Orbital bony reconstruction with pre-sized and pre-contoured porous polyethylene - titanium implants

Introduction:

The orbit is a confined space with well-defined bony topographic anatomy. Trauma to either the globe or adjacent rim can cause fracturing of the orbital walls via increased hydraulic pressure, buckling, or a combination of both.¹ Bony prolapse into the ethmoid and/or maxillary sinuses can result in displacement of orbital tissues including the medial and inferior rectus muscles and their adjacent fascial septae. The result is a variety of functional and cosmetic sequelae. Large unmitigated bony orbital deformities may result in profound enophthalmos, hypoglobus, and intractable diplopia.²

Complex orbital floor and medial wall defects involving the inferomedial strut or the posterior bony ledge present reconstructive challenges due to loss of intraoperative orbital anatomical landmarks and adjacent bony support.³ A posterior orbital ledge composed of the palatine bone is a key landmark in floor fracture repair surgery. Reconstructive orbital implants should extend to this posterior ledge to provide proper anatomical reconstruction of the floor with its normal upward slanting configuration of 18-22 degrees.⁴ Failure to find the posterior ledge or its bony remnant may result in poor fracture repair outcomes as the implant will often be positioned within the maxillary sinus posteriorly.³ Previously described reconstructive options include the use of autologous bone grafts as well as a variety of different alloplastic materials. Alloplastic implant catergories include non-porous sheets (silastic, nylon, silicone), porous sheets (porous polyethylene or PPE), titanium mesh, or a hybrid of titanium mesh with incorporated PPE.^{2, 3,5–11} Hybrid PPE - titanium implants combine the advantages of pre-existing implant materials including excellent rigid support with opportunity for bony fixation with a porous overlay allowing for fibrovascular ingrowth by sinus mucosa while improving ease of implant placement with a reduction in sharp edges following intraoperative modification.⁷ Additionally, the PPE barrier layer on the orbital side reduces the risk of orbital adherence syndrome (OAS), a possible complication following the use of bare titanium mesh implants^{10, 12–14} Historically, hybrid implants have been manufactured as flat sheets requiring surgeon manipulation and contouring; an often time consuming and error prone process.^{3, 5} Recent advances in orbital implants include pre-operative modeling using a 3D printed fracture template^15–18 as well as manufacturer pre-sized and contoured titanium mesh implants.¹⁹

Commercially available pre-sized and pre-contoured PPE - titanium implants (Medpor Titan 3D MTB Orbital Floor Implant, Stryker Craniomaxillofacial, Kalamazoo, MI) are designed to re-establish normal orbital floor and medial wall anatomy and are modeled after 300 anatomically averaged orbits (mean age 59, range 11-95; 92% Caucasian, 1% Middle-Eastern; 59.3% male). ²⁰ 3D Titan implants are available in two sizes, small and large, and are otherwise analogous in design and contour (Figure 1). Even if the posterior ledge is absent, these implants provide the normal upward slant of the orbital floor relative to the orbital rim recreating a normal anatomical configuration. This reconstructive option was FDA approved in 2015²⁰ and this is the first study to report clinical outcomes and experiences with this implant in a large series of patients. Herein, we also describe a technique to insert full-sized implants without inferior oblique disruption. Of note, none of the authors have any financial or other relationship with the manufacturer of the studied implant and this research was not financially supported by the manufacturer.

Figure 1.

Example hybrid PPE-titanium right sided small and large implants with their respective dimensions.

Methods

This retrospective study was approved by the Institutional Review Board of the University of Miami and included all patients undergoing orbital fracture repair with the 3D Titan MTB implant from January 2016 to June 2018 at one institution. All facets of this study were conducted in accordance with the tenets of the Declaration of Helsinki. Patients were identified via an operating room reporting process that tracks implant use based on specific model codes as well as a list of patients with specific fracture related CPT codes who underwent surgery during the study period. The medical records were reviewed in compliance with the Health Insurance Portability and Accountability Act (HIPAA) and informed consent was obtained for the use of all patient clinical photographs. Charts were reviewed for demographic data, mechanism and date of injury, time from injury to repair, previous orbital fracture repair, CT fracture characteristics, past ocular history, surgical approach including implant modifications, clinical exams including motility and exophthalmometry, pre and post-surgical symptomatology, and post-operative complications including bleeding, infection, and implant revision. Subsequent related surgical procedures were also identified, including strabismus surgery.

Surgical Technique

The surgical approach was similar in all cases including a transconjunctival incision with preseptal dissection to the inferior orbital rim. A swinging eyelid approach consisting of a lateral canthotomy with inferior cantholysis was utilized when necessary to improve surgical access. Additionally, a transcaruncular approach to the medial wall was utilized for medial wall defects. A limited subperiosteal dissection of the anterior face of the inferior orbital rim was performed in order to provide additional bone for later fixation. Forced duction testing was performed at the start of each case and following implant placement. The inferior oblique was left intact except for two cases in which it was intentionally divided and reanastamosed at the end of the case as described by Alameddine et al.²¹ Herniated orbital contents were carefully freed with a hand over hand technique using a Frazier suction, Freer elevator, and malleable retractors. Intraoperative measurements were used to assess the distance from the orbital rim to the posterior most aspect of the orbital floor. The goal in all adult patients was to place a large sized implant, especially those with large concomitant medial wall fractures. The small size implant was chosen when the distance from rim to posterior floor was less than 35mm. The medial wall or wing portion of the implant was shortened intraoperatively when there was no associated medial wall fracture or the medial wall fracture was small and inferior in order to limit unnecessary intraoperative dissection. Insertion of small sized implants and large sized implants with trimmed medial wings was generally straightforward. However, inserting a full size large implant without disrupting the inferior oblique required introduction of the medial wing portion first with the orbital contents retracted followed by a 90 degree rotation around the inferior oblique muscle (Video 1). Implant positioning was confirmed intraoperatively using the available bony margins and then the implant was fixated to the caudal portion or anterior face of the rim using the available fixation holes and two titanium screws. Both self-drilling and self-tapping screws were utilized depending on surgeon preference. The two distal most fixation eyelets were often removed leaving the fixation holes overlying the caudal portion of the rim and two proximal eyelets which were bent downwards over the anterior face of the rim. Intraoperative feasibility was used to determine which fixation points were utilized.

Results

A total of 33 patients and 34 orbits were identified (mean age 43 ± 16 years, 70% male). All patients were treated by an ASOPRS-trained oculoplastic surgeon via a transconjunctival incision with or without (79%) a swinging eyelid approach and transcaruncular medial orbitotomy as needed. Common traumatic mechanisms included assault (36%), falls (18%), and motor vehicle accidents (21%). Other orbital reconstructions were performed following maxillectomy, as part of treatment for mucoceles, or due to silent sinus syndrome. Two patients were anophthalmic with unrepaired orbital fractures, and orbital fracture repair with implant was performed to address orbital volume deficiency and anophthalmic socket syndrome. All patients had large orbital floor deformities and 74% had additional defects of the medial wall. Most orbits had an absent posterior ledge (82%) and disrupted inferomedial strut (68%). Symptomatic diplopia (73%, n=24) and enophthalmos (89%, mean of 3.7 ± 2.1mm) were common pre-operative findings. A large number of cases were revisions of previous failed surgeries (44%, Figure 2) and nearly all revisional patients had pre-operative symptomatic diplopia as well as extraocular motility deficits (92%) with positive forced ductions (83%). Most orbits (82%) received size large implants. The medial wing was trimmed in 59% of implants, particularly in patients with isolated floor fractures. Post-operatively, mean follow-up was 7.8 ± 6.7 months (range 7 days to 21 months, 17 patients had minimum 24 weeks of follow-up). All patients had improved globe positioning, with reduction of enophthalmos (mean post-op relative Hertel asymmetry of 0.92 ± 1.39mm, mean change in Hertel of 2.44 ± 1.93mm) and hypoglobus. Seven patients (21%) had persistent post-operative diplopia in primary or reading gaze; 5 responded to prism therapy and two patients required strabismus surgery. Six patients with persistent diplopia were revisional cases (86% of all patients with persistent diplopia) and these cases were believed to have intrinsic muscle damage based on decreased extraocular motility, negative forced duction testing, and anatomical globe positioning. One patient required revision surgery at post-op week 2 due to a delayed retrobulbar hemorrhage from Valsalva, and one patient required implant explantation due to implant exposure and infection in the setting of maxillectomy and adjuvant radiation for maxillary sinus squamous cell carcinoma. No implants were revised due to implant malposition or restriction of extraocular motility. Two patients presented with pre-existing lid retraction and ectropion as a result of previous surgeries. No other post-operative lid malpositions were observed in the remaining patients.

Figure 2.

Pre-operative CT orbits in the coronal (A) and sagittal (B) plane on the left side. The manually bent implants are placed at disparate vertical levels with suboptimal approximation of the normal orbital floor topography (A). The left implant is posteriorly displaced downwards away from the apex into the maxillary sinus (B) allowing residual orbital expansion and symptomatic enophthalmos. Post-operative scans (C, D) showing bilateral implant replacement with pre-sized and pre-contoured PPE - titanium implants that were modified by removing the medial wing. In the coronal plane the implants conform well to the expected bony anatomy and in the sagittal plane on the left side the implant is directed posteriorly at an appropriate vector for full fracture reduction.

Discussion

Complex bony orbital deformities with loss of the posterior ledge or inferomedial strut are often challenging surgical cases as a result of the loss of both anatomical landmarks and absence of adjacent bony support (Figure 3). Revisional surgery adds another layer of anatomical complexity and increased operative time removing scarred implants.^{2, 3,11, 22} Favorable outcomes have been reported via a number of different approaches as well as using a variety of implant materials.^{2, 8, 11, 23} Modern techniques often include the use of a transconjunctival approach, with or without swinging eyelid augmentation, as well as the use of alloplastic implant materials such as nylon, PPE, and titanium mesh.⁸ Implants may be supplied as flat sheets requiring intraoperative modification by the surgeon or pre-operative modification using a 3D printed template.^{5, 15}

Figure 3.

This patient presented with a complex fracture of the left orbital floor, medial wall, and inferomedial strut complaining of symptomatic diplopia due to globe positioning. Post-operative scans in the coronal, axial, and sagittal planes show orbital floor and medial wall reconstruction using a pre-sized and pre-contoured PPE - titanium implant.

Cho et al. reported outcomes for a cohort of patients using the previous generation of PPE - titanium implants which were supplied as sheets and required intraoperative manipulation. Pre-operative defect measurements were used to size the implants which were then manually bent and contoured. Although diplopia mostly improved in this cohort and no new diplopia was created after fracture reduction, 18% of patients still required subsequent revision due to misplacement of the implant creating inadequate reduction in two patients and compressive optic neuropathy in the other.³

Recently, implants have been designed based on normative anatomical modeling and supplied pre-sized and formed to fit an average orbit. ¹⁹ A cadaveric study showed the advantage of pre-contoured implants over implants requiring intraoperative bending in terms of accuracy of anatomical reconstruction.⁷ Pre-sized and pre-contoured implants may shorten operative time and reduce the risk of revision by decreasing the number of implant manipulations that must be made intraoperatively. These devices may be supplied fully sized and shaped by the manufacturer or sized and contoured based on a 3D printed patient model, which requires additional pre-operative planning and preparation.^15–18

Although analogous anatomically modeled bare titanium implants now exist¹⁹, there remains a concern of the so called orbital adherence syndrome (OAS).^{10, 12–14} OAS is considered rare but its true incidence is unknown and it has been reported even with the most modern precontoured implants¹²; for this reason these implants are mostly avoided by oculoplastic surgeons.⁸ PPE coating overlying the titanium mitigates the risk of OAS by providing a smooth barrier surface on the orbital side that prevents cicatricial tethering of the orbital tissue while a porous matrix on the sinus mucosal side additionally allows for fibrovascular ingrowth and bio-integration.⁵ In comparison to bare titanium implants, hybrid implants provide smoother implant edges following intraoperative modification due to the PPE coating. PPE or other materials can be layered on top of bare titanium implants, but this modification does not address the sharp titanium cut edges, and adds another layer of expense.⁶

This is the first series to report outcomes for a commercially available pre-sized and precontoured PPE - titanium implant in a large group of patients. Our patient cohort is distinctive in that a preponderance of patients (45%) were revisional cases. Patients presenting for revisional surgery had a higher rate of pre-operative (92%) and post-operative diplopia (50%) as compared to patients who had no previous surgery (72% pre-op and 6% post-op). Additionally, nearly all patients had large complex bony defects involving the posterior ledge and/or inferomedial strut. We further chose to use this implant in patients who had undergone cancer reconstruction, patients with silent sinus syndrome, and patients with volumetrically expanded anophthalmic sockets due to unrepaired chronic fractures.

In most cases (82%), we chose to utilize the size “large” implant and would recommend this for the average adult patient; the small size would be more appropriate for children or smaller adults and sizing may be checked intraoperatively by measuring from the rim to the most posterior aspect of the floor defect. In isolated floor fractures without a posterior ledge or reduced adjacent bony support we elected to use this implant given its normative modeling of the angle between the rim and posterior orbit. Additionally, by trimming the medial wing but leaving intact the upward curved portion corresponding to the inferior aspect of the medial wall the intact inferomedial strut became a useful anatomical landmark for implant orientation. For fixation, numerous options are provided including along the rim and in the form of distal eyelets which may be bent downwards over the face of the maxilla. We often trimmed the most distal eyelets to facilitate easier implant placement while minimizing the dissection over the face of the maxilla. We utilized both the holes overlying the caudal portion of the rim as well as those overlying the face of the rim depending on intraoperative feasibility. In a tight lower lid, the holes overlying the rim superiorly were often easier to access however care must be taken to ensure that the implant is not positioned too far forward on the rim which rotates the posterior aspect of the implant upwards into the orbital apex. Both self-drilling and self-tapping titanium screws can be used depending on surgeon preference and we typically placed two screws that were 4-5mm in length.

The full large size implant is sizeable relative to the tight operative space afforded by an inferior transconjunctival incision; however, a swinging eyelid approach was only utilized in 21% of cases as we preferred to keep the lateral canthus intact whenever possible. Additionally, several groups have reported induced post-operative diplopia as a result of iatrogenic inferior oblique disruption ^{24, 25} and so generally we prefer to keep the inferior oblique muscle attached, which effectively creates two operative spaces and prohibits straightforward implant insertion. In two cases the inferior oblique was intentionally divided and reattached and this was due to surgeon preference – neither patient had any untoward sequelae from this approach. To keep the inferior oblique attached, we utilized a technique for full implant insertion that involves introduction of the intact medial wing of the implant first followed by a 90 degree rotation around the intact inferior oblique muscle (Video 1). This allows maximum floor and medial wall fracture reduction with minimal normal tissue disruption. None of our patients required implant revision due to poor implant positioning. Additionally, our average post-operative relative Hertel difference was less than 1mm (0.92 ± 1.39mm) with improved globe positioning in all patients.

Post-operative complications were nominal and included a small retrobulbar hemorrhage after coughing and sneezing which required drainage as well as removal of the implant in a patient who had undergone previous maxillectomy. The maxillary sinus mucosa overlying the implant eroded several times resulting in a smoldering infection which responded well to antibiotic therapy and implant removal. Interestingly, this patient initially maintained improved globe positioning after implant removal but subsequently developed recurrent hypoglobus.

This study is limited by its retrospective nature and selection bias due to surgeon preference in choosing the appropriate implant for each case. Prior to the introduction of this new implant, we routinely utilized standard flat PPE - titanium implants that required intraoperative sizing and bending (Stryker Titan & OFW implants) for complex fracture repair as well as dual porous polyethylene or Supramid sheets for combined orbital floor and medial wall fracture repair with an intact medial strut and adequate bony support including a posterior ledge. Although our overall complication rate was low, our findings are limited by short duration of follow-up a common issue in studies of predominantly trauma patients.

Supplementary Material

Supplemental Video File

Video This video demonstrates insertion of an unmodified “large” size hybrid PPE-titanium implant via a transconjunctival incision without disinsertion of the inferior oblique or disruption of the lateral canthus. A transcaruncular incision has also been made in order to reduce herniated contents from the ethmoid sinus. The inferior orbit is retracted with a malleable and the implant is inserted medial wing first. The medial wing and implant are then rotated 90 degrees around the intact inferior oblique muscle. The margins of the implant are checked and soft tissue is freed as necessary using a Freer elevator.

Figure 4.

This patient presented for fracture revision after previous implant placement on the right side via a subtarsal approach with subsequent right hypoglobus, exotropia, 2mm enophthalmos, superior sulcus deformity, lid retraction, and symptomatic diplopia. (A) After implant revision using a pre-sized and pre-contoured PPE-titanium implant via a transconjunctival approach his globe positioning and superior sulcus deformity are improved and he is orthophoric with no diplopia in any direction of gaze. Pertinently, at the time of this visit he still had residual right lower retraction from his previous surgery as well as residual traumatic ptosis with right brow elevation imparting an appearance of residual hypoglobus and sulcus hollowing. (B)