Intraoperative ultrasound in pediatric brain tumors: does the surgeon get it right?

Intraoperative ultrasound (iUS) is a valuable tool-inexpensive, adds minimal surgical time, and involves minimal risk. The diagnostic predictive value of iUS is not fully characterized in Pediatric Neurosurgery. Our objective is to determine if surgeon-completed iUS has good concordance with post-operative MRI in estimating extent of surgical resection (EOR) of pediatric brain tumors.

We reviewed charts of all pediatric brain tumor resections (single institution 2006-2013). Those with iUS and postoperative imaging (<1 week) were included. The surgeon's estimation of the EOR based on iUS and the post-operative neuroimaging results (gold standard) were collected, as well as information about the patients/tumors.

Two hundred two resections were reviewed and 58 cases were included. Twenty-six of the excluded cases utilized iUS but did not have EOR indicated. The concordance of interpretation between iUS and post-operative MRI was 98.3%. Of 43 cases where iUS suggested gross total resection, 42 were confirmed on MRI (negative predictive value (NPV), 98%). All 15 cases where iUS suggested subtotal resection were confirmed on MRI (positive predictive value (PPV), 100 %). Agreement between iUS and post-operative imaging had an overall Kappa score of 0.956, signifying almost perfect agreement.

The results from this study suggest that iUS is reliable with both residual tumor (PPV-100%) and when it suggests no residual (NPV-98%) in tumors that are easily identifiable on iUS. However, tumors that were difficult to visualize on iUS were potentially excluded, and therefore, these results should not be extrapolated for all brain tumor types.