Measurement of hormones
Introduction
The role of accurate and reliable laboratory testing is particularly important for patients with potential endocrine disorders. The revolution which has taken place in the past 50 years in the methodology of hormone measurement is thus of considerable significance to this patient group. It is difficult to imagine that not too long ago common hormone measurements, such as thyroid function tests, took more than a week to produce. Now we live in a world where same day turnaround is the norm for the high throughput commonly requested tests. This is largely due to advances in the way hormones are measured and results delivered to the practising clinical endocrinologist.
Measuring hormones has always been a challenge as most circulate at extremely low concentrations, typically in the pico- (10–12) or nanomolar (10–9) range, and often in a milieu of closely related and potentially interfering compounds making great demands on method sensitivity and specificity. The most common procedures currently used are immuno- and immunometric assays but gas chromatography mass spectrometry (GCMS) and high-performance liquid chromatography (HPLC) also have a place. Liquid chromatography mass spectrometry (LC-MS/MS) is rapidly gaining acceptance for a limited number of hormone measurements.
It is not the aim of this chapter to provide precise detail on hormone measurement methodology but rather to overview general principles and applications of methods in current use. Attention is drawn to preanalytical and analytical problems which could have significant clinical consequences if not recognized.
Antibody-based methods
It was in late 1950s and early 1960s that it was first demonstrated that specific antibodies could be used to detect hormones and this discovery revolutionized clinical endocrinology. The first immunoassays were described by Yalow and Berson (1) and Ekins (2) for the measurement of insulin and thyroxine, respectively. Immunoassay is now the most widely applied technique for measuring hormones in biological samples. Nowadays, immunoassays are more likely to be developed within the diagnostics industry than by academic experts, with increasing emphasis on methods suitable for large, throughput automated platforms. Manual ‘in house’ or commercial kit procedures, however, remain prominent for the more specialist, lower throughput hormone measurements and are usually performed in specialized regional clinical laboratories.
The basic requirements for immunoassay are an antibody (or antibodies) to the analyte to be measured, a labelled form of the analyte (competitive immunoassay) or a labelled second antibody to the analyte (noncompetitive immunoassay). Procedures for separating antibody-bound tracer from unbound tracer and a means for detecting the tracer are also required. To facilitate separation, antibodies can be attached to solid surfaces, such as polystyrene reaction tubes or microtitre plates, plastic beads, or cellulose particles, thus allowing the unbound portion to be removed by a wash procedure. Commercial methods frequently use magnetized particles, which simplify separation on automated platforms.
Competitive immunoassay
Hormone immunoassays rely on high specific activity labels, such as radioisotopes, to reveal the products of the hormone–antibody reaction. The ‘first generation’ immunoassays—in common use from the 1960s to the mid-1980s—relied almost exclusively on the inclusion of a trace amount of radiolabelled hormone in a reaction mixture comprising the test sample (serum, urine, or saliva) and a limited amount of antibody (or other binding agent). The analytical principle governing these methods (termed RIA) involve ‘competition’ between labelled and unlabelled hormone molecules for the antibody present (Fig. 1.7.1). After incubation, the proportion of labelled analyte decreases as the concentration of the analyte being measured increases. Such assays are therefore often referred to as ‘competitive’ or ‘displacement’ assays. To avoid problems related to handling of radioactivity and the limited shelf-life of radiolabelled reagents these have now been largely, but not yet completely, superseded by labels employing fluorescent or chemiluminescent substances or enzymes.
One requirement for competitive immunoassay is that there should be no interference from circulating binding proteins which could participate in competition with the labelled analyte. Since many small-molecular-weight hormones, such as steroids, bind with high affinity to circulating binding proteins, the traditional approach was to separate them from the binding protein by extracting into an organic solvent such as diethyl ether. This has the added benefit of also removing water soluble, potentially cross-reacting, conjugated steroids. Unfortunately solvent extraction is a labour-intensive step that is difficult to automate. The introduction of simple, direct immunoassays for the measurements of steroids in unextracted serum or plasma was therefore a significant advance. In direct steroid immunoassays steroids are displaced from binding proteins by a chemical agent ‘in situ’ and, ideally, these agents should not affect antibody-binding characteristics, but this is not always achieved (3). Although the advances in measuring steroid hormones directly have progressed the introduction of automated steroid measurements on large, fast throughput automated immunoassay platforms they do place great demand as illustrated later, on antiserum specificity as potentially cross-reacting steroid conjugates are not removed.
Noncompetitive immunoassays
In the late 1960s, methods relying on radiolabelled antibodies (termed immunoradiometric assays (IRMAs)) were first described (4, 5) As with competitive immunoassays as the labelled antibody procedure evolved a whole range of nonradioactive labels (enzymes, fluorofloures, chemiluminescent) were introduced. Since in these assays no competition occurs between labelled and unlabelled analyte for antibody-binding sites these immunometric assays have also been termed noncompetitive immunoassays or ‘sandwich’ assays. They rely on the detection of occupied antibody-binding sites to which the analyte has bound (Fig. 1.7.2). The amount of analyte bound to the first antibody is detected by the binding, and formation of a ‘sandwich’, with another antibody to which a label is attached. These assays are suitable for analytes of large molecular size (that is, of a molecular weight of approximately 1000 Da and above). For hormones of smaller molecular size (and thus incapable of binding simultaneously to two antibodies) the competitive approach continues to be generally employed.
Immunometric procedures require high concentrations of unlabelled antibody of known specificity. The full potential of the immunometric assay was, therefore, only realized with the introduction of in vitro monoclonal antibody procedures by Köhler and Milstein in 1975 (6). Further improvements were later made by the introduction of high activity labels for attachment to the second antibody in the ‘sandwich’ with remarkable improvement in sensitivity. In the 1970s, time-resolved fluorometric immunoassay methodology, now known as DELFIA (7, 8), was developed. Based on the use of lanthanide chelate fluorophors and labelled monoclonal antibodies, this was the first of many ‘ultra-sensitive’ nonisotopic immunoassay methodologies. The same approach has subsequently been adopted by many manufacturers using other high specific activity non-isotopic labels, as reviewed by Kricka (9). The use of such labels in immunoassays of noncompetitive design revolutionized the immunodiagnostic field towards the end of the twentieth century and underlies attempts to further improve assay sensitivities.
Very high sensitivity is clearly of particular importance in the case of certain hormones, for example, thyroid-stimulating hormone (TSH). Serum concentrations in hyperthyroid individuals not only fell below the limit of detection of the original radioimmunoassay methods used in the 1970s and 1980s, but were essentially indistinguishable from normal values. Ultrasensitive TSH methods (in combination with free thyroxine assays) are now widely used in the laboratory diagnosis of thyroid dysfunction. An equally important consequence of the development of ultrasensitive immunoassays has been a major reduction in assay performance times, resulting in the emergence of the automated immunoanalysers that now dominate the field. Total incubation times in the order of minutes are typical, replacing the hours or days characterizing first generation ‘competitive’ methodologies.
In summary, hormone immunoassay methods now in common use (and widely available as kits—usually incorporated in the menus provided by immunoanalyser manufacturers) are of both competitive single-site and noncompetitive two-site design. The former approach is generally adopted for the assay of hormones of small size (such as steroid and thyroid hormones), the latter for the assay of hormones of large molecular size (e.g. polypeptide and glycoprotein hormones). High specific activity nonisotopic labels, yielding higher sensitivities in assays of noncompetitive design, have largely replaced radioisotopic labels. Though their use does not significantly improve the sensitivities of competitive methods, the longer shelf lives of labelled reagents and other such practical benefits are also factors contributing to the general abandonment of radioisotopic labels by immunoassay kit manufacturers.
Free (nonprotein bound) immunoassays
In the case of those hormones (such as thyroid and steroid hormones) present in blood in free and protein-bound forms, it is widely accepted that the free hormone concentration measured under equilibrium conditions in vitro constitutes the determinant of the hormone’s physiological activity. This concept, termed the ‘free hormone hypothesis’, derives primarily from observations that, in subjects in whom serum binding protein concentrations are ‘abnormal’, overall hormone effects correlate closely with the free hormone concentration. Despite doubts about the validity of the free hormone hypothesis (10, 11). Direct measurement of free thyroid and steroid hormones by equilibrium dialysis or centrifugal ultrafiltration is technically challenging and generally unavailable outside specialized research laboratories. Free hormone immunoassay methods have therefore been developed that rely on the basic physicochemical principle, which is that exposure of a small amount of antihormone antibody to a test serum sample results in occupancy of antibody-binding sites to an extent that reflects the ambient free hormone concentration in the sample (Fig. 1.7.3). Occupancy of binding sites can be determined in three different ways.
♦ The ‘labelled hormone, back-titration’ approach (‘two-step’ free hormone immunoassay) which relies on determination of unoccupied antibody-binding sites (the antibody being generally linked to a solid support) by their exposure to labelled hormone following removal of the test serum.
♦ The ‘labelled hormone analogue’ approach (‘single-step’ free hormone immunoassay), which obviates these sequential operations by the use of a labelled hormone analogue that must, in principle, be totally unreactive with serum proteins (though retaining the ability to react with antibody).
♦ The ‘labelled antibody’ approach (likewise a ‘single-step’ immunoassay), which also relies on the use of a hormone analogue. The analogue used in labelled antibody techniques, however, is coupled to a solid support, such attachment apparently contributing to a reduction of analogue binding to serum proteins. For this and other reasons, labelled antibody assay kits appear to conform more closely to the principles governing valid analogue-based free hormone immunoassays, and generally yield correct and clinically reliable results.
The main current application of free hormone immunoassay is for the measurement of free thyroxine. Undoubtedly reliable measurement of circulating ‘free’ thyroxine by immunoassay is a better diagnostic test than total thyroxine. Although a few commercial kits do exist for measuring free steroid concentrations, these have not received general acceptance. In contrast, in the case of testosterone, an index of the free hormone concentration is often calculated. For females the free androgen index (total testosterone/sex hormone-binding globulin × 100) (12) is generally used but this is not valid for males, in whim ‘free’ testosterone can be derived from measured circulating total testosterone, sex hormone-binding globulin, and albumin concentrations based on the binding constants of testosterone to these circulating proteins (13).
An alternative to measuring the circulating free hormone concentration is to measure the hormone in saliva. Saliva measurements have been developed for a number of steroids and have the potential to provide a convenient and noninvasive assessment of the serum ‘free’ steroid concentrations. Salivary concentrations of unconjugated steroids reflect those for free steroids in serum although concentrations may differ because of salivary gland metabolism. The use of salivary assays for both research and routine purposes has recently been reviewed in detail by Wood (14). Measuring salivary cortisol late evening is now an accepted and sensitive screening test for Cushing’s. Salivary 17-hydroxyprogesterone and androstenedione assays are valued as noninvasive for home monitoring of hydrocortisone replacement in patients with the 21-hydroxylase deficiency variant of congenital adrenal hyperplasia. The diagnostic value of salivary oestradiol, progesterone, testosterone, dehydroepiandrosterone, and aldosterone testing is compromised by rapid fluctuations in salivary concentrations of these steroids.
Methods based on chromatographic separation followed by nonimmunological detection
Hormone analysis by gas chromatography and HPLC rely on chromatographic separation followed by a variety of detection procedures. As indicated in Table 1.7.1 these methods are usually restricted to measurement of nonprotein hormones such as steroids and related metabolites, vitamin D metabolites, catecholamines, and metabolites. Prior to chromatographic separation an initial purification step may be required: the simplest is extraction into an organic solvent(liquid/liquid extraction) but alternatively solid phase extraction (SPE) is becoming more popular. SPE can be used to effect sample extraction, concentration, and purification. The most commonly used hormone measurement procedure is reverse-phased SPE (polar liquid phase, nonpolar modified solid phase) which involves capture of the analyte from a liquid phase onto silica microparticles coated with sorbent packed into syringes or cartridges. One of the earliest descriptions of the use of SPE for hormone measurement is a method for purification of urinary steroids prior to gas chromatographic analysis described by Shackelton and Whitney in 1980 (15). Although it is not possible to automate liquid/liquid extraction procedures, recent advances in laboratory robotics allow automation of the solid phase extraction step off-line (16) or online, utilizing column-switching techniques, linked to either HPLC or LC-MS/MS (17).
Table 1.7.1 Procedures used to commonly measure hormone and hormone metabolites
Procedure | Sample type | Hormones measured |
|---|---|---|
Immunoassay | Serum/plasma | Thyroid hormones |
Urine | Steroid hormones | |
Saliva | Specific hormone-binding proteins | |
Filter paper dried blood spots | Protein and peptide hormones | |
Vitamin D metabolites | ||
Cortisol | ||
Cortisol, 17- hydroxyprogesterone | ||
Thyroid-stimulating hormone, 17-hydroxyprogesterone | ||
High-performance liquid chromatography | Serum/plasma | Vitamin D metabolites |
Urine | Steroid hormones | |
Catecholamines and metabolites | ||
Gas chromatography mass spectrometry | Urine | Steroid metabolites |
Tandem mass spectrometry (LC-MS/MS) | Serum/plasma | Vitamin D metabolites |
Urine | Testosterone | |
Filter paper dried blood spots | Adrenal steroid profiles | |
Cortisol | ||
17-hydroxyprogesterone |
Based on the knowledge of the chemical nature of both the hormone to be purified and any interferants, if present, the polarity of the organic solvent (for liquid/liquid extraction procedures) or type of solid phase and polarity of the eluting solvent (for SPE) can be selected to ensure adequate purification. The amount of purification and concentration required is governed, to some extent, by the efficiency of the next chromatographic stage and the sensitivity of the final detection stage.
In the chromatographic stage, in order to achieve separation a dynamic equilibrium distribution between compounds in a mobile phase such as a flowing gas or liquid and the stationary phase is established. As the analyte is propelled by the mobile phase over the stationary phase chromatographic separation is achieved with those compounds preferentially distributed in the mobile phase passing more quickly through the system than those preferentially distributed in the stationary phase. The stationary phase is contained in a fused silica (gas chromatography) or steel tube (HPLC) and a gas or liquid flow is maintained by the application of pressure. With careful optimization of the polarity of the stationary and/or the mobile phase very complex separations can be achieved.
Extraction and chromatographic separation will inevitably lead to loss of analyte necessitating the need for inclusion of an internal standard to correct for procedural losses. The internal standard should be indistinguishable from the analyte during the process of extraction and purification and the choice of internal standard is to some extent dictated by the quantification procedure. For example, for procedures employing mass spectrometry deuterated internal standards are ideal being chemically identical yet detectable by virtue of increased mass. This procedure generally compensates for any matrix related effects and is commonly termed isotope dilution mass spectrometry. When other types of detection systems such as light absorption, fluorescence, and electrochemical properties are used, an internal standard is usually selected which has similar chemical properties to the analyte but is not present in biological samples.
Gas chromatography mass spectrometry
GCMS has been established over several decades as an important procedure for measuring hormones. The combination of gas chromatography with mass spectrometry exploits the high-resolving power of gas chromatography to separate closely related molecules and the ability of mass spectrometry to provide precise data for identification and quantification of the separated substances. The two prerequisites are volatility and thermal stability of the compounds to be separated. This limits GCMS to measurement of compounds with a molecular weight of less than 800 Da such as steroids and thyroid hormones. Furthermore, derivatization is often required to increase volatility and thermo stability of the analyte.
During gas chromatography a liquid sample is evaporated at high temperature and the volatile constituents blown through a hollow flexible silica capillary column, to which is coated or bonded a liquid stationary phase. In most GCMS systems, the gas chromatography column passes through a vacuum seal delivering the separated molecules into the ion source of the mass spectrometer. Here, under vacuum, the molecules are bombarded with either electrons (electron impact ionization) or charged ions (chemical ionization) resulting in molecular instability and production of positively charged fragments. The mass spectrometer is able to use differences in the mass-to-charge ratio (m/z) of these ionized fragments to separate and detect each fragment. In essence the pattern of fragments provides a ‘fingerprint’ for the molecule under investigation allowing positive identification and quantification.
Unfortunately GCMS methods require laborious sample preparation limiting use for routine hormone measurements. Such methods have, however, provided a valuable tool for establishing reference methods for steroid and thyroid hormones where sample throughput is not an issue (18). GCMS has been more widely used in the specialized endocrine laboratory for profiling urinary steroid metabolites for both routine and research purposes by methods adapted from those first introduced by Shackleton in 1986 (19). The measurement of urinary steroid metabolites aids the diagnosis of a number of inherited disorders of the synthesis and metabolism of adrenal steroids, and steroid-producing tumours. (20) The procedure is particularly valuable in identifying the site of the enzyme defect in congenital adrenal hyperplasia. An example of a urinary steroid profile in a case of untreated late-onset congenital adrenal hyperplasia is shown in Fig. 1.7.4a. The full fragmentation pattern for each steroid metabolite permits absolute identification of the metabolite. An example of a fragmentation pattern for one of the abnormally elevated metabolites (11-oxo-pregnanetriol) is shown in Fig. 1.7.4b.
Tandem mass spectrometry
The relatively new procedure of tandem mass spectrometry is beginning to make a significant impact on hormone measurement. When linked to HPLC this procedure is commonly abbreviated to LC-MS/MS. As in the early days of immunoassay there is currently a flurry of LC-MS/MS method development activity within specialist endocrine laboratories. The revolution got going in the 1990s with the introduction of atmospheric pressure ionization (API) and electrospray ionization procedures allowing the first clinical applications which were in the areas of neonatal screening and therapeutic drug monitoring. Compared with GCMS sample preparation for LC-MS/MS is more straightforward and can be applied to thermo labile compounds. Instead of analytes being ionized in a gas phase they are ionized in liquid phase which is much more appropriate for biological samples. As with GCMS, mass spectrometry is used to identify, characterize, and quantitate but by utilizing two quadrupole mass filters, separated by a collision cell, far greater specificity is achieved (Fig. 1.7.5). The first quadrupole mass filter (QMF1) selects ions sharing identical mass-to-charge ratio (m/z), all other ions are filtered out. In the collision cell the selected ions are fragmented into characteristic product ions by collision with a neutral gas (e.g. argon) and transmitted to the second quadrupole mass filter (QMF2). QMF2 is set to filter out all but the selected fragment ion. Thus, one defined ‘daughter ion’ from one defined ‘parent ion’ finally reaches the detector. The selection of masses by QMF1 and QMF2 can be changed within milli-seconds enabling a large number of different mass transitions to be monitored in parallel allowing multianalyte quantification. Illustrated in Fig. 1.7.6 are the output scans for a method developed for measurement of vitamin D metabolites by isotope dilution LC-MS/MS (16). After semiautomated SPE sample preparation analysis was performed using an automated LC-MS/MS system. The tandem mass spectrometer was used in positive ion mode with electrospray source and a stable isotope, hexadeuterated 25-hydroxyvitamin D3, used as internal standard to correct for procedural losses. The multiple-reaction monitoring transitions (parent → daughter) selected for quantification of 25-hydroxyvitamin D3, 25-hydroxyvitamin D2, and hexadeuterated 25-hydroxyvitamin D3 were m/z 401.35→159, 413.30→83, and 407.35→159, respectively. A more detailed account of principles of LC-MS/MS procedures and endocrine applications is provided in a two reviews by Vogeser and colleagues (21, 22).
Currently the methodology has been applied to steroids, vitamin D metabolites and catecholamines with greatest impact, so far, in the routine measurement of testosterone (23), vitamin D metabolites (16, 24) and urinary free cortisol (25). The ability to simultaneously measure a number of analytes is of value both for routine and research purposes. As previously illustrated (see Fig. 1.7.4) a number of vitamin D metabolites can be measured and likewise serum adrenal steroid profiles with simultaneous measurement of five, or more, steroids can be performed. A few further examples of methods developed to simultaneously measure hormones are testostosterone;dihydrotestosterone (26), cortisol;cortisone (27), and plasma-free metanerphine;normetanephrine (28).
The main obstacles to the use of LC-MS/MS in the clinical laboratory are the high initial cost of instrumentation (£150 000–250 000) and the high level of expertise required to develop methods and run the equipment. To some extent the high initial cost can be offset by the low cost of reagent compared to diagnostic immunoassay kits and as LC-MS/MS becomes more widely used expertise in this area is rapidly increasingly within both specialist and routine clinical laboratories.
Assay performance, interferences, and errors
Within the laboratory assay performance is assessed both internally and through external quality assessment schemes. It is important that each laboratory documents analytical accuracy, precision (both within and between batch) and reference ranges for all procedures. For procedures developed ‘in house’, or adapted from commercial procedures, more extensive evaluation on analytical and functional sensitivity and specificity are required. Comparison of results obtained by other methods and performance in external quality assessment schemes provide reassurance. Ideally methods should be compared to established reference methods but few are available in the endocrine field. Although this section focuses on analytical problems it is important to be aware that errors can also occur before and after the analytical process. Some of the causes of error from all three categories are summarized in Table 1.7.2.
Table 1.7.2 Errors related to laboratory testing
Preanalytical | Analytical | Postanalytical |
|---|---|---|
Inappropriate patient preparation (fasting, posture, time of last medication, stress) | Poor analytical performance Incorrect assay standardization | Incorrect reference range |
Dynamic function test performed incorrectly | Antibody interference in immunoassay | Incorrect interpretation Incorrect units |
Incorrect name on sample | Drug interference | Computer error during processing of reports |
Inadequate detail with request (e.g. clinical details age, sex, ethnicity, pregnancy, stage of menstrual cycle) | High dose ‘hook’ effect in immunometric assay Poor immunoassay specificity | |
Wrong test requested | Ion suppression in LC-MS/MS | |
Inappropriate sample type (serum, heparinized or EDTA plasma) | ||
Sample collected at inappropriate time of day | ||
Illegible handwriting on request form |
The accuracy of hormone assays may be compromised by a variety of interfering substances. The interference may be positive or negative and may vary in magnitude depending on the concentration of interfering substance in the sample. Often samples circulated by external quality assessment schemes are selected to address such issues.
Immunoassays are prone to interference from endogenous binding proteins and drugs. This has been mentioned earlier in the case of steroid hormones and can also occur with other methods for hormone measurement. For instance growth hormone binding protein can interfere with the measurement of growth hormone to different extents in different immunoassays (29). In free thyroxine assays, drugs, including phenytoin, carbamazepine, and salicylate, compete with thyroid binding to serum-binding protein and may increase the free thyroxine concentration. The most common drug interference in cortisol immunoassays is with prednisolone, which cross-reacts, to varying degrees, with all cortisol antisera.
The presence of antibodies in a serum sample can cause numerous problems in immunoassay. The effect of the interference will depend on the type of assay used and the site where the antibody binds to the analyte. Interference may lead to either falsely elevated or decreased values (30). An example of a common interferant is macroprolactin. Prolactin mostly circulates in monomeric form with a molecular weight 23 KDa but in a few individuals a large antigen-antibody complex of molecular weight greater than 100 KDa, commonly called macroprolactin, is present which may be detected as prolactin in the assay. It is therefore good laboratory practice to further investigate samples giving unexpected elevated prolactin results for the presence of macroprolactin. This can he easily achieved by precipitating macroprolactin in the serum sample with a precipitation agent, such as polyethylene glycol, and remeasuring prolactin on the supernatant (31). Less common and more difficult to recognize is interference from endogenous antibodies. Heterophilic antibodies can be associated with autoimmune and other inflammatory diseases. Interference may occur in both competitive and noncompetitive assays, but the latter is more common. The same is true if specific human antianimal antibodies which are present in some individuals in response to prior immunizations. Monoclonal antibody-based immunometric assays are especially sensitive to the presence of heterophilic and antianimal antibodies, which interfere by linking the capture to the detection antibody, causing false-positive results. Commercial heterophilic, antibody-blocking reagents can be used to minimize the effect of this type of interference or, as described for macroprolactin, antibodies may be removed by polyethylene glycol precipitation.
A significant and potential extremely dangerous problem can also occur specifically in immunometric assays if an exceptionally high concentration of the hormone being measured is present which simultaneously binds both the capture and detecting antibodies. This prevents the formation of the required complexes with capture antibody, analyte, and detecting antibody producing an incorrect low result in a sample that actually contains extremely high concentrations of analyte. This type of interference is commonly known as the high-dose ‘hook’ effect. For immunoassay procedures a simple test to indicate the presence of interference is to measure the sample over a range of dilutions. Interference is likely if a nonlinear response is obtained.
One major advantage of LC-MS/MS over immunoassay is improved specificity and this has been demonstrated most vociferously in relation to the measurement of testosterone. Although many nonextraction immunoassay methods perform satisfactorily in males, measurement in females and children is fraught with problems. Interferences related to the presence of incompletely blocked binding proteins and conjugated interfering steroids can cause falsely elevated results. In 2003, Taieb et al. (32) reported on the measurement of female testosterone by using 10 direct commercially available immunoassays compared with an isotope-dilution GCMS reference method. They concluded that most nonextraction immunoassays showed a large positive bias. Such was the extent of the problem that it prompted a hard hitting editorial in Clinical Chemistry, ‘Immunoassays for testosterone in women: better than a guess?’ (33). Although the exact nature of the interference is unknown there is some evidence that implicates the adrenal steroid dehydroepiandrosterone sulfate, which circulates at extremely high concentrations (µmol/l) compared with testosterone (nmol/l). It is, however, probable that other conjugated steroids and also binding protein-related interferences play a part and that different direct immunoassays are affected to different degrees depending on the specificity of the antibody used. Recognizing this problem, the Endocrine Society in the USA commissioned a panel of experts to look into the issue, which has now published a position statement on the utility, limitations, and pitfalls in measuring testosterone (34). They concluded that ‘direct’ immunoassay procedures are too insensitive and inaccurate to measure testosterone in the plasma of women and children and recommend that assays after solvent extraction and chromatography, followed by mass spectrometry or immunoassay, are likely to furnish more reliable results. These findings have accelerated the progress of LC-MS/MS as the method of choice for measuring testosterone in the clinical laboratory.
Another area where LC-MS/MS has led to improved diagnostic accuracy is the measurement of 17-hydroxyprogesterone in neonates. Transient elevation of both unconjugated and conjugated Δ-5 adrenal steroids produced by the fetal zone of the adrenal cortex early in life, especially in neonates born prematurely, cause positive interference in most direct 17-hydroxyprogesteorne immunoassays but not in more specific LC-MS/MS procedures. There are, however, situations where poor immunoassay specificity may actually be advantageous. For instance to correctly assess vitamin D status it is important to measure both 25-hydroxyvitamin D2 and 25-hydroxyvitamin D3 in the same sample. It is claimed that a number of commercial immunoassays achieve or partially achieve this but there is at least one example of an extremely specific commercial immunoassay that only detects 25-hydroxyvitamin D3 with the consequence that patients who are switched to vitamin D2 are not correctly assessed (35). Of course the best solution, as described earlier, is to measure both metabolites simultaneously by LC-MS/MS. It is, however, worth mentioning that LC-MS/MS is not totally free from analytical problems. Ion suppression can affect the quantitative performance of a mass detector. This can be caused by the presence of nonvolatile compounds such as salts, ion-pairing agents, endogenous compounds, and drugs/metabolites. This problem can usually be minimized by modification of reagents or chromatographic conditions (36).
To end on a rather sobering thought, it is important for the clinician to realize that half of all errors in the diagnostic process are not related to methodology at all, but occur before the sample is analysed. In fact 20% of these preanalytical errors are related to sample collection (Table 1.7.2). Even in hospitals where there is a heightened awareness of these problems, there is a prevalence of 1% preanalytical errors. These effects can be of sufficient magnitude to alter the analysis enough to create situations for clinical errors. Most problems can be prevented by clear instructions and documented policies for sampling. Some issues are relatively straightforward such as collecting the sample into the correct blood tube at the correct time of day and ensuring that samples are transported to the laboratory fast enough and at the correct temperature. If in doubt contact your local laboratory for current protocols.
In addition samples can and do get mixed up and mislabelled. In some instances this is easily identifiable if a totally inappropriate result is obtained, for instance, a male testosterone concentration in a female patient, but often differences from previous results can be more subtle. It is important when unexpected results are obtained that the possibility of preanalytical, analytical error and postanalytical error is thoroughly investigated. The first step is often to repeat the test. This could be followed by arranging for the measurement to be performed by a different method or procedure. Whenever specimens with interfering substances are identified, other laboratory data and clinical information on the patient, especially any acute or chronic disease and medications should be obtained. This information may provide clues to the cause of the interference which can be investigated in more detail in the laboratory.
Since clinical endocrinology is so dependent on laboratory investigation it is important that a close working relationship is built up between the endocrine clinician and the clinical laboratory scientist. A climate of mutual respect and close collaboration should ensure that problems are recognized and attended to promptly and that procedures are developed that are fit for purpose.
References
1. Yalow RS, Berson SA. Assay of plasma insulin in human subjects by immunological methods. Nature, 1959; 194: 1648–49.
Find This Resource
2. Ekins RP. The estimation of thyroxine in human plasma by an electrophoretic technique. Clin Chem Acta, 1960; 5: 453–9.
Find This Resource
3. Ratcliffe WA. Direct (non-extraction) serum assays for steroids. In: Hunter WM, Corrie JET, eds. Immunoassays for Clinical Chemistry 1983. Ediburgh: Churchill Livingstone, 1983: 401–9.
Find This Resource
4. Wide L, Bennich H, Johansson SGO. Diagnosis of allergy by an in-vitro test for allergen antibodies. Lancet, 1967; 2: 1105–7.
Find This Resource
5. Miles LEH, Hales CN. Labelled antibodies and immunological assay systems. Nature, 1968; 219: 186–9.
Find This Resource
6. Köhler G, Milstein C. Continuous cultures of fused cells secreting antibody of pre defined specificity. Nature, 1975; 256: 495–7.
Find This Resource
7. Marshall NJ, Dakubu S, Jackson T, Ekins RP. Pulsed light, time resolved fluoroimmunoassay. In: Albertini A, Ekins RP, eds. Monoclonal Antibodies and Developments in Immunoassay. Amsterdam: Elsevier, 1981: 101–8.
Find This Resource
8. Soini E, Lövgren T. Time-resolved fluorescence of lanthanide probes and applications in biotechnology. Anal Chem, 1987; 18: 105–54.
Find This Resource
9. Kricka LJ. Trends in immunoassay technologies. J Immunoassay, 1993; 16: 267–71.
Find This Resource
10. Robbins J, Rall JE. Thyroid hormone transport in blood and extravascular fluids. In: Gray CH, James VHT, eds. Hormones in Blood. London: Academic Press, 1979: 575–688.
Find This Resource
11. Tait JF, Burstein S. In vivo studies of steroid dynamics in man. In: Pincus V, Thimann KV, Astwood EB, eds. The Hormones. New York: Academic Press, Vol V, 1964: 441–57.
Find This Resource
12. Nanjee MN, Wheeler MJ. Plasma free testosterone–is an index sufficient?. Ann Clin Biochem, 1985; 22: 387–90.
Find This Resource
13. Vermeulen A, Verdonck L, Kaufman JM. A critical evaluation of simple methods for the estimation of free testosterone in serum. J Clin Endocrinol Metab, 1999; 84: 3666–72.
Find This Resource
14. Wood P. Salivary steroid assays–research or routine?. Ann Clin Biochem, 2009; 486: 183–96.
Find This Resource
15. Shackleton CH, Whitney JO. Use of Sep-pak cartridges for urinary steroid extraction: evaluation of the method for use prior to gas chromatographic analysis. Clin Chim Acta, 1980; 107: 231–43.
Find This Resource
16. Knox S, Harris J, Calton L, Wallace AM. A simple automated solid-phase extraction procedure for measurement of 25-hydroxyvitamin D3 and D2 by liquid chromatography-tandem mass spectrometry. Ann Clin Biochem, 2009; 46: 226–30.
Find This Resource
17. Xu RN, Fan L, Rieser MJ, El-Shourbagy TA. Recent advances in high-throughput quantitative bioanalysis by LC-MS/MS. J Pharm Biomed Anal, 2007; 44: 342–55.
Find This Resource
18. Thienpont LM, Van Nieuwenhove B, Stöckl D, Reinauer H, De Leenheer AP. Determination of reference method values by isotope dilution-gas chromatography/mass spectrometry: a five years’ experience of two EuropeanReference Laboratories. Eur J Clin Chem Clin Biochem, 1966; 34: 853–60.
Find This Resource
19. Shackleton CH. Profiling steroid hormones and urinary steroids. J Chromatogr, 1986; 379: 91–156.
Find This Resource
20. Shackleton CH. Mass spectrometry in the diagnosis of steroid-related disorders and in hypertension research. J Steroid Biochem Mol Biol, 1993; 45: 127–4.
Find This Resource
21. Vogeser M, Seger C. A decade of HPLC-MS/MS in the routine clinical laboratory - goals for further developments. Clin Biochem, 2008; 41: 649–62.
Find This Resource
22. Vogeser M, Parhofer KG. Liquid chromatography tandem-mass spectrometry (LC-MS/MS)–Technique and Applications in Endocrinology. Exp Clin Endocrinol Diabetes, 2007; 115: 559–70.
Find This Resource
23. Turpeinen U, Linko S, Itkonen O, Hämäläinen E. Determination of testosterone in serum by liquid chromatography-tandem mass spectrometry. Scand J Clin Lab Invest, 2007; 68: 50–7.
Find This Resource
24. Maunsell Z, Wright DJ, Rainbow SJ. Routine isotope-dilution liquid chromatography-tandem mass spectrometry assay for simultaneous measurement of the 25-hydroxymetabolites of vitamins D2 and D3. Clin Chem, 2005; 51: 1683–90.
Find This Resource
25. McCann SJ, Gillingwater S, Keevil BG. Measurement of urinary free cortisol using liquid chromatography-tandem mass spectrometry: comparison with the urine adapted ACS:180 serum cortisol chemiluminescent immunoassay and development of a new reference range. Ann Clin Biochem, 2005; 42: 112–8.
Find This Resource
26. Shiraishi S, Lee PW, Leung A, Goh VH, Swerdloff RS, Wang C. Simultaneous measurement of serum testosterone and dihydrotestosterone by liquid chromatography-tandem mass spectrometry. Clin Chem, 2008; 54: 1855–63.
Find This Resource
27. Vogeser M, Groetzner J, Küpper C, Briegel J. The serum cortisol:cortisone ratio in the postoperative acute-phase response. Horm Res, 2003; 59: 293–6.
Find This Resource
28. de Jong WH, Graham KS, van der Molen JC, Links TP, Morris MR, Ross HA, et al. Plasma free metanephrine measurement using automated online solid-phase extraction HPLC tandem mass spectrometry. Clin Chem, 2007; 53: 1684–93.
Find This Resource
29. Fisker S, Edrup L, Orsko. Influence of growth hormone binding protein estimation in different immunoassays. Scand J Clin Lab Invest, 1998; 58: 373–81.
Find This Resource
30. Jones AM, Honour JW. Unusual results from immunoassays and the role of the clinical endocrinologist. Clin Endocrinol (Oxf), 2006; 64: 234–44.
Find This Resource
31. Sadideen H, Swaminathan R. Macroprolactin: what is it and what is its importance?. Int J Clin Pract, 2006; 60: 457–61.
Find This Resource
32. Taieb J, Mathian B, Millot F, Patricot MC, Mathieu E, Queyrel N, et al. Testosterone measured by 10 immunoassays and by isotope-dilution gas chromatography-mass spectrometry in sera from 116 men, women, and children. Clin Chem, 2003; 49: 1381–95.
Find This Resource
33. Herold DA, Fitzgerald RL. Immunoassays for testosterone in women: better than a guess?. Clin Chem, 2003; 49: 1250–1.
Find This Resource
34. Rosner W, Auchus RJ, Azziz R, Sluss PM, Raff H. Utility, limitations, and pitfalls in measuring testosterone: an Endocrine Society position statement. J Clin Endocrinol Metab, 2007; 92: 405–13.
Find This Resource
35. Cavalier E, Wallace AM, Knox S, Mistretta VI, Cormier C, Souberbielle JC. Serum vitamin D measurement may not reflect what you give to your patients. J Bone Miner Res, 2008; 23: 1864–5.
Find This Resource
36. Annesley TM. Ion suppression in mass spectrometry. Clin Chem, 2003; 49: 1041–44.
Find This Resource