Clinical Chemistry
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Clinical Chemistry 48: 269-277, 2002;
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Web of Science (21)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Tarkkinen, P.
Right arrow Articles by Lövgren, T.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Tarkkinen, P.
Right arrow Articles by Lövgren, T.
Related Collections
Right arrow Clinical Immunology
Right arrow Proteomics and Protein Markers
Right arrow Endocrinology and Metabolism
Right arrow Automation and Analytical Techniques
(Clinical Chemistry. 2002;48:269-277.)
© 2002 American Association for Clinical Chemistry, Inc.

Ultrarapid, Ultrasensitive One-Step Kinetic Immunoassay for C-Reactive Protein (CRP) in Whole Blood Samples: Measurement of the Entire CRP Concentration Range with a Single Sample Dilution

Piia Tarkkinen1a, Tom Palenius2b and Timo Lövgren1

1 Department of Biotechnology, University of Turku, Tykistökatu 6A, 6th Floor, FIN-20520 Turku, Finland.

2 PerkinElmer Life Sciences/Wallac Oy, PO Box 10, FIN-20101 Turku, Finland.

aAuthor for correspondence. Fax 358-2-333-8050; e-mail piia.tarkkinen{at}utu.fi.b Current address: AboaTech Ltd, FIN-20520 Turku, Finland.


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
Background: Recently, measurement of very low concentrations of C-reactive protein (CRP) has gained popularity as a potential new means for predicting the risk of future cardiac complications. In this study, we demonstrate the feasibility of a kinetic, one-step microparticle assay for quantitative determination of extremely low and high CRP concentrations in the limited timeframe typical for point-of-care testing.

Methods: A noncompetitive, kinetic CRP immunoassay was developed that uses individual, porous microparticles as the solid phase. The microparticles were covalently coated with a monoclonal capture antibody, and the monoclonal detection antibody was labeled with europium. The one-step binding reaction was stopped by washing after 2 min of incubation, and the fluorescence signal of individual particles was measured.

Results: The analytical detection limit (mean of zero calibrator + 3 SD) was 0.00016 mg/L CRP. Clinical samples were diluted 400-fold before assay to cover the CRP concentration range of 0.064–1200 mg/L. The assay correlated well with the Dade Behring N High Sensitivity CRP assay (for 0–10 mg/L, r = 0.969, Sy|x = 0.68, n = 54; for 0–350 mg/L, r = 0.969, Sy|x = 11.7, n = 100). The within- and between-run CVs based on calculated concentrations were, respectively, 9–16% and 14% at 0.11 mg/L, 4.5–12% and 8.2% at 4.2 mg/L, and 3.5–6.3% and 4.4% at 105 mg/L, with a CV <15% at 0.2 mg/L and above.

Conclusions: Use of the kinetic microparticle approach combined with time-resolved fluorometry allows ultrasensitive quantification of CRP in whole blood in 2 min with a linear assay range spanning more than four orders of magnitude.


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
The development of highly sensitive assays (1)(2)(3)(4)(5)(6) for C-reactive protein (CRP) 1 (7)(8) has rapidly increased recently because numerous studies (9)(10)(11)(12)(13)(14)(15)(16) have indicated the usefulness of CRP as a prognostic factor of future cardiac complications. In these studies, CRP concentrations have been shown to increase slightly years before onset of the actual disease. The association extends to apparently healthy individuals with no traditional symptoms of heart-related disease, emphasizing the significance of high-sensitivity CRP (hs-CRP) as a marker for long-term risk assessment. It has also been shown that cardiac panels combining the measurement of hs-CRP as well as the more common risk factors can better predict the future occurrence of a cardiovascular event than can models based solely on traditional prognostic markers (17)(18)(19). The clinical range of CRP concentrations useful for predicting the risk of future heart-related disease is, however, much lower than the range used in the traditional application of CRP as a highly sensitive marker of inflammation. Although the detection limit for the latter is usually ~5 mg/L, measurement of concentrations relevant to heart-related risk prediction requires a detection limit at least 10 times lower (1)(12), thus giving rise to a need for a new generation of CRP assays.

Quantitative measurement of CRP as a marker of inflammation is already among the most common tests performed in clinical laboratories (20), making it a strong candidate for point-of-care testing (POCT) as well. Consequently, the turnaround times of the assays have become increasingly important, and CRP assays suitable for POCT have been developed at a rapid pace (21)(22)(23). However, although the overall sensitivity of CRP assays has improved considerably because of the new applications described above, the present high-sensitivity assays facilitate CRP measurements only in clinical laboratory rather than in POC conditions. Correspondingly, the assays suitable for POCT generally use detection technologies that do not allow the detection of low analyte concentrations. Therefore, the present POC assays for CRP are well suited for the detection of, e.g., bacterial or viral infections or for follow-up of antiinflammatory therapy, but are inadequate for detecting CRP at the concentrations required for heart disease risk assessment.

In this study, we explored the possibility of using a kinetic one-step assay concept to develop a CRP assay capable of measuring a wide range of concentrations and suitable for POC situations. Porous microparticles served as a mobile, high-capacity solid phase to facilitate efficient collection of analyte molecules in a limited time. The high specific activities of lanthanide chelates (24)(25)(26) and the superior sensitivity and dynamic range of time-resolved fluorometry (27)(28)(29)(30) allowed sensitive detection of both very low and high analyte quantities in submicroliter volumes of whole blood with a single sample dilution. Because of the sensitive detection technology, results could be obtained long before the immunoassay reached equilibrium without compromising assay performance.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
microparticles
The immunoassays were performed with highly porous, 60-µm acrylate microparticles purchased from SINTEF Applied Chemistry. The monosized particles were polymerized by a swelling method (31) and contained active surface epoxy groups for covalent immobilization of antibodies. The radii of the particle pores varied from <5 to 500 nm, producing a total surface area of 5.4 µm2 per particle. According to the manufacturer, pores with a radius >30 nm accounted for ~83% of the total pore volume.

covalent coating of the microparticles
The epoxy-surfaced microparticles were covalently coated with a monoclonal, CRP-specific capture antibody purchased from Medix Biochemica. The coating was performed batchwise, using ~1 x 104 microparticles per reaction. For immobilization, 15 ng of antibody/particle was added in 24 nL of phosphate-buffered saline (8.0 g/L NaCl, 0.2 g/L KCl, 1.2 g/L Na2HPO4 · 7 H2O, and 0.2 g/L KH2PO4), pH 7.2. The reaction was incubated for 20 h at room temperature in a Rotamix mixer (Heto Lab Equipment). After coating, the particles were washed three times with 2 mmol/L Tris-HCl, pH 7.5, containing 0.1 g/L Tween 20. The coated particles were stored at 4 °C.

labeling of the detection antibody
A monoclonal CRP-specific antibody (purchased from HyTest Ltd.) was labeled with the intrinsically fluorescent Eu3+ chelate of 4-[2-(4-isothiocyanatophenyl)ethynyl]-2,6-bis{[N,N-bis(carboxymethyl)amino]methyl}pyridine (PerkinElmer Life Sciences/Wallac Oy). The reaction was carried out overnight at 4 °C with a 75-fold molar excess of the chelate in 50 mmol/L NaHCO3, pH 9.8. Excess label was removed by gel filtration on a Superdex 200 HR 10/30 column (Pharmacia Biotech) with Tris-saline-azide buffer (6.1 g/L Tris, 9.0 g/L NaCl, and 0.5 g/L NaN3), pH 7.5, as elution buffer. The fractions containing the antibody were pooled, and diethylenetriamine pentaacetic acid-treated bovine serum albumin (PerkinElmer Life Sciences/Wallac Oy) was added to a final concentration of 1 g/L. The labeling degree of the antibody was 5.5 Eu3+/IgG molecule.

crp calibrators and clinical samples
The CRP calibration material was acquired from HyTest Ltd. According to the manufacturer, the material had been produced from human pleural fluid or plasma, and its purity was >98%. Two hundred EDTA-whole blood samples collected for routine CRP measurements were obtained from the Turku University Central Hospital. The hematocrit (Hct) of each sample was determined by centrifugation (Biofuge Hemeo; Heraeus Instruments). From each of the 100 samples used in the correlation studies, a small volume (200 µL) of whole blood was set aside for the microparticle assay, and the remainder was centrifuged for 15 min at 2600g (Centrifuge 5804 R; Eppendorf) to separate the plasma. The whole blood samples were analyzed fresh, whereas the plasma samples were stored deep-frozen for 1 week at -70 °C before being assayed in the Dade Behring BN II nephelometer with the Dade Behring N High Sensitivity CRP assay. In the microparticle assays, DELFIA Assay Buffer (PerkinElmer Life Sciences/Wallac Oy) was used as the diluent for both the calibrators and clinical samples.

assay kinetics
All immunoassays were performed with individual microparticles coated with an anti-CRP capture antibody as solid phase. The kinetics of the different steps of the sandwich-type CRP immunoassay were determined using an assay volume of 330 nL/particle (3 particles/µL), the sample volume being one-third of the total volume. The assays were performed with CRP calibrator and detection antibody concentrations of 0.5 and 9.3 mg/L, respectively, and the concentrations were kept invariable for each assay condition. The binding reactions were performed in DELFIA Assay Buffer containing 3.5 g/L Tween 20. For practical reasons, the incubations were performed using a batch of 21 microparticles in an assay volume of 7 µL/tube. The reactions were incubated at room temperature in a vortex-type mixer (Scientific Industries) for different time periods (0–1150 min), after which the reaction was interrupted by washing twice with 200 µL of DELFIA Wash Solution (PerkinElmer Life Sciences/Wallac Oy). The particles were transferred and dried on a Cyclopore track-etched polycarbonate membrane (Whatman Polyfiltronics) to measure the signal of individual microparticles in a time-resolved microfluoro-meter (30). For each time point, four replica particles were measured.

In addition to the one-step kinetics (i.e., analyte and detection antibody added simultaneously in the reaction), the kinetics of the analyte and the detection antibody binding reactions were studied separately as well. These assays were performed essentially in the same way as above, but with the analyte and tracer incubated sequentially and the missing reagent being replaced by DELFIA Assay Buffer.

kinetic crp immunoassay
Effect of microparticle and detection antibody concentrations.
To examine the performance of a kinetic microparticle immunoassay, we studied the effect of microparticle concentration, using 1, 10, or 100 particles/µL (i.e., assay volume of 1000, 100, or 10 nL/particle, respectively). The total assay volumes were adjusted to 6–20 µL to facilitate manual handling of the reactions and to include at least 20 microparticles per reaction. Accordingly, the reactions were performed using the following three combinations: 20 particles/20 µL, 60 particles/6 µL, and 600 particles/6 µL. The detection antibody concentration (6.2 mg/L) and the ratio of sample volume to total volume (one-third) were kept constant for each assay condition. Experiments to assess the influence of different detection antibody concentrations were performed with 3 microparticles/µL (i.e., assay volume of 330 nL/particle). Three detection antibody concentrations, 6.2, 31.0, and 124.0 mg/L, were studied at variable analyte concentrations. All reactions were performed in a one-step format, i.e., by dispensing the detection antibody and the analyte simultaneously onto the particles. Otherwise, the same assay procedure as described under "Assay Kinetics" was used with an incubation time of 2 min.

Calibration curve and clinical samples.
To avoid variations in incubation times and temperatures that could distort the quantitative interpretation of the results, the clinical samples and the CRP calibrators used for generating the calibration curve were run together in same assay cycles. The 2-min assays were performed essentially in the same way as described above, with a total assay volume of 330 nL/particle and a detection antibody concentration of 6.2 mg/L. The analytical detection limit of the kinetic assay was calculated as the mean signal of the zero calibrator + 3 SD.

comparison method
The Dade Behring N High Sensitivity CRP assay was used as the comparison method in this study. The assay was performed in the Dade Behring BN II nephelometer with the manufacturer’s reagents as directed. According to the manufacturer, the assay can be used to measure CRP concentrations in the range of ~0.18–1100 mg/L.


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
assay kinetics
The principle of the noncompetitive microparticle CRP assay is shown in Fig. 1 . The assay consists of a monoclonal antibody immobilized on the microparticle surface (Fig. 1 , component 1), the analyte (component 2), and another monoclonal antibody labeled with an intrinsically fluorescent europium chelate (component 3). The kinetics of the assay components were studied separately (Fig. 2 ). As could be predicted, the binding of the detection antibody to the microparticle-bound capture antibody-analyte complex was the fastest reaction, reaching equilibrium in ~20 min. The rapid kinetics can be explained by the 15-fold molar excess of the antibody compared with the amount of analyte used. The kinetics could be further improved by increasing the antibody excess, although the effect on background signal and assay sensitivity should also be taken into consideration. The analyte-binding step reached equilibrium at ~240 min. The one-step kinetics, however, were unexpectedly slow, not reaching plateau even after 20 h of incubation.



View larger version (52K):
[in this window]
[in a new window]
  		Figure 1. Principle of the CRP microparticle immunoassay.

The noncompetitive assay consists of a monoclonal antibody immobilized on the microparticle surface ( (1)) and another monoclonal antibody labeled with an intrinsically fluorescent europium chelate ( (3)). The analyte, CRP ( (2)), is a pentameric protein with a molecular mass of 114 kDa. Assay components are not shown in actual proportion.



View larger version (18K):
[in this window]
[in a new window]
  		Figure 2. Kinetics of the noncompetitive CRP immunoassay.

In addition to the one-step kinetics ({triangleup}), the kinetics of the analyte (•) and the detection antibody ({square}) binding reactions were also determined. In the one-step reaction, all assay components shown in Fig. 1Up (1, 2, and 3) were incubated simultaneously. The binding kinetics of the analyte were studied by incubating components 1 + 2 for different times; component 3 was incubated afterward for a constant time. The detection antibody binding kinetics were studied by incubating components 1 + 2 first for a constant time; component 3 was then incubated in a separate step for different times. All assays were performed in a total volume of 330 nL/particle at CRP and detection antibody concentrations of 500 µg/L (one-third of the total volume) and 9.3 mg/L, respectively. For practical reasons, the incubations were performed using a batch of 21 microparticles/7 µL.

As demonstrated previously, the number of microparticles per microliter is in linear proportion to the assay kinetics (32). The same effect was observed here as well: increasing the number of particles per microliter by a factor of 10 correspondingly increased assay kinetics 10-fold (data not shown). However, assuming that the correlation between the number of particles and the kinetics remained constant, generating a microparticle CRP assay that reached equilibrium in ~2 min would require reducing the assay volume per particle by a factor of 600 (Table 1 ). In other words, instead of 3 microparticles/µL, 1800 particles/µL would be required to reach the desired kinetics, an approach clearly not feasible. Because the aim of the study, however, was to develop a rapid assay for measuring CRP under POC situations, another alternative, a kinetic approach, was taken under investigation.


View this table:
[in this window]
[in a new window]
  		Table 1. Calculated effect of microparticle number on one-step CRP kinetics.

kinetic crp immunoassay
Effect of microparticle concentration.
In common endpoint immunoassays, the sensitivity of an assay is strongly dependent on the sample volume or, in other words, on the total amount of analyte per reaction. In this study, we investigated the effect of the total analyte quantity on another kind of assay approach, a kinetic microparticle application, where the one-step binding reaction was interrupted before reaching equilibrium. We studied the effect of sample amount by varying the assay volume per particle but keeping the ratio of sample to assay volume (i.e., analyte concentration) constant. The proportion of sample was invariably one-third of the total volume, and a similar 2-min incubation time was used for each assay. Fig. 3 shows the fluorescence intensities at different CRP concentrations in total assay volumes of 10, 100, and 1000 nL/microparticle (i.e., 600 particles/6 µL, 60 particles/6 µL, and 20 particles/20 µL, respectively).



View larger version (17K):
[in this window]
[in a new window]
  		Figure 3. Mean fluorescence intensity measured at different CRP concentrations at an assay volume of 10 (•), 100 ({blacksquare}), or 1000 nL ({diamond}) per microparticle.

The one-step reactions were performed using the following three combinations: 600 particles/6 µL (10 nL/particle), 60 particles/6 µL (100 nL/particle), and 20 particles/20 µL (1000 nL/particle). The 2-min incubation time, the detection antibody concentration (6.2 mg/L), and the one-third proportion of sample to total volume were kept constant for each assay condition.

Effect of detection antibody concentration on the high-dose hook phenomenon.
Because no major difference was found between the assay volumes studied (10–1000 nL/particle), we selected a 330-nL volume for further analysis because of the relative ease of handling and our earlier experience with this assay volume. Subsequently, we studied the dynamics of the assay and the extent of high-dose hook interference typical of quantitative one-step assays (33) in detail, using various amounts of the detection antibody. The dynamic range of the kinetic microparticle assay was more than four orders of magnitude (Fig. 4 ) at each of the detection antibody concentrations studied (6.2, 31.0, and 124.0 mg/L). Furthermore, the signal obtained at a CRP concentration of 20 mg/L was, under each assay condition, at least equivalent to the signal achieved at a CRP concentration of 2 mg/L. Calculated from the most heavily bending curve (i.e., with a detection antibody concentration of 6.2 mg/L), CRP concentrations <=12 mg/L had no effect on result interpretation at an assay range of 0–3 mg/L CRP. To cover the CRP concentration range of 0.064–1200 mg/L, clinical samples were diluted 400-fold before analysis.



View larger version (17K):
[in this window]
[in a new window]
  		Figure 4. Effect of detection antibody concentration on the high-dose hook phenomenon.

Fluorescence intensity of individual microparticles was measured at different CRP concentrations at detection antibody concentrations of 6.2 (•), 31.0 ({blacktriangleup}), and 124.0 ({circ}) mg/L and a total assay volume of 330 nL/particle. For practical reasons, the incubations were performed using a batch of 21 microparticles/7 µL.

Calibration curve and clinical samples.
The calibration curve for the 2-min kinetic microparticle assay was established using an assay volume of 330 nL/particle. A typical curve based on four replicate particles is shown in Fig. 5 together with a within-assay precision profile. The analytical detection limit of the assay was 0.00016 mg/L CRP, calculated as the mean signal of the zero calibrator + 3 SD. In addition, the calibration curve was linear up to 3 mg/L CRP, i.e., more than four orders of magnitude.



View larger version (17K):
[in this window]
[in a new window]
  		Figure 5. Calibration curve (•) and precision profile ({circ}) for the one-step kinetic microparticle CRP assay using an assay volume of 330 nL/particle.

The time-resolved fluorescence intensity was measured for four individual microparticles at different CRP concentrations (0–3.0 mg/L). The dashed line indicates the analytical detection limit of the assay, 0.00016 mg/L (mean background + 3 SD). Incubations were performed using a batch of 21 microparticles/7 µL and a detection antibody concentration of 6.2 mg/L.

The within- and between-run variations for EDTA-whole blood samples are shown in Table 2 . Because of the poor stability of whole blood samples, the assays were performed on 2 sequential days, two runs per day, with four replicate particles per assay. As could be predicted, the highest imprecision was seen at the lowest CRP concentrations. The functional detection limit of the assay, i.e., the lowest clinical CRP concentration that could be measured with a CV <15%, was estimated to be 0.2 mg/L.


View this table:
[in this window]
[in a new window]
  		Table 2. Assay variability.1

We studied the correlation between the microparticle (y; EDTA whole blood) and the Dade Behring N High Sensitivity CRP (x; EDTA plasma) assays with 100 patient samples, using both Hct-corrected and uncorrected concentration values obtained from the microparticle assay (Fig. 6 ). In addition to the total CRP concentration range, the hs-CRP range from 0 to 10 mg/L was analyzed separately. The slopes for the uncorrected concentration values (Fig. 6A ) were close to those expected based on plasma (and thus also analyte) displacement by red blood cells in the sample (for hs-CRP, y = 0.604x, Sy|x = 0.458; for total CRP, y = 0.571x, Sy|x =7.733), yielding lower concentrations measured in the microparticle assay compared with the Dade Behring assay. After Hct correction (Fig. 6B ), however, the slope was largely normalized (for hs-CRP, y = 0.933x, Sy|x =0.680; for total CRP, y = 0.852x, Sy|x =11.735). The intercepts for the Hct-corrected hs-CRP and total-CRP curves were 0.075 ± 0.137 (r = 0.969; n = 54) and 2.135 ± 1.396 mg/L (r = 0.969; n = 100), respectively (P <0.0001).



View larger version (24K):
[in this window]
[in a new window]
  		Figure 6. Correlation between the kinetic microparticle immunoassay using EDTA-whole blood samples (y axis) and the Dade Behring N High Sensitivity CRP assay using corresponding plasma fractions (x axis).

Linear regression analysis was performed both on directly measured concentrations (A) and after correction with the respective Hct values (B). The insets show the correlation at CRP concentrations <10 mg/L. For uncorrected concentrations (A), the slopes for the low-range (0–10 mg/L CRP) and the wide-range (0–350 mg/L CRP) curves were 0.604 ± 0.022 and 0.571 ± 0.014 (Sy|x = 0.458 and 7.733), respectively, with y-intercepts of 0.024 ± 0.092 (r = 0.966; n = 54) and 1.227 ± 0.920 (r = 0.970; n = 100) mg/L, respectively. For Hct-corrected concentrations (B), the slopes for the low- and wide-range curves were 0.933 ± 0.033 and 0.852 ± 0.022 (Sy|x = 0.680 and 11.735), respectively, with y-intercepts of 0.075 ± 0.137 (r = 0.969; n = 54) and 2.135 ± 1.396 (r = 0.969; n = 100) mg/L. For all curves, P was <0.0001.

In the whole blood samples used in the correlation studies (n = 100), the Hct values ranged from 0.21 to 0.47. In addition, a mean Hct value was calculated from a total of 200 EDTA whole blood samples collected for routine CRP analysis at the local hospital (mean Hct, 0.357; range, 0.21–0.50). The CRP concentrations obtained with individual Hct or the mean Hct values for correction were compared with each other to determine the bias between the correction methods. Linear regression analysis (Fig. 7A ) and difference analysis (Fig. 7B ) both showed excellent agreement between the two methods for Hct correction, although some deviation of single patient samples was noted. For CRP concentrations <5 mg/L, the regression analysis yielded a slope of 1.053 ± 0.016 (Sy|x =0.135) and a y-intercept of -0.079 ± 0.033 mg/L (r = 0.996; n = 41). The corresponding values over the entire range of CRP concentrations (0–250 mg/L) were as follows: slope, 1.040 ± 0.006 (Sy|x = 2.947); y-intercept, -0.221 ± 0.355 (r = 0.998; n = 100). The difference analysis revealed no bias at CRP concentrations of 0–5 mg/L and a bias of 1.05 mg/L over the entire CRP concentration range (0–250 mg/L).



View larger version (24K):
[in this window]
[in a new window]
  		Figure 7. Comparison of two methods for Hct correction using linear regression analysis (A) and difference analysis (B) in 100 whole blood samples.

The mean Hct value was calculated from 200 EDTA whole blood samples collected for routine CRP measurements. The insets show the correlation at CRP concentrations <5 mg/L. In the linear regression analysis (A), the slopes for the low-range (0–5 mg/L CRP) and the wide-range (0–250 mg/L CRP) curves were 1.053 ± 0.016 and 1.040 ± 0.006 (Sy|x =0.135 and 2.947), respectively, with y-intercepts of -0.079 ± 0.033 (r = 0.996; n = 41) and -0.221 ± 0.355 (r = 0.998; n = 100) mg/L, respectively. The quintile cutoff values suggested by Rifai and Ridker (34) are indicated by dashed lines, and the one sample classified into differing quintiles is indicated by an arrow. The difference analysis (B) showed mean biases of 0.0 mg/L for CRP concentrations <5 mg/L and 1.05 mg/L for the entire CRP concentration range (0–250 mg/L; solid lines), with 95% limits of agreement of -0.29 to 0.30 mg/L and -5.81 to 7.90 mg/L (dashed lines), respectively.

Because the risk of a future coronary event has been found to increase as a function of hs-CRP concentration (9)(10)(11)(12)(13)(14)(15)(16), we evaluated the effect of fixed Hct correction on cardiovascular risk assessment, using the quintile cutoff values of 0.7, 1.2, 2.0, and 3.9 mg/L suggested by Rifai and Ridker (34). In the population with CRP <5 mg/L (n = 41), one sample with a Hct of 0.28 was classified into a different quintile based on a slightly overestimated CRP concentration because the Hct correction based on the individually determined or predetermined values produced CRP concentrations of 3.7 and 3.9 mg/L, respectively (Fig. 7AUp , inset).


   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
Most quantitative immunoassays depend on separating bound from free detection molecules once an equilibrium binding state has been achieved. In many cases, incubating the reaction to or almost to equilibrium is necessary to gain sufficient signal and assay sensitivity. However, using a label technology sensitive enough to produce a sufficient signal from fewer analyte molecules facilitates the introduction of a kinetic assay format with considerably shorter incubation times. In the noncompetitive microparticle CRP assay, the low detection limits achievable with lanthanide chelates allowed the binding reaction to be stopped long before it reached plateau, at the same time retaining an excellent analytical detection limit of 0.00016 mg/L. Although an even shorter incubation time could have been used, the present 2-min reaction time was selected for practical reasons.

Another important advantage of using a stable lanthanide chelate as a label was that unprocessed whole blood could be used as the sample material in the one-step assay without disturbance from red blood cells or EDTA, which is a strong chelating agent and a commonly used anticoagulant in blood samples. Because of the extremely sensitive assay concept, the EDTA-whole blood samples used in the correlation studies were diluted 400-fold before being assayed to cover a CRP concentration range of 0.064–1200 mg/L. The range could have easily been shifted by altering the dilution factor, but the wide dynamic range of the microparticle assay eliminated the need for multiple sample dilutions. In addition, a high dilution factor is very likely to have a positive contribution in reducing any blood-derived assay interference from the whole blood samples.

The effect of Hct correction in whole blood samples became evident in the correlation study between the microparticle and the Dade Behring N High Sensitivity assays. The lower slope obtained for the uncorrected concentration values was directly related to Hct and the displacing effect of red blood cells in the sample. Thus, after correction for Hct, the slope was normalized. The overall correlation between the microparticle and the Dade Behring assays was good, but with some deviation seen at higher CRP concentrations. The best slopes between the two assays were seen at concentrations <10 mg/L, and samples above this limit gave systematically lower concentrations in the microparticle assay compared with the Dade Behring assay. In the latter assay, CRP concentrations <10 mg/L are assayed at an initial sample dilution of 1:20, whereas samples with high concentrations are automatically diluted 1:100 (>10 mg/L) or 1:400 (>50 mg/L), which may have an effect on assay linearity. In the microparticle assay, all samples were assayed at a 1:400 dilution.

Only a minor bias was noted between concentrations corrected by use of an individually determined Hct value or a predetermined mean Hct value. As was also discussed by Pettersson et al. (35), typical variations in Hct values have fairly insignificant effects on assay outcomes, rendering individual Hct determinations unnecessary in the case of many test analytes. However, the use of a predetermined mean Hct value unavoidably leads to slightly overestimated concentrations when the Hct is low and, correspondingly, slightly underestimated concentrations when the Hct is high. As a result, incorrect classification of single patient samples with low or high Hct values is possible, especially near cutoff concentrations. In this study, however, only 1 patient sample (n = 41) was classified into a different quintile based on a slightly overestimated CRP concentration. For markedly increased concentrations, the bias becomes even more insignificant.

The effect of the high-dose hook phenomenon on the one-step kinetic microparticle assay was also insignificant. Considering the predilution of clinical samples, only a CRP concentration >4800 mg/L would affect result interpretation, and this concentration is far higher than the concentrations seen in clinical situations. In fact, CRP concentrations in the range of 100-1000 mg/L are rare and indicate a very serious clinical condition (36)(37)(38). An assay that also covers this range, however, facilitates sensitive follow-up of the effect of medication, e.g., in field circumstances, where training in instrument operation and parallel diagnostic methods may be limited. In conclusion, because the clinical CRP concentration range for evaluating the risk of future cardiac complications is ~0.5–5 mg/L and concentrations >5 mg/L indicate a bacterial or viral infection, the microparticle assay developed can be used for any application involving measurement of CRP concentrations in clinical samples.

We have previously demonstrated the clinical functionality of a customary endpoint assay that uses similar microparticles coated with Fab fragments specific for prostate-specific antigen (PSA) (39). When we used the same assay volume (330 nL/particle) as in this study, the equilibrium of the one-step assay was reached in 240 min. Slower kinetics were observed for the one-step CRP assay, where reaching equilibrium took >1200 min. The difference in analyte-binding kinetics between the two assays was, however, smaller (160 min for PSA; 240 min for CRP). The kinetic differences are likely to originate directly from the much larger size of CRP (114 kDa) compared with PSA (33 kDa), which not only slows the diffusion rate in the microparticle pores, but also makes entering the pores more difficult because of steric hindrance. The extensive differences in the analyte and the one-step kinetics of the CRP assay can be explained by the increased molecular size as well: whereas in the first case the analyte and detection antibody diffuse separately in the reaction mixture and microparticle pores, in the latter case they form a liquid-phase complex before entering the pores. Because of the large molecular excess of the detection antibody and the pentameric molecular form of CRP, the complex may become notably large compared with single analyte and antibody molecules. In addition to the size-dependent effects mentioned above, this may also slow down and complicate the solid-phase binding reaction. If the current CRP assay were to be performed as an endpoint measurement, however, the slow kinetics could at least to some extent be improved by use of microparticles with a larger pore size or, alternatively, use of antibody fragments instead of intact antibodies for antigen capture and detection (39)(40). However, to reach a 2-min incubation time with the current assay design would require reducing the assay volumes to 3.0 and 0.6 nL/particle in the PSA and CRP assays, respectively. In addition to poor assay sensitivities, this would lead to impractical assay volumes or to the use of a notably higher number of microparticles per assay. Given that individual microparticles are measured, the customary equilibrium/endpoint assay is not feasible for POCT.

The kinetic assay format allowed us to generate a clinically functional, ultrasensitive CRP assay with an incubation time of only 2 min. Another key aspect of the assay developed was the independence of assay performance from microparticle concentration. Only a 1.5-fold change in the net signal was detected regardless of the 100-fold difference in the surface-to-volume ratio, obviously because of the low percentage of sample bound during the 2-min incubation time. In other words, because none of the assays was near equilibrium and the amount of bound analyte remained lower than the amount of unbound analyte throughout the 2-min incubation period, moderate changes in microparticle concentration or, more accurately, in the total assay volume per particle were irrelevant. Thus, the kinetic microparticle assay measured the initial rate of analyte association rather than the total amount of the antigen.

The revolution in assay and detection technologies has recently facilitated the development of genuinely portable instrumentation for extralaboratory testing, and commercial, portable POC CRP assays with extremely short turnaround times have already been developed. For example, the 2-min reflectometric NycoCard® CRP Whole Blood test (Axis Shield) (21) and the turbidimetric, 3-min QuikRead® CRP whole blood/serum/plasma assay (Orion Diagnostica) (23) represent pioneering quantitative test systems suitable for extralaboratory testing. In the flow-through NycoCard assay, the results are corrected with an assumed Hct value of 0.40, and the assay also facilitates visual interpretation of results. Both assay procedures, however, require manual dispensing of the assay-specific reagents and the sample. Because of the detection technologies used in these assays, the detection limits of the assays are high, 10 mg/L for the NycoCard and 8 mg/L for the QuikRead, and the measurement ranges cover less than two orders of magnitude. Thus, neither of these assays comes close to the sensitivity or the dynamic range seen in the 2-min kinetic microparticle approach. In fact, the difference in analytical detection limits between the microparticle and the above-mentioned assays is more than 30 000-fold.

Although many assay steps were performed manually in the present study, total automation of the assays can be accomplished. The fact that the kinetic assay concept described can easily bear changes in assay volumes or microparticle concentration gives much more flexibility in adjusting the assay conditions with an instrument platform, and it also enables more robust production of the assay itself. In addition, the batchwise procedure for coating the particles increases the ease and flexibility of production. Because it is substantially faster to perform than most of the current high-sensitivity immunoassays, the kinetic assay format should be well suited for POC purposes. Because assay systems for extralaboratory use require at least some extent of automation to reduce the risk of human error and to increase the accuracy of the results, discontinuing the reactions before equilibrium should not present a notable additional challenge for instrumentation. The assay-to-assay variation typical of manually performed kinetic assays could easily be avoided by exact device adjustment and programming of the incubation times and temperatures.


   Acknowledgments
 
This study was partially supported by the Defense Advanced Research Projects Agency (DARPA), USA. We gratefully acknowledge the skillful assistance of Virve Kokko in performing the microparticle assays for the correlation study, and Prof. Kim Pettersson for helpful discussions concerning the analysis of clinical samples. We also thank Seppo Laine and Marjo Leponiemi at the Tampere University Hospital for analyzing the EDTA-plasma samples in their Dade Behring BN II nephelometer. Finally, we warmly thank the Turku Police Department and especially investigator Simo Takomo for rescuing the manuscript and several years of other work by recovering our stolen computer within 20 h.


   Footnotes
 
1 Nonstandard abbreviations: CRP, C-reactive protein; hs-CRP, high-sensitivity CRP; POCT, point-of-care testing; Hct, hematocrit; and PSA, prostate-specific antigen.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Macy EM, Hayes TE, Tracy RP. Variability in the measurement of C-reactive protein in healthy subjects: implications for reference intervals and epidemiological applications. Clin Chem 1997;43:52-58.[Abstract/Free Full Text]
  2. Ledue TB, Weiner DL, Sipe J, Poulin SE, Collins MF, Rifai N. Analytical evaluation of particle-enhanced immunonephelometric assays for C-reactive protein, serum amyloid A, and mannose binding protein in human serum. Ann Clin Biochem 1998;35:745-753.
  3. Wilkins J, Gallimore JR, Moore EG, Pepys MB. Rapid automated high sensitivity enzyme immunoassay of C-reactive protein [Technical Brief]. Clin Chem 1998;44:1358-1361.[Free Full Text]
  4. Eda S, Kaufman J, Molwitz M, Vorberg E. A new method of measuring C-reactive protein, with a low limit of detection, suitable for risk assessment of coronary heart disease. Scand J Clin Lab Invest Suppl 1999;230:32-35.[Medline] [Order article via Infotrieve]
  5. Borque L, Bellod L, Rus A, Seco ML, Galisteo-Gonzalez F. Development and validation of an automated and ultrasensitive immunoturbidimetric assay for C-reactive protein [Technical Brief]. Clin Chem 2000;46:1839-1842.[Free Full Text]
  6. Rifai N, Tracy RP, Ridker PM. Clinical efficacy of an automated high-sensitivity C-reactive protein assay. Clin Chem 1999;45:2136-2141.[Abstract/Free Full Text]
  7. Pepys MB. C-Reactive protein fifty years on. Lancet 1981;21:653-657.
  8. Gewurz H, Mold C, Siegel J, Fiedel B. C-Reactive protein and the acute phase response [Review]. Adv Intern Med 1982;27:345-372.[Web of Science][Medline] [Order article via Infotrieve]
  9. Kuller LH, Tracy RP, Shaten J, Meilahn EN. Relation of C-reactive protein and coronary heart disease in the MRFIT nested case-control study. Multiple Risk Factor Intervention Trial. Am J Epidemiol 1996;144:537-547.[Abstract/Free Full Text]
  10. Tracy RP, Lemaitre RN, Psaty BM, Ives DG, Evans RW, Cushman M, et al. Relationship of C-reactive protein to risk of cardiovascular disease in the elderly. Results of the Cardiovascular Health Study and the Rural Health Promotion Project. Arterioscler Thromb Vasc Biol 1997;17:1121-1127.[Abstract/Free Full Text]
  11. Ridker PM, Cushman M, Stampfer MJ, Tracy RP, Hennekens CH. Inflammation, aspirin, and the risk of cardiovascular disease in apparently healthy men. N Engl J Med 1997;336:973-979.[Abstract/Free Full Text]
  12. Ridker PM, Cushman M, Stampfer MJ, Tracy RP, Hennekens CH. Plasma concentration C-reactive protein and risk of developing peripheral vascular disease. Circulation 1998;97:425-428.[Abstract/Free Full Text]
  13. Ridker PM, Buring JE, Shih J, Matias M, Hennekens CH. Prospective study of the C-reactive protein and the risk of future cardiovascular events among apparently healthy women. Circulation 1998;98:731-733.[Abstract/Free Full Text]
  14. Koenig W, Sund M, Frohlich M, Fisher H-G, Lowel H, Doring A, et al. C-Reactive protein, a sensitive marker of inflammation, predicts future risk of coronary heart disease in initially healthy middle-aged men: results from the MONICA (Monitoring Trends and Determinants in Cardiovascular Disease) Augsburg Cohort Study, 1984 to 1992. Circulation 1999;99:237-242.[Abstract/Free Full Text]
  15. Danesh J, Whincup P, Walker M, Lennon L, Thomson A, Appleby P, et al. Low grade inflammation and coronary heart disease: prospective study and updated meta-analyses. BMJ 2000;321:199-204.[Abstract/Free Full Text]
  16. Roivainen M, Viik-Kajander M, Palosuo T, Toivanen P, Leinonen M, Saikku P, et al. Infections, inflammation, and the risk of coronary heart disease. Circulation 2000;101:252-257.[Abstract/Free Full Text]
  17. Ridker PM, Glynn RJ, Hennekens CH. C-Reactive protein adds to the predictive value of total and HDL cholesterol in determining risk of first myocardial infarction. Circulation 1998;97:2007-2011.[Abstract/Free Full Text]
  18. Morrow DA, Rifai N, Antman EM, Weiner DL, McCabe CH, Cannon CP, Braunwald E. C-Reactive protein is a potent predictor of mortality independently and in combination with troponin T in acute coronary syndromes: a TIMI 11A substudy. Thrombosis in Myocardial Infarction. J Am Coll Cardiol 1998;31:1460-1465.[Abstract/Free Full Text]
  19. Ridker PM. Evaluating novel cardiovascular risk factors: can we better predict heart attacks [Review]?. Ann Intern Med 1999;2:986-988.
  20. Dahler-Eriksen BS, Lassen JF, Lund ED, Lauritzen T, Brandslund I. C-Reactive protein in general practice—how commonly is it used and why?. Scand J Prim Health Care 1997;15:35-38.[Web of Science][Medline] [Order article via Infotrieve]
  21. Urdal P, Borch SM, Landaas S, Krutnes MB, Gogstad GO, Hjortdahl P. Rapid immunometric measurement of C-reactive protein in whole blood. Clin Chem 1992;38:580-584.[Abstract/Free Full Text]
  22. Hobbs FD, Kenkre JE, Carter YH, Thorpe GH, Holder RL. Reliability and feasibility of a near patient test for C-reactive protein in primary care. Br J Gen Pract 1996;46:395-400.[Web of Science][Medline] [Order article via Infotrieve]
  23. Laitinen H, Mäkinen P, Raussi J. Evaluation of the new QuikRead system using dried reagents for measuring C-reactive protein in whole blood [Abstract]. 12th IFCC European Congress of Clinical Chemistry 1997;:A32.
  24. Hemmilä I, Dakubu S, Mukkala VM, Siitari H, Lövgren T. Europium as a label in time-resolved immunofluorometric assays. Anal Biochem 1984;137:335-343.[Web of Science][Medline] [Order article via Infotrieve]
  25. Soini E, Lövgren T. Time-resolved fluorescence of lanthanide probes and application in biotechnology [Review]. CRC Crit Rev Anal Chem 1987;18:105-154.
  26. Hemmilä I. Lanthanides as probes for time-resolved fluorometric immunoassays [Review]. Scand J Clin Lab Invest 1988;48:389-399.[Web of Science][Medline] [Order article via Infotrieve]
  27. Ekins RP, Dakubu S. The development of high sensitivity pulsed light, time-resolved fluoroimmunoassays. Pure Appl Chem 1985;57:473-482.
  28. Lövgren T, Pettersson K. Time-resolved fluoroimmunoassay, advantages and limitations. Van Dyke K Van Dyke R eds. Luminescence immunoassay and molecular applications 1990:233-253 CRC Press Boca Raton, FL. .
  29. Dickson EF, Pollak A, Diamandis EP. Ultrasensitive bioanalytical assays using time-resolved fluorescence detection [Review]. Pharmacol Ther 1995;66:207-235.[Web of Science][Medline] [Order article via Infotrieve]
  30. Lövgren T, Heinonen P, Lehtinen P, Hakala H, Heinola J, Harju R, et al. Sensitive bioaffinity assays with individual microparticles and time-resolved fluorometry. Clin Chem 1997;43:1937-1943.[Abstract/Free Full Text]
  31. Ugelstad J, Berge A, Ellingsen T, Schmid R, Nilsen TN, Mørk PC. Preparation and application of new monosized polymer particles. Prog Polym Sci 1992;17:87-161.
  32. Härmä H, Lehtinen P, Takalo H, Lövgren T. Immunoassay on a single microparticle: the effect of particle size and number on a miniaturized time-resolved fluorometric assay of free prostate-specific antigen. Anal Chim Acta 1999;387:11-19.
  33. Fernando SA, Wilson GS. Studies of the ‘hook’ effect in the one-step sandwich immunoassay. J Immunol Methods 1992;151:47-66.[Web of Science][Medline] [Order article via Infotrieve]
  34. Rifai N, Ridker PM. Proposed cardiovascular risk assessment algorithm using high-sensitivity C-reactive protein and lipid screening. Clin Chem 2001;47:28-30.[Free Full Text]
  35. Pettersson K, Katajamäki T, Irjala K, Leppänen V, Majamaa-Voltti K, Laitinen P. Time-resolved fluorometry (TRF)-based immunoassay concept for rapid and quantitative determination of biochemical myocardial infarction markers from whole blood, serum and plasma. Luminescence 2000;15:399-407.[Web of Science][Medline] [Order article via Infotrieve]
  36. Eskola J, Holmberg C, Koskimies O. C-Reactive protein—an indicator of generalized bacterial infection in children with nephrotic syndrome. Acta Paediatr Scand 1986;75:846-848.[Web of Science][Medline] [Order article via Infotrieve]
  37. Gosling P, Dickson GR. Serum C-reactive protein in patients with serious trauma. Injury 1992;23:483-486.[Web of Science][Medline] [Order article via Infotrieve]
  38. Povoa P, Almeida E, Modeira P, Fernandes A, Mealha R, Aragao A, Sabino H. C-Reactive protein as an indicator of sepsis. Intensive Care Med 1998;24:1052-1056.[Web of Science][Medline] [Order article via Infotrieve]
  39. Härmä H, Tarkkinen P, Soukka T, Lövgren T. Miniature single-particle immunoassay for prostate-specific antigen in serum using recombinant Fab fragments. Clin Chem 2000;46:1755-1761.[Abstract/Free Full Text]
  40. Eriksson S, Vehniäinen M, Jansén T, Meretoja V, Saviranta P, Pettersson K, Lövgren T. Dual-label time-resolved immunofluorometric assay of free and total prostate-specific antigen based on recombinant Fab fragments. Clin Chem 2000;46:658-666.[Abstract/Free Full Text]



The following articles in journals at HighWire Press have cited this article:


Home page
 Clin. Chem.Home page
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]

Home page
 Clin. Chem.Home page
P. von Lode, J. Rainaho, and K. Pettersson
Quantitative, Wide-Range, 5-Minute Point-of-Care Immunoassay for Total Human Chorionic Gonadotropin in Whole Blood
Clin. Chem., June 1, 2004; 50(6): 1026 - 1035.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Web of Science (21)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Tarkkinen, P.
Right arrow Articles by Lövgren, T.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Tarkkinen, P.
Right arrow Articles by Lövgren, T.
Related Collections
Right arrow Clinical Immunology
Right arrow Proteomics and Protein Markers
Right arrow Endocrinology and Metabolism
Right arrow Automation and Analytical Techniques

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS