Steven R. Peacey The need for normal adrenal functioning continues into old age. After decades of speculation, there is now firm evidence that the patterns of basal and stimulated levels of cortisol secretion are substantially unchanged in healthy older adults. Subtle age-related changes have been described related to the metabolism of adrenal hormones, and morphologic features such as nodules appear quite commonly in the aging adrenal glands. Their importance arises from much readier and often serendipitous recognition as advanced imaging techniques are more widely used. It is therefore relevant to open this chapter with a summary of the physiologic and biochemical actions of adrenal steroids, mechanisms controlling their secretion, and techniques available for assessing the function and anatomy of the adrenal glands. Of the multitude of steroids found in the adrenal cortex, only the secretions of cortisol and aldosterone have undisputed and vital endocrine roles. The distinction between glucocorticoid and mineralocorticoid hormone actions is based on physiologic observations, backed by differential effects on critical enzyme systems in target tissues. Cortisol (hydrocortisone) is the natural glucocorticoid of humans and most other mammals, but not the rat, which is unable to synthesize cortisol and uses corticosterone instead. It has long been recognized that cortisol, especially in high doses, has mineralocorticoid properties, and this has led to the widespread use of dexamethasone (a synthetic glucocorticoid), effectively without mineralocorticoid properties, as the benchmark glucocorticoid.1 This practice has passed from laboratory experiments to clinical investigation, as will be discussed later. There are reasons to question the validity of such assumptions, although the pragmatic clinical tests have proven value. There was previously a tendency to subdivide the actions of glucocorticoids into those seen at low doses and termed physiologic and those seen with high doses, typically causing cushingoid side effects, termed pharmacologic. There is no sound scientific basis for this differentiation because new effects are not seen with high doses, although the clinical sequelae are striking. The term glucocorticoid derives from the effects on carbohydrate metabolism—antagonism of insulin action, promotion of hepatic glycogen synthesis, and participation in the defenses against hypoglycemia. It may affect resource use by virtue of tissue differences in response to the key glycolytic enzyme, phosphoenolpyruvate carboxykinase.2 Glucocorticoids have many other actions, often permissive in nature. These include vascular and renal responses affecting control of blood pressure and extracellular water content. Other critical roles include actions on protein and lipid synthesis and complex interactions with the immune system. In addition, there is the well-recognized but poorly characterized function played by enhanced glucocorticoid secretion in combating stress. The stimuli recognized as stressful and capable of evoking enhanced cortisol secretion are numerous; these include fever, trauma, hemorrhage, plasma volume depletion, hypoglycemia, and even psychological disturbance. A unifying hypothesis is thus hard to formulate but, with regard to inflammatory processes, it is now widely believed that the role of glucocorticoids is to curtail the effects of rapidly responding cytokine and acute phase protein production; if protracted, this could be potentially damaging.3 The action of aldosterone is ostensibly simpler, operating primarily via renal mechanisms to control extracellular sodium and potassium levels, with secondary consequences on fluid balance and blood pressure. The effects of mineralocorticoids on other tissues such as the colon, brain, and pituitary have been documented, but their significance is much less certain. The secretion of aldosterone and its circadian rhythm are maintained in older adults, despite a decrease in tonic levels of renin, its principal regulator.4 The adrenal cortex also synthesizes androgens. These include androstenedione and dehydroepiandrosterone; much of the latter is conjugated and secreted as the sulfate. The function of adrenal androgens remains obscure, although much has been made of the phenomenon in childhood of the so-called adrenarche, when enhanced amounts are made from about the age of 7 years. By contrast with cortisol production, there is a well-documented fall in adrenal androgen production in older adults to as little as 5% of young adult levels, with decreased adrenocorticotropic hormone (ACTH) responsiveness, which has been termed the adrenopause.5 Apart from effects on body hair, it is not at all clear what function the secretion of adrenal androgens serves in normal adults. It has been postulated that the decline in dehydroepiandrosterone levels is partly responsible for the increased atherogenesis and, hence, cardiovascular disease in older adults, but evidence has not supported this hypothesis.6 It appears more likely that dehydroepiandrosterone has an immunomodulatory and possibly antioncogenic action. Dehydroepiandrosterone replacement in older adults increases natural killer cell cytotoxicity and has been claimed to improve the sense of physical and psychological well-being dramatically.7 The effects of hormones on tissues depend on the distribution of specific receptors. Advances in knowledge have simultaneously clarified aspects of steroid hormone action and led to paradoxes that await definitive resolution. Steroids are lipophilic and readily enter cells; steroid receptors are intracellular. The classic model of steroid action is that steroid hormones bind to cytoplasmic receptors, forming activated complexes that are translocated to the nucleus where specific genes are activated, leading eventually to protein products as the end point of hormone influence.8 A similar pattern was proposed for the structurally dissimilar thyroid hormones. Molecular cloning techniques have not only revealed that all steroid hormone receptors show strong homologies to each other and the proto-oncogene c-erb-A, but that the latter actually appears to be a thyroid hormone receptor. All these receptors share homologies in the hormone and the DNA-binding domains and can be regarded as constituting a superfamily of genes whose products are transcriptional regulatory proteins evolved from a common ancestor gene.9 The steroid hormone receptor is bound to a protein complex containing the heat shock proteins hsp 90, hsp 70, and hsp 65. Exposure to steroid hormone leads to dissociation of the receptor from the complex so that the receptor is able to bind the hormone.10 Along with the classic genomic mechanisms, the cortisol-glucocorticoid receptor complex can interact with other transcription factors such as nuclear factor-κB and also nongenomic pathways via membrane-associated receptors and second messengers.11 The new complex of hormone-plus receptor adopts a different molecular conformation, exposing the DNA-binding domain of the receptor. Thus far, the generalized scheme for steroid hormones applies to glucocorticoids. When it comes to identifying the molecular basis for mineralocorticoid and glucocorticoid actions, difficulties arise. The type 1 receptor—originally considered to bind mineralocorticoids with higher affinity than glucocorticoids—shows no such distinction with more modern techniques. There is a marked relationship shown at the molecular level as well.12,13 A possible explanation for the failure of the great molar excess of cortisol to swamp the type 1 receptor with regard to aldosterone binding has been suggested for tissues such as kidney, gut, and salivary glands. These tissues possess a potent 11-hydroxysteroid dehydrogenase enzyme system, which rapidly converts cortisol to cortisone, and cortisone does not bind measurably to the receptor.14 Acting through the genome, glucocorticoids enhance several key metabolic enzymes, such as hepatic tyrosine aminotransferase15 and tryptophan oxygenase.16 In addition to this classic mode of action, it has also been suggested that many of the actions of glucocorticoids on the immune system are mediated by a specific protein product formerly termed lipocortin, now termed annexin A1, which acts as a second messenger.17 This has multiple sites of action, especially inhibiting polymorphonuclear leukocyte trafficking, reduction of proinflammatory cytokines, and stimulation of antiinflammatory cytokines. Cortisol secretion is under the immediate control of pituitary ACTH secretion; this acts to promote the conversion of cholesterol to pregnenolone by the removal of the six-carbon fragment from the cholesterol side chain. These steps occur within the mitochondrion. A complex cascade of cytochrome P450 variants has been implicated as steroidogenesis proceeds, shuttling from mitochondrion to endoplasmic reticulum and back. The chronic effects of ACTH affect many more steps in steroidogenesis than just cholesterol side chain cleavage.18 Physiologic control of ACTH secretion involves three major areas—circadian rhythms, stress, and negative-feedback inhibition by cortisol. ACTH is synthesized as part of a large 31-kDa precursor polypeptide, pro-opiomelanocortin.19 This is cleaved and the major fragments, including ACTH and β-endorphin, are usually cosecreted in equimolar proportions. The stimulus to ACTH release is from the hypothalamus via the hypothalamopituitary portal vessels conveying corticotropin-releasing factors.20 These are a complex of polypeptides, the major constituent of which is a 41-residue moiety, corticotropin-releasing hormone (CRH). However, this alone has less potent ACTH-releasing properties than crude hypothalamic extracts. It has been shown that vasopressin (AVP) and probably other unidentified compounds act synergistically with CRH.21 The secretion of these corticotropin-releasing factors appears to be driving pulses of ACTH and cortisol, in turn. The circadian rhythm is composed of pulses of varying amplitudes and frequency, with a nadir reached at midnight, but the onset of activity is at about 3 to 4 AM, reaching a peak at 8 to 9 AM. The pulses of ACTH and cortisol decrease in size and frequency thereafter, although there is often a secondary rise at about lunchtime, which seems to be related to food ingestion.22 As mentioned earlier, there is a formidable array of apparently unrelated stressors that can stimulate the release of ACTH and cortisol. There has been preliminary evidence that the relative importance of CRH, AVP, and oxytocin varies according to the stimulus.23 When inflammation is involved, there is growing evidence for interleukin-1, interleukin-6, and tumor necrosis factor having the capability to stimulate the hypothalamic-pituitary-adrenal axis, thus providing a loop to suppress their own production.24 Reports that have suggested extrahypothalamic production of ACTH secretagogues lack confirmation of authenticity or physiologic significance. Negative feedback of cortisol on ACTH production constitutes a sensitive homeostatic regulatory mechanism. The sites of negative feedback include not only the ACTH-producing cells of the anterior pituitary itself, but also higher centers, including the hypothalamus and CA3 field of the hippocampus.25 Aldosterone is produced by the distinct outer part of the adrenal cortex, the zona glomerulosa. In humans, this is found in cell clusters rather than in a distinct zone. The main regulation of aldosterone is via the renin-angiotensin-aldosterone system (RAAS). The stimuli to renin release from the juxtaglomerular cells of the kidney are low renal perfusion pressure, sodium depletion, and hypokalemia, although hyperkalemia acting directly on the zona glomerulosa is a more potent stimulus for aldosterone release than hypokalemia. Renin acts on renin substrate or angiotensinogen, released into the circulation from the liver, to form angiotensin-I. This decapeptide is converted to the octapeptide angiotensin-II by angiotensin-converting enzyme (ACE), which is of widespread distribution, but is most importantly found in the pulmonary bed.26 Angiotensin-II, apart from being a powerful arteriolar vasoconstrictor, stimulates aldosterone secretion from the adrenal cortex. Aldosterone, as noted, acts powerfully to retain salt (and obligatorily water), but promotes kaliuresis, hence closing the homeostatic feedback loop. There are other minor influences recognized as acting on aldosterone secretion, including ACTH, dopamine, and serotonin. Numerous studies indicate that basal, circadian, and stimulated cortisol secretion remains intact well into older age.27–33 This is particularly important with regard to the ability to withstand stress, and the cortisol response to exogenous ACTH has been shown to be normal in older adults following myocardial infarction.34 There are well-documented changes in the metabolism of corticosteroids, with an age-related decrease in the catabolism of cortisol.35,36 Because of the intact negative feedback mechanisms, there is a commensurate reduction in the cortisol production rate. Aldosterone secretion is also normally well preserved in healthy older adults.37 The recognized decline in adrenal androgen production has been referred to earlier.30,38–41 Tests of adrenal function in older adults are for the reasons noted, mainly those established for the younger adult population. The diminishing reliance on urinary collection is beneficial for practical reasons and also means that some of the physiologically irrelevant changes alluded to earlier will not prove distracting. The key to successful and safe investigation is careful selection. To investigate possible adrenal insufficiency, the basal measurement of greatest value is the plasma cortisol level, measured at the circadian peak at 8 to 9 AM. For most laboratories, a plain clotted sample is required to measure cortisol. Measurement of the midnight cortisol level is uninformative. Random plasma cortisol level measurements at other times of the day are usually worthless, although an undetectable plasma cortisol level (<50 nmol/L)—in the afternoon, for example—might warrant further investigation. If the 9 AM cortisol level is less than 150 nmol/L, the diagnosis of adrenal insufficiency is a strong possibility and, if greater than 450 nmol/L, the patient is normal. For values in between, a formal test of adrenal reserved should be performed. These tests are described here. The short Synacthen test (SST) is the simplest and most widely used test. A direct assessment of adrenal reserve can be made by measuring the plasma cortisol level before and 30 minutes after the intramuscular or intravenous administration of 250 µg of tetracosactrin, synthetic ACTH(1-24). Failure of the 30-minute cortisol level to peak above 500 nmol/L may indicate adrenal insufficiency. This test is an excellent choice when primary adrenal insufficiency is suspected, but may also be used to detect secondary adrenal insufficiency due to the fact that chronic ACTH deficiency often leads to adrenal atrophy or adrenal downregulation; thus, rapid cortisol release in response to Synacthen is impaired.42 It is important to be aware that false-negative results may occur when using the SST in individuals with secondary adrenal insufficiency. This relates to incomplete downregulation of the adrenals, occurring in some individuals with ACTH deficiency. By performing this test at 9 AM, the possibility of false-negative results can be highlighted when such individuals have a relatively low basal 9 AM cortisol level but have a normal peak cortisol level. However, if an individual fails to achieve a satisfactory peak cortisol (>500 nmol/L), this does not indicate whether this is primary or secondary adrenal insufficiency, and further investigation is required. Traditionally a long Synacthen test was performed to distinguish between these two possibilities.43 However, given the widely available measurement of ACTH, the distinction is made by measuring ACTH level son two samples approximately 30 minutes apart (due to the pulsatile nature of ACTH) at around 9 AM. A low or normal-range ACTH level indicates secondary adrenal insufficiency, and should prompt further pituitary investigations, whereas a raised ACTH level indicates primary adrenal insufficiency. In malnourished individuals or those with a low protein state (e.g. low albumin level), a falsely low measurement of cortisol can occur related to the reduced level of cortisol-binding globulin (CBG), to which cortisol is mostly bound and that accounts for much of the measured total cortisol—bound and free cortisol.44 If secondary adrenal insufficiency is suspected, a direct assessment of the hypothalamic-pituitary-adrenal axis is required. In those younger than 65 years, the insulin stress test (IST) is still considered the gold standard test (for ACTH and growth hormone [GH] reserve), whereby achieving a peak cortisol of more than 500 nmol/L during insulin-induced hypoglycemia of the IST (<2.2 mmol/L) is considered normal. However, contraindications to the test include a 9 AM plasma cortisol level less than 100 nmol/L, a history of epilepsy, and known ischemic heart disease. Due to the potential for occult ischemic heart disease in older adults, the test is not commonly used, although it has been used successfully.45 The glucagon stimulation test (GST) is an alternative direct test of the pituitary-adrenal axis (as well as a test of GH reserve) and can be used at any age and in patients with epilepsy and ischemic heart disease.46 Following a basal cortisol measurement in the fasting state, 1 mg of glucagon is given intramuscularly and the cortisol level measured at 150 and 180 minutes. A peak cortisol level of over 500 nmol/L is considered normal47 and has been closely correlated to cortisol responses during the IST.48 One disadvantage is the difficulty interpreting unusual results that are sometimes seen with the GST, whereby the basal cortisol is satisfactory but subsequent cortisol levels become reduced during the test, leading to an reputation for unreliability as noted by by some endocrinologists. Primary but not secondary adrenal insufficiency is associated with reduced aldosterone production. Blood can be sampled at any time of the day for aldosterone and renin, but the exact laboratory requirements should be checked. A plasma sample separated within 30 minutes is often required, implying that such measurements cannot be performed at all facilities. The finding of raised renin and low aldosterone levels confirms mineralocorticoid insufficiency. Adrenal hyperfunction usually means cortisol excess or Cushing syndrome. Conventional methods of investigation are used first to establish the presence of the syndrome. This is a simple and reliable test. Its value derives from the fact that at normal levels of plasma cortisol, most is bound to high-affinity CBG.49,50 The free cortisol level (thought to be the biologically active fraction) is generally small and is readily excreted in the urine. Because the capacity of CBG is limited and can be saturated with even minor degrees of cortisol hypersecretion, there tends to be a nonlinear and marked rise in urinary-free cortisol when excess cortisol is produced. As always, the cumbersome nature of a 24-hour urine collection may prove a more awkward test for some older patients. The 1-mg overnight dexamethasone suppression test is widely used as a screen for cortisol excess in an outpatient setting. Dexamethasone, 1 mg, is given orally between 11 PM and midnight, followed by blood sampling for cortisol at 9 AM the following morning. A cortisol level less than 50 nmol/L excludes Cushing syndrome. However, some normal individuals do not suppress to this level (false-positives), and a higher cutoff of less than 140 nmol/L is sometimes used.51,52 If this higher cutoff value is chosen, there will be a corresponding increase in false- negative results. Further evaluation is often required with a low-dose dexamethasone suppression test, and this must be performed as an inpatient procedure. Dexamethasone, 0.5 mg, is given orally every 6 hours for 48 hours, and the cortisol level should be suppressed below 50 nmol/L at 48 hours. If Cushing syndrome is diagnosed, further evaluation is required to differentiate between adrenal, pituitary, and ectopic ACTH causes. Measurement of the ACTH level at 9 AM is the initial step; an undetectable result indicates an adrenal cause for the excess cortisol level. A normal or raised ACTH level is due to an ACTH-secreting pituitary lesion (Cushing disease) or ectopic ACTH production from another source, commonly a lung tumor.53 The latter two situations are differentiated using a combination of pituitary and lung imaging, along with the current gold standard test of inferior petrosal sinus sampling for ACTH. Use of the high-dose dexamethasone suppression and corticotropin-releasing hormone (CRH) tests are currently used infrequently.54 In cases of adrenal carcinoma, it is not unusual to have mixed patterns of steroid excess. Virilization in women is not uncommon, and plasma testosterone level is raised. A striking rise in the dehydroepiandrosterone sulfate level is characteristic of adrenal carcinoma,55 and this large production of a weak androgen may greatly increase urinary 17-oxosteroid excretion.56
Adrenal and Pituitary Disorders
Disorders of the Adrenal Glands
Physiologic Responses to Adrenocortical Steroids
Glucocorticoids
Mineralocorticoids
Adrenal Androgens
Biochemical Actions of Steroid Hormones
Regulation of Adrenal Function
Regulation of Glucocorticoid Production
Regulation of Aldosterone Production
Adrenocortical Function in Normal Aging
Investigation of Adrenal Function
Glucocorticoid Deficiency
Short Synacthen Test.
Insulin Stress Test.
Glucagon Stimulation Test.
Mineralocorticoid Deficiency
Glucocorticoid Excess
24-Hour Urine-Free Cortisol Test.
Dexamethasone Suppression Tests.
Related posts:
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
Adrenal and Pituitary Disorders
87
