Proceedings of the British stereotactic and functional Neurosurgery group, Glasgow, 30 – 31 October 2006

Click to increase image sizeClick to decrease image size

Image-guided craniotomy and survival in patients with glioblastoma multiforme

N. H. Simms & L. T. Dunn (Department of Neurosurgery, Institute of Neurological Sciences, Glasgow, UK)

Objective: Image guidance systems are well established in clinical practice in the United Kingdom. Objective quantifiable evidence of direct benefit for patients is scarce. We have sought to investigate whether the use of image guidance is associated with improved survival in patients undergoing craniotomy for glioblastoma multiforme in a single institution.

Design: Retrospective review of all patients undergoing craniotomy for a first diagnosis of glioblastoma multiforme between August 1998 and May 2005. Cases were identified from neuropathology records cross-referenced with data from the national cancer registry. Case notes were reviewed and relevant data extracted.

Subjects: Two-hundred-and-one adult patients undergoing craniotomy for glioblastoma. 22 had an image-guided craniotomy.

Outcome measures: The primary outcome measure was survival following surgery (census date 31 December 2005). Secondary outcome measures were length of surgery, duration of surgery and new postoperative neurological deficits. Survival from date of surgery was obtained from national cancer registry and mortality data. Survival curves for patient undergoing image-guided (IG) or standard craniotomies (SC) were compared using the Kaplan – Meier method. Other variables in the two groups were compared using χ² or t-tests as appropriate.

Results: Median survival in the IG group was 335 days compared with 251 days in the SC group (χ² = 0.4722; p = 0.4920). Mean length of surgery was 3.6 h in the IG group and 3.1 in the SC group. Mean postoperative hospital stay was 8.1 days in the IG group and 6.4 days in the SC group (p = 0.0414). New postoperative deficits were identified in five patients in the IG group and 16 patients in the SC group (χ²; NS).

Conclusions:Survival following surgery in the IG group was not significantly greater than that in the SC group, although the number of patients in the IG group was small suggesting that the study may be underpowered to detect a clinically relevant difference. Use of the image guidance system did not lengthen the operative procedure significantly. Postoperative stay in the IG group was longer, but there was no statistically significant difference in the frequency of postoperative complications between the two groups.

The complication rate of intrathecal baclofen pump systems in the adult population. a comparison between multiple sclerosis and spinal injury patients

R. Dubey, C. Mathieson, R. Johnston & M. Fraser (Institute of Neurological Sciences, Southern General Hospital, 1345 Govan Road, Glasgow, UK)

Objectives: This paper discusses the complication rate of intrathecal baclofen pump insertion for two adult populations: multiple sclerosis (MS) and spinal cord injury (SCI) patients.

Design: Retrospective case note review.

Subjects: Forty-eight patients (25 SCI and 23 MS) were included in this study. All patients had a baclofen pump inserted by a single surgeon between 1987 and 2005.

Outcome measures: The number and type of complications were recorded as well as the number of admissions for system assessment.

Results:In total, the MS group required 35 revisional operations compared with 26 in the SCI group. The most common complications were spinal catheter issues followed by infection then reservoir related problems. Seventy per cent of admissions for assessment of system function led to operative intervention. Of the whole cohort, only 35% of patients required no further surgery following implantation during their follow-up period (average 4.5 years).

Conclusions: The majority of patients fitted with an intrathecal pump will require revisional surgery at some stage. This is an important factor when obtaining consent for these patients. The complication rates between these two groups are fairly similar, but the MS patients may be more prone to catheter fracture and disconnection.

Demonstration of activation and somatotopy in the human sensory nociceptive trigeminal system using 3T-based fMRI

C. Hayhurst, S. Basu, P. R. Eldridge, T. R. K. Varma, T. Nurmikko, A. Zaman & N. Roberts (Walton Centre for Neurology and Neurosurgery, Liverpool, UK)

Objective: The anatomical substrate of the neuronal network involved in pain perception is described as the pain matrix and reorganization of the matrix is known to occur in neuropathic pain states. The objective is to demonstrate somatotopy, and prove the hypothesis that reorganization of the pain matrix with fMRI in individual volunteers and patients with neuropathic pain.

Materials and methods:Eleven healthy volunteers underwent fMRI on a 1.5 Tesla platform and data analysed at group level has demonstrated the pain matrix and somatotopy. The experiment has now been transferred to a 3 Tesla scanner to demonstrate this at an individual basis. Eleven volunteers and four patients with facial pain including a case of active trigeminal neuralgia have been scanned and data analysed.

Results:In volunteers at 1.5T group analysis has demonstrated somatotopically organised activations in the brain stem, ventroposterior thalamus and sensory neocortical areas. At 3T not only are the findings replicated, but are proven to be demonstrable on an individual rather than group analysis. The imaging is shown to be practical in individual patients and early analysis of somatopy presented. In addition a single patient was imaged during an attack.

Conclusion: 3T fMRI can demonstrate somatotopically-organized components of the trigeminal sensory system at an individual basis. The translation of 1.5T-based group result to 3T-based individual analysis in volunteers and patients proves the robustness of the methodology for further application to investigate reorganisation of somatotopy in neuropathic pain states.

Brain shift during deep brain stimulation: how to quantify the effect?

E. Campbell & L. T. Dunn (Institute of Neurological Sciences, Southern General Hospital, Glasgow, UK)

Objective: To devise a model to measure the amount of brain shift that occurs during deep brain stimulation.

Design: Observational study.

Subjects: Four patients undergoing deep brain stimulation for Parkinson's Disease or Dystonia.

Methods:Comparison of pre- and postoperative CT scans during deep brain stimulation with the stereotactic frame still applied. Using neuronavigation software, nine targets were selected and their location calculated in a Cartesian system. The degree of displacement for each target, between pre- and postoperative scans was calculated using Pythagoras' theorem.

Results: Deformation of the position of all targets was demonstrated in all patients, but was greatest in superficial cortical targets.

Conclusions: A simple model to estimate brain displacement, but needs to be applied to a greater number of subjects to determine its accuracy.

Long-term requirements for levodopa in Parkinson's patients who had STN stimulation

S. Akmal, A. Taha, Y. C. Gan, R. Mitchell & H. Pall (Department of Neurosurgery, Queen Elizabeth Hospital, Edgbaston, Birmingham, UK)

Objective: To determine whether STN stimulation alters the long-term requirement of levodopa in patients with Parkinson's disease postsurgery.

Design: Retrospective review of patients' notes and clinic letters.

Subjects and methods:Patients who underwent surgery from 1999 to 2004 with the minimum follow-up of 2 years. The levodopa equivalent dose is then calculated for each patient at 1, 2, 3 and 4 years after surgery. The postoperative levodopa doses were then calculated as the percentage of the preoperative levodopa doses. The mean for the dosages and percentages were then obtained for the time periods above.

Outcome measures: Total levodopa equivalent dosage at the time period as stated above, the percentage of the preoperative levodopa dosage after surgery.

Results: Of the 34 patients (27 males and seven female), 16 had 4 years, seven had 3 years and 11 had the minimum 2 years follow-up. The mean levodopa equivalent dosage immediately presurgery was 1864 and this was reduced to 852 at 1 year postsurgery, 752.8 at 2 years, 784.9 at 3 years and 918.7 at 4 years The mean percentage doses were 51.8% at 1 year postsurgery, 50.4% at 2 years, 72% at 3 years and 78.6% at 4 years.

Conclusions: The results showed that the levodopa requirements decrease dramatically after surgery in the first year and slowly rise in the subsequent years after surgery. At 4 years they have not reached the preoperative dose.

Does STN recording during placement of stimulator leads correlate to the accuracy of lead placement and useful voltage range?

A. Taha, Y. C. Gan, R. Mitchell, P. Mocroft & H. Pall (Department of Neurosurgery, Queen Elizabeth Hospital, Edgbaston, Birmingham, UK)

Objective: To correlate STN recording during the placement of stimulator leads to the accuracy of lead placement and the range of useful voltage.

Design: Retrospective study reviewing patients' operative notes and case notes. STN recordings were reviewed retrospectively by Dr Mocroft, consultant neurophysiologist.

Subjects and methods: Thirty-two patients, 24 males and eight females, underwent operations to insert STN stimulators from July 1999 to July 2005. Data was obtained for both sides of each patient. The length in millimetres and quality of STN recording (good, moderate and poor) were determined. Details of the electrode used [electrode 1, 2 (left side), 5 and 6 (right side) are central electrodes] and useful voltage range (the range of voltage from having initial beneficial effects to side effects) were obtained when the stimulator for each patient was trialled and switched on 6 weeks after surgery. The electrode with the maximum voltage range was then selected as the optimal electrode. The maximum voltage range and minimum voltage range for each side for each patient were noted, the mean values obtained and correlated to the STN recordings.

Outcome measures: Length of STN recording, quality of STN recording, the electrode with maximum range of useful voltage, the maximum and minimum range of useful voltage for each side.

Results: Of the 64 STN recordings, 35 were good, 15 moderate and 14 poor quality. For good quality STN recordings, 83% of the electrodes used were central; moderate quality, 73.3% central; poor quality, 78.6% central. For useful voltage, good quality recordings gave a mean maximum voltage range of 2.8 V and mean minimum voltage range of 1.30; moderate recordings, 3.5 and 1.31 V and poor, 3.5 and 1.31 V. For length of STN recording, between 5 and 7 mm, 85.7% were central electrodes; 3 – 5 mm, 78.8% central and 1 – 3 mm 70% central electrodes. For useful voltage, between 5 and 7 mm, the mean maximum and minimum voltage range were 2.8 and 1.1 V; 3 – 5 mm, 3.1 and 1.3 V, and 1 – 3 mm, 3.7 and 1.9 V, respectively.

Conclusions: The length of STN recording seemed to correlate well with the accuracy of electrode placement in the STN nucleus, less so the quality of STN recording. Both, however, do not correlate well with the range of useful voltage and hence indirectly with early side effects of stimulation.