Objective: The objective was to investigate the changes in ventricular volume in hydrocephalic children following successful endoscopic third ventriculostomy (ETV). Materials and methods: Using segmentation techniques, serial measurements of ventricular volume were performed using the MRI scans of 13 hydrocephalic children who had successful ETV between 1999 and 2002 to monitor ventricular response. All patients remained asymptomatic, did not require shunting and demonstrated radiological evidence of stoma patency on phase contrast cine MR, throughout the follow-up period extending from 1 to 3.5 years. There were 6 boys and 7 girls with a mean age at operation of 76 months (range 0.1-196 months). Imaging was obtained preoperatively, 1 week, 3 months, 6 months, 12 months and 24 months postoperatively. Each volume measured was divided by the corresponding average normal volume for sex and age, to calculate the "x Normal" ventricular volume (xN). The patients were divided into two groups for analysis: those children having large ventricular volumes at presentation (>5xN) and those with moderate initial volumes (<5xN). Results: The mean preoperative volume was 207 cm(3) (11.9xN) while the mean volumes at 1 week, 3 months, 6 months, 12 months and 24 months were 120 cm(3) (6.7xN), 104 cm(3) (5.7xN), 119 cm(3) (6.8xN), 146 cm(3) (7.8xN) and 185 cm(3) (10.3xN) respectively, for the entire group. Nine patients had large preoperative ventricular volumes while 4 patients presented with moderate volumes. The pattern of change in ventricular size varied between the large and small volume groups. For the majority of patients presenting with large volumes (>5xN), ventricular size decreased significantly until 3-6 months following ETV, after which the volume change levelled off. In some patients, a slight increase in volume was observed after this period. Patients presenting with moderate initial volumes had a much less steep reduction in ventricular size in the 3-6 months following ETV, after which the volume appeared to stabilise or fall slightly. However, the final volume in both groups remained higher than normal, especially in the large presentation volume group (mean x N volumes at 12 months: large preoperative volume group = 9.8xN, moderate preoperative volume group = 2.4xN). Conclusion: In response to ETV, ventricular volume falls to a value lower than preoperatively but higher than the normalised value for age and sex. All patients appeared to have supranormal volumes in the long term, with the volume stabilising at 3-6 months. This contrasts with shunted patients who continue to exhibit declining ventricular volumes after 6 months. The observation that the final volumes are much higher than normal (especially in the large volume group) implies that the absorptive mechanism works less well in these patients in comparison to normal subjects and it thus appears that successful ETV produces a state of compensated communicating hydrocephalus. The long-term neurocognitive consequence of persistently enlarged ventricles may require further evaluation.
