Shunt migration in ventriculoperitoneal shunting: A comprehensive review of literature
Correspondence Address: Source of Support: None, Conflict of Interest: None DOI: 10.4103/0028-3886.253968
Source of Support: None, Conflict of Interest: None
Keywords: Shunt displacements, shunt complications, ventriculoperitoneal shunt
Cerebrospinal fluid (CSF) shunt-related complications are broadly classified into the mechanical/nonmechanical types., The mechanical complications include obstruction, disconnection, and migration. The nonmechanical complications include infection and distal compartment-related complications (pseudocyst formation, ascites, and pleural effusion)., Shunt dysfunction is considered to be the most common neurosurgical complication encountered. The mechanical complication of shunt migration is roughly noted in 1 in 1000 patients who have undergone a shunt procedure. It is relatively uncommon in comparison with the other shunt-related complications (such as infection or shunt obstruction).
Definition of shunt migration
Migration may be broadly defined as “translocation of the part/whole of the shunt system (proximal/distal catheter/reservoir/valve) from the compartment where it was intended to be to a new compartment which may be associated with/without shunt dysfunction.”
Types of shunt migration/displacement
Shunt migration can be classified based on:
The terms cranial (upward) and caudal (downward) would refer only to situations where there is migration of the entire shunt system (total) and strictly not to migration of proximal/distal catheter. In general, a caudal migration is more common than a cranial migration. The incidence of caudal migration is around 10%., Caudal migrations are usually asymptomatic, whereas cranial migrations are usually symptomatic. Shunt migrations may also be classified based on their clinical presentation (symptomatic: shunt dysfunction/shunt infection, or asymptomatic).
A literature search was performed using PubMed Central for all articles containing the term “ventriculoperitoneal shunt migration.” All articles published through 2016 were included in this review. The inclusion criteria were (1) publications in English and (2) publications based on human subjects. Articles which dealt with lumboperitoneal shunt migrations and ventriculoatrial (VA) shunt migrations were excluded. The bibliography search was also done independently by the senior author to find any missed out articles. Also, duplication of articles was carefully looked out for and excluded.
The variables analyzed were (1) the site of migration, (2) the age of the patient, (3) the presentation of the patient, (4) the association with shunt dysfunction, (5) the association with shunt infection, (6) the time interval between the last shunt placement and migration, (7) the number of previous shunt revisions, and (8) and the shunt system used.
Defining the variables
Migration site: This was divided into the following nine categories: (1) intracranial (ventricular and subdural), (2) subgaleal, (3) chest wall/thorax, (4) breast, (5) cardiac/intravascular, (6) abdominal wall (including anterior and posterior abdominal wall [PAW] migration), (7) genitourinary (scrotum, bladder, and perineum), (8) bowel (stomach, small bowel, large bowel, and rectum), and (9) miscellaneous (intrahepatic, neck, and any other part not categorized above). In a few patients, one shunt system may have migrated into multiple sites. In such patients, the migration site included all the categories into which the migration had taken place.
Age of the patient: This criterion broadly divided the patients into the pediatric (<18 years) and adult patients.
Clinical presentation: This was broadly defined as symptomatic or asymptomatic. Symptomatic patients were with reference to the site of migration/shunt function. Asymptomatic patients were defined as those in whom migration was an incidental radiological finding. Extrusion of the shunt was included in the symptomatic (migration site related) presentation. When an article quoted growth of microorganism in the CSF/shunt system, this was accepted as evidence of shunt infection.
Shunt revision: If an article mentioned about the number of shunt revisions a patient had undergone, then this number of revisions was mentioned. The time interval between the last shunt revision and shunt migration was noted.
Based on PubMed Central search, we found 288 articles. Among them, 111 articles were excluded because the abstracts did not mention the site of migration, or the article did not include any case with shunt migration. Those articles which dealt with peritoneal shunt migration but did not mention whether it was lumbar or ventricular peritoneal shunt migration were also excluded. Based on the bibliographic search, 127 articles were found. Among them, 15 were excluded because the full article was not in English or they did not define any migration. Twenty-six articles were further excluded because of overlap (a few had more than one patient with various migration sites; a few had more than one shunt system in the same patient migrating to various sites). Also, articles in which age was not mentioned in any of the patients in the case series were also excluded, which was found to be the case in 24 articles. In total, 239 articles were reviewed in this narrative review. From these articles, we found a total of 396 shunt migrations [Figure 1].
The most common site of migration was found to be perforation of the bowel in 139 patients, followed by migration to the genitourinary region (n = 70), abdominal wall (n = 57), chest/thorax (n = 32), intracranial site (n = 30), cardiac/intravascular system (n = 28), breast (n = 13), subgaleal plane (n = 12), and miscellaneous sites (n = 15; liver = 7; neck = 6; lower limbs = 2). Among the patients who had shown a genitourinary region migration of the shunt, scrotal migration occurred in the majority of the patients (n = 55) and only the remaining 15 patients had migration to the bladder or perineum. Though for ease of classification, migration to the genitourinary region was taken to include migration to scrotum, bladder, and perineum, scrotal migration will be dealt separately due to its unique presentation and management strategies.
With respect to the age, 282 patients belonged to the pediatric age (71.2%) group and 112 patients were adults (28.8%). The age groups ranged from the youngest being 4 days old to eldest being 87 years old., About 92.3% of migration to the breast (n = 12) occurred in adult patients, and 91.6% of the patients with subgaleal migration (n = 11) were children. A total of 182 (56.3%) patients were male and 141 (43.7%) were female, with no mention about the gender being found in 73 patients.
Overall, 260 (71.6%) patients presented with symptoms related to the site of migration and 95 (26.2%) with evidence of shunt dysfunction. In 18 patients, shunt migration was an incidental finding (approximately 5%), and the details of site of migration could not be obtained in 33 patients. Among the various sites, migration to the chest (78.8%) and scrotum (86.5%) were the ones to present with local symptoms. Cranial migration to the intracranial (70%) and subgaleal location (58.3%) had maximal incidence of shunt dysfunction. Shunt infection was noted in 51 (approximately 14%) patients only, out of whom migration to the bowel was responsible for infection in approximately 80% (n = 40) of shunt infections.
On evaluation for the number of shunt revisions, 169 (66.8%) patients had no shunt revision, 84 (33.2%) had one or more revisions, and in 179 patients, there was no mention of shunt revision. In merely 88 (20.4%) patients, there was a mention of the shunt system that was used. An abandoned distal catheter migration was noted in 12 patients with 75% (n = 9) of them presenting with symptoms of distal site migration and 50% (n = 6) being associated with a hollow viscous (bowel or bladder) perforation. Information about the time interval between the last shunt that had been performed and shunt migration was available in 334 (77.5%) patients and this time interval ranged from 24 h to 38 years.,
The discussion to follow will be about each migration site in detail.
Migration into the bowel
This was found in approximately 35% patients. Abdominal complications following a ventriculoperitoneal shunt (VPS) included a pseudocyst formation, a subcutaneous CSF collection, an obstruction, or shunt disconnection and migration. Though abdominal complications contribute to 25–30% of VPS-related issues, among them, gastrointestinal (GI) perforation is responsible for only 0.1–0.7% of complications., Still, this is the most common migration site for a VPS. Till 2016, 139 patients with migration into the bowel have been identified. The first report of migration of the shunt into the bowel was published in 1966 by Wilson and Bertand in two patients following a lumbo-peritoneal shunt procedure. The first case report in a live patient was by Rubins et al., in 1972. The largest case series of bowel perforation was by Vinchon et al., in 2006, reporting about 19 patients. It was also Vinchon et al., who introduced the term BPPC (bowel perforation by peritoneal catheter). The largest review till date has been by Odebode et al., in 2007 who reviewed around 119 patients with bowel perforation.
Migration into the GI tract has been observed to occur in children (80.6%) more often than in adults (19.4%). This may be due to the weaker GI wall and excessive peristaltic activity noted in children., A few papers have quoted that male patients are more prone to developing GI migration,, and on analyzing the data (from the data available in 103 patients), it was found that around 60% of the patients were males. Regarding the presentation, it was noted that abdominal symptoms (transanal extrusion, transoral extrusion, pain, vomiting, peritonitis, and watery diarrhea) were the most common (n = 117; 74%) symptoms, followed by shunt dysfunction which occurred in 21.5% (n = 34) [a few patients were having abdominal symptoms with shunt dysfunction]. Transanal/transoral presentation may be associated with other abdominal symptoms (n = 24) or it may be an asymptomatic extrusion (n = 52). Any patient who has undergone a VPS having an abdominal symptom should be evaluated vigilantly for BPPC, with Masouka et al., even suggesting a computed tomography (CT) abdomen/endoscopy for evaluation., Asymptomatic presentation (in the context of BPPC, an asymptomatic patient would refer to an absence of abdominal or cranial symptoms and with no incidental radiological findings in patients with transanal or transoral presentation) was more commonly noted in children (88.8%) than in adults. It has been reported that before bowel perforation, there could be an episode of transient shunt dysfunction when the catheter adheres to the bowel.
Transanal presentation was the most common manifestation (n = 55). Among these patients, 84% were asymptomatic (n = 46) and 16% were symptomatic (with abdominal or shunt-related symptoms). Only 14 patients with a transoral presentation have been reported, among whom 85% were symptomatic and 15% were incidental [Table 1]. Peritonitis was noted only in 7 (5%) patients and meningitis was the presentation in 21 (13.3%) patients. Peritonitis is considered to be less common than meningitis because of the omental and fibrous adhesions that are formed around the catheter which walls off the infection and prevents the spread of infection inside the peritoneum. Infection of the shunt was identified in 40 (around 30%) patients, with this incidence being highest among the shunt migrations to any location. Amongst the infectious agents, Escherichia coli was the most commonly grown bacteria (24 of the 40 patients), and overall, gram-negative bacteria were noted in 31 (77%) patients. Any patient who has undergone a VPS, presenting with gram-negative meningitis or ventriculitis, has to be evaluated for perforation of the bowel.
The information regarding which part of the GI tract was perforated was available only in 75 patients. Among the sites of GI tract involvement, the most common was the colon (n = 42; 56%) followed by the stomach (n = 21; 28%), and the least common site was the small intestine (n = 12; 16%). It may be proposed that parts of GI tract which are relatively fixed are easier sites for catheter perforation than is the more motile small bowel. It was also noted that among the patients with a transoral presentation, every patient had a gastric perforation except the one who had jejunal perforation. In a similar fashion, most of the transanal presentations had perforation of the lower GI tract (especially the colon), except one patient who had gastric perforation.
Details about the various shunts used were available only from 31 patients. Among the articles which mentioned about the shunt catheter used, Raimondi peritoneal catheter (n = 11), Pudenz (n = 4), Chhabra (n = 4), or a silicon-based catheter (n = 8) was used most often.,,, Also, it was mentioned by Plummer et al., that the use of silicone catheters may induce a foreign body reaction which may be the inciting event for chronic abdominal inflammation. In the entire shunt migration review, there were 12 abandoned distal catheters which had migrated, among which 50% were from the BPPC group (which formed the individual site with the highest incidence). Hence, an abandoned distal catheter in the peritoneum may be a risk factor for bowel perforation, and any abdominal complaint in a patient with orphaned catheter would need immediate evaluation. When evaluating for the role of shunt revision, it was noted that only around 30% patients had a previous history of shunt revision among the patients with BPPC.
The time interval between the previous shunt surgery and bowel perforation was found to vary from as short as 10 days to as long as 14 years with an average of 19.3 months (median duration 10 months)., Park et al., in his review, have also mentioned that the average time interval for a patient to develop BPPC is around 18.7 months. Hence, perforation and migration into the bowel is a chronic process and is unlikely to be an acute event. Park et al., also suggested that if a perforation occurred in less than 3 months, it is more in favor of a direct injury during placement of the distal catheter. On evaluation of patients presenting in <3 months following the placement of their shunt, 16 patients were identified but there was no mention of the peroperative perforation event except in one article. The age of presentation with this complication is summarized in [Figure 2].
The basic pathophysiology behind bowel perforation based on the abovementioned factors can be explained as follows. The catheter rubs against a less motile bowel constantly and gradually adheres to the bowel as a fibrous tract forms around the catheter and bowel (during this period, there may be a transient shunt dysfunction). With further friction between them, the catheter perforates and enters the bowel. As there is a well-formed tract, there is no spillage of bowel contents. Previous infections/surgery/radiation (radiation colitis) makes the bowel more weak and sticky and increases the chance of perforation. Also, in young children with previous shunt infection, the hyperperistaltic activity, restricted abdominal space, or the weaker bowel wall makes them more prone for perforation.
When suspecting a bowel perforation, a CT abdomen/pelvis may be the ideal investigation as it helps in identifying the mucosal thickening as well as inflammation and the distal catheter inside the bowel. To identify the site of entry into the bowel, a shuntogram may be suggested., More recently, an upper GI endoscopy/sigmoidoscopy/colonoscopy has been used for identifying the site of entry and also for removal of the perforated part of the catheter.,, Any sign of shunt dysfunction would warrant an imaging of the brain. In a CT scan of the brain, a finding of pneumocephalus suggests the presence of BPPC. Though shunt infection was observed in only 30% of the migration (in the BPPC group), it would be appropriate to perform a shunt tap to rule out infection.
The ideal management which has been suggested by most authors would be disconnection of the distal end and its removal. This may be done blindly by pulling the shunt from the extrusion site (transanal/transoral) or under the guidance of endoscopy. Endoscopy has an additional advantage of inspecting the site of perforation for any leak., Placement of fibrin glue has also been suggested to obliterate the shunt–bowel fibrous tract., Following the shunt removal, the patient would be advised to avoid an oral diet for a couple of days to provide time for the healing of the perforated site. Most authors have suggested externalization of the shunt following its disconnection to treat the assumed infection for a couple of weeks and then replacing it in the abdomen. Laparotomy has a very limited role, unless there is peritonitis or intestinal obstruction. Some authors also suggest a bowel resection-anastomosis at the site of perforation. Laparoscopy has also been suggested by Mandhan et al., for establishing the location of the site of perforation as well as for determining the feasibility for the direct removal of the distal catheter. In patients with shunt infection, the externalization of the proximal shunt system would continue for 2–3 weeks till the infection settles, following which a new distal tube, which is shorter in length and with a softer tip,, may be placed in the peritoneum. The protocol suggested is shown in [Figure 3].
This occurs in approximately 18% patients. This region includes scrotal migration (14%) and migration to the bladder or perineum (4%). Both are dealt as separate entities as the management of both of them is will be different.
The incidence of involvement of this site is in the range of approximately 4%. This compartment of migration includes migration to the bladder and perineal structures (vulva/vagina). There have been 16 case reports till date predominantly contributed by bladder migration (n = 11) and transvaginal migration (n = 4). There is only one case report of migration through the vulva by Nagulic et al., in 1996. The oldest case report was from Patel et al., in 1973 reporting on transvaginal migration of the shunt. The first case report on bladder perforation was from Mevorach et al., in 1992.
From the data available in 11 patients (on bladder perforation), it was noted that more than 70% (n = 8) of the patients were children and the rest were adults [Table 2]. The incidence of bladder perforation was equally distributed among male and female patients. The presentations among bladder perforation were transurethral extrusion (most common; n = 5), recurrent urinary tract infections (n = 3), lower urinary tract symptoms (n = 2), and urinary incontinence (n = 1). It is not possible to come to a conclusion on whether the infection from the shunt had spread to the urinary system or vice versa. In four patients with bladder perforation, a bladder stone formation was identified. This was believed to be due to foreign body reaction of the bladder to the catheter or supersaturation of urine leading to stone formation.,
Transvulval/transvaginal migration, as the name suggests, presents through the vulva or per vagina. Transvulval migration is believed to occur following migration of the shunt through the posterior wall of the inguinal canal, passing down to the perineum. Transvaginal migration occurs following perforation of the pouch of Douglas.
Management of bladder perforation depends on the presence of infection. Its presence may warrant a shunt exteriorization, whereas its absence may permit an immediate replacement of the distal end of the shunt. All patients need a cystoscopy-guided removal of the distal catheter after its disconnection at the abdomen or the neck, followed by its replacement with a new distal end or its exteriorization. Also, most authors suggested the catheterization of bladder after surgery to allow for a spontaneous healing of the perforation site.,, If associated with bladder stone, a cystolithopaxy or a lithotripsy may be needed., In the presence of transvaginal migration of the shunt, it was suggested that the distal catheter be disconnected at the abdominal end followed by shunt exteriorization. An antibiotic cover should also be initiated for 2–3 weeks and then the distal end replacement into the abdomen may be performed. The protocol suggested is illustrated in [Figure 4].
Scrotal migration (approximately 14%)
The first patient with scrotal migration was reported by Ramani et al., in 1974 following which 55 such patients have been reported. Majority (n = 54) of the patients have been children with only one adult patient being reported by Rehm et al., in 1997. Even among children, most (84.8%) of the migrations occur in children <2 years of age. There has been only one report of herniation of the shunt through the canal of Nuck in a female infant. The largest series of 10 patients with an inguinoscrotal migration was by Pandey et al., in 2010.
The term “clinical inguinal manifestations” (CIM) was described by Celik et al., referring to the myriad manifestations in the inguinoscrotal region following performance of a ventriculoperitoneal shunt (migration, penetration, hydrocele, and hernia). VPS migration into the scrotum (86.5%) predominantly manifests as scrotal swelling, which is reducible. Very rarely, there have been reports of shunt dysfunction /shunt infection  in this location. Most of the migrations are toward the right scrotum [Figure 5] similar to the CIM noted in the series by Celik et al. Most of the shunt migrations occurs within the first 6 months of performing a ventriculoperitoneal shunt [Figure 6] (mean duration: 8.3 months; median duration: 5 months) with a range from less than 24 h to 48 months. There have been four patients in whom migration had occurred within 24 h of performing the VPS.,,, There has been only one case report of scrotal penetration by the distal catheter, which was associated with shunt infection.
The following theories were proposed to explain why the peritoneal catheter migrates to the scrotum: a patent processus vaginalis (PV), aided by peristaltic movements; and, due to raised intra-abdominal pressure.,, Normally, the PV covers the testis and later develops into tunica vaginalis, which separates from the abdominal cavity [Figure 7]. Rowe et al., had suggested that 60–70% of infants (<3 months of age), 50–60% of infants (up to 1 year of age), 40% of children from 2 to 10 years of age, and 15–30% of adults have a patent PV, which communicates with the abdominal cavity. The CSF from the peritoneum thus leaks out through the path of least resistence. Along with the CSF, the catheter is also carried into the scrotum via the patent PV which is described by Wong et al., as the “trough effect.” This movement of the catheter into the scrotum is aided by the peristaltic movement of the intestines. As children have a small abdominal cavity and a higher chance of a patent PV, the incidence of scrotal migration of the catheter is more than seen in adults., The additional length of the catheter was not regarded as an important factor, but this statement has been proven to be false by Ram et al.
In the case of an acute presentation, the importance of Doppler imaging was stressed by Bristow et al., to differentiate a shunt migration from torsion of the testis. The importance of sonography rather than the X-ray images (to avoid radiation exposure) in establishing the diagnosis of scrotal migration of a shunt was proposed by Karaosmanoglu et al., in 2008. An early diagnosis and treatment of the scrotal migration is recommended to avoid incarceration of the shunt catheter that may later on lead to shunt dysfunction/shunt extrusion., Repositioning of the displaced catheter into the abdominal cavity without repair of the sac (PV) was associated with recurrence. Indeed, bilateral obliteration of the sac was suggested  once this complication occurs. Opening the sac and reducing the catheter into the abdominal cavity was suspected to be associated with the risk of infection. Therefore, milking out the catheter before opening the sac was suggested., Laparoscopic repositioning of the catheter and repair of the hernia sac was suggested by Potineni et al. Ward et al., had proposed a protocol for the management of scrotal migration in which he had suggested performing analysis of the fluid in the sac for ruling out infection and had also suggested a baseline and repeat imaging of the brain postoperatively to rule out shunt dysfunction. The proposed protocol of management would be as given in [Figure 8].
Abdominal wall migration (approximately 14%)
This compartment includes migration into the abdominal wall, which may be anterior, posterior, or umbilical. There have been 57 patients reported with this complication. The earliest case report was by Adeloye et al., in 1973 in which a VPS had extruded through the umbilicus. Until 2016, among the 57 patients, 40 had anterior abdominal wall migration (AAW), 10 had an transumbilical (TU) migration, and 7 had an PAW migration of the shunt tube. The largest case series was published by Abode-Iyamah et al., in 2016, with 16 patients being reported.
The age-wise breakup of this complication is shown in [Table 3]. When analyzed site wise, >90% of TU/PAW migration occurred in children, and 87.5% of AAW migrations occurred in adults. It is postulated that in children, the umbilicus is a weak point; also, the PAW is weaker than the anterior wall in resisting the increased intra-abdominal pressure (especially in children with a myelomeningocele), and hence, they are prone for migration at these sites. On analyzing the symptoms, overall 55% patients presented with local symptoms or signs (including swelling, tenderness, erythema, discharge, or extrusion), 25% had shunt dysfunction, 10% had shunt infection, and 10% had asymptomatic TU extrusion. Interestingly, all of the cases of shunt infection were noted in TU extrusions, and all of the cases of shunt dysfunctions were noted in AAW migration. Extrusion through the intact skin was observed in 15 patients (10 umbilical; 3 PAW; 2 AAW) but infection was associated only with TU migration. Though more cases are needed to unequivocally determine this fact, it may be suggested that TU migrations are more prone to develop infections (in 60% patients), and AAW migrations are more prone to develop shunt dysfunctions (in 40% patients).
The time interval after the last shunting (an overall figure for all three types [AAW + TU + PAW] of abdominal wall extrusions) was available in only 35 patients. Among them, 22 (62.8%) of the migrations occurred within 3 months of the last shunt placement and the rest beyond that with the earliest after 10 days  and the latest after 14 years [Figure 9]. In their review, Abode-Iyamah et al., had also quoted that 68.8% of their migrations occurred within 3 months.
It may be suggested from the trend that migration into the abdominal wall is a relatively early process following placement of a shunt, which may be due to a sudden increase in abdominal pressure that the cavity is not able to accept. Thus, the shunt tube is pushed out.
Why does a shunt catheter migrate through the umbilicus? This question has not been dealt with by the articles which describe such patients. From the available data, it is believed to be due to an increase in abdominal pressure in children following a shunt placement, which leads to breakthrough at the weakest point that is the umbilicus (this constitutes the direct path). As an alternative pathway, there may be migration to the embryological umbilical remnant and indirectly through the umbilicus (this constitutes the indirect path), as described in a case report by Kella et al., (where the shunt tip migrated into the urachal remnant). The shunt may also migrate along an indirect path with an artificially created connection with the umbilicus, as described by De Jong et al., in a patient in whom a Mitrofanoff procedure (an appendicovesicostomy) was performed. Migration to the back (PAW) may also be due to a poorly developed or a weak back musculature which gives way when there is increase in intra-abdominal pressure.
Several theories explain the AAW migration. The most common theory related to the presence of obesity in patients.,,, In the review of 50 patients, 33 had their body mass index (BMI) mentioned. Among the 33 patients, 88% (n = 29) were obese (with a BMI >30 kg/m 2). Nakahara et al., in his case report, had proposed that the normal intra-abdominal pressure increases from 0 to around 8 mmHg in an obese individual, which is further increased following placement of a VPS, and hence, the distal tube is pushed out. Nagasaka et al., had proposed another interesting theory called the “fat pad shift.” According to this theory, excess abdominal wall fat moves downward when a patient changes from the supine to sitting or standing position, which in turn pulls the catheter outside the abdomen. Hydrogel catheters were introduced to reduce the tissue damage and infection, but these catheters were found to have minimal frictional resistance and, hence, easily pulled out of the abdomen., The other theories proposed are the conduction of a vigorous exercise; straining, coughing, or sneezing that lead to a sudden increase in abdominal pressure; a longer length of the catheter; and the lack of purse-string sutures at the peritoneum. This last point (purse string suture) may not hold true, because in the series by Abode-Iyamah et al., all the patients who reported with shunt migration had purse string sutures being applied during surgery. Another factor analyzed and found to be significant by Abode-Iyamah et al., was a history of previous shunt revision. In the review, only 26 patients (out of a total of 70 patients) had mentioned whether the procedure being conducted was shunt revision or a de novo shunt placement, and among them 21 (84%) had shunt revisions. Whether or not this fact is significantly associated with abdominal wall extrusion of the shunt needs further evaluation.
The management protocol proposed to manage abdominal wall migration is given in [Figure 10].
To avoid shunt migration to the abdominal wall, several novel techniques have been attempted. Use of laparoscopy for distal catheter placement has been proposed to reduce migration as it is associated with a smaller scar and the shunt placement is performed under direct vision., Using a polypropylene mesh to weave the catheter around it, and to increase the friction, has been tried by Morrison et al., and Abode-Iyamah et al., and found to be successful., Nagasaka et al., had advised a subfascial tunneling (i.e., performed above the rectus abdominis muscle) to avoid the fat pad shift related migration.
Intracranial/Subgaleal migration (approximately 11%)
This section combines the intracranial and subgaleal migration of the shunt system. These types of shunt migrations are referred to as “upward migration” based upon the description by Mclone et al., in 1989 (not to be confused with the term cranial/upward migration used in this article). Incidence of complete intracranial/subgaleal migration was reported to be 0.1–0.4% of the total shunt-related complications. There have been 42 patients among whom 30 (72.7%) had an intracranial migration and 12 (27.3%) had a subgaleal space migration. The earliest report of intracranial migration was by Scott et al., in 1955, and the earliest migration into the subgaleal space was reported in 1994 by Heim et al. Intracranial migration includes migration into the ventricle (n = 24), the brain parenchyma (n = 4), or the subdural space (n = 2). The author of this paper has also described a case of subdural migration of the shunt.
It is to be noted that 90.5% of the patients with an intracranial/subgaleal migration belong to the pediatric age group. This may be related to the shorter distance that a catheter has to travel upwards in a child than in an adult. Most of the intracranial/subgaleal shunt migrations present as shunt dysfunction (n = 28, 66.6%) and there has been no report regarding shunt infection in this group. Local symptoms in the form of subgaleal swelling (n = 11, 26.2%), seizures, and visual disturbances  (intraparenchymal migration) have been reported. Most (69.2%) migrations occur within 3 months after the shunt placement and there has been a declining trend after that [Figure 11]. Most migrations occurring within 3 months may favor the mechanism of sudden fall in intracranial pressure (sucking effect), sudden increase in intra-abdominal pressure (pushing effect), faulty fixation, and a retained memory of the tube. There have been case reports of migration occurring within 3 days  of shunt placement to up to 10 years  post-shunting (the median duration being 2.25 months).
With respect to the patients, children are more prone than adults because of the shorter distance the catheter has to travel from the caudal to cranial end, and the growth spurt in them which favors the pulling out of the shunt., A malnourished child is at a higher risk because the subcutaneous fat is inadequate to hold the tube in position., Similarly, in a child with cortical atrophy, the brain matter cannot hold the proximal catheter inside the brain effectively., Repeated head movement in children (flexion-extension) produces a windlass effect/ratchet movement which helps in pulling the catheter upward.,, Also, when a child lies prone, the repetitive head lifting (extension) also aids in migration of the shunt into the intracranial/subgaleal space. The negative pressure in the ventricles (due to the open anterior fontanelle) causes a siphon/sucking effect and a positive pressure in the abdomen (that is filled with CSF) creates a pushing effect that aids in migration of the shunt.,,, Considering the shunt systems that were utilized, shunts without reservoirs/unishunt systems are more prone to developing a shunt migration.,, Another important factor is the retained memory of the shunt tube, which may be responsible for the recoiling after placement.,, The surgical technique may also help in the upward migration in the following ways:,,, Placement of a large burr hole or a large dural opening encourages shunt migration; an occipital burr hole has a shorter distance from the abdomen than a frontal burr hole; a larger subgaleal pocket,, or multiple subcutaneous passages encourage shunt displacements; and failure to anchor the shunt system to the pericranium leads to shunt migration due to the lack of tethering of the shunt tube. Few authors have also suggested anchoring the shunt system at the abdominal end in order to avoid intracranial/subgaleal migration., Some authors have suggested avoiding an occipital burr hole. They favor the frontal location of the burr hole. On evaluation based on which shunt system was utilized, data were available from only 22 patients. It was found that 68.2% (n = 12) had either a Chhabra (cylindrical shunt) or a unishunt system. In order to avoid upward shunt migration, the use of a reservoir is suggested rather than a unishunt/valveless shunt system, and anchoring the shunt tubing under the pericranium is considered as an important step.
For a complete intraventricular migration of the shunt tube, a craniectomy or an endoscopic retrieval may be planned.,,, The existing shunt system is removed because it is considered to be prone to migration. When the new shunt system is placed, certain general principles have been suggested by various authors to avoid the migration again.,,,,,,,,,,,, These include selecting a new entry point; placement of a new shunt system; the use of a subcutaneous, large-sized reservoir; the creation of minimum subcutaneous passages and subgaleal pockets; the passage of the shunt catheter under the pericranium; and the making of a small burr hole and a small dural opening. At times, there may be asymptomatic migration, which may be observed without intervention unless shunt dysfunction develops., The protocol suggested for symptomatic intracranial/subgaleal migration is given in [Figure 12].
Chest/Thoracic migration of the VPS (approximately 8%)
The first patient with a distal catheter migration into the thoracic cavity was reported in 1977 by Obrador et al., in a 14-month old child. After that report, 32 such migrations have been reported. About 81.3% (n = 26) of the patients are children with the rest being adults (n = 6; 18.7%). This group includes migration to the thoracic cavity (n = 28) and extrusion through the chest wall (n = 4).
Taub and Lavyne in 1994 had classified the thoracic shunt-related complications of VPS as: Sustaining an intraoperative trauma while placing the shunt (leading to lung injury and/or pneumothorax), pleural effusion in association with CSF ascites, and catheter migration into the thoracic cavity (leading to pleural effusion/tension hydrothorax/pneumonia/broncho-pleural fistula). Migration inside the thoracic cavity presents predominantly as acute/chronic respiratory distress depending on the pathology in nearly 78.8% patients, and less often as shunt dysfunction/infection in approximately 15.1% patients. On the other hand, migration to the chest wall manifests as a swelling/blister which ruptures and there is extrusion of the shunt tube. These patients are often associated with shunt dysfunction (in 50% patients) or infection (in 25% patients). Following a VPS, migration into the thoracic cavity occurs as soon as within 5 days  to as long as 16 years  after the last surgery (average duration: 33.6 months; median duration: 5 months). Looking at the trend, most migrations occur in the first 6 months after surgery with a decline in incidence as time progresses [Figure 13].
Intrathoracic migration of the distal catheter is divided into the transdiaphragmatic (through the diaphragm) or the supradiaphragmatic (through the chest wall) migration by Taub and Lavyne  [Figure 14]. The spear-headed Heyer–Schulte subcutaneous passer was discontinued due to the risk of erroneous tunneling and the intrathoracic migration of catheter.,,, The positive pressure in the abdomen and the negative intrathoracic pressure favors migration into the thoracic cavity., The role of the direction of tunneling in penetration of the thoracic cavity was raised by Patibandla et al., but only two papers have mentioned the direction of tunneling (both cranial-to- caudal), and hence, this fact could not be evaluated.,
Congenital defects in the diaphragm include the foramen of Morgagni/Bochdalek and the right xipho-costal point (from where the superficial epigastric vessels enter). Migration of the catheter has been noted through each of these points.,,, Chronic abdominal infection weakens the diaphragm and the inflammation helps in anchoring the catheter to the diaphragm, following which perforation takes place due to constant friction., The stiffness of the catheter, the incision close to the costal margin, the bellow-like effect of the diaphragm, and the positive intra-abdominal pressure further increase the risk of perforation. Also, the anchoring of the distal catheter at the abdominal end, or the placement of the incision near an old scar, can prevent the free movement of the catheter inside the abdominal cavity and promote its constant rubbing against the diaphragm.
The importance of the lateral view X-ray imaging of the chest is to analyze the course of the distal catheter and also to look at the vertebro-costal angle (foramen of Bochdalek)., With the advent of CT scan of the chest, the role of lateral X-ray has decreased though the latter may be used for screening purposes. Only 15% of thoracic cavity migration of the shunt is associated with shunt dysfunction, and so, the role of CT scan of the brain is limited, unless there are “red flag” signs visible.
Management of shunt extrusion would be based on the presence or absence of shunt dysfunction/infection. Repositioning of the catheter is associated with a theoretical risk of transmitting latent infection (one case report exists of a pseudocyst developing after 3 years). However, repositioning is a much simpler procedure as compared to its replacement. Replacement of the catheter, on the other hand, is a relatively more tedious procedure but has a lesser chance of shunt infection. In patients with shunt infection (incidence: 4%), exteriorization of the catheter and treatment with antibiotics is suggested. In patients with associated defects in the diaphragm, repositioning the shunt into the peritoneum will need a diaphragmatic repair also, or the shunt may have to be converted to a VA shunt. Doh et al., and Rahimi et al., had also suggested that avoiding long distal catheters, anchoring at the abdominal end, or incision near the previous scar may help in reducing the chance of migration into the thorax., The protocol for management of thoracic migration is presented in [Figure 15].
Cardiac/Intravascular migration (7%)
Migration to the vascular compartment includes migration to major vessels or cardiac migration. Majority of the group is contributed by cardiac migration (n = 27) and there is one case report of migration to the inferior vena cava. The earliest migration to the heart was reported by Morell et al., in 1994 in a 12-year old child. An excellent review of most migrations to the heart was provided by Carrasco et al., in 2015, in which he had reviewed 26 patients with intracardiac migration. He had also proposed the terminology “cardiac migration of peritoneal catheter.” Out of the 28 patients reviewed, 19 were adults and 9 were children.
The clinical presentation was a shunt dysfunction (n = 10; 35.8%), presence of cardiac symptoms (n = 8; 28.6%), or other local symptoms (n = 6; 21.4%) like swelling/pain/tenderness; and the discovery was incidental in six patients. There was no mention regarding shunt infection in most of the articles. Shunt migration into the heart took place anytime between 5 days and 4 years with an average duration of 7.9 months and median duration of 2 months. Most of the migrations occurred within 12 months (in 20 patients, out of the data available in 23 patients), especially within the first 3 months (n = 15; more often in children) [Figure 16].
Two theories have been proposed regarding migration of the shunt to the heart., Either an unrecognized damage to the internal jugular vein (IJV) while tunneling occurs or a chronic erosion of the IJV occurs due to close proximity of the shunt to the vessel. Following the entry into the IJV, migration to the heart is favored by the presence of a negative intrathoracic pressure (that sucks the tube) and the venous flow to heart (that pushes the tube). In most shunt migrations that occur within the first 3 months of life, the role of chronic erosion of the vessel appears less likely. Only one case report has mentioned about cranial-to-caudal tunneling during placement of the shunt that may be responsible for shunt migration  and two case reports mention a profuse bleeding during the tunneling procedure., It is also equally difficult to conclude whether or not injury to the IJV could have occurred during the subcutaneous passage of the catheter.
A CT scan of the chest would be advisable to confirm the diagnosis of migration into the heart followed by an echocardiogaphy (to assess heart function) and a CT angiography (to assess the flow in cardiac vessels). As around 36% of these patients were associated with shunt malfunction, imaging of the brain may also be suggested. Once a diagnosis is arrived at, surgical intervention is essential as cardiac migration may develop into an unstable cardiac condition. The basic principle of surgical intervention is the removal of the distal catheter from the heart followed by its repositioning into the peritoneum or its conversion into a VA shunt., During the withdrawal of the shunt, catheter arrhythmia may occur which needs cardiac monitoring. Also, if the catheter is knotted/tangled, it may damage the valve, and hence, withdrawal should be done under echocardiographic/fluoroscopic monitoring., At times, the catheter may adhere to the IJV at its point of entry into the vascular compartment, and to remove the catheter, a venotomy may be necessary. In case the catheter is entangled with a valve or cannot be pulled out easily, an endovascular intervention will be the ideal approach.,
Breast migration (approximately 3%)
Breast-related complications with a VPS can be classified as follows:, Breast CSF pseudocyst, shunt migration, CSF galactorrhea, and shunt obstruction. Among them, shunt migration is the commonest complication seen. The earliest breast-related complication in a patient with VPS was retrograde CSF tracking noted by Nakano et al., in 1994 and by Morón and Barrow (following ventriculo-pleural shunting). In the last decade, a rise in breast-related complications have been noted (especially shunt migration), probably related to the increase in breast implants and mammoplasty. There have been 13 case reports of VPS distal catheter migration into the breast until 2016, following the first report in 2005 by Spector et al.
Shunt migration to the breast reported till date has been in adults, with only one report of migration in a 13-year-old female child. All the patients presented with asymmetrical swelling of the breast (100%) with tenderness/erythema noted in 25% of the patients (n = 3). Nipple discharge (may be due to damage to the lactiferous duct) was observed in only one patient. Only one (7.6%) patient had shunt dysfunction with none of the patients having shunt infection. Eight (70%) patients with shunt migration had a preexisting breast implant. Hence, it may be suggested that the presence of an implant definitely increases the risk of migration into the breast.,,, Migration to the breast has been reported to occur from 1 month , to 36 years  after the VPS procedure, with most (85%) migrations occurring after <6 months of shunting (with a median duration of 2 months).
The mechanism leading to migration of the distal catheter to the breast may be summed up as follows: The pulling effect of an implant or an inadvertent passage through the implant, along with the recoiling nature of the shunt tube, promoted by the pushing effect of the raised abdominal pressure (cough/sneeze/strain/obesity) favors migration to the breast. There has also been a case report of migration to the breast following vigorous breast manipulation. This may also be a propagating factor [Figure 17].
Migration to the breast may be avoided by eliciting a careful history regarding the presence of implants and making a guarded tunneling (away from the convexity of the breast). In a patient with VPS, if an implant is planned, it would be advisable to place the implant deeper to the pectoralis major muscle (in the submuscular plane) to avoid complications. A few authors have also suggested anchoring at the abdominal end to avoid the pushing/pulling effect.
Miscellaneous (approximately 4%)
This includes migration to the liver (n = 7), neck (n = 6), and to the thigh (n = 2). The earliest case report on hepatic migration was by Wolbers et al., in 1987. This focused on hepatic perforation in a 29-year-old female patient.
Migration to the neck is said to follow the same mechanism as upward migration and is related to factors like an excessive subcutaneous tunneling, a very long length of the catheter, and the windlass effect. The proximal catheter may also descend to the neck if associated with poor anchoring to the pericranium or due to the lack of a reservoir. Migration to the thigh is considered to be very rare with only two case reports focused on this site. This is said to occur following migration via the femoral canal.
Most migrations are noted in children but adults have a greater tendency for migration to occur to the heart/breast/abdominal wall. The most common site for VPS migration is to the gastrointestinal system. Most migrations present in the first year following a VPS, beyond which the chances of migration decrease gradually. In any patient, who has undergone a shunt procedure and presents with non-neurological symptoms/signs, shunt migration should be considered. Migration of the shunt to the bowel, heart, as well as the intracranial and the subgaleal space is associated with >20% chance of shunt dysfunction. Any extrusion (irrespective of the site) where the shunt catheter is exposed to the external environment is associated with around 50% chance of shunt infection. Most shunt migrations need surgical management in the form of shunt replacement/shunt repositioning either immediately or after a course of antibiotics. Shunt migration is an avoidable complication that requires intricate surgical techniques of management. The complication needs to be focused on in more details. This focus at present is lacking, as highlighted by this review.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
[Figure 1], [Figure 2], [Figure 3], [Figure 4], [Figure 5], [Figure 6], [Figure 7], [Figure 8], [Figure 9], [Figure 10], [Figure 11], [Figure 12], [Figure 13], [Figure 14], [Figure 15], [Figure 16], [Figure 17]
[Table 1], [Table 2], [Table 3]