Brain tumors in children under 3 years of age: recent experience (1987-1997) in 39 patients

Brain tumors in children under 3 years of age differ in clinical presentation and pathological behavior from those in older patients. In this study, we reviewed data from 39 children (24 males and 15 females) under 3 years of age who were treated for intracranial brain tumors since the introduction of magnetic resonance imaging. The purpose was to assess correlations between clinicoradiological and treatment-related factors, and their impact on prognosis. The following factors were analyzed: sex, age, duration of symptoms, intracranial hypertension on admission, tumor location, surgical removal and histology. Associations between these factors and correlations with prognosis were determined using bivariate analyses (chi(2) test) and Kaplan-Meier survival curves. Collins' concept of a period of risk recurrence was tested. Mean follow-up was 41 months (range 0-136). In March 1998, 20 children were still alive (51.2%) with a mean survival time of 65 months (range 2-136). The incidence of supratentorial tumors was significantly higher in children less than 1 year old (p = 0.027). Lateral tumors were 9/10 (90%) supratentorial versus only 7/26 (27%) midline tumors (p = 0.001). Outcome (dead or alive) was significantly better (p = 0.037) for low-grade astrocytomas (9/12 = 75% survival) when compared to ependymomas (2/6 = 33%) and primitive neuroectodermal tumors (3/12 = 25%). Total tumor removal was achieved in 20 cases and was associated with a better outcome (65 vs. 33% survival; p = 0.049). Survival analysis confirmed a worse prognosis for children with ependymomas and primitive neuroectodermal tumors (p = 0.011) and revealed a worse survival for children with intracranial hypertension on admission (p = 0.047). Total tumor removal was associated with a longer survival, although not significantly (p = 0. 077). Finally, we found no exceptions to Collins' law.