Nodular osteochondrogenic activity in soft tissue surrounding osteoma in neurogenic para osteo-arthropathy: morphological and immunohistochemical study
© Youssefian et al; licensee BioMed Central Ltd. 2004
Received: 30 December 2003
Accepted: 25 November 2004
Published: 25 November 2004
Neurogenic Para-Osteo-Arthropathy (NPOA) occurs as a consequence of central nervous system injuries or some systemic conditions. They are characterized by bone formation around the main joints.
In order to define some biological features of NPOAs, histological and immunohistological studies of the soft tissue surrounding osteoma and Ultrasound examination (US) of NPOA before the appearance of abnormal ossification on plain radiographs were performed.
We have observed a great number of ossifying areas scattered in soft tissues. US examination have also shown scattered ossifying areas at the early stage of ossification. A high osteogenic activity was detected in these tissues and all the stages of the endochondral process were observed. Mesenchymal cells undergo chondrocytic differentiation to further terminal maturation with hypertrophy, which sustains mineralization followed by endochondral ossification process.
We suggest that periosteoma soft tissue reflect early stage of osteoma formation and could be a model to study the mechanism of osteoma formation and we propose a mechanism of the NPOA formation in which sympathetic dystony and altered mechanical loading induce changes which could be responsible for the cascade of cellular events leading to cartilage and bone formation.
Neurogenic Para Osteo-Arthropathies (NPOA) occurs in patients with brain or spinal cord injury, hemiplegias, various encephalopathies, tetanus  or neurological disregulation . In this process, new bone named "osteoma" forms in extraskeletal areas which in normal condition do not ossify. NPOA were first described by Dejerine and Cellier  from observations of medullary wounded soldiers. They proposed the term NPOA, though other terms are used, such as: neurogenic osteoma, ossifying myositis in paraplegic, ectopic ossification, heterotopic ossification, etc. NPOAs have also been described as complications of many systemic diseases , acute pancreatitis, toxic syndromes and others .
The first clinical manifestations are local inflammatory signs, tumefaction and progressively limited range of motion of the involved joint region. Those appear between the second and tenth weeks after the onset of the pathological condition . Despite anti-inflammatories treatment to prevent NPOA , excision of the newly formed bone called "osteoma", is the only known therapy.
As shown by radiographic and scintigraphic observations, heterotopic bone formation evolves from an early appearance of soft tissue densification and attenuation of the muscle signal to a mineral signal . After 6 months, osteoma rarely increases in amount, but some further maturation occurs. As an assumption based on the fading of technetium fixation, the lesion is supposed to be mature after 1 to 1.5 years [9, 10]. Hence, the process of NPOA formation seems to be frozen at the time of osteoma mineralization.
Very little is known about the pathophysiology of NPOA formation. Assuming such a freezing of the process of NPOA formation and an involvement of the periosteoma tissues in the reported relapses following surgery, we postulated that the periosteoma soft tissues could show some of the very early stages of the NPOA formation. We performed histological, histochemical and immunohistochemical studies of soft tissues dissected from the periphery of osteomas. We used samples of varying age lesions and searched for the main osteogenic and chondrogenic markers: alkaline phosphatase (ALP) activity, type I collagen and osteocalcin (OCN) for the bone [11–13], and type II collagen, sulfated and acid glycosaminoglycans, type X collagen and Vascular Endothelial Growth Factor (VEGF) for the cartilage . In the light of our results, we propose a model of NPOA formation.
a)Specimen processing and histochemicals
The 28 specimens were obtained from 27 patients undergoing orthopedic surgery for osteoma excision. NPOA's were localized on: elbows (7), hips (18) and knees (3). The time from the neurologic insult ranged from 5 months to 216 months. The initial conditions were: 11 Brain Injuries (BI), 3 Spinal Cord Injury (SCI), 1 BI plus SCI, 4 strokes, and 9 patients sustained coma of various etiology (legionellose, anoxia, toxic condition, pneumonia, suicide attempt using neuroplegic).
The anti-human monoclonal mouse antibodies against type II collagen (NeoMarkers Inc., CA, USA), OCN (Interchim, Montlucon, France), anti-human polyclonal goat antibody against type I collagen (Santacruz Biotechnology, CA, USA), anti-rat polyclonal rabbit antibody against type X collagen (Calbiochem-Novabiochem, CA, USA) and anti-human polyclonal rabbit antibody against VEGF (Santacruz Biotechnology, CA, USA) were used at 1 mg/ml. Peroxidase-conjugated goat anti-mouse was purchased from Immunovision Technologies (CA, USA), peroxidase-conjugated goat anti-rabbit from Dako (Dako Corporation, CA, USA) and peroxidase-conjugated donkey anti-goat from Santacruz Biotechnology (CA, USA). They were used at a 1/20, 1/50 and 1/100 dilution, respectively. Frozen sections post-fixed with acetone were treated with hyaluronidase (5 mg/ml in Tris Buffered Saline) one hour at 37°C. For type X collagen immunostaining slides were treated first 15 minutes at 37°C with hyaluronidase (5 mg/ml in PBS), washed two times in PBS, then 15 minutes with chondroitinase (2 U/ml in PBS), and washed two times in PBS. Immunohistochemistry was performed using ABC method. Briefly, endogenous peroxydase activity was eliminated with 0.3% H2O2 until total clearing of oxygen bubbles. Non-specific protein binding was performed with 10% non-immune serum, same host as secondary antibody, in PBS with 1% Bovin Serum Albumin (Sigma, Saint Quentin Fallavier, France). Sections were then incubated with primary antibodies for 1 hour at room temperature, or 24 hours at 4°C with anti-type I collagen. Excess antibody was removed by washing the sections with PBS. Sections were incubated 1 hour with horseradish peroxydase-labeled secondary antibody diluted in PBS. 3-3'diaminobenzidine (DAB) solution (Dako Corporation, CA, USA) was then added in order to obtain staining. Sections were counter-stained with hematoxylin-eosin or nuclear red/eosin, dehydrated, and mounted with Mountex medium (Microm, France). Controls were systematically performed omitting the primary antibody. Slides were examined by light microscopy using a Leica DMR microscope (Leica Micro-Systems, Reil-Malmaison, France).
c)Ultrasound (US) Examination
Most of the patients were referred to Raymond Poincaré teaching hospital, at the time of surgery. US examination was performed when NPOA was clinically suspected in five patients whose rehabilitation has begun. Five hips were explored by US in two BI and three SCI. Linear 8 to 15 MHz and sectorial 4 MHz transducers (Sequoia Acuson-Siemens Erlangen) were used. A sectorial low frequency transducer was used because in the hip area NPOA can be very large and very deep especially in the gluteus area compared to the subcutaneous plane. US examination was combined with color and energy Doppler. In all cases a plain film was obtained the same day as the US examination.
a)Histological and immunohistological studies
Bands of eosinophilic material stained partially by Von Kossa underlined some borders of the mineralized cartilage. These osteoid-like bands were lined by cells which morphologically appeared to be osteoblast-like cells.
To confirm the osteoblastic nature of these cells, immunolabelling for OCN was performed (Fig 5c). In front of the osteoblastic-like cells a consistent matricial OCN immunoreactivity was evident onto the eosinophilic matrix underlining the mineralized cartilage. We have also observed deposition of OCN into woven bone formed adjacent to cartilage.
In order to document the collagenic composition of these tissues and confirm their osseous or cartilaginous nature, immunolabelling for type I, II and X collagens were performed.
Conspicuous immunolabelling for type X collagen was observed in most of the hypertrophic chondrocytes and in their matrix. The immunoreactivity became more intense nearby the mineralization front. In addition the matrix of the fibrous and non cartilage-like tissue around some mineralized areas was unexpectedly labelled (Fig 5d). Control samples in which primary antibody incubation was omitted were clearly negative (data not shown).
Distribution of type II collagen was limited to the matrix of non hypertrophic and prehypertrophic chondrocytes with high ALP activity (Fig 5e). Nevertheless some type II collagen immunoreactivity could sometimes be detected in the hypertrophic areas.
In order to determine the chronology of events at work in the described endochondral ossification, we performed serial cryosectioning of samples in which a cortical bone followed soft tissue. These samples seem to be appropriate to have all stages of osteogenesis
To further confirm the process of endochondral ossification, we decided to search for VEGF expression. Immunolabelling of these tissues with VEGF monoclonal antibody, showed a labelling of the hypertrophic chondrocytes as well as an intense labelling of activated osteoblats lining the osteoid. Some clusters of rounded cells also expressed VEGF in the fibrous part of these preparations (Fig 8f).
c)Ultrasound examination and digital radiographs of suspected NPOA
Classical zone phenomenon previously described in the literature [15, 16] was visible with a central hypo echoic area surrounded by small (less than one centimeter) hyper echoic nodules with posterior attenuation (Fig 9b). At this stage color Doppler examination showed increasing angiogenesis with the appearance of large vessels in the tumor mass (Fig 9c).
The zone phenomenon became visible on the second US examination performed one week later (Fig 9d).
The opacity of the early ossification became slightly visible on plain films only two weeks after the US detection of zone phenomenon (Fig 9e).
NPOA pathogenesis is still poorly understood, and the exact environmental and humoral conditions underlying the ossifying process are not clear. In this study we postulated that periosteoma soft tissues display interrupted early stages of osteoma formation, which could help us to understand the chronology as well as the mechanism of osteoma formation. Thus, histological and immunohistological experiments were performed on 28 specimens. Moreover, ultrasound examination of suspected NPOA tumor on five patients permitted to follow osteoma formation in the early stages before ossification.
Histological studies have shown varying amount of muscle and connective tissue degeneration in which some areas underwent reorganization. Islands of cartilage, woven bone, and mature lamellar bone were a constant finding in most of specimens, whatever the estimated age of the studied lesion. The spatial organization of chondrocytes was reminiscent of the epiphyseal growth plate or of the fracture callus organization . In the developmental pathway leading to skeletogenesis, undifferentiated mesenchymal cells pass sequentially through at least four differentiation stages: committed mesenchymal cells which produce type I collagen and possibly basal level of type II collagen, quiescent chondrocytes, then proliferating chondrocytes characterized by the synthesis of a large amount of type II collagen and sulfated proteoglycans, and ultimately hypertrophic chondrocytes characterized by the synthesis of type X collagen. Then these hypertrophic chondrocytes allow the mineralization of the matrix elaborated and induce vascular invasion by releasing VEGF [18, 19]. Studies of our specimens showed the same sequence of events. These results suggest that endochondral osteogenesis is the major pathway in the NPOA bone formation. Nevertheless in view of some features suggesting bone deposition without any cartilage scaffold we cannot exclude the occurrence of membranous bone formation.
Some sections showed a high expression of type X collagen in the hypertrophic chondrocytes areas, and unexpectedly in non differentiated mesenchymal tissue. As this labelling did not occur in other fibrocellular areas it was unlikely to be produced by a binding of the antibody to some other matricial component of the extra cellular matrix. It was shown, that type X collagen is not only associated with chondrocyte proliferation and hypertrophy, but also with resting chondrocytes, cells at the border of the perichondrium and resting cartilage of the fetal femoral head .
The finding of still degenerating muscular fibers and early chondro-osteogenesis accompanied by heavy ALP activity in large parts of the soft tissues in old lesions (till 8 years) is singular. Except when serial sectioning was performed the studied specimens were carefully dissected from the osteoma during the surgery or at the fixation time. Therefore the extra-osteoma localization of these tissues can without any doubt be assumed. In one study  "recent POA" was described as a kind of fibrocellular tissue including vascular stasis, overabundance of micro vessels, myolysis, edematous fibrocellular tissue, with chondrogenesis, osteogenesis and lamellar bone apposition on mineralized structures. This description of "recent POA" is perfectly in agreement with our description of the periosteoma tissues. However, we found each of these elements notwithstanding the advanced age of some lesions. Most of our specimens contained clusters of ALP positive cells in the undifferentiated fibrous connective tissue, suggesting the presence of preosteogenic cells. This fact and the presence of cartilage and bone at varying stages of maturity are indicative of a persistent chondro-osteogenic activity in these tissues. This point sounds of interest as regards to contingencies of heterotopic bone formation relapse following surgical excision. The occurrence of relapses was reported to correlate neither to the classical criteria of osteoma maturation nor to the amount of heterotopic bone left after excision nor to the age of the lesion [22–24]. The occurrence of relapses could be linked to the activation of the still process.
It was claimed that osteoma develops in the periphery of a muscle in which some myofibers undergo degeneration [25, 26], and that osteoma involves the muscles. However the endochondral process of bone formation described here is in agreement with the results of various bone induction experiments in muscle [27–30]. Moreover, US examination of suspected NPOA tumor at the early stage showed huge focal disorganisation of muscle surrounding the hip joint and disappearance of normal longitudinal muscular striation of the psoas iliaque muscle replaced by masses with heterogeneous echostructure. This finding argues in favour of an intramuscular beginning of the process. We could not set up the place and part of the muscle degeneration process in the heterotopic bone formation. Though the process described here resembles by some aspects the Fibrodysplasia Ossificans Progressiva (FOP) where endochondral ossification was demonstrated in muscle and adjacent connective tissue [31, 32]. These reports combined with our finding of endomysium-perimysium hyperplasia in the degenerating muscles with ALP activity and US result, could suggest a role for the muscular tissue and especially mesenchymal cells from endomysium and perimysium in the setting of the heterotopic bone formation process. Vascular disorders, such as vascular disruption or compression  and venous stasis together with a cascade of inflammatory reactions including release of enzymes from necrotic tissues and α-adrenergic mediated vasoconstriction, lead to the formation of hypoxic zones beneath the nervous injury.
On the other hand it has been shown that hypoxia promotes chondrocyte differentiation [34, 35]. In our specimens, high amount of thrombotic vessels were indicative of the hypoxic status of these tissues. Color Doppler examination of tumor during the first clinical signs of NPOA, showed an avascular area in which some vessels appeared. These signals became more intense one week later. This tends to confirm hypoxic status of the initial lesion. This suggests that a timely defined hypoxic condition in tissue induces chondrocytes differentiation.
The volume of the tumor is acquired during the very early stages of NPOA. After tumefaction there is no real change in the tumor size . Moreover chondrocytes in the hypertrophic stage increase volume 10 times . So the chondrocyte hypertrophy could be the cause of the tumefaction and so determines the final size of the tumor. Then, the related vascular sign occurs and color Doppler showed the first sign of increasing angiogenesis.
On tissue section, VEGF immunolabelling revealed a more intense expression by osteoblasts and osteoblast-like cells, in addition to its expected expression by hypertrophic chondrocytes. VEGF induces neoangiogenesis and then endochondral ossification occurs. The restoration of normoxic conditions promote the onset of lamellar ossification and hamper any other de novo cartilaginous differentiation. US examination of the same NPOA tumor one week later showed scattered ossifying nodules. This nodular activity is in accordance with our histological finding in periosteoma soft tissue. Extense of the hypoxic area determines the size of each nodule. Areas with important hypoxia induce larger cartilage zones which could join together after ossification but small nodules remain scattered in periosteoma soft tissue. These results confirm that, periosteoma soft tissue has the same pattern of early stage of osteoma ossification, and could be a model of ossification for further studies.
Moreover, NPOA occurs in neurologically deficient patients with altered mechanical loading. Mechanical loading is of pivotal importance to the development, function and repair of all tissues in the musculoskeletal system. In nonfunctional joints, as it is the case of these patients, the absence or reduction of intermittent hydrostatic pressure in the articular cartilage could permit cartilage degeneration and the progressive advance of the ossification These mechanical influence could indeed shed light on the finding that osteomas only occur near the main joints.
Moreover, Carter and associates have also shown that intermittently applied shear stresses (or strain energy) promote endochondral ossification and that intermittently applied hydrostatic compression inhibits or prevents cartilage degeneration and ossification. Thus, the imbalance of these forces among these patients can promote endochondral ossification of the cartilage nodules in areas of high shear (deviatoric) stresses.
Urist demonstrated that the induced endochondral bone is resorbed once the inductive agent has disappeared . We have not seen many osteoclasts nor other multinucleated cells in ours preparations, and the literature does not report on NPOA regression except in people under 15 years of age . Therefore it would be interesting to study the regulation of osteoclasts and the remodelling in such model.
We therefore propose a model of the NPOA lesion formation. The sympathetic hyperactivity causes major changes in the peripheral vascular dynamics. As related some of these changes end in vascular stasis and/or thrombosis . Edema follows on in the connective tissue which sustains some amount of necrosis. The trauma, subsequent neurological conditions and perhaps systemic factors [40, 41] induce major changes in these tissues. Regenerating celles under low oxygen pressure/high dilatational hydrostatic forces [34, 42, 43] transmogrify themselves into chondrogenic cells. Then cartilage differentiation gets moving on until hypertrophy of the chondrocytes and cartilage matrix mineralization. The cell hyperplasia and hypertrophy of the chondrocytes could account for the solid swelling clinically noticed soon after the onset of the clinical signs of NPOA. Like they did in the developing limb, some competent cells lining the cartilage rudiments undergo the osteoblastic differentiation and lay down osteoid on the cartilage. Concomitantly the cartilage hypertrophy induces angiogenesis, osteoid deposition, and some extent of cartilage resorption. Remodelling of the mineralized cartilage and woven bone occurs. The osteoma could then control the process and inhibit any further bone formation.
Some questions remain which deserve further studies. Why do NPOAs form only around the main joints? Why are they not resorbed as does any intramuscularly implanted bone graft . What freezes the osteoma bone growth and the process of bone formation? Studies are ongoing in order to find some clues about the regulation of this heterotopic bone formation.
In conclusion, our results demonstrate that periosteoma soft tissues are a replica of the early stages of osteoma formation, and could be used as a model for NPOA formation. We propose also a mechanism for osteoma formation in which hypoxia is a major cause of nodular osteoinduction and chondrocyte differentiation. Combination of hypoxia and applied shear stresses induce endochodral ossification. Finally our results indicate implication of different types of mesenchymal cells in NPOA formation but US examination support specially muscular origin hypothesis.
- Chantraine A, Nusgens B, Lapiere CM: Biochemical analysis of heterotopic ossification in spinal cord injury patients. Paraplegia. 1995, 33: 398-401.View ArticlePubMedGoogle Scholar
- Ebinger T, Roesch M, Kiefer H, Kinzl L, Schulte M: Influence of etiology in heterotopic bone formation of the hip. J Trauma. 2000, 48: 1058-1062. 10.1097/00005373-200006000-00010.View ArticlePubMedGoogle Scholar
- Dejerine A, Cellier A: Paraosteoarthropathies of paraplegic patients by spinal cord lesion. Clinical and roentgenographic study. Clin Orthop. 1991, 263: 3-12.PubMedGoogle Scholar
- Richards AM, Klaassen MF: Heterotopic ossification after severe burns: a report of three cases and review of the literature. Burns. 1997, 23: 64-68. 10.1016/S0305-4179(96)00074-5.View ArticlePubMedGoogle Scholar
- Taly AB, Nair KP, Kumar MV, Jayakumar PN, Vasudev MK, Ravishankar D, Kalaivant PL, Padankatty BS, Murali T: Heterotopic ossification in non-traumatic myelopathies. Spinal Cord. 1999, 37: 47-49. 10.1038/sj.sc.3100751.View ArticlePubMedGoogle Scholar
- Hernandez AM, Forner JV, de la FT, Gonzalez C, Miro R: The para-articular ossifications in our paraplegics and tetraplegics: a survey of 704 patients. Paraplegia. 1978, 16: 272-275.View ArticlePubMedGoogle Scholar
- Kjaersgaard-Andersen P, Nafei A, Teichert G, Kristensen O, Schmidt SA, Keller J, Lucht U: Indomethacin for prevention of heterotopic ossification. A randomized controlled study in 41 hip arthroplasties. Acta Orthop Scand. 1993, 64: 639-642.View ArticlePubMedGoogle Scholar
- Bressler EL, Marn CS, Gore RM, Hendrix RW: Evaluation of ectopic bone by CT. AJR Am J Roentgenol. 1987, 148: 931-935.View ArticlePubMedGoogle Scholar
- DeLee J, Ferrari A, Charnley J: Ectopic bone formation following low friction arthroplasty of the hip. Clin Orthop. 1976, 53-59.Google Scholar
- Nilsson OS: Heterotopic ossification. Acta Orthop Scand. 1998, 69: 103-106.View ArticlePubMedGoogle Scholar
- Caplan AI: Embryonic development and the principles of engineering. Novartis Found Symp. 2003, 249: 17-25. Discussion 25-33, 170-4, 239-41. ReviewView ArticlePubMedGoogle Scholar
- Bruder SP, Caplan AI: First bone formation and the dissection of an osteogenic lineage in the embryonnic chick tibia is revealed by monoclonal antibodies against osteoblasts. Bone. 1989, 10: 359-375. 10.1016/8756-3282(89)90133-6.View ArticlePubMedGoogle Scholar
- Stein GS, Lian JB: Molecular mechanisms mediating proliferation/differentiation interrelationship during progressive development of the osteoblast phenotype. Endoc Rev. 1993, 14: 424-442. 10.1210/er.14.4.424.View ArticleGoogle Scholar
- Caplan AI: Mesenchymal stem cells: building blocks for molecular medicine in the 21st century. Trends Mol Med. 2001, 7: 259-64. 10.1016/S1471-4914(01)02016-0.View ArticlePubMedGoogle Scholar
- Cassar-Pullicino VN MCMBDAHMCIWPRGEMW: Sonographic diagnosis of heterotopic bone formation in spinal injury patients. Paraplegia. 1993, 31: 40-50.View ArticlePubMedGoogle Scholar
- Carlier RY MDDP: Edited by: . 1995, Paris, 78-8. L'échographie dans l'exploration des para-ostéo-arthropathies neurogènes (POA). , Masson, Les paraostéorarthropathies neurogènes.
- Campbell JT, Kaplan FS: The role of morphogens in endochondral ossification. Calcif Tissue Int. 1992, 50: 283-289. 10.1007/BF00296294.View ArticlePubMedGoogle Scholar
- Horner A, Bishop NJ, Bord S, Beeton C, Kelsall AW, Coleman N, Compston JE: Immunolocalisation of vascular endothelial growth factor (VEGF) in human neonatal growth plate cartilage. J Anat. 1999, 194 ( Pt 4): 519-524. 10.1046/j.1469-7580.1999.19440519.x.View ArticleGoogle Scholar
- Chen Q, Johnson DM, Haudenschild DR, Goetinck PF: Progression and recapitulation of the chondrocyte differentiation program: cartilage matrix protein is a marker for cartilage maturation. Dev Biol. 1995, 172: 293-306. 10.1006/dbio.1995.0024.View ArticlePubMedGoogle Scholar
- Nerlich AG, Kirsch T, Wiest I, Betz P, von der MK: Localization of collagen X in human fetal and juvenile articular cartilage and bone. Histochemistry. 1992, 98: 275-281. 10.1007/BF00270010.View ArticlePubMedGoogle Scholar
- Rossier AB, Busat P, Infante F, Zender R, Courvoisier B, Muheim G, Donath A, Vasey H, Taillard W, Lagier R, Garbiani G, Baud CA, Pouezat JA, Very JM, Hachen HJ: Current facts on para-osteo-arthropathy (POA). Paraplegia. 1973, 11: 36-78.View ArticleGoogle Scholar
- Denormandie P, Viguie G, Denys P, Dizien O, Carlier R: Results of excision of heterotopic new bone around the elbow in patients with head injuries. Chir Main. 1999, 18: 99-107.PubMedGoogle Scholar
- Denormandie P, Rech C, Gleizes V, Dizien O, Feron JM: Les ParaOstéo-Arthorpathies neurogènes du genou. A propos d'une série de 14 cas. 1999, 74ême réunion annuelle SOFCOT, Paris, 9-10-11-12 novembre 1999:Google Scholar
- Denormandie P, Mailhan L, Chaory K, Carlier R, Denys P, Bussel B: Résultats analytiques et fonctionnels du traitement chirurgical de 55 ParaOstéoArthropathies Neurogènes de hanches chez 37 patients. 1999, 74ême réunion annuelle SOFCOT, Paris, 9-10-11-12 novembre 1999:Google Scholar
- Chantraine A, Minaire P: Para-Osteo-Arthropathies. A new theory and mode of treatment. Scand J Rehabil Med. 1981, 13: 31-37.PubMedGoogle Scholar
- Chantraine A, Very JM, Baud CA: Etude histologique et biophysique des ossifications ectopiques post-traumatiques. J Traumatol Sport. 1989, 6: 128-132.Google Scholar
- Selye H, Lemire Y, Bajusz E: Induction of bone, cartilage and hemopoietic tissue by subcutaneously implanted tissue diaphragms. Roux Archiv. 1960, 151: 572-585. 10.1007/BF00577813.View ArticleGoogle Scholar
- Selye H, Mahajan S, Mahajan RS: Histogenesis of experimentally induced myositis ossificans in the heart. Am Heart J. 1967, 195-201. 10.1016/0002-8703(67)90148-2.Google Scholar
- Mizuno K, Mineo K, Tachibana T, Sumi M, Matsuabara T, Hirohata K: The osteogenic potential of fracture haematoma subperiosteal and intramuscular transplantation of the haematoma. J Bone Joint Surg Br. 1990, 72: 822-829.PubMedGoogle Scholar
- Urist MR, Hay PH, Dubuc FL, Büring K: Osteogenic competence. Clin Orthop. 1969, 64: 194-220.PubMedGoogle Scholar
- Kaplan FS, Tabas JA, Gannon FH, Finkel G, Hahn GV, Zasloff MA: The histopathology of Fibrodysplasia Ossificans Progressiva. J Bone Joint Surg Am. 1993, 75A: 220-230.Google Scholar
- Gannon FH, Glaser D, Caron R, Thompson LDR, Shore EM, Kaplan FS: Mast cell involvemant in Fibrodysplasia Ossificans Progressiva. Human Pathol. 2001, 842-848. 10.1053/hupa.2001.26464.Google Scholar
- Gallien P, Nicolas B, Le Bot MP, Robineau AS, Rivier I, Sarkis S, Brissot R: Paraostéoarthropathies et compressions vasculaires. Rev Rhum Ed Fr. 1994, 61: 823-828.PubMedGoogle Scholar
- Hall BK: HYpoxia and differentiation of cartilage and bone from common general cells in vitro. Life Sci. 1969, 8: 553-558. 10.1016/0024-3205(69)90255-0.View ArticlePubMedGoogle Scholar
- Schipani E, Ryan HE, Didrickson S, Kobayashi T, Knight M, Johnson RS: Hypoxia in cartilage: HIF-1alpha is essential for chondrocyte growth arrest and survival. Genes Dev. 2001, 15: 2865-2876.PubMedPubMed CentralGoogle Scholar
- Farnum CE, Lee R, O'Hara K, Urban JP: Volume increase in growth plate chondrocytes during hypertrophy: the contribution of organic osmolytes. Bone. 2002, 30: 574-581. 10.1016/S8756-3282(01)00710-4.View ArticlePubMedGoogle Scholar
- Carter DR OTEFDPSDJ: Influences of mechanical stress on prenatal and postnatal skeletal development. Clin Orthop. 1987, 219: 237-250.PubMedGoogle Scholar
- Wong M CDR: Mechanical stress and morphogenetic endochondral ossification of the sternum. J Bone Joint Surg Am. 1988, 70: 992-1000.PubMedGoogle Scholar
- Garland DE, Shimoyama ST, Lugo C, Barras D, Gilgoff I: Spinal cord insults and heterotopic ossification in the pediatric population. Clin Orthop. 1989, 303-310.Google Scholar
- Feldman Z, Contant CF, Pahwar R, Goodman JC, Robertson CS, Narayan RK, Grossman RG: The relationship between hormonal mediators and systemic hypermetabolism after severe head injury. J Trauma. 1993, 34: 806-816.View ArticlePubMedGoogle Scholar
- KURER MHJ, Khoker MA, Dandona P: Human osteoblast stimulation by sera from paraplegic patients with heterotopic ossification. Paraplegia. 1992, 30: 165-168.View ArticlePubMedGoogle Scholar
- Brighton CT, Fox JL, Setzer D: In vitro growth of bone and cartilage from rat periosteum under various oxygen tension (abstract). 22nd Annual ORS 1976;jannuary 28-30 :p 51. 1976Google Scholar
- Bassett: Influence of oxygen concentration and mechanical factors in differentiation of connective tissues in vivo. Nature. 1961, 190: 460-View ArticlePubMedGoogle Scholar
- Vargervik K: Critical sites for new bone formation. in Bone Grafts and Bone Substitutes. 1992, M.B.Habal ,A.H.Reddi eds; W.B.Saunders:PHILADELPHIA ,Penn: 112-120.Google Scholar
- The pre-publication history for this paper can be accessed here:http://www.biomedcentral.com/1471-2474/5/46/prepub