Influence of Different Implants on the Biomechanical Behavior of a Tooth-Implant Fixed Partial Dentures: A Three-Dimensional Finite Element Analysis

Dental implants have been in use for several decades. However, limitations continue to exist in the area of implant therapy. When inadequate bone support is an issue, several surgical methods are available to resolve the problem. These methods involve high biological risk, complex clinical procedures, increased costs, and additional clinical time. The patient may refuse such surgical treatments for these reasons. Anatomical limitations of the posterior edentulous area, osseointegration failure, and the economic or physiological condition of the patient may dictate the need for tooth-implant supported fixed partial denture.^1–5 

The threshold for tactile perception between natural teeth and implants is significantly different.^4–6  Natural teeth have a physiological horizontal mobility of 50 to 200 μm, while the mobility of an osseo-integrated implant is <10 μm. This difference in mobility between natural teeth and an implant system can lead to biomechanical problems.⁷  When teeth and implants are connected by means of a prosthesis, the risk to teeth is intrusion, while the risk to implants is overloading.^7,8  Problems such as loss of osseointegration, horizontal bone loss, and technical complications (including loss of retention, screw loosening, and loss or fracture of abutments and implants) have also been reported.^9–12 

From a theoretical perspective, a nonrigid connection between a tooth and an implant appears to be advantageous because biomechanically, it enables physiological movement of the tooth under occlusal forces.¹  However, clinical experience correlates this type of connection with intrusion of the natural teeth,¹³  which reduces abutment support. This leads to an increase in cantilever stresses to the implant and supporting bone and may permit migration of the opposing teeth.⁷  With these complications, the rigid connection has proven to be more successful than a nonrigid connection. It is often advocated to prevent tooth intrusion under a connected restoration,^8–11  which minimizes the occurrence of mechanical failures.¹ 

The implant system may influence the connection of external hexagon (EX) implants compared to natural tooth movement.^14,15  Internal hex implants increase the stability of the prosthesis; however, at the same time, they can increase stress concentration.³  Implants with internal connections, such as internal hex and Morse-taper (MT), have been used with tooth-implant connected prostheses.^3,9,16,17 

Finite element analysis (FEA) is an established, non-invasive tool for studying the biomechanics and influence of mechanical forces on biological systems. It is a technique that provides valuable insight to stress patterns in the implant and peri-implant bone and predicts the success of implants in clinical conditions.¹⁸ 

The goal of the present study was to analyze the biomechanical behavior of rigid and nonrigid tooth-implant supported fixed partial dentures. Different implant systems were used to observe the load distribution over teeth, implants, and adjacent bone using three-dimensional FEA. The null hypothesis was that neither the type of implant, EX or MT, nor the type of attachment, rigid vs nonrigid, has an influence on the biomechanical behavior of tooth-implant supported fixed partial dentures.

For this study, a 3D model of the posterior part of an edentulous mandible was modeled with the cortical and trabecular bone, coronoid and condyle processes, and foramen mentalis. A model was constructed for computer simulations using Rhinoceros 4.0 SR8 software. A simulation of tooth loss of the first and second right molars was created with an implant placed in the second right molar and a prepared tooth with simulated periodontal ligament (PDL) in the second right premolar (Figure 1).

An artificial second premolar tooth obtained from a dental mannequin was scanned using a microtomograph (SkyScan, Kontich, Belgium). This technique yielded a geometric model. After tooth modeling, an axial reduction of 1.5 mm and 6 degrees of inclination was made, simulating a complete metal-ceramic crown. A simplified 0.2 mm simulated periodontal ligament (PDL) was created according to a previous report.¹⁹ 

Two types of implants (Neodent, Curitiba, Brazil) and their respective abutments were selected: an EX implant (3.75 × 13 mm) with a 4.1 mm universal abutment; and an MT implant (3.75 mm × 13 mm) with a single-body universal abutment. The geometry of the implants and their components were provided by the implant manufacturer (Neodent) and transformed into a 3D format. Screw geometry was obtained by simplification of the original design.

A metal-ceramic framework was constructed using the superficial scanning of the occlusal surface of the artificial teeth obtained from a dental mannequin. Microtomography (Kontich) was used. Ceramic-coated nickel-chromium-titanium alloy was used as the structure material. Two types of partial fixed prosthesis were studied: rigid and nonrigid. The nonrigid connection was used an intracoronal attachment (2.5 mm) matrix positioned distally to the second right premolar, and the matrix was placed into the pontic.

The simulations were executed using the following four models:

1. Model 1: EX implant and second premolar connected by a rigid connection.

2. Model 2: MT implant and second premolar connected by a rigid connection.

3. Model 3: EX implant and second premolar connected by a nonrigid attachment positioned on the distal side of the second premolar.

4. Model 4: MT implant and second premolar connected by a nonrigid attachment positioned on the distal side of the second premolar.

The MT implants were located 2 mm deep inside the crestal bone. The EX implants were placed immediately supraosseously, as recommended in a previous report.²⁰ 

Mesh generation and data processing were performed using the 3D FEM analysis package (ANSYS, Inc, Canonsburg, Pa). All geometric designs were exported for a finite element preprocessing phase (Workbench 12, ANSYS). The following four assumptions were made:

1. The PDL was assumed to be a viscoelastic material, and the other materials were considered as isotropic, linear elastic, and homogeneous.²¹ 

2. Complete adhesion was assumed between the implant and the bone provided by osseointegration (ie, accepted as 100%) and screw torque, respectively.

3. Screw and implant were assumed to be a single structure. The mechanical properties of the materials are listed in Table 1.^22–25 

4. All models were assumed to be of a single structure without gaps.

A static load of 50 N to the premolar and 100 N to the implant²⁶  were applied distributed at occlusal points resulting in satisfactory occlusion. The forces used corresponded to the forces that are applied to a full removable maxillary prosthesis. The friction coefficient (μ) was set at 0.5 to the attachment and 0.16 between the titanium and the MT implant surfaces.²⁷ 

A linear equation was adopted for the elastic properties, in which the deformation of the body is proportional to the force applied. The properties were assumed to be constant and isotropic, and all models were assumed to be of one body with total absence of mismatches for the static and linear analysis.

The equivalent von Mises stress analyses was used to study implants, abutments, and frameworks. To observe the trabecular and cortical bone, maximum and minimum principal stresses were performed using graphical visualization color maps.²⁴ 

The minimum stress values at the PDL are shown in Figure 2. When the rigid connection was used, there was compression at the apical region of the tooth and maximum stress values, represented tensile loads. At the third medial and crestal regions, the tooth experienced intrusion in the distal direction. In the EX implant (Figure 2a) there was a tendency toward inclination (ie, flexion). For the MT implants (Figure 2b), there was a distribution of maximum stress values and there was no movement. Forces were annulled without tensile loads apically, and intrusion did not occur. The minimal stress distribution on the PDL of nonrigid models (with attachment) exhibited minimal stress values similar to models with a rigid connection. On the EX implants (Figure 2c), there was prevalent tensile stress in the distal region of the PDL and compressive loads at the apical region. This indicates possible intrusion of the tooth. In the MT model (Figure 2d), there was better stress distribution, and the forces were annulled.

When the rigid connection was used in both the MT (Figure 3) and EX (Figure 3b), there was a higher incidence of compressive forces detected at the cervical region corresponding to the cortical bone. In the MT implants (Figure 3b), tensions were higher because they were placed 2 mm below the bone crest. A lack of maximum stress values was observed in both implants (ie, MT and EX). There was no tension at the cancellous bone adjacent to the threads of the MT or EX implants. Minimum stress distribution was similar. EX (Figure 3c) exhibited prevalent stress levels at the distal region of the PDL, with compression at the apical region, flexion of the PDL, and intrusion of the tooth. The MT implants (Figure 3d) exhibited better maximum stress distribution. Minimum stress occurred at the cortical bone, with lack of stress at the trabecular bone in both implants (ie, MT and EX).

The von Mises stress distribution is shown in Figure 4. Failure occurs when tension exceeds the strength of the metallic material. It is an important parameter to consider, given that stress affects the prosthesis framework. When the rigid connection was used in 2 models, stress was predominant in the connection region of the implants and in the abutment/pontic area (Figure 4a and b). When the nonrigid connection was used, there was higher stress at the attachment region (Figure 4c) and an accumulation of stress in the connection between the MT implants and abutment/pontic area (Figure 4d).

One of the most important factors in the success of osseointegrated implants is how the implant reacts to the loads to which they are subjected. Peri-implant bone remodeling can be influenced by the stress that results from the incidental loads.^24,26  The present study evaluated the generation of stress in a tooth-implant supported fixed partial denture with two different connectors, rigid versus non rigid, and two different types of implants, EX and MT. The results suggest that the implant type did have an influence on the biomechanical behavior of the rehabilitation. However, the rigid and nonrigid connections had no influence on the results that were analyzed using the FEA method.

Occlusal forces affect the bone surrounding an oral implant. Compression loads tend to maintain the integrity of the bone-implant interface, tension loads tend to separate this interface.^24,28  The strain is dependent on the mechanical properties of the bone. A given load may affect various bones differently.^29,30  Cortical bone is harder than the trabecular bone, which is more porous and more brittle.^24,29,30  As the bone is considered a brittle material,²⁴  maximum and minimum stresses at the peri-implant bone were analyzed. These stresses on the bone near the tooth were also analyzed.

In Model 1, in which the EX implant and second premolar were connected rigidly, there was compression (minimum principal stress) at the apical region and tension at the middle-third of the root and at the crestal bone. This indicates intrusion of the tooth in the distal direction, with flexion of the PDL. These results are important because compression tends to maintain the integrity of the bone. However, tension tends to separate the interface and is more destructive.^24,28 

In Model 2, the MT implant and second premolar were connected rigidly, and they experienced better load distribution, with accumulation of loads in the PDL: the tooth did not experience intrusion. When the minimum and maximum principal tensions were analyzed at the implant site, there was a greater tendency toward compressive tension at the crestal area regardless of the implant type (ie, EX or MT). There was an absence of tension in the adjacent trabecular bone-to-implant threads in both implant types. This was an important finding as continuous fatigue over long periods of time can result in a stress fracture.²⁵ 

In Model 3, the EX implant and second premolar were connected by a nonrigid attachment positioned on the distal side of the second premolar. The same distribution of tension was observed, and this resulted in tooth intrusion. When the MT implant was connected by a nonrigid attachment, there was an absence of tension and compression at the cortical bone, and an absence of tension in the trabecular bone.

These results are important due to the structure of the bones. Cortical bone is harder and more compact, more capable of resisting stress than is the trabecular bone, which is more porous and brittle.²⁹  Therefore, the load could be transferred to the bone with the greatest ability to tolerate stress. There was concentration of stress in the bone surrounding the implant neck, as observed in other studies.^25,30,31 

Different results were found by Tsouknidas et al²⁵  who verified that a nonrigid connection showed higher stress values compared to the rigid connection. These differences possibly occurred because the loads were different (200 to 230 N). Moreover, the authors do not mention the type of alloy that was used to make the metallic structure. Our study used the Ni-Cr-Ti alloy, which is extremely hard.

It is important to emphasize that bone quality may affect biomechanical behavior when the tooth and implant are connected.³²  The MT implants exhibited better behavior with lower principal stress values because the conical interfaces resist lateral loads. This aids the distribution of stress in bone by reducing the effects of bending and permitting settlement of the locking system, thus improving stability.³³  This occurs due to the wide connection area and the increased resistance to lateral loads, in which the lateral wall dissipates the load and promotes superior stress distribution.³⁴ 

The bone level of the implant can also interfere with stress distribution over the bone. MT implants are placed 1 mm to 2 mm deep, which decreases the magnitude of stress, thus improving the stress pattern.³⁵  The subcrestal placement of implants had a positive impact on crestal bone remodeling. The EX is placed supracrestally, which interferes with stress distribution.²⁰ 

FEA is an important process used to study stress patterns in implants and their components. FEA is also used to study the peri-implant bone and the role of biomechanical conditions on the implant-bone interface as well as the stress patterns in these areas. This process will lead to improvements in implant design and placement techniques.¹⁸  FEA allows simulation of the interface and interaction between the implant and its adjacent osseous tissue. The von Mises stress values are utilized to analyze the mechanical effect on the implants and peri-implant trabecular/cortical bone due to occlusal forces.³⁶  If the von Mises stress values exceed the yield strength of a metallic material, failures occur. This is important for interpreting the stress that occurs within the metallic fixed partial denture material.³⁷ 

The PDL should be included in FEA models of the masticatory apparatus because it plays an important role in load transfer from the teeth to the alveolar bone.³⁸  As a consequence of continuous pressure on the PDL, a disturbance of blood flow and cell death in the area of compression (described as hyalinization) are observed. Resorption of the hyalinized tissue by macrophages and bone resorption by osteoclasts results in intrusion of teeth.³⁸  The PDL exhibits a non-linear, time-dependent viscoelastic behavior^34,40,41  and plays a major role in controlling tooth movement under an external force. It also influences the remodeling process in alveolar bone.⁴²  The presence of the PDL provides positive dental stability; it absorbs the occlusal force without transmitting it to the bone, and hyalinization occurs.²⁵  The tooth-to-implant connection may cause a larger portion of the masticatory load to be transferred to the implant. It is more rigidly anchored to the bone than to the tooth,⁴³  and implant movement cannot be detected.²⁵  A tooth rigidly connected to an implant shares load support with the implant; as the load increases, the tooth is increasingly involved.⁴⁴  The existence of supporting structures with different mechanical properties creates a complex biomechanical environment.²⁵ 

The models used in the present study relied on several assumptions regarding the simulation structures: They were all assumed to be homogeneous and isotropic, and to possess linear elasticity.⁴⁵  Minimum stress distribution demonstrated that Models 1 and 3 (EX) displayed stress, which was prevalent at the distal end of the PDL, and compression at the apical region of the PDL. This suggests possible flexion and intrusion of the tooth. The presence of equilibrium in the maximum stress distribution at the distal and mesial PDL in Models 2 and 4 (both MT) suggests there was no movement of the tooth and stress was balanced. This could be associated with the rigid behavior exhibited in the conical region. This was due to the MT connection and the demonstrated superior abutment stability that was related to the least stress concentration of the abutment screw.⁴⁶  In a taper connection, the load is resisted mainly by the taper interface that prevents the abutment from tilting off. This process permits stable retention of its position by frictional forces. MT implants do not cause tooth intrusion; they are characteristically stable and promote strength.⁴⁷ 

Comparison of a rigid framework and implants revealed that MT and EX had higher stress incidence in the cortical bone. This was observed mainly in the MT, which was placed 1 mm below the crestal bone. In both cases, there was no maximum stress at the trabecular bone adjacent to the threads of implants. In the EX connection, there was no form or positive locking, and the lateral force were absorbed mainly by the abutment screw.³² 

The peak of stress in the bone supporting a dental implant appears in the dental area near the level where the implant begins to attach to the bone.^45,48  Occlusal force exerted from the implant to the bone is primarily concentrated on the cortical bone. High stress concentration or distribution in peri-implant cortical bone should be avoided to maintain long-term function of implant loading.³⁶ 

The particular stress/strain distribution observed in each implant-abutment connection will result in different stress/strain patterns in the bone, mainly at the area where the bone nears the implant crest.⁴⁸ 

Advantages of the MT connection are better distribution of microstrain in all the implants, decreasing of the microgap, and better stability of the joint. This may result in higher stress values of the prosthetic components.^33,47 

It is important to address the limitations of the present study. FEA has certain limitations because it is an in silico, computerized method in which clinical conditions may not be completely replicated. In this study, 100% osseointegration was assumed. However, real-life osseointegration ranges between 40% and 60%.²⁵  Results were obtained using linear analysis, and the PDL was assumed to be a viscoelastic material. The others were considered isotropic, linear, and homogeneous. The results are only qualitative; however, the literature has demonstrated that the results of FEA can be extrapolated to clinical scenarios, albeit with caution.⁴⁹  Future research should be supplemented with clinical evaluation to accurately analyze the biomechanical behavior of the tooth-implant connection.

Despite the limitations of the in-silico study, the results were promising for the indication of a mixed prosthesis when supported by natural tooth and implant. This is especially true when a Morse taper implant is used. It prevents the intrusion of the natural tooth and demonstrates superior load distribution as well as the absence of tension in the trabecular bone in both prosthesis connection types.

Within the limitations of this study, it can be concluded that—regardless of the type of prosthesis attachment (ie, rigid or nonrigid)—intrusion of the tooth will tend to occur when the implant has an EX connection. When MT implants were used, there was no intrusion of the tooth regardless of whether the prosthesis had a rigid or nonrigid connection. MT implants demonstrated better load distribution, with an absence of tension in the trabecular bone in both prosthesis connection types. Finally, regardless of the implant type (MT or EX), there was a greater tendency toward compressive loads at the crestal area.

Abbreviations

3D:

three dimensional

EX:

external hexagon

FEA:

finite element analysis

MT:

Morse taper

PDL:

periodontal ligament

This study was supported by Grant #2011/11190-0 from Sao Paulo Foundation Center (FAPESP) and Coordination for the Improvement of Higher Education Personnel (CAPES).

The authors declare no conflict of interest related to this study.