The bone-implant interface--nanoscale analysis of clinically retrieved dental implants

Nanomedicine: Nanotechnology, Biology, and Medicine10 (2014) 1729–1737

Case Study

The bone-implant interface – nanoscale analysis of clinically retrieveddental implants

Furqan Ali Shah, BDS, MSca,b, Bengt Nilson, DDSc, Rickard Brånemark, MD, MSc, PhDd,e,Peter Thomsen, MD, PhDa,b, Anders Palmquist, MSc, PhDa,b,⁎

aDepartment of Biomaterials, Sahlgrenska Academy at University of Gothenburg, Göteborg, SwedenbBIOMATCELL VINN Excellence Center of Biomaterials and Cell Therapy, Göteborg, Sweden

cTandvårdshuset, Vetlanda, SwedendCentre of Orthopaedic Osseointegration, Sahlgrenska University Hospital, Göteborg, Sweden

eDepartment of Orthopaedics, Sahlgrenska Academy at University of Gothenburg, Göteborg, Sweden

Received 27 February 2014; accepted 30 May 2014

nanomedjournal.com

Abstract

Evaluation of the fine structure of the bone-implant interface in humans is a prerequisite for a deepened understanding of structure–function relationships with nano-modified biomaterials. In this study, three clinically stable, yet retrieved, laser-modified dental implantswere evaluated using histological and interface ultrastructural analyses. The cumulative results for all threads containing intact tissue showedremodeled Haversian bone with bone area and bone-implant contact in excess of 85% and 80%, respectively. Collagen fibrils, laid downparallel to the surface oxide layer, were mineralized by plate-like crystallites of stoichiometrically relevant (Ca/P ratios 1.30-1.67) bone-apatite. An overlap of titanium, oxygen, calcium and phosphorus signals indicated the gradual intermixing of bone-apatite and the nano-rough surface oxide. These results suggest that bone bonding to nano-textured titanium implant surfaces is promoted in human jaw-bone afterfunctional loading.

From the Clinical Editor: In this study, newly developed and laser-modified titanium dental implants demonstrate strong evidence forimplant-osseo integration basen on the surface and chemical analysis of three clinically stable dental implants.© 2014 Elsevier Inc. All rights reserved.

Key words: Dental implant; Nano; Osseointegration; Ultrastructure; Electron microscopy; Bone bonding

Bone remodeling is a coupled homeostatic process in whichosteoclasts resorb while osteoblasts produce new bone, to repair,balance and adapt the bone mass for mechanical loading1 and it

Potential conflict of interest: Rickard Brånemark owns stock inBrånemark Integration AB. The company was not involved in the studydesign, data acquisition, interpretation, writing and submission of the article.

Funding information: This study was supported by the SwedishResearch Council (grant K2012-52X-09495-25-3), the BIOMATCELLVINN Excellence Center of Biomaterials and Cell Therapy, the RegionVästra Götaland, an ALF/LUA grant, the IngaBritt and Arne LundbergFoundation, the Dr. Felix Neubergh Foundation and the Hjalmar SvenssonFoundation. FAS holds a PhD position financed by the Swedish GovernmentStrategic Funding of Materials Science Area of Advance, provided toChalmers and Department of Biomaterials, University of Gothenburg. Thegrant providers were not involved in the study design, data acquisition,interpretation, writing and submission of the article.

⁎Corresponding author at: Department of Biomaterials, SahlgrenskaAcademy atUniversity of Gothenburg, Box 412, SE-405 30Göteborg, Sweden.

E-mail address: [email protected] (A. Palmquist).

http://dx.doi.org/10.1016/j.nano.2014.05.0151549-9634/© 2014 Elsevier Inc. All rights reserved.

Please cite this article as: Shah FA, et al, The bone-implant interface – nanos2014;10:1729-1737, http://dx.doi.org/10.1016/j.nano.2014.05.015

is a prerequisite for osseointegrated metal implants to supportfunctional loading in the long term. Bone consists ofhierarchically organized2 organic type-I collagen matrix andinorganic poorly crystalline carbonated hydroxyapatite, orbioapatite. While organization at larger (milli-) scales isinfluenced by the mechanical feedback loop, micro- to nanoscaleorganization is self-assembly driven.3

Nanoscale structure dictates the mechanical properties ofbone.4 Collagen fibrils are derived from tropocollagen moleculesextruded by osteoblasts. The axial stagger of adjacent collagenmolecules within the fibril produces a characteristic 67 nm cross-striated pattern consisting of 40 nm-long gap zones and 27 nm-long overlap zones. In gap zones, mineralization of this collagenmatrix is initiated5 to form typically plate-like intrafibrillarcrystallites6 with crystallographic c-axes approximately parallelto fibril long axes. Extrafibrillar mineral coats collagen fibrils7

and makes up a major fraction of total bone-apatite.Biomechanical fixation and enhanced bone formation around

osseointegrated implants have been made possible by micro-

cale analysis of clinically retrieved dental implants. Nanomedicine: NBM

Figure 1. Surface morphology of the laser-modified implant. (A) SEM image of a complete thread with machined and selectively laser-ablated regions.(B) Micro-topography consisting of 1-10 μm diameter globules of resolidified metal in the laser-ablated region. (C) Micro-CT reconstruction of the implantsurface from a retrieved implant showing surface structure and laser-ablation tracks similar to those seen in (A). (D) SEM image showing nano-texturesuperimposed on surface micro-topography.

1730 F.A. Shah et al / Nanomedicine: Nanotechnology, Biology, and Medicine 10 (2014) 1729–1737

topography modification. Today, nano-technology is emergingin the field of implantology; and nano-structured and hierarchicalsurface treatments are often considered,8 mimicking the differentlength scales of bone constituents and structures. Controlledmicro- and nanoscale surface modification of titanium surfacesmay be achieved by laser ablation resulting in improved earlybone formation and biomechanical fixation in experimentalstudies.9 And indeed, nanoscale surface features modulatecellular behavior and differentiation.10 In an experimentalmodel, high-resolution transmission electron microscopy andelectron tomography studies have described a functionallygraded interface characterized by the gradual intermixing ofbone with the surface oxide layer in relation to nano-structuredlaser-modified implants.11 However, due to the limited numberof published ultrastructural analyses of clinically retrievedimplants, our understanding of the bone-bonding process inhuman bone is limited. Clinically used methods for evaluation ofbone healing around implants such as radiography and resonancefrequency analysis are only indicators of oral implant stability.These techniques provide little, if any, information about themicro- and nanoscale structure of bone formed adjacent toosseointegrated implant surfaces. Therefore, in order to under-stand the process of osseointegration in human subjects, it isessential to study well-integrated implants with surroundingbone tissue.

Long-term osseointegration studies in humans are commonlyrestricted to relatively few cases at a time for reasons includingdifficulty in controlling healing durations between individualpatients prior to retrieval, age and gender-associated differences,underlying systemic conditions and the use of medication,complicating factors such as history of smoking, and otherbiological variations.

Here, we present data on the fine structure and compositionof the bone-biomaterial interface from three nanotextured

dental implants, retrieved from the maxilla of a patient withbruxism. All implants were well integrated with human bone,undergone functional loading, and explanted due to mechanicalfailure of the material attributable to challenging functioningconditions, rather than biological failure (e.g., implant loosen-ing due to lack of osseointegration or periimplantitis). Casessuch as these are extremely few and present a great opportunityto study osseointegration.

Methods

Partly laser-modified dental implants (BioHelix™, Bråne-mark Integration AB, Mölndal, Sweden) were used. In short,screw-shaped commercially pure titanium implants prepared byturning (machining) were ablated in the thread valleys, using aQ-switched Nd:YAG laser (Rofin-Sinar Technologies Inc.,Plymouth, USA) operating in multimode, at an infraredwavelength of 1064 nm and spot size 100 μm, in ambient air.

Surface analysis

The implant surface morphology was analyzed using a LeoUltra 55 (Leo Electron Microscopy Ltd, UK) scanning electronmicroscope (SEM) operating at 5 kV accelerating voltage. Anew sterile-packaged implant was kindly provided by themanufacturer for this analysis.

Case history

The patient was a 66-year-old Caucasian female, without anyknown systemic complicating conditions, with a dental history ofsevere bruxism and missing the right maxillary canine and bothpremolars. Three fixtures (diameter: 3.5 mm; length: 13 mm)were placed using a one-step surgical technique and good

Figure 2. Light optical micrographs of the ground sections. (A)Overview of one of the implants, showing a large amount of bone tissue surrounding the implant,and trephine-associated damage to some threads. (B)Many threads are almost completely filled with mature bone tissue. (C) Osteocytes are lined up parallel tothe lamellar direction, thus confirming osteonal bone. (D) And (E) osteocytes and canalicular networks close to the implant surface.

1731F.A. Shah et al / Nanomedicine: Nanotechnology, Biology, and Medicine 10 (2014) 1729–1737

primary stability was achieved. An implant-supported definitiveprosthesis (fixed bridge) was installed with 0.32 Nm torque after1.5 months of healing. Fractures were detected in all threefixtures on routine radiographic evaluation, after 47 months inservice. All implants were judged as clinically stable. Theimplants were removed by trephining. The specimens wereimmersed in formalin and sent to Biobank 513, University ofGothenburg. Clinical radiographs and panoramic views areprovided as supplementary data (Figure S1).

Sample preparation

Implants with surrounding tissue were processed forhistology. Following fixation, followed by dehydration in agraded series of ethanol and embedding in plastic resin (LRWhite, London Resin Company, UK), a central ground sectionwas prepared by sawing along the long axis of the implant. Thefinal ground sections (15-20 μm in thickness) were stained withtoluidine blue for histology and histomorphometry.

The remaining parts of the implant–tissue blocks were usedfor micro-CT (Skyscan 1172, Bruker MicroCT, Belgium) andelectron microscopy. The blocks were polished and sputter-coated with palladium for backscattered electron scanningelectron microscopy (Ultra 55 FEG SEM, Leo ElectronMicroscopy Ltd., UK). Further, electron transparent (100 nmthick) samples of the intact bone-implant interface were preparedusing an in-situ lift-out technique in a focused ion beam system(Strata DB235 FIB/SEM; FEI Company, The Netherlands) for

transmission electron microscopy (Tecnai G2 20, FEI Company,The Netherlands).

Analysis

HistologyHistological examination was performed using light optical

microscopy (Nikon Eclipse E600). The tissue morphologyaround the implant was evaluated by determining inflamma-tion, the presence of osteoblasts, osteocytes and osteoclasts,bone remodeling and the osteocytic network close to theimplant surface.

HistomorphometryHistomorphometry was performed by measuring the amount

of bone in direct contact with the implant and the bone areaconfined within the threads. This was done using lightmicroscopy and backscattered electron scanning electronmicroscopy (BSE-SEM), where quantification was performedby regular manual tracing and semi-automated segmentation andedge detection using Adobe Photoshop CS5.1 and ImageJ(imagej.nih.gov/ij). All threads with intact tissue, along the entireimplant were considered for quantification.

Micro-CTThe micro-CT system was set at a resolution of 2.9 μm, with

a scan rotation step of 0.4 degrees for full 360-degree acquisitionand an acceleration voltage of 85 kV. The rotation was along thelong axis of the implant. Reconstruction and visualization were

Table 1Quantitative histomorphometry of the retrieved implants (mean ± standard deviations are presented; n = number of intact threads evaluated for each implant).

Method Implant 1 Implant 2

n BIC (%) Bone area (%) n BIC (%) Bone area (%)

LM manual 17 82.4 ± 20.3 83.7 ± 20.7 11 83.3 ± 16.1 93.1 ± 6.1LM semi automated 17 83.4 ± 19.4 84 ± 21 11 87.3 ± 14.4 94.6 ± 5.5SEM semi automated 14 77.3 ± 23 79.9 ± 16.2 6 93.7 ± 6.1 84 ± 12Micro-CT 18† 88.98 ± 5.5

† The number of vertical 2D slices used for quantification. Each slice comprised of 6 contiguous intact threads.

1732 F.A. Shah et al / Nanomedicine: Nanotechnology, Biology, and Medicine 10 (2014) 1729–1737

performed using associated PC programs (NRecon, Imageviewer, CTAn, CTVox, CTVol; Bruker MicroCT, Belgium).

Ultrastructural interface analysisTEM and scanning TEM (STEM) were performed in both

bright field and high angle annular dark field (HAADF) mode inorder to visualize the bone morphology, surface oxide and bulkmetal. Further, site-specific chemical analysis was performedwith a nanoprobe in STEM mode using energy dispersive X-rayspectroscopy (EDX), where both line profiles across the interfaceand mapping of the interface zone were performed. Selected areaelectron diffraction (SAED) was performed to investigate thecrystalline orientation of the bioapatite in bone.

Results

Surface analysis

The SEM analysis of native implant surfaces showed the dualsurface structure created during the laser modification. The upperpart of the flank and the top of the thread retained the as-machined surface. The lower part of the thread valley showedlaser ablation tracks perpendicular to the machining direction(Figure 1). At higher magnification, a microstructure is visible,created by globules of resolidified material 1-10 μm in size. Athigh resolution, distinct nanoscale irregularities are present, witha size range up to 300 nm, forming fairly high roughness on thislength scale. In order to confirm the surface morphology of theactual clinically retrieved implants with the implant batch usedfor surface characterization, a segmentation of the implant part ofthe micro-CT set was taken out for visualization. It showed largesimilarities to the low-resolution SEM.

Histological evaluation

Light microscopy of the ground sections showed damage tothe implants originating from the trephine retrieval process,where some threads on two implants were partially destroyed.The third implant was almost completely destroyed, leaving onlya few intact threads.

The general histological picture was dominated by a largeamount of mature bone tissue around and in direct contact withthe implant surface (Figure 2). Many of the threads were almostentirely filled with bone tissue showing lamellar bone organizedinto Haversian systems. The orientation of concentric lamellaeclose to the implant surface was principally parallel, indicatingthat the remodeling processes occurred from the implant surface

and outwards. Multiple osteocytes were observed close to theimplant surface, oriented co-parallel to the lamellae, in additionto canaliculi networks predominantly perpendicular to theimplant surface. Larger Haversian canals were also observed,with traces of on-going remodeling, where some areas displayeda jagged surface, while other areas displayed lamellae ofincreasing blue color, indicating bone resorption and boneformation respectively. No morphological signs of adverseevents, such as the presence of inflammatory cell infiltrates,were detected.

Histomorphometry of ground sections demonstrated a largeproportion of mineralized bone within the threads (bone areaN80%) and in direct contact (bone-implant contact N80%) withthe implant surface (Table 1). Regrettably, the trephine hadmarkedly damaged the third implant during the retrievalprocedure and was not possible to study histomorphometrically.

Similar amounts of bone (N80%) occupying the threads werealso observed using morphometry applied to another part of theimplant tissue block using BSE-SEM (Table 1). In addition, asanalyzed by micro-CT, similarly high degrees of contact andamounts of bone were demonstrated in different sections of theimplant-tissue block. However, a variation could be seen,depending on the level of the section due to porosities in thebone (Figure 3).

Ultrastructural interface analysis

Mineralized bone tissue in direct contact with the laser-modified surface was demonstrated by TEM (Figure 4). Close tothe implant surface, a well-defined orientation of bone tissue wasobserved. A characteristic 67 nm cross-striated pattern ofcollagen was observed approximately perpendicular to theimplant surface, indicating an orientation of collagen fibrilsparallel to the implant surface. This observation was consistent atdistances of up to 0.5-1 μm from the surface oxide (TiO2). Atgreater distances, this parallel orientation of collagen appeared tobe obscured. However, collagen fibrils appeared to be organizedinto 1-3 μm diameter concentric bundles, that have previouslybeen described as rod-like structures.12 Collagen fibrils reachedup to 100 nm from the implant surface, closely following thesurface micro-topography. In one of the TEM samples, circularfeatures in the 200-400 nm size range were found 1-2 μm fromthe surface, suggestive of transectioned, osteocyte canaliculi.

Chemical analysis by STEM-EDX elemental mapping of theinterface zone showed an overlap of Ti, O, Ca, and P signals(Figure 5). The overlap of Ti and O signals indicates the surfaceTiO2 layer, while Ti, Ca and P overlap shows the gradual

Figure 3. Micro-CT analysis showing a large amount of bone ingrowth into the implant threads. (A) Cross-sectional view at the vertical level indicated as ahorizontal line in (B). (B) Longitudinal sections from locations designated i-v in (A) showing a variation in the amounts of bone in different areas. (C) Bonearea varying between 74% and 95% around the implant.

1733F.A. Shah et al / Nanomedicine: Nanotechnology, Biology, and Medicine 10 (2014) 1729–1737

intermixing of bone mineral with the surface oxide. Further, line-scans revealed stoichiometrically relevant atomic ratios of Caand P (1.30 b Ca/P b 1.67) for bone-apatite (bioapatite), close tothe implant surface and up to 800 nm away from the surface.

In some areas, circular holes 50-70 nm in size were observed.On higher magnification of these areas, an open, lacy structurecomprised of electron-dense mineral structures, with lengths of59.01 ± 9.11 nm (n = 15) and thicknesses of 5.05 ± 0.61 nm(n = 30), could be seen interspersed between and positioned atminor angles to electron-lucent holes (Figure 6). STEM-EDXspot-scans showed C/Ca of 1.33 ± 0.50 over the electron-denseareas and 3.33 ± 0.95 over the electron-lucent holes, confirmingthe latter as collagen. In addition, Ca/P = 1.33 ± 0.15 wasobserved for the mineral structures, confirming the presence ofbone-apatite. SAED showed a preferential alignment of apatitecrystallographic c-axis with respect to the orientation of collagenfibrils (Figure 6, A and C). The (002) and (004) reflections

formed two arcs subtending angles of ~36° centered on the fibrildirection. Almost no (002) and (004) reflections were observedfor transversally sectioned collagen fibrils, indicating that theseplanes were oriented normal to the plane of the image.

Discussion

In spite of several studies in vitro and in animals, theultrastructure of the contact zone between a nano-texturedimplant material and the biological environment in humans islargely unknown. Histological and ultrastructural analysis ofretrieved, functionally loaded human implants is thereforecritical. In the present case study, three dental implants wereretrieved from one patient after 47 months due to the mechanicalfailure of the implant material, which permitted in-depth analysis

Figure 4. (A) TEM sample prepared using the in-situ lift-out technique, showing bone tissue (black asterisk) in close contact with the metal implant (whiteasterisk). (B) STEM image of the sample seen in (A). (C) Bone tissue in contact with the implant surface. Collagen fibrils immediately adjacent to the implantsurface appear to follow the microscale contour. Several canaliculi (arrowheads) are detected. (D) Higher magnification of a canaliculus (asterisk) ≈5 μm fromthe surface oxide. Groups of collagen fibrils oriented parallel to the image plane, and are laid down with respect to the canaliculus. Some are also seen runninginto the plane of the image (arrowhead).

1734 F.A. Shah et al / Nanomedicine: Nanotechnology, Biology, and Medicine 10 (2014) 1729–1737

of the bone tissue interfacing the implant surface, at both micro-and nano-length scales.

By combining three different morphological analytical tools(histology, BSE-SEM and micro-CT) and evaluating differentparts of the implant-tissue block, strong evidencewas provided thatthe analyzed nano-textured implants had osseointegrated. Thepresent morphometric values are in agreement with resultsobtained from retrieved, clinically stable, machined dentalimplants (of a similar design to the implants in the present study)where values between 56-85% for bone-implant contact (BIC), and79-95% for bone area were shown.13 The authors evaluatedseven implants retrieved from four patients aged 52-70 years, after1-16 years of function, but no discrimination was made betweenthe different bone-types (mandible and maxilla). The highest BICand bone area were reported for one implant in the mandibleretrieved after 16 years, while the implant retrieved from themaxilla showed the least BIC.

Important general conclusions in relation to these morpho-metric observations are that (i) nano-textures superimposed on amicron-scale topography promote osseointegration in humans atleast to the same level as the micron-scale textures of clinically

used, similarly designed oral implants and (ii) irrespective ofimplant surface texture, the amount of mineralized bone incontact with and surrounding these surfaces varies considerably,i.e. the bone-implant interface is heterogeneous and adapts tolocal environmental biological and mechanical conditions.

The morphology of the tissue plays a vital role inaccommodating load transfer from the implant to the surround-ing bone tissue, where osteocytes are the mechanosensingcomponent. Areas of active remodeling at a distance and close tothe implant surface were commonly detected. Further, numerousnetworks of osteocyte canaliculi were observed adjacent to theimplant surface. In agreement with previous observations inbone of different species (without implants),3 osteocytes werealigned parallel to the lamellar direction and the canalicularprojections directed approximately perpendicularly. Further, tothe authors' knowledge, the present observations show, for thefirst time, that (i) bone lamellae are arranged concentricallywithin the implant thread, (ii) lamellae close to the implantsurface follow the contour and curvature of the thread and (iii),while those away from the implant surface were still co-paralleland concentric, they appeared more flattened, as if following a

Figure 5. (A) STEM image showing a collagen-banding pattern adjacent to the implant surface. The black square indicates area for chemical analysis.(B) Elemental maps showing the distribution of titanium, oxygen, calcium and phosphorus as outlined in (A). (C) STEM-EDX line-scan across the bone-implant interface. (D) Intermixing between the bone and the oxide (TiO2) channels, which extend perpendicularly outwards from the bulk metal.

1735F.A. Shah et al / Nanomedicine: Nanotechnology, Biology, and Medicine 10 (2014) 1729–1737

larger curvature. These observations indicate that remodeling,i.e. the conversion of initial woven bone toward a cortical structure,started at the implant surface and was directed outwards.

Another interesting ultrastructural finding was the observationof collagen fibrils running parallel to the implant surface. Inaddition, osteocyte canaliculi were observed, perpendicular to theplane of the section, reaching 1 μm from the implant surface.These ultrastructural findings are largely in agreement withprevious observations from early14 and late healing in a rabbitin-vivomodel15 and unloaded, retrieved clinical dental implants.16

Previously, stereo high-voltage TEM experiments have shown thatthe osteocyte process reaches the implant surface and bends in aparallel fashion following the implant surface,17 similar to whathas been observed for the interface with an adjacent osteon.3

At higher magnification, entangled contact was observedbetween the tissue and the surface oxide layer. Accommodatingthe numerous ionic substitutions within the apatite lattice, forinstance, CO3

2− for hydroxyl or phosphate groups, F− and Cl− forOH−, Mg2+ and Na+ for Ca2+ ions, and HPO4

2− (acid phosphate)for PO4

3− (orthophosphate), atomic ratios of 1.30 b Ca/P b 1.67generally suggest the presence of bone-apatite. Interestingly, thefinding of a co-existence of Ca, P, Ti and O, within the surfaceirregularities, suggests intimate contact between bone-apatite and

the surface oxide on the nanoscale, indicating a bonding betweenthe nano-textured surface and human bone. Further analyses arerequired in order to understand the true molecular nature of thisbone-bonding process, with regard to both the cellular activity19

and the true atomic scale composition at the implant interface usinge.g. atom probe tomography.20

In this study, it is important to remember that only a fewspecimens were retrieved from one patient and only a handful ofultrastructural observations were possible due to the nature of thesample preparation techniques that were used. Nevertheless, thetopographical complexity and rapidly changing contour of therelevant implant surface enabled a number of key observationswith regard to tissue organization in relation to the micro- andnano-surface structure. The nanoscale features of the surface oxide(TiO2) layer appear as narrow channels extending perpendicularlyoutwards from bulk titanium. However, the fibrillar component ofthe extracellular matrix closest to the surface oxide was neverfound to approach the surface perpendicularly, i.e. parallel anddirected toward the oxide “channels”. Likewise, the fibrillarcomponent was consistently observed to follow the micro- andsubmicron scale surface contour, thus suggesting a topography-mediated organization of the extracellular matrix. Several 1-3 μmdiameter cross-sectioned bundles of collagen fibrils were also seen,

Figure 6. TEM micrographs of the ultrastructure of bone around implants: (A) Collagen fibrils laid down parallel to the implant surface running along imageplane. (B)Open-lacy structure of transversely sectioned collagen fibrils surrounded by mineral platelets. (C)And (D) SAED patterns obtained from area shownin (A) as a white ring and (B) respectively. (E) TEM and STEM (inset) image of a region similar to that shown in (B), where STEM-EDX analysis showed highand low C/Ca ratios over collagen fibrils (black asterisk) and mineral platelets (white asterisk), respectively.

1736 F.A. Shah et al / Nanomedicine: Nanotechnology, Biology, and Medicine 10 (2014) 1729–1737

as close as 300-400 nm from the surface. The appearance of thesebundles can be described as round, low z-contrast (black) featuresapproximately 50-65 nm in diameter, separated from one anotherby plate-like structures of high z-contrast (white), approximately60-70 nm long and 5-7 nm wide. These measurements are inagreement with the simplified model of bone mineral and collagenorganization, based on measurements of longitudinal and cross-sections of cortical bone, described by McNally et al.18

The present study shows the formation of mature, mineralizedbone in the interface zone and within the threads of laser-modified human dental implants. The first bone formed at asurgical wound-healing site is generally believed to be of woventype. As it undergoes gradual remodeling, a well-definedhierarchical structure is achieved. For biomechanically loadedimplants, this remodeling process is a function of the quality ofinitial healing around the implant and the mechanical stimulitransferred to the surrounding bone tissue.

Conclusions

The present study reveals the structural connection betweenfunctionally loaded bone and laser-modified implants at both the

micro- and nano-length scales. A large amount of remodeled,osteonal bone was observed within the implant threads and aroundthe implant. At the ultrastructural level, a functionally gradedinterface was demonstrated characterized by the gradual intermix-ing of bone with the nanoscale surface oxide layer, extending tohighly ordered mineralized collagen fibrils contouring the micro-topography. Although the present results have to be verified inadditional human studies, with a larger number of implants, thepresent observations indicate that functionally loaded, nano-textured titanium implants promote bone formation and remodel-ing, resulting in bone bonding to the nano-textured surface.

Acknowledgments

The authors would like to thank Lena Emanuelsson andBirgitta Norlindh for their excellent technical assistance.

Appendix A. Supplementary data

Supplementary data to this article can be found online athttp://dx.doi.org/10.1016/j.nano.2014.05.015.

1737F.A. Shah et al / Nanomedicine: Nanotechnology, Biology, and Medicine 10 (2014) 1729–1737

References

1. Huiskes R, Ruimerman R, van Lenthe GH, Janssen JD. Effects ofmechanical forces on maintenance and adaptation of form in trabecularbone. Nature 2000;405:704-6.

2. Fratzl P, Weinkamer R. Nature's hierarchical materials. Prog Mater Sci2007;52:1263-334.

3. Kerschnitzki M, Wagermaier W, Roschger P, Seto J, Shahar R, DudaGN, et al. The organization of the osteocyte network mirrors theextracellular matrix orientation in bone. J Struct Biol 2011;173:303-11.

4. Taton TA. Nanotechnology. Boning up on biology. Nature2001;412:491-2.

5. Wang Y, Azais T, Robin M, Vallee A, Catania C, Legriel P, et al. Thepredominant role of collagen in the nucleation, growth, structure andorientation of bone apatite. Nat Mater 2012;11:724-33.

6. Landis WJ, Hodgens KJ, Song MJ, Arena J, Kiyonaga S, Marko M, et al.Mineralization of collagen may occur on fibril surfaces: evidence fromconventional and high-voltage electron microscopy and three-dimensionalimaging. J Struct Biol 1996;117:24-35.

7. Alexander B, Daulton TL, Genin GM, Lipner J, Pasteris JD,Wopenka B,et al. The nanometre-scale physiology of bone: steric modelling andscanning transmission electron microscopy of collagen–mineral struc-ture. J R Soc Interface 2012;9:1774-86.

8. Davies JE, Mendes VC, Ko JC, Ajami E. Topographic scale-rangesynergy at the functional bone/implant interface. Biomaterials2014;35:25-35.

9. Brånemark R, Emanuelsson L, Palmquist A, Thomsen P. Boneresponse to laser-induced micro- and nano-size titanium surface features.Nanomedicine 2011;7:220-7.

10. Dalby MJ, Gadegaard N, Tare R, Andar A, Riehle MO, Herzyk P, et al.The control of human mesenchymal cell differentiation using nanoscalesymmetry and disorder. Nat Mater 2007;6:997-1003.

11. Grandfield K, Gustafsson S, Palmquist A. Where bone meetsimplant: the characterization of nano-osseointegration. Nanoscale2013;5:4302-8.

12. Reznikov N, Shahar R, Weiner S. Three-dimensional structure of humanlamellar bone: the presence of two different materials and new insightsinto the hierarchical organization. Bone 2014;59:93-104.

13. Sennerby L, Ericson LE, Thomsen P, Lekholm U, Astrand P. Structureof the bone-titanium interface in retrieved clinical oral implants.Clin Oral Implants Res 1991;2:103-11.

14. Palmquist A, Emanuelsson L, Sjövall P. Chemical and structural analysisof the bone-implant interface by TOF-SIMS, SEM, FIB and TEM:experimental study in animal. Appl Surf Sci 2012;258:6485-94.

15. Palmquist A, Emanuelsson L, Brånemark R, Thomsen P. Biomechanical,histological and ultrastructural analyses of laser micro- and nano-structuredtitanium implant after 6 months in rabbit. J Biomed Mater Res B ApplBiomater 2011;97B:289-98.

16. Palmquist A, Grandfield K, Norlindh B, Mattsson T, Branemark R,Thomsen P. Bone-titanium oxide interface in humans revealed bytransmission electron microscopy and electron tomography. J R SocInterface 2012;9:396-400.

17. Steflik DE, Corpe RS, Lake FT, Young TR, Sisk AL, Parr GR, et al.Ultrastructural analyses of the attachment (bonding) zone between boneand implanted biomaterials. J Biomed Mater Res 1998;39:611-20.

18. McNally EA, Schwarcz HP, Botton GA, Arsenault AL. A model for theultrastructure of bone based on electron microscopy of ion-milledsections. PLoS One 2012;7:e29258.

19. Omar OM, Lenneras ME, Suska F, Emanuelsson L, Hall JM, PalmquistA, et al. The correlation between gene expression of proinflammatorymarkers and bone formation during osseointegration with titaniumimplants. Biomaterials 2011;32:374-86.

20. Gordon LM, Joester D. Nanoscale chemical tomography of buriedorganic-inorganic interfaces in the chiton tooth. Nature 2011;469:194-7.