Clinical Chemistry / GADOLINIUM CONTRAST AGENT INTERFERENCE
Gadolinium Magnetic Resonance Contrast Agents
Produce Analytic Interference in Multiple Serum Assays
Kerry A.S. Proctor, MD,1 Lokinendi V. Rao, PhD,2 and William L. Roberts, MD, PhD1
Key Words: Gadodiamide; Gadoversetamide; Gadoteridol; Gadopentetate dimeglumine; Omniscan; Optimark; Magnevist; Prohance
DOI: 10.1309/MGA3LC4X8CGLX9CH
Abstract
Gadolinium magnetic resonance contrast agents
are known to interfere with some clinical chemistry
tests, particularly colorimetric assays for serum
calcium. We studied the effects of 4 agents,
gadodiamide, gadoversetamide, gadopentetate
dimeglumine, and gadoteridol, for interference with
multiple serum assays.
Gadodiamide and gadoversetamide produced
clinically significant negative interference with
colorimetric assays for serum angiotensin-converting
enzyme, calcium, and zinc. These agents produced
clinically significant positive interference in magnesium
and total iron binding capacity assays and both positive
and negative interference in iron assays. Gadopentetate
dimeglumine produced a negative interference with iron
assays, and gadopentetate dimeglumine and
gadoteridol produced negative interference with a
colorimetric zinc assay.
Caution should be exercised when using
colorimetric assays for angiotensin-converting enzyme,
calcium, iron, magnesium, total iron binding capacity,
and zinc in serum samples from patients who have
recently received magnetic resonance contrast agents.
In general, gadodiamide and gadoversetamide are more
likely to produce a clinically significant interference
than gadopentetate dimeglumine and gadoteridol.
Likewise, certain analytic methods are more prone to
interference, while others not affected.
282
282
Am J Clin Pathol 2004;121:282-292
DOI: 10.1309/MGA3LC4X8CGLX9CH
Interfering substances are an important potential source
of error in laboratory analyses that can be present in otherwise normal clinical specimens. While accuracy and precision are assessed routinely, interfering substances often are
not suspected or recognized. Substances known to interfere
with laboratory assays arise from endogenous (bilirubin,
protein, lipids, hemoglobin) and exogenous (therapeutic and
recreational drugs and their metabolites) sources. An awareness of potential interference can help reduce the risk of
reporting erroneous laboratory values that could affect patient
care adversely.
Contrast media are used commonly to enhance T 1 weighted images in magnetic resonance (MR) imaging.
Gadolinium (Gd), a lanthanide ion with 7 unpaired electrons
and an especially long electronic relaxation time, is widely
used as a contrast agent.1 Because it is intrinsically toxic, in
part by blocking calcium channels, Gd must be chelated with
an appropriate ligand to allow clinical use. 2 These Gd
complexes can be linear or macrocyclic. Gd contrast agents
that are used clinically are known to be powerful chelators
and have been reported to interfere with laboratory assays.
An early report described artifactual decreases in serum
calcium measurements after the administration of gadodiamide. 3 This interference was noted when colorimetric
methods of calcium analysis were used, such as o-cresolphthalein complexone (OCPC), methylthymol blue (MTB)
and Arsenazo III methods but not with inductively coupled
plasma–atomic emission spectroscopy or ion-selective electrode methods. These authors postulated that gadodiamide
dissociates under the acidic conditions of the colorimetric
assays, causing the chromophore to bind with the free ligand,
thereby causing apparent decreases in measured calcium as
© American Society for Clinical Pathology
Clinical Chemistry / ORIGINAL ARTICLE
fewer chromophore-calcium complexes were available for
measurement. Negative interference by gadodiamide with
OCPC and MTB serum calcium assays was confirmed.4
These authors presented evidence that the cause of the interference was due to dissociation of Gd3+ ions from their
ligand and complexation with the chromophore used in the
calcium assay. Recently, a large retrospective study showed
spurious decreases in serum calcium concentrations in
patients following gadodiamide administration.5 In addition,
these authors studied 3 other commonly used MR contrast
agents, gadoversetamide, gadopentetate dimeglumine, and
gadoteridol, and found that gadoversetamide interfered with
a single OCPC calcium assay, but gadopentetate dimeglumine and gadoteridol did not. This study did not examine
any other calcium methods.
Gd MR contrast agents have been reported to show
interference with other analytes. Increases in serum iron and
bilirubin concentrations were observed in healthy volunteers
after gadopentetate dimeglumine administration and attributed to mild hemolysis during the venipuncture.6 Published
data also indicated that angiotensin-converting enzyme
(ACE) activity could be inhibited by gadodiamide and
gadopentetate dimeglumine due to a transmetallation
effect.7,8 These authors observed in vivo and in vitro effects
that were strongest with linear Gd complexes.
Despite these scattered reports, to the best of our knowledge, there are no detailed reports about the potential effects
of various Gd contrast agents on common analytes measured
in the clinical laboratory. At the University of Utah Hospital,
Salt Lake City, it is estimated that more than 700 MR scans
are done per month and that in more than half of these,
patients receive Gd contrast agents to enhance imaging
before the procedure. Hospital inpatients, in particular, are
more likely to have blood work done after having received a
Gd contrast agent and, therefore, might be most prone to
have misleading laboratory results.
The goal of the present study was to assess potential
analytic interference by 4 commonly used Gd contrast
agents for multiple analytes and multiple analyzers. Experimentally, we focused on interference with analytic
methods using serum samples supplemented with the
agents in vitro. Because different institutions might
encounter different combinations of contrast agents and
analytic analytes, we attempted to evaluate multiple
routine methods for selected methods that we thought
might be the most subject to clinically significant interference, the divalent cations in particular. We hope that by
increasing awareness about Gd interference and by
providing a more complete list of affected analytes and
specific methods of analysis that are free of interference,
we can help reduce the number of potentially erroneous
laboratory results generated.
Materials and Methods
Four commercially available Gd contrast agents were
studied: gadodiamide (Omniscan, Amersham Health,
Princeton, NJ); gadoversetamide, (Optimark, Mallinckrodt, St
Louis, MO); gadopentetate dimeglumine (Magnevist, Berlex,
Wayne, NJ); and gadoteridol (Prohance, Bracco, Princeton,
NJ). All 4 agents have similar volumes of distributions, peak
concentrations, dosing recommendations, and renal clearance.
The Gd concentration of each agent for intravenous
administration is 0.5 mol/L. A 10-fold dilution with deionized water was performed to a Gd concentration of 50
mmol/L. Pooled human serum was supplemented with the 4
contrast agents to a final concentration of 0.5 mmol/L, corresponding to the estimated peak serum concentration
achieved in vivo for each agent at a standard dose of 0.1
mmol/kg. From the initial 0.5-mmol/L concentrations,
further dilutions were made with serum to Gd concentrations
of 0.25, 0.125, and 0.0625 mmol/L to simulate the decreases
seen in vivo as the agents are cleared from the body. All
studies using samples from human subjects were approved
by the institutional review board of the University of Utah
Health Sciences Center, Salt Lake City.
Initially, 41 analytes were assayed in duplicate using a
Modular P chemistry analyzer in samples containing 0.5mmol/L concentrations of the 4 agents ❚Table 1❚ . All
reagents were used according to the manufacturer’s instructions. Nonsupplemented serum served as a control. Six additional analytes were studied using other analyzers (Table 1).
Any analytes showing clinically significant differences in the
presence of a contrast agent were rerun, using serial dilutions
of the agent.
Calcium was studied further by using the AU 400 and
Dimension RxL analyzers (Table 1), both of which have an
OCPC method of analysis. Two instruments that use Arsenazo III dye methods for calcium measurement, the AU 400
(both OCPC and Arsenazo III methods are available from
Olympus) and Vitros 950, an ion-selective electrode method
on the Synchron LX-20, and inductively coupled plasma
mass spectrometry (ICP-MS) also were used (Table 1).
Ionized calcium was measured by using the Bayer model
865 blood gas analyzer (Table 1).
Total iron binding capacity (TIBC) was studied with the
Dimension RxL analyzer. Iron assays were performed on the
Vitros 950 and Synchron LX-20 analyzers and by ICP-MS.
The Dimension RxL and Vitros 950 iron and TIBC assays
use ferene. The Modular P and Synchron LX-20 iron and
TIBC assays use ferrozine. Magnesium was studied with the
Dimension RxL, Modular P, Synchron LX-20, and Vitros
950 analyzers, which use MTB, xylidyl blue, calmagite, and
formazan, respectively, and by ICP-MS. Zinc was assayed
using a manual colorimetric kit (Table 1) and by ICP-MS.
Am J Clin Pathol 2004;121:282-292
© American Society for Clinical Pathology
283
DOI: 10.1309/MGA3LC4X8CGLX9CH
283
283
Proctor et al / GADOLINIUM CONTRAST AGENT INTERFERENCE
❚Table 1❚
Methods and Analytes Used*
Method
Modular P, Roche Diagnostics, Indianapolis, IN
Advia Centaur, Bayer Diagnostics, Tarrytown, NY
AU 400, Olympus, Melville, NY
AxSYM, Abbott Diagnostics, Abbott Park, IL
Bayer 865, Bayer Diagnostics
Dimension RxL, Dade Behring, Deerfield, IL
ICP-MS
IMMULITE 2000, Diagnostic Products Corp, Los Angeles, CA
Manual colorimetric (Wako Diagnostics)
Synchron LX-20, Beckman Coulter, Brea, CA
Vitros 950, Ortho Clinical Diagnostics, Raritan, NJ
Analyte
Albumin
Aldolase
Alkaline phosphatase
Amylase
Amylase, pancreatic
α1-Antitrypsin
Angiotensin-converting enzyme (reagent from Trinity Biotech, St Louis, MO)
Alanine amino transferase
Aspartate amino transferase
Bilirubin, direct
Bilirubin, total
Calcium
Carbon dioxide, total
Ceruloplasmin
Chloride
Cholesterol
Complement C3
Complement C4
Creatine kinase
Creatinine
Fructosamine
Glucose
γ-Glutamyltranspeptidase
Haptoglobin
HDL-cholesterol
Iron
Lactate dehydrogenase
LDL-cholesterol (reagent from Genzyme Diagnostics, Cambridge, MA)
Lipase
Lipoprotein (a) (reagent from Wako Diagnostics, Richmond, VA)
Magnesium
Phospholipids
Phosphorus
Potassium
Protein, total
Sodium
TIBC
Transferrin
Triglycerides
Urea nitrogen
Uric acid
Folate
Thyroxine
Thyroid stimulating hormone
Vitamin B12
Calcium OCPC
Calcium Arsenazo
α-Fetoprotein
Ionized calcium
Calcium
Iron
TIBC
Calcium
Copper
Iron
Magnesium
Zinc
Parathyroid hormone
Zinc
Calcium
Iron
Magnesium
Calcium
Iron
Magnesium
ICP-MS, inductively coupled plasma mass spectrometry; HDL, high-density lipoprotein; LDL, low-density lipoprotein; OCPC, o-cresolphthalein complexone; TIBC, total iron
binding capacity.
* Reagents were obtained from the manufacturer of the analyzer except as noted.
284
284
Am J Clin Pathol 2004;121:282-292
DOI: 10.1309/MGA3LC4X8CGLX9CH
© American Society for Clinical Pathology
Clinical Chemistry / ORIGINAL ARTICLE
The percentage of recovery of the analyte in the presence of contrast agent was calculated by taking the average
of supplemented serum run in duplicate, dividing by the
average of the nonsupplemented serum run in duplicate, and
multiplying by 100. No correction was made for dilution
with the contrast agents because this introduced less than 1%
error. Significant findings were defined as any interference
greater than 2% for sodium, greater than 5% for calcium and
chloride, and greater than 10% for all other analytes. The
SEM was calculated for each data point for any analyte that
exhibited a clinically significant effect.
Results
We studied the effects of all 4 MR contrast agents on all
analytes listed in Table 1. The results of clinically significant
interference are summarized in ❚Table 2❚. Interference was
found initially for 3 assays on the Modular P analyzer. When
compared with control results, significant differences in
recoveries were seen for ACE (26% and 32%), calcium
(80% and 79%), and TIBC (127% and 129%) in serum
supplemented with 0.5-mmol/L concentrations of gadodiamide and gadoversetamide, respectively. For these 3
analytes, serial dilutions of the contrast agents were made to
examine the concentration dependence of the interference.
Results of these experiments for ACE are shown in ❚Figure
1❚, for calcium in ❚Figure 2❚, and for TIBC in ❚Figure 3❚.
Because gadodiamide- and gadoversetamide-supplemented samples showed significant interference with calcium
measurements on the Modular P instrument, which uses the
OCPC method of analysis, further investigation was performed
with the AU 400 and Dimension RxL analyzers that both use
an OCPC method. Calcium recoveries of 76% and 78% for the
AU 400 and 91% and 91% for the Dimension RxL analyzers
were observed with 0.5-mmol/L concentrations of gadodiamide and gadoversetamide, respectively. With increasing dilutions, this interference was shown to decrease in a near linear
manner (Figure 2). Two analyzers that use Arsenazo III dye
methods for calcium analysis (AU 400 and Vitros 950), an ionselective electrode method (Beckman Coulter Synchron LX20), an ionized calcium method, and ICP-MS showed no interference with calcium measurements by any of the 4 agents.
In addition to the interference with TIBC measurements on the Modular P, interference was demonstrated on
the Dimension RxL analyzer with recoveries of 149% for
gadodiamide and 121% for gadoversetamide (0.5-mmol/L
concentrations, Figure 3). Owing to the increases in TIBC
observed, we studied iron recoveries, which had not shown a
significant interference when assayed on the Modular P
analyzer. Significant decreases were observed for the
Synchron LX-20 (75% and 78%, at 0.5-mmol/L concentrations,
❚Table 2❚
Analytes Showing Clinically Significant Interference From a Contrast Agent*
Analyte
ACE
Calcium
Iron
Magnesium
TIBC
Zinc
Method
Modular P
AU 400 (Arsenazo)
AU 400 (OCPC)
Bayer 865 (ionized)
Dimension RxL
ICP-MS
Modular P
Synchron LX-20
Vitros 950
Dimension RxL
ICP-MS
Modular P
Synchron LX-20
Vitros 950
Dimension RxL
ICP-MS
Modular P
Synchron LX-20
Vitros 950
Modular P
Dimension
ICP-MS
Manual colorimetric
Gadodiamide
(Omniscan)
Gadoversetamide
(Optimark)
Gadopentetate
Dimeglumine (Magnevist)
Gadoteridol
(Prohance)
26†
101
76†
95
91†
98
80†
100
102
101
107
99
75†
118†
98
100
105
117†
105
127†
149†
94
15†
32†
101
78†
95
91†
99
79†
101
102
101
107
97
78†
114†‡
95
96
100
124†
100
129†
121†
97
26†
91
97
96
98
95
96
100
100
97
96
104
97
88†
89†
95
96
100
100
98
97
96
99
42†
91
97
97
98
96
97
101
99
97
99
105
99
98
97
95
96
95
98
95
96
95
97
87†
ACE, angiotensin-converting enzyme; ICP-MS, inductively coupled plasma mass spectrometry; OCPC, o-cresolphthalein complexone; TIBC, total iron binding capacity.
* Data are given as the percentage of recovery, which is shown for a 0.5-mmol/L concentration of each contrast agent unless otherwise indicated. For product information, see the
“Materials and Methods” section and Table 1.
† This recovery represents a clinically significant interference as defined in the “Materials and Methods” section.
‡ This recovery is for a gadoversetamide concentration of 0.25 mmol/L.
Am J Clin Pathol 2004;121:282-292
© American Society for Clinical Pathology
285
DOI: 10.1309/MGA3LC4X8CGLX9CH
285
285
Proctor et al / GADOLINIUM CONTRAST AGENT INTERFERENCE
A
B
100
Modular P ACE Recovery (%)
Modular P ACE Recovery (%)
100
75
50
25
0
75
50
25
0
0
0.1
0.2
0.3
0.4
0.5
Gadodiamide Concentration (mmol/L)
0
0.1
0.2
0.3
0.4
0.5
Gadoversetamide Concentration (mmol/L)
❚Figure 1❚ Effects of gadodiamide (A) and gadoversetamide (B) on the angiotensin-converting enzyme (ACE) assay. The error
bars represent the SEM. In some cases, the solid square for the data point is as large as or larger than the error bar. For product
information, see the “Materials and Methods” section and Table 1.
respectively), and increases in iron recoveries were observed
for gadodiamide and gadoversetamide on the Vitros 950
(118% at a 0.5-mmol/L concentration and 114% at a 0.25mmol/L concentration, respectively). Iron recovery was 88%
and 89% with a 0.5-mmol/L concentration of gadopentetate
dimeglumine on the Synchron LX-20 and Vitros 950
analyzers, respectively ❚Figure 4❚. Dilutions revealed a dosedependent effect. No interference with iron measurements by
the Dimension RxL or ICP-MS methods was observed.
Although magnesium measurements were not affected
by any of the 4 agents on the Modular P analyzer, we studied
magnesium measurements with additional analyzers because
magnesium is a divalent cation with chemical properties
similar to calcium. Magnesium recoveries were not affected
by any of the 4 agents on the Dimension RxL or Vitros 950
analyzers. However, 0.5-mmol/L concentrations of gadodiamide and gadoversetamide led to increased magnesium
recoveries on the Synchron LX-20 analyzer of 117% and
124%, respectively ❚Figure 5❚. No interference was shown
for magnesium when quantified by ICP-MS.
Because ACE, a zinc-dependent metallopeptidase,
initially showed interference on the Modular P, we measured
zinc with a manual colorimetric assay in samples supplemented with a 0.5-mmol/L concentration of all 4 contrast
agents. Significant interference was observed for all 4 agents
with recoveries as follows: 15% for gadodiamide, 26% for
gadoversetamide, 42% for gadopentetate dimeglumine, and
87% for gadoteridol ❚Figure 6❚. The interference was dose
dependent but not linear. No interference was shown for zinc
when measured by ICP-MS.
286
286
Am J Clin Pathol 2004;121:282-292
DOI: 10.1309/MGA3LC4X8CGLX9CH
Total bilirubin showed borderline interference with 86%
recovery for gadodiamide, gadoversetamide, and gadopentetate dimeglumine that we suspected might be secondary to
the low bilirubin concentration (0.7 mg/dL [12.0 µmol/L])
present in the original serum pool. Therefore, the experiment
was repeated on a serum pool with a total bilirubin concentration of 8.7 mg/dL (148.8 µmol/L). Supplementation with
each of the 4 agents at a concentration of 0.5 mmol/L did not
produce a clinically significant interference (recoveries,
105%-106%), suggesting the original findings might be due
to poor assay precision at a low bilirubin concentration. No
clinically significant interference was noted for any of the
immunoassays evaluated on dedicated immunoassay
analyzers that were evaluated.
Discussion
Our results show that Gd MR contrast agents can
produce analytic interference, both positive and negative,
with assays performed in the clinical laboratory, including
ACE, calcium, iron, magnesium, TIBC, and zinc. Gadodiamide and gadoversetamide produced the most interference, although iron measured in samples supplemented
with gadopentetate dimeglumine showed interference and a
manual colorimetric zinc assay showed interference with
gadodiamide, gadoversetamide, gadopentetate dimeglumine, and gadoteridol. All of the analytes affected are
endogenous divalent cations or related to divalent cations in
some manner.
© American Society for Clinical Pathology
Clinical Chemistry / ORIGINAL ARTICLE
A
B
95
100
AU 400 Calcium Recovery (%)
AU 400 Calcium Recovery (%)
100
90
85
80
95
90
85
80
75
75
0
0.1
0.2
0.3
0.4
0.5
0
Gadodiamide Concentration (mmol/L)
C
D
Dimension RxL Calcium Recovery (%)
Dimension RxL Calcium Recovery (%)
0.3
0.4
0.5
100
95
90
85
80
95
90
85
80
75
75
0
0.1
0.2
0.3
0.4
0
0.5
0.1
0.2
0.3
0.4
0.5
Gadoversetamide Concentration (mmol/L)
Gadodiamide Concentration (mmol/L)
F
100
Modular P Calcium Recovery (%)
100
Modular P Calcium Recovery (%)
0.2
Gadoversetamide Concentration (mmol/L)
100
E
0.1
95
90
85
80
75
95
90
85
80
75
0
0.1
0.2
0.3
0.4
Gadodiamide Concentration (mmol/L)
0.5
0
0.1
0.2
0.3
0.4
0.5
Gadoversetamide Concentration (mmol/L)
❚Figure 2❚ Effects of gadodiamide (A, C, E) and gadoversetamide (B, D, F) on AU 400, Dimension RxL, and Modular P
o-cresolphthalein complexone calcium methods. The error bars represent the SEM. In some cases, the solid square for the data
point is as large as or larger than the error bar. For product information, see the “Materials and Methods” section and Table 1.
Am J Clin Pathol 2004;121:282-292
© American Society for Clinical Pathology
287
DOI: 10.1309/MGA3LC4X8CGLX9CH
287
287
Proctor et al / GADOLINIUM CONTRAST AGENT INTERFERENCE
A
B
150
Dimension RxL TIBC Recovery (%)
Dimension RxL TIBC Recovery (%)
150
140
130
120
110
140
130
120
110
100
100
0
0.1
0.2
0.3
0.4
0
0.5
C
D
150
0.2
0.3
0.4
0.5
140
150
Modular P TIBC Recovery (%)
Modular P TIBC Recovery (%)
0.1
Gadoversetamide Concentration (mmol/L)
Gadodiamide Concentration (mmol/L)
130
120
110
100
140
130
120
110
100
0
0.1
0.2
0.3
0.4
0.5
Gadodiamide Concentration (mmol/L)
0
0.1
0.2
0.3
0.4
0.5
Gadoversetamide Concentration (mmol/L)
❚Figure 3❚ Effects of gadodiamide (A, C) and gadoversetamide (B, D) on total iron binding capacity (TIBC) measurements on
the Dimension RxL and Modular P analyzers. The error bars represent the SEM. In some cases, the solid square for the data
point is as large as or larger than the error bar. For product information, see the “Materials and Methods” section and Table 1.
Because interference observed for calcium, iron,
magnesium, and zinc was not seen with all methods tested,
contamination of the contrast agents with any of these
elements is unlikely. All analytes with which interference
was found, except ACE, use chromophore dye methods of
analysis. The Gd contrast agents can dissociate, and a
complex can form between Gd3+ ions and the assay chromophore, and/or a complex between divalent cations and the
dissociated ligand can form.4 Interference can occur when
the Gd3+ binds with the chromophore or the contrast agent
ligand binds with the analyte being measured. For the
former, the interference might be positive or negative, while
for the latter, it would be expected to be negative. It is noteworthy that for the Vitros 950 iron method, gadodiamide and
288
288
Am J Clin Pathol 2004;121:282-292
DOI: 10.1309/MGA3LC4X8CGLX9CH
gadoversetamide produced clinically significant positive
interference, while gadopentetate dimeglumine produced a
clinically significant negative interference. The exact mechanism of this interference is unclear.
The ACE assay measures the rate of conversion of N-[3(2-furyl)acryloyl]-L-phenylalanylglycylglycine to furylacryloylphenylalanine and glycylglycine spectrophotometrically
under alkaline conditions. Because ACE depends on zinc for
activity, it is likely that the dissociated contrast agent ligand
binds zinc and reduces the rate of reaction, giving a falsely
low ACE activity. Interference was observed for gadodiamide and gadoversetamide. No interference was seen with
gadopentetate dimeglumine. This is in contrast with results
from an earlier study that found inhibition of ACE by
© American Society for Clinical Pathology
Clinical Chemistry / ORIGINAL ARTICLE
A
B
125
Synchron LX-20 Iron Recovery (%)
Synchron LX-20 Iron Recovery (%)
125
115
105
95
85
75
115
105
95
85
75
0
0.1
0.2
0.3
0.4
0.5
0
Gadodiamide Concentration (mmol/L)
C
D
0.3
0.4
0.5
115
115
105
95
85
105
95
85
75
0
0.1
0.2
0.3
0.4
0
0.5
F
125
105
95
85
0.2
0.3
0.4
0.5
125
Vitros 950 Iron Recovery (%)
115
0.1
Gadodiamide Concentration (mmol/L)
Gadopentetate Dimeglumine Concentration (mmol/L)
Vitros 950 Iron Recovery (%)
0.2
125
Vitros 950 Iron Recovery (%)
Synchron LX-20 Iron Recovery (%)
125
75
E
0.1
Gadoversetamide Concentration (mmol/L)
115
105
95
85
75
75
0
0.1
0.2
0.3
0.4
Gadoversetamide Concentration (mmol/L)
0.5
0
0.1
0.2
0.3
0.4
0.5
Gadopentetate Dimeglumine Concentration (mmol/L)
❚Figure 4❚ Effects of gadodiamide (A, D), gadoversetamide (B, E), and gadopentetate dimeglumine (C, F) on Synchron LX-20
and Vitros 950 iron assays. The error bars represent the SEM. In some cases, the solid square for the data point is as large as
or larger than the error bar. For product information, see the “Materials and Methods” section and Table 1.
Am J Clin Pathol 2004;121:282-292
© American Society for Clinical Pathology
289
DOI: 10.1309/MGA3LC4X8CGLX9CH
289
289
Proctor et al / GADOLINIUM CONTRAST AGENT INTERFERENCE
A
B
130
Synchron LX-20 Magnesium Recovery (%)
Synchron LX-20 Magnesium Recovery (%)
130
125
120
115
110
105
125
120
115
110
105
100
100
0
0.1
0.2
0.3
0.4
0.5
Gadodiamide Concentration (mmol/L)
0
0.1
0.2
0.3
0.4
0.5
Gadoversetamide Concentration (mmol/L)
❚Figure 5❚ Effects of gadodiamide (A) and gadoversetamide (B) on the Synchron LX-20 magnesium assay. The error bars
represent the SEM. In some cases, the solid square for the data point is as large as or larger than the error bar. For product
information, see the “Materials and Methods” section and Table 1.
gadopentetate dimeglumine that was comparable to that
observed for gadodiamide.5 One possible explanation is that
the concentration of free ligand in gadopentetate dimeglumine preparations has been reduced from previous levels.
Gadodiamide and gadoversetamide were the only 2
agents that showed interference with the calcium measurements. The AU 400, Dimension RxL, and Modular P instruments all use an OCPC method for calcium analysis. These
assays operate under alkaline conditions, and it is possible
that the Gd3+ binds to the OCPC dye. Consequently, less
chromophore is available to bind with calcium ions and a
negative interference results. Our results for the Modular P
calcium method are consistent with those in a recent report.5
We did not find a clinically significant interference with the
Arsenazo III methods of calcium analysis on the AU 400 and
Vitros 950. It is interesting that these assays are performed
under acidic conditions. A previous study that examined the
Kodak Ektachem method (now Vitros) found a negative
interference of greater than 5%, but this was with a gadodiamide concentration that was 3 times that used in our study.3
We used serum supplemented with the peak concentration of
each agent that would be achieved in vivo with a standard
dose (0.1 mmol/kg). Higher concentrations can be achieved
clinically, as in MR angiography, so that more analytes and
methods could be affected and the magnitude of the effect
could be greater than we observed.
In the iron assays, the Synchron LX-20 uses ferrozine as
the chromophore dye; it is interesting that the Modular P also
uses ferrozine, but no effect was seen on iron measurement
with that platform. The Dimension RxL uses the ferene chromophore, and no interference was detected with this assay. All
290
290
Am J Clin Pathol 2004;121:282-292
DOI: 10.1309/MGA3LC4X8CGLX9CH
iron methods studied include a step to reduce ferric to ferrous
iron, which then is available to bind with the chromophore. All
iron assays studied occur under acidic test conditions, making
dissociation of the Gd contrast agent unlikely. The Dimension
RxL, Modular P, and Vitros 950 use ascorbic acid as a
reducing agent, whereas the Synchron LX-20 uses hydroxylamine and thioglycolate. It is possible that these 2
compounds unique to this assay account for the negative interference seen in this assay. The Vitros 950 uses a sulfonamide
chromophore that is different from the others and could cause
the positive interference seen with the iron measurement on
that assay; we are unsure why there was negative interference
with gadopentetate dimeglumine in the same assay.
The only magnesium measurement method demonstrating interference was the Synchron LX-20, in which positive interference was seen with gadodiamide and gadoversetamide. All methods use a chromophore to bind with
magnesium that then is measured by spectrophotometry. No
effect was seen with the Dimension RxL, which uses MTB;
the Modular P, which uses xylidyl blue; or the Vitros 950,
which uses formazan dye. The Synchron LX-20 uses
calmagite dye. This assay is performed under alkaline conditions, as is the xylidyl blue assay on the Modular P analyzer.
Whereas the Dimension RxL, Modular P, and Vitros 950 all
include compounds to chelate calcium that potentially might
interfere with the assay, no calcium chelator is mentioned in
the Synchron LX-20 package insert. Perhaps the alkaline
conditions of the Synchron LX-20 assay could promote association between Gd 3+ and the chromophore, creating a
colored complex resembling a magnesium-chromophore
complex and causing a positive interference.
© American Society for Clinical Pathology
Clinical Chemistry / ORIGINAL ARTICLE
A
B
100
100
75
Zinc Recovery (%)
Zinc Recovery (%)
75
50
25
25
0
0
0
0.1
0.2
0.3
0.4
0.5
0
Gadodiamide Concentration (mmol/L)
C
0.1
0.2
0.3
0.4
0.5
Gadopentetate Dimeglumine Concentration (mmol/L)
D
100
100
75
75
Zinc Recovery (%)
Zinc Recovery (%)
50
50
25
50
25
0
0
0
0.1
0.2
0.3
0.4
0.5
Gadoteridol Concentration (mmol/L)
0
0.1
0.2
0.3
0.4
0.5
Gadoversetamide Concentration (mmol/L)
❚Figure 6❚ Effects of gadodiamide (A), gadopentetate dimeglumine (B), gadoteridol (C), and gadoversetamide (D) on the manual
zinc assay. The error bars represent the SEM. In some cases, the solid square for the data point is as large as or larger than the
error bar. For product information, see the “Materials and Methods” section.
Positive interference was seen with TIBC on the Dimension RxL and Modular P analyzers with gadodiamide and
gadoversetamide, while both agents produced negative interference for the iron assay on the Synchron LX-20 and positive interference in the Vitros 950 iron assay. It is interesting
that gadopentetate dimeglumine produced negative interference in the Vitros 950 iron assay. Previously, increases in
serum iron and bilirubin concentrations observed after administration of gadopentetate dimeglumine were attributed to
slight hemolysis during venipuncture.6 The positive interference we observed for iron was not secondary to RBC breakdown because it occurs in the absence of RBCs. We also did
not observe increased bilirubin values, as was reported by the
same authors. Both TIBC assays use chromophores that bind
to ferrous iron. In the first step of both assays, free MR
contrast agent ligand could bind some excess added ferrous
iron under alkaline conditions, causing less iron to be available to bind to the unsaturated transferrin sites. This then
could result in a higher unbound iron binding capacity
measurement and an elevation in TIBC as seen here.
While no interference was seen with the ICP-MS method
of zinc analysis, negative interference was seen with all 4
contrast agents on a manual colorimetric assay. The kit used
for these measurements used formazan dye and an alkaline
buffer. Binding of the formazan dye by the Gd3+ under alkaline conditions could account for this negative interference.
Am J Clin Pathol 2004;121:282-292
© American Society for Clinical Pathology
291
DOI: 10.1309/MGA3LC4X8CGLX9CH
291
291
Proctor et al / GADOLINIUM CONTRAST AGENT INTERFERENCE
The greatest amount of interference was observed with gadodiamide and gadoversetamide. The large error bars on Figure
6 result from the manual nature of this assay.
Of all the affected analytes, calcium is the most likely to
produce a result that leads to unnecessary treatment.5 This
could adversely affect patient care. If the calcium concentration must be measured before elimination is complete, we
recommend administering an agent other than gadodiamide
or gadoversetamide, using a non-OCPC method or
measuring ionized calcium. It is noteworthy that the effect of
calcium has been documented to occur after in vivo administration of a contrast agent.3 It also is important to recognize
potential interference with other analytes. One limitation of
our data for additional analytes is that it was all generated in
vitro. Further verification of our findings should be
conducted using data from patients or research subjects from
whom serum samples are obtained before and after administration of the contrast agent. The possibility exists after in
vivo administration of a contrast agent that the formation of
metabolites could alter an agent’s interference profile.
Based on our in vitro data, additional warnings of the
interference described herein seem to be necessary for Gd
contrast agents and in vitro diagnostic assays. Currently, the
package inserts for gadodiamide and gadoversetamide state
that these products interfere with calcium measurements by
some colorimetric assays. The Modular P calcium assay
package insert includes Gd agents as potential causes of
interference and mentions gadodiamide specifically but does
not mention gadoversetamide.
Based on these findings, we recommend using care
when interpreting results of the analyte-method combinations listed in Table 2 for patients who recently have received
Gd contrast agents, particularly gadodiamide and gadoversetamide. It is estimated that with normal renal function, the
elimination half-life of the agents is approximately 90
minutes. The package insert for gadodiamide recommends
waiting 12 to 24 hours between contrast agent administration
and blood specimen collection to ensure that the contrast
agent has been cleared. A longer waiting period might be
necessary for patients with renal insufficiency. The average
half-life of gadodiamide in patients with severely reduced
renal function is 34 hours.9
292
292
Am J Clin Pathol 2004;121:282-292
DOI: 10.1309/MGA3LC4X8CGLX9CH
From the Departments of Pathology, 1University of Utah Health
Sciences Center, Salt Lake City, and 2UMASS Memorial Medical
Center, Worcester, MA.
Supported by the ARUP Institute for Clinical &
Experimental Pathology.
Address reprint requests to Dr Roberts: ARUP Laboratories,
500 Chipeta Way, Salt Lake City, UT 84108.
Acknowledgments: Gadoversetamide and gadopentetate
dimeglumine were graciously provided by Mallinckrodt and
Berlex, respectively.
References
1. Runge VM, Wells JW. Update: safety, new applications, new
MR agents. Top Magn Reson Imaging. 1995;7:181-196.
2. Bourne GW, Trifaro JM. The gadolinium ion: a potent
blocker of calcium channels and catecholamine release from
cultured chromaffin cells. Neuroscience. 1982;7:1615-1622.
3. Normann PT, Froysa A, Svaland M. Interference of
gadodiamide injection (OMNISCAN) on the colorimetric
determination of serum calcium. Scand J Clin Lab Invest.
1995;55:421-426.
4. Lin J, Idee J-M, Port M, et al. Interference of magnetic
resonance imaging contrast agents with the serum calcium
measurement technique using colorimetric reagents. J Pharm
Biomed Anal. 1999;21:931-943.
5. Prince MR, Erel HE, Lent RW, et al. Gadodiamide
administration causes spurious hypocalcemia. Radiology.
2003;227:639-646.
6. Niendorf HP, Seifert W. Serum iron and serum bilirubin after
administration of Gd-DTPA-dimeglumine: a pharmacologic
study in healthy volunteers. Invest Radiol. 1988;23(suppl
1):S275-S280.
7. Corot C, Hentsch AM, Curtelin L. Interaction of gadolinium
complexes with metal-dependent biological systems. Invest
Radiol. 1994;29(suppl 1):S164-S167.
8. Corot C, Idee JM, Hentsch A-M, et al. Structure-activity
relationship of macrocyclic and linear gadolinium chelates:
investigation of transmetallation effect on the zinc-dependent
metallopeptidase angiotensin-converting enzyme. J Magn
Reson Imaging. 1998;8:695-702.
9. Joffe P, Thomsen HS, Meusel M. Pharmacokinetics of
gadodiamide injection in patients with severe renal
insufficiency and patients undergoing hemodialysis or
continuous ambulatory peritoneal dialysis. Acad Radiol.
1998;5:491-502.
© American Society for Clinical Pathology