In aiao study on hydroxyap
`trabecular architecture for b
`
`atite scaffolds with
`one repair
`
`Mark R. Appleford,l Sunho Oh,1 Namsik Oh,t Joo L. Ongl
`lDepartrnent of Biamedicøl Ertgineering, Llniuersity of Texns øt Snn Antonio, Søt Antonio, Texøs 78249
`zDepartment of Dentistry, Inha Llniaersity Hospital, Incheon, Sotúh Korea
`
`Received 21 March 2007; revised 22 December 2007; accepted 25 Febmarr' 2008
`Published online 13 May 2008 in Wiley InterScience (wr,r,w.interscience.wiJey.com). DOI: 1,0.1,002/ibm.a.32049
`
`Abstract: The objective of this research was to investigate
`the bone formation and ar-rgio-conductive potential of hy-
`droxyapatite (HA) scaffolds closely matched to trabecular
`bone in a canine segmental defect after 3 and 12 weeks post
`implantation. Flistomorphometric comparisons were made
`between naturally forming trabecular bone (conh'ol) and
`defects implanted with scaffolds fabricated with micro-size
`(M-HA) and nano-size HA (N-HA) ceramic surfaces. Scaf-
`fold architecture was similar to trabecular bone formed in
`control defects at 3 weeks. No significant differences were
`identified between the two HA scaffolds; however, signifi-
`cant bone in-growth was obsen'ed by 12 rveeks u'ith 43.9 +
`4.1o/o and 50.4 + 8.8% of the cross-sectional area filled with
`mineralized bone in M-HA and N-HA scaffolds, respec-
`tively. Partially organized, lamellar collagen fibrils 'r,r'ere
`
`identified by birefringence under cross-polarized light at
`both 3 and 12 weeks post implantation. Substantial blood
`vessel infiltration was identified in the scaffolds and corn-
`pared with the distribution and diameter of vessels in the
`surrounding cortical bone. Vessels were less numerous but
`significantly larger than native cortical Haversian and Volk-
`mann canals reflecting the scaffold architecture where open
`spaces allowed interconnected channels of bone to form.
`This study demonstrated the potential of trabecular bone
`modeled, highly porous and interconnected, HA scaffolds
`for regenerative orthopedics. O 2008 Wiley Periodicals, Inc.
`I Biomed Mater Res 89A:1019-1027,2009
`
`Key words: hydroxyapatite; scaffolds; bone regeneratìon;
`vascularization; animal study
`
`INTRODUCTION
`
`Bone regeneration supported by artificial scaffold-
`ing requires a substantial healing period before fr,mc-
`tional restoration of natural tissue is realized. Porous
`scaffold designs take advantage of natural bone in-
`growth to accelerate the osteoconduction and inte-
`gration between implant and bone. Characteristics
`such as material composition, surface features, pore
`morphology, and overall architecture are a few of
`the parameters that determine implant success.
`With careful control of these factors, scaffolds can L're
`biocompatible, osteo- and angio-conductir,,e while
`encouraging a natural bone arrangement with haver-
`sianlike features durir-rg bone regeneration.
`Among the many biomaterials used for manufac-
`turing scaffolds, the use of hydroxyapatite (HA) has
`demonstrated excellent performance for bone recon-
`struction.l t HA is a well established material for
`
`Correspondence for J. L. Ong; e-mail: Anson.Ong@utsa.edu
`Contract grant sponsor: National lnstifutes of Health;
`contract grant number: RO14R46581
`
`@ 2008 Wiley Periodic¿ls, lrlc.
`
`bone repair and very comparable to natural apatite
`providing a strong bio-mechanical interlock with
`host tissue.e'tu The composition of HA exhibits a va-
`riety of useful properties such as closely matched
`composition to bone, negligible imnunoreactivity,
`osteoconductivity, and the availability of local cal-
`cium and phosphorous for surrounding cells. Addi-
`tionally, recent research on HA has also shown that
`nano-based surfaces offer impror.ed characteristics
`for bone formation.ll
`In addition to n-raterial composition, the impor-
`tance of pore morphology ¿n¿ architecture has been
`the focus of scaffold designs to irnprove clinical
`results.l2-16 Scaffolds have been manuiaclrred from
`HA using techniques such as porogen leaching, tex-
`tile, solid free form fabrication, and template coating.
`Each of these methods offers advantages and disad-
`vantages for scaffold design with ongoing research
`particularly airned at the identification of a mini-
`mum pore size necessary for tissue infiltration.TT 7e
`The macroscopic arrangement of pore designs and
`the importance of surface area and shape of ceramic
`materials have also been implicated in
`activities such as differentiation into the
`lineage.2o
`
`Baxter Healthcare Corp., et al. v. Millenium Biologix, IPR2013-00590, Exhibit 1174, p. 1
`
`

`
`r 020
`
`APPLEFORD ET AL
`
`Open, interconnected porosity is essential for tissue
`infiltration as it guides cell migration, proliferation
`and bone cell differentiation to mechanically connect
`implant -with bcne. The long-term success for clinically
`applied scaffolds may be assisted by the overall archi-
`tecture of open channel designs since functional blood
`supply provides a foundation for tissue growth.21
`Incomplete pore corulections can obstruct vascular
`infiltration and ultimately constrain tissue formation
`in clinically applied biomaterials. In addition to tissue
`survival, blood vessels have a substantial role in coor-
`dinating the biochemical activitl' that guide bone cell
`behavior including remodeling.22 Successful scaffold
`designs must balance these factors to support vessel
`formation for functional bone tissue formation.
`In this study, it was hypothesized that scaffolds
`coated with nano-size HA (N-HA) would induce a
`different rate of bone regeneration when compared
`to scaffolds coated with of micro-size HA (M-HA).
`Bone forrnation and angio-concluctive potential of a
`scaffold closely matched to trabecular bone was
`investigated in a segmental defect model of a canine
`after 3 and '12 weeks post implantation. Histological
`sections were compared between the naturally form-
`ing trabecular bone (control) and defects implanted
`with scaffolds. Additionally, collagen arrangement
`was evaluated using cross-polarized microscopy.
`
`MATERTALS AND METHODS
`
`Scaffold preparation
`
`HA scaffolds were produced using a template-coating
`technique. Polvurethane sponges (EN Murral', Denver,
`CO) u,ere coated n,ith FIA por,r'der (TAL Materials, Arur
`Arbor, MI), irr a distilled u'ater-based slurry. Binders used
`r,r'ith the slurry to irnprove sintering and to stabilize the
`scaffold structure included 3'l. high rnolecular weight pol-
`y'r,in1'l alcohol, l"/" v/t, carboxymethylcellulose,
`lo/o v/tr
`ammonium polyacrylate dispersant, and 3o/o v/v N,N-
`dimethylformamide drying agent. Coated sponges were
`vacuum-dried overnight before sintering to 1230"C for 3 h
`in a ìrigh-tempelatllre furnace (Thermolpre, Dubr,rque,
`Iowa). Scaffolds r,r'ere trvice coated with HA slurry and
`resintered. Twice-coated scaffolds demonstrated nticro sur-
`face features and were designated M-HA. A sol-gel coat-
`ing of nano-size HA was prepared from calcium nitrate in
`methanol using a previously reported p.ocess.t3 A stoìchi-
`ornetric Ca/P ratio of I.67 was obtained ar-rd solution a¡¡ed
`for 7 days befole inrrnersion coating onto scaffolds and
`sintered to 600'C for 1 h. Sol-gel coated scaffolds demon-
`strated nano surface features and were designated N-HA.
`Composition purity rvas validated using X-ray diffractior-r
`analysis (D8 Aclvance, Bruker Axs Inc., Madison, WI).
`Final scafiold dimensions were ø 5 mm and length of 5
`mm as shown in Figure 1. Samples were sterilized by eth-
`¡4ene oxicle gas before implantation.
`
`Iotnnnl of Biontedical Matcrials Rcsaarch Pnrt A
`
`Figure 1. C¡'linclrical, ilterconnected 5 X 5 mm2 FIA scaf-
`folds for defect repair in the dog mandible, scale bar mea-
`sures 20 rnm.
`
`Scaffold characterization
`
`Scaffolds rvere imaged by scanning electron microscopy
`(SEM) to observe macro and micro features. Histomorph-
`ometry lt'as performed on cross-sectional slìdes prepared
`from representative samples Ernbedding was performed in
`one-component photo-curing resin (Exakt 7200 \¡LC, Okla-
`homa City, OK), and thin sectioned using a precision
`miclosar.t' (Buelrler, Lake Blufi IL). Sections \ /ere progres-
`sively polished to 1200 grit paper and adhered to glass
`slicles using a methvl methacrl4ate resin (Surgipath Medi-
`cal Ind , Richmond, lL). Sections r.r'ere imaged at B0 and
`200X magrification n,ith a digital caûrera (Qhnaging, Bur-
`naby, Canada) on a Nikon TE300 microscope (Nikon, Mel-
`r.ille, NY). An analysis of the scaffold lvas perfolmed using
`bone histomorphometry soflware (Bioquant Osteo, Nash-
`ville, TN). Two longitudinal cross-sections were prepared
`from each scaffold. Palametels r.r'ere defined from tladi-
`tional bone histomorphometry guidelines and the following
`characteristics were recorded: percentage of bone volume
`[BV/TV - bone volume/totaÌ r,olume x I00'/.], ratio of
`bone surface perimeter to total volume IBS/TV - bone sur-
`face length/total r'olume], ratio of bone surface perimeter
`to bone volume IBS/BV : borre surface length/bone vol-
`umel, trabecular thìckness [Tb.th. : (4/1..199) x (BSIBV)],
`trabecular number [Tb.n. : (a/ ù x (BVITV)) 0 s/GU.tt-r.)]
`and trabecular separation [Tb sp. = (a/¡r) x (TVIBV) - 1)
`X (Tb.th.)]. The folrnulae were derived rvith lespect to tra-
`becular bone tissue including both rod- and platelike ge-
`ornetry.24 In the preceding calculations, bone variables are
`leplaced rn'ith scaffolcl r'ariables for the reported scaffold
`characterization. The use of these measures allows for com-
`parisons between known values of primarily rod-like tra-
`beculal structures from other anatomical locations.25
`
`Surgical ptocedure
`
`Ten, 2-year-old adult male foxhound dogs, weighing
`between 20 Io 25 kg were used for' this studl'. The 10 dogs
`r.vere cared for in compliance with DOD proglams and
`NIH publication #86-23, Guide for the Care and Use of
`Laboratory Animals. Prior to expelirnentation, the protocol
`
`Baxter Healthcare Corp., et al. v. Millenium Biologix, IPR2013-00590, Exhibit 1174, p. 2
`
`

`
`HYDROXYAPATITE SCAFFOLDS WITH TRABECULAR ARCHITECTURE FOR BONE REPAIR
`
`7027
`
`\ /as evaluated and approved by the IACUC of The Uni-
`versity of Tennessee Health Science Center at Memphis to
`ensure that the policies, standards and guidelines for the
`proper use, care, handling and treahnent of animals were
`obsen'ed.
`Two defect locations were created in both the left and
`right aspects of the mandible. The periosteum was
`retracted to expose the mandible surgical locations below
`the forward premolar teeth which were removed in con-
`junction with a separate study. A treptrine drill bit was
`used to create S-mm defects through the buccal side of the
`cortical bone. Defects r,l'ere left unfilled as an empty con-
`trol, filled with M-HA or N-HA scaffolds. Randomization
`in tl-te order of in-Lplant placement was performed in each
`animal. As illustrated in Figure 2, the clefects were intro-
`duced into the mandible with image analysis performed in
`a coronal (frontal) plane of view.26
`
`Surgical evaluation and histomorphometry
`
`Animals were sacrificed at 3 and 12 weeks post surgery.
`Mandibles were collected ancl placed into 10% neutral
`buffered formalin. Micro-CT sections \ /ere generated to
`observe macro tissue to implant integration and to assess
`any abnormalities ir the surrounding tissue. Embeddìng
`and sectioning rvere performed follou'ing the same proce-
`dure for control scaffolds. Two sections of each implant
`representing the central region were used for quantitative
`histology as shown in Figule 3, while remaining sections
`lt'eLe obserwed for qualitative features. Thin sections were
`stained for connective tissues with Paragon (toluidine blue
`and basic fuchsin) and calcified bone tissue with Alizarin
`Red. Quantitative histomorphometry was performed on
`stained sections in Bioquant Osteo by color ttuesholds and
`direct measurement of area values. Tissue and bone forma-
`tion data represent the percentage of those tissues within
`each total region of interest. Tissue was classified as all
`biological material staining wíth Paragon and Aliza¡in Red
`whereas mineralized bone u'as classified by material stain-
`ing onlv with Alizarin Red. As defined, partiallv mineral-
`ized osteoid was rìot ircluded in the bone classification.
`All tissue-scaffold measurements lr.'ere performed inside
`the perimeter of the scaffold. All porous spaces were
`excluded from tissue or bone measurenìents except for
`osteocyte lacunae and other featr,rres in the micron range.
`
`ð
`(h)
`
`{a}
`
`f
`Figure 2. Representative image of the dog mandible
`showing (a) the location of the introduced defects below
`the premolar teeth and (b) coronaÌ or frontal cross section
`of the mandible with defect traversing the lateral aspect of
`the cortical bone, mandible image adapted from Boyd
`et a1.26
`
`(}!¡rlrr liaclltns lirr llìslonurr¡rhonrcir¡
`
`Qrrrrlifuf ire
`Prr¡FhÈr¡rl
`!É*lic*s
`
`Figure 3. Histological preparation scheme showing cross
`sections through an implanted scaffold vvith assignment of
`quantitative slides from the center and qualitative descrip-
`tions from the outer sections.
`
`Vessels were identified in calcified sections within the scaf-
`folds at 12 weeks. Vessel distribution and diameter were
`compared with r¡essels counted from the same animal at a
`location greater than 5 nrm from the defect. Haversian
`canals, Volkmaru-r's canals, and small vessels were quanti-
`fied with visibly interconnected channels counted as one.
`Currently forming osteons and resorbtion spaces were not
`included in the measurement. Natural trabecular bone
`parameters were also reported from the control (nonscaffold)
`group using the same histomorphomtery values as
`described n ScffiId chnracterizøtiott.
`
`Statistical analysis
`
`Statistical calculations 'were performed with SigmaStat
`softr,r'are (Systat, Point Richmond, CA). Significance
`between groups was anaÌyzed by one-way ANOVA with
`Tukey pairwise multiple comparisons. Significance levels
`were set at p < 0.05.
`
`RESULTS
`
`Material characterization
`
`Overall scaffold architecture shown in Figure
`4(a,b) demonstrates the open interconnected features
`of the rodlike strut design. Open channels were
`arranged with isotropic geometry and rounded-edge
`triangular strut morphology. M-HA and N-HA sur-
`faces were observed for scaffolds with and without
`sol-gel coatings, respecti\¡ely [Fig, 4(c,d)]. Sol-gel
`coatings formed a thir-r layer of precipitated HA crys-
`tals on the bulk scaffold increasing the SEM visual-
`ized surface roughness and nano-texturing. Scaffolds
`averaged an average of 77.4% porosity as measured
`by histomorphometry (Table I).
`
`ln aioo analysis
`
`Following surgical resection, mandible-block speci-
`mens imaged by micro-CT were examined for
`
`Journal of Biomedical Moterials Reseøt'ch Pnrt A
`
`Baxter Healthcare Corp., et al. v. Millenium Biologix, IPR2013-00590, Exhibit 1174, p. 3
`
`

`
`1022
`
`APPLEFORD ET AL
`
`¡.:. "
`
`I¡
`
`Figure 4. SEM pirotographs of the HA scaffolds (a) bulk macloporisity, (b) N-HA scaffold strut morphology, (c) M-HA,
`arrd (d) N-HA surface features. Scale bar dimensions of 1000, 40,2, and2 ¡tn:..
`
`delayed- or ïìonunion with the surrounding bone tis-
`sue. At the time of recovery, there was no e\¡idence
`of inflammation or rejection around the scaffolds or
`open defects. An intact periosteum was found cover-
`ing the surgery location and macroscopic bone for-
`nation could be identified in the specimens. Micro-
`CT, after 3 weeks post surgery, revealed soft and
`hard tissue within tl-re control clefect and trotl-r scaf-
`fold designs, Figure S(a-c). After L2 weeks implanta-
`tion, micro-CT obserr.ations indicated the formation
`of dense cortical l¡one r,r.,ithin both scaffolds, lthereas
`a thin cortical shell blidge r,r,as observed forming
`across the control defects [Fig. s(d-f)].
`After three weeks, trabecular bone could tre
`observed forming inside the control defect by histol-
`
`ogy. Tlris bone rvas analyzed and compared with
`scaffold values with data represented in Table I. It
`was observed that scaffold properties closely
`matched the properties of newly formed trabecular
`bone at 3 r'eeks 'r,r'ith respect to bone surface to bone
`voltrnre ratio (35.6 -r 5.6,25.4 t 7.6 mm 1), trabecu-
`lar number (5.+ + 0.2, 5.1 + 1.2 mm r), separatìon
`(400 * 68,342 + 64 prm), and thickness (98 t 15, 138
`1 36 ¡rnr), respectively.
`Histomorphometry evaluation revealed tissue infil-
`tration into both scaffolds at 3 weeks. New bone for-
`mation progressed from the originaì bone edge into
`the scaffold and from the marrow cavity. Ner,r'ly
`formed and n-Lineralized bone rt'as observecl infiltrat-
`ing both scaffolds u,'ith seni-random collager.r fiber
`
`TABLE I
`Histomorphometry Parameters of HA Scaffold Compared to Newly Formed Trabecular Bone in the Defect Location
`3 Weeks Post Surgery
`HA Scaffold
`
`Histomorphornetry Parlameter'
`
`Aìrbrer¡iation
`
`3 Week Bone
`
`Units
`
`mmt
`tnm l
`mml
`Fm
`Inn
`
`5.4
`1.7
`7.6
`1.2
`64
`Jt)
`
`+++ I ;
`
`36.4
`9.7
`25.4
`5,1
`342
`138
`
`++++++
`
`BV/TV
`Bc¡ne vohrme,/tot¿rl volunre
`22.6
`Bone strrface/ total r,'oÌume
`BS/TV
`7.6
`Bone sulface/bone volume
`BSlBV
`3,+.6
`l-b.N
`Trabercu.ìar number
`5.4
`'Ib Sp
`Tr¿rbec ul¿rr sep irratíon
`466
`Trabeculal thickness
`Tb Th
`98
`HA scaffolds are structurally identical for M-HA and N-HA surfaces, p < 0.05
`
`5.É,
`0.7
`5.6
`0.2
`68
`15
`
`lountnl of Biomeclic¡l l¡4otetiøIs Rcscarth Port A
`
`Baxter Healthcare Corp., et al. v. Millenium Biologix, IPR2013-00590, Exhibit 1174, p. 4
`
`

`
`HYDROXYAPATITE SCAFFOLDS WITH TRABECULAR ARCHITECTURE FOR BONE REPAIR
`
`7023
`
`.-
`
`(d)
`
`(fl
`Figure 5. Coronal orientation, micro-CT images of control defect (a, d), implanted M-HA (b, e), and N-HA (c, f) scaffoìds
`after 3 and 12 r.t'eeks post surgery in the canine mandible. Radial defect was created from the lateral surface r,r'ith conrrol
`bone bridging initiated by 12 weeks. Scafiold groups show earlv union and bridging r,r'ith dense cortical tissue found at L2
`weeks, arror¡' shows cortical bridge formed in control defect.
`
`,t
`
`orientation observed by cross-polarized light. Con-
`trol and N-HA scaffolds after 3 weeks are shown in
`Figure 6(a,b) with N-HA demonstrating no morpho-
`logical differences compared to M-HA scaffolds.
`Quantification of this area showed an early majority
`of fibrous tissue,61.8 -'. 6.0 and 72.7 -r 2.8%, inil-
`trating the M-HA and N-HA scaffolds, respectively.
`At 3 weeks, 4.4 'r 2.6'k mineralized bone formation
`in M-HA scaffolds and 7.2 -r 6.6'k in N-HA scaffolds
`(Table II) had occurred. Normalized percent values
`of mineralized to total tissue rvere recorded with 7.0
`r 3.5"k conversion in M-HA scaffolds and 10.1 r
`9.57o com'ersion in N-HA scaffolcls (Table III). Colla-
`gen organization was also found in a partially lamel-
`lar organization coating the surface of the scaffolds,
`Figure 7(a,b).
`By 12 weeks post surgery, histomorphometry eval-
`uations of the untreated defects concurred with the
`micro-CT observations indicating the formation of a
`thin cortical bridge across the opening. Tissue infil-
`tration progressed throughout both scaffold designs
`with few pores left unfilled. Mineralized bone forma-
`tion represented 43.9 -r 4.7"k of the M-HA scaffolds
`and 50.4 r 8.8"/" of the N-HA scaffolds (Table II)
`and conversion from connective tissue into mineral-
`ized bone were correspondingly recorded at 59,0 +
`3.5 and 65.3 | 11.3% (Table III). Figure 6(c,d) shows
`the visible and cross-polarized light distribution of
`tissue formed inside a N-HA scaffold also represen-
`tative of M-HA scaffold morphology. Open channels
`
`remain unfilled at the center of the scaffold pores
`representing the blood and nutrient supply to the
`newly formed tissues. Collagen patterning by 12
`r¡'eeks has visibly organized into interlaced strands
`wrapping around the scaffold struts, FigureT(c,d).
`Blood '.'essel conduction and diameter within the
`scaffolcls were reported and compared with blood
`vessels found in the adjacent tissue at a distance of 5
`mm from the defect site. Within the cortical bone tis-
`sue these consisted predominantly of Hat ersian and
`Volkmann's canals. The control bone averaged 76.9
`-r 3.4 vessels/mtn2 with an average diametei of 19.0
`+ 0.7 ¡rm. As shown in Figure 7(c), vessels within
`HA scaffolds were observed. Vessels found within
`tlre scaffolds had a less frequent distribution of 7,7
`r- 1.4 and 7.6'+ 2.7 vessels,/mm2, with a larger aver-
`age diameter oI 23.3 + 2.7 and 23.0 +' 0.4 pm in M-
`HA and N-HA scaffolds, respectively (Table IV). No
`significant differences were fould betr.t'een vessel
`distribution or diameter between the two HA scaf-
`folds. However, significant difference in vessel distri-
`bution and diameter were observed between HA
`scaffolds and control bone (p < 0.05).
`
`DISCUSSION
`
`This study investigated the bone forming and
`angio-conductive capability of a scaffold design
`
`Jounml of Biomedical Mntcrials Rescnrch Pttt A
`
`Baxter Healthcare Corp., et al. v. Millenium Biologix, IPR2013-00590, Exhibit 1174, p. 5
`
`

`
`1024
`
`APPLEFORD ET AL.
`
`r.,¡
`
`Figure 6. Bone tissue cross section of cor-rtrol defect stained with Paragon for connective tissr-re (r,iolet) and Alizaril Red
`for mineralized bone tissue (red) after (a) 3 and (b) 12 r,r'eeks post surgeÐ'and N-HA scaffold shown after (c) 3 and (d) 12
`weeks r,r'ith scaffold in black, x80 original magnification. Lateral edge orientated upward.
`
`ìS.
`
`ô
`
`\
`
`LT
`
`¡l
`
`(d)
`
`<'#s^
`
`ê
`
`Figure 7. Bone tissue cross section of N-HA scaffold under phase contrast after (a) 3 and (c) 12 weeks post surgery with
`correspondir.rg cross-polatized light rnicrographs shown in (b,d) r'epresenting birefringence of collagen strands, x200 (S,
`scaffold; M, mineralizecl bone; C, collagen birefi'ingence; V, vessel).
`
`lottrnnl of Eiomedicøl Mntetials Rescarch Prtrt A
`
`Baxter Healthcare Corp., et al. v. Millenium Biologix, IPR2013-00590, Exhibit 1174, p. 6
`
`

`
`HYDROXYAPATITE SCAFFOLDS WITH TRABECULAR ARCHITECTURE FOR BONE REPAIR
`
`1,025
`
`TABLE II
`Total Tissue Area and Mineralized Bone Area Percentages in M-HA and N-HA Scaffolds
`at 3 and 12 Weeks Post Surgery
`Area of Total Tissue (%)
`
`Area of Mineralized Bone (%)
`
`N-HA
`
`66
`88
`
`++
`
`72
`504
`
`M-HA
`4.4 + 2.6
`43.9 + 4.r
`
`N-HA
`72.7 + 2.8
`77.3 + 4.5
`
`M.HA
`
`60
`31
`
`++
`
`61 B
`744
`
`Time Post
`Surgery (Weeks)
`
`3
`12
`
`closely approximating trabecular bone. A sol-gel
`coating of nano-crystalline HA r¡'as also investigated
`for improvements related to its nano-texturing and
`naterial properties at 3 and 12 weeks post implanta-
`tion into the canine mandible. Although significant
`evidence exists in the literature defining the morpl-ro-
`logical properties of natural borre,75'27 it is seldom
`related to bone formation in scaffolds. As the natural
`process in segmental bone repair progresses througl-r
`a wo\¡en and trabecular bone-like phase to a mature
`cortical structure, this research focused on using a
`scaffold patterned on trabecular bone. The presence
`of material cun'atures is known to be a substantial
`influence on cell and tissue behavio¡ and that careful
`control of these architectures may lead to enhanced
`bone formation.2s As observed in this study, surface
`to volume ratio was designed to provide nulnerous
`locations for early cell attachn-rent with strut dimen-
`sions (thickness, 98 r 15 ¡rm) smaller than their
`in aíao counterparts (1gA * 36 pm). Rationale for this
`arrangement was to permit cell-layering around the
`struts with an organized collagen network. Although
`the defect size used in this study was not critical in
`character, as evidenced by the bridging in the con-
`trol after 12 weeks, the use of this defect size was
`ideal for comparing the naturally formed trabecular
`bone with the replacement scaffold. Additionally,
`comparisons of tissue formation were not possible
`between control defects and scaffold groups because
`the original defect location could not be accurately
`located to ensure that controls would match the
`exclusive center-section criteria for quantitative his-
`tomorphometry.
`
`TABLE III
`Relative Mineralized Tissue Formation Compared to
`Total Tissue in M'HA and N-HA Scaffolds at 3 and
`12 Weeks Post surgery
`
`Time Post
`Surgery (Weeks)
`
`3
`12
`
`Mineralized/Tota1 Tissue (%)
`
`M-HA
`7.0 + 35
`59.0 + 35
`
`N-HA
`10.1 + 9.5
`65.3 + 77.3
`
`At the early time point of 3 weeks post surgery,
`the interconnected scaffolds of 77.4 -r 5.6/" porosity
`r,r'ere immediately filled with comective tissue. Gen-
`eration of osteoid was already established at this
`time, and 4.4 'r 2.6 and 7.2 -r 6.6'k ftrlly mineralized
`tissue was observed in M-HA ancl N-HA scaffolds,
`respectively. Histology revealed that the mineraliza-
`tion front was initiated from contact with existing
`bone, from the rnarrow ca\zity, and clirectly on the
`surface of the scaffold struts reflecting the strong
`osteoconductive characteristics of material and archi-
`tecture. Osteoblasts were identified at 3 weeks mak-
`ing use of the scaffold arrangement by depositing
`newly formed bone in circular lamella from the scaf-
`fold surface. Cross polarized light revealed bundled
`collagen fiber organization by luminous birefrin-
`gence that was characteristic of wo\¡en and partially
`lamellar bone at this early time point.
`After 72 weeks post surgery, nearly complete min-
`eralízed bone formation r,r'as observed in both HA
`scaffolds. This represents the majority of the avail-
`able space for bone tissue as the scaffold itself occu-
`pies 22.6"/" of the area. At 12 weeks, the mineralized
`bone showed full interconnection with the scaffold
`and small nutrient chalnels were observed tunneling
`througl'rout the tissue. When compared to the sur-
`rounding cortical bone tissue, intra-scaffold vessels
`were fould to be less numerous but significantly
`larger than Haversian and Volkmann's canals. Since
`the scaffold design ultimately prevents complete
`osteon formation, mineralized bone must be sup-
`plied with nutrients by channels twisted around
`struts. This vessel arrangement could be related to
`
`TABLE IV
`Blood Vessel Distribution (per mm) and Diameter (¡rm)
`Within 12 Week M-HA, N-HA Scaffolds and Natural
`Cortical Bone
`
`Bloc¡d Vessel
`Parameter
`
`Cortical
`M-HA
`N-HA
`Bone
`7.7 + 1.4
`+16.9 + 3.4
`7.6 + 2.7
`Vessels/mm
`23.3 + 2.7
`23.0 + 0.4
`+19.0 + 0.7
`Vessel diameter (pm)
`Values represent Haversìan, Volkmanl's canals and
`blood vessels only.
`*Denotes statistical difference from each scaffold group,
`p < 0.05.
`
`lourrnl of Biomedical Mnterials Resench Pøt A
`
`Baxter Healthcare Corp., et al. v. Millenium Biologix, IPR2013-00590, Exhibit 1174, p. 7
`
`

`
`1026
`
`APPLEFORD ET AL.
`
`the scaffold architecture where open spaces permit
`less nurnerous but larger, interconnected, bone-
`blocks to form.
`Collagen organization within the scaffolds at 12
`rveeks showed lamellarlike organization with bun-
`dles woven around and between the struts. Visual-
`ized cross-sections at the center of scaffold pores
`demonstrated a circular collagen arrangement with a
`Haversian-like vessel at its center. Although the scaf-
`fold architecture cannot permit true osteons to form
`in a straight path, it was observed that osteocytes
`and their surrounding collagen matrix did form in
`concentric rings to the rressels. This arrangement
`allowed a semi-lamellar pattern to form althougl-r the
`distribution of collagen bundles was partially tan-
`dom. With no evidence of necrosis, even at the cen-
`ter of the scaffolds, this implant arrangement dem-
`onstrated the functionality of open-channel designs
`for both osteo- and angio-conduction.
`
`CONCLUSION
`
`This study demonstrated the potential of porous,
`interconnected, HA scaffolds with morphological
`similarity to natural trabeculae for bone regenera-
`tion. Distinct strut features and open channels
`allowed for successful mineralized bone and vascu-
`lar infiltration throughout both scaffold designs. Par-
`tially organized, lamellar collagen fibrils were identi-
`fied by birefringence under cross-polarized light at
`both 3 and 12 weeks post implantation.
`
`References
`1 Arinzelr TL, Tran T, Mcalary J, Daculsi C. A comparative
`study of biphasic calcìum phosphate ceramics for human
`mesenchymal stem-cell-induced'bone formation Biornaterials
`2005;26:3631--3638.
`2. Dong l, Kojima H, Uemura'I', Kikuchi M, Tateishi 'l', Ï'anaka
`J. In r,:ivo evaluation of a novel porous hvdroxyapatite to sus-
`t.ril osteogenesis of transplanted bone marrou,-derivecl osteo-
`ì:last cells J Biomed Mater l(es 2007;57:208-276.
`3. Flautl'e B, Anselme K, Delecourt C, Lrt l, Hardouin P, Des-
`camps M Histological aspects in bone regeneration of an
`assocíation with porous hydrox¡,apatite ard bone marto\l'
`cells. J Mater Sci Mater Med 1999;10:811-814.
`4. Gauthier O, Mulle¡ R, r'on Stechow D, Lamy B, Weiss P. Bou-
`ler JM, Aguado E, Daculsi G. In vir¡o bone regeneration u'ith
`injectible calcinm phosphate biomaterial: A three-dimensionaL
`mico-computed tt-rmographic, biomechanical and SEM study
`Biornaterials 2005 ;26 :5 4M-5 453
`5 Kon E, Muraglia A, Corsi A, Bianco P, Marcacci M, Martin I,
`Boyde A, Ruspantiní I, Chistolini P, Rocca M, Giardino R,
`Cancedda R, Quarto R. Autologous bone marrorv stromaì
`cclls loadecl onto porous hydroxyapatite celamic accelerate
`bone repair in critical-size defects oi sheep long bones.
`f Biomed Matel Res 7999;49:328 337.
`
`lon'nnl of Biomeclical Matø ials Research Port A
`
`6. Koshino'I', N4urase T, 'l'akagi T, Saito'I'. Ner.r'bone formation
`around porous hydroxyapatite r,r'edge ímp'lantecl in opening
`u'edge high tibial osteotomy in patients r.vith osteoarthritis
`Biomaterials 2001, ;22:1,579 -1582.
`7. LrlX, Gallur A, Flauhe B, Anselrne K, Descamps M, Thierry
`B, Hadouin P Comparative study of tissue reactions to cal-
`ciurn phosphate ceramics among cancellous, cortícal, and
`medullar bone sites in r"rbbits J Biomed Mater Res 1998;
`42:357-367.
`8. Mastrogiacoma M, Sca¡;Jione S, Martinetti R, DoÌcini L, Bel-
`trame F, Cancedda R, Quarto R Role of scaffold intemal
`structure on in vivo bone formation in macro¡rorous calcium
`plrosplrate bioceramics. Bi om a telial s 2006 ;27 :323(1 3237 .
`9 Posner AS. Bone mineral on the molecular level. Fed Proc
`1973;32:1933-7937.
`10. Richard M, Aguada E, Cottrel M, Daculsi G Ultrastructu'al
`and electron diffraction of the bone-ceramic interfacial zone
`in coral and biphasic CaP implants. Calcif Tissue Int 1998;
`62:437442.
`11 Arts JJ, Verdonschot N, Schreurs BW, Buma P The use of a
`bioresorbable nano-cryatlline hydroxyapatite paste in acetab-
`ular bone imp"rction grafting. Biomaterials 2006;27 :'1110-11,1,8.
`12. De Oliveira JF, De Aguiar PF, Rossi AM, Soares GA EÍfect of
`process parameters on the characteristics of porous calcium
`phosphate ceramics for bone tissue scaffolds Artif Organs
`2003;27:4O6-41.1..
`13. Lu JX, Flautre B, Anselme K, Hardouin P, Gallur A, Des-
`camps M, Thierry, B. Role c¡f interconnectic¡ns in porous bio-
`ceramics on bone recolonization in vitro and in vivo. J lvfater
`Sci Mater Med 1999;10:1ll-120
`14. Mankani MH, Kuznetsov SA, Fowler B, Kingman A, Robev
`PC. In vivo bone formation by human bone marrou' stromal
`cells: Effect of carrier particle size and shape. Biotechnol Bio-
`eng 2007;72:96-107
`15. Navarro M, deì Valle S, Martinez S, Zeppetelli S, Ambrosio
`L, Planell JA, Ginelrra MP Nerv Íracropolous calcium phos-
`phate glass ceramic for guicled bone regeneration Biomateri-
`als 200 4;25 : 4233 - 4241..
`16 Septìr'eda P, Bimer JC, Rogero SO, Higa OZ, Bressiani JC,
`Production of porous hvd¡oxyapatite by the gel-casting oÍ
`foams and cytotoxic evaluation, J Biomed Matel Res 2000;
`50:27-34.
`'17 Gauthier O, Bouler JM, Aguado E, Pilet P, Daculsi G. Macro-
`porous biphasic calcitLm phosphate ceramics: lnf-luence of
`macropore diameter and macropolosity percentâge on bone
`ingroi,r'th, Biomaterials 1996;19:133-139
`18. Chang BS, I-ee CK, Hong KS, Yourr HJ, Ryu HS, Churrg SS,
`Park KW Osteoconduction at porous h.vdroxyapatite with
`various pore configurations. Biomaterials 2OO0 ;21:1291-1298.
`19, Karageorgiou V, Kaplan D. Porosity of 3D biomatelial scaf-
`f olds and osteogenesis. Biomaterials 2005 ;27 :547 4-5 497
`20. Habibovic P, Yuan H, van der \¡alk C, Meijer G, varr Blitters-
`wijk C, de Groot K. 3D microelrvironmellt as essential ele-
`ment ior osteoincìuction by biomaterials. Biomaterials 2005;
`26:3565-3575.
`21. Bovde A, Corsi A, Quarto R, Ca;rcedda R, Bianco P. Osteo-
`conduction in large macroporous hydroxlrap¿1i1s ceramic
`implants: Er.idence for a complimentar)¡ integration and dis-
`integratiorl mechanism. Bone 7999 ;24:579-589.
`22 Barou O, Mekraldi S, Vico L, Boi','in G, Alexanclre C, LaÍage-
`Proust MH. Relationships between trabecular bone rernodeling
`and Lrone vascularization: A quantitative study. Bone 2002;
`30:604-61.2
`23 )'ou C, Oh S, Kim S. Influences of heating condition and sub-
`strate-surf.rce roughness on the characteristics of sol-gel-
`derived hydroxyapatíte coatings. J Sol Gel Sci Tech 2OO'1;21-:
`49 54.
`
`Baxter Healthcare Corp., et al. v. Millenium Biologix, IPR2013-00590, Exhibit 1174, p. 8
`
`

`
`HYDROXYAPATITE SCAFFOLDS WITH TRABECULAR ARCHITECTURE FOR BONE REPAIR
`
`1027
`
`24. Parhft AM, Drezner MK, Glorieux FH, Kanis JA, Malluche H,
`Meunier