@article{JLPM11369,
author = {Evelina La Civita and Felice Crocetto and Matteo Ferro and Daniela Terracciano},
title = {The end of the single-marker era: multimarker algorithms and artificial intelligence in prostate cancer diagnosis—a narrative review},
journal = {Journal of Laboratory and Precision Medicine},
volume = {0},
number = {0},
year = {2026},
keywords = {},
abstract = {Background and Objective: Prostate-specific antigen (PSA) testing has long represented the cornerstone of prostate cancer (PCa) screening and diagnosis; however, its limited specificity has contributed to substantial overdiagnosis and overtreatment, particularly within the PSA “gray zone”. The increasing clinical complexity of PCa and the heterogeneity of disease biology highlight the need for diagnostic approaches that move beyond single-marker strategies. The objective of this review is to provide a clinically oriented synthesis of the transition from PSA-based testing to integrated diagnostic models combining circulating biomarkers, multiparametric magnetic resonance imaging (mpMRI), and artificial intelligence (AI) for improved risk stratification.Methods: A narrative review was conducted through a structured literature search of PubMed/MEDLINE, Scopus, and Web of Science from database inception to November 2025. Search terms included combinations of “prostate cancer”, “PSA”, “Prostate Health Index”, “PHI”, “4Kscore”, “Stockholm3”, “Proclarix”, “IsoPSA”, “biomarkers”, “liquid biopsy”, “multiparametric MRI”, “mpMRI”, “artificial intelligence”, “machine learning”, “neural networks”, and “risk stratification”. Original studies, meta-analyses, clinical guidelines, and high-quality reviews focused on biomarker performance, integrated diagnostics, and AI-based models were included. Articles not directly related to PCa diagnosis or lacking full-text data were excluded. Study selection and synthesis were performed through independent screening and consensus discussion among the authors.Key Content and Findings: Emerging evidence supports the superior clinical utility of multimarker algorithms, including prostate health index (PHI), 4Kscore, Stockholm3, Proclarix, and IsoPSA, in improving the detection of clinically significant prostate cancer (csPCa) while reducing unnecessary biopsies. mpMRI has enhanced lesion characterization but remains limited by cost, inter-reader variability, and equivocal findings. Integrated approaches combining biomarkers and mpMRI demonstrate improved diagnostic accuracy compared with either modality alone. Furthermore, AI-driven models incorporating biomarker data, clinical variables, and imaging parameters show promising performance, with higher discrimination for csPCa and more personalized risk prediction. These models consistently report improved area under the curve (AUC), sensitivity, and specificity relative to single-marker approaches.Conclusions: Current evidence supports a paradigm shift toward integrated, multimodal diagnostic pathways that combine circulating biomarkers, imaging, and AI to optimize PCa detection and risk stratification. Such approaches may reduce overdiagnosis, minimize unnecessary invasive procedures, and enable more personalized and equitable diagnostic strategies aligned with the principles of precision medicine.},
issn = {2519-9005}, url = {https://jlpm.amegroups.org/article/view/11369}
}