "no touch" enucleation for choroidal melanoma. Abstract. Invest Ophthalmol Vis Sci 1996;37:S243.
92.	Shields JA, Shields CL. Intraocular Tumours: A Text and Atlas. Philadelphia: W.B. Saunders; 1992. p 117-36.
93.	Seddon JM, Egan KM, Gragoudas ES. Epidemiology of uveal melanoma. In: Ryan SJ, editor. Retina. 2nd ed. St.Louis: CV Mosby; 1994. p 717-24.
94.	Shields JA, Shields CL, De Potter P, Cu-Unjieng A, Hernandez C, Brady LW. Plaque radiotherapy for uveal melanoma. Int Ophthalmol Clin 1993;33:129-35.   [PUBMED]
95.	Packer S, Stoller S, Lesser ML, Mandel FS, Finger PT. Long term results of Iodine-125 irradiation of uveal melanoma. Ophthalmology 1992;99:767-74.   [PUBMED]
96.	Robertson DM, Earle J, Kline RW. Brachytherapy for choroidal melanoma. In: Ryan SJ, editor. Retina. 2nd ed. St. Louis: CV Mosby; 1994. p 773-84.
97.	Cruess AF, Augsburger JJ, Shields JA, Donoso LA, Amsel J. Visual results following cobalt plaque radiotherapy for posterior uveal melanoma. Ophthalmology 1984;91:131-36.   [PUBMED]
98.	De Potter P, Shields CL, Shields JA, Cater JR, Brady LW. Plaque radiotherapy for juxtapapillary choroidal melanoma: visual acuity and survival outcome. Arch Ophthalmol 1996;114:1357-65.   [PUBMED]
99.	Lommatzsch PK, Alberti W, Lommatzch, Rohrwacher F. Radiation effects on the optic nerve observed after brachytherapy of choroidal melanomas with 106Ru/106Rh plaques. Graefe's Arch Clin Exp Ophthalmol 1994;232:482-87.
100.	Vrabec TR, Augsburger JJ, Gamel JW, Brady LW, Hernandez Z, Woodleigh R. Impact of local tumour relapse on patient survival after cobalt-60 plaque radiotherapy. Ophthalmology 1991;98:984-88.
101.	Augsburger JJ, Gamel JW, Sardi VF, Greenberg RA, Shields JA, Brady LW. Enucleation versus cobalt plaque radiotherapy for malignant melanomas of the choroid and ciliary body. Arch Ophthalmol 1986;104:655-61.
102.	Adams KS, Abramson DH, Ellsworth RM, Haik BG, Bedford M, Seddon J. Cobalt plaque versus enucleation for uveal melanoma: comparison of survival rates. Br J Ophthalmol 1988;72:494-97.
103.	De Potter P, Shields CL, Shields JA, Cater JR, Tardio DJ. Impact of enucleation versus plaque radiotherapy in the management of juxtapapillary choroidal melanoma on patient survival. Br J Ophthalmol 1994;78:109-114.
104.	Augsburger JJ, Gamel JW, Shields JA, Markoe AM, Brady LW. Post-irradiation regression of choroidal melanomas as a risk factor for death from metastatic disease. Ophthalmology 1987;94:1173-77.
105.	Guthoff R, Haase J, von Domarus D, Draeger J, Lauritzen K. Das refgressionsverthalten des alderhautmelanomas nach strahlentherapie: ein neuer prognostischer parameter? Klin Monatsbl Augenheilkd 1990;196:6-10.
106.	Glynn RJ, Seddon JM, Gragoudas ES, Egan KM, Hart LJ. Evaluation of tumour regression and other prognostic factors for early and late metastasis after proton irradiation of uveal melanoma. Ophthalmology 1989;96:1566-73.
107.	Augsburger JJ, Mullen D, Kleineidan M: Planned combined I-125 plaque irradiation and indirect ophthalmoscope laser therapy for choroidal melanoma. Ophthalmic Surg 1993;24:76-81.
108.	Finger P. Microwave plaque thermoradiotherapy for choroidal melanoma. Br J Ophthalmol 1992;76:358-64.
109.	Shields CL, Shields JA, De Potter P, Singh AD, Hernandez C, Brady LW. Treatment of non-resectable malignant iris tumours with custom designed plaque radiotherapy. Br J Ophthalmol 1995;79:306-12.
110.	Shields CL, Shields JA, De Potter P, Quaranta M, Freire J, Brady LW. Plaque radiotherapy for the management of choroidal metastasis. Arch Ophthalmol 1997;115:203-9.
111.	Shields JA, Shields CL. Intraocular Tumours: A Text and Atlas. Philadelphia: W.B. Saunders; 1992. p 393-419.
112.	Kincaid MC, Grossniklaus HE. Immunohistochemical staining in ophthalmic pathology. Ophthamol Clin North Am 1995;8:17-24.
113.	Seregard S, Oskarsson M, Spangberg B. PC-10 as a predictor for prognosis after antigen retrieval in posterior uveal melanoma. Invest Ophthalmol Vis Sci 1996;37:1451-58.
114.	Gamel JW, Mclean IW, Greenberg RA, Lichenstein SJ. Computerized histologic assessment of malignant potential: a method for determining the prognosis of uveal melanomas. Hum Pathol 1982;13:893-97.
115.	Marcus DM, Minkouitz JB, Wardell SD, Albert DM. The value of nucleolar organizer regions in uveal melanoma. Am J Ophthalmol 1990;110:527-34.
116.	Flocks M, Gerende JH, Zimmerman LE. The size and shape of malignant melanoma of the choroid and ciliary body in relation to prognosis and histologic characteristics: a statistical study of 210 tumours. Trans Am Acad Ophthalmol 1955;59:740-58.
117.	McLean IW, Foster WD, and Zimmerman LE. Prognostic factors in small malignant melanoma of choroid and ciliary body. Arch Ophthalmol 1977;95:48-58.
118.	de la Cruz Jr PO, Specht CS, McLean IWL. Lymphocytic infiltration in uveal malignant melanoma. Cancer 1990;65:112-15.
119.	Vit VV. Prognostic role of morphologic characteristics of the immune response in uveal melanoblastomas of various cellular types (in Russian). Arkh Patol 1983;45:25-30.
120.	Folberg R, Rummelt V, Parys-Van Ginderdeuren R, Hwang T, Woolson RF, Pe'er J. The prognostic value of tumour blood vessel morphology in primary uveal melanoma. Ophthalmology 1993;100:1389-98.
121.	Kanda S, Cochran AJ, Lee WR, Morton DL, Irie RF. Variations in the ganglioside profile of uveal melanoma correlate with cytologic heterogeneity. Int J Cancer 1992;52:682-87.
122.	Rummelt V, Folberg R, Rummelt C, Gruman LM, Hwang T, Woolson RF. Microcirculation architecture of melanocytic nevi and malignant melanoma of the ciliary body and choroid: a comparative histopathologic and ultrastructural study. Ophthalmology 1994;101:718-27.
123.	Sakamoto T, Sakamoto M, Yoshikawa H, Hata Y, Ishibash T, Ohnishi Y. Histologic findings and prognosis of uveal malignant melanoma in Japanese patients. Am J Ophthalmol 1996;121:276-83.
124.	Gamel JW, McLean IW, McCurdy JB. Biologic distinctions between cure and time to death in 2892 patients with intraocular melanoma. Cancer 1993;71:2299-305.
125.	Wang M, Sheilds JA, Donoso LA. Subclinical metastasis of uveal melanoma. Int Ophthalmol Clin 1993;33:119-27.
126.	Donoso LA, Folberg R, Naids R Augsburger JJ, Shields JA, Atkinson B. Metastatic uveal melanoma: hepatic metastasis identified by hybridoma-secreted monoclonal antibody MAb8-IH. Arch Ophthalmol 1985;103:799-801.
127.	Smith R, Selby P, Southgate J, Pittman K, Bradley C, Blair GE. Detection of melanoma cells in peripheral blood by means of reverse transcriptase and polymerase chain reaction. Lancet 1991;338:1227-29.

Figures

  [Figure - 1], [Figure - 2], [Figure - 3], [Figure - 4], [Figure - 5], [Figure - 6], [Figure - 7], [Figure - 8], [Figure - 9], [Figure - 10], [Figure - 11], [Figure - 12]
 
 

Tables

  [Table - 1], [Table - 2], [Table - 3]