Clinical Chemistry 45: 388-393, 1999;
(Clinical Chemistry. 1999;45:388-393.)
© 1999 American Association for Clinical Chemistry, Inc.
Automated Homogeneous Immunoassay for Gentamicin on the Dimension Clinical Chemistry System
Tie Q. Weia,
Victor P. Chu,
Alan R. Craig,
James E. Duffy,
David M. Obzansky,
Daniel Kilgore,
Ignazio S. Masulli,
Connie M. Sanders and
John C. Thompson
a Author for correspondence. Fax 302-631-7487; e-mail weitq{at}dadebehring.com
 |
Abstract
|
|---|
Background: Monitoring of the concentration of gentamicin in serum
and plasma during therapy is widely recommended and practiced in
hospitals. Our aim was to develop a homogeneous immunoassay based on
particle-enhanced turbidimetric inhibition immunoassay technology to
quantify gentamicin on the Dimension® clinical chemistry
system.
Methods: Assay performance was assessed on each of the Dimension
models in a 15-instrument interlaboratory comparison study. A
split-sample comparison (n = 1171) was also performed between the
gentamicin methods on the Dimension system and the Abbott®TDx® analyzer, using multiple reagent and calibrator lots
on multiple instruments.
Results: The Dimension method was linear to 25.1 µmol/L (12.0
µg/mL) with a detection limit of 0.63 µmol/L (0.3 µg/mL).
Calibration was stable for 30 days. The within-run imprecision (CV) was
<1.3%, and total imprecision ranged from 1.8% to 3.2% between 4.2
µmol/L (2.0 µg/mL) and 16.7 µmol/L (8.0 µg/mL) gentamicin.
Linear regression analysis of the results on the Dimension method (DM)
vs the Abbott TDx yielded the following equation: DM =
0.98TDx - 0.42; r = 0.987. Minimal
interference was observed from structurally related compounds such as
sagamicin, netilmicin, and sisomicin.
Conclusion: The monoclonal antibody used in this method has
similar reactivities toward the individual gentamicin subspecies C1,
C1a, and C2, thus providing analytical recovery not significantly
dependent on relative subspecies concentrations. © 1999 American
Association for Clinical Chemistry
|
Introduction
|
|---|
Because excess accumulation of gentamicin in blood can lead to
renal tubular necrosis and degeneration of the auditory nerve
(1) monitoring the concentration of this drug during therapy
is widely recommended and practiced in hospitals (2)(3).
Clinical laboratory trends toward workstation consolidation are
apparent, and this has increased the importance of performing
therapeutic drug monitoring
(TDM)1
assays on
high-throughput, multipurpose clinical chemistry analyzers. A
particle-enhanced turbidimetric inhibition immunoassay (PETINIA) was
thus developed for the measurement of gentamicin on the
Dimension® clinical chemistry
system.2
Features
and performance of this assay are discussed here.
In the assay, gentamicin linked to latex particles reacts with a
F(ab')2 fragment of an anti-gentamicin monoclonal antibody
to form aggregates that increase the turbidity of the solution. Free
gentamicin from the sample competes for the antibody fragment, thereby
decreasing the rate of particle aggregation. Thus, the rate of
aggregation is inversely proportional to the concentration of
gentamicin in the sample. The rate of turbidity increase is measured
bichromatically at 340 and 700 nm and converted to analyte
concentration using a logit transformation (4). The
particles are 65 ± 10 nm in diameter and composed of a
polystyrene core and a polyglycidylmethacrylate shell. The particles
were prepared according to procedures described previously
(5).
The assay was designed for robustness relative to several performance
attributes. This included relative consistency of analytical recovery
for the C1, C1a, and C2 subspecies (6) in pharmaceutical
gentamicin preparations. Thus, we used a monoclonal antibody that
reacts similarly with all of these subspecies. F(ab')2 fragments were used as a precaution to prevent
nonspecific binding, which can occur when rheumatoid factor binds to Fc
fragments of intact antibodies (7)(8)(9)(10). As an additional
precaution, operating parameters were included to prevent the reporting
of results from the very rare presence of nonspecific aggregation that
can arise from unusual serum constituents. Because the Dimension system
encompasses a family of instrument models, assay performance was
assessed on each of these in a 15-instrument interlaboratory comparison
study.
Although the main purpose here is to describe assay features and
analytical performance, we also describe results of work done to shed
light on the mechanism of the agglutination reaction used. Craig
(11) proposed previously that the immunospecific binding of
antibody fragments to analyte-modified particles changes the colloidal
properties of the latex particles, thus triggering agglutination
without the need for bridging of particles by the bivalent antibody.
Strong evidence for the validity of this mechanism was reported by
Thompson et al. (12) in a detailed study of the kinetics and
mechanism of reactions involved in a PETINIA test for phenytoin. Data
on the gentamicin assay described here suggest a similar mechanism.
|
Materials and Methods
|
|---|
Gentamicin FlexTM reagent cartridges were
from Dade Behring Inc. These contained particle reagent, antibody
reagent, and buffer. The antibody reagent is a F(ab')2 fragment prepared by standard procedures (13).
The parent monoclonal antibody was derived from a hybridoma cell line
resulting from fusion of spleen cells of BALB/c mice immunized with
gentamicin-bovine serum albumin and mouse myeloma cells, using standard
techniques (14). The particle reagent is described above.
Specimens (serum, plasma) were from several hospitals.
DADE® IMMUNOASSAY (COMPREHENSIVE TRI-LEVEL),
DADE TDM, BIORAD® LIQUICHEKTM TDM and
LYPHOCHEKTM TDM, CIBA-CORNING TDM, HYCOR
SENTRY® COMBO-TROL controls were used. For recovery and
linearity studies, U.S. Pharmacopeia (USP) gentamicin sulfate was dried
under reduced pressure not exceeding 5 mmHg at 110 °C for 3 h
and added immediately to gentamicin-free serum pools to give the
reported concentrations. For stability studies, Beckman Gentamicin
Calibrator containing gentamicin concentrations of 0.0, 2.1, 4.2, 8.4,
16.7, and 25.1 µmol/L (0.0, 1.0, 2.0, 4.0, 8.0, and 12 µg/mL) was
used. Calibration was performed with Dimension system Drug Calibrator
II (Dade Behring Inc.), a multianalyte calibrator that includes
gentamicin concentrations of approximately 0.0, 3.1, 6.0, 12.5, and
25.1 µmol/L (0.0, 1.5, 3.0, 6.0, and 12.0 µg/mL). The calibrator
bottle values are assigned by a master pool anchored to USP gentamicin
sulfate (lot J) using two Dimension instruments. Gentamicin subspecies
C1, C1a, and C2 were purified using HPLC (Waters Model 510) and a
DuPont ZORBAX® Sil column (21.2 x 25 cm)
as described by Anhalt et al. (15). The mobile phase used
was H2O:methanol:diethylamine (60:40:0.5, by
volume).
After 1.0 mL of gentamicin sulfate (Sigma Chemical Co.; lot no.
42H06105) was injected, the subspecies were separated in the column by
an isocratic running program at a flow of 5.0 mL/min. A split-sample
correlation study was performed in our laboratories and two clinical
hospital laboratories. Serum specimens used in our laboratories were
acquired from several hospital laboratories, shipped on dry ice, and
subsequently stored frozen at -20 °C until thawed before use.
Clinical specimens used in hospital laboratories were treated the same
way. Analyses were performed on both the Dimension system and
TDx® analyzer on the same day, once a sample was
thawed.
More than 6000 samples were screened for human anti-mouse antibodies,
using murine monoclonal antibodies with different specificity in
sandwich assays as proposed by Vaidya and Beatty (16). Nine
positive samples from the screening test were selected randomly for the
interference testing in this study. Also used was a serum sample that
showed the most severe murine heterophilic antibody interference among
all the positive samples we collected. The interlaboratory comparison
was done with a full-factorial design, in which individual instruments
and days of the study were the experimental factors.
Dade® IAC TDM controls, three serum pools, and
one lot of calibrator were run in five replicates per day over 5
consecutive days on each of 15 calibrated Dimension instruments,
including the XL, AR, ES, and SMS models. The instruments were
physically located in five separate laboratories. We used
JMP® statistical software (SAS Institute, Inc.)
to analyze the data.
Processing of gentamicin assays on the Dimension system, as directed by
the system software, is depicted and described briefly in Fig. 1
. Data shown were captured in a nonroutine processing mode, in
which absorbances are monitored continuously. Operating principles of
the Dimension system have been published previously (17).
View larger version (16K):
[in this window]
[in a new window]
|
Figure 1. Absorbance change over time in the gentamicin assay.
Absorbance was measured continuously (nonroutine mode of instrument
operation) on a Dimension system model AR at 340 and 700 nm
wavelengths. R1, R2, and
R3 indicate the times at which the instrument measures
these absorbances during routine assay processing. The lines show the
measured difference in the two absorbances over time for 0.0 and 25.1
µmol/L (12 µg/mL) gentamicin calibrators. In the routine operating
mode, particle reagent (PR) and buffer are first added
to the cuvette, followed by water, and mixed ultrasonically. Sample (3
µL) is added, followed by water, and the contents mixed
ultrasonically. R1 and R2 measurements are performed to detect any
unusual nonspecific agglutination that might occur and which would be
flagged as errors. Antibody reagent is then added, and a final
measurement (R3) is made at a fixed time. The measured
bichromatic R2 absorbance is subtracted from that for R3, the
difference being inversely proportional to the concentration of
gentamicin in the sample.
|
|
Experiments to explore the response of the assay over a wide range of
antibody concentrations were done as described earlier for another
assay system (11).
|
Results and Discussion
|
|---|
The clone used to produce the monoclonal antibody in this assay
was selected from a number of hybridoma fusions based on several
criteria important to analytical performance. Characteristics screened
included ability to agglutinate gentamicin-modified latex particles,
antibody specificity (minimum cross-reactivity with other
aminoglycoside drugs), and similarity of reactivities of various
gentamicin subspecies. (Gentamicin is a fermentation product containing
a mixture of antibiotics including gentamicin, C1, C1a, and C2
subspecies.) Fig. 2
shows a typical calibration curve, which indicates
agglutination more than sufficient for application in a clinical assay.
The specificity characteristics of the assay are indicated in Table 1
, and the responses for gentamicin subspecies are presented in
Table 2
along with data for the Abbott® TDx
analyzer. Results show similarity of the two assays in this respect.
View larger version (21K):
[in this window]
[in a new window]
|
Figure 2. Representative gentamicin calibration curve.
Data were obtained on a Dimension system, model AR.
Points show means of duplicate determinations for
calibrators with five concentrations of gentamicin. The logit curve fit
was used to obtain calibration coefficients for the slope, intercept,
C2, and C3 with a fixed C4 term (0.5).
|
|
View this table:
[in this window]
[in a new window]
|
Table 1. Specificity1
of the Dimension
system gentamicin method: Reactivity to structurally related compounds,
other drugs, and physiological
substances.
|
|
In the course of our investigations, several patient specimens
containing high rheumatoid factor titers interfered with the assay
using the intact monoclonal antibody. F(ab')2
fragments prepared by pepsin digestion showed no such interference;
therefore, this step was incorporated into the manufacturing process
for the antibody reagent. The particle reagent formulation was
optimized by adjusting the concentration of gentamicin used in the
conjugation to achieve various densities of analyte analog on the
particle surface; the formulation was selected for precise analytical
results across the range of the calibration curve. The curve shape and
the overall range of absorbance changes for the optimized reagents are
shown in Fig. 2
. An error routine was incorporated into the software,
as described in the caption of Fig. 1
. If agglutination was to occur
before antibody is added, indicating a sample substance interacting
with particles, the assay result would be aborted. We did not observe
such a case in these studies on this particular analyte.
Both within-run and total precision were excellent for serum and plasma
determinations, as summarized in Table 3
. The data were obtained using a Dimension system (model AR) and
are representative of the precision observed for all four instrument
models used in this testing. Although precision on individual
instruments provides important information about the assay, it does not
indicate the total method variability such as might be observed in a
multisite proficiency survey. We thus performed an interlaboratory
comparison study as described above. The results are reported in Table 4
. The overall SD, which may be taken as a realistic predictor of
the variability expected in College of American Pathologists (CAP) or
other multianalyte surveys, indicate very good multilaboratory
performance with one reagent and one calibrator lot.
The results of split-sample studies, shown in Fig. 3
for the subject assay in comparison with the Abbott TDx
analyzer, show very good correlation. For maximum robustness of the
comparisons investigated, we used multiple lots of reagents and
multiple instruments and calibrators, as detailed in the caption of
Fig. 3
. The entire study occurred over 5 months and entailed five or
more calibrations for each instrument.
View larger version (24K):
[in this window]
[in a new window]
|
Figure 3. Comparison of gentamicin results as reported on the
Dimension system with results from the Abbott TDx assay.
The data represent 1171 duplicate determinations for 761 clinical
specimens. Each set of duplicates was performed on one of five
Dimension systems used in the total study (three in our laboratories
and two in separate clinical hospital laboratories). Three Abbott TDx
analyzers were used across the study. Three reagent lots and three
calibrator lots were used for the Dimension system, and two calibrator
lots and three reagent lots were used for TDx analyzers. Values for the
line are as follows: r = 0.987; n = 1171;
Sy|x = 0.78; slope = 0.98 ± 0.005;
intercept = -0.42 ± 0.039.
|
|
The Bland-Altman form of the difference plot (18)(19) is
also provided in Fig. 4
to show the measure of agreement between the two methods. It is
apparent that there is no obvious relationship between the differences
and measured concentrations. The mean difference and the SE of the mean
differences were calculated to be -0.57 and 0.023 µmol/L,
respectively. At the 95% confidence interval, this is a significant
but relatively minor systematic bias of -0.61 to -0.52 µmol/L
(-0.29 to -0.25 µg/mL).
View larger version (43K):
[in this window]
[in a new window]
|
Figure 4. Difference plot of the data used in Fig. 3
with the mean difference (bold line) and SD of the mean
difference (thin line).
Both 2 SD and 3 SD lines are shown
because >1000 determinations were used. SEM represents
the standard error of the mean differences.
|
|
A direct comparison of serum results with plasma results was done on
drug-free specimens to which gentamicin sulfate had been added. This
approach was used to demonstrate the equivalence of the two sample
matrices because of the lack of availability of matched draws from
patients being administered gentamicin. This study, which included the
anticoagulants sodium EDTA, lithium heparin, and potassium oxalate,
showed the equivalence of the two specimen types. The regression
statistics obtained were as follows: sodium EDTA result, 1.00 x
serum result + 0.04 (n = 21); lithium heparin result, 1.00 x
serum result - 0.04 (n = 22); and potassium oxalate result,
1.01 x serum result + 0.02 (n = 20). Actual patient plasma
specimens containing gentamicin, when compared using the
Dimension system and the Abbott TDx analyzer,
gave correlation slopes not statistically different from the
correlations with serum specimens.
The accuracy of the method was further evaluated by recovery of added
USP gentamicin standard (carefully dried) and by results for CAP survey
samples. By the addition technique, we found 102%, 105%, and 107%
recovery for 4.2, 8.4, and 16.7 µmol/L (2.0, 4.0, and 8.0 µg/mL) of
gentamicin. CAP survey samples Z-11 to Z-15 (1995) were tested on the
Dimension system model AR; the results were within evaluation criteria
for peer group means. For example, the means of five replicates of
samples measured on the Dimension system model AR were 15.9, 9.6, 2.9,
14.2, and 12.3 µmol/L (7.6, 4.6, 1.4, 6.8, and 5.9 µg/mL); the
means of these same samples in the CAP survey were 15.3, 8.8, 2.7,
13.6, and 12.1 µmol/L (7.3, 4.2, 1.3, 6.5, and 5.8 µg/mL),
respectively, for all the methods from >3700 clinical
laboratories.
The limit of detection was 0.63 µmol/L (0.3 µg/mL) when defined as
the concentration corresponding to 2 SD above the 0.0 µmol/L value
(n = 20). Linearity was assessed by fitting the data to a
quadratic model and by a test of significance of the coefficient of the
second-degree term (20). Linearity was thus found to extend
beyond 25.1 µmol/L (12 µg/mL; data not shown).
The stability of the reagents is important to the shelf life and
calibration interval of the product. Reagent cartridges, calibrated and
periodically measured over 90 days, showed a maximum rate of change of
5% over a 30-day period of testing for the four highest calibrator
concentrations. The zero-concentration calibrator showed no drift
outside the limit of detection [0.63 µmol/L (0.3 µg/mL)]. Based
on this, a 30-day calibration interval was assigned. In continuing
studies extending over 1 year using 30-day calibration intervals, the
overall drifts for all calibrators were <5%; thus, a shelf life of at
least 12 months was obtained for this method.
During the investigation of the assay chemistry reported here, we
performed experiments that allow us to comment on the possible
mechanism of the agglutination reaction. It is commonly believed that
latex agglutination immunoassay reactions proceed through linking of
particles by the two arms of the bivalent antibody, and this is
certainly true for some assays. In such cases, one expects a
bell-shaped curve for a plot of the rate of turbidity increase vs
antibody concentration. This is so because high antibody concentrations
would saturate all antigenic sites on particles, disallowing
cross-linking. We tested this, as shown in Fig. 5
. Even at molar excesses of antibody to particle of 350-fold, no
decrease in reaction rate was seen. Thus, the reaction shows behavior
similar in this respect to that of a phenytoin PETINIA assay described
recently (11). In that study, several of the authors of this
study proposed that bivalent bridging is, in fact, not necessary for
particle agglutination and confirmed this by demonstrating strong
immunospecific agglutination imparted by a (monovalent) F(ab) fragment.
We proposed that a rapid, immunospecific initial reaction of antibody
with antigenic sites on the particle triggers a second, rate-limiting
step of agglutination because of decreased colloidal stability imparted
by the presence of protein (IgG) on the particle surface. We believe
that is the case here, as well. Clearly, such behavior requires careful
selection of reaction conditions to avoid interferences from serum
constituents that could upset colloidal stability. The practicality of
finding such conditions is demonstrated by the excellent analytical
performance reported here. The PETINIA technology has been adapted to
six other therapeutic drugs on the Dimension system.
View larger version (17K):
[in this window]
[in a new window]
|
Figure 5. Rate of turbidity increase vs F(ab')2
concentration.
The y-axis is the rate of absorbance change at 340 nm
wavelength, as measured on a Cobas Bio® Analyzer (Roche
Diagnostics). The rates were determined over 30 s, starting
40 s after reaction initiation. The absorbance-vs-time
relationships were linear in this interval. The x-axis
gives the molar concentration of antibody F(ab')2 in the reaction
mixture. The particle concentration is 2 nmol/L.
|
|
We believe the addition of the gentamicin assay enhances the
utility of this system for highly accurate and precise monitoring
of therapeutic drugs in laboratory settings where workstation
consolidation is advantageous.
|
Acknowledgments
|
|---|
We thank P. Thammana for generating the monoclonal antibody against
gentamicin and David Hudson for technical support on calibrator work.
|
Footnotes
|
|---|
Dade Behring Inc., Glasgow Business Community, P.O. Box 6101, Newark, DE 19714-6101.
1 Nonstandard abbreviations: TDM, therapeutic drug monitoring; PETINIA, particle-enhanced turbidimetric inhibition immunoassay; USP, U.S. Pharmacopeia; and CAP, College of American Pathologists. 
2 2 Dimension® is a registered trademark of Dade Behring Inc. Abbott® and TDx® are registered trademarks of Abbott Laboratories. BIORAD®, BIORADTM, LIQUICHEKTM, and LYPHOCHEKTM are trademarks of Bio-Rad Laboratories. ZORBAX® is a registered trademark of E.I. DuPont deNemours Co., Inc.
|
References
|
|---|
-
Keys TF, Kurtz SB, Jones JD. Renal toxicity during therapy with gentamicin or tobramycin. Mayo Clin Proc 1981;56:556-559.
[ISI][Medline]
[Order article via Infotrieve]
-
Schentag JJ. Aminoglycosides. Evans W Schentag J Jusko W eds. Applied pharmacokinetics 1980:174-209 Applied Therapeutics San Francisco, CA. .
-
Zaske DE, Cipolle RJ, Strate RG. Gentamicin dosage requirements: wide interpatient variations in 242 surgery patients with normal renal function. Surgery 1980;87:164-169.
[ISI][Medline]
[Order article via Infotrieve]
-
Dudley PA, Edwards P, Ekins RP, Finney DJ, McKenzie IGM, Raad GM, et al. Guidelines for immunoassay data processing. Clin Chem 1985;31:1264-1271.
[Abstract/Free Full Text]
-
Litchfiled WJ, Craig AR, Frey WA, Leflar CC, Looney CE, Lluddy MA. Novel shell/core particles for automated immunoassays. Clin Chem 1984;30:1489-1493.
[Abstract/Free Full Text]
-
Burtis C, Ashwood E. Tietz textbook of clinical
chemistry, 2nd ed. Philadelphia, PA: WB Saunders, 1994:1131pp..
-
Kahn M. The history of rheumatoid factor. Rev Rhum Engl Ed 1995;62:7-8.
[Medline]
[Order article via Infotrieve]
-
Boscato LM, Stuart MC. Heterophilic antibodies: a problem for all immunoassays. Clin Chem 1988;34:27-33.
[Abstract/Free Full Text]
-
Hansen HJ, LaFontaine G, Newman ES, Schwartz MK, Malkin A, Majozisik K, et al. Solving the problem of antibody interference in commercial "sandwich"-type immunoassays of carcinoembryonic antigen. Clin Chem 1989;35:145-151.
-
Nahm HM, Hoffmann JW. Heteroantibody: phantom of the immunoassay. Clin Chem 1990;36:6.
-
Craig AR. A novel mechanism for latex agglutination
assays [Abstract]. 67th Annual CIC Conference. Ottawa, Canada:
Chemical Institute of Canada, 1984:M57–05..
-
Thompson JC, Craig AR, Davey CL, Newman DJ, Lonsdale ML, Bucher WJ, et al. Kinetics and proposed mechanism of the reaction of an immunoinhibition, particle-enhanced immunoassay. Clin Chem 1997;43:2384-2389.
[Abstract/Free Full Text]
-
Mariani M, Gamagna M, Tarditi L, Seccamani E. A new enzymatic method to obtain high yield F(ab)2 suitable for clinical use from mouse IgG1. Mol Immunol 1991;28:69-77.
[ISI][Medline]
[Order article via Infotrieve]
-
Koehler G, Milstein C. Continuous cultures of fused cells secreting antibody of pre-defined specificity. Nature 1975;256:495-497.
[Medline]
[Order article via Infotrieve]
-
Anhalt JP, Sancilio FD, McCorkle T. Gentamicin C-component ratio determination by high-pressure liquid chromatography. J Chromatogr 1978;153:389-493.
-
Vaidya HC, Beatty BG. Eliminating interference from heterophilic antibodies in a two-site immunoassay for creatine kinase MB by using F(ab')2 conjugate and polyclonal mouse IgG. Clin Chem 1992;38:1737-1742.
[Abstract/Free Full Text]
-
Knight D, Singer R, White JM, Fraser CG. Laboratory evaluation of the DuPont Clinical Chemistry System. Clin Chem 1988;34:1899-1903.
[Abstract/Free Full Text]
-
Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of agreement. Lancet 1986;i:307-310.
-
Peterson PH, Stockl D, Blaabjerg O, Pedersen B, Birkenose E, Thienpont L, et al. Graphical interpretation of analytical data from comparison of a field method with a reference method by use of difference plots. Clin Chem 1997;43:2039-2046.
[Abstract/Free Full Text]
-
Burnett RW. Quantitative evaluation of linearity. Clin Chem 1980;26:44-46.
The following articles in journals at HighWire Press have cited this article:
|
|
|
|
 
N. Isoherranen and S. Soback
Determination of Gentamicins C1, C1a, and C2 in Plasma and Urine by HPLC
Clin. Chem.,
June 1, 2000;
46(6):
837 - 842.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
