| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Technical Briefs |
Departments of
1
Urology,
2
Clinical Chemistry and Laboratory Medicine,
3
Medical Informatics and Biomathematics, University of Münster, Albert-Schweitzer Strasse 33, D-48129 Münster, Germany
4
Department of Urology, Zentralklinikum Augsburg, D-86009 Augsburg, Germany
5
Department of Clinical Chemistry, Klinikum Grosshadern, Ludwig-Maximilians-University, Marchioninistrasse 15, D-81366 Munich, Germany
aauthor for correspondence: fax 49-251-834-8492, e-mail Semjono{at}uni-muenster.de
The major form of prostate-specific antigen (PSA) in serum is complexed to 
1-antichymotrypsin (1)(2). This form is present in a higher proportion of men with prostate cancer (PCa) than in men with benign prostate (2). The ratio between the unbound free form of PSA and total PSA is lower in patients with PCa. Assays for total PSA that preferentially detect free PSA overreport the total PSA in sera of patients with benign prostate compared with patients with PCa (3). Thus, it was recommended that assays for total PSA should measure the immunologically detectable forms of PSA on an equimolar basis (4). The clinical relevance of these considerations remains controversial (5)(6)(7)(8)(9)(10)(11), although differences between total PSA assays are well documented (3)(12)(13)(14)(15)(16)(17)(18), but it has yet to be discovered whether the molar response characteristics of total PSA assays have a clinically relevant influence on the clinical diagnostic performance of the assay. Recently, the Assay Comparison Study for PSA of the Working Group on Laboratory Diagnostics of the German Urological Association (19) evaluated 25 total PSA assays and 11 free PSA assays using identical serum samples. During the course of the study, three nonequimolar assays were identified, which subsequently underwent major modifications by the manufacturers, thus challenging us to test the performance before and after modification using aliquots of identical sera.
Serum aliquots of 338 men with total PSA concentrations of 2–30 µg/L (mean, 9.2 µg/L; median 7.5 µg/L; SD, 6.6 µg/L) as determined by the Tandem-E® (Hybritech) assay reagent set were selected from the Assay Comparison Study for PSA (19) in accordance with practices and ethical standards of the Committee on Ethical Issues of the University of Münster and the Declaration of Helsinki, including informed consent of the participants. All sera were validated by clinical data, and the prostate status of all men (age range, 45–88 years; mean, 64 years; median, 65 years) was confirmed by digital rectal examination and transrectal ultrasound. In men with a total PSA 
4 µg/L as determined by the Tandem-E reagent set or suspicion for PCa at digital rectal examination, prostatic status was confirmed histologically after prostate biopsy. The study population consisted of 62 men without clinical evidence of PCa, 110 patients with histologically benign prostate, and 166 patients with untreated PCa. The PSA concentrations as determined by the Tandem-E® reagent set in the 166 patients with PCa (mean, 12.2 patients; median, 11.1 patients; SD, 6.5 patients) and the 172 men without PCa (mean, 6.4 patients; median, 4.1 patients; SD, 5.3 patients) differed as expected.
All sera were aliquotted within 3 h and stored at -70 °C. Each aliquot underwent only one freeze-thaw cycle. After thawing, the samples were processed within 8 h. Determinations of free PSA with the Tandem-R free PSA reagent set and total PSA with the Tandem-E reagent set were performed to evaluate the equimolar or nonequimolar response characteristics of the assays under investigation. As defined by Graves (3), a nonequimolar-response assay reports quantities of free PSA as greater than the same amount of PSA complexed to 1-antichymotrypsin. For normalization of the assays and their modifications, the Hybritech assays were chosen because they have been investigated most intensively in this context (5)(7)(8)(9)(10)(16)(20).
Three assays were used to determine total PSA concentrations before (the Abbott AxSYM® PSA; the Chiron ACS® PSA 1; the bioMérieux Vidas® PSA) and three were used to determine total concentrations after [the Abbott AxSYM Total PSA (the AxSYM Total PSA is currently sold only outside the US); the Chiron ACS PSA 2; the bioMérieux Vidas TPSA] modifications by the manufacturers. Abbott replaced a polyclonal–monoclonal antibody combination with a monoclonal–monoclonal combination, whereas bioMérieux substituted the previous monoclonal–monoclonal antibody combination with a new monoclonal–monoclonal combination. Chiron did not change the polyclonal–monoclonal antibody format.
Clinical performance was evaluated by comparison of the areas under ROC curves (21) using the program MedCalc Version 6.0 (www.medcalc.be).
In accordance with the inclusion criteria, total PSA concentrations were 2–30 µg/L in the Tandem-E assay. The free/total PSA ratio was 0.02–0.49 (Tandem-R free/Tandem-E). All three investigated assays showed a marked nonequimolar response before modification (Fig. 1, A, C, and E
). After modification, the assays by Abbott and bioMérieux produced an equimolar response (Fig. 1, A and C
), whereas the assay modification by Chiron did not lead to equimolar assay characteristics (Fig. 1E
). In each assay, the modifications led to a decrease of the total PSA concentrations reported. Modification of the Abbott and bioMérieux assays produced a small but statistically significant (P 
0.001) improvement in the area under the ROC curve (Fig. 1B, D
); however, no improvement in the area under the ROC curve was observed for the modification of the Chiron assay (Fig. 1F
). At 90% sensitivity, the specificity of the Abbott assay increased from 48% (95% confidence interval, 40–55%) to 54% (95% confidence interval, 46–62%) and of the bioMérieux assay from 45% (95% confidence interval, 38–53%) to 52% (95% confidence interval, 44–59%). At 90% sensitivity, both assay reformulations protected 11 of the 172 patients without PCa from unnecessary biopsy.
|
There has been a long-lasting and intense discussion about the clinical relevance of equimolar or nonequimolar measurement of the immunologically detectable forms of PSA by various assays (3)(5)(6)(7)(8)(9)(10)(11)(16)(22). Although for theoretical considerations, equimolar assays should show superior performance in the discrimination between benign prostate and PCa, this has not yet been shown in a clinical setting (17). Attempts to standardize assays by recalibration with certain standard reference materials were made (23)(24), but could not be expected to transform nonequimolar into equimolar assay characteristics (25). Even in equimolar assays, the results of recalibration with such reference substances were shown to be disappointing (20)(26), although a reduction in interassay variability may be achieved to some extent (27).
The existing differences between various assays for PSA create substantial problems in the interpretation of PSA concentrations. Not only did the upper limit of the previous Chiron assay (ACS PSA 1) need to be nearly doubled to achieve the same specificity for the differentiation between benign prostate and PCa as the Hybritech assay (13), but high apparent PSA values were observed in sera with a high proportion of free PSA, i.e., sera from patients without PCa (16). Recently Brawer et al. (28) reported a substantial equivalence of the new recalibrated Chiron assay (ACS PSA 2) with the Hybritech method (Tandem-R PSA), on the basis of regression analysis in various ranges of total PSA concentrations and on comparison of the number of cancers that exceeded chosen total PSA decision limits in the investigated assays. With the use of a decision point of 4.0 µg/L, 1 of 44 patients with PCa would have been missed if the ACS PSA 2 was used instead of the Tandem-R assay. Unfortunately, the number of patients without PCa who would have undergone unnecessary biopsies at this decision point has not been reported because recalibration of a nonequimolar assay theoretically is a trade-off between too many patients with PCa being missed and too many men with benign prostate undergoing unnecessary biopsies.
In the present study, the assay modifications performed by Abbott and bioMérieux produced a slight improvement of diagnostic performance as determined by ROC analysis (Fig. 1, B and D
). The modification of the Chiron assay, which did not change the monoclonal–polyclonal antibody format, produced substantially lower total PSA concentrations found in identical samples, but did not improve the discriminative diagnostic power of the assay.
Although the diagnostic improvements gained by reformulation of the Abbott and bioMérieux assays were relatively small, they were statistically significant (P 0.001). These data provide justification for seeking equimolar assays over nonequimolar assays.
Acknowledgments
This study was financially supported by the producers of all the participating assays in equal parts. We greatly acknowledge the inexhaustible technical assistance of Francoise Flammang, Beate Pepping-Schefers, Sabine Plonka, Karin Hofmann, and Ashley Steele.
References
1-antichymotrypsin. Clin Chem 1991;37:1618-1625.1-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.The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  F. H. Jansen, M. Roobol, C. H. Bangma, and R. H. N. van Schaik Clinical Impact of New Prostate-Specific Antigen WHO Standardization on Biopsy Rates and Cancer Detection Clin. Chem., December 1, 2008; 54(12): 1999 - 2006. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Stephan, M. Klaas, C. Muller, D. Schnorr, S. A. Loening, and K. Jung Interchangeability of Measurements of Total and Free Prostate-Specific Antigen in Serum with 5 Frequently Used Assay Combinations: An Update Clin. Chem., January 1, 2006; 52(1): 59 - 64. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. W. Roddam, C. P. Price, N. E. Allen, and A. M. Ward Assessing the Clinical Impact of Prostate-Specific Antigen Assay Variability and Nonequimolarity: A Simulation Study Based on the Population of the United Kingdom Clin. Chem., June 1, 2004; 50(6): 1012 - 1016. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Wesseling, C. Stephan, A. Semjonow, M. Lein, B. Brux, P. Sinha, S. A. Loening, and K. Jung Determination of Non-{alpha}1-Antichymotrypsin-complexed Prostate-specific Antigen as an Indirect Measurement of Free Prostate-specific Antigen: Analytical Performance and Diagnostic Accuracy Clin. Chem., June 1, 2003; 49(6): 887 - 894. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
