Uveal Melanoma: Evidence for Efficacy of Therapy
Uveal Melanoma: Evidence for Efficacy of Therapy
The surest way of achieving local tumor control is by enucleation, although this does not entirely exclude the risk of orbital recurrence. This may be the best option in a patient presenting with a blind, painful eye.
Of all the eye-conserving forms of treatment, proton beam radiotherapy is the modality associated with the lowest overall risk of local tumor recurrence (3.5% local recurrence at 5 y, 5% at 10 y). This is because adjustments can be made to the treatment plans even if tantalum markers are not positioned or measured accurately. The greatest risk of treatment failure after proton beam radiotherapy is with large ciliary body tumors, which tend to spread diffusely in a circumferential manner, so that it is prudent in these cases to administer treatment with a 4 mm lateral safety margin instead of the usual 2.5 mm. The risk of recurrence is also increased with diffuse iris melanomas, because of clinically invisible spread around the angle and seeding to distant parts of the anterior chamber (Fig. 1). In such cases, local recurrence is avoided by irradiation of the entire anterior segment, as far posteriorly as the ora serrata. Such extensive treatment is usually well tolerated, although glaucoma may be a problem in some patients. Tube shunts have been utilized in these patients with few complications. Central tumor recurrences have been reported after proton beam radiotherapy of iris melanomas. Whether these have occurred because some tumors are radioresistant or whether a narrow Bragg peak has been placed at the wrong depth is uncertain. Recently, transpalpebral proton beam radiotherapy as a means of avoiding collateral damage to the upper eyelid margin has been shown to avoid eyelid and ocular surface complications without compromising local tumor control (Fig. 2).
(Enlarge Image)
Figure 1.
A diffuse iris melanoma is seen extending into the angle and ciliary body temporally, with additional seeding of the inferior anterior chamber angle. Proton beam radiotherapy was administered to the entire anterior segment and ciliary body, up to the ora serrata.
(Enlarge Image)
Figure 2.
A, Vertical side elevation through the center of a radiation field of a typical proton beam radiotherapy patient. This shows the effect on the isodoses of modeling the upper eyelid, where its thickness is included in the maximum range calculation (version 3.6 of the EyePlan software). The numbered circles refer to the tantalum markers that delimit the size and position of the tumor. B, Transpalpebral proton beam radiotherapy for choroidal melanoma, sparing the eyelid rim. From Konstandinidis et al.
The local tumor recurrence rate after iodine-125 plaque brachytherapy as reported by the Collaborative Ocular Melanoma Study (COMS) was 10% [95% confidence interval (CI), 8–13] of 657 patients at 5 years. Marginal tumor recurrence is most likely with larger posterior tumors, particularly if they have an irregular shape. This is because they are difficult to localize. For smaller tumors, recurrence is 3% at 7 years. Another risk factor is increased tumor thickness, because of the way in which the dose of radiation diminishes with depth, especially with ruthenium plaques, which emit β-radiation. To obviate these risks, guidelines suggest positioning the plaque so that it physically overlaps the entire tumor margin by at least 2 mm. This method inevitably increases the risk of collateral damage to optic nerve and fovea so that there is a tendency to reduce the radiation dose as much as possible, which, however, increases the chances of a central recurrence.
The second author (B.E.D.) has therefore developed techniques and instruments for enhancing local tumor control with ruthenium applicators while reducing irradiation of the optic disc and fovea. Briefly, these methods involve the use of a plaque template, which has perforations through which the globe is transilluminated while performing binocular indirect ophthalmoscopy (Fig. 3). This procedure enhances the accuracy of plaque placement, making it possible to safely position the plaque eccentrically, with its posterior edge aligned with the posterior tumor margin. Because the plaque is then located further from the optic disc and fovea, the risk of damaging these structures is decreased despite delivering a higher dose of radiation (ie, >350 Gy to sclera and >85 Gy to the tumor apex). Other authors advocate intraoperative ultrasonography and the use of other isotopes, such as iodine and palladium, which have a greater range. A problem with such isotopes is that they deliver higher doses of radiation to healthy ocular tissues, especially in the case of iodine-125.
(Enlarge Image)
Figure 3.
Technique for ruthenium plaque placement. The longitudinal basal tumor diameter is measured with B-scan echography so that the intended plaque location can be determined (A). A fundus drawing is made (B). Tumor is localized by transillumination (C). Tumor margins and intended location of the anterior plaque edge are marked on the sclera (D). The plaque template is placed on the sclera so that its anterior edge lies over the relevant ink mark (E). The template is pressed against the eye to create scleral dimples, indicating suture entry and exit points (F). The plaque template has grooves and perforations to guide and position the transilluminator (G). The plaque position is confirmed with indirect ophthalmoscopy (H). The template is replaced by the plaque, which is secured with the same lug sutures (I). The rectus muscles are reattached to the sclera (J). From Damato and Singh.
Recently, radiosurgery and stereotactic radiotherapy have been used to treat choroidal melanomas. These modalities can be considered in patients unsuitable for plaque brachytherapy due to large-sized or peripapillary or very posteriorly located tumors, in which tantalum marker placement is technically challenging. High rates of local tumor control have been reported (95.9% after 5 y and 92.6% after 10 y with stereotactic radiotherapy). Long-term complications are similar to proton beam and plaque brachytherapy and include optic neuropathy, radiation retinopathy, cataract, exudative retinal detachment, and neovascular glaucoma. The long-term overall eye retention rate is approximately 70% to 75%. To date, stereotactic radiotherapy has not been investigated in the treatment of iris melanoma, probably because it is likely to cause severe eyelid damage with such tumors.
Transscleral resection, also known as "eyewall resection" or "exoresection," was associated with higher rates of local tumor recurrence (6% to 57%) than radiotherapy in initial studies. To some extent this is because the tumors are relatively large and therefore more likely to be aggressive, with more extensive diffuse extension. The incidence of local treatment failure has diminished considerably (to 5% to 10%) following the introduction of adjunctive ruthenium plaque radiotherapy, particularly with the use of a 25 mm ruthenium applicator (Fig. 4). The plaque is left in place for about 1 day so that a dose of 100 Gy is delivered to a depth of 1 mm. An advantage of this approach is that it eliminates the need for wide surgical margins, thereby reducing ocular morbidity.
(Enlarge Image)
Figure 4.
Exoresection of a choroidal melanoma with ruthenium plaque placement. Disinsertion of medial rectus muscle (A). Creation of posteriorly hinged partial thickness scleral flap (B). Dissection of scleral flap (C). Closure of vortex vein with bipolar cautery (D and E). Limited core vitrectomy (F). Biopolar cautery to short posterior ciliary arteries (G). Deep sclerotomy (H). Division of deep sclera around tumor margins (I). Choroidotomy (J). Separation of tumor from ciliary epithelium by blunt dissection (K). Division of choroid around tumor margins (L). Balanced salt solution injection to reinflate eye (M). Insertion of plaque template (N). Replacement of template with radioactive plaque (O). Reapproximation of conjunctiva, after reinsertion of medial rectus (P). From Damato and Singh.
Endoresection or "transretinal resection" (Figs. 5,6) is controversial because of intuitive concerns that the piecemeal tumor removal with a vitreous cutter will disseminate viable tumor cells around the eye and into episcleral tissues, causing local tumor recurrence and metastatic disease. Long-term follow-up studies have shown rates of local tumor recurrence of ≤5%, similar to plaque brachytherapy and proton beam radiotherapy. The greatest danger is with residual tumor in the choroid, which can spread extraocularly and can also grow through the retinotomy to seed around the vitreous cavity. Konstantinidis et al reported the largest long-term case series of endoresection (71 patients) to date, with a local recurrence actuarial rate of 3.7% at 10 years. Although none of those patients suffered systemic complications, a case of a fatal air embolism after endoresection has been reported. The postulated mechanism was exposure of open choroidal and tumor vasculature to infused air at high pressure. To avoid this rare but serious complication, the surgical technique has been modified. Instead of performing an air-fluid exchange, perfluorocarbon liquid is used to flatten the retina. Direct perfluorocarbon liquid-silicone oil exchange can then be performed.
(Enlarge Image)
Figure 5.
Endoresection of a choroidal melanoma. After a complete pars plana vitrectomy, tumor is removed through a retinotomy (A). Heavy liquid (such as perfluoro-n-octane [PFO]) is used to flatten the retina (B). Endolaser is applied to destroy any remaining tumor cells and for retinopexy (C). Silicone oil is then placed for taponade (D). From Damato.
(Enlarge Image)
Figure 6.
A, Preoperative fundus photograph of a massive choroidal melanoma, with hand-motion vision. B, Ten months' postoperative fundus photograph after endoresection, silicone oil removal, and cataract extraction with intraocular lens placement. Vision was 20/30
Photocoaguation was associated with high rates of local tumor recurrence and has been superseded by transpupillary thermotherapy. This involves heating the tumor for 60 seconds, with a 3-mm diode laser. Recurrence rates of up to 20% have been reported so that this treatment is suitable only for very small melanomas and for melanocytic tumors of uncertain malignancy, that is, "suspicious nevi" (Fig. 7). Photodynamic therapy can be effective with some amelanotic choroidal melanomas, but fails in approximately 50% of cases so that it should be administered only when the patient accepts such a high risk of local recurrence.
(Enlarge Image)
Figure 7.
Superotemporal choroidal melanoma in the right eye before transpupillary thermotherapy (A) and after treatment (B). From Damato.
There is a need for better methods for describing local tumor recurrence (eg, distinguishing central from marginal recurrence), and also for differentiating possible from definite treatment failure. Outcome analyses should take account of salvage therapies, which ultimately determine whether or not vision and the eye are conserved.
Local Tumor Control
Enucleation
The surest way of achieving local tumor control is by enucleation, although this does not entirely exclude the risk of orbital recurrence. This may be the best option in a patient presenting with a blind, painful eye.
Proton Beam Radiotherapy
Of all the eye-conserving forms of treatment, proton beam radiotherapy is the modality associated with the lowest overall risk of local tumor recurrence (3.5% local recurrence at 5 y, 5% at 10 y). This is because adjustments can be made to the treatment plans even if tantalum markers are not positioned or measured accurately. The greatest risk of treatment failure after proton beam radiotherapy is with large ciliary body tumors, which tend to spread diffusely in a circumferential manner, so that it is prudent in these cases to administer treatment with a 4 mm lateral safety margin instead of the usual 2.5 mm. The risk of recurrence is also increased with diffuse iris melanomas, because of clinically invisible spread around the angle and seeding to distant parts of the anterior chamber (Fig. 1). In such cases, local recurrence is avoided by irradiation of the entire anterior segment, as far posteriorly as the ora serrata. Such extensive treatment is usually well tolerated, although glaucoma may be a problem in some patients. Tube shunts have been utilized in these patients with few complications. Central tumor recurrences have been reported after proton beam radiotherapy of iris melanomas. Whether these have occurred because some tumors are radioresistant or whether a narrow Bragg peak has been placed at the wrong depth is uncertain. Recently, transpalpebral proton beam radiotherapy as a means of avoiding collateral damage to the upper eyelid margin has been shown to avoid eyelid and ocular surface complications without compromising local tumor control (Fig. 2).
(Enlarge Image)
Figure 1.
A diffuse iris melanoma is seen extending into the angle and ciliary body temporally, with additional seeding of the inferior anterior chamber angle. Proton beam radiotherapy was administered to the entire anterior segment and ciliary body, up to the ora serrata.
(Enlarge Image)
Figure 2.
A, Vertical side elevation through the center of a radiation field of a typical proton beam radiotherapy patient. This shows the effect on the isodoses of modeling the upper eyelid, where its thickness is included in the maximum range calculation (version 3.6 of the EyePlan software). The numbered circles refer to the tantalum markers that delimit the size and position of the tumor. B, Transpalpebral proton beam radiotherapy for choroidal melanoma, sparing the eyelid rim. From Konstandinidis et al.
Brachytherapy
The local tumor recurrence rate after iodine-125 plaque brachytherapy as reported by the Collaborative Ocular Melanoma Study (COMS) was 10% [95% confidence interval (CI), 8–13] of 657 patients at 5 years. Marginal tumor recurrence is most likely with larger posterior tumors, particularly if they have an irregular shape. This is because they are difficult to localize. For smaller tumors, recurrence is 3% at 7 years. Another risk factor is increased tumor thickness, because of the way in which the dose of radiation diminishes with depth, especially with ruthenium plaques, which emit β-radiation. To obviate these risks, guidelines suggest positioning the plaque so that it physically overlaps the entire tumor margin by at least 2 mm. This method inevitably increases the risk of collateral damage to optic nerve and fovea so that there is a tendency to reduce the radiation dose as much as possible, which, however, increases the chances of a central recurrence.
The second author (B.E.D.) has therefore developed techniques and instruments for enhancing local tumor control with ruthenium applicators while reducing irradiation of the optic disc and fovea. Briefly, these methods involve the use of a plaque template, which has perforations through which the globe is transilluminated while performing binocular indirect ophthalmoscopy (Fig. 3). This procedure enhances the accuracy of plaque placement, making it possible to safely position the plaque eccentrically, with its posterior edge aligned with the posterior tumor margin. Because the plaque is then located further from the optic disc and fovea, the risk of damaging these structures is decreased despite delivering a higher dose of radiation (ie, >350 Gy to sclera and >85 Gy to the tumor apex). Other authors advocate intraoperative ultrasonography and the use of other isotopes, such as iodine and palladium, which have a greater range. A problem with such isotopes is that they deliver higher doses of radiation to healthy ocular tissues, especially in the case of iodine-125.
(Enlarge Image)
Figure 3.
Technique for ruthenium plaque placement. The longitudinal basal tumor diameter is measured with B-scan echography so that the intended plaque location can be determined (A). A fundus drawing is made (B). Tumor is localized by transillumination (C). Tumor margins and intended location of the anterior plaque edge are marked on the sclera (D). The plaque template is placed on the sclera so that its anterior edge lies over the relevant ink mark (E). The template is pressed against the eye to create scleral dimples, indicating suture entry and exit points (F). The plaque template has grooves and perforations to guide and position the transilluminator (G). The plaque position is confirmed with indirect ophthalmoscopy (H). The template is replaced by the plaque, which is secured with the same lug sutures (I). The rectus muscles are reattached to the sclera (J). From Damato and Singh.
Stereotactic Radiotherapy
Recently, radiosurgery and stereotactic radiotherapy have been used to treat choroidal melanomas. These modalities can be considered in patients unsuitable for plaque brachytherapy due to large-sized or peripapillary or very posteriorly located tumors, in which tantalum marker placement is technically challenging. High rates of local tumor control have been reported (95.9% after 5 y and 92.6% after 10 y with stereotactic radiotherapy). Long-term complications are similar to proton beam and plaque brachytherapy and include optic neuropathy, radiation retinopathy, cataract, exudative retinal detachment, and neovascular glaucoma. The long-term overall eye retention rate is approximately 70% to 75%. To date, stereotactic radiotherapy has not been investigated in the treatment of iris melanoma, probably because it is likely to cause severe eyelid damage with such tumors.
Eyewall Resection
Transscleral resection, also known as "eyewall resection" or "exoresection," was associated with higher rates of local tumor recurrence (6% to 57%) than radiotherapy in initial studies. To some extent this is because the tumors are relatively large and therefore more likely to be aggressive, with more extensive diffuse extension. The incidence of local treatment failure has diminished considerably (to 5% to 10%) following the introduction of adjunctive ruthenium plaque radiotherapy, particularly with the use of a 25 mm ruthenium applicator (Fig. 4). The plaque is left in place for about 1 day so that a dose of 100 Gy is delivered to a depth of 1 mm. An advantage of this approach is that it eliminates the need for wide surgical margins, thereby reducing ocular morbidity.
(Enlarge Image)
Figure 4.
Exoresection of a choroidal melanoma with ruthenium plaque placement. Disinsertion of medial rectus muscle (A). Creation of posteriorly hinged partial thickness scleral flap (B). Dissection of scleral flap (C). Closure of vortex vein with bipolar cautery (D and E). Limited core vitrectomy (F). Biopolar cautery to short posterior ciliary arteries (G). Deep sclerotomy (H). Division of deep sclera around tumor margins (I). Choroidotomy (J). Separation of tumor from ciliary epithelium by blunt dissection (K). Division of choroid around tumor margins (L). Balanced salt solution injection to reinflate eye (M). Insertion of plaque template (N). Replacement of template with radioactive plaque (O). Reapproximation of conjunctiva, after reinsertion of medial rectus (P). From Damato and Singh.
Endoresection
Endoresection or "transretinal resection" (Figs. 5,6) is controversial because of intuitive concerns that the piecemeal tumor removal with a vitreous cutter will disseminate viable tumor cells around the eye and into episcleral tissues, causing local tumor recurrence and metastatic disease. Long-term follow-up studies have shown rates of local tumor recurrence of ≤5%, similar to plaque brachytherapy and proton beam radiotherapy. The greatest danger is with residual tumor in the choroid, which can spread extraocularly and can also grow through the retinotomy to seed around the vitreous cavity. Konstantinidis et al reported the largest long-term case series of endoresection (71 patients) to date, with a local recurrence actuarial rate of 3.7% at 10 years. Although none of those patients suffered systemic complications, a case of a fatal air embolism after endoresection has been reported. The postulated mechanism was exposure of open choroidal and tumor vasculature to infused air at high pressure. To avoid this rare but serious complication, the surgical technique has been modified. Instead of performing an air-fluid exchange, perfluorocarbon liquid is used to flatten the retina. Direct perfluorocarbon liquid-silicone oil exchange can then be performed.
(Enlarge Image)
Figure 5.
Endoresection of a choroidal melanoma. After a complete pars plana vitrectomy, tumor is removed through a retinotomy (A). Heavy liquid (such as perfluoro-n-octane [PFO]) is used to flatten the retina (B). Endolaser is applied to destroy any remaining tumor cells and for retinopexy (C). Silicone oil is then placed for taponade (D). From Damato.
(Enlarge Image)
Figure 6.
A, Preoperative fundus photograph of a massive choroidal melanoma, with hand-motion vision. B, Ten months' postoperative fundus photograph after endoresection, silicone oil removal, and cataract extraction with intraocular lens placement. Vision was 20/30
Phototherapy
Photocoaguation was associated with high rates of local tumor recurrence and has been superseded by transpupillary thermotherapy. This involves heating the tumor for 60 seconds, with a 3-mm diode laser. Recurrence rates of up to 20% have been reported so that this treatment is suitable only for very small melanomas and for melanocytic tumors of uncertain malignancy, that is, "suspicious nevi" (Fig. 7). Photodynamic therapy can be effective with some amelanotic choroidal melanomas, but fails in approximately 50% of cases so that it should be administered only when the patient accepts such a high risk of local recurrence.
(Enlarge Image)
Figure 7.
Superotemporal choroidal melanoma in the right eye before transpupillary thermotherapy (A) and after treatment (B). From Damato.
There is a need for better methods for describing local tumor recurrence (eg, distinguishing central from marginal recurrence), and also for differentiating possible from definite treatment failure. Outcome analyses should take account of salvage therapies, which ultimately determine whether or not vision and the eye are conserved.
Source...