Fibrovascularization of Porous Polyethylene Orbital Floor Implants in Humans

Porous polyethylene (Medpor; Porex Surgical Inc, Newnan, Ga) orbital implants are increasingly popular and used commonly as sheets, blocks, or spheres for volume replacement in the anophthalmic socket and for orbital wall repair in orbital wall fractures. Synthetic orbital implants are generally less expensive than natural coral implants and are also biocompatible and nontoxic, with interconnecting pores and channels.^1-3

Fibrovascular tissue growth from adjacent orbital tissue into spherical porous polyethylene orbital implants is well established and has been demonstrated^2,4 using several techniques (histopathologic findings, ⁵ technetium isotope scanning, ^6,7 computed tomography, ⁸ and magnetic resonance imaging^1,7,8).Vascularization usually occurs from the periphery of the implant toward the center of the sphere and aids integration of the implant into host tissues. This is believed to reduce infection, extrusion, and exposure of the implant.

Porous polyethylene sheets or blocks are increasingly being used as orbital floor implants (especially in augmenting orbital volume postenucleation and in reconstructive surgery following orbital trauma^3,9)with studies demonstrating fibrovascular tissue in-growth into Medpor orbital floor implants in an animal model.¹⁰ We demonstrate histologic evidence of fibrovascular tissue growth into the anterior aspect of porous polyethylene orbital floor implants in humans in 3 cases in which the implant required trimming.

A 36-year-old man had undergone right enucleation of a blind eye as a result of trauma, 10 years after injury. A secondary spherical orbital implant(18-mm porous polyethylene) was placed 8 years later. One year later, the orbit required further volume augmentation. Seven layers of stacked 0.85-mm-thick porous polyethylene sheets (total height, 5.95 mm) were inserted into the subperiosteal space on the orbital floor via a conjunctival incision. The surgery was performed by an oculoplastic trainee, and the periosteal closure to the arcus marginalis may have been incomplete, as 5 months later the patient had a prominent bump from the anterior edge of the stack, giving an irregular contour to the lower eyelid. This remained stable, and 1 year after insertion, the orbital floor implant was trimmed via a conjunctival approach for aesthetic reasons. The anterior few millimeters of the implant was excised by cutting down with a 15 Barde Parker blade.

A 56-year-old man had a blunt injury resulting in a left orbital floor fracture, diplopia, and enophthalmos. Maxillofacial surgeons repaired the fracture by inserting a porous polyethylene block into the subperiosteal space via a conjunctival incision. His diplopia and enophthalmos improved. However, postoperatively he had lower eyelid retraction, and anterior displacement of the floor implant was noted 6 months after surgery, resulting in an irregular inferior orbital margin contour. Two years after insertion, the anterior protruding edge was excised via a conjunctival incision, and the adhesions causing the eyelid retraction were simultaneously freed.

A 35-year-old man was shot in his right orbit. He had inferior and medial orbital wall fractures, and his eye was enucleated with a primary spherical orbital implant placed without repair of the fractures. Two years later, he was referred to the Oculoplastic and Orbital Service (The Western Eye Hospital, London, England) for treatment of his volume-deficient socket. Stacked porous polyethylene sheets were used to repair the floor defect and augment the orbital volume. Eight 1.5-mm-thick sheets (total height, 12 mm) were inserted via a conjunctival incision by an oculoplastic trainee. Periosteal closure to the arcus marginalis was difficult due to previous eyelid trauma, and the stack displaced anteriorly within 1 week of surgery. Five months later, the anterior edge of the stack was trimmed via a conjunctival incision.

Specimens of the anterior trimmed edges of the orbital floor implants from all 3 cases underwent histopathologic analysis.

The histopathologic findings in the 3 cases were similar and showed particles of refractile foreign material (porous polyethylene) when viewed with polarized light (Figure 1).Throughout all 3 specimens, there was fibrovascular tissue between the particles of foreign material, within the pores of the polyethylene implant (Figure 2). The vessels were predominantly lined by endothelial cells only (lymphovascular channels); a few, however, contained smooth muscle in their walls (consistent with small venules and arterioles) (Figure 3). There was a mild chronic inflammatory foreign body giant cell response, more marked in the specimen removed only 4 months after insertion.

Porous polyethylene orbital implants are commonly used as spheres for primary and secondary orbital implants following evisceration or enucleation. They are also used as orbital floor implants to correct postenucleation volume deficiency and for orbital wall reconstruction. Porous polyethylene floor implants are inserted into the subperiosteal space via a transconjunctival or subciliary skin approach. A single sheet is placed on the floor (0.4 mm, 0.85 mm, or 1.5 mm thick) where there is a fine fracture without significant orbital volume change. Several stacked sheets or a carved block are placed on the floor to augment volume for the correction of enophthalmos following orbital floor fracture and as secondary volume augmentation after spherical orbital implant. Usually, the orbital septum is resutured to the arcus marginalis at the orbital rim, and the need for a channeled implant or fixation with a plate and screws is reserved for large fractures without adequate bony support. The perceived advantage of porous polyethylene over nonautogenous materials(silicone block, bioactive glass, lactic acid polymer, and polydioxanon) is that it is nonabsorbable and has the potential for fibrovascular integration and hence adds volume and has long-term stability. Autogenous calvarial bone grafts do integrate but have variable resorption and involve a second surgical site.

There have been several reports^2-8 describing fibrovascular tissue growth into the pores of spherical polyethylene implants and an animal study¹⁰ of fibrovascular ingrowth into orbital floor implants for fractures. To date, this has not been confirmed in humans. The present small series reports fibrovascular tissue growth into pores of porous polyethylene orbital floor implants in humans—2 following fracture and 1 for postenucleation socket syndrome.

The anterior edge of the orbital floor implants was trimmed between 5 and 24 months after insertion. Anterior displacement had occurred earlier, between 1 week and 6 months after insertion, but had then remained in a stable position (Table 1). It is likely that fibrous integration of floor implants has commenced by 5 to 6 months, as occurs for spherical porous polyethylene implants. Until integration occurs, if there is inadequate anterior support from the orbital septum, these implants can displace.

These 3 cases confirm that even in the relatively avascular subperiosteal space, porous polyethylene will become vascularized. This is theoretically more likely following fracture in which sinus mucosa may be in contact with the implant but less likely when the implant is just slipped subperiosteally to augment orbital volume with intact bony walls. Two of the 3 cases had old orbital fractures. The histopathologic findings confirmed vascularization of the anterior 5 to 7 mm of implant but did not analyze deeper vascularization since only the anterior protruding edge was trimmed. The anterior edge may be more likely to vascularize because of its proximity to the eyelid soft tissues. The height of the sheet stacks in cases 1 and 3 was approximately 6 mm and 12 mm and was equally vascularized throughout the height. All cases had associated giant cell inflammation.⁵

Our study used histologic analysis of the anterior edge of Medpor porous polyethylene orbital floor implants to assess fibrovascular tissue growth into the pores of the implant. In all 3 cases, there was evidence of fibrovascular tissue in the pores of the sheet implants, consistent with that of spherical orbital implants despite the difference in shape and location of the synthetic implants.

This study is the first histologic description to our knowledge in humans of fibrous ingrowth of porous polyethylene implants in this location. We show integration of the synthetic implant with host tissue when the polyethylene implant is placed in contact with the floor of the orbit for augmenting volume and improving cosmesis in the enophthalmic socket. It remains unclear whether the deeper portion of the orbital floor implants are also histologically vascularized. Only removing the entire anteroposterior length of the implant for histopathologic analysis would confirm this.

Corresponding author: Jane M. Olver, FRSC, FRCOphth, Oculoplastic and Orbital Service, The Western Eye Hospital, Marylebone Road, London NW1 5YE, England (e-mail: janeolver@aol.com).