«1 Authors’ Note 2 Introduction 3 Uveal melanoma 3.1 Precursor lesions (indeterminate lesions) 3.2 Primary choroidal and ciliary body melanoma 3.3 ...»
Clinical Practice Guidelines for the Management of
Melanoma in Australia and New Zealand
Ocular and Periocular Melanoma: Supplementary Document
1 Authors’ Note
3 Uveal melanoma
3.1 Precursor lesions (indeterminate lesions)
3.2 Primary choroidal and ciliary body melanoma
3.3 Iris melanoma
3.4 Prognosis of treated uveal melanoma patients
3.5 Follow-up of treated uveal melanoma patients
3.6 Metastatic uveal melanoma
4 Conjunctival melanoma
4.1 Precursor lesions
4.2 Primary conjunctival melanoma
4.3 Prognosis of treated conjunctival melanoma patients
4.4 Follow-up of treated conjunctival melanoma patients
4.5 Treatment of node-positive disease
4.6 Metastatic conjunctival melanoma 5 Eyelid melanoma 6 Orbital melanoma 7 Metastatic melanoma and periocular tissues 8 Psychological aspects of ocular melanoma Management of Ocular and Periocular Melanoma: Supplementary Document 1 Authors’ Note The Australian Cancer Network, in accordance with the National Health and Medical Research Council (NHMRC), has recently redeveloped the evidence-based Clinical Practice Guidelines for the Management of Melanoma. A previous edition of the guidelines published in 1999 had been used widely throughout Australia and New Zealand and translated into several languages for overseas use. The current guidelines, published by the Australian Cancer Network and approved by the NHMRC as a handbook and a webbased resource in 2008, contain a chapter and supplementary document on ocular and periocular melanoma (see www.cancer.org.au/skincancerguides, Ocular Melanoma; Chapter 24). The supplementary document appears below. This is the first NHMRC compilation of evidence-based guidelines for the management of ocular melanoma.
NHMRC guidelines are created by members of an expert working group from a broad range of disciplines each contributing a chapter. The guidelines are compiled according to NHMRC standards of quality and process. The aims are dual: to create concise recommendations for the management of ocular and periocular melanoma in the Australasian context and to grade the strength of the supporting evidence according to the level of the evidence base. Details regarding the process of the formation and grading of the guidelines may be found on the NHMRC website.
Uveal, conjunctival and cutaneous melanomas are malignant proliferations of melanocytes derived from neural crest cells. Despite similar origins they have different epidemiology, molecular biology, pathogenesis and metastatic pattern. This supplementary document focuses on evidence-based aspects of management. For detailed information regarding the basic science, epidemiology and pathogenesis of ocular, periocular and cutaneous melanoma, the reader is referredto the complete NHMRC document and the many excellent comprehensive reviews in the literature.
The two principal types of primary ocular melanoma are uveal and conjunctival melanoma. Of the two, uveal melanoma is more common, with an incidence of four to six per million per year in most western countries.2 90 percent of uveal melanomas arise in the choroid, seven percent in the ciliary body and three percent in the iris.3 Based on these data, the predicted Australian incidence is 90–100 cases per year of choroidal melanoma.4 The frequency of conjunctival melanoma is one in 20 to one in 40 that of uveal melanoma, with an incidence of only 0.1–0.3 per million in most western populations.5 Periocular melanoma includes eyelid and orbital melanoma. Primary eyelid melanoma has a similar incidence to conjunctival melanoma6 whilst primary orbital melanoma is exceedingly rare. Melanoma metastatic to the eye or periocular sites is recognised, but also very rare.
The Collaborative Ocular Melanoma Study (COMS), undertaken in 1985, is the only source of level II evidence for the management of primary ocular melanoma. Participants were subdivided into three trials for small, medium and large choroidal melanoma. This trial has confirmed the role of brachytherapy in the treatment of medium-sized and small choroidal tumours, and has shown that pre-enucleation irradiation for large tumours has no survival benefit.7-9 Another recent advance in management is the identification of several risk factors for malignancy of small choroidal tumours.10 This has led to improved clinical detection with more accurate selection of cases.
The management of ocular and periocular melanoma is a rapidly changing and expanding field. The aim of these guidelines is to highlight the various issues involved in the treatment of ocular and periocular melanoma.
Management of Ocular and Periocular Melanoma: Supplementary Document 3 Uveal melanoma
3.1 Precursor lesions (indeterminate lesions) Uveal melanoma originates either de novo or from degeneration of an existing naevus, but the proportion of each is unknown. Six percent of individuals have choroidal naevi, of which one in 5000 is estimated to undergo malignant degeneration.11 Much effort has been directed at determining which choroidal naevi have malignant potential or are small melanomas and require treatment, however diagnosis remains controversial. Specific precursor lesions for ciliary body and iris melanoma are less well studied, however are presumed to be similar to those for the choroidal melanoma.
In general, tumour growth over a short period (months) is considered a hallmark of choroidal melanoma.
One arm of the COMS followed 204 patients with small choroidal melanocytic lesions (3mm thick, 10mm in diameter) over five years to discover that 31% demonstrated growth, which has been found to be associated with a 3.2 relative risk of malignancy.10,12,13 However, naevi may also demonstrate growth, usually slowly over a period of years.
Additionally a number of qualitative risk factors including tumour thickness (2mm), orange pigment, absence of drusen, visual symptoms, proximity to the optic disc, subretinal fluid and absence of adjacent retinal pigment epithelial changes have been observed to be predictive of tumour growth.10,12,13,14 Lesions which display such features have an elevated potential for malignant transformation and have been termed indeterminate lesions or naevomas (see table 1).
The management of indeterminate lesions remains controversial and may vary considerably from one centre to another. At some centres, definitive treatment for indeterminate lesions is preferred while at others, indeterminate lesions (especially if involving only one or two risk factors) are observed more often (usually three to six monthly) for growth using clinical photography and ultrasound. Transvitreal biopsy is used in some centres to improve diagnostic accuracy.15 Treatment methods in relation to indeterminate lesions are similar to those for small melanoma (see Primary choroidal and ciliary body melanoma, below).
Uveal melanoma is more common in people with light coloured eyes with increasing frequency in older age groups, and with no sex predilection. Unlike cutaneous melanoma, the link to sunlight exposure is weaker. Certain conditions predispose to uveal melanoma, including oculodermal melanocytosis, neurofibromatosis type 1, familial cutaneous melanoma and cutaneous dysplastic naevus syndrome.
3.2 Primary choroidal and ciliary body melanoma
The clinical presentation of choroidal and ciliary body melanoma depends on the size and location of the lesion. Clinical diagnosis aided by ultrasound, which has characteristic features for uveal melanoma and can detect extraocular extension, has been demonstrated to have a very high level of accuracy in experienced hands.16 Ultrasound biomicroscopy is particularly useful for the evaluation of ciliary body melanoma. Fluorescein angiography may provide useful ancillary information, while indocyanine green may be useful to detect complex microvascular patterns predictive of tumour growth.17 CT or MRI imaging studies are useful to exclude extraocular extension and are used to plan radiotherapy at some centres.3,5 In addition, a systemic evaluation is usually performed at presentation to assess for metastatic disease.
Biopsy is performed more liberally at some centres than others. It is generally indicated if other tests have failed to elucidate the diagnosis. The reported diagnostic accuracy varies depending on tumour size and location, diameter of needle, sampling technique and other factors.18 With better techniques, fewer Clinical Practice Guidelines for the Management of Melanoma in Australia and New Zealand 4 complications and greater understanding of the importance of cytogenetics, some authors feel that it will become more common in the future, particularly if the cytogenetic profile is used in treatment planning.19 There are risks, however, in terms of complications (e.g. vitreous haemorrhage) and the theoretical risk of spreading the tumour extra-ocularly. Sampling variability due to tumour cell heterogeneity with respect to the cytogenetic analysis of tumours also needs to be recognised.
The management of choroidal melanoma has changed considerably in recent years. Until the late 1970s, enucleation was the treatment of choice.20,21 Today, eye-conserving plaque radiotherapy is the most common treatment. Other forms of treatment include periodic observation, transpupillary thermotherapy, charged particle irradiation, local tumour resection, enucleation and exenteration.14,22,23 The choice of treatment depends upon the size, location and growth of the tumour; the visual acuity, intraocular pressure and retinal changes in the affected eye; the patient’s age, general health, wishes and psychological status; and the status of the other eye.3,5,23 The various treatment options are outlined in Table 2 and are described in detail in the literature.
3.3 Iris melanoma
Ninety percent of iris melanomas are discrete circumscribed lesions with an excellent prognosis.40 The remainder are diffuse or ring variants with a poorer prognosis. At many centres, suspicious circumscribed lesions are observed and intervention initiated if there is documented growth. Ultrasound biomicroscopy and clinical photography are particularly useful for documenting growth. Besides growth tendency, other factors that may initiate treatment include large size, multifocal tumours, tumour-related glaucoma, spontaneous hyphaema, involvement of the subjacent ciliary body or extrascleral extension. Primary treatment is usually surgical resection or various forms of radiotherapy (e.g. plaque brachytherapy or charged particle therapy). Enucleation remains the treatment of choice for tumours unsuited to conservative modalities. The majority of metastatic events occur when drainage procedures are inadvertently performed for tumour-induced raised intraocular pressure or ring melanomas, or when the iris melanoma is an extension of ciliary body or diffuse uveal melanoma.
3.4 Prognosis of treated uveal melanoma patients Metastatic uveal melanoma is the cause of death in approximately 50% of all patients with the disease.
The prognostic factors associated with this are as follows:
Size: Choroidal and ciliary body tumours are primarily divided into small (3mm height, 5-10mm diameter), medium (3-5mm height, 10-15mm diameter) and large (5mm height, 15mm diameter) subgroups.47 Metastatic rate increases with the size of tumour, with the largest tumour diameter being more important than depth.48 Small, medium and large tumours have 16%, 32% and 53% 5 year all-cause mortality, and 35%, 60% and 81% 10-year all cause mortality.7 In addition, tumour size was found on meta-analysis to be the most important prognostic factor for survival in the COMS medium-sized tumour trial.7 Ciliary body involvement: Ciliary body tumours present later, larger and are more likely to invade the angle.49 Ring melanoma involving the ciliary body and iris metastasis frequently has a poor prognosis.50 In contrast, iris melanomas have a lower incidence of systemic metastases than other uveal melanomas, and have a 2% long-term tumour-related mortality.2,5 Cell type: Two melanoma cell types are common: epithelioid and spindle cell. Epithelioid cells have a worse prognosis than spindle cell, and a mixed pattern has an intermediate prognosis.51 Management of Ocular and Periocular Melanoma: Supplementary Document Histopathology: Further histological poor prognostic markers include extravascular matrix patterns consisting of closed periodic acid-Schiff positive loops in the tumour, increased microvascular density, higher cell proliferation rate/mitoses, lymphocytic tumour invasion, and macrophage tumour invasion.3 Human Lymphocyte Antigen (HLA) class I and II expression: Expression of HLAs on melanoma cell membranes are associated with a poor prognosis, and may indicate a protective role of Natural Killer lymphocytes.52 Chromosomal abnormalities: Genomic changes of uveal melanoma have been studied using comparative genomic hybridisation.53 Loss of chromosome three occurs early, and is highly predictive of metastatic risk. Other genetic changes have also been found to have prognostic significance and genetic profiling of tumours is now able to provide reliable prognostic information.
Local tumour recurrence after conservative treatment: Local tumour recurrence after conservative treatment is associated with a fourfold increase risk of metastasis. This may reflect a causal link between recurrence and metastasis, or a more aggressive cell line.3
3.5 Follow-up of treated uveal melanoma patients