How should we define patients who are at high risk of death from prostate cancer?

Andrew Vickers

This article has no abstract so we have provided the first paragraph of the full text.

Several biomarkers for prostate cancer exist, and, in general, the higher a patient's level of a biomarker, the worse their prognosis. From this trivial premise, two straightforward conclusions follow. First, for any particular biomarker, patients who have a biomarker level above any randomly chosen cutoff value will have a worse prognosis than patients with a biomarker level below that value. Let's take a PSA level of, say, 12.38 ng/ml (after all, there is no reason to believe that biology has a special affinity for round numbers such as 2). Quick analysis of some data I have to hand reveals that patients with a PSA level of 12.38 ng/ml or above have a higher risk of biochemical recurrence (hazard ratio 3.32; P <0.001) than those with a PSA level below this cutoff value. This finding does not mean, however, that a PSA level of 12.38 ng/ml should be used as a threshold to determine which patients receive adjuvant chemotherapy. Second, because there are lots of different biomarkers for prostate cancer, a threshold applied to just one biomarker is unlikely to separate patients into two homogenous risk groups: all prognostic factors should be considered when assessing prostate cancer risk. For example, in this paper by D'Amico et al., the authors suggest a 2 ng/ml/year cutoff value for PSA velocity; however, we would expect a patient with a PSA velocity of 2.1 ng/ml/year with an organ-confined, Gleason 6 tumor to do better than a patient with a PSA velocity of 1.9 ng/ml/year, a Gleason 9 tumor and a positive lymph node.

ARTICLE NAVIGATION - ARTICLE TOOLS

• Full text
• Download PDF
• Send to a friend
• PDA services
• Permissions

ISSUE NAVIGATION - THIS ISSUE

• Table of Contents
• Previous article
• Next article