Many pre-clinical models of atrophic non-union do not reflect the clinical scenario, some create a critical size defect, or involve cauterization of the tissue which is uncommonly seen in patients. Atrophic non-union is usually developed following high energy trauma leading to periosteal stripping. The most recent reliable model with these aspects involves creating a non-critical gap of 1mm with periosteal and endosteal stripping. However, this method uses an external fixator for fracture fixation, whereas intramedullary nailing is the standard fixation device for long bone fractures. OBJECTIVES. To establish a clinically relevant model of atrophic non-union using intramedullary nail and (1) ex vivo and in vivo validation and characterization of this model, (2) establishing a standardized method for leg positioning for a reliable x-ray imaging. Ex vivo evaluation: 40 rat's cadavers (adult male 5–6 months old), were divided into five groups (n=8 in each): the first group was fixed with 20G intramedullary nail, the second group with 18G nail, the third group with 4-hole plate, the fourth group with 6-hole plate, and the fifth group with an external fixator. Tibiae were harvested by leg disarticulation from the knee and ankle joints. Each group was then subdivided into two subgroups for mechanical testing: one for axial loading (n=4) and one for 4-point bending (n=4) using Zwick/Roell® machine. Statistical analysis was carried out by ANOVA with a fisher post-hoc comparison between groups. A p-value less than 0.05 was considered statistically significant. To maintain the non-critical gap, a spacer was inserted in the gap, the design was refined to minimize the effect on the healing surface area. In vivo evaluation was done to validate and characterize the model. Here, a 1 mm gap was created with periosteal and endosteal stripping to induce non-union. The fracture was then fixed by a hypodermic needle. A proper x-ray technique must show fibula in both views. Therefore, a leg holder was used to hold the knee and ankle joints in 90º flexion and the foot was placed in a perpendicular direction with the x-ray film. Lateral view was taken with the foot parallel to the x-ray film. Ex vivo: axial load stiffness data revealed that intramedullary nails are significantly stronger and stiffer than other devices. Bending load to failure showed that 18G nails are significantly stronger than 20G, thus it is used for the in vivo experiments. In vivo: final iteration revealed 3/3 non-union, and in controls with the periosteum and endosteum intact but with the 1mm non-critical gap, it progressed to 3/3 union. X-ray positioning: A-P view in supine position, there was an unavoidable degree of external rotation in the lower limb, thus the lower part of the fibula appeared behind the tibia. To overcome this, a P-A view of the leg was performed with the body in prone rather, this arrangement allowed both upper and lower parts of the fibula to appear clearly in both views. We report a novel model of atrophic non-union, the surgical procedure is relatively simple and the model is reproducible

Introduction: Atrophic nonunion is a well recognised complication of long bone fractures. Clinical trials show that BMP-2 accelerates healing and reduces nonunion in open tibial fractures. We are interested in a natural small molecule that has been previously demonstrated to stimulate angiogenesis in vivo. Our aim is to assess the two treatments in the prevention of nonunion. The small animal model we used is a non-critical size defect of the tibia deprived it of its blood supply by surgical stripping of the periosteum and curetting of the local endosteum thus closely reflecting the clinical situation. The outcomes were measured by radiographic assessment and histology. Methods: Wistar rats were treated with either the angiogenic molecule (0.1% or 0.003%), BMP-2 or vehicle alone (PBS) soaked in a type I collagen sponge. All animals underwent a 2mm osteotomy, stripping of the periosteum and endosteum proximally and distally for the length of the diameter of the tibia. Fluorescent markers were injected at 2 weekly intervals. The rats were sacrificed at 8 weeks. Both tibiae were disarticulated; fixator and soft tissues were removed and AP and lateral X-rays were taken. Subjective assessment of the healing on X-ray was carried out in two ways; using a radiographic scoring system and by grey scale analysis. The samples were embedded, sectioned and stained for new bone formation. Results: Bridging or potential to bridge was seen in a number of animals on x-ray. Bridging or potential to bridge was judged to be present in 72.22% of the BMP-2 group and 66.67% of the high dose group compared to 22.22% of the control group. Histological analysis is being performed to confirm these findings. Discussion: Atrophic nonunion is a serious clinical complication, unfortunately BMP-2 is a highly costly treatment option and therefore alternative molecular therapies are much sought after. We describe here an angiogenic molecule has some potential in preventing formation of nonunion

Introduction. what size of defect is optimal for creating an atrophic nonunion animal model has not been well defined. Our aim in this study was to establish a clinically relevant model of atrophic nonunion in rat femur by creation of a bone defect to research fracture healing and nonunion. Materials and methods. We used 30 male Fischer 344 rats (aged 10–11 weeks), which were equally divided into six groups. The segmental bone defects to a single femur in each rat were performed by double transverse osteotomy, and different sized defects were created by group for each group (1 mm, 2 mm, 3 mm, 4 mm, 5 mm and 6 mm). The defects were measured and maintained strictly by using an original external fixator. The periosteum for each defect was stripped both proximally and distally. Thereafter, these models were evaluated by radiology and histology. Radiographs were taken at baseline and at intervals of two weeks over a period of 8 weeks. Atrophic nonunion was defined as a lack of continuity and atrophy of both defect ends radiologically and histologically at eight weeks. Results. In the 1 mm defect group, all defects had healed. In the 2 mm group, one-fifth remained atrophic nonunion. In the 3 mm group, three-fifths had atrophic nonunion, and all of the defects of groups of 4 mm and over were atrophic nonunion. Conclusion. This study showed that we were able to predictively produce an atrophic nonunion animal model by creating defects of at least 4 mm in the Fischer 344 rat femur

Atrophic non-unions are usually attributed to impaired blood supply but the events that lead to atrophic non-union remain poorly understood. Recent studies. 1,. 2. have shown that vascularity is not reduced in established non-unions but these studies have not examined vascularity at an early stage. The aims of this study were to: 1) develop and validate a clinically relevant small animal model of atrophic non-union and 2) test the hypothesis that the vessel density of atrophic non-unions reaches that of normal healing bones but at a later time point. Twenty eight adult female Wistar rats underwent application of a novel circular frame external fixator to the right tibia under general anaesthesia. The fixator construct was standardised, with eight needles that were drilled through the skin into the proximal and distal metaphyses of the tibia. An osteotomy was performed with a 1mm burr under irrigation. The periosteum was removed on 14 of the 28 animals using a scalpel and the intramedullary canal was curetted. Both insults were performed proximally and distally for a distance equivalent to 1 diameter of the tibia. A 1mm gap was introduced at the osteotomy site and the wound was closed. Once the animal had recovered it was allowed unrestricted weight bearing. Anteroposterior X rays were performed every 2 weeks. Animals were killed at 1, 3, 8 and 16 weeks. Callus areas were measured from X rays using an image analysis system. The average callus area was calculated for each rat every 2 weeks as an indicator of callus production. Specimens were fixed, decalcified, embedded in paraffin wax and 6 ìm sections were stained with H& E. Vascularity was assessed immunohistochemically with monoclonal antibody against smooth muscle actin. The total number of blood vessels in the interfragmentary gap was counted. At 8 and 16 weeks post-osteotomy all animals where stripping and curetting had been performed went on to an atrophic non-union. All animals where this was not performed went on to unite successfully. Histological observations support these radiological findings. Significantly less callus formed in the non-unions than in those that united. There were significantly fewer vessels in the non-unions at week 1 compared to the controls but, by 8 weeks the blood vessel density in the established atrophic non-unions had reached the same level as the vessel density during normal healing. An atrophic non-union model that closely resembles the clinical situation has been developed and validated in rats. The results support the hypothesis that the number of vessels in atrophic non-unions reaches the same level as in those that unite but at a later time point. It is concluded that diminished vessel density within the first 3 weeks may prevent fractures from uniting

We describe a new surgical treatment of atrophic nonunion of the clavicle. The nonunion is excised by cuts at 45° to the long axis and repair uses 3.5 mm pelvic reconstruction or dynamic compression plates, with a lag screw to provide interfragmentary compression. The site is grafted with cancellous bone.

We have been successful in all seven patients, with early return to normal function. The consequent narrowing of the shoulder girdle is fully acceptable for appearance and function.

The role of mesenchymal stem cells (MSCs) in enhancing healing process has been examined with allogeneic and xenogeneic cells in transplantation models. However, certain factors might limit the use of allogeneic cells in clinical practice, (e.g. disease transmission, ethical issues and patient acceptance). Adipose tissue represents an abundant source for autologous cells. The aim of this study was to evaluate adipose-derived autologous cells for preventing non-union.

Adults male Wistar rats (n=5) underwent a previously published surgical procedure known to result in non-union if no treatment is given. This consisted of a mid-shaft tibial osteotomy with peri/endosteal stripping stabilised by intramedullary nail fixation with a 1mm gap maintained by a spacer. During the same operation, ipsilateral inguinal subcutaneous fat was harvested and processed for cell isolation. After three weeks in culture, the cell number reached 5×106 and were injected into the fracture site.

At the end of the experiment, all tibias (injected with autologous fat-MSCs) developed union. These were compared with a control group injected with PBS (n=4) and with allogenic (n=5) and xenogeneic (n=6) cell transplantation groups. The amount of callus was noticeably large in the autologous cell group and the distal-callus index was significantly greater than that of the other groups, P-value =<0.05, unpaired t-test, corrected by Benjamini & Hochberg.

We report a novel method for autologous MSCs implantation to stimulate fracture healing. Local injection of autologous fat-MSCs into the fracture site resulted in a solid union in all the tibias with statistically significantly greater amounts of callus.

In atrophic non-union models, a minimally invasive technique is used to deliver stem cells into the fracture site via percutaneous injection. This technique is significantly affected by a backflow leakage and the net number of cells might be reduced. The Z-track method is a technique used in clinical practice for intramuscular injections to prevent backflow leakage.

We evaluated the potential of the Z-track injection technique for preventing cell loss in non-union models by determining the behaviour of observable marker fluids. Firstly, toluene blue stain was used as an injection material to allow visual detection of its distribution. Rat's cadaver legs were used and tibias were kept unbroken to ensure intact skin and overlying soft tissue. Technique includes pulling the skin over the shin of tibia towards the ankle and injection of the dye around the mid-shaft. The needle was then partially pulled back, the skin was returned to its normal position and a complete extraction of the needle was followed. Secondly, a mixture of contrast material and toluene blue was used to allow direct visual and radiological detection of the injected material into the fracture site. Ante-grade nailing of tibia via tibial tuberosity was carried out followed by a 3 point closed fracture. Injection was performed into the fracture gap similarly to the steps above. X-rays were taken to visualise the location and distribution of the injected material.

Observation revealed no blue stain could be detected over the skin, X -rays revealed that the radiopaque dye remained around the tibia with no escape of the material into the superficial layers or onto the skin surface. Therefore, the number of cells delivered and maintained at a target site could be increased by the Z-track method and therefore, the therapeutic benefit of stem cell injections could be optimised with this simple technique.

Appropriate in vivo models can be used to understand atrophic non-union pathophysiology. In these models, X-ray assessment is essential and a reliable good quality images are vital in order to detect any hidden callus formation or deficiency. However, the radiographic results are often variable and highly dependent on rotation and positioning from the detector/film. Therefore, standardised A-P and lateral x-ray views are essential for providing a full radiological picture and for reliably assessing the degree of fracture union.

We established and evaluated a method for standardised imaging of the lower limb and for reliably obtaining two perpendicular views (e.g. true A-P and true lateral views). The normal position of fibula in murine models is posterolateral to the tibia, therefore, a proper technique must show fibula in both views. In order to obtain the correct position, the knee joint and ankle joints were flexed to 90 degrees and the foot was placed in a perpendicular direction with the x-ray film. To achieve this, a leg holder was made and used to hold the foot and the knee while the body was in the supine position. Lateral views were obtained by putting the foot parallel to the x-ray film. Adult Wister rat cadavers were used and serial x-rays were taken.

A-P view in supine position showed the upper part of the fibula clearly, however, there was an unavoidable degree of external rotation in the whole lower limb, and the lower part of the fibula appeared behind the tibia. Therefore, a true A-P view whilst the body was in the supine position was difficult. To overcome this, a P-A view of the leg was performed with the body prone position, this allowed both upper and lower parts of the fibula to appear clearly in both views. This method provides two true perpendicular views (P-A and lateral) and helped to optimise radiological assessment.

There is a growing trend towards using pre-clinical models of atrophic non-union. This study investigated different fixation devices, by comparing the mechanical stability at the fracture site of tibia bone fixed by either intramedullary nail, compression plate or external fixator. 40 tibias from adult male Wistar rats' cadavers were osteotomised at the mid-shaft and a gap of 1 mm was created and maintained at the fracture site to simulate criteria of atrophic non-union model. These were divided into five groups (n=8 in each): the first group was fixed with 20G intramedullary nail, the second group with 18G nail, the third group with 4-hole plate, the fourth group with 6-hole plate, and the fifth group with external fixator. Tibia was harvested by leg disarticulation from the knee and ankle joints, the soft tissues were carefully removed from the leg, and tibias were kept hydrated throughout the experiment. Each group was then subdivided into two subgroups for mechanical testing: one for axial loading (n=4) and one for 4-point bending (n=4). Statistical analysis was carried out by ANOVA with a fisher post-hoc comparison between groups. A p-value less than 0.05 was considered statistically significant. Axial load to failure data and stiffness data revealed that intramedullary nails are significantly stronger and stiffer than other devices, however there was no statistically significant difference axially between the nail thicknesses. In bending, load to failure revealed that 18G nails are significantly stronger than 20G. We concluded that 18G nail is superior to the other fixation devices, therefore it has been used for in-vivo experiments to create a novel model of atrophic non-union with stable fixation.

We treated 15 patients with atrophic nonunion of a diaphyseal fracture of the humerus with an associated bony defect using an autogenous cancellous bone graft and a plate to bridge the defect. There were nine men and six women with a mean age of 48 years. The mean length of the bony defect was 3 cm.

At a mean follow-up of 30 months only one fracture failed to unite. This suggests that, in the presence of a well-vascularised envelope of muscle, the application of an autogenous cancellous bone graft in conjunction with a bridging plate represents a good alternative to more demanding surgical techniques.

This review is aimed at clinicians appraising preclinical trauma studies and researchers investigating compromised bone healing or novel treatments for fractures. It categorises the clinical scenarios of poor healing of fractures and attempts to match them with the appropriate animal models in the literature.

We performed an extensive literature search of animal models of long bone fracture repair/nonunion and grouped the resulting studies according to the clinical scenario they were attempting to reflect; we then scrutinised them for their reliability and accuracy in reproducing that clinical scenario.

Models for normal fracture repair (primary and secondary), delayed union, nonunion (atrophic and hypertrophic), segmental defects and fractures at risk of impaired healing were identified. Their accuracy in reflecting the clinical scenario ranged greatly and the reliability of reproducing the scenario ranged from 100% to 40%.

It is vital to know the limitations and success of each model when considering its application.

Over 1 million fractures occur each year in the UK. Approximately 5-10% of these fractures have problems with healing. The treatments used for these patients often have a poor outcome and are associated with increased morbidity and disability. Application of synthetic peptides such as thrombin degradation peptide (TP508) has been shown to accelerate fracture repair in a closed rat femoral fracture model. Controlled release of TP508 using microspheres has been shown to enhance repair of articular cartilage defects and stimulate bone formation in segmental defects in rabbits. The aim of this study was to determine whether TP508 could bring about healing in an established fracture non-union model.

A validated rat model of fracture non-union was used. The model was created and left for 8 weeks in order to represent a clinically equivalent model of a non union of a fracture. Rats were randomised into two treatment groups receiving 10microg and 1microg doses of TP508 diluted in 50microL of microspheres and delivered directly to the non union site using percutaneous injection 8 weeks after surgery. The control group received no treatment. At 16 weeks post-surgery, osseous bridging was assessed both radiographically and histologically.

Radiographically there was no difference between the control and two treatment groups. However, histomor-phometric analysis demonstrated that bone formation increased by 43.9% in animals that received high dose of TP508 compared to the control animals. The analysis also indicated that administration of the low dose of TP508 increased the amount of bone formation compared to the control by 9.9 %.

Administration of TP508 has been shown to enhance healing of segmental defects in both critically and noncritically sized defects. However, in our model which is an established fracture non-union model, TP508 did not manage to achieve full osseous union. It has been suggested that the action of this peptide is concentration and environment dependent possibly indicating that TP508 might be less effective when administered in a chronic situation such as that associated with the established non-union fracture. However, even in this sub-optimal situation an increased amount of bone formation was observed.

Introduction: Currently used small animal models of a critical size defect do not sufficiently simulate the biologically unreactive situation in an atrophic non-union. Furthermore, models using intramedullary nails are of little, and poorly standardised, biomechanical stability. This is a characteristic known to promote callus formation though, rather leading to a hypertrophic non-union.

The aim of this study was to establish an atrophic non-union model in the rat femur under well defined biomechanical conditions and with minimised interactions between the processes in the healing zone and the implant by using external fixation.

MATERIALS AND METHODS: 80 male Sprague Dawley rats were randomly divided into two groups (non-union vs. control). All animals received an osteotomy (app. 0.5 mm gap) of the left femur, stabilised with a custom made external fixator. In the non-union group the periosteum was cauterised 2mm distal and proximal of the osteotomy, and the bone marrow was removed. X-rays were performed once weekly. Animals were sacrificed at 14 or 56 days post-operation. At both time points the femurs of 16 animals of each group underwent histological/histomorphometrical and immunhis-tochemical analyses (PMMA or paraffin embedding). Additionally at 56 days 8 animals of each group were tested biomechanically. The maximum torsional failure moment and the torsional stiffness were determined in relation to the intact femur. Post-mortem x-rays were evaluated in a descriptive manner.

RESULTS: At 14 days the histology and radiology showed considerable mineralised periosteal callus in the control group, while the non-union group only showed very little periosteal callus, distant to the osteotomy. At 56 days the control group was completely, or at least partially, bridged by mineralised callus. The non-union group did not show a bridging of the osteotomy gap in any of the animals, moreover the bone ends were resorbed and the gap widened. The relative mean torsional stiffness was significantly larger (p< 0.001) in the control group compared to the non-union group (136.2±34.5% vs. 2.3±1.2%). In the non-union group no maximal torsional failure moment could be detected for the osteotomised femurs. In the control group it was 134.2±79.1%, relative to the intact femur.

DISCUSSION: The cauterisation of the periosteum and the removal of the bone marrow, in combination with a high stiffness of the external fixator may create an atrophic non-union under well defined biomechanical conditions and with minimised interactions between the healing zone and the implant. This model will allow better standardised investigations on the subject of atrophic non-unions.

There is a growing trend towards using pre-clinical models of atrophic non-union. This study investigated different fixation devices, by comparing the mechanical stability at the fracture site of tibia bone fixed by either intramedullary nail, compression plate or external fixator. 40 tibias from adult male Wistar rats' cadavers were osteotomised at the mid-shaft and a gap of 1 mm was created and maintained at the fracture site to simulate criteria of atrophic non-union model. These were divided into five groups (n=8 in each): the first group was fixed with 20G intramedullary nail, the second group with 18G nail, the third group with 4-hole plate, the fourth group with 6-hole plate, and the fifth group with external fixator. Tibia was harvested by leg disarticulation from the knee and ankle joints, the soft tissues were carefully removed from the leg, and tibias were kept hydrated throughout the experiment. Each group was then subdivided into two subgroups for mechanical testing: one for axial loading (n=4) and one for 4-point bending (n=4).

Statistical analysis was carried out by ANOVA with a fisher post-hoc comparison between groups. A p-value less than 0.05 was considered statistically significant. Axial load to failure data and stiffness data revealed that intramedullary nails are significantly stronger and stiffer than other devices, however there was no statistically significant difference axially between the nail thicknesses. In bending, load to failure revealed that 18G nails are significantly stronger than 20G. We concluded that 18G nail is superior to the other fixation devices, therefore it has been used for in-vivo experiments to create a novel model of atrophic non-union with stable fixation.

We present our clinical experience in treating atrophic non-union of long bones by injecting, percutaneously, autologous bone marrow aspirate concentrated as a source of progenitors stem cells

Bone marrow aspirated from the iliac crest contains progenitor cells that can be used to obtain bone-healing of non-union. However, its efficacy appears to be related to the number and concentration of progenitors in the graft. The last three-year period, 11 patients (8 men-3 women) with established atrophic non-union were treated in our department. In all cases, the gap between the fragments was smaller than 5 mm. A constant volume of 60+60 ml of marrow were aspirated from both iliac crests and centrifuged for 15 minutes aiming at the increase of concentration of progenitor-mononucleotide cells. An average volume of 20 ml (+/− 2) concentrated bone marrow was injected percutaneously, under C-arm, at the site of non-union. The graft contained an average of 272.64 x 10(6)/ ml mononucleotide cells. The evaluation of treatment was based on the clinical and radiological findings after 3, 6, 9 and 12 months. However, prior to administration of bone marrow stem cells, there was no case with evidence of ongoing deep sepsis.

Bone union was obtained in 10 out of 11 patients (full weight bearing, callus formation in 3 out of 4 cortices). In one case a second operation was needed due to impaired indications of treatment. However, in all cases, there were no signs of local or systematic complications.

Percutaneous concentrated bone marrow grafting is an efficient and safe method, for treating atrophic non-unions, with a minimal invasion technique. Contraindications for the above technique are a gap larger than 5 mm and a preexisting angular and axial deformity.

We retrospectively reviewed eight children with idiopathic chondrolysis (IC) of the hip and nine with atrophic tuberculosis (TB) of the hip treated over the 10 years 1990 to 1999. Both conditions present with a stiff hip and radiographic joint space narrowing. Our aim was to delineate clinical, radiological and histological differences between the two conditions, thereby obviating the need for biopsy in IC, which could worsen the prognosis.

In the IC group all patients were girls. Their mean age was 12 years (11.5 to 13). They presented with a flexion abduction and external rotation deformity of the hip. Chest radiographs were normal in all patients, and all except one had an ESR below 20. The Mantoux was negative in six of the eight. Radiographs showed joint space narrowing and osteopoenia, but the subchondral bony line remained present. Four of the eight had a synovial biopsy, which showed non-specific chronic synovitis. The cartilage looked pale and lustreless. In one hip the cartilage was biopsied and showed cartilage necrosis.

In the TB group, five of the nine patients were boys. The mean age was 7 years (5 to 13.5). The only constant hip deformity was flexion. Chest radiographs were normal in all patients. In all patients the ESR was below 20 and the Mantoux was positive. Hip radiographs showed osteopoenia with loss of the subchondral bony line. Peri-articular lytic lesions were present in all patients except one. Histology of synovial biopsy showed caseous necrosis in all hips, and seven of the nine had a positive culture for TB. Macroscopically the cartilage looked normal, and in one hip the cartilage biopsy was histologically normal.

We confirmed that in IC the joint space narrowing is due to cartilage necrosis. We postulate that in atrophic TB the loss of subchondral bone due to subchondral erosion gives the impression of joint space narrowing. We also concluded that IC was a diagnoses per se and not by exclusion, and that biopsy was not required.

Purpose: Athrophic non unions constitute a major problem in orthopaedic trauma. The main probably cause of atrophic non union is damage of the vascular system and dysfunctional regeneration of the vasculature at the area of the fracture. The most important hormonal pathway controlling angiogenesis is VEGF (Vascular Endothelial Growth Factor). The use of VEGF for enhancing bone healing in atrophic non unions could be a very promising solution for the future. An interesting alternative to the use of VEGF is the use of Erythropoietin (Epo). VEGF has been also reported to interact with Endothelial Progenitor Cells (EPCs). Our scope is to identify a possible new role for Epo as a valid substitute for VEGF through the clarification of the molecular and cellular pathways of fracture healing.

Methods: A survey was conducted via internet (Med-line - Pubmed, Cochrane database, Scopus) and relevant textbooks.

Results: It has been reported that Epo could induce increased chemotaxis, migration of Mesenchymal Stem Cells (MSCs), but also activation of Metaloproteinase - 9 and production of pro-angiogenic factors. These effects on MSCs could explain the observation that Epo could be very useful in the treatment of wound healing and burn healing in animal studies. It has been that Epo could express receptors at the chondrocytes, but also induce better bio-mechanical strength, callus formation, histomophometric image and increased bone density at the treated with Epo animals when compared with control animals. It is also worthy to note that the Epo has been found to stimulate neo-vascularisation in vivo, differentiation of endothelial cell lines towards a vascular pathway and improvement of cardiac function through EPCs and VEGF, implying Epo also in the differentiation and chemotaxis of the circulating EPCs. We should not forget that the transformation of EPCs in mesenchymal cells (i. g. myoblasts) has already been demonstrated.

Conclusions: The consequences of these observations could be very interesting: EPCs have been reported to enhance neo-vascularisation and angiogenesis at the region of the fracture. All these imply a novel role for EPCs in combination (or even replacing the rare) MSCs under the influence of VEGF and Epo for the enhancement of fracture vascularisation and healing enhancement. Further studies should clarify this new field in basic orthopaedic, trauma and bone metabolism science.

During fracture repair, a number of growth factors and cytokines are present at elevated levels at the fracture site such as Transforming Growth Factor Beta (TGF-), Fibroblast Growth Factor (FGF) and Platelet Derived Growth Factor (PDGF). The aim of the study was to investigate the presence of these growth factors in healing fractures and fracture non-unions, in order to test the hypothesis that atrophic non-unions express a lower level of growth factors than hypertrophic non-unions and healing fractures.

Biopsies were taken from the fracture site of 23 patients (mean age 46) with uninfected non-unions, 12 patients with hypertrophic (mean 13.8 months after fracture) and 11 patients with atrophic (mean 16.5 months after fracture). A comparison group of biopsies from early fracture callus (one to four weeks after fracture) in five patients with healing fractures was also included. Five-micron paraffin sections were immunohistochemically stained for TGF-, FGF-II and PDGF. Growth factors were then assessed in six different cell types.

Fibroblasts, endothelial cells and macrophages were found to express TGF-, FGF-II and PDGF in all three-fracture groups. Osteoblasts, osteoclasts and chondrocytes were not present in the healing fracture group. The growth factor expression in osteoblasts, osteoclasts and chondrocytes in the non-union groups were found to be variable, however, the expression of these growth factors appeared to be less in the atrophic non-unions than hypertrophic non-unions.

The expression of these growth factors was found to be less in the atrophic non-union group than the hypertrophic non-union group in osteoblasts, osteoclasts and chondrocytes. These results may have relevance for new therapies that can be aimed at delivering growth factors to treat fracture non-unions. By further investigation of the differential expression of these growth factors it may be possible to determine which factors are likely to stimulate fracture healing.