Keywords
virtual surgical planning - three-dimensional printing - orbital fracture - orbital
reconstruction - point-of-care
Virtual surgical planning (VSP) and three-dimensional (3D) printing have gained greater
utility in the field of orbital reconstruction. Three-dimensional printing of customized
models of patients' orbital anatomy provides the surgical team with an additional
view of the defect and surrounding structures. The basic process includes converting
medical images from various modalities (computed tomography [CT], magnetic resonance
imaging, etc.) into digital files that can be printed on 3D printers. Both bone and
soft tissue have been successfully printed from low-cost desktop 3D printers.[1] The main limitation in the process of 3D printing is outsourcing to third-party
companies, making the process less efficient and less feasible in the setting of acute
orbital fractures. Incorporating VSP and 3D printing at the point-of-care offers the
advantage of having VSP and 3D printing expertise in proximity to the surgeons at
the treating hospital. When expedited treatment is needed, the process from imaging
to printing patient-specific models can be accomplished in 24 to 48 hours.[2]
Overview of the Traditional Approach to Acute Orbital Trauma
Overview of the Traditional Approach to Acute Orbital Trauma
Types of Orbital Fracture and Indications for Surgery
Although there is no consensus classification system in use for trauma to the orbital
skeleton, orbital traumas can be divided into pure and impure fractures. Pure orbital
fractures involve a single wall of the bony orbit such as the medial wall or the floor.
Impure orbital fractures involve segmental lower orbital rim and adjacent facial bones,
such as nasomaxillary, zygomatic, maxillary, and nasoethmoid bones. Blowout fractures
involve the medial wall or floor of the orbit, and often result in expansion of the
orbital volume. Blow-in fractures are often caused by high impact damage to the lateral
wall or roof and can decrease the orbital volume.[3] Indications for surgical repair of orbital fractures include trap door mechanism
with soft tissue incarceration, ocular motility restriction, persistent diplopia,
globe malposition, or significant enophthalmos. The goals of fracture repair are restoring
orbital volume and unrestricted extraocular muscle function, maintaining symmetric
globe position with the contralateral side, and preventing long-term sequalae and
facial disfigurement.[4]
[5]
[6]
[7]
Surgical Approaches to the Orbit
Current techniques used to gain access to the orbit emphasize exposure while resulting
in a more cosmetically concealed scar and preservation of normal eyelid function and
position. The choice of orbitotomy incision type varies based on the specific fracture
pattern, extent of exposure needed, and associated soft tissue injuries. Approaches
to the lower eyelid include subciliary, subtarsal, and transconjunctival. For a subciliary
approach, the incision is made below the lash line and can be skin only or skin and
muscle. The transconjunctival approach is often preferred as it does not produce a
visible scar, has low complication rates, and can be combined with a lateral canthotomy
via either a retroseptal or preseptal approach for increased exposure. The medial
wall is most commonly accessed by a transcaruncular approach, which can be combined
with a transconjunctival approach. The orbital roof and lateral wall can be approached
via an extended upper blepharoplasty excision. Each method of orbital exposure carries
its own advantages and limitations which must be taken into consideration when choosing
an incision. Regardless of the approach chosen, repair proceeds with exposure of the
orbital rim and bony wall in a subperiosteal plane, identification and protection
of neurovascular structures, reduction of herniated soft tissue, and visualization
of the entire orbital defect[8] ([Fig. 1]).
Fig. 1 Transcaruncular approach to the medial orbit may be necessary to access the medial
wall and can be combined with a transconjunctival lower eyelid incision to access
the floor when necessary.
Types of Orbital Implants
Several alloplastic materials are available for reconstructing simple, single wall
defects, such as Medpor (Stryker, MI), polytetrafluoroethylene (PTFE), silicone, preshaped
titanium plates (Synthes/DePuySynthes, KLS Martin, Stryker), titanium mesh, and resorbable
materials such as polydioxanone sulfate membrane (Ethicon, Norderstedt, Germany).[8]
[9] Bone grafts were used in the past but have been associated with unpredictable resorption
rates and increased donor site morbidity, thus limiting their current use in favor
of alloplastic material.[10]
[11] Medpor is a nonresorbable, porous polyethylene implant that is easy to shape and
has had similar results as bone tissue with fewer infection rates.[9] Titanium mesh implants have been used since 1992 to correct orbital wall fractures
of moderate to large size and have the advantage of being radio-opaque and radiologically
visible. The main complication of titanium mesh is difficult removal in cases of infection
due to tissue growing within the gaps of the mesh. PTFE and silicone are not commonly
used because of risk of extrusion that can occur up to 21 years after placement. Resorbable
materials are used for pediatric patients due to their low immune reactivity and high
biocompatibility. Patient-specific implants (PSIs) are particularly useful in complex
orbital fractures and provide a precise implant based on the contralateral orbit.
While these alloplastic materials offer biocompatibility, stability, and safety, one
disadvantage they all share is cost, which will be discussed later.[8]
[12]
[13]
Advantages of PSIs
PSIs are used for apical, skull base fractures, defects that are too large or complex
for prefabricated implants, and when additional bulk is needed to correct orbital
volume or globe position. The normal orbital cavity is virtually planned by mirroring
the normal side onto the fracture side. After normal orbital volumes are calculated
and the contour and thickness of the implant are designed, the implant can be 3D printed.
PSIs can be made using various materials, including titanium, Medpor, and polyetheretherketone
(PEEK). PEEKs are highly biocompatible, durable, and fatigue resistant, but only a
few studies have reported their use in orbital implant reconstruction.[14] The efficiency of PEEK PSIs was compared with that of prebent titanium implants
in the reconstruction of posttraumatic orbital wall fractures. Postoperative diplopia
was seen in 17.9% of patients treated with PEEK PSIs and 29.4% of patients with prebent
implants. Intraoperative time was shorter in the PEEK PSI group, with an average of
54.25 minutes, compared with 82.9 minutes in the prebent implant group. Finally, the
average difference in orbital volume between the fractured and normal orbits was 0.74 cm3 in the PEEK PSI group, which was lower than 1.9 cm3 in the control group.[15] PSIs are also used in pediatric cases. Akiki et al described the case of a 7-year-old
girl who was in a motor vehicle collision and experienced an orbital blowout fracture
involving the medial wall and floor of the orbit. A 3D model was used as the defect
was large and provided the surgeons with a better understanding of the space it would
have to fill. Resorbable material was used for this case because she had not reached
skeletal maturity. At the 18 months' follow-up, the authors reported no complications
and normal eye projection.[16]
Three-dimensionally printed PSIs and models have been shown to be anatomically accurate.
Schön et al demonstrated that PSIs were accurate within a 1-mm range on postoperative
CT scan.[17] In this study, the accuracy of the 3D orbital reconstruction was determined via
image fusion of the postoperative CT and the preoperative virtual plan, and measuring
the absolute maximal distances in the axial, coronal, and sagittal planes.[17] Blumer et al found a mean difference of 0.6 mm between the virtual and surgical
reconstructions, with a mean maximal distance of 3.4 mm. In this study, the accuracy
of the reconstruction was determined by superimposing the postoperative 3D image onto
the preoperative virtual reconstruction.[18] These two studies calculated surgical error using different techniques and so cannot
be directly compared, but both demonstrated that PSIs are able to obtain precise results.
In another study, the accuracy of the 3D models and two-dimensional CT images was
compared and found to have a difference of < 1 mm that was not statistically significant.[19] The accuracy of these models translates into fewer postoperative complications.
Other advantages to custom 3D-printed implants are shortened intraoperative times,
reduced length of anesthesia, and greater precision in restoration of orbital volume
to match the unaffected orbit.[9] In addition, the revision rates of orbital reconstruction using 3D-printed models
and PSI were lower compared with reconstruction with standard implants.[20]
One of the challenges to using PSIs has been cost. The cost of standard implants depends
on the material used. Medpor Titan Mesh (Stryker) prices ranged from $20,000 to $29,000,
3D preformed implants (Stryker) ranged from $42,000 to $50,000, MatrixORBITAL (Synthes,
PA) ranged from $33,000 to $40,000, and RapidSORB (Synthes) ranged from $35,000 to
$39,000.[21] While no systemic review has directly compared the cost of PSIs and preformed standard
orbital implants, several studies point to the increased cost of PSIs as a limitation.[22]
[23]
[24] However, 3D printing technology is becoming more widespread and low-cost 3D printers
have been shown to be reliable. PSIs can cost between $2,000 and $14,000, but their
accuracy and low revision rates must also be taken into account when considering long-term
costs.[25]
Posttraumatic Orbital Deformities
Posttraumatic Orbital Deformities
Posttraumatic orbital anatomy can be altered due to scar tissue, atrophy of orbital
fat, and bony remodeling of fractures, making exposure of orbital wall defects challenging.
Inadequate visualization of the orbital wall defect and important landmarks may result
in overcontouring or malpositioning of implants.[11]
[20] Long-term complications can result from orbital fractures as well as orbital reconstruction
because of inaccurate reduction of fractures, orbital volume enlargement, or hardware
malposition. Diplopia can develop due to extraocular muscle restriction, scarring,
or plate malposition. Eyelid retraction, ectropion, entropion, and enophthalmos are
common indications for secondary reconstruction of posttraumatic deformities.[4]
[20] Vision loss can occur due to optic nerve compression during orbital fracture reconstruction
or from postoperative retrobulbar hemorrhage. While retrobulbar hemorrhage is a rare
complication, it presents the patient and the surgeon with an emergency. In a review
of orbital wall repair cases from 1983 to 1994, retrobulbar hemorrhage was reported
in 0.32% of 1,240 cases, with 50% of those cases resulting in permanent vision loss.[26] In a later review, retrobulbar hemorrhage was seen in 0.17% of 1,180 cases from
2006 to 2011, with 50% of those cases also resulting in permanent blindess.[27] More recently, retrobulbar hemorrhage was encountered in 1.15% of 261 cases from
2011 to 2019. Of those cases, 33% had permanent vision loss.[28]
Timing of Orbital Fracture Repair
Timing of repair of orbital fractures is also an important factor when discussing
potential postoperative complications because delaying repair can lead to permanent
damage.[29] The Burnstine criteria have been used to determine optimal timing of orbital fracture
repair. Immediate intervention is recommended for early enophthalmos (> 2 mm) or hypoglobus
causing facial asymmetry, diplopia with CT evidence of entrapped muscle or periorbital
tissue associated with a nonresolving oculocardiac reflex, or a “white-eyed blowout
fracture” in pediatric patients. Intervention is often delayed in patients with orbital
floor fractures to allow for resolution of initial edema. Surgical intervention within
2 weeks is recommended if more than 50% of the orbital floor is depressed, causing
latent enophthalmos or globe ptosis, or if there is progressive infraorbital hypesthesia.
In cases with minimal diplopia or enophthalmos and good ocular motility, observation
is preferred.[30] However, there are other factors besides the size of orbital floor involvement that
deserve consideration when evaluating whether to intervene surgically. Alinasab et
al proposed a new algorithm based on herniated orbital volume and other CT scan findings
for the prediction of late visible deformities, such as superior sulcus deformity,
hypoglobus, and enophthalmos. This algorithm had 83% accuracy, which increased to
91.5% if patients were followed up at interval times. Based on the new criteria, patients
have high risk of late visible deformity in an inferior blowout fracture with < 1 mL
herniation and a ratio between fracture area and orbital wall area of > 42%, or a
fracture area of > 2.3 cm2; inferior blowout fracture with > 1 mL herniation and > 3 cm distance from inferior
orbital rim to posterior edge of the fracture; or inferomedial wall fracture with > 0.9 mL
of herniation.[31] A recent review also found five factors that contribute to the development of delayed
enophthalmos, including linear measurements of the fracture, involvement of specific
intraorbital structures, rounding of the inferior rectus muscle, orbital fracture
area, and orbital volumetric changes.[32]
Enophthalmos
Enophthalmos occurs when there is inadequate restoration of orbital volume, resulting
in a discrepancy between orbital soft tissue volume and volume of the affected bony
orbital cavity. This may result from soft tissue herniation into a paranasal sinus,
muscular or periorbital atrophy, fat tissue necrosis, and orbital implant malposition.[4]
[10]
[11] Medial orbital wall and combined medial-inferior orbital wall fractures have the
highest association with enophthalmos. He et al reported that out of 71 patients with
enophthalmos, 76% had medial wall fractures and 53% had a combined medial-inferior
wall fracture.[33] In surgical correction of enophthalmos, the goal is to restore orbital anatomy,
volume, function, and aesthetic appearance.[10]
VSP and 3D Printing in Management of Posttraumatic Enophthalmos
In a study by Dvoracek et al, 9 patients with acute orbital floor or wall fractures
were seen at the Children's Hospital of Pittsburg, with 7 patients presenting with
preoperative enophthalmos. CT scans were obtained and reconstructed using Mimics Medical
v21.0 and 3-Matic Medical v13.0 (Materialise, Leuven, Belgium). Three-dimensional
models of the affected side and mirrored unaffected side were printed from an in-house
Formlabs Form 2 stereolithography printer (Formlabs, Somerville, MA), using 0.05-mm
layer thickness and Dental SG Resin. The models served as templates for titanium plates
that were to be used for reconstruction. An average of 10.4 hours was spent preparing
the model and 60 seconds in plate bending.[34] This is compared with obtaining models from third-party companies and having a delivery
delay of 2 to 3 days.[35] Maximum material costs were $21 per patient. Intraoperative time was decreased by
approximately 50%. Of the 7 patients who started out with enophthalmos, 6 had resolution
and 1 had persistent enophthalmos at 4 months of follow-up. The difference in orbital
volume postoperatively between the affected and unaffected eye was insignificant.[34]
Implant Malposition
The most common indication for revision orbital surgery is implant malposition with
an associated clinical symptom, such as globe malposition, vision changes, ocular
motility restriction, or diplopia.[35] In a retrospective study of patients who underwent reconstruction of orbital fractures,
6.5% of 232 reconstructions required revisional surgery. The need for revision was
highest in more complex fractures, such as midfacial fractures that involved the rim,
and was associated with implant malposition. The study also analyzed how the implant
material affected the scoring of implant position after reconstruction (good, acceptable,
poor). Materials used were patient-specific milled titanium implant (PSIs) (Planmeca
Ltd), bioactive glass (BAGS53P4 BonAlive Biomaterials Ltd), polylactide acid and/or
polyglycolic acid polymer (PLA/PGA) (Synthes, Stryker), manually bent titanium mesh
(Synthes/DePuySynthes, Stryker), and preformed 3D titanium plates (Synthes/DePuySynthes,
KLS Martin, Stryker). Eighty-four and seven percent of the PSIs received a score of
“good,” and 100% of both the bioactive glass and PLA and/or PGA received a score of
“good.” This is compared with the 77.2% of the manually bent titanium mesh and 50%
of the preformed 3D titanium plates that received a score of “good.” Patients that
underwent reconstruction with PSIs had a revision rate of 3.4%, compared with 12.9%
with the preformed 3D titanium plates. PSIs and resorbable materials resulted in better
positioning and lower revision rates compared with other materials.[20]
Case 1: VSP and 3D Printing in Management of Bilateral Enophthalmos and Globe Malposition
A 27-year-old male presented to our institution after multiple prior facial surgeries
following an all-terrain vehicle accident 4 years prior. He continued to suffer from
diplopia and enophthalmos despite previous orbital reconstruction with Medpor wedge
implants bilaterally, which had both migrated anteriorly at time of presentation.
He underwent CT imaging with 3D reconstruction, which was used to calculate intraorbital
volumes. Previous studies have shown the average orbital volume of the male patient
to be 24.9 mL.[36] The left orbital volume in this patient was found to be 31.4 mL, whereas the right
orbital volume was found to be 30.8 mL, explaining the patient's enophthalmos and
diplopia. The CT scans were imported into Mimics software (Materialise, Belgium).
The bones of the face were segmented and transferred to 3-Matic (Materialise, Belgium).
A STL file was made and transferred into Pro-plan (Materialise, Belgium). The models
were printed in-house using material jetting on an Objet 500 (Stratasys, MN). The
patient underwent removal of prior implants with placement of new PSIs bilaterally,
resulting in significant improvement of enophthalmos and diplopia after surgery ([Fig. 2]).
Fig. 2 Virtual measurement of intraorbital volumes prior to surgical management of enophthalmos,
globe malposition, and diplopia. The calculated left orbital volume was 31.4 mL, and
the right orbital volume was 30.8 mL, both significantly larger than the average male
orbital volume (see text).
VSP and 3D Printing in Reconstruction of Orbital Fractures
VSP and 3D Printing in Reconstruction of Orbital Fractures
The anatomy of the orbit is complex, and reconstruction can be difficult as visualization
of the surgical field is limited. Noncustom implants often require shaping and trimming
by the surgeon at the time of reconstruction to fit the patient's orbital anatomy,
as well as repeated placement and removal to confirm correct size and shape. This
extends the operative time and may increase the risk of infection.[37] VSP and 3D printing has allowed surgeons to plan corrected orbital volumes of the
fracture side based on normalized anatomy. A 3D-printed model of the normalized side
can be used to contour various types of orbital plates to reestablish accurate orbital
anatomy and volume. Patient-specific 3D-printed models allow surgeons to visualize
the deformity and bend titanium implants or do mock surgery in complex cases beforehand.
The accuracy of the models also ensures correct contouring and positioning of the
implant, and so decreases the risk of revisional surgery due to malpositioning.[38]
In a case series of three patients with enophthalmos due to medial orbital wall fractures,
3D reconstruction of the CT images of the defect created precise models that were
used as a template for the creation of an iliac crest bone graft. Benefits were both
immediate, by decreasing case complexity and operative time, and long-term, by minimizing
the need for future corrections. All reported patients had correction of their enophthalmos
and good postoperative outcomes.[39]
Point-of-Care (In-House) VSP and 3D Printing
Three-dimensional printing at the point-of-care offers the advantage of team learning,
increased efficiency, and an expedited process of VSP and 3D printing. Hatz et al
compared the accuracy of a low-cost desktop 3D printer with a professional-grade 3D
printer and found that the mandible models that were created were comparable in accuracy.
There are many techniques for 3D printing anatomical models, with the most common
being fused filament fabrication (FFF) and selective laser sintering (SLS). While
SLS technology has shorter printing times and higher printer resolution, which makes
it better at printing fine anatomical structures, they are more expensive. FFF technology
is cheaper, costing less than $3,000 USD, and was shown to produce suitable and accurate
anatomical models requiring only minimal adjustment intraoperatively. One difference
noted is that models from FFF printers are made with material that cannot be steam
sterilized and so required special sterilization before use in surgery. Models printed
with SLS printers can undergo steam sterilization and do not require further sterilization
before use in the operating room.[35] At our institution, Ultimaker 3D printers are used, with cost ranging from $3,000
to $5,000 for a reliable model.
Printing PSIs in-house has been shown to be more cost effective than outsourcing to
third parties to print. The average cost of industry-printed PSI was $1,678, whereas
the average cost of printing PSIs in-house averaged $236. The cost breakdown of printing
in-house came out to be $34.50 for software and disposable fees, $43.80 for segmentation,
$11 for materials, $65.60 for print time fees, and $20.50 for production.[40] Thus, printing at the point-of-care can further decrease the cost of PSIs and increase
their availability to patients.
Case 2: Point-of-Care VSP and 3D Printing in Acute Orbital Trauma with Inferior Rim
Comminution and Floor Blowout Fracture
A 17-year-old male presented with a left orbital blowout fracture and comminuted anterior
maxillary fractures after sustaining a hit by a golf ball. Symptoms included diplopia,
and physical exam was notable for enophthalmos and contour depression of the left
inferior orbital rim and maxilla. After a maxillofacial CT was performed ([Fig. 3]) VSP was completed using similar methods described in case 1. Two 3D printed models
were created, one being a model of the patient's skeletal deformity ([Fig. 4A]) and the other a reconstruction of the left orbital floor, inferior rim, and anterior
maxilla using a mirror image of the normal right orbital and maxillary anatomy ([Fig. 4B]). The process of VSP and 3D printing of the two models was accomplished at the point-of-care.
The surgical procedure was completed through a combined transconjunctival and transoral
approach. The 3D printed model was utilized intraoperatively to contour the orbital
floor implant ([Fig. 4C]). The 3D printed model was used to contour the inferior orbital rim plate pre-operatively.
The pre-contoured plate to the perfected anatomy ([Fig. 5A]) was sterilized and used intraoperatively as a guide to help reduce and fixate the
bony fragments ([Fig. 5B]). The patient had a successful outcome with restoration of globe and bony symmetry
at 13 months follow-up.
Fig. 3 (A) Maxillofacial CT with 3D reconstruction demonstrating the comminuted nature and
contour deformity of the left inferior orbital rim and anterior maxilla. (B) and (C) demonstrate the contour deformity of the left rim and anterior maxilla prior to
and after open reduction and internal fixation of fractures using VSP and 3D printing.
Fig. 4 (A) Printed 3D model of the traumatized facial skeleton showing left orbital floor blowout
fracture with inferior rim comminution. (B) Printed 3D model at the point-of-care after VSP with left orbito-maxillary anatomy
created by mirroring the normal right side onto the left side. (C) 3D printed model used intraoperatively to contour the orbital floor implant.
Fig. 5 An illustration of VSP using a precontoured plate based on perfected anatomy. (A) Plate fixated medially and laterally on unaffected bone prior to reduction of rim
fractures. The plate was used as a guide to reestablish accurate symmetry and contour
with the right side. (B) After reduction and fixation of the inferior rim fractures.
3D Printing for Resident Education on Orbital Anatomy
3D Printing for Resident Education on Orbital Anatomy
Three-dimensionally printing models offer an additional platform in resident surgical
education. These models serve as reusable visual teaching aids to enhance hands-on
learning experiences, such as live surgeries and cadaver dissections. They have been
shown to enhance visual-spatial skills by providing immediate feedback, improve memory
of procedures, and allow for preparation with a realistic model prior to the day of
surgery.[41] The use of 3D-printed models in the teaching of orbital anatomy is especially helpful
as orbital anatomy is complex and there is a restricted field of view during surgery
that makes intraoperative teaching difficult. Vatankhah et al performed a study in
which 24 ophthalmology residents in years 1 and 2 at Mashhad University of Medical
Sciences in Iran were randomized into two groups of learning. One group trained with
traditional methods and the other group with 3D-printed models of fractures and congenital
abnormalities. Pretest and posttest scores were compared with measure knowledge enhancement
3 and 14 days later. The posttest scores of students who learned with 3D-printed models
were higher than the scores of students in the traditional learning group. Interestingly,
the use of 3D models in teaching was more effective in year 1 residents than year
2 residents, as evidenced by their posttest scores.[41] Three-dimensional models for resident education have been shown to improve residents'
learning in a concrete way by improving test scores but also stimulate interest and
curiosity in their field of study ([Fig. 6]).
Fig. 6 Three-dimensional (3D) printed models enhance resident education. The surgical anatomy
is reviewed and when necessary, plates are bent preoperatively or mock surgery can
be performed ahead of surgery in complex cases.
Conclusion
VSP and 3D printing have advanced the field of orbital fracture reconstruction by
providing surgeons with precise anatomical models for preoperative planning and decreasing
postoperative complications. Integrating VSP and 3D printing at the treating hospital
has further cut down time to operation by eliminating third-party outsourcing and
decreased the overall cost. Low-cost desktop 3D printers were shown to be comparable
in accuracy with more expensive professional-grade 3D printers. As VSP and 3D printing
become more widely used and accessible, the treatment and outcome of acute complex
orbital fractures will be elevated as a new gold standard emerges.
