Surgical precision and clinical outcomes of 3D preformed titanium mesh in orbital blowout fracture reconstruction: a comparative study from Thailand
Article information
Abstract
Background
Orbital floor fractures are frequently associated with head trauma, and the complex three-dimensional (3D) structure of the orbital cavity poses a significant challenge for accurate repair of orbital wall defects. Accordingly, this study aimed to compare preoperative titanium mesh bending using 3D-printed models with conventional intraoperative bending techniques, focusing on the accuracy of orbital geometry reconstruction and intraoperative efficiency, with the goal of determining potential advantages in optimizing surgical outcomes.
Methods
This study presents a comparative analysis between a prospective preoperative bending (3D pre-bent) of titanium mesh and retrospective data from conventional intraoperative bending. Clinical and radiographic evaluations of orbital geometry, along with intraoperative efficiency metrics, were assessed at both preoperative and postoperative stages.
Results
Sixty-four patients were included, with 32 in the 3D pre-bend titanium mesh group and 32 in the conventional bending group. Mean operative time was shorter in the 3D pre-bend group (81.07± 13.04 minutes) than in the conventional group (96.07± 4.46 minutes). The 3D pre-bend group achieved more accurate reconstruction, with orbital volume and height correction rates of 96.78% ± 4.91% and 100.62%± 3.77%, respectively, versus 94.88% ± 4.33% and 96.08% ± 5.44% in the conventional group. Patient and surgeon satisfaction was higher in the 3D pre-bend group.
Conclusion
Pre-bending titanium mesh using a 3D-printed model significantly reduces intraoperative time and improves orbital height correction. Additionally, this technique enhances both patient and surgeon satisfaction. However, long-term clinical outcomes remain comparable between groups.
INTRODUCTION
Maxillofacial injuries represent a significant global health concern, with their epidemiology varying by region. The estimated global incidence of maxillofacial fractures ranges from 20 to 60 cases per 100,000 individuals annually [1,2]. The etiology of these injuries differs across geographical areas. In Asia and Africa, maxillofacial fractures are predominantly attributed to traffic accidents, particularly in regions with inadequate traffic regulations. Conversely, in Europe and North America, assaultrelated fractures are more prevalent. In Latin America, interpersonal violence and sports-related injuries are the leading causes of maxillofacial trauma [3,4]. Orbital floor fractures are broadly categorized into pure blowout fractures, which involve isolated fractures of the orbital floor, and impure blowout fractures, which are associated with orbital rim fractures and often occur in conjunction with midfacial fractures, such as those involving the zygomaticomaxillary complex. Blowout fractures typically result from a sudden increase in infraorbital pressure, leading to fractures of the orbital floor or medial wall. The incidence of orbital floor fractures ranges from approximately 4% to 16% of all maxillofacial fractures, with variations depending on the study population and the predominant etiologies of facial trauma in different regions [5,6].
The primary objective in the treatment of orbital floor and medial wall fractures is to select an appropriate material for defect coverage, thereby preventing the descent of orbital contents into the maxillary or ethmoidal sinuses and ensuring the restoration of orbital volume. Orbital floor reconstruction can be achieved using autologous grafts, resorbable alloplastic materials, or non-resorbable alloplastic implants. Ideal reconstruction materials should possess key characteristics, including resistance to infection, minimal tissue reactivity, and cost-effectiveness [7,8]. Titanium mesh is widely recognized as a standard material for orbital floor reconstruction due to its favorable biomechanical properties and global acceptance. Its key advantages include excellent biocompatibility, a low extrusion rate, and the ability to be easily contoured to fit the patient’s orbital anatomy. However, intraoperative challenges may arise, particularly when using a transconjunctival approach, which provides limited surgical exposure. This restriction can make precise intraoperative contouring of the titanium mesh challenging, often resulting in prolonged operative time to achieve an optimal fit tailored to the patient’s anatomy [9-11].
Three-dimensional (3D) simulation and printing technology have gained widespread recognition and are increasingly being integrated into various medical fields, particularly in craniofacial surgery [12-15]. The authors aim to investigate the application of preoperative bending techniques by utilizing 3D-printed models for the pre-contouring of titanium mesh, comparing this approach to conventional intraoperative bending. The study will assess the impact of preoperative bending on key parameters, including intraoperative efficiency, as measured by operative time, and surgeon satisfaction, with the goal of determining its potential advantages in optimizing surgical outcomes.
METHODS
Study design and setting
This study is a comparative analysis of two patient groups: one undergoing prospective preoperative bending (pre-contouring) of titanium mesh and the other receiving conventional intraoperative bending. The prospective cohort was recruited from January 2024 to December 2024, while retrospective data were extracted from medical records spanning January 2023 to December 2023. The study was conducted at Phramongkutklao Hospital, Bangkok, Thailand.
Ethical approval
The study received approval from the Royal Thai Army Medical Department and the Ethics Committee (IRBRTA 657/2567) and was registered with the Thai Clinical Trials Registry (TCTR 202503260006). Written informed consent was obtained from all participants for study inclusion and publication of photographic documentation.
Patient selection and sample size calculation
This study included patients aged ≥20 years diagnosed with blow-out fractures who underwent orbital floor reconstruction using titanium mesh. Patients with concurrent maxillofacial fractures, a history of prior orbital surgery on the affected side, or underlying bleeding disorders were excluded. The sample size was determined using a clinical superiority design formula.
Orbital floor and orbital fracture area
The area of each orbital floor was calculated using a two-dimensional computed tomography (CT)-based method by measuring distances from contiguous coronal CT slices. The ethmoid-maxillary suture was identified as the medial anatomical reference point for all subsequent measurements. The extent of the fracture was determined by measuring the distance from this reference point to the medial and lateral fracture boundaries in each CT slice where the fracture was visible [16]. The length of the orbital floor in each slice was quantified by measuring the distance from the medial anatomical landmark to the lateral aspect of the orbital floor. The anterior–posterior extent of the fracture was established by correlating the CT slices displaying the fracture with those delineating the anterior and posterior orbital floor borders. The infraorbital rim was designated as the anterior border, while the posterior border was defined by the tip of the orbital pyramid. The width of the orbital floor or fracture was multiplied by the CT slice thickness (1.0 mm) to estimate the cross-sectional area. By summing the areas obtained from each slice, the total orbital floor area and total fractured area were determined.
Orbital volume measurements
The orbital volume is defined as the space extending from the outer orifice of the optic canal at the orbital apex to the anterior orbital rim. Accurate measurement of orbital volume is essential for assessing the extent of orbital fractures and evaluating postoperative reconstruction outcomes [17]. In this study, orbital volume was calculated using a CT-based volumetric analysis. The process involved measuring the cross-sectional area of each CT scan slice containing the orbit. Each individual section was analyzed to determine its respective area, which was then multiplied by the CT slice thickness to obtain the volume contribution of that particular section. The total orbital volume was subsequently derived by summing the volumes of all slices extending from the orbital apex to the anterior rim. This method ensures a precise and reproducible assessment of orbital volume, allowing for direct comparisons between the fractured and contralateral healthy orbits (Fig. 1) [18].
Computed tomography image showing the measurement of orbital volume and orbital height. (A) Preoperative and (B) postoperative images.
Orbital height measurements
Orbital height was measured using a specialized algorithm that relied on bony reference points unaffected by the fracture to ensure accurate and reproducible measurements. In the sagittal plane, the measurement process began laterally, selecting the first slice in which the superior orbital fissure became visible. The total orbital floor length, extending from the infraorbital rim to the posterior ledge, was measured in this plane and subsequently bisected. At the midpoint of the orbital floor, the image plane was reoriented to a coronal view. In the coronal plane, orbital height was measured at three locations: the central region, as well as 5 mm medial and 5 mm lateral to the center. To ensure consistency and comparability, matching sequences were identified across preoperative, fractured, and postoperative reconstructed orbits so that measurements were obtained from identical image sections (Fig. 2).
Three-dimensional (3D)-printed model generated from the mirror image of the unaffected side to simulate the orbital floor of the affected side. (A) Mirror from the unaffected side. (B) Fracture side. (C) 3D printing model of mirror from the unaffected side. (D) 3D printing model of fracture side.
The reconstruction rates of orbital volume and orbital height were determined based on the method described by Wi et al. [19], using the formula: [1−(A−B)/B]×100%, where A represents the reconstructed orbital volume or height, and B corresponds to the volume or height of the contralateral, non-traumatized orbit. This approach allowed for a standardized assessment of reconstruction accuracy by quantifying the extent to which postoperative orbital dimensions approximated the normal anatomy (Fig. 1).
Model 3D preparations
High-resolution craniofacial CT scans (≤1 mm slice thickness) were exported in DICOM (Digital Imaging and Communications in Medicine) format and processed using Mimics Research v24.0 (Materialise NV) for 3D reconstruction. Segmented bone models were converted to STL files and refined using 3-matic v16.0 for virtual surgical planning and contour correction. The mirrored unaffected orbit served as a reference for reconstructing the affected side. Finalized models were printed using a desktop 3D printer (Creality 10S Pro) with polylactic acid filament, a biocompatible material, at a layer thickness of 0.1 mm and 100% infill density. After printing, each model was cleaned, ultraviolet-light cured, and sterilized before use for preoperative simulation and pre-bending of titanium mesh to ensure precise anatomical adaptation during orbital reconstruction (Fig. 2).
Implant preparations
Preoperatively, standard flat titanium orbital fixation plates with a thickness of 0.2 mm (OsteoMed) were contoured to conform to the shape of the mirrored unfractured orbital model and subsequently evaluated on the fractured-side model. Necessary adjustments were made to optimize the fit, ensuring precise adaptation to the orbital wall and complete reconstruction of the defect. Excess material was trimmed as required to achieve an optimal anatomical fit (Fig. 3). The 3D-printed models were sterilized through autoclaving following the manufacturer’s guidelines, while the titanium plates underwent standard sterilization protocols prior to surgical implantation.
Prebending of titanium mesh using a three-dimensional (3D)-printed model of the mirror image of the unaffected side.
Intraoperative procedures
Sterilized 3D-printed models and preformed titanium plates were positioned on the back table. Standard surgical approaches to the orbit were utilized, and the orbital fracture walls were fully exposed using a subperiosteal dissection. Any herniated or entrapped orbital soft tissue was manually reduced. The preformed titanium plate was then placed within the orbit, and its fit was assessed. Adjustments were made as necessary to ensure optimal anatomical conformation. Once satisfactory positioning was achieved, the plate was secured to the orbital rim using titanium screws, followed by layered closure of the periorbital tissues. In the conventional group, the preoperative modeling process was omitted. Instead, intraoperative contouring of the titanium implant was performed using a freehand technique, and placement was carried out without the aid of a 3D-printed model (Fig. 4).
Intraoperative orbital floor augmentation using a prebend titanium mesh and a three-dimensional (3D)-printed guide model. (A) Prior to placement of the titanium mesh. (B) Following the placement of the titanium mesh.
Clinical outcome measurements
Patients underwent comprehensive clinical evaluations preoperatively, at 1 week postoperatively, and subsequently at 1, 3, and 6 months. Assessments included measurement of enophthalmos (>2 mm) using a Hertel exophthalmometer, evaluation of diplopia, limitations in ocular gaze motility, and midface paresthesia. Satisfaction scores (0–10) were recorded by the operating surgeon on the day of surgery and by patients at 1 month postoperatively. Radiographic assessment involved noncontrast CT of the facial bones, performed preoperatively and at 1 week postoperatively. CT scans were acquired with a slice thickness of 1.0 mm to evaluate the extent of the orbital wall defect. Orbital volume measurements were compared between the fractured and non-fractured orbits to assess the accuracy of reconstruction.
Statistical analyses
Statistical analyses were conducted using the Stats Kingdom online platform (https://www.statskingdom.com/index.html). For comparisons of demographic variables and outcome parameters between groups, the two-sample t-test calculator (Welch t-test) was utilized. A p-value <0.05 was considered statistically significant. Patient and surgeon satisfaction scores were analyzed and reported as percentages.
RESULTS
Patient demographics and injury characteristics
A total of 64 patients were equally divided into two groups: 3D pre-bent titanium mesh (n=32) and conventional intraoperative bending (n=32). The mean age was 43.71±18.96 years in the 3D pre-bent group and 40.93±19.54 years in the conventional group. Average weight, height, and BMI were 70.14± 14.70 kg, 166.14±7.35 cm, and 25.25±4.05 kg/m² in the 3D pre-bent group, compared to 69.00±8.08 kg, 169.57±8.10 cm, and 24.13±3.50 kg/m² in the conventional group. Most participants were female, accounting for 19 patients (59.4%) in the 3D pre-bent group and 21 patients (65.5%) in the conventional group. The side of injury was evenly distributed in the 3D prebent group, with 16 cases (50.0%) on each side. In the conventional group, 17 cases (53.1%) involved the right side and 15 cases (46.9%) involved the left side. No statistically significant differences were observed between groups (Table 1).
Demographic characteristics
Operative outcomes
The mean orbital floor area was 4.62±0.70 cm² in the 3D prebent titanium mesh group and 4.30±0.69 cm² in the conventional group, with no significant difference. However, the orbital floor defect was significantly larger in the 3D pre-bent group (2.23±0.55 cm²) than in the conventional group (1.81±0.41 cm²). The mean time from trauma to surgery was 9.93±5.92 days in the 3D pre-bent group and 11.57±5.54 days in the conventional group, with no significant difference. The transconjunctival approach was used in 18 cases (56.3%) in both groups. The subciliary approach was performed in nine cases (28.1%) in the 3D pre-bent group and eight cases (25.0%) in the conventional group. The midtarsal approach was used in five cases (15.6%) in the 3D pre-bent group and six cases (18.7%) in the conventional group (Table 1).
Clinical and radiologic outcomes
Operative time was significantly shorter in the 3D pre-bent group compared with the conventional group (81.07±13.04 minutes [73.22–89.34] vs. 96.07±4.46 minutes [94.03–98.55], respectively).
The length of hospital stay was comparable between groups, with a mean duration of 5.21±3.45 days in the 3D pre-bent group and 5.50±2.88 days in the conventional group. Preoperative findings and postoperative outcomes assessed at 1 week, 1 month, 3 months, and 6 months showed no statistically significant differences between the two groups (Table 2).
Clinical and outcome measurement
In both groups, the preoperative orbital volume difference relative to the healthy orbit was increased, measuring 1.67±0.86 cm³ in the 3D pre-bent group and 1.48±0.89 cm³ in the conventional group. Postoperatively, this difference decreased following orbital floor reconstruction, with values of 0.54±0.85 cm³ in the 3D pre-bent group and 0.88±0.77 cm³ in the conventional group, though the difference between groups was not statistically significant. Similarly, the preoperative orbital height difference relative to the healthy orbit was 0.36±0.25 cm in the 3D pre-bent group and 0.32±0.21 cm in the conventional group, with no significant intergroup difference. However, postoperative orbital height correction was significantly greater in the 3D pre-bent group (0.02±0.01 cm) compared to the conventional group (0.13±0.05 cm; p=0.014), indicating improved accuracy with the 3D pre-bending technique (Table 3).
Orbital parameter measurements
In the 3D pre-bent group, the orbital volume correction rate closely approximated the healthy orbital volume at 96.78%±4.91% (95% CI, 94.45–98.67), compared to 94.88%±4.33% (95% CI, 93.78–96.23) in the conventional group, with no statistically significant difference between the groups. Conversely, the orbital height correction rate was significantly higher in the 3D pre-bent group, achieving 100.62%±3.77% (95% CI, 98.12– 102.46) to 96.08%±5.44% (95% CI, 93.76–99.23) in the conventional group, indicating a statistically significant improvement in orbital height reconstruction with the 3D pre-bending technique (Table 4).
Reconstruction rates of the orbital volume and height
Satisfaction outcomes
Patient-reported satisfaction was higher in the 3D pre-bent titanium mesh group, with a mean score of 8.79±0.97 compared to 7.93±0.73 in the conventional group. Similarly, surgeon satisfaction was greater in the 3D pre-bent group, with a mean score of 9.64±0.50 versus 8.43±0.64 in the conventional group, indicating a preference for the pre-bending technique in both patient and surgeon assessments (Table 5). Example case of 3D prebend orbital floor augmentation (Fig. 5).
Patient and surgeon satisfactions
A 72-year-old woman presented with left enophthalmos, limited upward gaze, and diplopia following a fall. Clinical examination and imaging confirmed a left pure orbital floor fracture. The patient underwent orbital floor reconstruction using a pre-bent titanium mesh designed from a three-dimensional (3D)-printed orbital model to achieve precise anatomical restoration. Preoperative documentation included a clinical photograph showing ocular asymmetry and enophthalmos (A). The titanium mesh was pre-contoured using a patient-specific 3D-printed orbital model (C) and subsequently placed intraoperatively to reconstruct the orbital floor (D). Preoperative documentation also included coronal (E), sagittal (G), and 3D computed tomography (CT) scans (I) that demonstrated the extent of the orbital floor defect. At 1-month follow-up, the patient exhibited significant improvement in globe position and ocular motility, with resolution of diplopia. This was evident clinically (B) and confirmed radiologically by postoperative coronal (F), sagittal (H), and 3D CT imaging (J), all demonstrating satisfactory orbital floor restoration and implant positioning.
DISCUSSION
The orbital floor is the most commonly fractured isolated or-bital bone in maxillofacial trauma. Additionally, fractures of the zygomaticomaxillary complex are frequently associated with orbital floor fractures [20,21]. Pure blowout fractures are often linked to low-impact trauma in the orbital and cheek regions, commonly resulting from physical assault, sports-related injuries, or low-speed vehicular accidents. The clinical presentation of pure blowout fractures varies depending on the degree of globe displacement and the extent of the bony defect. Patients may be asymptomatic, with fractures detected only through imaging, or may exhibit severe globe displacement, leading to enophthalmos, diplopia, and restricted ocular movement due to extraocular muscle entrapment [22,23].
The primary goal of treatment for orbital floor fractures is to restore both form and position. Augmentation with additional materials is required when the defect exceeds 1 cm², utilizing either allografts or alloplastic materials [24,25]. Among alloplastic options, titanium is widely accepted as a standard material for orbital floor reconstruction. However, the complex 3D anatomy of the orbital cavity and the limited surgical field present significant challenges in achieving precise defect repair. Inadequate reconstruction may lead to complications such as diplopia, restricted ocular movement, enophthalmos, or exophthalmos [26]. The conventional approach to orbital floor reconstruction involves intraoperative bending of titanium mesh to fit each patient’s anatomy. However, complete exposure of the orbital floor, particularly the posterior region with its conical shape, is often challenging, potentially compromising the accuracy of reconstruction. Advances in technology, including virtual surgical planning and 3D printing, have enabled more precise and innovative techniques for orbital wall reconstruction, improving surgical outcomes. Orbital floor reconstruction using virtual planning software and 3D-printed facial bone models has demonstrated effective and satisfactory outcomes [27,28]. Maxillofacial, otorhinolaryngologic, and plastic surgeons have also utilized computer-assisted navigation systems, which employ a mirror image of the intact contralateral orbit to guide reconstruction [29]. While these advanced techniques offer significant clinical advantages, their widespread adoption remains limited due to high costs and extended manufacturing times. Therefore, a surgeon’s thorough understanding of normal orbital floor anatomy remains essential for achieving optimal reconstruction outcomes.
In this study, the 3D pre-bent titanium mesh group was designed using CT scan data to generate 3D-printed models of the mirrored, unaffected orbit, requiring no further modifications. Some demographic differences were observed between study groups due to variations in enrollment timing. Operative time was significantly shorter in the 3D pre-bent group; however, hospital length of stay showed no significant difference. The reduced operative time may benefit patients by minimizing anesthesia exposure and promoting faster recovery. The unchanged hospital stay duration can be attributed to the institutional practice of discharging patients only after postoperative swelling has subsided. During clinical evaluation and followup, no significant differences were observed between the 3D pre-bent titanium mesh group and the conventional intraoperative bending group. However, V2 paresthesia appeared to resolve more rapidly in the 3D pre-bent group (Table 2), possibly due to differences in surgical approach. In most cases, the 3D pre-bent group underwent a transconjunctival approach, whereas the conventional group predominantly received external skin incisions, including subciliary and midtarsal approaches. At long-term follow-up (6 months), no postoperative complications such as enophthalmos, diplopia, restricted upward gaze, or persistent paresthesia were observed.
Orbital parameter measurements included orbital volume and orbital height. The results demonstrated significantly greater accuracy in orbital height correction in the 3D pre-bent group. This finding may be attributed to the limitations of the conventional method, where achieving complete adaptation of the titanium mesh to the orbital floor is challenging. Consequently, the 3D pre-bent group achieved an orbital height correction rate of 100.62%±3.77%, closely approximating the healthy orbital height, compared to 96.08%±5.44% in the conventional group, a statistically significant difference. However, no significant differences were observed between groups in orbital volume correction. Satisfaction scores were significantly higher in the preoperative 3D pre-bent group from both patient and surgeon perspectives. Patients reported greater satisfaction due to reduced anesthesia-related complications and the absence of visible scarring from the transconjunctival approach. Surgeons favored the 3D pre-bent technique due to its advantages in preoperative planning, as the 3D-printed model facilitated patient education and informed consent. Additionally, the pre-bent titanium mesh improved intraoperative efficiency by aiding precise placement over the orbital wall defect, contributing to a shorter operative time.
Regarding the cost and processing time of the 3D model in the 3D pre-bent group, the authors observed that the in-house 3D printing process, including the fabrication of the model, pre-bending of the titanium mesh, and sterilization, required approximately 4 to 6 hours per case. Additionally, the material cost for producing the 3D-printed model and pre-bent titanium mesh was estimated to be approximately 15 USD per case, reflecting the financial and logistical considerations associated with implementing this technique in clinical practice. One of the primary limitations of this study is its mixed-methodology design, in which the retrospective nature of the control group may introduce potential bias. Future research should focus on minimizing bias by employing more rigorous study designs, such as randomized controlled trials [30], and incorporating a more diverse patient population across different ethnicities and geographical regions to enhance the applicability and validity of the findings in global clinical practice.
The findings of this study indicate that preoperative contouring of titanium mesh using patient-specific 3D printed anatomical models allows surgeons to achieve more accurate anatomical reconstruction than conventional intraoperative bending techniques. This approach also results in a significant reduction in operative time and yields higher satisfaction among both patients and surgeons. However, the long-term clinical outcomes remain comparable between the two methods.
Notes
Conflict of interest
No potential conflict of interest relevant to this article was reported.
Funding
None.
Acknowledgments
The completion of this undertaking was not possible without the participation and assistance of all patients and the many people whose names may not be enumerated. To all families, relatives, and others who, in one way or another, shared their support, either morally, financially, or physically, we thank you.
Ethical approval
The study received approval from the Royal Thai Army Medical Department and the Ethics Committee (IRBRTA 657/2567).
Patient consent
All participants provided written informed consent for the publication of the case and related images.
Author contributions
Conceptualization: Chatchai Pruksapong. Data curation: Chatchai Pruksapong, Navin Klongkitkarnkha, Akaradech Attainsee, Chairat Burusapat, Suttisun Jankajorn, Nutthapong Wanichjaroen, Nuttadon Wongprakob, Rapeepat Sapruangthong, Pantip Suwansa-ard. Formal analysis: Chatchai Pruksapong, Navin Klongkitkarnkha. Methodology; Project administration; Visualization: Chatchai Pruksapong. Writing-original draft : Chatchai Pruksapong, Navin Klongkitkarnkha. Writing–review & editing: Chatchai Pruksapong. All authors read and approved the final manuscript.
Abbreviations
3D
three-dimensional
CT
computed tomography