1997 by Humana Press Inc. A[t rights of any nature, whatsoever, reserved. 0163-4984/97 / 5803-0209 $ll.25. Trace Element Status of Hemodialyzed Patients.
9 1997 by Humana Press Inc. A[t rights of any nature, whatsoever, reserved 0163-4984/97/ 5803-0209 $ll.25
Trace Element Status of Hemodialyzed Patients MICHAEL KRACHLER, ~ G E R H A R D W I R N S B E R G E R , 2 AND KURT J .
IRGOLIC *'~
1Institute for Analytical Chemistry, Karl-Franzens University, Graz, Austria; and 2Department of Internal Medicine, Division of Nephrology, Karl-Franzens University, Graz, Austria Received February t996; Accepted October 1996
ABSTRACT The trace elements Ba, Bi, Cd, Co, Cs, Cu, Hg, La, Mn, Mo, Pb, Rb, Sb, Sn, Sr, TI, and Zn were determined by inductively coupled plasma mass spectrometry in plasma samples of 68 hemodialysis patients. The same elements (with exception of La and Mn) were also determined in whole blood after mineralization with high-purity nitric acid/hydrogen peroxide in a closed-pressurized microwave system. The accuracy and precision was checked by analyzing two Seronorm "whole blood" reference materials. All samples were contaminated with barium (heparinized tubes) and the plasma samples with tin (collection tubes). The concentrations for Bi, Hg, Pb, Rb, Sb, and Sr in whole blood were within the literature ranges for healthy adults. All of the concentrations for Co, and some of the concentrat-ions for Cd, Cs, T1, and Zn were higher than the high limits of the normal ranges. Approximately 14% of the Cu concentrations were lower than the low limit of the normal range. The Mo and Sn concentrations are difficult to evaluate, because the normal ranges appears to be unreliable. All concentrations for Cd, Co, Mo, Pb, Sn, and Sr and some of the concentrations for Cu (15%) and Mn (75%) in the plasma samples were higher than the high limits of the normal ranges. The concentrations for Rb tended to be lower than the normal range. To establish unequivocally the causes for elevated and reduced concentrations of trace elements in whole blood and plasma of dialysis patients, all fluids in the dialysis process must be investigated. Index Entries: Trace elements; hemodialysis; whole blood; plasma; ICP-MS.
*Author to whom all correspondence and reprint requests should be addressed. Biologgcal /'race Element Research
209
Vof. 58, 1997
210
Krachler, Wirnsberger, and hgolic
INTRODUCTION Patients with chronic renal failure requiring treatment by hemodialysis are at risk of developing trace element imbalances (1-3). The occurrence and consequences of such imbalances (except for aluminum) in patients with end-stage renal failure have received little attention (3-5). Trace element disturbances caused by medication, the uremic state, the dialysis process, and the quality of the water used for dialysis may contribute to clinical abnormalities in dialyzed uremic patients (6,7). Dialysis equipment has been identified as a source of trace element contamination, for instance with Zn and Ni that may lead to acute intoxications (8,9). The identification of aluminum as cause of dialysis encephalopathy (10) has drawn the attention to other potential sources of contamination with trace elements such as water and salts, from which dialysis fluids are prepared (11). Blood as the medium for the transport of trace elements, plasma, and serum are easily collected and are convenient samples, in which trace elements can be determined. Concentrations of these elements provide information about the trace element status of a patient and can be the basis for appropriate clinical interventions. To improve our knowledge about the role of trace elements in biological systems, to find correlations between excesses and or deficiencies of trace elements and clinical symptoms of hemodialyzed patients, and to explore the effects of interactions among trace elements requires a versatile, reliable, and rapid method for the simultaneous determination of toxic and essential trace elements in easily collected samples that reflect the trace element status. Inductively coupled plasma mass spectrometry ([CP-MS) promises to be such a method with detection limits equal to or lower than those of other techniques (12). A review of the capabilities of ICP-IVIS for the quantification of trace elements in body fluids (including serum) and tissues was published recently (13). We report here the application of tCP-MS to the simultaneous determination of trace elements in blood and plasma of hemodialyzed patients to establish their trace element status.
MATERIALS AND METHODS Chemicals and Reagents Water was first prepurified by deionization, then distilled twice in a quartz distillation unit, and finally purified with a cartridge purification system (18 M.Q.cm, Barnstead, NANOpure, Boston, USA). All glassware was cleaned by soaking for 24 h in concentrated HNO3 followed by rinsing three times with high-resistivity water. Stock solutions (1000 mg element/L) for each of the 17 elements to be determined in the digests by ICP-MS were prepared by diluting the contents of Titrisot ampoules
Biological Trace Element Research
Vol. 58, 1997
Trace Element Status of Hemodialyzed Patients
211
(Merck, Darmstadt, Germany) to 1L with NANOpure water. By combination and appropriate dilution of aliquots of these stock solutions with 0.29 M high-purity nitric acid, standard solutions were obtained. Standard 1 contained Cu, Rb, and Zn; Standard 2 contained Ba, Bi, Cd, Co, Cs, Hg, La, Mn, Mo, Pb, Sb, Sn, Sr, and Tt; and Standard 3 contained Ga, In, and Re. Aliquots of Standard 1 were mixed and diluted with 0.29 M nitric acid to produce calibration solutions of 20, 50, or 100 gg/L. From Standard 2, calibration solutions with 2, 5, or 10 ~tg/L were prepared similarly. These calibration solutions were spiked to 50 ~tg/L with Ga, In, and Re for the internal calibration of the ICP-MS response. The digestions were carried out with high purity hydrogen peroxide (30%, Supraput@, Merck) and concentrated nitric acid purified by subboiling distillation in an all-quartz distillation unit. Whole blood and serum "second-generation reference materials" (SeronormTM Trace Elements Whole Blood I + II, Batch No. 203052 ยง 203056 and SeronormTM Trace Elements Serum, Batch No. 010017) were purchased from Nycomed, Oslo, Norway.
Mineralization A closed-pressurized, high-performance microwave digestion unit (MLS 1200 MEGA, Milestone, Leutkirch, Germany) equipped with a rotor for 10 Teflon vessels designed for pressures up to 30 bar was used for mineralizations. The vessels are equipped with a pressure release system to prevent explosions. The magnetron operates at pouters of 250 or 1000 W (unpulsed). The 1000-W stage can be pulsed with power cycles from 0.2 to 5 s to deliver time-averaged power in 10-W steps in the range from 10 to 990 W.
Determination of Trace Elements An inductively coupled plasma mass spectrometer (VG PlasmaQuad 2 Turbo Plus, VG Elemental Ltd., Winford, UK) equipped with a Meinhard concentric glass nebulizer, a double-pass, Scott-type spray chamber (water cooled, 0~ and a Gilson Minipuls-3 peristaltic pump were used to determine trace elements. The operating conditions for the ICP-MS are summarized in Table 1.
Collection of Blood Blood samples (2 mL) for trace element determinations were taken from all patients undergoing hemodialysis three times a week before dialysis with a PrecisionGlide Vacutainer| (Becton Dickinson, France) into a 2-mL Sterile Lithium Heparin Vacutainer| (Becton Dickinson, France). Plasma was obtained by collecting 7 mL blood in a SST| Gel and Clot Activator Vacutainer@ (Becton Dickinson, France) and immedi-
Biological Trace Element Research
VoL 58, 1997
Krachler, Wirnsberger, and trgolic
2t2
TabIe t Operating Conditions for the 1CP-MS Plasma
rf power
Argon gas flows
Cooling gas Auxiliary gas Nebulizer gas Nebulizer
Nebulization Ion sampling Vacuum Measurement
Sample cone Skirmner cone Expansion Intermediate Analyzer Channels/ainu Dwell time Mass region Data aquisition
Forward 1.40 kW Reflected < 5 W 13.5 L min -l 1.1 L rain -1 0.84 L rain -1 Meinhard concentric glass nebulizer type: SB-30-A3, uptake = 1.0 mL rain -I Nickel, orifice 1 mm diameter Nickel, orifice 0.7 mm diameter 1.6 mbar 1.0 x 10-4 mbar 2.1 x 10-6 mbar 24 320 ~s 7-210 B Scanning mode
ate centrifugation for 10 min at 1800g at 4~ The centrifuged plasma was transferred into 2-mL sterile tubes (Nalgene| UK). All samples were stored at 4~ in a refrigerator until they were mineralized.
Digestion of Whole Blood, Plasma, and S e r u m Blood, plasma, or the freshly reconstituted reference materials were gently swirled to obtain complete homogenization. Approximately 1 g (range: 0.9-1.1 g) of whole blood or plasma was poured into a weighed Teflon digestion vessel. The vessel containing the sample was weighed to 0.1 mg. Concentrated nitric acid (1.5 mL) and 30% hydrogen peroxide (0.5 mL) were added. The vessels were shaken to mix the contents, then closed, fixed in the rotor, and digested under the conditions given in Table 2. After completion of the digestion the rotor was placed into a water bath and cooled for 20 min. The solutions were transferred quantitatively into 10-mL volumetric flasks. Aliquots (50 ~tL) of Standard 3 (Ga, In, Re) were added and the flasks were filled to the mark with NANOpttre water. The solutions were analyzed for trace elements with ICP-MS.
RESULTS AND DISCUSSION Graphite furnace atomic absorption spectrometry (GFAAS) has been extensively used for the determinations of single elements in biological samples because of the low detection limits of this technique. Inductively coupled plasma atomic emission spectrometry (ICP-AES) has for several elements higher detection limits than GFAAS but provides multielement capabilities. Inductively coupled plasma mass spectrometry (ICP-MS), a Biological Trace Element Research
Vol. 58, 1997
213
Trace E l e m e n t S t a t u s o f H e m o d i a l y z e d Patients Table 2 Optimized Digestion Program for the Mineralization of Whole Blood, Serum, and Plasma Step 1 2 3 4 5 6 7 8 9 10 11
Reagent conc. HNO3 30 % H202 --~ ---~ --
-
Total Volurne:
Volume, mL
Microwave power, W
1.5 0.5 ---~, -----
--250
-
-
-
-
300 400 500 600
-
2.0 mL
Total Time:
Time, min
2.0 0.5 5.0 0.5 10.0 0.5 5.0 4.0 2.0 27.5 rain
multi-element technique with very, low detection limits (0.01-0.1 n g / m L for most of the elements), unites the advantages of low detection limits of GFAAS with the multi-element approach of ICP-AES. For optimal performance of the ICP-MS instrument, as many components of the matrix as possible should be removed. The necessity to completely oxidize the organic matrix of biological samples prior to the determination of trace elements by instrumental methods to achieve accurate and reproducible results is widely recognized (14,15). The organic (but not the inorganic) matrix can be r e m o v e d by oxidative conversion to carbon dioxide and water. Closed-pressurized microwave digestion can be advantageously used for the mineralization of biological samples (16). To keep blank levels low, high-purity reagents required for sample preparation (acidification, dilution, mineralization) m u s t be used sparingly. Consequently, the very low detection limits claimed for ICP-MS can be approached only w h e n the concentrations of total dissolved solids in the solutions to be analyzed are kept to a m i n i m u m . High concentrations of inorganic and organic solids change the properties of the argon plasma, cause deposits to f o r m on the central tube of the plasma torch, m a y lead to clogged orifices in the cones of the ICP-MS, and will certainly increase the detection limits. Contamination during sample preparation must be minimized to obtain reliable results. Versieck et al. (5) discuss the problems of sampling and contamination troubling determinations of trace elements in h u m a n plasma and serum.
Microwave-Assisted Digestion For the determination of trace elements, sample manipulations must be m i n i m i z e d to reduce the chances for contamination. To free blood, Biological Trace Element Research
Vol. 58, 1997
214
Krach ler, Wirnsberger, and Irgolic
serum, and plasma from the organic matrix, a suitable digestion procedure has to be employed. In recent years, microwave digestion became popular, because of the many advantages it offers for sample preparation. Microwave digestion has better reproducibilit}, is more accurate, and requires less time than conventional digestions on hot plates in open beakers. To keep blank levels as low as possible, small volumes of concentrated nitric acid (1.5 mL) purified by sub-boiling distillation and high purity hydrogen peroxide (0.5 mL) were used for the mineralization of the samples. The selection of appropriate digesting reagents that are compatible with ICP-MS and the optimization of digestion parameters were described earlier (16). The optimized procedure for the mineralization approximately of 1 g whole blood, plasma, and serum yielded colorless, homogeneous digests within 30 min (Table 2).
S i m u l t a n e o u s D e t e r m i n a t i o n of Trace Elements by ICP-MS Although [CP-MS has lower detection limits for most elements than other multi-element techniques, not all essential and toxic elements that should be determined can be quantified by low-resolution, quadrupole 1CP-MS (17). Interferences from atomic and molecular ions produced in the plasma from argon and matrix constituents of the aspirated solutions plague the quantification of several elements (for example, 51V/35C1160;
75As/40Ar35Cl; 77Se/40Ar37C1; 80Se/40Ar2; 82Se/I2C35C12; 52Cr/40Ar12C) when low-resolution ICP-MS systems with quadrupole mass spectrometers are used. To minimize interferences, as many components of the matrix as possible should be removed, for example by microwave digestion (16). High resolution ICP-MS can overcome some of the polyatomic interferences by separating the signals of the analyte and the interfering species (17). Other elements, such as Fe (in blood) or Na (in blood and plasma), would require further dilution of the samples to achieve concentrations that do not cause detector-overflow. Consequently, we chose elements for our study that are free from spectral overlaps and can be quantified simultaneously by ICP-MS.
Quality Assurance To assess the accuracy of the analyses, two whole blood reference materials (one unspiked and one spiked) were analyzed with every sampie batch. In the unspiked material only three concentrations (Cd, Hg, Pb; four if the "less than" value for cobalt is included) are certified. In the spiked whole blood reference material the spikes produce concentrations for Bi, Co, Hg, Mo, Sb, and T1 much higher than the normal ranges in whole blood and, therefore, are not very useful for quality control. Better reference materials for whole blood, serum, or plasma are unfortunately not commercially available. The average concentrations of 16 Biological Trace Element Research
Vol. 58, 1997
Trace Element Status of Hemodialyzed Patients
215
elements determined in the digests of the whole blood reference materials with ICP-MS are summarized in Table 3. The experimental concentrations for Cd, Co, and Pb in the unspiked whole blood are within the certified ranges. The concentrations for several elements (Bi, Cd, Co, La, Mo, TI) in the unspiked whole blood have high relative standard deviations (17% to 56%), because the concentrations are close to the method detection limits (Table 3). The concentration for Hg is below the detection limit of 2.2 ~tg/L, although 5 ~tg/L are certified. Mercury is known to produce low signals when determined by ICP-MS (18,19). The concentrations of Sb and Sn are below the method detection limits. The concentration for Ba at 137 ~tg/L is approx 100 times higher than the normal range (0.47-2.4 ~tg/L). The source for this contamination with Ba could be the anticoagulant heparin (20). The results obtained with the Seronorm spiked whole blood are in acceptable agreement with the expected concentrations with exception of Hg and Bi (Table 3). The low value for Hg is an artifact of the ICP-MS method. The low concentration for Bi could be caused by the incomplete dissolution of Bi upon reconstitution of the reference material with water. Bismuth is known to easily form basic salts of low solubility in weakly acidic and neutral aqueous systems.
Trace E l e m e n t s in Whole Blood of H e m o d i a l y z e d Patients The median, 50% quartile, and the observed range for each of the 15 trace elements determined in the whole blood of 68 hemodialyzed patients are juxtaposed in Table 4 with concentration ranges for healthy adults. The averages for most elements are close to the median values suggesting a Gauss distribution of the concentrations. Only for Co (average 5.16, median 4.72) and Mo (7.69, 6.43) are the averages considerably higher than the medians. The concentrations for Bi, Hg, Pb, Sb, and Sr are within the literature ranges for healthy adults, whereas the concentrations for Cd and Co were considerably higher. Approximately 26% of the concentrations for Cs and 15% for Zn exceeded the high limits for healthy individuals. Contamination from heparin employed to prevent the coagulation of blood (20) elevated the Ba concentrations to approx 100 times the normal values in blood. Consequently, barium concentrations from blood treated with heparin are useless for monitoring the influence of hemodialysis on the barium concentration. Concentrations for Cu and Rb were in the lower regions of the normal ranges. The concentration range of 1.2 to 15 ~tg/L for molybdenum in the hemodialyzed patients has only a small region at the lower end of the normal range (5 to 157 ~tg/L) in common. The results of determinations of Mo in blood of healthy persons and patients with coronary heart disease (range: < 0o63-6.17 .ug/L) (23) suggest that the normal range might be too wide and that dialysis patients may have elevated concentrations Biological Trace Element Research
VoL 58, 1997
216
Krachler, Wirnsberger, and kgolic
ddddd~ddddddddd~ ,S r~
0 G
o4
O~
o
0
o 11 0 r.~
>,,.~
+1 +1 +1 +1 +1 +1 +l +l +1 +I +i + 1 ~
0
+t +1 ~
.~
o
r~
r~ r...,l
