Met-Score: A 45-Gene Signature May Predict Prostate Cancer Metastatic Risk
A study published in the Journal of Translational Medicine has identified a 45-gene expression signature, called Met-Score, that could help predict which initially localised prostate cancers are more likely to develop metastases.
Researchers analysed six independent cohorts comprising 1,000 patients, including 306 who developed metastatic disease. They identified 27 genes with increased expression and 18 with reduced expression among patients who subsequently developed metastases. These genes were combined into a molecular score designed to capture the biological characteristics associated with metastatic progression.
The model was subsequently evaluated in two independent cohorts without further retraining. In the Johns Hopkins cohort (239 patients, 93 metastatic events), its predictive accuracy, measured by the area under the curve (AUC), reached 0.75 at five years and 0.71 at ten years. In the Durham Veterans Affairs cohort (555 patients, 40 events), the AUC was 0.79 at both time points.
Importantly, Met-Score remained associated with metastasis-free survival after adjustment for pathological Gleason grade. Exploratory analyses also suggested that it might improve risk discrimination among patients with Gleason 7 disease, where conventional grading can conceal considerable biological differences.
Additional analyses found that the gene signature was more strongly expressed in metastatic tumours and could distinguish newly diagnosed metastatic disease from localised prostate cancer in biopsy datasets, with AUC values ranging from 0.93 to 0.98. Experimental findings also suggested that the underlying molecular programme could respond to therapeutic intervention.

Leave a Reply
Want to join the discussion?Feel free to contribute!