| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Point-of-Care Testing |
1 Department of Biotechnology, University of Turku, Turku, Finland. 2 Innotrac Diagnostics Oy, Turku, Finland.
aAddress correspondence to this author at: Department of Biotechnology, University of Turku, Tykistökatu 6A, 6th Floor, FIN-20520 Turku, Finland. Fax 358-2-333-8050; e-mail piia.vonlode{at}utu.fi.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Methods: We developed a simple, dry-reagent, all-in-one immunoassay for the quantitative measurement of hCG in whole blood, plasma, or serum. The noncompetitive assay equally measures intact, nicked, and hyperglycosylated hCG as well as nonnicked and nicked hCG ß-subunit with a rapid and simple procedure consisting of a 5-min, one-step incubation and, subsequent to washing, the measurement of time-resolved fluorescence directly from a wet well surface.
Results: The assay had a detection limit (background + 3 SD) of 0.4 IU/L hCG. The within-run CV was <15% down to 2 IU/L, and the assay was linear to 6000 IU/L. The within- and between-run CVs in heparinized whole blood and plasma were 
10% throughout the measured range (4.0–4400 IU/L). The mean (95% confidence interval) difference between whole blood and plasma was –42 (–24 to –61)% without hematocrit correction and 6.5 (–14 to 27)% with hematocrit correction (n = 106). Regression analysis with the Diagnostic Products IMMULITE® 2000 hCG method yielded the following: slope (SD), 1.02 (0.01); y-intercept (SD), –6 (10) IU/L; Sy|x = 99 IU/L (n = 124; range, 1.6–4746 IU/L; r = 0.995).
Conclusions: Combined with the fully automated instrumentation, the 5-min, dry-reagent assay allows quantitative and reproducible determination of hCG in whole blood while sustaining the speed and simplicity of conventional rapid assays.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
and ß, with the -subunit similar in structure to that of the corresponding subunits of the other glycoprotein hormones, luteinizing hormone, follicle-stimulating hormone, and thyroid-stimulating hormone (TSH). After conception, the hCG concentration in maternal serum increases rapidly, with a doubling time of 1.5 days during the first weeks (3).
Although immunoassays specific for intact hCG are available, most commercial assays detect the ß-subunit of the hormone (hCGß; either free or bound to the -subunit), which dictates the specific biological activity of the intact hormone. The ability of hCG assays to also detect some irregular hCG molecular forms has become the subject of increasing discussion. Proper detection of hyperglycosylated hCG and the nicked forms of both hCG and hCGß is considered increasingly important because in early pregnancy and in trophoblastic disease these hCG variants may be the principal forms present in circulation (4)(5).
Simple and rapid hCG assays have been developed for both POC and home testing (6)(7)(8), and the use of these and other rapid tests is steadily increasing. Several manufacturers provide automated readers to decrease operator-related variability (9)(10), and completely automated POC immunoanalyzers (11)(12) have also been made commercially available, providing fully quantitative test results with wide measurement ranges, low detection limits, relatively short (
15 min) turnaround times, and most importantly, analytical quality similar to that of central laboratory methods. To allow fuller use of the relatively costly instruments, it would be desirable for them to measure multiple analytes, such as hCG, C-reactive protein (13)(14), prostate-specific antigen (15), and TSH (16), in addition to the currently prevailing cardiac markers (11). For hCG, assay sensitivity need not be extremely high for conventional pregnancy detection, but sensitive and quantitative assays can also contribute dramatically to the prompt diagnosis and management of urgent disorders, such as ectopic pregnancy in patients with acute abdominal pain or vaginal bleeding (17)(18), as well as threatened abortion and gestational trophoblastic disease (4)(19)(20). However, women have been diagnosed with ectopic pregnancies or cancer based solely on false-positive hCG results, with devastating consequences (21)(22)(23). The occasional false-positive results produced by many commercial hCG assays are considered a major problem.
We have described (24) a novel nonadentate europium chelate that facilitates sensitive and reproducible measurements without need for signal enhancement and development steps, including the drying of the assay surface before measurement. Here we describe a quantitative, dry-reagent hCG assay that uses the novel chelate as label in combination with direct time-resolved fluorescence measurement from a wet assay well surface.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
3.0 pmol. Serum and heparin- or EDTA-anticoagulated whole-blood and plasma samples were obtained from volunteers of the staff at the Department of Biotechnology, University of Turku. Before their use in the precision, linearity, recovery, and correlation studies, we added variable volumes of pooled human pregnancy plasma (pool concentration, 47 540 IU/L hCG). The serum samples used for the method-comparison study were kindly provided by Dr. Henrik Alfthan (Helsinki University Central Hospital, Helsinki, Finland), where the hCG concentrations had been determined with the Diagnostic Products Corporation IMMULITE® 2000 hCG assay (27).
The plasma and serum samples used in this study were either analyzed fresh or stored frozen at –20 or –70 °C until use (no longer than 1 month). The samples were stable in both freezing temperatures. All whole-blood samples were analyzed fresh. The hematocrit (Hct) values for the whole-blood samples were determined by centrifugation (Biofuge Haemo; Heraeus Instruments). All procedures followed in this study were in compliance with the Helsinki Declaration of 1975 as revised in 1996.
other reagents
The monoclonal anti-hCGß capture (clone 5008) and detection (clone 057-10043) antibodies were acquired from Medix Biochemica and OEM Concepts, respectively. Cross-reactivity with other glycoprotein hormones was studied by use of luteinizing hormone, follicle-stimulating hormone, and TSH (all from Scripps Laboratories). The Heterophilic Blocking Tubes were purchased from Scantibodies Laboratory Inc. All other reagents used were of analytical grade.
labeling of antibodies
The detection antibody was labeled with europium in 0.1 mol/L borate buffer, pH 9.0, with a 200-fold molar excess of the nonadentate [2,2',2'',2'''-{[2-(4-isothiocyanatophenyl)ethylimino]bis(methylene)bis{4-{[4-(-galactopyranoxy)phenyl]ethynyl}pyridine-6,2-diyl}bis(methylenenitrilo)}tetrakis(acetato)] europium(III) chelate, which was synthesized as described previously (24). The reaction was carried out overnight at room temperature. The labeled antibody was separated from excess free chelate on a Superdex 200 HiLoad 26/60 gel filtration column (Pharmacia Biotech) with Tris-saline-azide buffer, pH 7.75, as elution buffer. The fractions containing the antibody were pooled, and the europium concentration was measured against a europium calibrator. The labeling degree of the antibody was 4.5 Eu3+ molecules per IgG molecule.
The capture antibody was biotinylated in 50 mmol/L NaHCO3, pH 9.8, by use of a 30-fold molar excess of the biotinylation reagent biotin isothiocyanate (provided by J. Rosenberg, Department of Biotechnology, University of Turku, Turku, Finland), which was first dissolved in a small volume of dimethylformamide. The reaction was carried out for 2 h at room temperature. The excess free biotinylation reagent was removed by use of a Superdex 200 HR 10/30 (Pharmacia Biotech) gel-filtration column with the Tris-saline-azide buffer, pH 7.75, for elution.
Finally, bovine serum albumin was added to a concentration of 1.3 g/L to the solutions containing the europium-labeled and biotinylated antibodies. The antibodies were filtered through 0.22 µm pore size disposable filters and stored at 4 °C.
preparation of the all-in-one dry-reagent wells
The assay-specific dry-reagent wells were prepared exactly as described previously (24). In brief, 400 ng of biotinylated capture antibody in a 50-µL volume was added to previously prepared streptavidin-coated wells. The wells were incubated overnight at room temperature and washed. The capture antibody was then covered with 40 µL of a protective solution, and the wells were dried overnight in a dry-condition cabinet. The europium-labeled detection antibody (200 ng/well) was applied on top of the protective layer in a 1-µL volume and immediately dried with a stream of warm air. The dry reagent wells were stored and protected from humidity at 4 °C until use. According to our stability studies, the wells are stable at 4 °C for at least 1 year.
immunoassay for HCG
The noncompetitive, one-step immunoassays for hCG were performed on a fully automated Innotrac Aio! Immunoanalyzer (Innotrac Diagnostics), which allows time-resolved fluorescence measurement directly from a solid phase. In the automated assay procedure, 10 µL of sample and 20 µL of the combined assay/wash buffer (5 mmol/L HEPES, 9 g/L NaCl, 0.1 mmol/L EDTA, 0.055 g/L Tween20, 1 g/L Germall II, 0.5 mmol/L CaCl2, pH 7.5) were dispensed in the dry reagent well and incubated for 5 min with shaking at 36 °C. Samples containing >6000 IU/L hCG were diluted 1:100 in the above-mentioned buffer before being assayed. After incubation, the wells were washed and aspirated a total of six times, with the last aspiration adjusted so that 35 ± 5 µL of the assay/wash buffer was left in the wells. The measurement of europium fluorescence from the well surface was carried out using the default measurement settings of the immunoanalyzer: excitation wavelength, 340 nm; emission wavelength, 615 nm; delay time, 400 µs; window time, 400 µs, cycling time, 1 ms; measurement time, 1 s (i.e., counts resulting from 1000 sequential excitations were integrated for each measurement).
comparison method
The Diagnostic Products Corporation IMMULITE 2000 hCG assay (27) was used as the comparison method in this study. According to the manufacturer, the noncompetitive chemiluminescent assay measures various forms of the hCG molecule, including intact, nicked, and hyperglycosylated hCG and nonnicked and nicked hCGß. The detection limit of the assay is 0.4 IU/L, with a calibration range up to 5000 IU/L. The assay was performed according to the manufacturer’s instructions, with 1:20 dilution of samples containing >5000 IU/L hCG.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
80% of steady state was reached in 5 min, which was chosen as the incubation time for the current assay. Whole-blood and plasma samples had virtually the same kinetics as the calibrators. Similarly, the assay kinetics were consistent throughout the linear assay range (concentrations from 3.1 to 4945 IU/L tested).
assay specificity and interference
Cross-reactivities with hyperglycosylated hCG, nicked hCG, free hCGß, nicked free hCGß, and hCGß-core fragment were determined against both the 4th hCG IS and the 1st hCG RR standard preparations (Table 1
). The measured concentrations of the two hCG standards differed slightly from each other; therefore, the percentages of cross-reactivity calculated for the different hCG variants were also dependent on the standard material used. However, because all of the studied hCG metabolites except the hyperglycosylated hCG were from the 1st RR series, the cross-reactivities calculated against the 1st hCG RR were likely to give more accurate results, showing almost equal detection of all molecular forms studied expect for the hCGß-core fragment, which was not recognized by the assay. On the basis of the above results, however, the current assay calibrated against the 4th IS will produce concentrations for unknown samples 20% higher than if the samples were measured against the new standard preparation. However, considering that most existing commercial assays are calibrated against the IS preparations as well, no actions were taken to recalibrate the 5-min assay.
|
Cross-reactivities were <0.002% with the other members of the glycoprotein hormone family, luteinizing hormone (measured up to 100 IU/L), follicle-stimulating hormone (measured up to 100 IU/L), and TSH (measured up to 500 mIU/L). No interference was seen with triglycerides up to 10 g/L and bilirubin up to 200 mg/L.
calibration curve and high-dose hook effect
A typical calibration curve for the 5-min hCG assay is shown in Fig. 1
together with a within-assay precision profile. The analytical detection limit of the assay, calculated as the mean background signal + 3 SD, was 0.38 IU/L. The mean (SD) background [196 ( (24)) fluorescence counts] was calculated by use of three apparently hCG-free heparinized whole-blood and plasma samples, each measured in 12 replicates on 2 separate days (total n = 144). The calibration curve was linear at least up to 6000 IU/L hCG, i.e., over more than four orders of magnitude.
|
A high-dose hook effect was observed at 100 000 IU/L hCG. For fully quantitative results, samples containing (or expected to contain) >6000 IU/L hCG were diluted 1:100 before assay. However, samples with hCG adding up to 200 000 IU/L and assayed without predilution always gave a calculated concentration >6000 IU/L. This means that in all applications other than cancer diagnostics, no false-negative results will occur even in the absence of dilution and that only samples giving results >6000 IU/L need to be diluted and assayed again. However, in patients with choriocarcinoma, hCG concentrations exceeding several million units per liter have occasionally been measured (28)(29). Although undiluted samples with concentrations as high as 1 000 000 IU/L still give a calculated result >1000 IU/L in the current assay, the risk of aberrant or false-negative hCG results is increased in patients suspected of having cancer, and such samples should preferably be measured diluted as well.
linearity and limit of quantification
To five heparinized whole-blood samples from males with hCG concentrations below the detection limit of the assay, we added different amounts of pooled endogenous hCG and diluted them 3- to 729-fold with the original whole-blood fractions as a diluent (Fig. 2
). The assay was linear (r = 0.998–1.000) throughout the measured range (0.26–453 IU/L; Fig. 2A
).
|
To assess the limits of quantification, we calculated CVs for the whole-blood samples with added hCG and studied the within-run imprecision profile of the lowest hCG concentrations (0.26–12 IU/L) in more detail (Fig. 2B
). As expected, the variation was highest for the samples containing very low amounts of hCG, with the CV reaching 36% at 0.26 IU/L (n = 6). The limit of quantification, defined as the lowest concentration measured with a CV 15%, was estimated to be 2 IU/L.
imprecision and analytical recovery
The precision and recovery of the rapid hCG assay were studied by adding known amounts of pooled endogenous hCG to apparently hCG-free heparinized whole-blood and plasma samples. The within- and between-assay imprecision was similar for whole-blood and plasma samples, ranging from 5.9% to 10% at the lower (4.0–19 IU/L) and from 4.5% to 7.3% at the higher (440–4379 IU/L) hCG concentrations (Table 2
). Although samples containing <4 IU/L hCG were not included in the precision studies, the earlier estimate of 2 IU/L for the limit of quantification (CV 15%; see Fig. 2B
) was calculated to be close to the accurate value based on the precision studies as well. The analytical recoveries for hCG concentrations of 2–4600 IU/L in whole blood and plasma were in the range of 92–109% and 90–102%, with mean recoveries of 101% and 95%, respectively (n = 30 per sample matrix).
|
correlation between whole blood and plasma
We studied the comparability of measurements in whole blood and plasma by adding different amounts of endogenous hCG to heparinized whole-blood samples taken from 15 apparently healthy male volunteers (mean of seven different hCG concentrations added to each sample). In the original 15 samples, the mean Hct value was 0.44 with relatively little variation among samples (range, 0.41–0.48). To increase the Hct range of the samples, we manipulated the plasma volume in 23 of 106 tests in such a way that in the end the samples variably contained somewhat less or more plasma than the original samples. The resulting mean Hct value of 0.45 was very close to the original but with a range now covering Hct values from 0.29 to 0.70.
Difference analysis of the two sample matrices (hCG range, 2.8–4235 IU/L in plasma; n = 106) was performed both with and without correction for the individual (including modified) Hct values (Fig. 3
). For whole-blood hCG concentrations not corrected for Hct, the mean difference compared with plasma was strongly negative as could be expected based on plasma displacement by blood cells, thus yielding lower concentrations measured in whole blood compared with plasma (mean difference, –42%; 95% confidence limits, –61% and –24%). The mean difference for whole-blood measurements corrected by use of the individual Hct values was 6.5% compared with plasma (95% confidence limits, –14% and 27%). The slightly positive difference was very likely in part attributable to subjective visual reading of the Hct capillaries.
|
Although measurements in whole-blood samples obviously provide the most accurate results when the results are corrected by use of individual Hct values, determination of Hct may not always be feasible in POC conditions. In these conditions, a predetermined mean Hct value from a certain reference group could possibly be used for Hct correction instead of an individually determined value (11)(14). To analyze the difference between these two correction methods, we used the mean Hct of 0.45 to correct the measured concentrations of all 106 samples. Compared with corrections based on individual Hct values, the mean difference was –0.7% with 95% confidence limits of –27% and 26%. The samples that were manipulated to have very low Hct values (range, 0.29–0.35; n = 7) gave the most strongly positive difference (19–30%). Correspondingly, samples manipulated to have extremely high Hct values (range, 0.61–0.70; n = 10) gave the most strongly negative difference (–29% to –45%). However, considering that the current assay is mostly to be used in women of child-bearing age, such high Hct values are not likely to occur in the target population.
For serum and EDTA whole-blood and plasma samples, linear regression analyses [in the assay with a 15-min incubation time and dry-mode detection (24)] gave the following equations: (EDTA whole blood) = 0.984 (heparin whole blood) + 2.850 IU/L (r = 0.999); (serum) = 0.983(heparin plasma) + 1.839 IU/L (r = 0.999); n = 45; range, 0.4–38 400 IU/L hCG (1:100 dilution when hCG is >6000 IU/L; n = 11), with individual Hct values used for correction of the whole-blood samples. Although the matrix correlation study was performed with our earlier 15-min dry-measurement assay principle, we do not expect any further variation between the different sample matrices with the current procedure. The regression equation relating results of the 5- and 15-min assay procedures using heparinized whole blood had a slope of 0.925 (95% confidence interval, 0.90–0.94) and y-intercept of 2.9 (–41 to 47) IU/L (Sy|x = 138 IU/L; r = 0.994; n = 106; range, 2.9–4303 IU/L).
method comparison
We compared the 5-min hCG assay with the Diagnostic Products IMMULITE 2000 hCG assay, using serum specimens containing 1.6–43 928 IU/L hCG (n = 140). For the 5-min assay, samples containing >6000 IU/L hCG (n = 12) were diluted 1:100 before analysis. Similarly, samples containing >5000 IU/L hCG (n = 16) were diluted 1:20 before analysis in the IMMULITE 2000 hCG assay.
In Fig. 4A
, the results of the comparison study are plotted logarithmically to better assess the whole range of the measured samples. Linear regression analysis of the entire concentration range yielded a slope (SD) of 1.146 (0.007) and y-intercept (SD) of –135 (51) IU/L (Sy|x = 574 IU/L; r = 0.997; n = 140). Similar analysis of samples not diluted before either assay [i.e., samples with hCG concentrations 5000 IU/L; n = 124) yielded somewhat better results: slope, 1.02 (0.01); y-intercept, –6 (10) IU/L; Sy|x = 99 IU/L; r = 0.995; inset in Fig. 4A
], possibly indicating that the sample dilutions in either assay may have been slightly inaccurate.
|
The highest proportional differences between the two methods were seen in samples containing <10 IU/L hCG. The relative difference of the 5-min assay in these samples varied from –32% to 122% (n = 13; Fig. 4B
) compared with the IMMULITE 2000 assay, whereas difference analysis of samples with >10 IU/L hCG showed a mean relative difference of only –1.0% with 95% confidence limits of –27% and 25% (inset in Fig. 4B
). The confidence limits were somewhat widened by a single sample giving a concentration of 57 IU/L in the IMMULITE 2000 assay and 102 IU/L in the 5-min POC assay: without this sample, the mean difference was –1.6%, and the 95% confidence limits were –23% and 20%. However, we detected no evidence of assay interference despite thorough investigations, and most likely the deviation was attributable to different detection of some hCG variants by the two assays.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The common goal for these and other evolving miniaturized assay systems is the development of automated and quantitative immunoassay techniques to improve the analytical performance of POC testing. Considering the additional challenges caused by the demand for whole blood as sample material, short turnaround times, and for instrumentation to be small in size and easy to use, the development of such methods has been difficult, particularly for the more experimental assay concepts, including immunosensors and some other homogeneous assay formats, although these methods are rapid and relatively simple to perform.
In this study, the goal was to develop a simple and rapid POC assay that relied on traditional noncompetitive assay techniques similar to those incorporated in immunoassays used for routine laboratory testing. The method with the one-step assay principle was published in 1996 by Lövgren et al. (33), who also used hCG as one of the model analytes. This assay was performed manually, however, with the time-resolved fluorescence detected in a special measurement solution requiring an additional 3-min incubation step before measurement. Since then, automated assay methods with direct surface measurement from a dried assay surface have been introduced (11). In a recent study (24), we described a novel nonadentate chelate that enables the omission of all signal development steps, including drying. The fluorescence signal is measured directly from the wet well surface immediately after washing.
In the current study, we placed high emphasis on decreasing the number and duration of assay steps to a minimum while retaining the analytical performance typical of laboratory methods. Although in our earlier study we used an incubation time of 15 min (24), 80% of steady-state values could be achieved within 5 min. Although selecting this incubation time remarkably shortened the turnaround time of the assay, no decrease in assay performance was detected. The detection limits of the 5- and 15-min assays were 0.4 and 0.9 IU/L, respectively, with linear ranges spanning four orders of magnitude. Similarly, the effect on assay precision was insignificant, and the between-assay CVs remained 10% throughout the measured range (4–4400 IU/L) despite the threefold shorter reaction time.
Although in the current approach we have eliminated as many process steps as possible, the performance characteristics of the novel assay can nevertheless be directly compared with those of the Diagnostic Products IMMULITE 2000 hCG assay used as the comparison method in this study. The two assays have the same detection limit (0.4 IU/L), with the upper measurements ranges of the IMMULITE and the 5-min assays extending to 5000 and 6000 IU/L, respectively, without predilution of samples. Furthermore, the correlation between the two methods for serum was very good in this area. Prediluted samples originally containing >6000 IU/L yielded slightly higher results in the 5-min assay, however, possibly because of slight inaccuracies in sample dilution in either assay; however, even for these samples the mean difference was <10% (n = 12). For a POC assay, these performance characteristics are exceptionally good. Considering that the automated dispensing, washing, and measurement steps take 2 additional min with the current instrumentation, fully quantitative and reproducible results can be obtained with a total turnaround time of only 7 min. Moreover, a simpler and smaller instrument is under development.
The primary sample type for the assay is, for obvious reasons, unprocessed whole blood. Similar to our previous studies using stable lanthanide chelates as labels (14)(24), we observed no unexpected differences between measurements in whole blood, plasma, or serum, indicating the absence of interference from compounds present in blood (e.g., erythrocytes) or from chelating agents such as EDTA. Because of the volume taken up by erythrocytes, however, the comparability of concentrations measured in whole blood and plasma was expectedly dependent on correcting the results for the Hct. Although the need for Hct correction is apparent to obtain corresponding results for whole blood and plasma, the correction method requires further consideration, in particular because the means for determining Hct in POC settings may be limited. As shown in this and one of our earlier studies (14), a predetermined mean Hct value is a practical solution to the problem. However, the feasibility of this correction method depends on the degree of accuracy that is needed for the measurement and should be evaluated individually for each marker. Considering the very high variation in individual hCG concentrations and the purpose for which the results will be used, the allowable inaccuracy for measurements of this marker is usually rather high. Therefore, in our opinion, the use of a mean Hct value instead of an individually determined value for correction of whole-blood hCG concentrations is in most cases adequate.
A recognized and serious problem affecting hCG immunoassays is the occurrence of false-positive hCG results, mostly attributable to heterophilic or anti-animal antibodies (34)(35)(36). Despite increasing awareness of this phenomenon, patients with falsely increased hCG concentrations are sometimes misdiagnosed with choriocarcinoma or gestational trophoblastic disease after an intra- or extrauterine pregnancy has been excluded (21)(22)(23), and the false positivity may not be detected until after all treatment has failed to reduce the hCG concentrations. As a matter of fact, an initial version of the all-in-one assay concept also used an antibody pair that was soon found to be susceptible to producing false-positive results with a frequency of 0.7% (3 in 436 samples; data not shown). With this antibody pair, eliminating the false positivity by adding native or denatured mouse immunoglobulins or different commercial blockers directly to assay reagents turned out to be very ineffective. Preincubating a false-positive sample with the Heterophilic Blocking Tube blocking agent (Scantibodies Laboratories Inc.), however, decreased the interference 10-fold, from 76 IU/L to 7.6 IU/L. This blocker, designed to eliminate interference caused by heterophilic antibodies, has been shown to be effective in reducing false-positive hCG results in other studies as well (37)(38). On the basis of the above experience, however, we strongly feel that too much confidence should not be placed on animal immunoglobulins or other blocking agents routinely added to the assay reagents to prevent possible interference because reformulating the assay reagents may have surprisingly little effect when an antibody pair prone to interference is used. In the all-in-one assay format, the best results were obtained when we changed the antibody combination to the current one. Since that change we have had no false positives, but as with all hCG immunoassays, this possibility should be considered each time persistently increased hCG values are encountered in the absence of detectable pregnancy or tumor.
Despite the use of hCG determinations for the diagnosis and management of trophoblastic and nontrophoblastic malignancies (4)(20) and ectopic pregnancies (17)(18) as well as for other applications, such as screening for Down syndrome (39) and follow-up of hCG clearance after induced or spontaneous abortion (40) or assisted reproduction (41), the primary use of the hCG assays is the conventional detection and monitoring of pregnancy. The current assay was also developed primarily for the latter purpose and therefore designed to detect intact and hyperglycosylated hCG as well as free hCGß, which are the principal forms of hCG in blood during (early) pregnancy (5)(42)(43). However, it has recently been shown that many hCG assays have poorer sensitivity for hyperglycosylated hCG despite that fact that it is the major molecular form produced in the first weeks of pregnancy and therefore the key molecule in sensitive pregnancy detection (5)(8). Furthermore, the variable detection of other hCG variants in addition to hyperglycosylated hCG has been shown to be a major cause of discordance among commercial assays (28)(44), and not all tests are appropriate for use in applications other than for pregnancy testing. The current 5-min assay was, however, found to detect almost equally intact, hyperglycosylated, and nicked hCG as well as nonnicked and nicked hCGß, thus improving the sensitivity of the assay for early pregnancy detection and potentially expanding the utility of the assay to more sophisticated POC testing strategies, including the diagnosis and management of trophoblastic disease. Similarly, the low limit of quantification of the novel assay may also be useful in the detection and management of different pregnancy-related disorders.
In summary, the automated dry-reagent hCG assay described here is a robust and reproducible POC method that facilitates fully quantitative measurements of five common hCG-related molecules directly in whole blood. Compared with traditional laboratory assays, several process steps have been eliminated to provide a simple assay procedure that includes only the 5-min one-step incubation, washing, and direct measurement of time-resolved fluorescence from a wet well surface. Consisting of disposable dry-reagent assay wells and a fully automated immunoanalyzer, the assay format combines the advantages of single-use manual test devices with the advantages of conventional laboratory assays, providing a simple and rapid assay with high reproducibility, sensitivity, and a wide measurement range. Moreover, the amount of data obtainable per assay could potentially be multiplied by use of the sharp and nonoverlapping emission profiles of the other lanthanides (45)(46). Although the current moisture-tolerant chelate structure has not yet been applied to the other lanthanide ions, the possible development of corresponding chelates for simple multiplex applications could eventually facilitate full use of time-resolved fluorometry in POC settings as well.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
and ß subunits of human chorionic gonadotrophin for immunoassay. J Endocrinol 1980;84:295-310.[ISI][Medline]
[Order article via Infotrieve]
subunits of human chorionic gonadotropin (hCG) in pregnancy sera. J Clin Endocrinol Metab 1984;58:1200-1202.[Abstract]
-fetoprotein and free ß subunit of human chorionic gonadotropin by dual-label time-resolved immunofluorometric assay. Clin Chem 1993;39:2084-2089.[Abstract]
-fetoprotein, human chorionic gonadotropin and estriol in serum of pregnant women by time-resolved fluoroimmunoassay. J Pharm Biomed Anal 1999;20:169-178.[CrossRef][ISI][Medline]
[Order article via Infotrieve]The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  S. Wittfooth, Q.-P. Qin, J. Lund, I. Tierala, K. Pulkki, H. Takalo, and K. Pettersson Immunofluorometric Point-of-Care Assays for the Detection of Acute Coronary Syndrome-Related Noncomplexed Pregnancy-Associated Plasma Protein A Clin. Chem., September 1, 2006; 52(9): 1794 - 1801. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
U.-H. Stenman Improving Immunoassay Performance by Antibody Engineering Clin. Chem., May 1, 2005; 51(5): 801 - 802. [Full Text] [PDF]
|
|
|
|
|
|
R. Vikstedt, P. von Lode, T. Takala, K. Irjala, O. Peltola, K. Pettersson, and P. Suominen Rapid One-Step Immunofluorometric Assay for Measuring Soluble Transferrin Receptor in Whole Blood Clin. Chem., October 1, 2004; 50(10): 1831 - 1833. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
) for the 5-min hCG assay.
