Ewing's sarcoma (ES) is the second most common primary malignant bone tumor in children and adolescents. Currently accepted clinical prognostic factors fail to classify ES patients' risk to relapse at diagnosis. We aimed to find a new strategy to distinguish between poor and good prognosis ES patients already at diagnosis. We analysed the gene expression profiles of 14 primary tumor specimens and six metastases from ES patients, using oligonucleotide microarray analysis. The over-expression of two genes was validated by quantitative PCR using the LightCycler system. We identified two distinct gene expression signatures distinguishing high-risk ES patients that are likely to progress from low-risk ES patients with a favorable prognosis of long-term progression-free survival. The microarray-based classification was superior to currently used prognostic parameters. Over-expressed genes in the poor prognosis patients included genes regulating the cell cycle and genes associated with invasion and metastasis, while among the downregulated genes were tumor suppressor genes and inducers of apoptosis. Our results indicate the existence of a specific gene expression signature of outcome in ES already at diagnosis, and provide a strategy to select patients who would benefit from risk-adapted improved therapy.

BACKGROUND. Maintenance of telomeres, in most instances by reactivation of telomerase, is obligatory for the indefinite proliferation of tumor cells. The objective of this study was to evaluate telomere length and telomerase activity (TA) as markers for progression and prognosis in neuroblastoma. METHODS. Primary tumor samples from 51 patients were analyzed for telomere length and TA and were correlated with known prognostic parameters and outcome. RESULTS. Telomere length had a highly significant correlation with prognosis (P = .007). Short telomeres were predictive of a favorable prognosis, whereas long or unchanged telomeres were predictive of a poor outcome. For the first time to their knowledge, the authors have shown that, within the high‐risk group patients, telomere length could define a favorable subgroup that had a progression‐free survival (PFS) rate of 86% compared with a PFS rate of 36% for patients with more adverse disease, which is the expected PFS rate for such patients (P = .04). In a multivariate analysis, telomere length was the most significant prognostic parameter (P = .032). TA was correlated significantly with outcome and with known prognostic factors. High TA and low TA were associated with adverse and favorable outcomes, respectively (P = .01). CONCLUSION. The results of this investigation suggested that telomere length is a highly significant prognostic parameter of clinical relevance in patients with neuroblastoma. In high‐risk patients, telomere length was the sole significant parameter that identified a group of patients who had a favorable prognosis. The authors suggest that telomere length should be included in the recommended diagnostic investigations for patients with neuroblastoma. Cancer 2006. © 2006 American Cancer Society.