Ostomy closure and the role of repeat endoscopic third ventriculostomy (re-ETV) in failed ETV procedures
Background: Endoscopic third ventriculostomy (ETV) has replaced shunt surgery for several indications. Failure of ETV secondary to restenosis can result in recurrence of symptoms of raised intracranial pressure. Objective: To analyze the rates of restenosis due to ostomy closure and factors resulting in failures and to assess the role of re-ETV in such cases. Materials and Methods: Re-ETV was performed after counselling and obtaining informed consent. The technique of re-ETV was essentially the same as in primary ETV. Video analysis of primary ETV was performed before selecting a patient for re-ETV. Factors analyzed included age, gender, etiology of hydrocephalus, cerebrospinal fluid (CSF) findings, presence of shunt tube and adequacy of ETV and bleeding at the time of ETV. Results: Thirty-two patients underwent re-ETV. The mean interval between the first ETV and re-ETV was 1.4 years (3 days to 2.9 years). Overall failure of ETV due to restenosis was 8.78%. The technical success rate of performing re-ETV was 93.2%. The overall clinical recovery following surgery was observed in 89% of the patients, three from early and 25 from delayed ETV failures. The radiological recovery was seen in 20 (63%) patients. The good flow of CSF via the re-ETV site was documented with cine mode magnetic resonance imaging (MRI) in seven patients. Unlike primary ETV, the success of re-ETV in children aged less than 2 years was 90% (P < 0.005). There were 56.25% failure of ETV in patients with previous infection or foreign body within the ventricle (P < 0.001). While the chances of restenosis were high in the procedure with some infections, the outcome was equally better. Gender of the patients and CSF findings had no influence on ostomy closure. Conclusions: re-ETV can be considered in carefully selected patients of failed ETV. It is more useful in delayed ETV failures and can be offered before a patient is advised VP shunt.
Keywords: Endoscopic third ventriculostomy (ETV), failed ETV, hydrocephalus, ostomy closure, re-ETV
Endoscopic third ventriculostomy (ETV) has made a mark in the treatment of hydrocephalus of varied etiology. ,,,,,,,,, It has become a substitute for ventriculo-peritoneal (VP) shunt and has been considered a treatment option in selected cases of hydrocephalus. ,,,,, It provides an opportunity for patients with hydrocephalus shunt-free life. , Failure of ETV can occur due to several factors, and inadequate ostomy results in early failure. The most significantly cited factor for failure of ETV is age of the patient. ,,,,, The cause for primary failure is the poor absorption of cerebrospinal fluid (CSF) in children aged less than 1-2 years. Delayed failure due to ostomy/stoma closure can happen after the initial success.  The incidence of ostomy or stoma closure and the possible risk factors have not been adequately addressed in the literature.  Moreover, there are no clear guidelines to tackle this issue. We analyzed our data of failure of ETV due to stoma closure and the role of re-ETV in such patients. We also analyzed various factors that can have an impact on closure of ETV.
Patients with hydrocephalus who were previously treated with re-ETV were included in the study. Criteria for failure or closure of ETV were based on clinical and radiological findings. All patients presenting with features of raised intracranial pressure (ICP) after a successful ETV were included for the analysis. Radiological, computed tomography (CT)/magnetic resonance imaging (MRI) features included increase in the size of ventricles with periventricular lucency as compared with previous scans. Phase contrast cine mode MRI was performed to further confirm closed ETV site in 10 patients. In the other patients, the diagnosis of closed ostomy was based on clinical and imaging features.
Case records of patients operated for ETV and the videos of the initial procedure were reexamined. The choice of re-ETV in early failure was only considered if the previous video showed an unsatisfactory ostomy during the first operation. The criteria for unsatisfactory/inadequate initial ostomy was analyzed by the senior author. The anatomical ostomy was considered patent if there was at least 3 F dilatation of floor of 3 rd ventricle with distinctly visible basilar artery or its branches with no second membrane. The to and fro movement of the margin of ETV with clear egress of irrigating fluid from the ostomy further added to confirmation of patency [Figure 1], [Figure 2], [Figure 3] and [Figure 4]. Based on these findings, if the initial anatomical ostomy was found to be satisfactory, the case was excluded for early re-ETV.
After obtaining informed consent, re-ETV was performed similar to the primary procedure.  Floor of third ventricle was perforated with blunt cautry tip in off position and was further dilated with 3 F fogarty catheter. The margins of stoma were coagulated. Success of re-ETV was defined if the patient had relief of clinical symptoms and not requiring a VP shunt. Factors analyzed included age, gender, CSF findings, size and adequacy of initial ostomy, indication of previous ETV, presence of shunt tube in situ and bleeding during the first surgery.
Statistical analysis was performed using SPSS version 11. Univariate analysis was conducted using Cox regression with age as continuous variable. The variables that were significant at the probability level of P <0.05 were then analyzed for multivariate Cox proportional hazard analysis.
During the years 2002-2009, the ETV procedure was performed in 296 (169 males and 127 females, age range: 7 days-64 years) patients. The follow-up period ranged between 1 and 7 years (mean 4.6 years). Thirty-four patients were analyzed in the re-ETV study, but two were excluded from the analysis as, technically, re-ETV could not be performed. The remaining 32 (14 children and 18 adults) patients underwent re-ETV. The time interval between the first ETV and re-ETV was 1.4 years (3 days-2.9 years). Among the children, 10 were less than 2 years of age. The procedure was clinically successful in 11 (78.5%) children and in 15 (83.3%) adults. The success of re-ETV in children less than 2 years of age was 90% (P < 0.005). The overall clinical recovery following surgery was observed in 89% of the patients (three early ETV failures and 25 delayed ETV failures). Radiological recovery of CSF flow was seen in 20 (63%) patients. Good CSF flow via the re-ETV site was documented with cine mode MRI in seven patients. Two patients had three ETV procedures and in both the patients hydrocephalus was secondary to tuberculosis meningitis.
Various causes of ETV failure are presented in [Table 1]. Impaired CSF absorption or delayed cerebral adoption was the major cause of ETV failure in early failures, and stoma closure was more often seen in delayed ETV failures. In the early part of the study, there was more selection errors probably related to learning curve. Technical failures included cases where ostomy was inadequate or second membrane was not looked for. Such errors were also greater during the early part of the study.
Failure of ETV was divided into two types: early within 7 days and late after 7 days.  In all, there were 83 (31.5%) patients with ETV failures. Early failure was observed in 51 (17%) patients. Of these, only six patients underwent re-ETV. The remaining patients were treated with VP shunt. Delayed failure was observed in 32 (11%) patients, and 26 of them had re-ETV. The remaining six patients were treated with VP shunt. Four patients refused consent for re-ETV and the other two had technical failure during re-ETV. In all, re-ETV was performed in 32 patients (six early and 26 late). True ostomy closure was observed in 26 (8.78%) patients. In the remaining three patients, it was related to technical failure as the second membrane remained unperforated during initial ETV.
Minor bleeding was observed in three patients at the initial ETV, which settled with irrigation. There were only two patients with moderate bleed. The bleeding had settled at the end of ETV but the floor of ostomy had residual blood. Both these patients presented with delayed failure but relatively early, 4 weeks and 6 weeks. Technically, it was possible to perform ETV in all the patients except in two patients in the delayed group, one each due to lost landmark and very thick floor. The stoma was closed in all the patients except in three patients in early failures [Figure 2]. In these patients, there was evidence of imperforated second membrane and they presented with early failures. The flimsy adhesions were noticed between basilar artery complex and floor of third ventricle in six (20.6%) patients. The technical success rate of re-ETV was 93.2%.
Analysis of various factors
Early vs. delayed failure
There were 51 patients with early ETV failures. Six patients were subjected to early re-ETV. In three patients, the cause of first ETV failure was presence of second membrane. The remaining had inadequate size of first ostomy. The success rate of re-ETV in these patients was 50%, two with second membrane and one with inadequate ostomy during first operation. Of the 32 patients with delayed failures, 26 patients had re-ETV, four patients refused consent and the other two had technical difficulties. The success rate of re-ETV in this group was 78% (P < 0.005). Inadequate ostomy was observed in the video review of six patients. All patients with delayed failure had stoma closure due to membranous tissue of variable thickness. Scarring was seen in four patients, and three patients demonstrated thin membranous arachnoid-like obstruction. The remaining had obstruction due to thick ependyma-like tissue similar to floor of third ventricle, whereas early failures had no or thin stenosis, but their number was too small for any statistical correlation.
There were 14 children and 18 adults. Among the children, 10 were less than 2 years of age. Clinical success was 78.5% in children and 83.3% in adults (P < 0.03). The success of re-ETV in children less than 2 years of age was 90% (P < 0.005).
There were 18 male and 14 females. Success of re-ETV in both the genders was nearly same (P < 0.04).
Initial etiology of hydrocephalus
The thickness of stoma closure was higher in patients where ETV was performed for previous infection, e.g. meningitis (14/32), or presence of concomitant VP shunt (4/32). There were 56.25% failures of ETV in patients with previous infection or foreign body within the ventricle (P < 0.001). There were 15 aqueduct stenosis, two brain stem glioma and one outlet obstructions. While the chance of restenosis is high in procedure with some infections, the outcome was equally better.
The CSF was considered normal if number of cells were less than 5/mm 3 with no polymorphs and normal biochemistry. There were seven patients with abnormal CSF (five with raised cell count and two with raised proteins). There was however no statistical correlation with restenosis and success of re-ETV (P < 0.04).
Presence of VP shunts in situ
In four patients, initial ETV was performed for failed VP shunt. Three had tuberculous meningitis and one had aqueductal stenosis. The shunt tube in these patients had not been removed due to adhesions. Although the number of cases is extremely few, it may be probable that presence of foreign body inside the ventricle may enhance the process of restenosis.
Size and adequacy of initial ETV
Adequacy of ETV was decided by the senior author as the per criteria described previously. Failure of ETV was 28% (P < 0.05) if the second membrane was not visible during first surgery or if there were residual floating margins of the floor of third ventricle covering the ostomy, or in close proximity, 18.8% (P < 0.01).
Bleeding during procedure
Two patients had moderate bleeding during ETV and presented with closure within 6 weeks with failed ETV.
Some neurosurgeons advocate ETV as the first treatment option for hydrocephalus failure that can be treated with VP shunt. ,,, Others believe that re-ETV should be tried in failed ETV before considering placement of shunt.  The role of ETV in infection has also emerged with significant success. ,,,,,, Additionally, the popularity of ETV has become so much that patients opt for ETV over VP shunt.
Several reports have documented ETV failures, and the reported rates varied between 0 and 100% depending on various inclusion and exclusion criteria. ,,,,,,,,, Like primary failure, the proposed risk factors for restenosis remains the same. Bleeding during procedure, adequacy of ETV and post-ETV infection can contribute to closure of stoma. Since its earliest descriptions, ,, stomal obstruction has been recognized as one of the possible mechanisms for ETV failure. ,,,,,,,,,,,,,,,,,, The possibility of reopening the stoma by repeating the procedure has been proposed as an alternative to shunt placement.  Cinalli et al. reported 66% success rate in six patients with repeat ETV. A large multicenter, retrospective series by Siomin et al. reported a success rate of 65% with repeat ETV. Age of the patient remains as one of the most debated factor for success of ETV. ,,,,,, While most reports consider younger age as a relative factor for poor outcome, ,, others differ with this view. , The consequence of these findings for decision-making as well as parental counselling is to try to weigh the age of the infant and its estimated impact on ETV success with other factors guiding the decision to perform ETV or shunt in the treatment of obstructive hydrocephalus.  The reported ETV success rates in the younger age group ranged between 20 and 57%, whereas metaanalysis of this study reported an overall success rate of 42.4%. ,,,,,,,, One study reported a five-fold increase in the risk of ETV failure in patients aged less than 6 months.  Gallo et al. felt that age cannot be considered the only parameter of the decision-making process, especially in these very young patients. Ogiwara et al. felt that ETV could be the first method of choice for hydrocephalus in children.  In some of the recent studies, age, etiology of hydrocephalus and other contributing factors may have not been shown to correlate with ETV success. , Age of the patient in our study had no influence on the outcome of re-ETV both in technical as well as clinical success. However, unlike primary ETV, age had no negative influence in the outcome of re-ETV. The explanation for this could be that once primary ETV is successful, there is likely to be greater response of re-ETV as the absorption of CSF is already normal in such patients. The failure of ETV due to restenosis is a correctable remedy and, hence, age will not have a major effect on the success of re-ETV. More data are however needed to validate this observation.
The outcome of ETV can be predicted. Kulkarni et al. recently developed and internally validated the ETV Success Score (ETVSS), a simplified means of predicting the 6-month success rate of ETV for a child with hydrocephalus. In a report on second ETV, Peretta et al,  observed that a second ETV can be performed successfully in 70% with a reasonable chance of restoring patency of the stoma and avoiding placement of an extrathecal shunt. In order to improve better functioning of ETV, some neurosurgeons place external drain in all cases. It is believed that it compensates for the altered CSF dynamics in the initial stages of ETV. They recommend that a cycle of one to three lumbar punctures should always be performed in patients who remain symptomatic and who show increasing ventricular dilatation after ETV; before ETV is assumed to have failed and an extracranial CSF shunt is implanted. ,, Other use choroid plexus fulgration in addition to ETV in infants with spina bifida.  As a practise, we do not routinely advocate drainage of CSF by either lumbar or ventricular catheter. We fear that flow of CSF across the stoma, which is a function of pressure gradient from third ventricle to pre-pontine cistern, can decrease with external drain resulting in malfunctioning of ETV and eventual restenosis.
For the proper functioning of ETV, the margins of the stoma need to be wide apart. In case the ostomy margins overhang in proximity of each other [Figure 4], it can be a potential area for membrane to grow over it, leading to restenosis. In fact, the margins may act as a bridge for regrowth of arachnoid or ependymal tissue. We therefore practise the coagulation of margins of ostomy that shrinks the tissue and pulls it apart, resulting in widening of the stoma. It further helps in better visualisation of distal area in the pre-pontine region and helps in identifying and fenestration of second membrane or adhesions or arachnoids tissue. Similarly, the presence of blood in the area, which can happen in some cases of intraoperative bleeding, can act as a cementing substrate for restenosis. It is therefore recommended that the settled blood at the ETV site should be washed or suctioned to prevent restenosis.
What causes the failure of ETV is poorly understood. It can be due to restenosis as observed in some studies or can be due to poor absorption. Early failures are due to poor absorption, usually seen in children. In these cases, the absorption of CSF mechanism is defective. Late failures after an initial successful ETV are invariably due to restenosis. Wagner et al. reported new arachnoid membranes in the basal cistern blocking CSF flow after ETV despite patent ventriculostoma. Improper CSF absorption in young infants is also postulated as a possible cause of ETV failure by several authors. ,,, There are consistent reports that age and sex of the patient has no influence in the success of ETV. ,,, Similar observations have been reported irrespective of etiology of the hydrocephalus. Both univariate and multivariate analysis confirm that there is no gender bias in terms of technical feasibility of performing ETV or re-ETV. It was initially believed that ETV is a relative contraindication for infective hydrocephalus. This is primarily based on the belief that communicating hydrocephalus would not respond to ETV. Recently, there has been a better understanding of the concept and pathophysiology of hydrocephalus. Grietz  demonstrated that CSF dynamic and flow is a function of cardiac cycle, and flow of CSF varies in systole and diastole. There have been several papers from Asian countries with significant success of ETV in tubercular meningitis. ,,, With the improvement of surgical skills, it is now possible to open the thick floor of the third ventricle. Expectantly, the rate of restenosis is a bit higher in infection. Presence of foreign body (e.g., shunt) can further accelerate and augment restenosis. It was further observed in our study that the degree of difficulty in operation of re-ETV in infection was no more than in initial ETV. What triggers the formation of arachnoidal membrane or ependymal growth resulting in closure of stoma remains unclear. It is further unclear whether the patient of infective hydrocephalus would present early with closure of stoma. Two patients in the delayed group presented relatively early (7 weeks and 8 weeks). Both of them had intermediate phase of tuberculous meningitis with hydrocephalus. Singh et al. observed that exudates were more pronounced in the intermediate phase, which makes the possibility of more failure in such cases. Regardless of the original cause of hydrocephalus, ETV has also been considered as an option for shunt-free life to the patients presenting with blocked shunts including the patients with meningitis or shunt infection. , The complication rate of ETV in blocked shunt or in meningitis is not significantly higher than in the primary procedure. ,, There can be a possibility of shunt-induced foreign body reaction resulting in restenosis.  In four patients, initial ETV was performed for failed VP shunt. The shunt tube in these cases had not been removed due to adhesions. Although the number of cases is extremely less, it may be probable that presence of foreign body inside the ventricle may enhance the process of restenosis.
In conclusion, second or re-ETV is a suitable option for the cases of stoma closure. It can be performed in all cases of hydrocephalus irrespective of the etiology. The stoma can get closed due to scarring, arachnoid or ependyma-like tissue. Restenosis is higher in infective hydrocephalus and in the presence of foreign body, e.g. shunt tube. Age had no major effect in the outcome of re-ETV. It is advisable that re-ETV should be considered for failed ETV before advising VP shunt in such cases.
Technical failures mean inadequate ETV or presence of second membrane. (*) patients who have given consent for re TV. Selection error: infants, Chiari malformation More Details, neural tube defects with hydrocephalus, dysplastic hydrocephalus, normal pressure hydrocephalus (see text for details). EVD, external ventricular drain; VP, ventriculo-peritoneal.
[Figure 1], [Figure 2], [Figure 3], [Figure 4]