104

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
Carsten Stephan, M.D.
Michael Lein, M.D.
Klaus Jung, M.D.
Dietmar Schnorr, M.D.
Stefan A. Loening, M.D.

BACKGROUND. Determining the ratio of free to total prostate specific antigen (fPSA to t-PSA, calculated as the percentage of f-PSA [f-PSA%]) in serum allows for
a clearer distinction between patients with prostate carcinoma (PCa) and patients
with benign prostate hyperplasia (BPH) than determining the level of t-PSA alone.
To find influencing factors on f-PSA%, the authors investigated prostate volume,
TNM classification, and tumor stage.

Department of Urology, University Hospital

Charite, Humboldt University Berlin, Berlin, Germany.

METHODS. The authors measured f-PSA and t-PSA in 36 men with untreated PCa
(tumor classification: T1, 2, 3pN0, M0), 44 patients with BPH, and 54 healthy
controls. Prostate volume was determined by transrectal ultrasound.
RESULTS. The median values of t-PSA and f-PSA% were 7.8 mg/L and 10.5% in PCa
patients, 4.3 mg/L and 20.8% in patients with BPH, and 1.4 mg/L and 23.6% in the
control group. Patients with PCa had a significantly lower proportion of f-PSA than
BPH patients and healthy men. There was no correlation of f-PSA% to TNM stage
or tumor grade. In PCa patients a significant positive correlation (correlation coefficient [r] 0.51, P 0.001) was found between f-PSA% and prostate volume,
whereas there was no significant correlation in BPH patients (r 00.27, P 0.05).
There was a significant difference in f-PSA% between PCa and BPH patients with
prostate volumes smaller than 40 cm3 (9.0% vs. 21.6%, P 0.01) but not between
patients in these 2 groups with prostate volumes exceeding 40 cm3 (15.1% vs.
18.2%, P 0.11).
CONCLUSIONS. Determining the ratio of f-PSA to t-PSA to discriminate between
PCa and BPH patients yields significant results only in men with a prostate volume
of less than 40 cm3. Cancer 1997; 79:1049. q 1997 American Cancer Society.

Supported in part by a grant from the Foundation of the Klee family (Michael Lein) and from
the Fonds der Chemischen Industrie (Klaus
Jung, project no. 400700).
The authors thank Silke Klotzek for her valuable
technical assistance.
This work includes parts of the doctoral thesis
of Carsten Stephan.
Address for reprints: Klaus Jung, Department of
Urology, University Hospital Charite, Humboldt
University Berlin, Schumannstrae 20/21,
10098 Berlin, Germany.
Received May 1, 1996; revision received August
15, 1996; accepted August 30, 1996.

KEYWORDS: prostate carcinoma, benign prostate hyperplasia, prostate volume,

prostate specific antigen, ratio of free to total prostate specific antigen, diagnostic
validity.

erum prostate specific antigen (PSA) exists in different molecular

forms.1 5 PSA is predominantly bound to a1-antichymotrypsin
(PSA-ACT) with approximately 70 90% of total PSA (t-PSA). Furthermore, serum PSA also binds in trace amounts to a1-antitrypsin and
a2-macroglobulin, whereas a2-macroglobulin encapsules all epitopes
of PSA.6 Approximately 10 30% of t-PSA is not bound to proteins and
is called free PSA (f-PSA). Recent publications have reported on the
use of molecular forms to distinguish between patients with prostate
carcinoma (PCa) and patients with benign prostate hyperplasia
(BPH).2,4,5 PCa patients have a lower ratio of f-PSA/t-PSA (f-PSA%)
than BPH patients. Therefore, the determination of f-PSA and calcula-

q 1997 American Cancer Society

/ 7b48$$0817

12-10-96 11:23:14

cana

W: Cancer

Free/Total PSA Ratio and Prostate Volume/Stephan et al.

tion of the f-PSA/t-PSA ratio have been suggested as

a useful diagnostic tool to differentiate between PCa
and BPH patients.2,4,5,7 10
Reasons for lower f-PSA% values in patients with
PCa compared with BPH patients have not been clarified until now. It was postulated that PCa cells, in
contrast to prostate cells of BPH, synthesize increased
amounts of a1-antichymotrypsin to form PSA-ACT
complexes early in the PCa cells.11,12 There are only a
few reports of other variables influencing f-PSA%.10,13
In the current study, prostate volume, tumor TNM
stage, and grade were analyzed as factors influencing
t-PSA and f-PSA%.

MATERIAL AND METHODS

Patients
The study included 134 men with total PSA values to
23 mg/L. The control group was comprised of 54
healthy men (median age, 54.3 years). Thirty-six patients had an untreated PCa (median age, 66.1 years).
Stage was assigned to each patient according to the
TNM system14 and ultimately confirmed pathologically (TNM stage: T1-3pN0M0; Grade G1-3). Forty-four
patients had BPH (median age, 71 years). Thirty-two
BPH patients underwent a transurethral resection and
there was histologic proof whereas in 12 patients the
diagnosis was made clinically. However, because there
were no differences between the two groups of BPH
patients, they were considered as one group.
Blood samples were taken before diagnostic procedures or 4 weeks (at the earliest) after digital examination, prostatic biopsy, and transrectal ultrasound.
No patients received any hormonal therapy before
blood sampling. Blood sampling in BPH patients was
made before transurethral resection. Blood samples
were centrifuged at 16001 relative centrifugal force (g)
for 10 minutes. The sera were frozen at 070 7C within
2 hours after collection and were not thawed (and
refrozen) before testing.
Methods
The size of the prostate was determined in all PCa and
BPH patients by transrectal ultrasonography (COMBISON 330; Fa. Kretz Technik, Zipf, Austria).
t-PSA and f-PSA measurements in serum were
performed with the IMMULITE PSA and Free PSA kits
(Diagnostic Products Corp., Los Angeles, CA; Cat. No.
LKPS1 and LKPF1). The assays were solid-phase, twosite sequential chemiluminescent immunometric tests
and were automatically handled on the IMMULITE
Automated Analyzer. The solid phase was a polystyrene bead coated with a polyclonal antibody specific
for PSA or with a monoclonal anti-PSA antibody specific only for f-PSA. For determining t-PSA, samples

/ 7b48$$0817

12-10-96 11:23:14

cana

105

TABLE 1
Total PSA, Free PSA, and Free PSA% in Healthy Persons, Patients
with Benign Prostate Hyperplasia, and Patients
with Prostate Carcinoma

Measurement

Controls
(n 54)

Benign prostate
hyperplasia
(n 44)

Prostate
carcinoma
(n 36)

Total PSA (mg/L)a

Free PSA (mg/L)a
Free PSA%a

1.4 (0.493.15)
0.29 (0.170.53)
23.6 (1345.8)

4.3b (1.313.6)
0.81b (0.381.85)
20.8 (11.438.3)

7.8b,c (3.9516.4)
0.8b (0.411.7)
10.5b,c (6.817.6)

PSA: prostate specific antigen; PSA%: percentage of free prostate specific antigen.
a
Values are given as median (and 1090 percentiles).
b
Significantly different from controls by at least P 0.01 (MannWhitney U test).
c
Significantly different from patients with benign prostate hyperplasia by at least P 0.01 (Mann
Whitney U test).

were incubated together with alkaline phosphataseconjugated monoclonal anti-PSA antibody and formed
a PSA-antibody sandwich complex. After removing the
unbound conjugate by centrifugal wash, the chemiluminescent substrate, a phosphate ester of adamantyl
dioxetane, was added. In the presence of alkaline
phosphatase, that substrate was cleaved into unstable
intermediate emitting photons that were measured by
the luminometer of the analyzer. For determining fPSA, f-PSA was bound to the polystyrene bead during
the first incubation period and an alkaline phosphatase-labeled polyclonal goat anti-PSA antibody was
bound to that formed complex of f-PSA/anti-f-PSA antibody in a second incubation cycle. The following
steps corresponded to those described for the determination of t-PSA.

Statistical Analysis
The percentage of free PSA (f-PSA%) was calculated as
the ratio of f-PSA to t-PSA multiplied by 100. The authors used the Mann-Whitney U test, variance analysis
(ANOVA), and the rank correlation coefficient according to Spearman with the statistical package Statgraphics, version 5.01 (Statistical Graphics Corp.,
Rockville, MD). The diagnostic validity was evaluated
by the receiver operation characteristic (ROC) curve
analysis.15 The GraphROC for Windows software was
used for calculations of areas under the curves.16

RESULTS
Table 1 shows the results of t-PSA and f-PSA measurements and for f-PSA% calculation. Patients with PCa
had a lower proportion of f-PSA than BPH patients
and healthy men. The PCa-free groups showed no differences in f-PSA% values.
TNM stage and grade had no influence on the

W: Cancer

106

CANCER January 1, 1997 / Volume 79 / Number 1

FIGURE 1. Relationship between the free/total prostate specific antigen

ratio and prostate volume in patients with prostate carcinoma (s) and
benign prostate hyperplasia (m). The vertical line separates patients with
prostate volume 40 cm3 from those with 40 cm3. The horizontal line
corresponds to the cutoff value of 15% for free/total prostate specific
antigen ratio.

ratio of f-PSA/t-PSA. The t-PSA was not correlated to

the grade (correlation coefficient [r] 0.3; P 0.08)
but to the TNM stage (r 0.43; P 0.05), whereas fPSA% had no correlation to either grading (r 0.04;
P 0.81) or TNM stage (r 0.09; P 0.58).
The authors were especially interested in the relationship of prostate volume to PSA. In BPH patients
the t-PSA showed a positive correlation to prostate
volume (r 0.66; P 0.0001). In contrast to this correlation, PCa patients showed no relationship between
prostate volume and t-PSA (r 0.01; P 0.93). However, PCa patients showed a significant positive correlation between f-PSA% and the prostate volume (r
0.51; P 0.001). Conversely, BPH patients showed no
relationship between f-PSA% and prostate volume (r
00.27; P 0.05). The contrast in the relationship
of t-PSA and f-PSA% to prostate volume in PCa and
BPH patients was noteworthy.
To accurately distinguish the gland volume as a
factor influencing PSA, the prostates were subdivided
into glands 40 cm3 and enlarged glands with volumes 40 cm3. The individual f-PSA% values with
respective prostate volume for BPH and PCa patients
are shown in Figure 1. The f-PSA% cutoff was marked
as 15% in this figure. This cutoff guaranteed a high
diagnostic efficiency in discriminating between PCa

/ 7b48$$0817

12-10-96 11:23:14

cana

and BPH patients, as the authors suggested in a previous paper.17 The analysis of these two volume groups
resulted in two important findings (Table 2): 1) BPH
patients with prostate volumes 40 cm3 and 40 cm3
did not differ in f-PSA% values (P 0.46). Prostate
volume had no influence on f-PSA% in BPH patients;
and 2) the f-PSA% values in PCa patients differed significantly in both volume groups (9% vs. 15.1%; P
0.002). This confirms the increase of f-PSA% with increasing prostate volume. Hence it follows that there
is a significant difference in f-PSA% (P 0.0001) in
BPH and PCa patients with volumes 40 cm3. If the
prostate volume exceeds 40 cm3, both groups do not
differ significantly in f-PSA% (18.2% vs. 15.1%; P
0.11). It should be noted that the evaluation of the
subset of patients only with PSA values within the
range of 4 10 mg/L resulted in nearly identical median
f-PSA% values as given for the total group in Table 2.
Whereas BPH patients with a prostate volume 40
cm3 (n 5) and 40 cm3 (n 13) had f-PSA% values
of 20.9% and 18.1%, respectively, PCa patients (n
14; n 8) had values of 9.1% (n 14) and 15.4%
(n 8), respectively. Thus, the different effect of the
prostate volume in BPH and PCa patients on the fPSA% values as described above does not depend on
the total PSA value within the range of 0 23 mg/L.
Only one PCa patient with a prostate volume of
40 cm3 exceeded the value of 15% f-PSA% (Fig. 1).
Of the BPH patients, only one patient was below this
cutoff limit. Thus, values for sensitivity and specificity
were high (93.8% and 96.2%, respectively). Compared
with f-PSA% calculation alone, this leads to an increase
of efficiency to 95%.
Figure 2 includes the ROC curves of t-PSA and fPSA% for all PCa and BPH patients (n 80) and the
f-PSA% curves for all PCa and BPH patients with prostate volumes 40 cm3 (n 42). The t-PSA curve for
the patients with prostate volumes 40 cm3 is not
indicated but used in calculations. The f-PSA% curve
of all 80 patients and the t-PSA curve of patients with
volumes 40 cm3 runs significantly above the t-PSA
curve of all patients (P 0.05). The f-PSA% curve of
patients with volumes 40 cm3 runs again significantly above the f-PSA% curve of all BPH and PCa
patients (P 0.05). It follows that the diagnostic validity of f-PSA% as a diagnostic tool to discriminate between PCa and BPH patients would be essentially improved if it is used in combination with prostate volume (40 cm3 and 40 cm3).

DISCUSSION
PSA is the most important serum parameter for early
detection of PCa.18 Increasing PSA concentrations are
also found in BPH, inflammatory prostatic disease, or

W: Cancer

Free/Total PSA Ratio and Prostate Volume/Stephan et al.

107

TABLE 2
Total PSA and Free PSA% in Dependence on Prostate Volume in Patients with Benign Prostate Hyperplasia
and Prostate Carcinoma
Benign prostate hyperplasia

Prostate carcinoma

Measurement

Vol 40 cm3
(n 16)

Vol 40 cm3
(n 28)

Vol 40 cm3
(n 26)

Vol 40 cm3
(n 10)

Total PSA (mg/L)a

Free PSA%a

2.4 (0.965.1)
21.6 (15.235.7)

6.2b (2.519.6)
18.2 (9.241.3)

7.6b (3.916.4)
9b (6.314.2)

7.9 (4.117.6)
15.1c (8.422.5)

PSA: prostate specific antigen; PSA%: percentage of free prostate specific antigen.
a
Values are given as median (and 1090 percentiles).
b
Significantly different from patients with benign prostate hyperplasia and prostate volume 40 cm3 by at least P 0.01 (MannWhitney U test).
c
Significantly different from patients with prostate carcinoma and prostate volume 40 cm3 by at least P 0.01 (MannWhitney U test).

FIGURE 2.

Receiver operating characteristic curves for total prostate

specific antigen ( l ), free/total prostate specific antigen ratio (fPSA%) ( s ), and free/total prostate specific antigen ratio combined
with prostatic volume 40 cm3 ( . ) for patients with prostate
carcinoma and benign prostate hyperplasia. The areas under the curves
for total prostate specific antigen, free/total prostate specific antigen ratio,
and the ratio combined with volume 40 cm3 were 0.708 (standard error
[SE]: 0.059), 0.862 (SE: 0.040), and 0.971 (SE: 0.024), respectively.

could be age-related or caused by rectal manipulation.

Therefore, various approaches including PSA velocity,
PSA density, age-related PSA reference intervals, and
the ratio of free to total PSA have been suggested to
improve the diagnostic validity of PSA.19 The detection
and measurement of different PSA forms in serum en-

/ 7b48$$0817

12-10-96 11:23:14

cana

ables better discrimination of PCa and BPH. A number

of reports regarding t-PSA, f-PSA, and the ratio of both
have demonstrated the possibility of improving discrimination between PCa and BPH.2,4,5,7 10 The current
investigation confirms these results but points out the
influence of prostate volume as a further parameter.
In 1995 Hammerer et al.20 described a dependence
of PSA on the three individual glandular zones of the
prostate. The correlation coefficients among serum
PSA and the transition zone, peripheral zone, and central zone were 0.934, 0.546, and 0.368, respectively.
Most BPH develops from the transition zone. Hyperplasia of the transition zone therefore is responsible
for an increase in the volume of the entire gland. The
current study results confirm this correlation, because
t-PSA from BPH patients strongly correlated with prostate volume. Contrary to this, PCa patients show no
correlation between these two parameters, but do
show the above mentioned significant correlation between f-PSA% and prostate volume, whereas BPH patients have no relationship between f-PSA% and volume.
The most important result of this article is the
demonstration of the significant positive correlation
between prostate volume and calculated f-PSA% in
patients with PCa. If the volume increases, the f-PSA%
also increases. PCa patients with an enlarged gland
often have BPH as well. This could lead to a decrease
in the influence of the PCa on f-PSA% in patients with
both conditions. An investigation by Stamey et al.21
regarding the influence of the PCa on f-PSA% showed
a general increase of bound PSA (PSA-ACT/PSA) with
increasing PCa volume. Patients with the largest PCa
volumes (12 17 cm3) had the smallest f-PSA% values.
This confirms the influence of PCa volume on the relation of PSA forms in the serum. However, the data of
Stamey et al.21 were based on prostates with only a

W: Cancer

108

CANCER January 1, 1997 / Volume 79 / Number 1

single peripheral carcinoma. In the current study, the

authors did not measure the tumor volume and were
unable to further substantiate the results of Stamey et
al.21 The authors have had experiences similar to those
of other authors in that the irregular shape and especially the multifocality of the PCa complicates the accurate determination of the tumor volume and calls
into question the approach of tumor volume measurement in practice.22,23 In addition, until now it has not
been clearly defined whether all foci, only the index
tumor nodule, or the apparent most significant lesion
should be included for the PCa volume determination.24
While this article was being prepared, a similar
report and correspondence on this topic were published.13,25,26 Catalona et al.13 also measured f-PSA% in
PCa patients with a prostate volume 40 cm3 and
enlarged glands ( 40 cm3) and found median values
similar to those presented here (vol 40 mL: 9% vs.
9.2%; and vol 40 mL: 15.1% vs. 15.9%). They used
a t-PSA range of 4.1 10 mg/L and applied Tandem E
PSA (Hybritech Inc., San Diego, CA), whereas the authors chose a t-PSA range of 0 23 mg/L and used
IMMULITE Diagnostic Products Corp. assays. Therefore, a comparison between the data obtained in these
two clinical studies remains problematic because of
the different assays used for f-PSA%17 and clinical differences. However, because the subset of patients in
the current study with PSA values of 4 10 mg/L showed
no differences compared with the total group of patients with the broader range of 0 23 mg/L and the
first data have been reported on the strong correlation
between Hybtritech and IMMULITE assays,27 the authors interpreted these results as mutual confirmation.
Catalona et al.13 also found no relationship between
TNM stage and grade with prostate volume, whereas
Filella et al.10 described a negative correlation between
TNM stage and volume. To obtain a specificity of 95%,
Filella et al.10 used an f-PSA% cutoff limit of 8%, but
did not consider prostate volume. With this cutoff the
sensitivity was only 44%. The various cutoffs recommended for f-PSA% perhaps result from the fact that
prostate volume was not taken into account as factor
influencing f-PSA%. For enlarged glands the f-PSA%
median values between PCa and BPH patients are not
significantly different.
In conclusion, the current study results show the
significant influence of prostate volume on f-PSA%.
Patients with prostate volumes of 40 cm3 have an
increased diagnostic value in determining f-PSA%.
Thus, the determination of f-PSA% to differentiate between PCa and BPH patients is important in men with
a normal sized prostate gland. By using t-PSA, f-PSA%,
and transrectal ultrasonography, the possibility of in-

/ 7b48$$0817

12-10-96 11:23:14

cana

creasing the sensitivity and specificity to approximately 95% may lead to a more effective detection of
PCa.

REFERENCES
1.

2.

3.

4.

5.

6.
7.

8.

9.

10.

11.

12.

13.

14.

15.

Christensson A, Laurell CB, Lilja H. Enzymatic activity of

prostate-specific antigen and its reactions with extracellular
serine proteinase inhibitors. Eur J Biochem 1990;194:755
63.
Christensson A, Bjork T, Nilsson O, Dahlen U, Matikainen
MT, Cockett ATK, et al. Serum prostate specific antigen complexed to a1-antichymotrypsin as an indicator of prostate
cancer. J Urol 1993;150:1005.
Wood WG, van der Sloot E, Bohle A. The establishment and
evaluation of luminescent-labeled immunometric assays for
prostate-specific antigen-a1-antichymotrypsin complexes in
serum. Eur J Clin Chem Clin Biochem 1991;29:78794.
Lilja H, Christensson A, Dahlen U, Matikainen MT, Nilsson
O, Pettersson K, et al. Prostate-specific antigen in serum
occurs predominantly in complex with a1-antichymotrypsin. Clin Chem 1991;37:161825.
Stenman UH, Leinonen J, Alfthan H, Rannikko S, Tuhkanen
K, Alfthan O. A complex between prostate-specific antigen
and a1-antichymotrypsin is the major form of prostate-specific antigen in serum of patients with prostatic cancer: assay
of the complex improves clinical sensitivity for cancer. Cancer Res 1991;51:2226.
Wu JT. Assay for prostate specific antigen (PSA): problems
and possible solutions. J Clin Lab Anal 1994;8:5162.
Lilja H. Regulation of the enzymatic activity of prostatespecific antigen and its reactions with cellular protease inhibitors in prostate cancer. Scand J Clin Lab Invest
1995;55(Suppl 220):4756.
Oesterling JE, Jacobsen SJ, Klee GG, Pettersson K, Piironen
T, Abrahamsson PA, et al. Free, complexed and total serum
prostate specific antigen: the establishment of appropriate
reference ranges for their concentrations and ratios. J Urol
1995;154:10905.
Prestigiacomo A, Stamey TA. Clinical usefulness of free and
complexed PSA. Scand J Clin Lab Invest 1995;55(Suppl
221):324.
Filella X, Alcover J, Molina R, Gimenez N, Rodriguez A, Jo J,
et al. Clinical usefulness of free PSA fraction as an indicator
of prostate cancer. Int J Cancer 1995;63:7804.
Bjork T, Hulkko S, Bjartell A, di SantAgnese A, Abrahamsson
PA, Lilja H. Alpha 1-antichymotrypsin production in PSAproducing cells is common in prostate cancer but rare in
benign prostatic hyperplasia. Urology 1994;43:42734.
Bjartell A, Bjork T, Matikainen MT, Abrahamsson PA, di
SantAgnese A, Lilja H. Production of alpha-1-antichymotrypsin by PSA-containing cells of the human prostate epithelium. Urology 1993;42:50210.
Catalona WJ, Smith DS, Wolfert RL, Wang TJ, Rittenhouse
HG, Ratliff TL, et al. Evaluation of percentage of free serum
prostate-specific antigen to improve specificity of prostate
cancer screening. JAMA 1995;274:121420.
Schroder FH, Hermanek P, Denis L, Fair WR, Gospodarowicz
MK, Pavone-Macaluso M. The TNM classification of prostate
cancer. Prostate 1992;20(Suppl 4):12938.
Zweig MH, Campbell G. Receiver-operating characteristic
(ROC) plots: a fundamental evaluation tool in clinical medicine. Clin Chem 1993;39:56177.

W: Cancer

Free/Total PSA Ratio and Prostate Volume/Stephan et al.

16. Kairisto V, Poola A. Software for illustrative presentation of basic
clinical characteristics of laboratory testsGraphROC for Windows. Scand J Clin Lab Invest 1995;55(Suppl 222):4360.
17. Jung K, Stephan C, Lein M, Henke W, Schnorr D, Brux B, et
al. Analytical performance and clinical validity of two free
prostate-specific antigen assays compared. Clin Chem
1996;42:102633.
18. Oesterling JE. Prostate specific antigen: a critical assessment
of the most useful tumor marker for adenocarcinoma of the
prostate. J Urol 1991;145:90723.
19. Bostwick DG. Prostate-specific antigen. Current role in diagnostic pathology of prostate cancer. Am J Clin Pathol
1994;102:S317.
20. Hammerer PG, McNeal JE, Stamey TA. Correlation between
serum prostate specific antigen levels and the volume of the
individual glandular zones of the human prostate. J Urol
1995;153:1114.
21. Stamey TA, Chen Z, Prestigiacomo A. Serum prostate specific antigen binding alpha 1-antichymotrypsin: influence of

/ 7b48$$0817

12-10-96 11:23:14

cana

22.

23.

24.
25.
26.

27.

109

cancer volume, location and therapeutic selection of resistant clones. J Urol 1994;152:15104.
Villers A, McNeal JE, Freiha FS, Stamey TA. Multiple cancers
in the prostate. Morphologic features of clinically recognized
versus incidental tumors. Cancer 1992;70:23138.
Aihara M, Wheeler TM, Ohori M, Scardino PT. Heterogeneity
of prostate cancer in radical prostatectomy specimens. Urology 1994;43:606.
Grignon DJ, Sakr WA. Pathologic staging of prostate carcinoma. What are the issues? Cancer 1996;78:33740.
Catalona WJ, Smith DS. Free serum prostate-specific antigen
and screening for prostate cancer. JAMA 1996;275:838.
Bangma CH, Kranse R, Schroder FH. Free serum prostatespecific antigen and screening for prostate cancer. JAMA
1996;275:8378.
Spitz J, Benz P, Jauch M, Eiffler E, Kollermann M. Critical
aspects related to the interpretation of the ratio free/total
PSA, estimated by different assay methods. Anticancer Res
1995;15:24001.

W: Cancer