Comparison of Conventional and Pontic Fixed Partial Dentures Over Implants Using the Finite Element Method: Three-Dimensional Analysis of Cortical and Medullary Bone Stress

The evolution of contemporary dentistry led to oral rehabilitation with high levels of resolution, principally with possibilities for implantology.¹  Dental implants can be used in upper or lower arches and for single to multiple missing teeth; they are very satisfactory and have high success rates.²  Posterior regions are highly affected by tooth loss, which impairs aesthetics and patients' function and reduces different aspects of quality of life.^3,4 

There are minimum requirements for adequate safety and predictability in implant restorations, such as good systemic and local conditions that allow surgical installation and successful osseointegration. These factors can result the return of masticatory function quality, which directly affects health as a whole.^5,6 

Currently, the majority of implant failures are the result of biomechanical complications after prosthetic loading. The risk of early failure of implants due to failed osseointegration has diminished due to the development of implant macro and microdesign.⁷  This phenomenon indicates the importance of the biomechanical behavior of the prostheses placed on osseointegrated dental implants.

Several factors are directly related to the biomechanical behavior of the prostheses. One is the presence of a favorable ratio between the crown and implant, which is much cited and is a fundamental characteristic of natural teeth; however, in implant prostheses, bone quality and the ability of the bone to transfer and dissipate the loads are just as important as this ratio.^8–10 

In addition, for rehabilitating posterior areas that have large masticatory loads, use of the maximum number of implants to support the prostheses is recommended. In addition, the use of splinted fixed partial dentures is indicated for better stabilization of the rehabilitation system. These procedures give the patient a complex rehabilitation system capable of supporting the functions of the dental arch.¹¹ 

However, variations are often present, such as the need to decrease the number of implants and use them, for instance, at the extremities of the rehabilitation^12–15 and variation in the types of constituent materials.

Thus, in a posterior mandibular region with 3-element dental absences, classic rehabilitation is expected to use 3 dental implants^16–18 ; however, the use of only 2 implants may also occur in rehabilitated regions containing 3 splinted dental elements and a pontic as the intermediate element. Although these prostheses are used empirically by professionals, more data and clarifications regarding their biomechanics are necessary for their correct indication and safe applicability^19–21  because there is no agreement on the ideal number of implants to provide adequate support for this situation.²² 

In this way, the objective of this study was to analyze, understand, and compare the differences in the stresses generated in the cortical and medullary bones according to the configuration of the 2 proposed types of fixed partial prostheses used: either conventional (on 3 implants) or pontic (on 2 implants).

To generate simulations by using the finite element method and visualize the stress present with the applied loads, three-dimensional models were developed.

Morse taper implants with a diameter of 4 mm and the same length were used as models for all cases (∅4 × 11 mm, Neodent, Titamax CM Cortical with 11.5 degrees of internal slope, Curitiba, Brazil), as were transmucosal components with 3.5 mm rotational calcinable cylinders because they were splinted with multiple prostheses of the same brand. The implants were positioned 2 mm subcrestal, as recommended by the manufacturer. For better reliability of the results, all the structures present in the models were created separately; that is, they had different parts, such as the periodontal ligament, pulp, dentin and enamel for the dental element, and the metal, ceramic and resin structures of the screw hole for the prostheses.

The drawings (CADs) were performed by the AnsysWorkbench 10.0 Software (Swanson, Analysis Systems, Inc, Houston, Tex), and the following experimental groups were formed: the regular splinted group (GR), with a conventional infrastructure on 3 regular-length Morse taper implants (4 × 11 mm), and the regular pontic group (GP), with a pontic infrastructure on 2 regular-length Morse taper implants (4 × 11 mm), as shown in Figure 1.

From both assembled models, the contour conditions were defined (fixation of the models in their basal area, oblique occlusal loading of each of the prosthetic elements of 45° in the buccal-vestibular direction, 365 N for molars and 200 N for premolars), and the physical characteristics of each structure present in the models also defined, as in previous studies (using the modulus of elasticity and Poisson's coefficient).¹  All elements that composed the models were considered isotropic, homogeneous, and linearly elastic.

The finite element mesh was generated for the 2 experimental groups (Figure 2); the GR had 205 988 nodes and 122 272 elements, and the GP had 172 800 nodes and 103 536 elements. Any incongruity in the generated mesh was checked, and then the simulations were performed in a virtual environment using the minimum principal stresses (TMiP) in cortical and medullary bones to obtain the stress data because it best represents the compressive stress of the support structures.^23,24 

The areas chosen for analysis of the results were the vestibular face of the cortical bone and the vestibular and internal/lingual surfaces of the medullary bone, the latter of which was obtained from a longitudinal section in the experimental models. The scales generated in the analyses were also standardized to allow qualitative and quantitative comparisons and to compare the experimental groups with each other.

The simulations were performed, and the results of the TMiP for the friable structures, specifically the vestibular face of the cortical bone and the vestibular and internal/lingual face of the medullary bone, were obtained. For better visualization and correlation of the qualitative and quantitative results between the experimental groups, the scales were standardized (Figure 3).

The Table shows the maximum absolute values of the TMiP (MPa). In Figure 3, the images of the simulations performed for the GR and GP experimental groups are presented. Images with colder colors represent higher levels of stress (negative scale).

Osseointegrable implants are directly attached to the bone tissue and all applied loads generate stress on the bone that can exceed its physiological limits²⁵ ; therefore, a splint is suggested to allow better stress dissipation and because splinted prostheses show better results than individual crowns.^26,27  Guichet et al²⁸  reported in their 2002 study that splinted prostheses dissipate stress better than individual prostheses, demonstrating that in cases of contiguous sequential toothlessness, a safer approach is use of a splinted prosthesis, even using an implant for each dental element. Corroborating the abovementioned studies, Bergkvist et al²⁶  also studied the distribution of stress in the bone surrounding implants supporting splinted and nonsplinted prostheses. In addition, reduced stress was found in the bone surrounding splinted implants compared to individual ones.

Considering the union of the crowns as the only rehabilitation option for prostheses with intermediate pontics, only the use of a splint-fixed partial denture on implants was simulated in the two experimental groups in this study.

The results showed that compressive stresses present in the bones (both cortical and medullary) were higher around the implants for the GP, mainly in the cervical (Figure 3, a2, b2) and apical (Figure 3, a3, b3) areas. For the GP (Figure 3, b2), there was a union of the radiated stresses in the medullary bone between the dental element and the anterior implant, which was different from findings for the GR group (Figure 3, a2).

Thus, some authors suggest that reducing the number of implants supporting a fixed partial denture would decrease financial and surgical costs²⁹ ; that may, in fact, lead to biomechanical damage to the system as a whole via this increase in stress. In the present study, no such thresholds were reached, but we did observe a tendency for increased stress as the number of support implants decreased.

On the other hand, considering the stresses on the cortical bone, it was observed that the placement of a prosthesis over 2 implants (GP) (Figure 3, b1) induced lower absolute values of cervical stress to the dental element compared to placement of a prosthesis on 3 implants (Figure 3, a1); the finding that the GR group did not encounter other types of stress found around the implants may be explained by some inconvenient peak generated by the mesh generation or even by areas of very acute bone contour surrounding the dental element, which would naturally concentrate more stress.

In agreement with the above findings regarding stress surrounding the implants, Hiskell et al³⁰  examined rehabilitation with two implants and a central pontic or three implants and found more favorable results regarding the stress and deformations of the structures when two implants were used instead of three. In specific cases, such as research in 2005 by Yokoyama et al³¹  in a context of total toothlessness, prostheses made in segments were found to generate greater stress than those made as a single part. However, it should be noted that there are other factors to be clinically analyzed, such as possible contraction tensions inherent in the deformation of large prosthesis structures.

There is, therefore, some divergence in the literature regarding the use of greater or fewer implants in oral rehabilitations and when one chooses the union or not when performing rehabilitations with implants.

In a study performed in 1998 by Stegaroiu et al,³²  models of fixed prostheses on two implants and a central pontic were compared with those of a fixed prosthesis of three joined elements placed on three implants using the finite element method as well as simulation. The results showed higher stress values in the support bone for the model with two implants and a central pontic, whereas smaller values were observed for the model of the fixed prosthesis with three united elements placed on three implants.³²  The results from Stegaroiu and colleagues are in agreement with the general findings of our present study. In the same work, the authors showed that under non-axial loads (such as those simulated in the present study), only splinted crowns supported by three implants were able to minimize the adverse effects of occlusal loading. Therefore, it is important to note the importance of occlusal adjustments and the need to refine this factor that is associated with chewing.

An external observation of the cortical bone, as described by Meirelles³³  in 2003 indicated a greater elasticity modulus that tends to concentrate higher stress levels; the results found in our present study indicate a larger stress range for the GR (Figure 3, a1) due to the union of the intermediate implant stress, but a higher absolute value for the posterior implant in the GP (Figure 3, b1). This demonstrates that the joining of more implants side by side may induce a greater range of stress compared to other methods; however, these stresses would be weaker than those resulting from placement of fewer implants, which increases stress, especially in the posterior area (shown in the GP group). Thus, it is necessary to decide whether to choose a greater range of stress with a lower absolute value (such as observed for the GR; Figure 3, a1) or a smaller range of stress with a higher absolute value in posterior areas (as observed for the GP; Figure 3, b1). This is a valuable observation for the extrapolation of results from a virtual environment to clinical practice.

Turning attention to the simulations and results of the medullary bone, it is clear that the use of only 2 implants with the pontic prosthesis (GP) leads to greater stress at more cervical threads of the implants (Figure 3, b2) than the use of 3 implants supporting a conventional fixed prosthesis (Figure 3, a2). The same fact can be affirmed when observing the stress in the cervical region of the implant, the bone crest, and the apical areas, where the highest absolute stress and range of stress were observed for the largest number of threads in the GP (Figure 3, b3).

It is important to note that even in the experimental group with 2 implants supporting a pontic prosthesis (GP), the stress did not reach levels that were damaging to the bone behavior under the occlusal loads, according to Toniollo et al¹  in 2012, Misch³⁴  in 1999, Iplikçioglu and Akça³⁵  in 2002, and Teixeira et al³⁶  in 2010. The maximum tolerated threshold (transition between elastic and plastic deformation) of the cortical and medullary bones is approximately 173 MPa and 10 MPa, respectively. This shows the real feasibility of using a rehabilitation system (GP) of only 2 regular-sized implants supporting a pontic prosthesis when necessary, such as in cases of financial or biological impossibility. This conclusion can be drawn from a biomechanical and bone tolerance point of view under the physiological occlusal loads applied in this study.

However, in this eventual use of this type of rehabilitation system (GP), the need for extreme attention to correct and judicious occlusal adjustment is emphasized, in addition to the ability to use implants of a regular length (11 mm) and place them in high-quality bone. Our assertions cannot be made safely without further study of the use of the GP model with inferior-quality bone, occlusal loads of greater magnitude or worse direction, or even with shorter implants. All of these conditions need to be better explored and clarified.

Therefore, studies are necessary to evaluate situations with implants of shorter length, prostheses of different dimensions, varied eventual occlusal adjustments, and varied bone quality (both more or less stiffness). In addition, the types of infrastructure used and the aesthetic coating material, should be studied and better discerned, as they can interfere with the stress dissipation and biomechanical behavior of the system.

Given the limitations of this study and the characteristics of the methodology used, following the proposed objective, it can be concluded that there are differences in the stress generated in support bones (cortical and medullary) according to the placement of the prosthesis (conventional or pontic) on 3 or 2 implants since the GP experimental group showed higher stresses, especially in the cervical and apical bone areas of the implants. However, these stresses given in the literature did not reach the values reported as injurious. Whenever possible, use of a greater number of implants in the rehabilitation has proven to be more advantageous for reducing the absolute values of stress generated.

Abbreviations

CADs:

computer-aided designs

GR:

regular splinted conventional group

GP:

regular splinted pontic group

TMiP:

minimum principal stress (compressive stress)

The researchers responsible for this work are grateful for the financial incentive (UniRV Researcher Grant n.18.2018.4.02) provided by the University of Rio Verde (UniRV) to the coordinating researcher involved in this project and for the financial incentive (PIBIC, National Council for Scientific and Technological Development (CNPq) given to the graduate student involved with this project.

The authors declare no conflict of interest.