Printer Friendly

Evaluation of a new method for the analysis of free catecholamines in plasma using automated sample trace enrichment with dialysis and HPLC.

The routine analysis of plasma free norepinephrine, epinephrine, and dopamine is fraught with analytical problems, primarily associated with low plasma concentrations and the relatively labor-intensive manual extraction procedure required to remove interfering compounds (1). The automation of such techniques using automated sequential trace enrichment of dialysates (ASTED) [3] has been applied successfully to the assay of urinary free catecholamines (2). The application to plasma, however, presented specific problems; in particular, the original ASTED system could not achieve regeneration and conditioning of both the dialyzer and the trace-enrichment cartridge (TEC). This has now been successfully resolved with the development of the second generation ASTED.XL[TM], which utilizes two rheodyne valves and two dilutors. The technique involves the dialysis of the plasma sample to remove protein and other macromolecules. The catecholamines are then trace-enriched onto a strong cation-exchange resin (3) and eluted onto the analytical column. Separation is achieved by reversed-phase ion-pair chromatography with coulometric detection of the catecholamines (4), whereas the TEC and dialyzer are regenerated for subsequent injections. Procedures reported previously for automated plasma catecholamine analysis have included a manual clean-up step or simple deproteinization (5) and then solid-phase extraction (6) followed by automated HPLC with electrochemical or fluorescence detection (7).

We here report the analytical evaluation of the fully automated analytical ASTED procedure for measuring plasma catecholamines. The analysis of plasma norepinephrine (NE) and epinephrine (E) was compared in healthy volunteers, patients being investigated for catecholamine-secreting tumors, patients with chronic renal failure, and in particular, assessment of the effect of regular hemodialysis in patients with end-stage renal failure.

Materials and Methods

Blood samples (10 mL) were collected into tubes containing lithium heparin as the anticoagulant (Sarstedt) and transported to the laboratory on ice within 30 min. Samples were centrifuged at 3500 rpm for 10 min at 4[degrees]C. Plasma was transferred into 2-mL polypropylene tubes (L.I.P.) and frozen immediately at -25[degrees]C. Analysis was then performed within 1 month of receipt of the sample. For longer periods, storage was at -70[degrees]C.

Samples were obtained from healthy volunteers (n = 31), patients with chronic renal failure (n = 15), end-stage renal failure patients on hemodialysis (n = 15), and hospital patients with hypertension being investigated for catecholamine-secreting tumors (n = 15). A group of healthy volunteer athletes (n = 10) was sampled before and after exercise to exhaustion, i.e., 70% of maximum [O.sub.2] utilization on a cycle ergometer as part of a study on the effects of exercise on glucose utilization and insulin production (8). This allowed us to obtain plasma samples (n = 69) with variable catecholamine concentrations within or greater than the upper limit of the reference interval. The end-stage renal failure patients were sampled before and after dialysis, and the hypertensive patients were sampled in a supine position after a 30 min rest period.

The collection of samples complied with the ethics standards of the Royal Liverpool University Hospital. The additional analyses were performed anonymously, in accordance with the ethics guidelines of the Hospital. The system used was the Gilson ASTED.XL (Anachem), a fully automated sample processing system. Incorporated into this was a Gilson 307 pump (Anachem) with a 10WSC pump head, a dialyzer block volume of 370 [micro]L with a 15 000 molecular weight cutoff cuprophan dialysis membrane and a hema-SB cationic TEC (Anachem). The optimized plasma sample volume was 740 [micro]L, taken as two separate aliquots of 370 [micro]L. The individual aliquots were dialyzed into 2 mL of 5 mmol/L ammonium phosphate buffer, pH 8.6, and the dialysate was applied to the TEC. To increase sensitivity, additional plasma aliquots could be used simply by changing the program controlling the ASTED processor. Electrochemical detection of the catecholamines was achieved using an ESA Coulochem II (ESA Analytical) fitted with a guard cell and a 5011 analytical cell. Oxidation of the analytes was carried out at +300mV in the guard cell followed by successively higher reduction potentials of -100 mV and -300 mV in the analytical cell and a sensitivity set at 20 nA.

The mobile phase consisted of 50 mL/L in 125 mmol/L diammonium hydrogen orthophosphate (BDH) containing 101 mg of heptane sulphonic acid (Sigma Chemical Co.) and 73 mg of EDTA (Sigma) adjusted to pH 3.5. A 15.0 x 0.46 cm Ultratechsphere 5 [micro]m ODS 2 column (HPLC Technology) was used at a flow rate of 1.5 mL/ min. The dialysis recipient solvent was 5 mmol/L diammonium hydrogen orthophosphate adjusted to pH 8.3. Dihydroxybenzylamine was used as the internal standard (IS) at a concentration of 60 nmol/L, 50 uL of which was added to 750 uL of sample.

The effect of the serum matrix on dialysis was assessed by comparison of peak areas after injection onto the complete system of an aqueous calibrator, a serum blank, and the same serum to which 10, 5, 20, or 5 nmol/L NE, E, dopamine (D), and IS had been added. The mean [+ or -] 1 SD recoveries (n = 5) were 83% [+ or -] 2%, 75% [+ or -] 2.5%, 95% [+ or -] 3.8%, and 91% [+ or -] 3%, respectively, which indicated a small loss because of the protein matrix effect. The relative recovery through the TEC without dialysis was compared with loop injection only using 100 [micro]L of an aqueous solution containing 10 nmol/L NE, 5 nmol/L E, 20 nmol/L D, and 5 nmol/L IS. The recoveries (mean [+ or -] SD; n = 4) were 89% [+ or -] 1.5%, 77% [+ or -] 1.5%, 91% [+ or -] 2%, and 94% [+ or -] 2.1%, respectively, indicating some loss on the TEC, in particular for E because of its reduced retention on the cation exchanger (2). The overall recovery or efficiency of the dialysis was assessed by comparison of the peak areas after injection of an aqueous calibrator (through the loop only) and a serum blank supplemented with 10 nmol/L NE, 5 nmol/L E, 20 nmol/L D, and 5 nmol/L IS taken through the complete system. The recoveries were 38% [+ or -] 2% for NE, 35% [+ or -] 1.5% for E, 19% [+ or -] 0.8% for D, and 34% [+ or -] 1.7% for IS (mean [+ or -] 1 SD; n = 4), indicating a relatively efficient dialysis.

The system was calibrated using standard additions of NE (Sigma) at 0, 5, 10, 15, and 20 nmol/L and E (Sigma) at 0, 4, 8, and 10 nmol/L to drug-free serum (Bio-Rad). The individual catecholamine responses were corrected to the IS response, and this ratio was related to known concentrations of each analyte. The regression analysis of n = 6 observations for NE and E was: y = 11.1 x [10.sup.5]x + 10, r = 0.998; and y = 12.5 x [10.sup.5]x + 8, r = 0.988, respectively, where y is the electronic signal in computed integration units (Gilson Unipoint Software). In routine use, the calibration was confirmed with 0 and 10 nmol/L NE and 0 and 5 nmol/L E. Calibrators were run every tenth sample, and controls were run every fifth sample. If the signal response deviated >15%, then recalibration would be carried out to ensure that the controls were always within 1 SD of the expected mean value. The analytical performance (Table 1), was monitored with commercially available Endocrine Controls 1 and 2 (Bio-Rad A and B). All solutions were prepared using doubly deionized water (Elga Products Ltd).

The alumina extraction (9) was used as the comparative reference procedure with 69 plasma samples obtained from the various groups outlined above. The catecholamines from a 1.0-mL plasma sample were desorbed from alumina with 400 [micro]L of 0.1 mol/L phosphoric acid (Sigma), and 370 [micro]L was injected for separation by HPLC as outlined for the ASTED procedure. The system was controlled and analytical integrations were carried out by Gilson UniPoint Software with a Gilson 506C interface module (Anachem).

All data were processed using Microsoft Excel for Office 95, and data were compared with standard linear regression analysis and differences between groups with parametric Tukey and nonparametric Kruskal--Wallis statistical analysis, using the ARCUS software as supplied by the University of Liverpool, Liverpool, UK.


Typical chromatograms obtained for automated plasma catecholamine analysis are shown in Fig. 1. The analytical performance given in Table 1 indicates good precision. The detection limit, under the conditions outlined and determined as 3 SD above the mean background signal noise, was 0.05 nmol/L for both E and NE with a 740-[micro]L plasma volume injected. The commonly prescribed beta blockers, angiotensin-converting enzyme (ACE) inhibitors, L-dihydroxyphenylalanine, and metabolites from caffeine or acetaminophen did not interfere in the assay.

Stability of the catecholamines on the system (i.e., on the autosampler rack at 20-22[degrees]C) was assessed using various control solutions and indicated a loss of activity of both NE and E of <1.0%/h in plasma and 3.0%/h in an aqueous solution. Plasma samples stored >72 h at 4[degrees]C decayed at 0.2%/h, over which time a 15% loss was considered significant. The routine practice, therefore, is to calibrate with serum-based calibrators and to perform the analysis in batches such that samples remain on the autosampler for periods no longer than 10-12 h. Analytical comparison with the reference alumina extraction procedure (Fig. 2) gave excellent agreement for NE and E over a wide range of concentrations, showing no significant difference between the two analytical procedures.


Plasma catecholamines varied little among the patient groups studied (Fig. 3). Plasma NE was significantly increased in the predialysis and hospital patients compared with the healthy controls, P <0.05, but pre- and postdialysis concentrations were not significantly different. E was relatively higher in the healthy volunteers. In the chronic renal failure patients, plasma NE was 2.0-5.8 nmol/L and E was 0.1-0.7 nmol/L; both were independent of creatinine clearance in the range <5-120 mL/min. In patients investigated for a catecholamine-secreting tumor, plasma NE and E concentrations similar to healthy controls would suggest exclusion of such a diagnosis. One patient (data not in Fig. 3) with increased NE (11.8 and 9.6 nmol/L), a strong clinical suspicion, and a negative clonidine suppression test was subsequently investigated with selective venous sampling, which identified a catecholamine-secreting tumor in the inferior vena fossa.

In the volunteers studied after exercise, the plasma NE rose from (mean [+ or -] 1 SD) 4.5 [+ or -] 2.5 nmol/L (range, 1.5-6.4 nmol/L) to 15.6 [+ or -] 5.6 nmol/L (range, 8.0-35.2 nmol/L), and E rose from 0.5 [+ or -] 0.3 nmol/L (range, 0.2-1.0 nmol/L) to 4.5 [+ or -] 1.5 nmol/L (range, 1.6-8.6 nmol/L).



We have shown that the fully automated quantitative analysis of the plasma free catecholamines NE, E, and D can be achieved without any sample pretreatment, using the newly developed ASTED.XL system. D was not usually measured under these conditions (being near the analytical detection limit); however, it can be seen at higher concentrations, and therefore more reliably quantified, after treatment with sulphatase (7) or when patients are on L-dihydroxyphenylalanine (10). This methodological advance has been made possible because of the incorporation of improved software to control two rheodyne valves that allow access to the two dilutors to regenerate the TEC and recondition the dialyzer block independently. In the ASTED configuration for urine analysis (2). dialysis was not required and only one dilutor was made available for regeneration of the TEC. However, when the ASTED.XL was used, both dialysis and trace enrichment could be controlled easily. The conditions were then optimized on 740 [micro]L of plasma. If measurement of concentrations below or near the detection limit is required, more sample, e.g., 1000 [micro]L of plasma or a longer dialysis time could be used.


We have been performing the procedure routinely for several years, having originally demonstrated that the system operates satisfactorily for plasma catecholamine analysis (11). To maintain good performance, reagents must be prepared fresh weekly, using pure water as outlined. Any loss of sensitivity other than that caused by old reagents can often be traced either to poor trace enrichment or to poor dialysis, although the dialyzer should remain effective for up to 6 months. The TEC and the analytical column should be functional for a similar period or a minimum of up to 2-3000 injections. The coulochem analytical cells have expected working lives of at least 2 years or 6000 injections. However, this is very much dependent on the number of samples analyzed, e.g., urine analysis should give a greater life expectancy because of a reduced sensitivity requirement. If loss of sensitivity is because of failure of the analytical cell, then regeneration (see manufacturer's instructions) might be a possibility; the best solution, however, is to invest in a new cell. The analytical cost is approximately 1.5 [pounds sterling]-2.0 [pounds sterling] (1 [pounds sterling] = $1.60) per test and compares favorably with the reagent costs for new RIAs and enzyme immunoassays (12) priced at 2 [pounds sterling] for each individual analyte.

The ASTED system can easily be adapted to run urinary catecholamines, the organic acids vanillylmandelic acid, 5-hydroxyindoleacetic acid, and homovanillic acid, the methylated catecholamine derivatives, and serotonin (13). The serotonin assay can, if required, be included in the plasma catecholamine analysis because its retention time on the chromatography system is ~20 min. However, because the system outlined is programmed to re-inject every 20 min, this would need to be changed to 30 min.

The ASTED procedure gives good analytical precision for the routine assay of plasma catecholamines, although previous studies using manual alumina clean-up procedures on 1-2 mL of plasma have reported equally good sensitivity and analytical precision (14). However, we have found that the ASTED system is much more sympathetic to general routine clinical applications, a sentiment confirmed by the UK external quality-control assessment scheme for urine catecholamines (15). We have been able to carry out several large studies, with up to 4000 analyses, involving the serial changes in plasma catecholamines after hip operations and in healthy individuals after strenuous exercise (8). Alternative automated procedures have been developed that require some manual pretreatment involving deproteinization followed by either individual solid-phase extraction with an automated processing device or precolumn o-phthaldehyde derivatization. Subsequent separation can use standard ion-pair chromatography with an electrochemical detection or ion-pair soap chromatography with fluorescence detection (6, 7). The latter procedure suffers from glutathione interference with the internal standard dihydroxybenzylamine and has thus caused difficulties in plasma assay reliability. The latter was also adapted to postcolumn derivatization with a fluorogenic agent, diphenylethylenediamine, but again without addition of an IS (16).

Of particular interest are the data now presented in patients with end-stage renal failure; the effect of dialysis on NE and E, D was not usually detected, i.e., the concentrations were generally <0.1 nmol/L (5). Previous reports have shown inconsistent findings, with either increased catecholamine concentrations or concentrations within the reference interval (17,18); however, the number of studies has been limited presumably because of difficulties in catecholamine analysis in renal failure. We have now shown that plasma NE was significantly increased in both patients on regular dialysis and those in varying stages of renal failure but not on dialysis compared with healthy controls. Although the range of results was similar, some of the end-stage renal failure patients would appear to be more stressed than others. Clearly, additional studies are required to explain the mechanism of this phenomenon and whether the stressed patients are more likely to have some form of autonomic nerve dysfunction associated with renal disease (19). E concentrations were generally the same in all groups and were not affected by dialysis, although concentrations were shown to be increased in the healthy volunteers. A particular benefit of the plasma assay for patients in end-stage renal failure is the ability to identify catecholamine-secreting tumors. This is highlighted by the recent report of an anuric patient presenting with paroxysmal hypertension suspected to be as a result of a dopamine-secreting tumor (20).

In conclusion, a procedure has been developed for the automated assay of plasma catecholamines. The assay is suitable for routine clinical applications, is relatively inexpensive to perform, and is sensitive and precise. This procedure will facilitate routine measurement of catecholamines and therefore, additional studies of changes in health and disease, in particular, renal disease and the investigation of catecholamine-secreting tumors in patients who are anuric.

We thank the medical staff of the renal unit of the Royal Liverpool and Broadgreen University Hospital, Liverpool, UK for data on patients and collection of blood samples during hemodialysis sessions.

Received June 5, 1998; accepted December 4, 1998.


(1.) MacDonald IA, Lake DM. An improved technique for extracting catecholamines from body fluids. J Neurosci Methods 1985;13: 239-48.

(2.) Green B, Cooper JDH, Turnell DC. An automated method for the analysis of urinary free catecholamines using ASTED and high pressure liquid chromatography. Ann Clin Biochem 1989;26: 361-7.

(3.) Wu AHB, Garnet TG. Preparation of urine samples for liquid chromatographic determination of catecholamines. Bonded phase phenylboronic acid, cation exchange resin and alumina absorbents compared. Clin Chem 1985;31:298-302.

(4.) Downes RGH, Bailey BA, Martin RG. Estimation of biogenic amines by HPLC and electrochemical detector. Chromatogr Rev 1984;11: 5-7.

(5.) Grossi G, Bargrossi AM, Lucarelli C, Paradisi R, Sprovieri C, Spovieri G. Improvements in automated analysis of catecholamine and related metabolites in biological samples by column-switching-high-performance liquid chromatography. J Chromatogr 1991; 541:273-84.

(6.) Kamahori M, Taki M, Watanabe Y, Miura J. Analysis of plasma catecholamines by high-performance liquid chromatography with fluorescence detection: simple sample preparation for pre-column fluorescence derivatisation. J Chromatogr 1991;567:351-84.

(7.) Yoshimura M, Komori T, Nakanishi T, Takahashi H. Estimation of sulphoconjugated catecholamine concentrations in plasma by high-performance liquid chromatography. Ann Clin Biochem 1993; 30:135-41.

(8.) MacLaren DPM, Reilly T, Campbell IT, Frayne K. Hormonal and metabolite responses to glucose ingestion with or without the addition of guar gum. Int J Sports Med 1994;15:466-71.

(9.) Bouloux P, Perret D, Barnes GM. Methodological consideration of plasma catecholamines by high-performance liquid chromatography with electrochemical detection. Ann Clin Biochem 1985;22: 194-203.

(10.) Dutton J, Copeland LG, Playfer JR, Roberts NB. Measuring L-dopa in plasma and urine to maintain therapy of elderly patients with Parkinsons disease with L-dopa and a dopa-decarboxylase inhibitor. Clin Chem 1993;39:629-34.

(11.) Roberts NB, Hodgkinson AJ, Dutton J, Hutchinson G. Automated analysis of plasma norepinephrine and epinephrine by ASTED HPLC. Proc XVI Int Cong Clin Chem 1996;B535:340.

(12.) Adrenaline and noradrenaline immunoassay kits. Gesellschaft fur Immunochimie und Immunobiologie. Hamburg, Germany: MBH.

(13.) Urinary free catecholamines. Gilson Asted applications guide, 2nd ed. Application 7 AG 7.1-7.6. Villers-le-Bel, France: Gilson S.A., 1990.

(14.) Hjemdahl P. Plasma catecholamines--analytical challenges and physiological limitations. In: Bouloux PMG, ed. Buillieres clinical endocrinology and metabolism, catecholamines. Bailliere Tindall, 1993:307-53.

(15.) Mackenzie F, Bullock DG, Davies JA. Specific target values and performance assessment in the UK NEQAS for urinary catecholamines. Proc XVI Int Congr Clin Chem 1996;100:75.

(16.) Os I, Nordby G, Lyngal PT, Eide I. Plasma vasopressin, catecholamines and atrial natriuretic factor during hemodialysis and sequential ultrafiltration. Scand J Urol Nephrol 1993;27:93-9.

(17.) Darwish R, Elias AN, Vaziri ND, Pahl M, Powers D, Stokes JD. Plasma and urinary catecholamines and their metabolites in chronic renal failure. Arch Intern Med 1984;144:69-71.

(18.) Prados P, Higashidate S, Imai K. A fully automated HPLC method for the determination of catecholamines on biological samples utilizing ethylenediamine condensation and peroxalate chemiluminescence detection. Biomed Chromatogr 1994;8:1-8.

(19.) Armengol NE, Amenos AC, Illa MB, Bertran JG, Ginesta JC, Fillat FR. Autonomic nervous system and adrenergenic receptors in chronic hypotensive haemodialysis patients. Nephrol Dial Transpl 1997;12:939-44.

(20.) Sollazzi L, Peritti V, Crea MA, Ballantyne R, Meo F, Sciarra M, et al. Anaesthetic management of phaeochromocytoma in a long term hemodialysed patient. Acta Anaesth Belg 1994;45:13-7.

[3] Nonstandard abbreviations: ASTED, automated sequential trace enrichment of dialysates; TEC, trace-enrichment cartridge; NE, norepinephrine; E, epinephrine; IS, internal standard; and D, dopamine.

John Dutton, [1] Andrew J. Hodgkinson, [1] George Hutchinson, [2] and Norman B. Roberts [1] *

[1] Department of Clinical Chemistry, Royal Liverpool and Broadgreen University, Hospital NHS Trust, Liverpool L7, 8XP, United Kingdom.

[2] Anachem Ltd, Luton LU2 0EB, United Kingdom.

* Author for correspondence. Fax 44 (1)51 706 5813.
Table 1. Summary of analytical performance.

 Level Mean, nmol/L CV, %
 n = 12 1 (a) 0.90 8.6
 n = 12 2 3.13 2.0
 n = 12 3 14.86 1.8
 n = 12 1 0.30 9.3
 n = 12 2 2.73 3.1
 n = 12 3 5.23 1.1
 n = 15 A (b) 1.73 9.0
 n = 15 B 8.15 3.1
 n=15 A 0.38 8.9
 n=15 B 6.44 4.2

(a) Levels 1, 2, and 3 were pooled human plasma with the range of
concentrations shown.

(b) Levels A and B are the Bio-Rad endocrine controls with quoted
means (range) for NE of 1.77 nmol/L (1.42-2.13 nmol/L) and 8.51
nmol/L (6.81-10.21 nmol/L) and for E of 0.49 nmol/L (0.39-0.58
nmol/L) and 7.13 nmol/L (5.7-8.56 nmol/L), respectively.
COPYRIGHT 1999 American Association for Clinical Chemistry, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1999 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:Endocrinology and Metabolism
Author:Dutton, John; Hodgkinson, Andrew J.; Hutchinson, George; Roberts, Norman B.
Publication:Clinical Chemistry
Date:Mar 1, 1999
Previous Article:Automated homogeneous immunoassay for gentamicin on the dimension clinical chemistry system.
Next Article:TE671 cell-based ELISA for anti-acetylcholine receptor antibody determination in myasthenia gravis.

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters