Abstract

There has been limited success in identifying prognostic biomarkers in prostate cancer. A partial explanation may be that insufficient emphasis has been put on clearly defining what type of marker or patient category a biomarker study aims to identify and how different cohort characteristics affect the ability to identify such a marker. In this article, the authors put forth the ABC model of prostate cancer, which defines 3 groups of patients with localized disease that an investigator may seek to identify: patients who, within a given time frame, will not develop metastases even if untreated (category A), will not develop metastases because of radical treatment (category B), or will develop metastases despite radical treatment (category C). The authors demonstrate that follow‐up time and prostate‐specific antigen screening intensity influence the prevalence of patients in categories A, B, and C in a study cohort, and that prognostic markers must be tested in both treated and untreated cohorts to accurately distinguish the 3 groups. The authors suggest that more emphasis should be put on considering these factors when planning, conducting, and interpreting the results from prostate cancer biomarker studies, and propose the ABC model as a framework to aid in that process.

Citation

Pettersson A*, Gerke TA*, Fall K, Pawitan Y, Holmberg L, Giovannucci EL, Kantoff PW, Adami H, Rider JR, Mucci LA. The ABC model of prostate cancer: a conceptual framework for the design and interpretation of prognostic studies. Cancer 2017; 123(9): 1490--1496. PMID: 28152172. PMCID: PMC5716345.