The undetected murder? Evaluation and validation of a practicable and rapid inductively coupled plasma-mass spectrometry method for the detection of arsenic, lead, and thallium intoxications in postmortem blood
Abstract
Homicide, suicide, or accident – elemental intoxication may be a cause in each of these types of deaths. Inductively coupled plasma mass spectrometry (ICP-MS) has emerged as the gold standard analytical method for toxic metal analysis in both clinical and forensic settings. An ICP-MS method was developed using a modified acidic workup for the quantitative determination of arsenic, lead, and thallium. Method validation focused on the assessment of linearity, between- and within-day precisions, limits of detection (LoD) and lower limits of quantification (LLoQ), and carryover. The method was applied to analysis of postmortem peripheral blood samples from 279 forensic cases for which orders for chemical–toxicological examination had been received from the public prosecutor's office. Using six-point and one-point calibrations (latter for rapid screening purposes), precisions and accuracies ranged from −4.8 to 5.8% and −6.4 to 7.5%. Analytical sensitivities for As, Pb, and Tl were 0.08, 0.18, and 0.01 μg/l (LoD) and 0.23, 0.66, and 0.03 μg/l (LLoQ), respectively. Observed postmortem peripheral blood concentrations were As, 1.31 ± 3.42 μg/L; Pb, 17.4 ± 13.1 μg/L; and Tl, 0.11 ± 0.07 μg/L (mean ± standard deviation [SD]). Elemental concentrations, determined in additional quality control samples, were in good agreement to those obtained with an external ICP-MS method based on alkaline sample processing. The current method is practicable and compatible with an ICP-MS system used for trace element analysis in an accredited medical laboratory. It allows for implementation of low-threshold investigations when metal intoxications are suspected in forensic routine.
1 INTRODUCTION
Apart from the standard diagnostic repertoire of forensic institutes, few laboratories have the facility to test for toxic metals. Atomic absorption spectrometry has been largely phased out, with inductively coupled plasma-mass spectrometry (ICP-MS) emerging as a superior detection method, due to its speed of detection, accuracy, flexibility, simplicity, and reproducibility, allowing for much faster and more comprehensive analysis.1 However, the uptake of ICP-MS has been much slower in the clinical and forensic landscape, likely due to its relatively high (running) costs and the requirement for more extensive in-house expertise.2 The consequence of specimens having to be sent to an external laboratory is that examinations for toxic metals are largely restricted to cases where there is a strong suspicion of heavy metal toxicity.
In recent years, several cases of heavy metal (e.g., thallium, mercury, lead) intoxication, some with lethal consequences, have been dealt with in Germany.3-5 These cases exemplify why investigating for toxic metals remains both relevant and important to consider. At the Institute of Legal Medicine of the University Hospital of Cologne, it was usual for only those cases in which there was a strong suspicion of metal intoxication that samples were sent to an external laboratory for toxic metal analysis. Questions arose as to whether intoxications might have been overlooked as a result of this approach and whether it would be possible to establish an analytical method suitable for routine use, enabling a broader, low-threshold, and more cost-effective investigation.
Trace elements and their concentrations play an important role in Clinical Chemistry and Laboratory Medicine.6 In Cologne, ICP-MS (carried out in an accredited medical laboratory setting) is primarily used to perform analyses for trace elements in human serum and plasma. A collaboration was thus formed to extend its application to the analysis of heavy metals in forensic specimens. An essential requirement was to establish a practicable and rapid “screening method” in postmortem blood samples for arsenic, lead, and thallium that could be utilized in forensic routine even in cases of unexplained death with low probability of metal poisoning.
2 MATERIALS AND METHODS
2.1 Chemicals and reagents
Ultrapure deionized water (18.2 MΩ cm) was obtained from the WernerEASYpure UV compact ultrapure water system (Wilhelm Werner GmbH, Leverkusen, Germany). Suprapur® nitric acid (69%) was purchased from Merck (Merck KGaA, Darmstadt, Germany). A volume of 580 mL of nitric acid (67%) was added to ultrapure water, and the solution then filled up to 2 L to obtain a nitric acid (20%) working solution. Tergitol™ 15-S-9 (100%) was obtained from SERVA (SERVA Electrophoresis GmbH, Heidelberg, Germany), and 1-propanol was purchased from VWR (VWR International GmbH, Darmstadt, Germany). ICP-MS tuning solution 10 ppm containing Be, Mg, Co, In, Ba, Ce, Tl, Pb, and Th (each at 10 mg/L) in 1% nitric acid and ICP-MS internal standard solution 100 ppm containing Li, Sc, Y, In, Tb, and Bi (each at 100 mg/L) in 5% nitric acid were obtained from Analytik Jena (Analytik Jena GmbH+Co. KG, Jena, Germany). Lyophilized whole blood calibrators (ClinCal®) and whole blood controls (ClinChek®) for trace elements were purchased from Recipe (Recipe Chemicals + Instruments GmbH, Munich, Germany). Fifteen- and 50-mL polypropylene tubes (PP tubes) were purchased from Greiner (Greiner Bio-One GmbH, Frickenhausen, Germany). Argon 4.8 gas for spectrometry was supplied by Linde (Linde GmbH, Pullach, Germany).
2.2 Postmortem samples
At autopsy, whole blood samples (from the femoral vein) were collected in 10-mL tubes and stored at 4°C for 8–11 months until analysis. Orders of the public prosecutor's office for chemical–toxicological examinations were available for the cases. These contain a basic order for extensive analyses and also include tests for heavy metals, however, which are not mandatory for each case. With one exception, there was no initial suspicion of metal poisoning.
2.3 Instrumentation
A PlasmaQuant MS system (Analytik Jena) was used to measure element concentrations of As, Pb, and Tl. The PlasmaQuant MS is a quadrupole-based mass spectrometer equipped with an integrated collision reaction cell (iCRC). A Precision Hydrogen generator (Peak Scientific, Inchinnan, Scotland, UK) was utilized for hydrogen (H2) generation, which served as reaction gas for spectral interference management in ICP-MS.
The system was equipped with a Cetac autosampler ASX-560 series (Teledyne CETAC Technologies, Omaha, USA) combined with an ENC series 560 DC autosampler enclosure (Teledyne Cetac Technologies) and a recirculating water chiller (Smart H150-3000, LabTech, Hopkinton, USA). A low flow glass concentric nebulizer (400 μL/min) combined with a double-pass Scott-type spray chamber (Analytik Jena) was utilized for sample introduction. The spray chamber was Peltier-cooled at 3°C. An argon-based plasma was generated inside a low-flow quartz torch with a 2.4-mm-inner-diameter injector (Analytik Jena). The plasma interface consisted of water-cooled, high-performance nickel sampler and skimmer cones and a set of three extraction lenses. Analyte ions were focused to the aperture of the quadrupole by integrated ion optics. All settings and iCRC gas flows were automatically optimized, and the system operated using ASpect MS software version 4.3.3 (Analytik Jena).
ICP-MS operating conditions were 1.24 kV generator power, 9 L/min plasma gas flow, 1.35 L/min auxiliary gas flow, and 1.02 L/min nebulizer gas flow. For the determination of As, Pb, and Tl, the system was operated either in H2 or no gas mode. Analytical details are summarized in Table 1.
| Analyte | Selected isotope(s) | Internal standard | ICP-MS mode |
|---|---|---|---|
| Arsenic (As) | 75As | 89Y | 2H, iCRC gas flow 80 mL/min |
| Lead (Pb) | 206Pb, 207Pb, 208Pb | 159Tb | No gas |
| Thallium (Tl) | 205Tl | 159Tb | No gas |
- Abbreiations: 2H, hydrogen; iCRC, integrated collision reaction cell; 159Tb, selected terbium isotope; 89Y, selected yttrium isotope.
2.4 Preparation of calibrators and controls
Whole blood calibrators and controls were reconstituted following manufacturer's instructions by adding 5 mL of ultrapure water. Specified mean concentrations of As, Pb, and Tl in whole blood controls (levels I–III) were As, 2.99, 5.37, and 10.2 μg/L; Pb, 35.0, 90.7, and 250 μg/L; and Tl, 0.81, 4.05, and 8.11 μg/L. Specified mean values of As, Pb, and Tl in the whole blood calibrator were 14.2, 341, and 10.5 μg/L, respectively (herein after referred to as level VI calibrator). For ICP-MS method development and validation experiments, fiveadditional calibrators were prepared by diluting defined volumes of reconstituted whole blood calibrator (level VI). Briefly, ultrapure water was added to whole blood calibrator sample aliquots into 15-mL PP tubes to final volumes of 3 mL, each. Overall obtained calibrator concentrations (levels I–VI) were As, 0.59, 2.84, 5.68, 8.52, 11.4, and 14.2 μg/L; Pb, 14.2, 68.2, 136, 205, 273, and 341 μg/L; and Tl, 0.44, 2.10, 4.20, 6.30, 8.40, and 10.5 μg/L. Calibrators (levels I–VI) and controls (levels I–III) were aliquoted by transferring sample volumes of 200 μL into 15-mL PP tubes, which were stored at −20°C until analysis.
2.5 Sample preparation
A sample preparation solution was arranged containing 1-propanol (20%, v/v), tergitol (0.5%, v/v), and internal standard solution (0.08%, v/v) in ultrapure water. 1-Propanol served as a carbon source to adjust the sample matrix of diluted calibrators for carbon, and tergitol served as a detergent. Frozen calibrators and quality controls were thawed at room temperature (RT) and briefly vortexed. Five-hundred microliters of preparation solution and 200 μL of nitric acid (20%) solution were added to volumes of 200 μL of calibrators, controls, and forensic blood samples as well as ultrapure water for each analytical run (serving as blank). Samples were then sonicated for 1 min and incubated for 30 min at RT. Thereafter, 4100 μL of ultrapure water were added and samples thoroughly vortexed. After centrifugation (10 min, 1945×g, RT), clear supernatants were transferred to 15-mL polypropylene tubes for ICP-MS analysis.
2.6 Method validation
The method was validated according to the Society for Toxicology and Forensic Chemistry (GTFCh) guidelines,7 which are based on international guidelines for bioanalytical method validation. For calculation of validation parameters, Valistat (Arvecon GmbH, Walldorf, Germany) software (version 2.0) was used.8
2.6.1 Linearity
Calibration curves were constructed using six calibrators. Linearity of calibration and variance homogeneity were checked by regression analysis of the instrument response (counts per seconds [c/s], proportionally adjusted to the relative internal standard response) to the calibrator concentrations and by performance of a Cochran test and a Mandel F test.
2.6.2 Precision and accuracy
To evaluate the accuracy, between-day, and within-day precision of the method, quality controls were analyzed in duplicate per levels I–III on eight different days and were quantified using the six-point calibration curve. Precision was calculated by the relative standard deviation (RSD, %), accuracy by the bias (%), and the 95% tolerance intervals that were determined.
2.6.3 Limits of detection and lower limits of quantification
The limits of detection (LoD) and lower limits of quantification (LLoQ) were determined by the DIN32645 procedure.7 Six blood concentrations around the expected limits were prepared by diluting reconstituted calibrator VI with ultrapure water. Obtained concentrations were As, 0.085, 0.156, 0.227, 0.298, 0.369, and 0.44 μg/L; Pb, 0.085, 0.222, 0.290, 0.358, and 0.426 μg/L; Tl, 0.011, 0.021, 0.032, 0.042, 0.053, and 0.063 μg/L. Samples were analyzed, and the linearity of these low-concentration calibration curves, as well as the presence of stragglers, tested on a 99% significance level.
2.6.4 Carryover
Sample residues within the ICP-MS system and the equipment may cause carryover of analytes in analyses of subsequent specimens. Sample carryover, expressed as % carryover, was therefore evaluated by calculating the signal response ratios of blank samples and preceding level VI calibrators (n = 6).
2.6.5 Evaluation of single-point calibrations
Between- and within-day datasets were reanalyzed by omitting calibrator levels I–V, to evaluate the validity of concentration results derived on the basis of single-point calibrations, using the blank and the certified whole blood calibrator (ClinCal®) only. Accuracy and precision results obtained in this manner were evaluated and compared with those obtained by using six-point calibrations.
2.6.6 Interlaboratory comparison
Assessment of the method's external validity was carried out by means of an interlaboratory comparison, performed with the Medical Laboratory Bremen (Bremen, Germany) using two lots of sheep blood (quality control low and high). Each lot was aliquoted into 7 samples and provided by the Laboratory in Bremen. In Bremen, samples were measured using a previously described ICP-MS method based on alkaline workup.9 As, Pb, and Tl concentration results were compared with those determined by the herein described method.
2.6.7 Application to forensic samples
The applicability of the current method to postmortem femoral blood samples was investigated by analyzing 279 forensic blood samples provided by the Institute for Legal Medicine in Cologne. Descriptive statistic analysis was performed using Prism version 10.1.0 (GraphPad Software, LLC, Boston, USA).
3 RESULTS
3.1 Method validation
3.1.1 Calibration range and linearity
No stragglers were observed at the 95% and 99% significance level. Results of the Cochrane test indicated homogeneity of variance at the 99% significance level across the entire calibration range, for all datasets. Mandel F test verified that the linear regression model could be used for generation of the As calibration curve. For Pb and Tl calibration, Mandel F test was passed from 68.2 to 341 μg/L and from 2.10 to 10.5 μg/L, indicating that a quadratic regression model would be better than linear regression. However, using linear regression, six-point calibration curves (blank and calibrators I to VI) showed very good coefficients of correlation (> 0.997), for all analytes, and precisions and accuracies obtained for all controls were very good (Table 2). Therefore, linear regression was considered sufficient and thus used for quantitative analyses across the entire calibration ranges, for all analytes.
| ClinCheck® mean value (control range) | Six-point calibration | One-point calibration | ||||
|---|---|---|---|---|---|---|
| Within-day precision (RSD, %) | Between-day precision (RSD, %) | Accuracy (bias, %) | Within-day precision (RSD, %) | Between-day precision (RSD, %) | Accuracy (bias, %) | |
| As (μg/L) | ||||||
| Level I 2.99 (2.39–3.59) | 3.0 | 3.0 | −4.8 | 3.0 | 3.4 | −6.5 |
| Level II 5.37 (4.29–6.44) | 3.7 | 3.7 | −1,5 | 3.8 | 4.4 | −3.3 |
| Level III 10.2 (8.14–12.2) | 3.1 | 4.9 | −1.1 | 3.2 | 6.1 | −2.9 |
| Pb (μg/L) | ||||||
| Level I 2.99 (2.39–3.59) | 2.8 | 2.8 | 2.7 | 2.3 | 3.1 | 6.3 |
| Level II 5.37 (4.29–6.44) | 3.2 | 3.2 | 0.0 | 3.1 | 3.4 | 7.4 |
| Level III 10.2 (8.14–12.2) | 3.7 | 4.3 | −2.7 | 2.9 | 4.3 | 5.2 |
| Tl (μg/L) | ||||||
| Level I 2.99 (2.39–3.59) | 3.1 | 3.1 | 4.6 | 3.1 | 3.1 | 6.3 |
| Level II 5.37 (4.29–6.44) | 3.0 | 3.0 | 5.8 | 3.1 | 3.4 | 7.4 |
| Level III 10.2 (8.14–12.2) | 2.8 | 3.0 | 3.5 | 2.9 | 4.3 | 5.2 |
- Imprecisions of the method for determining arsenic (As), lead (Pb), and thallium (Tl), determined by using a six-point calibration (left) or a single-point calibration (right).
3.1.2 Precision and accuracy
All tests were passed. Overall, analytical imprecision was < 5% (six-point calibration). As, Pb, and Tl concentrations determined in quality controls (levels I–III) were close to the specified concentrations and remained well within the control ranges indicated by the manufacturer (Table 2).
3.1.3 LoD and LLoQ
The low calibration curves for As, Pb, and Tl passed the test for linearity, and no stragglers were observed. LoD for As, Pb, and Tl was 0.08, 0.18, and 0.01 μg/L, respectively. LLoQ for As, Pb, and Tl was 0.23, 0.66, and 0.03 μg/L, respectively.
3.1.4 Carryover
No significant carryover was observed. Mean carryover for As, Pb, and Tl was 0.15, 0.12, and 0.01%, respectively.
3.1.5 Single-point calibration
All tests for precision and accuracy were passed. No relevant differences were found compared with results determined by applying a six-point calibration. Analytical imprecision was from −6.5 to 7.4% (Table 2).
3.1.6 Between-laboratory comparison
Overall, concentration results for As, Pb, and Tl determined in Bremen and Cologne were in good agreement. In pool 1 (containing low concentrations of As, Pb, and Tl), differences of laboratory results were from −12.2 to −15.5%, and in pool 2 (containing higher concentrations of As, Pb, and Tl), differences were from −1.1 to 4.1% (Table 3).
| Medical LaboratoryBremen | Institute for Clinical Chemistry, University Hospital Cologne | |||||
|---|---|---|---|---|---|---|
| As (μg/L) | Pb (μg/L) | Tl (μg/L) | As (μg/L) | Pb (μg/L) | Tl (μg/L) | |
| Quality control low | 2.4 | 9.8 | 2.0 | 2.1 | 8.3 | 1.7 |
| Quality control high | 10.6 | 67.9 | 9.5 | 10.5 | 65.1 | 9.3 |
- Interlaboratory comparison results for arsenic (As), lead (Pb), and thallium (Tl) concentrations. Displayed are the target values determined in the Medical Laboratory Bremen (left) and the mean concentrations of seven measurements, conducted in Cologne (right).
3.2 Application to forensic samples
Elemental concentrations of As, Pb, and Tl were determined in 279 independent postmortem femoral blood samples. Blood concentrations of As, Pb, and Tl were 1.31 ± 3.42 μg/L, 17.4 ± 13.1 μg/L, and 0.11 ± 0.07 μg/L (mean ± standard deviation [SD]), respectively (Figure 1). Descriptive statistics analysis data are shown in Table 4.
| As (μg/L) | Pb (μg/L) | Tl (μg/L) | |
|---|---|---|---|
| Minimum | 0.07 | 1.6 | 0.02 |
| 25% Percentile | 0.22 | 8.9 | 0.069 |
| Median | 0.49 | 14 | 0.1 |
| 75% Percentile | 1.1 | 21 | 0.14 |
| Maximum | 46 | 106 | 0.59 |
| Range | 46 | 104 | 0.57 |
| 1% Percentile | 0.094 | 2.5 | 0.024 |
| 99% Percentile | 15 | 68 | 0.37 |
| Mean | 1.3 | 17 | 0.11 |
| Std. deviation | 3.4 | 13 | 0.071 |
| Std. error of mean | 0.2 | 0.79 | 0.0042 |
- Statistical analysis data for As, Pb, and Tl concentrations obtained from 279 independent post-mortem femoral blood specimens.
4 DISCUSSION
4.1 Choice of method
Postmortem blood matrices differ significantly from the whole blood and serum of living persons, which are normally used for the analysis of heavy metals in a medical laboratory. Therefore, the existing method at the Institute of Clinical Chemistry had to be adapted to the postmortem blood sample matrix. The initial strategy was to adopt the method published by Heitland and Köster.9 In this method, alkaline processing is used for sample preparation, for which 200 μL of blood is mixed with a solution containing 0.02% octylphenol ethoxylate (Triton-X-100) and internal standards. Next, 150 μL of a 20% (v/v) ammonia solution is added, and finally, the solution filled up to 5 mL. According to the authors, this avoids or minimizes the precipitation of proteins, which is particularly advantageous for whole blood samples. Initial experiments with this preanalytical method yielded very clear extracts, even with postmortem whole blood samples. However, acidic sample preparation was already established at the Institute of Clinical Chemistry, and the measurement of alkaline forensic samples was not compatible with the existing validated routine analysis. Such an approach would necessitate a switch from acidic to alkaline environment on each run, with unforeseeable consequences to the validated clinical method and routine laboratory diagnostics, for which conformity with regulation (EU) 2017/746 on in vitro diagnostic medical devices has been established.10 In addition, as Triton-X-100 is included in the list of substances of concern under this regulatory framework,11 a slight modification was carried out by replacing it with tergitol as an alternative detergent.
For these reasons, we switched to acidic sample processing, with a recent publication on postmortem whole blood lending itself to being adapted for this purpose.12 However, the described sample processing was carried out under extreme conditions (digestion with hydrofluoric acid and nitric acid in polytetrafluoroethylene tubes), which we did not want to introduce into the laboratory, for health and safety reasons.
A less aggressive acidic preparation, already in use for serum and plasma samples at the Institute of Clinical Chemistry, was therefore adapted. Small changes were made to the preparation, as the corpuscular components of the whole blood had to be broken prior to analysis. In a first step, nitric acid (10%) was added to the specimens instead of nitric acid (1%). In addition, ultrasound treatment was used, and the sample was then further diluted. Moreover, less sample volume was used (200 instead of 250 μL), and the digested suspension was centrifuged. As a result, sufficiently clear and measurable extracts were obtained, without the need to change the general method parameters or the equipment setup when switching from postmortem blood to serum or plasma matrices and vice versa.
4.2 Choice of analytes
To investigate the basic suitability of the method for postmortem samples, we first focused on the elements arsenic, lead, and thallium. Mercury and antimony were also considered relevant from a toxicological point of view. However, because no mean values were provided in the commercial whole blood calibrator, and routine quantification of mercury is affected by pronounced memory effects,13 these elements were omitted in this first step. In the future, based on the current approach, further method development is planned by successive inclusion of other elements of interest, such as antimony, cadmium, and mercury in order to establish a qualitative screening method for a broad panel of toxic elements.
4.3 Validation
The method has been validated according to guidelines of the GTFCh.7 All specifications could be met. With respect to physical properties, only stable metal isotopes were selected.14-16 As the recommended storage time of the whole blood calibrator after reconstitution (30 days when stored below −18°C) was not exceed, stability testing was omitted from the validation procedure. The analytes are stable over the aforementioned time period and storage conditions, as specified by the manufacturer. All samples were measured in a timely manner after processing. Spike-and-recovery experiments were not performed as no analyte was either spiked to a diluent for the creation of calibrators, nor to the sample matrix (whole blood). Instead, the reconstituted whole blood calibrator was diluted by using ultrapure water. Selectivity was assured by control of significant spectral interferences for 75As, such as 40Ar35Cl+ and 40Ca35Cl+, by utilizing the iCRC mode for quantification of arsenic.17 Because, in the natural state, the abundance of lead isotopes is not uniform, the sum of the isotopes 206Pb, 207Pb, and 208Pb was used for quantification of lead (Table 1).17 Nonspectral interference from the matrix, which could interfere with thallium quantification, was minimized by dilution of the sample matrix.18 The interlaboratory comparison using pooled sheep blood demonstrated the external validity of the established method (Table 3).
4.4 Single-point calibration
For the validation of the quantitative method, a CE-certified calibrator from Recipe (ClinCal®, Ref 9943, Lot 2501) was used. After reconstitution with ultrapure water, it was diluted accordingly in order to obtain the additional five calibration points, as at least five calibration points are required for forensic analysis.7 The linearity, the correctness, and the precision achieved with this approach were very satisfactory (Table 2). However, the procedure did not correspond to the goal of devising a rapid and low-threshold screening analysis method. Consequently, we tested whether a high correctness and precision were also achieved when measuring with only the highest (undiluted) calibration point and a blank sample. This latter approach was demonstrated during validation (Table 2), confirming that the commercially available whole blood calibrator, on its own, is suitable for rapid routine analysis.
4.5 Results of measurements
In total, postmortem blood samples from 279 individuals were retrospectively analyzed (Figure 1, Table 4). For comparison, the study by Söderberg et al. may serve as a reference, in which As, Pb, and Tl concentrations were reported in postmortem blood specimens from 120 cases of death by hanging.12 The median As concentration observed in our study was 0.49 μg/L, which is markedly lower than the median reported in the Söderberg study (2.25 μg/L). In the Söderberg study as well as in our study, total arsenic (i.e., all arsenic species) was quantified. A factor potentially contributing to this difference may be related to dietary intake of fish and seafood, which is found more frequently in the Swedish population.19 In the bloodstream, arsenic is distributed between the plasma and the erythrocytes, where it is bound to the globin of hemoglobin.20 It is possible that, due to the less aggressive acidic digestion used in our method, disruption of erythrocytes and hemoglobin is less efficient when compared with standard protocols. In contrast, median Pb and Tl concentrations observed in the current study (14.0 and 0.1 μg/L) were comparable with those reported by Söderberg et al. (15.8 μg/L and 0.12 μg/L). Likewise, most of the lead found in erythrocytes is bound to proteins, the primary binding ligand being delta-aminolevulinic acid dehydratase.21 All postmortem blood samples were analyzed following storage times of up to 11 months at 4°C, following parameters concluded by Tanvir et al., in their study on obtaining reliable As, Tl, and Pb test results from stored blood specimens.22
4.6 Limitations
Some limitations of our study have to be considered. The commercial whole blood calibrator contains elemental concentrations well below the respective toxic concentrations,23 as it is designed to quantify concentrations of metals in the normal population. Blood concentrations above the calibrator VI could only be extrapolated. For an exact quantitative determination of intoxications, dilution of the sample is required. Further, disruption of erythrocytes may be less efficient due to the milder acidic digestion applied during sample preparation. Therefore, elements distributed into erythrocytes may be found at lower concentrations. However, our data indicates that the method is sufficient for the detection of As, Pb, and Tl intoxications in postmortem blood. Furthermore, the current approach for sample processing is focused on practicability and not on the purity of the sample solution. High sample throughput may result in significant contamination of the sample introduction system, plasma torch, and the interface region. Therefore, regular maintenance of the ICP-MS system is required when measuring larger sample quantities. It can be assumed that the use of sample filters (for example, Amicon® Ultra Filter [50 kDa], Merck, Darmstadt, Germany, or Vivaspin® Centrifugal Concentrators, Sartorius, Göttingen, Germany) in a final purification step would lead to even cleaner extracts, which would further reduce pollution in the system.
5 CONCLUSION
We here established a practicable and sufficiently accurate ICP-MS method for the determination of As, Pb, and Tl in whole blood which is compatible with routine ICP-MS diagnostics in an accredited medical laboratory. The current method and approach serve as a prelude to implementing low-threshold investigations for metal intoxications in forensic routine.
ACKNOWLEDGMENTS
We kindly thank Dr. June Mercer-Chalmers-Bender for thoughtful and professional English revision and corrections. We kindly thank Dr. Peter Heitland (Medizinisches Labor Bremen) for providing sheep blood samples and for participating in the between-laboratory comparison. We thank Dr. Franz Lehmann and Dr. Sebastian Faßbender (Analytik Jena) for kind and competent technical and scientific support. No funding was received for conducting this study.
CONFLICT OF INTEREST STATEMENT
The authors have no relevant conflicts of interest to disclose.
Open Research
DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy, ethical, or legal restrictions.
