Joshua Akin (Presenter)
UC San Diego Health System
Bio: Josh Akin is a CLS Specialist in the Clinical Mass Spectrometry Laboratory, Center for Advanced Laboratory Medicine at UC, San Diego Health System. His research interests include clinical mass spectrometry, trace metal analysis, and clinical trial design. He is currently pursuing a masters degree in Clinical Research at UC San Diego.
Authorship: Joshua Akin, Judy Stone, Robert Fitzgerald
UC San Diego Health System
| Short Abstract We developed a quantitative method for analysis of As, Cd, Hg and Pb in clinical whole blood samples using ICP-MS. We evaluated previously published strategies for ICP-MS internal standard selection, acid wash concentration and matrix-matching calibrators. Internal standardization can help correct for non-spectral interferences. Rhodium and Iridium performed optimally for all four analytes. Wash cycle duration proved more effective than acid rinse concentration for reducing Mercury memory effects and preventing carryover. We showed that four commonly used whole blood anticoagulants can be suitable for matrix-matching calibrators. Development of analytical methods for ICP-MS can be challenging and several factors must be considered in the validation phase. |
Long Abstract
Introduction:
Inductively coupled plasma mass spectrometry (ICP-MS) is an instrumental method for simultaneous determination of trace elements that is used in the clinical laboratory. We developed an assay for measurement of Arsenic (As), Cadmium (Cd), Mercury (Hg) and Lead (Pb) in whole blood using ICP-MS. In this study, we evaluate several previously published strategies for ICP-MS method optimization and contrast their utility in a clinical diagnostic setting.
Background:
Currently, the most powerful technique for measurement of trace metals in biological matrices is ICP-MS. ICP-MS allows for simultaneous determination of trace elements, while providing the dynamic range and throughput needed for a clinical laboratory. Development of analytical methods for ICP-MS can be problematic for inexperienced laboratorians and several factors must be considered in the validation phase. Whole blood samples present several challenges to method development, namely matrix effects and greater variance from small differences in pipetting technique. Matrix-matching calibrators is complicated by the anticoagulant used, as well as the volume required for routine production. Fortunately, most clinical laboratories have an ample stock of sample blood with various anticoagulants. The use of an internal standard can help correct for non-spectral interferences. Internal standards are typically matched to analytes based upon their mass and ionization potential. In addition, maintaining the elements of interest in solution poses additional challenges to sample preparation, introduction, and analysis. Nitric acid is commonly used in ICP-MS analysis due to its oxidizing ability and relative freedom from most chemical and spectral interferences. The degree of ionization is dependent upon plasma temperature and first ionization energy. The plasma temperature decreases with increasing concentrations of acid. Therefore Mercury, which has the highest first ionization potential of these analytes, has the tendency to be most affected by changes in acid concentration. This is compounded by the well documented "mercury memory" effect. Mercury adheres to spray chamber walls, remaining as a vapor in the chamber. Acid rinse solutions prevent mercury adsorption, but volatility still persists. To overcome this, gold or dichromate is typically added, which prevents mercury volatilization and adsorption losses. Proper wash-out using HNO3 therefore is dependent upon minimizing carryover or “memory effect” while optimizing signal
Materials and Methods.
An aqueous stock solution of As, Cd, Hg, and Pb was prepared from certified reference material in nitric acid (0.5%). Whole blood calibrators at five concentrations covering the analytical range of interest were prepared each day of analysis by spiking blank whole blood with the aqueous stock solution. 0.25 mL of whole blood is added to 4.25 mL of sample diluent (0.5% HNO3, 0.005% Triton X-100, 1mg/L gold) in 15 mL plastic conical tubes for a 1:20 dilution. 0.25 mL of internal standard (50 ug/L Ge, 1.67 ug/L Rh and 125 ug/L of Ir) is added to all samples except for the double blank. Three UTAK lyophilized controls are also included in each batch. All samples are placed on a rocker for 10 minutes, and then centrifuged at 3,000 RPM for 10 minutes.
An Agilent 7700 ICP-MS was used to acquire data, with subsequent analysis using MassHunter software. Sample was introduced to the ICP-MS by a ASX-500 auto sampler, followed by a double-walled spray chamber with Micro Mist Nebulizer via peristaltic pump. The instrument was operated in no gas mode for Hg, Pb and Ir, and gas mode for As, Cd, Rh, Ge and Ir with He flowing at 3.8 ml/min. Standard ALS rinse is 5% HNO3 and 0.5% HNO3 probe rinse with 3 minutes for both rinses. All reagents were of trace metal grade and 18Ω water was generated using an ELGA PureLab Ultra ultrafiltration unit.
Since concentration of acid used in the wash step has been documented to have an effect on ionization efficiency and memory effect, we evaluated various concentrations of acid for the wash step. We evaluated the effect of aqueous calibrators as compared with whole blood calibrators to determine if the matrix had an effect on accuracy of the measurement. We also wanted to determine if expired blood bank units (easily obtained) collected with acid citrate dextrose (ACD) anticoagulant could be used in place of EDTA (generally used for patient specimens) or other anticoagulants. Finally we evaluated the effect of using different internal standards on accuracy and precision.
To assess the effect of nitric acid rinse port concentration on analyte counts, four different concentrations of nitric acid were prepared at concentrations of 0.5%, 1%, 3% and 5%. A calibration curve, along with two blanks and three QC levels were analyzed at each acid concentration for a total of four separate batches. Each batch was contrasted to assess metal washout performance, particularly for mercury, which has a documented “memory effect” in previous studies.
We compared four different anticoagulants used for matrix-matching the calibrators. Four samples were collected in different anticoagulant tubes from the same individual. Four separate calibration curves were prepared, along with two blanks and three QC levels for each whole blood anticoagulant. These specimen were also analyzed to verify that endogenous levels of analytes were below the limits of quantification (<0.5 ug/dL Pb, <0.5 ug/L Hg, <10.5 ug/L Cd and <0.5 ug/L As).
Selection of internal standards was based upon published literature, as well as the molecular weight and ionization potential of the analytes of interest. For internal standards, we investigated Germanium (Ge), Iridium (Ir), Indium (In) and Rhodium (Rh). The target was to normalize the counts per second (CPS) of the internal standards to the highest calibrator for the matching analyte.
Results:
Analysis of blanks, calibration curve and QC using the four different concentrations of wash showed little difference during analysis. QC material analyzed using the different washes had low CV values (<10%). Mercury counts were also unaffected by varying the concentration of rinse acid. The rinse cycle between samples and the duration of nitric washes was critical for optimal washout, to avoid carryover and to minimize memory effects.
Evaluation of appropriate whole blood samples for matrix-matching the calibrators showed subtle differences in CPS that were dependent upon the anticoagulant used. CPS values were similar between ACD and Citrate, as well as EDTA and Heparin (CV <13% for all calibrators). This may be due to the citrate constituent shared in the ACD and Citrate anticoagulants. Calibrations between the matrices were similar, with calibration values showing <10% bias between the different anticoagulants. QC analysis using the different anticoagulants was also similar with a bias of < 10% for all the anticoagulant blood matrices.
Of the four internal standards that were evaluated, only two showed sufficient reproducibility between runs as well as over the duration of longer runs (>6 hours). Initial recovery studies showed that indium performed poorly, with CV values >20%. Germanium was initially chosen as the internal standard for As due to a similar ionization potential, but suffered from drift over the length of the run, as well as variability with different patient specimens. When using Ge, QC values for As varied as well, which suggested that the internal standard was not tracking with the analyte. When the internal standard for As was changed to Rh, QC precision improved (CVs < 10%) and results were within range (<10% bias). By using Iridium as internal standard for Pb and Hg, we were able to demonstrate good accuracy (<10% bias) and precision (<10% CV) when measuring QC samples over time (n=20).
Results of CAP PT samples for blood lead showed agreement within our established analytical measuring range. Preliminary patient comparisons with values obtained from a reference laboratory also had good agreement.
Discussion:
We developed a quantitative method for analysis of As, Cd, Hg and Pb in clinical whole blood samples using ICP-MS. Direct dilution methods, such as this, allow for a relatively rapid sample preparation, but increase the amount of dissolved solids entering the instrument, which demonstrates the need for internal standardization. Using an internal standard significantly improves analytical accuracy and precision, although there are no single physical or chemical properties that should be considered when choosing the appropriate internal standard. In our hands, Rhodium and Iridium performed optimally for all four analytes in whole blood.
Acid concentration of the wash and sample rinse are important for elemental solubility and ionization, as well preventing memory effects. Interface cones should always be pre-conditioned with sample matrix prior to analysis to assure consistency during sample infusion. Comparing different nitric acid concentrations of probe rinse did not significantly affect the background counts nor contribute to any additional variability in the measurement of mercury. Wash cycle duration and rinse protocol showed the most improvement in reducing the memory effect of mercury and other potential carryover issues.
For ICP-MS analysis of clinical samples, calibration matrix should match as much as possible to the sample matrix. Our results show that four commonly used whole blood anticoagulants can be suitable for As, Cd, Hg and Pb analysis in analytical ranges suitable for clinical diagnostics. Such suitability allows for the use of supplementary whole blood that may be at the laboratories disposal.
References & Acknowledgements:
References:
Goullé JP, Saussereau E, Mahieu L, Guerbet M. Current role of ICP-MS in clinical toxicology and forensic toxicology: a metallic profile. Bioanalysis. 2014 Aug;6(17):2245-59.
R. J. Bowins and R. H. McNutt Electrothermal Isotope Dilution Inductively Coupled Plasma Mass Spectrometry Method for the Determination of Sub-ng ml-' Levels of Lead in Human Plasma, J. Anal. At. Spectrom., 1994, 9, 1233–1236
Jian L, Goessler W, Irgolic KJ. Mercury determination with ICP-MS: signal suppression by acids. Fresenius J Anal Chem. 2000 Jan;366(1):48-53. PubMed
Vanhaecke F, Vanhoe H, Dams R, Vandecasteele C. The use of internal standards in ICP-MS. Talanta. 1992 Jul;39(7):737-42.
Fong BM, Siu TS, Lee JS, Tam S. Determination of mercury in whole blood and urine by inductively coupled plasma mass spectrometry. J Anal Toxicol. 2007 Jun;31(5):281-7.
Bolann BJ, Rahil-Khazen R, Henriksen H, Isrenn R, Ulvik RJ. Evaluation of methods for trace-element determination with emphasis on their usability in the clinical routine laboratory. Scand J Clin Lab Invest. 2007;67(4):353-66.
| Description | Y/N | Source |
| Grants | no | |
| Salary | no | |
| Board Member | no | |
| Stock | no | |
| Expenses | no |
IP Royalty: no
| Planning to mention or discuss specific products or technology of the company(ies) listed above: | no |