Most cases of shunt malfunctions occur due to occlusion (blockage) of the proximal ventricular catheter. In these instances, pumping of the shunt will show a valve that is slow to refill, or does not refill at all. An imaging (CT or MRI) scan will show if the ventricles are of normal size and if the shunt is working properly.
In a majority of cases, infection is the cause of shunt malfunction when a distal blockage is suspected. A preoperative CSF specimen from a shunt tap should be obtained to exclude this possibility. The proximal system can be tested by ensuring free flow of CSF, whereas the distal system can be tested by runoff using a manometer.
The signs and symptoms of shunt malfunction are the same as for hydrocephalus itself: headaches, nausea, vomiting, irritability, change in behaviour or intellectual performance, etc.
In most cases of hydrocephalus shunt malfunction, the diagnosis is obvious because of the apparent signs of elevated intracranial pressure, such as headaches, vomiting and lethargy. In about 30% there will only be subtle signs of deterioration, with neuropsychologic, cognitive, and behavioural symptoms heralding the shunt dysfunction.
In infants an ultrasound examination can be used as a diagnostic tool if the skull has not yet fused. A CT scan is used for diagnosing older children and adults. The size of the ventricles can be compared to the "normal values". The person's size, age, and growth progress (in children) will indicate whether the ventricles are too large.
Once there is a suspicion of a shunt dysfunction, a CT scan or MRI scan is used to compare the ventricular size and show the most definitive signs of a malfunction. This is only useful if a previous scan can be used for comparison. In cases where the symptoms of a shunt malfunction are present but the scanning shows no evidence, the next step involves a shunt tap test.
A shunt tap is performed after washing the skin over the shunt with a sterile antibacterial solution. After placing a small needle through the skin and into the shunt, the spinal fluid pressure can be measured. Fluid is withdrawn to test for infection and to see if symptoms improve temporarily.
There is a heightened risk of shunt blockage in the first few months following placement. The proximal end (the end within the ventricle of the brain) is the most likely to be blocked. The blockage is usually caused by chloroid plexus or blood generated during the placement of the catheter.
The catheter exerts a suction effect which can draw choroid plexus, blood, and debris towards the holes in the catheter, and this contributes to obstruction. Shunt designs, such as those with siphon control devices or flow regulated valves, may alleviate this cause of late shunt occlusion.
Distal end occlusions are less common than proximal ones. If the used shunt has a distal slit valve, progressive debris accumulation may lead to its occlusion. Ventriculo-atrial shunts may occlude due to thrombus formation if the distal end moves out of the atrium. Ventriculo-peritoneal shunts will malfunction if the peritoneum loses its absorptive capacity. This will typically present an increasing abdominal girth. An abdominal ultrasound will show intraperitoneal fluid collection.
To prevent proximal catheter blockage, the catheter needs to lie anterior to the foramen of Monroe. Having an occipital placement with a catheter long enough to reach the front of the ventricular system, or a frontal catheter placement, will help hinder hydrocephalus shunt malfunctions.
It is important to design the incision so that no part of the shunt hardware or shunt system lies under the incision line. The use of the frontal catheter placement is sometimes useful when the symptomatic slit ventricle syndrome occurs. The occipital placement needs to be far enough posteriorly, so that one does not risk placing the catheter through the internal capsule. When an error in placement of the ventricular catheter has been made, and this is detected on the postoperative scan, an assessment of the condition of the patient should be made before deciding on elective revision. Suboptimally placed ventricular catheters do not have to be automatically revised.
There are two situations, however, which have a high incidence of leading to shunt failure, and elective revision should be considered. The first is the situation where the catheter is barely in the ventricle, so that when ventricular decompression occurs the catheter would actually become intraparenchymal.
The second situation is a catheter that is located in the anterior part of the temporal horn where subsequent shunt occlusion is very likely to occur.
Intraventricular endoscopes can be used to accurately place ventricular catheters, to fenestrate cysts, and to reconnect loculated ventricles. Intra-operative ultrasound can also be used to assist catheter placement and locate cysts.
When blockage of a ventricular catheter is found, the catheter is usually sticking to the choroid plexus and it can be difficult to remove. These malfunctions are usually associated with the quick development of elevated intracranial pressure, so that a rapid shunt revision should be performed shortly after the patient arrives at the hospital. The revision is performed, replacing the occluded catheter with a right angle ventricular catheter of an appropriate chosen length. This is then connected to the rest of the shunt assembly with a single straight connector.
In cases where the ventricular catheter is stuck, several manoeuvres may help free the catheter and avoid intraventricular haemorrhage. First, the ventricular catheter can be grasped with a haemostat and rotated. This may free the catheter from the underlying choroid plexus, and the resistance my give way suddenly. If the twisting of the catheter does not free it, the next step is to place the stylette down the shunt catheter and touch the Bovie cautery to the stylette. This sometimes will coagulate the choroid plexus at the tip of the catheter and release it. If there is still firm resistance, the catheter should be left in place, a new burr hole should be made next to the existing one, and a new ventricular catheter should be utilised. Intraventricular endoscopy may be particularly useful in this situation. The adherent choroid plexus can be coagulated and freed from the ventricular catheter using the working channel of the endoscope.
If a proximal revision needs to be performed in the face of relatively small ventricles, there are special considerations that will ensure the safety of this operation. It usually occurs in the setting of shunt dependency where the ventricles were slit-like when the patient was well, but have dilated only slightly with the shunt malfunction. It is frequently wise to place a tube at a frontal site leaving the blocked occipital catheter in place.
A second alternative is to attempt to slide a new ventricular catheter down the same tract following the removal of the old ventricular catheter. No attempt should be made at placing the catheter with a stylette or brain needle, since it may veer off the tract making it difficult to hit the small ventricle. If the catheter does not find its way into the ventricle, then a single attempt with a brain needle or stylette in the ventricular catheter can be made to try to make an opening in the small ventricle. If this is unsuccessful however, repeated attempts should not be made, and the operation should be aborted. The patient should be watched closely for signs of elevated intracranial pressure, and receive a series of CT scans. Subsequently, a proximal shunt revision should be performed when the ventricle has dilated. In this setting it may then be useful to utilise the frontal shunt placement, where it can be easier to hit a small ventricular system.
If when performing a proximal revision, and on placing the new ventricular catheter, one obtains blood tinged CSF, then an effort should be made to clear the ventricular system of this bleeding (which originates from the choroids plexus). The new catheter should be attached to a three-way stopcock and irrigated with saline until it begins to clear. This often requires patience, as it may require up to 20 minutes of gentle irrigation with saline. Once the CSF does clear, a new ventricular catheter should be placed so that there is no risk of a catheter being left in place which is occluded with a blood clot. These patients obviously need to be watched closely for signs of shunt failure within the first few hours after a revision, in which case a temporary ventriculostomy may be needed until the blood clears.
Most distal malfunctions that are not associated with a short catheter are due to shunt infections. In the face of a distal malfunction, one should look carefully at the CSF prior to shunt revision, to make sure that an infection is not present. The presence of an abdominal pseudocyst detected on abdominal ultrasound or CT scanning should be considered a shunt infection until proven otherwise. It is especially important to allow the fluid to be analysed in the laboratory for longer than usual, so as to look for diptheroids which may not grow on the culture medium in the first two days.
Distal shunts have been known to erode into various abdominal viscera. There are reports of shunts eroding into the intestine, the bladder, the vagina, and even protruding from the anus. Most of these complications were seen with the spring loaded distal shunt tubing, which now should be avoided.
The second most common cause of shunt failure is disconnection or fracture of the shunt. Disconnection may occur at any site of connection along the course of the tubing. This is usually related to improper technique (loose ligature), or excessive strain along the shunt tube between two points of fixation. The incidence of disconnection has been decreased by the introduction of one-piece shunts. Shunts are made from silicone rubber. Silicone is very flexible but may deteriorate in time after implantation. Calcification of the outer wall of the shunt may produce fixation to the subcutaneous tissue, most commonly in the neck or along the chest. This fixation may lead to stretching of the tube and result in fracture or disconnection. Palpation along the shunt tubing might not reveal the fracture as the examiner will feel a firm, fibrous sheath that envelopes the shunt. Definitive diagnosis is made by visualising the disconnection or fracture of the shunt on radiographs.
Mechanical hydrocephalus shunt malfunction will usually present with the classic signs of increased intracranial pressure: headaches, nausea, vomiting, and possibly papilledema. Infants will demonstrate irritability, full fontanelle, and rapidly increasing head circumference. Diagnosis is confirmed by comparing a radiograph study (cranial ultrasound, CT, MRI), obtained while the child is symptomatic, with a baseline study obtained previously. The comparison involves documenting an increase in the size of the ventricles. If the diagnosis of shunt malfunction remains unclear, or a previous study is unavailable for comparison, a shunt function study is performed. The test is run by injecting a small amount of Technetium (a radioisotope) into the reservoir of the shunt. By following its flow using scintillation detectors, the test can demonstrate the presence or absence of flow within the shunt.
Over-drainage is a more difficult problem. Although a higher pressure shunt will initially solve the problem, it is not usually a long term solution. Studies have shown that the use of an 'anti-siphon device' (a small button inserted into the shunt tubing) will often solve the problem, but this does not always work. Some shunts have these built-in, but neurosurgical opinion varies as to whether they should be used. To change a valve pressure it is necessary to remove the valve and insert another. The 'programmable' or adjustable shunt is intended to allow adjustment of the working pressure of the valve without operation. The valve contains magnets that allow the setting to be changed by laying a second magnetic device on the scalp. This is useful where the need for a valve of a different pressure arises. The adjustable valve is no less prone to over-drainage than any other and it cannot be used to treat this condition.
It has long been believed that a raised protein level in the CSF will block the shunt. In babies with hydrocephalus shunting has been delayed until the protein level has fallen. Recent research has shown that a raised CSF protein level has no ill-effect on shunt function, nor does it increase the risk of infection, and there is now no reason for delay unless blood is also present.
Over-drainage may lead to a variety of problems such as subdural haematoma, post-shunt craniosynostosis, and slit ventricle syndrome. As postural changes and patient height are partially related to the over-drainage, these problems, with the exception of craniosynostosis, are seen mostly in children and adolescents rather than infants.
Both subdural haematoma formation and the occurrence of post-shunt craniosynostosis are caused by "craniocephalic disproportion." In paediatrics, this is the result of a large head created by a hydrocephalic brain. The head is unable to decrease in size to the same extent that the brain does when the hydrocephalus is treated, and this creates a space between the inner surface of the skull and the outer surface of the brain. A subdural haematoma is formed when the veins bridging between the brain and skull become stretched and tear. This may occur spontaneously or be the result of relatively trivial injury. Some haematomas may be asymptomatic, filling the space created by decompression of the hydrocephalus and resolve themselves spontaneously. If the haematoma is causing symptoms (headaches, vomiting, somnolence) treatment options include a craniotomy and evacuation of the haematoma, or placement of a shunt into the subdural space.
Post-shunt craniosynostosis is the result of apposition and overlapping of the cranial sutures in an infant, following decompression of hydrocephalus. Surgical intervention is usually not necessary, but, when indicated, requires a reshaping of the entire skull.
The slit ventricle syndrome is usually seen after the shunt has been in place for several years. It is characterised by chronic or recurring headaches, and by slit-like ventricles shown by CT. The slit ventricle syndrome is not a single pathologic entity; but it is a complex symptom with several etiologies. The evaluation and management of the slit ventricle syndrome requires a systematic and comprehensive approach.
The initial step in the evaluation of the slit ventricle syndrome is to ensure patency of the shunt using the radioisotope shunt function study discussed above. If the shunt is occluded, a shunt revision is performed.
If the shunt is functional, the next step is to monitor the intracranial pressure. This is achieved by the placement of a small fiberoptic transducer into the outer surface of the brain so as to measure the pressure over several hours or days. The child or parents record when the headaches occur, and this is correlated with their pressure recordings. Consistently high intracranial pressure, or frequent waves of high pressure of a functioning shunt, imply a reduction of the buffering reserve normally afforded by CSF. Increased intracranial pressure, which would result from a rise in intracranial blood volume, is normally prevented by increased absorption of CSF. A constant decrease in the amount of CSF in the intracranial space (from over shunting) precludes this protective mechanism and allows increased intracranial pressure and headaches with increases in blood flow.
These children invariably have a relatively small head circumference and thick skull. The preferred treatment is a cranial expansion. This will increase the intracranial volume and allow for the normal changes in blood flow. Headaches may occur with evidence of low intracranial pressure, a condition termed intracranial hypotension. These headaches usually occur in the upright position and are relieved by lying down. They are due to over-drainage by the shunt and are treated by the addition of an anti-siphon device to the shunt system.
Under-drainage, in which the fluid is not removed quickly enough and the symptoms of hydrocephalus return, is one of the most common problems. It is usually due to blockage of the upper or lower tubes of the shunt tissue, though it can be due to the shunt breaking, or due to its parts becoming disconnected from each other. It is rarely due to the valve itself, which usually continues to function in the same way for years. Pressure may sometimes build up rapidly resulting in loss of consciousness, and treatment is required as an emergency.
However, in most cases the onset is more gradual and can follow a minor illness such as a cold. Headaches increase in frequency and severity, often worse on waking in the morning. Vomiting and dizziness also occur, and sometimes other symptoms which vary from patient to patient. In these cases the parents or carers will be able to recognise the symptoms from previous episodes. Specialist hospital staff are now fully aware of the various presentations of 'blocked shunt', but non-specialists and family doctors may not be.
Shunt blockage can also have much more subtle consequences and the headaches may be infrequent, with the main problem being behavioural deterioration. In older children this might take the form of increased irritability, 'laziness', poor or disruptive school performance, or even more antisocial activity. This may be very difficult to distinguish from the usual teenage angst but, if there is any reason to suspect that the deterioration in behaviour is not 'normal', assessment must be carried out by an experienced educational psychologist with knowledge of hydrocephalus.
The basis for the effects of high CSF pressure has been explained in the previous article. If the shunt is to blame, a dramatic improvement can result from appropriate treatment, though this form of shunt problem is particularly difficult to diagnose. It may be necessary to monitor CSF pressure, often over 24 hours. This can be done using a pressure monitor in the scalp connected to a recorder. In this way pressure can be recorded during sleep and for changes in posture. Scans showing the size of the ventricles are particularly useful if they can be compared to previous scans, though in someone with clear symptoms of either high or low CSF pressure they may also serve to support the diagnosis.
In the case of over-drainage, the shunt allows CSF to drain from the ventricles more quickly than it is produced. If this happens suddenly, usually soon after the shunt is inserted, then the ventricles in the brain collapse, tearing delicate blood vessels on the outside of the brain and causing a haemorrhage ('subdural haematoma'). This can be trivial or it can cause symptoms similar to those of a stroke. If the over-drainage is more gradual, the ventricles collapse gradually to become slit-like ('slit ventricles'). This often interferes with shunt function causing the opposite problem, namely high CSF pressure, to reappear. Unfortunately the slit ventricles do not always increase in size again, producing the situation where there is very high CSF pressure with headaches, vomiting, etc. but very small ventricles on scan.
Epilepsy has been associated with shunt-treated hydrocephalus. However, it is not thought to be directly related to the placement of the shunt, but rather to the original cause of the hydrocephalus. Epilepsy is a broad subject which I hope to elaborate on in the future.
The rate of shunt infection is about 10-15%, and 95% of infections will occur in the first 5 days after surgery.
The signs of shunt infection may include fever, neck stiffness, light sensitivity (also called photophobia), headaches, or signs of hydrocephalus shunt malfunction. Shunt infections can present with signs of meningitis and ventriculitis. In addition, signs of septicaemia or peritonitis can be seen, depending on the type of shunt. The skin may redden over the area of the shunt and tubing, or the wounds may be reddened and/or draining pus. Most commonly the bacteria responsible are those that reside normally in the skin of the patient i.e. staphylococcus species. Distal shunt malfunctions frequently accompany shunt infections. In VA shunts blockage due to infection is rare, and many months or years can go by before the infection becomes apparent. During this time there will be tiredness, irritability, poor appetite, various aches and pains, skin rashes, and other signs, but all of these can be due to common disorders. A blood test will usually reveal anaemia. This is important although, on its own, it is not a specific indication of infection. Blood cultures and even CSF cultures can be negative. Later, blood may appear in the urine due to secondary kidney damage.
Diagnosis is made by culture of the wound if there is drainage or, more commonly, by culture of the cerebrospinal fluid within the shunt. Fluid is sampled by inserting a needle into a reservoir of the shunt. The major risk is low- approximately one percent. Factors such as time from the last surgery, and the presence of other possible causes of the fever, are taken into consideration.
Most shunt infections are caused by either Staph-epidermidis or auerus. Both bacteria are normally found on the skin and are presumably implanted at time of surgery. Infection usually becomes apparent within the first few months following surgery: 50 percent by two months and 90 percent by six months after surgery.
Shunt infections occur with in 2-8% of shunt operations. This is usually associated with a general infection following surgery.
Although statistics are difficult to find, one review article found that the expected shunt infection rate ranged between 10-20%.
Overall, between 5 and 15% of shunts can be expected to become infected over their lifetime. Of these infections, 70% are diagnosed within one month after surgery, and close to 90% by six months.
The treatment of shunt infections is somewhat controversial. The standard treatment is to remove all of the shunt hardware system, usually done within 2 days of the diagnosis, and treat with appropriate antibiotics. Approximately 10 days later, a new shunt system can be placed. While the child is on the systemic antibiotics, a temporary ventriculostomy may be necessary to control the hydrocephalus.
Once the infection is cleared, the new VP shunt is placed in the same site unless there are skin abnormalities. In such cases it is necessary to place the shunt on the opposite side. After a shunt has been externalised for approximately 7 days and CSF cultures have cleared, a completely new VP shunt is placed and antibiotics are continued for an additional two days. The antibiotics can then be discontinued.
If however there are signs of peritonitis, or distal shunt malfunction, then the shunt should be removed quickly. A septic child may be too great an anaesthetic risk until the vital signs have stabilised.
In cases where a nonfulminating shunt infection is diagnosed within the first month after a shunt insertion, an attempt can occasionally be made to treat the shunt infection with intravenous and intrathecal antibiotics without removing the shunt system. The families need to understand that this method of treatment may fail, and that ultimately the shunt would have to be removed. However, this method does provide a way of saving the shunt system and successfully treating many early shunt infections. This method of treatment must be abandoned if cultures do not become sterile within several days, or if more aggressive signs of infection are noted.
Antibiotics have not been shown to be of benefit for this purpose, and other measures often have only a temporary effect. The care and expertise of the surgical team is one of the most important factors in reducing the rate of infection to a minimum. However, even in the best of hands infection still occurs. One of our recent developments has been a process which makes shunts resistant to bacterial infection, and we hope that the current clinical trials will show that it is capable of reducing shunt infection by more than 80%.