- Email: [email protected]
The effect of static bone strain on implant stability and bone remodeling
Bone 49 (2011) 783–789
Contents lists available at ScienceDirect
Bone j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / b o n e
The effect of static bone strain on implant stability and bone remodeling Anders Halldin a, b,⁎, Ryo Jimbo a, d, Carina B. Johansson c, Ann Wennerberg a, d, Magnus Jacobsson a, b, Tomas Albrektsson e, f, Stig Hansson b a
Department of Prosthodontics, Faculty of Odontology, Malmö University, Malmö, Sweden Astra Tech AB, Mölndal, Sweden Dental Materials, Department of Prosthodontics, Institute of Odontology, The Sahlgrenska Academy, University of Gothenburg, Sweden d Department of Biomaterials/Handicap Research, Gothenburg University, Gothenburg, Sweden e Department of Biomaterials, Gothenburg University, Sweden f Department of Dental Materials Sciences & Technology, Malmö University, Sweden b c
a r t i c l e
i n f o
Article history: Received 21 April 2011 Revised 15 June 2011 Accepted 4 July 2011 Available online 14 July 2011 Edited by: David Burr Keywords: Static bone strain Removal torque In vivo experiment Biomechanical FE simulation Bone remodeling
a b s t r a c t Bone remodeling is a process involving both dynamic and static bone strain. Although there exist numerous studies on the effect of dynamic strain on implant stability and bone remodeling, the effect of static strain has yet to be clarified. Hence, for this purpose, the effect of static bone strain on implant stability and bone remodeling was investigated in rabbits. Based on Finite Element (FE) simulation two different test implants, with a diametrical increase of 0.15 mm (group A) and 0.05 mm (group B) creating static strains in the bone of 0.045 and 0.015 respectively, were inserted in the femur (group A) and the proximal tibia metaphysis (groups A and B respectively) of 14 rabbits to observe the biological response. Both groups were compared to control implants, with no diametrical increase (group C), which were placed in the opposite leg. At the time of surgery, the insertion torque (ITQ) was measured to represent the initial stability. The rabbits were euthanized after 24 days and the removal torque (RTQ) was measured to analyze the effect on implant stability and bone remodeling. The mean ITQ value was significantly higher for both groups A and B compared to group C regardless of the bone type. The RTQ value was significantly higher in tibia for groups A and B compared to group C while group A placed in femur presented no significant difference compared to group C. The results suggest that increased static strain in the bone not only creates higher implant stability at the time of insertion, but also generates increased implant stability throughout the observation period. © 2011 Elsevier Inc. All rights reserved.
Introduction Bone is a unique material in its ability to adapt its mass and structure in response to the loads to which it is exposed [1,2]. The loads induce strains in the bone, and the modeling and remodeling stimulus has been found to be dependent on strain magnitude, strain frequency [3], and strain rate [4]. The bone mass can either be maintained by a relatively small number of loading cycles with high strains, for example 4 loading cycles per day with a strain of 2000 microstrain [5], or by a great number of loading cycles with low strains, for example 10,000 loading cycles per day with a strain of 400 microstrain [6]. It has also been found that the osteogenic effect of strains increases with increasing strain rate [4]. The insertion of an implant with a certain torque means that static strains will be induced in the surrounding bone. The magnitude of these static strains depends on the bone anatomy, the bone quality, the
Abbreviations: RTQ, Removal torque; ITQ, Insertion torque; ROI, Region of interest. ⁎ Corresponding author at: Department of Prosthodontics, Faculty of Odontology, Malmö University, Malmö, Sweden. Fax: +46 40 6658503. E-mail address: [email protected] (A. Halldin). 8756-3282/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.bone.2011.07.003
osteotomy preparation, the implant design, and on the implant surface topography. While it is well known that dynamic strains are the driving force behind bone modeling and remodeling [7,8], our knowledge about the effect of static strains is limited. In a study performed on rats, dynamic loads applied 18 s per day gave rise to new bone formation while static loads of the same magnitude and duration did not modulate bone modeling and remodeling [9]. In a study using dogs, osseointegrated implants with a static lateral load, superimposed on functional dynamic loads, resulted in a densification of the bone adjacent to the implant and in an increase in bone-to-implant contact [10]. Directly after implant installation, the implant stability depends solely on mechanical interlock for stabilization. According to SzmuklerMoncler et al. [11] micro-motions should be less than 50–150 μm in order to avoid fibrous tissue encapsulation and implant instability. In the classic implant installation protocol, dental implants were loaded after 3 months in the mandible and after 6 months in the maxilla [12]. It has later been suggested that early and even immediate loading of dental implants can be practiced with predictable results, if the implants exhibit a certain minimal primary stability [13]. Clinically, the stability of an implant is often evaluated by measurement of the insertion torque [14]. The minimum insertion
784
A. Halldin et al. / Bone 49 (2011) 783–789
torque for application of immediate loading has in different studies varied between 20 and 48 Ncm [15–23]. Concerns have been raised that excessively high insertion torque may give rise to ischemia and pressure necrosis in the bone [24]. However, ischemia in the rat tibia for 4.5 h was found to cause periosteal proliferation instead of bone death [25]. It was suggested that transient ischemia is an important factor in fracture healing. Kalebo et al. [26] reported bone resorption after 6 h of ischemia in the tibia of rabbits. In a study using turkeys, it was found that a static bending load on the ulna, maintained during 8 weeks, which created an estimated maximum strain of ~0.002, resulted in similar bone loss as functional deprivation alone [27]. Perren et al. [28] inserted compression plates in the tibia in sheep and observed that a pressure at the screw sites of about 40 MPa did not result in pressure necrosis and a steep fall in pressure, but rather a gradual decrease in pressure, which was explained by haversian remodeling. Assuming a modulus of elasticity of 12 GPa for cortical bone, it can be suggested that a static strain of 0.0033 did not provoke necrosis. Based on results from animal studies of internal fixation of fractures, Perren et al. [28] later concluded that cortical bone tolerates 60 MPa without necrosis. Translated to strain, this indicates that a static strain of 0.005 does not provoke necrosis. Carter et al. [29] reported a yield strain of human cortical bone of 0.0068. The strain levels reported above are lower than the yield strain and consequently within the linear elastic portion of the stress–strain curve of cortical bone. Beyond the yield strain damage occurs in the form of microcracks in the bone [30]. The ultimate strain of cortical bone depends on several parameters such as age, quality and loading rate. McElhaney reported that ultimate strains of human cortical bone were between 0.01 and 0.02 with lower values for high strain rates [31]. Currey reported an ultimate strain of 0.022 for human adult femur [32]. In the age span of 20–80 years, McCalden et al. [33] reported ultimate strains, for human femora, ranging from 0.01 to 0.04 with the lower value for higher ages. The aim of the present study was to investigate the effect of predetermined static strains between the yield strain (~0.007) and ultimate strain (~0.02) and beyond the ultimate strain (~0.02) on implant stability and bone remodeling in a rabbit model after 24 days. The static strains were induced during implant insertion, by an increase in implant diameter. In addition, simulation of the mechanical events at the bone–implant interface by means of finite element analysis (FEA) was conducted.
Materials and methods Implants and surface topographical characterizations Screw shaped implants, manufactured from commercially pure titanium (grade 4), were used. The test implants comprised of three different portions; one cutting portion, one transition portion with a gradual increase in diameter and one condensation portion (Fig. 1). Two types of test implants with tight tolerances were used; one where the increase in diameter from the cutting region to the condensation region was 0.15 mm (group A) and another with a diameter increase of 0.05 mm (group B). The control implants (group C) had no diameter increase. In order to reduce a possible grinding effect, that may occur in the bone during implant insertion, the implants in this study had a turned surface. Topographical analysis of the implant surface was performed using optical interferometry (MicroXAMt, PhaseShift, Tucson, AZ, USA). From each group, three implants were randomly selected and measured at 5 areas each (5 areas on the thread flanks). Surface roughness parameters were calculated after errors of form and waviness had been removed with a 50 × 50 μm-sized Gaussian filter and the following 3D parameters were selected: Sa (μm)= the arithmetic average height deviation from a mean plane; Sds (mm − 2) = the density of summits; and Sdr (%) = the
Fig. 1. Schematic figure of the implant with different regions. The condensation region has an increase in diameter compared to the 3.5 mm diametrical cutting region. The diametrical increase creates static strain in the bone. In the present study group A implants had a diameter increase of 0.15 mm and group B implants had a diameter increase of 0.05 mm. The length of the transition region was 1 mm and of the condensation region 1.5 mm.
developed surface ratio. (For further specifications, see Wennerberg [34]). Surgical procedures and insertion torque Fourteen New Zealand white mature rabbits (female, approximately 10 months old) were used in this study. The study was approved by the local animal ethical committee at Gothenburg University. Animals were anesthetized with intramuscular injections of fentanyl and fluanison (Hypnorm Vet, Janssen Farmaucetica, Belgium) at a dose of 0.5 ml per kg of body weight and intraperitoneal injections of diazepam (Stesolid, Dumex, Denmark) at a dose of 0.25 mg per animal. Before surgery, the shaved skin of the rabbit was carefully washed with a mixture of 70% chlorhexidine and 70% ethanol. Local anesthesia with 1.0 ml of 5% lidocaine (Xylocain, AstraZeneca, Sweden) was injected subcutaneously in the surgical site. A single dose of prophylactic antibiotic (Borgal, Intervet, Boxmeer, Netherlands) was administered; 0.5 ml per kg body weight. All 14 rabbits received 0.5 ml of an analgesic (Temgesic, Reckitt and Coleman, England) at a concentration of 0.3 mg/ ml on the day of operation and 3 days thereafter. Each animal received three implants in each hind leg; two in the proximal tibia metaphysis, and one in the femoral condyle. Test implants were always inserted in the left leg and control implants in the right leg. Group A implants were installed in the femoral condyle and proximally in the proximal tibia metaphysis. Group B implants were installed distally in the proximal tibia metaphysis (Table 1). The bone bed was gently drilled with a final burr diameter of 3.3 mm corresponding to the core diameter of the cutting portion of the implant. When the implant was inserted, the cutting features created a threaded cavity in the bone, which was congruent with the implant shape at the cutting portion. When the transition portion entered the bone, it created Table 1 Illustration of implant placements and the number of implants for each group. Group A represents an increase in implant diameter of 0.15 mm. Group B represents an increase in implant diameter of 0.05 mm. Group C represents no increase in implant diameter. Place
Left
Right
Control implant
N
Test implant
N
Femur Tibia prox Tibia dist
C C C
8 11 11
A A B
8 11 11
A. Halldin et al. / Bone 49 (2011) 783–789
a gradual increase in the strains in the surrounding bone without cutting. When the condensation portion entered the bone, a predetermined bone condensation was obtained and the installation stopped when the condensation section had penetrated the bone. The implants were inserted with W&H implant unit (Elcomed, W&H SA-310, Burmoos Austria) with the rotation speed set at 20 revolutions/min. During implant placement, the drilling unit measured the insertion torque (ITQ) during the complete installation procedure. Thereafter all data were exported to Excel. In order to illustrate the mean ITQ over time the data were adjusted to the same starting point defined as when the torque of 50 Nmm was reached and did not fall below this value. One rabbit did not survive surgery due to anesthetic complications. Two rabbits were euthanized before full time due to healing complications.
Removal torque analysis After 24 days, all animals were euthanized with an overdose of pentobarbitalnatrium (60 mg/ml, Apoteksbolaget AB, Stockholm, Sweden). The skin above the implants was incised, and all implants were subjected to removal torque (RTQ) tests. The peak RTQ was recorded with a computerized control RTQ device, in which the values were transmitted at a frequency of 100 Hz to the computer via a control box [35]. The square shaped implant head was connected to this device, and an increasing reverse torque was applied until failure of the bone– implant interface occurred. The RTQ unit was preset to stop after 10 s and the remaining part of the implant remained in site. The first peak value of resistance to reverse torque rotation was assumed to represent the RTQ value. The RTQ data were adjusted to the same starting point defined as when the RTQ of 50 Nmm was reached.
785
Preparation of undecalcified cut and ground sections After euthanization, the samples were retrieved with the removal torque tested implants remaining in situ and were immersed in 4% neutral buffered formaldehyde. A randomized selection of samples with equal amount of test and control specimens (femur n = 8, tibia proximal n = 7, tibia distal n = 6) were subjected to undecalcified cut and ground sectioning [36,37] with the Exakt equipment (Exakt Apparatebau, Norderstedt, Germany). In brief, the samples were processed by dehydration in ethanol from 70% up to 100% followed by infiltration in diluted resins and finally embedded in pure resin (Technovit 7200 VLC, Kulzer, Germany) and cured using UV light. The cured blocks were cut along the implant axis and a 200 μm thick section (from the mid part of each implant) was taken. The sections were ground to 15 μm with SiC papers of various grain sizes. The surface was not polished [37]. The sections were stained with the routine staining i.e. toluidine blue mixed in pyronin G allowing identification of old and new bone using a light microscope. Finally the section was cover-slipped with ordinary mounting media.
Histological observations The prepared sections were subjected to histomorphometric analyses. The bone area was quantified in a region of interest (ROI) using a light microscope (Eclipse ME600, Nikon, Japan) and software (Image Analysis 2000, Tekno Optik AB, Sweden). The ROI was defined as the area, which theoretically presented the highest bone strains according to the theoretical FEA calculations. An illustration in Fig. 2 demonstrates the strain levels within the ROI comprising 6 threads (1.32 × 0.88 mm). Bone area percentage was calculated inside the ROI and both the area of new and old bone was measured.
Fig. 2. Results from the 2D axisymmetric FEA simulation of cortical bone with the radial displacement of the implant surface, for group A 0.075 mm (corresponds to 0.15 mm diametrical increase) and group B 0.025 mm (corresponds to 0.05 mm diametrical increase). In the threaded portion the maximum principal strain for group A implant reaches 0.045 and for group B 0.016. The rectangle illustrates the strain level within the ROI (~1.32 × 0.88 mm).
786
A. Halldin et al. / Bone 49 (2011) 783–789
Simulated strain levels The bone was simulated as a linearly elastic isotropic material with a Young's modulus of 12 GPa and a Poisson's ratio of 0.3 [38]. The simulated maximum principal strain in the surrounding bone for group A implants was 0.045 (Fig. 2), which is beyond the ultimate strain of human cortical bone of approximately 0.02 [31]. For group B implants the principal maximum strain obtained was 0.015 (Fig. 2) which is between the yield strain and ultimate strain of human cortical bone. Results Interferometry measurements The mean surface roughness (SD) measured by interferometry was as follows; Sa = 0.23 μm (SD 0.04), Sds = 128,740 mm − 2 (SD 11,976), and Sdr = 4.0% (SD 0.7). Insertion torque measurements Fig. 3. Mean insertion torque over time for different implant groups and locations.
Statistical analysis Data, assumed to be normally distributed using standard statistical methods, were analyzed with t-test.
Theoretical calculations Finite element analysis Prior to the animal experiment a 2-dimensional axisymmetric finite element model of the condensation portion of the implant and the surrounding bone was developed to calculate the theoretical maximum principal strain. The implant and the bone were modeled in a CAD software (Pro/Engineer, PTC Corporate Needham, MA, USA) and then transferred into the finite element software ANSYS 12.01 (ANSYS, Inc. Canonsburg, PA, USA). The strain in the bone was induced by radial displacement of the implant surface by 0.075 mm and 0.025 mm simulating a diameter increase of 0.15 mm representing group A and 0.05 mm representing group B respectively.
For all implant groups the mean ITQ as a function of time is shown in Fig. 3. There are 5 different sections identified in the curves. 1) In the first section there is an increase in torque as the cutting portion of the implant cuts a thread in the coronal cortical bone. 2) In the second section of the curve the torque is reduced due to the fact that there is already a cut thread pattern in the bone that matches the implant thread. 3) When the transition portion enters the bone there is a gradual increase in ITQ. 4) When the condensation portion of the implant enters the bone the insertion torque levels out. 5) An additional increase in ITQ may occur if the cutting portion of the implant cuts a thread in the contralateral cortical bone (Fig. 3). Theoretically the condensation part should penetrate the bone beyond 10 s, which is in agreement with the ITQ measurement. Therefore the maximum torque beyond 10 s was assumed to represent the torque at the condensed section. The results and statistical analysis of ITQ measurement for groups A and B are presented in a box plot (Fig. 4) and Table 2. In all sites, the mean ITQ value for groups A and B was significantly higher than that of the corresponding group C (p b 0.001) (Table 2).
Fig. 4. Box plot of the mean insertion torque over time, asterisk representing statistical significance between test and control.
A. Halldin et al. / Bone 49 (2011) 783–789 Table 2 Mean insertion and removal torque for all groups. An asterisk represents significant difference between test and control. Place
ITQ
RTQ
Femur Tibia prox Tibia dist Femur Tibia prox Tibia dist
Test
Control
Implant
Mean torque Nmm (SD)
Implant
Mean torque Nmm (SD)
A A B A A B
398 442 343 251 260 230
C C C C C C
106 100 149 209 168 172
(87) (51) (82) (61) (69) (53)
(42) (38) (52) (78) (78) (53)
p-Value t-test
787
Table 3 Mean difference between ITQ and RTQ for Group A compared to Group B.
Tibia
ITQ–RTQ (A) Nmm (SD)
ITQ–RTQ (B) Nmm (SD)
p-Value t-test
183 (82)
134 (70)
0.226
b0.0001⁎ b0.0001⁎ b0.0001⁎ 0.248 0.009⁎ 0.010⁎
Table 4 Percent of old and new bone areas within the ROI for all groups. Asterisk represents significant difference between test and control. Place
Removal torque tests
Old
The results and statistical analysis of the RTQ measurement for groups A and B are presented in a box plot (Fig. 5) and Table 2. The mean RTQ value for groups A and B placed in tibia was significantly higher than that of the corresponding group C (p = 0.009 and 0.01 respectively). For group A implants in the femoral condyle the difference in mean RTQ compared to the corresponding group C implants was not statistically significant (p = 0.248) (Table 2). Difference between installation torque and removal torque In tibia there was no significant difference between the difference between ITQ and RTQ for group A compared to group B (Table 3). Histological observations No statistical difference of bone area at the left side and the right side of the implant was detected neither for old nor for new bone (p = 0.724 and p = 0.534 respectively). For group A, placed in tibia, the area of old and new bone was significantly different compared to the corresponding group C (p = 0.005). The remaining sites showed no significant difference between old and new bone (Table 4). Discussion In the present study the effect of different static strains, at the time of insertion, on removal torque after 24 days of healing was evaluated. While several studies have investigated how dynamic load affects
New
Femur Tibia prox Tibia dist Femur Tibia prox Tibia dist
Test
Control
Implant
% bone in ROI
Implant
% bone in ROI
A A B A A B
47.6 56.0 76.1 22.3 19.6 12.5
C C C C C C
54.7 43.7 76.0 24.0 26.1 10.4
(9.4) (7.8) (4.4) (7.4) (6.4) (2.7)
(13.4) (7.1) (9.0) (3.5) (4.6) (3.5)
p-value t-test 0.128 0.005⁎ 0.968 0.568 0.005⁎ 0.068
bone remodeling [5,7,8,39], only a limited number of studies have analyzed how static strain affects bone remodeling. Currey found that the yield strain and ultimate strain of cortical bone from different species of mammals did not exhibit large variations [32]. Hence, it is assumed that the yield strain and ultimate strain values for cortical bone of rabbit do not deviate substantially from those of human cortical bone. In order to minimize the surface roughness of the implant that could possibly affect the outcome, turned implants were used [40]. It is likely that a turned surface has a slight grinding effect on the bone and the calculated strains might represent a slight overestimation. The maximum insertion torque beyond 10 s represents the maximum condensation applied to the surrounding bone. The maximum insertion torque for both test implants was significantly higher compared to those of the control implants, confirming that the condensation feature on the test implants created strains in the bone. Higher strains means higher interface stresses and a higher frictional force which increases the insertion torque for the test implants. After the implant installation procedure the bone surrounding the test implants was thus in a state of prestress. If the cutting portion of the implant cuts a thread in the contralateral cortical bone, a further increase in the ITQ occurs. As shown in Fig. 3 the tibial distal control
Fig. 5. Box plot of removal torque, asterisk representing significant difference between test and control.
788
A. Halldin et al. / Bone 49 (2011) 783–789
implants have a slightly increased ITQ at the end, a phenomenon which may be detectable for the control implants but not for the test implants. The analysis was consequently performed with the maximum ITQ beyond 10 s, neglecting this phenomenon. When an implant is inserted the bone is affected by the surgery. According to Eriksson et al. [41] a 0.5 mm border zone becomes necrotic due to the surgical trauma. Slaets et al. [42] inserted implants press-fit in the tibial diaphysis of rabbit. In the implant border zone, osteocytic lacunae devoid of osteocytes were found. This region had a maximum extension of 400 μm 28 days after implant surgery. The damaged bone was invaded by basic multicellular units the activities of which peaked after 2–4 weeks but were still ongoing 6 weeks after implant insertion. However it has been suggested that necrotic bone can serve as structural support during remodeling [43]. In the present study, the tibial implants were supported by cortical bone. The group A and group B implants generated strains, which theoretically exceeded the ultimate strain and the yield strain of cortical bone respectively. Including the dimension tolerances for the implant the strain levels for group B are between yield and ultimate strain and the strain levels for group A are beyond ultimate strain of cortical bone. The removal torque of groups A and B indicated however that, there was no extensive bone resorption at 24 days after surgery. On the contrary, the removal torque of the experimental implants was higher than that of the control implants, which were designed not to produce static strains in the bone. It is of great interest that the highest removal torque value was obtained for group A implants, for which the strains induced by far exceeded the ultimate strains of cortical bone. It is suggested that the increased removal torques in the tibia for group A and B implants compared to group C are due to remaining prestress. It is further suggested that there is a slow gradual decline in the bone stresses produced at implant insertion and that this decline consists of two components, where one component is a visco-elastic stress relaxation of the bone and the other is the normal remodeling by basic multicellular units whereby prestressed bone is replaced by new haversian systems, which are not prestressed [44]. Even though no significance was obtained the difference between ITQ and RTQ seems to be greater for group A compared to group B. This may indicate that the increased pressfit decreases more rapidly with higher static strain due to visco-elastic behavior and/or increased bone remodeling rate. According to Roberts et al. [45], the cortical bone of rabbit remodels three times faster than human cortical bone. Thus 24 days in rabbit bone would correspond to 72 days in human and providing prestressed state over the initial healing phase. In a study in sheep, Cordey et al. found that the remaining traction forces, exerted by cortical bone screws, after 1 and 5 weeks amounted to 72% and 62% of the original traction force respectively [46]. They found that the decrease in traction force as a function of time followed a logarithmic curve. In a study observing fracture healing in sheep using internal fixation, small areas of bone which had undergone plastic deformation were not removed by surface resorption but by internal remodeling [47]. In the current study, it is also noticeable that the differences in stability between test and control implants decreased at 24 days, both in absolute and relative terms. One interpretation of this may be that the retention of the control implants increases with time while the retention of the test implants decreases with time and that their retention graphs after some time will merge. The histomorphometric evaluation showed that for the femoral and distal tibial sites, no significant differences were detected in area of old bone or new bone between test and control implants. In the proximal tibia however group A implants had significantly larger old bone area than the control implants. Due to the limitations of the current study, it is difficult to interpret this phenomenon. However it may suggest that the increase in condensation did not adversely affect the amount of old bone, and therefore the press fit on the implant remained at 24 days.
Conclusions This study indicates that increased static bone strains do not affect an extensive bone resorption but provides higher installation torque, and a higher removal torque after 24 days in rabbit bone. Acknowledgments This work was supported by the Swedish Research Council (6212010-4760). We would like to acknowledge the assistance of Ms Petra Johansson and Ms Maria Hoffman for the animal surgery and sample preparation and Mr Stefan Johnsson at Astra Tech for preparing all implants. References [1] Wolff J. Das Gesetz der Transformation der Knochen. Berlin: Verlag von August Hirschwald; 1892. [2] Huiskes R, Ruimerman R, van Lenthe GH, Janssen JD. Effects of mechanical forces on maintenance and adaptation of form in trabecular bone. Nature 2000;405: 704–6. [3] Qin YX, Rubin CT, McLeod KJ. Nonlinear dependence of loading intensity and cycle number in the maintenance of bone mass and morphology. J Orthop Res 1998;16: 482–9. [4] Mosley JR, Lanyon LE. Strain rate as a controlling influence on adaptive modeling in response to dynamic loading of the ulna in growing male rats. Bone 1998;23: 313–8. [5] Rubin CT, Lanyon LE. Regulation of bone formation by applied dynamic loads. J Bone Joint Surg Am 1984;66:397–402. [6] Lanyon LE, Hampson WG, Goodship AE, Shah JS. Bone deformation recorded in vivo from strain gauges attached to the human tibial shaft. Acta Orthop Scand 1975;46:256–68. [7] Stanford CM, Brand RA. Toward an understanding of implant occlusion and strain adaptive bone modeling and remodeling. J Prosthet Dent 1999;81:553–61. [8] Rubin CT, Sommerfeldt DW, Judex S, Qin Y. Inhibition of osteopenia by low magnitude, high-frequency mechanical stimuli. Drug Discov Today 2001;6: 848–58. [9] Turner CH, Owan I, Takano Y. Mechanotransduction in bone: role of strain rate. Am J Physiol 1995;269:E438–42. [10] Gotfredsen K, Berglundh T, Lindhe J. Bone reactions adjacent to titanium implants subjected to static load. A study in the dog (I). Clin Oral Implants Res 2001;12:1–8. [11] Szmukler-Moncler S, Salama H, Reingewirtz Y, Dubruille JH. Timing of loading and effect of micromotion on bone–dental implant interface: review of experimental literature. J Biomed Mater Res 1998;43:192–203. [12] Adell R, Lekholm U, Rockler B, Branemark PI. A 15-year study of osseointegrated implants in the treatment of the edentulous jaw. Int J Oral Surg 1981;10:387–416. [13] Attard NJ, Zarb GA. Immediate and early implant loading protocols: a literature review of clinical studies. J Prosthet Dent 2005;94:242–58. [14] Schincaglia GP, Marzola R, Scapoli C, Scotti R. Immediate loading of dental implants supporting fixed partial dentures in the posterior mandible: a randomized controlled split-mouth study—machined versus titanium oxide implant surface. Int J Oral Maxillofac Implants 2007;22:35–46. [15] Donati M, La Scala V, Billi M, Di Dino B, Torrisi P, Berglundh T. Immediate functional loading of implants in single tooth replacement: a prospective clinical multicenter study. Clin Oral Implants Res 2008;19:740–8. [16] Schincaglia GP, Marzola R, Giovanni GF, Chiara CS, Scotti R. Replacement of mandibular molars with single-unit restorations supported by wide-body implants: immediate versus delayed loading. A randomized controlled study. Int J Oral Maxillofac Implants 2008;23:474–80. [17] Testori T, Galli F, Capelli M, Zuffetti F, Esposito M. Immediate nonocclusal versus early loading of dental implants in partially edentulous patients: 1-year results from a multicenter, randomized controlled clinical trial. Int J Oral Maxillofac Implants 2007;22:815–22. [18] Lindeboom JA, Frenken JW, Dubois L, Frank M, Abbink I, Kroon FH. Immediate loading versus immediate provisionalization of maxillary single-tooth replacements: a prospective randomized study with BioComp implants. J Oral Maxillofac Surg 2006;64:936–42. [19] Merli M, Merli A, Bernardelli F, Lombardini F, Esposito M. Immediate versus early non-occlusal loading of dental implants placed flapless in partially edentulous patients. One-year results from a randomised controlled trial. Eur J Oral Implantol 2008;1:207–20. [20] Cannizzaro G, Leone M. Restoration of partially edentulous patients using dental implants with a microtextured surface: a prospective comparison of delayed and immediate full occlusal loading. Int J Oral Maxillofac Implants 2003;18:512–22. [21] Cannizzaro G, Leone M, Esposito M. Immediate versus early loading of two implants placed with a flapless technique supporting mandibular bar-retained overdentures: a single-blinded, randomised controlled clinical trial. Eur J Oral Implantol 2008;1:33–43. [22] Ostman PO, Hellman M, Sennerby L. Direct implant loading in the edentulous maxilla using a bone density-adapted surgical protocol and primary implant stability criteria for inclusion. Clin Implant Dent Relat Res 2005;7(Suppl 1):S60–9.
A. Halldin et al. / Bone 49 (2011) 783–789 [23] Misch CM. Immediate loading of definitive implants in the edentulous mandible using a fixed provisional prosthesis: the denture conversion technique. J Oral Maxillofac Surg 2004;62:106–15. [24] Bashutski JD, D'Silva NJ, Wang HL. Implant compression necrosis: current understanding and case report. J Periodontol 2009;80:700–4. [25] Svindland AD, Nordsletten L, Reikeras O, Skjeldal S. Periosteal response to transient ischemia. Histological studies on the rat tibia. Acta Orthop Scand 1995;66:468–72. [26] Kalebo P, Johansson C, Albrektsson T. Temporary bone tissue ischemia in the hind limb of the rabbit. A vital microscopic study. Arch Orthop Trauma Surg 1986;105: 321–5. [27] Lanyon LE, Rubin CT. Static vs dynamic loads as an influence on bone remodelling. J Biomech 1984;17:897–905. [28] Perren SM, Huggler A, Russenberger M, et al. The reaction of cortical bone to compression. Acta Orthop Scand Suppl 1969;125:19–29. [29] Carter DR, Caler WE, Spengler DM, Frankel VH. Fatigue behavior of adult cortical bone: the influence of mean strain and strain range. Acta Orthop Scand 1981;52: 481–90. [30] Winwood K, Zioupos P, Currey JD, Cotton JR, Taylor M. The importance of the elastic and plastic components of strain in tensile and compressive fatigue of human cortical bone in relation to orthopaedic biomechanics. J Musculoskelet Neuronal Interact 2006;6:134–41. [31] McElhaney JH. Dynamic response of bone and muscle tissue. J Appl Physiol 1966;21:1231–6. [32] Currey JD. Tensile yield in compact bone is determined by strain, post-yield behaviour by mineral content. J Biomech 2004;37:549–56. [33] McCalden RW, McGeough JA, Barker MB, Court-Brown CM. Age-related changes in the tensile properties of cortical bone. The relative importance of changes in porosity, mineralization, and microstructure. J Bone Joint Surg Am 1993;75:1193–205. [34] Wennerberg A, Albrektsson T. Suggested guidelines for the topographic evaluation of implant surfaces. Int J Oral Maxillofac Implants 2000;15:331–44.
789
[35] Johansson CB, Jimbo R, Stefenson P. Ex vivo and in vivo biomechanical test of implant attachment to various materials: introduction of a new user-friendly removal torque equipment. Clin Implant Dent Relat Res 2010, doi:10.1111/j.17088208.2010.00296.x. [36] Donath K, Breuner G. A method for the study of undecalcified bones and teeth with attached soft tissues. The sage-schliff (sawing and grinding) technique. J Oral Pathol 1982;11:318–26. [37] Johansson CB, Morberg P. Importance of ground section thickness for reliable histomorphometrical results. Biomaterials 1995;16:91–5. [38] Reilly DT, Burstein AH. Review article. The mechanical properties of cortical bone. J Bone Joint Surg Am 1974;56:1001–22. [39] Goodship AE, Lanyon LE, McFie H. Functional adaptation of bone to increased stress. An experimental study. J Bone Joint Surg Am 1979;61:539–46. [40] Albrektsson T, Wennerberg A. Oral implant surfaces: part 1—review focusing on topographic and chemical properties of different surfaces and in vivo responses to them. Int J Prosthodont 2004;17:536–43. [41] Eriksson RA, Albrektsson T, Magnusson B. Assessment of bone viability after heat trauma. A histological, histochemical and vital microscopic study in the rabbit. Scand J Plast Reconstr Surg 1984;18:261–8. [42] Slaets E, Carmeliet G, Naert I, Duyck J. Early cellular responses in cortical bone healing around unloaded titanium implants: an animal study. J Periodontol 2006;77:1015–24. [43] McKibbin B. The biology of fracture healing in long bones. J Bone Joint Surg Br 1978;60-B:150–62. [44] Sedlin ED. A rheologic model for cortical bone. A study of the physical properties of human femoral samples. Acta Orthop Scand Suppl 1965;Suppl 83:1–77. [45] Roberts WE, Turley PK, Brezniak N, Fielder PJ. Implants: bone physiology and metabolism. CDA J 1987;15:54–61. [46] Cordey J, Blumlein H, Ziegler W, Perren SM. Study of the behavior in the course of time of the holding power of cortical screws in vivo. Acta Orthop Belg 1976;42 (Suppl 1):75–87. [47] Perren SM, Rahn BA. Biomechanics of fracture healing. Can J Surg 1980;23:228–32.
Contents lists available at ScienceDirect
Bone j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / b o n e
The effect of static bone strain on implant stability and bone remodeling Anders Halldin a, b,⁎, Ryo Jimbo a, d, Carina B. Johansson c, Ann Wennerberg a, d, Magnus Jacobsson a, b, Tomas Albrektsson e, f, Stig Hansson b a
Department of Prosthodontics, Faculty of Odontology, Malmö University, Malmö, Sweden Astra Tech AB, Mölndal, Sweden Dental Materials, Department of Prosthodontics, Institute of Odontology, The Sahlgrenska Academy, University of Gothenburg, Sweden d Department of Biomaterials/Handicap Research, Gothenburg University, Gothenburg, Sweden e Department of Biomaterials, Gothenburg University, Sweden f Department of Dental Materials Sciences & Technology, Malmö University, Sweden b c
a r t i c l e
i n f o
Article history: Received 21 April 2011 Revised 15 June 2011 Accepted 4 July 2011 Available online 14 July 2011 Edited by: David Burr Keywords: Static bone strain Removal torque In vivo experiment Biomechanical FE simulation Bone remodeling
a b s t r a c t Bone remodeling is a process involving both dynamic and static bone strain. Although there exist numerous studies on the effect of dynamic strain on implant stability and bone remodeling, the effect of static strain has yet to be clarified. Hence, for this purpose, the effect of static bone strain on implant stability and bone remodeling was investigated in rabbits. Based on Finite Element (FE) simulation two different test implants, with a diametrical increase of 0.15 mm (group A) and 0.05 mm (group B) creating static strains in the bone of 0.045 and 0.015 respectively, were inserted in the femur (group A) and the proximal tibia metaphysis (groups A and B respectively) of 14 rabbits to observe the biological response. Both groups were compared to control implants, with no diametrical increase (group C), which were placed in the opposite leg. At the time of surgery, the insertion torque (ITQ) was measured to represent the initial stability. The rabbits were euthanized after 24 days and the removal torque (RTQ) was measured to analyze the effect on implant stability and bone remodeling. The mean ITQ value was significantly higher for both groups A and B compared to group C regardless of the bone type. The RTQ value was significantly higher in tibia for groups A and B compared to group C while group A placed in femur presented no significant difference compared to group C. The results suggest that increased static strain in the bone not only creates higher implant stability at the time of insertion, but also generates increased implant stability throughout the observation period. © 2011 Elsevier Inc. All rights reserved.
Introduction Bone is a unique material in its ability to adapt its mass and structure in response to the loads to which it is exposed [1,2]. The loads induce strains in the bone, and the modeling and remodeling stimulus has been found to be dependent on strain magnitude, strain frequency [3], and strain rate [4]. The bone mass can either be maintained by a relatively small number of loading cycles with high strains, for example 4 loading cycles per day with a strain of 2000 microstrain [5], or by a great number of loading cycles with low strains, for example 10,000 loading cycles per day with a strain of 400 microstrain [6]. It has also been found that the osteogenic effect of strains increases with increasing strain rate [4]. The insertion of an implant with a certain torque means that static strains will be induced in the surrounding bone. The magnitude of these static strains depends on the bone anatomy, the bone quality, the
Abbreviations: RTQ, Removal torque; ITQ, Insertion torque; ROI, Region of interest. ⁎ Corresponding author at: Department of Prosthodontics, Faculty of Odontology, Malmö University, Malmö, Sweden. Fax: +46 40 6658503. E-mail address: [email protected] (A. Halldin). 8756-3282/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.bone.2011.07.003
osteotomy preparation, the implant design, and on the implant surface topography. While it is well known that dynamic strains are the driving force behind bone modeling and remodeling [7,8], our knowledge about the effect of static strains is limited. In a study performed on rats, dynamic loads applied 18 s per day gave rise to new bone formation while static loads of the same magnitude and duration did not modulate bone modeling and remodeling [9]. In a study using dogs, osseointegrated implants with a static lateral load, superimposed on functional dynamic loads, resulted in a densification of the bone adjacent to the implant and in an increase in bone-to-implant contact [10]. Directly after implant installation, the implant stability depends solely on mechanical interlock for stabilization. According to SzmuklerMoncler et al. [11] micro-motions should be less than 50–150 μm in order to avoid fibrous tissue encapsulation and implant instability. In the classic implant installation protocol, dental implants were loaded after 3 months in the mandible and after 6 months in the maxilla [12]. It has later been suggested that early and even immediate loading of dental implants can be practiced with predictable results, if the implants exhibit a certain minimal primary stability [13]. Clinically, the stability of an implant is often evaluated by measurement of the insertion torque [14]. The minimum insertion
784
A. Halldin et al. / Bone 49 (2011) 783–789
torque for application of immediate loading has in different studies varied between 20 and 48 Ncm [15–23]. Concerns have been raised that excessively high insertion torque may give rise to ischemia and pressure necrosis in the bone [24]. However, ischemia in the rat tibia for 4.5 h was found to cause periosteal proliferation instead of bone death [25]. It was suggested that transient ischemia is an important factor in fracture healing. Kalebo et al. [26] reported bone resorption after 6 h of ischemia in the tibia of rabbits. In a study using turkeys, it was found that a static bending load on the ulna, maintained during 8 weeks, which created an estimated maximum strain of ~0.002, resulted in similar bone loss as functional deprivation alone [27]. Perren et al. [28] inserted compression plates in the tibia in sheep and observed that a pressure at the screw sites of about 40 MPa did not result in pressure necrosis and a steep fall in pressure, but rather a gradual decrease in pressure, which was explained by haversian remodeling. Assuming a modulus of elasticity of 12 GPa for cortical bone, it can be suggested that a static strain of 0.0033 did not provoke necrosis. Based on results from animal studies of internal fixation of fractures, Perren et al. [28] later concluded that cortical bone tolerates 60 MPa without necrosis. Translated to strain, this indicates that a static strain of 0.005 does not provoke necrosis. Carter et al. [29] reported a yield strain of human cortical bone of 0.0068. The strain levels reported above are lower than the yield strain and consequently within the linear elastic portion of the stress–strain curve of cortical bone. Beyond the yield strain damage occurs in the form of microcracks in the bone [30]. The ultimate strain of cortical bone depends on several parameters such as age, quality and loading rate. McElhaney reported that ultimate strains of human cortical bone were between 0.01 and 0.02 with lower values for high strain rates [31]. Currey reported an ultimate strain of 0.022 for human adult femur [32]. In the age span of 20–80 years, McCalden et al. [33] reported ultimate strains, for human femora, ranging from 0.01 to 0.04 with the lower value for higher ages. The aim of the present study was to investigate the effect of predetermined static strains between the yield strain (~0.007) and ultimate strain (~0.02) and beyond the ultimate strain (~0.02) on implant stability and bone remodeling in a rabbit model after 24 days. The static strains were induced during implant insertion, by an increase in implant diameter. In addition, simulation of the mechanical events at the bone–implant interface by means of finite element analysis (FEA) was conducted.
Materials and methods Implants and surface topographical characterizations Screw shaped implants, manufactured from commercially pure titanium (grade 4), were used. The test implants comprised of three different portions; one cutting portion, one transition portion with a gradual increase in diameter and one condensation portion (Fig. 1). Two types of test implants with tight tolerances were used; one where the increase in diameter from the cutting region to the condensation region was 0.15 mm (group A) and another with a diameter increase of 0.05 mm (group B). The control implants (group C) had no diameter increase. In order to reduce a possible grinding effect, that may occur in the bone during implant insertion, the implants in this study had a turned surface. Topographical analysis of the implant surface was performed using optical interferometry (MicroXAMt, PhaseShift, Tucson, AZ, USA). From each group, three implants were randomly selected and measured at 5 areas each (5 areas on the thread flanks). Surface roughness parameters were calculated after errors of form and waviness had been removed with a 50 × 50 μm-sized Gaussian filter and the following 3D parameters were selected: Sa (μm)= the arithmetic average height deviation from a mean plane; Sds (mm − 2) = the density of summits; and Sdr (%) = the
Fig. 1. Schematic figure of the implant with different regions. The condensation region has an increase in diameter compared to the 3.5 mm diametrical cutting region. The diametrical increase creates static strain in the bone. In the present study group A implants had a diameter increase of 0.15 mm and group B implants had a diameter increase of 0.05 mm. The length of the transition region was 1 mm and of the condensation region 1.5 mm.
developed surface ratio. (For further specifications, see Wennerberg [34]). Surgical procedures and insertion torque Fourteen New Zealand white mature rabbits (female, approximately 10 months old) were used in this study. The study was approved by the local animal ethical committee at Gothenburg University. Animals were anesthetized with intramuscular injections of fentanyl and fluanison (Hypnorm Vet, Janssen Farmaucetica, Belgium) at a dose of 0.5 ml per kg of body weight and intraperitoneal injections of diazepam (Stesolid, Dumex, Denmark) at a dose of 0.25 mg per animal. Before surgery, the shaved skin of the rabbit was carefully washed with a mixture of 70% chlorhexidine and 70% ethanol. Local anesthesia with 1.0 ml of 5% lidocaine (Xylocain, AstraZeneca, Sweden) was injected subcutaneously in the surgical site. A single dose of prophylactic antibiotic (Borgal, Intervet, Boxmeer, Netherlands) was administered; 0.5 ml per kg body weight. All 14 rabbits received 0.5 ml of an analgesic (Temgesic, Reckitt and Coleman, England) at a concentration of 0.3 mg/ ml on the day of operation and 3 days thereafter. Each animal received three implants in each hind leg; two in the proximal tibia metaphysis, and one in the femoral condyle. Test implants were always inserted in the left leg and control implants in the right leg. Group A implants were installed in the femoral condyle and proximally in the proximal tibia metaphysis. Group B implants were installed distally in the proximal tibia metaphysis (Table 1). The bone bed was gently drilled with a final burr diameter of 3.3 mm corresponding to the core diameter of the cutting portion of the implant. When the implant was inserted, the cutting features created a threaded cavity in the bone, which was congruent with the implant shape at the cutting portion. When the transition portion entered the bone, it created Table 1 Illustration of implant placements and the number of implants for each group. Group A represents an increase in implant diameter of 0.15 mm. Group B represents an increase in implant diameter of 0.05 mm. Group C represents no increase in implant diameter. Place
Left
Right
Control implant
N
Test implant
N
Femur Tibia prox Tibia dist
C C C
8 11 11
A A B
8 11 11
A. Halldin et al. / Bone 49 (2011) 783–789
a gradual increase in the strains in the surrounding bone without cutting. When the condensation portion entered the bone, a predetermined bone condensation was obtained and the installation stopped when the condensation section had penetrated the bone. The implants were inserted with W&H implant unit (Elcomed, W&H SA-310, Burmoos Austria) with the rotation speed set at 20 revolutions/min. During implant placement, the drilling unit measured the insertion torque (ITQ) during the complete installation procedure. Thereafter all data were exported to Excel. In order to illustrate the mean ITQ over time the data were adjusted to the same starting point defined as when the torque of 50 Nmm was reached and did not fall below this value. One rabbit did not survive surgery due to anesthetic complications. Two rabbits were euthanized before full time due to healing complications.
Removal torque analysis After 24 days, all animals were euthanized with an overdose of pentobarbitalnatrium (60 mg/ml, Apoteksbolaget AB, Stockholm, Sweden). The skin above the implants was incised, and all implants were subjected to removal torque (RTQ) tests. The peak RTQ was recorded with a computerized control RTQ device, in which the values were transmitted at a frequency of 100 Hz to the computer via a control box [35]. The square shaped implant head was connected to this device, and an increasing reverse torque was applied until failure of the bone– implant interface occurred. The RTQ unit was preset to stop after 10 s and the remaining part of the implant remained in site. The first peak value of resistance to reverse torque rotation was assumed to represent the RTQ value. The RTQ data were adjusted to the same starting point defined as when the RTQ of 50 Nmm was reached.
785
Preparation of undecalcified cut and ground sections After euthanization, the samples were retrieved with the removal torque tested implants remaining in situ and were immersed in 4% neutral buffered formaldehyde. A randomized selection of samples with equal amount of test and control specimens (femur n = 8, tibia proximal n = 7, tibia distal n = 6) were subjected to undecalcified cut and ground sectioning [36,37] with the Exakt equipment (Exakt Apparatebau, Norderstedt, Germany). In brief, the samples were processed by dehydration in ethanol from 70% up to 100% followed by infiltration in diluted resins and finally embedded in pure resin (Technovit 7200 VLC, Kulzer, Germany) and cured using UV light. The cured blocks were cut along the implant axis and a 200 μm thick section (from the mid part of each implant) was taken. The sections were ground to 15 μm with SiC papers of various grain sizes. The surface was not polished [37]. The sections were stained with the routine staining i.e. toluidine blue mixed in pyronin G allowing identification of old and new bone using a light microscope. Finally the section was cover-slipped with ordinary mounting media.
Histological observations The prepared sections were subjected to histomorphometric analyses. The bone area was quantified in a region of interest (ROI) using a light microscope (Eclipse ME600, Nikon, Japan) and software (Image Analysis 2000, Tekno Optik AB, Sweden). The ROI was defined as the area, which theoretically presented the highest bone strains according to the theoretical FEA calculations. An illustration in Fig. 2 demonstrates the strain levels within the ROI comprising 6 threads (1.32 × 0.88 mm). Bone area percentage was calculated inside the ROI and both the area of new and old bone was measured.
Fig. 2. Results from the 2D axisymmetric FEA simulation of cortical bone with the radial displacement of the implant surface, for group A 0.075 mm (corresponds to 0.15 mm diametrical increase) and group B 0.025 mm (corresponds to 0.05 mm diametrical increase). In the threaded portion the maximum principal strain for group A implant reaches 0.045 and for group B 0.016. The rectangle illustrates the strain level within the ROI (~1.32 × 0.88 mm).
786
A. Halldin et al. / Bone 49 (2011) 783–789
Simulated strain levels The bone was simulated as a linearly elastic isotropic material with a Young's modulus of 12 GPa and a Poisson's ratio of 0.3 [38]. The simulated maximum principal strain in the surrounding bone for group A implants was 0.045 (Fig. 2), which is beyond the ultimate strain of human cortical bone of approximately 0.02 [31]. For group B implants the principal maximum strain obtained was 0.015 (Fig. 2) which is between the yield strain and ultimate strain of human cortical bone. Results Interferometry measurements The mean surface roughness (SD) measured by interferometry was as follows; Sa = 0.23 μm (SD 0.04), Sds = 128,740 mm − 2 (SD 11,976), and Sdr = 4.0% (SD 0.7). Insertion torque measurements Fig. 3. Mean insertion torque over time for different implant groups and locations.
Statistical analysis Data, assumed to be normally distributed using standard statistical methods, were analyzed with t-test.
Theoretical calculations Finite element analysis Prior to the animal experiment a 2-dimensional axisymmetric finite element model of the condensation portion of the implant and the surrounding bone was developed to calculate the theoretical maximum principal strain. The implant and the bone were modeled in a CAD software (Pro/Engineer, PTC Corporate Needham, MA, USA) and then transferred into the finite element software ANSYS 12.01 (ANSYS, Inc. Canonsburg, PA, USA). The strain in the bone was induced by radial displacement of the implant surface by 0.075 mm and 0.025 mm simulating a diameter increase of 0.15 mm representing group A and 0.05 mm representing group B respectively.
For all implant groups the mean ITQ as a function of time is shown in Fig. 3. There are 5 different sections identified in the curves. 1) In the first section there is an increase in torque as the cutting portion of the implant cuts a thread in the coronal cortical bone. 2) In the second section of the curve the torque is reduced due to the fact that there is already a cut thread pattern in the bone that matches the implant thread. 3) When the transition portion enters the bone there is a gradual increase in ITQ. 4) When the condensation portion of the implant enters the bone the insertion torque levels out. 5) An additional increase in ITQ may occur if the cutting portion of the implant cuts a thread in the contralateral cortical bone (Fig. 3). Theoretically the condensation part should penetrate the bone beyond 10 s, which is in agreement with the ITQ measurement. Therefore the maximum torque beyond 10 s was assumed to represent the torque at the condensed section. The results and statistical analysis of ITQ measurement for groups A and B are presented in a box plot (Fig. 4) and Table 2. In all sites, the mean ITQ value for groups A and B was significantly higher than that of the corresponding group C (p b 0.001) (Table 2).
Fig. 4. Box plot of the mean insertion torque over time, asterisk representing statistical significance between test and control.
A. Halldin et al. / Bone 49 (2011) 783–789 Table 2 Mean insertion and removal torque for all groups. An asterisk represents significant difference between test and control. Place
ITQ
RTQ
Femur Tibia prox Tibia dist Femur Tibia prox Tibia dist
Test
Control
Implant
Mean torque Nmm (SD)
Implant
Mean torque Nmm (SD)
A A B A A B
398 442 343 251 260 230
C C C C C C
106 100 149 209 168 172
(87) (51) (82) (61) (69) (53)
(42) (38) (52) (78) (78) (53)
p-Value t-test
787
Table 3 Mean difference between ITQ and RTQ for Group A compared to Group B.
Tibia
ITQ–RTQ (A) Nmm (SD)
ITQ–RTQ (B) Nmm (SD)
p-Value t-test
183 (82)
134 (70)
0.226
b0.0001⁎ b0.0001⁎ b0.0001⁎ 0.248 0.009⁎ 0.010⁎
Table 4 Percent of old and new bone areas within the ROI for all groups. Asterisk represents significant difference between test and control. Place
Removal torque tests
Old
The results and statistical analysis of the RTQ measurement for groups A and B are presented in a box plot (Fig. 5) and Table 2. The mean RTQ value for groups A and B placed in tibia was significantly higher than that of the corresponding group C (p = 0.009 and 0.01 respectively). For group A implants in the femoral condyle the difference in mean RTQ compared to the corresponding group C implants was not statistically significant (p = 0.248) (Table 2). Difference between installation torque and removal torque In tibia there was no significant difference between the difference between ITQ and RTQ for group A compared to group B (Table 3). Histological observations No statistical difference of bone area at the left side and the right side of the implant was detected neither for old nor for new bone (p = 0.724 and p = 0.534 respectively). For group A, placed in tibia, the area of old and new bone was significantly different compared to the corresponding group C (p = 0.005). The remaining sites showed no significant difference between old and new bone (Table 4). Discussion In the present study the effect of different static strains, at the time of insertion, on removal torque after 24 days of healing was evaluated. While several studies have investigated how dynamic load affects
New
Femur Tibia prox Tibia dist Femur Tibia prox Tibia dist
Test
Control
Implant
% bone in ROI
Implant
% bone in ROI
A A B A A B
47.6 56.0 76.1 22.3 19.6 12.5
C C C C C C
54.7 43.7 76.0 24.0 26.1 10.4
(9.4) (7.8) (4.4) (7.4) (6.4) (2.7)
(13.4) (7.1) (9.0) (3.5) (4.6) (3.5)
p-value t-test 0.128 0.005⁎ 0.968 0.568 0.005⁎ 0.068
bone remodeling [5,7,8,39], only a limited number of studies have analyzed how static strain affects bone remodeling. Currey found that the yield strain and ultimate strain of cortical bone from different species of mammals did not exhibit large variations [32]. Hence, it is assumed that the yield strain and ultimate strain values for cortical bone of rabbit do not deviate substantially from those of human cortical bone. In order to minimize the surface roughness of the implant that could possibly affect the outcome, turned implants were used [40]. It is likely that a turned surface has a slight grinding effect on the bone and the calculated strains might represent a slight overestimation. The maximum insertion torque beyond 10 s represents the maximum condensation applied to the surrounding bone. The maximum insertion torque for both test implants was significantly higher compared to those of the control implants, confirming that the condensation feature on the test implants created strains in the bone. Higher strains means higher interface stresses and a higher frictional force which increases the insertion torque for the test implants. After the implant installation procedure the bone surrounding the test implants was thus in a state of prestress. If the cutting portion of the implant cuts a thread in the contralateral cortical bone, a further increase in the ITQ occurs. As shown in Fig. 3 the tibial distal control
Fig. 5. Box plot of removal torque, asterisk representing significant difference between test and control.
788
A. Halldin et al. / Bone 49 (2011) 783–789
implants have a slightly increased ITQ at the end, a phenomenon which may be detectable for the control implants but not for the test implants. The analysis was consequently performed with the maximum ITQ beyond 10 s, neglecting this phenomenon. When an implant is inserted the bone is affected by the surgery. According to Eriksson et al. [41] a 0.5 mm border zone becomes necrotic due to the surgical trauma. Slaets et al. [42] inserted implants press-fit in the tibial diaphysis of rabbit. In the implant border zone, osteocytic lacunae devoid of osteocytes were found. This region had a maximum extension of 400 μm 28 days after implant surgery. The damaged bone was invaded by basic multicellular units the activities of which peaked after 2–4 weeks but were still ongoing 6 weeks after implant insertion. However it has been suggested that necrotic bone can serve as structural support during remodeling [43]. In the present study, the tibial implants were supported by cortical bone. The group A and group B implants generated strains, which theoretically exceeded the ultimate strain and the yield strain of cortical bone respectively. Including the dimension tolerances for the implant the strain levels for group B are between yield and ultimate strain and the strain levels for group A are beyond ultimate strain of cortical bone. The removal torque of groups A and B indicated however that, there was no extensive bone resorption at 24 days after surgery. On the contrary, the removal torque of the experimental implants was higher than that of the control implants, which were designed not to produce static strains in the bone. It is of great interest that the highest removal torque value was obtained for group A implants, for which the strains induced by far exceeded the ultimate strains of cortical bone. It is suggested that the increased removal torques in the tibia for group A and B implants compared to group C are due to remaining prestress. It is further suggested that there is a slow gradual decline in the bone stresses produced at implant insertion and that this decline consists of two components, where one component is a visco-elastic stress relaxation of the bone and the other is the normal remodeling by basic multicellular units whereby prestressed bone is replaced by new haversian systems, which are not prestressed [44]. Even though no significance was obtained the difference between ITQ and RTQ seems to be greater for group A compared to group B. This may indicate that the increased pressfit decreases more rapidly with higher static strain due to visco-elastic behavior and/or increased bone remodeling rate. According to Roberts et al. [45], the cortical bone of rabbit remodels three times faster than human cortical bone. Thus 24 days in rabbit bone would correspond to 72 days in human and providing prestressed state over the initial healing phase. In a study in sheep, Cordey et al. found that the remaining traction forces, exerted by cortical bone screws, after 1 and 5 weeks amounted to 72% and 62% of the original traction force respectively [46]. They found that the decrease in traction force as a function of time followed a logarithmic curve. In a study observing fracture healing in sheep using internal fixation, small areas of bone which had undergone plastic deformation were not removed by surface resorption but by internal remodeling [47]. In the current study, it is also noticeable that the differences in stability between test and control implants decreased at 24 days, both in absolute and relative terms. One interpretation of this may be that the retention of the control implants increases with time while the retention of the test implants decreases with time and that their retention graphs after some time will merge. The histomorphometric evaluation showed that for the femoral and distal tibial sites, no significant differences were detected in area of old bone or new bone between test and control implants. In the proximal tibia however group A implants had significantly larger old bone area than the control implants. Due to the limitations of the current study, it is difficult to interpret this phenomenon. However it may suggest that the increase in condensation did not adversely affect the amount of old bone, and therefore the press fit on the implant remained at 24 days.
Conclusions This study indicates that increased static bone strains do not affect an extensive bone resorption but provides higher installation torque, and a higher removal torque after 24 days in rabbit bone. Acknowledgments This work was supported by the Swedish Research Council (6212010-4760). We would like to acknowledge the assistance of Ms Petra Johansson and Ms Maria Hoffman for the animal surgery and sample preparation and Mr Stefan Johnsson at Astra Tech for preparing all implants. References [1] Wolff J. Das Gesetz der Transformation der Knochen. Berlin: Verlag von August Hirschwald; 1892. [2] Huiskes R, Ruimerman R, van Lenthe GH, Janssen JD. Effects of mechanical forces on maintenance and adaptation of form in trabecular bone. Nature 2000;405: 704–6. [3] Qin YX, Rubin CT, McLeod KJ. Nonlinear dependence of loading intensity and cycle number in the maintenance of bone mass and morphology. J Orthop Res 1998;16: 482–9. [4] Mosley JR, Lanyon LE. Strain rate as a controlling influence on adaptive modeling in response to dynamic loading of the ulna in growing male rats. Bone 1998;23: 313–8. [5] Rubin CT, Lanyon LE. Regulation of bone formation by applied dynamic loads. J Bone Joint Surg Am 1984;66:397–402. [6] Lanyon LE, Hampson WG, Goodship AE, Shah JS. Bone deformation recorded in vivo from strain gauges attached to the human tibial shaft. Acta Orthop Scand 1975;46:256–68. [7] Stanford CM, Brand RA. Toward an understanding of implant occlusion and strain adaptive bone modeling and remodeling. J Prosthet Dent 1999;81:553–61. [8] Rubin CT, Sommerfeldt DW, Judex S, Qin Y. Inhibition of osteopenia by low magnitude, high-frequency mechanical stimuli. Drug Discov Today 2001;6: 848–58. [9] Turner CH, Owan I, Takano Y. Mechanotransduction in bone: role of strain rate. Am J Physiol 1995;269:E438–42. [10] Gotfredsen K, Berglundh T, Lindhe J. Bone reactions adjacent to titanium implants subjected to static load. A study in the dog (I). Clin Oral Implants Res 2001;12:1–8. [11] Szmukler-Moncler S, Salama H, Reingewirtz Y, Dubruille JH. Timing of loading and effect of micromotion on bone–dental implant interface: review of experimental literature. J Biomed Mater Res 1998;43:192–203. [12] Adell R, Lekholm U, Rockler B, Branemark PI. A 15-year study of osseointegrated implants in the treatment of the edentulous jaw. Int J Oral Surg 1981;10:387–416. [13] Attard NJ, Zarb GA. Immediate and early implant loading protocols: a literature review of clinical studies. J Prosthet Dent 2005;94:242–58. [14] Schincaglia GP, Marzola R, Scapoli C, Scotti R. Immediate loading of dental implants supporting fixed partial dentures in the posterior mandible: a randomized controlled split-mouth study—machined versus titanium oxide implant surface. Int J Oral Maxillofac Implants 2007;22:35–46. [15] Donati M, La Scala V, Billi M, Di Dino B, Torrisi P, Berglundh T. Immediate functional loading of implants in single tooth replacement: a prospective clinical multicenter study. Clin Oral Implants Res 2008;19:740–8. [16] Schincaglia GP, Marzola R, Giovanni GF, Chiara CS, Scotti R. Replacement of mandibular molars with single-unit restorations supported by wide-body implants: immediate versus delayed loading. A randomized controlled study. Int J Oral Maxillofac Implants 2008;23:474–80. [17] Testori T, Galli F, Capelli M, Zuffetti F, Esposito M. Immediate nonocclusal versus early loading of dental implants in partially edentulous patients: 1-year results from a multicenter, randomized controlled clinical trial. Int J Oral Maxillofac Implants 2007;22:815–22. [18] Lindeboom JA, Frenken JW, Dubois L, Frank M, Abbink I, Kroon FH. Immediate loading versus immediate provisionalization of maxillary single-tooth replacements: a prospective randomized study with BioComp implants. J Oral Maxillofac Surg 2006;64:936–42. [19] Merli M, Merli A, Bernardelli F, Lombardini F, Esposito M. Immediate versus early non-occlusal loading of dental implants placed flapless in partially edentulous patients. One-year results from a randomised controlled trial. Eur J Oral Implantol 2008;1:207–20. [20] Cannizzaro G, Leone M. Restoration of partially edentulous patients using dental implants with a microtextured surface: a prospective comparison of delayed and immediate full occlusal loading. Int J Oral Maxillofac Implants 2003;18:512–22. [21] Cannizzaro G, Leone M, Esposito M. Immediate versus early loading of two implants placed with a flapless technique supporting mandibular bar-retained overdentures: a single-blinded, randomised controlled clinical trial. Eur J Oral Implantol 2008;1:33–43. [22] Ostman PO, Hellman M, Sennerby L. Direct implant loading in the edentulous maxilla using a bone density-adapted surgical protocol and primary implant stability criteria for inclusion. Clin Implant Dent Relat Res 2005;7(Suppl 1):S60–9.
A. Halldin et al. / Bone 49 (2011) 783–789 [23] Misch CM. Immediate loading of definitive implants in the edentulous mandible using a fixed provisional prosthesis: the denture conversion technique. J Oral Maxillofac Surg 2004;62:106–15. [24] Bashutski JD, D'Silva NJ, Wang HL. Implant compression necrosis: current understanding and case report. J Periodontol 2009;80:700–4. [25] Svindland AD, Nordsletten L, Reikeras O, Skjeldal S. Periosteal response to transient ischemia. Histological studies on the rat tibia. Acta Orthop Scand 1995;66:468–72. [26] Kalebo P, Johansson C, Albrektsson T. Temporary bone tissue ischemia in the hind limb of the rabbit. A vital microscopic study. Arch Orthop Trauma Surg 1986;105: 321–5. [27] Lanyon LE, Rubin CT. Static vs dynamic loads as an influence on bone remodelling. J Biomech 1984;17:897–905. [28] Perren SM, Huggler A, Russenberger M, et al. The reaction of cortical bone to compression. Acta Orthop Scand Suppl 1969;125:19–29. [29] Carter DR, Caler WE, Spengler DM, Frankel VH. Fatigue behavior of adult cortical bone: the influence of mean strain and strain range. Acta Orthop Scand 1981;52: 481–90. [30] Winwood K, Zioupos P, Currey JD, Cotton JR, Taylor M. The importance of the elastic and plastic components of strain in tensile and compressive fatigue of human cortical bone in relation to orthopaedic biomechanics. J Musculoskelet Neuronal Interact 2006;6:134–41. [31] McElhaney JH. Dynamic response of bone and muscle tissue. J Appl Physiol 1966;21:1231–6. [32] Currey JD. Tensile yield in compact bone is determined by strain, post-yield behaviour by mineral content. J Biomech 2004;37:549–56. [33] McCalden RW, McGeough JA, Barker MB, Court-Brown CM. Age-related changes in the tensile properties of cortical bone. The relative importance of changes in porosity, mineralization, and microstructure. J Bone Joint Surg Am 1993;75:1193–205. [34] Wennerberg A, Albrektsson T. Suggested guidelines for the topographic evaluation of implant surfaces. Int J Oral Maxillofac Implants 2000;15:331–44.
789
[35] Johansson CB, Jimbo R, Stefenson P. Ex vivo and in vivo biomechanical test of implant attachment to various materials: introduction of a new user-friendly removal torque equipment. Clin Implant Dent Relat Res 2010, doi:10.1111/j.17088208.2010.00296.x. [36] Donath K, Breuner G. A method for the study of undecalcified bones and teeth with attached soft tissues. The sage-schliff (sawing and grinding) technique. J Oral Pathol 1982;11:318–26. [37] Johansson CB, Morberg P. Importance of ground section thickness for reliable histomorphometrical results. Biomaterials 1995;16:91–5. [38] Reilly DT, Burstein AH. Review article. The mechanical properties of cortical bone. J Bone Joint Surg Am 1974;56:1001–22. [39] Goodship AE, Lanyon LE, McFie H. Functional adaptation of bone to increased stress. An experimental study. J Bone Joint Surg Am 1979;61:539–46. [40] Albrektsson T, Wennerberg A. Oral implant surfaces: part 1—review focusing on topographic and chemical properties of different surfaces and in vivo responses to them. Int J Prosthodont 2004;17:536–43. [41] Eriksson RA, Albrektsson T, Magnusson B. Assessment of bone viability after heat trauma. A histological, histochemical and vital microscopic study in the rabbit. Scand J Plast Reconstr Surg 1984;18:261–8. [42] Slaets E, Carmeliet G, Naert I, Duyck J. Early cellular responses in cortical bone healing around unloaded titanium implants: an animal study. J Periodontol 2006;77:1015–24. [43] McKibbin B. The biology of fracture healing in long bones. J Bone Joint Surg Br 1978;60-B:150–62. [44] Sedlin ED. A rheologic model for cortical bone. A study of the physical properties of human femoral samples. Acta Orthop Scand Suppl 1965;Suppl 83:1–77. [45] Roberts WE, Turley PK, Brezniak N, Fielder PJ. Implants: bone physiology and metabolism. CDA J 1987;15:54–61. [46] Cordey J, Blumlein H, Ziegler W, Perren SM. Study of the behavior in the course of time of the holding power of cortical screws in vivo. Acta Orthop Belg 1976;42 (Suppl 1):75–87. [47] Perren SM, Rahn BA. Biomechanics of fracture healing. Can J Surg 1980;23:228–32.