A Novel 3D-Printed Titanium Prosthesis with Osseointegrated Dental Implants for Reconstruction of Segmental Mandibular Defects

Abstract

Introduction

Segmental mandibular defects present complex reconstructive challenges and are associated with significant functional and aesthetic morbidity. While vascularized bone flaps remain the standard for reconstruction, they are associated with donor-site morbidity, prolonged operative times, and limitations in predictable dental rehabilitation. This proof-of-concept study evaluated the mechanical feasibility, in a synthetic bench-top model, of a patient-specific, 3D-printed titanium mandibular prosthesis incorporating conventional dental implants for both dental rehabilitation and biological anchorage via osseointegration.

Methods

The prosthesis was digitally designed and fabricated in Ti6Al4V-ELI titanium alloy by powder-bed fusion additive manufacturing. A 25 mm critical-sized segmental defect was created in synthetic polyurethane mandible models, and prototype constructs were assembled with osteosynthesis screws and conventional dental implants engaging the residual mandible. Static load-to-failure testing was performed on a single prototype construct. Dynamic cyclic loading at 40–60 N and 2 Hz for 500,000 cycles was performed on three prototype constructs and a control reconstructed with a conventional 2.5 mm reconstruction plate under an identical protocol.

Results

The prosthesis demonstrated accurate fit and reproducible assembly. Static testing resulted in failure at 213.9 N, occurring within the mandible model rather than in the prosthesis, fixation screws, or implant-prosthesis interface. All prototype constructs and the control completed 500,000 cycles without catastrophic failure, visible deformation, or loosening of fixation screws. The change in per-cycle displacement range over testing was small in all constructs and showed no progressive increase, indicating an absence of appreciable plastic deformation. Reverse torque testing demonstrated no implant loosening.

Conclusion

This proof-of-concept study supports the mechanical feasibility of a novel 3D-printed titanium mandibular prosthesis under defined loading conditions in a synthetic bench-top model. Osseointegrated dental implants are intended to provide biological anchorage in addition to mechanical fixation, although osseointegration was not assessed in this model. This design concept warrants further translational investigation.

Introduction

Mandibular segmental defects arise most commonly from oncologic resection, trauma, and advanced inflammatory disease. Loss of mandibular continuity results in marked impairment of mastication, speech, swallowing, airway support, and facial aesthetics, leading to a significant reduction in quality of life. The main goals of reconstruction are to restore form, maximize function and quality of life while minimizing morbidity.1

Microvascular reconstruction using vascularized bone flaps has become the standard of care, with the fibula free flap being recognized as the gold standard.2 Despite excellent reported success rates, fibula reconstruction is associated with inherent limitations including donor-site morbidity, geometric mismatch with the native mandible, prolonged operative times, and challenges in dental rehabilitation.3 While the use of dental implants in mandibular reconstructions with fibula flaps has been reported as early as 1989, there are several challenges in achieving predictable dental rehabilitation including unfavorable crown–implant ratios, compromised emergence profiles, and increased susceptibility to peri-implant disease.4–9

Recent developments in virtual surgical planning, computer-aided design/computer-aided manufacturing (CAD/CAM) technologies, and additive manufacturing have enabled the design and fabrication of patient-specific implants for mandibular reconstruction. Custom reconstruction plates designed to fit the patient’s anatomy allow customization to replicate the complex shapes and curves of the mandible.10,11 While custom reconstruction plates improve accuracy and have been shown to have higher stability and rigidity than manually bent conventional reconstruction plates, they still require the use of a bone graft.11–18

Patient-specific mandibular prostheses fabricated from titanium alloys offer an alternative approach that may obviate the need for bony reconstruction altogether. Various designs have been proposed from custom titanium mesh trays, to endoprostheses and 3D titanium prostheses incorporating different design elements.19–27 Published reports of 3D titanium prostheses largely involve extended temporomandibular joint replacements (eTJR), which have different biomechanics from segmental defects. In an eTJR, the load is transferred through a condylar component articulating against a fossa component across an unconstrained contact interface, with the only rigid interface being where the prosthesis is anchored to the mandibular body or ramus.28,29 A segmental prosthesis is instead loaded through an intact native temporomandibular joint and transfers the bending and torsional forces of mastication across two bony interfaces, producing a more complex and less predictable distribution of stress.30–32 A few reports on 3D titanium prostheses incorporating dental implants for the reconstruction of segmental defects have been published but still rely on mechanical anchorage of osteosynthesis screws and often require the use of bone grafts.33–36

This study proposes a novel patient-specific, 3D-printed titanium mandibular prosthesis for the reconstruction of segmental mandibular defects incorporating conventional dental implants for both biological anchorage via osseointegration and functional dental rehabilitation. The biological anchorage of osseointegrated implants is hypothesized to provide long-lasting, stable anchorage with the potential for supporting bone to adapt to functional stresses over time. The primary aim of this proof-of-concept study was to evaluate the mechanical feasibility of the proposed design using bench-top biomechanical testing in preparation for further translational work.

Materials and Methods

This proof-of-concept study consisted of two main phases:

1. Design and fabrication of a patient-specific mandibular prosthesis.

2. Biomechanical evaluation of the prototype constructs using a synthetic mandible model.

Design and Fabrication of Prosthesis

The prosthesis was conceptualized as a 3D-printed titanium construct designed to reconstruct a critical-sized segmental mandibular defect (Figure 1). The design incorporated a few key features including a reduced height of the mandible prosthesis relative to the alveolus to facilitate soft tissue closure, proximal and distal reconstruction plate elements for initial mechanical stability, and proximal and distal bony ledges over healthy native mandible for axial support and incorporating channels for the insertion of osseointegrated implants to engage the native mandible.

Figure 1.

Conceptual design of mandibular prosthesis (not drawn to scale). The prosthesis is shown in red, osseointegrated implants shown in blue and mandible outlined in black.

The stereolithography (STL) file of synthetic polyurethane mandible model (Synbone AG, Davos, Switzerland) was digitally modified to create a segmental defect within the mandibular body measuring 25 mm anteroposteriorly with additional 8 mm wide offset cuts to receive the ledges of the prosthesis, leaving 15 mm of residual mandibular height at the resection margins to receive the osseointegrated implants (Figure 2). The prototype prosthesis was designed using CAD software (Geomagic Freeform Plus, Hexagon AB, Stockholm, Sweden) in accordance with the design concepts (Figure 3).

Figure 2.

Virtual surgical planning of the segmental defect on the mandible model measuring 25 mm in length with 8 mm offset cuts anteriorly and posteriorly, preserving 15 mm of residual mandibular height at the resection margins.

Figure 3.

CAD design of prototype prosthesis.

The prosthesis was fabricated in medical-grade Ti6Al4V-ELI titanium alloy via powder-bed fusion additive manufacturing. A combination of micro-rough and 1 mm-spaced lattice surfaces was used, with the lattice surface at bone contacting surfaces to facilitate bony ingrowth, and micro-rough surfaces at soft tissue contacting surfaces to facilitate soft tissue growth and adhesion.37,38

Commercially available dental implants (Standard Plus SLA, Ø4.1 mm, Straumann Group, Basel, Switzerland) were selected. Implant angulation was planned to balance axial loading, prosthodontic emergence, and prosthesis thickness. Implant length at each site was determined during virtual planning by selecting the longest available implant that could be fully contained within the bony envelope of the mandible model without perforation of the cortex. Although 15 mm of mandibular height was preserved at the resection margins, the taper of the mandible toward the inferior border, combined with the planned implant angulation, restricted the depth to which the implant could be placed while remaining completely within bone. A 14 mm implant was therefore used anteriorly and a 12 mm implant posteriorly.

The defect was created in polyurethane mandible models (Synbone AG, Davos, Switzerland) using a custom designed cutting guide. The mandible models were obtained from the same batch to improve consistency. The models are fabricated from polyurethane foam with a cancellous inner core and a harder outer shell simulating cortical bone. This mandible model was chosen due to its anatomical shape and comparable biomechanical properties to cadaveric bone as described in other biomechanical studies.39–42

The prosthesis was then assembled and secured to the mandible model with bicortical osteosynthesis screws (2.4 mm Titanium MatrixMandible locking screws, self-tapping, 14 mm, Synthes, Paoli, PA, USA) (Figure 4). Dental implant osteotomies were performed using sequential standard drills from the manufacturer (Straumann Group, Basel, Switzerland), and the dental implant was inserted into the prepared osteotomy site using a torque wrench until the polished collar of the implant contacted the prosthesis surface. Implant osteotomies were intentionally underprepared to a final diameter of 3.5 mm relative to the 4.1 mm implant diameter, in accordance with conventional dental implant placement protocol, to optimize primary stability. Insertion torque exceeded 35 N cm for all implants.

Figure 4.

Assembled prototype prosthesis with mandible model.

Biomechanical Evaluation of Prototype

Biomechanical testing was performed using a universal testing machine (Instron ElectroPuls E1000, Instron, Norwood, MA, USA). A custom fixture jig was used to secure the mandible model at the bilateral condyles, and the load was applied at the mandibular incisal region (Figure 5a).

Figure 5.

(a) Setup of biomechanical testing with prototype specimen mounted on custom jig and loading anvil at incisors. (b) Attachment of EM sensors to implant adaptors and the reference position.

Static load-to-failure testing was performed on a single prototype model. The specimen was preloaded with a force of 10 N at the central incisors and increased at a rate of 10 N/s until failure as indicated by a sudden increase in displacement of the anvil or structural collapse of the specimen.

Dynamic cyclic testing was performed on three prototype specimens and one control specimen reconstructed with a 2.5 mm manually contoured conventional reconstruction plate (matrixMandible 12-hole reconstruction plate, 2.5 mm, Synthes, Paoli, PA, USA) secured with the same osteosynthesis screws. A 20 N preload at the central incisors was applied to stabilize the specimen, followed by cyclic loads of 40–60 N at 2 Hz for 500,000 cycles continuously without interval. The loading protocol was selected to reflect reported masticatory function after segmental mandibular resection rather than physiological bite force in an intact mandible. Marunick et al reported anterior occlusal forces of 34–50 N in patients following partial mandibulectomy with reconstruction, persisting up to 20 months postoperatively, and reduced bite force following mandibular surgery is well documented and attributed to muscle splinting, surgical division of the muscles of mastication, and loss of muscle attachment.43,44 A frequency of 2 Hz approximates the physiological chewing rate. The target of 500,000 cycles was selected to represent a sustained period of functional loading well beyond the critical early healing interval during which osseointegration is established. The specimen was visually inspected at the end of the testing for any signs of failure such as cracks, deformation or loosening of fixation screws. The dental implants were also checked for loosening with a reverse torque test.45 The displacement of the loading anvil was recorded throughout the testing and the change in displacement between cycles was compared as a proxy for assessing plastic deformation of the specimen with an increase in displacement between cycles indicating plastic deformation. The stability of the implants was assessed using electromagnetic (EM) sensors (trakSTAR, Ascension Technology Corporation, Burlington, VT, USA) secured to each implant adaptor and a reference sensor placed into a prepared slot in the last tooth in the mandible model to compare the relative motion of each implant with each other and the reference (Figure 5b). This reference point was chosen to avoid compromising the integrity of the model and as close to the implants as possible to minimize the effect of flexure of the mandible model on the relative positions.

The number of specimens was limited (static testing, n = 1; dynamic testing, n = 3 prototype and n = 1 control construct), consistent with the exploratory, proof-of-concept design of the study. No formal statistical analysis was undertaken due to the specimen number limiting inference and results are therefore reported descriptively.

Results

Static Load-to-Failure Testing

Static testing resulted in failure at 213.9 N. Failure occurred in the mandible model at the posterior corner of the resection margin rather than within the prosthesis, fixation screws, or implant–prosthesis interface (Figure 6). No visible damage, screw loosening or prosthesis deformation was observed. The dental implants remained firmly in position and did not show any loosening when tested with the torque wrench, remaining engaged with the mandible model.

Figure 6.

Mechanism of failure of mandible model on static load-to-failure testing at 213.9 N.

Dynamic Cyclic Testing

Both the control specimen and all three prototype specimens completed 500,000 cycles of dynamic loading without catastrophic failure. Visual inspection at the conclusion of testing did not demonstrate any cracks, distortion, or permanent deformation of the titanium prosthesis. The prostheses remained well adapted to the mandible models, and there was no evidence of loosening of fixation screws. Reverse torque testing confirmed that the dental implants remained firmly seated, with no detectable loosening or change in position.

Instron anvil displacement data were analyzed as a surrogate measure for plastic deformation of the construct. Table 1 summarizes the maximum and minimum displacement values recorded during the initial and final 1000 cycles of testing. Across all specimens, the difference between maximum and minimum displacement per cycle remained largely unchanged over the course of testing. The change in per-cycle displacement range was small, with a maximum difference of 0.044 mm observed in prototype specimen 1, indicating an absence of appreciable plastic deformation. In the control specimen, the per-cycle displacement range changed by 0.197 mm over the course of testing without progressive increase, also indicating an absence of appreciable plastic deformation.

Table 1.

Displacement of Instron Anvil at Initial and End of Test

	Prototype 1	Prototype 2	Prototype 3	Control
Initial minimum displacement (mm)	−6.955	−3.883	−3.430	1.466
Initial maximum displacement (mm)	−5.709	−3.620	−3.132	1.952
Initial Difference (mm)	1.246	0.263	0.297	0.486
End minimum displacement (mm)	−9.220	−3.952	−3.537	1.385
End maximum displacement (mm)	−7.929	−3.716	−3.261	1.674
End Difference (mm)	1.290	0.236	0.276	0.289
End difference – Initial Difference (mm)	0.044	−0.027	−0.021	−0.197

Notes: Initial: first 1000 cycles of testing; End: last 1000 cycles of testing. All measurements are recorded to 3 decimal points. Control: Mandible model reconstructed with conventional 2.5 mm reconstruction plate. Displacement values for the control are reported with positive values due to differences in zeroing, however, differences in displacement are still directly comparable.

Relative inter-sensor displacement was analyzed as a measure of implant micromotion during cyclic loading (Table 2; full coordinate data in Supplementary Tables S1–S4). In prototype specimens 1 and 3, the change in relative position between the implant sensors and the reference sensor over 500,000 cycles did not exceed 0.11 mm and 0.78 mm respectively in any axis. Prototype specimen 2 demonstrated changes of up to 10.50 mm when start and end time points were compared. However, the absolute coordinates of all three sensors, including the reference sensor fixed within the mandible model which was not itself load-bearing, shifted concurrently, a pattern inconsistent with movement of an individual implant. Data sampled at approximately 250,000 cycles showed maximum changes of 0.22 mm, comparable to the other specimens. Taken together with intact visual inspection and the absence of implant loosening on reverse torque testing, this pattern is most consistent with dislodgement or displacement of the sensor assembly rather than true implant instability.

Table 2.

Maximum Change in Relative Inter-Sensor Position Over Cyclic Loading

Specimen	Time Points Compared	Maximum Change in Relative Position in Any Axis (mm)	Reverse Torque Test	Visual Inspection
Prototype 1	1k vs 500k cycles	0.11	No loosening	Intact
Prototype 2	1k vs 500k cycles	10.50a	No loosening	Intact
Prototype 2	1k vs 250k cycles	0.22	NAb	NAb
Prototype 3	1k vs 500k cycles	0.78	No loosening	Intact

Note: ^aAttributed to sensor dislodgement rather than actual implant movement; ^bReverse torque testing and visual inspection were performed at the conclusion of testing (500k cycles) only.

Overall, dynamic cyclic testing indicates that the prototype construct withstood the simulated post-resection masticatory loading protocol in a synthetic model without mechanical failure, appreciable deformation, or loss of implant stability. These findings characterize mechanical behavior in a bench-top model only and do not reflect potential in vivo behavior.

Discussion

This proof-of-concept study demonstrates the mechanical feasibility in a bench-top model of a novel patient-specific, 3D-printed titanium mandibular prosthesis incorporating conventional dental implants for combined mechanical fixation and biological anchorage. The concept was developed in response to persistent challenges in segmental mandibular reconstruction, particularly donor-site morbidity associated with vascularized bone flaps, geometric limitations of grafted bone, and the difficulty of achieving predictable dental rehabilitation.3,6–8,46,47 By integrating dental implants into a custom titanium prosthesis designed to engage native mandibular bone, this approach aims to provide a stable reconstructive solution while facilitating functional dental rehabilitation.

Existing patient-specific reconstructive strategies for segmental mandibular defects have largely focused on CAD/CAM reconstruction plates or custom titanium mesh constructs.19–26 While these approaches improve intraoperative accuracy and rigidity, they remain dependent on adjunctive bone grafting and rely primarily on mechanical fixation with osteosynthesis screws. In contrast, patient-specific mandibular prostheses seek to reconstruct the defect primarily through the implant itself. Prior reports of modular endoprostheses and 3D-printed titanium prostheses have demonstrated feasibility but have been limited by complications such as screw loosening, soft tissue dehiscence, and uncertainty regarding long-term biological integration.23,26 The present design attempts to mitigate these limitations by incorporating conventional dental implants that engage the native mandible, thereby introducing the potential for osseointegration-based biological anchorage in addition to screw fixation.

Implant trajectory represented a key design consideration in the development of the prosthesis. Digital simulation of implant positions during the design phase demonstrated that near-axial implant placement could compromise prosthesis thickness and prosthodontic emergence, whereas increased angulation to engage maximal bone introduced significant off-axis loading. An intermediate implant angulation (approximately 10° to the vertical axis of the mandible) was therefore selected reflective of the anatomical and biomechanical constraints encountered in segmental mandibular reconstruction. This reflects a common clinical scenario in which ideal implant positioning must often be balanced against the availability of bone.

The mechanical testing results support the mechanical feasibility of the proposed design in a bench-top model under a simulated post-resection masticatory load. Static load-to-failure testing resulted in failure within the mandible model at 213.9 N, far exceeding the expected masticatory forces, rather than in the prosthesis, fixation screws, or implant-prosthesis interface. While this suggests that the mechanical stability of the prosthesis is favorable at the simulated load, it may also suggest excessive stress concentration at the bone osteotomy-implant interface resulting in crack propagation. The pattern of failure through the posterior corner of the resection is consistent with the biomechanical and finite element analysis findings of Murakami et al.48 Possible improvements to the design may include extending the posterior plate arm of the prosthesis and beveling the osteotomies to avoid sharp resection corners to reduce stress concentration. These potential design improvements can be further evaluated with finite element analysis in subsequent work.

Dynamic cyclic testing using loads of 40–60 N for 500,000 cycles demonstrated that all prototype specimens and the control specimen were able to withstand the simulated post-resection masticatory loading without catastrophic failure, progressive deformation, or loss of implant stability.43,44 The change in per-cycle displacement range across testing was small in all specimens, suggesting that the prosthesis provides at least comparable dimensional stability under repeated loading to the current standard fixation method, although the single control specimen precludes any statistical comparison.

Electromagnetic sensor analysis further supported the stability of the implant-prosthesis construct during cyclic loading. Relative implant movements in two of the three prototype specimens remained within a narrow range, indicating minimal micromotion under repeated loading, which is important for osseointegration to avoid fibro-osseous integration.49 Further stability may be achieved clinically by considering a period of intermaxillary fixation, followed by a soft, non-chew diet following reconstruction, reducing forces on the implant during the critical period of osseointegration. Although one specimen demonstrated larger changes in relative sensor coordinates at the end of testing, these findings were not corroborated by visual inspection or reverse torque testing, and mid-cycle sensor data were more consistent with the other specimens. This suggests that the observed discrepancy was more likely related to sensor displacement rather than true implant or prosthesis instability.

There are some limitations to this study. The number of specimens tested was limited, consistent with the proof-of-concept nature of the study. Formal finite element analysis was not performed, as the primary objective was to establish mechanical feasibility under clinically relevant loading conditions rather than to optimize design geometry.

While the use of a synthetic polyurethane mandible model allows for consistency and reproducibility in mechanical evaluation, it does not replicate the complex properties of native bone such as anisotropy related to bone grain orientation or the viscoelastic behavior conferred by collagen and water content, nor does it allow assessment of osseointegration or soft tissue response.50 Loading conditions were simplified and localized to the incisal region and do not fully replicate the complex, multidirectional forces encountered during mastication. In contrast to polyurethane models, native bone can remodel in response to functional loading, allowing adaptation to repetitive stresses over time.51 This is particularly relevant in the context of fatigue loading, as cyclic masticatory forces in vivo are distributed over prolonged periods rather than applied continuously with an average meal consisting of approximately 660 chews.52

Biological factors which would be crucial to clinical translation could not be assessed in this bench-top study including the reliability of the osseointegrated implants to provide sufficient anchorage for the prosthesis. Another critical biological factor is the soft tissue coverage of both the prosthesis and the incorporated osseointegrated implants, with previous reports of titanium mandibular prostheses reporting significant soft tissue complications with exposure and infection of the prostheses.23,26,53 Peri-implant tissue maintenance is also important to minimize the risk of peri-implantitis which can result in loss of integration of the osseointegrated implants in the long term, which is often a challenge in reconstructed mandibles due to limited keratinized gingiva.8,9 This design aims to mitigate these concerns with micro-rough surfaces at soft tissue contacting regions and a fitted seal of the osseointegrated implant surface to the prostheses to reduce ingress of oral contaminants to the implant-bone interface but will require in vivo testing to evaluate the effectiveness of these measures.

Stress shielding is another recognized potential biological phenomenon arising from the discrepancy in elasticity between titanium and bone causing inefficient load transfer between the prosthesis and bone and subsequent bone resorption.54–56 The incorporation of osseointegrated implant may help to directly transfer masticatory loads to the native bone to mitigate this phenomenon. A porous lattice throughout the prosthesis may be considered in future designs to more closely match the elasticity of bone to reduce this effect.37

Despite these limitations, the findings of this proof-of-concept study provide a rationale for further investigation and represent the first phase of a staged translational development pathway. Future work should focus on refinement of prosthesis geometry to reduce stress concentrations, incorporation of porous titanium interfaces to enhance biological integration and more closely match the modulus of elasticity of bone, and evaluation of soft tissue behavior. In vivo studies will be necessary to assess osseointegration, long-term stability, potential effects of stress shielding and functional outcomes, as well as to define appropriate clinical indications for this reconstructive approach before first-in-human evaluation.

Conclusion

This proof-of-concept study demonstrates the mechanical feasibility of a patient-specific, 3D-printed titanium mandibular prosthesis incorporating osseointegrated dental implants for segmental mandibular reconstruction. Three prototype constructs withstood 500,000 cycles of simulated post-resection masticatory loading at 40–60 N and 2 Hz without mechanical failure, appreciable plastic deformation, screw loosening, or detectable implant loosening. These findings represent preliminary mechanical validation in a synthetic bench-top model. Further translational investigation, including evaluation in in vivo models, will be required to establish clinical effectiveness, osseointegration and biological anchorage of the dental implants, peri-implant soft tissue health, long-term durability under physiological multidirectional loading, and biocompatibility of the final device before this concept can be considered as a clinical reconstructive option for segmental mandibular defects.

Funding Statement

This project was supported by grants from the Research, Innovation and Enterprise (RIE) 2020 Centre Grant Seed Funding Programme from the National Research Foundation through the National Dental Centre Research Fund, as well as from National Additive Manufacturing Innovation Cluster @ NTUitive (NAMIC@NTUitive) and the National University of Singapore Faculty of Dentistry.

Abbreviations

CAD/CAM, Computer-aided design/computer-aided manufacturing; EM, Electromagnetic; eTJR, Extended temporomandibular joint replacement; STL, Stereolithography.

Data Sharing Statement

The raw data supporting the conclusions of this study are available from the corresponding author upon reasonable request.

Author Contributions

All authors made a significant contribution to the work reported, whether that is in the conception, study design, execution, acquisition of data, analysis and interpretation, or in all these areas; took part in drafting, revising or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work.

Disclosure

The authors declare no conflicts of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript; or in the decision to publish the results.