Hydrocephalus encompasses a range of conditions. Hydrocephalus shunt technology: 20 years of experience from the Cambridge Shunt Evaluation Laboratory

Transcription

1 J Neurosurg 120: , 2014 AANS, 2014 Hydrocephalus shunt technology: 20 years of experience from the Cambridge Shunt Evaluation Laboratory Technical note Aswin Chari, B.M. B.Ch., Marek Czosnyka, Ph.D., Hugh K. Richards, Ph.D., John D. Pickard, M.Chir., F.Med.Sci., and Zofia H. Czosnyka, Ph.D. Department of Neurosurgery, Cambridge University Hospital National Health Service Trust, Cambridge, United Kingdom Object. The Cambridge Shunt Evaluation Laboratory was established 20 years ago. This paper summarizes the findings of that laboratory for the clinician. Methods. Twenty-six models of valves have been tested long-term in the shunt laboratory according to the expanded International Organization for Standardization 7197 standard protocol. Results. The majority of the valves had a nonphysiologically low hydrodynamic resistance (from 1.5 to 3 mm Hg/[ml/min]), which may result in overdrainage related to posture and during nocturnal cerebral vasogenic waves. A long distal catheter increases the resistance of these valves by 100% 200%. Drainage through valves without a siphon-preventing mechanism is very sensitive to body posture, which may result in grossly negative intracranial pressure. Siphon-preventing accessories offer a reasonable resistance to negative outlet pressure; however, accessories with membrane devices may be blocked by raised subcutaneous pressure. In adjustable valves, the settings may be changed by external magnetic fields of intensity above 40 mt (exceptions: ProGAV, Polaris, and Certas). Most of the magnetically adjustable valves produce large distortions on MRI studies. Conclusions. The behavior of a valve revealed during testing is of relevance to the surgeon and may not be adequately described in the manufacturer s product information. The results of shunt testing are helpful in many circumstances, such as the initial choice of shunt and the evaluation of the shunt when its dysfunction is suspected. ( Key Words hydrocephalus shunt cerebrospinal fluid hydrodynamics adjustable valve nonadjustable valve Hydrocephalus encompasses a range of conditions with an array of clinical symptoms, abnormal brain imaging, and derangements to CSF flow. 26 Shunting of CSF is one of the main treatment options. According to the International Organization for Standardization ([ISO] 7197; a standard that specifies the safety and performance requirements of shunts for hydrocephalus), a shunt is defined as an artificial connection between 2 compartments in the body, and a number of different shunt technologies are currently in clinical use ( iso.org/iso/catalogue_detail.htm?csnumber=38403). The shunt treatment industry has grown rapidly and had reached US $1B/year in Despite significant progress in the technology of shunting, recent studies have shown that shunt survival has not improved since the Abbreviations used in this paper: ICP = intracranial pressure; ISO = International Organization for Standardization; UK = United Kingdom. 1960s, 25 and that shunt revision rates have not improved despite the advent of adjustable shunts, 17 although more recent evidence suggests improvement in revision rates in adult patients with these types of shunts. 22 The recent example of Poly Implant Prothèse (PIP) breast implants 13 emphasizes the importance of objective testing of implantable devices. In addition to such safety issues, a detailed knowledge of the hydrodynamic properties of shunts is helpful for a number of reasons. First, it is important to recognize that the properties of a shunt may influence patient response to CSF shunting. It may also influence the choice of shunt; it is increasingly recognized that matching the patient s individual CSF hydrodynamic properties to that of the shunt may yield better outcomes and decrease complications. 7,27 Additionally, this information provides a basis for the testing of in vivo shunt function when patients present with shunt problems and complications. 21 More than 2 decades ago, Aschoff and colleagues 3,15 697

2 A. Chari et al. established the first shunt evaluation laboratory in Europe, in Heidelberg, Germany. The Cambridge Shunt Evaluation Laboratory was established by the senior author (Z.H.C.) and based on the experience of a medical student, Helen Adams, at the Wessex Neurological Centre, and a prototype of the first computer-controlled rig was constructed in Cambridge in The laboratory was established thanks to a grant from the Department of Health, and over 20 years, 26 shunts have been evaluated using the ISO 7197 standard protocol. Under the initial grant ( ), all shunts in use in the United Kingdom (UK) (16 types) were systematically evaluated, with blue reports printed by the Medical Devices Agency. New devices were tested as they appeared in the marketplace (or as prototypes or subsequent major redesigns of previously tested shunts), and these results have been published in the scientific press. This study updates these earlier findings 5 8 with data from newer shunt models, and summarizes these results in comparative tables. As our understanding of the pathophysiological mechanisms behind various types of hydrocephalus improves, it is hoped that these data will be useful to the clinician in deciding which shunt to use and how to assess its performance in vivo, 9,14 on a case-by-case basis. Methods Eighteen fixed-pressure and 8 adjustable hydrocephalus shunts (Table 1) have been tested in the Cambridge Shunt Evaluation Laboratory based on an expanding ISO 7197 standard since The hydrocephalus shunt testing rig has been described in detail in previous studies 5,6 and is illustrated in Fig. 1. Measurement was controlled by a standard IBMcompatible personal computer with software designed in-house, which precisely measures flow through the shunt and differential pressure. All samples used in the evaluation program were provided by manufacturers and were sterile and in the original packages. Three shunts of the same type were filled with deionized and deaerated water and mounted in 3 identical rigs. Pressure-flow performance curves were tested over a minimum of 28 days. The performance of shunts in altered conditions was subsequently studied. These included changing both the outlet level (23 cm, according to the ISO 7197 standard) and the depth at which the valve was submerged in the water tank (between 1 cm and 10 cm, reflecting variable external pressure); influence of a distal drain; bath temperature; pulsatile component of inlet pressure; and presence of small particles in the reagent (small red cell [10-mm] and larger tissue [25-mm] debris). With these maneuvers we were able to determine whether the shunt was susceptible to undesired alterations in CSF drainage after implantation. The valve s durability was tested by comparing the pressure-flow performance at the beginning and end of the protocol, which took approximately 40 days and involved daily testing as recommended by the international standard. In addition, the durability to shock waves of up to 200 mm Hg (simulating the maximal CSF pressure increase provoked by coughing), reversal pressure of the same magnitude, and behavior in a static magnetic field (low-field magnets [ mt] shifted above the valve and MRI studies) were also tested. For every shunt tested, operating pressure for a flow rate of 0.3 ml/min (equal to CSF formation rate) with a full-length distal catheter and hydrodynamic resistance for completely opened valve (with and without distal catheter) were measured, the clinical implications of which are described in the discussion. Hydrodynamic resistance was defined as and calculated by assessing the increment in differential pressure with increments of steady flow through the shunt within a range of ml/min; it is expressed using the units mm Hg/(ml/min). This can be calculated for pressures greater than the valve opening pressure and is a well-defined parameter for valves with a relatively linear pressure-flow performance curve. This value can then be compared with physiological resistance to CSF outflow of 5 10 mm Hg/(ml/min). 1 For every shunt and every performance level, socalled critical pressures were calculated. These are the thresholds above which, in an adequately functioning shunt, pressures should not increase during a constantrate infusion study. Critical pressures were measured as operating pressures (increased by 5 mm Hg: credit for averaged abdominal pressure for nonobese patients in horizontal position) for constant flow through through the valve with distal catheter, mimicking infusion study flow rates of 1 ml/min and 1.5 ml/min. 26 All results presented in the tables are means of the 3 valves tested in each rig. All valves had similar results reproduced with each one, and any instances of discrepancy between the 3 valves tested are clearly indicated. Results All data included in this section are results of experiments performed in the Cambridge (UK) Shunt Evaluation Laboratory over the last 20 years. The 26 valves are shown in Table 1 along with their functional mechanisms, which are elucidated as follows: 11 1) classic differential valve when opened according to differential pressure (inlet minus outlet) presents lowresistance channel for CSF drainage; 2) adjustable as above, plus opening pressure may be adjusted in vivo by external magnet; 3) gravitational classic differential valves that change shunt s performance (increase opening level) in vertical body position (these may be fixed pressure or adjustable); and 4) flow regulating not pressure but flow through the valve is stabilized, usually matching average CSF production rate (0.35 ml/min). Some shunts are integrated with siphon-preventing devices: membrane, gravitational, or flow regulating. Valves themselves may have different constructions: ball on spring, silicone membrane, miter, distal slit, or proximal slit. Basic hydrodynamic parameters, including so-called critical pressure levels for shunt testing in vivo (see Methods for definition) are shown in Table 2. For the adjustable valves with pressure levels (operating pressures for flow 698

3 Laboratory evaluation of hydrocephalus shunts TABLE 1: Shunts tested in the Cambridge Shunt Evaluation Laboratory since 1997* Shunt Name Manufacturer Functionality Construction Delta Valve Medtronic PS Medical cd + sp silicone membrane Low Profile Valve Heyer-Schulte (now Integra) cd + sp silicone membrane Pudenz Flushing Valve w/ ASD Integra cd + sp silicone membrane In-Line Valve Heyer-Schulte (now Integra) cd miter Contour Flex Radionic Medical Products cd silicone membrane Holter Valve Codman cd proximal slit Hakim Precision Valve Codman cd ball on spring Accu-Flo Codman cd silicone membrane Omnishunt Integra cd ball on spring Unishunt Codman cd distal slit CSF Flow Control Valve Medtronic PS Medical cd silicone membrane Hakim Valve Integra cd ball on spring CSF Lumboperitoneal Shunt Medtronic PS Medical cd distal slit SinuShunt CSF Dynamics cd miter Orbis Sigma Valve Integra flow-regulating moving diaphragm Diamond Valve Phoenix flow-regulating diamond aperture Dual Switch Aesculap-Miethke grav diaphragm & spring PaediGAV Aesculap-Miethke grav ball on spring Sophy Sophysa adj ball on spring Hakim Adjustable Codman adj ball on spring Strata Medtronic PS Medical adj + sp ball on spring Strata NSC Medtronic PS Medical adj ball on spring Polaris Sophysa adj ball on spring ProSA + mininav Aesculap-Miethke adj + grav ball on spring ProGAV Aesculap-Miethke adj + grav ball on spring Certas Codman adj + sp ball on spring * Adj = adjustable; ASD = antisiphon device; cd = classic differential shunt; grav = gravitational; sp = membrane or other (Siphon- Guard) siphon-preventing mechanism included. No longer in production. Recently recalled by manufacturer. 0.3 ml/min) and critical pressures that can be expressed as a linear function of setting, the relevant formulae are shown in Table 3. General findings are discussed in summary below. Pressure-Flow Performance For different construction mechanisms pressure-flow performance curves may have different shapes, from completely linear (above opening pressure) to absolutely nonlinear (for example flow-regulating valves like the Orbis Sigma or the Diamond). Convergence of pressure-flow measurement points forming the performance curve is better in ball-on-spring valves than in silicone-membrane valves in long-term evaluation. In some valves wide hysteresis of the performance curve can be noticed, suggesting that measured differential pressure is dependent on direction of the change of flow (increasing or decreasing) through the shunt (Fig. 2). Silicone-membrane valves show particularly significant hysteresis, whereas ball-onspring valves are less susceptible. Hydrodynamic Resistance This characteristic could not be evaluated for the Orbis Sigma or Diamond valves, which are flow-regulating valves, having theoretically infinite resistance within the flow-regulating operational range. The majority of the shunts show low resistance to flow (as low as 1.05 mm Hg/[ml/min]; see Table 2), which is substantially lower than physiological resistance to CSF outflow and is likely to result in overdrainage of CSF. Exceptions are the Medtronic Lumboperitoneal Shunt, the Codman Uni shunt, the SinuShunt, and, to some extent, the Holter Valve, the latter two of which have been discontinued. The resistance of the Unishunt, however, may be strongly affected by conditions at the distal end (that is, in the peritoneal cavity). Influence of Pulse Amplitude of Inlet Pressure (Intracranial Pressure) Any repetitive variations of proximal pressure have a tendency to decrease the nominal operating pressure of 699

4 A. Chari et al. Fig. 1. Diagram of the shunt testing rig. The shunt being tested is submerged in a water bath at a constant temperature at a defined depth (h). The working fluid (deionized and deaerated water) is supplied by the cylinder or infusion pump. A pulse pressure of controlled amplitude created by the pulse pressure generator can be added to the static pressure. A model of residual resistance to CSF outflow can be added before the device to study the shunt s performance in conditions mimicking the in vivo environment. Pressure before the shunt is measured with a proximal pressure transducer. Fluid flowing through the shunt is collected in a container placed on the electronic balance. Measurement is controlled by a standard IBM-compatible personal computer that reads and zeroes the balance periodically (every 15 seconds) to calculate the flow rate. This enables us to measure the weight of the outflowing fluid incrementally, which cancels the influence of fluid vaporization from the outlet container. The computer analyzes the pressure waveform from the pressure transducer and controls the rate of the infusion pump. The effects of changes in atmospheric pressure are compensated by using the atmospheric pressure transducer. Three-way stopcocks enable switching between different branches of the testing tubing. ICM is a software program. shunts with unidirectional valves. This may lead to overdrainage in situations in which there is a high respiratory magnitude or regular vasogenic intracranial pressure (ICP) waves (B waves 26 ). Influence of a Distal Catheter Long distal catheters increase resistance of the majority of classic differential valves toward normal physiological values. It is important to remember that the resistance of a catheter is the inverse of the fourth power of its inner diameter, and it is directly proportional to its length (Poiseuille law). Therefore, a 1-m-long catheter with a 1-mm inner diameter has a resistance of approximately 4 mm Hg/(ml/min), whereas a similar-length catheter of 1.2-mm inner diameter has a resistance of approximately 2 mm Hg/(ml/min). By comparison, the resistance of the typical ventricular catheter is no greater than 1 mm Hg/ (ml/min). Overdrainage of CSF Negative outlet pressure decreases operating pressure by the same value in all valves without a siphon-preventing mechanism, with the exception of Orbis Sigma and Diamond valves. When the resistance of the shunt system is low (4 6 mm Hg/[ml/min]), a negative outlet pressure of -15 mm Hg may accelerate the drainage rate to a nonphysiological value of 2 4 ml/min. Overdrainage may also occur when a low-resistance valve is subjected to pulsatile pressure (exceptions are Orbis Sigma, Diamond, and valves fitted with Codman SiphonGuard). Another rarely mentioned cause of overdrainage may be pumping of the proximal reservoir of the shunt (Fig. 3), which may be performed in emergency departments when shunt dysfunction is suspected. Influence of Devices Preventing Overdrainage These devices may be integrated with valves or chosen as accessories for shunt systems. Membrane devices may be blocked by external pressure, as shown below. Membrane and gravitational devices prevent only posture-related overdrainage. Flow-regulating devices, like SiphonGuard or Orbis Sigma valves, reduce overdrainage related to both posture and nocturnal vasocycling (spontaneous changes in CSF pressure overnight related to fluctuations either in arterial blood pressure or cerebral blood flow). However, they may result in raised ICP if the CSF formation rate is greater than 0.3 ml/min (Orbis Sigma), or when the device is locked in a high resistance state (SiphonGuard). This peculiar behavior of flow regulators is not always well reflected in shunt documentation. External Pressure All valves with membrane siphon-preventing devices are sensitive to external pressure. The external pressure (x) exerted by tense skin or scar on the skin increases operating pressure of the valve by a value of x. It means 700

5 Laboratory evaluation of hydrocephalus shunts TABLE 2: Numerical values of resistance, operating pressure, and critical levels for nonadjustable valves and for adjustable valves in which they cannot be expressed as formulae* Resistance (mm Hg/[ml/min]) Pressure (mm Hg) Shunt Name Levels w/o Distal Catheter w/ Distal Catheter Operating; for 0.3 ml/min Flow Critical; for 1.5 ml/min Infusion Critical; for 1 ml/min Infusion Delta Valve Low Profile Valve low med high Pudenz Flushing Valve w/ ASD low med high In-Line Valve low med high Contour Flex low med high Holter Valve low med high Hakim Precision Valve (Codman) very low low med-low med-high high Accu-Flo low med high Omnishunt low med high Unishunt regular low NA regular med NA regular high NA elliptical low NA elliptical med NA elliptical high NA CSF Flow Control Valve standard low standard med standard high contoured low contoured med contoured high bur hole low bur hole med bur hole high (continued) 701

6 A. Chari et al. TABLE 2: Numerical values of resistance, operating pressure, and critical levels for nonadjustable valves and for adjustable valves in which they cannot be expressed as formulae* (continued) Resistance (mm Hg/[ml/min]) Pressure (mm Hg) Critical; for 1.5 ml/min Infusion Shunt Name Levels w/o Distal Catheter w/ Distal Catheter Operating; for 0.3 ml/min Flow Critical; for 1 ml/min Infusion CSF Flow Control Valve (continued) button low-low button low button med Hakim Valve (Integra) very low low med high very high CSF Lumboperitoneal Shunt 1 level Orbis Sigma Valve 1 level very high very high 7 to SinuShunt 1 level Diamond Valve 1 level very high very high Dual Switch Valve horiz 5 cm H 2 O horiz 10 cm H 2 O horiz 13 cm H 2 O horiz 16 cm H 2 O PaediGAV horiz 4 cm H 2 O horiz 9 cm H 2 O horiz 15 cm H 2 O ProSa w/ MiniNAV horiz MiniNAV 0 cm H 2 O horiz MiniNAV 5 cm H 2 O horiz MiniNAV 10 cm H 2 O horiz MiniNAV 15 cm H 2 O vert NA NA NA Sophy adj (8 levels) see Table 3 see Table 3 see Table 3 Polaris 30 mm H 2 O mm H 2 O mm H 2 O mm H 2 O mm H 2 O Hakim Adjustable w/o Siphon adj (18 levels) see Table 3 see Table 3 see Table 3 Guard Strata adj (5 levels) see Table 3 see Table 3 see Table 3 Strata NSC ProGAV adj (0 20 cm H 2 O) see Table 3 see Table 3 see Table 3 Certas adj; 7 levels w/o SiphonGuard see Table 3 see Table 3 see Table 3 1 w/ SiphonGuard see Table 3 see Table 3 see Table 3 * Horiz = horizontal; med =medium; NA = not applicable; vert = vertical. Critical levels for testing in horizontal position. 702

7 Laboratory evaluation of hydrocephalus shunts TABLE 3: Characterization, operating, and critical pressures for adjustable valves, expressed as formulae Pressure (mm Hg) Operating; for Flow 0.3 ml/min Critical; for Infusion 1 ml/min Critical; for Infusion 1.5 ml/min Characteristic setting (in mm H 2 O)/ setting (in mm H 2 O)/ setting (in mm H 2 O)/13.56 Sophy: 3 main settings: 50, 110, & 170 mm H 2 O (coded low, med, & high); the other 5 settings (65, 80, 95, 130, & 150 mm H 2 O) are intermediate, which means less reliable for adjustment see Table 2 see Table 2 see Table 2 Polaris: 5 settings: 30, 70, 110, 150, & 200 mm H 2 O; hydrodynamic resistances increase w/ settings setting (in mm H 2 O)/ setting (cm H 2 O)/ setting (in mm H 2 O)/1.356 Hakim Adjustable: operational pressure may be programmed proportionally to the position of the magnet-driven rotor from 3 to 20 cm H 2 O in steps of 1 cm H 2 O (18 settings). 4 performance level performance level performance level Strata: 5 settings: 0.5, 1, 1.5, 2, & 2.5; equivalent to opening pressures 1.5, 4.2, 6.2, 8.2, & 10.4 mm Hg see Table 2 see Table 2 see Table 2 Strata NSC: 5 settings: 0.5, 1, 1.5, 2, & 2.5; equivalent to operating pressures 1.2, 2.1, 6.1, 9.7, & 12.6 mm Hg. Critical values for infusion tests in Table 2 ProGAV: settings from 0 to 20 cm H 2 O, w/ steps of 1 cm H 2 O in horiz position setting (cm H 2 O)/ setting (in cm H 2 O)/ setting (in cm H 2 O)/ setting setting setting 1.7; 2.9 setting 1.1 (w/ SiphonGuard) Certas*: 7 settings: 1 7 (pressure range mm H 2 O); 8th setting equivalent to shunt being switched off see Table 2 NA NA ProSa (adjustable only in vertical position): gravitationally compensating pressure counteracting overdrainage may be adjusted from 0 to 40 cm H 2 O magnetically * Recalled by manufacturer. that scar over a membrane-type antisiphon device that increases pressure on the device s membrane makes the shunt drain less than it would if there were no scar over the device. This a static change. Valve Adjustability All adjustable valves can be reset in vivo by applying an external magnetic field. Most valves cover a range of operating pressures from 0 to 20 cm H 2 O (0 to 15 mm Hg). The number of steps varies from 5 to 20 (Fig. 4, Table 3). In almost all valves the levels are equally spaced, except for the Codman Certas Valve, in which the last (8th) step is very high (above 40 cm H 2 O), with the intention of switching off the valve in vivo. In all valves except the Codman Hakim Adjustable Valve, verification of the setting may be conveniently performed without the need for imaging, by using an external compass placed over the valve. Both measurement and adjustability may be affected if the valve rotates under the skin. External Magnetic Field Magnetic fields can have an undesirable influence on adjustable valve settings. The Sophy, Strata, and Codman- Hakim Adjustable valves can be readjusted by relatively weak fields (approximately 40 mt). Newer valves (Polaris, ProGAV, ProSA, and Certas) have mechanisms intended to prevent accidental readjustments, even in MRI machines (up to 3T). All new valves tested were safe in MRI units up to 3T (translational and torque forces are safe, and heating is minimal [< 1 C]), but cause significant distortion of the MR image. An MR imaging study of the brain may therefore be unhelpful in these patients; however, MRI of other areas is possible. Particulate Matter Small particles (10 mm, mimicking red blood cells) in the reagent tend to increase shunts resistance or block them permanently. Large particles (25 mm) can block shunts, but commonly tend to open them permanently (suspend the balls above surface of cone or the membrane above the outlet orifice see Fig. 5). However, it is not clear whether microspheres mimic the presence of particles in CSF. Temperature Variations Variations in temperature (30 C 40 C) have no meaningful effect on shunt functioning. Reflux of CSF Reflux was not seen in any of the tested valves. Junction Stability All junctions were free from leaks and breaks when tested according to the ISO standard. Accompanying Documentation Booklets are provided with shunts as a standard. They differ widely, particularly in the selection of technical parameters provided. Values or ranges of operating 703

8 A. Chari et al. Fig. 2. Examples of different pressure-flow performance curves. A: The almost linear (above shunt s opening pressure) curve of a ball-on-spring valve. Its slope is equivalent to an inverse of the shunt s hydrodynamic resistance. The pressure flowperformance curve shows a lower gradient (greater resistance) after connection of distal catheter. B: The nonlinear pressureflow performance of a flow-regulating Orbis Sigma Valve. C: A Certas Valve with SiphonGuard showing wide hysteresis. Arrows mark the direction of change in flow through the shunt system. Fig. 3. Chart showing pressures recorded proximal to the shunt with a large reservoir. The reservoir pressure decreased quickly to -70 mm Hg with the onset of pumping of the reservoir. This experiment was performed in the laboratory. 704

9 Laboratory evaluation of hydrocephalus shunts Fig. 4. Chart showing distribution of operating pressures in the adjustable Codman Certas Valve. pressures are usually provided. Hydrodynamic resistance is never provided, and usually the term resistance is confused with opening or operating pressure. Discussion Summary and Clinical Relevance of Results Tables 2 and 3 provide a concise summary of data from our evaluation of 18 fixed-pressure and 8 adjustable shunts at the Cambridge Shunt Evaluation Laboratory. The data validate and expand on the details supplied by the manufacturers. These data are useful in clinical practice in a number of different circumstances. The first and currently most relevant clinical application of laboratory data is in interpretation of in vivo shunt testing. The critical pressure for various infusion rates 4,15 and the hydrodynamic resistance can be compared with results of infusion tests into shunt prechambers described previously 8,16,21,26 to assess potential for shunt dysfunction. A constant-rate infusion test performed into shunt prechamber or lumbar space can reveal whether a shunt is fully patent, or partially or fully blocked. During such a test, with a ml/min constant infusion, a critical level of pressure should not be exceeded (Fig. 6). This level can be expressed by a formula for shunts that have a linear pressure-flow performance curve (for levels above the valve s opening pressure): critical pressure (mm Hg) = operating pressure (mm Hg) + (resistance infusion rate) + 5 mm Hg. This so-called critical pressure has previously been shown to be the best guide for shunt functioning, with a > 90% positive predictive value. 8 It is evaluated under the assumption that abdominal pressure is not more than 5 mm Hg. In obese patients or pregnant women, this value should be increased accordingly. 23 For shunts with nonlinear performance curves (for example Orbis Sigma or Diamond valves) the formula above does not apply. A software package commonly used in our unit for monitoring, ICM+ ( contains a full database created from these tests and performs the comparisons semiautomatically. The other clinical applications, namely in the diagnosis of hydrocephalus and use of the hydrodynamic properties of the patient and shunt to guide shunt choice, are novel concepts that require further investigation prior to routine clinical use. However, the library of postmarketing surveillance data provided by the Cambridge Shunt Evaluation Laboratory will be helpful in evaluating these hypotheses and trying to improve outcomes in patients with hydrocephalus. The point about the utility of MRI studies is particularly important to note in the context of shunt malfunc- Fig. 5. Chart showing the difference between pressure-flow performance in normal conditions (A) and after injection of large (25-mm) microspheres (B) in the Heyer-Schulte Low Profile Shunt. Flow increased immediately following injection and closing pressure decreased to 0 mm Hg. h = hours. Fig. 6. Example of an infusion study in a shunt-treated patient from the hydrocephalus clinic at our institution. Baseline pressure may be above or below operating pressure. It increases after the start of the infusion (gray area). If it rises above the critical pressure assessed for every shunt (operating pressure + [resistance of the valve with distal catheter infusion rate] + 5 mm Hg), the shunt is underdraining. Details of in vivo testing of shunt function via infusion study have been previously reported (Czosnyka et al. 8 and Weerakkody et al.). 705

10 A. Chari et al. tion, especially given the recent evidence that shunt surveillance decreases unscheduled visits to the emergency department and clinic. 4 Despite the advent of new lowdose protocols for CT evaluation of shunts, 14 the impact of CT scans on the incidence of leukemia and brain tumors 20 makes the ability of a shunt to be evaluated using the MRI modality crucial to the future of shunt development. Study Limitations The methodology is not without its shortcomings. The system created here aims to model the conditions in vivo, but given the limited and rapidly progressing understanding of CSF hydrodynamics in health, the model may not accurately reflect the in vivo situation in health and disease. For example, many of the measurements are made at steady state. Recent findings suggest that ICP fluctuates according to respiration and with vasogenic waves (accounting for nocturnal vasogenic cycling), and although these were tested for and possibly account for part of the overdrainage experienced in vivo, the integration of this into the model might alter the measurements of hydrodynamic resistance or opening pressure to reflect the behavior in vivo more accurately. Although efforts at in vivo modeling have been made in a past study, 21 the main limitation of the present study is the lack of data correlating these in vitro approaches to actual shunt function in vivo. It is also worthwhile to stress that the methodology used in our shunt laboratory is only one of multiple possible options for shunt testing. Other laboratories using different methods may produce equally precise or better results. In general, any independent testing of the properties of shunts seems to be useful in improving understanding of the management of hydrocephalus. The Past, Present, and Future of Shunt Treatment Shunt placement for the treatment of hydrocephalus is a relatively recent concept, with the first effective shunts being reported by Nulsen and Spitz in Since then the industry has grown considerably, with more than 3000 shunt operations being performed yearly in the UK, according to data from the UK Shunt Registry. Despite significant progress in the technology of shunt treatments, studies have shown that shunt survival has not improved since the 1960s, 25 and that shunt revision rates have not improved despite the advent of programmable shunts, 17 although there is some evidence to show the efficacy of these devices in adults. 22 Table 1 lists all shunts tested in our laboratory over a long span of time. It contains all original prototypes but also designs that were introduced secondarily to improve shunt performance. Some of the types have been discontinued (Holter Valve and SinuShunt). The adjustable Strata Valve has a version without a siphon-control device (Strata NSC), leaving the option for using the shunt without or with another type of siphon-control device. The first adjustable valve historically (Sophy) has been rereleased as the adjustable, MRI-resistant (in terms of accidental readjustment in a strong magnetic field) Polaris valve. Randomized controlled trials of certain shunts have shown that blockage occurs in 31.4% of patients at 1-year follow-up, whereas overdrainage occurs in 3.5% of patients. 9 These high rates are largely because of our currently limited understanding of the hydrodynamics of CSF in health and disease. It is therefore not known whether the shunt properties are optimal at creating a near-physiological system that is stable over time and stable when subject to a variety of internal and external forces that could potentially alter CSF drainage. In 2000, Drake and colleagues 10 surmised that 1-year failure rates could be as low as 5% by Recent data show that we are still some way from achieving this. 22 The future of shunt treatment depends on advancements in two main fields. First, it relies on a more in-depth understanding of CSF hydrodynamics, both in health and disease. 24 It also relies on the understanding of the hydrodynamic properties of the shunts themselves. Given recent advances in the mathematical modeling of CSF dynamics, it is reasonable to envision a future in which shunts are selected on a case-by-case basis, choosing the appropriate shunt based on the patient s native CSF hydrodynamic properties and how this might be altered in disease. Current clinical practice suggests that we are still some way off from being able to predict outcomes from a patient s CSF hydrodynamic characteristics 2,28 or shunt choice. 12 Unlike pharmacological agents, medical devices are not subjected to the rigors of the clinical trial process before introduction to the marketplace. It is therefore important that their performance in the laboratory and in vivo is carefully monitored. Apart from shunt evaluation in the laboratory, an effective way of achieving this is through national registries. The UK Shunt Registry has been active since May 1994, and now holds data on over 65,000 shunt and shunt-related procedures. The valves and shunts currently used in the UK appear to work to the manufacturers specifications, and hardware failures are rare. Recent years have seen the increased popularity of adjustable valves and the introduction of antibiotic-impregnated catheters. The performances of both of these products are continuously being evaluated by the UK Shunt Registry 22 and, in combination with these data from the Cambridge Shunt Evaluation Laboratory, should provide robust data on the efficacy of various shunt systems in specific pathophysiological circumstances. Conclusions The behavior of a valve revealed during testing may not be adequately described in the manufacturer s product information. The results of shunt testing are useful both in the choice of shunt and during shunt evaluation when a patient presents with shunt dysfunction. Disclosure Prof. Pickard was in the past a member of the Scientific Advisory Board of Codman and Medtronic PS Medical. Dr. M. Czosnyka signed in the past agreements for paid lectures with Integra and Codman. He is also a consultant for Codman and for Johnson and Johnson, and has received clinical or research support for this study from Codman, Medtronic, Miethke, Sophysa, and Integra. Dr. Z. Czosnyka received clinical or research support for this study from 706

11 Laboratory evaluation of hydrocephalus shunts Johnson and Johnson, Medtronic, Sophysa, Integra, and Miethke. The ICM+ software ( is licensed by University of Cambridge (Cambridge Enterprise Ltd), and Dr. M. Czosnyka has financial interest in a fraction of licensing fee. Financial support was provided by the Department of Health (UK) Medical Device Agency, National Institute for Health Research (UK). Research grant agreements between manufacturers (Codman, Medtronic, Aesculap, and Sophysa) and the University of Cambridge were signed for a fee of approximately 6000 per evaluation program. Author contributions to the study and manuscript preparation include the following. Conception and design: M Czosnyka, Chari, Pickard, Z Czosnyka. Acquisition of data: M Czosnyka, Z Czosnyka. Analysis and interpretation of data: M Czosnyka, Chari, Richards, Z Czosnyka. Drafting the article: Chari. Critically revising the article: all authors. Reviewed submitted version of manuscript: all authors. Approved the final version of the manuscript on behalf of all authors: M Czosnyka. Statistical analysis: Richards. Administrative/technical/material support: Pickard, Z Czosnyka. Study supervision: M Czosnyka, Pickard. References 1. Albeck MJ, Børgesen SE, Gjerris F, Schmidt JF, Sørensen PS: Intracranial pressure and cerebrospinal fluid outflow conductance in healthy subjects. J Neurosurg 74: , Anile C, De Bonis P, Albanese A, Di Chirico A, Mangiola A, Petrella G, et al: Selection of patients with idiopathic normalpressure hydrocephalus for shunt placement: a single-institution experience. Clinical article. J Neurosurg 113:64 73, Aschoff A, Krämer P, Benesch C, Klank A: Shunt-technology and overdrainage a critical review of hydrostatic, programmable and variable-resistance-valves and flow-reducing devices. Eur J Pediatr Surg 1 (Suppl 1):49 50, Chern JJ, Muhleman M, Tubbs RS, Miller JH, Johnston JM, Wellons JC III, et al: Clinical evaluation and surveillance imaging in children with spina bifida aperta and shunt-treated hydrocephalus. Clinical article. J Neurosurg Pediatr 9: , Czosnyka M, Czosnyka Z, Whitehouse H, Pickard JD: Hydrodynamic properties of hydrocephalus shunts: United Kingdom Shunt Evaluation Laboratory. J Neurol Neurosurg Psychiatry 62:43 50, Czosnyka Z, Czosnyka M, Richards H, Pickard JD: Hydrodynamic properties of hydrocephalus shunts. Acta Neurochir Suppl 71: , Czosnyka Z, Czosnyka M, Richards HK, Pickard JD: Laboratory testing of hydrocephalus shunts conclusion of the U.K. Shunt evaluation programme. Acta Neurochir (Wien) 144: , Czosnyka ZH, Czosnyka M, Pickard JD: Shunt testing in-vivo: a method based on the data from the UK shunt evaluation laboratory. Acta Neurochir Suppl 81:27 30, Drake JM, Kestle JR, Milner R, Cinalli G, Boop F, Piatt J Jr, et al: Randomized trial of cerebrospinal fluid shunt valve design in pediatric hydrocephalus. Neurosurgery 43: , Drake JM, Kestle JR, Tuli S: Cerebrospinal fluid shunt technology. Clin Neurosurg 47: , Drake JM, Sainte Rose C: The Shunt Book, ed 1. Cambridge, MA: Blackwell Science, Haberl EJ, Messing-Juenger M, Schuhmann M, Eymann R, Cedzich C, Fritsch MJ, et al: Experiences with a gravity-assisted valve in hydrocephalic children. Clinical article. J Neurosurg Pediatr 4: , Horton R: Offline: A serious regulatory failure, with urgent implications. Lancet 379:106, Jończyk-Potoczna K, Frankiewicz M, Warzywoda M, Strzyżewski K, Pawlak B: Low-dose protocol for head CT in evaluation of hydrocephalus in children. Pol J Radiol 77:7 11, Kremer P, Aschoff A, Kunze S: Therapeutic risks of anti-siphon devices. Eur J Pediatr Surg 1 (Suppl 1):47 48, Malm J, Lundkvist B, Eklund A, Koskinen LO, Kristensen B: CSF outflow resistance as predictor of shunt function. A longterm study. Acta Neurol Scand 110: , Notarianni C, Vannemreddy P, Caldito G, Bollam P, Wylen E, Willis B, et al: Congenital hydrocephalus and ventriculoperitoneal shunts: influence of etiology and programmable shunts on revisions. Clinical article. J Neurosurg Pediatr 4: , Nulsen FE, Spitz EB: Treatment of hydrocephalus by direct shunt from ventricle to jugular vein. Surg Forum 2: , Patwardhan RV, Nanda A: Implanted ventricular shunts in the United States: the billion-dollar-a-year cost of hydrocephalus treatment. Neurosurgery 56: , Pearce MS, Salotti JA, Little MP, McHugh K, Lee C, Kim KP, et al: Radiation exposure from CT scans in childhood and subsequent risk of leukaemia and brain tumours: a retrospective cohort study. Lancet 380: , Petrella G, Czosnyka M, Smielewski P, Allin D, Guazzo EP, Pickard JD, et al: In vivo assessment of hydrocephalus shunt. Acta Neurol Scand 120: , Richards H, Seeley H, Pickard JD: Are adjustable valves effective in all ages of patient? Data from the UK Shunt Registry. Cerebrospinal Fluid Res 7 (1 Suppl):S40, 2007 (Abstract) 23. Samuels P, Driscoll DA, Landon MB, Ludmir J, McKrisky PJ, Mennuti MT, et al: Cerebrospinal fluid shunts in pregnancy. Report of two cases and review of the literature. Am J Perinatol 5:22 25, Schuhmann MU, Sood S, McAllister JP, Jaeger M, Ham SD, Czosnyka Z, et al: Value of overnight monitoring of intracranial pressure in hydrocephalic children. Pediatr Neurosurg 44: , Stein SC, Guo W: Have we made progress in preventing shunt failure? A critical analysis. J Neurosurg Pediatr 1:40 47, Weerakkody RA, Czosnyka M, Schuhmann MU, Schmidt E, Keong N, Santarius T, et al: Clinical assessment of cerebrospinal fluid dynamics in hydrocephalus. Guide to interpretation based on observational study. Acta Neurol Scand 124:85 98, Williams MA, McAllister JP, Walker ML, Kranz DA, Bergsneider M, Del Bigio MR, et al: Priorities for hydrocephalus research: report from a National Institutes of Health-sponsored workshop. J Neurosurg 107 (5 Suppl): , Woodworth GF, McGirt MJ, Williams MA, Rigamonti D: Cerebrospinal fluid drainage and dynamics in the diagnosis of normal pressure hydrocephalus. Neurosurgery 64: , 2009 Manuscript submitted October 8, Accepted November 19, Please include this information when citing this paper: published online January 3, 2014; DOI: / JNS Address correspondence to: Marek Czosnyka, Ph.D., Academic Neurosurgical Unit, Addenbrooke s Hospital, Hills Rd., Cambridge CB2 0QQ, UK. mc141@medschl.cam.ac.uk. 707