To the Editor.
—A PSA test should be viewed as a risk assessment for prostate cancer over a continuum of PSA values. Our article reported that (1) in the PSA range of 2.6 to 4.0 ng/mL, 22% of 317 volunteers with a benign-feeling prostate gland had cancer detected, and (2) measurement of percent-free PSA (%FPSA) could avoid 18% of the unnecessary biopsies while detecting 90% of cancers. These cancers would not be detected using traditional biopsy criteria. We hypothesize that earlier detection will result in improved outcomes. To this end, we are conducting longitudinal outcomes studies.Dr Moul points out the contradiction that Carter et al1 consider biannual testing sufficient in men with PSA levels less than 2.0 ng/mL and that diagnosing cancer at levels of 4.0 to 5.0 ng/mL is sufficient for detecting curable cancer, while we recommend biopsy in men with PSA levels higher than 2.6 ng/mL. Moul believes that