Rapid, Cost-Effective, and Routine Biomedical Analysis Using ICP-MS

November 2, 2005
Bill Spence, Martin Nash, Simon M. Nelms, Phil Shaw
Special Issues

Volume 0, Issue 0

Elemental analysis in biological samples generally is achieved using flame atomic absorption spectrometry (AAS) and graphite furnace AAS (GFAAS). Flame AAS is fast, easy-to-use, and economical, but insufficiently sensitive for assays such as Se in serum and Pb/Cd in whole blood. These measurements require use of the more sensitive GFAAS. Inductively coupled plasma-mass spectrometry (ICP-MS), despite its low detection limit capabilities and wide elemental range, has had relatively little impact to date on biomedical analysis because of the popularly held conception that it is complex to use and expensive. In recent years, the instrumentation has been simplified and purchase, running, and maintenance costs have fallen. As a result, clinicians are becoming more interested in ICP-MS, although the perception that it is still much more expensive than GFAAS remains. This article provides a comparison of the costs of ICP-MS and GFAAS for biomedical sample analysis and illustrates the performance of ICP-MS for..

Cost Analysis

To allow a direct comparison between inductively coupled plasma-mass spectrometry (ICP-MS) and graphite furnace atomic absorption spectrometry (GFAAS), this cost analysis focuses on the measurement of six key elements of biomedical interest (Al, Cu, Zn, Se, Cd, and Pb) in each of the samples supplied to the laboratory, in the first year of operation following purchase. The data used incorporate purchase, running, and maintenance costs are based upon current values. Figure 1 shows how the cost per sample for ICP-MS and GFAAS varies with increasing sample throughput. It shows that the cost per sample at low sample throughput for one quadrupole ICP-MS is around three times that of one GFAAS and also illustrates that ICP-MS, because of its higher productivity, becomes the more cost-effective technique when three or more GFAAS instruments are required to meet the workload. On the other hand, where less than three furnace systems are required, this technology is more cost effective than ICP-MS.

Table I. Pb in blood analysis results (data in ppb, dilution corrected)

Example Applications

Each of the following example applications were carried out using an XSeriesII ICP-MS system (Thermo Electron Corporation, Winsford, Cheshire, UK).

Figure 1. Cost per sample versus sample throughput for graphite furnace AA and ICP-MS (first year of operation, six elements per sample).

Rapid analysis of Pb in whole blood. This analysis is important because Pb exposure leads to cognitive function problems, especially in children. Despite the disbanding of lead-based paints in recent years, young children living in older homes in particular remain vulnerable to ingestion of Pb through chewing items coated with such paints. In addition, occupational exposure must be monitored for those who are employed in foundries or smelting works, for example. The levels that need to be quantified typically are down to 50–100 ppb Pb in the undiluted sample. In low sample-throughput laboratories, this analysis generally is performed using GFAAS instrumentation, because for smaller workloads, this methodology is more cost-effective to use. For laboratories processing upwards of 100 samples/day, ICP-MS becomes more attractive.

Table II. Water and dialysis fluid results (data in ppb, dilution corrected)

For this study, Seronorm (Sero, Billingstad, Norway) and Bio-Rad (Hercules, CA) blood-certified reference materials were measured together with two U.K. NEQAS (National External Quality Assessment Scheme) reference samples. All the samples were diluted 1:50 with 0.1% ammonia, 0.01% Triton X-100 solution and Rh (10 ppb) was used as the internal standard. Standard addition calibration was used based on an initial 1% HNO3 blank, followed by a low Pb level blood zero addition standard, then subsequent spikes of this blood standard at 100, 200, 400, 600, and 800 ppb Pb. A Burgener Ari Mist nebulizer (Bergener Research, Mississauga, ON, Canada) and standard injector diameter torch were used. An autosampler "probe to wash early" function was applied to maximize sample throughput. The total analysis time (uptake/wash and three repeats/sample) was equal to 1.2 min, corresponding to a throughput of 51 samples/h. The analysis results are shown in Table I.

Table III. Serum sample results (data in ppb, dilution corrected)

The results in Table I show good agreement between the measured and expected Pb values. It is noteworthy that the expected values cover a wide range, which highlights the existing accuracy problems with this assay. Most of the results in the expected data set were derived from furnace methods and in many cases, might have been affected by contamination problems.

Measurement of Al in clinical samples. Determination of Al is important because chronic exposure to elevated levels of this element is known to cause dementia, memory loss, and impaired cognitive function (distinct from Alzheimer's disease). Individuals with kidney failure accumulate Al in the blood, so renal patients particularly are at risk. The types of samples that must be measured routinely are the water used to prepare the dialysis fluids, the dialysis fluids themselves, and patient's serum. Most clinical laboratories use GFAAS for this analysis because for the sample throughput required, this technology is sufficient. However, laboratories that have ICP-MS instrumentation already have incorporated ICP-MS technology successfully into their routine set of assays to help keep it cost effective.

Table IV. Spike recovery test results (data in ppb)

For this study, six water samples, five dialysis fluid solutions, two certified human sera materials (UTAK), and three bovine sera (measured in duplicate) were measured. The water and dialysis fluids were diluted 1:10 and the serum samples diluted 1:20 with 2% (v/v) HNO3 . Ga (5 ppb) was used as the internal standard. Calibration solutions were prepared at blank, 1, 5, 10, and 20 ppb Al. Spike recovery tests were performed on one specimen of each sample type, at a spike level of 5 ppb Al. The relative accuracy of the method was assessed using the UTAK reference sera.

In this work, a Burgener MiraMist nebulizer was used in conjunction with a standard injector diameter torch. The total analysis time (uptake, wash, and 3 repeats/sample) was 2.5 min. This included a qualitative survey scan across the full mass range. The Al calibration and figures of merit are shown in Figure 2 (parameters are shown in the table above Figure 2). The results of the water and dialysis fluid analysis are shown in Table II.

Good agreement was achieved between the consensus and measured results, and the data obtained were consistent with the known added amounts of Al. The serum sample results are shown in Table III. Excellent agreement was achieved between the consensus/reference data and the measured ICP-MS results. The spike recovery test results are presented in Table IV. Table IV shows that quantitative spike recovery results were obtained for all three sample types, illustrating the robustness of the method.

Figure 2. Al calibration and analytical figures of merit.

Measurement of Ag in serum samples. This analysis is important because Ag can enter the human body as a result of its use in some water treatments as a bactericide. It is recognized to have antibiotic properties and for this reason some salves and plasters containing Ag are used for various injuries including burns. For this analysis, only ICP-MS can routinely provide the required detection capability.

In terms of its deleterious effects on health, airborne Ag particles can cause nasal irritation and occupational exposure can affect kidney function. The Ag+ ion inhibits the function of certain enzymes, which has been shown to lead to Se depletion. Chronic Ag exposure causes "argyria" (a non-fatal but permanent blue/grey skin staining). This condition has been recognized in the media as a result of the well-known case of the U.S. politician, Stan Jones, who, in 2002, developed the condition after consuming considerable amounts of colloidal Ag that he was using as a health treatment.

Figure 3. Ag isotope calibrations and figures of merit.

In this study, two serum samples (Seronorm) were analyzed. The samples were diluted 1:20 with 1% (v/v) HNO3 and Rh internal standard (0.5 ppb) was added. A Burgener Ari Mist nebulizer and standard injector torch were used in conjunction with high-sensitivity interface cones fitted to the ICP-MS. Calibration solutions were prepared from blank to 0.1 ppb Ag. Spiked sera were prepared at 0.1, 0.2, 0.5, and 1.0 ppb before 1:20 sample dilution, to evaluate the spike recovery performance at different Ag levels. The total analysis time was 4 min (including uptake, wash, three repeats/sample and a survey scan from m/z 102 to 111). The Ag calibration and figures of merit are shown in Figure 3 (parameters are shown in the table above Figure 3). The analysis results are given in Table V.

The samples were diluted 1:20 for analysis, so the actual Ag concentrations measured in the spiked samples were only 5, 10, 25, and 50 ppt. Despite the very low level of Ag measured, Table V shows that excellent precision and spike recoveries were achieved.

Table V. Ag in serum analysis results (data in ppb)


This article has shown that ICP-MS is ideal for cost-effective, high sample-throughput biomedical analyses such as Pb determination in blood and Al in renal patient samples. It also provides the flexibility for multielemental analysis down to ultratrace levels. For lower sample throughput laboratories, with a limited range of elements at trace levels, GFAAS remains more cost-effective.