A novel treatment stratification system for current low- and intermediate-risk neuroblastoma patients using gene expression-based classification

To improve risk estimation of low- and intermediate-risk neuroblastoma patients, we determined gene expression profiles of 709 primary neuroblastoma samples using 44 K oligonucleotide-microarrays. A prognostic gene expression-based classifier was generated using a training set of 75 samples and evaluated on the remaining 634 test set patients. In the low- and intermediate subgroup of patients (n = 413), the classifier predicted event-free (EFS) and overall survival (OS) accurately (EFS, 0.846 ± 0.020 vs. 0.376 ± 0.071; OS 0.996 ± 0.004 vs. 0.699 ± 0.070; both p < 0.001). In univariate Cox regression models based…