The authors present the clinical results of their method of customized reconstruction of orbital wall defects using titanium mesh or sheet. High resolution computed tomography (CT) data are imported and processed to create a three-dimensional (3D) image which is used to reconstruct the orbital defect. Mirror imaging of the air in the contralateral maxillary sinus is used to overcome artefact defects in the floor. A stereolithographic model is constructed, from which titanium mesh or sheet is shaped and sized to the required contours for implantation. Twenty-two patients were treated using this technique from 2003 to 2008. Postoperatively 10 patients reported early resolution of their diplopia. Six patients noticed significant improvement of their symptoms with mild residual diplopia in one direction only and at the extremes of gaze at final review. One patient required ocular muscle surgery. Enophthalmos resolved in eight of the nine cases. No patients developed enophthalmos or diplopia as a postoperative complication. The use of titanium mesh for orbital floor reconstruction has been shown to be safe and effective. Customized titanium implants accurately reproduce orbital contours thus restoring orbital volume. This reduces operative time and improves the functional and aesthetic outcomes of post-traumatic orbital reconstruction.
The use of high resolution computed tomography (CT), with the ability to examine the scans routinely in three planes, has provided a better understanding of the three-dimensional (3D) structure of the bony orbit. This combined with the use of rapid prototype models from CT data allows the construction of a more accurate customized implant for post-traumatic reconstruction of orbital wall defects . In this article, a retrospective review of orbital wall fractures treated using customized titanium mesh or sheet implants is described. The authors’ recently described method of virtually reconstructing the orbital floor defect prior to rapid prototype model construction is also briefly discussed .
Materials and methods
Indications for using the technique included severe comminution, large defects with little or no posterior bony support and secondary reconstruction.
Patients were scanned using high resolution CT with a slice thickness of 0.5 mm. The raw CT information obtained was imported in DICOM (Digital Imaging and Communications in Medicine) file format and processed to create a 3D image using Mimics software (Materialise NV, Leuven, Belgium). Despite the ultra-thin CT slices, the orbital floor may, in some areas, be less than 0.5 mm thick . This is too thin to be captured by the scan, resulting in the phenomenon of pseudo-foramina when viewed in 3D format on both the injured and unaffected orbital floors. To overcome this obstacle, the captured air volume of the contralateral maxillary sinus was identified to create a virtual 3D image of that sinus. This modified CT was exported as an STL file into another software package, Freeform ® (SensAble, USA) which allowed intuitive 3D digital manipulation with biofeedback. The position of the mid sagittal plane was then established in order to be able to mirror image the sinus morphology to the defect side ( Fig. 1 ). The superior contour of the 3D sinus volume thus recreates the shape of the orbital floor .
The resulting virtual model was used to construct a stereolithographic model ( Fig. 2 ) using rapid prototyping. This model was used to enable the shaping of the titanium mesh in the laboratory to the required contours to fit the virtually repaired orbital defect. The pre-shaped titanium mesh was then sent to be sterilized for use in the operating theatre. Alternatively the model was used to make an implant of the desired shape and size using titanium sheet as previously described . This was carried out by taking an impression of the affected orbit on the stereolithographic model with the now reconstructed orbital wall defect. A hard stone plaster model was poured, which was subsequently used to fabricate the customized implant from commercially pure titanium using pressure flasks. Two different titanium thicknesses of 0.25 and 0.5 mm were available, pressed over the cast using 6000 PSI for 24 or 48 h, respectively. The resulting plate was trimmed, smoothed and polished. Multiple venting holes were created in some cases, according to the designing surgeon’s preference, and a flange was extended over the orbital rim to facilitate implant positioning and to prevent it from being displaced posteriorly, with a hole for subsequent screw fixation. The implant was pacified overnight in concentrated nitric acid for decontamination. The final plate was sent for sterilisation to be used in the operating room. The shortest achievable time from the decision to operate to the implant being sterile and ready to use was 4 days with the rate-limiting step being mainly the logistic problems of the construction of the stereolithographic model at another offsite institution.
The two alternative incisions used for access were the mid-eyelid blepharoplasty or the conjunctival incision, depending on operator preference. Where the conjunctival incision was used in the presence of floor and medial defects, a post-caruncular extension was performed to access the medial wall. Orbital dissection was performed in the normal way, exposing and delineating the defect. Herniated orbital contents were retrieved and the plate was inserted and anchored with one or more screws depending on plate design and position. Forced duction testing was performed in all cases and closure was performed after confirming that there was no restriction or tethering of the globe.
A retrospective review of 22 cases treated at the authors’ institution using customized titanium implants for orbital wall reconstruction was carried out. A form was designed for data collection and patients’ case notes were examined, looking at emergency cards, outpatient clinic notes, operating records and correspondence. The collected information was entered into a spreadsheet programme (MS Excel ® , Microsoft USA) for further processing. The results were reviewed and analysed.
Twenty-two patients were treated using this technique between 2003 and 2008 ( Figs 3 and 4 ). A summary of the results is presented in Table 1 . Seventeen patients presented with diplopia, nine of whom also had enophthalmos. Two patients had enophthalmos with hypoglobus but no diplopia. Postoperatively, 10 patients reported early resolution their diplopia. Six patients noticed significant improvement of their symptoms with mild residual diplopia, in one direction only and at the extremes of gaze, at final review. One patient required ocular muscle surgery. Enophthalmos resolved in eight of the nine cases. No patients developed enophthalmos or diplopia as a postoperative complication.
|Age||Sex||Diagnosis (site and indication)||Aetiology||Preop enophthalmos||Preop diplopia||Time from trauma to surgery in days||Incision||Implant profile||Postop enophthalmos||Postop diplopia||Complications||Follow-up in months|
|1||23||M||Floor comminution with malar fracture||Assault||Yes||Yes||12||Lower blepharoplasty||0.5 mm plate||No||Resolved 30/7||Scleral show, settled later||16|
|2||30||M||Large floor defect with malar fracture||Assault||No||No||7||Lower blepharoplasty||Mesh||No||No||Scleral show||19|
|3||29||M||Floor comminution||Sports||No||Yes||13||Lower blepharoplasty||Mesh||No||Resolved 8/12||None||8|
|4||18||F||Floor and medial wall secondary reconstruction||RTA||Yes||Yes||114||Lower blepharoplasty||0.5 mm plate||Resolved||Improved||Mild residual diplopia||3|
|5||18||M||Large floor defect||Assault||No||yes||31||Conjunctival||Mesh||No||Improved||None||2|
|6||64||M||Floor and malar fracture||Work-related||No||No||17||Lower blepharoplasty||Mesh||No||No||None||5|
|7||24||F||Large floor defect||Assault||No||Yes||37||Lower blepharoplasty||Mesh||No||Improved||None||2|
|8||27||M||Large floor defect||Assault||No||Yes||20||Lower blepharoplasty||Mesh||No||Resolved 1/7||None||3|
|9||36||M||Large floor and medial wall defect||RTA||Yes with hypoglobus||No||62||Lower blepharoplasty||0.5 mm plate||Resolved||No||Ectropion (also had severe lid injuries)||5|
|10||15||M||Large floor defect||Sport||No||Yes||10||Lower blepharoplasty||0.5 mm plate||No||Resolved 1/7||None||20|
|11||29||M||Large medial wall defect||Sport||No||Yes||34||Lower blepharoplasty and medial ethmoidotomy||0.25 mm plate||No||Resolved 1/7||Temporary ectropion, resolved||6|
|12||32||F||Floor and medial wall secondary reconstruction||RTA||Yes with hypoglobus||Yes||2 years||Lower blepharoplasty||0.5 mm plate||Improved||Improved||Discomfort and palpable plate rim||38|
|13||44||M||Large floor defect||Sports||Yes||Yes||73||Conjunctival||0.25||Resolved||Required ocular muscle surgery||None||24|
|14||49||M||Large floor and medial wall defect||Assault||Yes with hypoglobus||Yes||253||Conjunctival with lateral canthotomy||Mesh||Resolved||Resolved||None||6|
|15||49||M||Large floor and medial wall defect||Assault||Yes with hypoglobus||No||68||Lower blepharoplasty||2 × 0.25 mm plates||Resolved||No||None||10|
|16||17||M||Large floor defect||Assault||Yes with hypoglobus||Yes||164||Conjunctival||0.25 mm plate||Resolved||Resolved 5 days postop||None||8|
|17||40||M||Large floor defect||Assault||Yes||Yes||234||Conjunctival||0.25 mm plate||Resolved||Resolved||Mild right globe elevation, subsequently settled||15|
|18||30||M||Comminution||Assault||No||Yes||27||Lower blepharoplasty||0.25 mm plate||No||Resolved||Deterioration in vision postop, resolved later||5|
|19||25||M||Large floor and medial wall defect||Fall||No||Yes||24||Lower blepharoplasty||0.25 mm plate||Resolved||No||None||6|
|20||29||M||Large floor defect||Sports||yes||Yes||90||Conjunctival||0.25 mm plate||Resolved||Resolved 7/7||None||3|
|21||19||M||Large floor defect||Assault||No||No||27||Lower blepharoplasty||0.25 mm plate||No||No||None||2|
|22||32||M||Delayed primary reconstruction||Assault||yes||Yes||331||Lower blepharoplasty||0.25 mm plate||Improved||Improved||None||4|