The effect of ICP on myelination and the relationship with neurodevelopment in infantile hydrocephalus

S212

Hydrocephalus

just prior 10 insertion. The device is inserted and left in place while a second image is recorded. Postoperatively, the specimen is subjectedto a high resolutionscan and dissection. These scans are co-registered to the intraoperative images. Deviations betweenthe intraoperative position and trajectory from that of the true position are reported as CUE. Results: CUE was measured for image guided lumbar pedicle screw insertion in pigs and in human cadavers. 2 human cadavers spines (20 pedicles)and one pig (12 pedicles) were tested using the protocol), The latter study used the methodsof CUE to study the effects of different registration techniques. The average CUE for the cadaver pedicles was 2.7 mm and 4.7". CUE obtained using surface fitling and paired points was 2.8 mm and 7.fr'. This was not significantly different from 2.9 mm and 8.4° measured for paired point registrationalone. Discussion and Conclusions: CUE is an unbiased method of error reporting by which IGS machines can be classified. These methods can be used to test and compare the accuracy of systems, instruments and techniquesdirectly with one another.

10-28-420 I Localizati~n-~ccuracy of an led-based neuronavlgancn-system

R. Steinmeier, J. Rachinger, M. Kaus, O. Ganslandt, R. Fahlbusch. Neurosurgical Department, Universityof Erlangen-Niirnberg, Erlangen, Germany Clinical accuracy of navigation-systems is of utmost importance. The overallaccuracy is influenced by (1) the technical accuracy of the system, (2) the recording procedure (3) distortion of image-data, and (4) intraoperative events (i.e. -brain-shiff). The aim of this study was to test the influence of different types of recording on the overall-accuracy. Methods: A cylindrical plexiglas-phantom (0 27 em) with 32 rods of different length was taken for navigation-targeting. 16 fklucials (Sofamor-Danek) were attached to the surface of the phantom forming 2 different pattern (4, 6, or 8 localized versus 4, 6, or 8 diffusely scattered): 9 were centrally located on the anterior side, 3 in a triangled shape on the upper base of the cylinder, and additional 2 plus 2 on each side of the cylinder. This model was scanned by MRI (Magnetom Open, Siemens) with 1 mm slices (3D-flash). Image-data were transferred to the navigation-system (Neurostation, Surgical Navigation Technologies). The localization-error was defined as the euclidian distance between targets defined in image-space and those detected in physical-space. Results: The mean localization-error using '1 centrally localized fiducials was: '1 = 4,1.99 ± 0.86 rnrn, '1 = 6, 2.18 ± 0.91 mm, '1 = 8,1.64 ± 0.61 mm. The error using the diffusely scattered fiducials was: '1 = 4, 2.44 ± 0.94 mm, '1 = 6, 2.40 ± 0.99 rnrn, '1 =8,1.93 ± 0.74 mm, '1 =10. 1.98 ± 0.68 mm. A significant localization-error-pattern could be detected showingan increaseof the error for rods located more distally to the camera. Conclusions: We conclude that 8 diffuselyscatteredfiducialsare optimalfor the fiducial-based registration process. Further work concerning the influence of image-data-modality and the influence of slice-thickness is just in progress in our group.

10-28-421

Thursday, 10 July 1997

loweda bettercomprehension of the relationshipbetween the computer-defined lesion borders and epileptogeniczone. Conclusion: In addition to allowing electrophysiological data to be registered with reference to a spatially recorded database, the ability of the MKM system to transfer image-defined lesion contours onto the microscopial operative site should improve our understanding of the lesionaVepileptogenic zone thus guiding the tailoring of the resection.

I0-28-4221 Aopen new frameless stereotactic system for use in the interventional MR R. Bernays1, S. Kollias 2 , N. Khan' • B. Romanowski 3, A. Valavanis 2 . Y. Yonekawa 1. 1Neurosurgical Clinic, 2 Neuroradiological Institute.

3Department of Radiology. University Hospital Zurich, Switzerland Introduction: Neurosurgical procedures with the IMR (interventional MR) require a new generation of surgical instruments. Even thin metallic so-called MR compatible instruments produce considerable image distortion resulting from magneticsusceptibilitydifferencesbetween the biopsy probe and the surrounding tissue. Also, a non-obstructive instrument guiding system is presently not available. Method: We have developeda set of polymer-based instrumentsto address these concerns. A new biopsy probe was developed using carbon fiber reinforced polyetheretherketone (PEEK). In addition we designed a burrhole-based instrumentguide made of polyoxymethylene (POM). To increase versatility, the partthat locksin to the burrholewas producedin three differentsizes. So optimal fixation is possible in the temporal region of the skull, a few millimeters thick, as well as in regions with skull bone over 5 and 10 mm thick. This new system was tested for susceptibilityartifacts and practicability in 5 cadaver studies and 9 patients with brain tumors. Results: Both biopsy cannula and the stereoguide produce no susceptibility artifacts. The stereoguide, when connected to the FPS (f1ashpoint system), allows real-time planning by locking the system to the selected orientation and guidingthe biopsyinstrumentalongthe plannedtrajectory. The 30"angle moving range of the stereoguide proved useful during the process. The biopsies from 9 patients with supratentorial lesions were all diagnostic and completed without adverse events. Discussion and Conclusion: Results from procedures on nine tumor patients support that the new frameless stereoguide, in conjunction with the FPS, is a powerful accessory for performing brain biopsies in the open IMR. The device allows accurate positioning of the biopsy probe and near real-time visualization during the open IMR biopsy procedure. These advances render open IMR a definite improvement over conventional systems.

14:00-16:30

Thursday. 10 July 1997

0-29

Hydrocephalus

I Experience with the "MKM"-system in the surgical treatment of intractable epilepsy

T. Czech, K. Ungersb6ck. W. Dietrich. A. Reinprecht, K. Hittmair 1 , K. Rossler, W.T. Koos. Department of Neurosurgery, 1 Department of Neuroradiofogy. Universityof Vienna MedicalSchool, Vienna, Austria Introduction: Frameless stereotactictechniquescombinedwith high resolution neuroimaging allow interactive image gUided procedures to be performed with high reliability. We report on our experience with the use of the Zeiss MKM system, a frameless stereotactic surgical microscope, in the surgical treatment of patients with intractable epilepsy. Patients and Methods: Between January 1996 and November 1996, 9 surgical procedures were performed in 5 epilepsy patients with the MKM mi· croscope, Resection of epileptogenic tissue was performed in 5 patients with prior chronic subdural electrode (SO) implantation in 4 of these patients. The planning of the operative procedure was based on CT (3 procedures) or MRI (6). Placement of the SO and resection were guided by transferring a contour of the lesion and individual sulci defined on the computer workstation to a display in the microscope ocular. Results: In 7 procedures in 4 patients a lesion contour was defined. Histologically a ganglioglioma was diagnosed in 2 patients and a dysplastic lesion in 2 patients. This could be superimposed with neurophysiological data obtained by chronic lnterlctal and ictal recording and cortical stimulation studies. In 1 patient without lesion brain surface structures and their relationship to electrode contacts could be defined. To avoid the need for implantedfiducial markersdrill burr holes on the bone marginwere used in rereferentiate the systemat the time of the second intervention.The spatially registeredelectrophysiological data at-

I0-29-4231 The effect of ICP on myelination and the relationship with neurodevelopment in infantile hydrocephalus p.w . Hanlo 1, R.H.J.M. Gooskens2 , M.S. van der Knaap" , M. van Schooneveld 2 , J.A.J. Faber" , W.P. Vandertop 1, CAF. Tulteken", J. Willemse2. 1 Opt. of Neurosurgery, 2 Dept. of Child Neurology, Utrecht University Hospital, 3 Dept. of Child Neurology. Free University Amsterdam, 4Centre for Biostatistics, UtrechtUniversity, The Netherlands Introduction: The process of myelin deposition can be affected by several intracranial disorders, including infantile hydrocephalus. It is probable, that the leveland durationof raisedintracranial pressure (Iep) are importantwith respect to the extent of the parenchymal damage and delay of brain maturation. Methods and Results: Nineteen hydrocephalic infants were follOWed-Up by MAl, neuro-developmental testing (NOT) and anterior fontanelle pressure (AFP)measurement. In 16 patients, with confirmed progressive hydrocephalus. shunt implantation was performed. In three patients with mild or non-progressive hydrocephalus, the decision regardingsurgical interventionwas postponed. Cerebrospinal fluid (CSF) volume decreasedgradually after shunt implantation. The averagedrop in mean AFP level. postoperatively, was 55 percent. Postoperatively, both the degree of myelination and NDT score increased equally and SUbstantially. There was a high correlation (r 0.80) between the myelination and NOTscores.The size of the CSF volume showed a poor correlation with the mean AFP,the degreeof myelinationand the NOT scores. There was, however, a significant correlation between tha mean AFP and the degree of myelination

=

Thursday, 10 July 1997

(r: 0.67) and also between the mean AFP and the NOT scores (r: 0.70). By means of multiple regression analysis. it was calculated that 69 per cent of the variation in NOT, was related to the influence of mean AFP and myelination. A more long-term follow-up (mean: 27 months) showed a significant correlation betweenthe progress of myelination and the developmentallevel (r: 0.78) . Most of the children with a severely delayed myelination, preoperatively, showed a recoveryof myelination, following CSF drainage. Conclusion: 1. raised ICP is relatedto developmental outcome, through the process of myelination, but cannot account entirely, for the delay in myelination and neurodevelopment; 2. the delay in myelinationcan be (partially) reversible; and 3. CSF volume is of minor importance regarding neurodevelopment.

I0-29-4241 changes Preop. and post v.p. shunt transcranial Doppler in patients with hydrocephalus: A prospective study of 35 patients AK. Mahapatra,A. Jindal. Departmentof Neurosurgery, Neuroscience Centre A.U.M.S., New Delhi-110029, India Transcranial doppler is a non invasive method to study cerebral blood flow velocities in basal blood vessels. Hydrocephalus causes displacement and stretchingof primary cerebral vessels, compromisingcerebral blood flow. Thirty five patients (16 children) were studied using TCD who had hydrocephalus (various etiologies other than SAH). Preoperatively 29 patients had high and 6 patients had normal PI value. Significant reduction (>0.12) in PI was seen in 28 patients and out of these 27 (96%) showed decreased ventricle size. Insignificant reduction was seen in 7 patientsand none of them showeda change in ventricular size. The decrease of PI correlated well with shunt function. TCD thus can be used in patients with hydrocephalus as a mean to determine shunt function.

S213

Hydrocephalus

Methods: From October 1993 to December 1996 82 programmable and 161 non-programmable valves were investigated retrospectively with respect to the underlying diseases: subarachnoidal and intracranial he,morrhage, inflammation, normal pressure hydrocephalus, aquaeductal stenoses, cerebellar infarction. subarachnoidal cyst. tumor induced hydrocephalus. posttraumatic hydrocephalus. The causes of malfunction were also looked for. Results: The distribution between programmable and non-programmable valves is shown in the following table:

SAH ICH Aqued. stenosis Inflammation

Progr.

Non-progr.

3 8 11 0

68

55 0

Postlr. hydroc. Tu. indu. hydroc. Subarch. cysl

Progr.

Non-progr.

6 17 6

24 0

7

3

In our series in 18 patient (7.8%) the implanted shunt valves had to be removedbecause of dysfunction. The different causes of valve removal will be described in detail. Discussion and Conclusion: Concerning the dysfunction rate the programmableand non-programmable Medos·Hakim-valve are very reliable easy in handling, allow a high CSF protein content and rarely lead to problems with overdrainage. There is a clear indication for using programmable valves in patient with NPH, aquaeductal stenoses and subarachnoidal cysts. For patients with SAH, ICH and inflammation the implantationof a non-programmable valve seems to be sufficient. In case of posttraumatic and tumor induced hydrocephalus the clear indication for the appropriate valve has yet to be defined.

I0-29-4271 2regulatmg year f.o"ow-up of 28 patients with a new flow shunt Newton Paes. UniversIty of Mogi oes Cruzes. Sao Paulo. Brazil

I0-29-4251

Shunt technology. What is the state-of-the-art?

A Aschoff, P. Kremer, C. Benesch, B. Hashemi, C. Schulte, St. Kunze. University of Heidelberg, Dept. Neurosurgery, Heidelberg, Germany 95 different valve constructions are available in 1996. Shunts must meet three requirements: 1. Accuracy and long-term-stability, 2. flow properties, to keep the ICP in the physiological ranges. independent on body position or ICP variations of vascular origin and 3. safety, e.g. no reflux. insusceptibility to extemal pressure. flectionson curve neck, body movements, magneticfields or increased CSF protein. In bench-tests we tested 383 valves, inclUding 52 different constructions of 15·producers; most were new, 145 explanted after 2 days-21 years. All specimen had resistance-tests concordant to ASTMIISO. Most probes submitted up to 35 additionalsubtests to measurethe long-term-stability (34 x 365 days, 116 x 14-90 days), the flow dependent on pressure, safety to reflux, susceptibility to (e.g. subcutaneous) extemal fluid pressure, vectorial forces, flection no curved neck,temperature, high CSF-protein(500mg/dl),air bubbles,pulsations, pumping manoeuvres, etc.. ·Programmable" valves were exposed to common and high magneticfields and tests of decentrationtolerance during adjustment. Gravitational valves were checked using varied angles of verticalisation and with simulated body movements. Only 36% of all valves met the manufacturer's specification ±20 mmH20 24% deviated ± 21- 50 mmH20, 15% ±51-80, 12% of ± 81- 135 and 13% 136-1446 mmH20 Ball-valveswere superior with 61% accurateresults. During long-term-lestsup to 90 days drifted 72%, during 365 days 89% of the probes. Only 10% of the valves offered approx. physiological flow-rates in horizontal and position, but most of them had safety problems. 32% of the valves were not reflux-safe, 29% susceptible to external fluid pressure (increase up to 3OOO%!), 50% to vectorial forces, and 56% to flections, which are typical on a curved head. The available optimum are ball-valves in combination with gravitational devices and kink-resistentcatheters with 0.7-0.8 mm internal diameter. Conclusion: 1. The majority of valves showed serious deficits in accuracy, long-term-stability, adequate flow-rates in upright and horizontal position and safety. 2. The clinically uneventful function of 80% of shunts is due to the adaptability of the patients, not a result of functional valves. 3. The available optimum are (adjustable) ball-valves with additional gravitational units and lowflow-catheters.

I0-29-4261 Indications and complications on 243 implantations of programmable and non-programmable Medos-Hakim valves C. Bonk, J. Hampl, M. Wiegleb, G. ReiB, S.-A May,G. Schackert. Department of Neurosurgery, Technical University Dresden, Dresden, Germany Introduction: The selection of the correct valve in the treatment of hydrocephalus is still a matter of debate. In the literature the valve dysfunction rate varies between 1.8-22.8% with an average of 8.3% (5820 implantations).

Clinical complicationsrelated to overdrainage of hydrocephalusshunts are well documented in medical literature. While the development of the anti-siphon mechanism by Portnoy was an important step in reduction of complications due to overdrainage. it introduced a new series of complications. Flow control mechanisms such as the Orbis Sigma also have reduced clinical complications due to overdrainage, however they have proven particularly susceptible to occlusion and complications related to underdrainage. The author presents design concepts and the first clinical results of a new flow regulated shunt designed with a dual pathway to ensure patient safety. The device reduces overdrainage and regulates the flow of CSF by way of its unique variable aperturedesign. Complications related to underdrainage are minimized by virtue of a secondfluid pathwayWhich is activatedat high pressures. ICP is maintained within physiologic limits regardlessof patient position. 28 patients were followed for up to 24 months, with pre and post surgical CT scans. Follow up CT scans were pertormedat 2 weeks, and I , 3, 6. 12, 18, & 24 months. Patient selection ranged from 2 months to 72 years, 4 (14%) adults, and 24 (86%) pediatric. 18 (64%) of the cases were initial implants, 4 (14%) were shunt revisions of obstructed systems, and 4 (14%) were revisions to correct overdrainage. The patients represented a variety of etiologies. 2 patients (7%) died due to causes unrelated to shunt function. 26 patients (93%) showed marked reduction of symptoms. Therewere no complications relatedto either under or overdrainage.

I0-29-4281

Clinical experience with the hydrostatic dual-switch valve

Christian Sprung 1 , Christoph Miethke1 , Hans-Axel Trost 2, WolfgangLanksch 1 , Dietmar Stolke2. 1 Neurosurgical Department, Virchow-Klinikum, Berlin, Germany, 2 Neurosurgical Department, University of Essen, Germany Despite the advantage01 ventriculostomy of the III. ventricle with the chance of restoringphysiological conditionsin casesof noncommunicating hydrocephalus, the majorityof hydrocephalic patients are still treated with valve-regulatedsilastic shunts, but the currently available valves are still far from pertect. Three types of valves are used: Conventional differential pressure valves with only one opening pressure for the lying position of the patient, adjustable valves and hydrostatic valves which take into account the physics in the drainage sysrem due 10 lhe changing posture of the patient. Whereas the design principles of differential pressure valves and adjustable devices involve the danger of overdrainage, the hydrostatic valves on the market have a tendency to clog. To overcome these main valve-related problems of shunting the dual-switch valve (DSV) has been introduced in 1995. In contrast to other devices the DSV consists of two different valve chambers in parallel, one for the supine and other for the standing position of the patient. Large area diaphragms were chosen to minimize the danger of clogging. A clinical trial was accomplished at the neurosurgical departments of the Virchow-Klinikum in Berlin and the University Hospital in Essen. comparing the new device with a conventional differential pressure valve. The results of this