| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Technical Briefs |
1 The University of Michigan, Clinical Ligand Assay Satellite Services [CLASS Laboratory and Study of Women’s Health Across the Nation (SWAN)], 1301 Catherine St., Medical Science I, M3232A, The University of Michigan, Ann Arbor, MI 48109;
2
Bayer Diagnostics, 333 Coney St., East Walpole, MA 02032
aaddress correspondence to this author at: Department of Pathology, Medical Science I, M3232A, The University of Michigan, Ann Arbor, MI 48109-0602; fax 734-936-7583, e-mail bengland{at}umich.edu
The availability of rapid, non-extraction-based estradiol-17ß immunoassays has been important in clinical settings where prompt turnaround time is required, including in vitro fertilization programs, and in large-scale research programs where large numbers of samples are processed. However, the inaccuracy of estradiol-17ß assay results has been recognized for some time, as indicated in the Textbook of Reproductive Medicine by L.R. Boots (1), in which he states: "Everyone measures estradiol levels and it is probably assumed that few if any problems exist with this assay.... [T]he most common use of estradiol levels is in relation to levels of estradiol during ovulation stimulation and every assay provides clinically relevant results. These results are clearly inaccurate but usually precise" (1). In the First and Second E2 International Workshops (2), problems associated with the analysis of estradiol-17ß were fully discussed. The First E2 International Workshop focused on the need for more specific, precise, and sensitive estradiol-17ß assays, and the Second E2 International Workshop explored various means of increasing accuracy in the measurement of estradiol-17ß. By consensus, the use of a panel of 22 samples confirmed by isotope dilution–gas chromatography–mass spectrometry (ID-GC-MS) (2) was made available for use in assay development. Indications of a lack of accuracy are also provided by examining survey programs, such as those offered by the College of American Pathologists, that reveal large differences in estradiol-17ß values obtained on the same sample.
Modern serum assays for estradiol-17ß claim detection limits as low as 40 pmol/L or lower, but they lack sufficient sensitivity for pediatric or postmenopausal specimens and are frequently unable to provide adequate analytical precision and accuracy in samples from human males. In postmenopausal women, circulating estradiol-17ß is usually <73 pmol/L, and concentrations 
18 pmol/L are common. Circulating estradiol-17ß in men is somewhat higher, but concentrations >147 pmol/L are rare. In prepubertal children, estradiol-17ß is even lower than in postmenopausal women (3). We describe a new method that uses ID-GC-MS-confirmed calibrators with a sensitivity requisite to measure estradiol-17ß in samples with low concentrations of this analyte.
The assay was developed on an ACS-180 system equipped with the manufacturer’s software and using E2-6 reagents supplied by Bayer Diagnostics. Calibrators and antibody were diluted in Bayer Multidiluent Buffer (MD-9) before the incubation scheme described below. Briefly, assay sensitivity is increased by allowing the estradiol-17ß in 400 µL of a sample (calibrator):MD-9 (1:1) mixture to bind to antibody in 400 µL of an anti-E2-6:MD-9 (1:3) mixture during a 30-min (37 °C) offline incubation. After the incubation, 200 µL of the calibrator or sample-antibody mixture is aspirated by the sample probe and added to the onboard incubation cuvette. Estradiol-17ß, labeled with dimethyl-acridinium ester (E2-DMAE), and mouse anti-rabbit immunoglobulin G conjugated to superparamagnetic particles (PMPs) are added to the incubation cuvette at probe 2. The E2-DMAE occupies free binding sites on the anti-E2-6, and the PMP-bound mouse anti-rabbit immunoglobulin G immobilizes the antibody. The immobilized anti-E2-6-labeled estradiol complex is pulled to the sides of the incubation cuvette by a set of rare earth magnets, and unbound E2-DMAE is washed out of the cuvette. Emission of light by the PMP-antibody-bound acridinium ester trace is then induced by the addition of hydrogen peroxide at a low pH to the cuvette, followed by the addition of sodium hydroxide, which releases the peroxide anion and initiates the light emission reaction. An inverse relationship exists between estradiol-17ß concentrations in the sample and relative light units detected by the ACS-180. Calibrators containing estradiol-17ß at 1.8, 3.7, 7.3, 14.7, 29.4, 45.9, 91.8, 183.5, 367.1, and 917.8 pmol/L were prepared from the Bayer estradiol-17ß Master Curve Material. Data analysis was performed using the StatLIA® statistical ligand immunoassay analysis software (Brendan Scientific Corp.) with the StatLIA four-parameter symmetrical logistic function.
The accuracy of the assay was evaluated by comparing two ID-GC-MS-confirmed samples, samples 19 and 20 (2), with Bayer-supplied estradiol-17ß calibration material. Within-run imprecision was evaluated by comparing results for samples obtained from premenopausal, perimenopausal, and postmenopausal women, assayed in duplicate in one of four independent assays. The reproducibility of the method was assessed by repeated measurements of low concentrations of estradiol-17ß in male serum. Five replicates of diluted (1:2) and undiluted samples were analyzed in duplicate in each of four different assays (20 duplicate results for each individual). The Fisher LSD (Protected t-test) statistical procedure revealed significant between-subject differences (P <0.001) but no significant within-subject differences. Aldosterone, cholesterol, cortisone, cortisol, danazol, dehydroepiandrosterone, dehydroepiandrosterone sulfate, estradiol, estradiol glucuronate, estriol, estrone, progesterone, corticosterone, testosterone, norethindrone, pregnenolone sulfate, 11-deoxycortisol, 17
-hydroxy-pregnenolone, and 17-hydroxyprogesterone at concentrations of 104–105 pmol/L demonstrated no significant cross-reactivity under assay conditions (<0.001% cross-reactivity). Recovery studies performed by adding 18, 92, and 184 pmol/L estradiol-17ß to MD-9 yielded a mean recovery of 104% (97%, 97%, and 117%, respectively), with the greatest variability observed on the addition of 18 pmol/L.
The clinical utility of this assay was confirmed by comparing daily serum estradiol-17ß concentrations in 11 cycling women (Fig. 1
) with daily urinary estrone conjugate concentrations measured by the method of Munro et al. (4). Daily results were indexed by the day of peak serum luteinizing hormone to permit comparison of estrone and estradiol-17ß results during luteal and follicular phases of the menstrual cycle.
|
The estradiol calibration curve (1.8–918 pmol/L) was linear on a logit-log plot, and the 50% point of 102 pmol/L (75 assays) had a CV of 6.1%. All points were significantly different from each other (ANOVA, P <0.001), and the lowest calibrator (1.8 pmol/L) was significantly different (P <0.01) from the 0.0 pmol/L calibrator.
Results from the two ID-GC-MS-confirmed samples (samples 19 and 20) compared favorably with calibrators from Bayer Diagnostics. Measured results for sample 19 (367 pmol/L) were 93%, 99%, and 91% of expected at dilutions of 1:1, 1:2, and 1:4, respectively. For sample 20 (367 pmol/L), they were 96%, 99%, and 94%, respectively, at the same dilutions.
Within-run CVs were <10% except at extremely low concentrations. Grouping of samples in 36.7 pmol/L bins revealed that all bins had CVs <10%, except in the 0.0–36.7 pmol/L bin, where the CV was 12.5%. ANOVA confirmed the uniformity of variance in all other bins. Cross-reactivity was <1% for estriol (0.28%), estrone (0.75%), 17-estradiol (0.04%), estradiol glucuronide (0.09%), and aldosterone (0.05%). All others showed <0.009% cross-reactivity in the assay. The low concentrations of estriol and estrone found in the serum of nonpregnant women render the cross-reactivity of these steroids inconsequential in the current assay. Recovery of estradiol-17ß added to serum pools with measured concentrations of 0, 20.93, and 102.42 pmol/L averaged 104%. Inter- and intraassay imprecision, evaluated with quality-control materials, showed similar CVs across the range of the assay with higher CVs at low concentrations of estradiol-17ß. In most cases the CVs were <10%; even in postmenopausal serum and serum from men, in which measured concentrations were <33 and <76 pmol/L, respectively, the CVs were <12.5%.
Serum estradiol-17ß and urinary estrone conjugates showed the cyclical changes expected in women with normal menstrual cycles: a dramatic increase in circulating concentrations during the midcycle preovulatory peak and secondary increases during the luteal and follicular phases. Creatinine-adjusted urinary estrone conjugates lagged behind the circulating concentration of the parent steroid by 
1 day throughout the cycle and appeared to lag by 
2 days during the midcycle increase in estradiol-17ß. The cause of the additional lag time noted during the preovulatory surge of luteinizing hormone and estradiol-17ß is not known.
Variability in estradiol-17ß measurement, as cited in the literature by Lee et al. (5) and revealed by a careful examination of the College of American Pathologists survey results, confirms that the performance characteristics of direct immunoassays for estradiol-17ß need to be improved, particularly with respect to analytical accuracy and detection limit. The use of weighed calibrators or the availability of a panel of samples with known concentrations of estradiol-17ß (2) could enhance accuracy of the various estradiol-17ß immunoassay methods. Assay sensitivity is particularly important in the measurement of estradiol-17ß in normal serum samples in which physiologic concentrations of total estradiol-17ß can be expected to range between 4 and 367 pmol/L.
Most commercially available estradiol-17ß immunoassays today have detection limits between 73 and 183 pmol/L (5). This makes it difficult or impossible to measure the circulating concentrations (6)(7)(8) of this analyte in postmenopausal women, men, and children The sensitive estradiol-17ß immunoassay method described here fully meets our need for rapid, accurate, and precise analysis of serum estradiol-17ß. The assay procedure uses an antibody with high-specificity binding characteristics, has no extraction requirements, and preserves most of the advantages of automation.
In summary, we report the development of a highly sensitive and precise semiautomated immunoassay that is capable of measuring estradiol-17ß concentrations in women during the perimenopausal and postmenopausal intervals and in healthy men and children.
Footnotes
1 current address: Future Diagnostics, 23 Brewster Rd., Arlington, MA 02476 
References
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  R. C. Thurston, M. R. Sowers, B. Sternfeld, E. B. Gold, J. Bromberger, Y. Chang, H. Joffe, C. J. Crandall, L. E. Waetjen, and K. A. Matthews Gains in Body Fat and Vasomotor Symptom Reporting Over the Menopausal Transition: The Study of Women's Health Across the Nation Am. J. Epidemiol., September 15, 2009; 170(6): 766 - 774. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. R. Sowers, H. Zheng, D. McConnell, B. Nan, S. D. Harlow, and J. F. Randolph Jr. Estradiol Rates of Change in Relation to the Final Menstrual Period in a Population-Based Cohort of Women J. Clin. Endocrinol. Metab., October 1, 2008; 93(10): 3847 - 3852. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. S. Lee, B. Ettinger, F. Z. Stanczyk, E. Vittinghoff, V. Hanes, J. A. Cauley, W. Chandler, J. Settlage, M. S. Beattie, E. Folkerd, et al. Comparison of Methods to Measure Low Serum Estradiol Levels in Postmenopausal Women J. Clin. Endocrinol. Metab., October 1, 2006; 91(10): 3791 - 3797. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. R. Sowers, M. Jannausch, D. McConnell, R. Little, G. A. Greendale, J. S. Finkelstein, R. M. Neer, J. Johnston, and B. Ettinger Hormone Predictors of Bone Mineral Density Changes during the Menopausal Transition J. Clin. Endocrinol. Metab., April 1, 2006; 91(4): 1261 - 1267. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. E. Swain, R. L. Dunn, D. McConnell, J. Gonzalez-Martinez, and G. D. Smith Direct Effects of Leptin on Mouse Reproductive Function: Regulation of Follicular, Oocyte, and Embryo Development Biol Reprod, November 1, 2004; 71(5): 1446 - 1452. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. W. Ayres, D. W. Carr, D. S. McConnell, R. W. Lieberman, and G. D. Smith Expression and Intracellular Localization of Protein Phosphatases 2A and 2B, Protein Kinase A, A-Kinase Anchoring Protein (AKAP79), and Binding of the Regulatory (RII) Subunit of Protein Kinase A to AKAP79 in Human Myometrium Reproductive Sciences, October 1, 2003; 10(7): 428 - 437. [Abstract] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
, arithmetic mean of 10 matched sets of urine samples collected daily throughout the cycle and assayed for the major urinary estrone metabolite, estrone glucuronide. LH, luteinizing hormone. Error bars, SD.