| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Abstracts of Oak Ridge Posters |
1
Spectral Diagnostics Inc., 135-2 The West Mall, Toronto, Ontario, M9C1C2 Canada
a author for correspondence: fax 416-626-3651, e-mail jash{at}spectraldiagnostics.com
Currently, the measurement of troponin I (TnI) can only be accomplished through the use of heterogeneous assays on closed-system automated analyzers. The development of this new and innovative latex technology will allow the measurement of TnI on a variety of turbidimetry-based open-system instruments, greatly enhancing clinical applicability of this test.
Determining the presence of TnI in the serum of patients is an important aid in the diagnosis of myocardial infarction. An advantage of TnI is its improved specificity for myocardial damage compared with creatine kinase-MB (1). In addition, there is strong evidence that future utilization of TnI will be for risk stratification to assist in the decision process for therapeutic intervention with glycoprotein II/IIIa inhibitors or low-molecular weight heparin (2)(3). In fact, the GUSTO trial, which should be completed soon, included TnI as one of the cardiac markers to be considered for risk stratification. The cardiac troponin complex is part of the contractile apparatus of the thin filament in striated muscle and consists of subunits C, T, and I. Different isoforms of TnI exist in the skeletal and cardiac muscles (fast skeletal, slow skeletal, and cardiac TnI). The distinct structural heterogeneity between these isoforms allows production of specific antibodies (4), which can be utilized by the latex assay to detect serum TnI in clinical conditions that involve myocardial damage. After acute myocardial infarction, damaged myocytes lose these proteins, and various forms of troponin (complexed, free, or fragments) appear in the blood (5). TnI concentrations become abnormal 4–8 h after the onset of chest pain, peak at 12–16 h, and remain increased for 5–9 days following an infarction.
We have developed an automated latex immunoassay for the detection of
TnI in serum with excellent sensitivity, precision, and stability. The
assay utilizes two monoclonal antibodies and one polyclonal antibody to
TnI. Antibodies were selected using Biacore analysis, epitope mapping,
and pairing studies. Each antibody was separately covalently bound to
200-nm supercarboxyl polystyrene latex particles using
1-ethyl-3(3-dimethylaminopropyl)carbodiimide hydrochloride in a
two-step coupling process. The various coupled antibodies were then
combined into one solution in a 1:1:1 ratio. Unbound antibody was
removed with a microgon system, which utilizes tangential flow. The
beads were then sonicated to redistribute the particles. When TnI is
present in the serum, the addition of a reaction buffer causes adjacent
beads to cross-link, increasing the turbidity of the solution. The
results are calculated in <9 min by the Cobas Mira from a stored
calibration curve generated with recombinant human cardiac TnI
calibrators using a logit/log4 calculation mode. Reaction buffer (140
µL), water (10 µL), and latex (45 µL of a 1.25 g/L solution) are
mixed and incubated at 37 °C for 125 s. Sample (30 µL) and
water (5 µL) are then added, and the absorbance change of the
reaction mixture is measured from 
1 min 25 s to 8 min after
the addition of sample. The rate of increase in turbidity, measured at
600 nm, is proportional to the concentration of TnI present in the
serum. The assay range was 0–25 µg/L. The detection
limit of the assay was 0.3 µg/L. This value was calculated
as 2 SD above the mean of 21 replicates of the zero calibrator. Using
NCCLS guidelines, intra- and interassay imprecision was determined
using both a high and low control over a course of 18 days with two
analytical runs per day. The intraassay imprecision (CV) for the
5 µg/L control was 6.8%, and its interassay CV was 7.5%. The 15
µg/L control gave an intraassay CV of 1.6% and interassay CV of
4.2%.
Stability studies were carried out on the latex particles, the assay buffer, and the calibrators by evaluating the performance of high and low controls with the test reagent at scheduled intervals with approved lots of reagents. The test reagent was considered unstable when the control results demonstrated a downward or upward shift of >10% from the day 0 value for the control and this variation was consistent on 2 or more consecutive days. Calibrators were found to be stable for 2 months when stored at room temperature. Latex reagent was stable for 21 days at 37 °C, and the reaction buffer was stable for 56 days at room temperature.
The assay was evaluated to determine whether antigenexcess
phenomenon (prozone effect) would cause false negatives. A false
negative occurs when a sample that has a high troponin concentration
produces a negative result instead of the high value. This phenomenon
was evaluated by preparing a high concentration of recombinant TnI in
storage buffer, making serial dilutions of this material, and testing
each sample. TnI concentrations up to 450 µg/L yielded a signal that
was above the limit defined by the measuring range, and concentrations
up to 4000 µg/L did not produce a false negative. Rheumatoid factor
added to a final concentration of 160 kIU/L to control and serum
samples positive for TnI did not interfere with the assay, nor did
hemoglobin (up to 5 g/L) or bilirubin (up to 400 mg/L). In
addition to the baseline concentrations of triglycerides and albumin
already present in the serum pools, additional amounts of triglycerides
(up to 250 mg/L) and albumin (up to 300 mg/L) when added to the pools
did not interfere with the assay. A comparison of assay values obtained
from testing serum samples using the latex TnI test and the Stratus TnI
immunoassay gave the following linear regression data: Latex TnI =
0.18 (Stratus) + 0.0; n = 281, r = 0.87; range,
0–37.2 µg/L (Fig. 1
).
|
At cutoffs for the diagnosis of acute myocardial infarction of 0.8 µg/L for the latex immunoassay as performed on the Cobas Mira and 1.5 µg/L for the Stratus TnI fluorometric enzyme immunoassay, the performance of the two assays was compared. There were 108 samples positive by the Stratus and 106 positive by the latex assay, which yielded a sensitivity of 98.1%. Of the 173 samples negative by the Stratus, 164 (95%) were negative by the latex assay. The overall agreement of 96% (270 of 280) demonstrates that the two products have equivalent performance for the determination of TnI in specimens from chest pain patients.
This represents the first report of a rapid, fully automated homogeneous latex immunoassay for TnI with good sensitivity and precision at low concentrations, limited prozone effect, and reagent stability comparable to the existing heterogeneous immunoassays.
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
