Serum calcium level is an important clinical index that can reflect pathophysiological states. Along with phosphorus level, serum calcium is associated with coagulation, maintenance of neuromuscular irritability, and bone turnover and is a predictor of cardiovascular disease risk . Moreover, calcium levels were found to be lower in post-menopausal than in pre-menopausal women . Despite its widespread use as a physiological indicator, the accuracy of serum calcium detection is far from perfect.
According to data from the National Electrolyte Trueness Verification Program launched by the National Center for Clinical Laboratory in China (NCCL), pass rates for low and high concentrations were 46% and 57%, respectively, among 137 laboratories in 2014 and 52% and 57%, respectively, among 186 laboratories in 2015, with 2% tolerable bias in WS/T 403-2012. The intra-laboratory coefficient of variation (CV) was small, and pass rates were all >90% based on an evaluation criterion of 2% tolerable imprecision in WS/T 403-2012. It is worth noting that calcium had the least accurate pass rate in tests of four electrolytes, which also included sodium, potassium, and magnesium. Therefore, establishing a suitable high accuracy method is needed for serum calcium measurement in China.
Isotope dilution mass spectrometry (IDMS) method is widely recognized as one of primary methods with high accuracy. To this end, we established an isotope dilution inductively coupled plasma mass spectrometry (ID ICP-MS) method for serum calcium measurement in China, and it will used for the establishment and assurance of measurement traceability through calibrating, evaluating routine methods and assigning values to target accuracy-based EQA project or calibrations.
Materials and methods
Reagent and serum samples
Nitric acid (69%, high purity) was obtained from Thermo Fisher Scientific (Waltham, MA, USA) and prepared as a dilute nitric acid (3% w/w) solution. The 42Ca isotope (42CaCO3, 97.8% 42Ca) was purchased from Cambridge Isotope Laboratories (Tewksbury, MA, USA), with certified abundances of 40Ca (1.75%), 42Ca (97.8%), 43Ca (0.03%), 44Ca (0.37%), 46Ca (<0.01%), and 48Ca (0.01%). First and second isotopic spike solutions were prepared by dilution with nitric acid using a gravimetric approach. Standard reference material (SRM) 3109a from the National Institute of Standards and Technology (NIST) was used for calibration (mass concentration: 9.819±0.019 mg/g). Serum samples for electrolyte trueness verification were prepared by China NCCL in 2014 and 2015 (lot numbers T201411, and T201512). SRM909bI, SRM909bII (lyophilized serum, stored at 4°C), and SRM909c (frozen serum, stored at −80°C) from NIST, and GBW09152 (reference material for determining inorganic composition of frozen serum) from the National Institute of Metrology, China, were used for accuracy assessment. Nine serum and seven aqueous samples were used in the reference method comparison program launched by China NCCL.
An Agilent 700X ICP mass spectrometer (Agilent Technologies, Santa Clara, CA, USA) was used in this study; the operation parameters are listed in Table 1. Solutions were prepared by the gravimetric method using a XS204 balance (max=220 g, d=10−4 g; Mettler Toledo, Giesen, Germany). Ultrapure water (>18.2 MΩ·cm) was obtained using a Milli-Q water purification system (Millipore, Billerica, MA, USA). A temperature-controlled electric hot plate (Shanghai New Extension Microwave Dissolution Testing Technology Co., Shanghai, China) was used to digest serum. A calibrated DMA 4500M densitometer (Anton Paar, Houston, TX, USA) was used for density measurement.
Preparation of calibration solution
Standard storage solution was prepared by dissolving 0.5 g SRM3109a in 3% nitric acid to obtain a 50-g solution, of which 3 g were diluted 12.5-fold to obtain the standard test solution, which was used in the reverse IDMS procedure. The first isotopic spike storage solution (13 μg/g) was prepared by dissolving 1 mg 42CaCO3 in 2-g high-purity nitric acid and diluting to 30 g with ultrapure water. On the day of measurement, 2 g of the first isotopic spike storage solution was diluted with ultrapure water to 20 g to obtain the second isotopic spike storage solution (1.3 μg/g). A predetermined mass of standard storage solution was added along with 0.5 g of the second isotopic spike storage solution to obtain the spiked calibration solution; this was diluted 80-fold with 3% nitric acid, in which 44Ca/42Ca was about 1.
Preparation of serum
Serum reference materials and samples were combined with the second spiked storage solution (containing 1.3 μg/g 42Ca) to obtain a 1:1 isotopic ratio, while serum concentration had been measured roughly by routine method before. The isotope ratio difference between the spiked sample and calibration solutions was within 5% . The serum was digested with 2.3 g of 69% nitric acid at 140°C on an electric hot plate for 2 h. Samples were then diluted 80-fold with ultrapure water after confirming that dilution multiples had little effect on the detection of isotopic ratio (Tables 2 and 3).
Preparation of blank
The blank solution was prepared three times using ultrapure water instead of serum and spiked solution and measured on 3 consecutive days. The background of the blank was low, i.e. 0.02%–0.09% and 0.08%–0.12% for the 44Ca- and 42Ca-spiked sample solution, respectively.
Sample concentration was calculated based on the 44Ca/42Ca isotope ratio. Spike concentration and the associated mass discrimination correction factor were determined by reverse IDMS. We then bracketed the spiked sample with a spiked calibration solution, and sample concentration was calculated based on the principles of IDMS. The detection sequence was as follows: (1) standard test solution→spike calibration solution; (2) spike calibration→two or three samples→spike calibration. The two steps were repeated over three rounds, and each step was repeated three times. The means of three rounds measurement was taken as the final concentration.
Comparison of methodologies
The International Federation of Clinical Chemistry and Laboratory Medicine and the German Society for Clinical Chemistry and Laboratory Medicine jointly launched the Ring Trial program in 2013. NCCL aimed to establish China’s own reference system of serum calcium measurement, and launched a reference method comparison program in China to lay the foundation of further work like reference material joint valuation. Some laboratories of the country which have been accredited to ISO/IEC 17025 and ISO 15195 established their own reference method. The Beijing Institute of Medical Device Testing (BIMDT) established a dynamic reaction cell and two-way ID ICP-MS method, and Ningbo Medical System Biotechnology Co. (Ningbo, China) has used ion chromatography to accurately determine total calcium concentration in human serum. NCCL also employed an ICP-MS-based method for detecting serum calcium  using aluminum as an internal standard. In the present study, we compared the performance of our IDMS method with those methods which have high-grade recommended measurement principle to explore their comparability.
Internal standard ICP-MS
This method, which is based on an internal standard principle, used the same instrument as our IDMS method (i.e. an Agilent 700X ICP mass spectrometer); H2 gas was used to reduce potential interference by 44Ca. Since aluminum is present at very low levels in normal human serum (about 0.0018 μg/mL) , serum samples were spiked gravimetrically with aluminum as an internal standard, digested with 69% ultrapure nitric acid, and diluted to the test concentration; 44Ca/27Al ratios were measured by ICP-MS in hydrogen mode. Samples were detected by the bracketing method.
Dynamic reaction cell and two-way ID ICP-MS
BIMDT also employed an ID ICP-MS method  (using an Elan DRC-e ICP mass spectrometer; Perkin-Elmer, Waltham, MA, USA), with 42Ca used for isotopic spiking. SRM915b was the standard; the spike isotope 42CaCO3 was obtained from Oak Ridge National Laboratory. Methane was used as the reaction gas at a rate of 0.9 mL/min. Sample 44Ca/42Ca ratios were measured and used to calculate serum calcium concentrations. The masses of the standard and serum sample were calculated in order to confirm that the isotopic ratio of 44Ca/42Ca was close to 1. Compared to standard isotope dilution, this method does not include an isotope spike calibration step (reverse IDMS); thus, serum calcium concentration was calculated solely based on the isotope ratio.
Ningbo Medical System Biotechnology Co. developed an ion chromatography method for detecting serum electrolytes  for which valuation transfer and traceability are lower and higher than those of ICP-MS and other routine methods, respectively. A simple sample treatment procedure (wet digestion) was used to remove proteins in human serum; an IonPac CS-16 analytical column to separate serum cations in a Dionex ICS 1100 instrument. Sample concentration was determined using an external standard based on peak area.
Evaluation of imprecision
We measured serum samples prepared for electrolyte trueness verification by NCCL in 2014 and 2015 (T1411 and T1512) with different concentrations. Three vials of each sample were measured every day for 3 days (Tables 5 and 6). Within-run variation, between-run variation, and intra-laboratory CV were 0.12%–0.19%, 0.07–0.09%, and 0.16%–0.17%, respectively.
Evaluation of accuracy
SRM909bI and SRM909bII (lyophilized serum stored at 4°C), SRM909c (frozen serum stored at −80°C) and GBW09152 were tested. Duplicate vials of each reference material were measured for three consecutive days (Table 7). The relative bias in our method was −0.02% to 0.29%, which was superior to that in ICP-sector field MS (−1.1% to 0.28%) .
Evaluation of recovery
Serum samples T1411 and T1512 were tested before and after adding known amounts of calcium gravimetrically which was equivalent to sample calcium; the recovery rate was calculated according to the change in calcium concentration. Triplicate vials of each sample were measured each day for 2 days. The recovery using our method was 99.98%–100.18% (Tables 8 and 9).
Assessment of uncertainty
Each operation during the procedure can increase uncertainty, especially weighing and detection steps (Table 10). Based on these sources of uncertainty, related standard uncertainty of sample concentration was μc,rel(C)=0.207% (expanded uncertainty, k=2). Consequently, related expanded standard uncertainty of sample concentration was 0.414%, which is lower than those of high-resolution (HR-) ICP-MS and collision-cell ICP-MS [9, 10].
Comparison of different methodologies
Internal standard ID ICP-MS
A total of 46 different serum samples were analyzed by both the internal standard method and ID ICP-MS. The absolute and relative biases were −0.46 to 0.38 μg/g and −0.45% to 0.49%, respectively. The scatter plot of ID ICP-MS results vs. relative bias indicated that the bias had a homogeneous distribution (Figure 1A). The mean of all biases was 0.01%. These results suggest that there is no systematic error associated with ID ICP-MS. A scatter plot of ID ICP-MS results vs. internal standard ICP-MS results showed that in the range of concentrations of the 46 serum samples, the internal standard method was linearly related to the ID ICP-MS method (Figure 1B); the equation of the line was y=0.99886x+0.11173, with a slope of 0.99886 (close to 1) and intercept of 0.11173 (close to 0); and the correlation coefficient was 0.99992. These results indicate that results obtained by the two methods are highly correlated.
Methods in other laboratories
NCCL launched a reference method comparison program in China; 16 samples were tested in various laboratories in 2015 and 2016, including nine serum and seven aqueous samples. Of the participating laboratories, BIMDT used the dynamic reaction cell method and two-way ID ICP-MS (Elan DRC-e) for sample measurements, while Ningbo Medical System Biotechnology Co. used ion chromatography. We compared all of these data with ours obtained by ID ICP-MS (Figure 2). The correlation coefficients were 0.99643 and 0.98930, respectively; slopes were 0.96974 and 0.92096, respectively; and intercepts were 0.04452 and 0.13809, respectively. This indicates that our ID ICP-MS method yields results that are consistent with those obtained by other reference methods.
We also compared results for aqueous samples, for which known preparation concentrations were taken as target values (Figure 3). IW201501–IW201504 were aqueous solutions prepared with SRM 3109a by the gravimetric method. Bias was determined to be −0.55% to 0.12% for ID ICP-MS as compared to −1.26% to 2.28% for Ningbo Medical System Biotechnology Co. and −3.85% to 2.14% for BIMDT; the standard deviations for ID ICP-MS were obviously smaller than BIMTD method and concentrations were closer to preparation values.
IDMS is widely recognized as a suitable reference method for clinical measurements which is considered as highly accurate for quantitative analyses. Some international laboratories have established IDMS reference methods to detect serum calcium level. Sources of interference include isobar, polyatomic ions (oxide, chloride, hydride, and argon compounds), and doubly charged ions. Three targeted technologies are used: cold plasma, collision cell, and high resolution detector technology. NIST  used cold plasma technology to reduce molecular ion interference of the Ar matrix (40Ar+) arising from plasma to below 1000 counts per second (CPS); molecular ion interference arising from serum matrices was eliminated by cation exchange chromatography. Octopole collision cell ICP-MS  used hydrogen reaction gas to resolve polyatomic interference of the Ar matrix, which was decreased by three orders of magnitude using a hydrogen collision cell . However, a blank correction is needed due to the high background signal (0.27%–0.44%). In high-resolution ICP-MS (HR-ICP-MS) , detector resolution is higher than 4500 CPS and can distinguish polyatomic ion interference from 44Ca, 42Ca, and 43Ca, although it cannot distinguish 40Ar+ interference of 40Ca from the Ar matrix. Thus, these methods have good performance but still have the problem of high matrix interference, background signal, high cost or complex operation.
In this study, we developed an accurate ID ICP-MS method for measuring serum calcium concentration. Doubly charged and polyatomic ions can be eliminated by shielded torch using the Agilent 7700X ICP-MS system, with an octopole collision cell in hydrogen mode. 40Ca is blocked by the intense background peaks of 40Ar sourced from plasma gas, which cannot be avoided. Therefore 44Ca/42Ca was detected as isotope pairs. Samples were diluted based on investigations of background signal and non-mass spectrum interference. This involved a simple serum digestion process that is more thorough than direct acid digestion , easier than microwave digestion , and does not require cation exchange chromatography  to estimate serum matrix interference. Moreover, it employs a single calibration solution for the whole measurement as well as a simplified injection scheme relative to internal standard ICP-MS. Although the preparation and calculations are somewhat complicated, a single test can be completed within 20 s, which is far less than the time required for ion chromatography. High accuracy (99.8%–100.29%), low imprecision (<0.20%) and low uncertainty (0.414%) qualify this method as a candidate reference method for serum calcium detection.
In addition, the ID ICP-MS method shows good comparability with three other reference methods in China (internal standard ICP-MS, ion chromatography, and two-way ID ICP-MS) and has better performance. In further study, this method can be used to establish and improve serum calcium reference system in China, including assessing routine method, assigning target value to EQA samples and China’s own reference materials.
Regmi P, Malla B, Gyawali P, Sigdel M, Shrestha R, Shah DS, et al. Product of serum calcium and phosphorus (Ca×PO4) as predictor of cardiovascular disease risk in predialysis patients. Clin Biochem 2014;47:77–81. Web of ScienceCrossrefGoogle Scholar
Bhattarai T, Bhattacharya K, Chaudhuri P, Sengupta P. Correlation of common biochemical markers for bone turnover, serum calcium, and alkaline phosphatase in post-menopausal women. Malays J Med Sci 2014;21:58–61. PubMedGoogle Scholar
Sargent M, Harte R, Harrington C. Guidelines for achieving high accuracy in isotope dilution mass spectrometry (IDMS). RSC Analytical Methods Committee, 2002. Google Scholar
Yan Y, Ge M, Ma R, Zhao H, Wang D, Hu C, et al. A candidate reference method for serum calcium measurement by inductively coupled plasma mass spectrometry. Clin Chim Acta 2016;461:141–5. CrossrefWeb of SciencePubMedGoogle Scholar
Li S, Wang J. Measurement of Calcium in Human Serum by Dynamic Reaction Cell and Two-way ID–ICP–MS. Chemical Analysis and Meterage, Vol. 24, No. 5, Sept. 2015. Google Scholar
Zou B, Zou J, Shen M, Zhang M, Wu L, Tu M. Establishment of reference measurement procedures for serum electrolytes based on ion chromatography. Lab Med 2015;30:1250–56. Google Scholar
Kramer U, Kress M, Reinauer H, Spannagl M, Kaiser P. Candidate reference measurement procedures for chloride, potassium, sodium, calcium, magnesium, and lithium by inductively coupled plasma (isotope dilution) sector field mass spectrometry (ICP-(ID) SFMS) in serum. Clin Lab 2013;59:1017–29. PubMedWeb of ScienceGoogle Scholar
Stürup S, Hansen M, Mølgaard C. Measurement of 44Ca:43Ca and 42Ca:43Ca isotopic ratios in urine using high resolution inductively coupled plasma mass spectrometry. J Anal At Spectrom 1997;12:919–923. CrossrefGoogle Scholar
Simpson LA, Hearn R, Merson S, Catterick T. A comparison of double-focusing sector field ICP-MS, ICP-OES and octopole collision cell ICP-MS for the high-accuracy determination of calcium in human serum. Talanta 2005;65:900–6. PubMedCrossrefGoogle Scholar
Murphy KE, Long SE. The accurate determination of potassium and calcium using isotope dilution inductively coupled ‘cold’ plasma mass spectrometry. J Anal At Spectrom 2002;17:469–77. CrossrefGoogle Scholar
Röhker K, Rienitz O, Schiel D. Ion chromatographic precision measurement procedure for electrolytes in human serum: validation with the aid of primary measurement procedures. Accredit Qual Assur 2004;9:671–7. CrossrefGoogle Scholar
About the article
Published Online: 2017-07-12
Published in Print: 2017-11-27
Author contributions: All the authors have accepted responsibility for the entire content of this submitted manuscript and approved submission.
Research funding: National High Technology Research and Development Program of China: 2011AA01A102, 2011AA02A116.
Employment or leadership: None declared.
Honorarium: None declared.
Competing interests: The funding organization(s) played no role in the study design; in the collection, analysis, and interpretation of data; in the writing of the report; or in the decision to submit the report for publication.