The Effect of Implants Loaded with Autologous Mesenchymal Stem Cells on the Healing of Canine Segmental Bone Defects*

BRUDER, SCOTT P. M.D., PH.D.†, BALTIMORE; KRAUS, KARL H. D.V.M.‡, NORTH GRAFTON, MASSACHUSETTS; GOLDBERG, VICTOR M. M.D.§; KADIYALA, SUDHA PH.D.†, BALTIMORE, MARYLAND

Journal of Bone & Joint Surgery - American Volume:
Article
Abstract

Bone marrow has been shown to contain a population of rare mesenchymal stem cells that are capable of forming bone, cartilage, and other connective tissues. We examined the effect of cultured autologous mesenchymal stem cells on the healing of critical-sized (twenty-one-millimeter-long) segmental defects in the femora of adult female dogs. Autologous mesenchymal stem cells were isolated from bone marrow, grown in culture, and loaded onto porous ceramic cylinders consisting of hydroxyapatite (65 per cent) and β-tricalcium phosphate ceramic (35 per cent). The animals were randomly assigned to one of three groups. In Group A (six dogs), a porous ceramic cylinder that had been loaded with autologous mesenchymal stem cells was implanted in the defect. In Group B (six dogs), a ceramic cylinder that had not been loaded with cells was placed in the defect. In Group C (three dogs), the defect was left untreated (no ceramic cylinder was implanted). Radiographs were made immediately after the operation and at four-week intervals. At sixteen weeks, the animals were killed, the involved femora were removed, and undecalcified histological sections from the defects and adjacent bone were prepared. Histological and histomorphometric studies were carried out to examine the healing of the defects and the formation of bone in and around the ceramic implants. Atrophic non-union occurred in all of the femora that had untreated defects, and only a small amount of trabecular bone formed at the cut ends of the cortex of the host bone in this group. In contrast, radiographic union was established rapidly at the interface between the host bone and the implants that had been loaded with mesenchymal stem cells. Numerous fractures, which became more pronounced with time, developed in the implants that had not been loaded with cells. Histological and morphometric analyses demonstrated that both woven and lamellar bone had filled the pores of the implants that had been loaded with mesenchymal stem cells; the amount of bone was significantly greater (p < 0.05) than that found in the pores of the implants that had not been loaded with cells. In addition, a large collar of bone (mean maximum thickness, 3.14 millimeters) formed around the implants that had been loaded with cells; this collar became integrated and contiguous with callus that formed in the region of the periosteum of the host bone. The collar of bone remodeled during the sixteen-week period of study, resulting in a size and shape that were comparable with those of the segment of bone that had been resected. Callus did not develop around the cortex of the host bone or around the defect in any of the specimens in the other two groups. CLINICAL RELEVANCE: Autologous cultured bone-marrow-derived mesenchymal stem cells that had been loaded onto porous ceramic cylinders elicited the healing of critical-sized segmental bone defects in dogs. It may be possible to exploit this technology to elicit the healing of bone defects in humans by using cells from bone marrow that has been aspirated from the iliac crest of the patient. This approach may provide an alternative to autologous bone-grafting and may be particularly useful when the number of endogenous mesenchymal stem cells is relatively small.

Author Information

†Osiris Therapeutics, 2001 Aliceanna Street, Baltimore, Maryland 21231. E-mail address for Dr. Bruder: sbruder@osiristx.com. E-mail address for Dr. Kadiyala: skadiyala@osiristx.com. Please address requests for reprints to Dr. Bruder.

‡Department of Surgery, Tufts University School of Veterinary Medicine, 200 Westboro Road, North Grafton, Massachusetts 01536.

§Department of Orthopaedic Surgery, Case Western Reserve University, 11100 Euclid Avenue, Cleveland, Ohio 44106.

Article Outline

The formation and repair of bone requires (1) a source of pluripotential mesenchymal stem cells capable of differentiating into osteoblasts; (2) growth and differentiation factors that direct these cells to migrate into the osseous defect, to proliferate, and to differentiate into osteoblasts; (3) a bioresorbable matrix, or scaffold, to support the attachment and migration of these cells into the osseous defect; and (4) angiogenesis and the formation of a vascular network throughout the newly formed bone.

Several experimental approaches have been utilized to elicit the formation of bone in segmental bone defects and to promote their healing. These approaches have included the implantation of osteoconductive extracellular-matrix scaffolds19,21,26,32,51 and the implantation of bone morphogenetic proteins in various matrices11,15,22,33,41,42,49.

We have developed an entirely different strategy that involves the direct implantation of pluripotential mesenchymal stem cells that have been loaded onto a porous ceramic cylinder2,5,7,12,16,18,28,29,37. This strategy is based on the hypothesis that such an approach decreases or eliminates the need for chemotaxis of osteoblast progenitor cells into the defect and their massive proliferation. One might expect the direct implantation of progenitor cells to lead to more rapid, uniform, and reliable healing of bone defects. A major challenge to this approach has been the identification of the proper type and source of cells for autologous cell therapy.

Progenitor cells with osteoblastic potential have been isolated from a variety of species, including humans, rats, rabbits, and dogs12,18,29,46. These postnatal, bone-marrow-derived cells are referred to as mesenchymal stem cells6 because they possess a high replicative capacity and are able to form bone, cartilage, tendon, muscle, fat, and hematopoietic-supportive stroma3,18,25,27,30,31,38,45,50. We previously used syngeneic mesenchymal stem cells to regenerate bone in critical-sized segmental defects of the femur in rats28. After the implantation of the cells on a porous cylinder of hydroxyapatite and β-tricalcium phosphate, new bone formed throughout the body of the implant by the direct conversion of mesenchymal stem cells into osteoblasts, without progression through a cartilaginous intermediate. That study showed that mesenchymal stem cells were superior to fresh bone marrow (also delivered on a ceramic cylinder)28. Experiments from our laboratories also have demonstrated that mesenchymal stem cells obtained from human bone marrow can elicit healing of critical-sized defects5. These and other preclinical studies2,5,7,8,28,40,46,50 provide the foundation for the study of mesenchymal stem-cell-based repair of musculoskeletal tissue defects in large-animal models.

The purpose of the current study was to evaluate the ability of cultured autologous mesenchymal stem cells to elicit repair at the site of a critical-sized segmental defect in the canine femur. A model for the resection of an osteoperiosteal segment and stabilization of the limb was developed, and the defect was left untreated (no cylinder was implanted), was filled with a ceramic cylinder that had been loaded with cultured autologous mesenchymal stem cells, or was filled with a ceramic cylinder that had not been loaded with cells. Bone formation and healing at the site of the defect were evaluated with use of radiographic, histological, and histomorphometric techniques.

Back to Top | Article Outline

Materials and Methods

Experimental Design

Fifteen skeletally mature female hounds were used in the present study. The dogs were housed at a facility accredited by the American Association for the Accreditation of Laboratory Animal Care, and all procedures were performed at that facility. The mean weight of the animals was 20.3 ± 1.1 kilograms, and the mean diameter of the mid-portion of the femoral diaphysis, as determined on radiographs, was 13.9 ± 0.6 millimeters. All of the animals had a unilateral resection of a twenty-one-millimeter-long osteoperiosteal segment of the femoral diaphysis, as will be described. The unique identification number for each dog was entered into a computer with use of software that randomly assigned each dog to one of three groups. In Group A (six dogs) the defect was filled with a porous ceramic cylinder that had been loaded with cultured autologous mesenchymal stem cells, in Group B (six dogs) the defect was filled with a ceramic cylinder that had not been loaded with cells, and in Group C (three dogs) the defect was left untreated (no ceramic cylinder was implanted). Bone marrow was collected for the isolation of mesenchymal stem cells from all animals in order to control for any systemic osteogenic response, as has been reported to occur in some animals after the aspiration of bone marrow1,14. The dogs were housed in individual pens, were provided with water and commercial dog food, and were allowed unrestricted weight-bearing and activity as tolerated postoperatively. Validation of the in vivo and in vitro osteogenic potential of mesenchymal stem cells from every dog was performed according to established techniques29. Radiographs were made postoperatively and at four-week intervals. At sixteen weeks, the animals were killed and the involved femora were removed for radiographic and radiomorphometric analysis. Specimens were subsequently processed for histological and histomorphometric evaluation.

Back to Top | Article Outline
Isolation and Cultivation of Canine Mesenchymal Stem Cells

Bone marrow was aspirated from the dogs sixteen days before creation of the segmental defect. The animals were anesthetized and, under sterile conditions, a 15-gauge Jamshidi needle was used to aspirate nine milliliters of bone marrow into a syringe containing one milliliter of heparinized (1000 units per milliliter) saline solution. The aspirates were placed in glass tubes, packed in insulated boxes containing a frozen cold pack, and sent by overnight courier service for delivery to the cell-culture facility on the next day.

When the aspirates arrived at the cell-culture facility, the mesenchymal stem cells were isolated from bone marrow and grown in culture as previously described29. In brief, the bone marrow was mixed with two volumes of complete medium consisting of 10 per cent fetal bovine serum (Gibco, Grand Island, New York) from a selected lot29,30, penicillin G (100 units per milliliter), streptomycin sulfate (100 micrograms per milliliter), and amphotericin B (0.25 microgram per milliliter) in low-glucose Dulbecco modified Eagle medium. After the cells had been counted, the nucleated-cell fraction of the bone marrow was enriched for mesenchymal stem cells by density gradient centrifugation over a Percoll cushion (1.073 grams per milliliter; Pharmacia Fine Chemical, Piscataway, New Jersey). The cells at the medium-Percoll interface then were collected, washed, and plated in T-185 culture flasks at 107 cells per flask. The cells were incubated at 37 degrees Celsius in a humidified 5 per cent carbon-dioxide environment. On the fourth day of culture, the non-adherent cells were removed along with the culture medium. Fresh medium was added to the adherent cells twice a week, and the cells were passaged on the ninth, tenth, or eleventh day by replating at 8 x 103 cells per square centimeter. Between the thirteenth and fifteenth days, the culture flasks were returned by commercial airline and ground transportation to the operative team at the animal-care facility. On arrival, the flasks were placed in a standard tissue-culture incubator, where they were maintained until the time of preparation of the ceramic implants on the sixteenth day of culture.

Back to Top | Article Outline
Preparation of the Ceramic Implants

Blocks of porous ceramic consisting of hydroxyapatite (65 per cent) and β-tricalcium phosphate (35 per cent), with a mean pore size of 200 to 450 micrometers, were provided by Zimmer (Warsaw, Indiana). The implants were prepared with a modification of methods that have been described previously12,28,29. In brief, the blocks of ceramic were cut into cylinders that were fourteen millimeters in diameter and twenty-one millimeters in length. A central canal, eight millimeters in diameter, was bored manually through the length of the entire cylinder with use of a series of needles and drill-bits. The cylinders were cleaned by sonication and rinsing in distilled water and sterilization with dry heat (220 degrees Celsius) for five hours. The cylinders then were coated with human plasma fibronectin (Gibco) by soaking in a solution containing 100 micrograms of fibronectin per milliliter for sixteen hours at 4 degrees Celsius. The cylinders were air-dried at room temperature overnight in a sterile biosafety cabinet and then were stored at 4 degrees Celsius.

The mesenchymal stem cells were loaded onto the ceramic cylinders as previously described28,29. In brief, the cylinders were placed in a five-milliliter suspension of cells (7.5 x 106 cells per milliliter) in serum-free Dulbecco modified Eagle medium. The loading vessel was tightly capped, and the cylinders were subjected to a vacuum with three bursts of five seconds each in order to remove any air within the pores of the ceramic and to facilitate the flow of fluid into the pores. The loading vessels were recapped loosely, placed in a tissue culture incubator for two hours, and gently agitated every thirty minutes until the time of the operative procedure. Residual non-adherent cells in the suspension medium were counted to determine the quantity of cells to be delivered with the ceramic into the defect. The control cylinders were treated identically, except that the incubation medium contained no mesenchymal stem cells.

Back to Top | Article Outline
Operative Procedure

The dogs were sedated with the intravenous administration of butorphanol (0.2 milligram per kilogram of body weight), acetylpromazine (0.05 milligram per kilogram of body weight), and glycopyrrolate (0.01 milligram per kilogram of body weight). Anesthesia then was achieved with the administration of thiobarbital (eight milligrams per kilogram of body weight to effect) and was maintained with use of isoflurane in oxygen, administered through an endotracheal tube. The limb to be operated on, chosen randomly, was shaved and prepared. An epidural injection of morphine solution (Duramorph, five milliliters of solution containing one milligram of morphine per milliliter) was administered and a fentanyl patch (delivering fifty micrograms per hour) was applied, at the time of induction of anesthesia, for postoperative analgesia. A lateral approach to the femur, between the biceps femoris muscle and the vastus lateralis muscle, was performed. A portion of the greater trochanter was removed to facilitate placement of the fixation plate. The periosteum was elevated only enough to allow the plate to lie directly on bone. An eight-hole, 4.5 by 135-millimeter lengthening plate (Synthes, Paoli, Pennsylvania) was contoured and applied to the lateral aspect of the femur. The plate then was removed, and a twenty-one-millimeter-long osteoperiosteal segmental cortical defect was made at the mid-portion of the diaphysis with use of an oscillating bone-saw that was continuously cooled by irrigation with saline solution. The plate then was reapplied, and a ceramic cylinder was implanted in the defect or the defect was left untreated. The cylinder was secured in the defect with two resorbable sutures that were placed circumferentially around the cylinder and the fixation plate. The wound then was closed. All of the animals were killed, sixteen weeks after implantation, by means of intravenous administration of sodium pentobarbital (eighty milligrams per kilogram of body weight).

Back to Top | Article Outline
Radiographic and Radiomorphometric Analysis

Before the operation, the dogs were anesthetized and anteroposterior and lateral radiographs of the femur and pelvis were made. Only anteroposterior radiographs were made subsequently. High-resolution radiographs of the femur were made at sixteen weeks postoperatively. Two of us (S. P. B. and S. K.) independently and blindly evaluated all of the radiographs with regard to the presence of osseous union at each host bone-implant interface (as indicated by obliteration of the transverse radiolucent line between the cylinder and the host bone that was seen immediately postoperatively), the presence of osseous callus overlying the implant, and the presence of periosteal new bone spanning the entire defect from the distal to the proximal host bone-implant interface. The time at which the callus around the implant was the thickest was also noted. Calipers were used to measure the thickness of the callus on the medial aspect of the implant in three locations: the proximal and distal host bone-implant interfaces and the middle portion of the implant. The mean thickness of the callus was calculated at four, eight, twelve, and sixteen weeks. Because of the plane of the radiographs and the presence of the fixation plate along the lateral aspect of the femur, only the callus around the medial portion of the implant could be evaluated.

Back to Top | Article Outline
Histological and Histomorphometric Analysis

After the radiographic analysis, the femora were cleaned of soft tissue and the fixation plates were removed. The central region of each femur, containing the implant and both host bone-implant interfaces, was excised and was fixed in 10 per cent buffered formalin. The specimens then were dehydrated, cleared, and embedded in methylmethacrylate. Two one-millimeter-thick cross sections were cut through the implant with use of a water-cooled saw (Isomet; Buehler, Lake Bluff, Illinois); each cut was made approximately three millimeters from the proximal or distal host bone-implant interface (Fig. 1). The remaining bone was reassembled, and a longitudinal section was cut in the sagittal plane. The central longitudinal section and the two cross sections from each femur were ground to a thickness of 100 micrometers, polished, and stained with toluidine blue O or MacNeal light green. Image-analysis software (Quantimet 500MC; Leica Instruments, Cambridge, United Kingdom) was used to determine the total area of the implant and to measure the areas occupied by ceramic, bone, and soft tissue. Approximately seventy-five microscopic fields in each sample were subjected to histomorphometric analysis. The percentage of the total area of the implant that was occupied by ceramic, the percentage of the area of the porous space that was occupied by bone, and the percentage of the area of the porous space that was occupied by soft tissue then were calculated20,21,28,42,48. The tissues within the medullary canal and external to the implant itself were not included in the measurements of area; however, the thickness of the callus around the implant was measured in all specimens. The mean thickness of the callus on the medial aspect of the implant was determined from measurements corresponding to three locations: the proximal and distal host bone-implant interfaces and the middle portion of the implant. In addition to the quantitative histomorphometric analysis, union at the host bone-implant interface was scored on a scale of 0 to 4 points, with 1 point assigned for union at each of the four interfaces between the femoral cortex and the implant. The independent variable for the histomorphometric analysis was the presence or absence of mesenchymal stem cells, and the dependent variable was the area of bone, expressed as the percentage of the area of available pore space. A Student t test was used to analyze the data after ensuring that they passed the normality test.

Back to Top | Article Outline

Results

Cultivation of Mesenchymal Stem Cells

All vials containing bone-marrow aspirates arrived at the cell-culture facility within twenty-four hours after collection. Both the primary cultures and the cultures of passaged mesenchymal stem cells from all of the animals displayed characteristic spindle-shaped morphology. The cultured cells were placed in flasks and then were returned to the animal-care facility, where they were combined with the ceramic cylinders on the morning of the operation. There were approximately thirty million cells, or fifteen million cells per milliliter of implant volume, on each cylinder when it was implanted in the defect.

Back to Top | Article Outline
Animal Model

All of the dogs were fully weight-bearing by one week, and they became active in their open pens. All of the wounds healed without infection, and there were no failures of fixation.

Back to Top | Article Outline
Radiographic Analysis

The three defects that were left untreated (Group C) did not heal during the sixteen-week period of study (Fig. 2). There was limited new-bone formation, first visible at eight weeks, at the cut ends of the cortices of the host bone. Little or no new bone had formed at these cut ends by sixteen weeks. All of the untreated defects had the radiographic appearance of atrophic non-union.

The ceramic implants with or without mesenchymal stem cells could be visualized easily because of the radiopacity of the material (Figs. 3 and 4). Immediately postoperatively, the implants had a granular appearance because the pores were empty. As new bone formed, the appearance of the implant became smoother and more radiopaque. A distinct radiolucent zone at the interface between the implant and the host bone also was visible on the immediate postoperative radiographs. The absence of this radiolucent transverse zone was considered to be an indication of union between the implant and the host bone. Union had occurred at ten of the twelve interfaces in Group B (implants that had not been loaded with cells) by sixteen weeks. However, a large number of cracks were seen within the ceramic at twelve weeks. Union was established more rapidly in Group A (implants that had been loaded with mesenchymal stem cells); by eight weeks, union had occurred at all twelve interfaces. The cracks that were seen in several of the specimens in this group at eight or twelve weeks became less evident as time progressed, and the fractures were stabilized by increasingly radiodense bone within the implant or surrounding the implant, or both.

The most striking radiographic observation was that a large osseous callus had developed around the periphery of five of the six implants that had been loaded with mesenchymal stem cells as well as around the adjacent host bone (Fig. 4). No callus was visible around any of the ceramic implants that had not been loaded with cells (Fig. 3) or the untreated defects (Fig. 2) during the sixteen-week period of study. The mean thickness of the callus around the implants that had been loaded with mesenchymal stem cells changed over time, reaching a maximum at eight or twelve weeks and decreasing by sixteen weeks (Fig. 5). The maximum thickness of the callus for each animal in Group A was measured, and the mean value was calculated to be 3.14 millimeters.

Back to Top | Article Outline
Histological Analysis

Histological examination of the untreated defects consistently demonstrated atrophic non-union with little or no new-bone formation. Trabecular bone appeared to arise from the cut ends of the host bone (the sites of the osteotomies) (Fig. 6). The central region of the defect was filled with fibrous tissue and skeletal muscle that had collapsed into the defect. There was no periosteal reaction and no callus around the host bone.

Solid union was evident between the host bone and the implants that had been loaded with mesenchymal stem cells (Group A) (Fig. 7, a and c). New bone was intimately associated with the ends of the implant: a continuous bridge of mineralized bone spanned the defect from the cut ends of the cortex into the porous ends of the implant (Fig. 7, c). In addition, there was uniform and substantial new-bone formation throughout the body of the implant, which was seen on both the longitudinal section (Fig. 7, a) and the cross sections (Fig. 7, b). Reestablishment of the marrow cavity was facilitated by the hollow core of the implant, which provided an open channel to the host medullary canal. Higher magnification showed that virtually all of the pores of the implant contained either woven or lamellar new bone and that the bone was in direct contact with the walls of the ceramic (Fig. 7, d). As had been seen on the radiographs, a substantial callus spanned the length of the defect and extended into adjacent areas surrounding the host bone. Cross sections showed that the callus was not restricted to the medial aspect of the femur (Fig. 7, b), as one might infer from the radiographs, and that there also was a substantial amount of new bone on the anterior and posterior surfaces of the implant.

Even though union was established at the ends of the ceramic implants that had not been loaded with cells (Group B) (Fig. 8, a and c), only a minimum amount of bone formed in the middle portion of the implant and most of the pores were filled with fibrous tissue (Fig. 8, a and d). Union at the host bone-implant interface appeared to be primarily due to the ingrowth of bone from the cut ends of the cortex of the host bone, with a concomitant endosteal spike of bone advancing into the hollow core of the implant. Although a modest amount of bone formed in the body of some of these implants, there also appeared to be a centrifugal osteoconductive process emerging from the spike of trabecular bone emanating from the cut end of the cortex and the medullary canal (Fig. 8, b). In general, bone formed only within the pores on the inner surface of the implant and not within the pores of the periphery. Unlike the specimens in Group A (implants that had been loaded with mesenchymal stem cells), but similar to the specimens in Group C (untreated defects), the specimens in Group B (implants that had not been loaded with cells) demonstrated no callus around the implant and no periosteal bone reaction on the host bone.

The area occupied by the ceramic material, expressed as a percentage of the total area of the implant, averaged approximately 35 per cent for Groups A and B (Table I); this value is consistent with an estimated pore volume of 65 per cent for this material. The area of bone in the pores of the implants in Group A was uniformly high (mean, 39.9 per cent); this area was significantly greater (p < 0.05) than that in Group B (mean, 24.0 per cent). The percentage of the total area that was occupied by soft tissue, including fibrous tissue and blood vessels, was calculated to be 60.1 per cent in Group A and 76.0 per cent in Group B. In addition to having a significantly greater amount of bone in the pores of the implant, the specimens in Group A also consistently exhibited a substantial callus around the implant (mean thickness, 1.56 millimeters on the medial surface of the implant at sixteen weeks). By sixteen weeks, the thickness of the callus in all of the specimens had decreased from the maximum, which had been apparent at eight or twelve weeks (Fig. 4). No callus could be detected at any time in Group B (implants that had not been loaded with cells) or Group C (untreated defects). As noted in the radiographic analysis, there was little difference between the two types of implants with respect to the prevalence of union at sixteen weeks. Evaluation of the histological features of one specimen in Group B was limited because it had virtually no structural integrity after it had been removed from the animal and could not be processed for histological analysis as an intact unit. Therefore, representative samples were obtained from throughout the implant in order to perform histomorphometric analysis.

Back to Top | Article Outline

Discussion

The results of the present study demonstrate that new-bone formation can be elicited in critical-sized segmental defects in dogs by the implantation of autologous mesenchymal stem cells. Previous studies have shown that these pluripotential cells are capable of forming bone in an in vitro environment3,25,28,29 and when they are implanted in an appropriate matrix at ectopic or orthotopic sites5,7,12,18,28-30. Under the conditions used in the present study, the implantation of mesenchymal stem cells led to the formation of new bone without progression through a cartilaginous intermediate; the new bone was distributed uniformly throughout the cell-matrix implant and became integrated with the host bone. This is the first report, to our knowledge, describing a successful autologous cell-based approach for the repair of osseous tissue in large animals.

The design and selection of an ideal carrier for the delivery of mesenchymal stem cells is based on several criteria. First, the material should allow for uniform loading and retention of cells. Second, the carrier should support rapid vascular ingrowth. Third, the matrix should be composed of radiolucent materials that are resorbed and replaced by bone as new bone is formed. Fourth, the material should allow or enhance osteoconductive bridging of host bone by the new bone. Finally, the cell-matrix combination should be easy for the physician to handle in a clinical setting. The form of ceramic (hydroxyapatite and β-tricalcium phosphate) that was used in the present study may not be the ideal delivery vehicle for the treatment of segmental defects because of its brittle nature and lack of completely interconnected pores. The radiopacity and slow resorption of this ceramic may make it difficult for new bone to be distinguished radiographically from the original ceramic. In the present study, radiographic interpretations corresponded favorably with the histological observations that were made long after the review of the radiographs.

Because this ceramic is not rapidly resorbed, the structural geometry of the material did not change over the sixteen-week period of study. Many of the ceramic implants had evidence of fractures at eight or twelve weeks; in the implants that had not been loaded with cells, these fractures became more pronounced with time. Thus, a porous ceramic material, such as the one used in the present study, although useful as an osteoconductive matrix that supports the differentiation of progenitor cells12,17,18,26,28,29, possesses little load-bearing capacity because it is brittle and tends to fracture in the center when it is used as a segmental cortical bone graft17,34. However, after the initial fracture, consolidation of the fracture line with increasing radiodensity was observed in the implants that had been loaded with mesenchymal stem cells, indicating that the cells were capable of restoring continuity to the implant.

The most important and fascinating aspect of the present study is the biological phenomenon whereby mesenchymal stem cells elicited new-bone formation at an osseous site. The only new bone associated with defects that had been treated with an implant that had not been loaded with cells was the bone that had formed as a result of the osteoconductivity of the porous ceramic and the trabecular bone in the medullary canal that was an extension of the outgrowth from the cut ends of the host bone. In contrast, bone formed throughout the entire network of pores in the implants that had been loaded with mesenchymal stem cells, and a thick collar of bone developed around five of those six implants. No such collar formed around any of the implants that had not been loaded with cells, and there was no periosteal reaction along the host bone in any of the specimens that had been treated with an implant that had not been loaded with cells or in those that had been left untreated. The collar, or callus, formed in a circumferential manner around the implant, but only the bone that had formed along the medial aspect of the implant was clearly visible on the anteroposterior radiographs. Measurements that were made over time demonstrated that the thickness of the callus generally reached a maximum between eight and twelve weeks and then decreased. The collar was contiguous with the new periosteal bone on the cortices of the host bone, both distally and proximally, thus providing additional support and osseous integration. Because mesenchymal stem cells undergoing osteogenic differentiation have been shown to secrete inductive factors that recruit new mesenchymal stem cells into the osteogenic lineage24, we believe that the periosteal response of the host bone could reflect, in part, the local paracrine effects of these factors synthesized by the implanted cells in the developing collar. Mesenchymal stem cells in the periosteum of the host bone4,35 respond to these factors by differentiating into osteoblasts and producing mature bone matrix. This explanation also accounts for the lack of new periosteal bone in the defects that had been treated with an implant that had not been loaded with cells or those that had been left untreated, as there was no callus in any of those specimens. Bone filled approximately 40 per cent of the available pore space in the ceramic implants that had been loaded with mesenchymal stem cells. This amount was comparable with that seen in studies from our laboratories in which either syngeneic rat cells were used28 or human mesenchymal stem cells were used in athymic rats5.

A number of investigators have studied the ability of fresh bone marrow to serve as a bone-grafting agent in the treatment of bone defects in rodents, dogs, and humans9,10,17,26,28,37,44,47. Although some of these investigations have indicated that bone marrow is capable of promoting new-bone formation, techniques for enriching the active component of bone marrow—namely, mesenchymal stem cells—are of primary importance2,7,9,36,47. In a direct comparison of the osteogenic potential of fresh bone marrow with that of mesenchymal stem cells, we demonstrated that, with a ceramic carrier, mesenchymal stem cells promoted faster and more extensive new-bone formation than fresh bone marrow did28. We attributed this finding to the 300-fold increase in the number of mesenchymal stem cells, which was achieved through the culture expansion process. Direct delivery of the cells that are responsible for the formation of bone allows the process of tissue formation to occur in a more rapid and uniform manner throughout the implant. The approach described in the present report offers the extraordinary advantage of increasing the number of progenitor cells at local sites where repair or regeneration of tissue is needed. This is especially important in the setting of clinical conditions that are associated with a decrease in bone marrow-derived osteoprogenitors, such as increasing age, osteoporosis, and other metabolic derangements13,23,39,43. This generalized decline leads to a diminution in the pool of cells available for repair of tissue and may account for impaired healing in these patients. The mesenchymal stem cells from such patients can be cultured and then delivered back to specific sites where augmentation of bone is needed in order to overcome the natural decline in the regenerative capacity of tissue.

In summary, the present study established the technical feasibility of using implants loaded with autologous mesenchymal stem cells to achieve site-specific new-bone formation in humans. The aspiration of bone marrow at a clinical center, the growth of cells at a remote cell-culture facility, and the return of the cell-matrix implants to the orthopaedic surgeon constitute the elements of a process for achieving bone repair in patients at distant sites. Because purified human mesenchymal stem cells retain their osteogenic potential after extensive growth in culture and cryopreservation3, autologous cell therapy may prove to be useful in association with a wide variety of clinical applications, including repair of segmental defects, spinal arthrodesis, fixation of prosthetic implants, and restoration of maxillofacial bone.

NOTE: The authors thank Robin Douglas, Diane Sterchi, Donna Brown, Marcia Black, Nancy Ricalton, Jessie Chou, James Cole, Deborah Miller, Harry Wotton, Amy Roscioli, and Dr. Neelam Jaiswal for expert technical assistance; Jennifer Kahler for drawing Figure 1; Dr. Carl Kirker-Head for helpful discussions; and Dr. Daniel R. Marshak for critical reading of the manuscript.

*One or more of the authors has received or will receive benefits for personal or professional use from a commercial party related directly or indirectly to the subject of this article. In addition, benefits have been or will be directed to a research fund, foundation, educational institution, or other non-profit organization with which one or more of the authors is associated. Funds were received in total or partial support of the research or clinical study presented in this article. The funding source was Osiris Therapeutics. Ceramic materials were provided by Zimmer, Incorporated.

Investigation performed at Osiris Therapeutics, Baltimore, and Tufts University School of Veterinary Medicine, North Grafton

1. Bab, I.; Gazit, D.; Muhlrad, A.; and Shteyer, A.: Regenerating bone marrow produces a potent growth-promoting activity to osteogenic cells. Endocrinology, 123: 345-352, 1988.
2. Bruder, S. P.; Fink, D. J.; and Caplan, A. I.: Mesenchymal stem cells in bone development, bone repair, and skeletal regeneration therapy. J. Cell. Biochem., 56: 283-294, 1994.
3. Bruder, S. P.; Jaiswal, N.; and Haynesworth, S. E.: Growth kinetics, self-renewal, and the osteogenic potential of purified human mesenchymal stem cells during extensive subcultivation and following cryopreservation. J. Cell. Biochem., 64: 278-294, 1997.
4. Bruder, S. P.; Horowitz, M. C.; Mosca, J. D.; and Haynesworth, S. E.: Monoclonal antibodies reactive with human osteogenic cell surface antigens. Bone, 21: 225-235, 1997.
5. Bruder, S. P.; Kurth, A. A.; Shea, M.; Hayes, W. C.; Jaiswal, N.; and Kadiyala, S.: Bone regeneration by implantation of purified, culture-expanded human mesenchymal stem cells. J. Orthop. Res., 16: 155-162, 1998.
6. Caplan, A. I.: Mesenchymal stem cells. J. Orthop. Res., 9: 641-650, 1991.
7. Caplan, A. I., and Bruder, S. P.: Cell and molecular engineering of bone regeneration. In Textbook of Tissue Engineering, pp. 603-618. Edited by R. Lanza, R. Langer, and W. Chick. Georgetown, Texas, R. G. Landes, 1997.
8. Caplan, A. I.; Fink, D. J.; Goto, T.; Linton, A. E.; Young, R. G.; Wakitani, S.; Goldberg, V. M.; and Haynesworth, S. E.: Mesenchymal stem cells and tissue repair. In The Anterior Cruciate Ligament: Current and Future Concepts, pp. 405-417. Edited by D. W. Jackson, S. P. Arnoczky, S. L-Y. Woo, C. B. Frank, and T. M. Simon. New York, Raven Press, 1993.
9. Connolly, J. F.; Guse, R.; Lippiello, L.; and Dehne, R.: Development of an osteogenic bone-marrow preparation. J. Bone and Joint Surg., 71-A: 684-691, June 1989.
10. Connolly, J. F.; Guse, R.; Tiedeman, J.; and Dehne, R.: Autologous marrow injection as a substitute for operative grafting of tibial nonunions. Clin. Orthop., 266: 259-270, 1991.
11. Cook, S. D.; Baffes, G. C.; Wolfe, M. W.; Sampath, T. K.; and Rueger, D. C.: Recombinant human bone morphogenetic protein-7 induces healing in a canine long-bone segmental defect model. Clin. Orthop., 301: 302-312, 1994.
12. Dennis, J. E.; Haynesworth, S. E.; Young, R. G.; and Caplan, A. I.: Osteogenesis in marrow-derived mesenchymal cell porous ceramic composites transplanted subcutaneously: effect of fibronectin and laminin on cell retention and rate of osteogenic expression. Cell Transplant., 1: 23-32, 1992.
13. Egrise, D.; Martin, D.; Vienne, A.; Neve, P.; and Schoutens, A.: The number of fibroblastic colonies formed from bone marrow is decreased and the in vitro proliferation rate of trabecular bone cells increased in aged rats. Bone, 13: 355-361, 1992.
14. Foldes, J.; Naparstek, E.; Statter, M.; Menczel, J.; and Bab, I.: Osteogenic response to marrow aspiration: increased serum osteocalcin and alkaline phosphatase in human bone marrow donors. J. Bone and Min. Res., 4: 643-646, 1989.
15. Gerhart, T. N.; Kirker-Head, C. A.; Kriz, M. J.; Holtrop, M. E.; Hennig, G. E.; Hipp, J.; Schelling, S. H.; and Wang, E.: Healing segmental femoral defects in sheep using recombinant human bone morphogenetic protein. Clin. Orthop., 293: 317-326, 1993.
16. Goshima, J.; Goldberg, V. M.; and Caplan, A. I.: The origin of bone formed in composite grafts of porous calcium phosphate ceramic loaded with marrow cells. Clin. Orthop., 269: 274-283, 1991.
17. Grundel, R. E.; Chapman, M. W.; Yee, T.; and Moore, D. C.: Autogeneic bone marrow and porous biphasic calcium phosphate ceramic for segmental bone defects in the canine ulna. Clin. Orthop., 266: 244-258, 1991.
18. Haynesworth, S. E.; Goshima, J.; Goldberg, V. M.; and Caplan, A. I.: Characterization of cells with osteogenic potential from human marrow. Bone, 13: 81-88, 1992.
19. Hollinger, J. O.; Brekke, J.; Gruskin, E.; and Lee, D.: Role of bone substitutes. Clin. Orthop., 324: 55-65, 1996.
20. Holmes, R. E., and Hagler, H. K.: Porous hydroxyapatite as a bone graft substitute in cranial reconstruction: a histometric study. Plast. and Reconstr. Surg., 81: 662-671, 1988.
21. Holmes, R. E.; Bucholz, R. W.; and Mooney, V.: Porous hydroxyapatite as a bone graft substitute in diaphyseal defects: a histometric study. J. Orthop. Res., 5: 114-121, 1987.
22. Hunt, T. R.; Schwappach, J. R.; and Anderson, H. C.: Healing of a segmental defect in the rat femur with use of an extract from a cultured human osteosarcoma cell-line (Saos-2). A preliminary report. J. Bone and Joint Surg., 78-A: 41-48, Jan. 1996.
23. Inoue, K.; Ohgushi, H.; Yoshikawa, T.; Okumura, M.; Sempuku, T.; Tamai, S.; and Dohi, Y.: The effect of aging on bone formation in porous hydroxyapatite: biochemical and histological analysis. J. Bone and Min. Res., 12: 989-994, 1997.
24. Jaiswal, N., and Bruder, S. P.: Human osteoblastic cells secrete paracrine factors which regulate differentiation of osteogenic precursors in marrow. Trans. Orthop. Res. Soc., 22: 524, 1997.
25. Jaiswal, N.; Haynesworth, S. E.; Caplan, A. I.; and Bruder, S. P.: Osteogenic differentiation of purified, culture-expanded human mesenchymal stem cells in vitro. J. Cell. Biochem., 64: 295-312, 1997.
26. Johnson, K. D.; Frierson, K. E.; Keller, T. S.; Cook, C.; Scheinberg, R.; Zerwekh, J.; Meyers, L.; and Sciadini, M. F.: Porous ceramics as bone graft substitutes in long bone defects: a biomechanical, histological, and radiographic analysis. J. Orthop. Res., 14: 351-369, 1996.
27. Johnstone, B.; Hering, T. M.; Caplan, A. I.; Goldberg, V. M.; and Yoo, J. U.: In vitro chondrogenesis of bone marrow-derived mesenchymal progenitor cells. Exper. Cell Res., 238: 265-272, 1998.
28. Kadiyala, S.; Jaiswal, N.; and Bruder, S. P.: Culture-expanded, bone marrow-derived mesenchymal stem cells can regenerate a critical-sized segmental bone defect. Tissue Eng., 3: 173-185, 1997.
29. Kadiyala, S.; Young, R. G.; Thiede, M. A.; and Bruder, S. P.: Culture expanded canine mesenchymal stem cells possess osteochondrogenic potential in vivo and in vitro. Cell Transplant., 6: 125-134, 1997.
30. Lennon, D. P.; Haynesworth, S. E.; Bruder, S. P.; Jaiswal, N. J.; and Caplan, A. I.: Human and animal mesenchymal progenitor cells from bone marrow: identification of serum for optimal selection and proliferation. In Vitro Cell. and Devel. Biol., 32: 602-611, 1996.
31. Majumdar, M. K.; Thiede, M. A.; Mosca, J. D.; Moorman, M. K.; and Gerson, S. L.: Phenotypic and functional comparison of cultures of marrow-derived mesenchymal stem cells (MSCs) and stromal cells. J. Cell. Physiol., 176: 57-66, 1998.
32. Martin, R. B.; Chapman, M. W.; Holmes, R. E.; Sartoris, D. J.; Shors, E. C.; Gordon, J. E.; Heitter, D. O.; Sharkey, N. A.; and Zissimos, A. G.: Effects of bone ingrowth on the strength and non-invasive assessment of a coralline hydroxyapatite material. Biomaterials, 10: 481-488, 1989.
33. Mayer, M.; Hollinger, J.; Ron, E.; and Wozney, J.: Maxillary alveolar cleft repair in dogs using recombinant human bone morphogenetic protein-2 and a polymer carrier. Plast. and Reconstr. Surg., 98: 247-259, 1996.
34. Moore, D. C.; Chapman, M. W.; and Manske, D.: The evaluation of a biphasic calcium phosphate ceramic for use in grafting long-bone diaphyseal defects. J. Orthop. Res., 5: 356-365, 1987.
35. Nakahara, H.; Bruder, S. P.; Goldberg, V. M.; and Caplan, A. I.: In vivo osteochondrogenic potential of cultured cells derived from the periosteum. Clin. Orthop., 259: 223-232, 1990.
36. Niedzwiedzki, T.; Dabrowski, Z.; Mistza, H.; and Pawlikowksi, M.: Bone healing after bone marrow stromal cell transplantation to the bone defect. Biomaterials, 14: 115-121, 1993.
37. Ohgushi, H.; Goldberg, V. M.; and Caplan, A. I.: Repair of bone defects with marrow cells and porous ceramics. Experiments in rats. Acta Orthop. Scandinavica, 60: 334-339, 1989.
38. Pittenger, M. F.; Mackay, A. M.; and Beck, S. C.: Human mesenchymal stem cells can be directed into chondrocytes, adipocytes and osteocytes [abstract]. Molec. Biol. Cell, 7: 305a, 1996.
39. Quarto, R.; Thomas, D.; and Liang, C. T.: Bone progenitor cell deficits and the age-associated decline in bone repair capacity. Calcif. Tissue Internat., 56: 123-129, 1995.
40. Saito, T.; Dennis, J. E.; Lennon, D. P.; Young, R. G.; and Caplan, A. I.: Myogenic expression of mesenchymal stem cells within myotubes of mdx mice in vitro and in vivo. Tissue Eng., 1: 327-343, 1995.
41. Schmitz, J. P., and Hollinger, J. O.: The critical sized defect as an experimental model for craniomandibulofacial nonunions. Clin. Orthop., 205: 299-308, 1986.
42. Stevenson, S.; Cunningham, N.; Toth, J.; Davy, D.; and Reddi, A. H.: The effect of osteogenin (a bone morphogenetic protein) on the formation of bone in orthotopic segmental defects in rats. J. Bone and Joint Surg., 76-A: 1676-1687, Nov. 1994.
43. Tabuchi, C.; Simmons, D. J.; Fausto, A.; Russell, J. E.; Binderman, I.; and Avioli, L. V.: Bone deficit in ovariectomized rats. Functional contribution of the marrow stromal cell population and the effect of oral dihydrotachysterol treatment. J. Clin. Invest., 78: 637-642, 1986.
44. Takagi, K., and Urist, M. R.: The role of bone marrow in bone morphogenetic protein-induced repair of femoral massive diaphyseal defects. Clin. Orthop., 171: 224-231, 1982.
45. Wakitani, S.; Saito, T.; and Caplan, A. I.: Myogenic cells derived from rat bone marrow mesenchymal stem cells exposed to 5-azacytidine. Muscle and Nerve, 18: 1417-1426, 1995.
46. Wakitani, S.; Goto, T.; Pineda, S. J.; Young, R. G.; Mansour, J. M.; Caplan, A. I.; and Goldberg, V. M.: Mesenchymal cell-based repair of large, full-thickness defects of articular cartilage. J. Bone and Joint Surg., 76-A: 579-592, April 1994.
47. Werntz, J. R.; Lane, J. M.; Burstein, A. H.; Justin, R.; Klein, R.; and Tomin, E.: Qualitative and quantitative analysis of orthotopic bone regeneration by marrow. J. Orthop. Res., 14: 85-93, 1996.
48. Wolff, D.; Goldberg, V. M.; and Stevenson, S.: Histomorphometric analysis of the repair of a segmental diaphyseal defect with ceramic and titanium fibermetal implants: effects of bone marrow. J. Orthop. Res., 12: 439-446, 1994.
49. Yasko, A. W.; Lane, J. M.; Fellinger, E. J.; Rosen, V.; Wozney, J. M.; and Wang, E. A.: The healing of segmental bone defects, induced by recombinant human bone morphogenetic protein (rhBMP-2). A radiographic, histological, and biomechanical study in rats. J. Bone and Joint Surg., 74-A: 659-670, June 1992.
50. Young, R. G.; Butler, D. L.; Weber, W.; Gordon, S. L.; and Fink, D. J.: Mesenchymal stem cell-based repair of rabbit Achilles tendon. Trans. Orthop. Res. Soc., 22: 249, 1997.
51. Zardiackas, L. D.; Teasdall, R. D.; Black, R. J.; Jones, G. S.; St. John, K. R.; Dillon, L. D.; and Hughes, J. L.: Torsional properties of healed canine diaphyseal defects grafted with a fibrillar collagen and hydroxyapatite/tricalcium phosphate composite. J. Appl. Biomater., 5: 277-283, 1994.
Copyright 1998 by The Journal of Bone and Joint Surgery, Incorporated