| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
From the Department of Surgery,* Division of Plastic Surgery, University of California, San Francisco, San Francisco, California; the Craniofacial and Skeletal Diseases Branch, National Institute of Dental and Craniofacial Research, National Institutes of Health, Bethesda, Maryland; the Materials Sciences Division, Lawrence Berkeley National Laboratory, and Department of Materials Science and Engineering, University of California, Berkeley, Berkeley, California; and the Department of Biomedical Engineering, State University of New York at Stony Brook, Stony Brook, New York
 |
Abstract |
|---|
 Top Abstract Materials and MethodsResultsDiscussionReferences |
|---|
Successful repair of calvarial defects in patients, whether arising from trauma, tumor resection, or congenital disorders, continues to be a major concern to reconstructive surgeons. The transplantation of bone autograft is the preferred method for reconstructing these areas when they are limited in size. When the volume of the defect is sizable, however, autograft often proves insufficient. Surgeons rely on bone graft extenders or alternatives, including osteoconductive matrices such as Interpore and MedPore. Unfortunately, these matrices experience only limited tissue incorporation and can suffer from significant rates of fracture, migration, and infection. As a result, a need exists for a technique for creating new bone for patients with sizable defects.
Our group has previously demonstrated the feasibility of closing calvarial defects in mice with isogeneic mouse BMSCs in conjunction with collagen matrices15 or with human BMSCs in conjunction with hydroxyapatite/tricalcium phosphate (HA/TCP) particles.16 However, no study has yet demonstrated the practicality of creating such BMSC transplants as large as those needed clinically. In this study we attempted to close calvarial defects in a large animal model using autologous BMSCs. The aims of this study were to: 1) demonstrate the feasibility of closing critical-size calvarial defects of a size comparable to those seen in patients; 2) demonstrate the practicality of scaling up cell processing and transplant preparation procedures; 3) create large enough transplants to complete noninvasive evaluations of new bone formation; and 4) create large enough transplants to complete biomechanical studies. BMSCs were harvested from healthy dogs, expanded in tissue culture, attached to HA/TCP particles, and autotransplanted into critical-sized calvarial defects. The radiographical, histological, and mechanical characteristics of the transplants were evaluated.
|
Materials and Methods |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
Bone marrow was harvested from the distal femur of six 9-month-old male mongrel dogs, in accordance with an approved NIH animal protocol (97-031). Multicolony-derived strains of BMSCs were obtained from the bone marrow in a manner previously described.17
Briefly, bone marrow cells were cultured in growth medium consisting of 
-minimal essential medium (Invitrogen, Grand Island, NY), 2 mmol/L L-glutamine, 100 U/ml penicillin, 100 µg/ml streptomycin sulfate (Biofluids, Rockville, MD), 10–8 mol/L dexamethasone (Sigma, St. Louis, MO), 10–4 mol/L L-ascorbic acid phosphate magnesium salt n-hydrate (Wako, Osaka, Japan), and 20% fetal bovine serum of a preselected lot (Equitech-Bio, Kerrville, TX). Cells were cultured at 37°C in an atmosphere of 100% humidity and 5% CO2.
Trypsin-released cells from passages three or four were pipetted into 50-ml polypropylene Falcon tubes (Becton, Dickinson, and Co., Franklin Lakes, NJ). Separately, HA/TCP particles (Zimmer, Inc., Warsaw, IN) of size range 0.5 to 1.0 mm were isolated using a sieve shaker (CSC Scientific, Fairfax, VA). Between 50 x 106 and 70 x 106 BMSCs were allowed to attach to each gram of HA/TCP particles. Control transplants consisted of HA/TCP particles that were moistened with growth medium but received no cells. The BMSC-free transplants were incubated with the same growth medium (including 20% fetal bovine serum) and for the same length of time (90 minutes) as that used for the BMSC transplants. The mixtures were incubated for 90 minutes at 37°C on a slowly rotating platform. After centrifugation of the mixtures at 1200 rpm for 60 seconds, the supernatant was discarded.
Each of the six dogs underwent creation of critical-sized bilateral fronto-parietal cranial defects via a bi-coronal scalp incision. Care was taken to elevate the periosteum for a distance of at least 1 cm from the defect. Each defect measured 35 mm in diameter, purposely chosen to be larger than the critical-size defect of 20 mm.18
In each dog, one defect was filled the BMSC + HA/TCP mixture and the contralateral defect was filled with control HA/TCP particles (Figure 1)
. Approximately 2 g of the appropriate HA/TCP mixture was needed for each defect.
|
. The extent of bone union between calvarial margin and the transplant was analyzed using histomorphometric techniques.
|
Colony-Forming Efficiency Assay
As part of in vitro testing, small subsets of the original bone marrow single cell suspensions were prepared for the determination of BMSC colony-forming efficiency (CFE) at the time of primary plating. For four of the animals, each of four 25-cm2 flasks was plated with 1 x 105 nucleated bone marrow cells. These were incubated at 37°C in an atmosphere of 100% humidity and 5% CO2. After 10 days, the flasks were rinsed with Hanks’ balanced salt solution (Invitrogen), the cells fixed with absolute methanol, and the adherent BMSC colonies stained with saturated aqueous solution of methyl violet (Sigma). Fibroblast colonies of greater than 50 cells were counted under a stereomicroscope and the CFE, or the number of colonies per 1 x 105 nucleated bone marrow cells, was calculated.
Noninvasive Monitoring of Transplants Using Quantitative Computed Tomography (qCT)
From 2 weeks to 18 months after transplantation, the dogs individually underwent CT scanning in a GE CTI (GE Medical Systems, Milwaukee, WI) at an energy of 80 kVp/200 mA, a slice thickness 1 mm, and an in-slice resolution of 0.625 mm. The 2-week time points represented baseline information about the transplants, whereas the later time points were timed to closely match tissue sampling. Each scan included a Mindways Model 2 phantom (Mindways Software, South San Francisco, CA). The bone mineral density (BMD) of each image slice of each transplant was obtained using QCT Pro v. 2.0.3 (Mindways Software) on a Dell XPS R450 (Dell Computer Corp., Round Rock, TX). BMD values were expressed in terms of mg/cc of K2HPO4 in distilled water, in which a BMD of 0 corresponded to the density of distilled water alone (no additional K2HPO4) and a BMD of >0 corresponded to nonaerated biological tissue.
Each transplant was represented by 7 to 27 slices, each 0.488 mm thick and bounded by an oval-shaped region of interest (ROI). The software provided the ROIs, whose size and shape could be determined by the investigator. In each slice of a transplant, the ROI was adjusted to match the size and shape of the transplant, which was typically round in cross-section. Because the ROI and the transplant silhouette were well matched to each other, the ROI encompassed nearly the entire (>90%) area of the transplant at each slice. The BMD of the center seven slices of a transplant, encompassing the center 2.9 cm of the transplant, were pooled; the mean of these individual BMD values was used as the overall BMD value for the transplant. All BMD values were calculated by the primary investigator (M.H.M.), who was blinded to the bone scores of the transplants.
In Vivo Noninvasive Ultrasound Testing of Transplants
We used an experimental ultrasound system that is capable of generating a noninvasive, high-resolution ultrasound wave to the calvarium of the dog.19-21
To transmit a signal to the calvarium, ultrasound transducers were designed in the frequency range of 0.5 MHz to 1 MHz. To obtain high-resolution signals associated with bone structure, the ultrasound signal was focused as a beam cylinder in a zone of 
0.5 to 1.0 mm in diameter and 2 to 3 mm in length. The transmitter was driven by pulse signals; these passed through bone and were then received by the receiver and amplifier unit and digitized at 25 MHz. The digitized waveform was then transferred to a personal computer (Dell Latitude, Austin, TX) for analysis. The control software was written using C++.
Reflect mode was used during the measurements, in which the waveform reflected from the bones’ periosteal and endosteal surfaces was captured by the same transducer. The time delay 
t reflects the temporal difference between acquisition of the reference signal and the bone signal, and reflects in part the velocity difference between the bone specimen and the coupling gel. The ultrasound wave velocity Vb in trabecular bone was calculated using the time of flight method, in which the thickness of the specimen w and ultrasound wave velocity in water Vw are known.
 |
 |
Ex Vivo Mechanical Testing of Transplants
A specially designed mechanical testing protocol was used to quantify bone strength in the control and experimental groups to characterize the differences in stiffness between BMSC-containing and BMSC-free transplants. From all six dogs, the entire calvarium, including transplant and surrounding intact normal calvarium, were harvested as a unit. Tissues were fixed in 4% paraformaldehyde for 24 hours, followed by storage in 1x phosphate-buffered saline (PBS) at –20°C until actual testing. Strips measuring 5 mm in width were cut from the center of each specimen; these incorporated normal calvarium at each end and transplant material in between. The strips were further sectioned using a slow (100 rpm) speed saw (TechCut II; Allied High Tech Products, Inc., Rancho Dominguez, CA) into beam-shaped specimens measuring 3 mm x 4 mm x 6.5 mm in size. The soft tissues were operatively removed from the calvarium and transplant. Such specimens were taken from three sites: 1) the margin of the transplant, near its interface with the normal calvarium; 2) the center section of the transplant; and 3) the normal calvarium at the ends. Before actual testing, the specimens were thawed for 1 hour and then soaked in PBS for 1 hour. The mechanical testing for measuring the elastic (compressive) modulus was then conducted in ambient air (25°C, 20 to 40% relative humidity), with the specimens being loaded in displacement control using a standard servo-hydraulic testing machine (MTS 810; MTS Systems Corp., Eden Prairie, MN) with a loading rate 0.015 mm/second (to simulate quasi-static loading conditions). The elastic (Young’s) modulus was then calculated from the load-displacement data obtained during such tests, with care being taken to eliminate the initial toe-in region. Three such measurements were made on each specimen. The operator was blinded to whether each specimen came from the experimental or control group. The data obtained were analyzed statistically using GraphPad InStat version 3.05 (GraphPad Software, San Diego, CA).
|
Results |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
Transplant Morphology
All calvaria exhibited good adherence of the overlying temporalis musculature to the transplant, at the time of both biopsy and transplant harvest. Examination of the transplants demonstrated a cohesive structure, with HA/TCP particles bound to each other throughout the transplant; no loose HA/TCP particles were noted in any transplant. Particles were found within the confines of the transplant, with no gross evidence of particle migration. Examination of the brain after calvarial removal showed the dura mater to be intact, and no particles were found deep to the dura. BMSC transplants were rigid on gross palpation and were firmly adhered to the adjacent normal calvarial margin. In contrast, BMSC-free transplants were not rigid and could be flexed with digital pressure while in situ. They exhibited a nonrigid binding to the surrounding normal calvarial edge.
Histologically, transplants without BMSCs contained aggregates of HA/TCP particles separated by fibrovascular tissue (Figure 2)
. Fibrovascular tissue also lay at the interface between the transplant and adjacent normal calvarium. In contrast, BMSC transplants contained abundant new lamellar cortical bone firmly attached to both particles and the adjacent calvarium (Figure 2)
. In transplants with more extensive bone (bone scores 
3), bone was accompanied by hematopoietic tissue and occasional adipocytes, all of which were spatially associated with the HA/TCP particles. Adipocytes were seen frequently among BMSC transplants, as visualized in Figure 2, B and C
. Very few adipocytes were seen in the BMSC-free transplants and were only seen in conjunction with new bone. Innervation of the transplants was not seen using standard histological techniques, including hematoxylin and eosin staining. There was abundant evidence of vascularization of both BMSC-free and BMSC transplants; gross examination of the transplants at the time of harvest suggested that vascularization was greater among the BMSC transplants.
|
Among the first set of three dogs, both calvarial sides were biopsied at 2 to 3 months after transplantation. Bone scores among transplants on the control side ranged from 0 to 2.0 (mean, 1.33) and those transplants in defects filled with BMSCs had bone scores of 2.0 to 3.0 (mean, 2.33) (Figure 3)
. At the time of harvest at 18 to 20 months, bone scores ranged from 2.0 to 3.0 (mean, 2.33) among BMSC-free transplants and 3.7 to 4.0 (mean, 3.90) among BMSC-inclusive transplants (P < 0.05). The second set of three dogs, harvested at 6 to 7 months, had bone scores of 3.0 to 3.7 (mean, 3.23) among BMSC-free transplants and scores of 3.3 to 4.0 (mean, 3.77) among BMSC transplants (Figure 3)
. Comparison of all BMSC-free transplants with all BMSC transplants demonstrated significant differences in cumulative bone scores (mean, 2.30 versus 3.33; P < 0.01).
|
. We attribute the better bone formation in BMSC transplants among the second set of dogs to our greater facility in handling these large volumes of cells and in placing these transplants in the animals. Bone Union
The degree and distribution of bony union, or that union seen between the transplant and surrounding calvarial margin, mirrored the extent of bone formation (Figure 4)
. Bone union could only be evaluated among fully harvested transplants, not transplant biopsies, because biopsy specimens were obtained from the center of the transplants. Bone union ranged from 25 to 50% (mean, 35%) among 6- to 7-month-old control transplants, whereas it ranged from 50 to 100% (mean, 74.3%) among 6- to 7-month-old BMSC transplants (Figure 5)
. Among transplants harvested at 18 to 20 months, bone union ranged from 11 to 33% (mean, 20%) in the control transplants, whereas it ranged from 67 to 94% (mean, 81%) in the BMSC transplants (P < 0.05). Comparison of all BMSC-free transplants with all BMSC transplants demonstrated significant differences in bone union (mean, 27% versus 78%; P < 0.01) (Figure 5)
.
|
|
CT images of the dog calvaria were obtained immediately after transplantation and coincident with the biopsies and calvarial harvests. Each image included a standard phantom for calibration purposes. Even without qCT analysis of the transplants, ossification of the BMSC-rich transplants and resorption of the BMSC-free transplants was evident by 6 months (Figure 6)
. The amount of bone formation within the transplants ranged from a bone score of 0 to a score of 4, and BMD values ranged from 443 to 838 mg/cc K2HPO4. The distribution of BMD relative to bone score is graphically represented in Figure 7
. Seven transplant specimens showed no or poor bone formation (bone scores of 0, 1, or 2), whereas the remaining 11 transplant specimens formed appreciable amounts of bone (bone scores of 3 or 4). A significant positive trend in BMD was detected (P < 0.0001) as a function of clinical bone score. The estimated relationship is given as BMD = 100.6 x (bone score) + 369.3 with correlation coefficient (r) = 0.8174.
|
|
3) bone formation among transplants with or without BMSCs (Figure 8)
. The BMD among BMSC-free transplants with either poor or appreciable bone formation averaged 501 and 681 mg/cc K2HPO4, respectively (P < 0.05), whereas the BMD among BMSC transplants with either poor or appreciable bone formation averaged 589 and 763, respectively (P < 0.005). When BMSC-free and BMSC-inclusive transplants were pooled, the mean BMD of poor bone formers and good bone formers was 526 and 733, respectively (P < 0.0005).
|
3) bone formation (Mankani MH, Kuznetsov SA, Robey PG, unpublished data), then 15 of the 18 specimens corroborated this relationship (Figure 7)
. The three exceptions included two transplants with bone score of 2 and BMD of 619 and 631, respectively, and one transplant with bone score of 3 and BMD of 580. In Vivo Ultrasound Testing of Transplants
Transplants from three of the animals underwent ultrasound testing before sacrifice. In each of the animals, there was a slight but not statistically significant difference in calculated elastic modulus E between the BMSC-containing transplant and the transplants that were free of BMSCs (Figure 9)
.
|
Transplants from all six animals underwent mechanical testing. In each of the animals, the BMSC-containing transplant was found to have both a higher bone score and elastic modulus than the control tissue from the same animal. Bone scores among BMSC-free and BMSC-containing tissues were 2.30 and 3.33, respectively (P < 0.05), whereas the overall Young’s modulus at the center of the same transplants was 0.24 and 0.59 GPa, respectively (P < 0.001) (Figure 10)
. The overall modulus of the tissue at the margins of the transplants was 0.24 and 0.90 GPa, respectively (P < 0.005). In contrast, the modulus for normal calvarial bone among the six dogs averaged 1.85 (±0.29) GPa.
|
|
Discussion |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
In this study, we transplanted culture-expanded autologous dog BMSCs into sizable calvarial defects, defects that were far larger than would be expected to spontaneously close. The cells had been combined with an osteoconductive ceramic, HA/TCP, just before transplantation. Each dog also received a HA/TCP transplant devoid of BMSCs in an identical contralateral calvarial defect. These BMSC-free HA/TCP transplants are a good control because they are commonly used by craniofacial surgeons to close bone defects and represent the standard of care. From periods of 2 to 20 months after transplantation, dogs were evaluated using histology, CT scanning, and ultrasound. The BMSC-containing transplants formed bone faster and more extensively than transplants containing only the HA/TCP osteoconductive matrix. The newly formed bone developed a union with the adjacent margin of calvarium, and the extent of bone union was significantly greater among the BMSC-containing transplants. BMSC-containing transplants maintained their cohesiveness, such that particles did not migrate from the transplant into the surrounding tissues. The transplants maintained their shape and size despite the absence of mechanical stimuli that promote remodeling and exhibited signs of continued new bone formation as late as 20 months. In contrast, BMSC-free transplants demonstrated zones of particle resorption and migration. Noninvasive testing of the transplants with qCT demonstrated that BMD increased in proportion to bone score, and that, with infrequent exception, a BMD of 600 served as a reliable threshold distinguishing poor bone formation from significant bone formation. Ex vivo mechanical testing of the transplants demonstrated that those with BMSCs had mechanical properties more closely matching that of normal bone than the BMSC-free transplants; additionally, such differences could be detected noninvasively with ultrasound.
This study was designed to closely prepare for a clinical trial. Dogs were specifically selected as the research subject because of their well-established history in bone graft studies, their size, and the characteristics of their BMSCs. Surgically induced calvarial defects in dogs are a well-established model for evaluating bone regeneration as well as the success of bone-grafting materials.18,24,25 The diameter of the calvarial defects created in this study is far larger than that which would be expected to spontaneously close. The overall size of the animals was also important, because it allowed for the nondebilitating harvest of bone marrow in amounts comparable to that expected clinically. Animal size also permitted CT scanning, qCT analysis, and ultrasound testing, which closely parallels that conducted in patients, and the size of the transplants was substantial enough to permit meaningful mechanical studies of the mature transplants. Dog BMSCs have characteristics that parallel those of human BMSCs. Their CFE values have been reported to range from 2 to 10 per 105 nucleated marrow cells,26,27 not greatly different from the CFE of 81 per 105 among the dogs in this study and close to the CFE of 21 to 77 per 105 of normal human BMSCs.28,29 Transplants of dog BMSCs combined with HA/TCP particles form cortico-cancellous bone when placed in immunocompromised mice, but transplants of dog BMSCs combined with bovine collagen sponges fail to form bone (Mankani MH, Kuznetsov SA, Robey PG, unpublished data); these findings closely match our experience with human BMSCs and are quite distinct from the characteristics of other species’ BMSCs.8 As would be planned clinically, autologous BMSCs were used in this study. Our experience with human and mouse BMSC transplants placed in mice suggest that the new bone found in these transplants originates from the implanted BMSCs and the hematopoietic tissues originate from the recipient.8,15,23 Although the transplanted BMSCs in this study were not labeled and therefore could not be tracked after transplantation, the profound differences in bone formation between BMSC-free and BMSC-containing transplants suggests that BMSCs were responsible for bone formation where they were placed.
A semiquantitative scale of bone formation was used in this study. When the bone scores reported on this scale have been compared to histomorphometric measurements of tissue sections, a correlation was previously observed between the bone score and the square root of the fraction of bone area to total transplant area (B/T) (r = 0.973).30
Bone scores of 0, 1, 2, 3, and 4 correlated with fractional bone areas of 0%, 1%, 3%, 8%, and 14%, respectively. The poor bone formation described in this study would be characterized by a fractional bone area of 
3%, whereas appreciable bone formation correlated with a fractional bone area of 8%. Thus, the differences in bone score noted between BMSC-free and BMSC-containing transplants parallels significant differences in transplant fractional bone area B/T.
Transfer of the BMSC transplantation techniques to patients requires the development of a noninvasive method for evaluating the extent of new bone formation among these transplants. HA/TCP has a high density that is comparable to that of cortical bone. Use of plain radiographs is insufficient because they fail to distinguish HA/TCP transplants that have no associated bone formation from those that do (Mankani MH, Kuznetsov SA, Robey PG, unpublished data). As an alternative to radiographs, we have previously demonstrated that the utilization of qCT techniques could be adapted to BMSC-HA/TCP transplants to describe the extent of bone formation within the transplants.31 Among transplants of human BMSCs in mice, we noted that a BMD of 600 mg/cc K2HPO4 serves to distinguish poor bone formation (bone scores of 0 to 2) from appreciable bone formation (bone scores of 3 or 4). In the present dog study, we note that all but 3 of 18 specimens in this large animal model corroborate findings from our mouse study, regardless of whether the transplants used BMSCs and regardless of their harvest date. Care was also taken to specifically use a clinical CT scanner rather than a unit used for investigational purposes. Although a research scanner such as MicroCT would undoubtedly have provided much better resolution and a finer slice thickness of these dog transplants, and would therefore have described the bone architecture better, it was important to us to demonstrate the feasibility of using qCT with clinically appropriate equipment and settings.
Transcutaneous ultrasound testing of bone is a relatively new and novel method for assessing the mechanical characteristics of this tissue.19,20,32 This represents the first report of its use to estimate the stiffness of bone-containing calcium phosphate transplants. Although this technique discerned a difference in modulus between the transplant types, the Young’s modulus values only approximately coincided with those of the gold standard study, ex vivo mechanical testing. We attribute the modest correlation with ex vivo testing to the heterogeneity of the transplants and to the likelihood that the areas sampled by ultrasound did not necessarily coincide with the areas evaluated ex vivo.
In summary, we have used an established bone defect model to demonstrate the feasibility of using large volume autologous BMSC transplants to close critical-sized calvarial defects, and we have confirmed that the extent of bone formation can be reliably estimated using noninvasive techniques. The success of such large transplants increases the practicality of transplanting culture-expanded BMSCs in patients.
|
Acknowledgements |
|---|
|
Footnotes |
|---|
Supported by the University of California–San Francisco Research Evaluation and Allocation Committee and the National Institutes of Health (Intramural Research Program of the National Institute of Dental and Craniofacial Research, DHHS; and grant no. 5R01 DE015633 to R.K.N. and R.O.R.).
Present address of B.S.: Sharon Regional Health System, Hermitage, PA.
Accepted for publication October 4, 2005.
|
References |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  S. A. Kuznetsov, M. H. Mankani, A. I. Leet, N. Ziran, S. Gronthos, and P. G. Robey Circulating Connective Tissue Precursors: Extreme Rarity in Humans and Chondrogenic Potential in Guinea Pigs Stem Cells, July 1, 2007; 25(7): 1830 - 1839. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. G. Robey and P. Bianco The use of adult stem cells in rebuilding the human face. J Am Dent Assoc, July 1, 2006; 137(7): 961 - 972. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
