Mass Spectrometry-Based Biomarkers to Detect Prostate Cancer: A Multicentric Study Based on Non-Invasive Urine...

Area of Science:

• Urology
• Oncology
• Biomarker Discovery
• Proteomics

Background:

• Prostate cancer (PCa) is a leading cancer diagnosis in men.
• Over-diagnosis and over-treatment due to the prostate-specific antigen (PSA) test lead to unnecessary biopsies.
• Current diagnostic methods for PCa often involve invasive procedures, highlighting the need for less invasive alternatives.

Purpose of the Study:

• To develop and validate a non-invasive biomarker model for prostate cancer detection using urine samples.
• To reduce the rate of unnecessary prostate biopsies by improving diagnostic accuracy.
• To establish a reliable method for PCa risk assessment based on urinary proteomics.

Main Methods:

• Proteomics analysis of urine samples from 970 patients across two clinical centers using capillary electrophoresis coupled to mass spectrometry.
• Development of a biomarker model by combining 181 significant peptides using a support vector machine algorithm on a training set.
• Independent validation of the model in separate patient cohorts.

Main Results:

• The developed biomarker model achieved an Area Under the Curve (AUC) of 0.81 in an independent validation set, outperforming current diagnostic standards.
• The model identified significant peptides from urine proteomics profiles for PCa detection.
• Validation demonstrated the model's effectiveness in distinguishing between patients with and without PCa.

Conclusions:

• The multi-dimensional biomarker model derived from urinary proteomics shows significant potential for improving prostate cancer diagnosis.
• This non-invasive approach can aid in guiding decisions regarding invasive biopsies, potentially reducing over-treatment.
• The findings support the clinical utility of urinary molecular markers in the early detection and management of prostate cancer.