Treatment options

When considering the treatment option for uveal melanoma, factors such as tumour size, activity, location, growth pattern, the patient’s general health, age and status of the fellow eye must be taken into consideration.Reference Sagerman and Alberti⁴Table 1 demonstrates the treatment options according to tumour size.

Table 1. Treatment options according to tumour size

*Used for tumours that show clinical risks for metastasis.

**Preferred method due to ocular intolerance of conservative methods.Reference Sagerman and Alberti^4,Reference Shields, Naseripour, Cater, Shields, Demirci, Youseff and Freire³⁰, Reference Augsburger^45–Reference Peyman, Juarez, Diamond and Raichand⁴⁸

Of these, episcleral plaque brachytherapy (EPB) and proton therapy are commonly accepted treatment options, as summarised in Table 2, and are more favourable than conventional radiotherapy due to their physical characteristics.Reference Wilson and Hungerford^5–Reference Munzenrider and Szechter⁷ EPB treats a small volume with an extremely heterogeneous dose distribution. A radioactive plaque is used to produce γ or β rays. The dose to the tumour base can be greater than or equal to the dose to the tumour apex. The plaque is placed directly over the tumour allowing a high dose to be delivered with rapid fall off in dose to adjacent normal tissue.

Table 2. Brachytherapy and proton therapy studies

Proton therapy involves highly conformal target volume shapes. Clinically useful protons are produced using a particle accelerator such as cyclotron or synchrotron. The high-dose region or “Bragg peak” is narrow due to the proton beams being monoenergetic; this allows for a uniform dose to be delivered to the entire tumour with minimal entrance and exit dose (Figure 1).

Figure 1. Depth-dose distribution in tissue of a single Bragg peak (grey), a spread-out Bragg peak (black), which consists of multiple Bragg peaks of different energy added together, and a 10 MV X-ray beam (dashed). Reference Yock and Tarbell⁴¹Both EPB and proton therapy offer improved conformality and preservation of normal tissue compared to conventional radiotherapy. Reference Sagerman and Alberti^4,Reference Landwehr, Umbach and Haase^42–Reference Delvin⁴⁴EPB, episcleral plaque brachytherapy.

Normal tissue morbidity and local tumour control

Due to the anatomical location of ocular melanomas, critical structures are abundant and in close proximity to each other. Radiation damage to key structures within the eye causes increased morbidity and consequent quality-of-life issues.Reference Melia, Moy, Reynolds, Hayman, Murray, Hovland, Earle, Kurinij, Dong, Miskala, Fountain, Cella and Mangione⁸ Accurate and highly localised treatment are paramount in order to achieve as little normal tissue morbidity as possible while providing good tumour control probability.Reference Regillo, Brown and Flynn⁹ Many studiesReference Damato, Kacperek, Chopra, Campbell and Errington^10–Reference Gündüz, Shields, Shields, Cater, Freire and Brady¹³ have investigated the normal tissue morbidity and local tumour control in patients treated with EPB or proton therapy for choroidal melanoma. Within this review, a comparison of the outcomes in patients treated with the two modalities was made to determine which treatment option is superior. There is currently a lack of clinical availability of proton therapy facilities, and this is largely associated with the cost of proton centres.Reference Thornton, Fitzek, Klein, Dickey, Sullivan, O’Ryan-Blair, Schreuder, Allgower, Mascia, Farr and Cameron¹⁴ Whether treatment outcomes can justify additional proton centres will also be considered.

Design limitations

Much of the literature reviewed consisted of studies that were single centre, retrospective, cohort studies. To generalise findings from a single centre and apply them to other centres can introduce bias into the findings. Retrospective studies use data that was not necessarily collected for research purposes; therefore, they may lack essential information. Reviewed articles were scrutinised for evidence of potential bias during the critical review process.

Reliability and validity

The reliability and validity of the literature reviewed in this research project was ensured by use of the CASP literature appraisal tool.¹⁵ By applying a series of questions to the literature, it was possible to assess the credibility, reliability and validity of the article.

Outcome measures

Ocular preservation, function and cosmesis are important factors for patients with a choroidal melanoma diagnosis.Reference Fuss, Loredo, Blacharski, Grove and Slater¹² In addition to this, other aims of therapy are to maintain local control, patient survival and good visual acuity (VA).Reference Dawson, Sagoo, Mehta, Comer, Hungerford and Lee¹⁶ VA is the ability of the eye to perceive details. Evidence indicates that these can be achieved with EPB or proton therapy, which are both well-recognised conservative methods of treatment.Reference Munzenrider^6,Reference Damato, Kacperek, Chopra, Campbell and Errington^10–Reference Gündüz, Shields, Shields, Cater, Freire and Brady^13,15–Reference Dendale, Rouic, Noel, Feuvret, Levy, Delacroix, Meyer, Nauraye, Mazal, Mammar, Garcia, D'Hermies, Frau, Plancher, Asselain, Schlienger, Mazeron and Desjardins¹⁷

Both of these treatment options involve the use of radiation therapy, and consequently, radiation damage to surrounding healthy tissue can result in acute and late effects, such as poor VA, ocular pain and discomfort.Reference Finger^18,Reference Anderson, Keith and Novak¹⁹ Late effects can develop months or years after treatment; these can be problematic with effects having permanent damage and resulting in partial or complete loss of function.Reference Anderson, Keith and Novak¹⁹ It is the level of normal tissue irradiation that ultimately results in frequency and severity of late effects.

Many authors have used radiation-related complication as an outcome measure to determine normal tissue morbidity.Reference Fuss, Loredo, Blacharski, Grove and Slater^12,Reference Gündüz, Shields, Shields, Cater, Freire and Brady^13,Reference Dendale, Rouic, Noel, Feuvret, Levy, Delacroix, Meyer, Nauraye, Mazal, Mammar, Garcia, D'Hermies, Frau, Plancher, Asselain, Schlienger, Mazeron and Desjardins^17,Reference Conway, Poothullil, Daftari, Weinberg, Chung and O'Brien²⁰ Outcome measures of radiation-related complications are presented as ocular complications and tumour events, such as recurrence, metastases and melanoma-related mortality. These outcome measures have been used for both proton therapy and EPB. Normal tissue morbidity can result in conditions that have an impact on the quality of life of a patient.Reference Finger^18,Reference Jensen, Petersen, Kline, Stafford, Schomberg and Robertson²¹ Normal tissue morbidity is relative to the level of radiation-related ocular complications, ocular function and VA. These are used as outcome measures to determine the normal tissue morbidity as a result of radiation treatment.Reference Dendale, Rouic, Noel, Feuvret, Levy, Delacroix, Meyer, Nauraye, Mazal, Mammar, Garcia, D'Hermies, Frau, Plancher, Asselain, Schlienger, Mazeron and Desjardins^17,Reference Conway, Poothullil, Daftari, Weinberg, Chung and O'Brien²⁰ VA loss has an impact on quality of life as it affects many activities of daily living.Reference Cocchiarella²² VA can be used as a single outcome measure or in addition to radiation-related complication measures to demonstrate normal tissue morbidity. Much of the data and results demonstrate VA as measured using Snellen or Bailey–Lovie charts (Figure 2).^23,24

Figure 2. Visual acuity chart.^23,24

Patient condition

Findings suggest that normal tissue morbidity was similar for both types of treatment.Reference Munzenrider⁶ Patients with predisposing medical conditions such as diabetes, retinal detachment or glaucoma were more likely to experience a greater degree of normal tissue morbidity, irrespective of modality of choice.Reference Conway, Poothullil, Daftari, Weinberg, Chung and O'Brien^20,25 Other influencing prognostic factors included male gender and a median age of 55 years.Reference Dendale, Rouic, Noel, Feuvret, Levy, Delacroix, Meyer, Nauraye, Mazal, Mammar, Garcia, D'Hermies, Frau, Plancher, Asselain, Schlienger, Mazeron and Desjardins^17,25

Tumour location

Tumour location was also found to be a determining factor. A tumour located near to the optic disc or fovea is more problematic and likely to result in greater ocular morbidity.Reference Damato, Patel, Campbell, Mayles and Errington^11,Reference Fuss, Loredo, Blacharski, Grove and Slater¹² With both EPB and proton therapy, ideally a 2–2.5 mm margin is required to ensure adequate tumour coverage.Reference Damato, Kacperek, Chopra, Campbell and Errington^10,Reference Damato, Patel, Campbell, Mayles and Errington¹¹ Although there was little evidence to indicate that either proton therapy or EPB predominantly results in greater normal tissue morbidity, some findings did suggest that anterior segment complication was higher when using proton therapy.Reference Fuss, Loredo, Blacharski, Grove and Slater^12,Reference Conway, Poothullil, Daftari, Weinberg, Chung and O'Brien²⁰ This occurs due to the beam having to pass through critical structures within the anterior segment.

VA comparison

The Collaborative Ocular Melanoma Study (COMS)²⁶ was a large-scale multicentre randomised controlled trials consisting of three arms. The second arm of the COMS compared the effectiveness of brachytherapy to enucleation for treatment of medium-sized choroidal melanomas. Patients with medium choroidal melanomas were randomised to iodine-125 (¹²⁵I) EPB or enucleation. COMS report no. 16 reported VA during 3 years after treatment in patients in the ¹²⁵I brachytherapy arm. The sample consisted of 657 randomly assigned patients, of which the VA data of 623 was included in the report. VA was determined using Snellen and Bailey–Lovie charts and standard lighting. A baseline VA was carried out prior to treatment, which was later to be used for comparison. A loss of six or more Snellen VA lines from baseline was defined as VA loss. All 623 patients were enrolled for at least 12 months following treatment; VA follow-up results of 93 and 80% at 2- and 3-year follow-up intervals, respectively, are tabulated below (Table 3).

Table 3. COMS visual acuity follow-up

COMS, Collaborative Ocular Melanoma Study.

The results showed that treatment using ¹²⁵I brachytherapy preserved a relatively high level of VA in over 50% of patients within the first year after treatment. This declined at an average rate of approximately two lines per year, with 43–49% of treated eyes having significant VA impairment 3 years after treatment.

In a single centre, retrospective analysis of a series of 136 patients treated with ¹²⁵I brachytherapy between 1990 and 2000, Lumbroso-Le Rouic et al. found that the mean VA at 5 years was 20/40; this is considerably higher level of VA than the findings in the COMS.Reference Lumbroso-Le Rouic, Charif Chefchaouni, Levy, Plancher, Dendale, Asselain, Solignac, Mazal and Desjardins²⁷ Sagoo et al.Reference Sagoo, Shields, Mashayekhi, Freire, Emrich, Reiff, Komarnicky and Shields²⁸ conducted a retrospective medical record review over 31 years. Although this study had a long review period, it had a limited number of participants. Thirty-seven consecutive patients were involved, who were monitored for long-term VA; the findings were similar to that in the COMS with VA of 20/200 or worse being observed in 62% of patients with 57% losing more than five Snellen VA lines.Reference Sagoo, Shields, Mashayekhi, Freire, Emrich, Reiff, Komarnicky and Shields²⁸

Damato et al.Reference Damato, Kacperek, Chopra, Campbell and Errington¹⁰ used VA as an outcome measure in 349 patients treated with proton beam therapy. This was a single centre study, which was carried out over a period of 10 years. Patients were prospectively and consecutively included. To measure VA, the Snellen chart was used, as well as counting fingers method. Baseline VA in the tumour-affected eye was logged prior to the start of treatment. Two hundred and twelve patients had 20/40 or better vision at this time. VA conservation of this level at follow-up is summarised as follows (Table 4):301 patients had 20/200 or better initial VA. This level of VA was conserved in 81.9% at 2 years, 61.1% at 5 years and 41.7% at 8 years. The study demonstrated that the greater the initial tumour height, the sooner loss of VA occurred.Reference Damato, Kacperek, Chopra, Campbell and Errington¹⁰

Table 4. Visual acuity conservation

Courdi et al. looked at VA following proton beam therapy for uveal melanoma.Reference Courdi, Caujolle, Grange, Diallo-Rosier, Sahel, Bacin, Zur, Gastaud, Iborra-Brassart, Hérault and Chauvel²⁹ This was a single centre prospective study reporting on 538 patients treated over a period of 5 years; the VA of 284 patients was adequately scored before and during follow-up. Patients were assessed according to a scale other than the Snellen or Bailey–Lovie chart to allow easy estimation of the change in VA following treatment. The scale ranged from 0 (being no light perception) to 14 (being 200/200 vision). The results showed that VA was initially improved in many patients immediately after treatment; however, it became gradually more impaired as length of follow-up time increased.

Local control

Local control is defined as tumour shrinkage and growth cessation.Reference Shields, Naseripour, Cater, Shields, Demirci, Youseff and Freire^30,Reference Gragoudas, Lane, Munzenrider, Egan and Li³¹ As an outcome measure, local control failure was considered to be tumour regrowth, recurrence, metastases or the need for enucleation.Reference Damato, Kacperek, Chopra, Campbell and Errington^10–Reference Gündüz, Shields, Shields, Cater, Freire and Brady^13,15–Reference Jensen, Petersen, Kline, Stafford, Schomberg and Robertson^21,Reference Sagoo, Shields, Mashayekhi, Freire, Emrich, Reiff, Komarnicky and Shields^28,Reference Gragoudas, Lane, Munzenrider, Egan and Li^31–Reference Sia, Harper, McAllister and Perry³⁶ The review suggested that overall proton therapy was slightly more favourable than EPB (see Table 5) in terms of local control rates at 5- and 10-year follow-up.Reference Fuss, Loredo, Blacharski, Grove and Slater^12,Reference Gragoudas, Lane, Munzenrider, Egan and Li^31,Reference Höcht, Bechrakis, Nausner, Kreusel, Kluge, Heese, Heufelder, Cordini, Homeyer, Fuchs, Martus, Foerster, Wiegel and Hinkelbein³⁷

Table 5. Enucleation free survival and local control rates

Med f/u, medical follow-up.

Choice of radioisotope

The local control outcome for proton beam therapy is consistent within the literature; however, EPB outcomes vary depending on radioisotope used. Radioisotopes such as gold-198 (¹⁹⁸Au) plaque and palladium-103 (¹⁰³Pd) plaque are less widely used within clinical practice, with ¹²⁵I being the most commonly used. Based on the findings in this review, local control rates for patients treated with ¹²⁵I are lower than those with 103^Pd or 198^Au. Further research should be carried out to provide a more sound evidence base for the use of 103^Pd and 198^Au, with a view to these becoming more routinely used over ¹²⁵I.

Modality choice

Proton beam radiotherapy and EPB are both effective treatment options in the management of choroidal melanoma.Reference Munzenrider⁶ The efficacy of each option, measured by comparing normal tissue morbidity and local tumour control, is largely dependent on factors such as tumour size and location, and in EPB, the radioisotope used.Reference Wilson and Hungerford⁵ The relatively low incidence of choroidal melanoma and the variety of treatment options make it difficult to gather data on long-term results for a large number of equally treated patients.Reference Egger, Schalenbourg, Zografos, Bercher, Boehringer, Chamot and Goitein³⁸ This limitation makes it difficult to establish conclusively which treatment method is most effective.

The literature does, however, suggest that EPB is most effective in the treatment of medium and large posterior choroidal melanoma,^25,Reference Shields, Naseripour, Cater, Shields, Demirci, Youseff and Freire³⁰ while proton therapy is most effective for small posterior tumours close to the optic nerve and fovea.Reference Damato, Kacperek, Chopra, Campbell and Errington¹⁰ Proton is also effective for tumours that are too large for EPB when the patient is unsuitable for surgery. However, given the variability in methods and patient selection criteria in the literature reviewed, a comparison of the results does not provide reliable support for this recommendation. In order to do this, a prospective, multicentre, randomised controlled trial would be necessary. Because treatment outcome is influenced by a variety of prognostic factors, many controls with highly specific inclusion and exclusion criteria would need to be considered in order to ensure reliability.

Clinical availability

The clinical availability of proton therapy is limited due to the small number of clinical sites with proton facilities; there are currently only 26 facilities in operation worldwide.³⁹ Inevitably, this has an impact on its viability as a treatment option. This limitation is largely associated with the cost of proton centres, including installation fees, equipment and software expenses and personnel to maintain the equipment.Reference Thornton, Fitzek, Klein, Dickey, Sullivan, O’Ryan-Blair, Schreuder, Allgower, Mascia, Farr and Cameron¹⁴ Had the findings within this study demonstrated more favourable outcomes in patients treated with proton therapy, a comprehensive cost analysis would have been suggested to consider costs arising as a result of treatment failure using EPB.

Conversely, ophthalmic EPB is a widely available, well-established, conservative treatment option for choroidal melanoma.Reference Finger^18,Reference Finger, Berson, Ng and Szechter³³ The increased commercial availability of this option compared to proton therapy unsurprisingly makes plaque brachytherapy first-choice option in ocular-conserving treatment for patients with choroidal melanoma.Reference Finger, Berson, Ng and Szechter³³ However, this alone should not govern the choice of modality.