Gene expression profiling identifies clinically relevant subtypes of prostate cancer

Area of Science:

• Oncology
• Molecular Biology
• Genomics

Background:

• Prostate cancer exhibits diverse clinical behaviors, from indolent to aggressive metastatic disease.
• Understanding the molecular basis of this heterogeneity is crucial for improved patient outcomes.

Purpose of the Study:

• To explore molecular variations underlying prostate cancer clinical heterogeneity.
• To classify prostate tumors based on gene expression patterns for potential prognostication and treatment stratification.

Main Methods:

• Gene expression profiling of 62 primary prostate tumors, 41 normal prostate specimens, and 9 lymph node metastases using cDNA microarrays.
• Unsupervised hierarchical clustering to identify tumor subclasses.
• Immunohistochemistry on 225 independent prostate tumors to evaluate MUC1 and AZGP1 as surrogate markers.

Main Results:

• Three distinct prostate tumor subclasses were identified based on gene expression patterns.
• Aggressive, high-grade, and advanced-stage tumors, including lymph node metastases, were enriched in specific subtypes.
• MUC1 expression correlated with increased recurrence risk (P = 0.003), while AZGP1 expression correlated with decreased recurrence risk (P = 0.0008).
• MUC1 and AZGP1 staining were independent predictors of recurrence in multivariate analysis.

Conclusions:

• Prostate tumors can be classified by gene expression profiles, offering a basis for improved prognostication.
• MUC1 and AZGP1 serve as valuable surrogate markers for predicting recurrence risk in prostate cancer.
• These findings support the potential for gene expression-based stratification in prostate cancer treatment.