| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Technical Briefs |
a author for correspondence: fax 619-536-8058, e-mail abblase{at}beckman.com
Prostate-specific antigen (PSA) has been established as a marker
to aid in detection and monitoring of prostate cancer
(1)(2). PSA in serum
exists predominantly in three forms: free, uncomplexed PSA; PSA
covalently complexed to 
1-antichymotrypsin (PSA-ACT);
and PSA covalently complexed to 2-macroglobulin (PSA-MG)
(3). Immunoassays available today
recognize free PSA and PSA-ACT but not PSA-MG, and their result for
"total PSA" refers to the sum of the free and ACT-bound forms of
PSA as measured by the immunoassay.
In general, the proportion of free PSA relative to PSA-ACT is lower in prostate cancer patients than in normal subjects or in patients with noncancerous prostatic disease (4). The percentage of free PSA ranges from 5% to 50% when total serum PSA is 4–10 µg/L (5)(6). However, increased serum concentrations of PSA do not necessarily indicate prostate cancer because such values can also occur in cases of benign prostate hyperplasia or prostatitis (7)(8)(9).
Assays of total serum PSA differ from one another in an important respect: They do not recognize the free and ACT-bound species of PSA equivalently (10)(11). "Equimolar-response assays" measure equal molar concentrations of free PSA and PSA-ACT equivalently; "skewed-response assays" measure these PSA forms differently (12)(13)(14).
The total PSA concentration measured by an equimolar assay depends on only the total concentration of free PSA plus PSA-ACT and is independent of their relative proportions. The interpretation of results reported by skewed-response assays, however, may be misleading because (a) the proportion of free PSA is generally higher in noncancer patients, and (b) the molar response of skewed-response assays for free PSA is typically higher than that for PSA-ACT. Because the average proportion of free PSA is higher in noncancer patients than in cancer patients (4), values reported by skewed-response PSA assays for noncancer patients may well be higher than predicted by equimolar PSA assays. Moreover, if the ratio of free PSA to total PSA changes between serial PSA measurements but the total PSA concentration is unchanged, the skewed-response assay might nonetheless report a change in the PSA value (15).
We compared PSA assays by use of a set of "defined PSA ratio" samples, prepared with specific proportions of free PSA and PSA-ACT. We purified PSA from seminal fluid as described by Sensabaugh and Blake (16) and incubated it with a sixfold molar excess of ACT (purchased from Athens Research and Technology, Athens, GA) for 18 h at 37 °C in Tris-buffered saline, pH 7.4. The PSA-ACT complex and the unreacted free PSA were isolated by hydrophobic interaction chromatography (16). The fractions corresponding to PSA-ACT complex and to free PSA were pooled separately. After buffer exchange with 100 mmol/L ammonium acetate buffer, the concentration of each pool was determined spectrophotometrically at 280 nm (17). We prepared four working concentrations of PSA, using as a diluent a bovine protein matrix containing no detectable PSA (PSA-R Zero diluent, provided in the Tandem®-R PSA kit; Hybritech, San Diego, CA). Mixtures of PSA-ACT and free PSA in various ratios were prepared by combining the free PSA solution with the corresponding PSA-ACT solution at the same working concentration. In all, 20 solutions were prepared, representing all possible combinations of four combinations (total PSA: 20, 10, 5, and 2.5 µg/L) at five molar ratios of free:complexed PSA (0:100, 25:75, 50:50, 75:25, and 100:0). Adjustments to the concentrations were based on the spectrophotometric values measured. Each solution was divided into aliquots, which were flash-frozen in liquid nitrogen and stored at -70 °C. Assigned concentrations were not based on immunoassay results.
Each of the 20 concentration–ratio combination samples was tested in duplicate at a single, independent laboratory with each of the following assays: Tandem-E PSA for the Photon ERA (Tandem PSA ERA), a semiautomated, dual-monoclonal immunoenzymometric assay (Hybritech); IMx PSA, an automated polyclonal/monoclonal enzymoimmunoassay formatted for use in the IMx system (Abbott Labs., Abbott Park, IL); ACS:180 PSA1 and ACS:180 PSA2, both automated polyclonal/monoclonal immunochemiluminometric assays formatted for use in the ACS:180 system (Chiron Diagnostics, East Walpole, MA); and AIA-PACK PA, an automated dual-monoclonal immunoenzymometric assay formatted for use with the AIA 600 or 1200 analyzers (Tosoh, Tokyo, Japan). The ACS:180 PSA2 assay was recently released as a recalibrated version of the PSA1 assay (18)(19).
For each assay format, we calculated the ratio of the determined PSA
concentration to the assigned concentration
(Rt =
[PSA]determined/[PSA]assigned)
for each sample (Table 1
). The data were subjected to general linear models (GLM) and
analysis of variance (ANOVA) routines in SAS (Statistical Analysis
Software, Cary, NC). Using the GLM procedure, we performed Duncan's
Multiple Range test to compare the results from each manufacturer.
According to this analysis ( = 0.05), Tandem PSA ERA and Tosoh AIA
results were not significantly different from each other, but IMx PSA,
ACS PSA1, and ACS PSA2 differed significantly
from each of the other assays. To determine the effect of free PSA
proportion and the effect of total PSA concentration on Rt,
we performed a separate ANOVA for each manufacturer's assay. The data
(Rt) were grouped according to the proportion
of free PSA in the sample (0%, 25%, 50%, 75%, and 100%, with each
group containing four data points) and the concentration of total PSA
(2.5, 5, 10, and 20 µg/L, with each group containing five data
points). At = 0.05, the effect of PSA concentration was not
significant for any of the assay formats: The P-values were
0.47 (Tandem PSA ERA), 0.74 (Tosoh AIA), 0.17 (IMx PSA), 0.37 (ACS
PSA1), and 0.26 (ACS PSA2). However, the
proportion of free PSA significantly affected Rt values for
all formats except Tandem PSA ERA: The P-values
were 0.40 for Tandem PSA ERA, 0.02 for Tosoh AIA, and <0.01 for IMx
PSA, ACS PSA1, and ACS PSA2.
|
The extent of this effect is illustrated in Fig. 1
, which shows that the PSA values reported by the Tandem PSA ERA
were unaffected by the proportion of free PSA. Although our analysis
suggested a statistically significant effect of the proportion of
free PSA on Rt for the Tosoh AIA-PACK PA assay, Fig. 1
shows that any such effect was barely observable. Thus, for all
practical purposes the Tandem PSA ERA and Tosoh AIA-PACK PA both
corresponded to the definition of an equimolar assay (12).
In contrast, values reported by the IMx PSA, ACS PSA1, and
ACS PSA2 assays increased with increasing proportions of
free PSA, corresponding to the definition of a skewed assay
(12).
|
These results reinforce earlier reports that IMx PSA (11) and ACS PSA1 (20)(21) are skewed-response assays; i.e., they measure the free PSA and PSA-ACT forms differently. Our study showed that ACS PSA2 also displays a skewed response. Although the recalibrated ACS PSA2 assay has been reported to give results in concordance with those determined with the Tandem PSA assay (18)(19), this study demonstrated that the former displays a degree of skewing similar to that of the IMx PSA assay.
The intent of this study was to investigate the analytical differences between various assays in a controlled manner. For this purpose, very defined samples were created by using purified material. Analyzing these samples offers an insight into one source of the differences in reported values between assays, even when the same samples are tested. Accordingly, under some circumstances, the disparities between assays seen here could translate into clinically significant differences. This possibility, however, requires documentation by further clinical study.
Acknowledgments
We thank Barb Allen and Kurt Norton for technical assistance, Steve Mikolajczyk for assistance in preparing the samples, and Bob Parson for assistance with the statistical analyses.
Footnotes
Hybritech Incorporated, PO Box 269006, San Diego, CA 92196-9006
References
1-antichymotrypsin bound PSA [Abstract]. Clin Chem 1994;40:1009.
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  S. A.R. Kort, F. Martens, H. Vanpoucke, H. L. van Duijnhoven, and M. A. Blankenstein Comparison of 6 Automated Assays for Total and Free Prostate-Specific Antigen with Special Reference to Their Reactivity toward the WHO 96/670 Reference Preparation Clin. Chem., August 1, 2006; 52(8): 1568 - 1574. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Semjonow, F. Oberpenning, C. Weining, M. Schon, B. Brandt, G. De Angelis, A. Heinecke, M. Hamm, P. Stieber, L. Hertle, et al. Do Modifications of Nonequimolar Assays for Total Prostate-specific Antigen Improve Detection of Prostate Cancer? Clin. Chem., August 1, 2001; 47(8): 1472 - 1475. [Full Text] [PDF]
|
|
|
|
|
|
R. J. Laffin, D. W. Chan, M. J. Tanasijevic, G. A. Fischer, W. Markus, J. Miller, P. Matarrese, L. J. Sokoll, D. J. Bruzek, J. Eneman, et al. Hybritech Total and Free Prostate-specific Antigen Assays Developed for the Beckman Coulter Access Automated Chemiluminescent Immunoassay System: A Multicenter Evaluation of Analytical Performance Clin. Chem., January 1, 2001; 47(1): 129 - 132. [Full Text] [PDF]
|
|
|
|
|
|
M. P. Fox, A. A. Reilly, and E. Schneider Effect of the Ratio of Free to Total Prostate-specific Antigen on Interassay Variability in Proficiency Test Samples Clin. Chem., August 1, 1999; 45(8): 1181 - 1189. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Nagasaki, M. Watanabe, N. Komatsu, T. Kaneko, J. Y. Dube, T. Kajita, Y. Saitoh, and Y. Ohta Epitope Analysis of a Prostate-specific Antigen (PSA) C-Terminal-specific Monoclonal Antibody and New Aspects for the Discrepancy between Equimolar and Skewed PSA Assays Clin. Chem., April 1, 1999; 45(4): 486 - 496. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. D. Cheli, M. Marcus, J. Levine, Z. Zhou, P. H. Anderson, D. D. Bankson, J. Bock, S. Bodin, C. Eisen, M. Senior, et al. Variation in the Quantitation of Prostate-specific Antigen in Reference Material: Differences in Commercial Immunoassays, Clin. Chem., July 1, 1998; 44(7): 1551 - 1553. [Full Text] [PDF]
|
|
|
|
|
|
P. C. Tewari, J. S. Williams, A. B. Blase, and R. L. Sokoloff Analytical Characteristics of Seminal Fluid PSA Differ from Those of Serum PSA Clin. Chem., January 1, 1998; 44(1): 191 - 193. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
