Answers are grounded in this paper and cite their sources.
The move by the JCEM to mandate measurement of sex steroids by mass spectrometry (MS) in manuscripts to be published from the beginning of 2015 (1) sparked vigorous debate among clinicians and scientists working within the diverse subspecialties of endocrinology. In response, The Endocrine Society Council assembled a Sex Steroid Assays Reporting Task Force (SSARTF) to review current reporting polices laid out by the Journal (and sister journals) and to recommend through the Publications Core Committee a clear policy on the future reporting of sex steroid hormone measurements. We commend the SSARTF in broadening the scope of their new guideline not only to encompass sex steroid hormone measurement but also to address the measurement of all steroid hormones more generally. Such an inclusive policy will undoubtedly propel the field toward optimized measurement procedures. With this goal in mind, we wish to make the case that the limitations inherent in immunoassay (IA) and the opportunities afforded by MS are not confined to the measurement of sex steroids, but rather the argument for measurement of cortisol and related metabolites is equally as compelling. Here, we present the case for cortisol measurement as an exemplar for the wider adoption of MS. IAs for serum steroid measurement are afflicted with cross-reactivity, to varying and often clinically significant degrees, with both endogenous and exogenous steroid hormones and their metabolites (2). Examples of analytical interference in commercial IAs include cross-reactivity in T assay by the structurally similar oral contraceptive pill component norethisterone (3) and the impact of both sample matrix and gender on the measurement of cortisol, likely attributable to variations in both cortisol-binding globulin concentration, and the accumulation of structurally related steroids in various physiological and pathophysiological conditions (4) acts to illustrate the inherent and persistent obstacles encountered by IA-based methodologies regardless of measurand, be it androgen or glucocorticoid. Similarly, indirect interference in many cortisol IAs has been demonstrated in patients receiving the 11β-hydroxylase inhibitor metyrapone for the medical management of Cushing's syndrome (5, 6). Metyrapone therapy blocks the conversion of 11-deoxycortisol to cortisol, causing a >10-fold increase in circulating concentration of this precursor (7, 8), which cross-reacts significantly in cortisol IA. The consequent positive interference carries the very real risk of inappropriate dose escalation and unrecognized hypoadrenalism. Salivary cortisol measurement used routinely in the investigation of Cushing's syndrome, and as a measure of “stress” in psychological research, is often measured by IA. However, there is a failure to recognize that, because of the abundance in parotid tissue of the enzyme 11β-hydroxysteroid dehydrogenase type II, which converts cortisol to cortisone, the latter is the predominate glucocorticoid in saliva at levels between 4- and 9-fold greater than cortisol. Cortisone cross-reactivity may therefore confound the accurate measurement of salivary cortisol in many IAs. Some of these problems inherent in IAs may be resolved by extracting samples to remove cross-reacting compounds and alleviate matrix effects before measurement, but importantly, such preanalytical extraction adds additional costs and complexity. In recent decades, the great advantage with IA has been speed and simplicity; sample prepurification seriously derogates from this and would therefore need to be evaluated in view of potential gains in diagnostic efficiency (9). These common and clinically important interferences in IA can be surmounted by the use of MS, and at present there are no feasible methodological alternatives that provide equivalent analytical robustness for steroid assay. As well as eliminating cross-reactivity, a further crucial practical advantage of MS is the dynamic range of measurement that can be achieved, comfortably exceeding that possible by conventional competitive IA methodology for the quantification of small molecules such as steroid hormones. For competitive IA, the typical dynamic range of the sigmoidal dose-response curve is no more than two orders of magnitude in measurand concentration. Assay manufacturers must therefore consider the clinical requirement for analytical sensitivity and also the expected (patho)physiological concentrations of measurand when designing these assays. Further methodological advantages of MS include the lack of susceptibility to interference from antireagent antibodies and the capacity to multiplex measurements for steroid profiling. In other words, a female androgen profile consisting of T, androstenedione, and dehydroepiandrosterone sulfate may be simultaneously measured in a single assay at negligible extra cost compared to measuring any of the steroids individually. From a practical standpoint, rapid MS assays for the routine measurement of serum cortisol are now available; the high level of specificity conferred by this technology has the added benefit of identifying samples that show evidence of exogenous steroid use. Recently, one such assay used in a routine diagnostics laboratory detected exogenous steroid use in 33% of serum samples, many of these samples giving a falsely elevated cortisol value by IA with the potential to affect patient management (10). Historically, the high cost of MS instrumentation has been prohibitive. However, the falling cost of hardware and low running costs for consumables is shaping MS as a viable option for routine diagnostic laboratories. To advocate for the measurement of sex steroids alone, the JCEM risks underutilization of clinical MS. As the portfolio of clinical MS applications expands, not least through advances to enable routine quantitative multiplexing alongside therapeutic drug monitoring, vitamin D and cortisol measurements, the economics of ensuring sample throughput supports the MS infrastructure of the laboratory, thus extending the utility of existing instrumentation to keep hardware busy provides economy of scale. It is important to acknowledge that MS, in common with IA, is liable to interference that can compromise the analytical integrity of the method, and factors including matrix effects, choice of internal standard and solvent, mass-transition selection, and ionic cross-talk must be considered during method validation (11). Matrix-dependent signal enhancement (ion enhancement) or suppression (ion suppression) can directly affect the quantitative performance of a mass spectrometer. These phenomena are commonly referred to as matrix effects, defined as the combined effect of all components of a sample other than the analyte on the measurement of the quantity of interest (12). There are many strategies to overcome matrix effects in MS-based assays; of particular note is the use of the internal standard approach whereby variation in the measurand signal can be balanced by an equivalent perturbation in the internal standard signal, thus correcting for the degree of ion suppression/enhancement. This topic has been comprehensively reviewed by Trufelli et al (13). The adoption of MS therefore needs to be undertaken in conjunction with robust standards for hormone measurement that recognize the need for assay standardization, traceability, and quality control. The JCEM preliminary stance on mandating MS for sex steroid assay may well have been contentious; arguments against a requirement for a specific methodology have focused on the potential for such a stand to stifle development of new methodologies by industry and academia. In advocating for MS, we would not wish to exaggerate the limitations of IA, and we recognize that in most clinical settings it is more than adequate to assist in clinical decision making. The JCEM is a journal with global impact that attracts manuscripts from around the world from researchers with very diverse infrastructure support. It cannot be the intention to exclude good science from the Journal because of inaccessibility to MS and a consequent use of well-validated IA. Bias and intra- and interassay coefficients of variation are issues with MS, and support for purist approaches to assay development focusing on traceability to a certified standard shouldn't be compromised as we move to new technologies. However, the development of assays traceable to high-order reference material and methodology does not detract from the fact that IA is susceptible to a plethora of analytical pitfalls that may be completely overlooked during standard assay validation procedures and may only come to light on postmarket surveillance. When possible, novel assays must be validated against a reference method, be assessed with certified reference material, or be compared against a well-validated procedure using clinical samples. The current “limitation” of MS for clinical use now appears not to be a cost or technical expertise issue, but rather one of calibration. A fundamental consideration in this respect is the potential matrix effects exerted by calibration materials. Any matrix effects in the calibrator should approximate those manifest by patient samples. Such matrix-matched calibrators are now becoming available. However, it will remain vital that commutability of these reference materials is thoroughly validated and demonstrable across methodologies, an approach that will necessitate open dialogue between all stakeholders, including manufacturers of reference material, the in vitro diagnostics (IVD) industry, and clinical laboratories. Furthermore, comprehensively established reference intervals and clinical validation of new assays for existing measurands is a prerequisite for optimal analytical and diagnostic assay performance. Currently, the two largest areas for diagnostic MS are therapeutic drug monitoring and vitamin D measurement, and this has been realized through the availability of commercial calibration sets. In contrast, the progression to MS assays for steroids is at present impeded by the scarcity of commercially available calibration material, and many diagnostic laboratories are unable or unwilling to produce calibrator and quality control materials for their MS assays. This is a real issue that requires addressing. The Joint Committee for Traceability in Laboratory Medicine database, which provides information for manufacturers and laboratories to establish and confirm traceability for routine methods, holds information concerning a number of reference preparations for cortisol in human serum that are available to manufacturers. There is currently a drive by manufacturers of mass spectrometers and MS-based assay kits to move into the diagnostics market. Manufacturers are required to register their devices with regulatory bodies. For example, in the United States this is overseen by the Food and Drug Administration, which requires a 510(k) submission or more rigorous premarket notification process. In Europe, the IVD Directive 98/79/EC applies. Assays that are developed and validated by laboratories “in-house” are currently exempted from the European Directive in its current form as long as they are produced within health-institution laboratories for use in that environment and are not subject to commercial transactions. However, in view of the ongoing revisions in this area, these in-house assays are likely to face greater interrogation by regulatory bodies in the near future, and rightly so. Indeed, laboratory-developed tests in the United States are regulated by the Center of Medicare and Medicaid Services, based on the Clinical Laboratory Improvement Amendments (CLIA) regulations, and laboratories are assessed against defined requirements as part of annual inspection for CLIA-certified laboratories. The role of external quality assessment (EQA) cannot be overstated as an essential tool for monitoring both short- and long-term analytical performance of diagnostic assays, and developers of MS-based assays must be held to operating by the same standards as the established commercial IA kit manufacturers. The utility of EQA is constantly evolving, and now may be an opportune time to push the debate on target level-based EQA—specifically, the use of MS target means to enable EQA performance evaluation against the best estimate of the true result. Currently, however, the main barrier to the use of MS target means for serum cortisol is the lack of participating laboratories. For example, less than 3% of laboratories participating in the UK National External Quality Assessment Scheme for steroid hormones measure serum cortisol by MS, limiting the robustness of target means. Nonetheless, as the role of EQA for steroid hormones expands, so too will its influence. It is evident that not all MS-based methods are the same (14, 15), although meaningful research should be supported by a high-quality, fully validated method, and evidence for this should be presented within the resultant publication. At this juncture, The Endocrine Society is best placed to move on journal standards, making it a requirement for manuscripts submitted to Endocrine Society journals to document assay calibration procedures and traceability. Such a move will promote transparency and greatly facilitate meaningful comparison of study data. This also presents an opportunity for The Endocrine Society to lead on the adoption of matrix-matched traceable commercial calibrator materials that may even be incorporated within ready-to-use steroid assay kits for IVD use, removing the need for a time-consuming method development and validation. In conclusion, well-validated IAs continue to serve us well in the diagnosis of most steroid-related clinical disorders. There are, however, notable exceptions that increasingly support the assertion that the era of steroid IA is drawing to a close. We must recognize that to reap the full benefits of MS-based assays for the diverse repertoire of clinically relevant steroids, there needs to be defined assay standards including designated reference material and participation in EQA schemes. The Endocrine Society has the opportunity to define and drive the adoption of best practice as the move to MS-based steroid assays gains momentum. Disclosure Summary: P.J.M. and P.J.T. have received research funding from Immunodiagnostic Systems PLC. The other authors have nothing to declare. external quality assessment immunoassay in vitro diagnostics mass spectrometry.
No takes yet. Share an insight, caveat, or question.
Monaghan et al. (2014) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: