Proceedings of the 154th Meeting of the Society of British Neurological Surgeons: Dublin, Ireland, October 2009

Click to increase image sizeClick to decrease image size

W4-02: Chronic injury reduces the production of new neurons from endogenous stem cells in the adult human brain – a microenvironment effect

M. J. Zaben¹, L. E. Sundstrom² & W. P. Gray^1,3 (Division of Clinical Neurosciences, University of Southampton¹, Capsant Neurotechnologies Ltd.² and Wessex Neurological Centre³)

Objective: Despite overwhelming evidence for the presence of stem cells in the adult brain, these cells do not produce neurons after brain injury. This is true even in the permissive niche of the hippocampus where new neurons, important for memory consolidation, are normally produced throughout life. We wished to examine the hypothesis that the injured environment reprograms these stem cells and see if this was reversible.

Methods: We isolated neural stem cells from normal cortex and from the sclerotic hippocampus in 5 patients undergoing epilepsy surgery. Stem cells were either grown as free-floating spheres in an ideal environment or within a 3D tissue (Hi-Spot) generated from the hippocampus or cortex, and their ability to generate neurons quantified using BrdU and Neu-N immunohistochemistry.

Results: We show that stem cells from the uninjured cortex and the sclerotic (injured) hippocampus generate new neurons with equal efficiency when grown under ideal free-floating conditions. Interestingly, the numbers of newly-born neurons generated by matched stem cells grown in sclerotic hippocampal Hi-Spots were significantly lower (5.9 ± 1.4 cells/mm²) than those of the cortical Hi-Spots (23.7 ± 3.9 cells/mm²) (p < 0.001).

Conclusions: These results strongly implicate the microenvironment in defective hippocampal stem cell function in areas of chronic brain injury, and importantly from the viewpoint of brain repair, show that the effect is reversible.

W4-03: Association between cerebral chemistry and outcome after traumatic brain injury: microdialysis study in 223 patients

I. Timofeev¹, P. Al-Rawi¹, J. Nortje², K. Carpenter¹, M. O'Connell¹, M. Czosnyka¹, P. Smielewski¹, D. K. Menon², A. K. Gupta², J. D. Pickard¹, P. J. Kirkpatrick¹ & P. Hutchinson¹ (Department of Neurosurgery, University of Cambridge¹ and Addenbrooke's Hospital, Cambridge, UK²)

Introduction: Secondary insults adversely influence outcome following severe traumatic brain injury (TBI). Monitoring cerebral extracellular chemistry with microdialysis has the potential for early detection of metabolic derangements. An association between common biochemical markers and neurological outcome following TBI in a large cohort of patients has not been conclusively established, and evaluating such a relationship was the aim of this study.

Methods: Prospectively collected observational neuromonitoring data from 223 patients with TBI were analysed. All patients were managed in a single tertiary neurosurgical centre and required intensive care management and neuromonitoring. Monitoring modalities included intracranial pressure (ICP), cerebral perfusion pressure (CPP), cerebrovascular pressure reactivity (PRx) and microdialysis markers (glucose, lactate, pyruvate, glutamate, glycerol, lactate/pyruvate ratio (L/P ratio). Outcome was assessed using the Glasgow Outcome Scale (GOS). Patient averaged values of parameters were analysed by univariate non-parametric methods and multivariate logistic regression.

Results: Microdialysis monitoring commenced on median [IQR] day 1[1;2] from injury and median [IQR] duration of monitoring was 4[2;7] days. During the initial 72 hours of monitoring median L/P ratio (p = 0.026) and lactate (p = 0.033) levels were significantly lower in patients with favourable outcome (GOS good recovery, moderate disability). During the total monitoring period levels of glutamate (p = 0.048), L/P ratio (p = 0.044), ICP (p = 0.006) and PRx (p = 0.01) were significantly higher in patients who died. In a multivariate logistic regression model, glucose, pyruvate, L/P ratio and PRx were significant independent predictors of mortality, together with age and GCS (p < 0.0001).

Conclusion:The results of this study suggest that extracellular metabolic markers are independently associated with outcome following TBI, with L/P ratio being the most consistent predictor. Whether treatment related improvement in biochemistry translates into better clinical outcome remains to be established.

W5-02: Analysis of 4751 Emergency Referrals at two Neurosurgical Centres – Were they all necessary?

N. Mukerji, A. Paluzzi, D. Holliman, C. Barrett, P. Mitchell & P. J. Kane (Newcastle General Hospital, Newcastle upon Tyne, UK)

Objective: EWTD compliance of the neurosurgical rotas in the UK is a serious problem in the absence of appropriate staffing levels. The predominant issue with compliance is the rest requirement for an on-call rota. We analysed the out-of-hours activity of the registrars via a continuously maintained database at two centres in the Northern Deanery.

Methods: Similar MS Access databases maintained at Newcastle General (NGH) and James Cook University (JCUH) hospitals were pooled and analysed. A total of 4751 referrals over the past two years from both centres were studied. The analysis was undertaken using SPSS (Ver 13.0). Out-of-hours was defined as the time period between 2100 and 0800 hours for the purpose of this study.

Results:The patterns of activity out-of-hours were similar at both centres. Out of hours activity accounted for nearly 12 and 16 percent of all referrals at NGH and JCUH respectively. Approximately 28 and 56 percent referrals out-of-hours came from hospitals within the same trust at NGH and JCUH respectively. A&E and medicine accounted for nearly 70% of out-of hours referrals at both centres. Nearly 80% of the referrals related to acute head or spinal trauma, intracerebral haematomas and hydrocephalus-related presentations. 7% of out-of-hours referrals at both centres related to newly diagnosed brain tumours, 7% and 14% of referrals related to non-specific diagnoses, usually simply seeking advice.

Conclusions:The majority of the out-of-hours referrals for the registrars were bona fide emergency neurosurgical cases, which are important for both service provision and training. Establishing better referral pathways for newly diagnosed tumours and guidelines for seeking neurosurgical advice within the parent hospital is desirable. This may help to reduce the out-of-hours workload but this alone is still unlikely to ensure EWTD compliance for on-call rotas.

W5-03: Telephone logs of Specialist Registrars in Neurosurgery

S. Mahmood¹, K. S. Manjunath Prasad² & P. Mitchell¹ (Newcastle General Hospital, Newcastle upon Tyne, UK¹ and James Cook University Hospital, Middlesbrough, UK²)

Objective: To assess the magnitude and patterns of telephone call times in the work schedule of Specialist registrars (SpR) in neurosurgery. Design Observational and retrospective.

Methods: Telephone logs of the handheld telephone sets carried by the eight neurosurgical SpRs in one unit were obtained from the hospital switchboard to cover the period between 01.10.07–31.03.08 and this formed the primary data. Ringtime was added to the talktime for outgoing calls. Monday to Friday, 8am to 5pm was considered as regular hours and the rest as other hours.

Results: During this six month period the total number of calls was 22783 of which incoming calls numbered 12071 while there were 10712 outgoing calls. The maximum number of incoming calls/day/ SpR was in the range of 39 to 78 while the range for outgoing calls was 36 to 102. The mean number of telephone calls received by the on call SpR in a 24 hour period was 36, while the number of calls made during the same period was 33. On an average, each call lasted 1.7 minutes. Typically, the on call SpR spent 137 minutes on the telephone (69 calls), while the mean time spent on telephone calls on non on call days was significantly less at 24 minutes (18 calls). 67% of the calls happened out of regular hours.

Conclusions: Time spent on the telephone by the on call neurosurgical SpR is not small. The cause is multifactorial. In an era of working time directives, targets and litigation, the time spent on the telephone is a significant work component and would have a bearing on timelines.

W5-04: A web-based referral system for neurosurgery. Comparison with the traditional paper-based system

S. M. Thennakon & C. M. Tolias (Kings College Hospital, London, UK)

Objective: Neurosurgical referrals are notoriously difficult to track and review due to the poor quality of the paper-based recorded information. We have recently introduced a web-based referral system to a number of our referring hospitals. We compare this system to the paper-based one.

Methods: We analysed retrospectively all the paper-based referrals made to our unit in March 2009 from the whole of our catchment area and compared them to a year's referrals through the web-based system (three hospitals). We investigated the patterns of information recorded in both systems and identified the advantages and shortcomings of each.

Results: 196 patients were referred using the online system, between the 1st April 2008 and the 1st June 2009, while 483 patients were referred in total to our unit in March 2009 using the traditional method. There were significant problems with legibility of recordings. Several fields of information were missing (24% of referrals had no radiological finding, 36% of referrals had not documented details of the admitting team, 21% of referrals had not documented a contact telephone number/bleep). In comparison 100% of documentation was achieved with the online referral system. However, there was approximately 50% penetrance in the best performing trust inspite multiple campaigns and significant delays in responding to the referrals.

Conclusion: The traditional paper-based referral system does not provide an acceptable level of documentation. A web-based referral system, allows unprecedented insight in the process, and has significant advantages. However, there are significant difficulties in overcoming cultural barriers.

W5-05: Neurophobia and undergraduate neuroscience training

J. A. Giles & A. Kirkby (Greater Manchester Neuroscience Centre, Salford, UK)

Objective: Neurophobia is described as a fear of clinical neuroscience in medical students. The study examined the influence of clinical neuroscience training (CNT) on neurophobia and the role of anxiety, to inform curriculum planning.

Methods: A prospective cohort study was performed using questionnaires administered to students (n = 163) at a neuroscience lecture before and after CNT. A questionnaire was devised based on previous research to assess students' reported interest, knowledge, skills, confidence and perceived difficulty in a range of specialties. The State Trait Anxiety Inventory was used to measure anxiety. Comparative analyses were performed to examine how perceptions of neuroscience differed from other specialties, and changes before and after CNT. Correlational analyses were performed between neuroscience perceptions and anxiety.

Results:Before training neuroscience was ranked most difficult (p < 0.001*) and most interesting (p = 0.002*). After training, neuroscience was reported as less difficult (p = 0.006*) but still more difficult than other specialties (p < 0.001*). Self-reported knowledge and confidence in neuroscience increased after training (p < 0.001*) but were still lower than other specialties (p = 0.006*). There were significant negative correlations between state anxiety and some aspects of neurophobia, which although reduced remained significant after training. (*Wilcoxon)

Conclusions: Neurophobia was present in students before neuroscience training. Training has potential to reduce neurophobia but did not eliminate it. Anxiety was associated with some aspects of neurophobia. Earlier educational intervention may be warranted to prevent neurophobia and enhance specialty recruitment.

W5-06: Testing of practical skills in neurosurgery national selection

I. D. Kamaly-Asl, H. C. Patel, G. A. Roberts, F. Patterson, L. Faulkes, M. P. Powell & R. J. Nelson on behalf of the Neurosurgery National Selection Board

Objectives: The neurosurgical national selection board is aiming to develop skills assessment methodologies that will be predictive of future technical aptitude in potential neurosurgical trainees.

Methods: Over the last 2 years, skills assessments have formed part of the national selection centre for neurosurgical trainees. The stations have generally been based upon global assessment rating scales and have included a suturing model, a mock up of a frameless stereotactic biopsy using the Brainlab autopilot biopsy system, evaluation of microscopic skills with bead transfer and an assessment of stereoscopic vision. Some of the assessments counted towards final scores others have been pilots. The results have been studied for distribution of scores, associations with seniority and associations between tests using linear regression analysis.

Results: Results for suturing, biopsy and microscope show a normally distributed spread of scores with biopsy and microscope showing a wider spread of scores. The suturing model has consistently been shown to correlate well with level of seniority/experience. The other tests do not show significant correlation with seniority and are therefore testing more innate rather than learnt skills. There are a variety of significant correlative associations between the tests but they do not all consistently correlate implying that they are testing different aspects of technical skills.

Conclusions: Our developing skills assessments have shown some correlation with the suturing model, a well validated test of learnt practical skills. However, these individual tests are clearly testing other technical abilities and time will tell which are most predictive of future technical aptitude.

W5-07: Brain biopsy in non-neoplastic neurological disease

C. E. Gilkes, C. M. Rice, E. Kidgell, R. J. Hardie & R. J. Edwards (Frenchay Hospital, Bristol, UK)

Introduction: Patients with non-neoplastic neurological disease undergo diagnostic brain biopsy when other means of obtaining a definitive diagnosis have been exhausted. The threshold for requesting a diagnostic brain biopsy is variable, reflecting variations in the perceived risk of the intervention & its diagnostic utility.

Objective: To review the indications for, complications & outcome of diagnostic brain biopsies performed 2003–2008.

Methods: Subjects were retrospectively identified using theatre log books & histopathology reports. Case records were analysed jointly by a neurologist & neurosurgeon.

Results: 54 biopsies were performed in 51 patients in whom the preoperative working diagnosis was not clearly malignancy; representing 12.5% of all (431) brain biopsies. The indications included focal lesions (atypical cystic & ring enhancing 6%), less specific diffuse change (white matter changes 49%, white and gray matter changes 6%, extensive oedema 4%, multiple spontaneous vascular events 6% & leptomeningeal enhancement 16%), atypical cerebral atrophy (6%) & progressive/acute neurological decline with unremarkable parenchyma (8%). The diagnostic yield was 69% (52% truly diagnostic, 17% suggestive). Combining clinical, radiological and biopsy information, a confident diagnosis was achieved in 84%; demyelinating disease (20%) & vasculitis (16%) were the most common. Biopsy changed management in 61%. Complications: transient worsening paresis (n = 1 (1.8%)) & CSF leak requiring repair (n = 1 (1.8%)). In retrospect, biopsy was difficult to justify in only 2 patients (4%).

Conclusion: Brain biopsy for non-tumoural conditions carries little risk & may have significant implications for diagnosis and patient management in carefully selected cases.

T2-02: Spina bifida – Chiari II – hydrocephalus complex: long-term cognitive and functional outcome

M. D. Jenkinson, C. Hayhurst, S. Clark, S. Campbell, P. Murphy & C. L. Mallucci (The Walton Centre for Neurology and Neurosurgery, Liverpool, UK)

Objective: Long-term in outcome spina bifida –Chiari II – hydrocephalus complex is poorly understood. Neurosurgical outcome measures are crude, and neuropsychological testing is an increasingly important tool in outcome assessment. Cognitive function, health, disability and lifestyle were investigated in a cohort of prospectively followed adults.

Methods: Patient interviews were conducted to ascertain clinical history (number of ventriculoperitoneal shunts; foramen magnum decompression) and current status (progressive; stable; chronic; under investigation; pending surgery). Validated lifestyle and hydrocephalus questionnaires were administered. Cognitive function was assessed via the Repeatable Battery of Neuropsychological Testing (RBANS).

Results: Twenty-one patients were assessed; mean age 33 years (range: 19–45). All had myelomeningocoele closure at birth. Seventeen were shunted before 6 months; one by 18 months; one at 20 years. Fifteen had less than 5 shunt revisions; one patient had over 100 shunt procedures. Four have functioning ETV; 17 functioning VP shunts. Eight had foramen magnum decompression: seven had stabilisation of symptoms; one had improvement. Patient status was classified as stable (n = 14), progressive (n = 2), chronic (n = 2) and under investigation (n = 3). Only eight are living independently. Seven are ambulant independently or with walking aids. Eleven drive a car. Four are in paid employment; five do voluntary work. All patients demonstrated significant impairment (averaging 7th percentile) in at least 1 cognitive domain, most commonly working memory. Average IQ was 80.

Conclusion: Despite intervention in childhood and adequate cerebrospinal fluid diversion the prognosis for independent living into adulthood without symptoms is poor. All patients have elements of cognitive impairment.

T2-03 Management of children with optic pathway and chiasmatic-hypothalamic tumours – an experience with 60 patients

J. R. Goodden¹, N. Thorpe², B. Pizer¹ & C. L. Mallucci¹ (Alder Hey Children's NHS Foundation Trust, Liverpool, L12 2AP¹ and Clatterbridge Centre for Oncology NHS Foundation Trust, Clatterbridge Road, Bebington, Wirral, CH63 4JY²)

Background: Optic pathway and chiasmatic-hypothalamic gliomas (OPCHG) are rare tumours, mainly presenting in childhood. Due to their location they are difficult to treat through surgery alone and often require combination therapy to achieve disease control. We present our experience in management of 60 of these unusual tumours.

Methods:Patients were identified from the Neuro-oncology database. Casenotes were retrieved and retrospectively reviewed. Data was collected regarding presenting symptoms, visual function, endocrine function, NF1 genotype, tumour size and location, neuroradiological imaging characteristics, histopathological findings, treatment (surgery, chemotherapy & radiotherapy), and control/relapse rates.

Results: Sixty patients were treated at AlderHey Children's Hospital between 1996 & 2008. Mean age at diagnosis was 7.9 years (range 3-months to 16½ years, median 6 years), with equal sex distribution. Follow-up ranged from 6-months to 13 years (mean 6.1 years, median 6.2 years). Treatment strategies included observation, and combinations of surgery (biopsy or debulking), chemotherapy and or radiotherapy. Tumour control rates and outcomes are reviewed.

Conclusions: Good long-term survival and functional outcomes can be achieved in children with OPCHG. Tumour control has been achieved through surgery or chemotherapy or radiotherapy in varied combinations. Surgery via a midline transcallosal approach can safely debulk the centre of third ventricular tumours whilst preserving hypothalamic and visual function. Due to high rates of long-term survivors we recommend observation or chemotherapy for initial treatment. Surgery has roles for diagnosis, tumour control, relief of mass effect, and optic decompression. We prefer to limit radiotherapy to older children due to late effects.

T2-04: Posterior fossa tumours in children: time to diagnosis

M. A. J. White, E. J. St George & J. Brown (Institute of Neurological Sciences, Southern General Hospital, Glasgow, Scotland)

Introduction: Paediatric posterior fossa tumours are common, responsible for approximately 70% of all childhood brain tumours. Initial presenting symptoms can be intermittent and vague delaying recognition until late in tumour progression.

Objectives: To determine the time interval between the first attributable symptom and radiological diagnosis in these paediatric tumours, comparing our findings with those in the established literature.

Methods: A retrospective case note review was performed of all children (n = 49) with posterior fossa tumours admitted in the last 5 years (m 23/26 f). Median age-6 (5 mths–16 yrs). Pathology included medulloblastoma (n = 16), pilocytic astrocytoma (n = 14), ependymoma (n = 5), brainstem glioma (n = 4) and others (n = 3). Patients with tumours not confined to the posterior fossa and notes with inadequate clinical information were excluded (n = 6) Results The mean time interval between symptom onset and radiological diagnosis was 224 days and between symptom onset to review by first physician was 62 days (1–365 days). The mean time interval between first physician review and radiological diagnosis was 162 days (6 hrs–730 days) The most common presenting symptoms were headache, nausea & vomiting, ataxia and oculomotor deficits. Most patients had seen an average of 2 specialists prior to ultimate diagnosis, tending to concentrate on specific individual symptoms.

Conclusions: To improve delays in diagnosis of this tumour group, more education/information may be required at several levels including parents, GP, A&E, and tertiary specialties. A protocol to identify these tumours in children with prolonged symptoms may be helpful as too often individual symptoms are the focus rather than the patient as a whole.

T2-05: The impact of electromagnetic navigated shunt placement on failure rates: A two centre pilot study

C. Hayhurst^1,2, M. D. Jenkinson^1,2, S. Clark^1,2, R. D. S. Nandoe³, T. Beems³ & C. L. Mallucci^1,2 (The Walton Centre for Neurology and Neurosurgery, Liverpool, UK¹, The Royal Liverpool Children's NHS Trust, Liverpool, UK² and University Medical Center, Nijmegen, The Netherlands³)

Objective: Up to 40% of shunts fail in the first year. The role of catheter position on failure rates has not been demonstrated. We conducted a prospective study of navigated shunt placement compared to standard blind shunts in two European centres, to assess impact on failure rates.

Methods: All patients undergoing de novo ventriculoperitoneal shunts were included, excluding slit ventricles. The first cohort underwent standard shunt placement using anatomical landmarks. Both centres subsequently adopted electromagnetic navigation for routine shunts forming the second cohort. On postoperative CT catheter position was graded from 1 –3 (1, optimal position free floating in CSF; 2, touching choroid or ventricular wall; 3, intraparenchymal). Episodes of shunt revision were recorded. Shunts were followed for minimum 6 months or until failure. Results: There were 41 blind shunts and 34 navigated shunts, including adults and children. 74% of navigated shunts were grade 1 versus 37% of blind shunts (p = 0.0014). There were no grade 3 placements in the navigated group, 8 in the blind group and 75% of these failed. 9 of 14 shunt failures in the blind group occurred within 4 weeks. 2 shunt failures occurred within 4 weeks in the navigated group, reducing the 30 day revision rate from 22% to 5.9% (p = 0.0485). Shunt failures within 30 days were due to proximal obstruction in 78% of blind shunts (7/9) and 50% of navigated shunts (1/2).

Conclusion: Using image guidance in shunt surgery can eliminate poor shunt placement, resulting in a significant reduction in early shunt revisions rates.

T2-06: Analysis of brain shift in deep brain stimulation

A. Borg¹, E. A. Petersen^1,2, E. M. Holl^1,3, I. Martinez-Torres¹, T. Foltynie¹, P. Limousin¹, M. I. Hariz¹ & L. Zrinzo¹ (Unit of Functional Neurosurgery, Sobell Department of Motor Neuroscience and Movement Disorders, Institute of Neurology, Queen Square, London¹, Department of Neurosurgery, University of Texas, Southwestern, Dallas TX, USA² and Department of Neurosurgery, Medical University, Graz, Austria³)

Objective: Brain shift is often considered a source of inaccuracy in functional neurosurgery. An image-guided approach requires a surgical technique that limits brain shift. The aim of this study was to quantify postoperative subcortical brain shift when using this approach. The effect of subcortical brain shift on clinical outcome was also analysed.

Methods: Images from 167 consecutive patients undergoing DBS electrode implantation were analysed. Accurate pre and postoperative stereotactic coordinates could be calculated from MRI for both anterior and posterior commissural points in 136 procedures allowing comparison between the two. The amount of pneumocephalus was also calculated.

Results:The mean scalar change in AC-PC length (AC-PC) was 0.6 ± 0.4 mm. There was no statistically significant difference in AC-PC between subgroups when considering anatomic target (p = 0.95), age (p = 0.64), gender (p = 0.59), unilateral or bilateral implantation (p = 0.15), or degree of pneumocephalus (p = 0.12). In 53 patients with Parkinson's disease undergoing bilateral STN DBS with at least 6 months of follow-up, the mean percentage improvement in UPDRS III score was 52% when comparing preoperative OFF drug and postoperative OFF drug ON stimulation. There was no correlation between magnitude of brain shift and clinical improvement in these patients (R² < 0.04). There was no symptomatic haemorrhage in any patient in this study.

Conclusions:The observed brain shift in this image-guided surgical approach was considerably less than that reported by series where MER was employed. The magnitude of brain shift observed in this study appears to have no bearing on clinical results in those patients undergoing STN DBS for Parkinson's disease.

T2-07: Visual Field Deficits after Selective Trans-Sylvian Amygdalohippocampectomy

S. M. Ross, A. Khan & W. P. Gray (Wessex Neurological Centre, Southampton, UK)

Objective: To identify and evaluate the incidence of visual field deficits (VFD) comparing different perimetry systems after selective Trans-sylvian Amygdalohippocampectomy (TAH) performed consecutively by one neurosurgeon and it's significance for driving.

Design: An analysis of pre- and post-operative Goldmann perimetry charts using the I4e stimulus and postoperative Humphrey and binocular Esterman visual fields. Subjects: Total of 39 patients, age range 5–55 years. Mean follow-up of 2 years 7 months. Results: 26 patients (67%) had postoperative visual field deficits on Goldmann and Humphrey perimetry. 13 patients (33%) had no visual field deficits. 37 out of 39 patients (95%) passed the Esterman test for driving. 2 patients failed the Esterman test, one was seizure free and the other was still having some seizures. 34 patients (87%) were seizure free and classified as Engel 1. 4 patients (33%) had some seizures and were Engel 2. 1 patient was classified as Engel 3 but passed the Esterman test.

Conclusions: There is a low incidence of visual field deficits precluding driving after a selective TAH with excellent seizure control. Both Goldmann and Humphrey perimetry give similar patterns, although Humphrey perimetry is more standardised. The Esterman test should be performed in all patients who wish to drive.

T2-08: Long term outcome of Intrathecal Baclofen Pump implantation in patients with advanced Multiple Sclerosis

A. A. Khan, I. Birks-Agnew, E. McGilloway, P. Bullock & D. Rushton (Kings College Hospital, London, UK)

Objective: To review the long term outcome in a group of severely disabled patients with Multiple Sclerosis (MS) following intrathecal Baclofen (ITB). Improved quality of life with reduced painful spasms and decreased spasticity are described as well as the complications of treatment.

Method: A retrospective study of patients with advanced MS who underwent ITB therapy between 1996 and 2007 from a total group of 132 patients undergoing pump implantation.

Results: 40 patients with MS underwent ITB. Mean age at diagnosis of MS was 31 years (range 16–46), and age at implantation of pump 46 (range 28–61). The majority (36) were wheelchair or bed bound. The Expanded Disability Status Scale was 8.0 (range 6.5–9). The average pre-operative oral Baclofen dose was 61mg/day (range 0–100), one year post-operatively the average intrathecal dose was 209μg/day (range 24–449). The average Modified Ashworth Scale improved from 4.6 (range 3–5) to 1.0 (range 0–4) and the Penn Spasm Frequency Scale from 3.6 (range 1–4) to 1.1 (range 0–2). Four patients had problems with the catheter delivery system due to migration. Five pumps were replaced due to expiration of battery, one due to flipping (pump rotation) causing non-delivery of Baclofen and one due to thinness of skin over the implant area.

Conclusions: ITB pump significantly improves the long term outcome in MS patients even with advanced disease. This series of patients provides an insight into the benefits of ITB in advanced MS where there is a sustained clinical improvement and low risk of complication.

T3-02: Is it appropriate for GPs to refer patients with presumed acute lumbar spinal pathology directly to the on call neurosurgical team?

A. Paluzzi & J. J. Nissen (Newcastle General Hospital, Newcastle upon Tyne, UK)

Background: the policy concerned with GP referrals of patients with presumed acute lumbar spinal pathology to the on call neurosurgical or orthopaedic team varies between centres in the UK. We debate whether it is appropriate for GPs to have direct access to tertiary referral centres for this type of referral.

Methods: A Microsoft Access database of urgent referrals to the on call neurosurgical team at NGH was analysed over a period of 9 months from August 2008 to April 2009. The analysis was undertaken using SPSS.

Results: 2672 emergency referrals were recorded over a period of 10 months, of which 355 (13.2%) were made by GPs. 143 referrals out of 355 (40%) were about patients with symptoms including low back pain, unilateral of bilateral sciatica, acute foot drop, urinary hesitancy, retention or incontinence. Out of the 143 patients referred, 11 (7.7%) underwent surgical intervention within 48 hours from the time of referral, 4 of them having radiological evidence of cauda equina compression and 7 of lateral disc protrusion.

Conclusions: Only 7.7% of the total number of patients referred with a working diagnosis of acute lumbar spinal pathology underwent surgery within 48 hours. We suggest that these patients are referred for assessment and the initial investigations to the local A&E department, acting as a filter for the tertiary referral centre.

T3-03: Reliability of clinical examination in diagnosing cauda equina syndrome

K. Balasubramanian, P. Kalsi, K. S. Manjunath Prasad & C. G. Greenough (Spinal surgery service, James Cook University Hospital, Middlesbrough, UK)

Objective: To assess the predictive value of clinical symptoms and signs in Magnetic Resonance Imaging (MRI) positive Cauda equina syndrome (CES).

Method: A retrospective observational study. Clinical files and MRI scans of 80 patients who underwent urgent MRI scans for suspected cauda equina syndrome were reviewed. Common presenting clinical symptoms and signs were assessed against a positive MRI scan. To prevent overestimation of statistical significance for small data, Yates Chi Square test was used to assess the significance of each clinical symptom and sign for a positive MRI scan.

Results: Of 80 patients with a clinical suspicion of CES, 15 patients had MRI confirmed cauda equina syndrome and underwent urgent surgery. Back pain, unilateral sciatica, bilateral sciatica, bladder incontinence, bladder retention, perianal sensory loss and previous surgical interventions were not statistically significant to suggest a cauda equina syndrome. We found that saddle anaesthesia is the only sign, which was statistically significant (p = 0.03) to predict a CES.

Conclusion: In our study, no clinical symptom or sign apart from saddle sensory deficit, could be relied upon to confirm CES. Hence those presenting with features of CES should undergo an urgent MRI scan for a definitive diagnosis.

T3-04: Out of hours MRI scanning for cauda equina syndrome (CES): what is the positive pick-up rate and what are the final diagnoses in those with a negative scan?

A. K. Demetriades, E. Broughton, J. Akinwunmi, G. Critchley, L.Gunasekera, J. S. Norris & C. Hardwidge (Hurstwood Park Neurological Centre, Brighton and Sussex University Hospitals)

Objective: Emergency MRI scanning services are subject to out-of-hours regulations. Regional neurosurgical centres are routinely relied upon to provide such a service in patients presenting with suspected acute CES. The aims were twofold; first, to assess the positive ‘pick-up’ rate of these referrals; what percentage led to an emergency operation. And second, to identify the final diagnosis of those patients with a ‘negative’ scan result.

Methods:All patients who were admitted in 2008 for an out-of-hours emergency MRI for possible CES were identified. Radiological results were analysed in relation to the presenting symptoms and signs. Patients who were returned to their referring hospitals because of no radiological proof of CES were traced and final diagnoses identified via communication with their referring physician and general practitioner.

Results: 33 patients were admitted for an out-of-hours MRI during the study period. 10 were proven to have acute CES and underwent an emergency operation. The male:female ratio was 1:1 and the average age 38.9 (34–46). 80% had all of radiculopathy, paraesthesia and sphincter disturbance; 2 patients had paraesthesia and sphincter loss in the absence of radiculopathy. 80% had acute back and leg pain. All had a lumbar disc prolapse on MRI and underwent a decompressive laminectomy and discectomy. 23 patients had a ‘negative MRI’. The male:female ratio in this group was 1:2.3 and the average age 53.1 (19–81). In this group there was a 70% incidence of incontinence, 60% paraesthesia, 40% radiculopathy, and 35% incidence of back and leg pain. While a small proportion of these remained undiagnosed (13%), eventual diagnoses fell into 4 categories:

• Neurological: Guillain-Barré, Multiple Sclerosis, transverse myelitis, central pontine myelinolysis.

• Malignant: sacral metastasis, sacral Ewing's sarcoma, rectal carcinoma.

• Psychiatric: functional disorder, dissociative personality disorder.

• Other: sacral fracture and rhabdomyolysis.

Conclusions: The positive pick up rate for out-of hours MRI for suspected CES was 30%. The eventual diagnosis in those with a ‘negative’ scan is varied but potentially serious and appropriate referrals should be made to investigate other neurological, malignant, psychiatric or other aetiologies.

T3-05: Forty Consecutive Cases of Medico-Legal Cauda Equina Syndrome : A Single Expert Evidence

N. V. Todd (Regional Neurosciences Centre, Newcastle General Hospital, Newcastle upon Tyne, UK)

Methods: A prospective longtitudional study of forty consecutive people who were referred to a single neurosurgical expert with cauda equina syndrome (CES).

Results:Age: median = 35, mean = 47. Male:female 1.4:1. Level and pathology: L5/S1 PLID 17 (42.5%), L4/5 13 (32.5%), L3/4 3 (7.5%), L2/3 2 (5%), stenosis 5 (12.5%). First presentation = (i) unilateral radicular pain = 10 (25%) (ii) bilateral radicular pain = 13 (32.5%) (iii) incomplete cauda equina syndrome (CESI) = 14 (35%) (iv) bladder paralysis with incontinence (CESR) = 3 (7.5%). 39 primary breaches of duty of care: orthopaedics 18 (46%), GP 8 (21%), A&E 7 (18%), neurosurgery 4 (10%), nursing 1 (3%), ambulance 1 (3%). 8 neurosurgical breaches (4 primary, 4 subsequent): (i) 4 failures to admit; (ii) 2 failures to investigate; (iii) 2 drill plunging. 11 (27.5%) had iatrogenic injury: unrecognised EDH = 4; inadequate decompression = 3; drill injury = 3; dural tear = 1; unknown = 1: long-term bladder paralysis in all cases. 28 patients were previously working; 6 (21%) returned to normal/modified work; 22 (79%) did not work again. 33 had compressive lesions causing CESR. 29 (88%) did not recover function. 7 patients were operated on within 24 hours of CESR, 4 recovered bladder function, 3 did not.

Conclusion: A unique though highly selected group. 93% presented with treatable lesions; only 7% presented with bladder paralysis. 7 patients with CESR were decompressed within 24 hours, 4/7 recovered bladder function. No patient with CESR treated beyond 24 hours recovered bladder function.

T3-06: How comprehensively is a patient's sciatica managed with analgesia before resorting to surgical decompression?

S. D. Freshwater, A. Steen, D. G. Currie & P. M. Brennan (Deparrtment of Neurosurgery, Aberdeen Royal Infirmary, Aberdeen, Scotland)

Objective: Sixty percent of disc related sciatica recovers spontaneously in 3 months. A Cochrane review found no significant difference in pain scores in comparisons of surgical and non-surgical treatment at 4 years. This suggests optimising non-surgical management, such as analgesia, may reduce need for surgery.

Methods: We surveyed pre-operative analgesic history of patients attending for elective management of sciatica. Patients rated their severity of back and leg pain, and adequacy of current and previous analgesia, using a visual analogue scale (0/no pain-10/severe). The WHO pain ladder was used to benchmark analgesia use.

Results:Fifty patients were surveyed. The average severity of back pain was 3.7, and leg pain 4.8. Average efficacy of current analgesia was 4.3. Only 26% of patients took paracetamol, NSAIDs and a weak opiate. Side effects stopped some taking medications that improved their pain. However, many patients with inadequate pain control had not tried analgesics other than their current regimen, despite average scores of severity of back pain of 4.2 and leg pain of 5.3; higher than the whole cohort. Further, only 22% of all patients had tried gabapentin, and 18% a strong opiate; both can help in sciatica. Those with most severe leg pain (>5) were as likely as those with less severe leg pain (<5) to have tried gabapentin or strong opiates (47v45%).

Conclusion: The WHO analgesia ladder provides a framework for titrating pain control in sciatica, since evidence-based guidance is lacking. Nevertheless, prescribing in patients appears haphazard. Prospective studies are needed to ascertain whether guidelines, based on the ladder, can improve pain control, in turn decreasing the requirement for surgical symptomatic control.

T3-07: The clinical value of early postoperative MRI after lumbar spine surgery

M. Crocker, T. Jones, P. Rich, B. A. Bell & M. C. Papadopoulos (St George's University of London, London, UK and St George's Hospital, London, UK)

Objective: MRI scanning has long been considered difficult to interpret in the early period following lumbar spine surgery, and hence of very limited value. We investigate the hypothesis that MRI scanning within six weeks of lumbar spine surgery cannot accurately diagnose neural compression in symptomatic patients.

Subjects: A retrospective series of 32 consecutive patients who underwent early postoperative MRI following lumbar discectomy or laminectomy for continued, worsening or new symptoms of neural compression.

Methods: The neuroradiologists' reports were evaluated for the degree of neural compression identified (none, low, moderate or severe) and confidence level (low, medium, high). These MRI findings were then compared to the patients' subsequent course and findings of any surgery performed.

Results: 19 of 29 scans (66%) were confidently predictive of the correct treatment pathway (reoperation with positive finding or conservative treatment with a good outcome) whereas 3/3 (100%) patients who had conservative management despite the MRI confidently suggesting compression had poor outcome. The MRI is highly likely to influence management: 11/14 (79%) of patients with scans suggesting neural compression progressed to revision surgery and 18/18 (100%) patients with no neural compression on MRI were managed conservatively.

Conclusions:Our data suggest that early MRI scanning after lumbar laminectomy or discectomy accurately detects neural compression at the surgery site in patients with continued or worsening symptoms.

T5-01: The Use of Solitaire Stent in the Treatment of Intracranial Aneurysms

F. Robertson, J. Madigan, J. Grieve, L. Watkins, S. Brew & N. Kitchen (The National Hospital for Neurology and Neurosurgery, London, UK)

Objective: To report our initial experience with stent placement for cerebral aneurysm treatment.

Methods: We report our series of 17 Solitaire (4 mm × 20 mm) stents in 16 aneurysms in 16 patients. These include 9 stents deployed in the context of subarachnoid haemorrhage (3 pre-planned/6 intraprocedural thrombotic complication salvage) and 7 elective aneurysm treatments (stent pre-planned)

Results: The stent was successfully deployed in 15/17 cases (88%) in vessels ranging from internal carotid artery (6 mm) to A2 segments (1.5 mm). We saw 2 stent displacements in 1 patient, early in series, one successfully retrieved without complication and one displaced into the aneurysm sac after detachment, both without clinical sequela. No thromboembolic complications were seen in relation to any stent. Two patients, both admitted with poor grade subarachnoid haemorrhage, died at 12 days and 14 days post-procedure, respectively. The aneurysm was completely occluded immediately in 14/16 cases (88%) and at 6 month angiographic follow-up in 3/5 (60%). The stent was successfully deployed to recanalise a vessel compromised by thrombus developing intraprocedurally in 6 cases with excellent angiographic results and no clinical sequela, obviating the need for intravenous Reopro in all cases. We also describe a novel application of the Solitaire stent in the treatment of 2 very wide-necked aneurysms.

Conclusions: In our experience, the Solitaire™ AB stent offers a significant step forward in the treatment of intracranial aneurysms in terms of its ease of positioning, deployment and retrieval, range of suitable vessel diameters and safety profile.

T5-02: Current practice in the treatment of cerebral aneurysms in a national neurosurgical centre post ISAT

J. Garry, E. Fenton, H. Hurley & D. Rawluk (Department of Neurosurgery, Beaumont Hospital, Dublin, Ireland)

Objective: The aim of this study was to assess current practice in the treatment of cerebral aneurysms in our unit post ISAT.

Methods: A retrospective analysis of patients who underwent clipping or coiling of cerebral aneurysms in the year 2008 was performed. Collected data included patients' age, sex, aneurysm location, aneurysm size and WFNS grade (if post haemorrhage) and a chart review was performed to assess periprocedural complications.

Results:244 aneurysms were treated. 149 (61%) were coiled and 95 (39%) were clipped. The mean age of coiled patients was 53.3 years (range 28–85 years). The mean age of clipped patients was 51 years (range 25–74 years). 97 patients were post subarachnoid haemorrhage. The male to female ratio was 42:58 in the clipped group and 37:63 in the coiled group. The periprocedural complication rate was 13.42% in the coiled group and 18.94% in the clipped group. 31 of the coiled patients were less than 40 years of age.

Conclusion: Although endovascular coiling is well established in the treatment of cerebral aneurysms, there remains a significant role for clipping (as our current practice demonstrates). Recent data has shown the long term benefits of clipping patients under the age of 40 and this should be further addressed.

T5-03: The use of SonoWand (3D-ultrasound) in intra-operative image guidance and clipping of cerebral aneurysms

R. C. Nordstrom, S. Akmal & K. S. O'Neill (Charing Cross Hospital, London, UK)

Objective: To assess the benefit of 3D intra-operative structural and doppler ultrasound (SonoWand) in cerebral aneurysm surgery.

Design: A prospective cohort study over four months.

Subjects: Patients with aneurysmal grade I/II subarachnoid haemorrhages (SAH) not amenable to coiling.

Method: All patients underwent CT angiography (CTA) pre-operatively for localisation and guidance. At dural opening, 3D ultrasound imaging and Doppler of the operative field was acquired to plan trajectory and dissection. This was updated intra-operatively regularly with Doppler flow to assess vessel patency pre and post-clipping. Post-surgery neurological state, vasospasm and 3 month Modified Glasgow Outcome Scores were recorded.

Results: A case series of eight males, mean age 53, was collected over four months. There were three MCA, two pericallosal and three ACOM aneurysms. Ultrasound demonstrated the position of the vasculature tree and aneurysms throughout surgery, accounting for shifts due to CSF leakage. There was no post-operative deterioration of pre-existing neurology, vasospasm and all made a good recovery with no incidences of stroke at three months.

Conclusion: Intra-operative 3D-ultrasound imaging is cost effective and user friendly. It allows for real-time structural neuronavigation. It is a valuable adjunct to aneurysm surgery, allowing aneurysm localisation and assessment of parent and daughter vessel patency pre and post-clipping.

T5-04: Early changes in the distensibility of the atheromatous carotid artery following aggressive statin therapy can be detected non-invasively using phase-contrast magnetic resonance imaging: a double blind study

S. P. S. Howarth, T. Y. Tang, S. Wallace, M. J. Graves, J. U. King-Im, A. Brown, Z. Li, I. Wilkinson, P. J. Kirkpatrick & J. H. Gillard (Addenbrookes Hospital, Cambridge, UK and Hope Hospital, Manchester, UK)

Background: Vascular distensibility and stiffness are closely related to cardiovascular and cerebral vascular outcomes. Decreased carotid distensibility has been associated with an increased atheroma burden. Magnetic resonance imaging (MRI) is a powerful tool for the non-invasive assessment of vascular anatomy and physiology. We report a double blind study using phase contrast imaging techniques to compare changes in carotid distensibility in subjects with carotid stenosis randomised to either high or low dose atorvastatin.

Methods and results: Subjects with moderate carotid stenosis and demonstrable inflammation as determined by ultra-small superparamagnetic iron oxide (USPIO) enhanced MRI were imaged at 1.5 Tesla before and 12 weeks post randomisation to either high (80 mg) or low (10 mg) dose atorvastatin. Real-time brachial blood pressure measurements were made during sequence acquisition and a central aortic pulse pressure was derived from off-line radial artery tonometry performed before imaging. Fasting serum lipid levels were measured at 0 and 12 weeks. There was a statistically significant increase in the distensibility coefficient and compliance coefficient of both the common carotid artery and the internal carotid artery in the group randomised to high-dose atorvastatin.

Conclusions:Phase-contrast MRI is both robust and feasible in the non-invasive assessment of changes in vascular distensibility over time with pharmacological intervention. High-dose statin therapy led to a significant improvement in the distensibility of the stenosed carotid artery and these changes were apparent after only 12 weeks therapy.

T5-05: Current treatment practice for cavernous malformations: international survery

R. Al-Shahi Salman¹, L. Chilton², A. D. Mendelow² & N. D. Kitchen³ (Division of Clinical Neurosciences, University of Edinburgh, Edinburgh, UK¹, Regional Neurosciences Centre, Newcastle General Hospital, Newcastle, UK² and Victor Horsley Department of Neurosurgery, The National Hospital for Neurology and Neurosurgery, London, UK³)

Introduction: From published case series of both conservative and actively treated cavernomas there would appear to wide variation in worldwide clinical practice. This survey aims to quantitate that uncertainty over optimum management.

Methods: We surveyed the following groups: ARUBA, ABN, SBNS, STICH II trialists, and the Joint Cerebrovascular Section of AANS/CNS. We asked these professionals to complete a brief, web-based survey (http://research.ncl.ac.uk/nctu/cavorrtt/Questionnaire.php), indicating what treatment they would recommend for various clinical scenarios involving a hypothetical adult with a cavernoma. We subdivided these scenarios by brain location of the cavernoma (lobar, deep, or brainstem) and mode of presentation (one haemorrhage, two haemorrhages or, epilepsy, or incidental).

Results: The 194 responses that were suitable for analysis were from a broad spread of countries: UK(60), USA(30), Germany(9), Spain(3), Canada(3), Italy(3), India(3), France(2), and we received single responses from 18 other countries. Of the 194 responses, 41 were from neurologists and 153 were from neurosurgeons, who between them saw a median of 5 new cavernoma patients per year. The treatment preferences were variable; surgery was advocated in 0–80% of cases, radiosurgery in 1–23%, conservative management in 2–92% and unsure in 7–25% depending on clinical scenario and cavernoma location. For example in patients with 2 haemorrhages from a brainstem cavernoma 59% of respondents advocated surgery, 14% radiosurgery, 7% were conservative and 20% unsure. The full range of results will be presented.

Conclusion: Considerable uncertainty exists amongst cerebrovascular experts as to optimum management of cerebral cavernomas. We thus argue for prospective studies to elucidate best treatment pathways.

T5-06: Intravenous Interleukin-1 Receptor Antagonist achieves experimentally neuroprotective cerebrospinal fluid concentrations within a therapeutic time window

J. Galea, S. Hulme, K. Ogungbenro, S. Scarth, M. Hoadley; A. King, L. Aarons, N. J. Rothwell, S. Hopkins & P. Tyrrell (Manchester Academic Health Science Centre)

Objective: Interleukin-1 (IL-1) is implicated in experimental ischaemic brain injury. Its naturally occurring antagonist (IL-1RA) is highly neuroprotective and can be administered therapeutically, showing few adverse effects and inhibiting the systemic acute phase response. A single dose regime pilot study showed that it penetrates cerebrospinal fluid (CSF) at experimentally-therapeutic concentrations, but this was achieved rather slowly. Determining the optimal protocol for rapid IL-1RA delivery in subarachnoid haemorrhage (SAH) is essential before testing biological efficacy.

Design: An open-labelled dose-ranging pharmacokinetic study.

Subjects: 25 patients with SAH and external ventricular drains inserted for clinical reasons.

Methods: Patients received intravenous (IV) IL-1RA as a bolus followed by a 4 hour infusion. Pharmacometric analysis of pilot data enabled identification of the administration regime that could achieve experimentally-therapeutic CSF IL-1RA levels within 30 min. Patients were allocated to five regimes, using IV boluses of 100 to 500 mg and IV infusions ranging from 1 to 10 mg kg⁻¹ h⁻¹. Plasma and CSF sampling was performed as informed by a D-optimal design. Analysis was done using nonlinear mixed effects modelling.

Results: IL-1RA plasma and CSF concentrations fell within predicted intervals. The regime leading to experimentally-therapeutic CSF concentrations of IL-1RA within 40 minutes was 500mg bolus followed by IV infusion at 10 mg kg⁻¹ h⁻¹.

Conclusion: Experimentally neuroprotective CSF concentrations in SAH patients was achieved within a reasonable therapeutic time window. Pharmacokinetic analysis suggests IL-1RA transport across the blood-CSF barrier is passive. Identification of this delivery regime allows further efficacy studies of IL-1RA in SAH.

T5-07: ICG Angiography: quality assurance for neurovascular surgery

R. A. Trivedi, T. Santarius, R. W. Kirollos & P. J. Kirkpatrick (Addenbrooke's Hospital, Cambridge, UK)

Objective: To describe the utility of indo-cyanine green (ICG) fluoroscopic angiography for vascular neurosurgical practice.

Methods: A retrospective review of all patients surgically treated for neurovascular disease, since the introduction of a microscope with capabilities of fluoroscopic angiography. Video recordings of ICG-angiography were analysed and correlation was made with post-operative clinical assessments.

Results:50 patients were included in the study over a 6 month period. 46 were treated for aneurysm, 2 for dural fistuale and 2 by EC:IC bypass for carotid occlusion. Video recordings were available for analysis in only 27 patients, due to microscope availability. ICG angiography confirmed vessel patency peri-operatively in 24/25 aneurysm patients and no lag time to fill in 1/2 EC:IC patients. Post-operative neurological deficits were noted in only 1/24 aneurysm patients in whom no ICG abnormality was noted. In 1 aneurysm patient, poor flow in the distal vessel was associated with a transient post-operative deficit. Post-op neurological deficit was evident in the EC:IC bypass patient, in whom there was a delay in transit time, and retrograde filling of the graft.

Conclusions: ICG angiography provides a reliable peri-operative quality assurance for clip placement in surgically treated aneurysms and graft patency in revascularisation patients.

T5-08: Intraoperative ICG Videoangiography as an aid to the Surgical Treatment of Spinal Vascular Malformations

D. C. Walsh & R. W. Gullan (King's College Hospital, London, UK)

Objective: Intra-operative Indocyanine Green (ICG) Videoangiography has become an established aid to cerebrovascular surgery. We describe the application of this technique in our series of microsurgically treated vascular malformations of the spinal cord.

Methods: Indocyanine green dye was administered as an intravenous injection and detected intraoperatively via an 800 nm filter on the surgical microscope. No special vascular access was required.

Results: 14 patients underwent surgery for vascular malformations of the spinal cord from Sept 2007 to June 2009. 11 were type 1 spinal dural AVFs, 2 were type 2 intramedullary lesions and one a probable type 4 perimedullary fistula. Two of the type 1 lesions were of a low flow subtype. In this sub-group obliteration of the fistula could most easily be demonstrated with the ICG technique. In type 2 and 4 lesions the ICG angiogram is used to delineate the nidus or fistulae and correlate with the pre-operative digital subtraction angiogram intra-operatively. Two rapidly deteriorating patients were explored surgically when diagnostic spinal angiography proved technically impossible. In these cases the ICG angiogram provided intraoperative direction on surgical strategy. No patient suffered a complication of dye administration. Intra-operative video is presented to illustrate the technique.

Conclusions: ICG Videoangiography is a time-efficient and safe alternative to intra-operative spinal angiography. It provides useful information on intra-procedural changes in haemodynamics and completeness of treatment.

T5-09: New “TRICKS” (Time Resolved Imaging of Contrast Kinetics Sequence) magnetic resonance angiography in the work-up of spinal vascular malformations

S. P. MacNally, A. Siddiqui, N. Deasy, T. Hampton, C. M. Tolias & D. C. Walsh (Kings College Hospital, London, UK)

Objective: Conventional catheter angiography is the gold standard investigation to evaluate suspected spinal vascular malformations but has significant drawbacks and potential morbidity. TRICKS (time resolved imaging of contrast kinetics sequence) magnetic resonance angiography provides dynamic information of the spinal vasculature. We describe our experience with this technique in spinal vascular malformations.

Methods: A retrospective analysis of prospectively collected data in 24 patients with suspected spinal vascular malformations. All patients underwent TRICKS MRA and selective catheter angiography. TRICKS images were reviewed by two neuroradiologists blinded to the catheter angiography results.

Results: Catheter angiography revealed 14 spinal dural AVFs, 5 spinal AVMs, 2 negative angiograms and 3 non-diagnostic studies that were later confirmed to have spinal AVMs at surgery. TRICKS was positive in 23 of 24 patients with one false positive and no false negatives. In the dural AVF group, TRICKS identified the side and precise segmental level in 8 and was accurate to within one segmental level in 6. Follow up TRICKS was performed in 9 of 12 post-operative dural AVF cases and revealed an absence of early venous filling in all. In the AVM group we found that because there were multiple feeders, it was difficult to accurately identify the number and levels of all feeders vessels.

Conclusions: TRICKS sequence is useful in identifying the approximate location of feeding vessels in suspected spinal vascular malformations. This allows shorter angiography procedures with reduced contrast load and radiation exposure. It may have a role in diagnosing low flow fistulae that are otherwise occult on MRI and assessing surgical success post-operatively.

F2-01: Body CT is non contributory and delays diagnosis in patients with GBM

A. A. Qureshi, M. R. Bailey & I. D. Kamaly-Asl (The Greater Manchester Neuroscience Centre, Salford, UK)

Objectives: To determine whether unnecessary scans were performed on patients with a radiological diagnosis of GBM and to compare radiological diagnosis with histological diagnosis.

Design: We retrospectively reviewed the minutes of 51 neuro-oncology MDT meetings from 2008 and 2009. Data was collected on all new referrals with radiological diagnoses of malignant primary brain tumours or brain metastases. A regional radiology database was used to identify those patients who underwent CT scans of the chest, abdomen and pelvis. Histological diagnoses were obtained for patients that underwent surgery. Microsoft excel was used for data storage and analysis.

Results: A total of 265 patients were identified. 33% of patients with a radiological diagnosis of GBM underwent staging body CTs. 94% of body scans in GBM patients revealed no additional relevant pathology. 89% of body CTs in patients with brain metastases demonstrated significant pathology. When compared to histological diagnosis the accuracy of radiological diagnosis for GBM was 92%, metastatic disease was 83%, and for primary brain tumours as a whole was 97.2%. Times from diagnosis to MDT discussion for GBM patients were significantly lower in those that did not have a body CT (average of 9.7 days for those with CT and 6.5 days for those without, t test p = 0.046).

Conclusions: The ability of our neuro-oncology MDT to correctly differentiate primary from metastatic brain tumours on radiology alone is high. Despite this, many patients with primary brain tumours undergo unnecessary staging CT scans that contribute little to their management and significantly delay their discussion in the MDT. More robust triggers for these investigations should be identified and addressed.

F2-02: Glioblastoma: Can we predict a patient's survival?

C. J. Parks, J. Law, C. H. Davis & G. Hall (Royal Preston Hospital, Preston, UK)

Objectives: In 2008, subanalysis of the Stupp trial data was performed to identify new prognostic factors for survival in patients with glioblastoma multiforme (GBM). An online prognosis predicting “GBM Calculator” was made available on the EORTC website intended for use when counselling patients and to give neuro-oncologists guidance when giving individual patients' prognoses. We tested the calculator against actual local survival data in order to validate its use in our patients.

Methods: All (n = 85) local patients diagnosed with WHO grade IV Glioma in 2006 and 2007 and who subsequently died had their details entered retrospectively into the GBM Calculator. The survival from date of diagnosis was calculated and compared to the median predicted survival. Patients were excluded if they could not be stratified on the calculator. The actual and predicted survivals were compared with a Chi square analysis and displayed on Kaplan-Meier curves with 95% confidence intervals.

Results: Within the first year following diagnosis the predicted median survival was significantly in excess of the actual survival (p < 0.0001). Some local patients survived in excess of two years despite the most optimistic predictions of 16.31 months. Almost 90% of patients would have received a significantly over optimistic prediction of survival if this calculator was used during this period.

Conclusions: Use of the GBM calculator would have misinformed the majority of our patients. This study would suggest that a unit contemplating the use of this potentially helpful tool should undertake a similar validation to insure that the information given to patients is accurate.

F2-03: Co-morbidity is the strongest pre-treatment prognostic factor for survival of patients with glioblastoma multiforme

V. Petrik, M. C. Papadopoulos & B. A. Bell (Academic Neurosurgery Unit, St.George's University of London, London, UK)

Objective: Patient's age and clinical performance status are common entry criteria for glioma trials and are widely accepted as the best predictors of survival in patients with glioblastoma multiforme (GBM). The aim of our study was to assess the effect of co-morbidity on overall survival of GBM.

Design: Prospective observational study.

Subjects: 91 patients (58 men, 33 women) consented for the study with an average age of 59.5 ± 1.0 (mean ± standard deviation) years, 63.7% had Karnofsky score = 70, 67% underwent craniotomy and excision of their tumour, 56% underwent postoperative radiotherapy and 19.8% chemotherapy.

Outcome measures: Co-morbidity was measured by the Charlson Index calculated by adding assigned weights for 19 co-morbidity conditions. Survival time was expressed in weeks from the time of diagnostic surgery and a Cox proportional hazard model was used for survival analysis.

Results: Overall median survival was 30 weeks. 53.8% of patients had Charlson index 2 indicating no associated co-morbidity. Age, Karnofsky score, Charlson index, extent of surgery and adjuvant oncological treatment were all strong predictors (p < 0.01) of survival in univariate analysis. In multivariate analysis of pre-treatment factors Charlson index was the strongest predictor of survival with hazard ratio 1.3806 (95%CI 1.0669 to 1.7865, p = 0.0147), followed by age and Karnofsky score with hazard ratios 1.0301 (95%CI 1.0056 to 1.0552, p = 0.0163) and 0.9949 (95%CI 0.9718 to 1.0107, p = 0.6744) respectively.

Conclusions: Assessment of co-morbidity should be incorporated into further glioma trials. The Charlson Index is a simple tool that can aid GBM management decisions in MDT settings.

F2-04: Serum Albumin Predicts Survival In Glioblastoma Multiforme

N. Borg, M. R. Guilfoyle & S. Thomson (Addenbrooke's Hospital, Cambridge, UK)

Objective: Serum albumin has been shown to correlate with outcome in several cancers, but whether it is a useful prognostic marker in glioblastoma multiforme (GBM) is unclear. We examined if preoperative serum albumin predicted survival in patients with GBM.

Methods: 685 patients with histologically proven GBM were identified from the regional cancer registry. Patient demographics, treatment details, and most recent preoperative blood results were collected. Serum albumin (normal range 30-50 g/L) was available for 549 patients and classified as low (<30 g/L, n = 82), lower-normal (30–40 g/L, n = 371) or upper-normal (40–50 g/L, n = 96). Survival was estimated using the Kaplan-Meier method and compared with the Log-rank test. Multivariate analysis was performed using a Cox regression model.

Results: Albumin levels were normally distributed with range 15–49 g/L and mean 34.6 g/L. Median survival was significantly shorter in patients with low serum albumin compared with normal (2.3 vs. 5.6 months, p < 0.001) and patients with serum albumin in the lower-normal range had significantly shorter median survival than those in the upper-normal range (5.1 vs. 8.8 months, p < 0.001). Of patients receiving adjuvant radiotherapy (n = 263/549) those with upper normal serum albumin (n = 78) had significantly longer median survival (11.0 vs. 7.2 months, p < 0.001). Multivariate Cox regression showed that age (Hazard Ratio 1.025, p < 0.001), debulking surgery (HR 0.72, p < 0.001), radiotherapy (HR 0.46, p < 0.001), chemotherapy (HR 0.60, p < 0.001), and serum albumin (HR 0.97, p < 0.01) were independent predictors of survival.

Conclusion: Serum albumin level appears to be an independent predictor of response to treatment and overall survival in patients with GBM.

F2-05: Boron Neutron Capture Therapy for newly diagnosed glioblastoma multiforme: An assessment of clinical potential

J. W. Hopewell¹, T. Gorlia², L. Pellettieri³, V. Giusti ⁴, B. H. Stenstam⁵ & K. Skald ⁶ (Green Templeton College, University of Oxford, UK¹, EORTC, Brussels, Beligium², Department of Neurosurgery, Gateborg, Sweden³, Department of Mechanical, Nuclear and Production Engineering, Pisa, Italy⁴, Nykaping Hospital, County of Sormland, Sweden⁵ and Hammercap Medical AB, Stockholm, Sweden⁶)

Objective: To assess the potential of Boron Neutron Capture Therapy (BNCT) with a 6 hour infusion of the boron carrier l-boronophenylalanine, as a fructose preparation (BPA-f), as first line therapy for newly diagnosed glioblastoma multiforme (GBM).

Methods: Patient survival data from a phase II study with BNCT was compared with data from two arms of a phase III study with conventional radiotherapy (RT) in the reference arm and RT plus concomitant and adjuvant temozolomide (TMZ) in the experimental arm. BNCT was also compared with small subgroups of these patients for which the methylation status of the MGMT (O6-methylguanine–DNA methyltransferase) DNA-repair gene was known. Differences in baseline characteristics, extent of salvage therapy and levels of severe adverse events were also considered in these comparisons.

Results: The results indicated that BNCT could offer a more effective treatment than RT alone and that it could be comparable with that for RT plus TMZ. When compared with a subgroup of patients with the unmethylated MGMT DNA-repair gene, that received RT/TMZ, a possible clinically relevant advantage of BNCT was suggested.

Conclusions: BNCT could offer prolonged survival for a proportion of patients with newly diagnosed GBM compared with the present standard treatment. BNCT is a single day treatment with mild side effects, which would offer an initial 6 weeks of good quality life, during the time when patients would otherwise receive daily treatments with RT and TMZ. It is suggested that the use of BNCT, with six hour infusion of BPA-f, should be explored in a stratifies randomize phase II trial in which patients with unmethylated MGMT DNA-repair gene are offered BNCT in the experimental arm with RT plus TMZ as the reference arm.

F2-06: “Long term survival” of glioma patients with “salvage” intra-arterial cisplatinum chemotherapy

E. Lekka & C. H. Davis (Royal Preston Hospital, Preston, UK)

Objective: We report the survival of 27 glioma patients who received “salvage” intra-arterial cisplatinum following recurrence after having received all possible treatment currently available.

Methods: We identified 27 patients during the period of January 1998 until April 2008 who received intra-arterial cisplatinum and retrospectively collected clinical and treatment data. Subgroups were formed according to initial histological grade and progression.

Results: A mean of 3.3 treatments were performed (range of 1–7). The overall survival was 50 months (range of 14–142). Survival from the first intra-arterial cisplatin therapy ranged from 1–19 months with a mean of 6.1. Eighteen out of the 27 patients survived more than 3 months following the start of administration of treatment. (N. B. 2 patients are still alive after 19 months each). The mean survival for patients with Glioblastoma Multiforme (GBM) was 5 months (range 2–9). Seven out of 9 patients lived more than 3 months. The mean survival of patients with initial grade 3 progressing to grade 4 was 8.3 months (range 2–19 months, 2 pts still alive). Seven out of the 10 patients lived more than 3 months. Grade 2 progressing to grade 4 patients had a mean survival of 4.75 months (range 1–17) with 4 out of 8 patients living more than 3 months.

Conclusions: Our 27 patients had a mean survival of 6.1 months after administration of cisplatinum chemotherapy. The treatment, although invasive, is very well tolerated (cisplatinum has little haematological toxicity and is suitable for patients pre-treated with chemotherapy). Although intra-arterial cisplatinum chemotherapy is rarely used nowadays for glioma patients, our results are encouraging and warrant further investigation on a larger scale, probably with chemosensitivity testing.

F2-07: The use of Gliadel® wafers in recurrent high grade glioma

C. E. Uff, D. I. McGregor, S. Levy, R. Bradford & L. W. Thorne (Royal Free Hospital, London, UK)

Introduction: Recent literature demonstrates the safety of Gliadel and TMZ and NICE guidelines (September 2007) approved radiotherapy (RT) and concomitant, and adjuvant, temozolamide (TMZ), and RT and Gliadel for primary HGG. It was not clear as to whether these were mutually exclusive, but it is commonly perceived to be the case and in consequence, all Gliadel procedures at our centre are carried out in patients with recurrent HGG after RT, however this is associated with a high complication rate¹. Due to lack of data, there are currently no NICE guidelines for Gliadel in recurrent HGG.

Methods: Retrospective case note review for audit.

Results: 30 patients with a diagnosis of recurrent HGG received Gliadel between January 2004 and January 2009. All had previously undergone radiotherapy. Mean age at diagnosis was 50 (25–66). Mean time from diagnosis to recurrence was 85 weeks (12–330) and mean post Gliadel survival 61 weeks (0–260, three patients alive to date) with 12 month survival at 37%, however 26% suffered complications. There was no relationship between longevity or high KPS, and age at diagnosis.

Conclusions: Compared to a recent trial of 32 recurrent HGG patients treated without Gliadel with 12 month survival at 26%², our results demonstrate clear survival advantage. We suggest that the increased complication rate is off-set by the increased survival.

F3-01: Analysis of multiple serum angiogenic factors in patients with glioblastoma suggests that TIMP-1 level predicts survival

M. Crocker, B. A. Bell, M. C. Papadopoulos & S. Zacharoulis (St George's University of London, London UK and The Royal Marsden Hospital, London, UK)

Objective: Angiogenesis is a hallmark of glioblastoma and is considered an important target for new chemotherapy regimes. The profiling of patients, tumours and treatment monitoring is therefore vital to the potential success of anti-angiogenic therapy. We investigated the hypothesis that serum angiogenic protein profiles show distinct patterns that are prognostic for survival.

Design: Laboratory investigation of samples from the tumour serum bank at St George's University of London.

Subjects: A cohort of 36 patients with histologically proven glioblastoma who had serum samples taken at the time of diagnosis. Information about the patients included basic demographics, performance status at time of diagnosis, surgical and adjuvant treatment, and survival time.

Methods: Samples were simultaneously assayed using protein arrays for 48 factors implicated in glioblastoma angiogenesis. These data were then subjected to statistical analysis including cluster and multivariate analysis.

Results: Cluster analysis was used to group patients according to similarities of their serum angiogenic profile. This produced reliable clusters with median survivals of 48 and 32 weeks respectively, although this did not reach statistical significance (p = 0.10). Multivariate analysis showed that established prognostic factors of young age, better presenting performance status, extent of resection and adjuvant therapy conferred significant survival advantage. The only serum protein independently associated with improved survival is Tissue Inhibitor of Metalloproteinase 1 (TIMP-1), with significance 0.01, lower levels of TIMP-1 predicted improved survival.

Conclusions: Serum analysis of angiogenic factors may be valuable as a prognostic marker in glioblastoma and this preliminary study reveals a possible role of TIMP-1 in this context.

F3-02: A feasibility study of Neighbourhood Tractography in the presence of paediatric brain tumours

C. S. Hill, J. Clayden, N. Kitchen, J. Bull, W. Harkness & C. Clark (UCL Institute of Child Health and Institute of Neurology, National Hospital for Neurology and Neurosurgery, London)

Objective: The aims of this study were to determine if the method of Neighbourhood Tractography (NT) is effective in producing true representations of white-matter tracts in the presence of distorting intracranial tumours in a paediatric cohort. This is the first work that has assessed the feasibility and accuracy of this technique in the presence of neoplasms. We also looked at how the process can be simplified and automated for use by neurosurgeons without prior tractography experience.

Background: Diffusion tensor imaging (DTI) is a technique that measures the directional dependence of water diffusion in tissues to track white matter bundles in vivo. NT is a novel technique of using candidate seed points (classically in a 7 × 7 voxel region) and matching the resulting candidate tracts to a reference pathway. The technique does not rely on ‘regions of interest’ and so avoids the associated reproducibility issues.

Methods: This study applied the NT technique to DTI data from a cohort of 17 patients with a variety of tumours. Graphical representations of the corpus callosum, corticospinal, cingulate and arcuate tracts were produced along with measures of structural integrity (fractional anisotropy, mean diffusivity) and goodness of fit.

Results and conclusion: Reconstruction of major white matter tracts were produced for all patients studied. Results showed significant differences in tract morphology depending on tumour characteristics and tract/site involvement when analysed with comparative statistical measures. Additionally, the presentation includes brief advice on how neurosurgeons can set up a NT system in their own department (or even on a home computer) and cautions regarding interpretation of tractography results.

F3-03: High Field Intraoperative MRI (1.5T) in Neurosurgical Operations – early single Centre Experience

N. D. Kitchen, M. S. A. Dherijha, M. Powell, J. Grieve, A. W. McEvoy, S. R. Wilson, M. Wilson, C. Micallef & T. Yousry (The National Hospital for Neurology and Neurosurgery, London, UK)

Objective: To analyse the contribution of high field (1.5T) intra-operative MRI in Neurosurgical procedures and share our early experience of its use in neurosurgical procedures.

Methods: Retrospective analysis of data of all operative cases performed in IMRI suite during the first six months. 17 patients had surgery in IMRI suite; 9 Brain Tumour, 2 Pituitary adenoma and 6 epilepsy patients. We divided surgical procedures in to three time periods; Anaesthetic, Surgical and Scanning time. All time periods were recorded.

Results: Total 28 MR scans were performed. 8 Scans were performed prior to surgery and 20 scans were performed intraoperatively in IMRI suite. The average scanning duration was 28 ± 7 mins, Average surgical time varies depending on the surgery. Average time for pituitary surgery was 40 ± 5 mins, brain tumour 120 ± 15 mins and epilepsy surgery was 240 ± 40mins. Intraoperative scans were done once surgeon was satisfied with their resections single intraoperative MR scans were performed in tumour and pituitary cases but in epilepsy surgery two intraoperative scans were done. No further resection was done in tumour and pituitary cases but in epilepsy surgery hippocampectomy was performed following intraoperative scans.

Discussion: IMRI is helpful in planning of neurosurgical procedures, however it is time consuming and primarily determined by imaging protocol.

Conclusion: Our initial experience demonstrates its useful modality to plan and teach neurosurgery. Patient outcome benefits cannot be known until prospective studies are conducted.

F3-04: Auditory Brainstem Implants, a National Centre's experience

C. Moran, I. Yousef, K. Walsh, K. Tambirajoo & D. Rawluk (Beaumont Neurosurgery, Beaumont Hospital, Dublin 9, Ireland)

Objective: To assess the functional outcome and impact on quality of life in neurofibromatosis type 2 (NF2) patients following insertion of an Auditory Brainstem Implant (ABI).

Methods: A retrospective cohort study. Subjects: Between 2002 and 2009, 9 patients, out of a cohort of 23 NF 2 patients (mean 25; range 11–43 yrs) were implanted with 24-electrode Nucleus device. Patients were selected for surgery based on disease status, personal motivation and degree of hearing loss. Intraoperative monitoring confirmed electrode contact ranging from a minimum 10 to a maximum 19, at the time of switch on electrode contact ranged from 4 to 19.

Results:All patients used their ABI for 10- 12 hours per day and continued use up to study date. This follow up ranged from 7 years to one month. Lip reading using implant showed significant improvement for single word assessment at one year post switch on. In 72% of patients, their perceived outcome of lip reading ‘met or exceeded’ their expectations, 80% of recipients had quality of life enhanced ‘significantly’ or ‘enormously’. In hearing environmental sounds this outcome met expectations in 29% of patients, and exceeded expectations in 43%. Results were obtained during routine outpatient follow up and the quality of life survey was completed at home by all recipients.

Conclusions: The benefit of ABI insertion for patients with NF2 was clearly demonstrated, their quality of life improved due to both enhanced ability to lip read and perception of environmental sounds. It is the unique state of sensorineural deafness that makes measuring an improvement in it difficult. Objective measurements are helpful but of even more relevance is the patients own perceptions, which this study attempts to explore.

F3-05: Radiotherapy for Atypical Meningiomas

R. J. Mair & T. A. Carroll (Royal Hallamshire Hospital, Sheffield, UK)

Introduction: Historically, adjuvant radiotherapy has been inconsistently used following surgical resection of WHO grade II/atypical meningiomas. With increased use of interval MR imaging and also access to stereotactic radiosurgery for growing tumour remnants/recurrences, there is a need to further clarify the role of such adjuvant radiotherapy for WHO grade II meningiomas.

Methods: We exploited the variable pattern of referral following surgery re adjuvant radiotherapy by individual consultant neurosurgeons within a large teaching hospital to explore any demonstrated benefit of such radiotherapy. We searched the neuropathology tumour database and identified the case notes of 78 patients with a histological diagnosis of WHO grade II/atypical meningioma.

Results:The mean age was 56 (range 19 to 84), with mean follow-up of 43 months (range 6 to 88 months). Forty three per cent of patients were female. A tumour remnant was left in 36% of patients. Thirty nine per cent overall underwent post-operative radiotherapy. As anticipated, there was a significant benefit with respect to disease recurrence/progression from surgery that did not result in leaving a tumour remnant (p < 0.005). The use of radiotherapy however demonstrated no significant difference in terms of reducing subsequent disease recurrence/progression.

Conclusions: We suggest that the routine use of radiotherapy in all operated patients with WHO II meningiomas should not occur. It will likely require larger numbers to clarify any overall benefit in patients with tumour remnants undergoing radiotherapy. It may be that radiosurgery to tumour remnants may be a more appropriate course in such patients.

F3-06: Radiosurgery Achieves Long Term Control of Petroclival Meningiomas

T. Flannery, H. Kano, L. D. Lunsford, S. Sirin, J. C. Flickinger, J. Gray & D. Kondziolka (Royal Hospitals Belfast, Northern Ireland and Center for Image-Guided Neurosurgery, University of Pittsburgh, USA)

Objective: Because of their critical location adjacent to brain, cranial nerve, and vascular structures, petroclival meningiomas remain a clinical challenge. We evaluated outcomes in 168 petroclival meningioma patients who underwent Gamma Knife radiosurgery (GKRS) during a 21-year interval.

Methods: GKRS was used either as the primary or adjuvant management for 168 petroclival meningiomas involving the region between the petrous apex and the upper two-thirds of the clivus. The most common presenting symptoms were trigeminal nerve dysfunction, balance problems, diplopia and hearing loss. The median tumor volume was 6.1 cm³ (range, 0.3–32.5 cm³) and the median margin dose was 13 Gy (range, 9–18 Gy).

Results: During the median follow-up interval of 72 months, neurological status improved in 44 patients (26%), remained stable in 98 patients (58%), and worsened in 26 patients (16%). Tumor volume decreased in 78 patients (46%), remained stable in 74 patients (44%), and increased in 16 patients (10%), all of whom underwent additional management. Eight patients had repeat radiosurgery, four underwent delayed resection, and four had fractionated radiation therapy. CSF diversion was performed in 7 patients (4%). Significant risk factors for tumor progression were tumor volume = 8 cm³, (p = 0.001) and male gender (p = 0.02).

Conclusions:In this 21-year experience, GKRS for petroclival meningiomas achieved freedom from the need for initial or further surgical resection in 98% of patients and was associated with a low risk of adverse radiation effects. We believe that radiosurgery should be the preferred initial option for patients with smaller-volume, symptomatic petroclival meningiomas.

F4-01: Traumatic brain injury admissions in a regional trauma centre from 2002 to 2009

D. Baxter, S. M. Joshi & N. Tai (The Royal London Hospital, Whitechapel, London, E1 1BB, UK)

Objective: To study the trends of adult traumatic brain injury (TBI) admissions to a trauma centre over 7 years.

Method: Admission details, TBI severity and GCS were obtained from the trauma registry for all patients admitted in the hospital between August 2002 to March 2009. Data was analysed using SPSS.

Results: 1554 TBI were recorded from the trauma data collection of 6037. 79.2% were male (mean age: 39.7 years). The majority of admissions at 24.9% were in the 2nd decade. Most admissions took place during weekend nights. There was increase in TBI incidence in all age ranges of patients, and an increasing trend in the number of patients presenting with all severities of TBI. The proportion of TBI in the severe, moderate and mild groups were 30.9%, 11.6% and 38.8% respectively (18.5% not recorded). The three commonest mechanisms of head injury were RTAs 48%, falls 27% and assaults 14%. There was a statistically significant difference in worse outcomes in presence of other injuries between the severe and mild TBI groups in all age groups

Conclusions: In line with the department's transformation into a regional trauma centre, this study reports a sustained increase across the seven-year period in the number of TBI admissions. These increases occurred across all age groups, among the male and female patient populations, and involved all severities and mechanisms. Our data can serve as comparisons for and guidance to new and developing regional trauma centres across the UK.

F4-02: Traumatic Brain Injury, it's changing demographics over six years

C. Moran, R. K. J. Murphy, A. S. Semanat, L. McEvoy, C. Ruane, M. O. Kelleher, C. Bolger & J. Phillips (Beaumont Neurosurgery, Beaumont Hospital, Dublin 9, Ireland)

Objectives: To compare current data collected on traumatic brain injury (TBI) with data published in The Phillips Report collated in 2003 and determine new trends in causes and outcomes of TBI. A prospective study.

Design and Subjects: The National Report on Traumatic Brain Injury (Phillips Report) was published 2008, this seminal study was commissioned to plan future TBI service development in the Ireland. Our audit compared a twelve week snapshot of patients from 2009 with that same time period in the Phillips report (2003).

Methods: All cases of traumatic brain injuries logged with Neurosurgical Advice Service (NAS) and admitted to Neurosurgical Unit (NSU) were included in the study. Demographic data such as age, gender, nationality and admission date were documented. ICD 10 codes for reporting external morbidity was used to assess the causal factors in an event of injury.

Results: Over 250 cases from each period were compared. Changes were recorded in the manner of injury. While fall remains to be the most common cause of TBI, the percentage has decreased from 60% (2003) to 49% (2009). TBI caused by assault showed the most significant increase from 20% (2003) to 29% (2009). Changes in age profile were also recorded. The median age at time of registration was 45 years (range <1 to 94 years) in 2003 but fell to 26 years (range <1 to 91 years) in 2009. Four cases of assault on women resulting in NITU admission was also reported, which represent an increase from 0% (2003).

Conclusions: More people are being admitted, surviving high velocity road traffic accidents, violence against women has come to the fore while alcohol and falls remain constant. Irish society continues to change reflected in the pattern of traumatic brain injury sustained by our population.

F4-03: Trauma network standards for referral of head injuries-can they achieved?

D. Holliman¹, C. Dubois¹, F. Hassan¹, C. Barrett¹, Y. Yoongkhong¹, P. Panaretos¹, A. Paluzzi¹, D. Bramley² & P. Mitchell¹ (Department of Neurosurgery, Newcastle General Hospital, Newcastle, UK¹ and Department of Accident and Emergency Medicine, Sunderland Royal Hospital, Sunderland, UK²)

Objective: Trauma networks have been developed in response to recommendations from the NCEPOD ‘Trauma:Who Cares?’ report (2007) regarding regional planning for the effective delivery of trauma services. Adherence to local network guidelines regarding standards for emergency department (ED) head injury management and referral for neurosurgical opinion was audited. Standards included time to head CT from arrival in ED (60 mins) and time for neurosurgical opinion (30 mins).

Methods:Head injury referrals to the neurosurgical registrar were audited for six weeks while a concurrent audit took place at referring EDs. Analysis of adherence to recommended standards was undertaken using completed referral proformas placed in on-call referral books and proformas completed in EDs.

Results: Referrals occurred on 36 out of 40 days (90%) resulting in data for 70 patients. During this period 10 proformas were completed by referring EDs. Average time to performing CT head after ED arrival was 51 mins (range 13–120 mins) while mean time to neurosurgical referral after CT was 43 mins (range 5–100mins). Definitive neurosurgical opinion was given in an average of 16 mins (range 1–90, median 9 mins). 86% (n = 60) of opinions were given within 30 mins. Imaging was not available for review in 8 out of the 10 referrals not opined upon within 30mins.

Conclusions: This study suggests that the majority of referrals can be imaged and opined upon within the pre-defined time frame. Delays usually involve obtaining CTs in referring EDs and difficulties with accessing imaging by neurosurgeons. It remains questionable whether it is necessary to opine on GCS 14–15 patients within 30 mins.

F4-04: Audit on the management of subdural haematomas in a neurosurgical unit: When is local conservative management a safe practice?

N. Haliasos, G. Prezerakos, S. Bavetta & H. L. Low (Queens Hospital, Romford, UK)

Objective: The notion of tertiary referring centre implies that a number of patients with acute or chronic subdural haematomas will be managed locally, to achieve effective resource allocation. Unfortunately there is a subgroup who will deteriorate and will need urgent transfer and surgical intervention. Our aim was to identify risk factors in this subgroup,that could refine and optimize our everyday decision process.

Methods: Retrospective record analysis of 339 patients events with subdural collections referred to our unit between January 2008–May 2009. We identified three groups. Patients managed conservatively in the local hospitals (I), patients admitted and operated as planned (II) and the group of patients who initially were managed conservatively but deteriorated acutely and required urgent transfer and surgical intervention (III).

Results: In 306 (90.2%), the initial management plan was not changed (groups I, II). Nevertheless, group III accounted for 33 patients (9.7%) Anticoagulation/antiplatelet treatment (p < 0.026), younger age (p < 0.046), were significant predictors in favor of surgical management. GCS and midline shift on first referral, unilateral or bilateral haematoma were not found to be significant.

Conclusions: The well established decision process based on the radiological findings and GCS, though is successful in providing safe management in 90.2% of cases, it does prove to be insufficient for a subgroup of patients, rendering them prone to increased mortality/morbidity risks. Measures like lowering the threshold for neurosurgical admission, or local frequent imaging might be necessary for these patients. Implications on the additional workload versus safe management balance are discussed.

F4-05: Infrascanner in Head Inury patients: Preliminary Experience

V. Mudigonda-Rao, C. K. Patel, J. L. Lang & R. Nannapaneni (University Hospital of Wales, Cardiff, UK)

Introduction: Infrascanner is a near infrared (NIR) based, non-invasive, portable device to detect intracranial hematomas. We are the first unit in the UK to trial it.

Design: Prospective, double blind study in a tertiary neurosurgical unit to evaluate and correlate Infrascanner findings vs. CT scans in head injury patients.

Method: Patients admitted to our hospital with head injury (GCS 13–15) requiring a CT scan were studied. Children (<16), ventilated patients, aggressive patients, those with large scalp hematomas or lacerations were excluded. The Infrascan operator and the neuroradiologist reporting the CT were both blinded. Data was collated and subjected to statistical analysis.

Results:66 patients (M: F 52:14, age range 18–94 years) were included since January 2009 (32 – assaults (49%), 21 – falls (31%), 8 – RTA (21%), 5 – other causes). All patients had a CT scan of the head and an Infrascan within 24 hours following admission. CT scan showed various types of hematomas in 25/66 patients (prevalence 37.8%). The Infrascanner detected the hematoma in 19/25 patients (sensitivity of 76%). Subsequent analysis showed all the undetected hematomas to be outside its detection range (i.e >2.5 cm from brain surface & < 3.5cc volume). CT scans were negative for hematomas in 41/66 patients; the Infrascanner correlated with this in 34/41 patients (specificity of 82.9%). The calculated PPV & NPV were 75% and 85.7% respectively.

Conclusion: The Infrascanner is envisaged as a screening adjunct in pre-hospital and acute clinical settings. Findings in our study compare well with published literature.

F4-06: Early decompressive craniectomy for traumatic brain injury may lead to a shorter stay in intensive care: A pilot study at King's College Hospital

N. K. Mazarakis, A. K. Demetriades, C. Jenkins, C. Pahl & C. Tolias (Department of Neurosurgery and Critical Care, King's College Hospital, London, UK)

Objective: The effectiveness of decompressive craniectomy (DC) is yet to be proven. DC has the advantage of lowering ICP, but can have serious complications. The aim of this pilot study was to evaluate the outcome of patients undergoing DC for raised intractable ICP following head injury at King's College Hospital.

Methods: From the patients recently admitted to intensive care following traumatic brain injury we reviewed 10 consecutive patients who underwent DC for raised intractable ICP. We excluded all patients who required initial evacuation for intracranial haematomas, as well as those whose records were incomplete or whose initial scans were not available.

Results: In those 10 patients we found that mean ICP dropped from 34 ± 2.8 preoperatively to 21.1 ± 1.1 mmHg postoperatively. Early DC (<24 hours) correlated significantly with a reduced length of stay in ICU (rpb = 0.72, p = 0.019). However, there was no correlation between the time of DC and GOS. While younger patients (<40 years old) tended to have marginally lower ICP values postoperatively (r = 0.67, p = 0.05), there was no correlation between the age of the patient and the timing of DC. There was no correlation between age and GOS (p > 0.05). Lower values of ICP following DC were associated with higher GOS at an average of 20.5 months after the operation, however, this did not reach statistical significance (rpb = 0.65, p = 0.06). Finally, there was no correlation between GCS prior to surgery and GOS post-operatively (p > 0.05).

Conclusions: While the results of the present pilot study are limited by the sample size, the reduced stay in ICU after early DC might have important clinical and financial implications. These preliminary results encourage further investigations with a larger sample.

F5-01: The natural history of cervical myelopathy

R. Murphy, I. Molinda, D. Kealy, L. McEvoy, M. Kelliher & C. Bolger (Department of Neurosurgery, Beaumont Hospital, Dublin, Ireland)

Introduction: To document the natural history of cervical myelopathy in patients who were placed on a neurosurgical waiting list for an anterior cervical discectomy. Long neurosurgical waiting lists meant that patients had to wait months to years for surgery.

Methods: Retrospective observational study was carried out assessing Myelopathy Index, SF-36, Visual Analogue Scores, and Neck Disability scores.

Results: Sixty-four patients who were waiting for 3 months or more for anterior cervical surgery were selected from the waiting list at a University based spinal neurosurgical unit. There were 37 females and 27 males, median age of 45 years (range 31–85). Mean time that the patients were on the waiting list was 20 months. (Range 3–60 months). 44 (69%) reported a general decline in their overall condition, 19 (30%) reported no change while 1 improved. 55 (86%) patients stated that pain was their principal complaint. Deterioration in arm function affected 45 patients. 20 patients reported deterioration in walking. 34 (53%) experienced worsening of sensory abnormalities in the upper limbs, such as numbness, tingling sensations, or pins and needles. Myelopathy index remained static 23 to 24. Neck disability increased on average 37% to 43%. SF 36 physical health scores deteriorated markedly; physical functioning 52% to 45%, Role limitations (physical) 39% to 35%. SF 36 Mental health parameters worsened; emotional 39% to 35%, vitality 36% to 32%, mental health 59% to 55%, social functioning 48% to 44%.

Conclusion: The majority of patients with cervical myelopathy tend to deteriorate subjectively and objectively while they await surgical intervention. Myelopathy indices remained broadly similar while there was a marked deterioration in patient's mental and physical health.

F5-02: Cancellous bovine bone substitution material like Tutobone is an efficient alternative to autologous bone in anterior cervical fusion surgery

S. K. Prakash, N. Mukerji, D. Bica & F. P. Nath (James Cook University Hospital, Middlesbrough, UK)

Objectives: This study was undertaken to assess the post operative radiological fusion rates in patients who had anterior cervical surgery using either autologous bone or ‘Tutobone’.

Methods: Case note review and review of post op cervical spine X rays done at two subsequent follow ups. This was a retrospective, observational study. Data were derived from a database of all anterior cervical surgeries performed at our centre over the past 10 years. Exclusion criteria included incomplete notes, undocumented diagnoses, complex cervical surgery, only discectomy with no fusion implant.

Results: An analysis of 152 patients who had anterior cervical discectomy and fusion with either tutobone (n = 92) or autologous bone (n = 60) is presented. These groups were well matched in terms of age, sex, preoperative symptoms and duration of follow up. Significantly greater number of patients in whom autologous bone was used had more than one level surgery (?2, p = 0.01). The median follow-up time was 1 year. The type of implant, age and the number of levels operated had no significant impact on the fusion and the median time to fusion was no different in the two groups (12 vs 10 months) on a survival model (Log rank test, p = 0.24).

Conclusion: From our data there is no evidence to suggest that autologous bone leads to significantly better or faster fusion and alternatives such as “Tutobone” may be used with similar results.

F5-03: Preoperative balloon embolisation of the Vertebral artery as an adjunct to resection of cervical spinal tumours

T. A. Elias, J. Madigan, F. Robertson, S. Brew & A. T. H. Casey (National Hospital for Neurology and Neurosurgery, London, UK)

Objective: Radical resection of extradural spinal tumours improves prognosis. The authors report on the management of extradural spinal tumours in the cervical region with preoperative angiography, balloon embolisation of the vertebral artery, and particle embolisation of the vascular supply to the tumour.

Methods: Patients with extradural tumours in the cervical spine, who underwent preoperative embolisation of the vertebral artery with detachable balloons, and selective particle embolisation of the vascular supply to the tumour, were included in this study. Surgical results, survival and procedural complications were evaluated.

Results: 10 patients with large or recurrent tumours in the cervical spine underwent radical excision after balloon embolisation of the vertebral artery between 2002 and 2008. The tumours included 4 chondrosarcomas, 1 chordoma, 1 each of haemangiopericytoma, aneurysmal bone cyst, malignant peripheral nerve sheath tumour, Pancoast's tumour and metastatic tumour. 4 patients underwent multiple procedures and all patients needed stabilisation of the spine after tumour resection. The average surgical blood loss after the embolisation was only 450 ml. Only 2 patients had procedure-related complications in the form of transient ischaemia and both resolved spontaneously. 5 patients died due to tumour progression or systemic metastases, at a mean of 18 months after the procedure. 5 patients are still alive, at a mean followup of 49 months.

Conclusions: Radical resection of spinal tumours in the cervical spine can be achieved with the aid of preeoperative vertebral artery embolisation, which is a safe technique with no permanent sequelae; thereby increasing the overall survival.

F5-04: Intramedullary enhancement and swelling of the cervical spinal cord. When is a tumour not a tumour?

B. Nurboja, K. M. David, A. Chaudhuri & D. Choi (The National Hospital for Neurology and Neursurgery, London, UK and Queens Hospital, Romford, UK)

Objective: We present a series of six patients with cervical myelopathy and a swollen cervical cord associated with intramedullary contrast enhancement, in whom it was difficult to decide whether the lesion was due to cervical spondylotic compression, or an intrinsic glial tumour. We describe the key features and suggest a management plan.

Methods: Retrospective review of six cases after 2 years follow-up, with cervical stenosis and MRI findings of high signal cord swelling and contrast enhancement, resembling an intrinsic cord tumour.

Results: All cases presented with cervical myelopathy without acute trauma and MRI revealed swelling of the spinal cord with extensive signal change and contrast enhancement suggesting a possible diagnosis of intrinsic cord tumour. There was coexisting evidence of cervical spondylosis and cord compression. Serological and cerebrospinal fluid analysis excluded demyelinating or inflammatory aetiologies. In all cases cervical decompression was performed without biopsy. All patients improved by a mean two years's follow up. Cord swelling and signal change improved with time, but enhancement persisted.

Conclusion: We present a group of patients with cervical cord swelling and enhancement in whom it was not clear whether the aetiology was one of intrinsic tumour or extrinsic spondylotic compression. We suggest that biopsy of the cord in these cases should be avoided in the first instance, and treatment should involve decompression of the cord and radiological surveillance thereafter. Biopsies should be reserved for those rare cases when decompression alone is insufficient to stop the progression of symptoms or radiological findings.

F5-05: Single Unit Experience in the Surgical Management of Symptomatic Thoracic Disc Protrusion

G. K. Prezerakos, N. Haliasos, J. E. Brecknell & K. M. David (Essex Neurosciences Centre, Queens Hospital, Romford, UK)

Objective: To assess the efficacy of current surgical management and intraoperative strategies of thoracic disc protrusion and to review our experience in Oldchurch/Queen's Hospital.

Methods: A retrospective case study of patients operated for thoracic disc protrusion via the transthoracic or the posterolateral costotransversectomy/transpedicular approach between 2000–2009. Patients case notes and theatre logbooks were used.

Results: Eighteen patients were studied. Mean age was 54 years; mean follow up was eleven months. Average duration of symptoms was 5.35 months. All patients had at least one feature of myelopathy. Radiculopathy was present in 11 cases (61%). Open thoracotomy was employed in ten patients, whilst a posterolateral approach was used in eight cases. Average length of stay was 12.8 days. Complications prolonging the length of stay were present in 4 patients. In all cases, progression of myelopathy was arrested, and significant symptomatic relief was achieved in 17 patients (94%). SSEPs were used in 10 cases, without significantly altering the complication rate, LOS or clinical outcome. In five cases of calcified disc protrusion, excision of the protrusion with resection of the adherent dura was employed. This did not alter the clinical outcome.

Conclusion: The use of transthoracic and posterolateral approaches in the management of thoracic disc protrusion yields very satisfactory results in terms of symptomatic recovery. The use of SSEPs does not significantly improve outcome. Excision of calcified discs with the adherent dura does not result in worse outcome but carries an increased complication rate. Considerations related with such an intraoperative strategy are discussed.

F5-06: Outcome results for the use of XLIF Cage in the management of acute thoraco-lumbar discitis

J. Timothy, K. Flynn, C. Derham, R. O'Kane, Y. Al-Tamimi & A. Rao (The General Infirmary at Leeds, Leeds, UK)

Objective: Management of acute bacterial discitis varies from conservative treatment with antibiotics and bed rest to aggressive debridement anteriorly and posterior fixation. The extreme lateral inter-body fusion (XLIF) technique is minimally invasive and provides direct access to the disc space lending itself to directly removing the source of infection, providing a microbiological diagnosis, and stabilizing the painful segment. It improves the speed of mobilization and reduces the length of hospital stay.

Methods: 7 patients over a 6-month period presented with thoraco-lumbar discitis. 6 patients underwent stand alone XLIF and one had XLIF with posterior fixation. Pre-operative back and leg pain was measured using the Visual Analogue Scale (VAS) and functional disability was measured using the Oswestry Disability Index (ODI). Outcomes were repeated at six weeks post operatively in 6 patients and 6 month post operative data is currently available for 2 patients. Samples of infected disc material were sent off for microbiological investigation in all cases.

Results: Pre-operative mean VAS score was 7.1 (5–8.2) for back pain and 4.3 (0.1–7.1) for leg pain. These had reduced to 3.5 and 1.68 at 6 weeks post operatively. Mean pre-operative ODI score was 64.3% (52–80%) this reduced to 45% (14–74%) at 6 weeks post operatively. In 5 cases, no growth or causative organism was cultured from disc material obtained during the procedure. Mycobacterium tuberculosis complex was isolated from one patient and staphylococcus aureus cultured from another patient. In two cases, bacterial DNA was found to be positive. Of these, Staphylococcus warneri/pasteuri was identified in one case. Patients were discharged back to the referring hospital within 3 days post-operatively.

Conclusions: Stand-alone XLIF is a minimally invasive technique that can be used to treat acute discitis. It allows direct debridement and stabilization of the painful level. It may help prevent post infection kyphosis and has a clinically significant effect on pain and disability. Although there is a low culture rate obtained from tissue samples obtained intra-operatively, this may be explained by the use of antibiotic treatment prior to referral. By allowing early mobilization of these patients after their procedure, hospital stay can be significantly reduced.

F5-07: Prospective evaluation of the “Wallis” lumbar interspinous dynamic stabilization system

R. K. J. Murphy, L. McEvoy, Z. Ali, C. Bolger & S. Young (Department of Neurosurgery, Beaumont Hospital, Dublin, Ireland)

Objective: The “Wallis” implant is indicated to stabilize symptomatic degenerative lumbar spine segments, relieving low back pain related to instability and thus delaying the need for irreversible, more invasive surgical management. The purpose of this study was to provide the first objective clinical evaluation of the “Wallis” lumbar dynamic stabilization system.

Methods: An independent prospective observational study was carried out utilising SF-36, Oswestry Disability Index (ODI) and visual analogue pain scores (VAS). Surgical pathologies in which this technique was used, the intra-operative and post-operative complications and length of post-op stay were recorded.

Results: 102 patients underwent Wallis insertion between June 2007–May 2009, Median age 51.5 (range 28–108). 94% of patients completed questionnaires and were followed up at 3, 6 and 12 month time points. ODI scores decreased from pre-op 39 to 27 at twelve months (p < .0016). VAS back pain scores decreased 59 to 36 (p < 0.0001). Leg scores decreased 50 to 39 (p < 0.0002). SF 36 scores improved significantly, physical functioning 46 to 59, physical health 30 to 54 and social functioning 47 to 68. 50% of patients believed their health to be better 12 months post-op. Pre-operatively 28% of patients were employed and working with 26% off work due to back problems. This rate increased steadily with 42% employed at 12 months. Two implants were removed, one due to non-benefit with subsequent arthrodesis and one due to infection. One superficial wound infection occurred.

Conclusions. The Wallis dynamic stabilization system provides a superficial and easily reversible surgical procedure with a lower complication rate than conventional athrodesis. Used in patients with painful degenerative lumbar conditions their quality of life objectively approached values of the age- and gender-matched general population.

F5-08: Intersegmental Decompression versus Laminectomy in the Surgical treatment of Lumbar Spinal Stenosis

T. Osman, C. Moran, Mo. Kelleher, L. McEvoy & C. Bolger (Beaumont Hospital, Dublin, Ireland)

Background context: Facet preserving intersegmental decompression offers a significantly less morbid alternative than historical laminectomies.

Objective: To evaluate the clinical outcomes of patients who underwent intersegmental decompression for focal lumbar spinal stenosis in comparison to those who underwent laminectomy.

Design: Retrospective Observational study.

Subjects: 62 patients divided into 2 groups, Group A (intersegmental) 37 patients, Group B (laminectomy) 25 patients. Median age 64. Average follow up 38 months.

Methods: Patient selection criteria: Neurogenic claudication or mechanical radiculopathy with no or tolerable low back pain. Outcome Measures: Visual analogue score (VAS) score, walking distance, Oswestry disability index (ODI) and length of hospital stay. Operative data included number of levels decompressed and complications.

Results: VAS improved in both groups postoperatively. Back pain; (A) 7.1 – 2.9: (B) 7.3 – 3.3. Leg pain; (A) 6.8 – 3.2: (B) 8.7 – 3.5. Subgroup analysis of pre and postoperative ODI; (A) 43.7 – 26.8: (B) 50 – 26.3. Postoperative length of stay was significantly shorter in the intersegmental group (A: B = 3.5:5.0 days). In Group A, 26 had single level decompression, 6 patients 3 levels and 5 patients 2 levels. In Group B, 23 single level laminectomy, 1 had 2 levels and 1 had 3 levels. 4 patients required redo decompression, 1 in group A and 3 in group B. Surgical complications occurred in 12: 11 CSF leaks (Group A -5, Group B -6) and 1 wound infection (Group A).

Conclusion: Comparable functional improvements can be achieved with focal intersegmental decompression compared with historical laminectomy.

F5-09: The Use of an Expandable Interbody Device (B-TWIN) For Back Dominant Lumbar Disc Disease

R. A. Collis, M. O.Kelleher, L. McEvoy & C. Bolger (Neurosurgical Unit, Beaumont Hospital, Dublin 9, Ireland)

Introduction: Several surgical options have been utilised to treat patients with back dominant lumbar disc disease. The purpose of our study was to compare the outcomes in patients who underwent lumbar fusion with an expandable interbody device (B-TWIN) using different surgical techniques (PLIF, TLIF or posterolateral screws alone).

Method: Observational study, retrospective analysis of prospectively collected data. Patients underwent a single level lumbar fusion. Group A: PLIF with B-Twin cage; Group B: TLIF with B-Twin cage and unilateral pedicle screw fixation and Group C: bilateral posterolateral screw fixation alone. Functional outcomes were assessed using: SF-36, Oswestry Disability Index (ODI), Distress and Risk Assessment Method scores (DRAM) and the visual analogue pain scores (VAS).

Results: There were 32 patients, 24 female and 8 male. Average age was 45 (range 33–63). Average follow up was 12 months (range 2–36). Level of spinal fusion was 2 L3/4, 11 L4/5 and 13 L5/S1. Mean hospital stay was 5.8 days. VAS improved in all 3 groups A 5.83 – 5; B 8 – 4.83; C 5.71 – 2.3. ODI improved in all 3 groups A 0.5 – 0.35; B 0.51 – 0.44, C 0.42 – 0.16. There was no statistical difference on comparison of the three groups. There were no operative complications. One patient broke her interbody device during a fall in the first post-operative week requiring a subsequent procedure.

Conclusion: Lumbar interbody fusions can safely be performed using an expandable interbody device. Good functional outcomes can be achieved in the majority of well selected patients.