Subconjunctival migration of silicone oil through a Baerveldt pars plana glaucoma implant

Extravasation of intraocular silicone oil through a sclerotomy into the subconjunctival space has been described.¹ Oil migration through Molteno and Ahmed implants has also been reported in the literature.^2,3 However, literature search revealed no cases of oil migration through a Baerveldt pars plana implant after a vitrectomy. We report such a case.

CASE REPORT

In May 2001, a 58 year old white man presented with dense vitreous haemorrhage, hyphaema, and neovascular glaucoma (intraocular pressure (IOP) was 55 mm Hg by applanation tonometry) associated with proliferative diabetic retinopathy in his phakic right eye (RE), and background diabetic retinopathy in his left eye (LE). Visual acuity (VA) was RE: counting fingers, LE: 20/20. There were mild cortical and nuclear sclerotic cataracts in both eyes. The high IOP along with severe headache and nausea was refractory to maximal medical therapy. A pars plana vitrectomy, endophotocoagulation, and placement of a Baerveldt pars plana glaucoma implant (Model BG-102–350, surface area: 350 mm² Pfizer Inc, New York, NY, USA) in the superior temporal quadrant allowed an immediate relief of severe ocular and systemic discomfort, and normalisation of the IOP to 14 mm Hg (applanation), RE, after surgery. Owing to the increased cataract, his general ophthalmologist performed cataract extraction by phacoemulsification and inserted a posterior chamber implant, RE, in October 2001. Subsequently, intraocular silicone oil tamponade for multiple retinal breaks including a large inferior relaxing retinotomy was required during a repeat vitrectomy for proliferative vitreoretinopathy, RE, in January 2002. Examination in May 2002 revealed disappearance of 50% of intravitreal oil because of its migration into the superior subconjunctival space via the Baerveldt shunt (fig 1). Patient complained of increasing ocular discomfort as a result of conjunctival inflammation and IOP rise to 30 mm Hg by applanation tonometry, associated with an enlarging superior conjunctival bleb with underlying infiltration of emulsified oil in the subsequent weeks. Application of dorzolamide hydrochloride-timolol maleate and brimonidine tartrate 0.2% ophthalmic solutions lowered the IOP to 18 mm Hg, RE. Removal of intraocular and subconjunctival silicone oil was performed on 28 May 2002. Surgical exploration showed widespread oil infiltration involving the posterior plate of the implant and the subconjunctival soft tissues. Extensive resection of swollen subconjunctival tissues infiltrated with oil droplets was performed (fig 2). The surgical dissection involved primarily the anterior subconjunctival tissues associated with most of the oil infiltration, and stayed away from the posterior orbital space where fibrous encapsulation around the implant plate was noted. The Baerveldt implant was not removed. The ocular inflammation subsided and the IOP was brought down to 16 mm Hg (applanation) without ocular hypotensive medications, RE, within 1 week after surgery. Ocular hypotensive medical therapy was no longer required afterwards. The VA was 20/200 and the IOP was 15 mm Hg (applanation) with complete retinal attachment, RE, 6 months later.

Figure 1.

 (A) External photograph of inflamed and elevated superior conjunctival bleb infiltrated with extravasated silicone oil from the vitreous cavity of RE. (B) Fundus photograph shows a high intraocular oil-fluid level as a result of extravasation of more than 50% of oil. Note retina is still completely attached despite loss of a large volume of oil.

Figure 2.

 High power photomicrograph of excised conjunctival tissue infiltrated with silicone oil shows a ring of epithelioid cells (large arrow), foreign body material (small arrows), and giant cell (arrowhead).

COMMENT

In recent years, Baerveldt pars plana glaucoma implants have become increasingly popular for control of refractory glaucoma in eyes with vitreoretinal complications that also require a pars plana vitrectomy.⁴ Frequently, silicone oil tamponade may also be indicated for such eyes. Emulsification of intraocular silicone oil usually takes many months after surgery to develop, the exact timing of which varies and depends on multiple factors, including the purity and viscosity of the oil.⁵ It is interesting that extraocular migration of silicone oil did not occur until 4 months after its placement, coincidental with the start of oil emulsification in this case. Despite the loss of intraocular oil, previous long term retinal tamponade with oil proved sufficient for maintaining retinal attachment after oil removal. The drainage tube was not removed or ligated during the second vitrectomy when silicone oil was inserted to avoid recurrent excessive rise of IOP after surgery in the absence of a patent drainage channel, potentially aggravated by reduced volume of the vitreous cavity for posterior aqueous flow due to the intravitreal silicone oil. Measures that may delay or prevent extraocular oil migration through a drainage tube include placement of the pars plana drainage tube in an inferior quadrant, replacement of the pars plana shunt with another tube shunt inserted into an inferior quadrant of the anterior chamber, and use of highly purified and super-viscous oil with lower tendency for emulsification. In addition, the patient is encouraged to sleep on the side of the drainage tube, since oil may rise from the dependent side and away from the tube. Eventually, emulsified oil droplets may find their way into the drainage tube for extraocular migration. However, this case shows that Baerveldt pars plana implant and silicone oil may coexist for a prolonged period for select cases. Silicone oil extravasation through a glaucoma shunt is not unique for a Baerveldt pars plana implant, but a phenomenon associated with other types of shunt implants as well, as shown by previous case reports.^2,3 To our knowledge, however, this is the first written report of silicone oil migration through the drainage tube of a Baerveldt pars plana implant.