Liquid chromatography-tandem mass spectrometry (LC-MS/MS) is often considered a gold standard measurement technique due to its relatively high degree of specificity compared to other analytical approaches, such as immunoassay. However, use of an MS-based method does not guarantee accuracy. Like their immunoassay counterparts, LC-MS/MS assays are susceptible to measurement errors arising from interferences and matrix effects.
Sources of Error
One potential major source of measurement error is calibration bias. Because only one Food and Drug Administration-approved MS-based method is currently available, most methods employed in clinical laboratories are laboratory developed tests. As a result, laboratories are responsible for preparing or purchasing calibrators and ensuring that the calibrator value assignments are correct. Differences in calibrator matrix, preparation technique, value assignment, or fit of the calibration curve equation can cause calibration biases that result in clinically significant differences among methods. In fact, even though different LC-MS/MS methods for the same analytes vary widely in use of equipment, sample preparation techniques, and internal standards, calibration bias has been found in numerous studies to be the major contributor of disagreement among LC-MS/MS methods.
For example, calibration bias caused significant disagreement among six routine LC-MS/MS assays for 25-hydroxyvitamin D (25(OH)D) (1). Use of a common calibrator reduced disagreement among methods from 30% to 9% and 18% to 7% for 25-OH vitamin D2 and D3, respectively. In another example, calibration bias was the major source of disagreement among LC-MS/MS methods for thyroglobulin even though the methods employed different sample preparation techniques and peptides used for quantification (2). Adjusting the results to eliminate bias significantly improved agreement among methods.
Harmonization and Standardization
One solution to the problem of method disagreement is harmonization. Results among different methods can be harmonized using common calibrators or by adjusting calibrator value assignments to produce the same results as a designated comparator method. Another way to achieve agreement among methods is through the process of standardization. Standardization ensures that the numerical results generated from routine laboratory methods are traceable, or linked, to a formal reference measurement system. A reference measurement system consists of a reference measurement procedure and reference materials (3).
Reference measurement procedures (RMPs) are higher-order methods with analytical performance parameters that meet defined criteria. RMPs require a large degree of analytical rigor and are usually developed by specialized metrological laboratories such as those at the Centers for Disease Control and Prevention (CDC), the National Institute of Standards and Technology (NIST), and the University of Ghent. The Joint Committee for Traceability in Laboratory Medicine (JCTLM) establishes analytical performance requirements for RMPs and provides a list of laboratories that maintain approved RMPs (http://www.bipm.org/en/committees/jc/jctlm/). LC-MS/MS is frequently the methodology of choice for RMPs because of its high degree of specificity and its ability to use stable isotope-labeled internal standards to account for differences in process efficiency among samples.
It is important to recognize that approved RMPs are very different than LC-MS/MS methods performed by routine clinical laboratories. RMPs are calibrated using pure reference standards traceable to the International System of Units, and sample preparation approaches are designed such that matrix interferences are negligible. Gravimetric measurements are used to prepare calibrator solutions and add internal standards. To achieve the most precise and accurate results possible, LC-MS/MS-based RMPs typically use a two-step process for quantification. First, the concentration of the analyte is initially estimated for each sample from a calibration curve. Then, the internal standard concentration for each sample is adjusted individually to closely approximate that of the analyte concentration in the sample in order to calculate the final result.
The second component of a reference measurement system consists of commutable reference materials (RMs). RMs are matrix-appropriate materials that have been value-assigned using a RMP. For use in standardization efforts, RMs must be commutable, which means that the measurement response of the analyte in the RM is the same as the response in patient samples when both the RM and the patient samples are measured using a routine method.
Resources for Laboratories
CDC has developed programs to help laboratories and manufacturers achieve standardization, including the Lipid Standardization Program, the Hormone Standardization Program, and the Vitamin D Standardization-Certification Program (http://www.cdc.gov/labstandards/lsp.html, http://www.cdc.gov/labstandards/hs.html). To support these programs, CDC has developed mass spectrometry-based RMPs for total cholesterol, low-density lipoprotein cholesterol, high-density lipoprotein cholesterol, triglycerides, 25(OH) D, testosterone, and estradiol. Future development plans include RMPs for free thyroxine, free triiodothyronine, parathyroid hormone, and glucose. If laboratories are enrolled in a standardization program, they will receive an initial set of single donor, commutable samples that have been value-assigned by the CDC RMP to enable the laboratory to establish and/or trace the method’s calibrator value assignments. Laboratories will then receive a second set of unknown test samples. If the laboratory’s values for the test samples fall within CDC’s acceptability criteria (based on performance criteria established from biological variation data), the laboratory will be listed on CDC’s website as standardized. Participating laboratories receive test samples on a quarterly basis to ensure they maintain standardization.
Results from standardization programs demonstrate successful improvement of comparability among methods. For laboratories participating in CDC’s Hormone Standardization Program, bias between routine LC-MS/MS assays and the RMP for testosterone decreased by approximately 50% from 2007 to 2011 (4). Questions regarding the CDC Standardization Programs may be emailed to [email protected].
NIST also has developed JCTLM-listed RMPs for 40 analytes, including LC-MS/MS-based RMPs for 25(OH) D, cortisol, and estradiol. External quality assessment programs such as the Vitamin D External Quality Assessment Scheme and the College of American Pathologists Accuracy Based Survey for Vitamin D use RMPs from NIST and CDC to support value-assignment for proficiency testing materials.
The analytical approaches conducted at NIST highlight another way that RMPs differ from routine laboratory methods. NIST dedicates a great deal of effort to determining the purity of primary RMs used to calibrate RMPs. For example, NIST employs quantitative nuclear magnetic resonance, moisture content analysis, and LC-ultraviolet to assess purity and assign mole values to primary materials. NIST-certified standard reference materials such as SRM 967a (Creatinine in Frozen Human Serum), SRM 971 (Hormones in Frozen Human Serum), and SRM 972a (Vitamin D Metabolites in Frozen Human Serum) are value-assigned using NIST RMPs and are available to clinical laboratories to assess accuracy. A detailed list of available SRMs can be found at www.nist.gov/srm.
Professional practice guidelines reinforce the importance of avoiding calibration bias through traceability. The Clinical Laboratory and Standards Institute (CLSI) C-62A guideline for quantitative LC-MS recommends that laboratories use calibrators with value assignments that are traceable to recognized RMPs, whenever available (5). For in-house-developed or commercial calibrators that are not traceable to an RMP, C-62A recommends using commutable reference materials to value-assign calibrators. ISO standard 17511 (6) and CLSI EP32-R (7) provide additional recommendations and technical details on how to implement calibrator traceability.
There are numerous ways for clinical laboratories to establish and monitor calibrator traceability to RMPs. They can participate in standardization programs or send their calibrators for value assignment to a laboratory that offers a JCTLM-listed RMP. Participating in accuracy-based proficiency testing surveys that use commutable materials (value assigned by RMPs) is another way to assess calibrator traceability. Alternatively, laboratories can assess the accuracy of their calibrator value assignments by using commutable RMs that have been value-assigned using an approved RMP.
For an initial evaluation of calibrator value assignments, laboratories can analyze the RMs as unknown samples using the laboratory’s routine method. If the measured values fall within the uncertainty of the RM concentrations from the certificate of analysis, the routine method’s calibration likely will produce values similar to those of the RMP. If the measured values fall outside of the uncertainty of the RM-assigned values, the laboratory can adjust its method calibrator values or calibration curve equation to recover the expected values (Fig 1). Laboratories should monitor calibration accuracy semi-annually and for new lots of calibrators.
Routine laboratories can achieve a high degree of accuracy by implementing LC-MS/MS methods, but attention must be paid to ensure calibrator accuracy and traceability. When available, participation in standardization efforts and accuracy-based proficiency programs or use of certified RMs is recommended. If standardization programs are not available, it is the responsibility of manufacturers and laboratories to work together to identify common calibrators or an agreed-upon comparator method to ensure harmonized results. These steps are necessary to ensure that LC-MS/MS methods employed by routine clinical laboratories meet accuracy expectations.
1) Yates AM, Bowron A, Calton L, et al. Interlaboratory variation in 25-Hydroxyvitamin D2 and 25-Hydroxyvitamin D3 is significantly improved if common calibration material is used. Clin Chem 2008;54:2082-83.
2) Netzel BC, Grant RP, Hoofnagle AN, et al. First steps toward harmonization of LC-MS/MS thyroglobulin assays. Clin Chem 2016;62:297-98.
3) Bunk DM. Reference materials and reference measurement procedures: An overview from a national metrology institute. Clin Biochem Rev 2007;28:131-37.
4) Vesper HW, Botelho JC, Wang Y. Challenges and improvements in testosterone and estradiol testing. Asian J of Andrology 2014;16;178–84.
5) CLSI. Liquid chromatography-mass spectrometry methods; Approved Guideline. CLSI document C-62A. Wayne, PA: Clinical and Laboratory Standards Institute; 2014.
6) ISO. In vitro diagnostic medical devices‑Measurement of quantities in biological samples‑Metrological traceability of values assigned to calibrators and control materials. ISO document 17511, Geneva, ISO 2003.
7) CLSI. Metrological traceability and its implementation: A report. CLSI document EP-32-R. Wayne, PA: Clinical and Laboratory Standards Institute; 2006.
Lorin M. Bachmann, PhD, DABCC, is an associate professor of pathology, co-director of clinical chemistry, and co-director of point-of-care testing at Virginia Commonwealth University Health System in Richmond, Virginia. +EMAIL: [email protected]
CLN's Focus on Mass Spectrometry is sponsored by Waters Corporation.