Neuroendoscopy for post-infective hydrocephalus in children
The treatment of hydrocephalus has changed in recent years with better imaging and introduction of endoscopic procedures as well as enhanced shunts. Indications of endoscopic third ventriculostomy (ETV) are now more refined with better quantification of outcome. This article reviews the current state of neuroendoscopy for infective hydrocephalus in children. The roles of third ventriculostomy as a primary procedure or after shunt malfunction, endoscopic interventions in multiloculated hydrocephalus and introduction of intraventricular lavage to salvage severely infected children are evaluated.
KeywordsHydrocephalus Paediatric CNS infection Meningitis Endoscopic third ventriculostomy (ETV) Choroid plexus coagulation (CPC) Ventricular lavage
Endoscopic third ventriculostomy (ETV) has been controversial for post-infective hydrocephalus (PIHC). Post-infectious aetiology is a low score criterion in ETV Success Score (ESS) . This seems logical due to the scarring of subarachnoid spaces and arachnoid villi with subsequent defective absorption leading to communicating hydrocephalus in most of these cases. Shunts have, therefore, been traditionally used as a primary surgical treatment for this variety of hydrocephalus. The favourable Ugandan experience  with endoscopy, however, has prompted many workers to revisit this indication, and we now have a better understanding of ETV as a primary procedure or after shunt failure. The addition of choroid plexus coagulation (CPC) has improved the outcome of ETV in some reports .
An unquestionable additional benefit of endoscopy is to open up CSF pathways in complex, often multiloculated, post-infective hydrocephalus with several procedures viz. septostomy, foraminoplasty, aqueductoplasty and fenestration of various membranes simplifying the hydrocephalus and, thereby, reducing the number of shunt implantations.
Endoscopic ventricular lavage (EVL) is being increasingly performed in severe cases of post-meningitic ventriculitis in some centres, with variable success in avoiding hydrocephalus, shunt implantation and delaying shunt blockage and is currently under investigation.
Pathophysiology of infective hydrocephalus
The effects of a CNS infection on CSF dynamics depend on the period of the primary infection (prenatal, neonatal or post-neonatal) as well as on the infectious agent (bacterial, viral, parasitic) . Leptomeningeal inflammation [7, 24], ependymal erosive necrosis with resultant ventriculomegaly and blockage of aqueduct by intraventricular inflammatory debris are the various causes for hydrocephalus attributed to prenatal toxoplasmosis infection . Cytomegalovirus (CMV) is the other prenatal infection leading to hydrocephalus, though the incidence is much less (10–15%)  and is mostly due to encephalomalacia causing volume loss. Leptomeningeal inflammation with thickening also contributes, and it is very rarely due to blockage of the aqueduct .
Neonatal infections are primarily bacterial and predominantly caused by coagulase negative staphylococci followed by other gram-positive and gram-negative organisms and result in communicating tetraventricular hydrocephalus frequently (> 50%) in early course of the disease [7, 12]. This may, however, subside with resolving infection, and if this persists, this may result in obstructive hydrocephalus in a delayed fashion due to aqueductal or 4th ventricular outlet blockage . Hydrocephalus is commonly associated with partially treated meningitis, occurring more commonly in neonates and infants, especially with group B streptococcus type III . In gram-negative infections, the ventriculitis and ependymal inflammation is much more common, frequently resulting in multiloculated hydrocephalus. Haemophilus influenzae is the causative agent of post-neonatal infections in most (> 60%) of cases , although there has been a decline with progressive vaccination. Hydrocephalus does not commonly follow (approx.10%), as basal meningitis is less common and choroid plexitis results in less CSF production . Gram-negative ventriculitis can progress to obstructive or communicating hydrocephalus in many cases. Tubercular meningitis (TBM) is a very common cause in children in Africa, India and other south Asian countries. Though predominance of basal meningitis commonly results in communicating hydrocephalus, recent reports suggest a large number of them may be obstructive at aqueduct or 4th ventricular outlet level or rarely because of mass effect of a strategically placed tuberculoma [13, 29, 41, 51].
Development of complex, multiloculated hydrocephalus
The pathophysiology of multiloculated hydrocephalus involves denuding of the ependymal cells with protrusion of underlying glial cells into the ventricular lumen. These cells then act as a nidus upon which fibro-glial septations can develop, leading to isolated compartments of CSF . It is a progressive process, with new septae usually continuing to develop, transforming the normal anatomical landscape of the ventricles from a single large system, into multiple non-communicating compartments . Further enlargement of compartments depends on inflammatory exudates with entrapment of CSF. Some loculations may not enlarge if the membranes are still permeable to CSF flow or if the initial insult has involved the choroid plexus, thus impairing CSF production . The chemical ventriculitis which follows intraventricular haemorrhage or post-infectious ventriculitis usually occurs in the neonatal period, and these early insults usually lead to the formation of septations within the ventricular system [1, 38, 43]. Other iatrogenic causes also play an important role, such as shunt infection, direct trauma to the ependymal cells during catheter insertion, over-drainage or even simply decompression of existing loculations which brings chronically inflamed ventricle walls together, allowing adherence and formation of new septae [17, 32, 43]. Multiloculated hydrocephalus may also occur in association with prematurity-related IVH, birth trauma, CNS tumours, head injury and other intracranial surgery [1, 28, 43, 56].
Role of imaging
Warf’s findings  included a small IV ventricle in 44% cases on pre-operative ultrasound imaging (which has correlated 100% with an obstructed aqueduct in patients undergoing ventriculoscopy) in patients with post-infective hydrocephalus (PIHC). Many patients had intraventricular septations on ultrasound, indicative of prior ventriculitis.
A simple technique of prediction of success of ETV was described by Figaji et al., where entry of air in the ventricular system by lumbar puncture route has been considered a negative predictor due to communicating nature of hydrocephalus in tuberculous meningitis patients . Air studies in non-communicating hydrocephalus show air typically trapped in the prepontine cistern; release of this air during an ETV can be helpful to identify entry into the cisternal space as these cases typically have cisterns filled with exudate and obscured anatomy.
It is easily available, inexpensive and quick, avoiding the need for sedation or anaesthesia and can be performed even in unstable patients with altered sensorium and is therefore the most common investigation in children. The presence of ventricular dilatation and periventricular ooze will suggest active hydrocephalus with meningeal enhancement on contrast administration, and basal exudates can be well visualised in tubercular meningitis (TBM). It may also demonstrate associated ischemia or infarction, cerebritis with parenchymal enhancement and tuberculoma, if any. However, CT scan cannot always differentiate between communicating and non-communicating hydrocephalus .
Dynamic invasive studies like CT scan with air encephalography or CT ventriculogram further aid in the determination of the level of obstruction and differentiation between communicating and non-communicating hydrocephalus. Since it is an invasive procedure, this technique is not widely used, especially due to the wider availability of MR imaging.
Multiloculated hydrocephalus may have a higher prevalence than appreciated and often overlaps with manifestations of shunt malfunction . It is often detected after a shunt revision demonstrating collapse of the drained ventricular compartment with persisting dilatation of the isolated compartments . Contrast CT ventriculography has also been used to assess communication between the various compartments pre- and post-operatively [2, 28].
MRI brain, MR angiogram and Cine MRI
Contrast-enhanced MR imaging demonstrates abnormal meningeal enhancement usually in the basal cisterns and also in other cisterns in severe cases. It also reveals information about hydrocephalus and tuberculomas. Vasculitis in the region of brainstem and basal ganglia is depicted better because of distinction between oedema and infarcts. It can also differentiate between communicating and obstructive hydrocephalus. Chugh et al. strongly advocated use of Cine MRI as a non-invasive tool of assessment and for comparison post-operatively in TBMH (TBM hydrocephalus) .
MRI is also considered as the preferred imaging modality for diagnosis of multiloculated hydrocephalus. It provides multiplanar views with higher spatial resolution and better sensitivity for revealing septations. Specific sequences like constant interference steady state (CISS), which is a T2-weighted gradient echo sequence, provide remarkable morphological detail. MR ventriculography (MRV) involves injection of contrast material directly into a cyst cavity to verify communication with the surrounding ventricular system also defining the margins of these compartments. The procedure allows more appropriate treatment planning but is invasive and requires multiple punctures for different compartments to be accessed and rotation of the patient’s head in all directions to spread the contrast material .
Endoscopic third ventriculostomy
It is now a well-established treatment modality for obstructive hydrocephalus with a success rate of 60–85% in most series [14, 23]. Figaji et al., Jonathan et al. and Husain et al. reported the first use of ETV in TBMH [14, 20, 23]. Raouf et al. reported successful outcome of ETV in 55.9% in 35 cases of hydrocephalus secondary to intracranial infection .
ETV causes diversion of CSF to previously inaccessible areas and clears exudates from the areas which had impaired absorption [20, 23]. It also decreases the transventricular pressure gradient and the demyelination of periventricular brain parenchyma, which could contribute to some symptoms of hydrocephalus [18, 35]. The improved CSF dynamics facilitate better penetration of antiTB drugs. Alteration in haemodynamics may also allow better drug delivery [20, 35].
Chugh et al. concluded that ETV should be considered as the first surgical option for CSF diversion in patients with TBMH. Cine MRI as a non-invasive tool of assessment and post-operative comparison was strongly recommended by them. They found that patients with chronic disease and longer duration of ATT administration responded well . On the contrary, patients with higher stage of illness and cisternal exudates as observed intraoperatively had a poorer outcome. Singh et al. reported a 77% success rate of ETV—60% early and 17% delayed. They reported the presence of a thin and transparent third ventricular floor to be a favourable prognostic indicator .
Choroid plexus coagulation to supplement ETV
Choroid plexus coagulation has been thought to be beneficial as it diminishes the CSF production and can thus increase the success rate of ETV. Warf et al.  conducted a prospective study in Uganda to determine if the outcome of bilateral CPC + ETV was superior to ETV alone. ETV was accomplished in 550 children of which 266 underwent combined ETV + CPC and 284 underwent ETV alone. Fifty eight percent (320 cases) of the children studied had post-infectious hydrocephalus. Overall, the success rate of ETV + CPC (66%) was superior to that of ETV alone (47%) among infants younger than 1 year of age (p < 0.0001). Although the difference was not significant for PIHC (62% compared with 52% success, p = 0.1607), a benefit was not ruled out (power = 0.3). For patients at least 1 year of age, there was no difference between the two procedures (80% success for each, p = 1.0000). The overall surgical mortality rate was 1.3%, and the infection rate was less than 1%.
Late follow-up analysis of the Ugandan study patients showed that there was no significant difference between the ETV + CPC group and the ventriculoperitoneal-shunt group in BSID-3 motor or language scores, rates of treatment failure (35 and 24%, respectively; hazard ratio, 0.7; 95% CI, 0.3 to 1.5; p = 0.24), or brain volume (z score, − 2.4 and − 2.1, respectively; estimated difference, 0.3; 95% CI, − 0.3 to 1.0; p = 0.12). This single-centre study involving Ugandan infants with post-infectious hydrocephalus showed no significant difference between endoscopic ETV + CPC and ventriculoperitoneal shunting with regard to cognitive outcomes at 12 months . The procedural success rates were not significantly different but failure occurred in 35% of the ETV + CPC group and 24% in the shunt group; there were two deaths, both randomised to the ETV + CPC group. Some improvement with the addition of CPC to ETV has been noted in infants with post-infective hydrocephalus by others as well . Recently, in a North American multicentre prospective study of infant hydrocephalus in which patients treated with ETV + CPC were matched with patients treated with ETV alone or ventriculoperitoneal shunting, the success rate for ETV + CPC was 36%, was significantly lower than for shunted patients and was not significantly different than for ETV alone, although the authors note that the study was not powered to show this .
ETV after shunt infection (secondary ETV)
The role of ETV as a primary treatment for hydrocephalus and as an alternative to shunt revision (secondary ETV) in infected and malfunctioning shunts was analysed by O’Brien DF and group . Seventy-four percent of cases in the primary ETV group and 70% in the secondary group had a successful outcome. The success rate of secondary ETV in mechanical shunt malfunction alone was 67% as compared to 79% in infected malfunctioning shunts. The aetiology of hydrocephalus was evaluated as a factor contributing to the success of ETV. In the primary group, infective aetiology led to a poor outcome. However, in the secondary group, 75% cases with infection had a successful outcome. They concluded that the success of secondary ETV, unlike primary ETV, is not origin-specific.
Thirty patients treated with third ventriculostomy for malfunctioning or infected shunts were retrospectively reviewed by Cinalli G et al. 76.7% experienced successful outcomes, resulting in shunt independence. They concluded that ETV is a valuable alternative to shunt revision in patients affected by obstructive hydrocephalus presenting with shunt malfunction or infection .
Shimizu et al. suggested that secondary ETV can be used in shunt infections. Although the procedure does not obviate the need for later shunt implantation, it is shown to delay the subsequent procedures . When performing a secondary procedure, some studies recommend removal or blockage of the malfunctioning shunt. This is shown to increase the longevity of the procedure .
In our experience, 36 patients underwent secondary ETV from 2004 to 2017, of whom, 15 had an infective aetiology as the cause of shunt malfunction. Our success rate for secondary ETV in these cases was 70%. The most important criterion to guide us to ETV was the radiological anatomy showing the triventricular enlargement as described by Cinalli et al. in 2012 .
Endoscopy in Neurocysticercosis
Endoscopic strategies in multiloculated hydrocephalus
Neuroendoscopic experience and good planning are critical. These are often difficult cases—the ventricular anatomy is usually severely distorted and anatomical landmarks are often unclear . Intraoperative decisions about where to create fenestrations are often more difficult in reality than the surgeon expects. Operative morbidity is increased in inexperienced hands by becoming disoriented in the ventricular system. A thorough pre-operative plan is essential, with the goals of the surgery clearly considered. Oftentimes, non-standard entry routes are chosen to maximise access to various compartments. This increases the chance of disorientation. Neuronavigation is invaluable as a pre-operative planning tool. This does however become less accurate with intraoperative brain shift due to CSF movement and real-time intraoperative image guidance with ultrasound or intraoperative MRI have been proven useful [33, 38, 46]. Flexible endoscopy may be valuable in navigating difficult trajectories, but visibility is often not as good, and it is even easier to become disoriented when not operating in a straight line. Intraventricular membranes are often tough to fenestrate and this may require the assistance of sharp dissection, cautery or laser. Wide fenestrations are necessary as the proinflammatory environment often persists for a prolonged time and closure of fenestrations is common. Where possible, it is worthwhile leaving the ventricular catheter traversing the fenestrations, with proximal perforations made on the catheter length where necessary.
The extent of neurological morbidity and developmental delay is determined by the underlying aetiology in multiloculated hydrocephalus, and the goal of surgery is therefore to limit the number of procedures and ideally facilitate insertion of a single ventricular catheter or even avoid catheter placement altogether, where possible [32, 37]. If a shunt is to be placed, it must be remembered that long procedures increase the risk of a shunt infection; therefore, unnecessary fenestrations of questionable benefit may not be in the patient’s best interest if these prolong the procedure unduly.
Intraventricular endoscopic lavage
The treatment of infective hydrocephalus in children, especially infants, remains a great challenge due to difficulty in eradicating the infective materials and inflammatory debris within the CSF pathway in a short period of time. External ventricular drainage (EVD) is frequently not effective and is associated with risks of secondary infection. The subsequent ventriculoperitoneal shunt is often complicated by recurrent blockage, repeated shunt infection and shorter shunt survival due to high level of protein and cell counts. An aggressive approach is therefore needed to facilitate the clearance and restoration of CSF flow [22, 39, 47].
ETV and septum pellucidotomy during lavage
In the case of aqueductal stenosis, endoscopic third ventriculostomy is best avoided at the same setting of lavage as this may contaminate the prepontine space. However, in the presence of the foramen of Monroe stenosis or thick adherent slough which could not be removed safely, endoscopic septum pellucidotomy is performed for access into the contralateral ventricle and creating pathway to communicate both lateral ventricles.
External ventricular drain is inserted on one side of the ventricle if CSF is clear at the end of procedure for post-operative control and subsequently challenge prior to removal. A plain control CT scan is done on day 1 after surgery to exclude complication of lavage and collapsed of ventricles.
ETV has been able to avoid shunts in selected cases of post-infective hydrocephalus. It has also proved to be quite successful in selected cases of shunt malfunction with infection or previous PIHC. However, the selection of cases is important, and an experienced surgeon is needed to achieve a high degree of success without complications. Other endoscopic interventions have contributed to clear up severe infection in ventricles (iVEL) and have created better chances of shunt survival with simplified insertion of a shunt by endoscopic fenestrations in multiloculated hydrocephalus. The latter has substantially reduced the prevalence of multiple shunt systems in patients. Though its primary use in removal of infective cysts (NCC) is unequivocal, its usefulness in CPC and iVEL is under evaluation. With all of these procedures, it is essential that the surgeon has sufficient neuroendoscopy experience as the anatomy is often difficult, visibility is often poor and stomas and fenestrations may be difficult to create.
Endoscopic third ventriculostomy has been primarily designed and written by Deopujari C and Samantray SK. The material for multiloculated hydrocephalus has been provided by Padayachy L, and the material for Intraventricular Lavage has been provided by Azmi A. Dr. Anthony Figaji has contributed to the manuscript and reviewed the text.
Compliance with ethical standards
Conflict of interest
On behalf of all authors, the corresponding author states that there is no conflict of interest.
- 2.Albanese V, Tomasello F, Sampaolo S, Picozzi P (1982) Neuroradiological findings in multiloculated hydrocephalus. Actaneurochirurgica 60(3–4):297–311Google Scholar
- 5.Chugh A, Husain M, Gupta RK, Ojha BK, Chandra A, Rastogi M (2009) Surgical outcome of tuberculous meningitis hydrocephalus treated by endoscopic third ventriculostomy: prognostic factors and postoperative neuroimaging for functional assessment of ventriculostomy. J Neurosurg Pediatr 3:371–377CrossRefPubMedGoogle Scholar
- 7.Ciurea AV, Coman TC, Mircea D (2004) Postinfectious hydrocephalus in children. In: Cinalli G, Meixner W, Sainte-Rose C (eds) Pediatric Hydrocephalus. Springer-Verlag Italia, Milano, pp 201–218Google Scholar
- 11.Di Rocco C, Cinalli G, Massimi L, Spennato P, Cianciulli E, Tamburrini G (2006) Endoscopic third ventriculostomy in the treatment of hydrocephalus in pediatric patients. Adv Tech Stand Neurosurg 31:171–173Google Scholar
- 13.Figaji AA, Fieggen AG (2013) Endoscopic challenges and applications in tuberculous meningitis. World Neurosurg 79(2 Suppl):S24.e9–S24.14Google Scholar
- 24.Kaiser G (1985) Hydrocephalus following toxoplasmosis. Z Kinderkir 40(Suppl. 1):10–11Google Scholar
- 27.Kulkarni AV, Riva-Cambrin J, Rozzelle CJ, Naftel RP, Alvey JS, Reeder RW, Holubkov R, Browd SR, Cochrane DD, Limbrick DD Jr, Simon TD, Tamber M, Wellons JC III, Whitehead WE, Kestle JRW (2018) Endoscopic third ventriculostomy and choroid plexus cauterization in infant hydrocephalus: a prospective study by the Hydrocephalus Clinical Research Network. J Neurosurg Pediatr 21(3):214–223CrossRefPubMedGoogle Scholar
- 33.Paraskevopoulos D, Biyani N, Constantini S, Beni-Adani L (2011) Combined intraoperative magnetic resonance imaging and navigated neuroendoscopy in children with multicompartmental hydrocephalus and complex cysts: a feasibility study. J Neurosurg: Peds 8(3):279–288Google Scholar
- 42.Siomin V, Cinalli G, Grotenhuis A, Golash A, Oi S, Kothbauer K, Weiner H, Roth J, Beni-Adani L, Pierre-Kahn A, Takahashi Y, Mallucci C, Abbott R, Wisoff J, Constantini S (2002) Endoscopic third ventriculostomy for patients with cerebrospinal fluid infections and/or hemorrhage. J Neurosurg 97:519–524CrossRefPubMedGoogle Scholar
- 44.Spennato P, Ruggiero C, Aliberti F, Nastro A, Mirone G, Cinalli G (2013) Third ventriculostomy in shunt malfunction. World Neurosurg 79(2 Suppl):S22.e21–S22.e26Google Scholar
- 51.Warf BC, MacGowen L (2001) Post infectious hydrocephalus: the most common cause of hydrocephalus in Uganda. Ifhsb (International federation for hydrocephalus and spina bifida)Google Scholar