Role of endoscopic third ventriculostomy at infected cerebrospinal fluid shunt removal

Cerebrospinal fluid shunt infection is distressing, especially in the pediatric population. Usually, infected CSF shunts are removed, and after temporary external CSF drainage, reinsertion of the CSF shunt is necessary. Unfortunately, it is not rare to encounter CSF reinfection after shunt renewal, and furthermore, the reinserted CSF shunt is at a considerable risk of malfunction. Endoscopic third ventriculostomy (ETV) is a potent option in managing CSF shunt infection, although ETV failure may occur more frequently when it is used to remove an infected shunt. The authors retrospectively evaluated CSF reinfection after using ETV during removal of infected CSF shunts; then the longevity of ETV and of successive reinserted ventriculoperitoneal shunts (VPSs) after ETV failure were also examined.

Children with shunted hydrocephalus were retrospectively reviewed, and data on their initial CSF shunt infections were extracted. Thirty-six children underwent VPS reinsertion (the VPS group), and 9 underwent ETV after removal of the infected CSF shunt (the ETV group). As the primary outcome, ETV efficacy against CSF reinfection within 6 months was analyzed by comparing the reinfection rates, and the risk factors for CSF reinfection were analyzed by logistic regression. The longevity of the reinserted shunt in the VPS group was calculated using the Kaplan-Meier method, which was compared with ETV longevity as the secondary outcome, and also with the longevity of reinserted VPSs in the ETV group after ETV failure as the tertiary outcome.

Reinfection of CSF was seen in 27.8% of children in the VPS group. Among 9 children in the ETV group, only 1 (11.1%) had CSF reinfection. However, logistic regression analysis failed to show that performing ETV was a significant factor protecting against CSF reinfection: the significant risk factors were younger age at reinsertion of VPS or ETV (p = 0.037) and a history of shunt revisions (p = 0.011). The longevity of reinserted VPSs in the VPS group was calculated to be 658 ± 166.3 days (mean ± SE). Longevity of ETV was compared in the analysis of the secondary outcome, which was 929.2 ± 511.1 days, and there were no significant differences between these durations. Only 2 ETVs stayed patent, and a VPS was eventually implanted in the other 7 children. The longevity of this reinserted VPS in the ETV group, calculated based on these 7 children, was 2011.1 ± 540.7 days, which was confirmed to be longer than that in the VPS group (p = 0.031).

Although the protective effect of using ETV during removal of an infected CSF shunt on reinfection is marginal, the ETV longevity can be considered equivalent to that of reinserted VPSs. Even if ETV failure occurs, the reinserted VPS has significantly better longevity than a VPS reinserted without using ETV, and use of ETV during infected CSF shunt removal can be considered a potent alternative or at least an adjunct to VPS reinsertion.