Tapping the radial side of the wrist normally elicits a reflex contraction producing elbow flexion, wrist extension and wrist radial deviation. An abnormal response, consisting of finger flexion when performing this manoeuvre is known as the inverted radial (supinator) reflex (IRR). The significance of this reflex in asymptomatic subjects is unknown.

To document the frequency of the IRR in an asymptomatic population and to identify any presymptomatic pathology in those subjects.

The study group consisted of patients and staff at the senior author's institution. Patients were taken from clinics where the complaints were of lower limb symptoms. Subjects were excluded if they had any history of neck pain or stiffness or if they had any subjectively abnormal sensation. The radial reflex was elicited with a tendon hammer. Those subjects with an IRR were asked to attend for a MRI scan of the cervical spine to investigate for any abnormality.

47 subjects were studied. There were 8 subjects who displayed an IRR. In 4 subjects the IRR was unilateral and in 4 bilateral. Seven subjects consented to further investigation by MRI. The average age of these patients was 36 years.

The MRI scans revealed normal appearances in 6 cases. There was no cord signal abnormality in any case.

The IRR occurred with a frequency of 17% in the study group. There was no significant cervical pathology identified in these subjects.

In young asymptomatic patients, the presence of an inverted radial reflex is of no diagnostic relevance.

AO Spine Reference Centre & Institute of Health & Biomedical Innovation, Queensland University of Technology, Brisbane, Australia

Traumatic spinal cord injury (SCI) is a devastating condition with no curative therapy. Pro-inflammatory therapy has been suggested recently to try and reduce the inhibitory glial scar and promote neural regeneration and healing. The aim of this study is to investigate the potential of sustained delivery of angiogenic/pro-inflammatory growth factors to reduce the secondary degeneration after spinal cord injury.

Adult male Wistar Kyoto rats (200-300g; 12-16weeks old) were subjected to cord hemisections via a T10 laminectomy. Animals were randomised to treatment or control groups after the spinal cord injury had been induced. Treatment consisted of implantation of a mini-osmotic pump capable of delivering 5 micrograms vascular endothelial growth factor (VEGF) and 5 micrograms platelet-derived growth factor (PDGF), via a catheter, to the site of the lesion, over 7 days(n=6). Control animals were subjected to either cord lesion only (n=6) or lesion plus mini-pump delivering PBS (phosphate-buffered saline) solution (n=6). Rats were sacrificed at one month and the spinal cords were harvested and examined by immunohistology, using anti-neurofilament-200 and anti-Glial Acidic Fibrillary Acidic Protein (GFAP) antibodies.

RESULTS: Active treatment spinal cords showed a higher level with aboration of the axonal filament through the defect and more dense neurofilament-200 staining at the lesion site compared to both control groups. The treatment also showed the elevated presence of activated microglia in the lesion, whilst distal to the lesion the microglia and astrocytes retained an unreactive phenotype.

Pro-inflammatory therapy in the rat spinal cord-injury model showed favourable histological findings after sustained delivery of PDGF and VEGF

Introduction: It is well known that the fate of biomaterials is determined by the distribution of proteins attached to the surface from the initial contact with blood or serum. This profile determines wether a material is inert, creates a foreign body response or is bioactive. Bioinert materials, such as polyethylene completely denature surface proteins, whilst materials inducing inflammatory responses are predisposed to complement protein attachment. Bioactive materials such autologous tissue grafts adsorb, but do not denature serum proteins such as fibronectin and Von Willebrand’s factor. This does not interfere with the healing cascade. This aim of this study is to prepare a synthetic bone graft substitute that activates the body’s autologous healing cascade by activating platelets, without activating a complement response through the controlled adsorption of serum proteins.

Methods: Polymers composed of varied concentration of acrylic acid (AA) and comonomers (methyl, ethyl and butyl methacrylates (MMA, EMA, BMA)) were prepared in glass vials by free radical polymerisation. Fresh blood was collected from a healthy donor and pipetted immediately into each chamber. Glass was used as a control. The chambers were incubated at 37o C for 2 hours. The surface morphology was examined using Scanning Electron Microscopy (SEM). Concentration of complement protein C5a and prothrombin fragments 1 and 2 were determined using commercial ELISA kits. Foreign body reaction (FBR) initiated by the biomaterial was estimated by counting leukocytes on clot sections using immunofluorescence.

Results: Extent of coagulation was correlated with plasma concentrations of Prothrombin fragments 1 and 2. These measurements show blood incubated with various polymers composed of different comonomers all promoted the formation of blood clots. It was found that the leukocyte population towards the interface of clot and polymer (AA:MMA) decreased with increasing surface acid concentration (65%AA:MMA 30 leukocytes/0.25mm2, glass 70 leukocytes/0.25mm2 (p< 0.05)). FBR is induced by the activation of complement system. The percentage of C5a concentration detected in blood incubated with various polymers composed of different comonomers relative to normal serum level of C5a (35ng/mL). No significant elevations of C5a were measured from polymer 65% AA:MMA and 65% AA:EMA. Glass induced vigorous complement response as expected. The synergistic combination of surface acid concentration and comonomers had a significant effect on extent of FBR. Increased acid concentration resulted in decreased C5a level with MMA and ET but increased level with BMA.

Discussion: The functional groups exposed on the surface of a material influence whether leukocyte or platelet activation is responsible for the subsequent physiological response. By modifying the combinations of surface acid concentrations and comonomers, we show that a biomaterial with an appropriate surface chemistry promotes the platelet plug formation and coagulation but down regulated foreign body reaction. This study shows that that a biomaterial with the appropriate surface chemistry to evoke the same coagulation response as damaged tissue, mediated through platelet activation and intrinsic and extrinsic coagulation, initiates the initial pathways of the bone healing cascade. This material is a realistic candidate for biomaterial induced bone regeneration.

Introduction: The biological activity of autologous grafts is due to a number of proteins (growth factors) that control bone cell differentiation, proliferation and expression. Several of these have been isolated including; bone morphogenetic proteins 2 and 7. These are commercially available and regularly used with the intention of accelerating fracture healing, repairing critical sized defects and combating bone mineral loss. Whilst it is commonly recognised that multiple growth factors are present at differing times in the healing cascade, the usual delivery, both in the clinic and the laboratory, is of one growth factor delivered over a very short and early time period. Commonly growth factors are delivered in solution or from a collagen sponge and are quickly metabolised in the proteolytic wound healing environment. The physiological need for BMPs is later than the acute delivery at the time of surgery. The aim of this study is to develop a granular protein delivery system that enables controlled release of multiple proteins at a variety of time points.

Methods: A series of homogenous polymer granules 8mm3 were prepared by photo-polymerising 12uL of mixtures of methacrylated adipic acid anhydride (MAAA) and methyl methacrylate (MMA) or MAAA and butyl methacrylate (BMA) with molar ratios ranging from 100- 55 % (MSAA). Into each granule 5ug of a model drug, carmoisine was loaded and 1%w/w of 2,2-dimethoxy-2-phenyl-acetophenone (DMPA) photoinitiator was added per granule. The granules were exposed to UV light at 390nm for 14 minutes. Multilayered granules were prepared photo-polymerising 4uL layers of different monomer compositions in a similar method to the single layered method above. The composition of the multilayered granules was chosen to optimise the release profile. Carmoisine release profiles were determined by UV-visible spectroscopy.

Results: Homogenous granules composed of 100% MAAA released 90% of their payload by 24hrs, those composed of 90:10 MAAA:MMA released by 48hrs those composed of 70:30 MAAA:MMA released by 80hrs those composed of 60:40 MAAA:MMA released by 170hrs those composed of 70:30 MAAA: BMA released by 288hrs and those composed of 60:40 MAAA:BMA released by 456hrs. The multilayered granule had a sustained release of the model drug over the test period of 19 days.

Discussion: The limitation of most drug delivery systems, such as microspheres or collagen, is poor control over the release profile. The drug is ether released instantly or well after it is required. This multilayered composite drug delivery system enables the controlled release of different bioactive compounds at different time points between 0 and 19 days. By altering the drug loading in each layer we were able to sustain the release of one compound over this time period. This technology enables us to switch compounds at a given time points for example delivery of angiogenic factors for one week, proliferative factors for the second week and differentiation factors for the third week. This technology enables the pre-programmed release of multiple growth factors at times in the healing cascade when they meet the physiological need. A controlled release of growth factors at the appropriate time should improve bone healing rates.

Introduction: Acute neurological damage from spinal cord injuries is believed to be localised, however it initiates a cascade of secondary events which usually leads to extensive and permanent neurological deficit. The secondary damage begins with the disruption of the blood-spinal cord barrier which unleashes a protracted inflammatory response. This prolonged inflammatory response is the catalyst for the secondary neurodegeneration and limited repair response that occurs in the chronic phase of a spinal cord injury. In this study it was proposed that the acute delivery of the angiogenic growth factors vascular endothelial growth factor (VEGF) and platelet derived growth factor (PDGF) would mediate inflammation and restore the blood spinal cord barrier. This would minimise the formation of glial scar and reduce the extent of secondary degeneration caudal and cranial to the lesion site.

Methods: Adult male Wistar rats (400g) were anesthetised. Complete laminectomies were performed at T10 and the animals were subjected to T10 hemisection. Animals were randomised to a treatment group (Lesion Control (LC), Gel Control (GC) and Angiogenic Gel (AG)) after the spinal cord was cut. Each treatment group had 6 animals sacrificed 3 months post injury. Sections were stained with antibodies to neurofilament 200, glial fibrillary acidic protein, smooth muscle actin (SMA), and fluorescent secondary antibodies and mounted with DAPI. The lesion size was measured from horizontal histological sections of the midline from 5 animals in each group using Axiovision version 4.6.1.0 (Carl Zeiss Imaging Solutions, Germany).

Results: The mean lesion size for the lesion control group was 2.09mm2, 1.97mm2 for the gel control group and 0.45mm2 for the active gel group. A t-test was used to confirm that the differences between the active gel and the two control groups were statistically significant (AG vs LC p= 0.021 AG vs GC p= 0.026). Histology showed a marked improvement of the morphology of the astrocytes in the treatment group over the control groups indicating that the treatment affected the population of reactive astrocytes. SMA staining showed an increased level of revascularisation in the treated lesions.

Discussion: Spinal cords do not heal because of prolonged inflammation which leads to secondary necrotic events, scar formation and the inhibition of regeneration. In this study we present a method for regulating the post lesion inflammatory signals, significantly reducing post-lesion scar formation. We propose the delivery of VEGF/PDGF significantly increases the permeability of the blood spinal cord barrier to neutrophils and macrophages and promotes angiogenesis observed in the lesion site. This may have two major effects on the progression of the spinal cord injury. Firstly, by increasing the initial influx of inflammatory cells it enables the faster removal of damaged tissue and phagocytosis of apoptotic cells thereby restoring the balance in favour of regulated inflammation and results in a finite and reduced inflammation time. Secondly, combination of VEGF and PDGF provides a robust angiogenic response and reduces ischemia, the population of reactive astrocytes and the capacity to form glial scars. These growth factors appear to moderate the secondary degenerative changes that result from the prolonged inflammation and thus promote the inherent capacity for regeneration.

Introduction: It has long been recognised that static plain x-rays are a sub-optimal method for the assessment of lumbar fusion. Blumenthal and Gil showed that radiographic assessment of fusion corresponded with operative findings only 69% of the time. Santos et al suggest that both plain x-rays and flexion/extension x-rays overestimate the fusion rate when compared to helical computed tomography (CT). To date there has been no correlation of CT assessment of fusion with surgical exploration. In this study we present an animal model of lumbar spine pseudarthrosis and compare three imaging modalities with micro-cut CT scanning and cadaveric assessment.

Methods: Approval was gained from the QUT animal ethics committee. Eleven mixed bred ewes were assigned to either a fusion group or an intentional pseudarthrosis (IP) group. A dorsal approach to the facet joints of L2/3 was made. The facet joints were destabilised by resecting the articulating surfaces with a rongeur. In the fusion group, the spinous processes of the destabilised segment were wired tightly together and a bone graft harvested from the iliac crest was placed into the joint space. In the IP group the bone graft bed was prepared similarly except that a small proportion of the articulating surface was left intact and a 1.0 cm2 roll of oxidised cellulose was placed into the facet joint space bilaterally. In the IP group the spinous processes were wired around an interspinous spacer which was later removed to create a similar degree of laxity in the fixation of each of the IP specimens. The animals were sacrificed at 6 months and static and dynamic lateral radiographs obtained. The spine was removed en bloc, and high speed fine cut (2mm) CT Scanning performed. The specimens were individually assessed for fusion by micro-cut CT scanning. Eight independent, blinded orthopaedic surgeons, were asked whether they considered the spine to be fused based on

plain x-ray

These results were correlated with a fusion rate based on the micro CT. The specificity and sensitivity of these radiological measures in diagnosing pseudarthrosis and inter-rater reliability using Fleiss’ Kappa scores for each method were calculated.

Results: For assessing pseudarthrosis identified by microCT the plain film sensitivity was 0.41 and the specificity was 0.47. For assessing pseudarthrosis with plain and flexion extension xrays the sensitivity was 0.55 and the specificity was 0.33. For assessing pseudarthrosis with plain flexion extension xrays and CT the sensitivity was 0.81 and the specificity was 0.88. The Kappa score for plain films was 0.15, for flexion extension was 0.07 and CT was 0.54.

Discussion: This study suggests that plain radiographs and flexion extension radiographs are an unreliable measure of posterior lumbar fusion. The current clinical gold standard for assessment of fusion (CT) was able to correctly identify non-union in 80% of cases. Whilst no alternatives to structural assessment of the fusion mass with CT currently exist it is important to recognise the limitations of this technique.

Introduction Anterior column reconstruction and fusion remains the gold standard of treatment for a number of spinal pathologies. One of the challenges of interbody fusions cages is the footprint of the cage reducing the surface area of endplate available for fusion. Biodegradable polymer implants will over time present a greater area for fusion and may help to reduce problems such as stress shielding, particulate debris and retained foreign body response. Resorbable cages have been have been prepared from a number of different materials, including inorganic composites (eg hydroxyapatite / tricalcium phosphate) and polymers (Poly L-lactide-co-D,L-lactide (PDLLA)). However all of the current options for interbody fusion have reported deficiencies or complications. The synthesis, mechanical properties, and degradation behaviour of two novel biopolymers are presented and the applicability for use as materials in interbody fusion devices is discussed.

Methods Methacrylated adipic anhydride (MAA) and methacrylated sebacic anhydride (MSA) pre-polymers were synthesized by melt condensation. Conversion of the acid to the anhydride was confirmed using 1H nuclear magnetic resonance (NMR) (Bruker, Alexandria, NSW) and FT- Infrared spectroscopy (Nicolet, Waltham MA). These pre-polymers were subsequently co-polymerized with methyl methacrylate (MMA) and 0.25 wt% benzoyl peroxide at 65oC for 16hrs and post-cured at 120oC under vacuum for 2 hrs to form biodegradable networks. The co-polymerization behaviour was monitored by FT-Raman spectroscopy. The compressive mechanical properties of the polymer were determined using an Instron 5567 (Bayswater Vic.). The polymer networks were degraded in phosphate buffered saline (PBS) with various amounts of MAA and MSA.

Results The formation of the pre-polymer was confirmed with the observation of NMR peaks at 5.8 and 6.2 ppm and FT-IR peaks at 1637cm-1. Copolymerization was followed with consecutive FT-IR acquisitions with 100% conversion achieved between 10 and 30 hrs depending on the ratio of MMA to MSA or MAA. Increasing the fraction of methacrylated anhydride slowed the reaction rate.

The compressive strength of the MAA and MSA based copolymers was measured as a function of anhydride concentration. Compressive strength for MMA increased (90±9 to 140±10 Mpa) in an approximately linear manner for MAA concentrations from 10 to 40 wt.% but decreased markedly for MAA concentration of 45% (62±14 Mpa). The compressive strength of MSA decreased exponentially for concentrations ranging from 10 to 45 wt.% (140±18 to 39±1 Mpa).

Discussion The use of poly-L-lactic acid in lumbar interbody cages has been shown to be mechanically feasible with the mechanical strength of the cage material reported to be 93 Mpa (1). The material described here has controlled mechanical properties in the required range as well as a degradation behaviour that lends itself better to spinal applications than current materials

Introduction The precise contribution of the posterior longitudinal ligament (PLL) and disc annulus in the burst fracture setting and their potential relative roles during intra operative reduction manoeuvres remains unclear. The anatomical attachments of the posterosuperior fragment most often associated with canal occlusion and potential neurological compromise are not well described in a reproducible model.

Methods Burst fractures were induced using a pendulum impact tester. The jig allowed for accurate positioning in all planes and for precise delivery of both the magnitude and vector of the impact force. This allowed for creation of fracture all three major groups of the AO classification. The A3 (burst fracture) was produced in 10 cadaveric sheep spines by delivering a neutral force vector on a physiologically flexed spine. The morphology of the fracture was confirmed by CT. Subsequent laminectomy was performed and the anatomical attachments of the large fragments were identified.

Results The PLL was identified following laminectomy in each case. In six of the ten spines there had been significant disruption of the longitudinal structure of the PLL .In a further two cases there had been stripping of the PLL from the posterior aspect of the vertebral body in association with the retropulsed canal fragment. Subsequent excision of the PLL from the posterior aspects of vertebral body and discs did not compromise the attachment of the retropulsed fragment to the disc annulus in any case.

Discussion This study confirms the anatomical relationship between disc fragment and disc annulus in the burst fracture setting. The strong attachment between fragment and disc facilitate rotation of the fragment about this hinge and into the canal. Subsequent intraoperative reduction of this fragment by restoration of disc height may require contribution both from this annular attachment and from tension set up in an intact PLL. The relative contributions of each of these structures in the reduction manoeuvre remains unclear.

Introduction Fluid dynamics in the intervertebral disc plays an important role in the overall mechanical function as well as nutritional supply for both the annulus fibrosis (AF) and the nucleus pulposus (NP). The apparent diffusion coefficient of water in intervertebral discs has been suggested to be related to the matrix composition as well as the structural integrity of the disc.

Methods Lesions were created in two intervertebral lumbar discs in 18 in-bred Sprague Dawley rats (approved by Animal Ethics Committee) using a 14G hypodermic syringe to induce degenerative change as described by Sobajima et al (1). Regenerative treatment involved the injection of multipotent stem cells isolated from the olfactory mucosa. These cells were injected either at the time of creation of the lesion or six weeks after creation of the lesion.

MRI experiments were undertaken at 25.0 ± 0.1 °C on a Bruker Avance NMR spectrometer (Bruker Bio-Spin, Rheinstetten, Germany) using a 7.0T vertical bore magnet system, equipped with a 1.1 T m-1 (110 G cm-1) gradient set and 15 mm ‘birdcage’ RF resonator. Specimens for testing were immersed in physiological saline inside a 15 mm NMR. Both multi echo and diffusion weighted images were acquired with a recycle time TR = 2 s and 8 averages using a 0.7 mm slice thickness, a field of view (FOV) of ca. 15 × 15 mm and a 128 × 128 matrix. For multi echo experiments the echo time was 5 ms with 64 echoes and for DT imaging a diffusion gradient duration δ = 2 ms and diffusion delay Δ = 12 ms. The diffusion tensor was calculated from the seven requisite diffusion-weighted images using in-house Matlab® code (The Mathworks, Natick, MA) written for the purpose.

Results Preliminary results of the multi echo images indicate that intervertebral disc degradation results in observable differences in the T2 of the NP ((mean/SD (ms)) control disc (63.2/3.4), untreated (49.5/1.4) and treated (49.7/3.4)). Diffusion tensor results show isotropic diffusion in the NP with anisotropic diffusion in the AF and observable differences in magnitude.

Discussion The use of high resolution MRI has been shown to provide a useful tool for quantifying the effects of regenerative treatments for degenerative disc disease.

Introduction Cement leakage into adjacent structures is the main complication during vertebroplasty. The majority of these leaks are asymptomatic, but pulmonary cement embolism has been reported to cause cardiovascular disturbances and even death (1,2). Furthermore, the use of calcium phosphate (CaP) cements for vertebroplasty may aggravate cardiovascular deterioration in the event of cement embolism by stimulating coagulation [3].

The cardiovascular effects of pulmonary cement embolism were investigated using an animal model.

Methods In 18 skeletally mature sheep, 2.0ml cement was injected into the pulmonary trunk during general anaesthesia (approved by Animal Ethics Committee). Three different cements were used: 1) PMMA (Simplex P, Stryker); 2) PMMA with 10% hydroxyapatite (PMMA & HA) (Vertecem, Synthes); 3) Experimental injectable CaP cement (Synthes). The following cardiovascular parameters were recorded continuously (endpoint: 60min post-injection): arterial, central venous, pulmonary arterial pressures and cardiac output. Blood gases and coagulation parameters (antithrombin, D-dimer, prothrombin fragments I & II) were measured pre-injection, 10, 30 and 60min post-injection. Postmortem, lungs were removed intact and submitted to computer tomography (CT) imaging.

Results There were no fatalities. After 1min, mean pulmonary arterial pressure had increased by 9%, 14% and 21% from pre-injection value in the PMMA, PMMA & HA and CaP group respectively. Differences in pulmonary arterial pressure between the three material groups were not statistically significant. Pulmonary arterial pressure stayed elevated for the duration of the experiment (i.e. 60min post-injection). There were no other significant changes in cardiovascular, blood gas or coagulation parameters from pre- to post-injection values. Three dimensional reconstructions of the CT images showed a tendency of the CaP cement to break up into multiple smaller pieces whereas the two other cements did not.

Discussion Cement embolism led to mild pulmonary hypertension in all material groups. Present results are in contrast to earlier reports (pig model) of fulminant cardiovascular deterioration after CaP cement embolism (3). Present changes were not as severe and there was no evidence of thromboembolism. This discrepancy may have been due to differences in the cement formulations or the animal model.

Pulmonary hypertension was more severe in the CaP cement group. This may have been due to the disintegration of the CaP cement resulting in blockage of more pulmonary vessels compared to the PMMA cements.

Introduction Far lateral disc prolapse (also known as foraminal or extreme lateral prolapse) make up 10% of all disc herniations. In addition, far lateral disc prolapses tend to affect more proximal levels more frequently than do prolapses in the posterolateral location and they are often associated with greater radicular symptoms than typical posterolateral herniations, most likely due to involvement of the dorsal root ganglion. Surgery for far lateral disc protrusions has been associated with a less favourable outcome, perhaps due to delays in diagnosis, inadequate preoperative imaging, and postoperative instability as a result of excessive bony and facet resection during the surgical approach

Methods Twelve patients with far lateral disc herniations operated on by the senior author (RPW) fulfilled the criteria of having both pre- and postoperative Oswestry Disability Index (ODI) scores recorded at each clinic visit. Results of these cases and those of a cohort of age and sex matched patients undergoing standard posterolateral discectomy undertaken by the same surgeon were analyzed. The presence of radiculopathy pre- and postoperatively, workers compensation status, return to work, length of stay and complications, as well as any prior intervention in the form of nerve root sleeve blocks or surgery were recorded

Results Both groups were well matched in terms of age and sex. Follow up ranged from 4 to 18 months. Herniations at more proximal levels (L2/3 and L3/4) were seen more frequently in the far lateral group than in the posterolateral group. Six patients in the far lateral group had preoperative nerve root sleeve blocks compared with one in the posterolateral group. Two patients in each group had had previous (different level) surgery. Patients in each group had similar preoperative ODI scores. Both groups demonstrated a reduction in the preoperative ODI compared with the preoperative score. The mean improvement was 24 (range −26 to +62) for the far lateral group and 22 (range −6 to +46). There was no significant difference between the groups

Discussion The results of this study are encouraging with respect to surgical treatment of far lateral discs. Recent literature has questioned the efficacy of surgical intervention for this pathology. These results show that with carefully selected patients results are comparable with standard posterolateral discectomy

Introduction Reported clinical results suggest that vertebroplasty is a safe and effective technique for providing pain relief. However, information about the long-term effect of PMMA on the adjacent intervertebral discs and the augmented bone is lacking. Adjacent intervertebral discs may be at higher risk of degeneration due to nutritional constraints. Bone loss in augmented vertebrae may occur due to mechanical stress-shielding or toxicological effects.

The aim of the present study was therefore to investigate the effect of PMMA augmentation on intervertebral disc and bone tissue after 6 and 12 months, using an animal model.

Methods In 12 skeletally mature sheep, 2.0ml PMMA (Simplex P) was injected into three lumbar vertebrae (approved by Animal Ethics Committee). Two injection holes were drilled into the middle of three vertebrae at a distance of 5.0mm from the cranial and caudal endplate and 1.0ml PMMA was injected into each hole. Four weeks before euthanasia, animals received an injection of tetracycline for bone labeling.

Postmortem, T1- and T2-weighted sagittal and axial MR images were taken prior to fixation in 80% ethanol. Spines were cut into specimens containing one intervertebral disc and half of the two adjacent vertebrae. The discs which were two levels above the first augmented vertebra served as controls. Microsections were stained with H& E, Goldner, Alcian blue-PAS and Safranin O. MRI signal intensity and morphology of discs were evaluated qualitatively. Histomorphological analysis of discs and endplates was conducted using published criteria [1]. Presence of bone remodeling, fibrous tissue and foreign body reaction in the vertebrae was also recorded.

Results There was no distinguishable loss of MRI signal intensity in the discs in between augmented vertebrae. Cement injection resulted in blocking 50–75% of the endplate lengths. Most discs that were in between augmented vertebrae showed signs of degeneration (chondrocyte proliferation, necrosis) after 6 (80%) and 12 months (88%). Inflammatory reaction to PMMA was observed in some specimens (approximately 25%). Cement had been covered with fibrous tissue in all augmented vertebrae, but tetracycline labeling revealed new bone formation in the vicinity of PMMA.

Discussion Augmentation of three adjacent vertebrae initiated degenerative changes of intervertebral discs in between two augmented vertebrae. This is in contrast to previous animal studies [2] where no degenerative changes after cementing endplates were observed. Current investigations were performed with the specific aim to block the endplates. Clinically, endplates may not get blocked as effectively. On the other hand, discs in older patients are nutritionally constrained due to end-plate calcification and even partial blockage may lead to degenerative changes as documented presently.

The risk of degenerative changes of intervertebral discs should be considered in patients undergoing vertebroplasty.

Introduction Polymethylmethacrylate (PMMA) has been widely used in orthopaedic procedures for fixation of joint replacements or enhancing the fixation of implants. However, the use of PMMA has been associated with cardiovascular deterioration and even death. More recently, PMMA has also been used for augmenting osteoporotic vertebral bodies which have fractured or are at risk of fracture. The main complication is PMMA leakage into adjacent structures. Transient hypotension and fatal fat embolism (FE) have also been reported.

The pathomechanism of cardiovascular deterioration after the injection of PMMA (i.e. FE) remains a highly controversial subject. The exact role of PMMA in the development of FE remains unclear. The aim of the present study was to elucidate the acute effects of injecting PMMA compared with bone wax into vertebral bodies on the cardiovascular system using an established animal model for vertebroplasty (VP) (Aebli, N, et al. Spine. 2002).

Methods In 8 skeletally mature mixed-bred ewes (2–4 years) 6.0ml PMMA (CMW3-Depuy) or bone wax (Bone Wax, Ethicon) were injected unilaterally, through an open approach into the L1 & L2 pedicles. Blood pressure, heart rate, and cardiac output were measured.

Results The major difference between the cardiovascular response of the PMMA and that of the bone wax group was the recovery in Pulmonary Artery Pressure (PAP) and Pulmonary Vascular Resistance (PVR). Three minutes post-injection, PAP had fully recovered to baseline values in the wax group. However in the PMMA group, PAP had only recovered by 52% after 3 min and fully recovered after 10 min.

Discussion The augmentation of vertebral bodies resulted in transient cardiovascular changes regardless of the material used. However, the recovery of PAP and PVR values took significantly longer with the PMMA group. The peak response was a result of pulmonary vasoconstriction triggered by a reflex reaction to the embolisation of bone marrow particles or by vasoactive cytokines. The peak response was therefore mainly associated with the increase in intraosseous pressure during the augmentation causing release of bone marrow contents into the and not the cement monomer. The cement monomer however plays a role in the cardiovascular complications during FE. The delayed recovery of PAP and PVR in the PMMA group may be due to a vasoconstriction effect of the cement monomer on the pulmonary vascular system.

Potentially serious cardiovascular complications may occur during VP regardless of the material used. The injection of PMMA may cause prolonged pulmonary hypertension during vertebro- and also arthroplasty. Continuous invasive cardiovascular monitoring may be required in patients with impaired cardiovascular and pulmonary function

Introduction Locking plates are the most used devices for achieving anterior cervical spinal fusion and offer considerable advantages such as faster and easier implantation and fewer implant-related failures than older plate systems. Recently polyaxial locking screws were introduced to make the implantation of these plates even easier by facilitating the implantation of the screw in all directions. However polyaxial screws may have the disadvantage of losing the angular stability with subsequent failure of the plate. The aim of this study was the radiological follow up of the patients with polyaxial screw and to compare them with the conventional looking plates.

Methods Patients underwent anterior cervical discectomy and fusion in which either ventral cervical locking plates or a polyaxial locking screw were used for indications including cervical spondylotic radiculopathy, disc herniation, trauma, and myelopathy. Patients underwent anterior cervical discectomy and interbody fusion and / or corpectomy. Preoperative and postoperative radiographic data included sagittal angle, translation, and settling of the graft.

Results One hundred and forty patients were investigated (mean age of 49 years) with an average follow up period of 21.5 months (range from 4 to 50 months). All underwent anterior cervical plate fusion as a component of the surgical treatment for symptomatic degenerative cervical spinal disease (55%) or for vertebral destruction caused by trauma (45%).100 (71%) of patients were treated with a conventional locking plate and 29 % with a plate with polyaxial screw fixation. Besides plate fixation, 4 of the 140 patients had a combined ventrodorsal fusion. In 46 cases (25%), one or more vertebral bodies were removed and replaced with either iliac bone graft (two levels, 21% of all cases) or fibula strutgraft (4%). In the group with conventional locking plates no patient had to be revised, 3 showed a subluxation (up to 1/3 of the vertebral body diameter) and 2 screw back out posteriorly without clinical relevance. In the group with the plate with polyaxial screw fixation two patient had to be revised (posterior stabilisation) because of subluxation due to loss of angular stability of the screws and one patient developed subluxation of 1/3 of the vertebral body also due to loss of angular stability.

Discussion The complication rate and the revision rate for anterior cervical discectomy and fusion with plates with and without polyaxial screw fixation were similarly low. Polyaxial screw may have the theoretical disadvantage of loosening with loss of the angular stability and subsequent failure necessitating revision.

Introduction Experimental heterotopic bone formation in the canine urinary bladder has been observed for more than seventy years without revealing the origin of the osteoinductive signals. In 1931, Huggins demonstrated bone formation in a fascial transplant to the urinary bladder. Through an elaborate set of experiments, it was found that proliferating canine transitional epithelial cells from the urinary system act as a source of osteoinduction.

Urist performed a similar series of experiments in guinea pigs as Huggins did in his canine model. After two weeks, mesenchymal cells condensed against the columnar epithelium and membranous bone with haversian systems and marrow began to form juxtapose the basement membrane. At no time was cartilage formation noted, only direct membranous bone formation. They also demonstrated the expression of BMP’s in migrating epithelium and suggested that BMP is the osteoinductive factor in heterotopic bone formation.

Method This study was approved by Institutional Animal Ethics Committee. Six dogs underwent a mid-line laparotomy incision followed by mobilisation of a right sided myoperitioneal vascularised flap based on an inferior epigastric artery pedicle. A sagittal cystotomy is made in the dome of the bladder and the vascularised flap was sutured in place with acryl absorbable, continuous suture. The animals were sacrificed at 6 weeks. The bladder samples were removed and assessed by histology and immunohistochemistry. Sections were incubated with optimal dilution of primary antibody for type I collagen, type III collagen, alkaline phosphatase (ALP), bone morphogenetic protein (BMP)-2 and –4, osteocalcin (OCN), osteopontin (OPN), bone sialoprotein (BSP).

Results The mechanism for bone formation induced by the epithelial-mesenchymal cell interactions is not clear. We were able to demonstrate mature lamellar bone formation 6 weeks after transplanting a portion of the abdominal smooth muscle into the bladder wall. The bone formed immediately adjacent to the proliferating transitional uroepithelium, a prerequisite for bone formation in Huggins’ model. Here we report evidence of cartilage formation and therefore endochondral ossification as well as membranous bone formation. This is very similar histologically to the process of endochondral ossification at the growth plate in the growing skeleton. We propose a mechanism for the expression of BMP by epithelial cells.

Discussion This study demonstrates transitional epithelium induced formation of chondrocytes and osteoblasts in muscle tissue. The sequential expression of bone matrix proteins was related to cell proliferation, differentiation and formation of endochondral and membranous bone. Further information regarding the molecular mechanism of bone formation induced by epithelial-mesenchymal cell interactions will improve understanding of cell differentiation during osteogenesis.

Introduction The influence of timing of surgery on functional outcomes following spinal cord injury remains controversial. Animal studies suggest that the rate, degree, and duration of cord compression are the principal determinants of spinal cord injury (SCI) severity and prognosis for recovery. Delamarter et al, (J Bone Joint Surg Am 1995) have shown that when experimental cord compression in dogs is relieved within 1 hour, full motor recovery can be achieved. It is suggested by some clinically based research that definitive surgical treatment for unstable injuries results in fewer sequelae than prolonged immobilization and allows more rapid entry into rehabilitation. It is however the timing of this surgery which remains controversial. It has been suggested that early surgical management promotes neurological recovery by limiting secondary damage caused by inflammation, oedema, ischemia and instability. To date few studies have found a link between neurological recovery and timing of surgery (Fehlings, et al; Spine 2001).

Methods Data was gathered retrospectively by chart review of patients referred to the Princess Alexandra hospital with spinal cord injury. Patients were age matched into high and low velocity groups. This data was studied to assess the effects of energy of injury and timing of surgical intervention on neurological outcome. Patients either had anterior, posterior, or combined surgery, external immobilization or traction depending on the preference of the treating surgeon.

Results A cohort of 43 patients all of whom had spinal cord injury was retrospectively studied. Of these, 21 had a high energy injury (eg. MVA) and 21 had a low energy injury (eg. rugby). 28 had anterior stabilization 7 had traction, 4 had external immobilization 2 had a combined anterior / posterior fixation and 1 had posterior stabilization. The data suggest that the prognosis for recovery following a spinal cord injury is unrelated to the energy involved. The low energy group improved on average 0.6 ASIA grades (SEM 0.16) while the high energy improved 0.7 ASIA grades (SEM 0.17). The timing of definitive intervention for patients with incomplete cord lesions was shown to significantly (p=0.029) effect ultimate functional outcomes. Those with early (within 8hrs) intervention improved an average of 1.4 ASIA grades (SEM 0.21) and those with late intervention improved 0.6 ASIA grades (SEM 0.19). This effect was present in both high and low energy injury groups.

Discussion The timing of definitive intervention for spinal cord injury is still controversial. However there is Class II evidence that early surgery can be done safely in a patient with spinal cord injury (Fehlings, et al; Spine 2001). The findings from this retrospective study suggest that early surgical intervention may improve neurological recovery.

Introduction The purpose of this prospective controlled study was to define indications and analyse the clinical and radiographic results of cages in the surgical treatment of traumatic cervical spine instability.

Methods 53 patients were treated by monosegmental anterior discectomy and interbody fusion using either autologous tricortical iliac crest bone graft and cervical spine locking plate (CSLP) (bone graft group, n= 26) or Syncage-C (Synthes) filled with autologous cancellous bone grafts and CSLP (cage group n=27). Indications for surgery were traumatic cervical spine instability were classified by the cervical fracture classification of Blauth et al¹ as B1, B2, B3, C2 or C3 fractures. Intraoperative parameters (operative time, blood loss radiation time and intra- and perioperative complications) were documented. Prior to surgery and at follow-up (6 and 12 months) evaluation included measurement of neck pain, shoulder/arm pain and Neck Pain Disability Index (NPDI). Neurological function was assessed using the ASIA scale. Radiographic evaluation included plain X-rays, flexion-extension views and CT-scans. Patient satisfaction was measured on a five-point Likert scale.

Results There was no statistically significant difference between the two groups in the demographic data. One patient in the bone graft group was not available for the 1-year follow-up evaluation; however, all patients were available for the 2-year follow-up. Operation time was significantly shorter (p< 0.05) in the cage group (67 +/− 6 min) than in the bone graft group (78 +/− 9 min). After 6 and 12 months there was no difference between both groups in pain or NPDI, neurological and overall outcome. The neurological improvement of the two groups was not statistically different. Although the cage group showed a trend for better maintenance of lordosis after 12 months, there was no statistically significant difference between groups in all radiographic parameters. There were no implant-related complications during the follow-up. General complications included one patient with eczema due to the stiff collar (cage group) and one patient with pneumonia (cage group). Complications associated with the harvesting of iliac crest bone grafts included 14 patients (9 patients in the bone graft group, 5 patients in the cage group) with prolonged pain (> 3 months) at the donor site, one superficial wound healing problem (bone graft group) which healed under conservative treatment and one hematoma (bone graft group) which required additional surgery.

Discussion Cages offer a valid alternative to a tricortical iliac crest bone graft in the surgical treatment of mono-segmental traumatic cervical spine instability. Although there was no significant difference between the cage and the bone graft group in the functional and radiographic outcome, less donor site morbidity and a shorter operation time make cages cost effective in this selected group of patients. Although the cages are expensive, less donor site morbidity, shorter operation time and reduced hospital stay might result in cost-effectiveness of this implant.

Introduction: Pelvic fixation is undertaken in order to restore stability to an unstable pelvis or correct severe scoliotic degeneration of the spine. Instability of the pelvic ring can result from resection of tumours, fractures of the pelvis or infection of the pelvic joints and bones. A number of methods for stabilising the pelvis have been described in the literature including the Galveston Reconstruction (GR)¹ and the triangular frame reconstruction (TFR)². These are associated with an improvement in functional ability, however failure of instrumentation or loosening often occurs.³ A recent mechanical analysis of these techniques has found the technique used in this hospital (GR) performed most poorly.²

Methods: A scoring system was developed from a retrospective analysis of 8 patients. The patients were categorised into two groups (high score and low score) based on age, presence of infection and serious non-associated comorbidities. A patient aged 60 years or over scored 5 points. Patients with bony infection scored 10 points. The presence of serious comorbidity including osteoporosis scored 5 points with minor comorbidities scoring 1 point.

Results: Eight patients who underwent pelvic fixation for varied indications (2 after resection of tumours, 1 fracture, 2 scoliotic degeneration, 3 for infection) were analysed. Three patients had a good functional improvement without loosening of screws beyond 1 year after surgery. These patients were otherwise healthy, relatively young and had no disease processes that affected local bone quality at the site of fixation or serious comorbidities. The other 5 patients all showed evidence of early screw loosening within one year. Of these patients, 2 had a number of serious comorbidities well recognised to compromise bone quality (osteoporosis, long term steroid use) and 3 had pre-existing extensive bony infection.

Discussion: Bone quality of the pelvic bones appears to be the primary predictor of long term functional outcome after pelvic fixation. The 5 patients who had a number of comorbidities well recognised to compromise bone quality all saw early screw loosening within 1 year. Since fixation of the pelvis requires extensive surgery necessitating both posterior and anterior approach and has a number of severe complications such as alteration of urinary, sexual and recto-sigmoid functions the benefit of pelvic fixation should be considered in light of these factors which appear to predict long term outcomes. Further prospective studies of patients undergoing pelvic fixation are required to validate our scoring system.

Introduction: Diaphyseal femoral allograft is well suited to anterior column reconstruction of the thoracolumbar spine due to its inherent structural properties and bio-compatibility. The Bridwell system of interbody fusion assessment¹ is based on plain x-rays and therefore lacks sensitivity. A new classification system of bony union is proposed using high-speed spiral CT imaging.

Methods: Twenty-six patients who underwent anterior thoracolumbar reconstruction for burst fracture using femoral allograft were followed for a minimum of 2 years. Each subject underwent high speed spiral CT scanning through the reconstructed region of the thoracolumbar spine and a classification system of graft to endplate union and central cancellous autograft incorporation was established.

The classification system reflects gradually increasing biological stability of the construct. Grade I (complete fusion) implies cortical union of the allograft and central trabecular continuity. Grade II (partial fusion) implies cortical union of the structural allograft with partial trabecular incorporation. Grade III (unipolar pseudarthrosis) denotes superior or inferior cortical non-union of the central allograft with partial trabecular discontinuity centrally and Grade IV (bipolar pseudarthrosis) suggests both superior and inferior cortical non-union with a complete lack of central trabecular continuity. Intra- and inter-observer error studies were carried out involving spinal surgeons, radiologists and trainees to examine reliability of the classification

Results: In this series 84% of cases demonstrated Grade I or Grade II characteristics. 1 case (4%) was identified as Grade IV. The classification showed good reliability with a kappa score of over 0.7

Discussion: Plain radiographs have always proved unsatisfactory for the accurate assessment of incorporation of grafts in the thoracolumbar spine. The use of CT imaging in the assessment of graft union has allowed a more accurate assessment of union. The classifi cation system presented allows a reproducible and relevant categorisation of allograft incorporation.

Introduction: Vertebroplasty using polymethylmethacrylate (PMMA) is an established technique in the treatment of osteoporotic fractures of the vertebra. Complications of vertebroplasty associated with PMMA leakage can include damage to the spinal cord. Previous studies have sought to investigate thermal changes in the paravertebral region, but used smaller volumes of cement than are used clinically¹, or used in vitro experimental techniques.² We have designed an in vivo sheep model to investigate the thermal changes after injection of clinically relevant volumes of PMMA, and to measure change in cord function associated with PMMA extrusion.

Methods: Five sheep were anaesthetised and 1.0ml of PMMA was injected into the spinal canal at the L1 level, with measurement of the temperature by thermocouple. The L2 to L5 vertebral bodies were then exposed and 9 thermocouples placed at points in and around the vertebra (superior and inferior endplate, disc above and below, central body, posterior wall, and spinal canal) to measure paravertebral temperature for a 10- minute period after injecting 6.0mls of PMMA. All animals were then humanely euthanased, and the T12 to L2 vertebrae harvested to examine the effect of temperature on the vertebral body and spinal cord using light microscopy.

Results: The experiments showed significant increases in the paravertebral temperature, especially at the end-plates (mean temperature 51.7°C, mean increase in temperature +16°C). This is contrary to studies using small cement volumes or in vitro conditions. Intradiscal and posterior wall temperature did not significantly rise. Spinal canal temperature reached a mean 75.4°C in the presence of “extruded” cement. Microscopic examination showed thermal damage to the spinal cord.

Discussion: The experiments indicate that neurological complications associated with vertebroplasty are likely to be thermally mediated, and that the analgesic effects of vertebroplasty are likely to be, at least in part, due to thermal damage to endplate neurological structures.