Single nucleotide polymorphism array analysis to predict clinical outcome in neuroblastoma patients

Purpose: Neuroblastoma (NB) is a heterogeneous tumor and demonstrates favorable or unfavorable outcomes. In Japan, a nationwide NB mass screening (MS) had been performed on 6-month-old infants for approximately 20 years, which might have detected almost all NB including regressing/maturing tumors. To clarify the heterogeneity of this tumor, we examined genetic alterations in the representative cases using genomewide microarrays. Methods: Genomic DNA was extracted from 198 NB tissue samples and paired blood samples including 76 MS-detected cases and analyzed by single nucleotide polymorphism arrays. Results: The single nucleotide polymorphism array classified the genetic aberrations into 4 types: whole gain/loss type, partial gain/loss type, MYCN-amplified type, and silent type. Most MS-detecting cases belonged to the whole gain/loss type, whereas unfavorable cases who died of disease showed partial gain/loss, MYCN-amplified, or silent types. Conclusions: Genomewide genetic analysis is useful to predict the outcome of patients. Although the cases whose tumors showed whole gain/loss may respond well to contemporary therapy, sparing intensive surgery, current therapeutic strategy may be insufficient for the subgroups with partial gain/loss, MYCN-amplified, or silent type. Validation of these results would provide new tools to predict clinical outcome of children with NB. (c) 2006 Elsevier Inc. All rights reserved.