Clinical Success Of Endoscopic Third Ventriculostomy For Pediatric Hydrocephalus

Background: Hydrocephalus is a common neurological disease in children.Ventriculoperitoneal shunt (VPS) was once considered as a routineapproach for hydrocephalus, but its complication is quite prevalent,especially in long-term follow-up. With technology advanced, endoscopicthird ventriculostomy (ETV) has gradually replaced ventriculoperitonealshunt and become the first choice for pediatric hydrocephalus. However,there still are some controversial issues. Endoscopic third ventriculostomyoverweighs ventriculoperitoneal shunt in treatment for pediatrichydrocephalus. Whether will the outcome of ETV change in different age oretiology groups? Which clinical or radiological factors will affect the ETVsuccess?Aims: To conduct a meta analysis for these two surgical approaches bycollecting and comparing the published trials that reported the outcome ofETV or VPS for pediatric hydrocephalus. A retrospective study wasconducted to analyze the data from our medical centre so as to determinewhether age or etiology will has an effect on the success of ETV and VPS.To conduct a regression analysis for ETV success to identify theindependent survival factors for ETV success.Methods: By searching related references, extracting the publishedoutcome data in every included study, a meta analysis was conducted inRevman5.2. A retrospective study was conducted on clinical andradiological data of pediatric patients in our center from2004to2013. AKaplan-Meier survival analysis was conducted according to age and etiology. A logistic regression was conducted by defining Evans Index, thirdventricular morphology index (TVMI) and age as continuous variables andetiology as categorical variables.Results: Eight studies and7177patients were included in this meta analysis.The population contribution in VPS and ETV is5654and1523, respectively.The result of meta analysis showed the risk of failure in ETV was lower thanthat in VPS (OR=0.60,95%CI0.20-1.86), however, the included studieshad great heterogeneity. In this study,428cases were collected with140forETV and288for VPS. The Kaplan-Meier survival showed age had aconsiderable effect on ETV success. Majority of failure in ETV was showedin infants and the failure seldom emerged in children. There is a significantdifference between these three group (p=0.005). Besides, in the infant group,the success of VPS is much higher than than in ETV, with statisticallysignificant (p=0.008). However, etiology has little effect on the success ofETV or VPS. Only in IVH group, the success of VPS is higher than that inETV (p=0.042). According to the regression formula, TVMI (p=0.047) andage (p=0.006) are included. Etiology (p=0.405) and Evans Index (p=0.394)are excluded from the regression formula. Hence, age and TVMI are theindependent prognostic factors for ETV success. The risk of ETV failurewill increase0.925times and0.955times when TVMI decreases0.01andone months younger, respectively.Conclusion: Despite of clinical heterogeneity, meta analysis demonstratesthere is a little higher risk of failure in ETV than that in VPS. Age has morecontributing effect on ETV than VPS. The younger patients undergoingETV, the more risk of failure. Etiology has little effect on outcome of thesetwo approaches. Age and TVMI are two independent prognostic factors forETV. The risk of ETV failure will become high if the age of patient isyoung and the TVMI is low.