Evolving role of myeloablative chemotherapy in the treatment of childhood brain tumours.

Primary brain tumours, a heterogeneous group of cancer that constitute the second most common cancer in childhood, were historically treated with neurosurgical resection and radiation therapy. Chemotherapy has proven to be beneficial for some histological types, which has since led to exploration of the role of high-dose chemotherapy and haematopoietic stem cell rescue. Patients with high-grade glial tumours, primitive neuroectodermal tumours and high-risk medulloblastoma usually fare poorly. The indicators of bad prognosis are metastatic status, extent of resection and age. Children <3 years at diagnosis carry worse prognosis. Rare cancers such as ependymoblastoma, atypical teratoid rhabdoid tumour and choroid plexus carcinoma have a dismal prognosis regardless of the above-mentioned indicators. The use of myeloablative therapy (MAT) has been investigated to improve the rate of long-term DFS, as well as to reduce and delay in the youngest children the use of the craniospinal irradiation associated with unacceptable late effects. We will overview the literature regarding patients with 'good and uncertain indications' to MAT. Ependymoma and brain stem tumours, for which the available data discourage the use of MAT, are excluded. Finally, we will summarize a single Institution experience (Giannina Gaslini Children's Hospital, Genoa) with MAT in the period 1997-2003.