31 Uveal Melanoma

SECOND EDITION

The GEC ESTRO Handbook of Brachytherapy

PART II: CLINICAL PRACTICE 31 Uveal Melanoma Luca Tagliaferri, Maria Antonietta Blasi, Carien Creutzberg, Gerd Heilemann, Erik Van Limbergen, Richard Pötter

Editors Erik Van Limbergen Richard Pötter

Peter Hoskin Dimos Baltas

Uveal Melanoma

3

THE GEC ESTROHANDBOOKOF BRACHYTHERAPY | Part II Clinical Practice Version 1 - 15/04/2020

31 Uveal Melanoma

Luca Tagliaferri, Maria Antonietta Blasi, Carien Creutzberg, Gerd Heilemann, Erik Van Limbergen, Richard Pötter

1. Summary 2. Introduction

3 3 4 4 5 5 8 8

9. Treatment planning

8

10. Dose, dose rate and fractionation

11 12 12 15 18 18

3. Anatomical topography

11. Monitoring

4. Pathology 5. Work up

12. Results

13. Adverse side effects 14. Key messages

6. Indications, contra-indications 7. Tumour and target volumes

15. References

8. Technique

1. SUMMARY

Uveal melanoma (UM) is the most common primary intraocular malignancy in adults and accounts for 5% of all melanomas and includes melanoma arising in the iris, ciliary body and choroid. The diagnosis is based on the ocular oncologist’s comprehensive clinical judgement taking into account results from the different diagnostic methods applied. Enucleation has been the treatment of choice for decades, until some European experience especially using Ruthenium-106 plaques and later the Collaborative Ocular Melanoma Study (COMS) demonstrated that mortality rates following brachytherapy for small andmedium-sized uveal melanoma did not differ from those following enucleation for up to 12 years after treatment. Multiple eye-conserving treatment options are now available for the majority of UM including transpupillary thermotherapy, trans-scleral resection, and radiotherapy using episcleral brachytherapy, charged particle radiotherapy (protons) or stereotactic photon radiotherapy. Enucleation remains an accepted treatment for large choroidal melanoma. Brachytherapy is the most commonly used conservative treatment for UM. The Clinical Target Volume (CTV) is determined based on pretreatment findings from ophthalmoscopy and ultrasound scan. Different radioisotopes can be used for ocular brachytherapy. Ruthenium-106 ( 106 Ru) and Iodine-125 ( 125 I) are the most widely used. Plaques loaded with Iodine-125 seeds can be used both for the treatment of small lesions and for the treatment of larger lesions. 106Ru plaques, based on a beta-emitter, are to be used for small and medium-sized lesions, generally less than 5-6mm in thickness, with the advantage of reduced dose to the lens and other parts of the retina. Dosimetry can be performed manually or by software that allows 3D processing of the dose. The use of 3D software for preplanning may improve treatment planning. Treatment results and morbidity are specific for the different isotopes Ruthenium-106 and Iodine-125. There is a considerable amount of data available, reflecting the broad experience with the 106 Ru eye plaque brachytherapy, largely used in Europe. 106 Ru brachytherapy represents a good treatment option for small andmedium-sizedmelanomas with local control rates between 78-95%. Based on AJCC staging for posterior uveal melanoma, 10-year metastatic rate is 12% for stage I, 29% for stage II, and 61% for stage III tumours. The risk of metastasis and death increases three-fold with each increasing melanoma stage. However, it has become increasingly apparent that genetic features of the primary uveal melanoma, such as chromosome 3 status and gene expression profile, are more sensitive than size as a predictor of metastasis and death. Enucleation is required in 4-6% of all patients, usually during the first years, either for tumour recurrence or toxicity. Although enucleation has been the gold standard for the treatment of recurrences, retreatment with episcleral plaque represents an excellent option, especially for marginal recurrences.

2. INTRODUCTION

higher incidence at 4.9 cases per million than do females at 3.7 cases per million. UM is diagnosed at older ages, with a progressively rising, age-specific, incidence rate that peaks near the age of 70 years. Uveal melanoma arises from melanocytes situated in the uveal tract of the eye, and can affect any part of the uveal tract, but choroidal melanoma is predominant (85-90%), while iris and ciliary body melanomas are far less frequent accounting for 9–15% of cases. Iris melanomas are associated with the earliest detection and overall best prognosis, while ciliary body melanomas are

Uveal melanoma (UM) is the most common primary intraocular malignancy in adults and accounts for 5% of all melanomas. It is most commonly found in light complexion Caucasians with an age-adjusted incidence of 4.3 per million people. In Europe, the incidence of uveal melanoma follows a north-to-south decreasing gradient ranging from 2 to 8 per million population. Males have a

Uveal Melanoma

4

THE GEC ESTROHANDBOOKOF BRACHYTHERAPY | Part II Clinical Practice Version 1 - 15/04/2020

associated with the worst prognosis. Around 40-50% of all patients diagnosed with UMwill developmetastasis, despite treatment, and in 90% of cases the liver is the target organ for metastasis, with survival time after metastasis averaging 6–12 months. Most UM are initially asymptomatic. As the tumour enlarges, it may cause distortion of the pupil (iris melanoma), blurred vision (ciliary body melanoma), or markedly decreased visual acuity caused by secondary retinal detachment (choroidal melanoma). Serous detachment of the retina may occur. If extensive detachment occurs, secondary closed-angle glaucoma occasionally develops. Treatment of primary uveal melanoma is either globe-conserving therapy or enucleation. Among radiation therapies, plaque brachytherapy has the lowest rate of treatment failures. The two most widely used forms of radiation therapy, Iodine-125 and Ruthenium-106 brachytherapy are both associated with a weighted average local recurrence rate of 4-15%; however the recent introduction of the intraoperative ultrasound plaque localization during brachytherapy has been shown to reduce the risk of local treatment failure. Stereotactic radiotherapy has similar rates local failure of 2-7.9%, charged particle radiation therapy, however, has a lower weighted average rate of local failure at 4.2% and this treatment modality should be discussed in tumour boards as alternative treatment for large, completely encircling juxtapupillary melanomas. Overall, surgical modalities have a higher rate of local failure compared with radiation modalities (18.6% vs 6.15%). Transpupillary Thermotherapy has the largest reported variation of failure from 0% to 55.6%, with a weighted average of 20.8%. The globe is composed of three tunicae. The outer coat consists of the opaque fibrotic sclera (containing mainly collagen with only few fibroblasts) with a thickness of 0.3 to 1.0 mm and a diameter of 24 mm (which tolerates extremely high radiation doses), to which the six ocular muscles controlling the eye movement attach, and the clear cornea. The uvea (middle coat) is formed by the choroid (thickness 0.1 - 0.3 mm), the ciliary body (thickness 2 mm), and the iris (thickness 0.5 - 3 mm), which represents the origin of different types of uveal melanoma. The inner coat consists of the retina including the retinal pigmented epithelium (thickness 0.1 mm). The vascular supply for the retina is derived from the central retinal artery entering the globe through the optic nerve. The vitreous body filling the inner part of the globe does not contain any vital cells. The optic nerve enters the globe at the optic disc (papilla of the optic nerve). The three tunicae of the globe stop at the border of the optic disc. The macula represents the point of the sharpest sight and is therefore most important for the overall function of the eye and is located temporal to the optic disc (3 mm). (Fig. 1) 3. ANATOMICAL TOPOGRAPHY

Fig. 1: ocular anatomy and uveal melanoma

tumour was located in the iris in 285 (4%), ciliary body in 492 (6%), and choroid in 7256 (90%) cases. Choroidal melanoma usually presents as a brown, elevated, well defined, pigmented subretinal mass. The major growth pattern (75%) is towards the inner part of the globe resulting in a dome- shaped configuration . At the same time, the tumour frequently spreads along the choroid or along the inner part of the sclera (up to 50%). The overlying retina is most often damaged for various reasons (e.g. through impaired blood supply). In more than 25% nodular tumours erode through Bruch`s membrane, which covers the choroid. Because of its elasticity this membrane cuts into the tumour, leading to a typical mushroom configuration , in which the surface of the mushroom may sometimes be larger than its base. In this configuration particularly, different forms of retinal tumour invasion are seen as well as cellular spread into the vitreous body. Choroidal melanoma can lead to a serous retinal detachment in its direct vicinity and far away at the caudal part of the fundus following gravity. The optic nerve is almost never infiltrated, even in juxtapapillary tumours. Extraocular extension is present at the time of diagnosis in 3–5.8% of the patients with uveal melanoma. A Diffuse configuration with infiltrating flat choroidal tumours is rare. The degree of pigmentation ranges from dark brown to totally amelanotic. Clinically, several lesions can simulate choroidal melanoma, includingmetastatic carcinoma, posterior scleritis, and benign tumours, such as nevi and hemangiomas. Iris melanoma can be circumscribed (90%) or diffuse (10%). In most cases, it is an incidental finding due to iris color changes (heterochromia) and pupil distortion (corectopia). It is most commonly located in the inferior quadrant, and because of its clinical appearance it is detected at an early stage. Diffuse iris melanoma has an infiltrative, flat, ill- defined growth pattern with confluent or multifocal iris involvement. The diagnosis of diffuse iris melanoma is challenging and is often delayed. The classic findings of diffuse iris melanoma include acquired hyperchromic heterochromia iridis and ipsilateral glaucoma. Ring melanoma of the anterior chamber is a distinct rare variant that manifests as circumferential, flat tumour growth confined to the trabecular meshwork and anterior chamber angle structures. It presents as refractory unilateral glaucoma simulating pigmentary glaucoma and can be diagnosed only by gonioscopy and ultrasonography.

4. PATHOLOGY

The most common site for uveal melanoma is the choroid. In a study of 8033 patients with uveal melanoma by Shields et al , the

Uveal Melanoma

5

THE GEC ESTROHANDBOOKOF BRACHYTHERAPY | Part II Clinical Practice Version 1 - 15/04/2020

Beside its nodular growth, melanoma of the ciliary body can also extend into the posterior part of the globe along the uvea and the sclera and into the anterior part towards the iris. Ciliary body melanoma is often diagnosed late as the lesion remains hidden behind the iris, and the patient seldom has any clinical symptoms until the lesion is large. The iridocorneal angle can be infiltrated rather early, whichmay lead to a secondary glaucoma. On the other hand, the lens may become displaced because of tumour growth.

Various clinical, histopathological, cytogenetic and gene expression features help in estimating the prognosis of uveal melanoma. The clinical features associated with poor prognosis in patients with uveal melanoma include older age at presentation, male gender, larger tumour basal diameter and thickness, ciliary body location, diffuse tumour configuration, association with ocular/oculodermal melanocytosis, extraocular tumour extension, and advanced tumour staging by the American Joint Committee on Cancer (AJCC) classification. The histopathological features suggestive of poor prognosis include epithelioid cell type, high mitotic activity, higher microvascular density, extravascular matrix patterns, tumour-infiltrating macrophages. Monosomy 3, 1p loss, 6q loss, and 8q gain and those classified as Class II by gene expression are predictive of poor prognosis of uveal melanoma. (Table 1, 2, 3, 4, 5)

5. WORK UP

There is general consensus that clinical diagnosis of uveal melanoma is sufficient and adequate for treatment. The diagnosis of uveal melanoma is based on the ocular oncologist’s comprehensive clinical judgement taking into account results from the different diagnostic methods applied. Histopathologic verification is not required. For experienced ophthalmologists the rate of false positive findings is nowadays reported to be below 1 - 2%. In doubtful cases, biopsy could be considered, although with an increasing risk of complications. Iris melanoma typically is visible through the cornea, and the diagnosis is carried out by slit-lamp biomicroscopy. Gonioscopy is useful to assess the involvement of the anterior chamber angle. For small tumours, anterior segment optical coherence tomography (AS-OCT) is useful with high-resolution imaging of anterior and lateral surfaces. Ultrasound biomicroscopy (UBM) assists in visualization of the posterior tumour extent. Choroidal and ciliary bodymelanomas commonly are diagnosed, measured, and assessed based on clinical examination, including slit-lamp biomicroscopy, ophthalmoscopy for detailed fundus evaluation, and ultrasonography (US). US is an important diagnostic tool used to define tumour extent and shape, and to measure tumour dimensions. Typical features of a posterior uveal melanoma on US include acoustic hollowing, choroidal excavation, and orbital shadowing. Additional methods, such as UBM imaging, OCT, autofluorescence, standard and wide-field photography, fluorescein and indocyanine green angiography, CT, PET/CT, and MR imaging may enhance the accuracy of assessment, especially in atypical cases. The large prospective randomized COMS trial demonstrated 99.6% accuracy in the diagnosis of medium-sized and large choroidal melanoma using standardized echography, fundus photography and fluorescein angiography. Smaller melanomas typically exhibit fewer diagnostic characteristics, making them more difficult to diagnose. A small uveal melanoma cannot be distinguished from a naevus. Clinical findings that may help to identify melanoma include the following: orange pigment on the tumour surface, subretinal fluid, tumour thickness of more than 2 mm, low internal reflectivity on ultrasound examination, and documented growth. Staging for uveal melanoma follows the AJCC Tumour-Node-Metastasis (TNM) staging system for eye cancer. In this classification iris melanoma is graded according to tumour extent, associated secondary glaucoma, and extraocular extension. Posterior uveal (ciliary body and choroid) melanoma is graded according to tumour basal diameter and thickness, ciliary body involvement, and extraocular extension. The T categories are defined by the tumour dimensions. (Table 1, 2, 3, 4, 5)

6. INDICATIONS, CONTRA-INDICATIONS

Enucleation has been the treatment of choice for decades, until European experience especially using Rutenium106 plaque, and later the Collaborative Ocular Melanoma Study (COMS) demonstrated that mortality rates following brachytherapy did not differ from mortality rates following enucleation for up to 12 years after treatment of patients with uveal melanoma. Although enucleation remains the accepted treatment for a large choroidal melanoma, multiple treatment options are now available for small and limited size UMwhichmay include observation, transpupillary thermotherapy (TTT), trans-scleral resection, and in particular radiotherapy using episcleral brachytherapy, charged particle radiotherapy or stereotactic radiotherapy. Brachytherapy is the most commonly used conservative treatment for the UM. Recent advances have expanded the use of plaque brachytherapy as well as improved the surgical technique, allowing more tumours to be treated effectively. Small melanomas can be treated based on the eye cancer specialist’s opinion, considering personal experience, clinical examination and evidence of growth although a discussion in the multidisciplinary Tumor Board is recommended, as early treatment provides best results. AJCC T1, T2, T3, and T4a-d uveal melanoma patients can be treated, after counseling about likely vision loss, eye retention, and local control outcomes. Patients with posterior melanomas, even those peripapillary and subfoveal, and those with exudative retinal detachments can be successfully treated, but typically have poorer resultant vision. They should be accordingly counseled. General exclusion criteria for brachytherapy include only tumours with T4e extraocular extension, basal diameters that exceed the limits of brachytherapy, blind painful eyes, and those with no light perception. Patients with these characteristics should always be evaluated in themultidisciplinary Tumour Board and brachytherapy could be indicated in selected cases. There is some controversy about treatment of certain uveal melanomas. For example, brachytherapy for tumours near, touching, or surrounding the optic disc. Optic nerve anatomy presents a unique obstacle for radioactive plaque placement. The optic disc face has a mean diameter of 1.8 mm and is surrounded by choroid. Thus, choroidal melanomas can touch or encircle the disc. Also, as the optic nerve exits the eye, it is enveloped by the

Uveal Melanoma

6

THE GEC ESTROHANDBOOKOF BRACHYTHERAPY | Part II Clinical Practice Version 1 - 15/04/2020

optic nerve sheath, thus expanding its total width to a 5- to 6-mm optic nerve sheath diameter. This difference creates an offset to proper plaque placement. Slotted plaques were devised with 8-mm openings. A slot allows the optic nerve sheath to enter the plaque carrier, thus more posteriorly locate the Iodine-125 seed sources and move the target volume into a normalized position surrounding the choroidal melanoma. It is important to note that plaque slots make dosimetry more complex. In these particular cases using Iodine-125 treatment, medical physicists must locate seed sources to both ‘‘fill-in’’ the gap created by the slot and cover the target volume. Ruthenium applicators with an opening for the optic nerve (COB) are also available. For large, completely encircling juxtapapillary melanomas proton beam radiotherapy is an alternative treatment. In addition, patients unsuitable for

brachytherapy or resection can be treated with proton beam radiotherapy or stereotactic radiosurgery. The ABS-OOTF recommends that plaque procedures should be performed in specialized medical centres with expertise in ophthalmic brachytherapy. Such centers should include a team composed of at least a subspecialty-trained plaque surgeon, a radiation oncologist and a medical physicist experienced in plaque brachytherapy. Communication between them is critical for any successful brachytherapy program.The ocular oncologist provides data related to tumour localization (choroid, ciliary body, iris) and basal diameters, thickness, distance from the sensitive areas of the fovea and optic disc. The plaque choice is made to encompass the tumour with margin on all sides; plaque apertures should exceed

TABLE 1 AJCC CLASSIFICATION OF POSTERIOR UVEAL MELANOMA (CHOROIDAL AND CILIARY BODY), T CATEGORY Thickness (mm) >15.0 4 4 4 4 4

4 4 3 3 3 2

4 4 4 4 4 4

12.1 to 15.0 9.1 to 12.0

3 3 2 1 1

3 3 2 1 1

3 3 2 1 1

3 3 2 2 1

3 3 3 2 2

6.1 to 9.0 3.1 to 6.0

≤ 3.0

≤ 3.0

3.1-6.0

6.1-9.0

9.1-12.0

12.1-15.0

15.1-18.0

>18.0

Largest basal diameter (mm)

TABLE 2 AJCC CLASSIFICATION OF POSTERIOR UVEAL MELANOMA (CHOROIDAL AND CILIARY BODY), T CATEGORY SUBCLASSIFICATION T Category T Criteria TX Primary tumor cannot be assessed T0 No evidence of primary tumor

TABLE 3 AJCC CLASSIFICATION OF POSTERIOR UVEAL MELANOMA (CHOROIDAL AND CILIARY BODY), REGIONAL LYMPH NODES AND DISTANT METASTASIS N Category N Criteria NX Regional lymph nodes cannot be assessed N0 No regional lymph nodes involvement N1 Regional lymph nodes metastasis or discrete tumor deposits in the orbit N1a Metastasis in one or more regional lymph node(s) N1b No regional lymph nodes are positive, but there are discrete tumor deposits in the orbit that are no contigious to the eye M Category M Criteria M0 No distant metastasis by clinical classification M1 Distant metastasis M1a Largest diameter of the largest metastasis ≤3,0cm M1b Largest diameter of the largest metastasis 3,1-8.0cm M1c Largest diameter of the largest metastasis ≥8,1cm

Tumor base ≤9mm with thickness ≤6mm Tumor base 9.1-12mm with thickness ≤3mm

T1

T1a

Tumor size category 1 without ciliary body involvement and extraocular extension

T1b

Tumor size category 1 with ciliary body involvement

T1c

Tumor size category 1 without ciliary body involvement but with extraocular extension ≤5mm in largest diameter

T1d

Tumor size category 1 with ciliary body involvement and extraocular extension ≤5mm in largest diameter

Tumor base ≤9mm with thickness 6.1-9mm Tumor base 9.1-12mm with thickness 3.1-9mm Tumor base 12.1-15mm with thickness ≤6mm Tumor base 15.1-18mm with thickness ≤3mm

T2

T2a

Tumor size category 2 without ciliary body involvement and extraocular extension

T2b

Tumor size category 2 with ciliary body involvement

T2c

Tumor size category 2 without ciliary body involvement but with extraocular extension ≤5mm in largest diameter

T2d

Tumor size category 2 with ciliary body involvement and extraocular extension ≤5mm in largest diameter

Tumor base 3.1-9mm with thickness 9.1-12mm Tumor base 9.1-12mm with thickness 9.1-15mm Tumor base 12.1-15mm with thickness 6.1-15mm Tumor base 15.1-18mm with thickness 3.1-12mm

T3

T3a

Tumor size category 3 without ciliary body involvement and extraocular extension

T3b

Tumor size category 3 with ciliary body involvement

T3c

Tumor size category 3 without ciliary body involvement but with extraocular extension ≤5mm in largest diameter

T3d

Tumor size category 3 with ciliary body involvement and extraocular extension ≤5mm in largest diameter

Tumor base 12.1-15mm with thickness >15mm Tumor base 15.1-18mm with thickness >12mm Tumor base >18mm with any thickness

T4

T4a

Tumor size category 4 without ciliary body involvement and extraocular extension

T4b

Tumor size category 4 with ciliary body involvement

T4c

Tumor size category 4 without ciliary body involvement but with extraocular extension ≤5mm in largest diameter

T4d

Tumor size category 4 with ciliary body involvement and extraocular extension ≤5mm in largest diameter

T4e

Any tumor size category with extraocular extension >5mm in largest diameter

Uveal Melanoma

7

THE GEC ESTROHANDBOOKOF BRACHYTHERAPY | Part II Clinical Practice Version 1 - 15/04/2020

TABLE 4 AJCC CLASSIFICATION OF POSTERIOR UVEAL MELANOMA (CHOROIDAL AND CILIARY BODY), STAGING When T is... And N is... And M is... Then the stage group is T1a NO MO I T1b-d NO MO IIA T2a NO MO IIA T2b NO MO IIB T3a NO MO IIB T2c-d NO MO IIIA T3b-c NO MO IIIA T4a NO MO IIIA T3d NO MO IIIB T4b-c NO MO IIIB T4d-e NO MO IIIC Any T N1 MO IV Any T N1 M1a-c IV

TABLE 5 AJCC CLASSIFICATION OF POSTERIOR UVEAL MELANOMA (CHOROIDAL AND CILIARY BODY), HISTOLOGIC GRADE G Category G Criteria

GX

Grade cannot be assessed

G1

Spindle cell melanoma (>90% spindle cells)

G2

Mixed cell melanoma (>10% epithelioid cells and <90% spindle cells

G3

Epithelioid cell melanoma (>90% epithelioid cells)

TABLE 1, 2, 3, 4, 5: Adapted from Amin MB, Edge SB, Greene F, et al., eds.: AJCC Cancer Staging Manual. 8th ed. New York, NY: Springer, 2017, pp 805–17 (ref 80)

Fig. 4: US A scan / the tumour thickness is the distance between Peak A (right arrow) and Peak B (left arrow)

Fig. 2: Fondus oculi with tumour

Fig. 5: GTV and CTV definition based on fundus oculi and eye section

Fig. 3: US B scan / Red line: diameter – Blue line: thickness

Fig. 6: GTV definition based on ultrasound exam (B scan on the left and A scan on the right)

Uveal Melanoma

8

THE GEC ESTROHANDBOOKOF BRACHYTHERAPY | Part II Clinical Practice Version 1 - 15/04/2020

the largest tumour diameter to create a tumour-free margin of safety to prevent geographic miss. Dose rate calculation and treatment plan approval represent the last phases of the workflow. All patients should be discussed in the Multidisciplinary Tumour Board before treatment.

sonographic measurement of the tumour thickness itself. In view of the specific growth pattern of uveal melanoma, there is debate over the target cells in uveal melanoma.The whole tumour volume and its microscopic extensions form the basis of the CTV, therefore the apex tumour dose represents the minimum target dose to tumour cells at the apex. In contrast, the uveal layer carrying the main blood supply is where tumour growth takes its origin and is continuously fed via the uveal blood supply. This layer receives a significantly higher dose, which is close to the scleral surface dose.

7. TUMOUR AND TARGET VOLUMES

The gross tumour volume (GTV) is determined based on pretreatment findings fromophthalmoscopy and ultrasound scan. The ophthalmologist provides a diagram of the fundus oculi with tumour border orientation and reports the distance from optic nerve and fovea. Basal diameters and apex height (thickness) of the disease are also carefully measured for gross tumour volume (GTV) definition. The use of both A and B ultrasound scans is strongly recommended. Tumour diameters are defined with B-scans while disease height assessment is evaluated with A-scans, and their information cannot be replaced by other US techniques, even if with higher nominal resolution. Sometimes MRI could be used if ultrasound cannot measure the tumour in a precise way. The pretreatment GTV measurements are mandatory for 3D pretreatment plan and to define the type and size of plaque. A confirmation of tumour diameters is done using the tumour’s transillumination shadow during the intervention in the operation theatre. (Fig. 2, 3, 4) Base diameter of the CTV The base diameter of the CTV is determined by the tumour base diameters . A safetymargin of 1 - 2mmfor subclinical disease is added in all directions, which accounts for microscopic spread along the ocular tunicae, mainly the uvea and the sclera. In some situations, uncertainties in tumour delineation (e.g. in the exact drawing of the tumour shadow during transillumination) or in plaque localization (e.g. in posterior pole locations) may be observed, which can be taken into account by adding an extra safety margin (PTV) to deal with these uncertainties.The amount of such a safetymargin should be determined individually according to the given situation. No safety margin for eyeball movements is taken into consideration, as the eye plaque is sutured tightly onto the outer sclera. Safety margin considerations have to be taken into account particularly if the tumour is located in the vicinity of critical structures, such as the optic disc, the macula, or the ciliary body. Height of the CTV (prescription point) The height of the CTV (prescription point – distance between apex of the tumour and the internal surface of the plaque) is determined by the tumour thickness plus 1mm (thickness of the sclera) or the measured distance from external scleral surface to tumour apex (A and B-scan). The thickness is usually taken from sonography findings (A-scan) with 1 mm added for the sclera. The thickness is measured at themost prominent point of the dome shaped tumour (tumour apex) with a line drawn towards the uvea which has to be perpendicular to its basis. For ciliary body tumours, the distance from the tumour apex and the surface of the plaque needs to be considered. (Fig. 5, 6) Therefore, in dome-shaped tumours these dimensions usually represent the maximum tumour thickness. Uncertainties in the determination of tumour target thickness are correlated to the

8. TECHNIQUE

The surgery is usually performed under general or local anaesthesia or intravenous sedation by a subspecialty-trained surgeon. The conjunctiva is opened at the limbus and the rectus muscles are isolated for traction with silk sutures to expose the quadrant (Fig.7). Localization of the tumour borders is achieved by transillumination placing a fiber optic light 180° away from the tumour. The tumour base shadows its subjacent sclera (Fig. 8). A sterile marking pen is used to mark the outline of the tumour border plus 2-3 mm free margin around the tumour base (Fig.9). If an extraocular muscle is overlying the tumour, the muscle should be detached for proper positioning of the plaque. The selection of the size of the plaque is based on the dimensions of the clinical target volume, which usually exceeds the GTV at each side by 1 - 2 mm. Furthermore, it will also take into account that there is an inactive edge at the outer margin of the plaque of 0.6 - 0.7 mm. The preplanning for iodine-125 eye plaques (See further section treatment planning) must be precise so that themanufactured plaque fulfils the demands for dimensions as outlined during the brachytherapy procedure. A template (dummy) of the same size and shape as the treatment plaque is used to position sutures on the sclera and to verify the position before placement of the radioactive plaque (Fig.10). The template (dummy) is then removed and replacedwith the radioactive plaque secured to the sclera with the preplaced sutures (Fig. 11). Any disconnectedmuscle is temporarily reattached to its insertion by using two preplaced sutures. A temporary knot is placed to hold themuscle to the rectus insertion.The conjunctiva is closed, and the eye is patched and, for iodine plaques, covered with a lead shield left in place throughout the treatment period. Plaque location can be confirmed intraoperatively by ultrasonography (Fig. 12). After a few days, the patient returns to the operating room for plaque removal. Under local anesthesia, the plaque is visualized, the anchoring sutures are cut, and the plaque is removed. The displaced muscle should be reattached into its insertion. (Fig. 7, 8, 9, 10, 11, 12, 13)

9. TREATMENT PLANNING

Different types of radioisotopes are available for ocular brachytherapy. Ruthenium-106 and Iodine-125 are the most widely used radioisotopes. Plaques loaded with Iodine-125 seeds can be used both for the treatment of small lesions and for the treatment of

Uveal Melanoma

9

THE GEC ESTROHANDBOOKOF BRACHYTHERAPY | Part II Clinical Practice Version 1 - 15/04/2020

larger lesions. Ruthenium-106 plaques, being a beta-emitter, are to be preferred for small lesions, generally less than 5-6mm, in view of their more favorable toxicity profile delivering a lower dose to lens and opposite retina); they can be used for up to 7-8 mm tumour thickness. Similarly, Strontium-90 applicators feature even higher dose gradients andmay be used for small lesions < 3.5mm. (Fig. 14) Historically, treatment planning has been performed manually based on depth-dose curves provided by the manufacturers with the dose prescribed to the apex. However, the use of 3D treatment planning software is recommended, which incorporates medical imaging, an anatomical volumetric model of the eye and full 3D dose calculation. The software enables definition of target volumes as well as organs- at-risk. This in turn provides dose-volume metrics, such as target coverage and dose to critical structures (e.g. fovea, optical disc and lens), which can be used to drive the treatment planning. It has been shown, that performing 3D treatment planning can help in reducing dose to the macula and optic disk. Currently, only one dedicated TPS for plaque based brachytherapy of uveal melanoma is available commercially (Plaque Simulator, Eye Physics LLC, Los Alamitos, USA). This allows single images from different modalities (i.e. CT, MRI, fundus photographs and US imaging) to be fused and overlays them on a 3D eye model. MRI based planning is under development. Additionally, Plaque Simulator can help in planning the plaque placement and surgery by givingmore accurate information on tumour and consequently suture location before the actual surgical procedure. However, because only single slices from CT or MRI are used to represent the patient-specific anatomy, there can be geometric limitations to its use. The need for a modern image-guided approach is widely discussed in the literature. Some have proposed using a conventional TPS (Pinnacle, Phillips Health Care, Hamburg, Germany) and CT imaging to provide more accurate representations of the patient eye geometries. Others pointed out the advantages of the superior contrast of MR images in defining the tumour and eye geometry and demonstrated its feasibility using another commercial TPS (Eclipse BrachyVision, VarianMedical Systems, Palo Alto, CA). MRI can help in assessing more accurate tumour dimensions which could lead to significant changes in treatment prescriptions. Additionally, another group has developed a stand-alone 3D TPS with a focus on automated treatment plan optimization routines. (Fig. 15, 16) As eye plaque brachytherapy with Ruthenium-106 and Iodine-125 are the most widely used applicators, a description of these two isotopes is given here in more detail. Ruthenium-106 eye plaque brachytherapy The isotope Ruthenium-106 has a half life of 367 days and decays to Rhodium-106 while emitting beta rays with a maximum energy of 39 keV. Rhodium-106 with a half life of 36 seconds, decays to Palladium-106 with a maximum energy of 3.5 MeV and a mean energy of 1.5 MeV providing the effective therapeutic irradiation. Description of Ruthenium-106 applicators Ruthenium-106 applicators were developed by Lommatzsch and Vollmar in the 60’s in East Berlin (1966) Lommatzsch PK. Opthalmologische Onkologie. 1999 Enke, Stuttgart. The 106 Ru is almost equally distributed on the concave surface of the shell shaped applicator. (Fig. 17, 18)

The isotope is deposited on a thin (0.2 mm) target foil made of silver, which in turn is bonded to a 0.7mm silver backing. Towards the eye ball a thin (0.1 mm) silver foil covers the Ruthenium-106; this has almost no absorption effect on the beta ray emission. On the other hand, the convex surface absorbs more than 95% of all radioactivity. This fact is important for radiation protection of the personnel in the operating theatre with regard to the handling of the applicator, as only minor safety procedures have to be undertaken. The concave irradiating side of the applicator should always be protected as electrons have a limited range in tissue but much greater in air. There is an inactive edge at the peripheral margin of the applicator which is reported to be about 0.7 - 0.8 mm. The external diameter varies between 15 and 20 mm and the spherical radius between 12 and 14 mm. There are usually two lugs on each applicator, by which it can be sutured to the sclera. Small grooves can be put onto the convex side of the applicator, enabling fixation of the applicator against the sclera by a suture across the applicator. There are applicators with notches, enabling positioning near the optic nerve and applicators for brachytherapy of ciliary body melanoma.The nominal activity varies (dependent on the size of the applicator) between 13 and 26 MBq or 0.35 and 0.7 mCi. (Fig. 19) Iodine-125 eye plaque brachytherapy Iodine-125 is a gamma emitter with a half-life of 60 days. 125 I decays exclusively by electron capture to an excited state of Tellurium-125 which spontaneously decays to the ground state with the emission of 35.5 kV gamma photons. Characteristic x-rays in the range of 27-35 kV are produced also due to electron capture and internal conversion. The half value layer for gold is only 0.025 mm, hence 0.5mmof gold is sufficient to absorb 99.95%of the incident gamma rays. The low energy of I-125 results also in a lesser radiation safety problem than using other gamma emitting sources, such as Co-60 or Ir-192. Description of the applicators Iodine-125 was introduced to treat ocular tumours at the end of the seventies (30). Presently 125 I is predominantly used in the US to treat uveal melanoma. (Fig. 20) In many centres Iodine-125 plaques are individually fabricated using multiple seeds imbedded in a concave metal plaque. The most prominent models available are the COMS plaque, the ROPES (RadiationOncology Physics and Engineering Services, Australia) plaque and the EyePhysics plaque. These plaques follow the same concave designwith fixed inner radius. However, non-standard dome shaped designs have been developed recently for the treatment of iris melanoma. Iodine seeds are usually delivered in titaniumencapsulation (0.05mm) containing between 0.5 – 20 mCi of I-125, the (outer) seed dimension is about 5 x 1 mm. Because of the presence of titanium and end welds the dose distribution around iodine seeds is highly anisotropic. The seeds are adhered to the concave portion of the plaque with an adhesive and at completion of treatment they are removed by dissolving the adhesive and are re-used. Instead of using an adhesive, silicon acrylic plaque inserts may also be used to accommodate seeds (4 - 18, depending on plaque size) at fixed positions. Many centres report the use of customdesigned inserts, some of which are manufactured on 3D printers. The major difficulties in the design of Iodine-125 seed eye plaques result from the need to have a thin device to slip over the surface of the eye and the relatively bulky physical dimensions of the seeds.

Uveal Melanoma

10

THE GEC ESTROHANDBOOKOF BRACHYTHERAPY | Part II Clinical Practice Version 1 - 15/04/2020

11a

11b

11c

Fig. 7: The conjunctiva is opened at the limbus and the rectus muscles are isolated for traction with silk sutures to expose the quadrant.

8a

8b

Fig. 11a, 11b and 11c: Template replaced with radioactive plaque

Fig. 8a and 8b: localization of tumour margins by transillumination

9a

9b

Fig. 12: Intraoperative US plaque position check / A poorly-positioned plaque - B well-positioned plaque

Fig. 9a and 9b: Outlined tumour borders plus 2-3 mm free margin

Well positioned plaque

Poorly positioned plaque

10a

10b

Fig. 13a: Follow-up images – based on the relationship between the tumour bed area and the actinic regression zone after treatment, it is possible to evaluate the right plaque position during the follow-up

Fig. 10a and 10b: Dummy plaque is used to position the sutures

Uveal Melanoma

11

THE GEC ESTROHANDBOOKOF BRACHYTHERAPY | Part II Clinical Practice Version 1 - 15/04/2020

The gold/titanium layer in the concave plaque absorbs radiation and hence minimizes doses to healthy structures in dorsal direction. The ability to attenuate 125 I with a thin layer of gold has allowed the production of rimmed plaques. The rim further reduces the lateral dose contributions to healthy structures of the eye and helps therefore to reduce ocular morbidity. On the other hand, higher plaque failures have been observed in thin tumours in close proximity to the optic nerve when using rimmed plaques. Similar to 106 Ru, applicator plaque diameters vary according to the dimension of the tumour. Cutouts are used for tumours next to the optic nerve.

The maximum tumour prominence as measured by ultrasound is taken and 1 mm is added for the sclera thickness to arrive at the target depth, or the measured distance from apex to scleral external surface is used. The dose rate per hour at the target depth (MinimumTarget Dose timeapp /hour) (MTD) is calculated based on the reference dose rate per minute at the target depth given in the calibration certificate of the manufacturer for the relative depth dose curve for a given 106 Ru applicator (MTD time0 /minute). This MTD time0 is multiplied with a factor representing the decay of the Ruthenium-106 since the time of production (factordecay) and multiplied by 60.The decay factor is looked up in a table or diagram indicating the decay of Ruthenium-106 over time:

MTD

/hour = MTD

/minute x Factor

x 60.

timeapp

time0

decay

The overall time of application (hours) is calculated by dividing the prescribed target dose (Gy) by the minimum target dose rate (Gy/hour):

10. DOSE, DOSE RATE, FRACTIONATION

At present, although there is no clear consensus about the required dose to the target, a dose to the tumour apex of at least 80-100 Gy is recommended. However, the minimum target (top) dose and the scleral surface dose should always be recorded and reported. As the critical structures to be spared fromexcessively high radiation doses (fovea, optic disc, choroid, retina) are located at the level of the tumour base, the doses at the tumour base are the most relevant for the assessment of brachytherapy-related morbidity Dosimetry of Ruthenium-106 eye plaque brachytherapy The usual recommended total dose of Ruthenium-106 brachytherapy is about 100 Gy, specified at the apex of the tumour which represents the CTV. The dose is calculated along the central axis of the applicator at the apex of the CTV. This minimum target dose is also to be given to the whole base of the CTV, taking carefully into account the dimensions of the applicator, of the active layer of Ruthenium-106 on its concave surface, and of the corresponding isodose distribution at the peripheral margins (inactive edge ~0.8 mm). The dose at the sclera varies significantly, somewhere between 200 and 1000 Gy in tumours which are 2 - 5 mm thick. This variation is due to the different tumour thickness treated. In the case of larger tumours, the scleral dose may even go beyond 1000 Gy, which is usually well tolerated. The dose rate on the day of application is dependent on the actual activity of the Ruthenium-106 applicator and the prescribed dose to the CTV, taking also into account the dose to the sclera, uvea and critical structures near the CTV, for example at the optical disc. The dose rate at the beginning of an active 106 Ru eye plaque is specified at the concave surface and is usually around 12 - 20 cGy per minute or 720 - 1200 cGy per hour which is mediumdose rate (MDR). Because of the rapid dose fall off, the dose rate changes with depth to 5 - 9 cGy per minute or 300 - 540 cGy per hour at 3 mm (43%) and to 2.4 - 4 cGy per minute or 150 - 240 cGy per hour at 5 mm (20%) which is a fall from MDR to low dose rate (LDR). As the half life time of 106 Ru is about 367 days, there is a significant change in dose rate over time towards lower dose rates. For a given applicator the dose rate at the beginning is specified by the manufacturer. A dose rate correction should therefore be used to account for decay with time (ref Marinkovic et al, Eur J Cancer 2016 and Br J Ophtalmol 2018).

app (hours) = ---------------------------- Prescribed MTD (Gy) MTD timeapp (Gy/hour).

Time

The scleral dose is calculated in the same way. For a prescribed dose of 100 Gy at the tumour apex, the duration of a 106 Ru application is somewhere between 1 and 7 days depending on the various factors mentioned. Dosimetry of iodine-125 applicators Patients are treated with dose rates between 50 - 100 cGy/hr. Apical doses between 70 - 150 Gy are given and doses at the tumour base are in the range between 200 - 700 Gy. Depending on tumour size treatment times vary between 30 and 300 hours. Due to the individual source seed arrangements in the gold plaque treatment planning for Iodine-125 is more complex. Treatment planning is usually performed using dedicated software. Model calculations take into account the active length of the seeds, scatter within the phantom and anisotropy of dose distribution from a single seed. However, approximations are made such as the seed is simulated by an unfiltered line source located at the geometrical centre of the seed. Other models used for dosimetric calculations are based on a point source assumption. These simplified models are sufficiently accurate for clinical calculations. Deviations of a few percent between TLD measurements and calculations have been reported in literature. The resulting isodose distribution when using multiple seed arrangements in rimmed or un-rimmed plaques can be irregular or asymmetric. When performing an individual plaque construction, the radiation physicist should be consulted and in order to avoid misalignment of the plaque, especially when using custom made plaques with asymmetric seed configuration, it is recommended that the physicist is present at the time of surgery in the operating room with isodose curves Dosimetric uncertainty and margins In recent years the dosimetric uncertainties of plaque treatments have been discussed, with the goal to validate the recommendations found in the literature. Some of these recommendations were made more than a decade ago. These studies aimed to provide more accurate models based on improved experimental measurements andmore sophisticatedMonte Carlo (MC) based dose calculations. While most of the studies were able to confirm the commonly

Uveal Melanoma

12

THE GEC ESTROHANDBOOKOF BRACHYTHERAPY | Part II Clinical Practice Version 1 - 15/04/2020

accepted margins, some have recommended that extra care is required for cases in which the lateral extent of the tumour is particularly large. Individual quality assurance of the surface dose rate homogeneity can help in avoiding critical underdosages in the target volume.

Specific patterns of tumour regression are typical for different tumour types and also vary with the treatment method. A rapid complete remission with a sharply delineated flat white scar may be achieved but more often, tumour regression takes place slowly resulting in a white scar at the margin and a prominent grey mass in the centre (“grey mouse”). Sometimes there is almost no change in tumour extension and height for years.Therapeutic decisions are usually taken if tumour regrowth is suspected because of tumour enlargement on ophthalmoscopy, ultrasonography, or fluorescence angiography or because of a decrease in ultrasonographic reflectivity. Follow-up after eye-conserving treatment is usually based on ophthalmoscopy, and ultrasound. Fluorescein angiography (FA) and optical coherence tomography (OCT) are useful for early detection of ischaemic changes leading to radiation retinopathy, maculopathy and optic neuropathy. Visual loss in the treatment area gradually occurs over the course of 2-3 years. Recent data have shown that use of exclusive plaque brachytherapy is preferred to brachytherapy plus routine use of adjuvant TTT in view of the better residual visual acuity . However, in cases of underdosage to the tumour area, especially for posterior disease, the use of adjuvant TTT could be taken in the account for improving tumour regression and local control. Eye plaque brachytherapy is mainly based on Ruthenium-106 and Iodine-125. In 2002, the Collaborative Ocular Melanoma Study reported that local treatment failure occurred in 10.3% of patients who underwent iodine 125 plaque brachytherapy. More recent studies revealed local recurrence rates of 4-15% at 5 years dependent on various factors (T-size category according to AJCC, presence of ciliary body involvement or extrascleral extension). These data are important, in that local tumour recurrence is a complication thought to be associated with an increased risk of uveal metastasis and thus decreased survival. Metastatic disease is demonstrated in 28% with local recurrence and the median time from local recurrence to metastasis is 6.8 years. The 5- and 10-year overall survival is 87% and 82% respectively. Based on AJCC staging for posterior uveal melanoma, 10-year metastatic rate is 12% for stage I tumours, 29% for stage II, and 61% for stage III. The risk of metastasis and death increased three-fold with each increasing melanoma staging. Enucleation is required in about 6% of all patients, the majority during the first years. The most common reason for secondary enucleation is tumour regrowth (51% of cases), followed by uncontrollable neovascular glaucoma (31%), scleral melting (7%) painful bullous keratopathy (2%) and painful hemolytic glaucoma (2%) . Management of Recurrences Radiation treatment by plaque brachytherapy has the lowest rate of treatment failures. Iodine-125 and Ruthenium-106 (106Ru) brachytherapy are both associated with a weighted average local recurrence rate of 4-15% and the recent introduction of the intraoperative ultrasound plaque localization during brachytherapy has been shown to further reduce the risk of local treatment failure. Stereotactic radiotherapy has similar rates of 2% charged particle radiation therapy has a lower weighted average rate of local failure at 4.2%. Overall, surgical modalities have a higher rate of local failure compared with radiation modalities (18.6% vs 6.15%). TranspupillaryThermotherapy has the largest reported variation in failure rates from0% to 55.6%, with a weighted average of 20.8%. It is possible to observe different kinds of recurrence and depending of their growth pattern, they may take the form of marginal, central, diffuse or distant recurrence or extrascleral

11. MONITORING

During the treatment time (applicator in place) only general rules of radioprotection have to be taken into consideration. For the beta emitter Ruthenium-106, the electrons are almost completely absorbed within the eye ball.There is an extremely small amount of bremsstrahlung background near the patient, which usually can be ignored. Each person in contact with such patient has to be allowed to come into contact with radiation and should be monitored for radiation protection based on the internal institutional policy and according to national radiation protection rules. For iodine-125, the low energy photons are not completely absorbed within the eye ball. Half value layers in water (or tissue) for Iodine-125 and Pd-103 are less than 2 cm. The half value layer for 125 I in lead is 0.025 mm, hence common lead shielding provides sufficient protection. However, for eye plaque brachytherapy the general radiation protection rules should always be taken into consideration for those frequently handling applicators: minimizing exposure time for personal and patient, maximizing distance (e.g. by using forceps) and using protection where possible (e.g. by wearing lead aprons). Specific acute side effects like swelling of the conjunctiva and acute exudative retinal detachment are transient and require symptomatic treatment only. Prophylactic systemic treatment with oral non-steroidal anti- inflammatory drugs may be indicated, while local treatment after surgery includes atropine and anti-inflammatory medications, in combination with a local antibiotic. After brachytherapy, patients are followed for local control, complications, and systemic disease every 3-6 months. Patients with posteriorly located tumours are at higher risk of radiation maculopathy and radiation optic neuropathy, and visual acuity is checked at each visit. These effects typically occur gradually within the first 3 years of follow-up. Similarly, most local tumour recurrence occurs during the first 5 years. In addition, patients should be periodically reexamined for evidence of metastatic disease. Imaging studies of the liver and biochemical tests of liver function are necessary as the baseline work up.

12. RESULTS

Treatment results and morbidity are specific for the different isotopes ruthenium-106 and iodine-125.