Anal Bioanal Chem (2005) 383: 476–482 DOI 10.1007/s00216-005-0010-0
SPECIAL I SSU E PAPER
Imre Varga . Ágnes Szebeni . Norbert Szoboszlai . Béla Kovács
Determination of trace elements in human liver biopsy samples by ICP–MS and TXRF: hepatic steatosis and nickel accumulation
Received: 30 March 2005 / Revised: 4 July 2005 / Accepted: 7 July 2005 / Published online: 13 September 2005 # Springer-Verlag 2005
Abstract Human liver biopsy samples, collected from 52 individuals, were analysed by inductively coupled plasma– mass spectrometry (ICP–MS) and total reflection X-ray fluorescence (TXRF) spectrometry in a retrospective study (i.e. patient selection and liver biopsy were not for the purpose of element analysis). The freeze-dried samples (typically 0.5–2 mg dry weight) were digested in a laboratory microwave digestion system and solutions with a final volume of 1 mL were prepared. The concentrations of Cr, Mn, Fe, Ni, Cu, Zn, Rb, and Pb were determined by use of a Thermo Elemental X7 ICP–MS spectrometer. TXRF measurements were performed with an Atomika Extra IIA spectrometer. Yttrium was employed as an internal standard, prepared by dissolution of 5N-purity yttria (Y2O3) in our laboratory. The accuracy was tested by analysis of NIST 1577a Bovine Liver certified reference material. The concentrations of Fe, Cu, Zn, and Rb determined in human liver biopsy samples were in good agreement with data
This paper was presented in part at the 2005 European Winter Conference on Plasma Spectrochemistry Budapest, Hungary. I. Varga (*) . N. Szoboszlai Department of Inorganic and Analytical Chemistry, L. Eötvös University, P.O. Box 32, 1518 Budapest, Hungary e-mail:
[email protected] Tel.: +36-1-2090555 Fax: +36-1-2090602 Á. Szebeni Ultrasound Laboratory, MI Central Hospital, Budakeszi út 48/B, 1121 Budapest, Hungary B. Kovács Department of Food Science and Quality Assurance, Debrecen University—Centre of Agricultural Sciences, P.O. Box 36, 4015 Debrecen, Hungary
published by other authors. The distribution of nickel in the samples was surprisingly uneven—nickel concentrations ranged from 0.7 to 12 μg g−1 (dry weight) in 38 samples and in several samples were extremely high, 36–693 μg g−1. Analysis of replicate procedural blanks and control measurements were performed to prevent misinterpretation of the data. For patients with steatosis (n=14) Ni concentrations were consistently high except for two who had levels close to those measured for the normal group. As far as we are aware no previous literature data are available on the association of steatosis with high concentration of nickel in human liver biopsies taken from living patients. Keywords Human liver biopsy . Hepatic steatosis . ICP–mass spectrometry . X-ray fluorescence analysis . Nickel Abbreviations AAS: Atomic-absorption spectrometry . AES: Atomic-emission spectrometry . CDLD: Chronic diffuse liver disease . CRM: Certified reference material . dw: Dry weight . INAA: Instrumental neutron-activation analysis . LOD: Limit of detection . MS: Mass spectrometry . NAFLD: Non-alcoholic fatty liver disease . SD: Standard deviation . ww: Wet weight . TXRF: Total reflection X-ray fluorescence
Introduction Chronic diffuse liver disease (CDLD) is very frequent [1–3]. In Hungary the prevalence of CDLD in an Internal Department of a General Hospital is 40%, in a Hepatology Department 49%. Incidence of the disease is increasing continuously, mostly because of the increase of obesity [4–7]. Ultrasonography is involved in the primary procedure for diagnosis of this patient group and in follow up of the patients. In CDLD the echogenicity and echodensity of the liver become higher (“bright liver”) than normal [1].
477
CDLD includes a variety of liver diseases with similar ultrasonographic signs. Nevertheless, two basic groups of bright liver can be differentiated according to attenuation: the low (group DI) and high (group DII) attenuation types. Quantitative attenuation measurement has been established on the basis of comparison of attenuation of patient’s liver with attenuation of a homogeneous tissue-equivalent reference phantom. By this method we can investigate correlation with other (e.g. clinical and histopathological) data [8–10]. Percutaneous liver biopsy provides important information with relatively low risk and in some cases of hepatic dysfunction is inevitable for diagnosis and staging. Non-alcoholic fatty liver disease (NAFLD) includes cases with steatosis alone and those with steatohepatitis. Most cases are associated with overweight or diabetes. Ultrasound or CT scan can diagnose steatosis. Van Hoek [11] and Choundry [12] stated that a liver biopsy is often needed to exclude other diseases and to assess inflammation and fibrosis. Collier confirmed the need for biopsy to predict progression of fibrosis in chronic hepatitis C virus infection [13]. Association between steatosis and AST, ALT, body weight, serum glucose, serum triglycerides, BMI, Gamma-GT, age, and unconjugated bilirubin was investigated by use of regression analysis [14]. The authors concluded that no reliable biochemical data could identify patients with severe chronic liver damage with sufficient sensitivity to avoid liver biopsy for diagnosis and staging of the disease. Inductively coupled plasma–atomic emission spectrometry [15], atomic-absorption spectrometry [16], total-reflection X-ray fluorescence spectrometry [17–19], inductively coupled plasma–mass spectrometry [20–24], proton-induced X-ray emission [25], and instrumental neutronactivation analysis [23, 25–26] have been used to determine metal concentrations in human blood and tissue samples. Many investigators studied elemental concentrations in connection with special liver diseases by analysis of biopsy samples. Villeneuve et al. [27] investigated the variability of hepatic iron in cases of hemochromatosis, alcoholic cirrhosis, and hepatitis C virus-related cirrhosis. Multiple biopsy samples were obtained at the time of liver transplantation (n=6) or at autopsy (n=2) and iron concentrations were measured by graphite furnace atomic-absorption spectrometry. Zinc concentration was determined by atomic-absorption spectrometry in liver biopsy samples from patients with hereditary hemochromatosis [28]. Hepatic zinc, copper, and manganese concentrations were measured in autopsy samples from 42 males and 31 females [29] and in biopsy samples from 49 infants with extrahepatic biliary atresia [30]. The primary objective of this work was to determine element concentrations in percutaneous human liver biopsy samples and to explore possible factors that can contribute to the formation of different ultrasonic images of liver in CDLD. Patients with Wilson’s disease and hemochromatosis—known to result in anomalous element accumulation—were excluded from this study.
Experimental Materials High-purity water from a Milli-Q apparatus and Suprapure nitric acid (Merck, Darmstadt, Germany) were used throughout the work. Polypropylene microvials (1.5 mL volume, used for sample storage) were cleaned with 0.5 mol L−1 nitric acid for 1 h then rinsed with high-purity water and dried in a clean bench at room temperature. PFA vials of 6 mL volume (Savillex, Minnetonka, MN, US) were used for microwave digestion. Standard solution containing yttrium was prepared by dissolution of highpurity (99.999%) Y2O3 in our laboratory. Y2O3 (1.270 g) was dissolved in 2.9 mL conc. hydrochloric acid by gentle heating. After cooling the volume was adjusted to 100 mL in a volumetric flask. Spike solutions for internal standardisation containing 0.1 mg dm−3 Y were freshly prepared every day of TXRF measurements. Sample preparation Percutaneous human liver biopsy samples were collected as part of the diagnostic procedure from 52 individuals, for whom clinical picture and laboratory data suggested CDLD, by using the disposable Hepafix-Menghini liver biopsy set (B. Braun Melsungen, Melsungen, Germany), in the MI Central Hospital Budapest. A portion of each biopsy material was used for histological examinations; that remaining was immediately weighed on a balance, freeze-dried in a Janetzki LGR 05 apparatus at −10°C and 0.01 MPa for 2×5 h, and stored for subsequent analysis in sealed polypropylene microvials. (The water content ranged between 40.9–65.3%.) Sample collection and the study were performed with the permission of the institution’s ethics committee. The freeze-dried human liver samples (typically 0.5– 2.0 mg dry mass) were weighed directly into 6-mL PFA vessels and 100 μL conc. nitric acid was added to each. Tightly capped vessels were placed in 120-mL digestion vessels containing 10 mL high-purity water. Four of the large vessels were placed in the carrousel of a CEM MDS2100 laboratory microwave digester. A maximum output power of 400 W was applied under controlled pressure, reaching 14 bar inside the large containers, for 15 min. The total volume of sample solutions was adjusted to 1 mL. Dilution was 1:20 for ICP–MS measurements. Instrumentation Inductively coupled plasma–mass spectrometry was performed with a Thermo Elemental X7 spectrometer using a 93% helium–7% hydrogen mixture as CCT (Collision Cell Technology) gas with a flow rate of 7.2 mL min−1. Forward
478
Na Mg K Ca Mn Fe Co Ni Cu Zn Rb Pb
2430±130 600±15 9260±70 120±7 9.9±0.8 194±20 0.21±0.05 a
158±7 123±8 12.5±0.1 0.135±0.015
ICP–MS measured±SD (μg g−1)
TXRF measured±SD (μg g−1)
2280±192 595±35 – 126±14 9.6±1.1 184±14 0.23±0.09 0.77±0.14 148±10 125±8 13.2±0.6 0.126±0.015
– – 11300±830 – 11.8±0.9 193±14 – n.d. 156±11 127±9 11.9±0.8 n.d.
Three independent replicates, concentrations in μg g−1 dry weight, elements are in order of increasing atomic number n.d. signifies concentration under the LOD a No reference value is given
rf power of 1410 W was applied with flow rates of plasma Ar 13 L min−1, auxiliary Ar 0.95 L min−1, and nebulizer Ar 0.84 L min−1. A normal flow, concentric type, pneumatic
Cu
Number of cases
12 8 4 0 0
30
60
90
120
Mn
12 8 4 0
150
0
5
Concentration (µg/g)
Number of cases
12
10
15
20
25
Concentration (µg/g)
Zn
8 4
Pb
12
0
8 4 0
0
100
200
300
400
500
0
Concentration (µg/g)
10
20
30
40
50
Concentration (µg/g)
Ni
12 Number of cases
Fig. 1 Concentration distribution of Cu, Mn, Zn, Pb, and Ni in 52 human liver biopsy samples. (Of 52 cases 49 are shown for Ni distribution because for three cases the concentration is outside the applied scale)
Number of cases
Certified values± certified range (μg g−1)
nebuliser was used for sample introduction with a sample uptake rate of 0.82 mL min−1. Instrument setting was performed in two steps. First, the settings were adjusted to maximize signal intensities for Co, In, and U without using CCT. In the second step flow rates of He and H2 gases were set using CCT to minimize possible polyatomic interferences. (The background was in the range of 80–900 counts.) The isotopes measured under these conditions were: 52Cr, 55Mn, 56Fe, 60Ni, 65Cu, 66Zn, 85Rb, and 208Pb. Total reflection X-ray fluorescence analysis was performed using an Atomika Extra IIA TXRF spectrometer equipped with line-focused X-ray tubes and an energy dispersive Si(Li) detector. Mo Kα 17.4 keV excitation energy and 1000 s data-acquisition live time were applied. K lines were used for Cr, Mn, Fe, Ni, Cu, Zn, and Rb, the L line was used for Pb determination. Sample solution (100 μL) was used to prepare specimens for TXRF analysis in the following manner. Sample solution (25 μL) was pipetted on to a previously siliconized quartz glass carrier and left to dry in a clean bench at 40°C. This procedure was repeated until 100 μL total volume of sample solution was reached. Finally 10 μL YCl3 solution containing 0.1 mg dm−3 Y was added as internal standard for quantification of TXRF measurements. A Perkin–Elmer Model 3110 atomic-absorption spectrometer equipped with deuterium arc background corrector, a Perkin–Elmer HGA-600 graphite furnace, and a
Number off cases
Table 1 Results from ICP–MS and TXRF analysis of NIST 1577a Bovine Liver CRM
8 4 0 0
30
60
90
120
150
210
240
270
300
330
Concentration (µg/g)
360
479 Table 2 Comparison of current results with literature data for normal subjects: concentration means—and (ranges) or ±SD—of elements in human liver Reference
Sample type
Weight
Cr (μg g−1)
This work
Biopsy
Dry
0.65 6.40 1160 (0.10–6.7) (4.12–10.8) (309–1850)
Mn (μg g−1)
Fe (μg g−1)
Ni (μg g−1)
Cu (μg g−1)
Zn (μg g−1)
Rb (μg g−1)
Pb (μg g−1)
5.4 28.8 168 6.1 2.8 (0.9–9.5) (11.0–70.4) (104–289) (3.2–11.5) (1.2–7.0)
Villeneuve Biopsy/ [27] autopsy Adams [28] Biopsy/ autopsy Subramanian Autopsy [33] Zhang [26] Autopsy Faa [34] Autopsy
Dry
Dry
0.53±0.34
1100±970
210±89
Dry Dry
0.36±0.39
705±321 1206 (184–3597)
186±41
Treble [29]
Autopsy
Dry
Orlowski [16] Benes [35]
Autopsy
Wet
Autopsy
Wet
–(447–2848) Dry
326±99
2.26 (0.22–12.9)
14.2 (1.7–104) 4.6
0.07
Results and discussion Selection and evaluation of analytical procedure A critical aspect of this study was the size of the samples, because sample size obtainable from living subjects is restricted. In this study percutaneous human liver biopsy
3.6
samples were collected from 52 individuals. Approximately half of each biopsy was used for histological examination, the remainder was immediately weighed on a balance and then freeze-dried. Thus 0.5–2 mg solid material was available for subsequent analysis. The choice of analytical methods was a key question—an adequate method should be capable of measuring very low concentrations simultaneously and sample consumption should be restricted to the minimum possible. Inductively coupled plasma-mass spectrometry has favourable detection limits and was selected for simultaneous determination of micro and trace elements. Total reflection X-ray fluorescence spectrometry is also a suitable and powerful technique for analysis of small-mass biopsy samples, because it requires small amounts of substance. The microwave digestion method used for human colon biopsy samples in previous
10000
1000 Concentration (µg/g)
Fig. 2 Box-whisker plot (median/quartiles/range) of nickel and iron concentrations in biopsy samples for three groups of patients separated by histology: normal (n=8), group DI (n=30), and group DII (n=14)
118.3 (38.5–231) 51.5 63
Perkin–Elmer AS-60 autosampler were used for verification measurement of nickel. Sample (10 μL) and Mg(NO3)2 (10 μL) modifier were pipetted on to the pyrolytically coated platform and a pretreatment temperature of 1400°C was applied for 30 s. The absorbance of nickel was measured at Ni I 232.0 nm by use of 2500°C as atomization temperature.
27.2±6.4
100
10
1
0,1 Ni: Normal
Ni: D-I
Ni: D-II
Fe: Normal
Fe: D-I
Fe: D-II
480
Distribution of elements in liver biopsy samples and comparison of results with literature data Concentration distributions of Cu, Mn, Zn, and Pb in 52 human liver biopsies are illustrated in Fig. 1. Data obtained by ICP–MS for Mn and means of data obtained by ICP– MS and TXRF for other elements were used to prepare the histograms. The frequency of elemental concentrations fit log–normal distribution as expected; the only exception was Ni, for which the distribution pattern was different. The hepatic concentrations of Mn, Fe, Cu, Zn, and Rb determined in this work were in good agreement with data published by other authors. Results for normal subjects are summarised in Table 2. Correlation between steatosis and Ni concentration Fig. 3 TXRF spectrum of a human liver biopsy having a Ni concentration below the detection limit under the applied conditions (mentioned in the text)
work [17] was slightly modified (described in detail above) yielding 1 mL final volume of sample solution. ICP–MS measurements were performed on 200-μL volumes diluted to 4 mL; 100-μL samples were used to prepare specimens for TXRF analysis. The remaining sample was retained for control measurements. Accuracy was tested by the analysis of NIST 1577a Bovine Liver certified reference material. Concentration means of three independent replicates (14–29 mg CRM was digested to prepare 2.5 mL solution) are illustrated in Table 1. The results obtained by both techniques were in good agreement with the certified (or consensus [26]) values. The homogeneity of NIST 1648 Urban Particulate Matter CRM and NIST 1577a Bovine Liver CRM have previously been demonstrated for selected elements by using milligram amounts of material to perform sequential leaching [31] and to accomplish in-situ digestion for TXRF analysis [32]. The results listed in Table 1 confirmed previous findings concerning the homogeneity of NIST 1577a CRM even down to several-milligram subsamples. Concentration data obtained by ICP–MS and TXRF for biopsy samples were in very good agreement for all the elements investigated except Mn. Chromium was below the calculated detection limit of TXRF (0.92 μg g−1 Cr corresponding to 1 mg irradiated sample and 1000 s measurement live time) for all the samples investigated. In TXRF the peak area calculation for Mn Kα at 5.9 keV was affected by a “wing overlap” from the adjacent Fe Kα peak at 6.4 keV causing small overestimation of Mn concentration if large excess of iron was present. A significant positive inter-element correlation was found between Mn and Fe concentrations both obtained by TXRF (r=0.8558, P=8.12×10−13) proving the suspected partial overlap. No Fe–Mn correlation was observed for ICP–MS (r=0.0166, P=0.901); Mn results for ICP–MS were therefore regarded as adequate.
The primary objective of this study was to measure element concentrations in human liver biopsy samples from patients with chronic diffuse liver disease and to explore possible correlations between medical diagnosis and concentrations of microelements. The patients in this retrospective study were separated into three groups according to attenuation of ultrasound and histology: normal group (n=8) and two groups with diffuse liver lesions (group DI, n=30 and group DII, n=14). Patients with histological results of steatosis alone or steatosis and cirrhosis all belonged to group DII. In contrast with the other elements investigated distribution of nickel in the samples was surprisingly uneven—nickel concentrations ranged from 0.7 to 12 μg g−1 (dry weight) in 38 cases and were in several cases extremely high, even as high as 693 μg g−1. The mean Ni and Fe concentrations were found to be significantly different between the groups. The mean nickel concentration in group DII was more than 10 times that in
Fig. 4 TXRF spectrum of a biopsy from a patient’s liver with histopathology results of steatosis. (The remarkably high Ni concentration is comparable to that of Zn)
481
either the normal or the DI groups (Fig. 2). In contrast, the opposite pattern was observed for iron. Mean iron concentration (574 μg g−1) in group DII was approximately half of those in other groups (1160 and 1123 μg g−1). For the other elements differences between groups were not significant at the level of P≤0.05. Sample contamination during autopsy was proposed as the major source of uncertainty in trace element analysis [36, 37]. Because of its ubiquitous occurrence in the environment and in the laboratory, determination of nickel in biological samples is difficult and often affected by contamination. Recommendations were outlined to eliminate possible contamination during sample collection [38]. The stainless steel needles used during venipuncture can contaminate blood with Cr, Mn, Fe, Co, and Ni. For measurement of these elements in blood an acceptable alternative is the use of a polypropylene intravenous canulla mounted on a trocar. The biopsy needles used in our study were made of steel but could not be substituted by polypropylene or Teflon utensils. To prevent misleading results control measurements were performed. All materials, solutions and tools used during sample collection and analysis were tested, the steel parts of disposable biopsy needles and scalpels were also digested and analysed. The disposable scalpel supplied with the Menghini-set was made of chromium alloyed steel without Ni content. The material of biopsy needle was found to be a standard stainless steel (main composition Cr:Fe:Ni/18:72:9), suggesting that during sample collection the biopsy samples can be contaminated with both nickel and chromium. In this work high Ni concentrations were exclusively observable in cases of steatosis. In neither normal nor diseased liver were elevated Cr levels observed (there was no significant difference between mean Cr levels of the normal and two disease groups) which precludes the possibility of substantial Ni contamination during sample preparation. In addition, no correlation between Ni and Cr concentration could be detected (r=0.1254, P=0.340) on the basis of our data. TXRF spectra of a biopsy from normal liver and a biopsy from a liver showing steatosis are illustrated in Figs. 3 and 4, respectively.
Conclusions In this paper results from ICP–MS and TXRF analysis of element distribution in 52 human liver biopsy samples taken from living patients are presented. The hepatic concentrations of Mn, Fe, Cu, Zn, and Rb determined in this work were in good agreement with the data published by other authors. Although it is usually recommended that biopsy samples taken by use of stainless steel needles are analysed for nonsteel metals only, nickel determination in liver biopsies may also give valuable information about irregularly increased levels of this metal. According to our results a preliminary decisive test with a threshold level of Ni could be established. Samples in which concentrations exceeded
15 μg g−1 Ni (on a dry-weight basis) all belonged to the group of patients with steatosis (group DII). For two subjects out of fourteen in group DII, however, the test gave false negatives results (i.e. histology showed steatosis but the Ni concentrations were below the threshold limit of the preliminary test). A possible explanation of the two false negative test results in the presence of steatosis could be that tissue accumulation depends not only on disease but also on ambient metal concentration, speciation, and uptake mechanism (inhalation, diet, occupational exposure). Although the possibility of contamination of our biopsy samples by elements from steel needles could not be completely discounted, substantial Ni contamination was disproved indirectly. Neither increased levels of chromium in group DII compared with other groups nor significant correlation between Cr and Ni concentrations in the samples was observed in the biopsies investigated. Exceptionally high values of Ni concentration in 12 of 52 samples are hard to interpret. The mechanism responsible for abnormalities in Ni accumulation is still obscure. Limitations of our study were that it was a retrospective analysis and no information about exposure to nickel compounds has yet been provided. For populations without significant occupational exposure approximately 100 μg day−1 dietary intake of nickel has been estimated [39]. The average nickel intake was found to be 170 (women) and 192 (men) μg day−1 during a three-year test in eastern Germany [40]. In addition, 14% (women) and 26% (men) retention of consumed nickel was also observed. The authors stated that adults store substantial amounts of nickel in the body gradually filling a “nickel pool”. A reduction of average nickel intake in Germany between 1992 and 1996 was reported [41] by the same research group. Denkhaus and Salnikow reviewed the literature data on nickel requirement, toxicity, and carcinogenicity. They concluded there was no correlation between nickel concentration and any type of cancer and that accumulation of Ni in tissues in some individuals mostly reflects lifetime exposure [42]. In contrast, our measurements revealed a probable correlation between Ni accumulation and hepatic steatosis. Acknowledgements This work was supported by Hungarian Scientific Research Fund (OTKA project T047047) and by Hungarian Academy of Sciences (Bolyai János Research Fellowship granted to Imre Varga).
References 1. Lonardo A, Bellini M, Tartoni P, Tondelli E (1997) Ital J Gastroenterol 29:351–356 2. Koch CW, Pauwels EKJ (1976) Radiol Clin 45:282–291 3. Vehmas T, Kaukiainen A, Immonen-Räihä P, Lohman M, Luoma K (2005) Ultrasound Med Biol 31:293–299 4. Lonardo A, Bellini M, Tondelli E, Frazzoni M, Grisendi A, Pulvirenti M, Della Casa G (1995) Am J Gastroenterol 90:2073 5. Mokdad AH, Ford ES, Bowman BA, Dietz WH, Vinicor F, Bales VS, Marks JS (2003) JAMA 289:76–79 6. Arterburn DE, Crane PK (2004) J Am Geriatr Soc 52:1907– 1912
482 7. Jolliffe D (2004) Int J Obes 28:4–9 8. Szebeni Á, Pintér E, Stotz G, Tolvaj G, Juhász M, Dávid K (1990) Acta Med Hung 47:163–170 9. Szebeni Á, Tolvaj G, Osztrogonácz K, Dávid K, Rischák K, Szollár L (2000) Ultrasound Med Biol 26(S2):A139 10. Szebeni A, Tolvaj G, Zalatnai A (2004) Folia Hepatol 8 (Suppl.):20 (in Hungarian) 11. van Hoek B (2004) Scand J Gastroenterol 39(56–59 Suppl.):241 12. Choundry J, Sanyal AJ (2004) Semin Liver Dis 24:349–362 13. Collier JD, Woodall T, Wight DGD, Shore S, Gimson AE, Alexander GJM (2005) J Viral Hepatitis 12:74–80 14. Papadia FS, Marinari GM, Camerini G, Murelli F, Carlini F, Stabilini C, Scopinaro N (2004) Obes Surg 14:952–958 15. Rahil-Khazen R, Bolann BJ, Myking A, Ulvik RJ (2002) J Trace Elem Med Biol 16:15–25 16. Orlowski C, Piotrowski JK, Kubow M (1996) Int J Occup Med Environ Health 9:255–263 17. Tar Z, Varga I, Török A, Ress P (1995) 7th Italian-Hungarian Symposium on Spectrochemistry Roma, Italy, Abstr 91 18. Viksna A, Lindgren ES, Kjellmer I, Bursa J (2002) J Trace Microprobe Tech 20:553–564 19. Marco LM, Hernandez-Caraballo EA (2004) Spectrochim Acta B 59:1077–1090 20. Muniz CS, Fernandez-Martin JL, Merchante-Gayon JM, Alonso JIG, Cannata-Andia JB, Sanz-Medel A (2001) Biol Trace Elem Res 82:259–272 21. Boulyga SF, Becker JS, Malechenko AF, Dietze HJ (2000) Mikrochim Acta 134:215–222 22. Garcia F, Ortega A, Domingo JL, Corbella J (2001) J Environ Sci Health A 36:1767–1786 23. Császma I, Andrási E, Lásztity A, Bertalan É, Gawlik D (2003) J Anal At Spectrom 18:1082–1087 24. Labat L, Dehon B, Lhermitte M (2003) Anal Bioanal Chem 376:270–273 25. Shenberg C, Feldstein H, Cornelis R, Mees L, Versieck J, Vanballenberghe L, Cafmeyer J, Maenhaut W (1995) J Trace Elem Med Biol 9:193–199
26. Zhang PQ, Chen CY, Horvat M, Jacimovic R, Falnoga I, Logar M, Li B, Zhao JJ, Chai ZF (2004) Anal Bioanal Chem 380:773–781 27. Villeneuve J-P, Bilodeau M, Lapage R, Côté J, Lefebvre M (1996) J Hepatol 25:172–177 28. Adams PC, Bradley C, Frei JV (1991) Clin Invest Med 14: 16–20 29. Treble RG, Thompson TS, Lynch HR (1998) BioMetals 11: 49–53 30. Bayliss EA, Hambidge KM, Sokol RJ, Stewart B, Lilly JR (1995) J Trace Elem Med Biol 9:40–43 31. Varga I, von Bohlen A, Záray G, Klockenkämper R (2000) J Trace Microprobe Tech 18:293–302 32. Marco LM, Capote T, Hernandez C, Greaves ED (2001) Spectrochim Acta B 56:2187–2193 33. Subramanian KS, Meranger JC, Burnett RT (1985) Sci Total Environ 42:223–235 34. Faa G, Sciot R, Farci AMG, Callea F, Ambu R, Congiu T, Van Eyken P, Cappai G, Marras A, Costa V, Desmet VJ (1994) Liver 14:193–199 35. Benes B, Jakubec K, Smid J, Spevackova V (2000) Biol Trace Elem Res 75:195–203 36. Versieck J, Speecke A, Hoste J, Barbier F (1973) Clin Chem 19:472–475 37. Versieck J, Barbier F, Cornelis R, Hoste J (1982) Talanta 29:973–984 38. Cornelis R, Heinzow B, Herber RFM et al (1995) Pure Appl Chem 67:1575–1608 39. Grandjean P, Nielsen GD, Andersen O (1989) In: Menné T, Maibach HI (eds) Nickel and the skin: immunology and toxicology. CRC Press, Boca Raton FL, pp 9–34 40. Anke M, Angelow L, Glei M, Müller M, Illing H (1995) Fresenius’ J Anal Chem 352:92–96 41. Seifert M, Anke M (1999) Trace Elem Electrolytes 16:17–21 42. Denkhaus E, Salnikow K (2002) Crit Rev Oncol/Hematol 42:35–56
