In Colombia, prostate cancer (PCa) ranks first in incidence among men. The prostate-specific antigen (PSA) test remains the most widely adopted primary screening biomarker for the presumptive diagnosis of PCa. However, elevated serum PSA levels may also occur in benign conditions, making it a suboptimal screening tool due to its low specificity, which often leads to unnecessary biopsies and overdiagnosis. Urine and serum emerge as non-invasive biofluids for biomarker discovery. Urine becomes enriched with prostate-derived cells and secreted proteins following digital rectal examination (DRE), which promotes the release of prostatic material into the urethra. Serum, on the other hand, contains circulating molecules released by the prostate and other tissues. The accessibility and molecular content of both biofluids make them ideal substrates for the identification of biomarkers aimed at the early detection of patients with suspected PCa.

The objective of this study was to identify candidate biomarkers in urine and serum from a cohort of patients with suspected prostate cancer. We previously performed transcriptomic analyses by RNA sequencing in a cohort of 75 participants with suspected PCa: 36 Gleason-positive cancer samples and 39 PCa-negative samples. Histopathological results according to Gleason Grade Group (GG) were distributed as follows: GG1 n=1, GG2 n=2, GG3 n=15, GG4 n=5, and GG5 n=13. The mean age of participants was 65 years, and PSA levels ranged from 4.34 ng/ml to 191 ng/ml. Blood samples, post-DRE urine samples, and prostate biopsies were collected from all participants according to clinical indication.

Differential expression analysis identified 29 genes as potential biomarkers, selected based on high PSA values, participant age, and histopathological diagnosis (Gleason score), of which 17 were verified by QuantiGene Plex and showed significant correlation with RNA-seq data (Pearson correlation, p < 0.005). Subsequently, urine and serum samples were processed for protein extraction. The 29 candidate proteins derived from gene expression analysis were evaluated at the protein level. Protein-level detection was conducted using liquid chromatography coupled with tandem mass spectrometry (LC–MS/MS), and label-free quantification (LFQ) was performed in MaxQuant. For the exploratory proteomic analysis, 5 participants were randomly selected, and both urine and serum samples were analyzed, totaling 10 samples. This exploratory phase was designed to assess protein detection feasibility prior to validation in a larger cohort of 286 participants.

Exploratory proteomic analysis identified 3 targets in serum: ALCAM, APOE, and EEF1A2; and 12 targets in urine: ACY1, ALCAM, AMACR, APOE, CLDN3, EEF1A2, ENTPD5, EPCAM, FASN, GOLM1, ISG15, and RAC3. Our findings demonstrate the successful translation from transcriptomic to proteomic detection, identifying 12 potential protein biomarkers for early prostate cancer detection. These biomarkers are currently being validated in a larger cohort with greater statistical power to assess their diagnostic performance and clinical utility. The identification of both urine-specific and serum-detectable targets provides flexibility for future clinical implementation and may complement PSA testing to reduce unnecessary biopsies. These results highlight the translational potential of combining multi-omic approaches for biomarker discovery in the Colombian population, although further validation is required to confirm their sensitivity, specificity, and clinical applicability.