DYSEMBRYOPLASTIC NEUROEPITHELIAL TUMOR - MORPHOLOGICAL, IMMUNOCYTOCHEMICAL, AND DEOXYRIBONUCLEIC-ACID ANALYSES IN A PEDIATRIC SERIES

OVERTREATMENT BY RADIOTHERAPY and/or chemotherapy for central nervous system tumors in infancy and childhood may be deleterious, so the reco gnition of surgically curable clinicopathological entities is mandator y. The dysembryoplastic neuroepithelial tumor is a complex multinodula r lesion consisting of glial nodules, associated with a specific glion euronal element and/or with focal cortical dysplasia, and occurring in young patients presenting with intractable, mostly complex partial, s eizures without neurological deterioration. We report on 14 patients; 9 were from a series of 600 pediatric patients with intracranial centr al nervous system tumors studied at a single institution from 1988 to 1993, and 5 were referred from other pediatric hospitals. Six tumors w ere frontal, six were temporal, one was parietal, and one was occipito parietal. Computed tomographic scans disclosed hypodense lesions with cystic appearances in 4 patients and slight focal postcontrast enhance ments in only 2 patients, whereas magnetic resonance imaging, availabl e for 7 of 14 patients, showed hypointense lesions in T1-weighted imag es and hyperintense lesions in T2-weighted images. Defonnities of the overlying cranium were also observed in five patients. The age range a t the time of surgery (excluding a 20-year-old male patient who underw ent surgery at the main pediatric hospital) was 2.6 to 13 years, with a mean of 6.68 years. The male to female patient ratio was 10:4, and t he duration of symptoms was 0.2 to 6 years. Monoclonal antibody immuno staining with neuronal specific enolase, neurofilament, microtubule-as sociated protein MAP2 (clone AP20; Sigma Chemical Co., St. Louis, MO), synaptophysin (Biogenex, San Ramon, CA), and beta-tubulin confirmed t he obvious neuronal component, whereas immunostaining with glial fibri llary acidic protein and S-100 confirmed the glial component. The prol iferative cell nuclear antigen labeling index range was 0.38 to 7.65%, with a mean of 2.1% +/- a standard deviation of 1.96, whereas MIB1 (I mmunotech, Marseille, France) (a clone recently developed from Ki-67) immunostained only sporadic nuclei. The deoxyribonucleic acid ploidy i n 11 of 14 patients (static cytometry) was diploid in all (deoxyribonu cleic acid index range, 0.92 to 1.08) except one, in whom it was aneup loid-triploid (deoxyribonucleic acid index, 1.52). In two patients, fo llow-up computed tomography failed to disclose significant growth, and the remaining patients have had no recurrence to date. Awareness of t his clinicopathological entity is essential for therapeutic and progno stic purposes.