Article Text
PDF
Abstract
Background Treatment modalities in iris melanoma include excision, plaque radiotherapy, photon or proton beam therapy and enucleation. In extensive tumours and diffuse seeding, radiotherapy remains as an alternative to enucleation.
Methods This study is a retrospective, consecutive, interventional, single-institutional case series. 54 patients with a diffuse and non-resectable iris melanoma diagnosed from September 1998 to June 2012 were included. A 68-megaelectron volt proton beam was used to treat the anterior segment with a total dose of 4×12.5 cobalt grey equivalent. The cases were evaluated for local tumour control, eye retention, functional outcome and local complications after treatment.
Results During a mean follow-up of 62.7 months (median 54.8 months, range 5.5–159.6 months), local tumour control was achieved in 96.3% of the patients. Cataract and glaucoma were the main complications developing after irradiation in 42.6% and 55.6%, respectively. In 34 of 44 patients (77.3%) who underwent cataract removal, a visual acuity of 20/40 or better following surgery was preserved. Enucleation was performed in three patients. The reason was suspected tumour recurrence in one and glaucoma in two. Hepatic metastasis occurred in one patient.
Conclusions As an alternative to enucleation, whole anterior segment fractionated proton beam radiotherapy offered excellent local tumour control in diffuse iris melanoma. Given the limited alternatives, the rate of complications appears acceptable and visual function could be preserved in the majority of the patients during follow-up.
- Iris
- Neoplasia
Statistics from Altmetric.com
Introduction
Uveal melanoma arises in the iris in only 2%–3% of the cases.1 Iris melanomas have been divided into circumscribed and diffuse forms, the term ‘diffuse’ referring to flat, infiltrating growth or to seeding through the anterior chamber.2 Diffuse seeding can involve the posterior and anterior chamber and may lead to hyperchromia and ipsilateral glaucoma. Metastasis of iris melanoma occurs in about 5% of the cases after 10 years1 but in diffuse iris melanoma it has been reported to occur in up to 13% at 6 years.2 Treatment modalities in iris melanoma include excision, plaque radiotherapy, photon or proton beam therapy and enucleation. In extensive tumours and diffuse seeding, radiotherapy remains as an alternative to enucleation. In 1998, we started to employ proton beam radiotherapy (PBR) in non-resectable iris melanoma and we decided to treat the whole anterior segment to include all potentially malignant seeds when signs of diffuse seeding were present. Most recently, Konstantinidis et al3 published their experience with whole anterior segment PBR in 12 patients. In the present study, we report our results in 54 patients with a mean follow-up of close to 5 years.
Patients and methods
All patients with a clinical or histological diagnosis of iris melanoma who were managed with whole anterior segment PBR between 13 October 1998 and 19 June 2012 were included in this single-centre, retrospective case series. The indication for this treatment was non-resectable anterior chamber involvement from extensive growth or seeding of an iris melanoma. We excluded patients with tumours extending past the ora serrata and also eyes that had prior tumour treatment (brachytherapy, iridectomy and iridocyclectomy).
A complete ophthalmologic examination, assessment of demographic features and informed consent was obtained before radiotherapy. The radiation dose was 4×12.5 cobalt grey equivalent (CGE) in 53 patients and 8×7.5 CGE in the first patient. The dose distribution was planned with EYEPLAN and OCTOPUS software.
A full iris aperture was used with a diameter of 18 mm in 41 cases (75.9%), 17 mm in 11 cases (18.5%), 20 mm in two cases (3.7%) and 14 mm in one case. The field had been extended in two patients towards a scleral or a ciliary body extension of the tumour indicated by high-frequency ultrasound and diaphanoscopy and to include a pre-existing filtering bleb in one patient. Localising markers were not placed except in one patient with ciliary body involvement. Irradiation was performed without pupillary dilation.
Follow-up examinations were made at 3-month to 12-month intervals during the first 3 years and then at 6-month to 24-month intervals. Screening for metastasis was recommended with 6-monthly liver ultrasound. The results of the latest liver ultrasound were collected at every follow-up visit. In 23 cases, the primary caring ophthalmologist provided the latest follow-up information. In these cases, the latest information on functional and tumour-related data was collected by a standard questionnaire. Additionally in these cases, the local ophthalmologist was asked to report on the known metastatic status of the patient.
Statistics included Kaplan–Meier estimate of local tumour control, eye retention, cataract and glaucoma. Glaucoma at presentation was analysed as a risk factor for a visual acuity of <20/200 with log rank (Mantel–Cox) test. Only an irreversible visual loss to <20/200 was noted as an event with respect to this analysis.
The tenets of the Helsinki declaration were respected. Ethics committee approval was obtained (reference number EA4/037/13).
Results
Between October 1998 and June 2012, 54 patients were included in this study. The mean patient age at time of treatment was 54.9 years (median 60 years, range 10–89 years). Twenty-six patients (48.1%) were women.
There was a significant loss of follow-up in our cohort, but we are able to present 5-year data in 59.3% and 3-year data in 83.3% of the patients which should be of some value in patient counselling. Median follow-up was 4.6 years (mean±SD 5.2±3.4 years, range 0.5–13.3 years). Of the 54 patients, 51 completed the follow-up for 1 year, 39 for 3 years and 24 for 5 years. One further patient was lost to follow-up (<3 months of follow-up) and, therefore, excluded from the analysis.
The contiguous tumour involved three clock hours or more circumferential extension in 32 cases (59.3%), 50 tumours were in contact with the anterior chamber angle and signs of diffuse tumour seeding to the anterior chamber angle or iris surface were present in 49 cases. In two cases, possible iatrogenic seeding after biopsy with a cutter was anticipated by full iris irradiation. Five cases showed evidence of ciliary body extension but the tumour was located primarily in the iris. None of the tumours exceeded the posterior ciliary band. According to the American Joint Committee on Cancer– Union for International Cancer Control TNM7 classification, tumours were T1 in 48 cases (88.9%; T1a in 24 cases, T1b in 13 cases, T1c in 11 cases), T2 in four cases (7.4%) and T3 in two cases (3.7%).
The diagnosis was made based on clinical and histopathological data. Twenty-two patients (40.7%) presented with documented growth, and in a total of 24 cases (44.4%), a biopsy was performed. The remainder was diagnosed on clinical factors (diffuse and/or multifocal involvement, tumour extension to the anterior chamber, scleral infiltration, tumour thickness >2 mm and secondary glaucoma).
Biopsy before radiotherapy was performed in 14 cases. In two of them, histopathology indicated a naevus but treatment was recommended for documented growth. In another 10 patients, a biopsy was done after radiotherapy at the time of lens exchange. The overall histopathology data demonstrated melanocytoma in one case, naevus in two cases, spindle-shape melanoma in 17, mixed cell melanoma in two cases and epithelioid cell melanoma in two cases.
The 5-year Kaplan–Meier estimates for local tumour control and eye retention were 94.7% and 95.1%, respectively (figure 1). Three eyes were removed for glaucoma with poor visual acuity at local hospitals 3.6, 5.3 and 0.9 years after PBR. In one of these cases local recurrence was suspected clinically at another service, but was not confirmed in histopathology. However, this case was evaluated as clinical recurrence in this analysis.
Kaplan–Meier estimates of enucleation-free rate (A), recurrence-free rate (B). glaucoma-free rate (C) and cataract-free rate (D).
Continued oncological surveillance was ensured in 31 patients who had in-house aftercare during the complete time of follow-up; in 23 patients who in the end preferred external post-treatment investigation, the metastatic status was known at least in 43% of the period (ie, of about the first 2 years of follow-up). One patient developed metastasis 20.2 months after PBR and 9 months after filtrating surgery. This patient also developed local recurrence with a flat posterior extension managed by ruthenium plaque therapy.
The change in visual acuity is presented in table 1. One year after proton beam irradiation visual acuity was 20/200 or more in all except one patient. After 3 years, 84.4% (38/45) of the patients retained visual acuity of 20/200 or better; 88.2% (15/17) retained this level of vision through 7 years after proton beam irradiation. Similarly, after an initial drop from 20/25 to 20/35 1 year after irradiation the mean visual acuity remained stable till year 7.
Visual acuity during follow-up
The mean intraocular pressure was 21 mm Hg (median 18 mm Hg, range 10–65 mm Hg) before treatment and 21 mm Hg (median 19 mm Hg, range 8–60 mm Hg) at the latest follow-up visit. Glaucoma was present in 14 patients (25.9%) before radiotherapy and in 30 patients (55.6%) after treatment. One patient presented initially after trabeculectomy and three other patients had received cyclodestructive treatment for unilateral glaucoma before PBR. Nineteen patients received glaucoma surgery after radiotherapy, cyclodestructive procedures (in nine cases), trabeculectomy (in four cases) or both (in five cases). One patient received an Ahmed-valve implantation. At the last follow-up, glaucoma was controlled in 68.5%, moderately uncontrolled (23–30 mm Hg) in 22.2% and grossly uncontrolled (≥30 mm Hg) in 9.3%.
All patients were phakic at radiotherapy. Radiation-induced cataract was detected in 23 patients (42.6%) after a mean of 1.8 years (median 2.0 years, range 0.2–3.1 years). In a total of 44 patients (81.5%) with radiation-induced cataract, as well as cataract that arose before PBR and became more compact afterwards, phacoemulsification together with an intraocular lens implantation has been performed to date.
Recurrent, but finally healed, corneal erosion occurred in two cases, in one of them resulting in a central corneal opacity representing the only case of a visually significant corneal opacity. During tumour regression, visibility and engorgement of tumour vessels arose in some patients. Transient hyphaema occurred in two patients managed by observation. Iris rubeosis occurred in only one patient. Engorgement of perilimbal conjunctival vessels developed frequently as well as subtle signs of iris atrophy, in one case with symptomatic mydriasis (figure 2).
(A) Proton dose distribution with an 18 mm full iris aperture. The 50% isodose is in line with the ora serrata. (B) Inferior nodular iris melanoma (clinical diagnosis) with seeding throughout the iris. Later, lens exchange and cyclodestruction were performed but visual acuity dropped to no light perception due to glaucoma. The tumour remained controlled for 9.5 years after irradiation. (C) Before and (D) 6.5 years after proton beam radiotherapy, tumour regression and telangiectatic changes of the tumour vessels that lead to transient hyphaema were present. (E) Nodular iris melanoma with diffuse seeding before irradiation. (F) 1.8 years after irradiation and 1 day after lens exchange and tumour biopsy confirming spindle cell melanoma.
Corneoscleral necrosis, radiation retinopathy, papillopathy or hypotony were not observed but one patient developed chronic cystoid macular oedema following cataract removal.
Discussion
In spite of the low-grade histology of most melanocytic iris lesions, Bechrakis and Lee demonstrated the dedifferentiation potential with progression from a spindle-shape to an epithelioid phenotype in several clinicopathological cases.4 When tumour dispersion in the anterior chamber is present, an irrefutable clinical differentiation into viable tumour seeds, melanomalysis, melanomacrophages or acellular pigment deposition may not be possible. Complete removal of such dispersions is not possible either and biopsy bears the risk of sample error. In addition, the extent of tumour dissemination in iris melanoma has been shown to be much greater in histopathology than suggested by biomicroscopy.4 Consequently, we applied PBR to the whole anterior segment, when anterior chamber dissemination was clinically evident or suspected.
Gragoudas et al5 demonstrated an equal effect in terms of tumour control comparing 50 CGE with 70 CGE proton beams in the treatment of posterior uveal melanoma. We selected a total dose of 50 CGE in four fractions to minimise potential radiation-induced complications.
Enucleation could be avoided in most of our patients. Local tumour control seems to be very high and the only well-documented case of recurrence probably represents a progression of a clinically undetected posterior choroidal extension.
Given a high-dose distribution to the entire anterior segment, the limited extent of local complications may be surprising. Cataract turned out to be a frequent and expected complication and was managed accordingly without signs of zonular insufficiency. The median time from PBR to cataract surgery was 26.9 months in this series.
The final visual acuity became poor (<20/200) in 10 patients (18.5%) mainly due to glaucoma. Secondary glaucoma increased from 28% before to 53% after radiotherapy, and pre-existing glaucoma was a risk factor for eventual visual loss (figure 3). Its management is complex, deserving a separate investigation. In general, it was non-neovascular and not associated with clinically evident significant melanomalysis. Trabecular scarring might contribute to postirradiation glaucoma in this series but it may also reflect the natural history of tumour-related glaucoma in part. Sharkawi et al6 used Baerveldt shunts effectively in a series of 31 patients after anterior segment PBR. We also could control the intraocular pressure with medical and surgical intervention in several patients, but as many patients lived at long distances and examinations were scheduled rather infrequently, glaucoma therapy might have been less rigorous than desirable. Early in this series, we have had a fairly cautious attitude towards filtering glaucoma surgery but we now consider filtrating surgery as safe given the apparently high rate of tumour control after PBR. The only patient who developed metastasis did so 9 months after filtrating surgery, the relationship of these two events remaining unclear. Radiotherapy of the almost entire ciliary body did not lead to hypotony in any case. The absence of radiation retinopathy and papillopathy is in accordance with the calculated dose.
Kaplan–Meier estimate of visual loss in patients presenting with or without glaucoma. PBR, proton beam radiotherapy.
Artificial tears were recommended in all patients during and after irradiation. Iritis was generally not present at the first follow-up examination after 3–6 months. Significant epitheliopathy was uncommon with signs of limbal stem cell insufficiency in one patient and transient corneal erosion during follow-up in two patients. This seems to be in contrast to the experience at the Jules-Gonin Hospital in Lausanne, where limbal stem cells are stored for reimplantation after irradiation with 4×15 CGE PBR to anticipate corneal decompensation.6 To our knowledge, results with regard to tumour control and visual function have not been published by this group.
A limited number of reports describe results from radiotherapeutic treatment of iris melanoma, where PBR proved to be an effective treatment of iris melanoma in some retrospective studies.7–9 Recurrence was reported in 0/21 at median 2.7 years,8 1/15 at median 3.3 years9 and 3/88 at median 2.7 years.7 Recurrence in these series most likely arose from clinically underestimated seeding.
Our rate of metastasis (1/54 or 1.9%) appears to be lower than that reported in the pivotal paper (13%) by Demirci et al.2 In a more recent paper (Konstantinidis et al3), no case of metastatic disease was noted in a somewhat comparable group of 13 patients at a median follow of 3.5 years. Prospective data on the metastatic potential of these higher-risk iris melanomas including larger patient samples are still awaited.
Although not exactly stated in these reports, in most cases <100% of the anterior chamber was treated in contrast to our study. Konstantinidis et al3 reported about 12 patients who received whole anterior segment PBR with a median follow-up of 3.5 years. In their series, 58.4 CGE (53.1 Gy) were given in four fractions and no local tumour progression was noted with 100% eye retention. Confirming biopsy was done in two patients and medical glaucoma treatment in 10. Two patients (17%) developed no light perception after treatment.
Shields et al10 reported results of custom-designed iodine plaque treatment for 38 non-resectable iris melanomas in 2003. The median plaque coverage of the cornea was 45%. In their series, 5-year Kaplan–Meier estimates for tumour recurrence and eye loss were 8% and 13%, respectively. Visual acuity in 16% was ≤20/200, 8% had rubeotic glaucoma and 33% had any type of glaucoma.
Petousis et al11 realised anterior segment irradiation for iris melanoma with a palladium plaque and a favourable course during 1-year follow-up.
Iodine-125 plaque treatment delivers a higher dose to the cornea, which is even higher in palladium-103 and ruthenium-106 brachytherapy.11 ,12 Brachytherapy generally exposes the patient to surgery and in-patient treatment, which is not needed in PBR. To our knowledge, there are no published case series on whole anterior segment brachytherapy in iris melanoma and, therefore, results are hardly comparable.
The retrospective design and the limited follow-up are shortcomings of this study. Endothelial cell count and quality of life-assessment were not performed either. Diagnosis was made clinically in the majority of cases, but it was confirmed in all cases but one (melanocytoma) that underwent biopsy at postirradiation cataract removal, and tumour regression occurred in most nodular tumour components.
We are aware of the fact that the approach of whole anterior segment irradiation in our cohort likely increases radiation-related morbidity but it was effectively aimed at maximising tumour control. Recurrence in posterior uveal melanoma can often be managed by additional focal destructive measures but recurrence in iris melanoma is often diffuse and enucleation was the management in all recurrent iris melanomas after irradiation treatment in several series.7 ,9 ,10 ,12 ,13 Most patients will need glaucoma treatment during follow-up and this often includes surgical therapy.
In conclusion, whole anterior chamber irradiation with a proton dose of 4×12.5 CGE is fairly well tolerated by most of the eyes. The functional results rely on effective control of glaucoma, which can be challenging. To further characterise the limitations of this approach, patients should be followed under a prospective regime.
Acknowledgments
The authors thank Stefan Dahm, Center for Cancer Registry Data, Robert Koch Institute, Berlin, Germany, for statistical support.
References
Footnotes
GDW and DC contributed equally.
Meeting presentation: The material has been presented in part at the meeting of the Deutsche Ophthalmologische Gesellschaft in Berlin, September 2012.
-
Contributors Design and conduct of the study: GDW, CH, DC, BK, MHF, NEB and AMJ; collection, management, analysis and interpretation of the data: GDW, CH, DC, BK, MHF, NEB, NL, LM, JH and AMJ; preparation, review or approval of the manuscript: GDW, CH, DC, MHF, NEB, NL, LM, JH and AMJ. All authors have contributed significantly to the conception and design, collection of data, as well as to the analysis and interpretation of data. All authors have drafted the article critically and have approved the version to be published.
-
Competing interests None.
-
Patient consent Obtained.
Ethics approval
-
Provenance and peer review Not commissioned; externally peer reviewed.
Linked Articles
- At a glance