Three-dimensional computer navigation in the reconstruction of complex unilateral orbital fractures: evaluation and review of applications

Article information

Arch Craniofac Surg. 2024;25(4):161-170

Publication date (electronic) : 2024 August 20

doi : https://doi.org/10.7181/acfs.2024.00143

Parampreet Singh Saini¹, Rajesh Kumar¹, Manu Saini², Tarush Gupta¹, Sunil Gaba^,1, Ramesh Kumar Sharma¹

¹Department of Plastic Surgery, Postgraduate Institute of Medical Education and Research, Chandigarh, India

²Department of Ophthalmology, Postgraduate Institute of Medical Education and Research, Chandigarh, India

Correspondence: Sunil Gaba Department of Plastic Surgery, Nehru Hospital, Postgraduate Institute of Medical Education and Research, 1 Room No. 43, Level II, Block D, Chandigarh 160012, India E-mail: drsgaba@gmail.com

Received 2024 March 19; Revised 2024 July 9; Accepted 2024 July 17.

Abstract

INTRODUCTION

Orbital fractures account for 25% of maxillofacial injuries presenting at tertiary care centers [1]. The most common mechanisms of injury are motor vehicle accidents and assaults. These injuries are easily noticeable and can have significant aesthetic and functional implications. Given the complexity of human facial anatomy, precise surgical planning is required; failure to adequately plan can result in suboptimal correction and functional disability. Current literature describes various techniques for restoring orbital anatomy [2].

Conventionally, complex maxillofacial reconstruction is planned using stereolithographic (STL) models of the patient. This approach is time-consuming and incurs additional costs, which limits its use in emergency surgeries. Despite computer-assisted planning, complications can still arise due to the lack of real-time feedback [3]. Defined surgical approaches also restrict the view of fracture morphology and the proximity of vital structures. Additionally, there is no universal implant that matches specific orbital dimensions. Common complications reported after orbital fracture repair include diplopia (20%–42.5% in cases of isolated orbital floor fractures; 86% in complex fractures involving multiple walls or the posterior third of the orbital floor), persistent enophthalmos (7%–27.5%), and infraorbital nerve hypesthesia (7%–59%) [4,5].

Restoration of orbital volume typically corrects globe projection and reduces the likelihood of residual diplopia. Proper placement of implants with adequate bony support helps prevent the delayed sagging of the mesh or plate and the recurrence of symptoms [6]. These essential surgical objectives have ushered in a new era of real-time surgical imaging that avoids radiation hazards by using infrared technology. Computed tomography (CT) image-guided navigation is one such modality that has demonstrated its importance across various surgical specialties. Computer-aided design/manufacturing software is available to create three-dimensional (3D) models of subjects and generate mirror images of anatomical subunits using Digital Imaging and Communications in Medicine (DICOM) image data [7]. These mirrored normal facial subunits are superimposed on the contralateral injured segment to plan accurate reductions and implant fixation. Intraoperative navigation, integrated with this virtually planned data, can provide real-time feedback to adjust the position of implants and the anatomical reduction of displaced fractured bony segments.

Over the past 20 years, various studies [8-14] have reported on the use of different modalities for preoperative data registration combined with intraoperative surgical navigation. These methods have been instrumental in identifying posterior bony landmarks and assisting with implant placement during anatomical orbital reconstruction. Most of these studies were retrospective analyses focused solely on pure orbital fractures, and none have prospectively examined their use in impure orbital fractures. Considering the available literature, we conducted a prospective cohort study design to evaluate the application of 3D-computer image-guided navigation in complex orbital reconstruction, incorporating minimally invasive preoperative planning.

METHODS

This single-center prospective cohort study was conducted by the department of plastic surgery in collaboration with the department of ophthalmology from January 2020 to July 2021, spanning 19 months. The study received approval from the Institutional Ethical Committee, adhering to the guidelines established by the Indian Council of Medical Research in 1994 and the modified Helsinki Declaration of 2008. Written informed consent was obtained from all participants for both the research and clinical photography.

Study procedure

Twelve patients who met the inclusion criteria (Table 1) [15] underwent surgery within 3 weeks of trauma due to soft tissue edema. One patient underwent delayed reconstruction 9 months after the injury. Two patients presented with bilateral orbital fractures, while the others had unilateral fractures. The selection of patients with bilateral fractures was justified, as the contralateral fracture was minimally displaced and did not require surgery. They do not have associated soft tissue injuries, and their outcome does not affect the study inclusion criteria. Two patients with pan-maxillofacial injuries were treated using the standard fixation protocol for associated fractures.

Table 1.

Inclusion and exclusion criteria

Each patient underwent a clinical evaluation, and CT scans were performed following the Brainlab software protocol. This protocol specified a slice thickness of 1 mm and a field of view that included the top and back of the nose, eyes, forehead, and entire face, excluding the headrest and table. The pitch was kept below 2, using a reconstruction algorithm optimized for soft tissue. Primary and secondary parameters were documented both preoperatively and postoperatively at the 6-month follow-up. The primary outcomes assessed included orbital volume, diplopia, enophthalmos, and facial proportions. CT-based measurements of orbital volume and enophthalmos were calculated using the auto-segmentation tool and grid scale provided by the Brainlab software, with preoperative planning facilitated by iPLAN cranio-maxillofacial (CMF) software (Brainlab). In some instances, manual corrections to the orbital volume were necessary because the replica image produced by the auto-segmentation did not accurately reflect the bony contours of the orbital cavity. Facial dimensions were analyzed using clinical photography according to the specified protocol (Fig. 1).

Fig. 1.

Standard views: (A) basal, (B) cephalic, (C) frontal, (D) oblique. (E) Left lateral profile view. (F) Right lateral profile view. (G) Smiling lateral view.

Diplopia grading was conducted using a customized scale (Table 2). Photographs were standardized for pixel count and magnification. Measurements of the facial width-to-height ratio (FWHR) and the palpebral fissure length ratio were recorded using a grid scale in CorelDraw software (Alludo) (Fig. 2). Enophthalmos was clinically evaluated using a standard Hertel exophthalmometer. Secondary outcomes included fracture morphology and clinical signs of tissue entrapment, such as extraocular motility disorder, neuropraxia, relative afferent pupil defect, lid retraction, lid lag, dystopia, and infraorbital hypesthesia. Fracture morphology was assessed using CT scan DICOM images, and forced-duction tests were conducted to evaluate tissue entrapment both preoperatively and postoperatively. A comprehensive ophthalmic evaluation was performed by a single ophthalmology expert to measure both primary and secondary outcomes. A general clinical assessment of the patient was also documented. The surgical procedure was virtually planned by overlaying a mirrored image of the normal contralateral auto-segmented orbit onto the injured segment, thus establishing an ideal unilateral reconstruction parameter (Fig. 3). Templated landmarks facilitated navigation through the planned bony orbital contour, assessment of dissection, identification of the posterior ledge, and anatomical molding of the surgical implant. Immediate complications, such as tissue incarceration, infection, hemorrhage, and optic neuropathy, as well as late complications including lid malposition, retraction, infection, and implant instability, were recorded. The need for secondary surgery was also documented as a percentage.

Table 2.

Diplopia charting and customized grading in all cardinal gazes

Fig. 2.

Sample photographic analysis. Facial analysis in a front-profile photograph to calculate the facial width-to-height ratio and palpebral fissure length ratio.

Fig. 3.

(A) Orbital volume measurements and mirroring and superimposition of the contralateral orbit. (B) Enophthalmos calculation based on a computed tomography scan. Radiological evaluation of the globe position relative to the interzygomatic line using measurement tools in Brainlab software.

Operative steps

Patients underwent surgery conducted by a single surgeon to eliminate any surgical bias. Nine periorbital fractures were repaired using the subciliary approach, while two were addressed using the infraorbital approach due to the presence of a traumatic wound or scar (Fig. 4). Other associated fractures were managed according to the standard CMF trauma protocol.

Fig. 4.

(A, B) Surface registration and implant placement under navigation.

The surgical procedure involves discreetly placing a reference marker (using a skull-post with beads) along the hairline and identifying key soft tissue anatomical landmarks such as the nasion, nasal tip, and supraorbital rim. Fracture alignment was guided by 3D computer navigation through a trial-and-error approach. During surgery, a titanium mesh was shaped to match the anatomical profile as suggested by the iPLAN CMF planned data. In one instance, a STL model was utilized to shape the plate, while in other cases, the plate was manually contoured with the assistance of navigation, using feedback from a laser pointer. The implant was carefully positioned at the posterior ledge and secured at the orbital rim, ensuring the safety of adjacent vital structures, including the optic nerve and infraorbital nerve (Fig. 5).

Fig. 5.

Intraoperative feedback to check the reduction accuracy.

After securing the implant, the alignment of bone fragments was confirmed by tracing the reconstructed orbital margins with a handheld device, guided by the pre-planned surgical data. The possibility of soft tissue entrapment was re-evaluated using a forced-duction test. Anatomical closure was meticulously performed with minimal dissection to reduce the risk of unintended scarring. Patients were monitored for 2 days postsurgery to identify any immediate complications before being discharged on the 3rd or 4th day, with a follow-up appointment scheduled in the outpatient department.

RESULTS

Demographic characteristics: The median age was 29 years (interquartile range [IQR], 22.5–38.5 years). Of the 12 patients, 10 were men and two were women (Table 3).

Table 3.

Distribution of fracture morphology

Dystopia

Three patients had inferior displacement of the globe (Table 4).

Table 4.

Distribution of inferior dystopia of the globe

Complications

No major complications were observed during the clinical procedures. One patient who underwent delayed reconstruction experienced temporary symptom relief, but ophthalmic complaints returned at the 6-month follow-up. Lower lid retraction, ranging from 1 to 2 mm, was identified in five patients. Additionally, one patient reported implant impingement near the infraorbital margin. A case illustration is shown in Fig. 6.

Fig. 6.

Case illustrating a 38-year-old patient with marked correction in dystopia and facial morphological profile after three-dimensional navigation-assisted orbital reconstruction: (A) 2 weeks after injury and (B) 6 months after reconstruction.

DISCUSSION

The orbit is a pyramidal structure with a broad rectangular base and a narrow apex. It is composed of seven bones, each contributing to a specific anatomical zone that influences both its function and aesthetics [16]. The orbital floor is thin, concave in the anterior two-thirds, and convex in the posterior third, resembling an italic “S” in the sagittal plane. The medial wall, which is the weakest, is formed by the perpendicular ethmoid lamina; it is convex and runs parallel to the sagittal plane. In contrast, the lateral walls are thick, shorter, and converge at a 45° angle, being more recessed with the rim aligning with the globe’s equator.

Traumatic disruption of this intricate conformation can lead to diplopia, enophthalmos, and dystopia. Therefore, anatomical reconstruction of the orbital walls is essential to restore ocular function and facial symmetry. This reconstruction begins at the orbit’s margins, which provide the reference framework, and proceeds with the reconstitution of its walls.

Conventional reconstruction techniques lack objective criteria for achieving precise conformation. Extensive tissue dissection is necessary to visualize the posterior ledge in a limited space, which is often misinterpreted, leading to improper implant fixation. Additionally, even after employing STL models to achieve the desired contouring of the orbital implant, there is no intraoperative feedback available to confirm the accuracy of the anatomical reconstruction. These limitations are effectively addressed using 3D navigation technology.

Our study demonstrated that 3D computer navigation-guided orbital reconstruction precisely restores orbital volume to match that of the contralateral normal orbit, exhibiting a high linear correlation. The study group was heterogeneous, consisting of 75% of patients with impure orbital fractures that involved the orbital margins. The sample size of our study was comparable to that reported in the current published literature regarding outcomes in complex orbital fractures (Table 5).

Table 5.

Comparison of the literature to the present study

The normal orbital volume is approximately 30 mL, with the globe occupying 7 mL of this space. When orbital walls are fractured, the densely packed contents of the orbit can herniate. The accuracy of calculating orbital volume in a traumatized eye is compromised by difficulties in identifying depressed and comminuted fracture walls, leading to variations among observers. In this study, to minimize potential bias, automated measurements were conducted using the Brainlab software’s auto-segmentation tool, which generated a 3D replica of the orbital cavity. Our study revealed that navigation-guided surgery corrects orbital volume within a range of 0.45 cm 3 difference. The difference in volume between the reconstructed and normal orbits was found to be consistent with findings reported in the existing literature (Table 6).

Table 6.

Comparison of orbital volume restoration in various studies

Diplopia following orbital trauma is primarily caused by increased orbital volume, muscle incarceration in fracture segments, nerve paresis, and differential globe projection. The displacement of the anterior orbital margin and floor traps the inferior rectus, while changes in the posterior volume impact globe suspension and projection, leading to ocular motility disorders. Subperiosteal release of the entrapped tissue from the fractured margin and resuspension of the periorbita to the anatomically reconstructed walls effectively correct diplopia without recurrence. This study demonstrated a statistically significant improvement in diplopia following surgical intervention. However, this finding has limitations, as perimetry, the gold standard for diplopia charting, was not performed. Perimetry examination is challenging in cases of acute traumatic maxillofacial injury because it requires the patient to position their face within the examination machine. The diplopia grading used in this study was customized, as it is not possible to tabulate diplopia in all cardinal gazes for comparative analysis. Bly et al. [17] reported similar results in their series, which utilized a case-control study design with navigation-guided orbital reconstruction and developed their own diplopia severity score.

Asymmetrical projection of the eyeball significantly influences a person’s appearance when facing forward. Enophthalmos is traditionally defined as the posterior displacement of the eyeball into the bony orbit. Alterations in the bony structure beyond the globe’s transverse axis, specifically involving the lateral wall, medial wall, and floor, markedly impact the normal projection of the globe. Contributing factors also include long-term consequences of orbital fat atrophy, loss of ligament support (specifically the Lockwood ligament), and scar contracture [23,24]. In this study, there was no significant improvement in enophthalmos. Several factors contributed to this finding. The majority of patients in the sample had an impure fracture, and all were assessed using a Hertel exophthalmometer by a single observer. The presence of an associated zygoma fracture introduced a measurement error due to the displacement of the lateral orbital rim, which served as a false reference point. Consequently, a CT scan was performed for each patient to evaluate the globe’s position relative to the inter-zygomatic line (Cannabis Index) [25], which is subject to the same limitation. Clinically evident, refractory enophthalmos has been noted in numerous studies even after optimal orbital reconstruction. It has been observed that despite adequate restoration of orbital volume, enophthalmos may persist in long-term follow-ups of complex orbital trauma. This persistence can be attributed to orbital fat atrophy, scar contracture, fibrosis of the injured extraocular muscles, or inadequate soft tissue repair.

The significance of aesthetics in orbital fracture reconstruction is not explicitly addressed in the present literature. Orbital fractures involving the zygomatic bone can significantly disrupt facial symmetry due to the loss of malar projection. If the reduction is not anatomical, it can lead to an increased face width and a sunken eye appearance. Despite these challenges, gross facial and orbital symmetry was consistently achieved during the study, except in cases with significant soft tissue damage. Over time, facial morphological features undergo changes due to delayed wound retraction and tissue atrophy, adding a temporal dimension to these alterations. When the repair of the overlying soft tissue does not align anatomically with the underlying bone injury, facial asymmetry may result. Damage to the compartmentalized orbital fat pad can lead to sunken eyelid deformities. Similarly, loss of the buccal fat pad, whether from trauma or surgical intervention, tends to make the cheekbones and jawline more prominent. The breakdown of superficial and deep fascial planes, along with the retaining ligaments, negatively impacts facial contour and expression. A displaced buccal fat pad can cause noticeable sagging of the face, an accentuated nasolabial fold, and apparent flattening in the infraorbital area. For aesthetic reconstruction, it is crucial to mobilize the soft tissues subperiosteally and suture them back into their anatomical positions across the face. Subperiosteal sutures should be placed anteriorly to anchor the periorbita to the reconstructed inferior orbital rim. The aponeurotic layer of the dissected cheek tissue should be plicated and secured to the robust zygomatic periosteum. A layered closure with proper canthal anchoring is essential to match the dimensions of the opposite eyelid. Extensive dissection and inadequate lid suspension can lead to scleral show due to lower lid retraction. Therefore, only precise, goal-directed surgical dissection combined with anatomical reconstruction can ensure symmetrical outcomes. Current navigation technology primarily guides bony reconstruction, leaving soft tissue adjustments to the discretion of the surgeon. Accurate surgical planning complemented by intraoperative navigation may reduce the extent of surgical dissection and enhance aesthetic results compared to conventional approaches.

No re-intervention was required for any patient, and no major complications were encountered during the perioperative period or in the subsequent follow-up period of 6 months to 1 year. During the surgical procedure, navigation feedback was used to accurately identify the posterior ledge of the maxillary slope. However, the implant chosen for reconstruction had certain limitations. Patients with complex orbital fractures involving more than two walls and comminuted fractures require adequate posterior bony support for contoured implants to prevent delayed sagging. Due to financial constraints, a patient-specific implant was not affordable; therefore, manual molding of the implant or pre-contoured mesh plate was employed to reconstruct the deficient walls. This approach may lead to human errors in judgment when matching the anatomical profile of the orbit. The use of a patient-specific implant and further validation by intraoperative CT scan would have objectively matched the reconstructed orbit, thereby eliminating registration and operative errors [26,27].

The study had some flaws, including the absence of a comparable control group, the use of a universal orbital implant, and undefined fracture morphology in relation to the orbital axis. Additionally, the proposed study group was subject to measurement errors, selection bias, and normal anatomical variance of the orbit. To establish a standardized treatment protocol, further randomized controlled trials with larger sample sizes and extended study periods are necessary.

Over the past two decades, navigation-guided surgery has gained widespread acceptance in various reconstructive procedures, though it still falls short in some critical areas. To improve predictability and outcomes, it is necessary to integrate real-time imaging and artificial intelligence for image analysis during surgery. This integration can enhance the clinical knowledge and skills of less experienced surgeons, bringing them closer to the level of expertise held by seasoned professionals in the field. A significant limitation of navigation technology is its inadequate guidance for soft tissues and the reliance on fixed registration reference points in cases involving mobile structures like the mandible. Furthermore, the assumption of facial symmetry is often incorrect, as research has demonstrated subtle variations in orbital volumes between the two orbits of an individual [28]. Despite concerns regarding cost and time, the superior outcomes and reduced revision rates provided by navigation surgery justify its use, especially in complex facial injuries caused by high-velocity trauma.

Article information Continued