- Email: [email protected]
Clinical and Histopathologic Review of 18 Explanted Porous Polyethylene Orbital Implants
Clinical and Histopathologic Review of 18 Explanted Porous Polyethylene Orbital Implants Jean Y. Chuo, MD,1 Peter J. Dolman, MD, FRCSC,1 Tony L. Ng, MD,2 Frank V. Buffam, MD, FRCSC,1 Valerie A. White, MD, FRCPC1,2 Purpose: To review the clinical and histopathologic features of porous polyethylene (PP) orbital implants requiring explantation. Design: Case series. Participants: Eighteen explanted PP orbital implants of 18 patients were studied. Methods: The charts and histopathologic findings were reviewed for all patients requiring explantation of PP orbital implants between 1997 and 2006 by 2 oculoplastic surgeons at the University of British Columbia. Main Outcome Measures: Clinical data obtained included patient demographics, the nature of the primary surgery, and the clinical presentation leading to eventual implant removal. The histopathologic data observed included the presence of anterior exposure, area of fibrovascular ingrowth, type of inflammation, and presence and type of bacterial colonies. Results: Nine (50%) of the 18 patients studied were referred from other surgeons. The balance represented 3.2% of all PP implants placed by the 2 surgeons. The procedures for the primary surgery were 12 enucleations (67%), 5 eviscerations (28%), and 1 secondary implant (5%). Clinical findings included anterior implant exposure and discharge in all cases. Histopathologic analysis was performed in all of the implants and showed less than 50% fibrovascular ingrowth in 16 implants (89%) and predominantly acute or mixed inflammation in 15 (83%). Foreign body giant cells were seen adjacent to the implant material in all cases. Bacterial colonies on gram stain were identified in 12 specimens (67%); overall, gram-positive cocci in clusters or chains were found in 10 implants (56%), and gram-negative bacteria were found in 1 (5.5%). Thirteen patients (72%) lived in locations distant from Vancouver, the surgical center. Conclusions: This article presents the largest review of explanted porous polyethylene orbital spheres. The findings suggest that anterior exposure allows bacterial colonization and the development of a heavy inflammatory infiltrate. Poor tissue ingrowth may limit the penetration of topical or systemic antibiotic therapy, leading to the necessity for explantation. Financial Disclosure(s): The author(s) have no proprietary or commercial interest in any materials discussed in this article. Ophthalmology 2009;116:349 –354 © 2009 by the American Academy of Ophthalmology.
Over the past 2 decades, porous implants have become the most popular choice to fill the anophthalmic socket after removing the globe. Common materials used today include coralline or synthetic hydroxyapatite (HA),1,2 porous polyethylene (PP),3–5 and aluminium oxide. These all share the benefits of allowing tissue ingrowth, reducing the risk of extrusion, and allowing pegging to improve the motility of the artificial eye. Their major drawback is their rough surface, promoting breakdown of the overlying tissue and leading to anterior exposure and possible infection.6 –10 Implant exposure has been reported to occur in 5% to 20 % of cases.1,2,11–14 Although erosions often can be repaired by advancing surrounding conjunctiva over a patch graft, placing a dermal fat or buccal mucosal graft, or advancing a Hughes lid flap, a significant portion may develop recurrent or larger areas of exposure or become infected (Fig 1A), necessitating implant removal (Fig 1B). This study is the largest clinical and histopathologic review of such explanted PP spheres. © 2009 by the American Academy of Ophthalmology Published by Elsevier Inc.
Patients and Methods The charts and pathology slides were reviewed for all patients requiring explantation of PP orbital implants by 2 oculoplastic surgeons (FVB and PJD) from 1997 through 2006. The clinical data obtained included demographics of the patients (age, gender, address), characteristics of the primary surgery (indications for surgery, method of globe removal), and clinical presentation leading to implant removal (symptoms, signs, and the time from implant insertion to removal). All implants were fixed in 10% neutral buffered formalin, processed through alcohols and xylene, and embedded in paraffin wax. Three-micrometer sections were stained with hematoxylin and eosin, gram stain for bacteria, and periodic acid– Schiff stain for fungi. Histopathologic features analyzed included presence of anterior erosion, relative areas of fibrovascular tissue ingrowth, type of inflammation, and the presence of bacteria on gram stain. The research protocol was approved by the University of British Columbia and Vancouver ISSN 0161-6420/09/$–see front matter doi:10.1016/j.ophtha.2008.09.022
349
Ophthalmology Volume 116, Number 2, February 2009
Figure 1. A, Photograph showing an infected orbital implant with anterior exposure. B, Photograph showing an explanted infected orbital implant.
General Hospital Ethics Committees before commencing the study. The surgical technique for enucleation was similar between the 2 surgeons. An 18- to 20-mm diameter PP implant was wrapped in donor sclera, with large windows left in the area of the medial and lateral rectus muscles for vascular ingrowth. Care was taken to seat the implant posteriorly in the orbit behind the posterior Tenon capsule and the recti muscles sutured through the windows onto the undersurface of the sclera in anatomic position. The anterior Tenon capsule was closed with interrupted sutures and the conjunctiva was closed with a locking absorbable suture. In eviscerations, after having removed the cornea and intraocular contents, one surgeon (PJD) routinely injected retrobulbar alcohol (1 ml) through the sclera inferotemporal to the optic nerve head for postoperative pain control.15 Relaxing incisions were made in the sclera to allow posterior placement of the implant and to permit coverage anteriorly by overlapping the sclera in a doublelayered closure: one surgeon (FVB) made radial incisions in the sclera posterior to the equator and around the optic nerve insertion, whereas the other (PJD) divided the scleral envelope into halves by cutting it from the supertemporal limbus to the inferonasal limbus, cutting around the optic nerve head and avoiding the extraocular muscles. The Tenon capsule and conjunctiva were closed in similar fashion to that in enucleations. In all cases, a rigid conformer was placed and an intermarginal suture was placed to close the eyelids and to prevent loss of the conformer or protrusion of edematous conjunctiva. In referred cases where the primary globe removal surgery was performed by another surgeon, the specific techniques of PP implantation were not known. However, it was known that in all but 2 of the referred enucleation cases, the implants were wrapped in sclera or Vicryl mesh (Ethicon Inc, Piscataway, NJ).
age of 46 years. Ten (56%) were male and 8 (44%) were female. Thirteen patients (72%) were living in locations distant from Vancouver, the primary referral center, in or outside the province of British Columbia; the remainder lived locally. Twelve implants were placed after enucleation (67%), 5 (28%) after evisceration, and 1 (5%) as a secondary orbital implant. All implants, other than the 5 placed in the evisceration cavity and 2 in enucleations performed elsewhere, were wrapped in donor sclera or Vicryl mesh. The original reason for removal of the globe was unsalvageable trauma in 8 cases (44%) and complications of intraocular surgery in 6 cases (33%). Significant medical history that may predispose a patient to poor wound healing included diabetes mellitus in 4 patients (22%) and smoking in 4 patients (22%). None of the patients had received previous radiation treatment. The interval from implantation to initial symptoms ranged from 3 to 86 months. Implant exposure was present in all patients. Other symptoms included discharge in 17 (94%), irritation and pain in 13 (72%), and bleeding in 5 (28%). Conjunctival cultures were obtained for 10 cases: of these, 3 grew Streptococcus viridans, 1 grew Staphylococcus aureus, 1 grew Peptostreptococcus, and 5 had no growth. Most of the patients already had been prescribed topical antibiotics at the time they were assessed by the 2 surgeons. All cases subsequently were
Table 1. Clinical Features in 18 Patients Who Required Porous Polyethylene Orbital Implant Removal
Age (yrs) Gender Clinical symptoms
Results A total of 18 explanted PP spheres were reviewed. Of these, 9 (50%) were referred from other comprehensive ophthalmologists. In referred cases, the type and date of primary surgery was identified, but specifics of the surgical technique were unknown. The remaining 9 cases represented 3.2% of all PP implants placed by the 2 surgeons during that interval. Table 1 presents the demographic and clinical features of the 18 patients. They ranged from 23 to 83 years of age, with an average
350
Original surgery
Time from surgery to symptoms (mos) Time from surgery to implant removal (mos)
Feature
Number
Average Range Male Female Discharge Irritation/pain Bleeding Enucleation Evisceration Second implant Average Range Average Range
46 23–83 10 8 17 13 5 12 5 1 35 3–86 40 4–168
Percentage (%)
55 45 94 72 28 67 28 5
Chuo et al 䡠 Review of Explanted Porous Polyethylene Orbital Implants treated with topical and oral antibiotic therapy for a minimum of 7 days. Although imaging studies (such as a gallium scan) of the infected implants would have been interesting to correlate with the histopathologic results, these were not obtained in these cases. All but 2 of the patients had at least 1 (and in several cases, multiple) attempted repairs of the erosions. This was performed by undermining of adjacent conjunctiva and Tenon capsule, advancement and layered closure with absorbable sutures, often overlying a scleral patch graft laid on the exposed implant. One patient had more than 9 mm of exposure and purulent discharge overlying the implant at presentation, and the implant was removed without attempted repair. The other patient lived in a remote community, and when her peg implant became painful, she removed the peg with pliers and then used a power drill to drill the implant to try to drain a possible source of infection. The interval from onset of symptoms to eventual implant removal ranged from 0 to 16 months, averaging 5 months. After being explanted, dermal fat grafts were placed in all but one case, where a polymethyl methacrylate sphere was placed. The dermal fat grafts generally healed well, with no increased incidence of fat atrophy noted at the 6-week follow-up at the time of prosthesis fitting, compared with a primary dermal fat graft. All patients reported immediate relief of pain, discharge, and bleeding. Table 2 displays the histopathologic findings for the 18 explanted PP spheres. All implants contained inflammatory infiltrates (Fig 2A); predominantly acute or mixed inflammation was found in 15 (83%) implants and mainly chronic inflammation was found in the remaining 3 implants. All implants had foreign body giant cells scattered around the PP trabeculae. Sixteen implants (89%) contained less than 50% fibrovascular ingrowth (Fig 2B). Gram-positive cocci in clusters or chains (Fig 2C) were identified in 10 implants (56%), gram-positive cocci and rods were identified in 1 implant (5.5%), and gram-negative bacteria were identified in 1 implant (5.5%).
Discussion Porous orbital implants have replaced solid spheres as the implant of choice for many ophthalmic surgeons because of their biocompatibility16 or because patients specifically request them. Tissue ingrowth through the pores allows biointegration, reducing the risk of extrusion and infec-
Table 2. Histologic Findings in 18 Porous Polyethylene Explanted Orbital Implants
Fibrovascular ingrowth
Inflammation Infection
Features
Number
Percentage (%)
ⱕ25% 26%–50% 51%–75% ⬎75% Mainly acute or mixed Mainly chronic Gram-positive cocci in clusters Gram-positive cocci in chains Gram-positive cocci in chains and clusters Gram-positive cocci and rods Gram-negative
7 9 1 1 15 3 5
39 50 5.5 5.5 83 17 28
2
11
3
17
1
5.5
1
5.5
tion.17 The use of peg coupling systems may enhance fine movements of the artificial eye. Although rare, infection of orbital implants tends to be difficult to control without implant removal. This review suggests that anterior exposure of implants allows bacterial entry and subsequent colonization. Although antibiotic therapy via topical, oral, intravenous, or direct injection is recommended before attempting surgical repair of the defect, this study found that poor tissue ingrowth in the implant may limit the penetration of antibiotic therapy. A recent study has confirmed that topical application of methylene blue does not penetrate nonvascularized PP spheres.18 Reports of infected porous implants are uncommon because the frequency of infection is low, ranging from 0% to 4.87% in different series with varying follow-up periods.17 Implant exposure is reported more commonly, with a rate of 5% to 53 % among various porous implants.19 In a recent review, Custer and Trinkaus19 found that implant removal was necessary in 29% of exposures, either secondary to chronic recurrent exposure despite surgical intervention or because of presumed implant infection. Implant exposure was found in 9 of 13 cases of definite, suspected, and probable implant infection in a series of removed implants by Jordan et al.20 In this series, implant exposure was present in all cases. It is likely that the exposures provided a route for bacterial penetration into the implant with subsequent colonization. It is generally accepted that porous implants are at the highest risk for infection before complete vascularization. At 4 weeks after implantation, histopathologic examination of a sclera-wrapped human HA implant by Shields et al14 showed fibrovascular tissue infiltrating its pores with the greatest tissue ingrowth near the region of the scleral windows, where extraocular muscles interface with the implant. Fibrovascular tissue was shown by Klapper et al21 to have progressed into the peripheral outer two thirds, but not the central third, of the implant at 3 months after implantation. A mild chronic inflammatory cellular infiltrate was noted with multinucleated giant cells. By 7 to 8 months, complete fibrovascular ingrowth into the center of the implant was observed by Nunery et al.13 Porous polyethylene implants have become more popular because of their low cost, ease of use, and presumed similar complication profile to HA. However, unlike HA implants, the rate of fibrovascular ingrowth into polyethylene implants have not been well documented and comparatively few studies have been published on their exposure and complication rates. Li et al22 found the exposure rate of polyethylene implants to be 9% in their retrospective series of 44 polyethylene and solid acrylic implants, which was comparable with published rates for HA implants. In a meta-analysis of studies on porous orbital implants, Custer and Trinkaus18 concluded that HA implants had a lower reported exposure rate (4.9%) than PP implants (8.1%), primarily related to the practice of inserting PP implants without any covering material, particularly in patients with retinoblastoma. The exposure rates were similar between the 2 implants (HA, 5.1%; PP, 4.2%) when the known patients with retinoblastoma were omitted from the pooled data. In this series, for all of the primary cases (9 of 18)
351
Ophthalmology Volume 116, Number 2, February 2009
Figure 2. A, Photomicrograph from central portion of implant showing heavy acute inflammation with poor fibrovascular ingrowth (stain, hematoxylin– eosin; original magnification, ⫻100). B, Low-power photomicrograph of an explanted orbital implant showing surrounding sclera with anterior erosion. The central light-colored portion consists of acute and chronic inflammatory cells (stain, hematoxylin– eosin; original magnification, ⫻1). C, High-power photomicrograph of gram stain showing numerous gram-positive cocci in clusters and chains (stain, gram; original magnification, ⫻1000).
352
performed by FVB or PJD, the implants were wrapped in donor sclera and none of the 18 cases previously had undergone irradiation. Despite the histopathologic evidence of fibrovascular ingrowth, concern exists regarding the effect of porous implants in the anophthalmic socket, particularly in the settings of implant exposure or infection. In an animal model, Reznick et al23 found that the HA implant interferes with the host’s ability to resolve bacterial infection in the underlying bone. Histopathologic studies of clinically exposed porous implants show a serosanguineous discharge with a profuse inflammatory infiltrate,6,13 a setting that is unfavorable for fibrovascular ingrowth. The lack of fibrovascular ingrowth results in avascularity in the exposed portion of the implant,6,13,24,25 which, in combination with a direct route to a nonsterile environment, leads to bacterial infection and colonization of the implant that often requires explantation. In this series, implant exposure was found in all cases. Inflammatory infiltration also was present in all implants, with predominantly acute or mixed inflammation in 15 implants and mainly chronic inflammation in the remaining 3 implants. Sixteen (89%) of the 18 implants had less than 50% fibrovascular ingrowth on histologic examination, consistent with the proposed theory that subsequent avascularity promotes exposure and resultant infection.6,13,24,25 Although erosions were believed to have resulted mainly from mechanical pressure between the porous implant material and the overlying prosthesis, poor fibrovascular ingrowth may have contributed to poor wound healing and a greater chance of implant exposure. Careful surgical techniques at the time of globe removal may help reduce the risk of exposure. Seating the implant deep in the socket is aided in enucleations by using a smooth surface wrap around the rough implant to prevent snagging of anterior tissues as it is pushed into the socket. In eviscerations, relaxing incisions in the posterior sclera to allow deep placement of the implant reduces pressure on the anterior scleral wound and also permits a double-layered closure of the sclera. Careful layered closure of the Tenon capsule and conjunctiva also is important. Both oculoplastic surgeons in this study routinely performed all these maneuvers. Fibrovascular ingrowth may be promoted in enucleations with large windows in the wrap material, increasing contact between the extraocular muscles and the porous material, and possibly by suturing the extraocular muscles more anteriorly than the traditional anatomic position. The latter technique also may provide additional cushioning between the delicate mucosal soft tissues and the rough implant material. Soaking the implants in antibiotics has been recommended to avoid possible introduction of bacteria at the time of implantation. Soaking PP implants in antibiotics does not result in penetration to the central core. A recent study has shown that the best technique to bring solution into the empty spaces of PP implants at the time of primary placement is to compress the solution into the implant within a large syringe.18 Although only 5 of the conjunctival bacterial cultures yielded positive results (perhaps because patients had al-
Chuo et al 䡠 Review of Explanted Porous Polyethylene Orbital Implants ready been treated with topical antibiotics), bacteria were identified on gram stain in 12 implants (67%). Bacterial clusters were situated primarily in the center of the implant in areas of microabscess formation, which also were the center of the avascular zone, suggesting that antibiotic penetration into these areas would be minimal. Seventy-two percent of the implants were removed from patients who lived in a distant location, despite the fact that local residents make up most of the cases requiring globe removal. The authors postulate that the distance made it difficult for these patients to seek medical attention and follow-up care in a timely fashion, resulting in long-standing implant exposures that subsequently led to established infections that were difficult to control without implant removal. It is also possible that ill-fitting prostheses may have been detected or treated less easily in remote communities. The authors recommend that patients be informed about the symptoms and signs of implant exposure and be encouraged to contact their local ophthalmologist or primary surgeon so that prompt medical intervention or surgical repair can be instituted. In summary, this is the largest review of explanted PP orbital spheres. With the results of this study, the authors postulate that implant exposure allows bacteria to penetrate into the center of the implant, which has limited fibrovascular ingrowth and acts as an immune sanctuary for infection. Techniques to limit implant exposure may include seating the porous implant deep in the socket, dividing the scleral envelope posteriorly in eviscerations, using an anterior covering to limit compression of the mucosal tissues between the implant and the prosthesis, advancing the extraocular muscles anteriorly to provide better anterior vascular coverage, and increasing the chance of fibrovascular ingrowth by increasing the surface area contact between the extraocular muscles and the implant through large scleral windows. The authors believe that patients, especially those living in remote communities, need to be informed that they must contact the surgeon as soon as possible if exposures form. Exposures need to be treated aggressively and promptly with cultures, systemic antibiotics, and surgical repair, especially in cases of enucleation and PP spheres. After symptoms of infection develop, one may try antibiotic injection into the implant,18 but implant removal likely will be required.
References 1. Massry GG, Holds JB. Coralline hydroxyapatite spheres as secondary orbital implants in anophthalmos. Ophthalmology 1995;102:161– 6. 2. Dutton JJ. Coralline hydroxyapatite as an ocular implant. Ophthalmology 1991;98:370 –7. 3. Berghaus A. Porous polyethylene in reconstructive head and neck surgery. Arch Otolaryngol 1985;111:154 – 60. 4. Blaydon SM, Shepler TR, Neuhaus RW, et al. The porous polyethylene (Medpor) spherical orbital implant: a retrospective study of 136 cases. Ophthal Plast Reconstr Surg 2003;19:364 –71. 5. Karesh JW, Dresner SC. High-density porous polyethylene (Medpor) as a successful anophthalmic socket implant. Ophthalmology 1994;101:1688 –95; discussion 1695– 6.
6. Remulla HD, Rubin PA, Shore JW, et al. Complications of porous spherical orbital implants. Ophthalmology 1995;102: 586 –93. 7. Jordan DR, Brownstein S, Jolly SS. Abscessed hydroxyapatite orbital implants: a report of two cases. Ophthalmology 1996; 103:1784 –7. 8. Glasgow BJ, Weinberg DA, Shorr N, Goldberg RA. Draining cutaneous fistula associated with infection of hydroxyapatite orbital implant. Ophthal Plast Reconstr Surg 1996;12:131–5. 9. Wilson MW, Wobig JL, Dailey RA. Infection of a porous polyethylene orbital implant with Capnocytophaga. Ophthal Plast Reconstr Surg 1998;14:398 – 402. 10. Christmas NJ, Gordon CD, Murray TG, et al. Intraorbital implants after enucleation and their complications: a 10-year review. Arch Ophthalmol 1998;116:1199 –203. 11. Jordan DR. Problems after evisceration surgery with porous orbital implants: experience with 86 patients. Ophthal Plast Reconstr Surg 2004;20:374 – 80. 12. Cheng MS, Liao SL, Lin LL. Late porous polyethylene implant exposure after motility coupling post placement. Am J Ophthalmol 2004;138:420 – 4. 13. Nunery WR, Heinz GW, Bonnin JM, et al. Exposure rate of hydroxyapatite spheres in the anophthalmic socket: histopathologic correlation and comparison with silicone sphere implants. Ophthal Plast Reconstr Surg 1993;9:96 –104. 14. Shields CL, Shields JA, De Potter P, Singh AD. Lack of complications of the hydroxyapatite orbital implant in 250 consecutive cases. Trans Am Ophthalmol Soc 1993;91:177– 89; discussion 189 –95. 15. Giligson A, Dolman PJ, Buffam F. Comparison of retrobulbar analgesics for evisceration. Ophthal Plast Reconstr Surg 2002; 18:258 – 60. 16. Custer PL. Enucleation: past, present, and future. Ophthal Plast Reconstr Surg 2000;16:316 –21. 17. Karsloglu S, Serin D, Simsek I, Ziylan S. Implant infection in porous orbital implants. Ophthal Plast Reconstr Surg 2006;22: 461– 6. 18. Badilla J, Dolman PJ. Methods of antibiotic instillation in porous orbital implants. Ophthal Plast Reconst Surg 2008;24: 287–9. 19. Custer PL, Trinkaus KM. Porous implant exposure: incidence, management, and morbidity. Ophthal Plast Reconstr Surg 2007;23:1–7. 20. Jordan DR, Brownstein S, Faraji H. Clinicopathologic analysis of 15 explanted hydroxyapatite implants. Ophthal Plast Reconstr Surg 2004;20:285–90. 21. Klapper SR, Jordan DR, Brownstein S, Punja K. Incomplete fibrovascularization of a hydroxyapatite orbital implant 3 months after implantation. Arch Ophthalmol 1999; 117:1088 –9. 22. Li T, Shen J, Duffy MT. Exposure rates of wrapped and unwrapped orbital implants following enucleation. Ophthal Plast Reconstr Surg 2001;17:431–5. 23. Reznick JB, Gilmore WC. Host response to infection of a subperiosteal hydroxylapatite implant. Oral Surg Oral Med Oral Pathol 1989;67:665–72. 24. Buettner H, Bartley GB. Tissue breakdown and exposure associated with orbital hydroxyapatite implants. Am J Ophthalmol 1992;113:669 –73. 25. Ainbinder DJ, Haik BG, Tellado M. Hydroxyapatite orbital implant abscess: histopathologic correlation of an infected implant following evisceration. Ophthal Plast Reconstr Surg 1994;10:267–70.
353
Ophthalmology Volume 116, Number 2, February 2009
Footnotes and Financial Disclosures Originally received: January 27, 2008. Final revision: August 29, 2008. Accepted: September 12, 2008. Available online: December 16, 2008. 1
2
Manuscript no. 2008-135.
Departments of Ophthalmology and Visual Sciences, Vancouver General Hospital and University of British Columbia, Vancouver, Canada.
354
Department of Pathology, Vancouver General Hospital and University of British Columbia, Vancouver, Canada. Financial Disclosure(s): The author(s) have no proprietary or commercial interest in any materials discussed in this article. Correspondence: Peter J. Dolman, Eye Care Centre, 2550 Willow Street, Section I, Vancouver, British Columbia, Canada V5Z 3N9.
Over the past 2 decades, porous implants have become the most popular choice to fill the anophthalmic socket after removing the globe. Common materials used today include coralline or synthetic hydroxyapatite (HA),1,2 porous polyethylene (PP),3–5 and aluminium oxide. These all share the benefits of allowing tissue ingrowth, reducing the risk of extrusion, and allowing pegging to improve the motility of the artificial eye. Their major drawback is their rough surface, promoting breakdown of the overlying tissue and leading to anterior exposure and possible infection.6 –10 Implant exposure has been reported to occur in 5% to 20 % of cases.1,2,11–14 Although erosions often can be repaired by advancing surrounding conjunctiva over a patch graft, placing a dermal fat or buccal mucosal graft, or advancing a Hughes lid flap, a significant portion may develop recurrent or larger areas of exposure or become infected (Fig 1A), necessitating implant removal (Fig 1B). This study is the largest clinical and histopathologic review of such explanted PP spheres. © 2009 by the American Academy of Ophthalmology Published by Elsevier Inc.
Patients and Methods The charts and pathology slides were reviewed for all patients requiring explantation of PP orbital implants by 2 oculoplastic surgeons (FVB and PJD) from 1997 through 2006. The clinical data obtained included demographics of the patients (age, gender, address), characteristics of the primary surgery (indications for surgery, method of globe removal), and clinical presentation leading to implant removal (symptoms, signs, and the time from implant insertion to removal). All implants were fixed in 10% neutral buffered formalin, processed through alcohols and xylene, and embedded in paraffin wax. Three-micrometer sections were stained with hematoxylin and eosin, gram stain for bacteria, and periodic acid– Schiff stain for fungi. Histopathologic features analyzed included presence of anterior erosion, relative areas of fibrovascular tissue ingrowth, type of inflammation, and the presence of bacteria on gram stain. The research protocol was approved by the University of British Columbia and Vancouver ISSN 0161-6420/09/$–see front matter doi:10.1016/j.ophtha.2008.09.022
349
Ophthalmology Volume 116, Number 2, February 2009
Figure 1. A, Photograph showing an infected orbital implant with anterior exposure. B, Photograph showing an explanted infected orbital implant.
General Hospital Ethics Committees before commencing the study. The surgical technique for enucleation was similar between the 2 surgeons. An 18- to 20-mm diameter PP implant was wrapped in donor sclera, with large windows left in the area of the medial and lateral rectus muscles for vascular ingrowth. Care was taken to seat the implant posteriorly in the orbit behind the posterior Tenon capsule and the recti muscles sutured through the windows onto the undersurface of the sclera in anatomic position. The anterior Tenon capsule was closed with interrupted sutures and the conjunctiva was closed with a locking absorbable suture. In eviscerations, after having removed the cornea and intraocular contents, one surgeon (PJD) routinely injected retrobulbar alcohol (1 ml) through the sclera inferotemporal to the optic nerve head for postoperative pain control.15 Relaxing incisions were made in the sclera to allow posterior placement of the implant and to permit coverage anteriorly by overlapping the sclera in a doublelayered closure: one surgeon (FVB) made radial incisions in the sclera posterior to the equator and around the optic nerve insertion, whereas the other (PJD) divided the scleral envelope into halves by cutting it from the supertemporal limbus to the inferonasal limbus, cutting around the optic nerve head and avoiding the extraocular muscles. The Tenon capsule and conjunctiva were closed in similar fashion to that in enucleations. In all cases, a rigid conformer was placed and an intermarginal suture was placed to close the eyelids and to prevent loss of the conformer or protrusion of edematous conjunctiva. In referred cases where the primary globe removal surgery was performed by another surgeon, the specific techniques of PP implantation were not known. However, it was known that in all but 2 of the referred enucleation cases, the implants were wrapped in sclera or Vicryl mesh (Ethicon Inc, Piscataway, NJ).
age of 46 years. Ten (56%) were male and 8 (44%) were female. Thirteen patients (72%) were living in locations distant from Vancouver, the primary referral center, in or outside the province of British Columbia; the remainder lived locally. Twelve implants were placed after enucleation (67%), 5 (28%) after evisceration, and 1 (5%) as a secondary orbital implant. All implants, other than the 5 placed in the evisceration cavity and 2 in enucleations performed elsewhere, were wrapped in donor sclera or Vicryl mesh. The original reason for removal of the globe was unsalvageable trauma in 8 cases (44%) and complications of intraocular surgery in 6 cases (33%). Significant medical history that may predispose a patient to poor wound healing included diabetes mellitus in 4 patients (22%) and smoking in 4 patients (22%). None of the patients had received previous radiation treatment. The interval from implantation to initial symptoms ranged from 3 to 86 months. Implant exposure was present in all patients. Other symptoms included discharge in 17 (94%), irritation and pain in 13 (72%), and bleeding in 5 (28%). Conjunctival cultures were obtained for 10 cases: of these, 3 grew Streptococcus viridans, 1 grew Staphylococcus aureus, 1 grew Peptostreptococcus, and 5 had no growth. Most of the patients already had been prescribed topical antibiotics at the time they were assessed by the 2 surgeons. All cases subsequently were
Table 1. Clinical Features in 18 Patients Who Required Porous Polyethylene Orbital Implant Removal
Age (yrs) Gender Clinical symptoms
Results A total of 18 explanted PP spheres were reviewed. Of these, 9 (50%) were referred from other comprehensive ophthalmologists. In referred cases, the type and date of primary surgery was identified, but specifics of the surgical technique were unknown. The remaining 9 cases represented 3.2% of all PP implants placed by the 2 surgeons during that interval. Table 1 presents the demographic and clinical features of the 18 patients. They ranged from 23 to 83 years of age, with an average
350
Original surgery
Time from surgery to symptoms (mos) Time from surgery to implant removal (mos)
Feature
Number
Average Range Male Female Discharge Irritation/pain Bleeding Enucleation Evisceration Second implant Average Range Average Range
46 23–83 10 8 17 13 5 12 5 1 35 3–86 40 4–168
Percentage (%)
55 45 94 72 28 67 28 5
Chuo et al 䡠 Review of Explanted Porous Polyethylene Orbital Implants treated with topical and oral antibiotic therapy for a minimum of 7 days. Although imaging studies (such as a gallium scan) of the infected implants would have been interesting to correlate with the histopathologic results, these were not obtained in these cases. All but 2 of the patients had at least 1 (and in several cases, multiple) attempted repairs of the erosions. This was performed by undermining of adjacent conjunctiva and Tenon capsule, advancement and layered closure with absorbable sutures, often overlying a scleral patch graft laid on the exposed implant. One patient had more than 9 mm of exposure and purulent discharge overlying the implant at presentation, and the implant was removed without attempted repair. The other patient lived in a remote community, and when her peg implant became painful, she removed the peg with pliers and then used a power drill to drill the implant to try to drain a possible source of infection. The interval from onset of symptoms to eventual implant removal ranged from 0 to 16 months, averaging 5 months. After being explanted, dermal fat grafts were placed in all but one case, where a polymethyl methacrylate sphere was placed. The dermal fat grafts generally healed well, with no increased incidence of fat atrophy noted at the 6-week follow-up at the time of prosthesis fitting, compared with a primary dermal fat graft. All patients reported immediate relief of pain, discharge, and bleeding. Table 2 displays the histopathologic findings for the 18 explanted PP spheres. All implants contained inflammatory infiltrates (Fig 2A); predominantly acute or mixed inflammation was found in 15 (83%) implants and mainly chronic inflammation was found in the remaining 3 implants. All implants had foreign body giant cells scattered around the PP trabeculae. Sixteen implants (89%) contained less than 50% fibrovascular ingrowth (Fig 2B). Gram-positive cocci in clusters or chains (Fig 2C) were identified in 10 implants (56%), gram-positive cocci and rods were identified in 1 implant (5.5%), and gram-negative bacteria were identified in 1 implant (5.5%).
Discussion Porous orbital implants have replaced solid spheres as the implant of choice for many ophthalmic surgeons because of their biocompatibility16 or because patients specifically request them. Tissue ingrowth through the pores allows biointegration, reducing the risk of extrusion and infec-
Table 2. Histologic Findings in 18 Porous Polyethylene Explanted Orbital Implants
Fibrovascular ingrowth
Inflammation Infection
Features
Number
Percentage (%)
ⱕ25% 26%–50% 51%–75% ⬎75% Mainly acute or mixed Mainly chronic Gram-positive cocci in clusters Gram-positive cocci in chains Gram-positive cocci in chains and clusters Gram-positive cocci and rods Gram-negative
7 9 1 1 15 3 5
39 50 5.5 5.5 83 17 28
2
11
3
17
1
5.5
1
5.5
tion.17 The use of peg coupling systems may enhance fine movements of the artificial eye. Although rare, infection of orbital implants tends to be difficult to control without implant removal. This review suggests that anterior exposure of implants allows bacterial entry and subsequent colonization. Although antibiotic therapy via topical, oral, intravenous, or direct injection is recommended before attempting surgical repair of the defect, this study found that poor tissue ingrowth in the implant may limit the penetration of antibiotic therapy. A recent study has confirmed that topical application of methylene blue does not penetrate nonvascularized PP spheres.18 Reports of infected porous implants are uncommon because the frequency of infection is low, ranging from 0% to 4.87% in different series with varying follow-up periods.17 Implant exposure is reported more commonly, with a rate of 5% to 53 % among various porous implants.19 In a recent review, Custer and Trinkaus19 found that implant removal was necessary in 29% of exposures, either secondary to chronic recurrent exposure despite surgical intervention or because of presumed implant infection. Implant exposure was found in 9 of 13 cases of definite, suspected, and probable implant infection in a series of removed implants by Jordan et al.20 In this series, implant exposure was present in all cases. It is likely that the exposures provided a route for bacterial penetration into the implant with subsequent colonization. It is generally accepted that porous implants are at the highest risk for infection before complete vascularization. At 4 weeks after implantation, histopathologic examination of a sclera-wrapped human HA implant by Shields et al14 showed fibrovascular tissue infiltrating its pores with the greatest tissue ingrowth near the region of the scleral windows, where extraocular muscles interface with the implant. Fibrovascular tissue was shown by Klapper et al21 to have progressed into the peripheral outer two thirds, but not the central third, of the implant at 3 months after implantation. A mild chronic inflammatory cellular infiltrate was noted with multinucleated giant cells. By 7 to 8 months, complete fibrovascular ingrowth into the center of the implant was observed by Nunery et al.13 Porous polyethylene implants have become more popular because of their low cost, ease of use, and presumed similar complication profile to HA. However, unlike HA implants, the rate of fibrovascular ingrowth into polyethylene implants have not been well documented and comparatively few studies have been published on their exposure and complication rates. Li et al22 found the exposure rate of polyethylene implants to be 9% in their retrospective series of 44 polyethylene and solid acrylic implants, which was comparable with published rates for HA implants. In a meta-analysis of studies on porous orbital implants, Custer and Trinkaus18 concluded that HA implants had a lower reported exposure rate (4.9%) than PP implants (8.1%), primarily related to the practice of inserting PP implants without any covering material, particularly in patients with retinoblastoma. The exposure rates were similar between the 2 implants (HA, 5.1%; PP, 4.2%) when the known patients with retinoblastoma were omitted from the pooled data. In this series, for all of the primary cases (9 of 18)
351
Ophthalmology Volume 116, Number 2, February 2009
Figure 2. A, Photomicrograph from central portion of implant showing heavy acute inflammation with poor fibrovascular ingrowth (stain, hematoxylin– eosin; original magnification, ⫻100). B, Low-power photomicrograph of an explanted orbital implant showing surrounding sclera with anterior erosion. The central light-colored portion consists of acute and chronic inflammatory cells (stain, hematoxylin– eosin; original magnification, ⫻1). C, High-power photomicrograph of gram stain showing numerous gram-positive cocci in clusters and chains (stain, gram; original magnification, ⫻1000).
352
performed by FVB or PJD, the implants were wrapped in donor sclera and none of the 18 cases previously had undergone irradiation. Despite the histopathologic evidence of fibrovascular ingrowth, concern exists regarding the effect of porous implants in the anophthalmic socket, particularly in the settings of implant exposure or infection. In an animal model, Reznick et al23 found that the HA implant interferes with the host’s ability to resolve bacterial infection in the underlying bone. Histopathologic studies of clinically exposed porous implants show a serosanguineous discharge with a profuse inflammatory infiltrate,6,13 a setting that is unfavorable for fibrovascular ingrowth. The lack of fibrovascular ingrowth results in avascularity in the exposed portion of the implant,6,13,24,25 which, in combination with a direct route to a nonsterile environment, leads to bacterial infection and colonization of the implant that often requires explantation. In this series, implant exposure was found in all cases. Inflammatory infiltration also was present in all implants, with predominantly acute or mixed inflammation in 15 implants and mainly chronic inflammation in the remaining 3 implants. Sixteen (89%) of the 18 implants had less than 50% fibrovascular ingrowth on histologic examination, consistent with the proposed theory that subsequent avascularity promotes exposure and resultant infection.6,13,24,25 Although erosions were believed to have resulted mainly from mechanical pressure between the porous implant material and the overlying prosthesis, poor fibrovascular ingrowth may have contributed to poor wound healing and a greater chance of implant exposure. Careful surgical techniques at the time of globe removal may help reduce the risk of exposure. Seating the implant deep in the socket is aided in enucleations by using a smooth surface wrap around the rough implant to prevent snagging of anterior tissues as it is pushed into the socket. In eviscerations, relaxing incisions in the posterior sclera to allow deep placement of the implant reduces pressure on the anterior scleral wound and also permits a double-layered closure of the sclera. Careful layered closure of the Tenon capsule and conjunctiva also is important. Both oculoplastic surgeons in this study routinely performed all these maneuvers. Fibrovascular ingrowth may be promoted in enucleations with large windows in the wrap material, increasing contact between the extraocular muscles and the porous material, and possibly by suturing the extraocular muscles more anteriorly than the traditional anatomic position. The latter technique also may provide additional cushioning between the delicate mucosal soft tissues and the rough implant material. Soaking the implants in antibiotics has been recommended to avoid possible introduction of bacteria at the time of implantation. Soaking PP implants in antibiotics does not result in penetration to the central core. A recent study has shown that the best technique to bring solution into the empty spaces of PP implants at the time of primary placement is to compress the solution into the implant within a large syringe.18 Although only 5 of the conjunctival bacterial cultures yielded positive results (perhaps because patients had al-
Chuo et al 䡠 Review of Explanted Porous Polyethylene Orbital Implants ready been treated with topical antibiotics), bacteria were identified on gram stain in 12 implants (67%). Bacterial clusters were situated primarily in the center of the implant in areas of microabscess formation, which also were the center of the avascular zone, suggesting that antibiotic penetration into these areas would be minimal. Seventy-two percent of the implants were removed from patients who lived in a distant location, despite the fact that local residents make up most of the cases requiring globe removal. The authors postulate that the distance made it difficult for these patients to seek medical attention and follow-up care in a timely fashion, resulting in long-standing implant exposures that subsequently led to established infections that were difficult to control without implant removal. It is also possible that ill-fitting prostheses may have been detected or treated less easily in remote communities. The authors recommend that patients be informed about the symptoms and signs of implant exposure and be encouraged to contact their local ophthalmologist or primary surgeon so that prompt medical intervention or surgical repair can be instituted. In summary, this is the largest review of explanted PP orbital spheres. With the results of this study, the authors postulate that implant exposure allows bacteria to penetrate into the center of the implant, which has limited fibrovascular ingrowth and acts as an immune sanctuary for infection. Techniques to limit implant exposure may include seating the porous implant deep in the socket, dividing the scleral envelope posteriorly in eviscerations, using an anterior covering to limit compression of the mucosal tissues between the implant and the prosthesis, advancing the extraocular muscles anteriorly to provide better anterior vascular coverage, and increasing the chance of fibrovascular ingrowth by increasing the surface area contact between the extraocular muscles and the implant through large scleral windows. The authors believe that patients, especially those living in remote communities, need to be informed that they must contact the surgeon as soon as possible if exposures form. Exposures need to be treated aggressively and promptly with cultures, systemic antibiotics, and surgical repair, especially in cases of enucleation and PP spheres. After symptoms of infection develop, one may try antibiotic injection into the implant,18 but implant removal likely will be required.
References 1. Massry GG, Holds JB. Coralline hydroxyapatite spheres as secondary orbital implants in anophthalmos. Ophthalmology 1995;102:161– 6. 2. Dutton JJ. Coralline hydroxyapatite as an ocular implant. Ophthalmology 1991;98:370 –7. 3. Berghaus A. Porous polyethylene in reconstructive head and neck surgery. Arch Otolaryngol 1985;111:154 – 60. 4. Blaydon SM, Shepler TR, Neuhaus RW, et al. The porous polyethylene (Medpor) spherical orbital implant: a retrospective study of 136 cases. Ophthal Plast Reconstr Surg 2003;19:364 –71. 5. Karesh JW, Dresner SC. High-density porous polyethylene (Medpor) as a successful anophthalmic socket implant. Ophthalmology 1994;101:1688 –95; discussion 1695– 6.
6. Remulla HD, Rubin PA, Shore JW, et al. Complications of porous spherical orbital implants. Ophthalmology 1995;102: 586 –93. 7. Jordan DR, Brownstein S, Jolly SS. Abscessed hydroxyapatite orbital implants: a report of two cases. Ophthalmology 1996; 103:1784 –7. 8. Glasgow BJ, Weinberg DA, Shorr N, Goldberg RA. Draining cutaneous fistula associated with infection of hydroxyapatite orbital implant. Ophthal Plast Reconstr Surg 1996;12:131–5. 9. Wilson MW, Wobig JL, Dailey RA. Infection of a porous polyethylene orbital implant with Capnocytophaga. Ophthal Plast Reconstr Surg 1998;14:398 – 402. 10. Christmas NJ, Gordon CD, Murray TG, et al. Intraorbital implants after enucleation and their complications: a 10-year review. Arch Ophthalmol 1998;116:1199 –203. 11. Jordan DR. Problems after evisceration surgery with porous orbital implants: experience with 86 patients. Ophthal Plast Reconstr Surg 2004;20:374 – 80. 12. Cheng MS, Liao SL, Lin LL. Late porous polyethylene implant exposure after motility coupling post placement. Am J Ophthalmol 2004;138:420 – 4. 13. Nunery WR, Heinz GW, Bonnin JM, et al. Exposure rate of hydroxyapatite spheres in the anophthalmic socket: histopathologic correlation and comparison with silicone sphere implants. Ophthal Plast Reconstr Surg 1993;9:96 –104. 14. Shields CL, Shields JA, De Potter P, Singh AD. Lack of complications of the hydroxyapatite orbital implant in 250 consecutive cases. Trans Am Ophthalmol Soc 1993;91:177– 89; discussion 189 –95. 15. Giligson A, Dolman PJ, Buffam F. Comparison of retrobulbar analgesics for evisceration. Ophthal Plast Reconstr Surg 2002; 18:258 – 60. 16. Custer PL. Enucleation: past, present, and future. Ophthal Plast Reconstr Surg 2000;16:316 –21. 17. Karsloglu S, Serin D, Simsek I, Ziylan S. Implant infection in porous orbital implants. Ophthal Plast Reconstr Surg 2006;22: 461– 6. 18. Badilla J, Dolman PJ. Methods of antibiotic instillation in porous orbital implants. Ophthal Plast Reconst Surg 2008;24: 287–9. 19. Custer PL, Trinkaus KM. Porous implant exposure: incidence, management, and morbidity. Ophthal Plast Reconstr Surg 2007;23:1–7. 20. Jordan DR, Brownstein S, Faraji H. Clinicopathologic analysis of 15 explanted hydroxyapatite implants. Ophthal Plast Reconstr Surg 2004;20:285–90. 21. Klapper SR, Jordan DR, Brownstein S, Punja K. Incomplete fibrovascularization of a hydroxyapatite orbital implant 3 months after implantation. Arch Ophthalmol 1999; 117:1088 –9. 22. Li T, Shen J, Duffy MT. Exposure rates of wrapped and unwrapped orbital implants following enucleation. Ophthal Plast Reconstr Surg 2001;17:431–5. 23. Reznick JB, Gilmore WC. Host response to infection of a subperiosteal hydroxylapatite implant. Oral Surg Oral Med Oral Pathol 1989;67:665–72. 24. Buettner H, Bartley GB. Tissue breakdown and exposure associated with orbital hydroxyapatite implants. Am J Ophthalmol 1992;113:669 –73. 25. Ainbinder DJ, Haik BG, Tellado M. Hydroxyapatite orbital implant abscess: histopathologic correlation of an infected implant following evisceration. Ophthal Plast Reconstr Surg 1994;10:267–70.
353
Ophthalmology Volume 116, Number 2, February 2009
Footnotes and Financial Disclosures Originally received: January 27, 2008. Final revision: August 29, 2008. Accepted: September 12, 2008. Available online: December 16, 2008. 1
2
Manuscript no. 2008-135.
Departments of Ophthalmology and Visual Sciences, Vancouver General Hospital and University of British Columbia, Vancouver, Canada.
354
Department of Pathology, Vancouver General Hospital and University of British Columbia, Vancouver, Canada. Financial Disclosure(s): The author(s) have no proprietary or commercial interest in any materials discussed in this article. Correspondence: Peter J. Dolman, Eye Care Centre, 2550 Willow Street, Section I, Vancouver, British Columbia, Canada V5Z 3N9.