[Skip to Content]
[Skip to Content Landing]
September 3, 1997

PSA Thresholds for Prostate Cancer Detection

Author Affiliations

Barberton Citizens Hospital Barberton, Ohio

JAMA. 1997;278(9):699-700. doi:10.1001/jama.1997.03550090023014

To the Editor.  —I concur with Dr Catalona and colleagues1 regarding the goal of increasing the detection of prostate cancer and was interested in their use of a lower serum PSA level, 2.6 ng/mL, on which to base prostate biopsy. However, in our current economic climate, we must turn a critical eye toward all new diagnostic testing, especially those tests that are expensive. The authors of this study have not shown convincing evidence of the need for an additional test, that of free PSA. They indicate a positive predictive value of 22% by using the PSA alone in their study population of 317 men who had total PSA levels between 2.5 and 4.0 ng/mL and who underwent needle biopsy. They then state,Using a free PSA cutoff of 27% as a criterio for biopsy would have detected 90% of the cancers and avoided 18% of unnecessary biopsies, yielding a positive predictive value of 24% in the men who underwent biopsy. Therefore, percentage of free PSA would provide