Steroids encompass a basic molecular skeleton of four rings of carbon atoms called perhydro-1,2-cyclopentenophenanthrene. The majority of endogenous steroids also possess two methyl groups at “angular” positions where two of these rings meet (1).
Steroid hormones are synthesized from a cholesterol backbone in a complex multi-enzyme process known as steroidogenesis (2,3). Steroidogenesis occurs in the mitochondria and smooth endoplasmic reticulum of cells of the adrenal cortex (Figure 1), placental syncytiotrophoblast, testicular Leydig, ovarian granulosa and theca (Miller and Bose, 2011).
The adrenal cortex is subdivided into three zones: the zona glomerulosa, which is responsible for the production of mineralocorticoids such as aldosterone; the zona fasciculata, which is responsible for the production of glucocorticoids such as cortisol; and the zona reticularis, which is responsible for the production of sex steroids such as dehydroepiandrostenedione (DHEA) and DHEA-sulfate (DHEAS), both of which, in the adrenals, are produced in far higher amounts than androstenedione and testosterone (Soldin and Soldin, 2009).
Mineralocorticoids and glucocorticoids are involved in the metabolism of electrolytes and sugars, respectively. Sex steroids influence the growth, development, differentiation and function of peripheral tissues of the female and male reproductive system, such as the mammary gland, uterus, vagina, ovary, testis, epididymis and prostate.
Steroid hormones are lipophilic which allows them to traverse easily across cell membranes. Steroids are present in the bloodstream in two fractions: the majority fraction, which is bound to plasma proteins; and the minority fraction, which is dissolved and free (4). The binding of steroids to water soluble proteins increases their half-life during transportation. However, only free hormones are able to penetrate cells in order to bind to specific cytoplasmic receptors and effect bioactivity. Steroids can act quickly, by binding to cell surface receptors; or slowly, by binding to cytoplasmic or nucleic receptors and ultimately activating gene transcription (1)(4).
Overall, there are more than 30 steroid hormones produced by the adrenal cortex (5). This method development project will investigate seven of these steroids, including testosterone, androstenedione, 17 hydroxyprogesterone (17 OHP), DHEAS, cortisol, 11 deoxycortisol (11 DOC) and 21 deoxycortisol (21 DOC).
Testosterone is a potent androgenic anabolic steroid which is regulated by pituitary-gonadal feedback involving luteinising hormone (LH) and, to a lesser degree, inhibins and activins (6). In males, approximately 95% of testosterone is produced in the testes and the rest from adrenal androgen precursors, through peripheral conversion of androstenedione in the liver, skin and adipose tissue (7). Testosterone is responsible for the development and maintenance of male reproductive organs and male secondary gender characteristics (8). In females, 25% of testosterone is synthesised in the ovarian theca cells, 25% by the adrenal gland, and the remaining 50% arises from peripheral conversion of testosterone precursors; DHEA and androstenedione (9). The main role in females is as an oestrogen precursor.
Testosterone is high in males during embryonic development when sexual differentiation occurs, the neonatal period and throughout the whole of adult sexual life. Male neonates have a rapid increase in testosterone, reaching maximum levels comparable to those in adolescents, between 30 and 60 days post-delivery. The levels then start to drop and no sexual differences are observed until puberty. In females, testosterone levels at birth are similar to those found in adult females (4).
The metabolism of testosterone mainly occurs in the liver by steroid catabolic enzymes, although metabolism is also present in other tissues, such as the skin (10). The majority of testosterone (40–60%) circulates bound, at high affinity, to sex hormone binding globulin (SHBG). A smaller fraction is bound, at low affinity, to albumin, and only 1–2% of total serum testosterone circulates in unbound form, known as free testosterone. The free and albumin-bound fractions contribute to bioavailable testosterone for androgen receptor activation (11).
Excessive testosterone production induces premature puberty in boys and virilisation in girls during childhood (12). During adulthood, excess testosterone production also results in varying degrees of virilisation including hirsutism, acne, oligo- or amenorrhea, or infertility in females. High testosterone can be due to androgen secreting tumours of adrenal and testicular/ovarian origin. Testosterone is also important in the investigation of newborns and infants presenting with ambiguous genitalia that may be as a result of congenital adrenal hyperplasia (CAH).
Decreased testosterone in females causes subtle symptoms. These may include some decline in libido and nonspecific mood changes. In males, it results in partial or complete degrees of hypogonadism. This is characterized by changes in male secondary sexual characteristics and reproductive function. The cause is either from primary, secondary or tertiary (pituitary/hypothalamic) testicular failure. In adult males, there also is a gradual, modest, but progressive, decline in testosterone production between the fourth and sixth decade of life. Since this is associated with a simultaneous increase of SHBG levels, bioavailable testosterone may decline more significantly than apparent total testosterone, causing nonspecific symptoms similar to those observed in testosterone deficient females (13). Opinions are however slightly divided, with evidence that after peak level at 19 years of age, the subsequent decline ends around 40 years of age after which it is maintained through to old age (14). It is notable that severe hypogonadism, consequent to aging alone, is rare.
Androstenedione (A4) is synthesised largely by the adrenal glands and regulated partially by adrenocorticotropic hormone (ACTH). A4 is also secreted by the testes and ovaries and synthesised in peripheral tissues through conversion of dehydroepiandrosterone (DHEA), which is catalysed by the enzyme 3β-hydroxysteroid dehydrogenase type 2 (15,16). A4 is an immediate precursor for testosterone. Peripheral conversion to testosterone is mediated by 17β-hydroxysteroid dehydrogenase type 5 (16)
The levels of A4 in blood exhibit a circadian rhythm synonymous with that of cortisol (17–19). Women have higher levels at mid-cycle than in the follicular phase of the menstrual cycle (20). Therefore, sampling time of the day and cycle should be taken into consideration when interpreting A4 results. A4 and testosterone can be converted peripherally to oestrogens, via aromatase, in both females and males. 80% of oestrogens present in males result from this process (21).
via aromatase (22).
In children, adrenal and gonadal tumours are uncommon, but many forms of CAH can increase serum A4 concentrations. The elevated levels can cause symptoms or signs of hyperandrogenism in women. Men are usually asymptomatic, but through peripheral conversion of androgens to oestrogens can occasionally experience mild symptoms of oestrogen excess, such as gynaecomastia. Most mild-to-moderate elevations in A4 are idiopathic. However, pronounced elevations may be indicative of androgen-producing adrenal or gonadal tumours.
17 hydroxyprogesterone (17 OHP) is a progestogen produced from 17 hydroxypregnenolone and progesterone, through the actions of 3β-Hydroxysteroid dehydrogenase and 17-hydroxylase respectively. 17 OHP is synthesised mainly in the adrenal glands and is controlled by ACTH. It is also synthesised to some degree by the gonads (23). 17 OHP is a relative agonist of the progesterone receptor, a partial agonist of the glucocorticoid receptor and an antagonist of the mineralocorticoid receptor. The main role of 17 OHP is as a precursor molecule for the synthesis of cortisol and sex steroids.
Serum 17 OHP concentrations are age-dependent, with peak levels observed during foetal life and at term (24). During the first week of life, serum 17 OHP levels fall 50%. A small transient increase occurs in male infants 30-60 days postnatally. The levels for both sexes remain constantly low during childhood, and then progressively increase during puberty, reaching adult levels. There is an increase in 17 OHP in the third trimester of pregnancy, mainly due to foetal adrenal and placental release (25). 17 OHP has a diurnal variation, with concentrations highest in the morning and lowest at night (19). 17 OHP concentrations also vary throughout the female menstrual cycle, with increases during the luteal phase. 17 OHP is bound to both corticosteroid binding globulin and albumin. It is converted to pregnanetriol which is excreted in the urine (26).
17 OHP is elevated in CAH. The majority of cases of CAH are due to 21-hydroxylase deficiency. A deficiency in the next step of cortisol biosynthesis, 11β-hydroxylase deficiency (11βOHD), also allows 17 OHP to accumulate, but in correlation with the immediate precursors 11-deoxycortisol (S) and 11-deoxycorticosterone (25,27). A deficiency of either 11- or 21-hydroxylase results in decreased cortisol synthesis, and feedback inhibition of ACTH secretion is lost. Consequently, there is increased pituitary release of ACTH which increases production of 17 OHP. However, if 17 α hydroxylase (which catalyses formation of 17 OHP from progesterone) or 3β hydroxysteroid dehydrogenase type 2 (which catalyses formation of 17 OHP formation from 17-hydroxypregnenolone) are deficient, 17 OHP levels are then low, with possible increase in progesterone or pregnenolone respectively.
Analysis of 17 OHP is also useful as part of a group of tests to evaluate females with hirsutism or infertility, both of which can result from adult onset CAH.
17 OHP has a diurnal variation in its production, with concentrations highest in the morning. As a consequence, dynamic function testing of the adrenal response to synthetic ACTH and subsequent simultaneous measurement of 17 OHP and cortisol helps in interpretation of results and the differentiation between subtypes of CAH. 17 OHP concentrations are affected in neonates by birth weight and gestational age; and there is assay interference by foetal steroids.
DHEAS is the most abundant steroid hormone in the circulation and is the major precursor of sex steroids (28). DHEAS results from conversion of dehydroepiandrosterone (DHEA) by sulfotransferase (29,30). It can be reconverted back to DHEA, through removal of the sulfate ester at carbon-3 by steroid sulfatase, and is consequently oxidized to androstenedione, the immediate precursor of testosterone or oestrone (31) (Figure 3).
DHEA and DHEAS are synthesised mainly by the zona reticularis of the adrenal cortex, which is the only source of these hormones for females. In males, these steroids are also produced by the testes (32). DHEA and DHEAS synthesis is partly controlled by ACTH (33).
The amount of DHEAS in the brain is higher than that in the adrenals, testes, and plasma, suggesting a neuroendocrine role of this hormone (29). During pregnancy, DHEAS and its 16-hydroxylated metabolites are secreted by the foetal adrenal gland in large quantities. They serve as precursors for placental production of the dominant pregnancy oestrogen, oestriol. Within weeks after birth, DHEAS levels fall by 80% or more and remain low until the onset of adrenarche. Following this, DHEAS output is highest between the ages of 20-30 years before a decline of 2% per year, resulting in 10–20% of the peak production by the eighth or ninth decade of life. The clinical significance of this age-related drop is unknown and trials of DHEAS replacement in the elderly have not produced convincing benefits. However, in young and old patients with primary adrenal failure, the addition of DHEAS to corticosteroid replacement has been shown in some studies to improve mood, energy, and sex drive (29).
The enzyme catalysed changes are shown in red (Goodman, 2009).
Elevated DHEAS levels can cause symptoms or signs of hyperandrogenism in women. Men are usually asymptomatic, but through peripheral conversion of androgens to oestrogens can occasionally experience mild oestrogen excess. Most mild to moderate elevations in DHEAS levels are idiopathic. However, pronounced elevations of DHEAS may be indicative of androgen-producing adrenal tumours. In small children, CAH due to 3β hydroxysteroid deficiency is associated with excessive DHEAS production. Lesser elevations may be observed in 21-hydroxylase deficiency and 11β-hydroxylase deficiency. In contrast, steroidogenic acute regulatory protein or 17α-hydroxylase deficiencies are characterized by low DHEAS levels. DHEAS appears to be involved in other diseases such as cardiovascular and immunological disorders, psychiatric pathology and degenerative diseases as well as ageing processes (34).
Cortisol is synthesised in the adrenal gland under the influence of ACTH, which is secreted by the anterior pituitary gland in response to corticotropin-releasing hormone (CRH) from the hypothalamus. Serum cortisol inhibits secretion of CRH and ACTH, which prevents excessive secretion of cortisol from the adrenal glands (35). Cortisol is the most clinically important glucocorticoid, accounting for 75% to 95% of all glucocorticoid activity.Cortisol is transported in plasma mainly bound to cortisol-binding globulin (CBG) and albumin. Normally, <5% of circulating cortisol is free (unbound) which is the physiologically active form (1).
Free cortisol diffuses across cell membranes at the target tissues and binds to the glucocorticoid α receptor (NR3C1). The cortisol–glucocorticoid receptor complex translocates to the nucleus and regulates gene expression by binding to the glucocorticoid response element of target genes (36).
Cortisol is measured for the investigation of conditions causing hypercortisolism and hypocortisolism. The most common cause of increased plasma cortisol levels in women is a high circulating concentration of oestrogen (e.g. oestrogen therapy, pregnancy) resulting in increased concentration of CBG. Another cause of hypercortisolism is Cushing’s syndrome, due to either primary adrenal disease (adenoma, carcinoma, or nodular hyperplasia) or an excess of ACTH (from a pituitary tumour or an ectopic source) hypercortisolism. ACTH-dependent Cushing’s syndrome due to a pituitary corticotroph adenoma is the most frequently diagnosed subtype. This is most commonly seen in women in the 3rd to 5th decades of life. The onset is insidious and usually occurs 2 to 5 years before a clinical diagnosis is made.
Hypocortisolism is seen in primary adrenal insufficiency (Addison’s disease), secondary adrenal insufficiency, pituitary insufficiency, hypothalamic insufficiency and congenital adrenal hyperplasia.
11-Deoxycortisol is a glucocorticoid steroid hormone and the immediate precursor of cortisol. It was first synthesized by Tadeusz Reichstein (37) and was referred to as Reichstein’s Substance. 11-Deoxycortisol acts as a glucocorticoid, though it is less potent than cortisol. It can be synthesized from 17α-hydroxyprogesterone.
11-Deoxycortisol is typically increased in conditions where ACTH is elevated, such as Cushing’s disease, ACTH-producing tumours and adrenocortical tumours. Deficiency of 11β-hydroxylase causes congenital adrenal hyperplasia. This defect accounts for about 5% of CAH cases (36).
11-Deoxycortisol is measured as part of the metyrapone test. This test allows the diagnosis of primary and secondary adrenal insufficiency. The results of overnight metyrapone testing correlate closely with the gold standard of hypothalamus-pituitary-adrenal (HPA) axis assessment i.e. insulin hypoglycaemia testing. Combining 11-deoxycortisol measurements with ACTH measurements during metyrapone testing further enhances the performance of the test. Metyrapone blocks the conversion of 11-deoxycortisol to cortisol, which then stimulates the adrenals to produce more 11-deoxycortisol. A blood level of 11-deoxycortisol which is not elevated after metyrapone administration indicates the presence of adrenal insufficiency (38). However, it does not differentiate between primary and secondary causes of adrenal failure. In patients with Cushing’s syndrome caused by ectopic ACTH or adrenal tumours, metyrapone does not change the ACTH or 11-deoxycortisol levels. Measurements of cortisol and ACTH have generally replaced the metyrapone test.
21-Deoxycortisol (21 DOC) is an intermediate steroid in the glucocorticoid pathway. 21 DOC results from hydroxylation of 17 alpha hydroxyprogesterone (17 OHP), through 11β hydroxylase (11βOH) activity (Figure 4). Due to the high levels of 17 OHP in 21 OHD, 21 DOC is also increased and can therefore be used as a supplementary marker for 21 OHD. Normal individuals will not have an increase in 21 DOC and is virtually non-functional in 11βOHD, supporting its use to discriminate between normal, 11βOHD and 21 OHD conditions (27).
11βOH is limited to the adrenal cortex in comparison to 21 OH. Consequently, 21 DOC is not affected by the same secretory pattern as 17 OHP and is not affected by puberty or menstrual cycle, making it an ideal marker to make therapeutic decisions in puberty (Tonetto-Fernandes et al., 2006).
21 DOC is the precursor of the urinary metabolite pregnanetriolone, which has been considered a hallmark analyte for confirming CAH in infancy (20).
21 hydroxylase deficiency leads to an increased production of steroid above the enzymatic block and depressed production below the enzymatic block. The biosynthesis of 21 deoxycortisol resulting from the conversion of excess 17 α hydroxyprogesterone via 11β hydroxylase is shown (39).
The major diagnostic utility of 21 DOC measurements lies in the diagnosis and differentiation of 21 OHD and 11βOHD types of CAH. The clinical features of both are similar, with the exception of the presence of hypertension and hyperkalaemia in 11 OHD presenting at a later stage of the disease. The measurement of 21 DOC can supplement or confirm 17 OHP and A4 measurements in the diagnosis of difficult cases of CAH (Tonetto-Fernandes et al., 2006). Furthermore, 21 DOC seems to discriminate between carriers of classic and non-classic forms of CAH after an ACTH-stimulation test where 17 OHP fails to do so (40).
Enzyme deficiencies in the steroid metabolic pathway disrupt normal flow from precursors to product, leading to imbalances resulting in excess or deficiency of particular steroids (41). Steroid metabolism abnormalities can result in hyperandrogenism and hypoandrogenism, which can present at any age. In addition to the investigation of diseases, steroids are also used in the monitoring of treatment. There are four main groups of disorders in which steroids are important for their investigation:
- Disorders of sex differentiation presenting in neonates with ambiguous genitalia.
- Disorders of sex development in children presenting with either precocious or delayed puberty
- Adult males presenting with hypogonadism
- Adult females presenting with hyperandrogenism
Disorders of sex development (DSD) are a wide range of conditions with diverse pathophysiologythat most often present in the newborn or the adolescent. Affected newborns usually present with atypical genitalia, whereas adolescents present with atypical sexual development during the pubertal years (42). Ambiguous genitalia in neonates can be due to congenital adrenal hyperplasia (CAH), androgen insensitivity syndrome (AIS) or 5 alpha reductase deficiency.
The most frequent forms of CAH are 21 hydroxylase (CYP21A2) and 11β hydroxylase (CYPB1) deficiency, accounting for 90% and 5% of cases of CAH, respectively (43).
The hormonal pattern and clinical manifestations of CAH result from hyperstimulation of the adrenal cortex by excessive production of ACTH; loss of negative feedback as a result of reduced cortisol levels; and the ultimate accumulation of cortisol precursors and androgens. The abnormality predisposes the female newborn to ambiguous genitalia, as well as precocious puberty, which may occur in both sexes (Tonetto-Fernandes et al., 2006). Approximately 50% of patients with classical congenital hyperplasia due to 21 OHD will have salt wasting crisis which can be fatal if not diagnosed promptly. This is particularly important in male neonates as they have normal genitalia at birth and CAH may be unrecognised leading to delayed diagnosis. The death rate in salt-wasting CAH is between 4 and 10% and it is thought to be under-reported in males due to underdiagnosed CAH (44).
Androgen insensitivity syndrome (AIS) is an autosomal recessive disorder resulting from a mutation of the gene encoding the androgen receptor. Patients present with different symptoms, from phenotypically normal males with impaired spermatogenesis, to phenotypically normal women with primary amenorrhea. Various forms of ambiguous genitalia have been observed at birth (45).
5α reductase deficiency is an autosomal recessive disorder where males have impaired virilisation due to defective conversion of testosterone to dihydrotestosterone (DHT). DHT is essential for normal male external genitalia (46).
Hyperandrogenism in females can manifest as polycystic ovary syndrome (PCOS), acne, alopecia, menstrual disorders, late onset CAH, androgen secreting neoplasm and infertility. PCOS is a heterogeneous disorder that, depending on how it is defined, affects between 10% and 20% of women of reproductive age. PCOS is characterized by the presence of a combination of hyperandrogenism, polycystic ovaries and ovulatory dysfunction. Women with PCOS also have an increased risk for insulin resistance, metabolic dysfunction, glucose intolerance and diabetes, and cardiovascular disease (33,47). PCOS should be considered a diagnosis of exclusion, as all definitions proposed note that to make the diagnosis, related or mimicking disorders should be excluded by testing or clinical evaluation, including at a minimum, androgen-secreting neoplasms, Cushing’s syndrome, adrenal hyperplasia, and thyroid and prolactin dysfunction (47).
Male hypogonadism results from either primary testicular failure or secondary to pituitary defects leading to low levels of testosterone. The most common causes of primary hypogonadism include: Klinefelter’syndrome, undescended testicles, mumps orchitis, haemochromatosis, injury to the testicles and cancer treatment. Secondary causes of hypogonadism include: Kallmann syndrome, pituitary disorders (e.g. pituitary tumours)(48).
The assessment of steroids is vastly performed by automated chemiluminescent (CLIA) or electro-chemiluminescent (ECLIA) immunoassay platforms and by semi-automated radioimmunoassays (RIA) techniques (49). Despite their simplicity and high throughput, these assays are subject to cross reactivity by other similar endogenous and exogenous steroids or metabolites and matrix effects (50). Moreover, immunoassay measurements tend to exhibit high variability at low concentrations with possible erroneous and misleading results (49). The performance of some immunoassay techniques have been recognised as suboptimal, resulting in recommendations to use mass spectrometry (MS), rather than immunoassays, for the measurement of sex steroids (51).
Currently, commercially available direct immunoassays for the measurement of testosterone lack sensitivity and accuracy for the concentrations found in women, children and men with hypogonadism (52). There is also lack of standardization (traceability). Testosterone immunoassays cross-react with the structurally similar oral contraceptive pill component, norethisterone. This is seen in Roche platforms, and to a smaller extent, in the Siemens ADVIA Centaur assay. DHEAS appears to interfere with some immunoassays for the measurement of testosterone in females, particularly in the Roche Testosterone assay (53). Such interference is not observed by liquid chromatography-tandem mass spectrometry (LC-MS/MS) (54). Despite the acknowledged superiority of LC–MS/MS for measuring female testosterone, there are still only a very small percentage of laboratories (18%) participating in the UKNEQAS scheme using this technology. This is an indication of the popularity of immunoassays despite the known limitations (51).
A4 is not routinely available on main clinical biochemistry analysers and the smaller niche immunoassays available overestimate A4 results compared to those reported by LC–MS/MS methods, according to UK NEQAS, Birmingham Quality (51).
17 OHP immunoassays are subject to interferences by other similar steroids, particularly 17-hydroxy pregnenolone, which tends to be very high in newborns and even medications such as spironolactone, due to poor antibody specificity (55,56). LC-MS/MS overcomes these issues of cross reactivity associated with the measurement of 17 OHP.
Immunoassays for the measurement of DHEAS demonstrate poor performance, as shown in a study comparing 7 commercially available immunoassays with the LC-MS/MS (34). Beckmann Access, Siemens Immulite 2000 and Diasorin Liaison, agreed well with LC-MS/MS, whereas Abbott Architect, Siemens Advia Centaur XP and Roche Modular measured higher DHEAS concentrations compared to LC-MS/MS, and Siemens Coat-a Count RIA measured lower concentrations than LC-MS/MS (34). This study reveals a wide variation of results encountered by immunoassays currently in use within UK laboratories.
Cortisol immunoassays are affected by interference from sample matrix and gender, likely attributable to variations in both CBG concentration and the accumulation of structurally related steroids in various physiological and pathophysiological conditions (57). Cortisol immunoassays are also prone to interference from 11 DOC in patients receiving the 11β-hydroxylase inhibitor metyrapone for the medical management of Cushing’s syndrome (58). Since metyrapone therapy blocks the conversion of 11 DOC to cortisol, it causes a >10-fold increase in circulating concentration of this precursor, which cross-reacts significantly with cortisol immunoassay. 11 DOC cross-reaction is a clinically significant and under-reported phenomenon. It leads to falsely high serum cortisol results and this may in turn lead to over-treatment with metyrapone and potentially hypoadrenalism (58,59). Cortisol results by immunoassays are also increased in patients receiving treatment with prednisolone or 6-methylprednisolone, and in patients with 21-hydroxylase deficiency due to elevated 21 DOC (60). The majority of the laboratories use automated cortisol immunoassay platforms. Although the variation for individual methods remain within narrow limits, the variation across all methods is considerably wider producing a wide dispersion of results with a sex specific bimodal distribution. Studies have shown LC-MS/MS has a good correlation with the GC-MS reference measurement procedure (61).
11 DOC can be measured using radioimmunoassays (RIA), but similar to other steroid immunoassays, these are also subject to cross-reactivity. Most of the RIA assays for 11 DOC require an initial solvent extraction step because of lack of specificity of the antisera (20).
Assays for 21 DOC are uncommon. According to an on-line search (http://www.assayfinder.com/), in Europe, there is only a non-accredited laboratory in France which officially performs analyses of 21 DOC and the methodology used is not stated. In the UK, King’s Viapath has developed a LC-MS/MS method for 21 DOC (62) but it is not yet used in the clinical investigation of CAH and sample preparation seems lengthy and manually intensive. Despite description of several RIAs for 21 DOC, none is commercially available. 21 DOC monoclonal antibodies are not cost effective and some have a high cross-reactivity with cortisol (63). Sample preparation involves ether-extraction and liquid chromatographic separation (HPLC) before RIA (63), therefore making it a labour intensive procedure. Moreover, this method has poor sensitivity and hence limited application in the measurement of 21 DOC concentrations within the normal range (63).
The poor performance of immunoassays may be ameliorated by applying pre-extraction steps to minimise cross reactivity but this would become labour intensive and would not necessarily guarantee that the results would be comparable to those of LC-MS/MS. Moreover, with immunoassays, each steroid is analysed separately and with very limited ranges, whereas multiplexing is possible with LC-MS/MS. Immunoassays also tend to have high variances at low concentration levels which can lead to error and misleading results (64).
Mass spectrometry (MS) allows the determination of the molecular mass of compounds by measuring the mass to charge ratio (m/z). Mixtures of analytes in samples can then be distinguished due to the separated signals from the different m/z values. Although no longer a novel technique, it is relatively new compared to other bioanalytical techniques and particularly as an application in the clinical environment. The early origins can be traced to J.J. Thomson as far back as 1897 (65). Since then, rapid steps have been made in progressing the technology such that, today, MS occupies an exceptional position amongst analytical techniques. This has mainly been in proteomics but is transitioning across other fields such as steroidomics (66), which is relevant to this study.
Among the many reasons for the advancement are, unparalleled sensitivity, low detection limits, analytical speed, and high-throughput, amongst others (67–71).
A mass spectrometer consists of three fundamental components (
Figure 5 ) (72):
- A source: encompasses a sample inlet divide that mediates the vaporisation of a solid or liquid bio-specimen into gaseous phase; and an ionisation device that ionises vaporised bio-samples;
- A mass analyser: an ion path that transitions bio-sample ions from the near-atmospheric pressure of the high vacuum, separates them from each other based on the m/z, and moves them towards a detector
- A detector: to record the individual ion counts (signal intensity) when contact occurs on the detector, which is then output to a PC monitor.
Ion source, used to vaporise/ionise the sample. Mass analyser, used to separate the gas phase ions by m/z ratio. Detector, to perceive the mass separated ions and measure their relative abundance. A tandem mass spectrometer has more than 1 mass analyser, 2 quadrupoles (Q), in this case: (Q1), the first mass analyser, where a separation of intact molecules occurs prior to fragmentation in the collision cell (Q2), followed by a separation of the fragment ions in the second mass analyser filter (Q3) (Grebe and Singh, 2011).
There are several different types of ion sources, mass analyser and detectors (Table 1) and these may also be coupled either gas (GC) or liquid chromatography (LC) (65).
LC MS/MS is the combination of two compatible selective techniques – Liquid chromatography (LC) and MS/MS – that allows analytes of interest in high complex mixtures to be isolated and measured. LC differentiates compounds by their physico-chemical properties on an analytical column and MS differentiates compounds by their m/z ratio. It is this dual selectivity that makes LC-MS/MS a powerful analytical tool, producing a technique that is highly specific and sensitive (74). The use of tandem mass spectrometers adds further advantages for characterisation of molecules due to the ability to perform fragmentation (75).
Although there are several ion sources available for LC-MS/MS (73), the most commonly used are electrospray ionisation (ESI) and atmospheric pressure chemical ionisation (APCI) (Figure 6). In ESI, the solvent-analyte flow from the LC passes into the source through a positively charged, very narrow capillary, and gets nebulised as microscopic, positively charged solvent-analyte droplets. These droplets fly towards the negatively-charged entry plate (cone), with solvent evaporating on the way, until they disintegrate in a Coulomb explosion, since the repulsive charge of their ionised components exceeds their surface tension (76). The individual analyte molecular ions then pass through the faceplate entry hole into the mass spectrometer. In APCI, the solvent-analyte stream from the LC is vaporised by heated nebuliser gas and the polar components of the solvent vapour are ionised by a high-current discharge from a corona needle. The solvent molecules in turn subsequently transfer their charge to ionisable analyte molecules, which pass through the faceplate entry hole into the mass spectrometer (72). The type of LC-MS/MS instrumentation used in this project is the Quattro triple quadrupole (Figure 7) which can perform quantitation and can accommodate both an ESI or APCI inlet probe (77).
(A) Electrospray ionisation; (B) Atmospheric pressure chemical ionisation (72).
(77). (1) Samples from LC are introduced at ambient pressure into source to generate ions which are sampled across (2), prior to separation by m/z at (3). The intact molecule ions are broken into smaller fragment ions at (4) prior to their separation by m/z at (5). At (6), the transmitted ions are detected by a photomultiplier detector and the signal is amplified and digitised prior to presentation at the PC monitor at (7).
The use of more selective techniques like gas chromatography combined with mass spectrometry (GC–MS) or liquid chromatography coupled to tandem mass spectrometry (LC-MS/MS) are generally considered the gold standard for steroid hormone measurements because of their accuracy and sensitivity (78).
Gas chromatography (GC) offers excellent chromatographic resolution and the benefit of multi-class profiling potential. However, this technique requires extraction and purification steps as well as derivatisation of steroids (64,79), making it burdensome. GC-MS reproducibility becomes an issue due to incomplete derivatization of compounds containing hydroxyl groups such as cortisol (79). Liquid chromatography (LC) separation is compatible with the use of the underivatised compound and therefore remains more advantageous to GC-based methods.
In GC-MS, the presence of other conjugates, especially steroid glucuronides can contribute to the signal of sulphated steroids. Moreover, most commercial sulfatase enzymes have glucuronidase activity too, and chemical acidification cannot distinguish between sulphates and glucuronides, thereby rendering the assay less specific (80). LC-MS has been shown to be the method of choice for the analysis of sulphated steroids without any modification (81,82). This is another clear evidence of the superiority of LC-MS over GC-MS as the technique of choice for this study.
LC coupled with tandem mass spectrometry (MS/MS) methods are now preferred for the direct measurement of steroid hormones. LC-MS/MS is particularly suited to the measurement of small molecules, and makes it possible to overcome the lack of specificity of routine immunoassays, at the same time offering the throughput required by large laboratories. Compared to traditional GC-MS, LC-MS/MS displays higher practicability, leaving room for implementation of routine applications, and a wide dynamic range for accurately measuring physiologic and diseased hormone levels, and also allowing multi-analytical profiling (79).
Simultaneous measurement of testosterone, androstenedione, 17 OHP, DHEAS, cortisol, 11 DOC and 21 DOC will be a useful tool to help elucidate the aetiology of ambiguous genitalia in neonates and children, hyperandrogenism and hypoandrogenism in adolescents and adults. To the knowledge of this study, this level of multiplexing is yet to be achieved. However, there is a demand for it since these tests are commonly requested together but analysed separately. A 7-steroid panel for simultaneous determination of these analytes would decrease staff time and sample volume requirement, particularly in paediatric samples, in a cost effective manner.
Currently in the laboratory at Imperial College Healthcare NHS Trust female testosterone, A4 and 17 OHP are measured simultaneously by LC-MS/MS. The remaining steroids are not routinely determined by LC-MS/MS. Cortisol measurement, which is indicated for patients receiving metyrapone treatment and those with 21 OHD, is performed on the Abbott Architect using a competitive chemiluminescent immunoassay, DHEAS is measured by Siemens Immulite immunoassay and 11 DOC is a send away test to specialist units. 21 DOC assay is not known to be part of any laboratory investigation, the only laboratory listed on AssayFinder for analysing 21 DOC is a non-accredited laboratory in France. Development of an LC-MS/MS multiplexed method for all seven steroids would provide improved specificity and sensitivity, due to the advantage over immunoassays, and also better opportunities for full assessment of disorders associated with steroids, due to the inclusion of 11 DOC and 21 DOC. Importantly, it would save resources and contribute toward a more cost effective service.
LC-MS/MS requires sample pre-treatment in order to achieve the necessary sensitivity. Protein precipitation and liquid-liquid extraction (LLE) are the most popular sample extraction methods used in the LC-MS/MS for steroids (64). Solid phase extraction (SPE) has more recently gained popularity in sample pre-treatment. In SPE, compounds of interest are extracted from a liquid mixture sample based mainly on their chemical affinity for the chemical coating on solid particles in chromatographic packing material (stationary phase). These bound compounds are then washed to remove contaminants and then subsequently eluted from the stationary phase material. One of the main advantages of SPE in the diagnostic laboratory, and hence the popularity, is the ease of automation, thereby avoiding lengthy and laborious manual preparation. In addition, SPE technology simplifies the complex sample matrix by removing interference with consequent ion suppression/enhancement by other components in the sample, as well as purification and trace concentration of the analyte of interest (83).
Water’s OASIS PRiME HLB SPE µ-elution plates, which were used in this study, utilise reversed-phase hydrophilic N-vinylpyrrolidone monomers and lipophilic divinylbenzene monomers on the stationary phase chromatographic packing material. These monomers improve retention of polar components and possess retention capacity threefold higher than the traditional silica-based SPE sorbents like C18. The water-wettable capacity enables retention of a wider polarity range of analytes and does not require pre-conditioning of the stationary phase. This technique removes 95% of the common matrix interferences such as salts, proteins and phospholipids (83).
An internal standard (IS) is a structurally similar analogue or stable isotopically labelled version of the molecule of interest. The IS will have a similar extraction recovery, MS ionization response and chromatographic retention time as the molecule of interest. It is included with calibration standards, QC samples and study samples at a known and constant concentration to correct for experimental variability during sample preparation and analysis. An appropriate IS will control for extraction, HPLC injection and ionization variability (84).
Chromatography conditions for method development tend to be based on literature review and then optimised during the research study. In the absence of previously described methods, the preliminary point would be the logD of the analyte, which can be obtained from several software packages available to calculate logF, such as the free online services provided by ChemIDplus at http://chem.sis.nlm.nih.gov/chemidplus/ (85). The logD of an analyte is a measure of its hydrophobicity at a given pH and is therefore usually related to its retention on a reversed phase column such as C18. The flow rate should be within the limits of the mass spectrometer. Also, any pH modifiers should be volatile. Typical additives include formic acid, acetic acid, ammonia solution, ammonium acetate and ammonium formate. In multi analyte measurement, it is important to ensure that good chromatographic separation and peak resolution, as well as adequately low analysis time is obtained.
Mass spectrometry tuning involves determining whether the analyte of interest will ionise in the mass spectrometer. It is usually carried out by introducing a constant stream of analyte into the ion source directly by using a syringe infusion pump. This allows the mass spectrometer parameters, such as gas flow and ionisation voltage, to be ‘tuned’ to the analyte, creating the optimum conditions for ionisation and therefore sensitivity. The optimum ionisation mode polarity (positive/negative) can then be determined, adducts noted and the most informative MS/MS fragments selected.
The chromatographic and mass spectrometry conditions were based on and optimised from multiple sources of methods previously published for steroid LC-MS/MS analysis (7,25,86–96).
The purpose of this method development involved the definition of analytes for testing and the design of the appropriate methodology. The validation of the developed testing procedure is to demonstrate fitness for the intended purpose. This study was based on international bioanalytical guidelines from the Clinical and Laboratory Standard Institute (CLSI), U.S. Food and Drug Administration (FDA), Clinical Laboratory Improvement Amendments (CLIA) and European Medicines Agency (EMA), which recommend examination of parameters including selectivity, linearity and lower limit of quantification (LLOQ), precision and accuracy, extraction recovery and matrix effect, and stability (97). The parameters examined for the 7-steroid panel in this study are summarised in Table 2.
|Parameters||Definition||Possible validation practice|
|Precision||Measure of agreement between replicates||Several approaches used by FDA, CLSI, CLIA and EMA. Usually measured using a minimum of five determinations per concentration. Subdivided into within-run, intra-batch precision (98).|
|Accuracy||Measure of agreement between a measured quantity and a true value||Replicate analysis of samples containing known amounts of the analyte.|
|The lowest concentration that can be measured with acceptable imprecision [inter-assay CV]. Lowest concentration that can be measured with <20% precision and 80-120 % accuracy; or Signal:noise ratio is >10.||Within-run LLOQ should be established using ≥5 samples independent of standards. Between-run LLOQ should be established using samples from 3 runs on ≥2 different days (99).|
|Specificity||Measure of a method to differentiate the analyte(s) of interest and IS from endogenous or exogenous components in the matrix or other components in the sample.||Evaluate a high concentration of potential interferent in sample matrix (ideally a patient sample) with and without the presence of analyte (98).|
|Carryover||Relates to residual products from previous or concurrent analyses that may be introduced into an assay||Assessed by running blank samples or ow concentration samples after a high concentration sample or calibration standard at the ULOQ (99).|
|Linearity||An assessment of the difference between an individual’s measurements and that of a known standard over the full range of expected values||9–11 concentrations should be analysed with 2–4 replicates each; linearity should be evaluated with the polynomial regression method; serial dilutions should be avoided (98).|
|Recovery||Measure of extraction efficiency of an analytical method. Reported as a percentage of the known amount of an analyte carried through the sample extraction and processing steps of the method.||Compare the analytical results for extracted samples at 3 concentrations (low, medium, and high) with un-extracted standards that represent 100% recovery (98).|
|Stability||Measure of chemical stability of an analyte in a given matrix under specific conditions for given time intervals||Experiments for freeze and thaw stability, short-term temperature stability, long-term stability, stock solution stability, and post-preparative stability (i.e. time in autosampler); experiments should be done with ≥3 aliquots of a low and high concentration (97).|
|Matrix effects||Measure of direct or indirect alteration or interference in response due to the presence of unintended analytes (for analysis) or other interfering substances in the sample.||Comparison of the peak area for matrix samples spiked with analyte post-extraction vs analyte spiked into neat solution tested at all points of the calibration curve (100).|
|Calibration curve||Measure of response of the instrument with regard to the concentration of analyte should be known, and should be evaluated over a specified concentration range.||Blank, zero, and 6–8 calibration standards, prepared in the same biological matrix as the samples, placed at the LLMI, ULMI, and any medically relevant decision point within the AMI (98) (99).|
|Quality Control||Measure of monitoring the performance of a bioanalytical method and to assess the integrity and validity of the results of the unknown samples analysed in an individual batch||≥3 QC concentrations (1 around 3× the LLMI, 1 in midrange, and 1 near the ULMI) tested in duplicate. The number of QC samples analysed during a batch should represent ≥5% of the total number of patient samples (98) (99) (98).|
|Dilutions||Measure of the dilution to confirm the absence of any impact on the measured concentration of the analyte.||≥5 separate replicates of each intended dilution should be verified (98).|
|Method comparison||Comparison to the new method results against the reference method or an acceptable method||Compare at least 40 samples across the analytical range.|
Matrix effects from a biological sample elicit a different response from a compound when analysed in a biological matrix compared to a standard solution. The difference may be described as suppression or enhancement according to whether the response is diminished or magnified. These unpredictable effects are a regular problem for API ionization sources and the signal adulteration is a significant issue for quantitation LC-MS/MS. Matrix effects are the modification of ionization efficiency of an analyte by the presence of co-eluting substances from the LC column. These effects are therefore inconspicuous in the LC chromatogram but have a damaging impact on accuracy and sensitivity of the method (101).
There are two common approaches to assess matrix effects: the post-column infusion method (102) which provides a qualitative assessment of matrix effects, identifying regions most likely to experience matrix effects; and the post-extraction method (100) which quantitatively assesses matrix effects by comparing the response of an analyte in neat solution to the response of the analyte spiked into a blank matrix sample that has been carried through the sample preparation process (100,103). Another approach to assess matrix effects is to use the precision of the calibration line slopes in five different replicates of a biological fluid as an indicator. The relative standard deviation (RSD) should be ≤3–4% for the method to be considered practically free from matrix effects (104). The method described by Matuszewski et al. (2003) was used in this method development.
Cite This Work
To export a reference to this article please select a referencing stye below:
Related ServicesView all
DMCA / Removal Request
If you are the original writer of this dissertation and no longer wish to have your work published on the UKDiss.com website then please: