Pheochromocytomas are chromaffin cell tumors in the adrenal medulla that can over-secrete catecholamines (adrenaline, noradrenaline, and dopamine) and metanephrines (metanephrine and normetanephrine). An excess of biogenic amines can lead to serious cardiovascular complications such as hypertensive crisis, myocardial infarction, and stroke [1]. Prompt diagnosis is therefore important.
The determination of catecholamines and metanephrines is important in biochemical screening and diagnosis of pheochromocytoma. Previous work suggests that the diagnostic sensitivity and specificity of metanephrine determination are higher than those for catecholamines [2]. The analysis of either plasma or urine samples is recommended as an initial screening test for pheochromocytoma. Complete 24-hour collection is needed for a urine specimen and an acid preservative must be added to the collection container. Despite the inconvenience to the patient, a urine sample is preferred to a plasma sample because plasma catecholamine and metanephrine levels can be greatly influenced by posture, sampling location, stress, and exercise. Reference intervals for plasma metanephrines are often inappropriately established [3]. The concentrations of catecholamines and metanephrines are also much higher in urine than in plasma, which makes measurements easier.
Liquid chromatography with electrochemical detection (ECD) is currently the most commonly used technique for determining urine catecholamines and metanephrines. However, ECD methods have the disadvantages of high background noise, inadequate sensitivity, and tedious sample preparation, and they are prone to interference by various drugs [4, 5]. The improved analytical sensitivity of modern liquid chromatography-tandem mass spectrometry (LC-MS/MS) has enabled urine catecholamine and metanephrine concentrations to be measured with better sensitivity and specificity using multiple reaction monitoring (MRM). In MRM, two unique mass transitions are monitored per analyte, enabling identity validation and quantitation. LC-MS/MS methods also have the advantages of less manual sample preparation and shorter running times. Many papers on catecholamine and metanephrine determination using LC-MS/MS have been published. The choice of chromatographic column is important because it affects the separation of adrenaline and normetanephrine, which are isobaric compounds. The separation was not optimal in some previous studies [6, 7]. In other studies, catecholamines and metanephrines were determined in separate runs [5, 8]. Sample extraction is another crucial step in catecholamine and metanephrine determination. The extraction efficiencies of catecholamines and metanephrines can be greatly affected by the sample pH and some published methods require pH adjustment of individual samples, which is time consuming [6, 9]. In this study, we report a simple LC-MS/MS method for determining urine catecholamines and metanephrines to overcome these problems.
Adrenaline (E), noradrenaline (NE), dopamine (D), metanephrine (M), and normetanephrine (NM) (purity≥98%) were purchased from Sigma-Aldrich (St. Louis, MO, USA). Deuterated internal standards (atom D%≥98.7%) were purchased from CDN Isotopes (Pointe-Claire, QC, Canada). Calibration standards and controls were purchased from Chromsystems (Munich, Germany).
Samples were treated using a modified version of a published procedure [8]. First, an internal standard solution (50 and 10 μg/L in 0.5% (v/v) formic acid) was added to the urine sample (500 μL) and the mixture was vortexed. Then 4.5 mL 0.1 mol/L ammonium acetate was added as a dilution buffer. The sample was applied to the conditioned cartridge after mixing.
Sample extraction was performed using a Biotage Rapid Trace automated solid-phase extraction (SPE) workstation as follows. An Oasis weak-cation-exchange (WCX) 1 mL cartridge from Waters Corporation (Milford, MA, USA) was conditioned using 1.5 mL methanol and 1.5 mL deionized water, loaded with 2.4 mL sample, washed with 1.5 mL 5% (v/v) methanol, and eluted with 1.2 mL 5% (v/v) formic acid.
The eluate (10 μL) was injected directly onto an Agilent ZORBAX Eclipse Plus Phenyl-Hexyl column (3.0 mm×100 mm, 1.8 μm) using a Waters ACQUITY UPLC I-class system and eluted with an isocratic flow of 0.2% (v/v) formic acid in water at a flow rate of 0.5 mL/min. The column was maintained at 25 ℃ and the sample compartment was maintained at 8 ℃. A typical chromatogram is shown in Fig. 1.
Catecholamines, metanephrines, and their corresponding internal standards were monitored with a Waters ACQUITY® XevoTM TQMS system in positive electrospray ionization mode using MRM. The mass transition and retention time parameters under the MS operating conditions (3 kV capillary voltage; 30 V cone voltage; 500 ℃ desolvation temperature; 1 000 L/h desolvation gas flow rate) are shown in Table 1.
Descriptive statistics, including the mean, standard deviation, coefficient of variation, and range, were used to present the overall data. Pearson correlation and linear regression were used to assess linearity. Comparisons with an existing HPLC-ECD method and another LC-MS/MS method were performed using Bland-Altman plots and Passing-Bablok regression.
The absolute recovery was assessed by comparing the peak areas of standards recovered from spiked urine sample with those of standard solutions. The absolute recoveries were 61.9% for adrenaline, 40.1% for noradrenaline, 80.8% for dopamine, 103% for metanephrine, and 97.2% for normetanephrine.
The relative recovery was assessed by comparing the peak areas before and after SPE of standards spiked into urine samples. The relative recoveries were 59.1% for adrenaline, 40.1% for noradrenaline, 81.5% for dopamine, 99.1% for metanephrine, and 92.2% for normetanephrine.
These data match the recoveries claimed by the Oasis WCX cartridge manufacturer [10]. The relatively low recovery of noradrenaline is considered sufficient for reliable measurements on the basis of the precision and lower limit of quantification data.
The precision was assessed by processing low (64.1 nmol/L for adrenaline, 367 nmol/L for noradrenaline, 1.248 μmol/L for dopamine, 790 nmol/L for metanephrine, and 1.459 μmol/L for normetanephrine) and high (264 nmol/L for adrenaline, 1.210 μmol/L for noradrenaline, 2.888 μmol/L for dopamine, 1.886 μmol/L for metanephrine, and 5.425 μmol/L for normetanephrine) urine controls 20 times a day for 20 days. The intra-day precision for all analytes ranged from 0.59% to 4.64%, and the inter-day precision ranged from 1.98% to 4.80%.
External quality assurance samples from the Royal College of Pathologists of Australasia (RCPA) quality assurance program (n=12) were assayed and showed excellent agreement with the target values (Table 2).
The analytical measurement range was determined from various dilutions of a high-concentration urine sample. The assay was linear up to 5.00 μmol/L for adrenaline, 5.00 μmol/L for noradrenaline, 6.10 μmol/L for dopamine, 5.60 μmol/L for metanephrine, and 3.46 μmol/L for normetanephrine.
The lower limit of quantification, defined as the concentration that consistently resulted in a signal/noise ratio greater than 10, was determined at 5 nmol/L for adrenaline, 5 nmol/L for noradrenaline, 12 nmol/L for dopamine, 6 nmol/L for metanephrine, and 7 nmol/L for normetanephrine.
No evidence of ion suppression was observed for any analytes at their corresponding retention times when the extracted urine sample was injected into the LC-MS/MS system together with post-column infusion of standards (1 μmol/L) at 10 μL/min.
Carry-over was estimated by running a high-spiked sample (264 nmol/L for adrenaline, 1.210 μmol/L for noradrenaline, 2.888 μmol/L for dopamine, 1.886 μmol/L for metanephrine, and 5.425 μmol/L for normetanephrine) three times, followed by running a low-spiked sample (10.0 nmol/L for adrenaline, 132 nmol/L for noradrenaline, 250 nmol/L for dopamine, 58 nmol/L for metanephrine, and 88 nmol/L for normetanephrine) three times. No significant carry-over was observed ( < 0.06%).
Comparison with an in-house HPLC-ECD method currently in use in our laboratory using urine samples (n=51) by Passing-Bablok regression gave slopes of 1.22, 1.00, and 0.89, y-intercepts of -10.0, 8.0, and 46.8, correlation coefficients (r) of 0.80, 0.97, and 0.99, and mean differences of -0.2, 4, and -142 nmol/L for adrenaline, noradrenaline, and dopamine, respectively. The in-house HPLC-ECD method could not be used for metanephrine and normetanephrine. The in-house HPLC-ECD method was performed as follows. The urine sample was first combined with an internal standard, namely 3, 4-dihydroxybenzylamine. The sample pH was adjusted to between 6.0 and 7.0 and then the sample was added to a cation-exchange-resin column. After washing with water, catecholamines were eluted from the column with boric acid and collected. The eluate was further purified by adsorption on alumina powder at pH 8.6, desorbed in 0.05 mol/L phosphoric acid, and quantified using reversed-phase HPLC with ECD. Fig. 2 shows Bland-Altman plots comparing the urine catecholamine concentrations obtained using HPLC-ECD and our LC-MS/MS method.
Comparison with an established LC-MS/MS method used by another laboratory (urine samples, n=50) by Passing-Bablok regression gave slopes of 1.00, 0.98, 0.97, 1.07, and 1.01, y-intercepts of 0, 1.01, 26.7, 3.99, and 1.84, correlation coefficients of 0.98, 1.00, 1.00, 0.99, and 0.99, and mean differences of 0.6, -2.7, -19.4, 9.4, and 3.7 for adrenaline, noradrenaline, dopamine, metanephrine, and normetanephrine, respectively. The established LC-MS/MS method used by the other laboratory is briefly described as follows. After spiking with deuterated internal standards, catecholamines and metanephrines were extracted from urine by SPE using a WCX cartridge (washed with 5% (v/v) acetonitrile and eluted with 0.5% (v/v) formic acid in 5% (v/v) acetonitrile), followed by separation on an Agilent Zorbax Eclipse AAA reversed-phase column at ambient temperature. The MRM mode was used for the determination of analyte concentrations. Fig. 3 shows Bland-Altman plots comparing the results for urine catecholamine and metanephrine concentrations measured using another established LC-MS/MS method and our LC-MS/MS method.
Our assay differs from previously reported methods in terms of the sample preparation method and chromatographic column used.
Sample preparation is important for removing interfering substances and concentrating biogenic amines. Diphenylboronic acid is commonly used for sample preparation in catecholamine determination, but this method is labor intensive and can cause significant ion suppression in metanephrine determination [11].
The use of a WCX cartridge for sample preparation in our study gave good recovery yields and there was no evidence of ion suppression. WCX cartridges used both reversed-phase and cation-exchange processes to improve retention of strong bases, with the removal of weak bases to minimize ion suppression. The eluate could be directly injected into the LC-MS/MS system without the need for solvent evaporation and reconstitution. Minimal hands-on sample preparation was required.
The chromatographic resolution of catecholamines and metanephrines was important because adrenaline and normetanephrine were isobaric compounds and shared the same mass transition, i. e., 184→166 (m/z). Quantitation was only feasible when these two amines could be separated. Some studies using reversed-phase C18 columns gave suboptimal separation of these two amines because of their highly polarities, and long run times were required [6]. Other studies used hydrophilic-interaction liquid chromatography, but the separation of metanephrine and normetanephrine was not optimal [12]. The phenyl-hexyl column used in our method gave complete baseline separation of all analytes within 3 min, and the linearity was excellent compared with those achieved in previous studies [5, 13]. These findings suggested that this new method enabled direct analysis of urine catecholamines and metanephrines in patients with pheochromocytoma, without the need for dilution, and was therefore suitable for routine clinical analysis. The phenyl-hexyl column also had the advantage of a simple mobile phase composition (0.2% (v/v) formic acid in water), which did not require any buffer or pH adjustment.
We validated an assay for the simultaneous ana-lysis of urine catecholamines and metanephrines using LC-MS/MS and automated SPE. This WCX SPE technique using phenyl-hexyl reversed-phase LC-MS/MS is a simple and robust method suitable for use in routine clinical analysis.
2017,
Vol. 35 Issue (10): 1042-1047
PDF
2017,
Vol. 35 Issue (10): 1042-1047
PDF