Child’s Nerv Syst (2000) 16:122–128 © Springer-Verlag 2000

ABSTRACTS

Hydrocephalus beyond 2000, Sydney, Australia, 8–10 March 2000

Magnitude and significance of CSF transport by extracranial lymphatic vessels. M. Johnston, M. Boulton, M. Flessner, I. Silver (Toronto) In adult sheep, approximately one-half of the CSF removed from the cranial vault transports through the cribriform plate into the nasal submucosa, where it is absorbed by cervical lymphatics. A close correlation exists between intracranial pressure and cervical lymphatic pressure and lymph flow rates. In addition, intracranial pressure accommodation in response to bolus infusions into the CSF compartment is impaired by extracranial obstruction of the cribriform plate. The cribriform plate was scraped and sealed with bone wax, and this procedure increased CSF outflow resistance 2.7-fold (P=0.006). Preliminary data from our group suggest that there may be an even greater role for lymphatics in CSF drainage in the fetus. If this is true, impaired CSF transport to, or through, extracranial lymphatics may be important in the development of pediatric hydrocephalus. Percutaneous endoscopic recanalization of catheter procedure – a study of 45 cases with long-term follow-up. J.V. Pattisapu, E. Trumble, K. Taylor, P.D. Howard, T. Kovacs, A. Arbogast (Orlando) Percutaneous endoscopic recanalization of the catheter (PERC procedure) has been performed at our institution in certain cases of proximal shunt malfunction since October 1996. This study reviews the present outcome in the children in whom this technique has been used, with an analysis of the critical factors affecting the outcome. A total of 45 PERC procedures were performed in 38 children (mean age 45 months) for proximal shunt malfunction; all procedures were performed in the operating room under general anesthesia as per the IRB protocol (as previously described). During the study period of 33 months, 220 cases of standard open shunt revisions were performed by the authors. The children were evaluated at 2, 4, 8, 12, and 24 weeks and every 6 months thereafter, with clinical assessments and CT/MRI studies. There were 11 failures in this study of 45 cases (76% success rate) with a mean follow-up of 13.3 months. Most malfunctions occurred before 5 months had elapsed, mainly in premature infants and children who had suffered previous infections. Five children underwent 7 repeat PERC procedures, and in 3 of these children the repeat attempt failed, making an open shunt operation necessary. There were no complications such as infection, bleeding, seizures, and heat damage. The PERC procedure has been very effective in selected cases of proximal shunt malfunction. This long-term study proves that this technique should be included in our armamentarium for shunt management. CSF shunt malfunction: a study of the spectrum of mechanisms and clinical presentation. C.E. Gilkes, A.J.W. Steers, R.A. Minns (Edinburgh) The objects of this study were to extrapolate and describe the multiple mechanisms of shunt malfunction and their frequency, timing and symptomatology, and to derive algorithms for investigation and management. All shunt malfunctions performed at the RHSC in Edinburgh during the period 1996–1998 inclusive were

studied retrospectively. There were 92 shunt malfunctions in 48 children. The statistical software ‘Statistica’ was used for analysis and comparison of survival. In all, 29 different malfunction mechanisms were derived, 25 of which were identified as contributing to the 92 malfunctions recorded during this period. The major categories of shunt failure were ‘fracture’ (92%) and ‘obstruction’ (49%). Survival analysis of ‘shunt age’ before revision revealed that 60% of these shunts had malfunctioned within a year of insertion. There were significant differences in timing between proximal and distal malfunctions and between infection and distal fracture. We identified 30 different symptoms and signs as presentations of shunt malfunction. The most common were vomiting, headache, drowsiness or lethargy, and behaviour change. The most common finding in the children with ventriculoperitoneal shunts was a CSF-filled swelling (decompensation), while lowered level of consciousness, bradycardia, bradypnoea and hypertension were the most common findings with cystoperitoneal shunts. Raised intracranial pressure (ICP) was found in 91% of those who had not decompensated, and preserved cerebral perfusion pressure in 49% of them. This study identified 29 comprehensive mechanisms of shunt malfunction, and it is hoped that the paper might provide a standard vocabulary. ICP measurement is the single most important investigation determining malfunction, although where CSF collections are evident ICP may be an unreliable way of detecting malfunction. Manifestations of impingement of the corpus callosum on the falx cerebri caused by hydrocephalus: a clinical and animal model study. J.R. Jinkins (San Antonio) The clinical features of patients with hydrocephalus include generalized reductions in coordinated motor and cognitive functions. A retrospective MRI review of subjects with criteria of hydrocephalus was undertaken to re-evaluate the specific imaging correlates of the signs and symptoms associated with this pathologic process. In addition, an animal model was developed for further study of this condition. Forty adults with hydrocephalus revealed by MR evaluation were carefully scrutinized in an effort to elucidate specific radiologic patterns of abnormality. Spin-echo MR techniques were used with T1 and T2 weighting in three orthogonal planes. MR criteria of hydrocephalus encompassed ballooned lateral ventricles to include the temporal horns, a pronounced upward elevation of the corpus callosum, and an outward expansion of the cerebral hemispheres at the expense of the subarachnoid space overlying the convexities. The animal model focused on a histological evaluation of the corpus callosum following impingement by a flat plastic blade surgically inserted into the interhemispheric fissure. The significant related morphologic change seen on MR in hydrocephalus was a localized dorsal flattening and thinning of the posterior body of the corpus callosum. Importantly, all but 3 of the 24 patients with this phenomenon manifested varying combinations of imbalance, gait disturbance, incontinence, short-term memory deficits, and global dementia. In the presence of hydrocephalus, but in the absence of this specific callosal configuration, only 1 of the remaining 126 subjects revealed symptoms that might suggest the presence of hydrocephalus (i.e., profound dementia). The animal model revealed callosal thinning (i.e., atrophy)

123

and gross axonal loss in the area of physical impingement. The structure responsible for this local callosal flattening and thinning in hydrocephalus is the rigid free surface of the falx cerebri as it impinges on the superior, caudal extent of the upwardly expanding corpus callosum and supracallosal hippocampal formation. This physical mechanical insult hypothetically causes variable axonal dysfunction, ranging from decreased to increased neurophysiologic activity. It is postulated that callosal impingement represents an active, dynamic partial hemispheric disconnection and accounts, in part, for the complex clinical state associated with hydrocephalus. The animal model studies demonstrate the histopathological changes of frank axonal loss and gross atrophy underlying the callosal impingement. The U.K. Shunt Registry. H.K. Richards, C.O. Kane, H. Seeley, M. Madakbas, P. Whitfield, J.D. Pickard (Cambridge) The United Kingdom Shunt Registry was inaugurated in October 1994, and following a pilot phase, has been active since May 1995. We now have data on over 10,000 operations performed on 7,400 patients. The aims of the Registry are: rapid accumulation of data to identify the most suitable devices for use in hydrocephalus patients by identifying variations between individual manufacturers and valve types; identification of substandard valves; facilitation of risk stratification in different diagnostic groups; and its use as a mechanism for market surveillance and to facilitate audit of standards of care. The Registry was initiated by contacting the clinical directors of all neurosurgery units throughout the United Kingdom and Eire and explaining the aims of the Registry, and asking for the name of a link person working in each unit’s operating theatre. This person, who was either a senior nurse or an operating department assistant, had to be prepared to take responsibility for ensuring that registration forms were completed correctly and returned to the Registry. The clinical directors were also asked to identify paediatric units they were aware of in which shunt insertion took place. A total of 61 neurosurgical and paediatric units were detected in this manner. This was followed by a visit to each centre by Registry personnel to explain form completion. Centres are also visited occasionally for audit and dissemination of results. The Registry form has undergone two refinements, but has always been kept to a single side of A4 paper. We have always felt that the filling in of Registry forms should be quick and simple, so that as much information as possible was collected as multiple-choice box ticking, with only a minimum of writing. We have always believed that anonymity is of paramount importance. Each centre can have ready access to its own results and to national averages, but not to results from other centres. Our data have enabled us to examine operative procedures, including antibiotic use, reasons for revision, and devices used. We also have a substantial data set on subsequent revisions. This data set has enabled us to calculate the annual overall revision rate at 25.8% (±1.6%). The revision rate has also been calculated for operations involving valve insertion or replacement. The calculated annual failure rate of valves is 16.2% (±1.5%). In vivo differential pressure measurement. G. Magram (Edison) The intraventricular and distal receptacle pressures were measured in patients at the time of their shunt insertion or revision. The pressures were measured through the proximal and distal tubing to determine the differential pressure across the valve. The amplitude and phase of the pressure waves were analyzed to improve understanding of the physiological parameters responsible for shunt flow. While the maximal amplitude of the distal receptacle pressure can exceed the intraventricular pressure during a Valsalva maneuver, the negative portion of the wave following the positive deflection can contribute to episodic shunt flow at a rate that exceeds the CSF production rate. The pathophysiological and therapeutic implications of these measurements will be reviewed.

Complex hydrocephalus and nocturnal intracranial hypertension. G. Magram (Edison) Hydrocephalus is most commonly categorized as either communicating or noncommunicating. A third category is being proposed, that of complex hydrocephalus. Complex hydrocephalus results when there is an increased resistance to the circulation at more than one level. In infancy, complex hydrocephalus needs to be differentiated from the slit ventricle syndrome, as both can result in intermittent or recurrent proximal shunt obstruction. Complex hydrocephalus is typically worsened when the sufferer is supine, whereas a slit ventricle syndrome is often alleviated when he or she is supine. It is important to recognize complex hydrocephalus in order to make adequate treatment of all levels of CSF circulatory resistance possible and allow a proper selection of candidates for III ventriculocisternostomy. Selected illustrative cases will be presented. Cerebrovascular adaptation and volume shift in chronic hydrocephalus. D.J. Skarupa, A.M. Booth, M.J. Johnson, I. Ayzman, A.S. Wood, J.J. Hoegler, M.G. Luciano (Cleveland) Because of observed cerebral blood flow changes in hydrocephalus, we set up the hypothesis that there is preferential compression of the cerebral vessels accompanying chronic tissue compression. In this study, 20 animals underwent successful surgical induction of hydrocephalus via cyanoacrylic glue injection into the IV ventricle. The degree and pattern of vascular compression were determined by quantitating the vessel density, diameter, and vascularity in three serial sections from multiple regions of the parieto-occipital lobe. These parameters were related to the degree of tissue compression as measured by postinduction MRI. We found overall vessel diameters decreased in periventricular regions (P=0.005) and capillary diameters decreased (P=0.042) in the white matter, while large vessel diameters (P=0.002) increased in superficial gray matter. Increases in both overall (P=0.042) and large-vessel (P=0.006) vascularity were observed in superficial gray matter. These findings suggest that there is no preferential compression in chronic hydrocephalus, but rather a vascular volume shift leading to preferential preservation of the vascular system in the gray matter. Finally, in the gray matter, overall density (P=0.003) and capillary density (P=0.004) increased with time after hydrocephalus induction. Increased vessel diameters and capillary density in gray matter may represent vascular adaptation to a hypoxic environment in chronic hydrocephalus. A new software for continuous computerized recording of intracranial mean and pulse pressure. C. Di Rocco, P. Santini, F. Velardi (Rome) In spite of the modern sophisticated tools for neuro-diagnosis, there are several clinical conditions characterized by an abnormal CSF dynamics that cannot be identified on the grounds of the morphologic findings with sufficient reliability to confirm or rule out that surgery is indicated. In such instances, the continuous recording of intracranial pressure (ICP) maintains its clinical relevance. Unfortunately, most of the available apparatuses for continuous ICP recording are hampered by difficulties in implanting the recording device and keeping it in place; furthermore, most of these systems require the patient to remain in bed and may cause discomfort, so possibly interfering with the physiological parameters being investigated. The authors have developed a software for prolonged ICP recording, which utilizes the Codman ICP express sensor and a desk-top computer. The system allows continuous recording of the ICP while the patient is relatively free to move. It complies with the main requirements for such apparatus: it is easy to use, it is reliable, its cost is relatively low and the data recorded can be saved, recalled and automatically analysed at any time.

124

Lateral ventricular asymmetry following CSF shunt surgery. S. Nomura, T. Nishizaki (Ube) After CSF shunt surgery for children with hydrocephalus, asymmetric lateral ventricles (ALV) may appear. In this study, children with ALV were characterized in comparison with children who had symmetric lateral ventricles (SLV) after surgery. ALV developed in 7 out of 26 children (27%), this incidence being significantly higher than that in adults (3.5%). ALV appeared in 1 out of 7 children under 1 year old and in 2 out of 4 children who were 1 year old. No ALV were seen in 5 children with long periods of ventriculomegaly. Out of 9 children with brain tumors, ALV were observed in 3. Five of the 6 children with SLV had tumors for longer than 2 years. Four children who developed ALV immediately after surgery were asymptomatic, and the degree of asymmetry was stable or improved. In contrast, the 3 children who developed ALV later suffered intermittent rises of intracranial pressure, necessitating shunt revision using a high-pressure valve or a valve with an anti-siphon device. ALV occurs occasionally in infants who have a short period of hydrocephalus. ALV caused by excessive elimination of CSF during shunt surgery improves spontaneously. Continuous overdrainage causes symptomatic ALV and necessitates shunt revision. Sequential proton MR spectroscopy studies of cat brain affected by kaolin-induced hydrocephalus. M.J. Kim, S.K. Hwang, Y. Chang, J.H. Hwang, Y.S. Kim, S.L. Kim (Taegu) The object of the study was to determine the sequential metabolic changes in experimental hydrocephalus and the clinical applicability to the diagnosis and prognosis of acute stage hydrocephalus using proton MR spectroscopy. Hydrocephalus was experimentally induced in 30 cats (2–3 kg body weight) by injecting 1 ml of sterile kaolin suspension(250 mg/ml) into the cisterna magna. Proton MRS was performed with a 1.5-T MRI/MRS unit (Vision Plus, Siemens) before this treatment and at 1, 3, 7, 14, 21, and 28 days after the kaolin injection. The NAA/Cr ratio of the sequential proton MRS in cats with kaolin-induced hydrocephalus reflects a metabolic aspect of the hydrocephalus at each stage, which may provide diagnostic information in acute-stage hydrocephalus. In addition, an initial fall of the NAA/Cr ratio and recovery in the late stage, when no lactate peak emerges, may suggest that the main insult of the parenchyma is not to the neuron itself but to the axon, which might be a good prognosis. Molecular biology and genetics of hydrocephalus. X. Cai, D. Bailey, J. Pattisapu (Orlando) Hydrocephalus is an etiologically heterogeneous disease. Its causes include trauma, subarachnoid hemorrhage, infection (meningitis), developmental abnormalities, and genetic changes. The pathogenesis of hydrocephalus, however, is still unknown. We review the recent advances in the study of the molecular biology and genetics of hydrocephalus and discuss the molecules concerned. Cell adhesion molecule L1 (L1CAM) is the only gene definitely known to cause human hydrocephalus. It is a single transmembrane molecule involved in neural cell adhesion, neurite outgrowth and pathfinding, neuronal migration and myelination, and memory and learning. Mutations in the L1CAM gene are highly variable, occurring throughout the molecule, and cause hydrocephalus and other neurological abnormalities. Transforming growth factor ß1 (TGF-ß1) is a multifunctional cytokine. Overexpression of TGF-ß1 in transgenic mouse brain has been shown to cause severe hydrocephalus. It is proposed that the mechanism is connected with increased production of extracellular matrix components. Also, TGF-ß1 is increased in some other neurological diseases. The forkhead / winged helix is a member of a large family of evolutionarily conserved DNA-binding proteins. It was recently reported that disruption of this gene caused severe hydrocephalus in mouse. The mechanism is unknown. Otx2 functions as a head organizer. It plays an important part in head morphogenesis. Mice that were heterozygous for a mutant Otx2

gene showed pronounced dilatation of lateral ventricles and a ballooned cerebrum. Histological analysis shows edematous change of the periventricular white matter. The mechanism of this is also unclear. Other molecules, such as FGF2, folic acid, are also discussed. Hydrocephalus in the H-Tx rat: a monogenic disease? X. Cai, G. McGraw, J.V. Pattisapu, L. von Kalm, S. Willingham, D. Socci, J.S. Gibson (Orlando) The H-Tx rat is a model of hydrocephalus with a poorly understood mechanism of inheritance. A polygenic mode of inheritance was assumed earlier, but mating data to support this hypothesis have not become available. We have analyzed data collected from eight generations of H-Tx rats and four generations of cross-matings of H-Tx rats and Sprague Dawley (SD) rats. We observed the hydrocephalus phenotype in 113 of 129 (87.60%) random brothersister matings, with males and females equally affected. In the first generation of the cross-matings there were no hydrocephalic rats among 124 pups (F1). However, subsequent brother–sister matings of F1 animals yielded a lower incidence of the disease in their pups (F2; 4.67% in hydrocephalic H-Tx rat/SD rat matings and 5.11% in normal H-Tx rat/SD rat matings) than the overall incidence observed for the H-TX rat colony (30.35%). The backcross matings between the F2 generations and the H-Tx rats yielded a higher incidence of hydrocephalus than did the crossmatings, but it was still lower than the incidence in the H-Tx rat colony. The data from random matings, cross-matings and backcross matings strongly suggest that the H-Tx rat is a homozygous carrier of an autosomal recessive gene. The invariable incidence in the H-Tx rats suggests that the gene is incompletely penetrated. The lower incidences in the cross-matings and in the back-cross matings may be attributed at least in part to the modified effect of the SD rat genome. Furthermore, the data clearly rule out sexlinked and polygenic modes of inheritance and provide further insight into the genetic inheritance of hydrocephalus. HR1 shunt versus other devices. C. Miethke, C. Sprung (Berlin) The strikingly large number of different designs of valves for the treatment of hydrocephalus is in contrast to the results of clinical studies, which reveal no significant differences for simple DP valves, the OSV, or the Delta valve. This observation raises the question of whether technical differences in valves really have any significant influence on the outcome after shunting or whether the cheapest shunt is possibly the best solution. Investigation of the valves available shows that the devices can be divided into three groups: DP valves with a discrete opening pressure based on the patient’s needs for adoption of the lying position, adjustable DP valves, which allow the possibility of readjustment of the performance of the valve, and hydrostatic valves, which take into account the changing physics in the drainage system with changes in the posture of the patient. Quite apart from differences in the quality of manufacturing, investigation of the valves in a computer-controlled test apparatus simulating the implantation of the shunt shows impressive differences. Both the valves in the first group and devices in the second group systematically produce unphysiological negative intraventricular pressures (IVP), even in the highest possible pressure range of the adjustable valves. On the one hand hydrostatic valves improve pressure control, but on the other some designs were found to have important specific disadvantages. So-called flow-controlled devices can lead to both under- and overdrainage. If the simulated CSF production is lower than the physiological production, the resulting IVP is extremely negative; if when the patient is in the lying position the production is higher, the IVP is pathologically increased and the patient is underdrained. The best results are seen with such gravity-assisted devices as the dual-switch valve, the Paedi-GAV, the ShuntAssistant and the Cordis GCA, or the horizontal-vertical valve. These devices are designed for physiological control of the IVP not only

125

when patients are lying down but also when they are in the upright posture. Both the in vitro results and numerous clinical results demonstrate that shunt therapy of hydrocephalus can be improved by the new generation of gravitational valves. Shunting to the sagittal sinus. S.E. Børgesen, N. Agerlin, F. Gjerris (Copenhagen) The results of shunting in patients with hydrocephalus are still highly unsatisfactory, in spite of the development of new shunting techniques. The complication rates following implantation of CSF shunts have decreased only marginally during recent years. In our database referring to 2,400 shunting procedures the complication rates can be seen to have remained unchanged over the years, resulting in a reoperation rate of 2.7 per patient and a ‘shunt survival’ of 50% at 5 years. About 50% of the complications may be ascribed to overdrainage of the CSF. Investigations of the cerebrospinal fluid (CSF) dynamics in patients and in normal subjects have enabled us to define normal values for intracranial pressure (ICP), resistance to outflow of CSF (Rout), CSF production rate, and pressure in the sagittal sinus. The values are derived from infusion tests and ICP measurements in more than 400 patients and in 79 normals. On the basis of these values and of our knowledge of CSF physiology and drainage, we have developed a shunt that drains CSF to the sagittal sinus and restores normal conditions for CSF dynamics. The shunt consists of a uni-directional valve, a ‘preliminary’ chamber, a resistance tube made of titanium, and a titanium tube conducting CSF into the sagittal sinus. The first results obtained with the use of the new shunt are very promising. It has an immediate effect on the clinical symptoms; it restores CSF dynamics (investigated with the shunt inserted); and the size of the ventricles is only gradually diminished. Slit ventricles have not yet been observed. Because of its physiological drainage of CSF, the shunt will probably have better survival rates. The shunt has proved easy to implant. The results from the first year of use of the shunt will be presented. The new shunt system for neonates (especially premature infants) – the neonate model ultra vs in-line valve. Y. Takahashi (Otaru) In neonatal hydrocephalic children it is not easy to adjust a shunt system. The complications arise because shunt systems are too large to be suitable for babies. This causes the necrosis and the leakage of cerebrospinal fluid (CSF) surrounding the shunt line, which lead in turn to the development of shunt infection and meningitis. In the past 6 years, we have placed 74 ventriculoperitoneal (VP) shunts in premature babies weighing under 2,500 g. We divided the cases into group A (using ultra vs in-line valve [UVS]: 22 cases) and group B (without use of the UVS: 52 cases). The complication rates by group are as follows: group A zero and group B 35%. This shows that the complication rate was significantly lower when the USV was used. We also obtained similar results in full-term babies. The shunt system is selected in each case according to the baby’s age and to our purpose and the future management strategy. The particular features of the UVS are that it offers pressure performance identical to that of the standard model, but has a maximum diameter of less than 4 mm and an overall length of less than 12 mm, making it ideal for patients requiring the lowest profile valve (low pressure 15–54 mmH2O). In view of the above advantages of UVS and our clinical results, my conclusion is that UVS is very effective for the treatment and prevention of shunt complications in premature and full-term babies. Withdrawal of the shunt – main purpose of clinical use of the programmable shunt system (specially ordered SOPHYR) and its effectiveness in pediatric hydrocephalus. Y. Takahashi (Otaru) The most important function of the programmable valve (PV) is to limit the shunt-dependent flow of cerebrospinal fluid by increasing the PV pressure to activate the regular circulation of cerebro-

spinal fluid for withdrawal of the shunt after cerebral development is achieved. In the work reported here, I studied the withdrawal of the shunt with PVs (MEDOSR and specially ordered SOPHYR). Before use of the PV, shunt removal was possible in 18 out of 57 patients (32%). After use of the PV was initiated, removal of the shunt was possible in 84 of 145 patients (58%). Thus, the rate of shunt removal was significantly higher after the use of the PV than before. Especially with specially ordered SOPHYR valves, the rate of shunt removal is gradually increasing. At present, it is possible to remove or withdraw the shunt system from more than half of our pediatric hydrocephalus patients who have had VP shunt operations earlier, following gradual and precise increase of the valve pressure on the PV. Careful observation is required during the period when the PV is on a high-pressure setting and for 6–12 months after the removal of the shunt system. MRI of fetal brain before and after in utero repair of meningomyelocele. L.T. Bilaniuk (Philadelphia) The purpose of this study was to show the utility of fetal MRI in monitoring the size of the ventricles and the subarachnoid spaces and the degree of hindbrain herniation following in utero repair of meningomyelocele. During a 16-month period, 36 women carrying fetuses diagnosed as having meningomyelocele underwent MR imaging and were evaluated in the Center for Fetal Diagnosis and Treatment at The Children’s Hospital of Philadelphia. Ten women agreed to fetal surgery to repair the meningomyelocele. MRIs utilizing an ultrafast sequence (HASTE) were performed before in utero surgery, every 3 weeks until birth, and then soon after birth. Multiple slices were obtained in 10–25 s, thus avoiding fetal motion. Preoperative fetal MRI demonstrated funneling of the posterior fossa structures, absent IV ventricle and subarachnoid spaces, and supratentorial hydrocephalus. Postoperative fetal MRI showed improvement of hindbrain herniation, with IV ventricle and subarachnoid spaces becoming evident. Supratentorial hydrocephalus increased slightly to moderately and then stabilized. Ultrafast MRI sequences permit rapid evaluation of the fetus and provide very good anatomic detail. Preliminary results indicate that in utero repair of the meningomyelocele reverses the hindbrain herniation and improves the CSF dynamics. Fetal MRI in the diagnosis of hydrocephalus. R.A. Zimmerman, L.T. Bilaniuk, J.V. Hunter (Philadephia) To demonstrate the utility of ultrafast MRI in the evaluation of the fetal brain for hydrocephalus, 35 infants with ultrasound diagnosis of “ventricular enlargement” were evaluated by ultrafast MRI performed on a 1.5-T MR. Fetal age ranged between 18 and 38 weeks of gestation. In almost all patients the three cardinal planes (axial, coronal, sagittal) could be obtained for the evaluation of the ventricular system. In this way, it was possible to assess the relative sizes of the lateral and III and IV ventricles. The categories of findings were: ventriculomegaly of unknown etiology in 3 cases, obstructive hydrocephalus in 3, hydrocephalus due to Chiari II malformation in 18, hydrocephalus associated with Dandy-Walker malformation in 4, and hydrocephalus ex vacuo in 2. Four patients had monoventricles due to holoprosencephaly and 1, a large arachnoid cyst mimicking a dilated occipital horn of the lateral ventricle. Ultrafast fetal MR allows evaluation of hydrocephalus in utero, giving information that can be used in therapeutic decision making. A retrospective study of 107 cases of hydrocephalus. L. DongHai, H. Tao, X. Geng-Sheng (Nanchang) The objective of this study was to analyze the causes of hydrocephalus and factors relating to the outcome. Clinical materials in 107 cases of hydrocephalus from 1991 to 1998 were analyzed; patients affected by various brain tumors were excluded from the study. There were 73 male and 34 female patients in the study,

126

with a mean age of 31.4 years. The causes of hydrocephalus included primitive aqueductal stenosis in 33 cases of obstructive hydrocephalus; there were 74 cases of communicating hydrocephalus, a history of meningitis in 9 cases, a history of intracranial tuberculosis in 11 cases, earlier mild to severe brain trauma in 16 cases, earlier tumor resection in 9 cases, a history of subarachnoid hemorrhage in 6 cases, and no definite known cause in 23 cases. In 95 cases, VP shunting operations were performed. Patients were divided into two groups according to preoperative ICP: group 1 with high pressure (53 cases), and group 2 with normal pressure (42 cases).The follow-up data from 1 to 5 years in 87 patients revealed that 42 cases in group 1 and 15 cases in group 2 were free of symptoms and signs, there was no change in 7 cases in group 2, and the others were improved. Shunting occlusion occurred in 16 cases, and reoperations had to be done in these patients. The relation between hydrocephalus and a history of infection, trauma or SAH is not definitive; the causes of hydrocephalus are complicated. Preoperative high pressure seems to be correlated with a good outcome. Shunt occlusion is one of the main complications. Magnetic resonance imaging assessment of endoscopic III ventriculostomy. M. Javadpour, M. Palaniappan, C. Mallucci, P. May, H. Carty (Liverpool) To evaluate the role of magnetic resonance imaging (MRI) in the assessment of suitability for and success of endoscopic III ventriculostomy, a retrospective review of a customised MR sequence in children with hydrocephalus was performed. The subjects were 20 consecutive patients, who underwent this sequence for III ventriculostomy over a 6-month period. The outcome measures used were the correlations between pre- and postoperative imaging and between both and the clinical outcome. A sagittal T2-weighted turbo spin-echo sequence of 12 slices 0.7 mm thick through the III ventricle was undertaken in each of the 20 patients. The demographics and the clinical and radiological outcomes are discussed. This highly specific imaging protocol for pre- and postsurgical assessment in endoscopic III ventriculostomy is a simple and effective method of determining the patency of the procedure. The role of fetal magnetic resonance imaging in paediatric neurosurgery. M. Javadpour, G.R. Mitchell, P. May, C. Mallucci, D. Pilling (Liverpool) Our object in preparing this paper was to present and discuss the expanding role of fetal magnetic resonance imaging (FMRI) in paediatric neurosurgery in the United Kingdom. The study took the form of a retrospective analysis of the clinical value of FMRI over a 1-year period in the diagnosis of CNS anomalies and in related counselling and presurgical planning. In the study period 40 fetal MRIs were performed as part of the Royal Liverpool Children’s Hospital FMRI service for systemic congenital anomalies. The six significant neurosurgical anomalies identified in this group are discussed and presented in the 6 selected patients. It is concluded that FMRI has an important role in the identification and treatment of CNS anomalies in paediatric neurosurgical practice. The role of fetal neurosurgery has yet to be determined, but in vascular and hydrocephalic anomalies the role is more clearly defined. Involvement of the neurosurgeon at an early stage is advocated. Three-dimensional composite SPECT/MRI images in hydrocephalus. T. Mito, I. Shibata, N. Sugo, T. Kano, M. Takano. H. Takahashi, J. Sugita (Tokyo) The three-dimensional (3D) SPECT imaging technique we have been studying and publishing on for the past few years is an analytical tool that permits visual expression of the cerebral circulation profile in various cerebral diseases. The greatest drawback of SPECT is that the limited precision of spatial resolution makes in-

tracranial localization impossible. In hydrocephalus, however, the key subject to be studied is the profile of cerebral circulation around the ventricles of the brain. This suggests that CT is a difficult technique for display of the cerebral ventricles in three dimensions, whereas MRI will be more useful. For this reason, we attempted to establish the profile of cerebral circulation around the cerebral ventricles by producing combined 3D images with SPECT and MRI. In patients who had shunt surgery for hydrocephalus, a difference between pre- and postoperative cerebral circulation profiles was assessed by combined 3D SPECT and MRI. The shunt system used in this study was an Orbis-Sigma valve of the type with automatic cerebrospinal fluid volume adjustment rather than the variable-pressure-type Medos valve currently in more common use, because this latter device requires frequent changes in pressure and a change in pressure may be detected after an MRI procedure. This composite 3D MRI and SPECT imaging technique made it possible not only to establish an intracranial position in SPECT, but also to assess the profile of cerebral circulation around the cerebral ventricles. Analysis of these 3D composite images permits quantitative expression of brain volume in SPECT and extensive elucidation of the cerebral circulation profile in morphological detail. This analysis is, therefore, considered to make a substantial contribution to the development of functional images. Autoregulation in normal-pressure hydrocephalus. Z. Czosnyka, M. Czosnyka, J. Copeman, P. Smielewski, S. Piechnik, H. Whitehouse, J.D. Pickard (Cambridge) We have studied the relationship between resistance to CSF outflow and cerebral autoregulation in 40 patients presenting with ventricular dilatation and clinical symptoms of normal-pressure hydrocephalus (NPH). A computerised CSF infusion test was performed, during which blood flow velocity waveform was recorded by transcranial Doppler ultrasonography and arterial blood pressure (ABP) was monitored with a Finapress finger cuff. The resistance to CSF outflow was calculated as an absolute increase in intracranial pressure (ICP; interpolated over vasogenic waves) divided by the infusion rate (1.5 ml/min in most cases). Dynamic autoregulation was assessed as a correlation coefficient between slow waves (period from 20 s to 2 min) in mean blood flow velocity and cerebral perfusion pressure. A zero or negative coefficient (Mx) signifies good and positively disturbed autoregulation. Mean ICP increased during the test from 6 mmHg to 20 mmHg (P<0.0001). Average mean flow velocity did not change significantly. Mx was significantly correlated with the resistance to CSF outflow (R=–0.41; P<0.01). Patients presenting with ventricular dilatation may have either low (atrophy) or increased (NPH) resistance to CSF outflow. Increased outflow resistance correlates with preserved autoregulation. Those with a low resistance suggestive of brain atrophy have severely disturbed autoregulation in the middle cerebral artery territory as assessed by transcranial Doppler ultrasonography. Usefulness of CSF shunting for ameliorating symptoms reminiscent of normal-pressure hydrocephalus in ischemic cerebrovascular diseases. T. Mima, Y. Horikawa, K. Kobayashi, K. Kishida, T. Mori, K. Mori (Nankoku, Niigata) It is not easy to differentiate ventricular dilatation caused by idiopathic normal-pressure hydrocephalus (NPH) from the cerebral atrophy often found in patients with cerebral ischemia. It is known that hypertension and the MRI finding of prominent leukoaraiosis (periventricular high intensity in T2-weighted images) are risk factors for idiopathic NPH. Therefore, ischemic cerebrovascular diseases with dilated ventricles and idiopathic NPH may share a common pathogenesis based on atherosclerosis. Two groups working within the Research Committee on Intractable Hydrocephalus sponsored by the Ministry of Health and Welfare of Japan have performed shunting operations in patients with

127

ischemic cerebrovascular diseases with the aim of ameliorating NPH-like symptoms. Horikawa and her colleagues have conducted shunting operations in 68 cases of Binswanger disease during the past 16 years. Their follow-up study has revealed symptomatic amelioration in 61 cases (89.7%), and in at least 23 cases the improvement persisted for longer than 3 years. Mori and his colleagues have shunted three patients with ischemic cerebrovascular diseases. They were affected by right carotid artery occlusion (for 4 years before the shunt operation), multiple lacunae (for 3 years), and right middle cerebral artery stenosis (for 4 years). All three patients showed moderate recovery, particularly in gait disturbance. These results indicate that CSF shunting can be a useful treatment when NPH-like symptoms are found in patients with Binswanger and other ischemic cerebrovascular diseases. Chronic cerebral hypoperfusion based on atherosclerosis may play at least a partial role in the pathogenesis of idiopathic NPH.

biopsy specimens taken through the burr hole before passage of the ventricular catheter in each case. All pathological specimens were reviewed by the same neuropathologist. Follow-up ranged from 60 days to 2 years (mean 180 days). Histopathological features consistent with AD were found in biopsies from 14 (43%) of the 32 patients (minimal in 3, mild in 3, moderate in 7, severe in 1). Upon follow-up, 85% of these patients demonstrated improvement in gait, whereas 15% demonstrated no significant improvement or even a decline in gait. One patient with significant subjective improvement in gait according to family members did not demonstrate any objective improvement on formal assessment. Of the non-AD patients, 16 (89%) demonstrated gait improvement on follow-up. Despite coexistent AD, patients with hydrocephalus and gait disturbance demonstrate significant gait improvement following VP shunting. Such improvement may thus ameliorate the otherwise progressive downhill course of AD.

Shunt testing in vivo: a method based on the data from the UK Shunt Evaluation Laboratory. Z. Czosnyka, J. Copeman, R. Taylor, M. Czosnyka, J.D. Pickard (Cambridge)

Treatment of normal-pressure hydrocephalus with the dualswitch valve – radiological and clinical outcome. C. Sprung, C. Miethke, K. Shakeri, W.R. Lanksch (Berlin, Kleinmachnow)

One objective of the UK Shunt Evaluation Laboratory was to perform an independent testing of the hydrodynamic performance of all hydrocephalus shunts and provide systematic reviewing for neurosurgeons and patients. Valves were subjected to long-term testing in a computer-controlled rig to evaluate their pressure–flow performance both at baseline and under conditions mimicking phenomena which may alter CSF drainage in vivo. The operating pressures (Poperating) and hydrodynamic resistances (R) of all types of valves currently in use in the UK have been evaluated (Codman: Hakim-Precision, Hakim-Programmable, Uni-Shunt, Accuflo, Holter; Medtronic PS Medical: Delta, Flow Control, Lumboperitoneal; NMT: Orbis-Sigma, Omni-Shunt, Hakim Valve; Heyer-Schulte: In-line, Pudenz-Flushing, LowProfile; Radionics: ContourFlex; Sophy Programmable). Thirty-eight (38) patients who had improved in the past following shunting but experienced recent (in 1998) recurrence of their clinical symptoms were admitted for a computerised infusion test through the shunt pre-chamber or Ommaya reservoir implanted prior to shunting. The criterion used to detect shunt underdrainage was an increase in ICP during constant infusion to a level above Poperating + R × Infusion rate + 5 mmHg. The validity of this formula has been confirmed in a laboratory study. Twenty-two (22) patients with a variety of types of hydrocephalus met the criteria for shunt malfunctions, and their shunts were revised. Twelve (12) of them improved clinically. Ten (10) were later (1998–1999) tested again because of lack of improvement, and shunt underdrainage was again shown in 7 cases. Only 3 patients who were shown in CSF infusion testing to be underdraining did not improve following shunt revision. Shunt testing in vivo has good predictive power (86%).

Hydrostatic valves have considerable advantages over conventional differential-pressure valves in terms of overdrainage, but are thought to have an inherent tendency to underdrain or clog. The main goal of the study was to evaluate these functional complications in a cohort of hydrocephalic patients with homogeneous etiologies. Out of a series of 100 patients shunted with the hydrostatic gravitation-assisted Dual-Switch Valve (DSV), 72 suffered from normal-pressure hydrocephalus (NPH). This collective comprised 25 cases of idiopathic NPH and 47 patients with secondary NPH following hemorrhage or meningitis. The clinical status and CT scanning were assessed prior to operation, 3 and 6 months after surgery, and in all cases of suspected complications. The restoration of ventricular size was determined by the reduction in the Evans index. The outcome in our series was characterized by excellent and good clinical results accompanied by an only minimal reduction or even none at all in ventricular size in the majority of cases. In only 2 cases of NPH did we see clinical and radiological signs of overdrainage. Thus, the results in our collective are at odds with those obtained in comparable series recorded in the literature as far as subdural effusions are concerned. Underdrainage was suspected in 7 patients but could be proven only in 3 cases by implanting a low-pressure DSV. The clinical and radiological outcome in our series gives strong evidence that with a low opening pressure of the valve chamber for the recumbent position, the DSV makes it possible to avoid overdrainage-related problems without increasing the danger of underdrainage even in NPH. But the ideal “physiological” intraventricular pressure after surgery remains a matter of controversy, especially in NPH cases.

Shunting procedures for normal-pressure hydrocephalus; does coexistent Alzheimer disease affect the outcome? A.Y. Mogilner, H.L. Weiner, M.A. Leonard, J. Golomb, D.C. Miller, J.H. Wisoff (New York) It is traditionally believed that patients with normal-pressure hydrocephalus have a lower likelihood of improvement following ventriculoperitoneal (VP) shunting if they have a coexisting dementia, such as Alzheimer disease (AD). Brain biopsies were performed on a series of patients undergoing VP shunt placement for idiopathic hydrocephalus to determine postshunting functional outcome for patients with and without AD. In all, 32 patients (age range 59–89 years, mean 75 years) underwent brain biopsy and VP shunting. All patients had evidence of hydrocephalus seen on CT/MRI scan, and also a clear gait disorder. Patients underwent standardized neuropsychological assessment and video/computerassisted gait evaluation before and after surgery. Patients underwent standard frontal or occipital placement of VP shunts, with

Is endoscopic III ventriculostomy of benefit in tumor-related aqueduct stenosis? K.Y.C. Goh, R. Abbott (New York) To determine the value of endoscopic III ventriculostomy in the management of noncommunicating tumor-related hydrocephalus, all patients with neoplastic lesions causing aqueduct stenosis underwent a III ventriculostomy with a 3-mm neuroendoscope, either as a definitive procedure or secondary to a resection. A successful end-point was defined as shunt independence. The results were analyzed in a retrospective manner. Over a 4-year period from 1993 to 1997, 63 patients (mean age 10.9 years [SEM 1.33], range 3 months to 49 years) underwent III ventriculostomy in addition to tumor biopsy or resection. There were 25 patients with posterior fossa lesions, 10 of whom had intrinsic brain stem lesions. A further 38 patients had supratentorial tumors, comprising 16 tectal lesions and 11 thalamic, 7 intraventricular, 3 pineal and 1 sellar tumors. There was no morbidity or mortality related to the III ventriculostomy. At follow-up after a mean of 11.43 months (SEM 1.31, range 1–39 months), 31 (31/63, 49%) patients re-

128

quired ventriculoperitoneal (VP) shunting while 32 remained shunt-free (32/63, 51%). When these results were compared with those obtained in a series of 43 III ventriculostomy patients with pure aqueduct stenosis, in 15 of whom (15/43, 35%) shunting was required, the difference in shunt requirements just achieved statistical significance (P=0.048). Endoscopic III ventriculostomy can reduce shunt requirements in tumor-related hydrocephalus, but appears to be more effective in pure aqueduct stenosis.

Evolution of the concept of hydrocephalus in Russia during the last few decades (1948–1998). B.L. Lichterman (Moscow) The first book on hydrocephalus and its surgical treatment was published in Moscow in 1948 by Andrei Arendt, who was head of the Pediatric Neurosurgery Department at the Burdenko Neurosurgery Institute. This monograph was based on 100 cases of hydrocephalus. New approaches to classification, preoperative evaluation and surgery were suggested. Hydrocephalus was classified according to (1) time of appearance (inherited or acquired), (2) clinical course (chronic or acute), (3) etiology, (4) dilatation of ventricular system or subarachnoid space (internal or external), (5) communication of CSF spaces: whether open (communicating) or closed (noncommunicating), (6) CSF production and resorption (hypersecretory or nonresorptive) and (7) stage of disease (progressive or stabilized). Simultaneous ventricular and lumbar punc-

ture was suggested to differentiate communicating and noncommunicating forms of hydrocephalus (Arendt’s test). Occlusive forms were treated by suprachiasmal perforation of the lamina terminalis, while lumboperitoneal drainage with omentum was propagated for communicating hydrocephalus. The overall mortality rate in Arendt’s series was 10%. Arendt’s approach was later developed by his successor Vanda Rostotskaya, who suggested her own classification in occlusive hydrocephalus (1966). All interventions were divided into (1) operations without intervention in the ventricular system (decompressive trepanation, fenestration of the tentorium) and (2) operations on the ventricular system (external ventricular drainage and subarachnoid ventricular drainage via section of the corpus callosum, porencephaly, insertion of a silver tube, ventriculosubdural drain, perforation of the lamina terminalis). She claimed in 1966 that “the idea of ventriculoperitoneostomy is interesting but due to technical problems this operation cannot be recommended for use in everyday practice.” The first ventriculoatrial shunts were not introduced into practice in Russia until the 1970s, and then only on a very limited scale. One of the pioneers of CSF shunting in Russia, Boris Simernitskii, introduced a modern classification of hydrocephalus in 1989. According to this, hydrocephalus is classified by two parameters: (1) form (external or internal, with the latter subdivided into closed and open, and open hydrocephalus further subdivided into hypersecretory and nonresorptive forms) and (2) stage: progressive, subcompensated and compensated (stabilized).