Abstract

Introduction

Osteoporotic vertebral fractures can cause severe vertebral wedging and kyphotic deformity. This study tested the hypothesis that kyphoplasty restores vertebral height, shape and mechanical function to a greater extent than vertebroplasty following severe wedge fractures.

Introduction

Vertebroplasty helps to restore mechanical function to a fractured vertebra. We investigated how the distribution of injected cement benefits both fractured and neighbouring vertebrae.

Background

Fracture of an osteoporotic vertebral body reduces vertebral stiffness and decompresses the nucleus in the adjacent intervertebral disc. This leads to high compressive stresses acting on the annulus and neural arch. Altered load-sharing at the fractured level may influence loading of neighbouring vertebrae, increasing the risk of a fracture ‘cascade’. Vertebroplasty has been shown to normalise load-bearing by fractured vertebrae but it may increase the risk of adjacent level fracture. The aim of this study was to determine the effects of fracture and subsequent vertebroplasty on the loading of neighbouring (non-augmented) vertebrae.

Introduction

Vertebral osteoporotic fracture increases both elastic and time-dependent ('creep') deformations of the fractured vertebral body during subsequent loading. The accelerated rate of creep deformation is especially marked in central and anterior regions of the vertebral body where bone mineral density is lowest. In life, subsequent loading of damaged vertebrae may cause anterior wedging of the vertebral body which could contribute to the development of kyphotic deformity. The aim of this study was to determine whether gradual creep deformations of damaged vertebrae can be reduced by vertebroplasty.

To investigate whether restoration of mechanical function and spinal load-sharing following vertebroplasty depends upon cement distribution.

Fifteen pairs of cadaver motion segments (51-91 yr) were loaded to induce fracture. One from each pair underwent vertebroplasty with PMMA, the other with a resin (Cortoss). Various mechanical parameters were measured before and after vertebroplasty. Micro-CT was used to determine volumetric cement fill, and plane radiographs (sagittal, frontal, and axial) to determine areal fill, for the whole vertebral body and for several specific regions. Correlations between volumetric fill and areal fill for the whole vertebral body, and between regional volumetric fill and changes in mechanical parameters following vertebroplasty, were assessed using linear regression.

For Cortoss, areal and volumetric fills were significantly correlated (R=0.58-0.84) but cement distribution had no significant effect on any mechanical parameters following vertebroplasty. For PMMA, areal fills showed no correlation with volumetric fill, suggesting a non-uniform distribution of cement that influenced mechanical outcome. Increased filling of the vertebral body adjacent to the disc was associated with increased intradiscal pressure (R=0.56, p<0.05) in flexed posture, and reduced neural arch load bearing (F_N) in extended posture (R=0.76, p<0.01). Increased filling of the anterior vertebral body was associated with increased bending stiffness (R=0.55, p<0.05).

Cortoss tends to spread evenly within the vertebral body, and its distribution has little influence on the mechanical outcome of vertebroplasty. PMMA spreads less evenly, and its mechanical benefits are increased when cement is concentrated in the anterior vertebral body and adjacent to the intervertebral disc.

We aimed (1) to determine the factors which influence outcome after surgery for CES and (2) to study CES MRI measurements. 56 patients with evidence of a sphincteric disturbance who underwent urgent surgery (1994-2002) were identified and invited to clinic. 31 MRIs were available for analysis and randomised with 19 MRIs of patients undergoing discectomy for persistent radiculopathy. Observers estimated the percentage of spinal canal compromise and indicated whether they thought the scan findings could produce CES and whether the discs looked degenerate. Measurements were repeated after two weeks.

(1) 42 patients attended (mean follow up 60 months; range 25–114). Mean age at onset was 41 years (range 24–67). 26 patients were operated on within 48 hours of onset. Acute onset of sphincteric symptoms and the time to operation did not influence the outcomes. Leg weakness at onset persisted in a significant number at follow-up (p<0.005). Bowel disturbance at presentation was associated with sexual problems (<0.005) at follow-up. Urinary disturbance at presentation did not affect the outcomes. The 13 patients who failed their post-operative trial without catheter had worse outcomes. The SF36 scores at follow-up were reduced compared to age-matched norms in the population. The mean ODI was 29, LBOS 42 and VAS 4.5.

(2) No significant correlations were found between MRI canal compromise and clinical outcome. There was moderate to substantial agreement for intra- and inter-observer reproducibility.

Introduction

Osteoporotic fracture reduces vertebral stiffness, and alters spinal load-sharing. Vertebroplasty partially reverses these changes at the fractured level, but is suspected to increase deformations and stress at adjacent levels. We examined this possibility.

Introduction

Vertebral osteoporotic fracture increases both elastic and time-dependent (‘creep’) deformations of the fractured vertebral body during subsequent loading. This is especially marked in central and anterior regions of the vertebral body, and could explain the development of kyphotic deformity in life. We hypothesise that vertebroplasty can reduce these creep deformations.

Purpose: To determine how cement volume during vertebroplasty influences:

stress distributions on fractured and adjacent vertebral bodies,

Methods: Nineteen thoracolumbar motion segments from 13 cadavers (42–91 yrs) were loaded to induce fracture. Fractured vertebrae received two sequential injections (VP1 and VP2) of 3.5cm3 of polymethylmethacrylate cement. Before and after each injection, motion segment stiffness was measured in compression and in bending, and the distribution of compressive “stress” in the intervertebral disc was measured in flexed and extended postures. Stress profiles yielded the intradiscal pressure (IDP), stress peaks in the posterior (SPP) annulus, and the % of the applied compressive force resisted by the neural arch (FN). Cement leakage and vertebral body volume were quantified by water-immersion, and % cement fill was estimated.

Results: Bending and compressive stiffness fell by 37% and 50% respectively following fracture, and were restored only after VP2. Depending on posture, IDP fell by 59%–85% after fracture whereas SPP increased by 107%–362%. VP1 restored IDP and SPP to prefracture values, and VP2 produced no further changes. Fracture increased FN from 11% to 39% in flexion, and from 33% to 59% in extension. FN was restored towards pre-fracture values only after VP2. Cement leakage, IDP and compressive stiffness all increased with %fill.

Conclusions: 3.5cm3 of cement largely restored normal stress distributions to fractured and adjacent vertebral bodies, but 7cm3 were required to restore load-sharing between the vertebral bodies and neural arch. Risks of cement leakage increased with %fill.

Conflicts of Interest: None

Source of Funding: This work was funded by Action Medical Research and The Hospital Saving Association Charitable Trust. Vertebroplasty materials were provided by Stryker.

Introduction: Vertebroplasty increases stiffness and partly restores normal load-sharing in the human spine following vertebral fracture. The present study investigated whether the mechanical effects of vertebroplasty are influenced by the distribution of injected cement.

Methods: Ten pairs of cadaver motion segments (58–88 yr) were loaded to induce fracture, after which one from each pair underwent vertebroplasty with polymethyl-methacrylate cement, the other with a resin (Cortoss). Various mechanical parameters were measured before fracture, after fracture and following subsequent vertebroplasty. Micro-computed tomography scans and plane radiographs (sagittal, frontal, and axial) obtained from each augmented vertebral body were analysed to determine percentage cement fill in the whole vertebral body and in selected regions. The relationship between volumetric fill obtained by micro-CT and areal fill obtained by radiography was investigated using linear regression analysis. Regression analysis also indicated whether changes in mechanical parameters following vertebroplasty were dependent upon cement distribution.

Results: Cement type had no significant influence upon regional fill patterns, so data from both cements were pooled for all subsequent analyses. Volumetric fill of the whole vertebral body was predicted best by areal fill in the sagittal plane (R2=0.366, P=0.0047). Restoration of intradiscal pressure and compressive stiffness following vertebroplasty were dependent upon volumetric cement fill both in the whole vertebral body (R2=0.304, P=0.0118 and R2=0.197, P=0.0499 respectively), and in the anterior half (R2=0.293, P=0.0137 and R2=0.358, P=0.0053).

Conclusion: Cement fill patterns can best be assessed radiographically from sagittal plane views. Placement of cement in the anterior vertebral body may help to improve mechanical outcome following vertebroplasty.

Introduction: The aim of this cadaver study was to examine how cement volume used in vertebroplasty influences the restoration of normal load-sharing and stiffness to fractured vertebrae.

Methods: Nineteen thoracolumbar motion segments obtained from 13 spines (42–91 yrs) were compressed to failure in moderate flexion to induce vertebral fracture. Fractured vertebrae underwent two sequential vertebroplasty treatments (VP1 and VP2) each of which involved unipedicular injection of 3.5ml of polymethyl-methacrylate cement. During each injection, the volume of any cement leakage was recorded. At each stage of the experiment (pre-fracture, post-fracture, post-VP1 and post-VP2) measurements were made of motion segment stiffness, in bending and compression, and the distribution of compressive stress across the disc. The latter was measured in flexed and extended postures by pulling a pressure transducer through the mid-sagittal diameter of the disc whilst under 1.5kN load. Stress profiles indicated the intradiscal pressure (IDP), stress peaks in the posterior annulus (SPP), and neural arch compressive load-bearing (FN). Measurements obtained after VP1 and VP2 were compared with pre-fracture and post-fracture values using repeated measures ANOVA to examine the effect of cement volume (3.5 ml vs. 7 ml) on the restoration of mechanical function.

Results: Fracture reduced compressive and bending stiffness by 50% and 37% respectively (p< 0.001) and IDP by 59%–85%, depending on posture (p< 0.001). SPP increased from 0.53 to 2.46 MPa in flexion, and from 1.37 to 2.83 MPa in extension (p< 0.01). FN increased from 11% to 39% of the applied load in flexion, and from 33% to 59% in extension (p< 0.001). VP1 partially reversed the changes in IDP and SPP towards pre-fracture values but no further restoration of these parameters was found after VP2. Bending and compressive stiffness and FN showed no significant change after VP1, but were restored towards pre-fracture values by VP2. Cement leakage occurred in 3 specimens during VP1, and in 7 specimens during VP2. Leakage volumes ranged from 0.5–3.0 ml, and were larger during VP2 than VP1.

Conclusions: Unipedicular injection of 3.5 ml of cement reversed fractured induced changes in IDP and SPP, but did not affect stiffness and neural arch load-bearing. Larger injection volumes may provide some extra mechanical benefit in terms of restoring stiffness and reducing neural arch loading, but these extra mechanical benefits can be at the cost of increased risk of cement leakage.

Introduction: Kyphoplasty is a modification of the basic vertebroplasty technique used to treat osteoporotic vertebral fracture. This study evaluated whether kyphoplasty conferred any short-term mechanical advantage when compared with vertebroplasty.

Methods: Pairs of thoracolumbar “motion segments” were harvested from nine spines (42–84 yrs). Specimens were compressed to failure in moderate flexion to induce vertebral fracture. One of each pair underwent vertebroplasty, the other kyphoplasty. Specimens were then creep loaded at 1.0kN for 2 hours to allow consolidation. At each stage of the experiment, motion segment stiffness in bending and compression was determined, and the distribution of compressive “stress” was measured in flexed and extended postures by pulling a pressure- sensitive needle through the mid-sagittal diameter of the disc whilst under 1.5kN load. Stress profiles indicated the intradiscal pressure (IDP), stress peaks in the posterior annulus (SPP), and neural arch compressive load-bearing (FN).

Results: Vertebral fracture reduced bending and compressive stiffness by 37% and 55% respectively (p< 0.0001), and IDP by 55%–83%, depending upon posture (p< 0.001). SPP increased from 0.188 to 1.864 MPa in flexion, and from 1.139 to 3.079 MPa in extension (p< 0.05). FN increased from 13% to 37% of the applied load in flexion, and from 29% to 54% in extension (p< 0.001). Vertebroplasty and kyphoplasty partially reversed these changes, and their immediate mechanical effects were mostly sustained after creep-loading. No differences were found between vertebroplasty and kyphoplasty.

Conclusion: Kyphoplasty and vertebroplasty are equally effective in reversing fracture-induced changes in motion segment mechanics. In the short-term, there is no mechanical advantage associated with kyphoplasty.

Purpose of the study: To determine if cement type, bone mineral density (BMD), disc degeneration and fracture severity influence the restoration of spinal load-sharing following vertebroplasty.

Methods: Fifteen pairs of thoracolumbar motion-segments (51–91 yrs) were loaded to induce fracture. Vertebroplasty was performed so that one of each pair was injected with Cortoss, the other with Spineplex. Specimens were then creep loaded at 1.0kN for 2 hours. At each stage of the experiment, stress” profiles were obtained by pulling a pressure-sensitive needle through the disc whilst under 1.5kN load. From these profiles, the intradiscal pressure (IDP), posterior stress peaks (SP_P), and neural arch compressive load (F_N) were determined. BMD was measured using dual photon X-ray absorptiometry. Severity of fracture was quantified from height loss.

Results: Fracture reduced IDP (p< 0.001) but increased SP_P and F_N (p< 0.001). Following vertebroplasty, these effects were significantly reversed, and in most cases persisted after creep-loading. However, no differences were observed between PMMA- and Cortoss-injected specimens. After fracture, decreases in IDP, and increases in SP_P and F_N, were greater in specimens with lower BMD or greater height loss (p< 0.05). After vertebroplasty, specimens with lower BMD showed greater increases in IDP, and those with more degenerated discs showed greater reductions in SP_P (p< 0.05).

Conclusions: Changes in spinal load-sharing following fracture were partially restored by vertebroplasty, and this effect was independent of cement type. The effects of fracture and vertebroplasty were influenced by BMD, disc degeneration, and fracture severity. People with more severe fractures, low BMD and degenerated discs may gain most mechanical benefit from vertebroplasty.

Introduction: We have shown that vertebroplasty increases stiffness and partly restores normal load-sharing in the human spine following vertebral fracture. The present study investigated how this restorative action is influenced by type of cement injected, bone mineral density (BMD), and fracture severity.

Methods: Fifteen pairs of thoracolumbar motion-segments (51–91 yrs) were loaded on a hydraulic materials testing machine to induce vertebral fracture. One from each pair underwent vertebroplasty with polymethyl-methacrylate (PMMA) cement, the other with a biologically- active resin (Cortoss). Specimens were then creep loaded at 1.0kN for 2 hours. At each stage of the experiment, bending and compressive stiffness were measured, and ‘stress’ profiles were obtained by pulling a pressure-sensitive needle through the disc whilst under 1.5kN load. Profiles indicated the intradiscal pressure (IDP) and neural arch compressive load (FN). BMD was measured using dual photon X-ray absorptiometry. Severity of fracture was quantified from height loss. Changes were compared using repeated measures ANOVA.

Results: Fracture reduced bending and compressive stiffness by 31% and 41% respectively (p< 0.0001), and IDP by 43%–62%, depending upon posture (p< 0.001). In contrast, FN increased from 14% to 37% of the applied load in flexion, and from 39% to 61% in extension (p< 0.001). Following vertebroplasty, these effects were significantly reversed, and in most cases persisted after creep-loading. No differences were observed between PMMA- and Cortoss-injected specimens. The decrease in IDP and increase in FN after fracture were correlated with BMD in flexion and with height loss in extension (p< 0.01). After vertebroplasty, restoration of IDP and FN in flexion were correlated with their loss after fracture (p< 0.01). The former was also related to BMD (p< 0.05).

Conclusions: Changes in spinal load-sharing following fracture were partially restored by vertebroplasty, and this effect was independent of cement type. The effects of fracture and vertebroplasty on spinal load-sharing were influenced by severity of fracture, and by BMD.

These findings suggest that people with more severe fractures and low BMD may gain most mechanical benefit from vertebroplasty.

Introduction: (1) Determine whether initial MRI findings correlate with clinical outcome.(2) Study the reproducibility of MRI measurements of large disc prolapses.(3) Estimate the ability to predict CES based on MRI alone.(4) Does CES only occur in degenerate discs?

Method: 31 patients with CES were identified and invited to attend clinic. 19 patients who underwent discectomy were identified. Digital photographs of all 50 MRIs were obtained. Observers: 1 Radiologist, 2 Spinal Surgeons and 1 Trainee did not know the number of patients in each group. Observers estimated the percentage spinal canal compromise on each view (0–100%), indicated whether they thought the scan findings could produce CES and commented on disc degeneration. Measurements were repeated after 2 weeks.

Results: 26 patients attended clinic – mean follow up 51 months (range 25–97). 12 of the 26 patients with CES had, on average, > 75% canal compromise. No significant correlations were found between MRI canal compromise and clinical outcome. Kappa values for intra-observer reproducibility of measurements ranged from 0.4–0.85 and inter-observer 0.63–5. Based on MRI, the correct identification of CES has sensitivity 68%, specificity 78%, positive predictive value 84% and negative predictive value 58%. Over 80% of the CES causing discs were degenerate.

Discussion: Canal compromise does not appear to predict clinical outcome. MRI measurement reproducibility has substantial agreement. CES is a clinical diagnosis supported by an MRI scan. In less clear cases the presence of a large disc on an MRI scan supports a diagnosis of CES (PPV 84%). CES occurs in degenerate discs.

Introduction: Current evidence suggests that CES should be operated within 48 hours from onset of sphincteric symptoms in order to maximise chances of recovery. Measurement reproducibility of large disc prolapses and clinical correlations have not previously been studied.

Objectives: (1) Determine whether initial MRI findings correlate with clinical outcome (2) Study the reproducibility of MRI measurements of large disc prolapses (3) Estimate the ability to predict CES based on MRI alone.

Study Design: 31 patients with CES were identified, the case notes reviewed and the patients invited to attend clinic. Outcome consisted of history and examination, and several validated questionnaire assessments. 19 patients who underwent discectomy for persistent radiculopathy were identified. None had sphincteric symptoms. All had a significant surgical target. Digital photographs of all 50 MRIs were obtained showing the T2 mid-sagittal image and the axial image with the greatest disc protrusion. The Observers: 1 Consultant Radiologist, 2 Consultant Spinal Surgeons and 1 SHO did not know the number of patients in each group. Observers estimated the percentage spinal canal compromise on each view and indicated whether they thought the scan findings could produce CES. Measurements were repeated after 2 weeks.

Results: 26 patients attended clinic mean follow up 51 months (25 to 97). As expected, the % canal compromise differed significantly between the two groups (p0.001). 12 of the 26 patients with CES had, on average, over75% canal compromise. No significant correlations were found between MRI canal compromise and clinical outcome. Canal compromise did predict whether the patient would fail their Trial Without Catheter (p0.05). Based on MRI alone, the correct identification of CES has sensitivity 68%, specificity 78%, positive predictive value 84% and negative predictive value 58%. Kappa values for intra-observer reproducibility ranged from 0.4 to 0.85 for sagittal compromise, axial compromise and correct prediction of CES. All three interobserver kappa values for these measurements were 0.64.

Conclusions: This is the largest radiological case series of CES with 4 years clinical follow up. Canal compromise on MRI does not appear to directly predict clinical outcome. Reproducibility of MRI measurements of large disc protrusions has substantial agreement. MRI could be of help in equivocal cases if the scan shows a large disc.

Objective: Determine factors influencing outcome after surgery for cauda equina syndrome with particular attention sphincteric recovery. Subjects:56 patients with evidence of a sphincteric disturbance who underwent urgent surgery between 1994 and 2002 were identified and invited for follow up.

Outcome Measures: History and examination, Oswestry Disability Index, Short Form 36, Visual Analogue Score, Low Back Outcome Score, Modified Somatic Perception Score, Modified Zung Depression Score, International Prostate Severity Score, Male Sexual Health Questionnaire and Sheffield Female Pelvic Floor Questionnaire.

Results: 42 patients attended with a mean follow up of 60 months (25 to 114 months). Mean age at onset was 41 years (24 to 67 years) with 23 males and 19 females. 25 patients had sudden onset of symptoms in less than 24 hours. 26 patients were operated on within 48 hours of onset. At presentation urinary retention was associated with acute onset of less than 24 hours (p0.01), leg weakness (p0.01), abnormal leg sensation (p0.05) and abnormal rectal tone (p0.05). Bilateral radiculopathy was associated with leg weakness (p0.005). All patients with abnormal rectal tone (21) had abnormal rectal sensation. At follow up significantly more females had urinary incontinence (p0.001) and bowel disturbance (p0.05), higher VAS scores (p0.05) and lower SF36 Pain and Energy scores (p0.05) than males. Urinary disturbance at presentation did not affect the outcomes. Bowel disturbance at presentation was associated with sexual problems (0.005) and abnormal rectal tone (p0.05) at follow up. Objective reduced perianal sensation at onset persisted in a significant number at follow up (21/32 patients; p0.05) as did leg weakness (14/23; p0.005). There was a weak association between delay to operation and bowel disturbance (p0.05) at follow up. Eight patients had faecal soiling and faecal incontinence at follow up and this was associated with sudden onset of symptoms, initial abnormal rectal tone and time to operation (p0.05). The SF36 scores at follow up were reduced compared to age matched norms in the population. The mean ODI was 29, LBOS 42 and VAS 4.5.

Conclusions: In our series the duration of symptoms and speed of onset prior to surgery appears to influence bowel but not bladder outcome two years after surgery. Based on the SF36, LBOS and ODI scores, patients who have had CES do not return to a normal status. Patient counselling about this would therefore be appropriate.

Introduction: Current evidence suggests CES should be operated < 48 hours from onset. MRI scanning is often not available 24 hours a day.

Objectives: (1) Determine whether MRI findings correlate with clinical outcome. (2) Study the reproducibility of MRI measurements of large disc prolapses. (3) Estimate the ability to predict CES based on MRI alone.

Study Design: 31 CES patients were identified,contactedand invited to follow up. Clinical outcome consisted of history and examination, and validated questionnaire assessments. 19 patients who underwent discectomy were identified. T2 mid-sagittal and axial digital photographs of all 50 MRIs were obtained. Observers did not know the number of patients in each group (1 Consultant Radiologist, 2 Consultant Spinal Surgeons and 1 SHO). They estimated the percentage spinal canal compromise on each view (0–100%) and indicated whether they thought the scan findings could produce CES. Measurements were repeated after 2 weeks.

Results: 26 patients attended clinic (mean follow up 51 months). There were no significant correlations found between MRI canal compromise and clinical outcome. Kappa values for the measurements ranged 0.52–0.85 and 0.61–0.75 for intra- and inter-observer reproducibility. Based on MRI alone correct identification of CES has sensitivity 67%, specificity 81%, positive predictive value 85% and negative predictive value 60%.

Conclusions: Canal compromise on MRI does not predict the outcome of patients with CES. Reproducibility of MRI measurements of large disc protrusions has substantial agreement. MRI could be of help in equivocal cases if the scan shows a large disc.

Numerous studies have examined the biomechanical properties of the vertebral body following PMMA cement augmentation for the treatment of osteoporotic vertebral body fractures. To date there is no published literature reporting the effects of Vertebroplasty on internal intervertebral disc biomechanics which in turn have been shown to reflect loading patterns of the vertebral column.

To study effects of PMMA cement augmentation of vertebral body fractures on intervertebral disc biomechanics using stress prolifometry to assess differential anterior and posterior vertebral column loading.

Eight cadaveric motion segments were individually loaded on a hydraulically powered materials testing machine under 1.5kN of axial compression. Following fracture induction the lower vertebral body underwent Vertebroplasty.

Profiles of the vertically acting compressive stress were obtained by pulling a pressure sensitive transducer along the mid-sagittal diameter of the intervertebral disc. “Stress profile” measurements were obtained before fracture, following fracture, and after vertebro-plasty both in extension and flexion.

Stress profiles were integrated over area to calculate the compressive force across the disc. The compressive load acting on the neural arch was calculated by subtracting the disc force from the applied 1.5kN load.

In flexed postures posterior column loading increased from 17.1% to 42.2% following fracture (p< 0.01) and then decreased significantly from 42.2% to 23.68% following vertebroplasty (p< 0.03). There was no significant difference between pre-fracture and post-vertebroplasty status (p=0.11). In extended posture, fracture produced increased posterior column loading 72.9% vs 51.8% (p< 0.005) and following vertebroplasty there was no significant change (p=0.2).

In moderate degrees of flexion, vertebroplasty produces normalisation of load bearing through the anterior vertebral column and hence offloads the posterior elements to a significant degree. This could be postulated, to partly account for the analgesic effect seen following vertebroplasty in the clinical setting.