Identification of a 5-gene signature panel for the prediction of prostate cancer progression

Br J Cancer. 2024 Dec;131(11):1748-1761. doi: 10.1038/s41416-024-02854-w. Epub 2024 Oct 14.

Abstract

Background: Despite nearly 100% 5-year survival for localised prostate cancer, the survival rate for metastatic prostate cancer significantly declines to 32%. Thus, it is crucial to identify molecular indicators that reflect the progression from localised disease to metastatic prostate cancer.

Methods: To search for molecular indicators associated with prostate cancer metastasis, we performed proteomic analysis of rapid autopsy tissue samples from metastatic prostate cancer (N = 8) and localised prostate cancer (N = 2). Then, we utilised multiple independent, publicly available prostate cancer patient datasets to select candidates that also correlate with worse prostate cancer clinical prognosis.

Results: We identified 154 proteins with increased expressions in metastases relative to localised prostate cancer through proteomic analysis. From the subset of these candidates that correlate with prostate cancer recurrence (N = 28) and shorter disease-free survival (N = 37), we identified a 5-gene signature panel with improved performance in predicting worse clinical prognosis relative to individual candidates.

Conclusions: Our study presents a new 5-gene signature panel that is associated with worse clinical prognosis and is elevated in prostate cancer metastasis on both protein and mRNA levels. Our 5-gene signature panel represents a potential modality for the prediction of prostate cancer progression towards the onset of metastasis.

MeSH terms

  • Aged
  • Biomarkers, Tumor / genetics
  • Disease Progression*
  • Disease-Free Survival
  • Gene Expression Profiling
  • Gene Expression Regulation, Neoplastic
  • Humans
  • Male
  • Middle Aged
  • Neoplasm Metastasis
  • Neoplasm Recurrence, Local / genetics
  • Neoplasm Recurrence, Local / pathology
  • Prognosis
  • Prostatic Neoplasms* / genetics
  • Prostatic Neoplasms* / pathology
  • Proteomics / methods
  • Transcriptome

Substances

  • Biomarkers, Tumor