collagen scaffolds implanted in the palatal mucosa · cross-linked collagen scaffolds and showed a...

Copyright @ 200 Mutaz B. Habal, MD. Unauthorized reproduction of this article is prohibited.8

Collagen Scaffolds Implanted in thePalatal Mucosa

Richard G. Jansen, DDS,* Anne M. Kuijpers-Jagtman, DDS, PhD,* Toin H. van Kuppevelt PhD,1

Johannes W. Von den Hoff, PhD*

Nijmegen, the Netherlands

Scar formation after repair of the cleft palate leads togrowth impairment of the upper jaw and midface.The implantation of a suitable scaffold duringsurgery may reduce this adverse effect. However,little is known about tissue reactions to scaffoldsimplanted in the oral cavity. Our goal was to analyzethe tissue reactions to cross-linked type I collagenscaffolds after submucoperiosteal implantation inthe palate of rats. Collagen type I scaffolds wereimplanted in the palate of 25 male Wistar rats.Groups of 5 rats were killed consecutively after 1, 2,4, 8, and 16 weeks and were processed for histologicand immunohistochemical analyses. After 1 and 2weeks, 3 rats from the sham group were also killed.On hematoxylin and eosinYstained sections, the celldensity and the number of giant cells weredetermined. Blood vessels, inflammation, and thepresence of myofibroblasts were detected by immu-nohistochemistry. An influx of inflammatory cellsstarted after 1 week but had completely subsidedafter 8 weeks. Myofibroblasts were observed withinthe scaffolds only in the first 2 weeks. Angiogenesisalready started after 1 week and showed a peak after4 weeks, slowly declining afterward. The scaffoldswere gradually integrated within the host tissue andonly elicited a mild and transient inflammatoryresponse. The scaffolds were biocompatible andseemed to be promising for future applications incleft palate surgery.

Key Words: Collagen, scaffold, wound healing, animalmodel, biocompatibility, in vivo test

Acleft palate is a frequently occurring con-genital malformation.1 Surgical closure ofthese clefts is indicated to overcome feed-ing and speech problems. However, the

existing surgical procedures lead to scar formation,which impairs the growth of the maxilla and thedevelopment of the dentoalveolar complex.2Y5 Apromising approach to reduce these problems maybe the implantation of a scaffold loaded with growthfactors to reduce scar formation. In the last decade,tissue-engineered scaffolds have already been deve-loped for the treatment of skin wounds. These scaf-folds, loaded with growth factors, enhanced woundhealing and reduced scar formation after implanta-tion in the skin.6,7 Particularly, collagen is widely usedto prepare these scaffolds because of its overall goodbiocompatibility and biodegradability in the skin.8

The immunologic biocompatibility of collagen in theskin has been demonstrated in animal models and inhumans.9,10 The low immunogenicity of a bovine col-lagen scaffold in humans is due to the high-sequencehomology between bovine and human collagen.11 Inaddition, the collagen used is predominantly in thenative triple-helical form.12 Porous type I collagenscaffolds implanted subcutaneously in rats wereshown to be biocompatible and to allow the ingrowthof cells.13 After implantation in the skin, the scaffoldsare gradually degraded and replaced by the regene-rating tissue.14 In addition, cross-linked collagenscaffolds preserved their integrity longer than non-cross-linked collagen scaffolds and showed adecreased inflammatory response, probably becauseof their reduced degradation.15 The implantation ofcollagen alone in skin was already found to reducescar formation in animals16 and to decrease woundcontraction in a human wound model.17

In summary, collagen scaffolds, either with orwithout growth factors, seem to be promising fortissue engineering purposes in the palate or elsewherein the oral cavity. However, it is not known whetherthey are suitable for intraoral applications, because thedermal and oral environments are different. In addi-tion, differences between oral and dermal fibroblastsin vitro have been reported.18Y20 Oral wound healing

599

From the Department of *Orthodontics and Oral Biology; and1Biochemistry, Nijmegen Centre for Molecular Life Sciences,Radboud University Nijmegen Medical Centre, Nijmegen, theNetherlands.

Address correspondence and reprint requests to JohannesW. Vonden Hoff, PhD, Department of Orthodontics and Oral Biology,Radboud University Nijmegen Medical Centre, PO Box 9101,Nijmegen 6500 HB, the Netherlands; E-mail: [email protected]


is generally reported to occur faster and to be morefetal-like,18,21 with less scarring thanwound healing inthe skin.22,23 In contrast, some researchers claim adelayed wound healing on the palate.24 However,most studies on intraoral wound healing are per-formed on buccal or periodontal wounds.25 Studiesfocused on palatal wound healing did show theformation of a rigid scar tissue tightly attached to thebone.26 Little is known about the tissue reactions tocollagen scaffolds in the oral cavity. One study inrabbits showed that the implantation of partly cross-linked atelocollagen in combination with a siliconsheet decreases scarring.27 However, the study con-tained only limited histologic analyses.

The aim of our study was to evaluate the tissuereaction to cross-linked collagen type I scaffolds aftersubmucoperiosteal implantation in the palate of rats.

MATERIALS AND METHODS

Animals

Thirty-one 5-week-old male Wistar rats (Harlan,Zeist, the Netherlands), weighing between 106

and 166 g, were used. All animals were kept undernormal laboratory conditions and were fed standardrat chow and water ad libitum. The rats had beenacclimatized to the animal housing facility for 1 weekbefore the start of the experiment. The experimentwas approved by the board for animal experiments ofRadboud University Nijmegen.

Collagen Scaffolds

Type I collagen was purified from bovine Achillestendon, and a chemically cross-linked scaffold wasprepared.28 Briefly, a 0.8% (w/v) type I collagensuspension in 0.5-mol/L acetic acid was shakenovernight at 4-C and homogenized on ice using aPotter-Elvehjem homogenizer. Air bubbles wereremoved by centrifugation at 250g for 10 minutes at4-C. The suspension was then slowly poured into aplastic mold (10 mL per 25 cm2), frozen in a bath ofethanol and solid CO2 (j80-C), and lyophilized in aZirbus lyophilizer (ZIRBUS Technology GmbH, BadGrund, Germany). Scaffolds were cross-linked using33-mmol/L 1-ethyl-3-(3-dimethyl aminopropyl) car-bodiimide and 6-mmol/L N-hydroxysuccinimide in50-mmol/L 2-morpholinoethane sulfonic acid, pH5.0, containing 40% ethanol (20 mL per 25 cm2) for4 hours at 22-C. Scaffolds were then washed with0.1-mol/L Na2HPO4, 1-mol/L NaCl, 2-mol/L NaCl,and MilliQ water, frozen in ethanol/CO2 again, andlyophilized.

Surgical Procedures

The rats were anesthetized with an intraperitonealinjection of 1-mL/kg body weight of ketamine (Nima-tek; Eurovet Animal Health, Bladel, the Netherlands)combinedwith an intraperitoneal injection of 0.25-mL/kg bodyweight of xylazine (Sedamun; Eurovet AnimalHealth, Bladel, the Netherlands). A standardizedtransversal incision of 3 mm was made in the palatalmucoperiosteum at the level of the contact pointsbetween the first and second molars. The mucoper-iosteum was then elevated caudally for 3 mm with aperiodontal probe to create a submucoperiostealenvelope. A circular collagen scaffold with a diameterof 3 mmwas placed in the envelope and a 10� 0-vicrylsuture was used to close it. The rats were medicatedpostoperatively with 0.02-mg/kg body weight bupre-norfine (Temgesic; Schering Plough, Brussels, Belgium)subcutaneously as an analgesic. The same procedurewithout placement of the scaffoldwas performed in thesham group.

Study Design

Groups of 5 rats were killed at 1, 2, 4, 8, and 16 weeksafter implantation and were processed for histologicanalyses of the wound tissue. Groups of 3 rats of thesham group (6 rats) were killed at 1 and 2 weeks afterimplantation.

Histologic Diagnosis

The rats were perfused using freshly prepared 4%paraformaldehyde in phosphate-buffered saline (PBS).After decapitation and removal of the skin andmandible, the heads were fixed in 4% paraformalde-hyde solution for 24 hours and thereafter decalcified in10% ethylenediaminetetraacetic acid in PBS at 4-C.Serial paraffin sections of 6 mm in size were cut in thefrontal plane and stained with hematoxylin and eosin(H&E, according to Delafield). Histomorphometricanalyses were performed blindly using an ocularmicrometer. The number of giant cells and the totalnumber of cells in the scaffolds were counted on 3sections (150-mm apart) for each wound. The totalnumber of cells in the scaffolds was counted with amicroscopic grid and was expressed as the number ofcells per square millimeter. At 1 and 16 weeks, the celldensity in the tissues surrounding the scaffolds wasalso quantified. The sections were also stained with theHerovici picropolychrome stain to analyze collagendeposition in the scaffolds.

Immunohistochemistry

Paraffin sections were collected on Superfrost Plusslides (Menzel-Glaser, Braunschweig, Germany),

THE JOURNAL OF CRANIOFACIAL SURGERY / VOLUME 19, NUMBER 3 May 2008

600


deparaffinated, and rehydrated. Before staining, theslides were rinsed in PBS for 10 minutes. aYSmoothmuscle actin staining was performed for the detectionof myofibroblasts, ED-1 staining was performed todetect inflammatory cells, and collagen type IVstaining was performed to detect blood vessels.

aYSmooth Muscle Actin Staining

Sections were treated with 3% H2O2 in methanol for10 minutes to block endogenous peroxidase and werethen rinsed in PBS. Then, the sections were preincu-bated in 10% normal donkey serum (ChemiconEurope, Hampshire, United Kingdom) in PBS. Afterpreincubation, the sections were incubated withmonoclonal mouse antiYaYsmooth muscle actin(Sigma Chemical Co, St Louis, MO) at a ratio of1:6400 for 60 minutes. After washing with PBS,detection was carried out using biotinylated donkeyantimouse IgG (Jackson ImmunoResearch Labora-tories, Inc, West Grove, PA) at a ratio of 1:100 for 60minutes and the avidin-biotin complex method(Vectastain ABC-Elite Kit; Vector Laboratories, Bur-lingame, CA). The presence of myofibroblasts wasscored on a scale from 0 to 3 on 3 sections (150-mmapart) for each wound.

0: No, or only a few myofibroblasts present;1: Groups of myofibroblasts around the scaffold, no

or only a few myofibroblasts in the scaffold;2: Groups of myofibroblasts around and in the

scaffold; and3: Myofibroblasts throughout the scaffolds and the

surrounding tissues.

ED-1 Staining

The ED-1 staining was performed similar to theaYsmooth muscle actin staining, but the sections wereincubated with monoclonal mouseYantiYrat ED-1(Serotec, DPC, Breda, the Netherlands) at a ratio of1:400 for 60 minutes. This antibody recognizes asingle-chain glycoprotein of 90 to 110 kd in size that isexpressed predominantly on the lysosomal mem-brane of myeloid cells.29 It mainly stains macrophagesand monocytes. The inflammatory response wasscored on a scale from 0 to 3 on 3 sections (150-mmapart) for each wound.

0: No, or only a few inflammatory cells present;1: Groups of inflammatory cells around the scaffold,

no or only a few cells in the scaffold;2: Groups of inflammatory cells around and in the

scaffold, and

3: Inflammatory cells throughout the scaffold and thesurrounding tissues.

Type IV Collagen Staining

The type IV collagen staining was performed similarto the aYsmooth muscle actin staining, but thesections were incubated with rabbit antiYcollagentype IV (Euro Diagnostica BV, Arnhem, the Nether-lands) at a ratio of 1:200 for 60minutes. The number ofblood vessels present in the scaffolds was quantifiedon 3 sections (150-mm apart) for each wound.

Statistics

All measurements were performed on triplicatesections from each wound. The cell density, thenumber of blood vessels, and the number of giantcells in the scaffolds were counted. The outcomes atall 5 time points were compared with a one-wayanalysis of variance. Significant differences werefurther analyzed by the Holm-Sidak method.

Differences in the scores for the number ofmyofibroblasts and the degree of inflammation at all5 time points were compared by a Kruskal-Wallis one-way analysis of variance on ranks. Significant differ-ences were further analyzed by the Dunn method.

The cell densities in and outside the scaffolds at1 and 16 weeks were compared using a paired t-test.P G 0.05 was considered significant.

RESULTS

General Histologic Findings

Normal Palatal Mucoperiosteum

The palatal bone is divided in half by the midpalatalsuture (Fig 1A). It is covered by the palatal mucoper-iosteum at the oral side. A molar is visible on eachside of the palate. The nasal septum divides the noseinto 2 nasal cavities, which are occupied by the nasalconchae. Below the nasal septum and just above thepalatal bone lies the nasopharyngeal tube. The palatalmucoperiosteum consists of 4 layers, namely, theepithelium, the lamina propria, the submucosa, andthe periosteum (Fig 1B). The epithelium is of theorthokeratotic stratified squamous type. The laminapropria consists of 2 sublayers, namely, a papillarylayer containing a relatively loose network of col-lagenous fibers and a deeper layer with more denselypacked collagenous fibers. The submucosa containsthe major arteries, veins, and nerves of the palate. Theperiosteum consists of 2 sublayers, namely, a fibrouslayer and an osteogenic layer. A cartilaginous tissue is

COLLAGEN SCAFFOLDS IN THE PALATE / Jansen et al

601


present in the midpalatal suture between the firstmolars.

Palatal Mucoperiosteum After Implantation

At 1 week after implantation, the collagen scaffoldscould easily be found between the submucosa and theperiosteum because of their specific fibrous structure(Fig 1C). Some inflammatory cells were presentaround the scaffolds, but only a few cells had invadedthem (Fig 1C). In one sample, incorporated foodparticles or sawdust had disturbed the healingprocess, and more inflammatory cells were found(not shown). Some giant cells were present at theedges of the scaffolds. At 2 weeks after implantation,more inflammatory cells were present around thescaffolds, and more cells had invaded them (Fig 1D).At 4 weeks after implantation, the number of giant

cells in and around the scaffolds seemed to beincreased (Fig 1E). The first blood vessels had alsoappeared in the scaffolds. After 8 and 16weeks, almostno inflammatory cells or giant cells were present inand around the scaffolds anymore (Fig 1F). Thecollagen bundles of the scaffolds had become muchthinner, and the scaffolds themselves seemed to havedecreased in size. In the next sections, some histologicaspects have been quantified and, where necessary,additional (immuno)histochemical staining has beenperformed.

Inflammation

Inflammation was scored on a scale from 0 to 3 on ED-1Ystained sections (Fig 2). ED-1 also stains giant cells,but these were counted separately on H&EYstainedsections (see later). After 1 week, inflammatory cells

Fig 1 Palatal mucoperiosteum of the rat. H&E staining of tissue sections of the palate from the sham group and theexperimental group at various time points. A, Sham rat 1 week after surgery. N indicates nasal septum; C, nasal cavity;S, midpalatal suture; B, palatal bone; M, secondmolar (bar, 625 mm). B, Sham rat 1 week after surgery. P indicates periosteum;SM, submucosa; LP, lamina propria; E, epithelium (bar, 200 mm). C, One week after implantation (bar, 100 mm). D, Twoweeksafter implantation (bar, 100 mm). E, Four weeks after implantation (bar, 100 mm). F, Sixteen weeks after implantation (bar,100 mm). Implanted scaffolds are indicated by an asterisk, and giant cells are indicated by an arrow.


602


were found only at the borders of the scaffolds(Fig 2A). After 2 and 4 weeks, inflammatory cellswere observed both in and around the scaffolds(Fig 2B). However, after 16 weeks, almost all inflam-matory cells had disappeared (Fig 2C). A significant

difference was only found between both 2 and 4weeks, with a median of 2.0, and 16 weeks, with amedian of 0.3 (Fig 2D). In the samples from thesham group, groups of inflammatory cells werepresent after 1 week, but 1 week later, they had

Fig 2 Inflammatory response. ED-1 staining of inflammatory cells at different time points. The number of inflammatorycells increased from 1 to 4 weeks and diminished afterward. A, One week after implantation. B, Four weeks afterimplantation. C, Sixteen weeks after implantation (bar, 100 mm). The scaffolds (asterisk) and inflammatory cells (arrow) areindicated. D, Inflammation scores. The box plots display the 25th, 50th, and 75th percentiles if available. Box plots represent 3to 5 animals. The asterisks denote a significant difference of 2 and 4 weeks with 16 weeks.

Fig 3 Cells within the scaffolds. The total number of giant cells in the scaffolds and the cell density were determined onH&EYstained sections. A, Number of giant cells within the scaffolds. The asterisk denotes significant differences between 4weeks and both 1 and 16 weeks. B, Cell density within the scaffolds (gray bars) and in the surrounding tissues (hatched bars).The asterisk denotes a significant increase after 8 and 16 weeks compared with 1 and 2 weeks. The pound sign denotes asignificant difference in the increase at 4 weeks compared to 1 week. The caret sign denotes a significant difference betweenthe cell density inside and outside the scaffolds at 1 week after implantation.


603


largely disappeared. The number of giant cells in thescaffolds was counted on the H&EYstained sections(Fig 1). Their number increased significantly from amean (SD) of 6.1 (3.8) per scaffold after 1 week to amean (SD) of 20.9 (11.4) per scaffold after 4 weeks(Fig 3A). After 16 weeks, the number of giant cellshad significantly reduced to a mean (SD) of 4.3 (1.9)per scaffold.

Cell Density

The cell density within the scaffolds was counted onH&EYstained sections (Fig 1). At 1 and 16 weeks, thenumber of cells outside the scaffoldswas also counted.At 1 week after implantation, a mean (SD) of 980 (223)cells/mm2 were present in the scaffolds. This numberincreased significantly until 8 weeks to a mean (SD) of

Fig 4 Collagen fibers. Herovici staining of the scaffolds (asterisk) shows newly formed collagen in blue and old collagen inred. After 2 weeks, newly formed collagen is present (arrows). A, One week after implantation. B, Two weeks afterimplantation (bar, 200 mm).

Fig 5 Myofibroblasts. Myofibroblasts were stained with an antibody against aYsmooth muscle actin. The number ofmyofibroblasts rises between 1 and 2 weeks and diminishes after 4 weeks. At 8 and 16 weeks, myofibroblasts were no longerpresent. A, One week after implantation. B, Two weeks after implantation. C, Four weeks after implantation (bar, 100 mm).Scaffolds (asterisks) and myofibroblasts (arrows) are indicated. D, Myofibroblast scores. If available, box plots display the25th, 50th, and 75th percentiles. Box plots represent 3 to 5 animals. The asterisk denotes significant differences between 8 and16 weeks, compared to 2 weeks.


604


1936 (144) cells/mm2. After 8 weeks, the cell densityremained stable. A significant difference was alsofound between 1 and 4weeks (980 [223] and 1627 [237]cells/mm2, respectively) (Fig 3B). At 1 week afterimplantation, a significant difference was foundbetween the cell density in the scaffold (980 [223]cells/mm2) and in the surrounding tissues (2220 [268]cells/mm2). Newly formed collagen was detected onsections stained by the Herovici method (Fig 4).Collagen depositionwas shown alongside the bundlesof the scaffolds at 2, 4, and 8 weeks.

Myofibroblasts

aYSmooth muscle actin was found in blood vesselsand in myofibroblasts (Fig 5). There were markeddifferences between the different time points. Afterthe first week, myofibroblasts were abundantlypresent at the borders of the scaffolds, but only afew myofibroblasts had invaded the scaffolds (Fig5A). One week later, the number of myofibroblasts inthe scaffolds was at its maximum. Myofibroblastswere found throughout the scaffolds (Fig 5B). At 4weeks, only a few myofibroblasts remained (Fig 5C).

At later time points, no myofibroblasts were found. Asignificant difference was shown between the myofi-broblast score after 2 weeks, with a median of 2.0, andthe score after both 8 and 16 weeks (median, 0.0) (Fig5D). In the sham groups, myofibroblasts were presentat 1 and 2 weeks.

Vascularization

The number of newly formed blood vessels within thescaffolds was counted on sections stained for type IVcollagen, which is present in all basement membranes(Fig 6). A gradual increase in the number of bloodvessels was seen until 4 weeks (Fig 6A and B). After 4weeks, a gradual decline was noted (Fig 6C). At 1week, a mean (SD) of 5.9 (2.7) blood vessels wascounted in the scaffolds (Fig 6D), which had increasedto a mean (SD) of 36.2 (8.2) blood vessels after 4weeks. After 16 weeks, the number had declined to amean (SD) of 14.3 (4.0) blood vessels (Fig 6C). Asignificant difference was found between the peak at 4weeks and all other time points (Fig 6D). There wasalso a significant difference between 1 week and both8 and 16 weeks. Not only that the number of blood

Fig 6 Blood vessels. Blood vessels were stained with an antibody against collagen type IV. Basement membranes were alsostained. The number of blood vessels increased until 4 weeks after implantation. A, One week after implantation. B, Fourweeks after implantation. C, Sixteen weeks after implantation. Bars, 100 mm. Scaffolds (asterisks) and blood vessels withinthe scaffolds (arrows) are indicated. D, Number of blood vessels within the scaffolds. The asterisk denotes significantdifferences between 4 weeks and all the other time points. The pound sign denotes significant differences between 8 and 16weeks, compared to 1 week.


605


vessels had changed over time but also that the widthof the blood vessels in the scaffolds appeared toincrease in time.

DISCUSSION

Our final aim was to develop a growth factordelivery system based on collagen scaffolds,

which can diminish wound contraction and scar for-mation after cleft palate surgery. In this first study,the tissue response to cross-linked porous collagentype I scaffolds without growth factors was evaluatedafter submucoperiosteal implantation in the palateof rats.

A characteristic finding in the current study wasthat the inflammatory response and the foreign-bodyreaction after implantation of the scaffolds was onlymild and subsided rapidly. In contrast, the total celldensity of the implanted scaffolds increased steadilyuntil 8 weeks, indicating that further tissue ingrowthwas based on fibroblasts and other noninflammatorycells. This was confirmed by the immunohistochemicalstaining for monocytic cells and macrophages. Pieperet al13 used similar collagen scaffolds for subcutaneousimplantation in rats. In contradiction to our results,they found higher numbers of giant cells and inflam-matory cells in the first weeks after implantation. Thismight be caused by the generally lower levels ofmacrophages, neutrophils, and T-cells present in oralwounds compared to skinwounds.30 Furthermore, oralwounds contain fewer proinflammatory factors such astransforming growth factor b1 and interleukin 6.23,30,31

However, these studies were performed on openwounds in the buccal mucosa, which differs fromimplantation in the palatal mucosa. A study describingthe implantation of collagen without telopeptides(atelocollagen) in the palate of rabbits did not reportthe severity of the inflammatory response.27 However,their histologic pictures shownomarked inflammatoryresponse up to 20weeks after implantation. A study onthe implantation of atelocollagen in the gingiva of ratsalso showed similar results to ours, although theauthors claim that atelocollagen is less antigenic thannormal collagen.32

The number of giant cells in our study increaseduntil 4 weeks after implantation, but after 16 weeks,almost all giant cells had disappeared, although thecollagen was still present. Newly formed collagen waslaid down against the fibrils of the scaffolds, indicatingthat the scaffolds were no longer recognized as aforeign body butwere integratedwithin the host tissue.In palatal scar tissue in rats, collagen bundles mainlyhave a parallel and transversal orientation.33,34 Becausethe fibrils in our scaffolds are oriented randomly, this

may finally result in a more random pattern of newlyformed collagen.

Further studies on the implantation of collagenscaffolds in the palate are not available, but someauthors have used other materials. Two studies onbiodegradable polymers implanted in the palate of dogslack extensive description of the inflammatoryresponse.35,36 Both studies report the formation of acellular capsule around the polymers, which pointstoward an extensive foreign-body reaction. Prematuresequestration of the polymers also occurred. This mightbe caused by the local accumulation of acidic degrada-tion products and the subsequent induction of foreign-body reactions.37,38

Overall, our collagen scaffolds were well detect-able in the specimens up to 16 weeks after implanta-tion. The scaffolds were lost in only 2 of 25 wounds.These scaffolds might have disappeared due toloosening of the suture. Although most scaffoldswere still present at all time points, they graduallybecame smaller, and the collagen bundles becamethinner. Fast degradation of a collagen scaffold mightevoke an excessive tissue reaction, because its degra-dation products are chemotactic to inflammatorycells.39Y41 Furthermore, the scaffolds need to stay inplace for some time to deliver growth factors in futurestudies. The fact that the cross-linked scaffolds largelymaintained their integrity is in agreement with a 10-week follow-up study on subcutaneous implantationin rats.13 That study showed that cross-linking of thecollagen increases scaffold stability. Other studies onthe implantation of collagen with different degrees ofcross-linking in the palate of rats or in skin wounds inmice also support this.42,43 In contrast, an intraoralstudy in rabbits showed that cross-linked collagenscaffolds are already largely resorbed after 10 days.27

This was probably due to the different properties ofthe scaffolds used, because they consisted of a cross-linked and a nonYcross-linked collagen layer. Inaddition, the cross-linking was performed with ultra-violet light that generally results in a considerablylower degree of cross-linking, compared with the 1-ethyl-3-(3-dimethyl aminopropyl) carbodiimidemethod.44

Vascularization is an essential process for theintegration of an implant. In our experiment, it startedafter 1 week. Studies using injectable bovine dermalcollagen in human skin showed no vascularization upto 13 weeks after implantation.45,46 This differencemight be caused by the porous structure of our collagenscaffolds. Our findings differ from those of Pieperet al,13 who used the same scaffolds as they describeangiogenesis mainly between 2 and 4 weeks. A fasteringrowth of blood vessels might be expected on the


606


palate compared to the skin, because oral woundhealing, in general, seems to proceed faster.23 After4 weeks, the number of blood vessels started to declinein our study, indicating the return to normal values.

Myofibroblasts invading the scaffolds mightcause their contraction similar to open wounds. Weonly found a transient influx of myofibroblastsbetween 1 and 2 weeks, including those of the shamgroups, which is in accordance with a study oncollagen scaffolds implanted in the palate of dogs.47

Myofibroblasts are related to both wound contractionand scar formation.48,49 In palatal wounds of rats,myofibroblasts are mainly present between 4 and 22days, with a peak intensity after 1 week.33 On purelymorphologic grounds, Dabelsteen and Kremenak50described the presence of myofibroblasts in openpalatal wounds in dogs between 1 and 2 weeks. Also,in skin wounds, myofibroblasts are abundantlypresent up to 2 weeks after wounding.51,52 In general,a comparable influx of myofibroblasts is found instudies on open wounds and on implanted collagenscaffolds. However, a collagen scaffold might inducea more random orientation of myofibroblasts, andthus a more random pattern of newly formed collagenfibers.

In conclusion, our study shows that cross-linkedcollagen scaffolds are highly biocompatible aftersubmucoperiosteal implantation. The scaffolds eli-cited only a mild and transient inflammatoryresponse. They were gradually integrated within thehost tissue, whereas the structure of the scaffoldsremained stable for at least 16 weeks. The scaffoldsseem to guide the newly formed collagen into a morerandom organization. The next step is to furtherinhibit the differentiation of myofibroblasts andsubsequent scar formation by loading the collagenscaffolds with growth factors.

REFERENCES

1. Murray JC. Face factsVgenes, environment, and clefts. Am JHum Genet 1995;57:227Y232

2. Wijdeveld MG, Grupping EM, Kuijpers-Jagtman AM, et al.Wound healing of the palatal mucoperiosteum in beagle dogsafter surgery at different ages. J Craniomaxillofac Surg 1987;15:51Y57

3. Ishikawa H, Nakamura S, Misaki K, et al. Scar tissuedistribution on palates and its relation to maxillary dentalarch form. Cleft Palate Craniofac J 1998;35:313Y319

4. Ross RB. Treatment variables affecting facial growth incomplete unilateral cleft lip and palate. Cleft Palate J 1987;24:5Y77

5. Liao YF, Mars M. Long-term effects of palate repair oncraniofacial morphology in patients with unilateral cleft lipand palate. Cleft Palate Craniofac J 2005;42:594Y600

6. Pieper JS, Hafmans T, van Wachem PB, et al. Loading ofcollagenYheparan sulfate matrices with bFGF promotes angio-

genesis and tissue generation in rats. J Biomed Mater Res2002;62:185Y194

7. Ruszczak Z, Friess W. Collagen as a carrier for on-site deliveryof antibacterial drugs. Adv Drug Deliv Rev 2003;55:1679Y1698

8. Klopper PJ. Collagen in surgical research. Eur Surg Res 1986;18:218Y223

9. DeLustro F, Smith ST, Sundsmo J, et al. Reaction to injectablecollagen: results in animal models and clinical use. PlastReconstr Surg 1987;79:581Y594

10. Keefe J, Wauk L, Chu S, et al. Clinical use of injectable bovinecollagen: a decade of experience. Clin Mater 1992;9:155Y162

11. Piez KA. Biosynthesis of alpha chains of collagen. J Cell Phys1970;76:389Y398

12. McPherson JM, Wallace DG, Sawamura SJ, et al. Collagenfibrillogenesis in vitro: a characterization of fibril qualityas a function of assembly conditions. Coll Relat Res 1985;5:119Y135

13. Pieper JS, van Wachem PB, van Luyn MJA, et al. Attachment ofglycosaminoglycans to collagenous matrices modulates thetissue response in rats. Biomaterials 2000;21:1689Y1699

14. Mikos AG, McIntire LV, Anderson JM, et al. Host response totissue engineered devices. Adv Drug Deliv Rev 1998;33:111Y139

15. HardinYoung J, Carr RM, Downing GJ, et al. Modification ofnative collagen reduces antigenicity but preserves cell compa-tibility. Biotechnol Bioeng 1996;49:675Y682

16. Shoshan S, Riebenfeld U, Neuman Z. Effect of enrichedcollagen solutions on scar formation in 3rd-degree burns inanimals after early surgical excision of slough. Burns 1977;3:153Y165

17. de Vries HJ, Zeegelaar JE, Middelkoop E, et al. Reduced woundcontraction and scar formation in punch biopsy wounds.Native collagen dermal substitutes. A clinical study. Br JDermatol 1995;132:690Y697

18. Lee HG, EunHC. Differences between fibroblasts cultured fromoral mucosa and normal skin: implication to wound healing.J Dermatol Sci 1999;21:176Y182

19. Stephens P, Davies KJ, Occleston N, et al. Skin and oralfibroblasts exhibit phenotypic differences in extracellular matrixreorganization and matrix metalloproteinase activity. Br JDermatol 2001;144:229Y237

20. Stephens P, Davies KJ, Al-Khateeb T, et al. A comparison of theability of intra oral and extra oral fibroblasts to stimulateextracellular matrix reorganization in a model of woundcontraction. J Dent Res 1996;75:1358Y1364

21. Okazaki M, Yoshimura K, Uchida G, et al. Elevated expressionof hepatocyte and keratinocyte growth factor in culturedbuccal-mucosaYderived fibroblasts compared with normal-skinYderived fibroblasts. J Dermatol Sci 2002;30:108Y115

22. Sciubba JJ, Waterhouse JP, Meyer J. Fine-structural comparisonof healing of incisional wounds of mucosa and skin. J OralPathol Med 1978;7:214Y226

23. Hakkinen L, Uitto VJ, Larjava H. Cell biology of gingivalwound healing. Periodontol 2000 2000;24:127Y152

24. Nooh N, Graves DT. Healing is delayed in oral compared todermal excisional wounds. J Periodontol 2003;74:242Y246

25. van Beurden HE, Von den Hoff JW, Torensma R, et al.Myofibroblasts in palatal wound healing: prospects for thereduction of wound contraction after cleft palate repair. J DentRes 2005;84:871Y880

26. Wijdeveld MG, Maltha JC, Grupping EM, et al. A histolo-gical study of tissue response to simulated cleft palatesurgery at different ages in beagle dogs. Arch Oral Biol 1991;36:837Y843

27. Fujioka M, Fujii T. Maxillary growth following atelocollagenimplantation on mucoperiosteal denudation of the palatalprocess in young rabbits: implications for clinical cleft palaterepair. Cleft Palate Craniofac J 1997;34:297Y308

28. Pieper JS, Oosterhof A, Dijkstra PJ, et al. Preparation andcharacterization of porous crosslinked collagenous matrices


607


containing bioavailable chondroitin sulphate. Biomaterials1999;20:847Y858

29. Damoiseaux JG, Dopp EA, Calame W, et al. Rat macrophagelysosomal membrane antigen recognized by monoclonal anti-body ED1. Immunology 1994;83:140Y147

30. Szpaderska AM, Zuckerman JD, DiPietro LA. Differentialinjury responses in oral mucosal and cutaneous wounds. JDent Res 2003;82:621Y626

31. Bodner L, Dayan D. Epithelium and connective tissue regen-eration during palatal wound healing in desalivated ratsVacomparative study. Comp Biochem Physiol A Physiol 1995;111:415Y419

32. Minabe M, Kodama T, Hori T, et al. Effects of atelocollagen onthe wound-healing reaction following palatal gingivectomy inrats. J Periodontal Res 1989;24:178Y185

33. Cornelissen AM, Stoop R, Von den Hoff HW, et al. Myofibro-blasts and matrix components in healing palatal wounds in therat. J Oral Pathol Med 2000;29:1Y7

34. Cornelissen AM, Maltha JC, Von den Hoff HW, et al. Palatalmucoperiosteal wound healing in the rat. Eur J Oral Sci 1999;107:344Y351

35. Leenstra TS, Kuijpers-Jagtman AM, Maltha JC. The healingprocess of palatal tissues after palatal surgery with and withoutimplantation of membranes: an experimental study in dogs. JMater Sci Mater Med 1998;9:249Y255

36. In de Braekt MM, Maltha JC, Kuijpers-Jagtman AM. Poly-(L-lactic) acid membranes in palatal surgery in beagle dogs:clinical and histologic evaluation. Cleft Palate Craniofac J 1995;32:290Y298

37. Minuth WW, Sittinger M, Kloth S. Tissue engineering: genera-tion of differentiated artificial tissues for biomedical applica-tions. Cell Tissue Res 1998;291:1Y11

38. Bergsma EJ, Rozema FR, Bos RRM, et al. Foreign-body reactionsto resorbable poly(l-lactide) bone plates and screws used for thefixation of unstable zygomatic fractures. J Oral Maxillofac Surg1993;51:666Y670

39. Chang C, Houck JC. Demonstration of the chemotactic proper-ties of collagen. Proc Soc Exp Biol Med 1970;134:22Y26

40. Houck J, Chang C. The chemotactic properties of the productsof collagenolysis. Proc Soc Exp Biol Med 1971;138:69Y75

41. Postlethwaite AE, Seyer JM, Kang AH. Chemotactic attractionof human fibroblasts to type I, II, and III collagens and collagen-derived peptides. Proc Natl Acad Sci U S A 1978;75:871Y875

42. Minabe M, Kodama T, Kogou T, et al. Different cross-linkedtypes of collagen implanted in rat palatal gingiva. J Periodontol1989;60:35Y43

43. Harriger MD, Supp AP, Warden GD, et al. Glutaraldehydecrosslinking of collagen substrates inhibits degradation in skinsubstitutes grafted to athymic mice. J Biomed Mater Res 1997;35:137Y145

44. Weadock K, Olson RM, Silver FH. Evaluation of collagencrosslinking techniques. Biomater Med Devices Artif Organs1983;11:293Y318

45. Stegman SJ. Injectable collagen. Plast Reconstr Surg 1987;80:86646. Kligman AM, Armstrong RC. Histologic response to intrader-

mal zyderm and zyplast (glutaraldehyde cross-linked) collagenin humans. J Dermatol Surg Oncol 1986;12:351Y357

47. Ophof R,Maltha JC, Von denHoff JW, et al. Histologic evaluationof skin-derived and collagen-based substrates implanted inpalatal wounds. Wound Repair Regen 2004;12:528Y538

48. Gabbiani G. The myofibroblast in wound healing andfibrocontractive diseases. J Pathol 2003;200:500Y503

49. Tomasek JJ, Gabbiani G, Hinz B, et al. Myofibroblasts andmechano-regulation of connective tissue remodelling. Nat RevMol Cell Biol 2002;3:349Y363

50. Dabelsteen E, Kremenak CR. Demonstration of actin infibroblasts of healing palatal wounds. Plast Reconstr Surg1978;62:429Y435

51. Greenhalgh DG. The role of apoptosis in wound healing. Int JBiochem Cell Biol 1998;30:1019Y1030

52. Huang NF, Zac-Varghese S, Luke S. Apoptosis in skin woundhealing. Wounds 2003;15:182Y194


608

collagen scaffolds implanted in the palatal mucosa · cross-linked collagen scaffolds and showed a...

Documents