| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Articles |
Departments of
1
Urology,
2
Laboratory Medicine and Pathobiochemistry, and
3
Pathology, University Hospital Charité, Humboldt University, Schumannstrasse 20/21, D-10098 Berlin, Germany.
a Author for correspondence. Fax 49-30-2802-1402; e-mail klaus.jung{at}charite.de
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
2-macroglobulin with the assay of free PSA (fPSA) and
the corresponding ratios to total PSA (tPSA) to improve the
differentiation between benign prostatic hyperplasia (BPH) and prostate
cancer (PCa). Methods: Serum samples were collected from 91 men without prostate disease and with normal digital rectal examination (controls), 144 untreated patients with PCa, and 89 patients with BPH. tPSA and cPSA were measured using the Bayer Immuno 1 system; fPSA and the additional tPSA were measured with the Roche Elecsys system.
Results: The median cPSA/tPSA, fPSA/tPSA, and fPSA/cPSA ratios were significantly different between patients with BPH and PCa (78.7% vs 90.7%, 25.5% vs 12.1%, and 36.8% vs 14.3%, respectively; P <0.001). No correlations of cPSA and its ratios to tumor stage and grade were found. ROC analysis showed that cPSA was not different from tPSA (areas under the curve, 0.632 vs 0.568), whereas the cPSA/tPSA ratio was similar to the fPSA/tPSA ratio in increasing discrimination between BPH and PCa patients with tPSA concentrations in the tPSA gray zone between 2 and 10 µg/L (areas under the curve, 0.851 vs 0.838).
Conclusions: Compared with tPSA, the fPSA/tPSA and cPSA/tPSA ratios both improve the differentiation between BPH and PCa comparably and are similarly effective in reducing the rate of unnecessary biopsies, whereas cPSA alone does not have any effect.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
PSA occurs in serum in different molecular forms (2)(3)(4).
Approximately 70–90% is bound to
1-antichymotrypsin (ACT), and a small
amount is complexed with 1-antitrypsin and
protein C. An additional portion of PSA that is complexed with
2-macroglobulin can be measured only if the
complex is opened and the PSA epitopes become accessible. Of the total
PSA (tPSA) in serum, 10–30% is not bound to serum proteins and is
called free PSA (fPSA). Numerous studies have demonstrated a
lower ratio of fPSA to tPSA in PCa patients, calculated as the
percentage of fPSA [reviewed in Ref. (5)]. This ratio has
been considered a promising tool for distinguishing between PCa and
BPH. It has also been shown that ACT-PSA and the corresponding
ACT-PSA/tPSA ratio improved the specificity and sensitivity for PCa
(2)(3)(6). However, several
analytical difficulties impair accurate ACT-PSA measurement
(7). For example, overrecovery of the ACT-PSA complex
results when anti-PSA antibodies are used on the solid phase to trap
ACT-PSA complex and when anti-ACT antibodies are applied to detect this
complex. This overrecovery is caused by the presence of ACT-cathepsin G
complex in serum, which also binds to the solid phase (8).
Various approaches have been suggested to eliminate these technical
problems (7)(8)(9). The use of monoclonal antibodies specific
against ACT-PSA with low cross-reactivities to the cathepsin G-ACT
complex, ACT, and fPSA (9), the addition of heparin
(8), and the application of special blocking reagents to
reduce the nonspecific binding of anti-ACT antibodies (7)
have been recommended. The recently introduced novel PSA assay that
measures all complexed PSA (cPSA) except PSA complexed with
2-macroglobulin also offers a promising
possibility to eliminate the technical obstacles for reliable
measurement (10).
We used this new test in patients with BPH and PCa to evaluate the assay with the following aims: (a) to evaluate the diagnostic performance of the assay in comparison with tPSA; (b) to compare cPSA and fPSA alone or as corresponding ratios to tPSA for differentiating between patients with PCa and BPH; and (c) to evaluate the relationship between cPSA and clinical characteristics of the patients.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Control group.
This group consisted of 91 men (median age, 54
years; range, 21–76 years) with normal digital rectal examinations.
The individuals were either patients hospitalized in our department or
attending our outpatient department because of nonprostatic diseases
(erectile dysfunction, hydrocele, stone disease without obstruction,
and infection).
BPH group.
This group included 89 untreated patients (median
age, 65 years; range, 49–85 years). The diagnosis of BPH was
established clinically by digital rectal examination and/or transrectal
ultrasonography. In 48 of these 89 patients, the BPH was histologically
confirmed either using tissue obtained by ultrasound-guided sextant
prostate biopsy or by transurethral resection of the prostate. Because
there were no differences in the fPSA/tPSA or cPSA/tPSA ratios between
the two BPH subgroups diagnosed by clinical or histological
examination, both groups were considered as one group.
PCa group.
The PCa group included 144 patients (median age, 65
years; range, 48–88 years) diagnosed histologically with blood samples
taken before different treatment regimens (radical prostatectomy,
radiotherapy, or hormonal therapy). The cancer stage was assigned
according to the TNM system, and the histological grade was classified
as grades 1, 2, and 3, as described in detail previously
(11). The pathological stages and grades of 75 patients were
as follows: pT2 pN0M0 (n = 51); pT3 pN0M0 (n = 24); G1
(n = 5); G2 (n = 45); and G3 (n = 25). The remaining 69
of the 144 patients were clinically staged with the following results:
T1 (n = 5; 2 with pN0M0 and 3 with N0M0); T2 (n = 34; 13 with
pN0M0, 19 with N0M0, 1 with pN1M0, and 1 with pN1M1); T3 (n = 30;
12 with pN0M0, 13 with N0M0, and 5 with pN1M0); G1 (n = 10); G2
(n = 40); G3 (n = 19).
sample collection
Blood samples were taken before diagnostic procedures,
transurethral resection of the prostate, or 4 weeks (at the earliest)
after digital rectal examination, prostatic biopsy, and transrectal
ultrasound to avoid possible errors caused by the release of PSA from
the prostate and the different elimination kinetics of their forms from
blood. The samples were collected in evacuated tubes (Monovette
03.1528; Sarstedt) and centrifuged at 1600g for 15 min at
4 °C after the blood was allowed to clot for 1 h at room
temperature. The sera were frozen at -80 °C within 2 h after
collection and were tested within 12 weeks. Female sera as negative
controls, in-house serum pools, and control sera from the producers of
the test kits and from Bio-Rad were used as control materials.
psa assays
tPSA was measured with both the Bayer Immuno 1 PSA Assay (product
no. T01-3450-5; Bayer Diagnostics, Tarrytown, NY) and the Roche Elecsys
PSA Immunoassay (product no. 1731262; Roche Diagnostics)
according to the instructions of the manufacturers. fPSA was measured
with the Roche Elecsys Free PSA Immunoassay (product no. 1820800; Roche
Diagnostics) on the Elecsys analyzer 1010. For determining cPSA, a
recently introduced immunoassay (product no. T01-3982-51) for the Bayer
Immuno 1 system was used (10). This assay is based on the
unique binding properties of the capture monoclonal antibody MM1 used
in the Bayer Immuno 1 assay for tPSA. That antibody fails to bind fPSA
in the presence of antibodies specific against epitope E, which is
exposed only in fPSA, so that all cPSA forms such as ACT-PSA and minor
forms except PSA complexed to 2-macroglobulin
are detected.
statistical analysis
Data were analyzed using the statistical software packages SPSS
8.0 for Windows (SPSS) and GraphPad Prism 3.00 for Windows (GraphPad
Software). The Kruskal–Wallis nonparametric ANOVA, the Mann–Whitney
U-test, and the calculation of rank correlation coefficients
according to Spearman (rS) were
performed. The software GraphROC 2.1 for Windows was used to analyze
the ROC. Patients for the ROC analysis were selected by the
randomization procedure used in SPSS 8.0. Regression analysis for
methodical evaluation was performed using the software EVAPAK 3.01 for
Windows according to Passing and Bablok (12). P
<0.05 was considered statistically significant.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
synoptically demonstrates the corresponding interassay
imprecision profiles, which show comparable reproducibility for all
four assays.
|
The Bayer Immuno 1 PSA assay for tPSA has shown to be an equimolar test (13). According to the manufacturer’s information, the Roche tPSA assay also measures on an equimolar basis. Data of a recent IFCC standardization study confirmed that the results obtained with the Roche test were comparable with the Tandem® test from Hybritech (14). The method comparison (12) of the two tPSA assays showed similar slopes and intercepts between the groups so that the regression line was estimated for all 324 samples combined. The equation (95% confidence intervals in parentheses), yRoche = 1.097 (1.072–1.136)xBayer + 0.057 (-0.003 to 0.124), showed that the tPSA concentrations measured with the Roche Elecsys assay were ~10% higher than those measured with the Bayer Immuno 1 assay. The values of fPSA plus cPSA in BPH and PCa patients related to the tPSA measured with the Bayer Immuno 1 assay and the Roche Elecsys assay were comparable between the patients but were different between the assay systems, being 111% vs 107% (Bayer) and 98.3% vs 101% (Roche), respectively. Therefore, the fPSA/cPSA ratio, although measured by two different assay systems, was used as an additional variable, whereas the fPSA/tPSA and cPSA/tPSA ratios were calculated using the data of the corresponding Bayer or Roche assays.
tPSA, fPSA, cPSA, AND THEIR RATIOS IN THE STUDY
GROUPS
Of the 324 subjects studied, 137 (84 controls, 36 BPH patients,
and 7 PCa patients) had tPSA values between 0 and 2 µg/L, 124 (7
controls, 47 BPH patients, and 81 PCa patients) had values between 2.01
and 10 µg/L, and 43 (6 BPH and 37 PCa patients) had values between
10.1 and 20 µg/L; of the remaining 19 PCa patients, 18 had tPSA
values between 20.1 and 100 µg/L, and 1 patient had a value >100
µg/L. Fig. 2
shows the scatter plots and medians for tPSA (only the Bayer
test), fPSA, cPSA, and the fPSA/tPSA (Roche tests), cPSA/tPSA (Bayer
tests), and fPSA/cPSA ratios. The mean ages of the BPH (65 years) and
PCa (65 years) patients did not differ but were somewhat higher than in
the controls (54 years). Significant differences between the groups
were shown by Kruskal–Wallis nonparametric ANOVA (Fig. 2
). PCa
patients showed higher tPSA and cPSA concentrations than controls and
BPH patients, whereas they were characterized by lower fPSA/tPSA and
fPSA/cPSA ratios and a higher cPSA/tPSA ratio, respectively (Fig. 2
, D-F).
|
relationship between cPSA AND OTHER PSA FORMS, TUMOR
STAGING, AND GRADING
The cPSA values were more closely related to the tPSA values
(corresponding rS = 0.959, 0.991, and
0.994 for controls, BPH patients, and PCa patients, respectively) than
to the fPSA values (rS = 0.542, 0.724,
and 0.647, respectively). As described previously for fPSA and
the fPSA/tPSA ratio (11), no correlations of cPSA
and the isoform ratios to the pathological tumor stage and the
histological grading were found. For example, the median values of the
cPSA/tPSA ratio in patients with stages of pT2 (n = 51) and pT3
(n = 24) or grades G1 (n = 5), G2 (n = 45), and G3
(n = 25) did not differ significantly (92.3% vs 93.3%, 91.3%,
91.7%, and 93%, respectively). Similarly, the cPSA/tPSA, fPSA/tPSA,
and fPSA/cPSA ratios were not different between patients with lymph
node stages of pN0 (n = 101) and pN1 (n = 7; 91.5% vs
96.9%, 12.1% vs 10.8%, and 14.2% vs 11.2%, respectively), whereas
tPSA and cPSA concentrations were significantly increased in patients
with metastatic lymph nodes (8.03 vs 26.9 µg/L and 7.1 vs 25.3
µg/L, respectively; P <0.001).
roc analysis and diagnostic validity
We performed ROC analyses in patients with BPH and PCa for the
entire tPSA range (0.33–365 µg/L) and for the particularly
characteristic range of overlapping tPSA concentrations in both groups
(2–10 µg/L). When the entire tPSA range was considered, areas under
the ROC curves for cPSA, fPSA, and the respective ratios did not show
any statistical differences compared with the area under the tPSA curve
[e.g., mean area under the curve ± SE, 0.870 ±
0.024 for tPSA (Bayer); 0.888 ± 0.022 for cPSA; 0.859 ±
0.028 for fPSA/tPSA]. To simulate the characteristics of overlapping
tPSA values in both groups of patients, equal numbers of corresponding
patients were randomly selected from the larger group to match the
number of patients in the smaller group at 1-µg/L tPSA intervals
within the tPSA range mentioned above, using the randomization
procedure of SPSS 8.0. Thus, 40 BPH and 40 PCa patients were selected.
A similar matching procedure was applied recently to avoid the
influence of tPSA as a confounding factor (15). The areas
under the ROC curves for tPSA (both assays), cPSA, and fPSA were not
significantly different (P >0.005; Fig. 3
A), but the areas under the curves of all ratios (Fig. 3B
) were
significantly higher (P 
0.05) than those of the PSA
concentrations. No significant differences (P >0.05)
between the areas under the curves of the fPSA/tPSA, cPSA/tPSA, and
fPSA/cPSA ratios were observed.
|
The diagnostic validity criteria sensitivity, specificity, and
efficiency of tPSA, cPSA, and their ratios at different decision limits
of the ROC curves are shown in Table 1
. cPSA alone did not improve the sensitivity or specificity
compared with tPSA to differentiate between patients with BPH or PCa.
However, the specificity and the sensitivity, respectively, increased
by ~30–50% compared with tPSA or cPSA if one of the three ratios,
fPSA/tPSA, cPSA/tPSA, or fPSA/cPSA, was used as discriminatory
indicator at the decision limit with the highest efficiency or a
sensitivity or specificity of 90%.
|
No statistical differences in the specificity and sensitivity,
respectively, were found at the selected 90% sensitivity or
specificity between the two ratios fPSA/tPSA and cPSA/tPSA
(P >0.05). On the basis of these data, the usefulness of
the two ratios was demonstrated by calcu-lating the true-negative and
false-positive results in BPH patients and the true-positive and
false-negative results in PCa patients, respectively (Table 2
). It is obvious that unnecessary biopsies could have
been avoided in ~65% of BPH patients, whereas ~8% of cancers
would had been missed.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The interassay imprecision for all assays was generally <5% (Fig. 1
)
and made it possible to compare the diagnostic validity of the various
assays and derivatives. It is of interest that although the Bayer and
Roche assays have been calibrated against the Stanford 90:10 reference
preparation (14), there were differences of ~10% between
tPSA concentrations measured by both assays. Reasons other than the
properties of the calibrator, e.g., the selectivity of antibodies, the
equilibration time during the measurement, and the format of the assay,
must be considered as influencing the final result.
Our findings clearly indicate that cPSA testing is not superior to tPSA
testing alone (Fig. 3
and Table 1
). These data are contradictory to the
results of Brawer et al. (19), who found that cPSA alone was
a better discriminator between BPH and PCa than tPSA or the fPSA/tPSA
ratio in the range between 4 and 10 µg/L. According to the
suggestions of those authors, the determination of cPSA could replace
the measurements of the two analytes tPSA and fPSA. However, both
practical experience and the theoretical background do not give reason
for such hope. Because cPSA strongly correlates with tPSA, a large
overlapping range of cPSA concentrations consequently exists between
PCa and BPH patients within in the gray zone of tPSA concentrations up
to 10 µg/L. Thus, although the cutoff may be narrowed when cPSA is
determined, the general problem of the overlapping PSA concentrations
characteristic of these two groups of patients cannot be solved by such
an approach. A cutoff for cPSA of 3.75 µg/L was chosen to
obtain a sensitivity for PCa detection equal to that achieved with the
tPSA cutoff limit of 4 µg/L (10); that choice may
emphasize that the high expectation concerning the sole determination
of cPSA is not very realistic. For example, a median cPSA/tPSA ratio of
~80% in BPH patients means that patients with tPSA concentrations
>4.7 µg/L generally exceed the limit of 3.75 µg/L for cPSA and
thus belong to the patient group with tPSA concentrations in the gray
zone. Two studies reported in abstracts were also unable to confirm
that cPSA was better than the fPSA/tPSA ratio in discriminating
biopsy-negative from biopsy-positive men
(20)(21). In addition, because the area under
the ROC curve for the percentage of fPSA did not show the typical value
higher than that for tPSA, it was assumed that Brawer et al.
(19) studied inappropriate cohorts (22).
Our results (Figs. 2
and 3
) confirm our own findings and the data of
numerous studies that the fPSA/tPSA ratio is statistically different
between patients with PCa cancer and BPH [reviewed in Ref.
(5)]. Because the diagnostic validity of that ratio is not
superior to tPSA over an expanded range (23), we included in
the ROC analysis only patients with tPSA concentrations between 2 and
10 µg/L. Selecting in our ROC curves the points at 90% sensitivity,
at the highest diagnostic efficiency, or at the equivalent point of
sensitivity and specificity, threshold values were 22.5%, 17.5%, and
19.2%, respectively (Table 1
). These cutoff points for the fPSA/tPSA
ratio roughly correspond to values given by other authors in the gray
zone of PSA between 2 and 20 µg/L (24)(25)(26)(27)(28), although
assay-specific compatibility of these values should be considered
(29). Thus, our results obtained with cPSA and the
corresponding ratios were based on an appropriate patient selection.
The matching approach as applied in our study avoided a possible
confounding effect related to unequal tPSA values in the cohorts and
should generally be used in such studies.
The cutoff point of 81.3% for cPSA/tPSA produced a diagnostic
sensitivity of 90% to detect PCa patients within the tPSA range
between 2 and 10 µg/L (Table 1
). Our results correspond to data shown
in a report on preliminary measurements of cPSA (10). Using
the same cPSA assay as the prototype test, those authors revealed that
all men with PCa had cPSA >77%. Brawer et al. (19) also
performed cPSA measurements using such a test but did not calculate the
cPSA/tPSA ratio. The ROC analysis (Fig. 3B
) and the specificity of the
cPSA/tPSA ratio achieved at a selected sensitivity of 90% (Tables 1
and 2
) compared with tPSA, cPSA, and other ratios underlines our view
that the cPSA/tPSA ratio can be accepted as a real alternative to the
fPSA/tPSA ratio for improving the differentiation between PCa and BPH
patients. The cutoffs shown in Table 1
are provisional because of the
limited number of patients investigated. However, the results show that
the two ratios, fPSA/tPSA and cPSA/tPSA, are similarly effective in
reducing the rate of unnecessary biopsies (Table 2
), as demonstrated
recently by other authors (15)(23). It can be
assumed that the use of the cPSA/tPSA ratio instead of the fPSA/tPSA
ratio is also helpful to prepare cancer probability curves
(30). While this report was being prepared, an abstract was
published that confirmed our results that the fPSA/tPSA and cPSA/tPSA
ratios are equal in their diagnostic validity (31). The
obstacles observed with ACT-PSA assays as found in our previous study
(18) and also by others (9) have apparently been
overcome the new cPSA assay.
We also evaluated the fPSA/cPSA ratio, although both components were
measured by the two assay systems of Bayer and Roche. From the
theoretical point of view, a better differentiation might be possible
because the proportion of fPSA and cPSA is opposite in both groups of
patients. Although the percentage of difference of the fPSA/cPSA ratio
between patients with BPH and PCa was significantly higher
(P <0.01) than that of the fPSA/tPSA ratio (Fig. 2
, D and
F), a better discrimination was not achieved (Fig. 3
and Table 1
). A
general difference of ~10% between both tPSA concentrations with
both assays perhaps partly explains the difference of the ratios
without the expected improvement of diagnostic power. However, Demura
et al. (32), using different assays but comparing variables
similar to those we studied, reported that the fPSA/cPSA ratio was the
most powerful tool for diagnosis of PCa.
The fPSA/tPSA ratio has been examined as a method to predict the final pathological stage of PCa. The relationships between this ratio and the stage, grade, and aggressiveness of PCa are controversial at this point in time [reviewed in Ref. (5)]. Our data could not confirm that the ratios of fPSA/tPSA, cPSA/tPSA, and fPSA/cPSA depend on tumor stage and grade. Therefore, none of these ratios can be recommended as a predictive indicator for final pathological staging of PCa.
In summary, compared with the determination of tPSA, the differentiation of patients with BPH and PCa with tPSA concentrations in the overlapping range between 2 and 10 µg/L could be equally improved with the ratios of free or complexed PSA to tPSA, whereas cPSA alone does not have any additional discriminatory power. Thus, the cPSA/tPSA ratio can be considered as an alternative to the fPSA/tPSA ratio.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
1-antichymotrypsin; tPSA, total PSA, both free and complexed forms; fPSA, free, noncomplexed form of PSA; and cPSA, complexed forms of PSA. 
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
1-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;51:222-226.
1-antichymotrypsin. Clin Chem 1991;37:1618-1625.
1-antichymotrypsin as an indicator of prostate cancer. J Urol 1993;150:100-105.
[Web of Science][Medline]
[Order article via Infotrieve]
1-antichymotrypsin. Clin Chem 1993;39:2098-2103.
[Abstract]
1-antichymotrypsin complex. Clin Chem 1995;41:1480-1488.
1-antichymotrypsin complex for the screening and management of prostate cancer. J Clin Lab Anal 1998;12:32-40.
[Web of Science][Medline]
[Order article via Infotrieve]
1-antichymotrypsin-prostate-specific antigen in patients with prostate carcinoma but not in those with benign prostate hyperplasia. Clin Chem 1999;45:422-425.
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  M. R. Cooperberg, J. W. Moul, and P. R. Carroll The Changing Face of Prostate Cancer J. Clin. Oncol., November 10, 2005; 23(32): 8146 - 8151. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Keller, H. Butz, M. Lein, M. Kwiatkowski, A. Semjonow, H.-J. Luboldt, P. Hammerer, C. Stephan, and K. Jung Discordance Analysis Characteristics as a New Method to Compare the Diagnostic Accuracy of Tests: Example of Complexed Versus Total Prostate-Specific Antigen Clin. Chem., March 1, 2005; 51(3): 532 - 539. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Nakamura, A. Scorilas, C. Stephan, K. Jung, A. R. Soosaipillai, and E. P. Diamandis The Usefulness of Serum Human Kallikrein 11 for Discriminating between Prostate Cancer and Benign Prostatic Hyperplasia Cancer Res., October 1, 2003; 63(19): 6543 - 6546. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Zhu, J. Leinonen, W.-M. Zhang, P. Finne, and U.-H. Stenman Dual-Label Immunoassay for Simultaneous Measurement of Prostate-specific Antigen (PSA)-{alpha}1-Antichymotrypsin Complex Together with Free or Total PSA Clin. Chem., January 1, 2003; 49(1): 97 - 103. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. A. Serdar, O. Oguz, A. Olgun, B. Seckin, S. Ilgan, A. Hasimi, M. Salih, F. Peker, and T. Kutluay Diagnostic Approach to Prostate Cancer using Total Prostate Specific Antigen-Based Parameters Together Ann. Clin. Lab. Sci., January 1, 2002; 32(1): 22 - 30. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Wiese, Y. Belosludtsev, T. Powdrill, P. Thompson, and M. Hogan Simultaneous Multianalyte ELISA Performed on a Microarray Platform Clin. Chem., August 1, 2001; 47(8): 1451 - 1457. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Stephan, K. Jung, M. Lein, P. Sinha, D. Schnorr, and S. A. Loening Molecular Forms of Prostate-specific Antigen and Human Kallikrein 2 as Promising Tools for Early Diagnosis of Prostate Cancer Cancer Epidemiol. Biomarkers Prev., November 1, 2000; 9(11): 1133 - 1147. [Abstract] [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
), 89 patients with BPH (
), and
144 patients with PCa (•). Median values of the respective groups are
shown as horizontal lines. Only tPSA values measured
with the Bayer Immuno 1 assay are presented. Significant differences
between the groups (Kruskal–Wallis nonparametric ANOVA) are indicated:
a, significantly different from controls;
b, significantly different from patients with BPH;
c, significantly different from patients with PCa.
), 0.643 ± 0.059; cPSA (
), 0.632 ± 0.062;
(P >0.05). (B), mean areas under the
curve ± SE: fPSA/tPSA (