- Email: [email protected]
Prediction at long-term condyle screw fixation of temporomandibular joint implant: A numerical study
Journal of Cranio-Maxillo-Facial Surgery 43 (2015) 469e474
Contents lists available at ScienceDirect
Journal of Cranio-Maxillo-Facial Surgery journal homepage: www.jcmfs.com
Prediction at long-term condyle screw fixation of temporomandibular joint implant: A numerical study A. Ramos a, b, *, R.J. Duarte a, b, M. Mesnard b a b
TEMA Center of Mechanical Technology and Automation, Department of Mechanical Engineering, University of Aveiro, Portugal Universit e de Bordeaux, Institut de M ecanique et d'Ing enierie, CNRS UMR 5295, Talence, France
a r t i c l e i n f o
a b s t r a c t
Article history: Paper received 13 September 2014 Accepted 16 February 2015 Available online 24 February 2015
The fixation of commercial temporomandibular joint (TMJ) implant is accomplished by using screws, which, in some cases, can lead to loosening of the implant. The aim of this study was to predict the evolution of fixation success of a TMJ. Numerical models using a Christensen TMJ implant were developed to analyze strain distributions in the adjacent mandibular bone. The geometry of a human mandible was developed based on computed tomography (CT) scans from a cadaveric mandible on which a TMJ implant was subsequently placed. In this study, the five most important muscle forces acting were applied and the anatomical conditions replicated. The evolution of fixation was defined according to bone response methodology focused in strain distribution around the screws. Strain and micromotions were analyzed to evaluate implant stability, and the evolution process conduct at three different stages: start with all nine screws in place (initial stage); middle stage, with three screws removed (middle stage), and end stage, with only three screws in place (final stage). With regard to loosening, the implant success fixation changed the strains in the bone between 21% and 30%, when considering the last stage. The most important screw positions were #1, #7, and #9. It was observed that, despite the commercial Christensen TMJ implant providing nine screw positions for fixation, only three screws were necessary to ensure implant stability and fixation success. © 2015 European Association for Cranio-Maxillo-Facial Surgery. Published by Elsevier Ltd. All rights reserved.
Keywords: Micromotions Screw fixation Strains TMJ implant stability Long term
1. Introduction The temporomandibular joint (TMJ) plays an extremely important role in daily activities; and since it is involved in many different everyday tasks such as communicating, feeding, or even sleeping, with up to 2000 motion cycles per day, it is consequently the most exercised joint in the human body (Guarda-Nardini et al., 2008; Tanaka and Koolstra, 2008). According to previous studies, 20%e 40% of the population presents with TMJ disorders. However, contrary to what would be expected, TMJ implants have been studied mostly in terms of clinical cases and without the use of numerical predictions (Solberg et al., 1979; Okeson, 1997; Hsu et al., 2006). Temporomandibular joint diseases affect almost half of the population during their lifetime, and although some of these
* Corresponding author. Biomechanics Research Group, University of Aveiro, 3810-193 Aveiro, Portugal. Tel.: þ351 234 370830; fax: þ351 234 370953. E-mail address: [email protected] (A. Ramos).
problems can be treated with drugs and physiotherapy, some individuals still need a TMJ implant. For this reason and because the TMJ implant success rate has been much lower than for total hip replacement or total knee replacement (van Loon et al., 1995; Hsu et al., 2011), this study attempted to investigate the best TMJ implant fixation position (Chase et al., 1995; Mishima et al., 2003; Mercuri and Giobbie-Hurder, 2004). Surgical treatment for a TMJ implant includes a total or partial condylectomy and the replacement of the disk by autografts or alloplastic materials (Chase et al., 1995) Nowadays, there are tree TMJ implants as options available on the market using screws to fixate the mandible part of the implant to the bone (Kanatas et al., 2012; Schuurhuis et al., 2012). Because the structures involved in this joint are very complex, the best implant fixation technique and the optimal number of screws remain uncertain. However, implant stability (Sidebottom and Gruber, 2013; Shen et al., 2014) is one of the most important and decisive factors in implant success (van Loon et al., 1995). The surgeon fixes, with screws, the mandible r implants as lateral plates in the TMJ implants to the bone (Bujta et al., 2014; Shen et al., 2014), normally using all or almost all of
http://dx.doi.org/10.1016/j.jcms.2015.02.013 1010-5182/© 2015 European Association for Cranio-Maxillo-Facial Surgery. Published by Elsevier Ltd. All rights reserved.
470
A. Ramos et al. / Journal of Cranio-Maxillo-Facial Surgery 43 (2015) 469e474
the available screw positions. However, this does not mean that the best implant stability is guaranteed (Hsu et al., 2006; Chowdhury et al., 2011; Hsu et al., 2011), and low levels of micromotion generated at the interface can hamper bone integration. The aim of this study was to analyze implant stability, simulating a long period of support as a function of the screw positions and strain distributions in the bone adjacent to the TMJ implant.
Table 1 Material properties.
2. Material and methods
based on a CAD mandible model that was used to construct the finite element model. Next, the stresses (s), strains (ε), and displacements (DL) were obtained. The Dε and DL were analyzed around the screw fixation to the bone. If the Dε were more than 15% higher than in the previous analysis, the analysis was stopped; otherwise the mean and maximal strain values were analyzed. If these results were higher than 4000 mε, we left the screw in place; if not, we removed the screw and the hole with the lowest strains and started once again the procedure shown in Fig. 3.
One clean cadaveric mandible, without teeth, of a 45-year-old woman was analyzed. The mandible geometry was obtained through CT images with a resolution of 0.780 78 0.25 mm. The model reconstruction was obtained using Simpleware software Scan IP and then converted to a solid model, using CAD software mes CATIA V5. In this process, both bone structures Dassault Syste were considered, the cancellous bone defined between 600 and 1300 HU and the cortical bone between 1300 and 1600 HU (Bujtar et al., 2010; Bujt ar et al., 2014). A commercial condylar implant (Christensen Prosthesis TMJ Implants, Inc., CO) with 9 screw holes was placed on the left ramus of the mandible (Ramos and Mesnard, 2014a,b). The implant was fixed with Ø 2.0-mm and 8-mm-long bicortical screws as in previous studies (Mesnard et al., 2011; Ramos et al., 2011, 2014). All of the screws had the same length and diameter (Fig. 1). The materials were considered isotropic and linear elastic (Table 1) in accordance with previous studies (Hsu et al., 2011; Mesnard et al., 2014). Finite element analyses were performed usmes software CATIA V5 simulation module ing the Dassault Syste with four-node tetrahedrons. We considered that both condyles were fixed in the X and Y directions and the incisor tooth fixed in the Y and Z directions, allowing only rotation of the model as shown in Fig. 2. The forces applied were based on previous studies (Ramos et al., 2010, 2011) with respect to the five most relevant muscle forces acting on the mandible and insertion points, as defined by magnetic resonance imaging (MRI). The actions and boundary conditions are presented in Fig. 2 and muscle action magnitude is shown in Table 2. The cortical and cancellous bone structures were considered to be glued, whereas the screweimplant and screwebone interfaces were considered to be in contact with a friction coefficient of 0.1 and 0.3, respectively. At their interface, the implant and the bone were modeled as surface-to-surface contact elements. In this study, seven different stages of evolution in screw fixation were considered. The algorithm used as a decisive factor was
Component
Material
E (GPa)
n
Cortical bone Cancellous bone Christensen implant Screws
Cortical bone Cancellous bone Titanium Titanium
13.0 1.6 110.0 110.0
0.3 0.3 0.3 0.3
Fig. 2. Loads and restrictions applied to the mandible.
Fig. 1. FEM of the implanted mandible.
A. Ramos et al. / Journal of Cranio-Maxillo-Facial Surgery 43 (2015) 469e474 Table 2 Muscle-loading actions. Muscle
Deep masseter Superficial masseter Medial pterygoid Temporalis Medial temporal
Ref.
DM SM MP T MT
Load (N) x
y
z
7.78 12.87 140.38 0.06 0.97
127.23 183.50 237.80 0.37 5.68
22.68 12.11 77.30 0.13 7.44
3. Results In the initial stage, we analyzed strain distribution with all the screws in place. In this case, we observed that screws #1, #6, #8, and #9 registered the highest strain values, whereas screws #2 and #3 registered the lowest strain values. Based on these results, and since the third screw did not influence the fixation performance,
471
we decided to remove screw #3 and then to re-evaluate the model behavior. Without screw #3, we observed that the mean strain values behaved differently. They increased in the adjacent bone around screws #1, #2, and #9. At this point, the greatest increase, about 8%e10%, was registered near screw #1. In contrast, the screw that registered the greatest decrease was screw #8. However, the screw that showed the lowest strain values was screw #5. Without screws #3 and #5, different behavior was observed in the other screw positions. Almost every hole increased its strain values except screw hole #7, in which there was a decrease. Screws #4 and #6 registered an increase of 12% and 10%, respectively; however, screw #2 presented the lowest strain values. The third step was therefore to remove screw #2 and to reevaluate the model. Without screws #2, #3, and #5, we observed a sharp decrease in the strain values in the bone surrounding screw #4 (~8%), but it was screw #8 that fulfilled the conditions for removal. At this stage, we also observed that in the bone
Fig. 3. Method to define screw fixation evolution.
472
A. Ramos et al. / Journal of Cranio-Maxillo-Facial Surgery 43 (2015) 469e474
21% and 30% in the bone adjacent to these screws. At this point, the bone surrounding screw #7 registered the highest increase. Micromotions at the boneeimplant interface were also measured to observe implant stability. Three different measuring stages were once more considered: an initial stage, a middle stage, and a final stage. In the initial stage, at the top of the implant we observed that micromotions were 25%e150% higher than in the final stage (Fig. 5) but with lower magnitude. We observed a rotation of the plate when some screws were removed. At the end of the implant, micromotions in the initial stage were 75% lower than in the final stage (Fig. 6). 4. Discussion
Fig. 4. Results of screw fixation evolution.
surrounding screw #7 there was a 3% increase in the principal strains, as at this stage this was one of the screws that was most responsible for the implant fixation, the others being screws #1 and #9 (Fig. 4). After removing half (i.e., four) of the screws, we observed that the bone surrounding screw #1 registered 688 mε for the ε1, and that screw #9 registered 969 mε. Positioned at the upper and lower extremities of the implant, these two screws had the main responsibility for load transfer, compensating for the screws that had been removed. In contrast, the bone surrounding screw #4 made only a small contribution to implant fixation and was thus the most likely candidate for removal. With just four screws (#1, #6, #7, and #9), we once more assessed which one could be removed next. Based on the principal strain results, we verified that the bone adjacent to screw #6 had the lowest strain values. This screw registered 513 mε and 777 mε for maximal and minimal principal strain, respectively. These values were 22% and 2% lower than the next lowest values registered in the bone surrounding screw #7, which indicated that screw #6 was the one to be removed. With only three screws (#1, #7, and #9), we analyzed the model behavior in terms of principal strains and decided to stop removing screws, since the strain values were reaching undesired values (>4000 mε) (Roberts et al., 2004), which could compromise the bone, causing hypertrophy or even fracture. Based on the results obtained with FEA (Finite Element Analysis), considering screws #1, #7, and #9 only, we can state that from the initial stage to the final stage there was a strain increase of between
The screws used to fixate the mandibular implant to the bone are subject to loads from daily activities, and, if these are excessive, there is a real chance that they may damage the bone, which could then lead to implant failure (Arabshahi et al., 2011). From the results obtained with FEA (Finite element Analysis), we can say that from the initial stage to the final stage, strains increased between 21% and 30%, improving the fixation around the screws positions. According to Roberts et al. (2004), strains lower than 200 mε cause atrophy, since the bone is not subjected to a relevant stimulation. On the other hand, if strains are higher than 4000 mε, this leads to fatigue failure. Based on the assumptions of Roberts et al. (2004), the ideal strains that guarantee optimal peak strain levels in the bone are between 200 and 2500 mε. Our results suggest that the mean strain values around the screw holes in the initial stage are within the ideal strain range; however, in some places around screw holes #1 to #7, there is a risk of atrophy. These results lead us to believe that some of the screws serve no purpose, and their removal could be supported by the remaining ones (Hsu et al., 2011). In the final stage, we left screws #1, #7, and #9, which are responsible for implant fixation and which guarantee implant stability. At this point, we observed that the problem of atrophy was out of the question, since the lower peak strain registered was about 240 mε. At this stage we noticed that the strains in a few places around the screw holes reached a peak of 2880 mε for screw #1, 2540 mε for screw #7, and 3180 mε for screw #9. These values were within the limit for bone ingrowth, but could eventually lead to local hypertrophy (Roberts et al., 2004). Comparing our results with those of others, we observe that Hsu et al. (2011) stated that three screws are more than enough to ensure a stable fixation of the implant. Also, Van Loon et al. (2000) reported, in their study of sheep, that a minimum of three screws rigidly connected to the TMJ condylar prosthesis provide a stable implant fixation, which corroborates our findings.
Fig. 5. Minimum principal strain at each stage: (1) Initial stage; (2) Middle stage; (3) Final stage.
A. Ramos et al. / Journal of Cranio-Maxillo-Facial Surgery 43 (2015) 469e474
473
Fig. 6. Micromotions at the bone-implant interface in the three different stages.
Although the studies by Hsu et al. (2011) and van Loon et al. (2000) present the same final results, confirming that just three screws are necessary to fixate the implant to the bone, it is important to note that the implants that they used were slightly different from ours, since they used a Christensen TMJ condylar prosthesis with 11 screw holes, whereas our implant had just nine screw positions and presents less contact surface with condyle. Arabshahi et al. (2011), who compared eight different TMJ implant solutions, state that, despite some differences, microstrain distribution in the bone at the screw locations is insignificant. These authors also register maximum strain at the first hole (~4031 mε) (Arabshahi et al., 2011). This is in agreement with our results; however, the highest strain registered is higher than in our study. Chowdhury et al. (2011), in their comparison of strain distribution for different screw dimensions and orientations in TMJ implants, state that the highest strains registered was between 3700 and 4400 mε for a parallel screw orientation, and between 4200 and 6300 mε for a zig-zag orientation, which means in positions #3, #6, and #9. These results are higher than ours, but nevertheless they are important for understanding that a parallel screw orientation gives lower strains than a zig-zag orientation, and thus confirm our results. To analyze implant stability, it is also important to evaluate what happens in terms of micromotions at the boneeimplant interface. Few studies have considered this point. One of these was conducted by Hsu et al. (2011) to evaluate the effect of screws on a TMJ condylar prosthesis. This author stated that the relative micromotions between bone and implant were higher with 3 screws in line than with just 2 screws. Our analysis stopped at the minimum of 3 screws because we also needed to consider strain distribution in the model, which, with less than 3 screws, is too high and could fracture the bone. We therefore decided to evaluate micromotions at three different stages: an initial stage with all screws; a middle stage, without screws #2, #3, and #5; and a final stage with just three screws, namely, #1, #7, and #9. We observed that in the initial stage the relative micromotions were not greater than 9 mm, which is lower than the threshold for bone ingrowth at the surface of the implant (Hsu, 2006; Chong et al., 2010; Hsu et al., 2011). On the other hand, in the final stage, with only 3 screws left, we observed that the micromotions increased to reach 14 mm. From these observations, we conclude that, by using just three screws, we achieve the desired range of micromotions, thus promoting bone ingrowth and adhesion in short time. Hsu et al. (2011)
state that the maximum micromotions achieved in their study were less than 6 mm, a figure that is lower than ours and prevents bone ingrowth.
5. Conclusions The results indicate that the minimum number of screws that should be used to fixate the implant to the bone is three, in positions #1, #7, and #9. We conclude that increasing the number of screws beyond three just slightly improves the implant stability; however, more screws would also contribute to reducing bone ingrowth and bone adhesion to the condyle component.
Acknowledgements The authors gratefully acknowledge FCT (Portuguese Foundation for Science and Technology) funding from the PTDC/EME-PME/ 112977/2009.
References Arabshahi Z, Kashani J, Kadir M, Azari A, Lee G: Influence of thickness and contact surface geometry of condylar stem of TMJ implant on its stability. In: 2011 International Conference on Physics Science and Technology (Icpst), vol. 22; 2011, 414e419, 2011 Bujtar P, Sandor G, Bojtos A, Szucs A, Barabas J: Finite element analysis of the human mandible at 3 different stages of life. Oral Surgery Oral Medicine Oral Pathology Oral Radiology and Endodontology 110(3): 301e309, 2010 r P, Simonovics J, V Bujta aradi K, S andor GKB, Avery CME: The biomechanical aspects of reconstruction for segmental defects of the mandible: a finite element study to assess the optimisation of plate and screw factors. Journal of CranioMaxillofacial Surgery 42(6): 855e862, 2014 Chase D, Hudson J, Gerard D, Russell R, Chambers K, Curry J, et al: The christensen prosthesisda retrospective clinical study. Oral Surgery Oral Medicine Oral Pathology Oral Radiology and Endodontics 80(3): 273e278, 1995 Chong D, Hansen U, Amis A: Analysis of bone-prosthesis interface micromotion for cementless tibial prosthesis fixation and the influence of loading conditions. Journal of Biomechanics 43(6): 1074e1080, 2010 Chowdhury A, Kashi A, Saha S: A comparison of stress distributions for different surgical procedures, screw dimensions and orientations for a temporomandibular joint implant. Journal of Biomechanics 44(14): 2584e2587, 2011 Guarda-Nardini L, Manfredini D, Ferronato G: Total temporomandibular joint replacement: a clinical case with a proposal for post-surgical rehabilitation. Journal of Cranio-Maxillofacial Surgery 36(7): 403e409, 2008 Hsu J, Lai K, Chen Q, Zobitz M, Huang H, An K, et al: The relation between micromotion and screw fixation in acetabular cup. Computer Methods and Programs in Biomedicine 84(1): 34e41, 2006
474
A. Ramos et al. / Journal of Cranio-Maxillo-Facial Surgery 43 (2015) 469e474
Hsu J, Huang H, Tsai M, Fuh L, Tu M: Effect of screw fixation on temporomandibular joint condylar prosthesis. Journal of Oral and Maxillofacial Surgery 69(5): 1320e1328, 2011 Kanatas A, Needs C, Smith A, Moran A, Jenkins G, Worrall S: Short-term outcomes using the Christensen patient-specific temporomandibular joint implant system: a prospective study. British Journal of Oral & Maxillofacial Surgery 50(2): 149e153, 2012 Mercuri L, Giobbie-Hurder A: Long-term outcomes after total alloplastic temporomandibular joint reconstruction following exposure to failed materials. Journal of Oral and Maxillofacial Surgery 62(9): 1088e1096, 2004 Mesnard M, Ramos A, Ballu A, Morlier J, Simoes J: Biomechanical analysis comparing natural and alloplastic temporomandibular joint replacement using a finite element model. Journal of Oral and Maxillofacial Surgery 69(4): 1008e1017, 2011 ~ es JA: Influences of implant condyle geometry on bone Mesnard M, Ramos A, Simo and screw strains in a temporomandibular implant. Journal of CranioMaxillofacial Surgery 42(3): 194e200, 2014 Mishima K, Yamada T, Sugahara T: Evaluation of respiratory status and mandibular movement after total temporomandibular joint replacement in patients with rheumatoid arthritis. International Journal of Oral and Maxillofacial Surgery 32(3): 275e279, 2003 Okeson J: Current terminology and diagnostic classification schemes. Oral Surgery Oral Medicine Oral Pathology Oral Radiology and Endodontics 83(1): 61e64,1997 Ramos A, Mesnard M: Comparison of load transfers in TMJ replacement using a standard and a custom-made temporal component. Journal of CranioMaxillofacial Surgery 42(8): 1766e1772, 2014a Ramos A, Mesnard M: Load transfer in Christensen® TMJ in alloplastic total joint replacement for two different mouth apertures. Journal of Cranio-Maxillofacial Surgery 42(7): 1442e1449, 2014b Ramos A, Completo A, Relvas C, Simoes J, Mesnard M, Ballu A, et al: The strain patterns of the mandible for different loadings and mouth apertures. Biodental Engineering, CRC press, pp: 133e137, 2010
Ramos A, Completo A, Relvas C, Mesnard M, Simoes J: Straight, semi-anatomic and anatomic TMJ implants: the influence of condylar geometry and bone fixation screws. Journal of Cranio-Maxillofacial Surgery 39(5): 343e350, 2011 ~es JA: Theoretical assessment of Ramos A, Mesnard M, Relvas C, Completo A, Simo an intramedullary condylar component versus screw fixation for the condylar component of a hemiarthroplasty alloplastic TMJ replacement system. Journal of Cranio-Maxillofacial Surgery 42(2): 169e174, 2014 Roberts WE, Huja S, Roberts JA: Bone modeling: biomechanics, molecular mechanisms, and clinical perspectives. Seminars in Orthodontics 10(2): 123e161, 2004 Sidebottom AJ, Gruber E: One-year prospective outcome analysis and complications following total replacement of the temporomandibular joint with the TMJ Concepts system. British Journal of Oral and Maxillofacial Surgery 51(7): 620e624, 2013 Schuurhuis JM, Dijkstra PU, Stegenga B, de Bont LGM, Spijkervet FKL: Groningen temporomandibular total joint prosthesis: an 8-year longitudinal follow-up on function and pain. Journal of Cranio-Maxillofacial Surgery 40(8): 815e820, 2012 Shen P, Zhang SY, Yang C, Yun B: Stability study of total temporomandibular joint replacement on sheep. Journal of Cranio-Maxillofacial Surgery 42(7): 1265e1270, 2014 Solberg W, Woo M, Houston J: Prevalence of mandibular dysfunction in young adults. Journal of the American Dental Association 98(1): 25e34, 1979 Tanaka E, Koolstra J: Biomechanics of the temporomandibular joint. Journal of Dental Research 87(11): 989e991, 2008 van Loon J-P, de Bont LGM, Boering G: Evaluation of temporomandibular joint prostheses: review of the literature from 1946 to 1994 and implications for future prosthesis designs. Journal of Oral and Maxillofacial Surgery 53(9): 984e996, 1995 van Loon J, de Bont L, Spijkervet F, Verkerke G, Liem R: A short-term study in sheep with the Groningen temporomandibular joint prosthesis. International Journal of Oral and Maxillofacial Surgery 29(5): 315e324, 2000
Contents lists available at ScienceDirect
Journal of Cranio-Maxillo-Facial Surgery journal homepage: www.jcmfs.com
Prediction at long-term condyle screw fixation of temporomandibular joint implant: A numerical study A. Ramos a, b, *, R.J. Duarte a, b, M. Mesnard b a b
TEMA Center of Mechanical Technology and Automation, Department of Mechanical Engineering, University of Aveiro, Portugal Universit e de Bordeaux, Institut de M ecanique et d'Ing enierie, CNRS UMR 5295, Talence, France
a r t i c l e i n f o
a b s t r a c t
Article history: Paper received 13 September 2014 Accepted 16 February 2015 Available online 24 February 2015
The fixation of commercial temporomandibular joint (TMJ) implant is accomplished by using screws, which, in some cases, can lead to loosening of the implant. The aim of this study was to predict the evolution of fixation success of a TMJ. Numerical models using a Christensen TMJ implant were developed to analyze strain distributions in the adjacent mandibular bone. The geometry of a human mandible was developed based on computed tomography (CT) scans from a cadaveric mandible on which a TMJ implant was subsequently placed. In this study, the five most important muscle forces acting were applied and the anatomical conditions replicated. The evolution of fixation was defined according to bone response methodology focused in strain distribution around the screws. Strain and micromotions were analyzed to evaluate implant stability, and the evolution process conduct at three different stages: start with all nine screws in place (initial stage); middle stage, with three screws removed (middle stage), and end stage, with only three screws in place (final stage). With regard to loosening, the implant success fixation changed the strains in the bone between 21% and 30%, when considering the last stage. The most important screw positions were #1, #7, and #9. It was observed that, despite the commercial Christensen TMJ implant providing nine screw positions for fixation, only three screws were necessary to ensure implant stability and fixation success. © 2015 European Association for Cranio-Maxillo-Facial Surgery. Published by Elsevier Ltd. All rights reserved.
Keywords: Micromotions Screw fixation Strains TMJ implant stability Long term
1. Introduction The temporomandibular joint (TMJ) plays an extremely important role in daily activities; and since it is involved in many different everyday tasks such as communicating, feeding, or even sleeping, with up to 2000 motion cycles per day, it is consequently the most exercised joint in the human body (Guarda-Nardini et al., 2008; Tanaka and Koolstra, 2008). According to previous studies, 20%e 40% of the population presents with TMJ disorders. However, contrary to what would be expected, TMJ implants have been studied mostly in terms of clinical cases and without the use of numerical predictions (Solberg et al., 1979; Okeson, 1997; Hsu et al., 2006). Temporomandibular joint diseases affect almost half of the population during their lifetime, and although some of these
* Corresponding author. Biomechanics Research Group, University of Aveiro, 3810-193 Aveiro, Portugal. Tel.: þ351 234 370830; fax: þ351 234 370953. E-mail address: [email protected] (A. Ramos).
problems can be treated with drugs and physiotherapy, some individuals still need a TMJ implant. For this reason and because the TMJ implant success rate has been much lower than for total hip replacement or total knee replacement (van Loon et al., 1995; Hsu et al., 2011), this study attempted to investigate the best TMJ implant fixation position (Chase et al., 1995; Mishima et al., 2003; Mercuri and Giobbie-Hurder, 2004). Surgical treatment for a TMJ implant includes a total or partial condylectomy and the replacement of the disk by autografts or alloplastic materials (Chase et al., 1995) Nowadays, there are tree TMJ implants as options available on the market using screws to fixate the mandible part of the implant to the bone (Kanatas et al., 2012; Schuurhuis et al., 2012). Because the structures involved in this joint are very complex, the best implant fixation technique and the optimal number of screws remain uncertain. However, implant stability (Sidebottom and Gruber, 2013; Shen et al., 2014) is one of the most important and decisive factors in implant success (van Loon et al., 1995). The surgeon fixes, with screws, the mandible r implants as lateral plates in the TMJ implants to the bone (Bujta et al., 2014; Shen et al., 2014), normally using all or almost all of
http://dx.doi.org/10.1016/j.jcms.2015.02.013 1010-5182/© 2015 European Association for Cranio-Maxillo-Facial Surgery. Published by Elsevier Ltd. All rights reserved.
470
A. Ramos et al. / Journal of Cranio-Maxillo-Facial Surgery 43 (2015) 469e474
the available screw positions. However, this does not mean that the best implant stability is guaranteed (Hsu et al., 2006; Chowdhury et al., 2011; Hsu et al., 2011), and low levels of micromotion generated at the interface can hamper bone integration. The aim of this study was to analyze implant stability, simulating a long period of support as a function of the screw positions and strain distributions in the bone adjacent to the TMJ implant.
Table 1 Material properties.
2. Material and methods
based on a CAD mandible model that was used to construct the finite element model. Next, the stresses (s), strains (ε), and displacements (DL) were obtained. The Dε and DL were analyzed around the screw fixation to the bone. If the Dε were more than 15% higher than in the previous analysis, the analysis was stopped; otherwise the mean and maximal strain values were analyzed. If these results were higher than 4000 mε, we left the screw in place; if not, we removed the screw and the hole with the lowest strains and started once again the procedure shown in Fig. 3.
One clean cadaveric mandible, without teeth, of a 45-year-old woman was analyzed. The mandible geometry was obtained through CT images with a resolution of 0.780 78 0.25 mm. The model reconstruction was obtained using Simpleware software Scan IP and then converted to a solid model, using CAD software mes CATIA V5. In this process, both bone structures Dassault Syste were considered, the cancellous bone defined between 600 and 1300 HU and the cortical bone between 1300 and 1600 HU (Bujtar et al., 2010; Bujt ar et al., 2014). A commercial condylar implant (Christensen Prosthesis TMJ Implants, Inc., CO) with 9 screw holes was placed on the left ramus of the mandible (Ramos and Mesnard, 2014a,b). The implant was fixed with Ø 2.0-mm and 8-mm-long bicortical screws as in previous studies (Mesnard et al., 2011; Ramos et al., 2011, 2014). All of the screws had the same length and diameter (Fig. 1). The materials were considered isotropic and linear elastic (Table 1) in accordance with previous studies (Hsu et al., 2011; Mesnard et al., 2014). Finite element analyses were performed usmes software CATIA V5 simulation module ing the Dassault Syste with four-node tetrahedrons. We considered that both condyles were fixed in the X and Y directions and the incisor tooth fixed in the Y and Z directions, allowing only rotation of the model as shown in Fig. 2. The forces applied were based on previous studies (Ramos et al., 2010, 2011) with respect to the five most relevant muscle forces acting on the mandible and insertion points, as defined by magnetic resonance imaging (MRI). The actions and boundary conditions are presented in Fig. 2 and muscle action magnitude is shown in Table 2. The cortical and cancellous bone structures were considered to be glued, whereas the screweimplant and screwebone interfaces were considered to be in contact with a friction coefficient of 0.1 and 0.3, respectively. At their interface, the implant and the bone were modeled as surface-to-surface contact elements. In this study, seven different stages of evolution in screw fixation were considered. The algorithm used as a decisive factor was
Component
Material
E (GPa)
n
Cortical bone Cancellous bone Christensen implant Screws
Cortical bone Cancellous bone Titanium Titanium
13.0 1.6 110.0 110.0
0.3 0.3 0.3 0.3
Fig. 2. Loads and restrictions applied to the mandible.
Fig. 1. FEM of the implanted mandible.
A. Ramos et al. / Journal of Cranio-Maxillo-Facial Surgery 43 (2015) 469e474 Table 2 Muscle-loading actions. Muscle
Deep masseter Superficial masseter Medial pterygoid Temporalis Medial temporal
Ref.
DM SM MP T MT
Load (N) x
y
z
7.78 12.87 140.38 0.06 0.97
127.23 183.50 237.80 0.37 5.68
22.68 12.11 77.30 0.13 7.44
3. Results In the initial stage, we analyzed strain distribution with all the screws in place. In this case, we observed that screws #1, #6, #8, and #9 registered the highest strain values, whereas screws #2 and #3 registered the lowest strain values. Based on these results, and since the third screw did not influence the fixation performance,
471
we decided to remove screw #3 and then to re-evaluate the model behavior. Without screw #3, we observed that the mean strain values behaved differently. They increased in the adjacent bone around screws #1, #2, and #9. At this point, the greatest increase, about 8%e10%, was registered near screw #1. In contrast, the screw that registered the greatest decrease was screw #8. However, the screw that showed the lowest strain values was screw #5. Without screws #3 and #5, different behavior was observed in the other screw positions. Almost every hole increased its strain values except screw hole #7, in which there was a decrease. Screws #4 and #6 registered an increase of 12% and 10%, respectively; however, screw #2 presented the lowest strain values. The third step was therefore to remove screw #2 and to reevaluate the model. Without screws #2, #3, and #5, we observed a sharp decrease in the strain values in the bone surrounding screw #4 (~8%), but it was screw #8 that fulfilled the conditions for removal. At this stage, we also observed that in the bone
Fig. 3. Method to define screw fixation evolution.
472
A. Ramos et al. / Journal of Cranio-Maxillo-Facial Surgery 43 (2015) 469e474
21% and 30% in the bone adjacent to these screws. At this point, the bone surrounding screw #7 registered the highest increase. Micromotions at the boneeimplant interface were also measured to observe implant stability. Three different measuring stages were once more considered: an initial stage, a middle stage, and a final stage. In the initial stage, at the top of the implant we observed that micromotions were 25%e150% higher than in the final stage (Fig. 5) but with lower magnitude. We observed a rotation of the plate when some screws were removed. At the end of the implant, micromotions in the initial stage were 75% lower than in the final stage (Fig. 6). 4. Discussion
Fig. 4. Results of screw fixation evolution.
surrounding screw #7 there was a 3% increase in the principal strains, as at this stage this was one of the screws that was most responsible for the implant fixation, the others being screws #1 and #9 (Fig. 4). After removing half (i.e., four) of the screws, we observed that the bone surrounding screw #1 registered 688 mε for the ε1, and that screw #9 registered 969 mε. Positioned at the upper and lower extremities of the implant, these two screws had the main responsibility for load transfer, compensating for the screws that had been removed. In contrast, the bone surrounding screw #4 made only a small contribution to implant fixation and was thus the most likely candidate for removal. With just four screws (#1, #6, #7, and #9), we once more assessed which one could be removed next. Based on the principal strain results, we verified that the bone adjacent to screw #6 had the lowest strain values. This screw registered 513 mε and 777 mε for maximal and minimal principal strain, respectively. These values were 22% and 2% lower than the next lowest values registered in the bone surrounding screw #7, which indicated that screw #6 was the one to be removed. With only three screws (#1, #7, and #9), we analyzed the model behavior in terms of principal strains and decided to stop removing screws, since the strain values were reaching undesired values (>4000 mε) (Roberts et al., 2004), which could compromise the bone, causing hypertrophy or even fracture. Based on the results obtained with FEA (Finite Element Analysis), considering screws #1, #7, and #9 only, we can state that from the initial stage to the final stage there was a strain increase of between
The screws used to fixate the mandibular implant to the bone are subject to loads from daily activities, and, if these are excessive, there is a real chance that they may damage the bone, which could then lead to implant failure (Arabshahi et al., 2011). From the results obtained with FEA (Finite element Analysis), we can say that from the initial stage to the final stage, strains increased between 21% and 30%, improving the fixation around the screws positions. According to Roberts et al. (2004), strains lower than 200 mε cause atrophy, since the bone is not subjected to a relevant stimulation. On the other hand, if strains are higher than 4000 mε, this leads to fatigue failure. Based on the assumptions of Roberts et al. (2004), the ideal strains that guarantee optimal peak strain levels in the bone are between 200 and 2500 mε. Our results suggest that the mean strain values around the screw holes in the initial stage are within the ideal strain range; however, in some places around screw holes #1 to #7, there is a risk of atrophy. These results lead us to believe that some of the screws serve no purpose, and their removal could be supported by the remaining ones (Hsu et al., 2011). In the final stage, we left screws #1, #7, and #9, which are responsible for implant fixation and which guarantee implant stability. At this point, we observed that the problem of atrophy was out of the question, since the lower peak strain registered was about 240 mε. At this stage we noticed that the strains in a few places around the screw holes reached a peak of 2880 mε for screw #1, 2540 mε for screw #7, and 3180 mε for screw #9. These values were within the limit for bone ingrowth, but could eventually lead to local hypertrophy (Roberts et al., 2004). Comparing our results with those of others, we observe that Hsu et al. (2011) stated that three screws are more than enough to ensure a stable fixation of the implant. Also, Van Loon et al. (2000) reported, in their study of sheep, that a minimum of three screws rigidly connected to the TMJ condylar prosthesis provide a stable implant fixation, which corroborates our findings.
Fig. 5. Minimum principal strain at each stage: (1) Initial stage; (2) Middle stage; (3) Final stage.
A. Ramos et al. / Journal of Cranio-Maxillo-Facial Surgery 43 (2015) 469e474
473
Fig. 6. Micromotions at the bone-implant interface in the three different stages.
Although the studies by Hsu et al. (2011) and van Loon et al. (2000) present the same final results, confirming that just three screws are necessary to fixate the implant to the bone, it is important to note that the implants that they used were slightly different from ours, since they used a Christensen TMJ condylar prosthesis with 11 screw holes, whereas our implant had just nine screw positions and presents less contact surface with condyle. Arabshahi et al. (2011), who compared eight different TMJ implant solutions, state that, despite some differences, microstrain distribution in the bone at the screw locations is insignificant. These authors also register maximum strain at the first hole (~4031 mε) (Arabshahi et al., 2011). This is in agreement with our results; however, the highest strain registered is higher than in our study. Chowdhury et al. (2011), in their comparison of strain distribution for different screw dimensions and orientations in TMJ implants, state that the highest strains registered was between 3700 and 4400 mε for a parallel screw orientation, and between 4200 and 6300 mε for a zig-zag orientation, which means in positions #3, #6, and #9. These results are higher than ours, but nevertheless they are important for understanding that a parallel screw orientation gives lower strains than a zig-zag orientation, and thus confirm our results. To analyze implant stability, it is also important to evaluate what happens in terms of micromotions at the boneeimplant interface. Few studies have considered this point. One of these was conducted by Hsu et al. (2011) to evaluate the effect of screws on a TMJ condylar prosthesis. This author stated that the relative micromotions between bone and implant were higher with 3 screws in line than with just 2 screws. Our analysis stopped at the minimum of 3 screws because we also needed to consider strain distribution in the model, which, with less than 3 screws, is too high and could fracture the bone. We therefore decided to evaluate micromotions at three different stages: an initial stage with all screws; a middle stage, without screws #2, #3, and #5; and a final stage with just three screws, namely, #1, #7, and #9. We observed that in the initial stage the relative micromotions were not greater than 9 mm, which is lower than the threshold for bone ingrowth at the surface of the implant (Hsu, 2006; Chong et al., 2010; Hsu et al., 2011). On the other hand, in the final stage, with only 3 screws left, we observed that the micromotions increased to reach 14 mm. From these observations, we conclude that, by using just three screws, we achieve the desired range of micromotions, thus promoting bone ingrowth and adhesion in short time. Hsu et al. (2011)
state that the maximum micromotions achieved in their study were less than 6 mm, a figure that is lower than ours and prevents bone ingrowth.
5. Conclusions The results indicate that the minimum number of screws that should be used to fixate the implant to the bone is three, in positions #1, #7, and #9. We conclude that increasing the number of screws beyond three just slightly improves the implant stability; however, more screws would also contribute to reducing bone ingrowth and bone adhesion to the condyle component.
Acknowledgements The authors gratefully acknowledge FCT (Portuguese Foundation for Science and Technology) funding from the PTDC/EME-PME/ 112977/2009.
References Arabshahi Z, Kashani J, Kadir M, Azari A, Lee G: Influence of thickness and contact surface geometry of condylar stem of TMJ implant on its stability. In: 2011 International Conference on Physics Science and Technology (Icpst), vol. 22; 2011, 414e419, 2011 Bujtar P, Sandor G, Bojtos A, Szucs A, Barabas J: Finite element analysis of the human mandible at 3 different stages of life. Oral Surgery Oral Medicine Oral Pathology Oral Radiology and Endodontology 110(3): 301e309, 2010 r P, Simonovics J, V Bujta aradi K, S andor GKB, Avery CME: The biomechanical aspects of reconstruction for segmental defects of the mandible: a finite element study to assess the optimisation of plate and screw factors. Journal of CranioMaxillofacial Surgery 42(6): 855e862, 2014 Chase D, Hudson J, Gerard D, Russell R, Chambers K, Curry J, et al: The christensen prosthesisda retrospective clinical study. Oral Surgery Oral Medicine Oral Pathology Oral Radiology and Endodontics 80(3): 273e278, 1995 Chong D, Hansen U, Amis A: Analysis of bone-prosthesis interface micromotion for cementless tibial prosthesis fixation and the influence of loading conditions. Journal of Biomechanics 43(6): 1074e1080, 2010 Chowdhury A, Kashi A, Saha S: A comparison of stress distributions for different surgical procedures, screw dimensions and orientations for a temporomandibular joint implant. Journal of Biomechanics 44(14): 2584e2587, 2011 Guarda-Nardini L, Manfredini D, Ferronato G: Total temporomandibular joint replacement: a clinical case with a proposal for post-surgical rehabilitation. Journal of Cranio-Maxillofacial Surgery 36(7): 403e409, 2008 Hsu J, Lai K, Chen Q, Zobitz M, Huang H, An K, et al: The relation between micromotion and screw fixation in acetabular cup. Computer Methods and Programs in Biomedicine 84(1): 34e41, 2006
474
A. Ramos et al. / Journal of Cranio-Maxillo-Facial Surgery 43 (2015) 469e474
Hsu J, Huang H, Tsai M, Fuh L, Tu M: Effect of screw fixation on temporomandibular joint condylar prosthesis. Journal of Oral and Maxillofacial Surgery 69(5): 1320e1328, 2011 Kanatas A, Needs C, Smith A, Moran A, Jenkins G, Worrall S: Short-term outcomes using the Christensen patient-specific temporomandibular joint implant system: a prospective study. British Journal of Oral & Maxillofacial Surgery 50(2): 149e153, 2012 Mercuri L, Giobbie-Hurder A: Long-term outcomes after total alloplastic temporomandibular joint reconstruction following exposure to failed materials. Journal of Oral and Maxillofacial Surgery 62(9): 1088e1096, 2004 Mesnard M, Ramos A, Ballu A, Morlier J, Simoes J: Biomechanical analysis comparing natural and alloplastic temporomandibular joint replacement using a finite element model. Journal of Oral and Maxillofacial Surgery 69(4): 1008e1017, 2011 ~ es JA: Influences of implant condyle geometry on bone Mesnard M, Ramos A, Simo and screw strains in a temporomandibular implant. Journal of CranioMaxillofacial Surgery 42(3): 194e200, 2014 Mishima K, Yamada T, Sugahara T: Evaluation of respiratory status and mandibular movement after total temporomandibular joint replacement in patients with rheumatoid arthritis. International Journal of Oral and Maxillofacial Surgery 32(3): 275e279, 2003 Okeson J: Current terminology and diagnostic classification schemes. Oral Surgery Oral Medicine Oral Pathology Oral Radiology and Endodontics 83(1): 61e64,1997 Ramos A, Mesnard M: Comparison of load transfers in TMJ replacement using a standard and a custom-made temporal component. Journal of CranioMaxillofacial Surgery 42(8): 1766e1772, 2014a Ramos A, Mesnard M: Load transfer in Christensen® TMJ in alloplastic total joint replacement for two different mouth apertures. Journal of Cranio-Maxillofacial Surgery 42(7): 1442e1449, 2014b Ramos A, Completo A, Relvas C, Simoes J, Mesnard M, Ballu A, et al: The strain patterns of the mandible for different loadings and mouth apertures. Biodental Engineering, CRC press, pp: 133e137, 2010
Ramos A, Completo A, Relvas C, Mesnard M, Simoes J: Straight, semi-anatomic and anatomic TMJ implants: the influence of condylar geometry and bone fixation screws. Journal of Cranio-Maxillofacial Surgery 39(5): 343e350, 2011 ~es JA: Theoretical assessment of Ramos A, Mesnard M, Relvas C, Completo A, Simo an intramedullary condylar component versus screw fixation for the condylar component of a hemiarthroplasty alloplastic TMJ replacement system. Journal of Cranio-Maxillofacial Surgery 42(2): 169e174, 2014 Roberts WE, Huja S, Roberts JA: Bone modeling: biomechanics, molecular mechanisms, and clinical perspectives. Seminars in Orthodontics 10(2): 123e161, 2004 Sidebottom AJ, Gruber E: One-year prospective outcome analysis and complications following total replacement of the temporomandibular joint with the TMJ Concepts system. British Journal of Oral and Maxillofacial Surgery 51(7): 620e624, 2013 Schuurhuis JM, Dijkstra PU, Stegenga B, de Bont LGM, Spijkervet FKL: Groningen temporomandibular total joint prosthesis: an 8-year longitudinal follow-up on function and pain. Journal of Cranio-Maxillofacial Surgery 40(8): 815e820, 2012 Shen P, Zhang SY, Yang C, Yun B: Stability study of total temporomandibular joint replacement on sheep. Journal of Cranio-Maxillofacial Surgery 42(7): 1265e1270, 2014 Solberg W, Woo M, Houston J: Prevalence of mandibular dysfunction in young adults. Journal of the American Dental Association 98(1): 25e34, 1979 Tanaka E, Koolstra J: Biomechanics of the temporomandibular joint. Journal of Dental Research 87(11): 989e991, 2008 van Loon J-P, de Bont LGM, Boering G: Evaluation of temporomandibular joint prostheses: review of the literature from 1946 to 1994 and implications for future prosthesis designs. Journal of Oral and Maxillofacial Surgery 53(9): 984e996, 1995 van Loon J, de Bont L, Spijkervet F, Verkerke G, Liem R: A short-term study in sheep with the Groningen temporomandibular joint prosthesis. International Journal of Oral and Maxillofacial Surgery 29(5): 315e324, 2000