• prostate specific antigen;
  • free prostate specific antigen;
  • prostate specific antigen transitional zone density;
  • transitional zone carcinoma;
  • incidental prostate carcinoma



This article summarizes the experience and results of different prostate carcinoma screening projects using total prostate specific antigen (PSA) as the initial test and different diagnostic tests to improve specificity.


The seven projects studied included 1) a mass screening study using PSA as the initial test in 21,079 volunteers; 2) an investigation of the usefulness of normal and age-referenced PSA cutoffs in 1618 men; 3) a PSA-based screening study of 2272 asymptomatic blood donors; 4) an investigation of the incidence and clinical significance of transitional zone carcinoma in 340 men with negative rectal examination findings and clearly visible prostatic zones on three-dimensional transrectal ultrasound; 5) determination of percent free PSA in one retrospective and two prospective screening studies to define the optimal range of total PSA and determine the appropriate cutpoints for percent free PSA within this range; 6) evaluation of the diagnostic benefit of PSA transitional zone density in 308 screening volunteers; and 7) a study of the impact of PSA-based screening on the percentage of incidental prostate carcinoma diagnosed in 1543 men undergoing transurethal resection of the prostate.


1) Of the 21,078 volunteers, 1618 (8%) had elevated PSA levels. Of these men, 778 (48%) underwent biopsies; 197 biopsies (25%) were positive for prostate carcinoma and 135 patients underwent radical prostatectomy. Ninety-five of the 135 pathologically staged lesions (70%) were found to be organ-confined. 2) A PSA cutoff of 2.5 ng/mL in men age 45-49 years and of 3.5 ng/mL in men age 50-59 years with normal digital rectal examination findings resulted in an 8% increase in both the number of biopsies (66 of 778) and the detection rate of organ-confined disease. 3) Of the 2272 men, 284 had elevated PSA levels and prostate carcinoma was detected in 62 men. All patients underwent radical prostatectomy and histologic examination revealed organ-confined disease in all but eight men. 4) Ninety-eight of 340 men (28.8%) had biopsies positive for carcinoma; 28 of these patients (28.5%) had carcinoma that originated in the transitional zone only. 5) In the retrospective study, receiver operating characteristic curve analysis showed that by using a percent free PSA of 18% as a biopsy criterion in men with an elevated PSA serum level, 37% of the negative biopsies could be eliminated although 94% of all carcinomas would still be detected. In the first prospective study, 106 of 158 men with elevated total PSA values between 2.5 and 10.0 ng/mL were further evaluated and 37 prostate carcinomas were detected. By using a percent free PSA of ≤22% as a biopsy criterion, 30% of the negative biopsies could be eliminated although 98% of the carcinomas would still be detected. In the second prospective study, 120 of 465 men with total PSA levels between 1.25 and 6.49 ng/mL, a percent free PSA of <18%, and normal digital rectal examination findings were further evaluated and 27 (22.5%) were found to have prostate carcinoma. 6) Receiver operating characteristic curve analysis for PSA transitional zone density showed that by using a PSA transitional zone density of >0.22 ng/mL/cc as a biopsy criterion, 24.4% of negative biopsies could be avoided without missing the detection of a single carcinoma. 7) In the prescreening era the incidence of T1a Grade 1 and 2 carcinomas was 3.1% and the incidence of T1a Grade 3 and T1b carcinoma was 2.3%, whereas in the years after the establishment of PSA-based screening the incidence was 4.6% and 1.03%, respectively.


These data suggest that PSA-based screening increases the detection rate of clinically significant and organ-confined tumors. Percent free PSA and PSA transitional zone density provide an additional diagnostic benefit over total PSA. Cancer 1997; 80:1818-29. © 1997 American Cancer Society.