Lack of benefit of endoscopic ventriculoperitoneal shunt insertion: a multicenter randomized trial.

OBJECT

Endoscopically assisted ventricular catheter placement has been reported to reduce shunt failure in uncontrolled series. The authors investigated the efficacy of this procedure in a prospective multicenter randomized trial.

METHODS

Children younger than 18 years old who were scheduled for their first ventriculoperitoneal (VP) shunt placement were randomized to undergo endoscopic or nonendoscopic insertion of a ventricular catheter. Eligibility and primary outcome (shunt failure) were decided in a blinded fashion. An intention-to-treat analysis was performed. The sample size offered 80% power to detect a 10 to 15% absolute reduction in the 1-year shunt failure rate. The authors studied 393 patients from 16 pediatric neurosurgery centers between May 1996 and November 1999. Median patient age at shunt insertion was 89 days. The baseline characteristics of patients within each group were similar: 54% of patients treated with endoscopy were male and 55% of patients treated without endoscopy were male; 30% of patients treated with and 26% of those without endoscopy had myelomeningocele; a differential pressure valve was used in 51% of patients with and 49% of those treated without endoscopy; a Delta valve was inserted in 38% of patients in each group; and a Sigma valve was placed in 9% of patients treated with and 12% of those treated without endoscopy. Median surgical time lasted 40 minutes in the group treated with and 35 minutes in the group treated without endoscopy. Ventricular catheters, which during surgery were thought to be situated away from the choroid plexus, were demonstrated to be in it on postoperative imaging in 67% of patients who had undergone endoscopic insertion and 61% of those who had undergone nonendoscopic shunt placements. The incidence of shunt failure at 1 year was 42% in the endoscopic insertion group and 34% in the nonendoscopic group. The time to first shunt failure was not different between the two groups (log rank = 2.92, p = 0.09).

CONCLUSIONS

Endoscopic insertion of the initial VP shunt in children suffering from hydrocephalus did not reduce the incidence of shunt failure.