Neuroblastoma (NB) remains an enigmatic cancer for which the cure rate for high-risk patients remains <30%, and there are no current methods to predict which of these patients will die from the disease. Wei et al. utilized cDNA microarray-generated gene expression profiles and artificial neural networks (ANNs) to predict the outcomes of patients with NB. They demonstrated that ANNs could ''learn'' the inherent prognostic signatures residing in the expression profiles of primary tumors at diagnosis, and accurately predict outcomes of independent test patients using only 19 genes. Notably, these genes can further separate high-risk patients, classified by the current Children's Oncology Group risk stratification, according to their outcomes. This shows promise to improve the survival rate of high-risk patients through tailoring individual therapy. For details, see the article by Wei et al. on page 6883 of the issue.

[Table of Contents]