Fracture Risk of Narrow Implant Components in Upper First Premolars: A Finite Element Analysis

Dental implants are a predictable treatment for tooth loss. However, a common clinical limitation is low bone availability for regular implants. Different treatment options were created to deal with these limitations. Among the most used treatment options are bone grafts, osteogenic distraction, short implants and narrow diameter implants (NDI), which may or may not be adequate depending on bone loss situation. When we find narrow bone crest, bone grafts are conventional treatment for most professionals. Even though it is a predictable treatment, as long as modern techniques and materials are used, this therapy has morbidity, time and cost as disadvantages, sometimes even leading patients to refuse such treatment. This situation stimulates search for less invasive treatment options, looking for better solutions and upgrading treatment acceptance by the patient [1-4].

Given this context, alternatives such as NDI are opportunes, as are other options. It is necessary that it is a predictable treatment, specially when are considered inherent characteristics of these implants, such as a possibility of lower resistance because of lower material volume. Some short duration clinical research demonstrated similar results compared to conventional implants, but exclusion criteria should be utilized on these papers, given the fact that is not known NDI treatment behavior on mechanically unfavorable situations. In order to make better prognostic and indications, it is opportune to study NDI limits. Considering that are possible risks on such research, it is ethically recommended that laboratory methods are used to verify human research and utilization [5].

The present study aims to evaluate through finite element analysis narrow implant components fracture risk in upper first pre-molars unitary prosthesis with different vestibular bone loss, including zero bone width around the implant or vestibular bone fenestration [6].

A 3d model built from a public available tomography was used as the base for model construction. In order to obtain different bone situations, CAD software Solidworks (Dassault Systems, Solidworks Corps, USA) was used, maintaining only #11, #12 and #13 teeth region and its adjacent bone. In order to obtain the desired bone conditions, teeth #12 was removed and its bone surface filled. Bone loss was determined by a 3 reference spots parabola, traced from #13 distal buccal region, varying through #12 middle section and ending at #11 mid buccal region. Trying to reproduce desired bone loss, the intermediate reference spot had variable positions, in order to represent 3, 3.6 and 4mm bone width, having as reference the implant platform, which is 1,5mm deep from bone surface - as recommended by the manufacturer. All bone buccal to the line determined was removed as far to tooth #12 apex. Angles were smoothed and cortical bone was established as 1,75mm, in order to represent a type III bone according to Lekholm and Zarb - defined as a thin cortical bone situation [7,8].

Implant manufacturer provided computer models of 3,5mm and 2,9mm diameter by 10mm height commercially pure titanium cone Morse (CM SIN Unitite - SIN Sistema de Implantes, SP, BR). They were positioned at bone crest centre and metal-ceramic unitary prosthesis were modelled on them. At the 3,5mm diameter model and 3mm wide bone crest, implant was placed with buccal fenestration. Titanium Ti-6Al-4V alloy abutments were used, with 1,5mm gingival height and 4mm long. Figure 1 sums up processing and table 1 sums up final model characteristics. Furthermore, it was simulated a 5mm food bolus covering chewing surfaces and a three-contact-point antagonist structure over the food bolus to standardize masticatory load [9].

Figure 1: Processing summarized and final model exemples with utilized implants (M3 and M4). A: tomography model, B: bone loss delimitation zone (X) and available bone (Y), C: example model with final bone remodeling and 3,5mm and 2,9mm implant models.

Table 1: Analysed models variables summary

Models were exported to finite element analysis software Ansys Workbench V19.1 (Ansys Inc., Canonsburg, PA, USA). Elasticity module ( measured in Gpa) and Poisson coefficient were: 69/0,3 for feldspathic porcelain, 218/0,33 for chrome cobalt alloy, 22,4/0,25 for zinc phosphate cement, 13,7/0,3 for cortical bone, 1,37/0,3 for medullar bone, 105/0,37 for grade IV titanium, 113,8/0,342 for grade V titanium (Ti-6Al-4V), 68,9/045 for periodontal ligament, 18,6/0,31 for dentine and 0,02157/0,499 for almond food bolus. It was simulated frictional contacts between titanium surfaces with a 0,2 friction coefficient. Contact between natural teeth and ceramic crown was configured as non frictional for convenience. Other contacts were considered adhered. Implants were simulated as Osseo integrated.

Each model had a two-step simulation. First step refers to pretension application, or abutment tightening. Ansys software “bolt pretension” resource was used, simulating different union strengths between components until a titanium proportionality tension peak of 65% was achieved on the screw, in order to make a balance of resistance and retention, with a peak error margin of 1%. Second step refers to masticatory loading. Two loads were applied on each model, both axial, perpendicular to bone crest and applied to the antagonist structure, with intensity of 120N and 240N. Rigid supports were added at areas where the segment would connect to the remaining maxilla. Simulations were not linear relating to contact.

In order to minimize created finite element mesh flaws, both first and second step had finite element mesh convergence analysis realized. 10 knots Tetrahedric quadratic elements were used (solid 187), and knots/element numbers varied from 944940/578201 to 1015578/623468. Models were solved (Windows 10 64 bits, Intel i & 6800k processor, 12gb RAM memory) and peri-implanter bone analysed.

Metallic infrastructure and ceramic presented the biggest lifespan of the analysed components. Considering that chrome, cobalt have a 710 MPa traction resistance, the highest peak measured in all models was only 14% at 120N and 27% at 240N. Looking at feldspathic porcelain results, considering 69, 74 MPa flexural resistance, the highest peak in all models was only 18% at 120N and 38% at 240N. As these structures are comparatively with a low fracture risk facing the other structures, the present study will focus on other results.

Implants and abutments were analysed by von Mises criteria, due to titanium ductile characteristics. Results were also proportionally displayed considering yield strength (maximum tension before plastic deformation) of 550 MPa for the implant grade four titanium and 880 MPa for abutment grade five titanium. Table 2 and 3, graphics 1 and 2, figures 2-5 sum up results [10].

Table 2: Implant peak values results according to von Mises criteria (at MPa) and their percentage related to grade IV titanium yield strength.

Implant diameter / bone crest thickness 

 Graphic 1: Implant result peaks according to von Mises criteria 

Figure 2: Implant distal and sliced result views under 120N load. Red color represents regions where implant would plastically deform under real conditions. 

Figure 3: Implant distal and sliced result views under 240N load. Red color represents regions where implant would plastically deform under real conditions.

Figure 4: Abutments distal and sliced result views under 120N load.

Figure 5: Abutments distal and sliced result views under 240N load. Red color represents regions where the abutment would plastically deform under real conditions.

Table 3: Abutment peak value results according to von Mises criteria (MPa measured) and its related percentage to grade V titanium alloy yield strength.

Implant diameter / bone crest thickness 

 Graphic 2: Abutment peak value results according to von Mises criteria

All peaks occurred on the vestibular platform/inner Morse cone angle. That happened due to load proximity, angle geometry that favors tension concentration and a tooth/abutment flexing tendency to vestibular caused by the cantilever. Red highlighted regions shown at figures 2 and 3 represent plastic deformed regions. At a 120N load a plastic deformation would occur at 2,9mm implant group (SD group), but it would be limited to the already mentioned angle, resulting in an angle rounding without structural commitment. On a 240N load 3, 5mm implants (LD GROUP) would suffer plastic deformation on the angle surface zone and at the inner cone cervical buccal portion, while SD group would show plastic deformation zones next to the implant sidewalls. Given this context, results show a smaller structural commitment risk difference at lower loads, but as it increases, it becomes more expressive, always showing a less risky option at LD group. SD group results at 240N suggests a significant smaller life span compared to other models and loads.

Buccal bone crest variation had a smaller impact on results, being more significant at SD group. That probably occurred due to implant/abutment interaction having a bigger role on results.

About abutments, tension peaks on LD group occurred at the top threads. This occurs mostly because of the screw pre-tension that causes 65% peaks at the region before masticatory load, while during load a abutment flexion occurs elevating the shown peaks. This behavior difference is due to the abutment rigidity difference. At LD group, this rigidity causes a more proportional and homogeneous flexion through the abutment extension, while at SD group the deformation concentrated on the region over the abutment. About extension, abutment screw threads tension concentration happened superficially at LD group, while at SD group it was deeper. For reference, Abutment diameter and area at platform height is approx. 2.5mm/4, 9 mm2 on 3,5mm Implant and 2mm/3.2mm2 on 2,9mm implant.

Fatigue phenomenon is related to both tension intensity and tension variation extent. As the abutment may undergo fatigue failure, this paper cannot state that on a 120N load the LD group abutment yield limit on masticatory cycle tensile variation from 65% to 85% would cause a fatigue process faster than the 0 to 65% oscillation that occurs on the SD group. However, on 240N load SD abutments presented worse results on tension intensity and oscillations, suggesting a significantly shorter life span [11].

Just as on implants, bone condition variation had a smaller role on abutment results, showing a smaller impact on SD group.

A similar to masticatory load intensity was evaluated (120N) and a higher muscular tone  correspondent load (240N). Although it isn’t possible to estimate an exact life span, results suggest that LD group models will show a medium to long life span, depending on the patient masticatory load. On the other hand, SD group results suggest a long life span at 120N load, However smaller than LD group, but shows structural failure at  short span  when enduring 240N loads. An in vitro paper analyzing two pieces unitary  CM NDIs, 2,9mm diameter, found a 130N fatigue limit on dynamic load test and concluded that those implants are suitable only for low masticatory loads, which is in agreement to the present paper. Other in vitro fatigue studies also reported worse outcome of NDIs abutments compared to those of conventional implants [12].

Considering that SD group showed the highest peaks and those are directly linked to cervical portion diameter, an alternative would be utilizing single body implants because they don’t present cervical narrowing. However, due to bone biocompatibility, these implants are generally made of grade 4 titanium, that shows a significant lower yield limit compared to grade 5 used on the analyses abutment. It is possible the construction fully made of grade 5 titanium, but there are doubts on bone biocompatibility. Other alternative would be a single titanium and zircon body NDI. Given this context, more studies are necessary to evaluate those variables on NDIs.

Authors know only a couple other papers that using finite element analysis had a screw tightening simulation before applying load on dental implant components. Articles that simulated tightening on SDI implants wasn’t find either. Even though less important on bone outcome analysis, tightening generally results on tension peaks ranging from 65% to 75% of the abutment yield limit, a significant implant tension and a cohesive strength between those components generally much stronger than masticatory loads. As an example, at a given dental implant in vitro study, applying a 20N/cm momentum force generates a 461,6 N component union force in average. Given this context, the  pre-tension simulation could significantly distort the evaluation of component performance. Other paper comparatively evaluated  3,3mm diameter cone morse SDIs made of grade 4 titanium or zircon/titanium alloys, but without pre-tension forces and considering its structures attached, making it as single-body. Implant results showed tension concentration at the surface coronal third and superior implant threads, in contrast to the present paper that tension concentrations were predominantly in the implant inner region [13-15].

Present research characteristics consider a slightly buccal cantilever, but it is likely that in a tooth centered implant, tension distribution could be more homogeneous to implant circumference and results could be more favorable, reducing observed tension peaks. It is recommended further research in order to investigate such hypothesis [16].

Considering analysed methods and models authors concluded that

• 2,9mm diameter implant components can be used with certain previsibility and long life span as long as used on selected patients and with low masticatory effort.
• 3,5mm diameter implant components possess a long life span on low masticatory effort patients and a medium to long life span on patients with masticatory efforts over average.
• Buccal Bone width showed a significantly less influence on abutment results compared to implant diameter.

1. Chavda S, Levin L (2018) Human Studies of Vertical and Horizontal Alveolar Ridge Augmentation Comparing Different Types of Bone Graft Materials: A Systematic Review. J Oral Implantol 44: 74-84.
2. Hjalmarsson L, Gheisarifar M, Jemt T (2016) A systematic review of survival of single implants as presented in longitudinal studies with a follow-up of at least 10 years. Eur J Oral Implantol 1: 155-162.
3. Narby B, Kronström M, Söderfeldt B, Palmqvist S (2008) Changes in attitudes toward desire for implant treatment: a longitudinal study of a middle-aged and older Swedish population. Int J Prosthodont. 21: 481-485.
4. Thoma DS, Cha JK, Jung UW (2017) Treatment concepts for the posterior maxilla and mandible: short implants versus long implants in augmented bone. J Periodontal Implant Sci 47: 2-12.
5. Lambert FE, Lecloux G, Grenade C, Bouhy A, Lamy M, et al. (2015) Less Invasive Surgical Procedures Using Narrow-Diameter Implants: A Prospective Study in 20 Consecutive Patients. J Oral Implantol 41: 693-699.
6. Vasco MA, Souza JT, Las Casas EB, de Castro e Silva AL, Hecke M (2015) A method for constructing teeth and maxillary bone parametric model from clinical CT scans. Computer Methods in Biomechanics and Biomedical Engineering: Imaging & Visualization. 3: 117-22.
7. Holmes DC, Diaz-Arnold AM, Leary JM (1996) Influence of post dimension on stress distribution in dentin. J Prosthet Dent 75:140-147.
8. Lekholm, U. Z. G. A. (1985) Patient selection and preparation. Tissue-integrated protheses: osseointegration in clinical dentistry. 199-209.
9. Matweb Titanium (2016) Ti-6Al-4V (Grade 5), Annealed Matweb, Material Property Data.
10. Matweb Titanium (2016) Grade 4 Matweb, Material Property Data.
11. Zarone F, Sorrentino R, Apicella D, Valentino B, Ferrari M, et al. (2006) Evaluation of the biomechanical behavior of maxillary central incisors restored by means of endocrowns compared to a natural tooth: a 3D static linear finite elements analysis. Dent Mater 22:1035-1044.
12. Quek CE, Tan KB, Nicholls JI (2006) Load fatigue performance of a single-tooth implant abutment system: effect of diameter. Int J Oral Maxillofac Implants 21: 929-936.
13. Oatis, D (2007) Analyzing bolt pretension in the ANSYS Workbench Platform. Ansys Advantage 1: 28-29.
14. Morris HF (1990) Properties of cobalt-chromium metal ceramic alloys after heat treatment. J Prosthet Dent 63: 426-433.
15. Giordano RA 2nd, Pelletier L, Campbell S, Pober R (1995) Flexural strength of an infused ceramic, glass ceramic, and feldspathic porcelain. J Prosthet Dent 73: 411-418.
16. Reis TAD, Zancopé K, Karam FK, Neves FDD (2019) Biomechanical behavior of extra-narrow implants after fatigue and pull-out tests. J Prosthet Dent 122: 54 e51-54 e56.