Prostate cancer is the most common noncutaneous malignancy in men in the United States. In 2025, an estimated 313,780 patients will be diagnosed with prostate cancer, including approximately 85% having nonmetastatic disease.1 Globally, it is the second most common cancer in men, accounting for nearly 1.5 million cases in 2022.2 Prostate cancer is clinically highly heterogeneous. Risk stratification, by providing an estimate of the clinical trajectory, helps guide management. Patients with a new diagnosis of prostate cancer are most commonly risk stratified using a combination of clinical and pathologic features, including the clinical tumor stage, serum prostate-specific antigen (PSA) level, the percentage of positive cores from the biopsy, and the Gleason grade group.

Formal risk stratification first began in the 1990s, with the introduction of the Partin Tables,3 Kattan Nomogram,4 and the 3-tiered D’Amico risk groups.5 While these and other efforts, such as the Cancer of the Prostate Risk Assessment (CAPRA) Score,6 have overall improved risk stratification, modern risk stratification systems7 still largely rely on the use of conventional clinical and pathologic variables. Despite improvement over time, these conventional risk stratification schemas have suboptimal performance,8 leading to undertreatment for certain patients and overtreatment for other (with associated toxicities). As such, novel biomarkers are needed to improve risk stratification and advance personalized medicine for patients with prostate cancer.

The goal of this review is to summarize advances in novel biomarkers for nonmetastatic prostate cancer, focusing on molecular, digital pathology, and imaging approaches, and discuss challenges and future directions.