Vol 9, No 3, May/June 2001 157 Regeneration is defined as the reconstitution or restoration of a body part, tissue, or substance, whether in response to injury or as a normal bodily process. Only two tissues in humans possess sig- nificant regenerative capacity— bone and liver. All other tissues, when damaged, heal with the for- mation of scar, leaving a mark of new fibrous connective tissue that replaces the injured structure. The limitation of scar tissue is that it does not possess the biomechani- cal, physical, and functional prop- erties of the original tissue. Thus, regeneration is a specialized repair process that confers a biologic privilege on those tissues that pos- sess it. There is a large amount of infor- mation known about bone regener- ation as it occurs in fracture heal- ing, which is a normal process in all vertebrate animals. Over the past several decades, methods of controlling bone regeneration (e.g., limb-lengthening procedures and technologies) have been developed. Subsequently, the cellular and mol- ecular bases for bone regeneration have been established, especially as regards the role of the bone mor- phogenetic proteins (BMPs). There is recent evidence for the clinical efficacy of at least one of these tech- nologies. The description of the pathophysiology of fibrodysplasia ossificans progressiva, a rare genet- ic disease characterized by the spontaneous formation of hetero- topic bone, highlights the immense capacity inherent for postnatal bone formation in human connec- tive tissues. Dr. Einhorn is Chairman and Professor, Department of Orthopaedic Surgery, Boston University School of Medicine, Boston, Mass. Dr. Lee is Orthopaedic Surgery Resident, Bowman-Gray School of Medicine, Winston- Salem, NC. One or more of the authors or the departments with which they are affiliated have received something of value from a commercial or other party related directly or indirectly to the sub- ject of this article. Reprint requests: Dr. Einhorn, Doctors Office Building, Suite 808, 720 Harrison Avenue, Boston, MA 02118. Copyright 2001 by the American Academy of Orthopaedic Surgeons. Abstract Bone is a biologically privileged tissue in that it has the capacity to undergo regeneration as part of a repair process. Fracture healing is the most com- mon and recognizable form of bone regeneration, but several other examples of bone regeneration have been observed in humans, suggesting that the abil- ity to regulate bone regeneration as a therapeutic tool should be possible. Historically, efforts at limb lengthening have led to procedures for regenerat- ing bone, such as the method of Ilizarov. This procedure, known as distrac- tion osteogenesis, has applications in a variety of skeletal conditions, includ- ing the restoration of large skeletal defects, the transport of bone in cases of severe trauma with bone loss, and the correction of skeletal deformities. Fibrodysplasia ossificans progressiva is an example of how an abnormal metabolic condition can be viewed as evidence for the capacity of humans to regenerate large amounts of bone if the cellular and molecular signaling events are altered. Elucidation of the cellular and molecular basis for bone regeneration in humans—particularly the role of the human genome in rela- tion to the expression of various growth factors and cytokines, such as the bone morphogenetic proteins—offers great potential for the treatment of orthopaedic conditions. Development of specific bone morphogenetic proteins as therapeutic substances to induce bone regeneration in patients is well under way. As methods for enhancing fracture healing, distraction osteogen- esis, and other procedures are refined, the development of protein- and gene- based therapies for regulating bone formation should lead to a new era of orthopaedic practice. J Am Acad Orthop Surg 2001;9:157-165 Bone Regeneration: New Findings and Potential Clinical Applications Thomas A. Einhorn, MD, and Cassandra A. Lee, MD Orthopaedic Research Society Special Article Fracture Healing Fracture healing is a form of bone regeneration, in that it results in functional bone tissue with all the properties that were originally pres- ent in the uninjured bone. There are four distinct tissue responses that can occur in fracture healing. These responses are produced by the bone marrow, the bone cortex, the perios- teum, and the external soft tissues. Depending on the manner in which the fracture is treated, these re- sponses can occur singly, or two or more can occur simultaneously. The bone marrow response begins with a loss of its normal architecture. In the region adjacent to the fracture hematoma, cellular components un- dergo reorganization into regions of high and low density. In the region of highest density, endothelial cells transform into polymorphic cells, which express an osteoblastic pheno- type. 1 These cells form bone within a few days after fracture. Interestingly, the bone marrow response occurs independent of the mechanical strain environment or the method by which the fracture is treated. 2 The cortical response is deter- mined by the type of fracture heal- ing that takes place. Two types have been recognized. In primary fracture healing, the cortex attempts to reestablish itself without the formation of a callus. This type of healing occurs only when the frac- ture is anatomically reduced and stabilized by rigid internal fixation. A tunneling resorptive response occurs, whereby new haversian sys- tems are established to allow pene- tration of blood vessels into the area of the fracture. Perivascular mes- enchymal cells and endothelial cells accompany these newly formed vessels and differentiate into osteo- progenitor cells. 3 In contrast, sec- ondary fracture healing results in the formation of a callus and in- volves the participation of the peri- osteum and external soft tissues. The cortex is enveloped by the process but is not involved in a direct response. This fracture heal- ing response is enhanced by motion and is inhibited by rigid fixation. 3 During fracture healing, the se- quential events of tissue develop- ment occur, leading to the regenera- tion of functional osseous tissue. The immediate response to injury includes hematoma formation, in- flammation and angiogenesis, carti- lage formation with subsequent cal- cification, and cartilage removal accompanied by bone formation. After this last step, bone remodeling begins; this leads to the restoration of the load-carrying capability of the bone. The surrounding soft tissue may also contribute to fracture healing. Rapid cellular activity and the de- velopment of an early bridging cal- lus help to stabilize the fracture fragments. This process, like the periosteal response, may be affected by mechanical factors and hindered by rigid immobilization. 3 Intramembranous ossification (the direct formation of bone from committed osteoprogenitor cells) contributes to the formation of a hard callus at the periphery of the fracture. Endochondral ossification (the indirect formation of bone from uncommitted mesenchymal cells) occurs adjacent to the fracture site and contributes to the formation of a soft callus. During this process, cells differentiate to chondrocytes; a cartilage anlage forms, undergoes calcification, and is ultimately re- placed by bone. The response to fracture injury involves disruption of the normal vasculature, infiltration of inflam- matory cells, and release of a multi- tude of cytokines and peptide signal- ing molecules. The first detectable factors released during this response are platelet-derived growth factor and transforming growth factor-β (TGF-β). 4 Other BMPs and their receptors that are also expressed are likely important in this reparative response. Macrophages and other inflammatory cells release proin- flammatory cytokines, such as inter- leukin-1, tumor necrosis factor-α, and interleukin-6. 5 These events of bone repair form the fundamental basis by which bone regeneration can be viewed as a naturally occurring clini- cal process. Limb Lengthening and Bone Transport The first successful attempt at thera- peutic human bone regeneration in humans was reported by Codivilla in 1905. As part of a strategy to lengthen shortened limbs, he created an osteotomy through the cortex of the femur and the tibia and induced tractional forces with the use of a calcaneal pin. In 22 cases, the gain in length was between 3 and 8 cm. In 1908, Magnuson reported suc- cessful human femoral lengthening, which was achieved by creating a median longitudinal step-cut osteot- omy. The proximal segment was fixed, and the distal segment was attached to a pulley-weight system that accomplished 2- to 3-inch lengthenings in 5 minutes. Once the desired length and alignment had been achieved, the fragments were fixed with screws. In 1913, Ombredanne was the first to use an external fixator for limb lengthening, but unfortunately complications of skin necrosis and infection arose. It was not until 1927 that Abbott introduced the concept of a latency period to pro- mote formation of bone prior to dis- traction. Current thinking suggests that the latency period provides time for the initial phases of bone repair to take place at the osteotomy site, resulting in a mechanically compliant callus, restoration of the blood supply by means of revascu- larization, and initiation of the bone regeneration sequence. 6 Bone Regeneration Journal of the American Academy of Orthopaedic Surgeons 158 The procedure of distraction os- teogenesis for bone regeneration was refined by Ilizarov. 6 Perhaps more than any other development in medical history, the Ilizarov method shows how bone regenera- tion is possible in humans. The so- called low-energy osteotomy of the cortex was suggested by Ilizarov to be critical to the success of the pro- cedure. Although it is possible to perform the osteotomy at any site, the metaphysis is ideal, in that it offers good stability because of the thin cortex and large trabecular sur- face and is endowed with excellent blood flow from an extensive sys- tem of collateral vessels. 6 The latency period prior to distraction ranges from 3 to 10 days (a shorter period for metaphyseal osteotomies and a longer period for diaphyseal oste- otomies). Distraction varies with respect to rate and rhythm, ranging from 0.5 to 2.0 mm/day and from one to four distractions per day. During distraction osteogenesis, angiogenesis precedes ossification, and bone is formed by intramem- branous ossification. Blood vessels are abundant where new bone is formed and sparse in regions of ma- ture bone. It has been established that the distraction rate affects the angiogenic response, and that a rate of 0.7 to 1.3 mm/day leads to opti- mal bone formation. 7 Because distraction osteogenesis involves gradual distraction with protection of adjacent joints, pa- tients have the ability to perform activities of daily living while un- dergoing extended lengthening pro- cedures. Several sites can be length- ened simultaneously to correct de- formities or to shorten the overall period of distraction. 8 Most impor- tant, patients with large skeletal defects who undergo this procedure can be treated without the need for bone grafting, internal fixation, or multiple operations. Although a number of problems and complica- tions are associated with this proce- Thomas A. Einhorn, MD, and Cassandra A. Lee, MD Vol 9, No 3, May/June 2001 159 dure, it is exceptionally effective when used as a means of bone re- generation. 8 An innovative method for the treatment of segmental defects caused by trauma, infection, or tumor resection was also devised by Ilizarov. 9 In this procedure, an oste- otomy is created proximal to the de- fect, and the intervening segment of bone is transported distally (Fig. 1). To be successful, the segment to be transported must possess an ade- quate blood supply so that bone for- mation can be induced at its trailing end and healing supported at its leading end. In addition, the micro- environment at the docking site must support healing. With Iliza- rov’s ring fixator, a bone segment can be transported in any direction with use of a system of pulling wires and transverse tension wires or half-pins. Multiple bone seg- ments can be transported in the same or opposite directions to facili- tate bone regeneration in the de- fect. 10 In some cases, autogenous bone grafting is necessary to en- hance healing at the docking site. The methods of limb lengthening and bone transport as described by Ilizarov and others have enjoyed substantial clinical success with regard to bone regeneration. This success vividly demonstrates the tremendous capacity for regenera- tion inherent in the human skeleton. Now that scientists possess the tools to investigate the molecular basis for these phenomena, it should be possible to develop more refined methods to produce and control regeneration of the skeleton. Figure 1 Lateral radiographs of the leg of a patient who underwent single-level proximal- to-distal transport because of bone loss after a gunshot injury. A three-ring apparatus was applied to the tibia, with a corticotomy at the proximal end and the transport ring at the dis- tal end; after transport, the transport fragment was docked with the bone on the opposite side of the defect. A, One month after osteotomy. B, At 3 months after osteotomy, the transport segment tilted posteriorly as the pins bent. C, At 6 months, the docking site was reduced in an open procedure, and the bone was grafted. D, Radiograph obtained shortly after removal of the apparatus at 10 months. Two years after removal, the anatomic and functional results were excellent. (Reproduced with permission from Paley D, Maar DC: Ilizarov bone transport treatment for tibial defects. J Orthop Trauma 2000;14:76-85.) A B C D Biologic Basis of Bone Regeneration In 1965, Urist observed that im- plantation of demineralized bone matrix at a heterotopic site led to the formation of a new ossicle with a hematopoietic marrow cavity. 11 In doing so, he introduced the con- cept of postfetal osteogenesis by a process known as bone induction. Over the course of the next 35 years, decalcified segments of di- aphyseal bone were implanted into muscle pouches in rats, 12 ulnar de- fects in rabbits, 13 lumbar sites in dogs, 14 and various sites in humans with certain skeletal disorders. 15,16 The process of bone induction be- gins with the formation of loose fibrous connective tissue, which is highly vascular and infiltrated with macrophages, lymphocytes, and fibroblasts. The process of endo- chondral ossification ensues, in which bone formation gives way to bone remodeling. Bone morpho- genetic proteins have been shown to exist within the bone matrix and to be responsible for this phenomenon. It is now known that the BMPs com- prise a family of molecules, each with its own function. Bone morphogenetic proteins are members of the TGF-β super- family of proteins but differ from other TGF-β family members in that some have more selective effects on bone. Bone morphoge- netic proteins are highly conserved from Drosophila (fruit fly) to humans and have been shown to induce proliferation and differentiation of mesenchymal stem cells to both chondrocytes and osteoblasts. Ge- netic and experimental evidence supports an even more diverse reg- ulatory role for BMPs in biologic processes, ranging from cell prolif- eration to apoptosis to differentia- tion to morphogenesis. They in- duce de novo bone formation by means of endochondral ossifica- tion. At high concentration, BMPs may form bone directly by intra- membranous bone formation. 17 The current concept of the role of BMPs is that they are key modula- tors of osteoprogenitor and mes- enchymal cells throughout the frac- ture healing process. Levels of BMP expression, particularly that of BMP-2, decrease as precursor cells mature. A transient spike in BMP expression occurs as mature chon- drocytes and osteoblasts lay down their respective extracellular matri- ces, but levels decrease during cal- lus remodeling. 18 Although mature osteoblasts and chondrocytes do not normally express large amounts of BMP, they do show increased ex- pression later in the course of frac- ture healing. Recent studies in rats have shown that, during fracture repair, chondrocytes and osteoblasts ex- hibit “up-regulated” expression of certain BMPs. Shortly after the fracture event, a small amount of those BMPs is released from the ex- tracellular matrix of bone. Osteo- progenitor cells in the adjacent periosteum differentiate in re- sponse to this initial release, and BMP-4 levels transiently increase. 19 Within this region, BMP-2 and BMP-4 appear to drive osteoprogeni- tor cells to mature into osteoblasts, as evidenced by up-regulation of Bone Regeneration Journal of the American Academy of Orthopaedic Surgeons 160 Definitions of Specialized Terms Autocrine Denoting the self-stimulating type of hormone function (i.e., the hormone is synthesized and released by an endocrine cell and binds to a receptor on a nearby cell of the same type) Down-regulation Development of a state in which there is a decrease in the number of receptors for a pharmacologic or physiologic substance on the cell surfaces in a given area, such that the cells in that area become less reactive to it Paracrine Denoting the type of hormone function in which the effects of a hormone are restricted to the local environment (i.e., the hormone is synthesized and released by an endocrine cell and binds to a receptor on a nearby cell of a different type) Transgene A gene that has been spliced into a strand of DNA Up-regulation Development of a state in which there is an increase in the number of receptors for a pharmacologic or physiologic substance on the cell surfaces in a given area, such that the cells in that area become more reactive to it Upstream Denoting a region of nucleic acid base sequences on the 5' side of a gene or region of interest BMP-2, BMP-4, and BMP-7 in the mesenchymal cells that infiltrate the fracture site. 20 By 7 to 14 days after fracture, BMP-2 and BMP-4 are at maximal levels in chondro- cyte precursors but at minimal lev- els in hypertrophic chondrocytes and osteoblasts. Once the fracture heals, overall BMP expression is re- duced. The precise mechanisms by which BMPs induce ectopic endo- chondral bone or even normal bone development are still unknown. It is possible that BMPs stimulate undifferentiated pluripotent stem cells to follow chondrogenic and osteogenic lineages over adipogenic or myogenic pathways. 21 Alterna- tively, BMPs may stimulate chon- drogenic and osteogenic lineages directly while inducing apoptosis in adipogenic and myogenic cells. 22 There is particular interest in the potential role of BMP-2 and BMP-7 as therapeutic molecules. Both have been isolated, sequenced, and syn- thesized by using recombinant DNA technology, and both are cur- rently under study in human clini- cal trials. Recombinant human BMP-2 (rhBMP-2) and osteogenic protein-1 (rhOP-1, which is analo- gous to rhBMP-7) have been used successfully to heal critical-sized defects (i.e., osseous defects that, by virtue of their size, will not heal spontaneously) in both the ap- pendicular and the craniomaxillofa- cial skeleton in various animal spe- cies. 13,18,23 However, for the rhBMPs to produce in vivo effects in hu- mans, they must be implanted in an adequate delivery system. Such a delivery system is essential to main- tain the concentration of BMP at the implantation site and to present the molecule to responding cells. In combination with a demineralized bone matrix carrier, rhBMP-2 is capable of inducing bone formation in a 5-mm rat femur defect in a dose-dependent manner. 23 Similar results were obtained with the related protein BMP-7 in 1.5-cm ulnar de- fects in rabbits. 13 These reports, as well as others, have generated en- thusiasm for the use of BMPs in clinical applications in which bone regeneration is needed. However, as this field of research enters its 36th year, a reliable BMP-based ther- apy has not yet become available. Use of BMP for Bone Regeneration The first study to demonstrate the clinical utility of a BMP in a critical- sized defect in humans tested the effectiveness of rhOP-1 combined with a type 1 collagen carrier 24 (Fig. 2). A randomized, double-blind Thomas A. Einhorn, MD, and Cassandra A. Lee, MD Vol 9, No 3, May/June 2001 161 Figure 2 Top, Radiographs showing a fibular defect after implantation of type 1 collagen at 4 months, 6 months, and 1 year. There was no substantial formation of new bone or bridg- ing at any time. Bottom, Radiographs showing a fibular defect after implantation of rhOP-1. There was substantial formation of bone with bridging at 4 months, more at 6 months, and bone formation and remodeling after 1 year. (Courtesy of Stryker Biotech, Hopkinton, Mass.) 4 mo 6 mo 1 yr 4 mo 6 mo 1 yr prospective study was conducted in 24 patients who underwent high tibial osteotomy in which a fibular defect was created to enhance the healing of the osteotomy and to serve as the implantation site for the test materials. First, the investi- gators validated the model of the critical-sized fibular defect by using demineralized bone matrix and untreated control defects. The untreated defects showed no pro- gression toward union, but in the demineralized bone matrix group, bone was formed in the defect from 6 weeks onward. In the second phase of the experiment, the investi- gators compared the osteogenic potential of rhOP-1 combined with a type 1 collagen carrier against type 1 collagen alone. There was no forma- tion of new bone when collagen alone was used; however, in the rhOP-1 group, all but 1 patient showed formation of new bone from 6 weeks onward. These findings suggest that rhOP-1 is osteogenic and capable of regenerating bone in humans. Use of Gene Therapy for Bone Regeneration Many of the diseases that orthopae- dic surgeons treat involve the fail- ure of molecular signals, including those arising from growth factors and cytokines. Deficiencies, includ- ing molecular signaling defects, are potentially correctable with gene therapy. Gene therapy has been attempted in heritable genetic dis- eases, as well as in acquired diseases. Most diseases, however, would require changes in many genes and gene products for expression to occur, and thus cannot be cured by substitution of one normal gene. To increase the efficiency of trans- ferring a gene into a cell, the DNA fragment encoding the therapeutic gene is often introduced within a delivery vehicle called a vector. Be- cause viruses have the ability to en- ter cells and manipulate the cellular machinery of the host, they have been used as vectors in gene therapy protocols. To make viral vectors, vi- ruses are modified to directly deliver the genetic material without the ability to replicate. The most com- mon viral vectors are retroviruses, adeno-associated viruses, adenovi- ruses, and herpes simplex viruses. The retrovirus is the best-developed viral vector. It is able to accommo- date up to 8 kilobases (kb) of genetic material, but inserts it at random locations in the host chromosome. Adeno-associated viruses are able to insert at specific sites and infect nondividing cells, but are able to accommodate only 4 kb of genetic material. Adenoviruses are nonin- tegrating viruses that show high ini- tial genetic expression, which rapidly tapers off. These viruses can infect both dividing and nondividing cells, but are immunogenic because they produce adenoviral proteins. Herpes simplex virus, unlike the other vec- tors, is able to accommodate ex- tremely large segments of genetic material. It can infect nondividing cells but can be cytotoxic and can show transient gene expression. 25 Successful gene therapy requires the gene to be expressed at an ap- propriate level, at the right time, and in the right place. This can be ac- complished with the help of so- called promoters. Promoters are regulatory regions in the DNA, usu- ally situated upstream of the gene, that can both up-regulate and down- regulate gene expression in response to temporal and environmental cues. The most common promoters used in gene therapy are borrowed from cytomegalovirus and simian virus 40. However, although these pro- moters are typically strong effectors of gene expression, they tend to shut down production quickly. Animal studies have shown that demineralized bone matrix, rhBMP-2, and rhBMP-7 can be used to repair critical-sized segmental defects under ideal laboratory conditions. However, these research models rarely mimic the clinical situation, in which defects are often large and healing is hampered by impaired vascularity and scar tissue in the de- fect. Current delivery system tech- nology is limited in that there is no control of the duration of the delivery of BMP. However, genetically ma- nipulated bone marrow cells could serve as an effective delivery vehicle. Lieberman et al 26 tested the effi- cacy of delivery of the BMP-2 gene to a critical-sized bone defect site by means of adenoviral transformation of autologous bone marrow cells ex vivo. Five groups of rats with critical- sized segmental femoral defects were treated with BMP-2–transformed bone marrow cells, rhBMP-2 in a demineralized bone matrix delivery vehicle, or three different types of control materials. Twenty-two of 24 defects in the gene therapy group and all of the defects in the rhBMP-2 group healed after 2 months, as mea- sured by radiographic criteria. However, while rhBMP-2 protein delivery and transformed bone mar- row cells showed equivalent effects in healing of the defects, those defects treated with genetically engineered cells showed advanced callus re- modeling (Fig. 3). Genetically engineered pluripo- tent mesenchymal stem cells have also been used to deliver the BMP-2 gene to a segmental defect. These cells express the transgene in the segmental defect, and the resultant protein affects responding cells in the microenvironment (paracrine effect). This strategy also induces a positive feedback signal to the cells themselves to produce more of the transgene (autocrine effect). Thus, use of cell-mediated gene transfer can induce both autocrine and paracrine activities. Using this approach, Gazit et al 27 compared the effects of BMP-2– engineered mesenchymal stem cells Bone Regeneration Journal of the American Academy of Orthopaedic Surgeons 162 with those of “wild-type” cells— specifically, nonprogenitor cells en- gineered to express BMP-2 and rhBMP-2 protein. Cells were deliv- ered on a collagen sponge to 2.5-mm radial defects in mice. Both types of cells were able to secrete the BMP-2 protein, thus exhibiting a paracrine function. However, the engineered mesenchymal stem cells also exhibited autocrine function by differentiating spontaneously into osteogenic cells. In contrast, wild-type cells differen- tiated only when exogenous rhBMP-2 was added. The pure protein caused new bone formation but did not bridge the gap as effectively as the BMP-2–producing cells did. In this model, engineered pluripotent mes- enchymal stem cells were shown to have greater therapeutic potential than engineered nonmesenchymal cells, nonengineered pluripotent mesenchymal cells, or purified rhBMP-2 protein. Lessons Learned From a Rare Disease The abundance of information from preclinical studies suggests that ani- mals are capable of musculoskeletal tissue regeneration, particularly the formation of cartilage and bone. However, the application of this in- formation to patient care has yet to be realized. Two lines of clinical evidence suggest that the human organism is fully capable of sub- stantial bone regeneration. The first is the observation that slow, steady distraction of an osteotomy, as cre- ated with use of the method of Il- izarov, can regenerate substantial amounts of new bone. The other de- rives from our growing knowledge about the rare but well-recognized metabolic disease fibrodysplasia ossificans progressiva (FOP). In patients with FOP, musculo- skeletal tissues ossify and form bone in orthotopic and heterotopic sites (Fig. 4). For example, injury or activation of undifferentiated mesenchymal cells in fascial planes will lead to the ossification of mus- cles; this has been observed in the biceps, iliopsoas, and other muscles of the appendicular skeleton. Shafritz et al, 28 in an immunohis- tochemistry study, showed that the lymphocytes of 11 of 12 patients with FOP demonstrated overex- pression of BMP-4, compared with only 2 of 26 control subjects. It was shown further that BMP-4 is the only member of the BMP family that demonstrates this effect, and that lymphocytes capable of BMP-4 expression circulate in the peripheral blood of patients with FOP. Thus, lymphocytes capable of expressing this morphogen may be recruited to sites of connective tissue injury, where they may release BMP pro- tein. Type IV collagen, a major com- Thomas A. Einhorn, MD, and Cassandra A. Lee, MD Vol 9, No 3, May/June 2001 163 A B C D E Figure 3 Radiographs showing critical-sized femoral defects 2 months after treatment with five different materials: A, BMP-2–producing bone marrow cells; B, rhBMP-2; C, β-galactosidase–producing rat bone marrow cells; D, noninfected rat bone marrow cells; and E, de- mineralized bone matrix alone. Note that the rhBMP-2–treated defects show lacelike trabecular bone filling the defect. Defects treated with the BMP-2–producing bone marrow cells showed a dense, coarse trabecular framework, which remodeled to form a new cortex. None of the other treatment groups showed healing. (Reproduced with permission from Lieberman JR, Daluiski A, Stevenson S, et al: The effect of regional gene therapy with bone morphogenetic protein-2-producing bone-marrow cells on the repair of segmental femoral defects in rats. J Bone Joint Surg Am 1999;81:905-917.) ponent of the basement membrane of endothelial and muscle cells, avidly binds BMP-4, resulting in a local increase in BMP-4 concentra- tion. At high concentrations, BMP-4 acts as a morphogen and is capable of up-regulating its own expression, which leads to the development of preosseous fibroproliferative lesions. These findings suggest that there is a definable human response to BMP-4 expression as long as that expression is delivered to the re- sponding cell in the appropriate way—in the case of FOP, by a lym- phocyte. Although the bone formed in this disease is unwanted, the observation that cell-mediated ex- pression of a morphogen leads to substantial bone regeneration in humans is compelling. Summary Orthopaedic surgeons tend to re- gard the use of molecular and gene treatment strategies as future pro- tocols for regeneration of the tis- sues that they treat every day— bone, cartilage, muscle, tendon, and ligament. However, the body does not naturally form tissues in an isolated fashion. Development of the human organism, particular- ly during embryogenesis, involves the simultaneous formation and modeling of several tissues and organs. It has recently been discov- ered that various BMPs affect not only bone and cartilage develop- ment, but also the formation of the kidneys, heart, skin, eyes, and other tissues. This suggests that BMPs are not entirely within the domain of the musculoskeletal system, but rather are a linkage of that system to others that constitute the human organism. The ability to under- stand and harness this power holds unlimited potential for the treat- ment of skeletal and nonskeletal in- juries and diseases. Indeed, the next generation of scientific discovery could yield substantial advances in patient care. Bone Regeneration Journal of the American Academy of Orthopaedic Surgeons 164 References 1. Brighton CT, Hunt RM: Early histo- logical and ultrastructural changes in medullary fracture callus. J Bone Joint Surg Am 1991;73:832-847. 2. McKibbin B: The biology of fracture healing in long bones. J Bone Joint Surg Br 1978;60:150-162. 3. Einhorn TA: The cell and molecular biology of fracture healing. Clin Orthop 1998;355(suppl):S7-S21. 4. Bolander ME: Regulation of fracture repair by growth factors. Proc Soc Exp Biol Med 1992;200:165-170. 5. Einhorn TA, Majeska RJ, Rush EB, Levine PM, Horowitz MC: The ex- pression of cytokine activity by frac- ture callus. J Bone Miner Res 1995;10: 1272-1281. 6. Ilizarov GA: The tension-stress effect on the genesis and growth of tissues: Part I. The influence of stability of fixa- tion and soft-tissue preservation. Clin Orthop 1989;238:249-281. 7. Li G, Simpson AHDW, Kenwright J, Triffitt JT: Effect of lengthening rate on angiogenesis during distraction osteo- genesis. J Orthop Res 1999;17:362-367. 8. Dagher F, Roukoz S: Compound tibial fractures with bone loss treated by the Ilizarov technique. J Bone Joint Surg Br 1991;73:316-321. 9. Aronson J, Johnson E, Harp JH: Local bone transportation for treatment of intercalary defects by the Ilizarov tech- nique: Biomechanical and clinical con- siderations. Clin Orthop 1989;243:71-79. 10. Aronson J: Experimental and clinical ex- perience with distraction osteogenesis. Cleft Palate Craniofac J 1994;31:473-482. 11. Urist MR: Bone: Formation by autoin- duction. Science 1965;150:893-899. 12. Wang EA, Rosen V, D’Alessandro JS, et al: Recombinant human bone mor- phogenetic protein induces bone for- mation. Proc Natl Acad Sci USA 1990; 87:2220-2224. 13. Cook SD, Baffes GC, Wolfe MW, Sam- path TK, Rueger DC, Whitecloud TS III: The effect of recombinant human os- teogenic protein-1 on healing of large segmental bone defects. J Bone Joint Surg Am 1994;76:827-838. 14. Lovell TP, Dawson EG, Nilsson OS, Urist MR: Augmentation of spinal fu- sion with bone morphogenetic protein in dogs. Clin Orthop 1989;243:266-274. 15. Glowacki J, Kaban LB, Murray JE, Folkman J, Mulliken JB: Application of the biological principle of induced osteogenesis for craniofacial defects. Lancet 1981;1:959-962. 16. Mulliken JB, Glowacki J, Kaban LB, Folkman J, Murray JE: Use of dem- ineralized allogeneic bone implants for the correction of maxillocraniofa- cial deformities. Ann Surg 1981;194: 366-372. 17. Wozney JM, Rosen V: Bone morpho- genetic proteins and their gene expres- sion, in Noda M (ed): Cellular and Mo- lecular Biology of Bone. San Diego: Aca- demic Press, 1993, pp 131-167. 18. Bostrom MP: Expression of bone mor- phogenetic proteins in fracture healing. Clin Orthop 1998;355(suppl):S116-S123. Figure 4 Lateral radiograph of the knee of an adult patient with FOP shows exten- sive ossification bridging the femur and the tibia. (Reproduced with permission from Kaplan FS, Strear CM, Zasloff MA: Radio- graphic and scintigraphic features of mod- eling and remodeling in the heterotopic skeleton of patients who have fibrodys- plasia ossificans progressiva. Clin Orthop 1994;304:238-247.) 19. Nakase T, Nomura S, Yoshikawa H, et al: Transient and localized expression of bone morphogenetic protein 4 mes- senger RNA during fracture healing. J Bone Miner Res 1994;9:651-659. 20. Onishi T, Ishidou Y, Nagamine T, et al: Distinct and overlapping patterns of localization of bone morphogenetic protein (BMP) family members and a BMP type II receptor during fracture healing in rats. Bone 1998;22:605-612. 21. Yamaguchi A, Katagiri T, Ikeda T, et al: Recombinant human bone morpho- genetic protein-2 stimulates osteoblas- tic maturation and inhibits myogenic differentiation in vitro. J Cell Biol 1991;113:681-687. 22. Katagiri T, Yamaguchi A, Komaki M, et al: Bone morphogenetic protein-2 converts the differentiation pathway of C2C12 myoblasts into the osteoblast lineage. J Cell Biol 1994;127(6 pt 1): 1755-1766. 23. Yasko AW, Lane JM, Fellinger EJ, Rosen V, Wozney JM, Wang EA: The healing of segmental bone defects, induced by recombinant human bone morphogenetic protein (rhBMP-2): A radiographic, histological, and biome- chanical study in rats. J Bone Joint Surg Am 1992;74:659-670. 24. Geesink RG, Hoefnagels NH, Bulstra SK: Osteogenic activity of OP-1 bone morphogenetic protein (BMP-7) in a human fibular defect. J Bone Joint Surg Br 1999;81:710-718. 25. Mulligan RC: The basic science of gene therapy. Science 1993;260:926-932. 26. Lieberman JR, Daluiski A, Stevenson S, et al: The effect of regional gene therapy with bone morphogenetic protein-2- producing bone-marrow cells on the repair of segmental femoral defects in rats. J Bone Joint Surg Am 1999;81:905-917. 27. Gazit D, Turgeman G, Kelley P, et al: Engineered pluripotent mesenchymal cells integrate and differentiate in re- generating bone: A novel cell-mediated gene therapy. J Gene Med 1999;1:121-133. 28. Shafritz AB, Shore EM, Gannon FH, et al: Overexpression of an osteogenic morphogen in fibrodysplasia ossifi- cans progressiva. N Engl J Med 1996; 335:555-561. Thomas A. Einhorn, MD, and Cassandra A. Lee, MD Vol 9, No 3, May/June 2001 165