A comparison of the free fraction of serum prostate specific antigen in men with benign and cancerous prostates: the best case scenario

J Urol. 1996 Aug;156(2 Pt 1):350-4. doi: 10.1097/00005392-199608000-00004.


Purpose: In most previous studies of free-to-total serum prostate specific antigen (PSA) ratios, the specimens from patients with prostate cancer or those with benign prostatic hyperplasia (BPH) have not been highly characterized. We compared preoperative sera from post-radical prostatectomy patients with clinically significant cancers of at least 2 cm.3 to sera from those with BPH and large, biopsy negative prostates.

Materials and methods: We used 2 different time resolved immunofluorometric assays for free and total PSA, and a combination of a chemoluminescent immunoassay for free PSA detection with an immunoradiometric assay for total PSA to measure free and total PSA. The serum ratios of free-to-total PSA in these assays were compared to those obtained previously from gel filtration studies. Sera from 51 men with prostate cancer volumes of 2 to 18 cm.3 were compared to those from 48 men with BPH and a mean prostate volume of 78 +/- 7 cm.3. The respective mean serum PSA levels plus or minus standard deviation were 10.0 +/- 6.3 and 8.9 +/- 7.2 ng./ml.

Results: Monoclonal assays for free PSA confirmed the previous study with gel filtration. For PSA 4 to 10 ng./ml., 94 to 95% of the men with prostate cancer were correctly diagnosed, with a cutoff of less than 15% for free-to-total PSA on immunofluorometric assay and less than 14% for chemoluminescent immunoassay with immunoradiometric assay. However, 46% (immunofluorometric assay) and 36% (chemoluminescent immunoassay and immunoradiometric assay) of men with BPH did not have enough free PSA for diagnosis of BPH (that is 36 to 46% false-positive rate).

Conclusions: For total PSA 4 to 10 ng./ml., the sensitivity of approximately 15% free-to-total PSA for prostate cancer is high (94 to 95%) but 36 to 46% of men with BPH and a large gland will not be correctly identified. For PSA 2 to 4 ng./ml., no ratio of percent free-to-total PSA discriminated BPH from prostate cancer.

Publication types

  • Comparative Study
  • Research Support, Non-U.S. Gov't

MeSH terms

  • Antibodies, Monoclonal
  • Humans
  • Immunoassay
  • Male
  • Prostate-Specific Antigen / blood*
  • Prostatectomy
  • Prostatic Hyperplasia / blood*
  • Prostatic Neoplasms / blood*
  • Regression Analysis
  • Sensitivity and Specificity


  • Antibodies, Monoclonal
  • Prostate-Specific Antigen