| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Articles |
1
Department of Biochemistry, Wythenshawe Hospital, Southmoor Rd., Manchester M23 9LT, UK.
2
Department of Biochemistry, Hope Hospital, Stott Lane,
Manchester M6 8HD, UK.
a Author for correspondence. Fax 0161-291-2125; e-mail juliewassell{at}hotmail.com
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Methods: Urine samples collected from patients on a variety of medications were grouped by specific drug type. The significance of any difference in the free catecholamine or total metanephrine concentrations in the different groups was assessed by one-way ANOVA. A group of patients receiving no medication was included as a control (no analytical interference). Additionally, analytical accuracy, detection limit, and precision were determined.
Results: No significant differences were found in the concentrations of free catecholamines or total metanephrines in urine from patients taking the medications investigated and the control group: P = 0.649 (fE), 0.221 (fNE), 0.149 (tM), and 0.170 (tNM). For free catecholamines, intraassay CVs were 4.6–18%; interassay CVs were 10–25%. For total metanephrines, intraassay CVs were 9.6–27%; interassay CVs were 5.8–22%. Detection limits were 0.009 and 0.027 µmol/L for fE and fNE and 0.119 and 0.346 µmol/L for tM and tNM, respectively.
Conclusions: None of the drugs examined in this study interfered in the measurement of free catecholamines or total metanephrines by these immunoassays. The technique is easier to use, requires less equipment, and is more accessible than HPLC. In combination, these assays are suitable as initial screening tests for pheochromocytoma.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
HPLC is the most commonly used technique for measuring fE, fNE, tM, and tNM in urine. Methods, however, vary between laboratories (2), with some procedures being prone to interference (7)(8)(9), often by antihypertensive agents (e.g., angiotensin-converting enzyme inhibitors, ß-blockers, calcium channel blockers, and diuretics) that patients being screened for pheochromocytoma are likely to be taking. For HPLC, interferents usually manifest as peaks coeluting, partially or totally, with the peaks of interest. The former are easily detected by inspection of the chromatogram. Total coelution is less easy to detect and can never be ruled out unless the method of detection relies on primary chemical properties of the analyte of interest (e.g., mass spectroscopy). Interference can also occur in the sample preparation method before chromatography (10).
Over the past 15 years, attempts have been made to develop immunoassays for the catecholamines and metanephrines (11)(12)(13). However, problems in raising antibodies against epinephrine and norepinephrine and their 3-O-methylated metabolites, metanephrine and normetanephrine, because of their small molecular size and the high susceptibility of the catecholamines to oxidation (14)(15) have prevented such assays from becoming available until recently (16). These new immunochemical methods should be easier to use than HPLC techniques and will be more accessible to laboratories, at the same time hopefully maintaining or improving assay sensitivity and specificity. However, the detection of interference in immunoassays is more difficult than for HPLC because no visual hardcopy of the measured end product is produced; a number is simply generated. It is therefore vital to assess the specificity these immunoassays. This is achieved by testing the reaction of the antibody with molecules structurally related to the intended analyte (17). However, cross-reactants with the antibody may be removed during the extraction procedure and separation stages of the immunoassay method. Thus, it is usually the method cross-reactivity that is evaluated. For the immunoassays used in this study, method cross-reactivity to related biogenic amines was established by the manufacturers for all four immunoassays (fE, fNE, tM, and tNM), with a maximum cross-reactivity of 3%.
The aim of this study was to determine whether drugs commonly
prescribed to subjects with symptoms of pheochromocytoma interfere with
the measurement of urinary free catecholamines and total metanephrines
by these new immunoassays. We took urine samples from patients on a
variety of medications (angiotensin-converting enzyme inhibitors,
ß-blockers, calcium channel blockers, diuretics, statins,
-methyldopa, and -receptor antagonists). By statistically
comparing fE, fNE, tM, and tNM excretion in these patients with that in
a nonmedicated population, we were able to examine potential
interference from genuine drug metabolites, such as may occur during
the use of these immunoassays in the clinical field. In addition, to
further elucidate the specificity of the immunoassays, we compared
results with those from an established HPLC method (18) that
in our laboratory has been shown to suffer little from visually
detectable interferences. To aid result interpretation, the precision,
accuracy, linearity, and sensitivity of each immunoassay were
determined. We used the RIA rather than the ELISA versions of the
methods because of the ready availability of a gamma counter in our
laboratory. The antigen-antibody system is identical in both types of
assay and should therefore be similarly affected by any interferences.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
ria methods for the determination of free urinary catecholamines
and total metanephrines
The RIA kits were manufactured by Immuno Biological Laboratories,
Hamburg, Germany, and supplied by Immunodiagnostic Systems Ltd. (IDS),
Tyne and Wear, UK. The AMICYL-TestTM KATCOMBI kit
contained all reagents, including calibrators and controls, for
the measurement of fE and fNE. The AMICYL-Test METCOMBI kit contained
all reagents for the measurement of tM and tNM.
The concentration of free catecholamines and total metanephrines in each urine was determined using competitive immunoassay with 125I-labeled antigen and polyethylene glycol-enhanced precipitation of the antibody-antigen complex. The primary antibody for all assays was raised against an acylated derivative of either metanephrine or normetanephrine. Therefore, not only did all samples require acylation, but for the catecholamine immunoassays, enzyme-catalyzed methylation to convert fE and fNE into free metanephrine and normetanephrine was also needed. Methylation was preceded by extraction of the catecholamines from the sample matrix (and removal of metanephrines) by borate affinity gels. No extraction step was required for the tM and tNM assays.
analytical procedure for free catecholamines (fE ANDfNE)
Extraction.
Catecholamines in samples, calibrators, and
controls were adsorbed onto borate-coated wells of a macrotiter plate.
After incubation (30 min at room temperature) and washing (twice with
doubly distilled water), catecholamines were eluted from the gel by
shaking with 300 µL of 0.05 mol/L HCl.
Methylation and acylation.
The extracted catecholamines were
aliquoted to duplicate tubes coated with Bolton-Hunter acylation
reagent (16), and the supplied methylation reagent (rat
liver catechol-O-methyltransferase with
S-adenosylmethionine as the coenzyme) was added. All tubes
were vortex-mixed and incubated in a water bath at 37 °C for 60 min.
RIA.
After incubation, 50 µL of the appropriate labeled
antibody (i.e., the 125I-derivative of acyl-M or
acyl-NM) was added to all tubes, followed by 50 µL of antiserum.
Samples were vortex-mixed, centrifuged (450g for 1 min at
4 °C), and incubated overnight at either room temperature (fE assay)
or 2–8 °C (fNE assay). The following day, precipitating antiserum
was added to all tubes, and after further vortex-mixing and a 15-min
incubation period at room temperature, the resulting precipitate was
pelleted by centrifugation (3000g for 15 min at 4 °C).
The supernatant was removed, and the amount of radioactivity in each
tube was determined.
Calculation of results.
A calibration curve of the percentage
of binding of label against log concentration of the calibrator was
fitted to the data for the calibrators using the spline-function
program in the Wallac Multicalc computer software (19). The
fE or fNE concentrations in subsequent samples were determined from
these calibration curves.
analytical procedure for total metanephrines
For the measurement of total (i.e., both conjugated and
unconjugated) metanephrine and normetanephrine, sulfate and glucuronate
conjugates were firstly acid hydrolyzed such that all metanephrines
were in their free form before acylation.
Acid hydrolysis.
Samples were heated for 20 min at 90 °C
with 0.1 mol/L hydrochloric acid. After cooling, 50 µL of each sample
was diluted 20-fold with assay buffer and vortex-mixed before
acylation.
Acylation.
Samples were added to duplicate coated acylation
tubes, vortex-mixed, and incubated for at least 15 min at room
temperature.
RIA.
The appropriate labeled antibody (i.e., 50 µL of
125I-derivative of acyl-M or acyl-NM) was added
to the acylated samples, followed by an equal volume of antiserum.
Samples were vortex-mixed, centrifuged (450g for 1 min at
4 °C), and incubated overnight at 2–8 °C. The following day,
precipitating antiserum was added to all tubes, and after vortex-mixing
and incubation (15 min at room temperature), the resulting precipitate
was pelleted by centrifugation (3000g for 15 min at
4 °C). The amount of radioactivity in each tube was determined. The
concentrations of tM and tNM in each sample were calculated by
comparison with a calibration curve as described for the catecholamine
assay.
quality control
Included in each batch of assays were four quality-control
samples, prepared by mixing urines from patients not receiving any
medications and pheochromocytoma patients to cover the entire
analytical range of the assays. These samples were stored as aliquots
at -20 °C.
assessment of analytical variables
Accuracy.
The accuracy of each immunoassay was assessed by
comparing RIA results with those from the HPLC method, using a paired
two-tailed t-test, Pearson correlation analysis, linear
regression, and difference plots (20)(21).
Linearity.
Three urine samples were serially diluted with
either pooled urine containing low concentrations of analytes or 0.05
mol/L HCl.
Detection limit.
The detection limit was defined as the
concentration of analyte corresponding the an activity 3 SD above the
response of the zero calibrator (n = 8) for each analyte
(22)(23).
Precision.
Intraassay precision was determined from the
results of duplicate analysis of samples from patients. The interassay
precision was calculated from analysis of the quality-control samples
included in every assay (22)(23).
hplc measurement of free catecholamines and total metanephrines
The HPLC measurement of urinary fE, fNE, tM, and tNM was a
modification of the method described by Green et al. (18).
To analyze total metanephrines, urine samples were first acid
hydrolyzed (2 mol/L HCl at 100 °C for 20 min) to convert conjugated
metanephrines into their free form before HPLC. A Gilson automated
ASTED sample processor (Anachem) allowed the automated online
extraction of analytes from urine by strong cation exchange. The HPLC
mobile phase (0.125 mol/L diammonium hydrogen orthophosphate, pH 3.3,
4.4 mmol/L heptane sulfonic acid, 2 mmol/L EDTA, and 2 mL/L methanol)
eluted the catecholamines and metanephrines from the ion exchanger
directly onto the analytical column (150 x 46 mm, packed with 5
µm particles of Spherisorb ODS2). Detection was electrochemical
(redox mode), using a coulometric detector with a dual analytical cell
(ESA Analytical). Quality-control material was prepared from commercial
preparations (Bio-Rad) and urine samples from patients with
pheochromocytoma.
investigation of potential immunoassay interferences
Urine samples (24-h or random) were collected from patients on a
variety of medications (angiotensin-converting enzyme inhibitors,
ß-blockers, calcium channel blockers, diuretics, statins,
-methyldopa, and -receptor antagonists). All samples were
acidified with HCl and stored at -20 °C. Urine pH was determined
before analysis; samples with pH >3 were discarded because of
potential catecholamine instability (14)(15).
Urine samples were grouped according to the specific type of drug being taken by the patient. Subjects in more than one group were not included in subsequent analysis. We used one-way ANOVA to assess the significance of any differences in the concentrations of free catecholamines and total metanephrines between the different drug types. Patients not on medication were included as a control group (no analytical interference in the immunoassay). However, it is possible that drug interference may have been negated by physiological effects of the drug in question on catecholamine metabolism. To investigate this possibility, the analysis was repeated using data from an HPLC method that has no known analytical interference (Reed P, Kane J, Weinkove C, unpublished laboratory observations). Significance based on these results would, therefore, suggest physiological effects of the drug in question on catecholamine metabolism.
statistical analysis
All statistical analysis was performed using Astute, Ver. 1.51
(DDU Software).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
). For tM, the immunoassay gave a mean concentration
significantly higher than the HPLC method (P <0.001; n
= 211); no such difference was found for tNM (Table 1
). Correlation
coefficients showed acceptable agreement between the two methods, with
values of 0.921 (fE), 0.808 (fNE), 0.822 (tM), and 0.896 (tNM).
However, the regression lines for the catecholamine immunoassays were
significantly different from 1.00 (P <0.01), with values of
0.842 and 0.627 for fE and fNE, respectively (Fig. 1
). Difference plots comparing RIA and HPLC results showed no
concentration-dependent bias (Figs. 1
and 2
).
|
|
|
Linearity.
Assays were linear up to 1.5 and 2.0 µmol/L for
fE and fNE, respectively. For the metanephrines, linearity was observed
up to 4.9 µmol/L for tM and 14.7 µmol/L for tNM. Linearity was not
affected by the diluent matrix (urine or 0.1 mol/L hydrochloric acid).
Detection limit.
For fE and fNE, the detection limits were
0.009 and 0.027 µmol/L, respectively; for tM and tNM, the detection
limits were 0.119 and 0.346 µmol/L, respectively.
Precision.
For the fE immunoassay, the intraassay CVs were
12–22% (0.021–0.65 µmol/L); the intraassay CVs for fNE were
4.6–14% (0.071–2.36 µmol/L) (Table 2
). The interassay CVs were 11–19% for fE (0.036–0.37
µmol/L) and 10–25% for fNE (0.39–1.40 µmol/L; Table 2
). The
intraassay CVs were 6.9–27% (0.15–5.07 µmol/L) for tM and
9.6–12% for tNM (0.49–16.4 µmol/L; Table 3
). The interassay CVs were 10–22% for tM (0.23–3.73 µmol/L)
and 5.8–16% for tNM (0.62–13.9 µmol/L; Table 3
).
|
|
investigation for potential immunoassay interferences
By one-way ANOVA, no significant differences were found between
the concentrations of free catecholamines or total metanephrines in
urines from patients taking the medications investigated and the
control group on no medication. The individual P values were
0.649 (fE), 0.221 (fNE), 0.149 (tM), and 0.170 (tNM; Tables
4 and
5). Additionally, no significant differences were found when the
analysis was repeated using the data obtained by HPLC
(P = 0.796 for fE, 0.246 for fNE; 0.259 for tM, and
0.251 for tNM).
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
).
However, a significant positive bias was observed for the tM
immunoassay (P <0.001; n = 211). Higher metanephrine
concentrations were also observed by Worthers et al. (24),
who compared the ELISA metanephrine assay with measurements by gas
chromatography with mass spectroscopy. These differences may be
attributable to variations in the calibration of the tM immunoassay.
For the catecholamine assays, although the mean concentrations of
fE and fNE agreed well with HPLC, the slopes of the regression lines
comparing HPLC and RIA results were significantly different from unity.
This may be attributable to the few high results exerting a
disproportionately strong influence on the regression line. In
addition, for fE, most of our samples contained only low analyte
concentrations, near the detection limit of the HPLC method (0.02
µmol/L); discrepancies here may have been caused by errors in the
HPLC measurement.
Precision.
This study involved a more rigorous assessment of
the precision of the immunoassays than previous publications
(13)(24).
The interassay CVs for tM and tNM from this study were similar to those reported by Worthers et al. (24), who used the ELISA version of the immunoassays. They found interassay CVs of 10% and 14% (mean metanephrine concentrations, 0.37 and 1.54 µmol/L) and 9% and 13% (mean normetanephrine concentrations, 1.09 and 4.73 µmol/L). Their results were based on the analysis of two samples 10 times by only two kits; in our study, four samples were analyzed up to 10 times by eight different kits.
For the catecholamine immunoassays, few other studies have reported
precision data. Manz et al. (16) found interassay CVs of 8%
and 13% (13.8 and 3.38 µmol/L, respectively) for fE, and 9% and
16% (33.2 and 8.74 µmol/L, respectively) for fNE. These compared
well with our CVs (Table 2
) despite the much lower analyte
concentrations used in this study.
The HPLC method showed better precision than the immunoassays, with between-batch CVs of 1–8% for both catecholamines and metanephrines. This is in agreement with published precision data for the measurement of these analytes by other HPLC methods (25)(26)(27). The better precision of HPLC may reflect the more automated nature of the technique compared with the immunoassays, suggesting that in this study, precision may improve with the use of automated work stations. In addition, a new version of the catecholamine kit is now available. Preliminary data from our laboratory have shown a marked improvement in intraassay CV (2.1–7.4% for fE at 0.021–0.65 µmol/L, and 1.8–6.4% for fNE at 0.071–2.36 µmol/L). Although insufficient assays were performed to assess interassay precision, it may be predicted that this will improve similarly.
identification of potential interferences
Immunoassay detection systems rely on the biological specificity
of antigen/antibody reactions. Therefore, substances with similar
structures to the catecholamines and/or metanephrines (e.g.,
-methyldopa and ß-blockers) may interfere in these assays. No such
interference was observed in this study. No significant difference was
found in fE and fNE concentrations (P = 0.649 and
0.221, respectively) or tM and tNM concentrations (P =
0.149 and 0.170, respectively) in urines samples from subjects in the
different drug groups and those on no drugs. A similar lack of
interference in catecholamine and metanephrine immunoassays has been
reported previously (11)(16)(24).
However, these studies examined only compounds structurally related to
the catecholamines rather than the clinically used drugs examined in
this study. In addition, in previous studies urine samples were
supplemented with the parent compound; therefore, interference by
metabolites was not investigated.
In addition to affecting antigen/antibody recognition, drugs and/or their metabolites could have interfered with other steps in the immunoassays, namely acylation, extraction with boronic acid, and methylation, with the last two steps applying only to the catecholamine immunoassay. It has been reported that the speed of the Bolton-Hunter acylation reaction depends on sample pH, with the reaction being slow at too low a pH, and denaturation of the acylation reagent occurring most rapidly at alkaline pH (28). This pH factor, although not examined in this study, is unlikely to have any significant effect on the RIAs because sample acidity should be sufficient to prevent denaturation of the acylation reagent, and the length of the acylation period was sufficient to allow for any pH-related variations in acylation rate.
With the additional extraction and methylation steps in the catecholamine immunoassay, it may be expected that these assays would be more prone to interference. Borate extraction has been shown to be better than other techniques because this material binds only compounds with cis-diol groups, as opposed to a requirement simply for diol groups in, for example, alumina extraction (10)(27)(29). This, coupled with the high specificity of immunological reactions, may be sufficient to prevent interferences at this stage of the analysis. Both calcium and an unidentified "inhibiting factor" have been shown to inhibit catecholamine-O-methyltransferase (30)(31). However, because only 25-µL samples of urine were used in the catecholamine immunoassay, these inhibitors are unlikely to have any significant effect.
In summary, overall these new immunoassays compared well with HPLC in the measurement of fE, fNE, tM, and tNM, and none of the drugs examined in this study showed significant interference. Therefore, in combination, they may provide a viable, accessible alternative to HPLC for the initial screening of subjects with suspected pheochromocytoma.
|
|
|
Acknowledgments |
|---|
|
Footnotes |
|---|
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  M. d'Herbomez, G. Forzy, C. Bauters, C. Tierny, P. Pigny, B. Carnaille, F. Pattou, J.-L. Wemeau, and N. Rouaix An analysis of the biochemical diagnosis of 66 pheochromocytomas Eur. J. Endocrinol., May 1, 2007; 156(5): 569 - 575. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. L. Brain, J. Kay, and B. Shine Measurement of Urinary Metanephrines to Screen for Pheochromocytoma in an Unselected Hospital Referral Population Clin. Chem., November 1, 2006; 52(11): 2060 - 2064. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. Charalampopoulos, E. Dermitzaki, L. Vardouli, C. Tsatsanis, C. Stournaras, A. N. Margioris, and A. Gravanis Dehydroepiandrosterone Sulfate and Allopregnanolone Directly Stimulate Catecholamine Production via Induction of Tyrosine Hydroxylase and Secretion by Affecting Actin Polymerization Endocrinology, August 1, 2005; 146(8): 3309 - 3318. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. L. Taylor and R. J. Singh Validation of Liquid Chromatography-Tandem Mass Spectrometry Method for Analysis of Urinary Conjugated Metanephrine and Normetanephrine for Screening of Pheochromocytoma Clin. Chem., March 1, 2002; 48(3): 533 - 539. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
