Prenatal prediction of need for ventriculoperitoneal shunt in open spina bifida.

OBJECTIVE

To investigate whether the need for ventriculoperitoneal shunting in neonates with open spina bifida can be predicted prenatally.

METHODS

This was a retrospective cohort study of all fetuses with open spina bifida identified at a single referral center between 1998 and 2012. Ultrasound records were reviewed and outcomes were ascertained from maternal, neonatal and pediatric records. The performance of screening was determined by receiver-operating characteristics (ROC) curve analysis.

RESULTS

We identified 124 cases of isolated open spina bifida, of which 48 were liveborn. Detailed postnatal follow-up was available for 39 cases, with an average follow-up time of 3.6 years. A shunt was inserted in 22 (56.4%) cases. The width of the posterior horn of the lateral ventricle (Vp) and the ratio between its width and that of the cerebral hemisphere (Vp/H) at diagnosis were significantly greater in cases that required a shunt compared with those that did not subsequently require a shunt (Vp: median 12.4 mm vs 7.7 mm, P < 0.001, and Vp/H: median 0.54 vs 0.33, P < 0.001, respectively). Similarly, at the last scan before delivery, Vp and Vp/H were significantly higher in those requiring a shunt (Vp: median 23.1 mm vs 8.2 mm, P < 0.001; and Vp/H: 0.54 vs 0.21, P < 0.001, respectively). Using Vp and Vp/H at the initial assessment, the rates of detection of fetuses requiring a shunt postnatally were 68.2% and 86.4% for a false-positive rate of 13%, with positive likelihood ratios of 6 and 7, respectively. All fetuses with a Vp of ≥ 12 mm or a Vp/H of ≥ 0.6 at the time of diagnosis required postnatal shunt insertion (positive predictive value = 100%).

CONCLUSIONS

In cases with open spina bifida the need for a postnatal shunt can be predicted prenatally. This novel observation should be useful in the prediction of outcome and therefore for accurate prenatal counseling and triaging cases for fetal surgery.