Percent free PSA as an additional measure in a prostate cancer screen

To assess the ability of the ratio of free prostate specific antigen to total prostate specific antigen (% fPSA) to aid in selection of subjects who would require follow-up studies.

Retrospective, nonrandomized analysis measuring total prostate specific antigen (tPSA) and free prostate specific antigen (fPSA) in serum from men enrolled in a community-based prostate cancer screening offered by the Cancer Outreach Program (COP), Christiana Care Health Systems (CCHS), Wilmington, DE. PARTICIPANTS Informed consent was obtained from 172 of the 231 eligible participants. Complete laboratory and follow-up data, excluding ethnic origin, was gathered from 157 participants; therefore, theoretical participation was 75% (172/231) while realized participation rate was 68% (157/231). Criteria for inclusion in this study included ambulatory men of age 40 and older with serum total PSA (tPSA) level >1.9 ng/mL and/or an enlarged or abnormal prostate by digital rectal examination (DRE). Subjects with a history of prostate cancer or prostatitis were excluded.

Laboratory tests to determine tPSA and fPSA were performed on serum samples obtained from consenting participants. Percent fPSA was calculated. Results of clinical findings with respect to each participant's DRE were recorded as normal, suggestive of a benign condition, such as benign prostate hyperplasia, or suspicious for prostate cancer. When available, transurethral ultrasound (TRUS) results and biopsy results were also noted. Each participant's results were evaluated and given one of the following diagnoses: Normal prostate (N), Benign Prostate Hyperplasia (BPH), or Prostate Cancer (PCa).

Prevalence of cancer 3/157 = 0.0191; of BPH = 0.688; of N = 0.293. Median values for tPSA for each of these groups were as follows: N, 2.9 ng/mL; BPH, 3.0 ng/mL; and PCa, 6.3 ng/mL; (p = 0.079). Median values for fPSA were as follows: N, 0.6 ng/ mL; BPH, 0.5 ng/mL; PCa, 0.5 ng/mL; (p = 0.51). Median values for % fPSA were as follows: N, 19%; BPH, 17%; and PCa, 9%; (p = 0.01). Medians were found to differ for % fPSA measurements, but not for tPSA or fPSA values. DRE screening results of 110 subjects were reported as not normal indicating either an enlarged (n = 97) or abnormal prostate gland (n = 13). Nine subjects had normal DRE results with serum tPSA level above 4.0 ng/mL. Using the combination of DRE and tPSA > 4.0 ng/mL as criteria for the recommendation of follow-up studies, 119 of the participants would have been advised to seek additional testing.

Using these two routine criteria, 119 (119/157; 76%) subjects would be candidates for follow-up procedures, such as transurethral ultrasound (TRUS) and/or sextant prostate biopsy. By adding % fPSA results of < 10% fPSA to tPSA results > 4.0 mg/mL as criteria for follow-up studies, specificity can be improved threefold with sensitivity unchanged. If an abnormal DRE suggestive of malignancy was included as part of the criteria, sensitivity of the diagnostic scheme would reach 100%. Using the triple diagnostic parameters of tPSA > 4.0 ng/mL, the ratio of free prostate specific antigen to total prostate specific antigen <10%, and an abnormal prostate DRE, 16 participants would be recommended for follow-up studies (16/157; 10%). This would eliminate 103 subjects from unnecessary and expensive testing.