Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2018 May 1.
Published in final edited form as: J Orthop Res. 2016 Jul 20;35(5):997–1006. doi: 10.1002/jor.23360

Time Course of Peri-Implant Bone Regeneration around Loaded and Unloaded Implants in a Rat Model

Shailly H Jariwala 1, Hwabok Wee 1, Evan P Roush 1, Tiffany L Whitcomb 2, Christopher Murter 1, Gery Kozlansky 1, Akhlesh Lakhtakia 3, Allen R Kunselman 4, Henry J Donahue 5, April D Armstrong 1, Gregory S Lewis 1,*
PMCID: PMC5800527  NIHMSID: NIHMS938563  PMID: 27381807

Abstract

The time-course of cancellous bone regeneration surrounding mechanically loaded implants affects implant fixation, and is relevant to determining optimal rehabilitation protocols following orthopaedic surgeries. We investigated the influence of controlled mechanical loading of titanium-coated polyether-ether ketone (PEEK) implants on osseointegration using time-lapsed, non-invasive, in vivo micro-computed tomography (micro-CT) scans. Implants were inserted into proximal tibial metaphyses of both limbs of eight female Sprague-Dawley rats. External cyclic loading (60 μm or 100 μm displacement, 1 Hz, 60 seconds) was applied every other day for 14 days to one implant in each rat, while implants in contralateral limbs served as the unloaded controls. Hind limbs were imaged with high-resolution micro-CT (12.5 μm voxel size) at 2, 5, 9, and 12 days post-surgery. Trabecular changes over time were detected by 3D image registration allowing for measurements of bone-formation rate (BFR) and bone-resorption rate (BRR). At day 9, mean %BV/TV for loaded and unloaded limbs were 35.5 ± 10.0 % and 37.2 ± 10.0 %, respectively, and demonstrated significant increases in bone volume compared to day 2. BRR increased significantly after day 9. No significant differences between bone volumes, BFR, and BRR were detected due to implant loading. Although not reaching significance (p = 0.16), an average 119 % increase in pull-out strength was measured in the loaded implants.

Keywords: Implant fixation, micro-computed tomography, mechanical loading, image registration, bone remodeling

Introduction

Early stable fixation of implants is paramount in the treatment of many orthopaedic disorders. Methods to increase peri-implant cancellous bone mass, especially in osteoporotic patients, are highly desirable clinically. However, in uncemented implants, direct bone contact to the implant and a robust biological fusion depend on the biological as well as mechanical environments of the implant. Following initial tissue differentiation, adaptation processes occur over time in the peri-implant region (1–4). Differentiation and adaptation can respond to mechanical loading, thereby affecting osseointegration (1). Many orthopaedic implants undergo mechanical loading, which is affected by post-surgical rehabilitation protocols. Multiple factors can affect peri-implant tissue stresses and micromotion of the implant relative to the surrounding bone (5, 6), including physical properties of bone tissue, mechanical interaction between bone and the implant, and the geometry of the implant (7–9). Understanding how mechanical loading interacts with normal bone healing is important for designing future implants and biomaterials.

Some in vivo mechanical loading studies have shown that immediate application of load during bone regeneration has a beneficial effect on osseointegration (5, 10–13). Bone chamber models have been employed (10, 12, 14) to measure de novo bone formation; however, these models do not incorporate adaption of pre-existent bone to mechanical loading. Zhang et al. (15) investigated effects of various controlled high- and low-frequency mechanical loading on peri-implant bone healing in a bicortical rat tibial implant model. Wazen et al. (6) emphasized that the interfacial strain magnitude is a critical factor in the biological response to implant micromotion. However, there are limited experimental models that explore the adaptive response of the bone to controlled mechanical loading (16). More recently, van der Muelen et al. (16) studied the adaptation of trabecular bone in response to mechanical loading in a rabbit distal femoral condyle model. They concluded that although 4 weeks of applied cyclic compressive loading increased trabecular bone volume fraction and enhanced cancellous bone integration into porous titanium foam implants, it also altered the trabecular architecture (17, 18).

Much of the literature related to loaded implant osseointegration is based on static analysis of single CT/micro-CT scans, histology, or pull-out/push-out tests performed on animals after euthanasia (4, 19). These methods do not allow investigation of time-varying, 3D bone remodeling processes occurring in the peri-implant region including temporal bone formation and resorption. Bone resorption rates have been determined by quantitative histomorphometry, but this technique is 2D in nature and is susceptible to large inter-observer variation (20). Time-lapsed in vivo micro-CT based studies have been performed previously for tracking changes in bone in response to osteoarthritis drug applications (21–26). Schulte et al. (27) characterized the relation between the mechanical environment and bone formation/resorption in tail vertebrae of mice using time-lapsed in vivo micro-CT and micro-finite element (FE) models calculating the local strain distribution. Li et al. (28) characterized spatio-temporal changes of peri-bone remodeling in mice using time-lapsed micro-CT. Kettenberger et al. (26) investigated fixation of zoledronate infused implants with in vivo longitudinal micro-CT scans. However, in vivo applications of micro-CT to study response of mechanical loading on peri-implant bone response have been limited (28, 29).

We have developed a novel rat model and imaging protocol that allows us to quantify the time course of peri-implant bone regeneration around loaded and unloaded implants positioned in metaphyseal bone. The analysis of the longitudinal micro-CT scans is based on image registration and analysis of consecutive scans of the same animal that result in identification of locations of resorbed and newly formed bone. Additionally implant fixation strength was quantified.

Methods

Implant

Sixteen cylindrical polyether-ether ketone (PEEK) implants having major diameter of 0.80 mm and 8.6 mm in length were fabricated using computer-controlled machining as shown in Figure 1(a). The bottom 2 mm-length portion of the implant had a grooved portion with three channels, each with depth of 0.15 mm. The grooved portion of the implant was coated at 2.5 mTorr pressure in an argon atmosphere with titanium (thickness 100 nm at 0.1 nm/sec) in a pulsed-dc magnetron sputtering system (CMS-18, Kurt Lesker Inc., Jefferson Hills, PA) operated at 200-W dc power. Three radiopaque glass microspheres (100 μm dia.) were fastened within the grooved portion of the implant, using translucent Momentive RTV 108 silicone adhesive, for use as landmarks for micro-CT image analyses and registration from various time points (Fig. 2(a)). These microspheres were fastened within holes precision machined into known 3D positions in the implant as part of the original implant fabrication process. The implant was held in place using a polycarbonate housing which included a housing plate, a nitrile o-ring, and a housing cap (Figure 1(a, b)). The designs of the implant and mechanical loading setup are based partly on literature (6, 11, 30) but modified to enable the loading of an implant within the proximal rat tibia. The grooved portion of the implant was implanted into proximal tibial metaphyses of both limbs of eight female Sprague-Dawley rats aged 12–14 weeks (250–325 g) (Charles River Laboratories, Wilmington MA) (Figure 1(c)), while the top larger diameter (3.5 mm) portion of the implant remained percutaneous, covered by the housing cap for external loading.

Figure 1.

Figure 1

Open in a new tab

(a) PEEK implant coated with titanium on its grooved portion, and the implant housing components. (b) Assembly of implant and its housing components. The o-ring was inserted between the housing plate and implant (to return the implant back to its initial position when under cyclic loading), the implant was inserted through the central hole in the housing plate, and the housing cap was then fastened. (c) Photograph of the implant fixed in the tibia with the grooved portion embedded in the proximal tibia metaphysis. The cap (here removed) remained percutaneous in vivo. The housing plate was fixed to the bone using screws. (d) Controlled cyclic mechanical loading applied in vivo to percutaneous implant up to 14 days post-surgery. Inset is magnified image of the rat limb loading set up with the loading (actuator) pin.

Figure 2.

Figure 2

Open in a new tab

(a) Three radiopaque microspheres (100 μm dia.) shown as A, B, & C, were embedded into the implants at precise locations for Micro-CT scan registration at various time points. (b) Preparation for an in vivo micro-CT scan of a rat under anesthesia with both the tibiae held by a customized jig. Inset is zoomed image showing the rat hind limbs held together at the knees by a velcro tie and ankles fastened by the customized jig for simultaneous imaging. (c) Novel customized rat restraint jacket was designed to prevent the rodents from disturbing the percutaneous implants. The rat jacket was built out of a polyethylene plastic exterior with a soft cloth like material, shown in black, in the interior. There were two shoulder straps designed to go around the arms of the rodents and attach to the back of the jacket using the velcro strips.

Animal model and surgical procedure

All animal procedures were approved by Institutional Animal Care and Use Committee of the College of Medicine, Pennsylvania State University. The surgery was performed under aseptic conditions by either A.A., an experienced orthopaedic surgeon, or G.L. with guidance from A.A. Implants were sterilized with 70 % ethanol and kept under UV light for 1 hour prior to surgery. All of the surgeries were performed under isoflurane anesthesia maintained at 1.5–3% by mask and with Buprenorphine-SR analgesia (1.2 mg/kg) injected subcutaneously prior to surgery. The animals were provided thermal support throughout the surgery. The skin covering the antero-medial side of the tibia was shaved and cleaned with betadine and 70% alcohol pads. A 3 cm skin incision was made, followed by blunt dissection of connective tissue to expose the proximal tibia. The housing plate was placed such that the implant would be positioned 2–3 mm below the proximal tibia growth plate, and fastened with two bicortical titanium screws (1.3 mm dia., 5–6 mm length) (Figure 1(c)). A unicortical 0.9-mm diameter hole was drilled transversely into the metaphyseal bone of the proximal tibia. Each implant was inserted with the nitrile o-ring (to facilitate mechanical loading) placed between the head of the implant and housing plate (Figure 1(c)), and the housing cap was screwed onto the housing plate (Figure 1(b)). The muscle and skin were closed with 4.0 Ethilon® nylon sutures (Ethicon, Inc.) around the housing cap and thermal support was not withdrawn until the rat was ambulatory after recovering from anesthesia. Post-surgery rats were fed with soft food and allowed to ambulate freely. Custom made restraint jackets (Figure 2(c)) and Elizabethan collars (Kent Scientific Corp, USA) were utilized to prevent self-trauma of the implants and incisions. Post-surgery, the animals were checked daily for swelling, exudate, or pain.

Mechanical loading

Controlled external cyclic loading was applied to one implant in each rat (n = 8) on days 2, 5, 7, 9, 12, and 14 via an electronically controlled mechanical test machine (Fig. 1(d)) (LM1 TestBench, Bose, Eden Prairie MN). The rat was anesthetized and the limb was supported under the loading actuator. The implant housing cap was fastened into the loading holder with two screws. The loading pin passed through a hole in the housing cap to permit displacement controlled loading of the implant. The actuator cyclically displaced the implant by either 60 μm (3 animals) or 100 μm (5 animals), at a rate of 1 Hz, for 60 seconds at each time point (5). The peak force (averaged across cycles) was measured using a 22 N load cell (Bose, Eden Prairie MN). The loaded implant was randomized between left and right limbs while the contralateral limb served as the unloaded control.

Micro-CT longitudinal imaging

Rat hind limbs were imaged in vivo with high-resolution micro-CT (VivaCT 40, Scanco, Brüttisellen, Switzerland) at 2, 5, 9, and 12 days post-surgery. The animals (n = 8) were kept under isoflurane anesthesia and provided thermal support during the scans, each of which lasted about 20 minutes. A customized rat holding fixture was designed (Fig. 2(b)) to image both limbs simultaneously and minimize the number of scans on each animal. The implant and surrounding bone were imaged using following scanning parameters: 12.5 μm voxel size at 70 kV, 145 μA 250 ms integration time, 420 slices.

3D image registration and analysis

Micro-CT scans from various time points from a single limb were adjusted for contrast using a consistent window and converted to 8-bit image files. Scans were imported into Avizo software (VSG US, Burlington MA). After transforming the scan from the first time point to a standardized 3D pose with the implant’s longitudinal axis at x = y = 0 and aligned with the z direction, the scans were registered together in 3D by a least-squares landmark-based registration using the 3 radiopaque microspheres within the implant. The registered stacks were then imported into a custom Matlab code (Mathworks, Natick MA). Image volumes were cropped to a 300x300 square matrix (3.75 x 3.75 mm), filtered using a Gaussian filter (31), and then registered with the image volume from the previous time point (Figure 3(a)). This registration used a Matlab rigid grayscale intensity-based transformation routine with mutual information metric and regular step gradient descent optimization.

Figure 3.

Figure 3

Open in a new tab

(a–c) show 2D slices taken from the 3D analysis, also visualized in (d). (a) Images were registered using landmark-based followed by intensity-based methods. Binarized overlapped scans of day 2 & day 5 are shown before and after the intensity-based registration. Green denotes resorbed bone areas; purple indicates new bone areas. Areas of quiescent (unchanged) bone are colored in white. The orange arrows indicate several areas that change and align due to registration. (b) Registered images cropped to get a peri-implant volume of interest, with 0.8 mm diameter central cylinder corresponding to implant location removed. (c) The volume of interest thresholded to calculate bone volume. (d) Three-dimensional (3D) visualization of the peri-implant bone volume indicating bone formation and resorption sites of a sample from the loaded group. The new bone and resorbed bone volumes were used to compute the BFR and BRR, respectively.

After the registration was complete, a peri-implant cylindrical volume of interest (VOI) twice the diameter of the implant (1.6 mm) and 1.25 mm in length (100 slices) was identified. This VOI was concentric with the axis of the implant as known from the microsphere coordinates. The implant cylinder volume (0.8 mm dia.) was subtracted to create a tubular VOI (Fig. 3(b)), which was then binarized with a level (threshold) determined from the lowest point in the trough between the two peaks of the image histogram having a bi-modal distribution (Fig. 3(c)). The above tubular VOI extended 0 – 400 μm radially, away from the implant surface. Additionally, this VOI was split into tubular VOI’s with 0 – 200 μm and 200 – 400 μm proximity to the implant for additional analysis (Supplementary Figure S-2). Bone volume fraction (bone volume (BV)/total volume (TV)) as shown in Figure 3(b, c) was determined utilizing the known voxel size. By comparing consecutive scans and subtracting each consecutive scan from the previous time point, voxel by voxel, resorbed, newly formed, and quiescent bone regions were identified. The bone-formation rate (BFR) was calculated as a ratio of new bone volume to total bone volume divided by the number of days between the two scans (%/day) (22). The bone-resorption rate (BRR) was calculated similarly but using resorbed bone.

Reproducibility was assessed for all the remodeling parameters (BV/TV, BFR, and BRR) shortly after animal sacrifice at the end of the experiment in two additional animals. Two unloaded rat limbs were scanned ex vivo 5 times each with a similar protocol used for the in vivo measurements with repositioning of the limbs between the scans. The repeated scans thus correspond to the last in vivo measurements at day 12 and were registered to their respective previous measurement scan at day 9. Parameters BV/TV, BFR and BRR were calculated. The reproducibility was characterized using precision errors (PE) as described in (32), therefore, the bone parameters were expressed both in absolute values (PESD) and as coefficients of variation (PE%CV) (28).

Implant pull-out test

Following euthanization at day 14, loaded and unloaded rat tibiae (n = 5) were explanted and stored in phosphate-buffered saline for immediate pull-out tensile testing (LM1 TestBench, Bose, Eden Prairie MN). The tibia was positioned flush against the bottom side of an aluminum plate, which was fastened to the mechanical test frame. After removing the housing cap, the exposed head of the implant was connected to the 222 N load cell (Interface, Inc.) via a custom-made clamp (Figure 4) and was then subjected to a constant pull rate of 5μm/sec. The alignment of the implant head was assured visually by capturing macro-lens images (before pull-out test began) using a Canon Ultrasonic 100 mm Macro lens camera (Fig. 4, Supplementary video VS1). The actuator force was measured and plotted against displacement, and the peak pull-out force was determined.

Figure 4.

Figure 4

Open in a new tab

Implant pull-out test set up following euthanization. Implants were pulled out from rat tibiae using the exposed implant head (appearing tan colored, following cap removal). The tibia (located beneath the field of view in this photo) was attached to an aluminum plate allowing the implant fixation strength to be tested. The implant head was carefully connected to the load cell using a custom-made clamp.

Histology

After pull-out testing, each rat tibia was dissected and fixed in 10% neutral buffered formalin solution for three days before dehydration in a series of ethanol solutions with ascending concentration (70%, 80%, 95% and 100%). Every specimen was defatted using xylene and xylene substitute (Clear-Advantage Cat. 24770), and then embedded by infiltration with methyl methacrylate Osteo-Bed Plus resin (Polysciences, Inc.). After polymerization for 48 hours at 34°C, each specimen was cut using an Isomet low speed saw (Buehler Ltd, USA) into slices of 200 μm thickness. The slices were glued onto glass slides using Cytoseal 60TM (Richard-Allan Scientific TM) and ground to approximately 50–60 μm thickness. Finally, the slices were stained with Sanderson’s rapid bone stain and van Geison counter stain (Dorn and Hart Microedge, Inc., Villa Park, IL). Images were taken with an upright light microscope (Nikon Optiphot2, Nikon Instruments).

Statistics

Bone volumes, BFR, and BRR were analyzed using a repeated-measures design with two repeated factors per animal: group (i.e., loaded, unloaded) and day of assessment (e.g., 2, 5, 9, and 12). For this repeated measures design, general linear models (GLM) with correlated errors (33) were fit to assess the effects of group, day, and the interaction of group and day on continuous outcomes (34). Pull-out tests also were analyzed using a GLM with correlated errors to compare the loaded and unloaded groups. The data for the different loading regimes (60 and 100 μm) within the loaded groups were lumped together for statistical analysis of the loaded group. All hypotheses tests were two-sided and all analyses were performed using SAS software, version 9.4 (SAS Institute Inc., Cary, NC). Significance was considered at a level of p < 0.05. For this study, a sample size of 8 animals would provide 82% power to detect a standardized effect size, i.e. the difference in means divided by the standard deviation, of 1.2 using a two-sided test having a significance level of 0.05.

Results

Influence of mechanical loading on peri-implant bone volume

Figure 5 depicts the peri-implant bone regeneration around a representative loaded implant. We saw no significant differences between bone volumes of loaded versus unloaded groups at each time point. Bone volumes and BFR were seen to be elevated at earlier time points but did not continue to increase after day 9 during the second week, and even decreased in some animals (Fig. 6(a, b)). On day 9, the mean %BV/TV for loaded and unloaded limbs were 35.5 ± 10.0 % (SD) and 37.2 ± 10.0 % (SD), respectively. By day 9 both the loaded and unloaded peri-implant regions demonstrated statistically significant increases in BV/TV, averaging 14.6% (p < 0.0001) and 19.0% (p < 0.0001) greater than on day 2, respectively. By day 12, the mean %BV/TV for loaded and unloaded limbs were 31.7 ± 10.9 % and 35.6 ± 13.0%, respectively.

Figure 5.

Figure 5

Open in a new tab

Micro-CT images of the peri-implant region of the same mechanically loaded rat tibia 2, 5, 9, and 12 days post-surgery. Scans are 3D registered.

Figure 6.

Figure 6

Open in a new tab

Peri-implant (a) bone volume (BV/TV), (b) bone formation rate (BFR), and (c) bone resorption rate (BRR) as functions of time (days) for unloaded and loaded hind limbs of rats. Asterisks (*) indicate statistical significance (p < 0.05) for BV/TV and BRR between time intervals considered, for both loaded and unloaded limbs.

Bone remodeling rates

The BFRs did not change significantly between time points for loaded and unloaded groups (Figure 6(b)). The BRRs (Fig. 6(c)) between day 2-day 5 were significantly lower than between day 9 –12 (p < 0.05) for both groups. The BRR was significantly different between intervals 5–9 and 9–12, and between intervals 2–5 and 9–12, for both loaded (p < 0.001) and unloaded groups (p < 0.05). The BRR for both loaded and unloaded groups increased significantly from interval 5–9 to interval 9–12 (p < 0.05). We saw no significant differences between bone formation and resorption rates of loaded versus unloaded groups at each time point.

The 200–400 μm VOI had significantly higher means (by paired t-test) for BV/TV, BFR, and BRR at most time points in comparison to 0–200 μm VOI closest to the implant (Supplementary Figure S-3).

Supplementary Table S-1 depicts the reproducibility results of the BV/TV, BFR and BRR for the last time interval (day 9–12). Precision error in terms of coefficient of variation, PE%CV, was 1.4 % for BV/TV, with BV/TV standard deviations of 0.7% and 0.3% across the five scans for the two different animals. PE%CV for BFR and BRR were 7.5% and 5.0%, respectively. The peak force during application of cyclic loading of rat limbs at each time point is shown in Figure 7(a).

Figure 7.

Figure 7

Open in a new tab

(a) Peak force measured during controlled-displacement cyclic loading of implants at each time point. Data are means ± standard deviations (n = 8). (b) Peak pull-out force data from loaded and unloaded rat implants 14 days post-surgery. Data are means ± standard deviations (n = 5 implants per group).

Pull-out strength

After 14 days, maximum pull-out force was 11.2 ± 4.6 N for loaded limbs (n = 5, 2 from 60 μm 3 from 100 μm displacement) and 5.1 ± 3.0 N for the unloaded limbs (Figure 7(b)). However, we did not observe significant differences (n = 5, p = 0.16) in pull-out strengths between the loaded and unloaded groups. Three each from loaded and unloaded groups had premature implant losses from their tibiae after day 12 and thus were not included in the pull-out strength analyses.

Using the 0 - 400 μm (1.6 mm outer dia.) VOI, we did not detect any correlation between the peri-implant BV/TV at the last observation point and the pull-out strength for both groups (p > 0.05). However, a statistically significant correlation was observed between BV/TV and pull-out strength for the unloaded group (r = 0.93, p = 0.02, 95% CI 0.27 to 0.99) when analyzing the VOI within 200 μm proximity to the implant (Supplementary Figures S-1 and S-2).

Histology

Histological observation did not reveal any obvious differences concerning bone formation between loaded and unloaded implants at 2 weeks post-surgery, as depicted in Figure 8. The newly synthesized bone formation was restricted to the peri-implant space (Fig. 8(a, b)) in both the loaded and unloaded groups.

Figure 8.

Figure 8

Open in a new tab

Representative histological sections of (a) an unloaded and (b) a loaded rat tibiae 14 days after surgery, post pull-out testing (at low magnification (4X)). (c) High-magnification (10X) image of the peri-implant region of loaded histological section in (b) (region within black box) showing the mineralized bone (varying shades of pink) adjacent to the implant. The sections were stained with Sanderson’s rapid bone stain and counterstained with Van gieson; osteoid seam and osteocytes are in light blue and osteoblasts in darker blue. Black box in (a) indicates peri-implant region. Scale bars, 1000 μm in (a) and (b), and 100 μm in (c). Black arrow indicates bone-implant interface with possible direct bone contact with implant prior to pull-out.

Discussion

This study presents a novel animal model for quantifying formation and resorption in mineralized peri-implant tissue across a series of time points in mechanically loaded implants. Implant loading did not significantly alter the peri-implant bone volumes, BFR, and BRR in comparison to unloaded implants. Although not reaching significance (p = 0.16), an average 119 % increase in pull-out strength was measured in the loaded implants compared to the unloaded implants in our study. Within 200 μm proximity to the implant, day 12 bone volume was significantly correlated with the pull-out force for unloaded, but not loaded implants. The histological analyses for both the groups showed that newly synthesized bone was deposited primarily in contact or around the implant, in-line with the micro-CT scan results.

The quantification of bone resorption and formation rates separately using micro-CT allows for direct time-dependent assessment of changes occurring in the peri-implant region (21, 26, 28). One of the significant advantages of this technique was that it allowed obtaining data from multiple time points without having to sacrifice animals compared to traditional 2D histology techniques. Our image analysis focused on trabecular bone surrounding the implants. Cortical bone thickness may play an important role in implant fixation (35) but was not measured due to difficulty identifying its boundary with the trabecular bone in some animals.

Previous literature has shown mixed effects of applied immediate implant loading on osseointegration in the peri-implant region (5, 36–38). Zhang et al. (15) reported enhanced osseointegration with immediate loading in comparison to delayed loading in a rat tibial implant model. In our study, we observed higher bone volumes at earlier time points up to 9 days post-implantation for loaded implants, in-line with previous findings (17, 36). However, a similar increase was observed in bone formation for our unloaded implants. The unloaded implants, although not loaded externally, were not completely unloaded as the rats were allowed to ambulate freely in their cages during the entire period of this study, resulting in mechanical stimulus applied throughout both hindlimbs.

In our study, implants were loaded at a frequency of 1 Hz consistent with loading due to ambulation (as in a knee replacement) (6, 39). Rat limbs were loaded 3 days per week in order to reduce anesthesia effects. Increased frequency of loading may have altered bone remodeling. Duyck et al. (39) applied static and dynamic loads at 1 Hz to bicortical implants in rabbit tibiae and observed that excessive cyclic loading (90 or 270 cycles/day at 1 Hz.) could result in bone loss lateral to the integrated implants (39). Vandamme et al. (5) studied the effect of micro-motion (30 and 90 μm displacements at 1 Hz, 3 times per week) in a rabbit model and concluded that the bone-to-implant contact was significantly higher for 90 μm implant displacement compared with 30 μm. However, no micro-CT scanning was performed. A recent study by Piccinini et al. (40) also reported implant losses in both loaded and unloaded groups in a bilateral rat percutaneous implant model, although unlike our study their study involved two diaphyseal (not metaphyseal) tibial implants with motion that brought the implant heads together.

It appears from literature that higher micromotion range can be deleterious resulting in an interfacial fibrous tissue formation, whereas a lower micromotion range may stimulate interfacial bone formation (41, 42). Wazen et al. (6) subjected implants in the mid-shaft of mouse tibiae to displacements of 150 or 300 μm, and concluded that bone remodeling was disrupted in the areas of higher strains. Pilliar et al. (41) also studied the influence of micromotion (150 μm) on bone ingrowth and concluded that micromotion >150 μm results in fibrous tissue encapsulation. The tolerated threshold of micromotion was found to lie between 50–150 μm (43) with up to 100 μm micromotion tolerated for implants with a bioinert surface (44). However, it is difficult to make direct comparisons of our results with previous literature, due to variation in the implant materials, animal models, and loading frequencies used.

The high bone volumes and BFRs observed in our study on day 9 did not change significantly after day 9, whereas BRRs were observed to increase for both loaded and unloaded limbs during this time period. This subsequent increase in BRRs could correspond to the bone-remodeling process that occurs after initial tissue differentiation. Slaets et al. (36, 45) studied histomorphometrically the bone regeneration with titanium implants in rabbit tibiae and observed simultaneous osteoclast and osteoblast activity from day 7 onwards. Van der Meulen et al. (17) used a rabbit model to study cancellous functional adaptation to mechanical loads and also observed a transient increase in bone mass only after 2 weeks. Interference of the early wound healing response associated with immediate implant loading may have occurred as reported in previous studies (46).

Both limbs received PEEK implants with a titanium (Ti) coating. Walsh et al. (47) reported that plasma-sprayed titanium coating on PEEK improved the bone-implant interface in comparison to plain PEEK. Furthermore, previous literature has reported that tissue response to mechanical loading is also dependent on the surface characteristics of the implant (7) and that hydroxyapatite (HA) coated implant surface induces more bone contact than a Ti surface (7, 42). Mouzin et al. (7) demonstrated that mechanical loading increased bone coverage and decreased fibrous tissue coverage around HA-coated implants, but conversely loading increased fibrous tissue coverage around Ti implants.

Use of in vivo micro-CT subjects the animals to ionizing radiation that may affect bone architecture (48, 49). However, Waarsing et al. (23) reported a radiation exposure of 0.4 Gy for a single 20-minute micro-CT scan of a rat hind limb had no significant deleterious effects on bone regeneration.

A primary limitation of the study is the limited sample size and power. Eight per group were analyzed with micro-CT parameters, but only five per group were ultimately tested for pull-out as implant loss was observed in some animals after day 12. In this study, to the best of our knowledge the implants were unperturbed by the rats, and we attempted to minimize any such disturbance due to the use of custom made restraint jackets and Elizabethan collars. We only investigated the bone re-modeling response up to 14 days of loading, thus focusing on the initial tissue differentiation phase and not long-term bone re-modeling. Our loading applied is not in the direction it is normally loaded in vivo such as in knee replacements, perpendicular to the lateral surface of the tibia (50), as this was deemed not feasible in a rat model. A limitation is that dynamic histomorphometry was not performed and it can detect bone formation at higher resolution than our micro-CT based method. However dynamic histomorphometry cannot assess bone formation and resorption at multiple time points in the same animal as possible with in vivo micro-CT based methods.

Supplementary Material

Supp FigS1

Figure S-1. Graph demonstrating significant correlation observed between peri-implant BV/TV [%] at the last observation point (day 12) and the peak pull-out force [N] of the unloaded group for the 0–200 μm proximity VOI around the implant.

Supp FigS2

Figure S-2. Peri-implant bone volume (BV/TV), BFR, and BRR for (a) 0–200 μm and (b) 200–400 μm VOI around the implant plotted as functions of time (days) for unloaded and loaded hind limbs of rats.

Supp FigS3

Figure S-3. Comparison of peri-implant (a) bone volume (BV/TV), (b) BFR, and (c) BRR for 0–200 μm and 200 – 400 μm radial VOI around the implant plotted as functions of time (days). The unloaded and loaded groups were lumped for the different regions for statistical analysis. Asterisks (*) indicate statistical significance (p < 0.05). Significantly higher BV/TV, BFR and BRR values were observed in most cases for the 200 – 400 μm VOI.

Supp TableS1
Supp VideoS1
Download video file (10.2MB, wmv)

Acknowledgments

The project described was supported by the National Center for Advancing Translational Sciences, Grant KL2 TR000126. We gratefully acknowledge assistance from Emmanuel M. Paul (Division of Musculoskeletal Sciences), Muhammad Faryad and Sema Erten (Department of Engineering Science and Mechanics), and the residents and staff in the Department of Comparative Medicine at Penn State. AL thanks the Charles Godfrey Binder Endowment at Penn State for ongoing support of his research activities. April Armstrong is a consultant for Zimmer, and Gregory Lewis has received implant materials from Depuy-Synthes for research, but these companies were not involved in the present research. There are no other conflicts of interest.

Footnotes

Author’s Contributions: Research design (SJ, HW, AK, HD, AA, & GL), mechanical design and implants (SJ, HW, ER, CM, AL, GL), surgeries (SJ, HW, ER, TW, AA, GL), data acquisition (SJ, HW, ER, GL), data analysis (SJ, AK, GL); data interpretation (SJ, AA, GL), manuscript preparation (SJ with AA, GL). All authors have read and approved the final manuscript.

References

  • 1.Liu X, Niebur GL. Bone ingrowth into a porous coated implant predicted by a mechano-regulatory tissue differentiation algorithm. Biomechanics and modeling in mechanobiology. 2008;7(4):335–44. doi: 10.1007/s10237-007-0100-3. [DOI] [PubMed] [Google Scholar]
  • 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(6787):704–6. doi: 10.1038/35015116. [DOI] [PubMed] [Google Scholar]
  • 3.Geris L, Vander Sloten J, Van Oosterwyck H. In silico biology of bone modelling and remodelling: regeneration. Philos Trans A Math Phys Eng Sci. 2009 May 28;367(1895):2031–53. doi: 10.1098/rsta.2008.0293. [DOI] [PubMed] [Google Scholar]
  • 4.Brånemark R, Öhrnell L, Nilsson P, Thomsen P. Biomechanical characterization of osseointegration during healing: an experimental in vivo study in the rat. Biomaterials. 1997;18(14):969–78. doi: 10.1016/s0142-9612(97)00018-5. [DOI] [PubMed] [Google Scholar]
  • 5.Vandamme K, Naert I, Geris L, Sloten JV, Puers R, Duyck J. The effect of micro-motion on the tissue response around immediately loaded roughened titanium implants in the rabbit. Eur J Oral Sci. 2007;115(1):21–9. doi: 10.1111/j.1600-0722.2007.00416.x. [DOI] [PubMed] [Google Scholar]
  • 6.Wazen RM, Currey JA, Guo H, Brunski JB, Helms JA, Nanci A. Micromotion-induced strain fields influence early stages of repair at bone–implant interfaces. Acta biomaterialia. 2013;9(5):6663–74. doi: 10.1016/j.actbio.2013.01.014. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Mouzin O, Søballe K, Bechtold JE. Loading improves anchorage of hydroxyapatite implants more than titanium implants. J Biomed Mater Res. 2001;58(1):61–8. doi: 10.1002/1097-4636(2001)58:1<61::aid-jbm90>3.0.co;2-s. [DOI] [PubMed] [Google Scholar]
  • 8.Van Oosterwyck H, Duyck J, Vander Sloten J, Van der Perre G, De Coomans M, Lieven S, et al. The influence of bone mechanical properties and implant fixation upon bone loading around oral implants. Clin Oral Implants Res. 1998;9(6):407–18. doi: 10.1034/j.1600-0501.1996.090606.x. [DOI] [PubMed] [Google Scholar]
  • 9.Simmons CA, Meguid SA, Pilliar RM. Differences in osseointegration rate due to implant surface geometry can be explained by local tissue strains. Journal of Orthopaedic Research. 2001;19(2):187–94. doi: 10.1016/S0736-0266(00)90006-8. [DOI] [PubMed] [Google Scholar]
  • 10.Geris L, Vandamme K, Naert I, Vander Sloten J, Duyck J, Van Oosterwyck H. Application of mechanoregulatory models to simulate peri-implant tissue formation in an in vivo bone chamber. J Biomech. 2008;41(1):145–54. doi: 10.1016/j.jbiomech.2007.07.008. [DOI] [PubMed] [Google Scholar]
  • 11.Leucht P, Kim J, Wazen R, Currey JA, Nanci A, Brunski JB, et al. Effect of mechanical stimuli on skeletal regeneration around implants. Bone. 2007;40(4):919–30. doi: 10.1016/j.bone.2006.10.027. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Vandamme K, Naert I, Geris L, Sloten JV, Puers R, Duyck J. Histodynamics of bone tissue formation around immediately loaded cylindrical implants in the rabbit. Clin Oral Implants Res. 2007;18(4):471–80. doi: 10.1111/j.1600-0501.2007.01339.x. [DOI] [PubMed] [Google Scholar]
  • 13.Rubin CT, McLeod KJ. Promotion of bony ingrowth by frequency-specific, low-amplitude mechanical strain. Clin Orthop. 1994;298(8):165–74. [PubMed] [Google Scholar]
  • 14.Duyck J, Vandamme K. The effect of loading on peri-implant bone: a critical review of the literature. J Oral Rehabil. 2014;41(10):783–94. doi: 10.1111/joor.12195. [DOI] [PubMed] [Google Scholar]
  • 15.Zhang X, Duyck J, Vandamme K, Naert I, Carmeliet G. Ultrastructural characterization of the implant interface response to loading. J Dent Res. 2014 Mar;93(3):313–8. doi: 10.1177/0022034513518345. [DOI] [PubMed] [Google Scholar]
  • 16.van der Meulen, Marjolein CH, Morgan TG, Yang X, Baldini TH, Myers ER, Wright TM, et al. Cancellous bone adaptation to in vivo loading in a rabbit model. Bone. 2006;38(6):871–7. doi: 10.1016/j.bone.2005.11.026. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.van der Meulen, Marjolein CH, Yang X, Morgan TG, Bostrom MP. The effects of loading on cancellous bone in the rabbit. Clinical Orthopaedics and Related Research®. 2009;467(8):2000–6. doi: 10.1007/s11999-009-0897-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Willie BM, Yang X, Kelly NH, Han J, Nair T, Wright TM, et al. Cancellous bone osseointegration is enhanced by in vivo loading. Tissue Engineering Part C: Methods. 2010;16(6):1399–406. doi: 10.1089/ten.tec.2009.0776. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Shah NJ, Hyder MN, Moskowitz JS, Quadir MA, Morton SW, Seeherman HJ, et al. Surface-mediated bone tissue morphogenesis from tunable nanolayered implant coatings. Sci Transl Med. 2013 Jun 26;5(191):191ra83. doi: 10.1126/scitranslmed.3005576. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Compston J, Vedi S, Stellon A. Inter-observer and intra-observer variation in bone histomorphometry. Calcif Tissue Int. 1986;38(2):67–70. doi: 10.1007/BF02556831. [DOI] [PubMed] [Google Scholar]
  • 21.Schulte FA, Lambers FM, Kuhn G, Müller R. In vivo micro-computed tomography allows direct three-dimensional quantification of both bone formation and bone resorption parameters using time-lapsed imaging. Bone. 2011;48(3):433–42. doi: 10.1016/j.bone.2010.10.007. [DOI] [PubMed] [Google Scholar]
  • 22.Kettenberger U, Ston J, Thein E, Procter P, Pioletti DP. Does locally delivered Zoledronate influence peri-implant bone formation?–Spatio-temporal monitoring of bone remodeling in vivo. Biomaterials. 2014;35(37):9995–10006. doi: 10.1016/j.biomaterials.2014.09.005. [DOI] [PubMed] [Google Scholar]
  • 23.Waarsing J, Day J, Van der Linden J, Ederveen A, Spanjers C, De Clerck N, et al. Detecting and tracking local changes in the tibiae of individual rats: a novel method to analyse longitudinal in vivo micro-CT data. Bone. 2004;34(1):163–9. doi: 10.1016/j.bone.2003.08.012. [DOI] [PubMed] [Google Scholar]
  • 24.Lan S, Luo S, Huh BK, Chandra A, Altman AR, Qin L, et al. 3D image registration is critical to ensure accurate detection of longitudinal changes in trabecular bone density, microstructure, and stiffness measurements in rat tibiae by in vivo microcomputed tomography (μCT) Bone. 2013;56(1):83–90. doi: 10.1016/j.bone.2013.05.014. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Powell KA, Latson L, Ibiwoye MO, Wolfman A, Grabiner MD, Zborowski M, et al. In vivo longitudinal assessment of bone resorption in a fibular osteotomy model using micro-computed tomography. Iowa Orthop J. 2005;25:123–8. [PMC free article] [PubMed] [Google Scholar]
  • 26.Kettenberger U, Latypova A, Terrier A, Pioletti DP. Time course of bone screw fixation following a local delivery of Zoledronate in a rat femoral model–A micro-finite element analysis. journal of the mechanical behavior of biomedical materials. 2015;45:22–31. doi: 10.1016/j.jmbbm.2015.01.007. [DOI] [PubMed] [Google Scholar]
  • 27.Schulte FA, Ruffoni D, Lambers FM, Christen D, Webster DJ, Kuhn G, et al. Local mechanical stimuli regulate bone formation and resorption in mice at the tissue level. PLoS One. 2013;8(4):e62172. doi: 10.1371/journal.pone.0062172. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Li Z, Kuhn G, von Salis-Soglio M, Cooke SJ, Schirmer M, Müller R, et al. In vivo monitoring of bone architecture and remodeling after implant insertion: The different responses of cortical and trabecular bone. Bone. 2015;81:468–77. doi: 10.1016/j.bone.2015.08.017. [DOI] [PubMed] [Google Scholar]
  • 29.Birkhold AI, Razi H, Duda GN, Weinkamer R, Checa S, Willie BM. Mineralizing surface is the main target of mechanical stimulation independent of age: 3D dynamic in vivo morphometry. Bone. 2014;66:15–25. doi: 10.1016/j.bone.2014.05.013. [DOI] [PubMed] [Google Scholar]
  • 30.Brunski JB, Currey JA, Helms JA, Leucht P, Nanci A, Wazen R. The implant and biological response. Osseointegration and Dental Implants. 2009:205–11. [Google Scholar]
  • 31.Bouxsein ML, Boyd SK, Christiansen BA, Guldberg RE, Jepsen KJ, Müller R. Guidelines for assessment of bone microstructure in rodents using micro–computed tomography. Journal of bone and mineral research. 2010;25(7):1468–86. doi: 10.1002/jbmr.141. [DOI] [PubMed] [Google Scholar]
  • 32.Kohler T, Beyeler M, Webster D, Müller R. Compartmental bone morphometry in the mouse femur: reproducibility and resolution dependence of microtomographic measurements. Calcif Tissue Int. 2005;77(5):281–90. doi: 10.1007/s00223-005-0039-2. [DOI] [PubMed] [Google Scholar]
  • 33.Diggle P, Heagerty P, Liang K, Zeger S. Analysis of longitudinal data. Oxford University Press; 2002. [Google Scholar]
  • 34.Fitzgmaurice G, Laird N, Ware J. Applied longitudinal analysis. Hoboken: 2004. [Google Scholar]
  • 35.Ruffoni D, Wirth AJ, Steiner JA, Parkinson IH, Müller R, van Lenthe GH. The different contributions of cortical and trabecular bone to implant anchorage in a human vertebra. Bone. 2012;50(3):733–8. doi: 10.1016/j.bone.2011.11.027. [DOI] [PubMed] [Google Scholar]
  • 36.Slaets E, Naert I, Carmeliet G, Duyck J. Early cortical bone healing around loaded titanium implants: a histological study in the rabbit. Clin Oral Implants Res. 2009;20(2):126–34. doi: 10.1111/j.1600-0501.2008.01623.x. [DOI] [PubMed] [Google Scholar]
  • 37.De Smet E, Jaecques S, Vandamme K, Vander Sloten J, Naert I. Positive effect of early loading on implant stability in the bi-cortical guinea-pig model. Clin Oral Implants Res. 2005;16(4):402–7. doi: 10.1111/j.1600-0501.2005.01156.x. [DOI] [PubMed] [Google Scholar]
  • 38.Duyck J, Vandamme K, Geris L, Van Oosterwyck H, De Cooman M, Vandersloten J, et al. The influence of micro-motion on the tissue differentiation around immediately loaded cylindrical turned titanium implants. Arch Oral Biol. 2006;51(1):1–9. doi: 10.1016/j.archoralbio.2005.04.003. [DOI] [PubMed] [Google Scholar]
  • 39.Duyck J, Naert I, Rønold HJ, Ellingsen JE, Van Oosterwyck H, Vander Sloten J. The influence of static and dynamic loading on marginal bone reactions around osseointegrated implants: an animal experimental study. Clin Oral Implants Res. 2001;12(3):207–18. doi: 10.1034/j.1600-0501.2001.012003207.x. [DOI] [PubMed] [Google Scholar]
  • 40.Piccinini M, Cugnoni J, Botsis J, Ammann P, Wiskott A. Peri-implant bone adaptations to overloading in rat tibiae: experimental investigations and numerical predictions. Clin Oral Implants Res. 2016 doi: 10.1111/clr.12760. [DOI] [PubMed] [Google Scholar]
  • 41.Pilliar R, Lee J, Maniatopoulos C. Observations on the effect of movement on bone ingrowth into porous-surfaced implants. Clin Orthop. 1986;208:108–13. [PubMed] [Google Scholar]
  • 42.Søballe K, Hansen ES, B-Rasmussen H, Jørgensen PH, Bünger C. Tissue ingrowth into titanium and hydroxyapatite-coated implants during stable and unstable mechanical conditions. Journal of Orthopaedic Research. 1992;10(2):285–99. doi: 10.1002/jor.1100100216. [DOI] [PubMed] [Google Scholar]
  • 43.Szmukler-Moncler S, Salama H, Reingewirtz Y, Dubruille J. Timing of loading and effect of micromotion on bone-dental implant interface: review of experimental literature. J Biomed Mater Res. 1998;43(2):192–203. doi: 10.1002/(sici)1097-4636(199822)43:2<192::aid-jbm14>3.0.co;2-k. [DOI] [PubMed] [Google Scholar]
  • 44.Brunski JB. Avoid pitfalls of overloading and micromotion of intraosseous implants. Dent Implantol Update. 1993 Oct;4(10):77–81. [PubMed] [Google Scholar]
  • 45.Slaets E, Carmeliet G, Naert I, Duyck J. Early trabecular bone healing around titanium implants: a histologic study in rabbits. J Periodontol. 2007;78(3):510–7. doi: 10.1902/jop.2007.060183. [DOI] [PubMed] [Google Scholar]
  • 46.Hollister SJ, Guldberg RE, Kuelske CL, Caldwell NJ, Richards M, Goldstein SA. Relative effects of wound healing and mechanical stimulus on early bone response to porous-coated implants. Journal of orthopaedic research. 1996;14(4):654–62. doi: 10.1002/jor.1100140422. [DOI] [PubMed] [Google Scholar]
  • 47.Walsh W, Bertollo N, Christou C, Schaffner D, Mobbs R. Plasma sprayed titanium coating to PEEK improves the bone-implant interface. The Spine Journal. 2014 doi: 10.1016/j.spinee.2014.12.018. [DOI] [PubMed] [Google Scholar]
  • 48.Klinck RJ, Campbell GM, Boyd SK. Radiation effects on bone architecture in mice and rats resulting from in vivo micro-computed tomography scanning. Med Eng Phys. 2008;30(7):888–95. doi: 10.1016/j.medengphy.2007.11.004. [DOI] [PubMed] [Google Scholar]
  • 49.Dare A, Hachisu R, Yamaguchi A, Yokose S, Yoshiki S, Okano T. Effects of ionizing radiation on proliferation and differentiation of osteoblast-like cells. J Dent Res. 1997 Feb;76(2):658–64. doi: 10.1177/00220345970760020601. [DOI] [PubMed] [Google Scholar]
  • 50.Yang X, Willie BM, Beach JM, Wright TM, van der Meulen, Marjolein CH, Bostrom MP. Trabecular bone adaptation to loading in a rabbit model is not magnitude-dependent. Journal of Orthopaedic Research. 2013;31(6):930–4. doi: 10.1002/jor.22316. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supp FigS1

Figure S-1. Graph demonstrating significant correlation observed between peri-implant BV/TV [%] at the last observation point (day 12) and the peak pull-out force [N] of the unloaded group for the 0–200 μm proximity VOI around the implant.

NIHMS938563-supplement-Supp_FigS1.tif (130.5KB, tif)
Supp FigS2

Figure S-2. Peri-implant bone volume (BV/TV), BFR, and BRR for (a) 0–200 μm and (b) 200–400 μm VOI around the implant plotted as functions of time (days) for unloaded and loaded hind limbs of rats.

NIHMS938563-supplement-Supp_FigS2.tif (259.9KB, tif)
Supp FigS3

Figure S-3. Comparison of peri-implant (a) bone volume (BV/TV), (b) BFR, and (c) BRR for 0–200 μm and 200 – 400 μm radial VOI around the implant plotted as functions of time (days). The unloaded and loaded groups were lumped for the different regions for statistical analysis. Asterisks (*) indicate statistical significance (p < 0.05). Significantly higher BV/TV, BFR and BRR values were observed in most cases for the 200 – 400 μm VOI.

NIHMS938563-supplement-Supp_FigS3.tif (145.5KB, tif)
Supp TableS1
NIHMS938563-supplement-Supp_TableS1.docx (18.4KB, docx)
Supp VideoS1
Download video file (10.2MB, wmv)

RESOURCES