Effect of cantilever length and alloy framework on the stress distribution in peri-implant area of cantilevered implant-supported fixed partial dentures

ABSTRACT Because many mechanical variables are present in the oral cavity, the proper load transfer between the prosthesis and the bone is important for treatment planning and for the longevity of the implant-supported fixed partial denture. Objectives To verify the stress generated on the peri-implant area of cantilevered implant-supported fixed partial dentures and the potential effects of such variable. Material and Methods A U-shaped polyurethane model simulating the mandibular bone containing two implants (Ø 3.75 mm) was used. Six groups were formed according to the alloy’s framework (CoCr or PdAg) and the point of load application (5 mm, 10 mm and 15 mm of cantilever arm). A 300 N load was applied in pre-determined reference points. The tension generated on the mesial, lingual, distal and buccal sides of the peri-implant regions was assessed using strain gauges. Results Two-way ANOVA and Tukey statistical tests were applied showing significant differences (p<0.05) between the groups. Pearson correlation test (p<0.05) was applied showing positive correlations between the increase of the cantilever arm and the deformation of the peri-implant area. Conclusions This report demonstrated the CoCr alloy shows larger compression values compared to the PdAg alloy for the same distances of cantilever. The point of load application influences the deformation on the peri-implant area, increasing in accordance with the increase of the lever arm.


INTRODUCTION
transmitted along the framework and the abutment to the surrounding bone where most part of it is absorbed at the expenses of bone deformation.
According to Frost 8 (2004), the bone reacts to forces according to the intensity of the tension. Bone responses to tension could then be divided in four intervals or windows: 1: The acute disuse window with tensions below 50 me (micrometer), resulting in bone loss because of an increase in the remodeling process; 2: The adaptation window with tensions between 50 me and 1500 me where physiological adaptation occurs with a balance between resorption and formation; 3: The mild overload window with tensions between 1500 me and 4000 me and where an increase in the modeling process occurs, improving bone structure; and 4: The pathologic overload window characterized takes place.
According to Chang, et al. 6 (2013), knowledge regarding the response of the peri-implant bone when the dental implant is excessively loaded is limited and the level of evidence is poor. With results, it is unclear whether occlusal overload might cause marginal bone loss or total loss of osseointegration to already osseointegrated dental implants when the applied load exceeds the biologically-acceptable limit. This biological limit is also unknown. Furthermore, higher remodeling activity of the peri-implant bone is found around implants subjected to high loading forces. The strain values that can actually cause biological changes are not completely known 30 . Certain hormones and biochemical agents can also change the system, causing changes to the limits of tolerance 8 .
distally extended lever arms present peculiar characteristics of force distribution since all the force applied in the posterior region of the cantilever is transmitted to the implants and consequently to the adjacent bone 29 such as Benzing, et al. 5 (1995) and Lewinstein, et al. 16 (1995), demonstrate that the increase of the cantilever arm promotes an increase in stress concentration around the terminal implant. A cantilever arm of 10-20 mm is considered acceptable depending on the quality of the bone where implants are placed 14,20,21,27 .
According to Benzing, et al. 5 (1995), the load application on the cantilever arm of an implantsupported framework produces deformation energy in the system that causes bending, depending on the differences of elastic modulus of several materials and components. Studies have demonstrated that the pattern of stress distribution among abutments depends, among other factors, on the alloy type used for framework 2,10,11 . According to some authors, Benzing, et al. 5 al. 9 (2001) and Duyck & Naert 7 (2002), a material resistance; frameworks made with rigid basic alloys suffer less deformation, being less prone to fatigue and, consequently, not overloading the screws. Some clinical 12 and laboratory 1,13,23,26,29 studies have used CoCr alloys for implant-supported prostheses frameworks.
The clinical success of osseointegrated implants stresses are transferred from the implant to the surrounding bone without generating forces of a magnitude that would jeopardize the longevity of implants and prostheses 25 . The force applied on the transmitted to the peri-implant area. However, the magnitudes of the resultant stresses, considering the elasticity of the bone, are underestimated. The aim of this in vitro study was to verify the mechanical stress generated on the peri-implant bone of an implant prosthodontic system when: (1) a load is applied at different cantilever lengths and (2) alloys of different elastic modulus (E) are used to fabricate the framework.

MATERIAL AND METHODS
A "U" shaped polyurethane model (PU, Axson -Cergy, St. Ouen l'Aumône, France) with the following dimensions: 100 mm in length, 13 mm in width, 19 mm in height, 46 mm in internal diameter, and 59 mm in external diameter was used to simulate the mandibular bone 18,19 . Two external hexagon Brånemark System ® 3.75 mm in diameter and 13 mm in length were embedded in the model during polyurethane's liquid pouring in a matrix. After polyurethane hardening, two multi-unit abutments (Nobel Biocare length were manually screwed into the implants. A previously calibrated electronic torque controller abutment screws to 20 Ncm torque. Ltd., Tokyo, Honshu, Japan) were bonded with cyanoacrylate on the surface of the polyurethane model on the distal (D), lingual (L), mesial (M), and buccal (B) sides of implant 1 (distal) and implant 2 (mesial), as can be seen in Figure 1. Strain gauges are able to measure the tension suffered by an object or structure with which it is in close deformation of a body when submitted to a given force that can be tensile (+) or compressive (-).
The strain gauges were connected to a data acquisition device (NIcDAQ-9172 -National Instruments Corp., Austin, Texas, USA) that sent a signal to a LabVIEW 8.1 program for Windows (National Instruments Corp., Austin, Texas, USA) installed in a computer were inputs from the eight strain gauges were analyzed.
Two frameworks simulating a cantilevered The frameworks were positioned in the PU model abutments and tested manually. As observed, only frameworks that adapted well to the abutments were to be approved for the tests. The two frameworks met this criterion and therefore there was no need to repeat the casts. Thus, titanium screws were tightened to 10 Ncm using an electronic torque controller (Nobel Biocare Torque The PU model was adapted and stabilized in a cylindrical steel base. The use of this rigid metallic base aimed at not interfering with the deformation of the PU model and not absorbing the load applied during the tests. Six test groups were formed (CoCr-5mm, PdAg-5mm, CoCr-10mm, PdAg-10mm, CoCr-15mm and PdAg-15mm), according to the alloy framework and to the point of load application.
Test specimens were taken to a Universal Testing Machine (model K-2000 MP -Kratos Equipamentos Industriais Ltda., São Paulo, SP, Brazil) and baseline developed on each strain gauge were carried out prior to the load application (reading precision of order 1X10 -6 ). Before initiating the readings on the deformation caused by loading the frameworks, the output of the measuring system was set to zero to separate from the deformation caused by abutment/prosthetic screw tightening. A round steel pre-determined reference point in the framework ( Figure 1). Thus, the testing machine was set to compression at a cross-head speed of 0.5 mm/ min until it reached 300 N and stopped for one minute. The 300 N load was used to run the test according to the maximal occlusal bite force values found by Akça, et al. 3 (2006) for implant-supported prostheses in opposition to natural teeth.
Deformation readings were taken at each one of the eight strain gauges for the duration of load application and 1 minute after load stabilization. Only the last 30 values of deformation were taken into account to ensure the maximum and stable levels of deformation were recorded for each site. Load application was repeated 5 times to calculate the mean and the standard deviation.
The two-way ANOVA statistical test was applied and PdAg) and the second variable being the peri- The Tukey test was applied to compare groups regarding the effect of two types of alloy. There   Table  1 and represented in Figure 2.
With the load applied on the cantilever for all six groups, the most relevant results of compression occurred on the distal (D1) and lingual (L1) sides of implant I (distal), and on the lingual (L2) side of implant 2 (mesial). Tension occurred on the buccal (B1) side of implant I.
According to Suedam, et al. 29 (2009), we can not sum the deformation suffered in every periimplant region of each implant and consider this value as deformation of the entirety because each component of the system prosthesis/abutment/ implant/bone can be found under various conditions of adaptation and load. As a result, a quantitative and qualitative evaluation of the results based on the statistical tests becomes necessary, which give us a biomechanical behavior view of the entire system involved with, and not only of the strain gauges or of the peri-implant regions individually.
The results of the Tukey test (Table 2) demonstrated the difference in the framework's elastic modulus influenced the intensity of deformation occurred in the peri-implant region, as can be noted in Table 1 and in Figure 2. The Pearson correlation test showed positive correlation (

DISCUSSION
Knowledge on the amount of mechanical stress generated in the peri-implant area when load is applied along the cantilever arm is essential for the planning, execution and longevity of the treatment with implant-supported prostheses.
This study showed the pattern of bone deformation generated by applying a static force of 300 N varied according to: 1-The position where the strain gages were located on the peri-implant region (D1, L1, M1, B1, D2, L2, M2 and B2); 2-The point of load application (with 5 mm, 10 mm and 15 mm cantilevers); 3-Implant position relative to load application (I1 and I2); 4-Type of alloy used for making frameworks (CoCr and PdAg).
For all studied groups the behavior was singular, with tension forces present in a larger degree in   4,5,13,21,23,28,29 in cantilevered prostheses, the most distal implants represent the fulcrum and, therefore, are subjected to compression forces while intermediary abutments suffer tension. In this study, the peri-implant regions were divided in four (D, L, M and B) allowing the observation that the distal and lingual sides of the most distal implant (I1) were subject to higher values of compression forces and that these values increased as the correlation test ( Table 2).
The numerical values expressed in tension and compression for both alloys are the result of framework behavior due to load application, type of deformation and consequently the tension transmitted to the bone 2,10,11,13,17,29 . Because of the lower elastic modulus of the palladium-silver alloy compared to the cobalt-chromium alloy, resistance, the results expressed in Table 1, and in Figure 2 demonstrated that when the load was applied to the CoCr alloy groups, larger compression values were recorded compared to PdAg alloy groups, for the same distances of cantilever. According to Rubo & Souza 23 (2009) and Suedam, et al. 29 (2009), the PdAg alloy deflects more, absorbing part of the load applied to the cantilever resulting compression forces of lower intensity being transmitted to the surrounding bone. On the compared to CoCr alloy, the PdAg alloy presented the largest values of tension on the buccal side of The load is transmitted to the surrounding bone where the most part of it is absorbed at the expense of deformation in bone structure, which is the less rigid structure in the system. Physiologic levels of tension serve also the purpose of bone remodeling. This mechanism would help maintain 22 . Nevertheless, mechanical overload can lead to biological failure 24 . When a pathological overload is applied to an osseointegrated implant, tension exceeds the physiological threshold tolerated by the bone and micro fractures may occur at the implant-bone interface. Repeated overload can lead to fatigue failure of the implant-bone interface, reducing periimplant bone density and leading to the formation of bone defects such as craters. The pathologic overload window of Frost's Theory represents this situation, when bone undergoes tensions above differently to occlusal overload promoting increased bone resorption, as demonstrated by Kozlovsky, et al. 15 (2007).
The measure of tension generated in periimplant area gave us the possibility of correlating these values with the bone remodeling theory 8 in an attempt to clarify the biological process that takes place in that area, considering an ideal clinical condition. According to the polyurethane model validation studies made by Moretti, et al. 19 (2011) and Miyashiro, et al. 18 (2011), the homogeneity of polyurethane (PU) could favor its use in biomechanical studies of force distribution on implant supported prostheses, aimed at establishing correlations between strains generated in the periimplant region and physiological strains as proposed by Frost's Theory.
Nevertheless, it is known that considerable differences exist between this study and clinically integrated implants. Although polyurethane can present similar elastic modulus to bone, other This study does not claim that the strains found in the polyurethane model matches precisely to the in-vivo situation but acknowledges the biomechanical process of load transmission in an attempt to understand how bone tissue processes these transmitted loads.
A strain diagram was used as a graphic representation of the deformation readings generated on each side of the peri-implant region.
shape with scales of 0 me to 7000 me, where readings of deformations generated on the distal, lingual, mesial and buccal sides of each implants of the groups CoCr-15 mm and PdAg-15 mm are visualized (Figure 3). In these diagrams, tensions (3) Cantilever arms smaller than 15 mm must be partial dentures.