Introduction

Although present brachytherapy is very effective in treating uveal melanoma, it has a marked pejorative effect on vision, especially for tumours near the fovea or optic nerve.1 Techniques to preserve vision are needed. Though endoresection has been used, this technique involves disrupting the retina, which provides a natural barrier to extension of the tumour.2 Recent preclinical studies suggest using intraocular silicone oil prior to iodine 125 (125I) brachytherapy for uveal melanoma attenuates the radiation dose to vital ocular structures, potentially reducing the development of radiation maculopathy and papillopathy.3 The authors report their own preclinical study results followed by clinical use of silicone oil as an adjunct to 125I brachytherapy in three patients with choroidal melanoma.

Materials and methods

For our Monte Carlo simulations using the EGSnrc user-code BrachyDose4 to study the effects of 125I radiation in an ocular model, silicone oil was used as a vitreous substitute and compared with water (Figure 1). A 16-mm COMS plaque fully loaded with Best 2301 125I seeds was positioned nasally on a right eye model. Our primary outcome was the radiation dose ratio (of silicone oil vs water) to ocular structures at various angles (ϕ) subtended from the ocular centre (Figure 2). Angle ϕ is 0 degrees at the anterior axis of the model and ϕ is 90 degrees on the plaque’s central axis near the plaque. The simulation results demonstrated that radiation dose ratios on the retina increased from 0.45 to 0.99 when moving from points diametrically opposed from the plaque (ϕ=−90 degrees) to the plaque’s central axis near the plaque (ϕ=90 degrees) (Figure 3). Of significance, intraocular silicone oil reduced the radiation dose to ocular structures at the anterior–posterior ocular axis to 65% of the dose without silicone oil. (Figure 3).

Figure 1
figure 1

Model used for dose calculations with BrachyDose shown in a plane through the eye’s centre. A 16-mm COMS plaque was fully loaded with Best 2301 125I seeds. Water phantom; silicone oil (100%) sphere to radius 1.08 cm. Water tumour modelled as ‘cylinders’ extending into the silicone oil sphere (height 5 mm from inner sclera).

Full size image
Figure 2
figure 2

Diagram showing the configuration considered for dose calculations and defining the angle ϕ used in some of the figures. Angle ϕ is defined as follows: 90 degrees is on the plaque’s central axis near the plaque; −90 degrees is on the plaque’s central axis at the opposite side of the eye to the plaque. Right eye assumed.

Full size image
Figure 3
figure 3

Ratio of doses at the radius of the retina (1.09 cm): doses for the configuration shown in Figure 1 (silicone oil sphere at eye centre; water elsewhere) to doses for water (no silicone oil). Both simulations include the plaque’s Modulay backing and Silastic insert as well as interseed effects.

Full size image

Silicone oil was then used in three cases of posteriorly located medium or large choroidal melanomas, in which each patient underwent microincision vitrectomy, endolaser, 1000-centistoke silicone oil placement, and 125I plaque with 80–85 Gy. Following determination of proper placement using either diaphanoscopy or indirect ophthalmoscopy, a dummy plaque was sutured in place and ultrasonography was used to determine proper placement. Once it was felt that the plaque would be sutured in proper position, a 23- or 25-gauge vitrectomy was performed. After an air–fluid exchange, the radioactive plaque was then sutured in place and then a silicone–air exchange was performed.

Case reports

Case 1

A 58-year-old woman presented with a 16 × 13 × 9 mm variably pigmented mushroom-shaped choroidal melanoma (T2bN0M0) with borders 0 mm nasal to the disc and 4 mm posterior to the ora in her left eye and vision of 20/40. She underwent 23-gauge vitrectomy, tumour biopsy, endolaser, silicone oil placement, and a 20-mm 125I plaque. Silicone oil was removed 5 months later, and on 2-year follow-up her tumour has decreased in size without any evidence of radiation retinopathy, but her visual acuity has declined to 2/200 due to a dense cataract.

Case 2

A 69-year-old woman had a 16 × 13 × 3 mm posterior multilobed pigmented choroidal melanoma (T2aN0M0) in her right eye with borders 5mm temporal to the fovea and extending to the ora at 9 o’clock, with vision of 20/30. She underwent 23-gauge pars plana vitrectomy with biopsy, endolaser, silicone oil placement, and a 20-mm 125I plaque. Silicone oil was removed 2 months later, and on 22 months’ follow-up her tumour remains regressed without evidence of radiation retinopathy, and most recent visual acuity of 20/40−1.

Case 3

A 58-year-old woman had an 11 × 8 × 2.5 mm oval pigmented choroidal melanoma (T2aN0M0) along the inferior arcade in her right eye that was 1.5 mm inferotemporal to the disc and 0.7 mm from the fovea with associated subretinal fluid. She underwent 23-gauge vitrectomy with biopsy, endolaser, silicone oil placement, and 16 mm 125I plaque placement. After 7 months, she was evaluated for silicone oil removal and cataract surgery, and was found to have an inferior rhegmatogenous retinal detachment with significant anterior vitreous contraction, requiring scleral buckle, vitrectomy, and endolaser. On 10 months’ follow-up she is 20/200+2 with tumour base regression and no radiation retinopathy.

Discussion

To date, none of our three patients have developed radiation retinopathy. Radiation retinopathy typically occurs 18 to 24 months after brachytherapy,5, 6 suggesting a potential protective effect of silicone oil in at least the two cases with longer follow-up. Nevertheless, each of these patients required additional operation for silicone oil removal, experienced cataract progression within, and one patient also developed a retinal detachment.

This study confirms that silicone oil attenuates radiation dose in vitro, and may protect against radiation retinopathy clinically in patients. However, we are concerned about the extensive surgery involved and would recommend limiting its use to only very special cases such as posterior pole melanomas in functionally monocular patients. It is critical to consider the risks of silicone oil use in these cases, including multiple surgeries and cataract progression. To our knowledge, there is no previous clinical data in the literature supporting silicone oil decreasing radiation retinopathy. Further studies are needed to elucidate the precise role of silicone oil as an adjunct to 125I brachytherapy in shielding radiation retinopathy.