Home About us Current issue Ahead of Print Back issues Submission Instructions Advertise Contact Login   

Search Article 
Advanced search 
Saudi Journal of Kidney Diseases and Transplantation
Users online: 1288 Home Bookmark this page Print this page Email this page Small font sizeDefault font size Increase font size 

Table of Contents   
Year : 2011  |  Volume : 22  |  Issue : 2  |  Page : 291-297
Utility of free prostate specific antigen serum level and its related parameters in the diagnosis of prostate cancer

1 Department of Radiology, Jordan University Hospital, Amman, Jordan
2 Department of Urology, Jordan University Hospital, Amman, Jordan
3 Department of Histopathology, Jordan University Hospital, Amman, Jordan

Click here for correspondence address and email

Date of Web Publication18-Mar-2011


We evaluated the role of free prostate specific antigen (f-PSA) serum level and its related parameters in detecting prostate cancer. This retrospective study was conducted between January 2006 and March 2008. Trans-rectal ultrasound guided prostate biopsy was performed for 107 patients who had total PSA (t-PSA) level of either >4 ng/mL with or without palpable nodule or ≤4 ng/mL with palpable nodule on digital rectal examination. The perfor-mance measurements for f-PSA, percent free PSA (%f-PSA) and free PSA density (f-PSAD) were determined and compared with those for t-PSA and total PSA density (t-PSAD). Descriptive statistics for all variables of interest were calculated, and receiver operating characteristic curves were generated. Nine patients (8.4%) had normal histology, 69 patients (64.4%) had benign disease and 29 patients (27.1%) had prostate cancer. The performance of f-PSA in PCa detection was better than other evaluated parameters. The largest area under the curve for patients in the gray area (t-PSA range 4.1-10 ng/mL) was for f-PSA, with a value of 0.64 and a sensitivity and specificity of 44% and 87%, respectively. For %f-PSA, these values were 0.59, 63% and 62%, respectively. For patients with a t-PSA level of 10.1-20 ng/mL, they were 0.68, 67%, and 81%, respectively, for f­PSA, and 0.64, 67%, and 76%, respectively, for %f-PSA. In conclusion, f-PSA serum levels performed better than free to total PSA ratio and t-PSA for prostate cancer screening. It is of clinical value which could affect the biopsy decision avoiding unnecessary interventions.

How to cite this article:
Haroun AA, Hadidy AS, Awwad ZM, Nimri CF, Mahafza WS, Tarawneh ES. Utility of free prostate specific antigen serum level and its related parameters in the diagnosis of prostate cancer. Saudi J Kidney Dis Transpl 2011;22:291-7

How to cite this URL:
Haroun AA, Hadidy AS, Awwad ZM, Nimri CF, Mahafza WS, Tarawneh ES. Utility of free prostate specific antigen serum level and its related parameters in the diagnosis of prostate cancer. Saudi J Kidney Dis Transpl [serial online] 2011 [cited 2021 Oct 18];22:291-7. Available from: https://www.sjkdt.org/text.asp?2011/22/2/291/77606

   Introduction Top

Early diagnosis of prostate cancer may lead to an increase in the number of patients who can be managed in time. Although prostate speci­fic antigen (PSA) is a specific marker for the prostate epithelium, it is not a prostate cancer specific as it can be found elevated in benign conditions that may affect the gland as well as in malignant lesions. [1] In addition, some authors found a correlation between a patient's age and PSA levels. [2]

PSA is now the commonest initial assessment method used in the early detection and follow-up of prostate cancer. However, its ability to distinguish between benign and malignant le­sions is particularly poor in the intermediate range of 4.1 and 10 ng/mL. There are several studies in the literature based on PSA, which have been done in an attempt to improve the efficiency of prostate cancer detection and to discriminate benign from malignant lesions. These studies used different PSA parameters such as total prostate specific antigen density (t-PSAD), [3] PSA velocity, [4] percent free pros­tate specific antigen (%f-PSA), [5],[6],[7] excess PSA, [8] and PSA transition zone density. [3],[5],[6],[7],[8] In the last few years, there have been several studies that used complexed PSA as another prostate can­cer predictor and the reported results were va­riable. [9],[10],[11] The purpose of all these studies was to reduce the number of patients requiring biopsy, particularly those in the gray area. However, it is still unclear which parameter is most useful in clinical practice.

The aim of our study was to enhance the sen­sitivity and specificity of prostate cancer de­tection, particularly in patients with t-PSA le­vel range of 4.1-20 ng/mL. For this purpose, we evaluated the utility of f-PSA serum level and its related parameters such as %f-PSA and free prostate specific antigen density (f-PSAD) and we compared them with total t-PSA and t-PSAD.

   Methods Top

This retrospective study was conducted bet­ween January 2006 and March 2008. It com­prised 107 patients who had a t-PSA level above 4 ng/mL with or without a palpable no­dule on digital rectal examination (DRE) or normal t-PSA level with palpable nodule. Sub­jects ranged in age from 50 to 85 years (mean 67 years ± 8). These patients were referred to our radiology department for trans-rectal ultra­sound (TRUS) guided prostate biopsy. Blood samples were taken at least 3 weeks before TRUS biopsy, stored at 4°C, and examined du­ring the first 24 hours. The t-PSA and f-PSA values were determined using the ARCHITECT total PSA and ARCHITECT free PSA assays (Abbott Ireland).

TRUS was performed on a prosound 5500 alloca machine using 7.5 MHz endocavitary probe. The prostate volume was calculated ac­cording to the prolate ellipsoid formula (0.52 × length × width × height).

Systemic sextant biopsy was performed in all patients in addition to two biopsies from the transitional zones. In addition, a targeted biopsy from any hypoechoic suspicious area in peri­pheral zones was also performed. According to t-PSA level, the patients were classified into four groups: group I with t-PSA value of ≤4 ng/mL; group II between 4.1 and 10 ng/mL; group III between 10.1 and 20 ng/mL; and group IV >20 ng/mL. The performance mea­surements for f-PSA, %f-PSA and f-PSAD were compared with those for t-PSA and t­PSAD. T-PSAD and f-PSAD were calculated by dividing the t-PSA, f-PSA levels by the transrectal ultrasound-estimated prostate vo­lumes (length × width × height × 0.5236).

   Statistical Analysis Top

Descriptive statistics were calculated for all variables of interest. The correlation between different PSA parameters was assessed using Pearson correlation coefficients. An analysis of variance (ANOVA) was carried out to assess differences in mean levels of the parameters between categories (malignant, benign and nor­mal subjects). A two-sample two-sided Student's t-test was utilized to assess differences bet­ween malignant and nonmalignant conditions.

An ANOVA was carried out to assess diffe­rences in mean levels of the parameters bet­ween the groups I-IV.

The diagnostic test characteristics were eva­luated using sensitivity, specificity and asso­ciated cut-points. Receiver operating charac­teristic (ROC) curves and their associated areas under the curve (AUCs) were also compared between tests using Mann-Whitney statistics.

Group I (≤4 ng/mL) yielded only one posi­tive and group IV (>20 ng/mL) yielded three positive measures, which were too few to cons­truct a meaningful analysis.

All analyses were carried out using SAS Version 9.1 (SAS Institute, Cary, NC, USA), and SPSS version 15.

   Results Top

According to histopathology results, there were nine patients (8.4%) with no abnormal findings, 69 patients (64.48%) had benign di­seases, and 29 patients (27.1%) had prostate gland malignancy.

[Table 1] shows the distribution of evaluated parameters according to histopathology results.
Table 1: Distribution of prostate specific antigen (PSA) (ng/mL) parameters in different groups according to histology results.

Click here to view

A positive correlation was found between t-PSA, and f-PSA, %f-PSA, t-PSAD and f-PSAD (r = 0.6, 0.07, 0.5, and 0.47, respectively). Analysis of variance indicated significant dif­ference (P ≤ 0.0001) in the mean values bet­ween groups on all measures except %f-PSA (P = 0.61) with group I being significantly lower than groups III and IV (but not II), group II being significantly lower than groups III or IV, and group III being lower than group IV. Although the means of t-PSA, f-PSA, and %f-PSA were higher in normal than in benign and malignant prostate diseases [Table 1], this was not statistically significant (P = 0.72, 0.56, 0.4, respectively). No significant difference across these categories in the total sample was also found for t-PSAD and f-PSAD (P = 0.49 and 0.81, respectively). Student's t-test did not also show a significant difference between pa­tients labeled as malignant or non-malignant (P = 0.8, 0.3, 0.2, 0.45, and 0.84 for t-PSA, f-PSA, %f-PSA, t-PSAD, and f-PSAD, respec­tively).

In group II patients, the specificity le-vels at a sensitivity of 90% for t-PSA, f-PSA, %f­PSA, f-PSAD and t-PSAD were 19, 21, 21, 6 and 6 at cut-off values of 9, 1.9, 0.27, 0.06, and 0.31, respectively. For group III patients, these values were 10, 0, 10, and 5 at cut-off values of 19.2, 5, 0.38 and 0.69, respectively.

Using Mann-Whitney statistics, the only pair found to differ significantly in AUC was f-PSA serum level and f-PSAD (P = 0.04), with f-PSA having a statistically larger AUC than f-PSAD.

   Discussion Top

PSA occurs in three major forms in blood; two of these are immunodetectable and the third form is not detectable with current immu­noassays for PSA. The major immunodetec­table form is PSA complexed with alpha-1­antichymotrypsin (PSA-ACT). Uncomplexed or free PSA is the other immunodetectable form. Total PSA is a term referring to the immu­nologically detectable forms. [12]

About 10-30% of t-PSA is not bound to pro­teins and is called f-PSA [13] its proportion is higher in normal subjects than in those with prostate cancer (PCa). [9] Collins et al [14] sugges­ted that f-PSA levels are largely dependent on the transitional zone which is contradictory to our observation. We propose that f-PSA is largely dependent on peripheral zone and our theory gains support from the following:

  1. The mean of f-PSA in our study group was higher in normal than in benign and malignant prostate diseases. This can be explained by the fact that the peripheral prostate is compressed by the enlarged transitional zone, so the normal epithelial cells in the peripheral zone will be­come under pressure, and in case of malig­nancy, these cells are destroyed, and the pro­duction of f-PSA will be reduced.
  2. In the literature, it is well established that %f-PSA is lower in malignant than in benign prostate conditions, [5],[13],[15] and this was also ob­served in our results.

The reason due to which %f-PSA is lower in malignant than in benign prostate conditions is still not well understood. Some authors attri­buted it to an increased amount of synthesized ACT by malignant cells. [16] This means that the complexed portion of t-PSA is also largely de­pendent on the peripheral zone. This theory cannot explain the elevation of complexed PSA serum level that was also observed in be­nign conditions in several reported studies. [5],[9],[17] However, we can conclude from these obser­vations that both free and complexed portions of PSA are largely dependent on the peripheral zone. The diseased epithelial cells (either be­nign or malignant) in the peripheral zone pro­duce less amount of f-PSA and more amount of complexed PSA, resulting in elevated t-PSA levels and reduction in f-PSA levels in both malignant and benign prostate conditions. We think that this may be the reason for the dis­crepancy in the reported results regarding the different related PSA parameters and their corresponding thresholds.

After reviewing the literature, we found that the role of f-PSA serum level in detecting PCa has not been evaluated, and the majority of studies were based on %f-PSA and other com­plicated parameters. The published threshold values for %f-PSA were widely variable, ran­ging from 0.10 to >0.25, [14],[18],[19] and there is still no consensus about a definite cut-off value which renders its clinical utility poor.

The performance of f-PSA serum level and then %f-PSA in PCa detection was better than the other evaluated parameters in patients of groups II and III. The statistical results demons­trated that the largest AUC for patients in the gray area was for f-PSA with a value of 0.64 (P = 0.04) and with a maximum combined sensitivity and specificity of 44% and 87%, respectively, at a cut-off point of 0.8 ng/mL.

The least performance indicator of the inves­tigated parameters in the present study was for f-PSAD with an AUC of 0.5. In comparing f­PSA and %f-PSA with t-PSA and t-PSAD on ROC curve, we found that they ran signifi­cantly above the t-PSA parameters [Figure 1]. Our results regarding the %f-PSA are in agree­ment with those of many other studies. [7],[14],[20] Some authors have questioned the clinical uti­lity of the %f-PSA [5],[21] as its performance was lower than other parameters such as t-PSAD, [20] PSA adjusted for the transition zone volume, [22] and complexed PSA. [10] However, lowering the cut-off points for biopsies increases the sen­sitivity but the specificity decreases which re­sults in an increase in the number of unnece­ssary biopsies. At 90% sensitivity which is usually ideal, all the investigated parameters in our study had a low specificity which is in agreement with other authors. [3],[9],[10]
Figure 1: ROC curves of t-PSA, f-PSA, %f-PSA, f-PSAD, and t-PSAD showing a larger area under the curve for f-PSA than the other parameters in patients of groups II (a) and III (b). The areas under the curves for f-PSA and %f-PSA are almost equal in all groups when combined together (107 patients) (c).

Click here to view

The threshold value was defined as a level at which high sensitivity (90%) is existent in com­bination with an acceptable specificity (approx. 50%), [5] which, to our knowledge, was not ob­tained for any reported PSA related parameter. The selected threshold value should not result in increasing the number of men falsely la­beled as malignant. From a practical point of view, we think of a reasonable compromise between both sensitivity and specificity at cer­tain cut-off points in order to reduce the unne­cessary biopsies. To date, the most accurate diagnostic tool for PCa detection is TRUS prostate biopsy, but with a sensitivity level of 67% and a cut-off value of 1.7 ng/mL for f­PSA, it would give the lowest number of un­necessary biopsies (81% specificity) in patients with group III (t-PSA range 10.1-20 ng/mL). The %f-PSA for the same patient group and with the same sensitivity and at a cut-off point of 0.11 yielded a specificity of 76%. There­fore, we recommend TRUS biopsy for patients in the gray area with an f-PSA value of 0.8 ng/ mL and for patients in group III with an f-PSA value of 1.7 ng/mL. The additional use of %f­PSA with cut-off values of 0.15 and 0.11 in patients of groups II and III, respectively, could be beneficial in selecting patients for prostate biopsy, and these values coincide with many reported recommendations. [7],[19],[22] The clinical use of f-PSA may improve the cost effectiveness of detecting PCa. It is evident that measuring different parameters usually requires more than one immuno-assay method and this is cost effective. According to World Health organi­zation reports, an estimated 513,000 cases and 255,000 deaths were attributed to prostate can­cer in 1999. [23] These statistics provide adequate rationale for the continued investigation into new and more effective diagnostic and manage­ment modalities. The limitation of the present study is the very limited number of patients having t-PSA levels lesser than 4 ng/mL and greater than 20 ng/mL, so that this could not be correctly evaluated statistically. They largely affected the different measurements when all patient groups were analyzed together.

The limitation of use of f-PSA is its stability in serum samples under a variety of sample collection and storage conditions. However, proper storage and manipulation according to the recommendations published by Woodrum et al [12] can overcome this limitation.

In conclusion, the most important finding in our study is the demonstration of the influence of PCa on f-PSA serum level. The results of this study enhance the utility of f-PSA in de­tecting PCa. It is superior to the other inves­tigated parameters, particularly the volume re­lated PSA. We believe that f-PSA serum level is a good PCa predictor and could affect the decision of biopsy, mainly in patients having t­PSA ranging from 4.1 to 20 ng/mL.

   References Top

1.Gümüş BH, Neşe N, Gündüz MI, Kandiloğlu AR, Ceylan Y, Büyuksü C. Does asympto-matic inflammation increase PSA? A histo-pathological study comparing benign and ma-lignant tissue biopsy specimens. Int Urol Nephrol 2004;36:549-53.  Back to cited text no. 1
2.Kamal B, Ali G, Taha S. Prostate specific antigen reference ranges in Saudi men. Saudi Med J 2003;24:665-8.  Back to cited text no. 2
3.Benson M, Whang I, Olsson C, McMahon D, Cooner W. The use of prostate specific antigen density to enhance the predictive value of intermediate levels of serum prostate specific antigen. J Urol 1992;147:817-21.  Back to cited text no. 3
4.Ito K, Yamamoto T, Ohi M, et al. Usefulness of prostate-specific antigen velocity in scree-ning for prostate cancer. Int J Urol 2002;9:316-21.  Back to cited text no. 4
5.Weckermann D, Maassen C, Wawroschek F, Harzmann R. Improved discrimination of pros-tate cancer and benign prostatic hyperplasia by means of the quotient of free and total PSA. Int Urol Nephrol 1999;31(3):353-9.  Back to cited text no. 5
6.Edrem E, Astü N, Akbal C, Bilen C, Ergen A, Özen H. The free-to-total serum prostatic specific antigen ratio as a predictor of the pathological features of prostate cancer. Int Urol Nephrol 2002;34:519-23.  Back to cited text no. 6
7.Yeniyol C, Bozkaya G, Çavuşoglu A, Arsln M, Karaca B, Ayder A. The relation of prostate biopsy results and ratio of free to total PSA in patients with a total PSA between 4-20 ng/mL. Int Urol Nephrol 2001;33(3):503-6.  Back to cited text no. 7
8.Kuligowska E, Barish M, Fenlon H, Blake M. Predictors of prostate carcinoma: Accuracy of grey-scale and color Doppler US and serum markers. Radiology 2001;220:757-64.  Back to cited text no. 8
9.Djavan B, Remzi M, Zlotta A, et al. Com-plexed prostate-specific antigen, complexed prostate-specific antigen density of total and transition zone, complexed/total prostate-spe-cific antigen ratio, free-to-total prostate-spe-cific antigen ratio, density of total and tran-sition zone prostate-specific antigen: results of the prospective multicenter European trial. Urology 2002;60:4-9  Back to cited text no. 9
10.Horninger W, Cheli C, Babaian R, et al. Com-plexed prostate-spcific antigen for early de-tection of prostate cancer in men with serum prostate-specific antigen levels of 2 to 4 nano-grams per milliliter. Urology 2002;60: 31-5.  Back to cited text no. 10
11.Brawer M, Bartsch G, Epstein J, Glodé L, Sakr W. Assays for complexed prostate-specific antigen and other advances in the diagnosis of prostate cancer. Rev Urol 2003;5:S10-6.  Back to cited text no. 11
12.Woodrum D, French C, Shamel B. Stability of free prostate- specific antigen in serum sam-ples under a variety of sample collection and sample storage conditions. Urology 1996;48 (6A):33-9.  Back to cited text no. 12
13.Stephan C, Lein M, Jung K, Schnorr D, Loening S. The influence of prostate volume on the ratio of free to total prostate specific antigen in serum of patients with prostate carcinoma and benign prostate hyperplasia. Cancer 2000;79: 104-9.  Back to cited text no. 13
14.Collins GN, Alexandrou K, Wynn-Davies A, Mobley S, O′reilly PH. Free prostate-specific antigen "in the field": A useful adjunct to standard clinical practice. BJU Int 1999;83: 1000-2.  Back to cited text no. 14
15.Jung K, Elgeti U, Lein M, et al. Ratio of free or complexed prostate-specific antigen (PSA) to total PSA: Which ratio improves differen-tiation between benign prostatic hyperplasia and prostate cancer? Clin Chem 2000;46:55--62.  Back to cited text no. 15
16.Björk T, Hulkko S, Bjartell A, di Snt′Agnese A, Abrahamsson PA, Lilja H. Production of alpha-1-antichemotrypsin by PSA-producing cells of the human prostate epithelium. Urology 1994;43:427-43.  Back to cited text no. 16
17.Okegawa T, Kinjo M, Watanabe K, Noda H, Miyata A, Murata A. The significance of free-to-complexed prostate-pecific antigen (PSA) ratio in prostate cancer detection in patients with a PSA level of 4.1-10.0 ng/mL. BJU Int 2000;85:708-14.  Back to cited text no. 17
18.Toubert ME, Guillet J, Chiron M, et al. Percentage of free serum prostate-specific antigen: a new tool in the early diagnosis of prostate cancer. Eur J Cancer 1996;32A(12): 2088-93.  Back to cited text no. 18
19.Luboldt HJ, Swoboda A, Borgermann C, Fornara P, Rubben H. Clinical usefulness of free PSA in early detection of prostate cancer. Onkologie 2001;24(1):33-7.  Back to cited text no. 19
20.Stephan C, Stroebel G, Heinau M, et al. The ratio of prostate-specific antigen (PSA) to prostate volume (PSA density) as a parameter to improve the detection of prostate carcinoma in PSA values in the range of <4 ng/mL. Cancer 2005;104:993-1003.  Back to cited text no. 20
21.Okegawa T, Noda H, Nutahara K, Higashihara E. Comparison of the various combinations of free, complexed, and total prostate-specific antigen for the detection of prostate cancer. Eur Urol 2000;38(4):380-7.  Back to cited text no. 21
22.Kang S, Bae J, Park H, Yoong D, Moon D, Kim J, Cheon J. Prostate-specific antigen adjusted for the transition zone volume as a second screening test: a prospective study of 248 cases. Int J Urol 2006;13:910-4.  Back to cited text no. 22
23.World Health Organization, authors. The World Health Report. 1999. Available at http://www.who.org/whr .  Back to cited text no. 23

Correspondence Address:
Azmi A Haroun
P. O. Box 460495, 11946 Amman
Login to access the Email id

PMID: 21422628

Rights and Permissions


  [Figure 1]

  [Table 1]

This article has been cited by
1 Laparoscopy-assisted orchiopexy versus laparoscopic two-stage fowler stephens orchiopexy for nonpalpable testes: Comparative study
Alzahem, A.
Urology Annals. 2013; 5(2): 110-114


    Similar in PUBMED
    Search Pubmed for
    Search in Google Scholar for
    Email Alert *
    Add to My List *
* Registration required (free)  

    Statistical Analysis
    Article Figures
    Article Tables

 Article Access Statistics
    PDF Downloaded706    
    Comments [Add]    
    Cited by others 1    

Recommend this journal