HYDROCEPHALUS

THE SURGICAL MANAGEMENT OF

PEDIATRIC HYDROCEPHALUS
James M. Drake, F.R.C.S.C. THE SURGICAL MANAGEMENT of hydrocephalus has undergone incredible changes
Division of Neurosurgery, over the past generation of neurosurgeons, including dramatic improvements in
Hospital for Sick Children, imaging, especially computed tomographic scanning and magnetic resonance imag-
University of Toronto,
Toronto, Canada ing, and remarkably innovative advances in cerebrospinal fluid valve technology,
complex computer models, and endoscopic equipment and techniques. In terms
Reprint requests: of overall patient outcomes, however, one could conclude that things are a little
James M. Drake, F.R.C.S.C.,
Hospital for Sick Children,
better, but “not much.” This frustrating yet fascinating dichotomy between techno-
555 University Avenue, logical advancements and clinical outcomes makes hydrocephalus, first described
Toronto, Canada, M5G 1X8. by the ancients, as one of the most understated and complex disorders that neuro-
Email: james.drake@sickkids.ca
surgeons treat. The challenge to the next generation of neurosurgeons is to solve
Received, May 19, 2007.
this vexing problem through better understanding of the basic science, improved
Accepted, October 25, 2007. computer models, additional technological advances, and, most importantly, a
broad-based, concerted multidisciplinary attack on this disorder. This review focuses
on the evolution of surgery for hydrocephalus over the last 30 years, the current
ONLINE
DIGITAL state of the art of hydrocephalus treatment, and what appear to be the most prom-
VIDEO
ising future directions.
KEY WORDS: Cerebrospinal fluid shunt, Endoscopic third ventriculostomy, Hydrocephalus, Outcomes,
Surgery

Neurosurgery 62[SHC Suppl 2]:SHC633–SHC642, 2008 DOI: 10.1227/01.NEU.0000296952.38148.49

Incidence resonance imaging provide reasonably detailed fetal brain

anatomy, detect malformations, have normative data for ven-

T
he incidence of pediatric hydrocephalus has declined in
many developed countries (13, 14). The incidence of open tricular size, and allow serial investigation during gestation
neural tube defects has dropped precipitously as a result of (Fig. 1) (65). Although attempts at in utero shunting have
largely been abandoned because of what appeared to be a lack
maternal folate supplementation (61), antenatal screening, and
of efficacy or failure to appreciate the most appropriate cases,
termination of pregnancy based on superior antenatal imaging
the quest for this type of surgery lingers (6, 8). In utero genetic
with ultrasound and magnetic resonance imaging. The incidence
testing by amniocentesis has largely been directed toward
of cerebrospinal fluid (CSF) shunting in open neural tube defects,
detecting chromosomal abnormalities in patients who are iden-
formerly reported to be as high as 90%, has also declined, possi-
tified as having a brain abnormality such as hydrocephalus;
bly as a result of a general, more conservative approach, and also
however, this is clearly an area of future growth. X-linked
the selection of lower-grade lesions for delivery with a lower hydrocephalus is now known to be part of the Crash spectrum
requirement for shunting (87). A prospective, randomized trial of with a mutation on chromosome Xq28 (92).
in utero repair is also under way (10, 23), in which the primary High-resolution magnetic resonance imaging of postnatal
outcome is the incidence of shunt-dependent hydrocephalus, life has provided clues to the etiology of hydrocephalus, which
thought to be possibly reduced by in utero repair (87). Similarly, in the past would have been labeled as idiopathic; these include
better perinatal management of prematurity has reduced the intraventricular hemorrhage, aqueduct stenosis, migrational
incidence of intraventricular hemorrhage (25). Antenatal screen- abnormalities, and so on. Serial imaging has allowed the evo-
ing, genetic testing, and pregnancy termination have also lution of hydrocephalus to be followed carefully, with the
reduced the incidence of other congenital malformations of the recognition that not only is ventriculomegaly often stable, but,
brain that cause hydrocephalus. particularly in intraventricular hemorrhage of prematurity,
often resolves (32). Extraventricular obstructive hydrocephalus
Diagnosis is now recognized to represent, almost universally, benign
Perhaps the greatest impact in hydrocephalus management pericerebral collections of infancy that are usually familial, re-
has been that of imaging. Antenatal ultrasound and magnetic solve with time, and almost never require treatment (56).

NEUROSURGERY VOLUME 62 | NUMBER 2 | FEBRUARY 2008 SUPPLEMENT | SHC633

DRAKE

A B

FIGURE 1. A, in utero magnetic resonance imaging scan showing a fetus

at 20 weeks gestation with hydrocephalus secondary to a posterior fossa
arachnoid cyst. Axial images reveal a posterior fossa arachnoid cyst dis-
placing structures to the left, occluding the aqueduct. B, more superiorly,
there is significant ventricular dilation.

FIGURE 2. Shunt complication-free survival curves showing a prospec-

The complex relationships among cysts, tumors, abscesses,
tive, randomized shunt design trial, which compared standard valves with
and the ventricular system are now easily appreciated, as are the Orbis Sigma Valve (Cordis, Miami, FL) and the Delta valve
the relationships between the ventricular walls and vascular (Medtronic, Goleta, CA). There were no differences between the three valve
structures in the cranial base, as must be considered when con- designs (from, Drake JM, Kestle J, Milner R, Cinalli G, Boop F, Piatt J,
templating an endoscopic third ventriculostomy. Changes in Haines S, Schiff SJ, Cochrane DD, Steinbok P, MacNeil N: Randomized
ventricular size using standardized measurements or com- trial of cerebrospinal fluid shunt valve design in pediatric hydrocephalus.
puter-generated, three-dimensional volumetric volume calcula- Neurosurgery 43:294–305, 1998 [20]).
tions can be used to detect small differences in size (46, 59).
Similarly, the resolution of periventricular edema, persistence despite the ability to adjust the valve for over- or under-
of flow voids through the floor of the third ventricle after endo- drainage, avoiding a shunt revision for that reason, the failure
scopic third ventriculostomy (ETV), is also accomplished easily. rates were no different (Fig. 3) (62). A single-arm prospective
study of an adjustable Delta valve, the Strata valve, suggested
Treatment
its failure rate was no different than any of the other previ-
The major changes in treatment have been the use of ously studied valves in clinical trials (Fig. 4) (43). Perhaps most
advanced CSF valve design, broad application of endoscopic astounding, when ventricular size was measured pre- and
techniques, including ETV, and a general, more conservative postoperatively, comparing three very different valves, there
approach given the recognition of the scale of CSF shunt com- was no difference over time, which indicates that the valves
plications. The important recognition that CSF shunt failure is were not performing as the engineers had predicted (85). The
a time-related event (68) and, if accounted for by actuarial tech- prospective studies have allowed accurate measurement of
niques, was depressingly poor, allowed for a much easier com- anticipated shunt failure—40% at 1 year, including an approx-
parison of shunt failure between centers. imate 10% shunt infection rate (20, 41, 43, 62).
New valve designs continue to appear with similar claims
CSF Shunts unsubstantiated by controlled data (60, 80). That neurosur-
The first spring, ball, and diaphragm valves were super- geons would embrace these reports, or use these valves,
seded by remarkably clever and innovative valve designs reflects a probable mixture of a “hope springs eternal” mental-
including anti-siphon devices (63), horizontal–vertical valves ity, gullibility, and a failure of the United States Food and Drug
(21), flow-controlled valves (67), adjustable valves (5), and Administration to regulate devices with the same scrutiny as
many more. Although each new valve was heralded as a sig- novel pharmaceuticals.
nificant advance that was more physiological in function with
improved results in uncontrolled studies, the prospective and CSF Shunt Infection
randomized trials told a very different story; they were all Shunt infection remains an intractable and serious complica-
negative. When the standard valves, a novel flow-controlled tion of CSF shunting. Although the initial reports in prospective
valve (Orbis Sigma; Cordis, Miami, FL), and an updated anti- studies of shunt infection rates of 8% were greeted with disbe-
siphon valve (Delta Valve; Medtronic PS Medical, Goleta, CA) lief, or comments about participating surgeons “not washing
were compared in a prospective, randomized trial, there was their hands,” innumerable studies since that time have
no difference (Fig. 2) (20, 40). Similarly, when the adjustable reported rates as high or higher (20, 41, 43, 62). Meta-analyses
(programmable) Medos valve (Codman/Johnson & Johnson, indicate that prophylactic antibiotics reduce the rate of shunt
Raynham, MA) was compared with all other designs, and infection (31, 51). Identification of risk factors has suggested

SHC634 | VOLUME 62 | NUMBER 2 | FEBRUARY 2008 SUPPLEMENT www.neurosurgery-online.com

 SURGICAL MANAGEMENT OF PEDIATRIC HYDROCEPHALUS

studies have had conflicting results (39, 73). An ongoing, but as

yet unsubstantiated, concern has been the selection of gram-
negative organisms or antibiotic-resistant organisms.
The treatment of shunt infection has generally been stan-
dardized to removal of shunt hardware, insertion of a tempo-
rary external ventricular drain allowed by reinternalization of
the shunt system. There is a great deal of variability in terms of
duration of antibiotic therapy, use of intrathecal antibiotics,
and CSF parameters consistent with sterilization. A recent
report reinvigorated the debate by reporting a high success
rate when leaving the shunt hardware in place and using
intrathecal plus systemic antibiotics, albeit through a separate
CSF reservoir, which in some cases needed to be implanted
before treatment, thereby negating the potential reduced surgi-
FIGURE 3. Shunt complication-free survival curves showing the prospec- cal impact of this technique (7). Alarmingly, the reinfection rate
tive, randomized Medos trial comparing the Medos valve (Codman, using the standard protocols is approximately 25% when meas-
Raynam, MA) with all other valve designs. There was no difference in out- ured by prospective observation (42, 50). CSF shunt infection
come, and the curves overlap (from, Pollack IF, Albright AL, Adelson remains perhaps the most vexing of complications. Antibiotic
PD: A randomized, controlled study of a programmable shunt valve ver- impregnation seems to be a promising technique, but the long-
sus a conventional valve for patients with hydrocephalus. Neurosurgery term results are pending.
45:1399–1411, 1999 [62]).
Shunt Insertion Technique
Several techniques for inserting shunts more accurately,
including the use of stereotaxis (53, 93), endoscopy, and ultra-
sound (75), have been developed, largely to prevent errant
catheters (which can end up in astonishing locations). The only
technique subjected to a prospective, randomized trial was endo-
scopic insertion using miniature scopes, which were passed
down the lumen of the ventricular catheter (41). The technique
was of no benefit, and if anything, the failure rates were higher
(Fig. 5). Interestingly, the final location of the shunt catheter was
no better when using the endoscope than when using traditional
landmarks, suggesting that either the catheter moved after
removal of the shunt scope, or surgeons were unable to take
advantage of the technology. In a post hoc analysis, a catheter
placed away from the choroid plexus had a lower failure rate,
FIGURE 4. Prospective cohort study showing the
suggesting that final catheter position within the ventricular sys-
Strata valve (Medtronic, Goleta, CA), a programmable tem remains an important concept. Stereotaxis has been used
version of the Delta valve. Complication-free survival primarily for patients with small ventricles. Frame-based and
rates appeared very similar to other prospective, ran- frameless systems and systems using magnetic field technology
domized studies, suggesting that the valve offered little (in which the tip of the ventricular catheter can be tracked) have
if anything in terms of improved outcome (from, Kestle all been described. An advantage of ultrasound is that the trajec-
JR, Drake JM, Cochrane DD, Milner R, Walker ML, tory of the catheter is observed in real time, and changes in ven-
Abbott R 3rd, Boop FA: Endoscopic Shunt Insertion tricular size, indentation of the ependymal wall, and even posi-
Trial participants: Lack of benefit of endoscopic ven- tion of the ventricular catheter holes can be visualized.
triculoperitoneal shunt insertion: A multicenter ran-
The widespread use of laparoscopy in general surgery has
domized trial. J Neurosurg 98:284–290, 2003 [41]).
gradually migrated to shunt surgery, in which distal shunt
catheters can be implanted under laparoscopic vision and frac-
that age, holes in the surgical gloves, and postoperative CSF tured catheters can be removed (35, 44, 72, 94). This technique
leak may be important (47). A follow-up study has shown some is perhaps best used in patients with dense peritoneal adhe-
evidence that double gloving may help reduce shunt infection sions or obese patients, and routine use may allow neurosur-
(18, 86). Shunts impregnated with antibiotics, typically geons to develop the expertise to use these techniques on their
rifampin and clindamycin, have been reported to reduce infec- own. Similarly, for the rarely used cardiac catheters, percuta-
tion in one prospective, randomized trial, although the trial neous fluoroscopic controlled access is relatively straightfor-
had a remarkably small number of patients and a very high ward, and it reduces the extent and difficulty of cutting down
infection rate in the control group (∼16%) (28). Uncontrolled and securing a tributary of the internal jugular vein (9, 15).

NEUROSURGERY VOLUME 62 | NUMBER 2 | FEBRUARY 2008 SUPPLEMENT | SHC635

DRAKE

been developed for children older than 5 years, validated, and

applied to a reasonably large cohort of shunt-dependent
patients (48, 49). Shunt infection and epilepsy appear to have
the greatest negative impact on quality of life, but there are
also factors that might not have been anticipated, such as dis-
tance from the treatment center. Additional work in this area,
for patients with ETV as well, is clearly important.

Endoscopic Techniques
(see video at web site)
The advent of high-resolution, charge-coupled device cam-
eras combined with improved light sources and the prolifer-
ation of endoscopic techniques in other areas of surgery has
FIGURE 5. Prospective, randomized trial showing the led to widespread application of these techniques in neuro-
use of an endoscope to place the intraventricular surgery. Germane to pediatric hydrocephalus are the ETV and
catheter during shunt insertion. Complication-free sur- cyst marsupialization. Although ETV was first attempted in
vival rate was not different with the use of the endo- the 1920s, it fell out of favor with the advent of CSF shunts in
scope (from, Kestle JR, Drake JM, Cochrane DD, the 1950s (1). A combination of the realization of the extent of
Milner R, Walker ML, Abbott R 3rd, Boop FA: Lack of
shunt complications, improved imaging to identify patients
benefit of endoscopic ventriculoperitoneal shunt inser-
with obstructive hydrocephalus, and improved endoscopic
tion: A multicenter randomized trial. J Neurosurg
98:284–290, 2003 [41]). technology has led to its reintroduction.
The operation is typically performed using a rigid endo-
scope, entering through a precoronal burr hole, and opening
the floor of the third ventricle anterior to the mamillary bodies
Outcomes and the basilar bifurcation posterior to the dorsum sellae.
Shunt failure has typically been defined as a return to the Typically, the floor is perforated with a blunt instrument, and
operating room to treat shunt obstruction or infection or the opening is expanded with either a balloon or forceps. On
another complication, i.e., subdural hematoma (20). Repeated occasion, in patients with a transparent floor of the third ven-
shunt failure has been less studied but appears to be at least a tricle, an opening is made directly onto the dorsum sellae using
frequent event in children, and the fourth shunt failure may electrocautery. This technique is used particularly when there is
happen at an even shorter time interval (84). The risk factors for little space between the basilar artery and the clivus. Patients
repeated shunt failure are not clearly defined but may be age, with large ventricles, an occluded aqueduct, a brainstem, and
etiology, and, interestingly, the time interval from the previous a basilar artery displaced posteriorly from the clivus seem to be
shunt failure. This raises the specter of some intrinsic reaction the best candidates. All of these tenets have been challenged,
to the shunt operation or hardware in some patients. however, as surgeons have pushed the limits in identifying
There is a poorly defined group of patients familiar to all neu- surgical candidates. For example, patients with posterior fossa
rosurgeons who are sometimes called “difficult shunt patients.” tumors in which the brainstem is applied to the clivus, patients
These are patients who, for no particular reason, reappear at the with slit ventricles and shunts, and even patients with commu-
emergency department and return to the operating room for nicating hydrocephalus have all been reported to have success-
what seem like interminable shunt problems. Many of them ful outcomes after this procedure.
have slit-like ventricles and chronic headache and are catego- This wide range of potential candidates has made patient
rized within the equally poorly defined slit ventricle syndrome selection difficult. In children, the effects of age and etiology on
(64). These patients are challenging, and no universal treat- outcome have been particularly controversial, with literature to
ment seems helpful. Intensive investigation to rule out an support age as the most important (38, 45, 88), etiology as most
unrecognized but lingering infection, and monitoring to char- important (4, 22, 24), and neither or both being important in
acterize the intracranial pressure and its relationship to symp- terms of successful outcome (11, 26, 76). In a large multicenter
toms are very important. Surgical restraint, when the solution study from Canada, when multivariate analysis was used, age
is not obvious, also seems wise because changing the valve to was the predominant determinant of outcome for young
determine whether that helps seems rarely to work. These patients, particularly neonates, having very high failure rates
patients may be candidates for shunt removal and placement of (Figs. 6 and 7) (19). This finding was supported by a large series
an ETV, as discussed subsequently, although the long-term effi- from a single center in Australia (38). In the Canadian ETV
cacy of this therapy is not established (3). series, 22% of the patients had preexisting shunts. The presence
The obvious devastation in the quality of life for chronically of a shunt did not affect the failure rate, indicating that shunt
ill shunt patients and their families has been a clarion call for an independence is a reasonable outcome in selected patients.
objective measure of this domain. Recently, such a measure has Some have argued that even with a high failure rate, an ETV is

SHC636 | VOLUME 62 | NUMBER 2 | FEBRUARY 2008 SUPPLEMENT www.neurosurgery-online.com

 SURGICAL MANAGEMENT OF PEDIATRIC HYDROCEPHALUS

FIGURE 6. Overall outcome from the Canadian col- FIGURE 7. Outcome from the Canadian collabora-
laborative study of endoscopic third ventriculostomy tive study was broken down by age. There is a pro-
defined failure as any subsequent operation for cere- gressive decline in outcome with decreasing age,
brospinal fluid diversion (from, Drake JM; the being lowest in children younger than 1 month of
Canadian Pediatric Neurosurgery Study Group: age (from, Drake JM; the Canadian Pediatric
Endoscopic third ventriculostomy in pediatric Neurosurgery Study Group: Endoscopic third ven-
patients: The Canadian experience. Neurosurgery triculostomy in pediatric patients: The Canadian
60:881–886, 2007 [19]). experience. Neurosurgery 60:881–886, 2007 [19]).

always better than a shunt and is, therefore, a reasonable implanted in all patients having this procedure, and that
approach. Given that the failure rates in younger patients are patients with imaging evidence of obstruction should be reop-
high, the risks of the surgery itself are higher than with a ven- erated even if they are asymptomatic (2, 57). Whether these
triculoperitoneal shunt, and failure requires insertion of a ven- suggestions will reduce the incidence of this unfortunate event
triculoperitoneal shunt or repeat ETV, this recommendation is unknown.
requires reflection. As an indication of just how wide open the Endoscopic cyst fenestration for loculated ventricles, intra-
indications for ETV are, a recent report from a single surgeon ventricular cysts, extracerebral arachnoid cysts, and tumor exci-
and center from Africa reported good success with ETV in chil- sion has been reported to be an effective way of treating the
dren when they used a flexible endoscope and coagulated the accompanying hydrocephalus (29). Perhaps the most clearly
choroid plexus at the same time on both sides of the ventricu- documented cyst fenestration has been for suprasellar arach-
lar system, even when this technique was applied to children noid cysts (58, 77, 89). High-resolution endoscopy has often
with such disorders as myelomeningocele (90, 91). revealed a “valve” opening in the base of the cyst adjacent to
The complications of ETV are not infrequent and may be the basilar artery (69). Perforation of the roof of these cysts,
serious, with perforation of the basilar artery being the most with or without an accompanying opening in the floor, has
dreaded (55). CSF leak, meningitis, hypothalamic injury, and been reported to be successful. For patients with loculated
cranial nerve injury have all been reported (83). The overall hydrocephalus and trapped ventricles, endoscopic fenestration
surgical complication rate appears to be approximately 10 to has either obviated the need for a shunt or reduced the need for
15%. A rare and initially unrecognized complication is late multiple shunts (27). Endoscopic cannulation of the aqueduct
rapid deterioration. In this event, patients appear to be doing for a trapped fourth ventricle, usually with a shunt stent, is an
well after the procedure (30, 34). They begin to complain of example (12, 66, 70, 74). Although colloid cysts are the most
headache and then rapidly deteriorate; without immediate common tumor to be excised endoscopically (33, 52, 71, 79),
access to neurosurgery, they lapse into unconsciousness, and other small tumors have been removed endoscopically in the
most die (Table 1). A recent report compiled 15 cases from the presence of hydrocephalus and without the need for subse-
literature and around the world (17). The patients were oper- quent shunting (78). Endoscopic biopsy, performed at the same
ated on at an average of 8.4 years and had been followed for an time as a CSF diversion procedure, i.e., an ETV, has also obvi-
average of 2.8 years. Thirteen patients died and, in all who had ated the need for an open approach to the tumors, depending
an autopsy or repeat ETV performed, the opening in the floor on the tumor type (16).
of the third ventricle was closed. The assumption that this clo-
sure was a recent event was disputed by several patients, one SUMMARY AND FUTURE DIRECTIONS
of whom had evidence of ETV obstruction for more than 4
years before late rapid deterioration. These findings have It is unlikely that dramatic improvement in the treatment
prompted some to suggest that an access reservoir should be of pediatric hydrocephalus will occur without a better under-

NEUROSURGERY VOLUME 62 | NUMBER 2 | FEBRUARY 2008 SUPPLEMENT | SHC637

DRAKE

TABLE 1. Summary of late rapid deterioration endoscopic third ventriculostomy casesa

Patient no.b Country Age (yr)/sex Etiology Previous shunt Follow-up (mo) ETV occluded Outcome
1 UK 3/M IVH No 4.0 Unknown Death
2 UK 11/M AS No 4.0 Unknown Death
3 Australia 4/M CH No 84.0 Yes Death
4 Canada 13/F TG No 42.0 Yes Death
5 UK 0.83/M AS Yes 7.0 Unknown Death
6 Canada 0.33/M AS No 24.0 Yes Death
7 France 13/F TG Yes 24.0 Yes Normal
8 USA 10.83/F TG No 30.0 Yes Death
9 UK 13.6/F TG No 60.0 Yes Death
10 UK 12.92/M CH Yes 57.0 Yes Death
11 France 3.25/M AS Yes 9.0 Yes Death
12 France 7.8/F AS No 94.0 Yes Disability
13 UK 4.83/M CH No 1.4 Yes Death
14 USA 9.92/M AS No 5.0 Yes Mild disability
15 USA 5.83/F AS Yes 1.2 Yes Death
16 Japan 7.92/F AS No 22.0 Yes Death
Average 7.6 Average 29.8

a
ETV, endoscopic third ventriculostomy; UK, United Kingdom; USA, United States; IVH, intraventricular hemorrhage; AS, aqueductal stenosis; CH, congenital hydrocephalus;
TG, tectal glioma.
b
Patients 1 through 7 were previously reported. Means for patient age and follow-up period were 7.6 years and 29.8 months, respectively.

standing of the fundamental pathogenesis of the disorder 4. Beems T, Grotenhuis JA: Is the success rate of endoscopic third ventricu-
and the surgical devices that are used to treat it. This will lostomy age-dependent? An analysis of the results of endoscopic third ven-
triculostomy in young children. Childs Nerv Syst 18:605–608, 2002.
require the challenging of conventional thinking. Perhaps the
5. Black PM, Hakim R, Bailey NO: The use of the Codman-Medos Program-
best and most recent example is the emerging evidence that mable Hakim valve in the management of patients with hydrocephalus:
arachnoid granulations have only a secondary role in CSF Illustrative cases. Neurosurgery 34:1110–1113, 1994.
absorption, the primary site being the nasal lymphatics (36, 6. Breeze AC, Dey PK, Lees CC, Hackett GA, Smith GC, Murdoch EM: Obstetric
37). Behind every CSF shunt design is a mathematical model and neonatal outcomes in apparently isolated mild fetal ventriculomegaly.
J Perinat Med 33:236–240, 2005.
that specifies and predicts the device performance (54). These 7. Brown EM, Edwards RJ, Pople IK: Conservative management of patients
models have failed to produce devices that have had any sig- with cerebrospinal fluid shunt infections. Neurosurgery 58:657–665, 2006.
nificant effect on clinical outcome when subjected to prospec- 8. Bruner JP, Davis G, Tulipan N: Intrauterine shunt for obstructive hydro-
tive, randomized studies (20, 62) and have been unable to cephalus—Still not ready. Fetal Diagn Ther 21:532–539, 2006.
9. Carol M, Robinson W, Harris BS: Percutaneous placement of ventriculoatrial
accurately predict such fundamentals as the change in the
shunts: Four-year case experience. Neurosurgery 18:348–349, 1986.
size of the ventricles (85). This is either because they have 10. Chescheir NC, D’Alton M: Evidence-based medicine and fetal treatment:
been overly simplistic or unrealistic (81). Developing better How to get involved. Obstet Gynecol 106:610–613, 2005.
mathematical models is of paramount importance if better 11. Cinalli G, Sainte-Rose C, Chumas P, Zerah M, Brunelle F, Lot G, Pierre-Kahn
shunt designs are to emerge, but this will take a concerted A, Renier D: Failure of third ventriculostomy in the treatment of aqueductal
stenosis in children. J Neurosurg 90:448–454, 1999.
and multidisciplinary effort (82). Perhaps the fundamental 12. Cinalli G, Spennato P, Savarese L, Ruggiero C, Aliberti F, Cuomo L, Cianciulli
lesson from the last 30 years is that hydrocephalus, which E, Maggi G: Endoscopic aqueductoplasty and placement of a stent in the
appears to be an alluringly simple problem of CSF accumu- cerebral aqueduct in the management of isolated fourth ventricle in children.
lation, is anything but. J Neurosurg 104 [Suppl]:21–27, 2006.
13. Cochrane DD, Kestle J: Ventricular shunting for hydrocephalus in children:
Patients, procedures, surgeons and institutions in English Canada, 1989–2001.
REFERENCES Eur J Pediatr Surg 12 [Suppl 1]:S6–S11, 2002.
14. Cochrane DD, Kestle JR: The influence of surgical operative experience on the
1. Abbott R: History of neuroendoscopy. Neurosurg Clin N Am 15:1–7, 2004. duration of first ventriculoperitoneal shunt function and infection. Pediatr
2. Aquilina K, Edwards RJ, Pople IK: Routine placement of a ventricular reser- Neurosurg 38:295–301, 2003.
voir at endoscopic third ventriculostomy. Neurosurgery 53:91–97, 2003. 15. Decq P, Blanquet A, Yepes C: Percutaneous jugular placement of ventriculo-
3. Baskin JJ, Manwaring KH, Rekate HL: Ventricular shunt removal: The ulti- atrial shunts using a split sheath. Technical note. Acta Neurochir (Wien)
mate treatment of the slit ventricle syndrome. J Neurosurg 88:478–484, 1998. 136:92–94, 1995.

SHC638 | VOLUME 62 | NUMBER 2 | FEBRUARY 2008 SUPPLEMENT www.neurosurgery-online.com

 SURGICAL MANAGEMENT OF PEDIATRIC HYDROCEPHALUS

16. Depreitere B, Dasi N, Rutka J, Dirks P, Drake J: Endoscopic biopsy for intra- scopic ventriculoperitoneal shunt insertion: A multicenter randomized trial.
ventricular tumors in children. J Neurosurg 106 [Suppl]:340–346, 2007. J Neurosurg 98:284–290, 2003.
17. Drake J, Chumas P, Kestle J, Pierre-Kahn A, Vinchon M, Brown J, Pollack IF, 42. Kestle JR, Garton HJ, Whitehead WE, Drake JM, Kulkarni AV, Cochrane DD,
Arai H: Late rapid deterioration after endoscopic third ventriculostomy: Muszynski C, Walker ML, Management of shunt infections: A multicenter
Additional cases and review of the literature. J Neurosurg 105 [Suppl]: pilot study. J Neurosurg 105 [Suppl]:177–181, 2006.
118–126, 2006. 43. Kestle JR, Walker ML; Strata Investigators: A multicenter prospective cohort
18. Drake JM: Does double gloving prevent cerebrospinal fluid shunt infection? study of the Strata valve for the management of hydrocephalus in pediatric
J Neurosurg 104 [Suppl]:3–4, 2006. patients. J Neurosurg 102 [Suppl]:141–145, 2005.
19. Drake JM; Canadian Pediatric Neurosurgery Study Group: Endoscopic 44. Kirshtein B, Benifla M, Roy-Shapira A, Merkin V, Melamed I, Cohen Z, Cohen
third ventriculostomy in pediatric patients: The Canadian experience. A: Laparoscopically guided distal ventriculoperitoneal shunt placement. Surg
Neurosurgery 60:881–886, 2007. Laparosc Endosc Percutan Tech 14:276–278, 2004.
20. Drake JM, Kestle J, Milner R, Cinalli G, Boop F, Piatt J, Haines S, Schiff SJ, 45. Koch D, Wagner W: Endoscopic third ventriculostomy in infants of less than
Cochrane DD, Steinbok P, MacNeil N: Randomized trial of cerebrospinal 1 year of age: Which factors influence the outcome? Childs Nerv Syst
fluid shunt valve design in pediatric hydrocephalus. Neurosurgery 20:405–411, 2004.
43:294–305, 1998. 46. Kulkarni AV, Drake JM, Armstrong DC, Dirks PB: Measurement of ventricu-
21. Drake JM, Sainte-Rose C: The Shunt Book. New York, Blackwell Scientific, 1995, lar size: Reliability of the frontal and occipital horn ratio compared to subjec-
p 103. tive assessment. Pediatr Neurosurg 31:65–70, 1999.
22. Etus V, Ceylan S: Success of endoscopic third ventriculostomy in children less 47. Kulkarni AV, Drake JM, Lamberti-Pasculli M: Cerebrospinal fluid shunt infec-
than 2 years of age. Neurosurg Rev 28:284–288, 2005. tion: A prospective study of risk factors. J Neurosurg 94:195–201, 2001.
23. Farmer DL, von Koch CS, Peacock WJ, Danielpour M, Gupta N, Lee H, 48. Kulkarni AV, Drake JM, Rabin D, Dirks PB, Humphreys RP, Rutka JT:
Harrison MR: In utero repair of myelomeningocele: Experimental pathophys- Measuring the health status of children with hydrocephalus by using a new
iology, initial clinical experience, and outcomes. Arch Surg 138:872–878, 2003. outcome measure. J Neurosurg 101 [Suppl]:141–146, 2004.
24. Feng H, Huang G, Liao X, Fu K, Tan H, Pu H, Cheng Y, Liu W, Zhao D: 49. Kulkarni AV, Rabin D, Drake JM: An instrument to measure the health status
Endoscopic third ventriculostomy in the management of obstructive hydro- in children with hydrocephalus: The Hydrocephalus Outcome Questionnaire.
cephalus: An outcome analysis. J Neurosurg 100:626–633, 2004. J Neurosurg 101 [Suppl]:134–140, 2004.
25. Fernell E, Hagberg G: Infantile hydrocephalus: Declining prevalence in 50. Kulkarni AV, Rabin D, Lamberti-Pasculli M, Drake JM: Repeat cerebrospinal
preterm infants. Acta Paediatr 87:392–396, 1998. fluid shunt infection in children. Pediatr Neurosurg 35:66–71, 2001.
26. Fritsch MJ, Kienke S, Ankermann T, Padoin M, Mehdorn HM: Endoscopic 51. Langley JM, LeBlanc JC, Drake J, Milner R: Efficacy of antimicrobial prophy-
third ventriculostomy in infants. J Neurosurg 103 [Suppl]:50–53, 2005. laxis in placement of cerebrospinal fluid shunts: Meta-analysis. Clin Infect
27. Fritsch MJ, Mehdorn M: Endoscopic intraventricular surgery for treatment of Dis 17:98–103, 1993.
hydrocephalus and loculated CSF space in children less than one year of age. 52. Lewis AI, Crone KR, Taha J, van Loveren HR, Yeh HS, Tew JM: Surgical
Pediatr Neurosurg 36:183–188, 2002. resection of third ventricle colloid cysts. Preliminary results comparing tran-
28. Govender ST, Nathoo N, van Dellen JR: Evaluation of an antibiotic-impreg- scallosal microsurgery with endoscopy. J Neurosurg 81:174–178, 1994.
nated shunt system for the treatment of hydrocephalus. J Neurosurg 53. Maher CO, Garrity JA, Meyer FB: Refractory idiopathic intracranial hyperten-
99:831–839, 2003. sion treated with stereotactically planned ventriculoperitoneal shunt place-
29. Greenfield JP, Souweidane MM: Endoscopic management of intracranial ment. Neurosurg Focus 10:E1, 2001.
cysts. Neurosurg Focus 19:E7, 2005. 54. Marmarou A, Shulman K, Rosende RM: A nonlinear analysis of the cere-
30. Hader WJ, Drake J, Cochrane D, Sparrow O, Johnson ES, Kestle J: Death after brospinal fluid system and intracranial pressure dynamics. J Neurosurg
late failure of third ventriculostomy in children. Report of three cases. 48:332–344, 1978.
J Neurosurg 97:211–215, 2002. 55. McLaughlin MR, Wahlig JB, Kaufmann AM, Albright AL: Traumatic basilar
31. Haines SJ, Walters BC: Antibiotic prophylaxis for cerebrospinal fluid shunts: aneurysm after endoscopic third ventriculostomy: Case report. Neurosurgery
A metanalysis. Neurosurgery 34:87–92, 1994. 41:1400–1404, 1997.
32. Hansen AR, Snyder EY: Medical management of neonatal posthemorrhagic 56. Ment LR, Duncan CC, Geehr R: Benign enlargement of the subarachnoid
hydrocephalus. Neurosurg Clin N Am 9:95–104, 1998. spaces in the infant. J Neurosurg 54:504–508, 1981.
33. Hellwig D, Bauer BL, Schulte M, Gatscher S, Riegel T, Bertalanffy H: 57. Mobbs RJ, Vonau M, Davies MA: Death after late failure of endoscopic third
Neuroendoscopic treatment for colloid cysts of the third ventricle: The expe- ventriculostomy: A potential solution. Neurosurgery 53:384–386, 2003.
rience of a decade. Neurosurgery 52:525–533, 2003. 58. Nitahara J, Dozor AJ, Schroeder SA, Rifkinson-Mann S: Apnea as a present-
34. Javadpour M, May P, Mallucci C: Sudden death secondary to delayed closure ing sign of hydrocephalus. Pediatrics 97:587–589, 1996.
of endoscopic third ventriculostomy. Br J Neurosurg 17:266–269, 2003. 59. O’Hayon BB, Drake JM, Ossip MG, Tuli S, Clarke M: Frontal and occipital
35. Jea A, Al-Otibi M, Bonnard A, Drake JM: Laparoscopy-assisted ventricu- horn ratio: A linear estimate of ventricular size for multiple imaging modal-
loperitoneal shunt surgery in children: A series of 11 cases. J Neurosurg 106 ities in pediatric hydrocephalus. Pediatr Neurosurg 29:245–249, 1998.
[Suppl]:421–425, 2007. 60. Paes N: A new self-adjusting flow-regulating device for shunting of CSF.
36. Johnston M: The importance of lymphatics in cerebrospinal fluid transport. Childs Nerv Syst 12:619–625, 1996.
Lymphat Res Biol 1:41–45, 2003. 61. Pitkin RM: Folate and neural tube defects. Am J Clin Nutr 85:285S–288S,
37. Johnston M, Zakharov A, Papaiconomou C, Salmasi G, Armstrong D: 2007.
Evidence of connections between cerebrospinal fluid and nasal lymphatic 62. Pollack IF, Albright AL, Adelson PD: A randomized, controlled study of a
vessels in humans, non-human primates and other mammalian species. programmable shunt valve versus a conventional valve for patients with
Cerebrospinal Fluid Res 1:2, 2004. hydrocephalus. Neurosurgery 45:1399–1411, 1999.
38. Kadrian D, van Gelder J, Florida D, Jones R, Vonau M, Teo C, Stening W, 63. Portnoy HD, Schulte RR, Fox JL, Croissant PD, Tripp L: Anti-siphon and
Kwok B: Long-term reliability of endoscopic third ventriculostomy. reversible occlusion valves for shunting in hydrocephalus and preventing
Neurosurgery 56:1271–1278, 2005. post-shunt subdural hematomas. J Neurosurg 38:729–738, 1973.
39. Kan P, Kestle J: Lack of efficacy of antibiotic-impregnated shunt systems in 64. Rekate HL: The slit ventricle syndrome: Advances based on technology and
preventing shunt infections in children. Childs Nerv Syst 23:773–777, 2007. understanding. Pediatr Neurosurg 40:259–263, 2004.
40. Kestle J, Drake J, Milner R, Sainte-Rose C, Cinalli G, Boop F, Piatt J, Haines S, 65. Rich P, Jones R, Britton J, Foote S, Thilaganathan B: MRI of the foetal brain.
Schiff S, Cochrane D, Steinbok P, MacNeil N: Long-term follow-up data from Clin Radiol 62:303–313, 2007.
the Shunt Design Trial. Pediatr Neurosurg 33:230–236, 2000. 66. Sagan LM, Kojder I, Poncyljusz W: Endoscopic aqueductal stent placement
41. Kestle JR, Drake JM, Cochrane DD, Milner R, Walker ML, Abbott R 3rd, Boop for the treatment of a trapped fourth ventricle. J Neurosurg 105 [Suppl 4]:
FA; Endoscopic Shunt Insertion Trial participants: Lack of benefit of endo- 275–280, 2006.

NEUROSURGERY VOLUME 62 | NUMBER 2 | FEBRUARY 2008 SUPPLEMENT | SHC639

DRAKE

67. Sainte-Rose C, Hooven MD, Hirsch JF: A new approach in the treatment of 92. Willems PJ, Vits L, Raeymaekers P, Beuten J, Coucke P, Holden JJ, Van
hydrocephalus. J Neurosurg 66:213–226, 1987. Broeckhoven C, Warren ST, Sagi M, Robinson D, Dennis N, Friedman KJ,
68. Sainte-Rose C, Piatt JH, Renier D, Pierre-Kahn A, Hirsch JF, Hoffman HJ, Magnay D, Lyonnet S, White BN, Wittwer BH, Aylsworth AS, Reicke S:
Humphreys RP, Hendrick EB: Mechanical complications in shunts. Pediatr Further localization of X-linked hydrocephalus in the chromosomal region
Neurosurg 17:2–9, 1991. Xq28. Am J Hum Genet 51:307–315, 1992.
69. Santamarta D, Aguas J, Ferrer E: The natural history of arachnoid cysts: 93. Woodworth GF, McGirt MJ, Elfert P, Sciubba DM, Rigamonti D: Frameless
Endoscopic and cine-mode MRI evidence of a slit-valve mechanism. Minim stereotactic ventricular shunt placement for idiopathic intracranial hyperten-
Invasive Neurosurg 38:133–137, 1995. sion. Stereotact Funct Neurosurg 83:12–16, 2005.
70. Schroeder HW, Gaab MR: Endoscopic aqueductoplasty: Technique and 94. Yu S, Bensard DD, Partrick DA, Petty JK, Karrer FM, Hendrickson RJ:
results. Neurosurgery 45:508–518, 1999. Laparoscopic guidance or revision of ventriculoperitoneal shunts in children.
71. Schroeder HW, Gaab MR: Endoscopic resection of colloid cysts. Neurosurgery JSLS 10:122–125, 2006.
51:1441–1445, 2002.
72. Schubert F, Fijen BP, Krauss JK: Laparoscopically assisted peritoneal shunt
insertion in hydrocephalus: A prospective controlled study. Surg Endosc COMMENTS
19:1588–1591, 2005.
73. Sciubba DM, Stuart RM, McGirt MJ, Woodworth GF, Samdani A, Carson B,
Jallo GI: Effect of antibiotic-impregnated shunt catheters in decreasing the T his article provides a concise yet comprehensive overview of hydro-
cephalus as it relates to neurosurgical practice today. In reading
this thoughtfully composed article, I am reminded of how far we have
incidence of shunt infection in the treatment of hydrocephalus. J Neurosurg
103 [Suppl]:131–136, 2005. come in the investigation and management of this condition over the
74. Shin M, Morita A, Asano S, Ueki K, Kirino T: Neuroendoscopic aqueductal past 35 years of my own practice. This is, of course, not to discount the
stent placement procedure for isolated fourth ventricle after ventricular shunt continuing frustrations that plague our efforts to manage patients with
placement. Case report. J Neurosurg 92:1036–1039, 2000. hydrocephalus, particularly children. To my mind, shunt infection,
75. Shkolnik A, McLone DG: Intraoperative real-time ultrasonic guidance of ven- especially its prevention, represents the single most important problem
tricular shunt placement in infants. Radiology 141:515–517, 1981.
with which we have made excruciatingly little progress. Reported
76. Siomin V, Cinalli G, Grotenhuis A, Golash A, Oi S, Kothbauer K, Weiner H,
Roth J, Beni-Adani L, Pierre-Kahn A, Takahashi Y, Mallucci C, Abbott R,
infection rates are virtually unchanged over the past three decades.
Wisoff J, Constantini S: Endoscopic third ventriculostomy in patients with This continues to have enormous implications for patients, physicians,
cerebrospinal fluid infection and/or hemorrhage. J Neurosurg 97:519–524, and society at large. The other continuing nemesis is, of course, shunt
2002. obstruction. Similarly, we have made only modest progress with this
77. Sood S, Schuhmann MU, Cakan N, Ham SD: Endoscopic fenestration and problem over the same three decades, despite notable improvements in
coagulation shrinkage of suprasellar arachnoid cysts. Technical note. shunt design and surgical technique.
J Neurosurg 102 [Suppl]:127–133, 2005. I fear that as long as we require hardware of any sort to effectively
78. Souweidane MM: Endoscopic management of pediatric brain tumors. treat so many cases of hydrocephalus, these two problems will con-
Neurosurg Focus 18:E1, 2005.
tinue to plague our best efforts. Over the years, attempts have been
79. Souweidane MM: Endoscopic surgery for intraventricular brain tumors in
made to find a treatment that does not require implanting hardware.
patients without hydrocephalus. Neurosurgery 57 [Suppl]:312–318, 2005.
80. Sprung C, Miethke C, Trost HA, Lanksch WR, Stolke D: The dual-switch Choroid plexectomy and pharmacological reduction in cerebrospinal
valve. A new hydrostatic valve for the treatment of hydrocephalus. Childs fluid (CSF) production came to naught, for sound physiological rea-
Nerv Syst 12:573–581, 1996. sons. The refinement of effective endoscopic techniques along with
81. Tenti G, Drake JM, Sivaloganathan S: Brain biomechanics: Mathematical mod- advances in imaging have led to a revival of ventriculocisternostomy.
eling of hydrocephalus. Neurol Res 22:19–24, 2000. Endoscopic third ventriculostomy (ETV) is now widely practiced and
82. Tenti G, Sivaloganathan S, Drake JM: Mathematical modeling of the brain: has been a godsend for many patients and their neurosurgeons.
Principles and challenges. Neurosurgery (in press). Although ETV has its limitations, in my opinion, this procedure has
83. Teo C, Rahman S, Boop FA, Cherny B: Complications of endoscopic neuro-
had a more profound effect on hydrocephalus management than any
surgery. Childs Nerv Syst 12:248–253, 1996.
other single advance over the past 35 years.
84. Tuli S, Drake J, Lawless J, Wigg M, Lamberti-Pasculli M: Risk factors for
repeated cerebrospinal shunt failures in pediatric patients with hydro- Paul H. Chapman
cephalus. J Neurosurg 92:31–38, 2000. Boston, Massachusetts
85. Tuli S, O’Hayon B, Drake J, Clarke M, Kestle J: Change in ventricular size and
effect of ventricular catheter placement in pediatric shunted hydrocephalus.
Neurosurgery 45:1329–1335, 1999.
86. Tulipan N, Cleves MA: Effect of an intraoperative double-gloving strategy
T his article represents an unbiased assessment of the state of the art
and science of hydrocephalus and its treatment. James Drake,
arguably one of the most knowledgeable authorities on this subject, has
on the incidence of cerebrospinal fluid shunt infection. J Neurosurg 104
[Suppl]:5–8, 2006. again delineated what is known, what is myth, and what remains to be
87. Tulipan N, Sutton LN, Bruner JP, Cohen BM, Johnson M, Adzick NS: The done in this field. Unfortunately, we do not understand the precise
effect of intrauterine myelomeningocele repair on the incidence of shunt- pathophysiology of each etiology of hydrocephalus; therefore, uni-
dependent hydrocephalus. Pediatr Neurosurg 38:27–33, 2003. formly successful treatments have eluded us. Thus, we are reminded
88. Wagner W, Koch D: Mechanisms of failure after endoscopic third ventricu- that our current treatment paradigms are mired in uncertainty and
lostomy in young infants. J Neurosurg 103 [Suppl]:43–49, 2005. habit. Although mathematical modeling may yield better mechanical
89. Wang JC, Heier L, Souweidane MM: Advances in the endoscopic manage- devices, thus far, randomized studies have demonstrated no frontrun-
ment of suprasellar arachnoid cysts in children. J Neurosurg 100:418–426,
ner in the hardware competition for “best shunt design.” Furthermore,
2004.
a rather stark and honest delivery of the facts on shunt infection, shunt
90. Warf BC: Comparison of endoscopic third ventriculostomy alone and com-
bined with choroid plexus cauterization in infants younger than 1 year of age: failure, and ETV failure are delivered. This article essentially sums up
A prospective study in 550 African children. J Neurosurg 103:475–481, 2005. approximately 50 years of scientific knowledge about hydrocephalus
91. Warf BC: Hydrocephalus in Uganda: The predominance of infectious ori- and its treatments. It does not tell you what shunt to use or when to
gin and primary management with endoscopic third ventriculostomy. perform an ETV, but it will provide you with a scientific basis with
J Neurosurg 102 [Suppl]:1–15, 2005. which to make those decisions. For those who treat hydrocephalus

SHC640 | VOLUME 62 | NUMBER 2 | FEBRUARY 2008 SUPPLEMENT www.neurosurgery-online.com

 SURGICAL MANAGEMENT OF PEDIATRIC HYDROCEPHALUS

(i.e., all neurosurgeons), this will become a must-read reference article. far, determined by the extent of the hemorrhage and the amount of
And for some, it will provide an impetus to push the field forward. parenchyma destruction rather than ventricular size and the control of
the hydrocephalus (6). In utero treatment of hydrocephalus, as noted
Richard G. Ellenbogen
by Dr. Drake, has been abandoned, because the risks of the procedure
Seattle, Washington
outweigh the benefits. Directing therapy at ventricular size does not
address the underlying problem that caused the hydrocephalus and
J ames Drake ends his overview by stating that “hydrocephalus, which
appears to be an alluringly simple problem of CSF accumulation, is
anything but.” It is and it isn’t. CSF drains in response to a pressure dif-
appears to have limited benefit.
Regarding the future, advances in imaging are likely to lead the
way. We should soon be able to noninvasively observe and monitor
ferential. If the resistance to drainage increases beyond a point, which
CSF movement, determine intracranial pressure, and automatically
probably varies with time in any given individual and by individual,
obtain the volume of CSF in the ventricles. There are, no doubt, tech-
CSF accumulates. That point is simple. Why and how much the resist-
niques awaiting discovery that we have yet to imagine, such as MRI
ance changes, the physiological response to pressure change, and the
was 30 years ago.
treatment of CSF accumulation is not so simple.
Hydrocephalus is a plumbing problem: this is the simple part.
Although there have been many new technological innovations in
Understanding the pathophysiology of CSF drainage is not so simple.
shunt hardware, the effectiveness of the shunting systems has not
For some time, the only CSF drainage pathway was thought to be via
dramatically improved over the past 30 years. There have been step-
one-way valves directly into the major venous sinuses. As cited by Dr.
wise gains, such as inserting enough peritoneal tubing at the time of
Drake, Johnston (3) and Johnston et al. (4) have done excellent work in
the initial insertion, even in a neonate, to avoid a lengthening proce-
addressing the role of the lymphatics in the CSF drainage pathways.
dure with growth. Another is the encasement of a barium stripe
We have also examined those pathways and their responses to pressure
within the wall of the tubing, making the tube radio-opaque, but pre-
(1, 2, 5, 7). Being able to manipulate CSF drainage pharmacologically
venting the barium in the silicone from coming in contact with the
might make it possible to treat hydrocephalus without hardware.
subcutaneous tissues and thereby preventing the reaction that accel-
Although we still have a long way to go in the practice of pediatric
erates biodegradation and subsequent tubing fracture. The major
neurosurgery, we do more good for more patients by treating their
problems of obstruction and infection still remain, the reduction of
hydrocephalus optimally than anything else.
which would greatly reduce shunt failure rates. Approximately four-
fifths of our revisions are to replace an obstructed ventricular catheter. J. Gordon McComb
A considerable focus of shunt hardware has and continues to be the Los Angeles, California
valve, in an attempt to regulate CSF drainage to as close to normal
physiological parameters as possible. In our practice, over- or under-
drainage is rarely a problem, with shunt revisions to change the pres- 1. Erlich SS, McComb JG, Hyman S, Weiss MH: Ultrastructure of the orbital
sure of the valve accounting for, at most, 1% of our revision patients. pathway for cerebrospinal fluid drainage in rabbits. J Neurosurg 70:926–931,
As it is the ventricular catheter that becomes partially obstructed 1989.
2. Erlich SS, McComb JG, Hyman S, Weiss MH: Ultrastructural morphology of
before becoming completely blocked and, therefore, becoming the
the olfactory pathway for CSF drainage in the rabbit. J Neurosurg 64:466–473,
site of increased resistance to CSF flow, what happens downstream at 1986.
the valve is far less relevant. An advance would be a valve that could 3. Johnston M: The importance of lymphatics in cerebrospinal fluid transport.
decrease the incidence of ventricular catheter obstruction. The addi- Lymphat Res Biol 1:41–45, 2003.
tion of an antisiphon device should, theoretically, help in this regard, 4. Johnston M, Zakharov A, Papaiconomou C, Salmasi G, Armstrong D: Evidence
but there is no definitive evidence to support such devices that of connections between cerebrospinal fluid and nasal lymphatic vessels in
decrease proximal obstruction. humans, non-human primates and other mammalian species. Cerebrospinal
The major advances in the past 30 years have been in imaging. To Fluid Res 1:2, 2004.
be able to noninvasively make the diagnosis, determine the anatomic 5. McComb JG, Davson H, Hyman S, Weiss MH: Cerebrospinal fluid drainage as
influenced by ventricular pressure in the rabbit. J Neurosurg 56:790–797, 1982.
abnormality (if any), surgically intervene, and then monitor the effect
6. McComb JG, Ramos AD, Platzker AC, Henderson DJ, Segall HD: Management
of treatment, be it with CSF shunt diversion, opening of an of hydrocephalus secondary to intraventricular hemorrhage in the preterm
obstructed pathway such as fenestration of a loculated ventricle, or infant with a subcutaneous ventricular catheter reservoir. Neurosurgery
bypassing an aqueductal stenosis with an ETV, all have been made 13:295–300, 1983.
possible by the advances in imaging. The best way to avoid shunt 7. Shen JY, Kelly DE, Hyman S, McComb JG: Intraorbital cerebrospinal fluid
problems is not to insert one. Being able to noninvasively monitor outflow and the posterior uveal compartment of the hamster eye. Cell Tissue
ventricular size has aided immeasurably in the continuing care for Res 240:77–87, 1985.
those with hydrocephalus.
The introduction of fetal ultrasound and magnetic resonance imag-
ing have added a new dimension, and now one can be aware of the
presence of hydrocephalus fairly early in gestation. In our practice, we
G iven James Drake’s expertise in the arena of intracranial CSF
dynamics and valve technology, he is perfectly suited to offer an
authoritative overview on the treatment of hydrocephalus. His com-
routinely see families who want to know the prognosis of their fetus, ments are concise and cover a spectrum of theoretical and practical top-
which has been diagnosed as having hydrocephalus. This is a little ics. Troubling to the neurosurgeon, however, are his conclusions that
like asking about the prognosis of having a fever. Attention is often underscore the disassociation between shunt technology and improve-
focused on the size of the ventricles, which can easily be measured, and ments in therapy. He justifiably points out that valve mechanics,
any change in size is noted. Overwhelmingly, the prognosis is deter- catheter design, and programmable devices have all fallen short of pre-
mined by what caused the hydrocephalus rather than the size of the dictive modeling or theoretical claims when studied in well-designed
ventricles. As an example, it is possible to control progressive hydro- clinical trials. Although that news is sobering, it does not distract from
cephalus that is subsequent to intraventricular hemorrhage in preterm what I believe to be the greatest advancement in the treatment of
neonates. The neurological outcome for this patient population is, by hydrocephalus over the past three decades, namely, the use of ETV for

NEUROSURGERY VOLUME 62 | NUMBER 2 | FEBRUARY 2008 SUPPLEMENT | SHC641

DRAKE

the treatment for noncommunicating hydrocephalus, a procedure I therapeutic challenge, despite programmable valves, flow-regulated
learned from him. I am appreciative not only of Dr. Drake’s overview valves, and on-off valves. For diagnosis, one must resort to shunt
on a vexing problem in neurosurgery, but also his continual objective exploration, Diamox (Lederle Pharmaceuticals, Pearl River, NY) trial,
assessment of purported advancements in CSF diversion. intracranial pressure monitoring, migraine medication trials, and even
subtemporal craniectomy. For treatment, one may have to consider
Mark M. Souweidane
lumboperitoneal shunt, fronto-orbital advancement, or psychotherapy,
New York, New York
depending on the presumed cause.
Another challenge is the long-term patient who has run out of dis-
J ames Drake has provided us with an evidence-based review of the
current treatment of hydrocephalus. For some, his conclusions may
engender a feeling of hopelessness. Lacking is the usual unfettered
tal shunt sites, or the complex patient with congenital heart disease and
an abdominal event that precludes the usual ventriculoperitoneal or
ventriculoatrial shunt. One must resort to unusual procedures, such as
enthusiasm for unproven but popular technical innovations: program-
ventriculofemoralatrial shunts, gallbladder shunts, or sagittal sinus
mable valves, endoscopic shunt placement, antibiotic-impregnated
shunts.
catheters, telesensors, gravity accumulators, and even ETV and choroid
The biggest problem remains recurring proximal obstruction that
plexectomy. None of these devices and procedures has been proven to
develops as a result of multicavitary hydrocephalus or choroid plexus.
be a real advance by a rigorous scientific trial. Their use is often driven
This may require cyst fenestration, which is most effective if performed
by poor studies, patient wishes, competition among neurosurgeons,
via craniotomy with microscopic techniques rather than endoscopy.
and industry-driven marketing. Dr. Drake has done a real service in
Despite the initial promise that ETV would eliminate the need for
putting all of this into perspective. He did not indulge in “how I do it”
shunts, this is clearly not the case. As we enter a new period in medi-
cookbook recipes, which are abundant elsewhere.
cine in which we will be accountable for cost, and only scientifically
The reality is that the biggest advance in the treatment of hydro-
proven technology will be reimbursed, all of us will be required to
cephalus since the original Spitz Holter valve was the computed tomo-
know what works, and what does not. Dr Drake deserves our thanks.
graphic scanner. The management of hydrocephalus remains extremely
complex, requiring judgment and experience rather than a new gizmo. Leslie N. Sutton
The so-called slit ventricle syndrome remains a serious diagnostic and Philadelphia, Pennsylvania

Illustration showing congenital hydrocephalus due to failure of the foramina of Luschka and
Magendie to open. (From: Dandy WE: The Brain. New York, Harper & Row, Publishers, 1969).

SHC642 | VOLUME 62 | NUMBER 2 | FEBRUARY 2008 SUPPLEMENT www.neurosurgery-online.com