shunt revision in hydrocephalus

Indian Journal of Pediatrics, Volume 72—October, 2005 843

Original Article

Correspondence and Reprint requests : Dr. Raj Kumar, AssociateProfessor, Department of Neurosurgery, Sanjay Gandhi PostGraduate Institute of Medical Sciences, Lucknow-226014, UP, India.Fax No.: 91(522) 2668129 and 2668017

Abstract. Objective : A retrospective analysis of 50 hydrocephalic children having a minimum follow-up of 6 months wascarried out to see their etiology, clinical features, complications, incidence of shunt revisions, outcome and the variation fromtheir Western counterparts. Methods : Clinical features, image findings and treatment of all the cases were recorded from theirdischarge summaries. Record of shunt revision complications and outcome was maintained by the principal author. The dataof all the cases were analyzed. Results : The age of children varied from 1 month to 12 yr (mean 2.2 yr). The most commonetiology of hydrocephalus was aqueductal stenosis in 18 (36%) children. Post infective hydrocephalus, either of post-tubercularmeningitis (TBM) or following bacterial meningitis, remained the cause in 15 children (30%). Congenital TORCH infection wasresponsible for 3 cases of hydrocephalus making infective etiology as the cause in 18 (36%) cases. Intra 4th ventricularneurocysticercus cyst caused blockade of CSF pathway in 2 children. 15 out of 50 children required shunt revision, either dueto infection (8,16%) or shunt obstruction (7, 14%). Multiple shunt revisions were required in 2 children only. These revisionswere required due to infection, obstruction or malfunction of the shunt. Conclusions : Infective etiology is responsible forhydrocephalus in significant number of children (36%). The possibility of TORCH infection, as a cause of hydrocephalus shouldbe considered even amongst the children of screened mothers during antenatal check-up. Pure intra 4th ventricularneurocysticercus cysts (without intraparenchymal cyst), though rare, can manifest with outlet obstruction. Incidence of shuntrevision using Chhabra’s medium pressure shunt is very high in children at an average follow up of 1.6 yr. Post infectivehydrocephalus is a major cause of delayed milestones, contributing to mental retardation.[Indian J Pediatr 2005; 72 (10) : 843-847] E-mail: [email protected]

Key words : Hydrocephalus; Etiology; Complication; Shunt revision

Shunt Revision in Hydrocephalus

Raj Kumar, Vinita Singh1 and Marakani V. K. Kumar

Department of Neurosurgery, 1Neuroanaesthesiology, Sanjay Gandhi Postgraduate Institute of Medical Sciences and1King George’s Medical University, Lucknow, India

Hydrocephalus has been the focus of more dedicatedstudy and investigation than perhaps any other conditionafflicting the human nervous system. For many centuriespatients with hydrocephalus had limited expectations forsurvival. With the development and refinements ofmodern cerebrospinal fluid (CSF) shunt, normal learningand intelligence is now possible and patients are able toenjoy full participation in all facets of life. CSF shuntdiversions though have many complications, the mostcommon being shunt obstructions and infection whichrequire the attention of neurosurgeons. The frequency ofshunt revisions in the 1st year and in the 10th yr have beenstudied in detail amongst the hydrocephalic childrenharbouring spinal dyspraphism1, where 30-40% childrenrequire at least one shunt revision during first years of life;of these 15 to 20% require multiple revisions and 85%undergo at least one shunt revision during first 10 yr.1

However, the incidence of shunt revision is not known inother hydrocephalic children. The present study is aretrospective analysis of hydrocephalic children who hadundergone CSF shunt diversion. The etiology, clinicalprofile and complication of shunt procedures are studied

in the Indian children. The frequency of shunt revisionsand their outcome has also been studied.

MATERIALS AND METHODS

Analysis of a cohort of 50 children of hydrocephalusretrospectively, who underwent a ventriculoperitonealshunt insertion during 1997 to 2001 in the NeurosurgeryDepartment at Sanjay Gandhi Post Graduate Institute ofMedical Sciences, India was done. The following detailswere ascertained from the case record files, dischargesummaries and follow-up register maintained by apediatric neurosurgeon : name, age, sex, etiology, clinicalprofile, imaging and treatment. Record of shunt revisions,complications and outcome at follow-up of each case wasmaintained in the computer database and follow-upregister.

Children having a minimum follow-up of 6 monthswere included in this analysis and those with an adequatefollow-up but inadequate records (etiology, procedure,complications and follow-up) were excluded. Childrenwith active signs of meningeal irritation, and tumors asthe cause of hydrocephalus were also excluded. Childrenincluded in the analysis underwent medium pressureChhabra’s shunt, either by the principal author himself orunder his direct supervision, particularly when the

Raj Kumar et al

844 Indian Journal of Pediatrics, Volume 72—October, 2005

procedure was performed electively. Chhabra’s shunt ismade up of implant grade silastic material and has aventricular end, a chamber and a peritoneal end. Thechamber has a spring in slit. The CSF flow is pressureregulated and flow rate type. There are around 40 holes inthe 1.5cm length of the radio-opaque ventricular end. Theperitoneal end is also radio-opaque. It is an economicallysound shunt used widely throughout India as well as inthe neighboring countries.

The clinical features of children in study aresummarized in table 1.

RESULTS

The fifty children included in the study comprised of 42males and 8 females with a ratio of 5.25: 1. The age at thetime of shunt insertion for the first time ranged from 1month to 12 yr (mean 2.2 yr). Follow-up ranged from 6months to 2 yr (mean 1.6 yr). Each child had undergonean initial CT scan. MRI was performed in cases wherethere was a suspicion of a structural lesion causinghydrocephalus. The causes of hydrocephalus amongst thechildren in this study are mentioned in table 2.

The children of TORCH infection were diagnosed bypresence of periventricular calcification (Fig 1). Thediagnosis was confirmed by serum and CSF ELISA.

Both the children of neurocysticercus cyst had historyof tonic clonic seizures along with raised intracranialpressure (ICP). Emergency shunt was inserted in one ofthese as he presented acutely with altered sensorium andfeatures of raised ICP, the cyst was enucleated at a laterdate in this child (Fig 2). An elective shunt was performedin the other child, where the cyst could not be defined atthe first presentation on CT scan. The 4th ventricular cyst

TABLE 1. Clinical Presentation (Signs and Symptoms)

Symptoms Number of patients

1. Headache 172. Vomiting 253. Increase in head size 284. Visual disturbances 75. Seizures 66. Fever 137. Premature with low birth weight 18. Delayed milestones 15

Signs1. Increased head circumference 332. Sunset sign 173. Frontal bossing 84. HMF Apathetic 3

Decorticate Posture with altered sensorium 15. Upward gaze paresis 216. Sensory /motor system examination. Hemiparesis/quadriparesis in 4 with exaggerated DTRs7. MMC/occipital encephalocele 7 (5 + 2)8. Fontanelle Bulging & tense in 339. Fundus examination Papilledema in 33 (Could be examined)

Abbreviations : MMC – Meningomyelocele, DTRs – Deep tendon reflexes, HMF – Higher mental function.

Fig 1. CECT showing hydrocephalus with periventricularcalcification in child with TORCH infection.

TABLE 2. Etiology of Hydrocephalus

Etiology n

1. Aqueductal stenosis 182. Congenital infection (TORCH) 33. Intraventricular hemorrhage 14. Post Meningitis

Post tuberculous meningitic 10Post (non tubercular) bacterial pathology 5

4. Spinal Dysraphism 7Meningomyelocele 5Occipital encephalocele 2

5. Dandy Walker syndrome 46. 4th ventricular outlet obstruction

(intraventricular neurocysticercus cyst) 2



was enucleated when the child had a suspected shuntmalfunction at 9 months, when CT-scan also confirmedthe diagnosis.

A single premature baby of 7 months gestational agehaving a birth weight of 1500gm was diagnosed to havehydrocephalus 1 month following birth due tointraventricular hemorrhage. An elective ventricular taprevealed xanthochromic CSF, suggesting an old bleed,further confirmed by biochemical and cytologicalanalysis. Within a period of 4 months, shunt revision wasrequired twice due to shunt blockade on him. The childremained mentally deprived at 1-year follow-up.

Two of four children with Dandy Walkermalformation had mental retardation at follow-up,though there was no evidence of structural anomalies ofthe cortex.

Complications

Fifteen of the fifty cases required shunt revisions during amean follow-up of 1.6 yr due to one or the other reason.Two of these 15 cases required multiple shunt revisions (1child of meningomyelocele had 5 shunt revisions in aperiod of 8 months and 1 with an intraventricularhemorrhage had two shunt revisions in a period of 4months). Shunt infection was the first and foremost causeof shunt revisions in 8 (16%) children (0.16 per case and0.14 per procedure). Shunt obstruction was the secondcause in 7(14%) children (abdominal end of the shuntgetting blocked in 2 and ventricular end in 5).

Three of 8 shunt infections developed wounddehiscence at 9 weeks in one child and at 2 months in theother 2 children. Scalp wound demonstrated growth ofstaphylococcus epidermidis and the peritoneal woundshowed E. coli and pseudomonas on culture. Two of the 8children developed infection of subdural hygroma alongwith shunt infection. Three of the 8 children presentedwith meningismus and fever for a few days to weeks

duration. Features of raised intracranial pressure andpericatheteral CSF filling due to occult leak fromventricular end were the manifestations in two otherchildren. One of these two had episodic tonic posturing ofthe body for a month along with left hemiparesis. Loss ofbilateral vision and secondary optic atrophy was noted inanother child of aqueductal stenosis who had undergoneventriculoperitoneal shunt 1.8 yr ago. All the 8 childrenwere treated either by exteriorization of shunt or byinstalling omaya reservoir, followed by intermittenttapping. External ventricular drainage was maintained tillCSF cytochemical analysis, microscopy and culturesensitivity proved an absence of infection. Antibioticswere given for 4-6 weeks (even if CSF was sterile) on thebasis of culture and sensitivity of CSF, shunt tubes orwound swabs. Both ends of the removed shunt tube weresubjected to culture and sensitivity in all cases of shuntremoval. In the event of sterile CSF, broad spectrumantibiotic like cefotaxim/amikacin and metronidazolewere given in an injectable form for 2 to 3 weeks, followedby broad-spectrum oral therapy for further 3 – 4 weeks. Insevere infection of CSF, intraventricular gentamicin wasgiven on alternate days. Reinstallation of a new shuntassembly was done only after two consecutive CSFsamples came out to be normal biochemically andcytologically.

Outcome

Thirty-five of the 50 children had no complicationfollowing shunt insertion and improvement was noted inneurological/functional status of the subjects at follow-up. Analysis of the neurological/functional outcomerevealed that 15 of the 50 cases continued to have delayedmilestones at follow-up and they required some sort offunctional assistance in their day-to-day activities. On thebasis of etiology, 3 of these 15 had a congenital infection,5 with post-tubercular meningitis, 2 with nontubercularbacterial infections, 2 with Dandy walker syndrome, 2had an associated spinal dysraphism and 1 had anintraventicular hemorrhage. Results reveal that childrenwith an infectious etiology (post-infective n=15,congenital infection n=3) had a higher incidence ofdependence on their families for their day-to-dayactivities in comparison to other etiologies ofhydrocephalus (10 of 18 infective etiology weredependent, 55.6%).

DISCUSSION

Most infants presenting with progressive hydrocephalusare treated either with ventriculoperitoneal shunt CSFdiversion or 3rd ventriculostomy. Although the risk inperforming a shunt is low, complications related to shuntdiversion are many, hence each child needs to be assessedindividually, weighing the risk of surgery vs no treatment.The outcome of shunt diversion depends on a number offactors which include concomitant medical problems, the

Fig 2. T1 wt saggittal MRI depicting large well defined hypointensecyst in 4th ventricle causing obstructive hydrocephalus.

Raj Kumar et al

846 Indian Journal of Pediatrics, Volume 72—October, 2005

degree of neurologic compromise at the time of surgery,the nature, etiology, duration and severity ofhydrocephalus, status of cortical mantle, structuralanomalies, timing of shunt etc1.

Etiological Factors and Their Relation to the Outcome

Amongst the etiological factors, congenital aqueductalstenosis accounts for 10% of all hydrocephalus cases inchildren.1 Aqueductal stenosis remained the mostcommon cause of hydrocephalus amongst children in thepresent study as well (n = 18, 36%). A higher incidencecan be explained on account of the exclusion criteria in thepresent study, where idiopathic cases, tumors etc as acause of hydrocephalus were excluded. None of thesechildren showed mental retardation. All children ofschool age attended school normally and were notdependent on their families for their day-to-day activities.

Tuberculosis is a serious health problem in thedeveloping countries. Most patients with central nervoussystem tuberculosis have an identifiable systemic disease.Hydrocephalus develops more frequently in thesepatients and almost always manifests itself within 4 - 6weeks of the disease2,3. In the present cohort ofhydrocephalic children, 10 (20%) had post TBMhydrocephalus. All cases responded well toantitubercular therapy with CSF diversion. The olderchildren manifested with mental retardation or delayedmilestones, probably due to extensive involvement ofbrain parenchyma and delayed shunting in them, andthey lagged behind their siblings in school at follow-up.

Hydrocephalus is an uncommon complication ofchildhood bacterial meningitis but occurs more frequentlyin neonates recovering from meningitis. In most cases, thecomplication develops insidiously over weeks to months.Severe neurologic sequelae such as hemiplegia/quadriplegia, hyperactivity and mental retardation arerelatively uncommon in these patients (less than or equalto 4% of all children recovering from bacterialmeningitis), although such sequelae are relatively morefrequent after neonatal meningitis1. Five of the 50 childrenin the present study had nontubercular meningitis as thecause for their hydrocephalus. One of these patients hadhemiplegia and 2 showed mental retardation at follow-upeven after shunting.

Cytomegalovirus, rubella, mumps, varicella, andparainfluenza virus can cross the placenta and have beenimplicated as a cause of congenital hydrocephalus.Congenital infection (TORCH) was the cause ofhydrocephalus in 3 children in the present study (6%). Allthese children showed delayed milestone at follow-up.

Neurocysticercosis is a leading cause of seizures,hydrocephalus and learning disability in children andadults.4,5 In the present study, 2 of 50 children, had 4th

ventricular outlet obstruction due to neurocysticercuscysts. Enucleation of the cyst from the 4th ventricle wasdone in 1 child who had undergone an emergency VPshunt at presentation, while in the other the cyst was

enucleated 9 months later, when he presented with shuntmalfunction. At follow-up both of them were doing wellin school and showed no mental retardation orneurological compromise.

The influence of the etiology of hydrocephalus onoutcome has been studied by various authors withdiffering conclusions, which no doubt reflects theheterogenesity of the underlying conditions.6,7 In thisanalysis etiology did influence the outcome significantly,as the children with hydrocephalus secondary toinfections were more than 3 times likely to requireassistance in their day to day activities (55.6% vs 15.6%),as compared to the other diagnostic categories (as shownin literature8). The present results, however, did notconfirm the finding of a higher incidence of poorintellectual outcome in aqueduct stenosis or non-communicating hydrocephalus.6

Complications of Shunt

Shunt obstruction has been found to be one of theimportant causes of shunt failure requiring revision. Themost common shunt malfunction was found to be at theventricular catheter as reported in literature9. In thepresent study mean follow-up period of 1.6 yr, 5 of 7(71%)shunt obstruction cases had a proximal block ofventricular end. Distal obstruction is seen principally ifthe distal end is placed in the pre-peritoneal space. Low-grade infection with intra-abdominal loculation orpseudocyst formation, disconnection, ingestion andwithdrawal of the catheter from the peritoneum are theother causes of distal malfunction10. Two children in thepresent analysis demonstrated block at the peritonealend, most probably with mental tissue, where the distaltubes started to drain once the peritoneal tips were cut.

Most centers report infection rates in the order of 5 to10%.1 These infections usually present within two monthsof shunt insertion, suggesting that most of these occurduring the surgery itself.11 The infection rate in thepresent study is 16% (n=8), (0.16 per case and 0.14 perprocedure) in the mean follow-up period of 1.6 yr. This isquite disappointing by the present standards12 and reflectthat the children under 6 months (n=21, 42%) aresignificantly at risk for infective complications. Accordingto etiology and clinical presentation the shunt infectionhas been divided into external and internal type. Thoughthe manifestations of both are different, a few of themmay present with atypical manifestations not coincidingwith either of the two. Two children in the present studypresented with atypical manifestations of shunt infectionlike diffuse pericatheteral swelling and globular swellingat the ventricular end13.

Organisms of low virulence generally cause infectionof CSF shunts. Staphylococcus epidermidis is the mostcommonly found organism in most series14. Staphylococcusaureus, Gram-negative bacilli, and mixed infections arethe next most common agents encountered. In the presentanalysis, 6 out of the 8 children showed shunt infection



due to staphylococcus species and the rest 2 had infectiondue to pseudomonas and E. coli species. Infection wasmanaged by the removal of the infected shunt assemblyand placement of omaya reservoir (for regular tapping ofCSF) or external ventricular drainage if need arose. Acourse of 4 to 6 weeks antibiotics was chosen on the basisof culture and sensivity of CSF, shunt tubes or woundswabs, and broad-spectrum antibiotics were administeredif cultures failed to reveal any pathogen. Intraventricularantibiotics were also administered in severe infections. Anew shunt assembly was installed once the CSF becameclear biochemically and microscopically on twoconsecutive testing. If the CSF is flowing through theshunt more quickly, the ability of the brain to expand, thecortical mantle, may remain unexpanded and subduralfluid collections may develop. In the present study, twochildren had subdural hygroma due to overdrainage ofCSF. This incidence might have been higher in the presentseries but we could document only two symptomaticcases mainly because of adventitious infection.

Being a retrospective analysis, the study suffers fromits own set of limitations. A short follow-up is anunacceptable compromise to the ideal situation of aprospective study. The validity of this study could also beimproved by recruiting large number of patients. Butpractically, this was possible only in a multicenter study,as follow-up is a real problem in the developing milieu.

CONCLUSION

The development of effective cerebrospinal fluid (CSF)shunts represents a landmark in neurosurgery. Apartfrom congenital aqueductal stenosis, post-infectivehydrocephalus remains the commonest cause ofhydrocephalus in developing countries. Treatable causelike TORCH infection in the mother may manifest withhydrocephalus in the newborn and this is accountable toinappropriate antenatal screening, immunization andignorance. Intra 4th ventricular neurocysticercus cyst mayrarely contribute to the blockade of CSF pathway withoutthe evidence of any other cyst in parenchyma or

ventricular system.Complications of shunt, like infection and obstruction

are quite disappointing and have a high incidence ofaround 30% to warranting shunt revisions. Delayedmilestones, mental retardation and dependence for day-to-day activities are more marked with the infectiveetiology of hydrocephalus.

REFERENCES

1. Mark S. Dias. Myelomeningocele. In Maurice Choux, ConcezioDI Rocco, AD Hockley, ML Walker, eds. Text book of PediatricNeurosurgery. 1st edn, Churchil Livingstone 1999; 43; 253– 269.

2. Farinha NJ, Razali KA, Holzel H, Morgan G, Novelli VM.Tuberculosis of the central nervous system in children : A 20year survey. J Infect 2000; 41 : 61-68.

3. Faulhauer K, Schmitz P. Overdrainage phenomena in shunttreated hydrocephalus. Acta Neurochir 1978; 45 : 89-101.

4. Lowichik A, Siegel JD. Parasitic infections of the centralnervous system in children : I. J Child Neurol 1995; 10 : 4 – 17.

5. Lowichik A, Siegel JD. Parasitic infections of the centralnervous system in children : II. J Child Neurol 1995; 10 : 77-87.

6. McCullough DC, balazer-Martin LA. Current prognosis inovert neonatal hydrocephalus. J Neurosurg 1982; 57 : 378-383.

7. Renier D, Saint-Rose C, Pierre-Kahn A, Hirsch JF. Prenatalhydrocephalus : outcome and prognosis. Childs Nerv Syst1998; 4 : 213-222.

8. Gasey ATH, Kimmings EJ, Kleinlugtebeld AD, Taylor WAS,Harkness WF, Hayward RD. The long-term outlook forhydrocepahlus in childhood. A ten-year Cohort Study of 155patients. Pediatr Neurosurg 1997; 27 : 63-70.

9. Drake JM, Kestle J, Milner R et al. Randomized trial ofcerebrospinal fluid shunt valve design in pediatrichydrocephalus. Neurosurgery 1998; 43 : 294 - 305.

10. Ginsberg HJ, Sum A, Drake JM et al. Ventriculoperitonealshunt flow dependency on the number of patent holes in aventricular catheter. Pediatr Neurosurg 2000; 33 : 7-11.

11. Pople IK, Quinn MW. Morbidity and outcome of shuntedhydrocephalus. Z Kinderchir 1990; 45 (Suppl I) : 29.

12. Choux M, Genitori L, Lang D, Lena G. Shunt implantation :Reducing the incidence of shunt infection. J Neurosurg 1992; 77: 875-880.

13. Raj Kumar. Uncommon manifestations of shunt infection.Indian J Pediatr 2003; 70(1) : 21-24.

14. Bayston R. Epidemiology, diagnosis treatment and preventionof CSF shunt infection. Nuerosurg Clin Am 2001; 12 (4) : 703-707.

shunt revision in hydrocephalus

Documents