VALUE OF POSTOPERATIVE SURVEILLANCE IMAGING IN THE MANAGEMENT OF CHILDREN WITH SOME COMMON BRAIN-TUMORS

The rationale for obtaining surveillance computerized tomography (CT) scans or magnetic resonance (MR) images in pediatric patients with bra in tumors is that early detection of recurrence may result in timely t reatment and better outcome. The purpose of this study was to investig ate the value of surveillance cranial images in a variety of common pe diatric brain tumors managed at a tertiary care pediatric hospital. A retrospective chart review was performed of children with astrocytoma of the cerebral hemisphere, cerebellum, optic chiasm/hypothalamus? or thalamus; cerebellar ol supratentorial high-grade glioma, supratentori al ganglioglioma; posterior fossa or supratentorial primitive neuroect odermal tumor (PNET); and posterior fossa ependymoma. Data were analyz ed to determine the frequency with which recurrences were identified o n a surveillance image and how the type of image at which recurrence w as identified related to outcome. In 159 children, 17 of 44 recurrence s were diagnosed by surveillance imaging. The percentage of recurrence s identified by surveillance imaging was 64% for ependymoma, 50% for s upratentorial PNET, 43% for optic/hypothalamic astrocytoma, and less t han 30% for other tumors. The rate of diagnosis of recurrence per surv eillance image varied from 0% to 11.8% for different tumor types. Only for ependymomas did there appear to be an improved outcome when recur rence was identified prior to symptoms. Our results indicate that, usi ng the protocols outlined in this study, surveillance imaging was not valuable in identifying recurrence of cerebellar astrocytoma or suprat entorial ganglioglioma during the study period, but was probably worth while in identifying recurrence of posterior fossa ependymoma and opti c/hypothalamic astrocytoma and, possibly, medulloblastoma. Surveillanc e protocols could be made more effective by individualizing them for e ach type of tumor, based on current data on the patterns of recurrence .