Prognostic impact of gene expression-based classification for neuroblastoma

Area of Science:

• Oncology
• Genomics
• Biostatistics

Background:

• Neuroblastoma is a pediatric cancer with variable outcomes.
• Accurate risk stratification is crucial for effective treatment planning.
• Current prognostic markers may not fully capture disease heterogeneity.

Purpose of the Study:

• To evaluate a predefined 144-gene expression-based classifier for clinical risk estimation in neuroblastoma.
• To assess the classifier's impact on cytotoxic treatment decisions.
• To validate the classifier's performance against existing prognostic strategies.

Main Methods:

• Microarray analysis of 440 neuroblastoma specimens (125 prospective).
• Classification using a 144-gene Prediction Analysis for Microarrays (PAM) classifier.
• Comparison of PAM classification with current prognostic markers and risk estimation.

Main Results:

• The PAM classifier significantly distinguished between favorable and unfavorable neuroblastoma patient groups (P < .001).
• It demonstrated robust discrimination in international cohorts and prospectively analyzed samples.
• PAM served as an independent prognostic marker in multivariate analyses (P < .001).

Conclusions:

• The 144-gene PAM classifier enables precise risk stratification for neuroblastoma patients.
• Gene expression profiling can enhance treatment decisions, particularly for non-high-risk patients.
• This approach contributes to personalized medicine strategies in neuroblastoma care.