a1-antichymotrypsin (PSA-ACT) complex. The rest of the complexes are present either in insignificant amounts or are not detectable immunologically.8
The amount of fPSA found in patients with prostate cancer is usually minimal, approximately between 0 percent to 5 percent, whereas a much higher percentage of fPSA is present in patients with benign prostate hyperplasia (BPH). Measuring fPSA alone, however, does not differentiate BPH from prostate cancer because of the much higher tPSA found in cancer patients than in BPH patients. Even at small percentages, the fPSA concentration in serum taken from cancer patients may be much higher than the concentration of fPSA found in BPH patients. Only when fPSA is expressed as % fPSA are the values of % fPSA higher in BPH than in cancer patients.
Assay for fPSA
The assay for free PSA usually follows the format of a sandwich type, solid phase enzyme immunoassay.9 Monoclonal antibodies specific to fPSA are used in these assays; they bind to the side of the fPSA molecule that's usually blocked by ACT in the PSA-ACT complex. Therefore, all fPSA assays using fPSA specific monoclonal antibody do not cross-react with the PSA-ACT complex.9
Because the fPSA concentration in the serum is very low, it's important to establish a sensitive fPSA assay capable of detecting at least 0.1ng/mL. Recent studies show that there were benefits to determining the % fPSA for people with tPSA in the normal range, such as in the tPSA ranges between 2.6ng/mL or 3ng/mL to 4ng/mL.10 In those specimens, the fPSA concentration may be as low as 0.1ng/mL. Therefore, the precision of the fPSA assay is important at this low concentration range if an accurate calculation of % fPSA at this level is desired.11
Since the concentration of fPSA will change in storage because of the continuous complexing of fPSA with protease inhibitors in the serum, the fPSA level will decline significantly in vitro within 24 hours unless the serum is kept frozen. It is recommended that the serum be separated within three hours and the sample either assayed within 24 hours or frozen for analysis at a later date.
PSA has a variety of applications, including:
1. differentiation of BPH from cancer of the prostate (CAP) during screening,
2. reducing biopsies for those with tPSA below 4ng/mL,
3. identification of tumor with normal tPSA,
4. as a prognostic marker and
5. differentiation of fPSA of prostate cancer from fPSA of BPH.
The following is a summary of each.
* Differentiation of BPH from CAP during screening. Because of its relative tissue specificity, PSA (or tPSA) was the first tumor marker recommended, in conjunction with digital rectal examination, for the screening of prostate cancer.12,13 Screening should be undertaken annually for men above 50 years old, since it's important to detect any highly curable, organ-confined tumors before they become metastasized.
When the tumor is still localized, the serum PSA level usually falls between 4ng/mL to 10ng/mL. Unfortunately, PSA is not specific for prostate cancer; a large percentage of patients with BPH also exhibit serum PSA in this concentration range.
In the past, biopsies were supposed to be performed for those who had serum PSA levels above 4ng/mL to confirm whether they had cancer or not. Consequently, BPH patients were often subjected to costly and painful biopsies. To reduce such unnecessary biopsies for BPH patients, the determination of % fPSA is recommended to help differentiate BPH from prostate cancer.14-16
Among the many new approaches to improve the specificity of the PSA assay, such as calculating PSA density, monitoring PSA velocity and applying the PSA reference range, the determination of % fPSA emerged as the best way to differentiate BPH from CAP. Free PSA can be expressed in several ways. Currently, the expression of fPSA as [fPSA/total PSA] x 100 is most frequently used. For the majority of the reports concerning the various cutoff values of % fPSA, the Hybritech assay kits and their values were most frequently used.15,16
As illustrated in Fig. 1, we found that there is a large overlap in % fPSA from cancer patients and patients with BPH. One must select two different cutoff points based on decisions regarding how many cancer patients to be identified and how many BPH patients one would like to exclude from the biopsy. The upper line is the cutoff point used to spare BPH patients from biopsy. Patients with % fPSA values greater than this cutoff point will not be recommended to receive biopsies. Therefore, the cutoff point should be chosen as to include a minimal number of cancer patients but a maximum number of BPH patients above this point.
The lower cutoff point identifies patients with prostate cancer. Below this point patients are considered to have cancer. As many cancer patients and as few BPH patients as possible should have % fPSA values below this cutoff point. Presently, approximately 23 percent and 7 percent for the upper and lower cutoff points are recommended by most investigators for the identification of prostate cancer and BPH patients.
Determination of % fPSA requires both fPSA and PSA assays. The value of both assays impacts the % fPSA determined. Because the concentration of serum fPSA is usually much lower than that of total serum PSA, an accurate determination of fPSA is critical to obtain an accurate % fPSA value. On the other hand, the value of tPSA and fPSA produced by various commercial kits on the same specimens are still not exactly identical with each other. It's important to specify the kits used when reporting the % fPSA value.
Our recent findings indicated that expressing % fPSA in terms of the PSA-ACT complex, such as [(fPSA/PSA-ACT) x 100], provides for better differentiation between BPH and CAP and eliminates most problems associated with the tPSA assays.17
* Reducing biopsies for those with tPSA below 4ng/mL. It was reported recently that approximately 20 percent of all patients with clinically significant organ-confined cancer have a normal tPSA value of 4ng/mL or less and 30 percent of those people with tPSA between 4ng/mL and 10ng/mL have micrometastasis.14-16 Therefore, to detect more organ-confined tumors at an early stage before they have the chance to become metastasized, the tPSA cutoff was extended to 2.5ng/mL, or to 3.0ng/mL,18 below the upper normal limit of 4ng/mL.16 Consequently, more unnecessary biopsies were carried out since there are more people with benign prostate disease with tPSA within the normal concentration range.
To reduce unnecessary biopsies, the % fPSA was applied to determine who should be spared from such a procedure. Using a % fPSA cutoff value of 27 percent, Catalona et al16 was able to detect 90 percent of cancers and avoid 18 percent of benign biopsies. In a prospective study, using 25 percent as a cutoff value for % fPSA, Catalona et al. Detected 91 percent of the cancers in "screened volunteers" whose tPSA concentrations fell between 2.6ng/mL to 4.0ng/mL, and 26 percent of the negative biopsies were eliminated.
* Identification of tumor with normal tPSA. Traditionally, 4ng/mL is the accepted upper normal cutoff for detecting prostate cancer. However, we now know that when the serum tPSA was below 4ng/mL it did not mean that the person was free of prostate tumor. Studies by Catalona et al15,16 and Vashi et al18 clearly indicated that about 20 percent to 22 percent of people with tPSA between 2.5ng/mL or 3.0ng/mL to 4ng/mL may have prostate tumor. Both groups concluded from their studies that % fPSA is capable of improving cancer detection in this tPSA range.
Percent fPSA of 19 percent and 23 percent has been used by Vashi et al and Catalona et al,16,18 respectively, to detect prostate cancer for people having tPSA in the normal range. Therefore, it is very likely that the determination of % fPSA will enhance cancer detection even when the tPSA is in the normal range. This also implies that lower % fPSA is frequently associated with prostate neoplastic disease.
On the other hand, our study provided direct evidence suggesting that almost 50 percent of our patients with biopsy-identified cancer were associated with lower than 10 percent of % fPSA.17 We obtained serial specimens from individual cancer patients and measured both % fPSA and tPSA. Many cancer patients with initial normal tPSA had low % fPSA. The tPSA eventually rose to elevated levels because of cancer progression but the % fPSA consistently remained at low values. In other words, it appears from our results that % fPSA is not only useful for identifying cancer when the tPSA is in the normal range, but it can also detect tumors at a much earlier stage (Fig. 2).
* As a prognostic marker. It was pointed out by Catalona et al14 that % fPSA could be treated as a prognostic marker for prostate cancer. In their study, patients with low % fPSA (< 4 percent) were 100 percent associated with prostate cancer, whereas patients with % fPSA greater than 25 percent only had a 10 percent chance of having a malignant prostate disease. Conceivably, low and high % fPSA are associated with a poor and good prognosis, respectively. Carter et al also suggested from their study that % fPSA is helpful in assessing the aggressiveness of the tumor.19 It's apparent that measuring fPSA is clinically more important than a simple measurement of tPSA.
* Differentiation of fPSA of prostate cancer from fPSA of BPH. Determination of % fPSA undoubtedly has improved the specificity of prostate cancer screening and spared almost 1/3 of all BPH patients from unnecessary biopsies. However, it would be most desirable if we could find a way to differentiate the fPSA of prostate cancer from fPSA of benign prostate tissue so that the overlap in % fPSA between prostate cancer and BPH could be further reduced and additional negative biopsies for BPH patients could be avoided. Capability of identifying tumor-specific fPSA will also make it possible to develop a specific diagnostic test for prostate cancer and improve the specificity of prostate cancer screening.
Our investigation of fPSA isolated from seminal plasma and LNCaP cells (a metastatic prostate tumor cell line) suggests that fPSA from benign and malignant sources of prostate tissue could differ. As shown in Fig. 3, different polyacrylamide electrophoretic patterns (PAGE pattern) were found for fPSA from LNCaP and seminal plasma. The LNCaP cell is a malignant prostate cancer cell line representing the malignant source of fPSA, whereas the seminal plasma represents the benign prostate tissue.
Fig. 3 shows that fPSA from LNCaP cells contains more inactive (in terms of protease activity) forms of fPSA that have slower electrophoretic mobility (the top two bands).20 Conceivably, if we can identify and remove bands of fPSA synthesized by benign prostate tissue (transition zone) from the total fPSA concentration of cancer serum before calculating the % fPSA, the % fPSA in cancer patients would be dramatically decreased and could be well differentiated from % fPSA derived from BPH. Alternatively, if we could identify the inactive fPSA isoform in patients' sera, we could have a specific assay for prostate cancer.
These different fPSA patterns can also be demonstrated on the fPSA elution profile of chromatofocusing chromatography. However, both PAGE and isoelectrofocusing require microgram concentrations of fPSA for detection that are impossible to obtain from patients' sera. Research is under way to determine whether differences in serum fPSA between cancer and serum BPH patients can be detected based on the chromatofocusing technique using an ultrasensitive fPSA assay (which could detect less than 0.1ng of fPSA) recently established.
Dr. Wu is professor of Pathology, University of Utah Health Science Center, and medical director of the Special Chemistry and Reagent Development Laboratories, ARUP, Salt Lake City.
1. Wang MC, Valenzuela LA, Murphy GP, et al. Purification of a human prostate specific antigen. Invest Urol 17:159, 1979.
2. Watt KWK, Lee PJ, M'Timkulu T, et al. Human prostate-specific antigen: Structural and functional similarity with serine proteases. Proc Natl Acad Sci USA 83;3166-3170, 1986.
3. Yu H, Diamandis EP, Levesque M, et al. Expression of the prostate specific antigen gene by a primary ovarian carcinoma. Cancer Res 1995;55:1603-1606.
4. Wang MC, Papsidero LD, Kuriyama M, et al. Prostate antigen: A new potential marker for prostatic cancer. Prostate 2:89, 1981.
5. Wang MC, Valenzuela LA, Murphy GP, et al. A simplified purification procedure for human prostate antigen. Oncology 39:1, 1982.
6. Christensson A, Laurell C, Lilja H. Enzymatic activity of prostate specific antigen and its reactions with extracellular serine proteinase inhibitors. Eur J Biochem 194:755-763, 1990.
7. Stenman UH, Leinonen J, Alfthan H, et al. A complex between prostate specific antigen and alpha 1-antichymotrypsin is the major form of prostate-specific antigen in serum of patients with prostatic cancer: Assay of the complex improves clinical sensitivity of cancer. Cancer Res 51:222-226, 1991.
8. Lilja H, Cockett ATK, Abrahamsson PA. Prostate specific antigen predominantly forms a complex with alpha 1-antichymotrypsin in blood. Cancer Supplement 70:230-234,1992.
9. Wu JT, Stephenson RA. Microplate assays for free PSA, PSA-ACT and total PSA: Assay characteristics, selection of calibrators and in vitro stability studies. Proceeding of the 1st International Consultation on Prostate Cancer. 159-168, 1997.
10. Catalona WJ, Colberg JW, Smith DS, et al. Measurement of percent-free PSA improves specificity for lower PSA cutoffs in prostate cancer screening. J Urol 155 (Suppl):422A, 1996.
11. Wu JT, Liu GH, Zhang P. Effect of denaturization and curve fitting on the accuracy of microplate assay for free PSA. In press.
12. Catalona W, Smith D, Ratliff T, et al. Measurement of prostate-specific antigen in serum as a screening test for prostate cancer. N Engl J Med 324:1156-1161, 1991.
13. Oesterling JE. Prostate specific antigen: A critical assessment of the most useful tumor marker for adenocarcinoma of the prostate. J Urol 145:907-923, 1991.
14. Catalona WJ, Smith DS, Wolfert RL, et al. Evaluation of percentage of free serum prostate-specific antigen to improve specificity of prostate cancer screening. JAMA 274:121-122, 1995.
15. Catalona WJ, Colberg JW, Smith DS, et al. Measurement of percent-free PSA improves specificity for lower PSA cutoffs in prostate cancer screening. J Urol 155(Suppl):422A, 1996.
16. Catalona WJ, Smith DS, Ornstein DK. Prostate cancer detection in men with serum PSA concentrations of 2.6 to 4.0 ng/mL and benign prostate examination.
17. Wu JT, Liu GH, Zhang P, et al. Monitoring percent free PSA in serial specimens: Improvement of test specificity, early detection and identification of occult tumors. J Clin Lab Anal 12:26-31, 1998.
18. Vashi AR, Oesterling JE. Percent free prostate-specific antigen: Entering a new era in the detection of prostate cancer. Mayo Clin Proc 72:337-344, 1997.
19. Carter HB, Partin AW, Luderer AA, et al. Percentage of free prostate-specific antigen in sera predicts aggressiveness of prostate cancer a decade before diagnosis. Urology 1997;49:379-84.
20. Wu JT, Lyons BW, Liu G, et al. Production of milligram concentrations of free prostate specific antigen from LNCaP cell culture: Difference between fPSA from LNCaP cell and seminal plasma. J Clin Lab Anal 12:6-13, 1998.