Introduction
Ventriculo-peritoneal shunt (VPS) placement is the most commonly performed procedure
for the treatment of hydrocephalus. Shunt malfunction occurs in more than 50% of children
within years after placement, with an 80% lifetime risk.[1] Malfunction may be due to obstruction, disconnection, infection, cerebrospinal fluid
(CSF) pseudocyst formation, and migration of the catheter.[2] The migration of the distal VPS catheter outside the peritoneal cavity is rare with
most cases reported in pediatric patients and an overall incidence ranging from 1.5
to 10%.[1]
[3] Distal shunt erosion into adjacent organs including the intestines, urinary bladder,
scrotum, heart, and the abdominal wall is well described. Migration of the VPS through
the vagina is rare with published reports in children on only 11 occasions. This review
aims to explore the symptoms of presentation, etiology and risk factors, timing, the
possible course of migration, management, and complications of transvaginal migration
of VPS in children by a detailed review of all published cases to date.
Methods
We reviewed the existing medical literature using PubMed, Google Scholar, Web of Science,
and Cochrane Library using the search terms: ventriculo-peritoneal shunt, VP shunt
migration, children, transvaginal migration, fallopian tube, uterus (see Appendix). The literature search was conducted with help of a librarian in October 2020.
Study Selection
Of the cases of VPS migrations that were retrieved, only those that reported the distal
VPS migration through the vagina, fallopian tube, or uterus in patients younger than
18 years old were included for review. The articles that mentioned adults, more than
18 years of age, and other genitourinary migration sites such as the anus, scrotum,
urinary bladder, or perineum were excluded.
The eligibility screening for the retrieved articles was done by two independent reviewers
first at the title and abstract level, and then at the full-text level if potentially
relevant. The references of the included articles had been also evaluated and articles
that met our criteria were included in the review ([Fig. 1]).
Fig. 1 Flow diagram showing the studies identification and selection
Results
A total of 203 papers were identified and after the duplicate exclusion, 93 abstracts
were screened. Abstracts were evaluated by the first author (SA) and 38 potentially
relevant abstracts were selected for full-text review and assessed for eligibility
by two authors (SA and SB). Finally, 11 articles were included in the review.
Discussion
Incidence
The overall incidence of migration of distal VPS varies from 1.5% to 10%.[1]
[3] A review by Harishchandra et al of 400 cases of shunt migration in both adults and
children revealed distal peritoneal catheter migration into the gastrointestinal tract
(35%), abdominal wall (14%), thoracic cavity (8%), cardiac/intravascular (7%), genitourinary
system (18%), intracranial (11%), and breast (3%) with the remaining 4% in miscellaneous
areas. Of the instances of genitourinary migration, the sites involved were the scrotum
(14%) and the bladder or perineum (4%).[4] A case series by Ezzat et al describes an incidence of 1.5% for the migration of
the distal VPS in children less than 12 years of age with the majority migrating to
the anus (40%) and scrotum (20%).[1] The most common site for VPS migration in both studies is the gastrointestinal system.[1]
[4] Migration of the VPS through the vagina is exceedingly rare, reported in only 3
adult patients and 11 children to date.[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
Etiology and Migration Course
The cause of migration of distal peritoneal catheter and perforation of visceral organs
is likely multifactorial. Localized inflammatory foreign body reaction or allergic
reaction between the shunt and the peritoneal membrane, pressure necrosis of the visceral
organ, peristalsis of bowel loops, abdominal wall contractions, and increased intraabdominal
pressure during respiration have been cited as contributing factors.[18] In cases of vaginal migration, the tip of the lower end of the shunt may become
adhered to the posterior vaginal wall in the pouch of Douglas due to inflammation
and chronic pressure by the stiff and sharp end of the shunt tip. Over time, the continuous
hammer effect of CSF pulsation could lead to perforation through the uterus or posterior
vaginal wall followed by transvaginal migration.[12] Migration can also occur through the fallopian tube likely from catheter friction
against the less-motile fallopian tube and gradual adherence with the formation of
a fibrous tract as a result of inflammation.[13]
[15] Of the cases reviewed in pediatrics, the authors have described the course of transvaginal
migration in only 5 out of 11 cases ([Table 1]).
Table 1
Summary of all cases of transvaginal migration in children
|
Cases (N = 11)
|
Age
|
Symptoms
|
Time of migration since the VP shunt placement
|
Risk factors
|
Possible course of migration
|
CSF culture
|
VP shunt catheter material
|
Technique of distal catheter placemen
|
Complications
|
|
Patel et al
1973
|
11 months
|
Vomiting, stiff neck, and fever
|
4 months
|
Lumbar myelomeningocele
One shunt revision in the past
|
Posterior vaginal wall
|
Enterococci
|
Raimondi Catheter
|
Not described
|
Shunt revision done in another session
|
|
Washington et al 2002
|
16 years
|
Headache, abdominal pain, clear vaginal discharge
|
15–18 months
|
Appendectomy, right ovarian cystectomy, two shunt revisions
|
Distal end stuck in the right fallopian tube with cystic loculation
|
Negative
|
Not reported
|
Not described
|
Laparoscopic salpingectomy and no shunt revision needed
|
|
Farrokhi et al
2007
|
16 months
|
Foreign body with VPS protruding through the vagina
|
6 months
|
Myelomeningocele
One shunt infection with revision in the past
|
Posterior aspect of bladder down to cul-de-sac and posterior wall of vagina
|
Not reported
|
Not reported
|
Not described
|
Not reported
|
|
Atlas et al
2012
|
14 months
|
Asymptomatic VP shunt tube extruding through the vagina
|
2 months
|
Prematurity and
one shunt revision in the past
|
Not described
|
Negative
|
Not reported
|
Not described
|
Shunt revision done in another session
|
|
Teegala et al
2012
|
6 months
|
Fever, neck, rigidity, and VP shunt tube protruding through the vagina
|
2 months
|
Dandy–Walker malformation
No shunt revisions
|
Not described
|
Negative
|
Not reported
|
Not described
|
Required endoscopic third ventriculostomy
|
|
Poillblanc et al
2012
|
17 years
|
Recurrent watery vaginal discharge
|
2 months
|
Porencephalic cyst
Cysto-peritoneal shunt
Multiple shunt revisions
Laparotomy with adhesion lysis 2 months prior
|
Perforation through the uterus
|
Negative
|
Not reported
|
Laparotomy
|
The uterine defect did not require any repair
Cystoatrail shunt was placed
|
|
Lotfinia et al
2016
|
24 months
|
Asymptomatic VP shunt tube protruding through the vagina
|
21 months
|
One shunt revision
|
Not described
|
Negative
|
Not reported
|
Not described
|
Received antibiotics for one week Shunt revision done in the same session
|
|
Chugh et al
2018
|
6 months
|
Asymptomatic with shunt tube protruding through the vagina
|
2 months
|
None
|
Not described
|
Negative
|
Not reported
|
Not described
|
Endoscopic third ventriculostomy
|
|
Chugh et al
2018
|
8 months
|
Asymptomatic with shunt tube protruding through the vagina
|
3 months
|
None
|
Not described
|
Negative
|
Not reported
|
Not described
|
Revision done in another session
|
|
Korulmaz et al 2019
|
16 years
|
Vomiting and foreign body in the vagina
|
6 months
|
Colostomy for anal atresia
One shunt revision in the past
|
Not described
|
Not reported
|
Not reported
|
Not described
|
Not reported
|
|
Lord et al
2020
|
6 years
|
White rubbery tube protruding from the vagina draining clear liquid
|
3 years
|
No shunt revisions in the past
|
Through the fallopian tube, uterus, and exit through the vagina
|
Streptococcus anginosus
|
Silicon (Medtronic)
|
Mini laparotomy
|
Spontaneous coiling
Asymptomatic ventriculitis requiring prolonged antibiotic course
Shunt revision in another session
|
Risk Factors
Incomplete or inconsistent data were reported among the 11 pediatric cases to uniformly
compare risk factors; however, some similarities give us clues as to the risk factors
that may account for the transvaginal migration of a VPS in children. Transvaginal
migration of distal VPS seems to be more common in younger children. Among the pediatric
cases, 63% (7/11) were less than 2 years old with a mean age of 12 months. An extensive
review by Harischandra et al of nearly 400 cases of migration of all distal VPS revealed
282 patients (71%) who belonged to the pediatric age group. Weaker musculature and
smaller volume of the intraperitoneal cavity may be contributing factors explaining
the increased incidence in children.
A history of previous abdominal surgeries appears to be another common denominator.
Of the nine cases reporting details, 77% (7/9) had a history of one or two shunt revisions
that were completed shortly prior to the episode of trans-vaginal migration. Two cases
report nonshunt-related intraabdominal procedures such as appendectomy and right ovarian
cystectomy and laparotomy with adhesiolysis.[12]
[13] In these cases, adhesions could have made the shunt tubing less mobile, promoting
adherence to the surrounding structures. Two of the eleven reports had a history of
lumbar meningomyelocele, which may have contributed to weaker musculature in the vagina
wall.[9]
[11]
The type of catheter material has also been implicated in predisposing to migration
and perforation in a variety of distal shunt migrations.[19]
[20] In the past, Raimondi spring coiled catheters were commonly associated with perforations
as reported in the case by Patel et al[11] but the introduction of softer and more flexible silastic silicone tubing has reduced,
but not eliminated this risk.[15]
[20] Rates of distal shunt migration for each type of tubing are not known.
The technique of distal catheter insertion may play a role in distal shunt migration
and failure.[21]
[22] A systematic review performed by He et al looking at outcomes based on laparoscopic
versus mini-laparotomy peritoneal catheter insertion of VP shunts revealed that laparoscopy-guided
placements were associated with a less-frequent occurrence of distal shunt failure
(distal shunt failure rate for laparoscopy vs. mini-laparotomy: 3.0% vs. 10.2%, p = 0.0003).[21] No pediatric cases of transvaginal migration reported the technique of distal catheter
placement. Controlling the length of the distal catheter during placement may help
mitigate the overall risk of migration.[23]
Clinical Presentation and Diagnosis
The migration of the distal shunt catheter typically occurs anywhere from hours to
years after the placement of the shunt system with most migrations presenting in the
first year, beyond which the chances of migration decrease gradually.[3]
[21] In this review of transvaginal migration in children, the time of migration ranged
from 2 months to 3 years of last shunt revision or initial placement of VPS, with
90% (10/11) of them occurring within an average of 6.3 months.
VPS malfunction can present in a myriad of ways such as headache, fever, vomiting,
and abdominal pain but can also be asymptomatic. The majority of the cases of transvaginal
migration presented with no symptoms aside from the extrusion of the shunt tubing
itself or leaking of clear fluid from the vagina. The diagnosis of transvaginal migration
is typically straightforward when the patients present with the catheter protruding
out through the vagina. Extrusion of the shunt through the vagina can be at risk of
infection by the urogenital or GI flora as it is close to the perineum and can present
with signs of meningitis.[11] In one case of asymptomatic presentation, there was culture-positive cerebrospinal
fluid (CSF).[15] Infectious complications must be considered in all cases of distal VP shunt migration
outside the peritoneum and especially with exposure to the external environment.
If the presenting symptoms are abdominal pain and/or profuse clear vaginal discharge
in a patient with VPS, a careful urogenital examination must be performed and imaging
may be required with shunt malfunction as part of the differential diagnosis until
malpositioning of the distal shunt has been ruled out. X-ray abdomen or shunt series
can be a useful test that can demonstrate the location of the distal shunt. Failure
to consider this diagnosis can result in a delay in diagnosis as seen in the case
by Washington et al where the distal shunt was stuck in the omentum or surrounding
structures causing leakage of CSF from the vagina but not full extrusion of the shunt
itself.[13]
Lord et al described the utility of ultrasound in these situations showing the VPS
catheter traversing a hypoechoic fluid-filled cavity, while Washington et al questioned
the specificity of the range of ultrasound when there are several small adnexal structures
obscuring the catheter tip.[13]
[15] Poillblanc et al described the utility of transvaginal ultrasound for a catheter
that has perforated the uterus but not extruded the vagina by identifying the tip
of the catheter in the uterus.[12] Computerized tomographic imaging can help to delineate VPS location with drawbacks
of radiation exposure and the possible need for sedation in young children. If an
abnormal connection between the VPS and the genitourinary structures is a concern
and there is no transvaginal migration, the clear vaginal discharge can be tested
for β-2 transferrin, a specific protein found only in CSF.
Management
The standard method of treatment in any shunt migration would be the removal of the
extruded shunt followed by shunt repositioning or other CSF diversion procedures either
immediately or after controlling any infection with a course of antibiotics. In some
cases, repair of a defect in the organs penetrated by the distal shunt is required.
Most children with transvaginal migration (63%) required replacement of the VPS in
another session after the peritoneal inflammation or infection subsided and none of
the patients required repair of affected viscera. Two of the patients had the shunt
replaced in the same session as there was no concern for infection.[8]
[13] The disconnection and removal of the migrated distal catheter can be performed by
pulling the shunt from the extrusion site after dividing the shunt intra-abdominally
either through a laparoscopic incision or an exploratory laparotomy, followed by externalization
of the proximal shunt. Blindly pulling the shunt from the extrusion site can be difficult
and dangerous, especially when there is coiling or twisting of the catheter around
bowel loops or other structures.[24]
[25] This can be avoided by careful review of the shunt course with an abdominal X-ray
to ensure sure there is no knotting or extensive looping in the abdomen prior to the
surgery. If there is knotting noted on the X-ray ([Fig 2]), care should be taken to involve pediatric surgery for laparotomy or laparoscopic
removal of the migrated distal shunt by relieving the knot. Removal of the entire
shunt may be required in cases of positive cerebrospinal fluid cultures with the temporary
placement of an external ventricular drain until the infection clears.
Fig. 2 X-ray lateral decubitus position showing the loop in the right side of the abdomen.
Image courtesy Lord et al15.
For high-risk or documented infections, the timing of shunt replacement is variable.
Treatment time is often guided by cultures of cerebrospinal fluid, which should be
negative for several days before the shunt revision is planned.[26] Of the documented pediatric cases, one had the shunt revised in the same session
but was given prophylactic antibiotics for 1 week.[8] Another case also had only one surgical intervention with no shunt revision required
because the distal shunt was never extruded but rather was stuck in the fallopian
tube.[13] In seven of the nine cases that reported follow-up, replacement of the shunt or
endoscopic third ventriculostomy was required later after a course of antibiotics.
Upon replacement, a new distal site other than the peritoneum should be considered
depending on risk factors for recurrence (i.e., adhesions). For initial and repeat
VPS placement, consideration should be given to using mini-laparotomy or other newer
laparoscopic procedures to decrease the risk of distal shunt migration.[22] To prevent further migration of the distal VPS in the future, endoscopic third ventriculostomy
can also be considered, if possible.[14]
[16]
Complications
There can be significant morbidity associated with the migration of the distal VPS
due to several complications. There are no deaths reported for any of the reported
cases of transvaginal distal shunt migration. However, migration of a VPS typically
requires multiple operative interventions and thus multiple anesthetics. Revision
of the shunt is almost always needed in cases of transvaginal migration. The majority
of children in this review required the shunt to be revised or replaced and two underwent
endoscopic third ventriculostomy.
Regardless of the site, any extrusion of the shunt with exposure to the external environment
is associated with an approximate 50% chance of shunt infection.[4] Among the reported pediatric cases, two of nine (with two cases not reporting culture
data) reported positive CSF cultures with enterococci and Streptococcus anginosus. Exposure to the outside environment and being near the urogenital and GI areas likely
put transvaginal migrations at a higher risk for infections. If there is a documented
CSF infection, the current guidelines recommend removing the VPS until CSF cultures
clear, often requiring an external ventricular drain (EVD) to manage the hydrocephalus
while awaiting resolution.[26] To achieve this, a prolonged hospital stay is needed to control the infection with
the placement of a peripherally inserted central catheter (PICC) for long-term antibiotic
administration. Finally, there can be delayed peritonitis resulting in significant
adhesions requiring adhesiolysis prior to shunt reinsertion.
During transvaginal migration of VPS, the distal catheter can erode into the posterior
vaginal wall, uterus, or fallopian tube. Laparoscopy is usually recommended in cases
of distal catheter migration to visualize the defect and determine if it requires
repair. Occasionally, the defect can be left alone to close by itself such as small
uterine defects due to the muscular nature of the uterine tissue.[12] If the urinary bladder wall is affected, cystoscopy must be performed with methylene
blue instilled into the bladder to visualize any urine extravasation into the peritoneal
cavity.
In two pediatric cases, the VPS entered the fallopian tube, and one required salpingectomy.[13]
[15] It is unclear if these complications could have an effect on future fertility or
create an increased risk of ectopic pregnancy. Gynecologic follow-up may be warranted
for further workup and monitoring.
Conclusion
VPS placement is a well-established procedure frequently employed in the management
of hydrocephalus. Although the placement of a VPS is a common procedure and considered
safe, 70 to 80% of the patients require at least 1 revision at some point in their
lives. Migration of the distal shunt seems to occur more commonly in children than
adults. In rare cases of transvaginal migration, we identified that previous nonshunt-related
abdominal operations and shunt revisions are consistent risk factors. We did not recognize
specific approaches to catheter placement or management that could have prevented
this complication. Ventriculitis necessitating shunt removal and therapies requiring
additional procedures and prolonged hospitalization are the major consequences identified.
Awareness of this unusual complication is very important among health care providers
taking care of children with VPS, especially general pediatricians and emergency care
health providers who are likely to be the first to encounter these children on initial
presentation. This review helps to raise awareness for early recognition and management
of transvaginal migration, which could limit complications and be potentially lifesaving.
