Clinical aspects

Description

Surgical principles. Shunt revision usually involves the exploration of a shunt with replacement of specific malfunctioning components. A wide skin prep is needed to maximize options for new incisions if needed. Compulsive skin preparation is imperative to avoid infection. One of the most common causes of obstruction is choroid plexus (or other tissue) occluding the proximal (ventricular) catheter. Coagulation of a stylet placed in the catheter can relieve the occlusion but other risks arise, and replacement of the catheter is usually needed.

There is evidence that endoscopic techniques or use of a laser filament within the shunt tubing may decrease the need to replace occluded ventricular catheters (16). Distal obstruction should always raise a concern of possible infection. The use of routine surgical practice guidelines can streamline the entire process, but treatment of a shunt malfunction is necessarily an individualized matter. Identified risk factors for first-time shunt malfunction include age younger than 6 months at first shunt placement, cardiac comorbidities, and use of an endoscope during proximal catheter placement (30). Additionally, obesity is a growing issue in society today and contributes to some unique complications at the time of ventriculoperitoneal shunt placement or revision. Several papers have been published to address avoidance of migration of the peritoneal catheter in morbidly obese patients. A technique of placing a stitch beside the peritoneal catheter, which extends through the peritoneum and the posterior and anterior sheath of the rectus abdomens muscle and subcutaneous fat, to eliminate the dead space available for migration has been described by Nakahara and colleagues (25). There are current attempts at generating an increased safety profile for ventriculoatrial shunts for obese individuals as an alternative to ventriculoperitoneal shunts (17).

A new technique for the revision of the distal catheter was described in which the previously placed distal catheter is exposed at its entry site into the abdomen and a soft guidewire is inserted down the distal catheter into the peritoneum (35). The distal catheter is then removed over the guidewire and a peel-away sheath and dilator are inserted over the guidewire, and the sheath is used to guide the new distal catheter into the peritoneum.

Goals for each patient must be individualized. In general, return or optimization of function, freedom from the symptoms of increased intracranial pressure, and minimization of risk of recurrent shunt problems lead the list of goals.

Indications

Urgent surgical shunt revision is indicated in any patient with clinical signs of malfunction. It should be emphasized that a shunt malfunction is a potentially life-threatening condition and should be treated as such. Asymptomatic malfunction may be found on routine evaluations either by detecting a break in the shunt tubing or by evidence of papilledema on clinical examination. The fibrous tract formed around the tubing can function as a conduit for CSF and, thus, prevent acute symptoms. These tracts are not reliable, and the shunts should be revised in a timely fashion. Asymptomatic individuals with first-time shunt revisions have decreased infection rate and subsequent shunt failure as compared with symptomatic individuals with first-time shunt revisions (38).

Contraindications

There are no true contraindications to shunt revision in the setting of shunt malfunction, other than an uncorrected bleeding diathesis. Signs of increased intracranial pressure must be addressed. Bleeding problems will need to be corrected if present. If there is an active infection, external ventricular drainage may be necessary until the infection is cleared. A suspected infection can sometimes be a relative contraindication to one-stage shunt revision, with relative favoring of externalization or partial removal of the shunt.

Outcomes

Outcomes analysis for shunt function is complicated by the heterogeneity of the disease process. The Shunt Design Trial examined a series of 344 patients prospectively and gives significant data on the value of different valve systems for the prevention of shunt complications (10). This study had a particular interest in the problem of overdrainage, along with focused interest on the siphon-resistant and flow-controlled valves currently on the market. The accumulation of extra-axial fluid as the ventricular system collapses, was noted in 3.4% of the patients studied. This may (but not invariably) require surgical treatment by drainage or revision of the valve. A subsequent review of randomized controlled trials examining clinical outcomes for various shunt designs endorsed no significant differences between fixed parameter and adjustable valves (15). Another overdrainage syndrome, which has been called, “slit ventricle syndrome,” but more appropriately should be called, “symptomatic small ventricles,” often presents recurrent problems with shunts in patients with ventricles that are small and apparently prone to relapsing shunt malfunctions. Although this is often thought of as a low-pressure type of overdrainage, when these shunts malfunction, the intracranial pressures are frequently high. A more important part of the pathophysiological process may be the loss of compliance in the ventricular system, resulting in symptomatic increased intracranial pressure without radiographic change in the size of the ventricle. This can be documented in many patients, with the use of mannitol in the inpatient setting. If the symptoms improve along the appropriate time scale expected with several repeat boluses of mannitol, treatment of the increased intracranial pressure is necessary.

Different approaches have been proposed to diminish overdrainage, and thereby reinflate the ventricles, allowing treatment of small ventricle symptoms or slit ventricle syndrome. These include the use of siphon prevention or flow control valves; the use of programmable, or adjustable, valves, which can be magnetically reset or “programmed” to have a different flow resistance (37); and subtemporal decompression or another type of cranial expansion. We have found subtemporal decompression extremely useful in the management of difficult cases. A well-controlled double-blinded study has not been attempted for this complex clinical problem.

Predictors of preventable shunt failure also include lack of endoscopy, recent shunt infection, shunt type, and center (Dave et all 2019). In pediatric procedures, one study indicated that the number of surgeons in the operating room was associated with shunt failure. In adults, the origin of hydrocephalus was a significant variable (02). In children undergoing a shunt revision, cardiac risk factors, chronic history of seizures, and a history of neuromuscular disease were independently associated with shunt failure (12). In a retrospective study of 74,552 pediatric patients with a CSF shunt, peritonitis, papilledema, and oculomotor palsies were the strongest indicators for shunt malfunction requiring a visit to the emergency department (28).

Adverse effects

The most common and serious risk of shunt surgery is shunt infection, which occurs in the operating suite from airborne bacteria, the patient’s skin flora, or a break in sterile technique. Minimizing traffic in the operating room (allowing no personnel change during the operation and limiting the number of people in the operating room), a careful prep, and an efficient surgery are all critical elements to reducing shunt infection. Intracranial hemorrhage during surgery is uncommon but can be serious. This typically occurs when the choroid plexus invades the ventricular catheter and results in bleeding when the ventricular catheter is removed. If it is particularly adherent and does not respond to stylet coagulation, it can be left in place, and a new catheter can be inserted through a new cranial burr hole. There is growing evidence that supports the use of an antibiotic-impregnated shunt to reduce risk of infection. In a study of 1605 patients in the United Kingdom, antibiotic shunts resulted in a lower rate of infection and reduced cost compared to the use of a standard shunt (22; 23). It is worth noting that the study found no difference in shunt failure rate between an antibiotic shunt and a standard shunt.

Although shunt revision is one of the most common pediatric neurosurgical procedures performed, there are relatively little data about the true incidence of shunt malfunction. Several studies have suggested a failure rate of 30% to 40% in the first year and approximately 50% in 2 years (14). A survey of members of the International Society for Pediatric Neurosurgery collected data on 773 operations (09). Twenty-nine percent of these required reoperations within 1 year. There was a slight increase in the revision rate after emergency operations (compared to elective), and a substantially higher failure rate for shunts inserted in the first 6 months of life (35% to 47%), compared with shunts initially placed in children over 6 months of age (14%). Piatt studied a database of 727 shunt operations over a 13-year period at one institution. Complex shunts (with multiple proximal catheters) had a higher failure rate than simple “linear” shunts, which themselves had a failure rate of 32% at one year (27). Similarly, he found a higher rate of failure in young children and also found that a short interval from one revision to the next had a higher risk of failure for the second operation. One multicenter evaluation of 5092 shunt operations over a 2-year period revealed a failure rate of 16.9% (861) within 90 days of index operation; of the failures, 35.7% were determined to be potentially preventable. The most common etiologies of failure were infection (44%) and proximal catheter malposition (27%) (08). Other factors that contribute to shunt failure or malfunction include etiology of hydrocephalus, type of hydrocephalus, and previous treatments before shunt surgery (29). Many of the factors responsible for shunt failure seem to be outside of the surgeon’s control.

Special considerations

Hydrocephalic women who become pregnant could develop symptoms of increased pressure, because of diminished intra-abdominal reserve in resorbing fluid and increased intra-abdominal pressure. Fortunately, this is an uncommon problem. Diagnostic imaging in this case would favor use of MRI over CT (to avoid fetal radiation), but surgical management in the case of true malfunction would be as in the nonpregnant patient. Endoscopic third ventriculostomy may be effective in this setting.

Clinical vignette

Patient 1. A 7-year-old girl with a history of hydrocephalus secondary to a perinatal intraventricular hemorrhage presented with a headache that progressed over 24 hours, together with nausea and vomiting. There was no papilledema, but a CT scan showed ventricular distention when compared with an earlier baseline symptom-free scan. At surgery, a proximal shunt obstruction was identified, and after replacement with a new ventricular catheter shunt function normalized and her symptoms resolved.

Patient 2. A 3-month-old infant with severe hydrocephalus from congenital aqueductal stenosis presented with irritability and fever; CT scan was performed.

Collapse of the ventricular system from overdrainage was noted, with bleeding into the subdural space. Overdrainage was treated without surgery by keeping the patient flat to avoid a gravitational siphoning effect. The subdural blood, a sign of overdrainage, resolved.

Scientific basis

The structure of a ventriculoperitoneal shunt. Virtually all current shunts for hydrocephalus are made from medical grade silastic, and all but a few (used in small infants or special circumstances) contain an in-line valve.

The ventricular catheter, valve, and peritoneal tubing are shown. There are many brands and varieties of shunt equipment, and none has proven to be superior over any other brand. Surgeons tend to have brand loyalty based on experience and their successes in using a particular valve. The one exception might be the programmable valve, which can be particularly useful in the setting of normal pressure hydrocephalus in adults. In these patients, the pressure of the valve can be manipulated without surgical intervention by the simple placement of a programming wand over the valve. A meta-analysis conducted by Li and colleagues demonstrated that the use of programmable valves is not only safe, but may actually reduce the revision rate and complications of under- or overdrainage, particularly in individuals less than 18 years of age with first-time shunt placement (20). Of note, tablet computers can change the setting of programmable shunts if held near the valve, and patients need to be educated about this possibility (36).

Most valves act as simple resistors and rectifiers to flow. This means that flow in the reverse direction is blocked by a check valve, and the presence of flow depends on sufficient elevation of pressure above the valve in the lumen of the shunt. Therefore, the absolute pressure within the intracranial end of the shunt system may depend significantly on the body position and drainage or, “siphoning,” of fluid in the distal part of the tubing, which is greatest in the upright position. This siphoning effect can be useful in premature infants with soft malleable skulls. Siphoning was recognized as a potential cause of symptoms years ago; thus, the “antisiphon” device was created. This device makes the flow dependent not only on the absolute difference in pressure between the upper chamber and the lower chamber, but it closes the valve or significantly increases the resistance to flow, if actual downstream pressure becomes subatmospheric. Correct functioning of this part of the device depends on the mobility of a membrane within it, along with the access of the diaphragm within the device to the external air pressure. Thus, such devices fail under thick scalp; if they become trapped in the closed position, they inhibit flow, thereby creating a situation otherwise identical to a shunt malfunction.

Liniger and colleagues found no difference in the development of slit ventricles or slit ventricle syndrome with flow control versus antisiphon valves (21).

Shunt failure and CSF physiology. The majority of symptoms related to untreated hydrocephalus are a result of increased pressure. Thus, the symptoms and signs are similar to those of hydrocephalus itself: headache, nausea, vomiting, upgaze paralysis, and papilledema. In a prospective study, drowsiness was found to be the best predictive symptom (05). In infants, a full fontanelle and head growth crossing percentiles would be added to the list. Usually, with higher pressure the ventricles get larger, but the quantitative degree varies significantly from patient to patient. However, the ventricles may not get larger at the time of a shunt malfunction in the patient with a poorly compliant brain. This situation is not common, but when it does occur, it is usually in patients with small ventricles, “slit ventricle,” or shunted arachnoid cysts. Up to 15% of pediatric patients have altered brain compliance and may not display enlargement of their ventricles in the presence of shunt failure and increased intracranial pressure (14). In some cases, shunt malfunction is difficult to diagnosis both clinically and by CT scan. Shuntograms may be useful in this regard. There are currently no validated prediction models for shunt malfunction in low-risk individuals based on clinical presentation alone (07).

There has been much discussion about “shunt” allergies. If this occurs, it is rare. The typical presentation of an allergy to silicone shunts includes abdominal pain and skin irritation or breakdown of the skin overlying the shunt. It is typically diagnosed via detection of antisilicone antibodies and can be treated by replacement of the silicone shunt with a polyurethane shunt (19). CSF eosinophilia has also been identified as a useful marker in determining shunt malfunction secondary to a hypersensitivity reaction or bacterial infection.

Loculation. Because the ventricular system can be hydrodynamically divided by septa into several compartments, it occasionally happens that an isolated part of the ventricular system enlarges, despite adequate shunting of a portion that does not communicate with this. This usually occurs in the setting of infection. Such cases of shunt failure may require the placement of more than one catheter or fenestration of the obstructing tissue endoscopically to make the system unified. The latter procedure permits drainage through a single catheter, which is often preferable. Endoscopic third ventriculostomies have proven to be efficacious for the management of childhood hydrocephalus due to congenital or acquired aqueductal stenosis, but are less effective in the management of hydrocephalus secondary to inflammatory or infectious etiologies (06).